diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,45673 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 6520,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00015337423312883436,
+      "grad_norm": 0.38905563950538635,
+      "learning_rate": 2.3076923076923077e-06,
+      "loss": 1.8499,
+      "step": 1
+    },
+    {
+      "epoch": 0.00030674846625766873,
+      "grad_norm": 0.4016835391521454,
+      "learning_rate": 4.615384615384615e-06,
+      "loss": 1.8378,
+      "step": 2
+    },
+    {
+      "epoch": 0.00046012269938650307,
+      "grad_norm": 0.37288936972618103,
+      "learning_rate": 6.9230769230769225e-06,
+      "loss": 1.855,
+      "step": 3
+    },
+    {
+      "epoch": 0.0006134969325153375,
+      "grad_norm": 0.37776508927345276,
+      "learning_rate": 9.23076923076923e-06,
+      "loss": 1.8468,
+      "step": 4
+    },
+    {
+      "epoch": 0.0007668711656441718,
+      "grad_norm": 0.35385963320732117,
+      "learning_rate": 1.1538461538461538e-05,
+      "loss": 1.8202,
+      "step": 5
+    },
+    {
+      "epoch": 0.0009202453987730061,
+      "grad_norm": 0.375183641910553,
+      "learning_rate": 1.3846153846153845e-05,
+      "loss": 1.7975,
+      "step": 6
+    },
+    {
+      "epoch": 0.0010736196319018406,
+      "grad_norm": 0.39178767800331116,
+      "learning_rate": 1.6153846153846154e-05,
+      "loss": 1.83,
+      "step": 7
+    },
+    {
+      "epoch": 0.001226993865030675,
+      "grad_norm": 0.3473685383796692,
+      "learning_rate": 1.846153846153846e-05,
+      "loss": 1.8651,
+      "step": 8
+    },
+    {
+      "epoch": 0.0013803680981595093,
+      "grad_norm": 0.3625858426094055,
+      "learning_rate": 2.076923076923077e-05,
+      "loss": 1.8234,
+      "step": 9
+    },
+    {
+      "epoch": 0.0015337423312883436,
+      "grad_norm": 0.3558821976184845,
+      "learning_rate": 2.3076923076923076e-05,
+      "loss": 1.8373,
+      "step": 10
+    },
+    {
+      "epoch": 0.001687116564417178,
+      "grad_norm": 0.3543674349784851,
+      "learning_rate": 2.5384615384615383e-05,
+      "loss": 1.8301,
+      "step": 11
+    },
+    {
+      "epoch": 0.0018404907975460123,
+      "grad_norm": 0.3688778579235077,
+      "learning_rate": 2.769230769230769e-05,
+      "loss": 1.8176,
+      "step": 12
+    },
+    {
+      "epoch": 0.001993865030674847,
+      "grad_norm": 0.33031773567199707,
+      "learning_rate": 2.9999999999999997e-05,
+      "loss": 1.8193,
+      "step": 13
+    },
+    {
+      "epoch": 0.002147239263803681,
+      "grad_norm": 0.37415361404418945,
+      "learning_rate": 3.230769230769231e-05,
+      "loss": 1.8045,
+      "step": 14
+    },
+    {
+      "epoch": 0.0023006134969325155,
+      "grad_norm": 0.3306610882282257,
+      "learning_rate": 3.461538461538461e-05,
+      "loss": 1.847,
+      "step": 15
+    },
+    {
+      "epoch": 0.00245398773006135,
+      "grad_norm": 0.33363091945648193,
+      "learning_rate": 3.692307692307692e-05,
+      "loss": 1.7906,
+      "step": 16
+    },
+    {
+      "epoch": 0.002607361963190184,
+      "grad_norm": 0.35496965050697327,
+      "learning_rate": 3.9230769230769226e-05,
+      "loss": 1.8476,
+      "step": 17
+    },
+    {
+      "epoch": 0.0027607361963190185,
+      "grad_norm": 0.3373720645904541,
+      "learning_rate": 4.153846153846154e-05,
+      "loss": 1.8224,
+      "step": 18
+    },
+    {
+      "epoch": 0.002914110429447853,
+      "grad_norm": 0.34787583351135254,
+      "learning_rate": 4.384615384615385e-05,
+      "loss": 1.8096,
+      "step": 19
+    },
+    {
+      "epoch": 0.003067484662576687,
+      "grad_norm": 0.34838318824768066,
+      "learning_rate": 4.615384615384615e-05,
+      "loss": 1.7741,
+      "step": 20
+    },
+    {
+      "epoch": 0.0032208588957055215,
+      "grad_norm": 0.33458781242370605,
+      "learning_rate": 4.846153846153846e-05,
+      "loss": 1.7897,
+      "step": 21
+    },
+    {
+      "epoch": 0.003374233128834356,
+      "grad_norm": 0.38390642404556274,
+      "learning_rate": 5.0769230769230766e-05,
+      "loss": 1.7789,
+      "step": 22
+    },
+    {
+      "epoch": 0.00352760736196319,
+      "grad_norm": 0.3306937515735626,
+      "learning_rate": 5.3076923076923076e-05,
+      "loss": 1.8112,
+      "step": 23
+    },
+    {
+      "epoch": 0.0036809815950920245,
+      "grad_norm": 0.3686997592449188,
+      "learning_rate": 5.538461538461538e-05,
+      "loss": 1.7402,
+      "step": 24
+    },
+    {
+      "epoch": 0.003834355828220859,
+      "grad_norm": 0.33693501353263855,
+      "learning_rate": 5.769230769230769e-05,
+      "loss": 1.7584,
+      "step": 25
+    },
+    {
+      "epoch": 0.003987730061349694,
+      "grad_norm": 0.3808348476886749,
+      "learning_rate": 5.9999999999999995e-05,
+      "loss": 1.7745,
+      "step": 26
+    },
+    {
+      "epoch": 0.004141104294478528,
+      "grad_norm": 0.36507943272590637,
+      "learning_rate": 6.23076923076923e-05,
+      "loss": 1.7665,
+      "step": 27
+    },
+    {
+      "epoch": 0.004294478527607362,
+      "grad_norm": 0.35144340991973877,
+      "learning_rate": 6.461538461538462e-05,
+      "loss": 1.8104,
+      "step": 28
+    },
+    {
+      "epoch": 0.004447852760736197,
+      "grad_norm": 0.3287273943424225,
+      "learning_rate": 6.692307692307693e-05,
+      "loss": 1.7794,
+      "step": 29
+    },
+    {
+      "epoch": 0.004601226993865031,
+      "grad_norm": 0.366516649723053,
+      "learning_rate": 6.923076923076922e-05,
+      "loss": 1.7886,
+      "step": 30
+    },
+    {
+      "epoch": 0.004754601226993865,
+      "grad_norm": 0.35005462169647217,
+      "learning_rate": 7.153846153846153e-05,
+      "loss": 1.7485,
+      "step": 31
+    },
+    {
+      "epoch": 0.0049079754601227,
+      "grad_norm": 0.34249061346054077,
+      "learning_rate": 7.384615384615384e-05,
+      "loss": 1.8158,
+      "step": 32
+    },
+    {
+      "epoch": 0.005061349693251534,
+      "grad_norm": 0.33375537395477295,
+      "learning_rate": 7.615384615384614e-05,
+      "loss": 1.7429,
+      "step": 33
+    },
+    {
+      "epoch": 0.005214723926380368,
+      "grad_norm": 0.34734371304512024,
+      "learning_rate": 7.846153846153845e-05,
+      "loss": 1.779,
+      "step": 34
+    },
+    {
+      "epoch": 0.005368098159509203,
+      "grad_norm": 0.33567336201667786,
+      "learning_rate": 8.076923076923076e-05,
+      "loss": 1.7624,
+      "step": 35
+    },
+    {
+      "epoch": 0.005521472392638037,
+      "grad_norm": 0.36983931064605713,
+      "learning_rate": 8.307692307692307e-05,
+      "loss": 1.7564,
+      "step": 36
+    },
+    {
+      "epoch": 0.005674846625766871,
+      "grad_norm": 0.3916028141975403,
+      "learning_rate": 8.538461538461537e-05,
+      "loss": 1.7686,
+      "step": 37
+    },
+    {
+      "epoch": 0.005828220858895706,
+      "grad_norm": 0.368770033121109,
+      "learning_rate": 8.76923076923077e-05,
+      "loss": 1.7422,
+      "step": 38
+    },
+    {
+      "epoch": 0.00598159509202454,
+      "grad_norm": 0.3694186508655548,
+      "learning_rate": 8.999999999999999e-05,
+      "loss": 1.7688,
+      "step": 39
+    },
+    {
+      "epoch": 0.006134969325153374,
+      "grad_norm": 0.3690738081932068,
+      "learning_rate": 9.23076923076923e-05,
+      "loss": 1.7952,
+      "step": 40
+    },
+    {
+      "epoch": 0.006288343558282209,
+      "grad_norm": 0.38482505083084106,
+      "learning_rate": 9.46153846153846e-05,
+      "loss": 1.7534,
+      "step": 41
+    },
+    {
+      "epoch": 0.006441717791411043,
+      "grad_norm": 0.379218727350235,
+      "learning_rate": 9.692307692307692e-05,
+      "loss": 1.7851,
+      "step": 42
+    },
+    {
+      "epoch": 0.006595092024539877,
+      "grad_norm": 0.38669463992118835,
+      "learning_rate": 9.923076923076922e-05,
+      "loss": 1.7677,
+      "step": 43
+    },
+    {
+      "epoch": 0.006748466257668712,
+      "grad_norm": 0.3666672706604004,
+      "learning_rate": 0.00010153846153846153,
+      "loss": 1.7737,
+      "step": 44
+    },
+    {
+      "epoch": 0.006901840490797546,
+      "grad_norm": 0.3732942044734955,
+      "learning_rate": 0.00010384615384615383,
+      "loss": 1.7951,
+      "step": 45
+    },
+    {
+      "epoch": 0.00705521472392638,
+      "grad_norm": 0.44426068663597107,
+      "learning_rate": 0.00010615384615384615,
+      "loss": 1.8048,
+      "step": 46
+    },
+    {
+      "epoch": 0.007208588957055215,
+      "grad_norm": 0.5392276644706726,
+      "learning_rate": 0.00010846153846153845,
+      "loss": 1.7756,
+      "step": 47
+    },
+    {
+      "epoch": 0.007361963190184049,
+      "grad_norm": 0.5342690348625183,
+      "learning_rate": 0.00011076923076923076,
+      "loss": 1.7605,
+      "step": 48
+    },
+    {
+      "epoch": 0.007515337423312883,
+      "grad_norm": 0.41496914625167847,
+      "learning_rate": 0.00011307692307692307,
+      "loss": 1.7722,
+      "step": 49
+    },
+    {
+      "epoch": 0.007668711656441718,
+      "grad_norm": 0.39706549048423767,
+      "learning_rate": 0.00011538461538461538,
+      "loss": 1.787,
+      "step": 50
+    },
+    {
+      "epoch": 0.007822085889570552,
+      "grad_norm": 0.39256054162979126,
+      "learning_rate": 0.00011769230769230768,
+      "loss": 1.7451,
+      "step": 51
+    },
+    {
+      "epoch": 0.007975460122699387,
+      "grad_norm": 0.3877248167991638,
+      "learning_rate": 0.00011999999999999999,
+      "loss": 1.7643,
+      "step": 52
+    },
+    {
+      "epoch": 0.00812883435582822,
+      "grad_norm": 0.39427489042282104,
+      "learning_rate": 0.0001223076923076923,
+      "loss": 1.7687,
+      "step": 53
+    },
+    {
+      "epoch": 0.008282208588957056,
+      "grad_norm": 0.40825018286705017,
+      "learning_rate": 0.0001246153846153846,
+      "loss": 1.7692,
+      "step": 54
+    },
+    {
+      "epoch": 0.00843558282208589,
+      "grad_norm": 0.4013516306877136,
+      "learning_rate": 0.0001269230769230769,
+      "loss": 1.7207,
+      "step": 55
+    },
+    {
+      "epoch": 0.008588957055214725,
+      "grad_norm": 0.38877299427986145,
+      "learning_rate": 0.00012923076923076923,
+      "loss": 1.7393,
+      "step": 56
+    },
+    {
+      "epoch": 0.008742331288343558,
+      "grad_norm": 0.616520345211029,
+      "learning_rate": 0.00013153846153846153,
+      "loss": 1.7749,
+      "step": 57
+    },
+    {
+      "epoch": 0.008895705521472393,
+      "grad_norm": 0.42086687684059143,
+      "learning_rate": 0.00013384615384615385,
+      "loss": 1.8068,
+      "step": 58
+    },
+    {
+      "epoch": 0.009049079754601227,
+      "grad_norm": 0.40906980633735657,
+      "learning_rate": 0.00013615384615384615,
+      "loss": 1.8052,
+      "step": 59
+    },
+    {
+      "epoch": 0.009202453987730062,
+      "grad_norm": 0.39890092611312866,
+      "learning_rate": 0.00013846153846153845,
+      "loss": 1.7028,
+      "step": 60
+    },
+    {
+      "epoch": 0.009355828220858895,
+      "grad_norm": 0.39110666513442993,
+      "learning_rate": 0.00014076923076923074,
+      "loss": 1.7618,
+      "step": 61
+    },
+    {
+      "epoch": 0.00950920245398773,
+      "grad_norm": 0.3864617645740509,
+      "learning_rate": 0.00014307692307692307,
+      "loss": 1.795,
+      "step": 62
+    },
+    {
+      "epoch": 0.009662576687116564,
+      "grad_norm": 0.43441009521484375,
+      "learning_rate": 0.00014538461538461537,
+      "loss": 1.7232,
+      "step": 63
+    },
+    {
+      "epoch": 0.0098159509202454,
+      "grad_norm": 0.4279692769050598,
+      "learning_rate": 0.0001476923076923077,
+      "loss": 1.7323,
+      "step": 64
+    },
+    {
+      "epoch": 0.009969325153374233,
+      "grad_norm": 0.4381473660469055,
+      "learning_rate": 0.00015,
+      "loss": 1.8478,
+      "step": 65
+    },
+    {
+      "epoch": 0.010122699386503068,
+      "grad_norm": 0.4402051866054535,
+      "learning_rate": 0.0001499999911174326,
+      "loss": 1.7174,
+      "step": 66
+    },
+    {
+      "epoch": 0.010276073619631901,
+      "grad_norm": 0.49763572216033936,
+      "learning_rate": 0.0001499999644697325,
+      "loss": 1.7212,
+      "step": 67
+    },
+    {
+      "epoch": 0.010429447852760737,
+      "grad_norm": 0.5003432035446167,
+      "learning_rate": 0.00014999992005690604,
+      "loss": 1.7893,
+      "step": 68
+    },
+    {
+      "epoch": 0.01058282208588957,
+      "grad_norm": 0.42017924785614014,
+      "learning_rate": 0.00014999985787896375,
+      "loss": 1.7918,
+      "step": 69
+    },
+    {
+      "epoch": 0.010736196319018405,
+      "grad_norm": 0.4818103611469269,
+      "learning_rate": 0.0001499997779359203,
+      "loss": 1.785,
+      "step": 70
+    },
+    {
+      "epoch": 0.010889570552147239,
+      "grad_norm": 0.5177863240242004,
+      "learning_rate": 0.00014999968022779466,
+      "loss": 1.7916,
+      "step": 71
+    },
+    {
+      "epoch": 0.011042944785276074,
+      "grad_norm": 0.37934622168540955,
+      "learning_rate": 0.00014999956475461,
+      "loss": 1.7589,
+      "step": 72
+    },
+    {
+      "epoch": 0.011196319018404908,
+      "grad_norm": 0.39523065090179443,
+      "learning_rate": 0.0001499994315163936,
+      "loss": 1.8217,
+      "step": 73
+    },
+    {
+      "epoch": 0.011349693251533743,
+      "grad_norm": 0.3827604055404663,
+      "learning_rate": 0.0001499992805131771,
+      "loss": 1.7992,
+      "step": 74
+    },
+    {
+      "epoch": 0.011503067484662576,
+      "grad_norm": 0.35017895698547363,
+      "learning_rate": 0.00014999911174499623,
+      "loss": 1.7357,
+      "step": 75
+    },
+    {
+      "epoch": 0.011656441717791411,
+      "grad_norm": 0.8097590208053589,
+      "learning_rate": 0.00014999892521189097,
+      "loss": 1.8286,
+      "step": 76
+    },
+    {
+      "epoch": 0.011809815950920245,
+      "grad_norm": 0.39429736137390137,
+      "learning_rate": 0.00014999872091390549,
+      "loss": 1.7896,
+      "step": 77
+    },
+    {
+      "epoch": 0.01196319018404908,
+      "grad_norm": 0.37513309717178345,
+      "learning_rate": 0.0001499984988510882,
+      "loss": 1.7518,
+      "step": 78
+    },
+    {
+      "epoch": 0.012116564417177914,
+      "grad_norm": 0.3957892954349518,
+      "learning_rate": 0.00014999825902349168,
+      "loss": 1.7365,
+      "step": 79
+    },
+    {
+      "epoch": 0.012269938650306749,
+      "grad_norm": 0.3707665801048279,
+      "learning_rate": 0.00014999800143117276,
+      "loss": 1.7204,
+      "step": 80
+    },
+    {
+      "epoch": 0.012423312883435582,
+      "grad_norm": 0.3641624450683594,
+      "learning_rate": 0.00014999772607419246,
+      "loss": 1.7719,
+      "step": 81
+    },
+    {
+      "epoch": 0.012576687116564417,
+      "grad_norm": 0.36411935091018677,
+      "learning_rate": 0.000149997432952616,
+      "loss": 1.7657,
+      "step": 82
+    },
+    {
+      "epoch": 0.012730061349693251,
+      "grad_norm": 0.3827705383300781,
+      "learning_rate": 0.00014999712206651276,
+      "loss": 1.7577,
+      "step": 83
+    },
+    {
+      "epoch": 0.012883435582822086,
+      "grad_norm": 0.3992520272731781,
+      "learning_rate": 0.00014999679341595645,
+      "loss": 1.736,
+      "step": 84
+    },
+    {
+      "epoch": 0.01303680981595092,
+      "grad_norm": 0.4533821940422058,
+      "learning_rate": 0.00014999644700102488,
+      "loss": 1.7884,
+      "step": 85
+    },
+    {
+      "epoch": 0.013190184049079755,
+      "grad_norm": 0.39199310541152954,
+      "learning_rate": 0.00014999608282180012,
+      "loss": 1.749,
+      "step": 86
+    },
+    {
+      "epoch": 0.013343558282208588,
+      "grad_norm": 0.37758469581604004,
+      "learning_rate": 0.0001499957008783684,
+      "loss": 1.8069,
+      "step": 87
+    },
+    {
+      "epoch": 0.013496932515337423,
+      "grad_norm": 0.3737950026988983,
+      "learning_rate": 0.00014999530117082023,
+      "loss": 1.7691,
+      "step": 88
+    },
+    {
+      "epoch": 0.013650306748466257,
+      "grad_norm": 0.3872266709804535,
+      "learning_rate": 0.00014999488369925028,
+      "loss": 1.7746,
+      "step": 89
+    },
+    {
+      "epoch": 0.013803680981595092,
+      "grad_norm": 0.35836392641067505,
+      "learning_rate": 0.00014999444846375742,
+      "loss": 1.7622,
+      "step": 90
+    },
+    {
+      "epoch": 0.013957055214723926,
+      "grad_norm": 0.37322524189949036,
+      "learning_rate": 0.00014999399546444477,
+      "loss": 1.6497,
+      "step": 91
+    },
+    {
+      "epoch": 0.01411042944785276,
+      "grad_norm": 0.446819543838501,
+      "learning_rate": 0.0001499935247014196,
+      "loss": 1.7638,
+      "step": 92
+    },
+    {
+      "epoch": 0.014263803680981594,
+      "grad_norm": 0.41678449511528015,
+      "learning_rate": 0.00014999303617479342,
+      "loss": 1.7239,
+      "step": 93
+    },
+    {
+      "epoch": 0.01441717791411043,
+      "grad_norm": 0.3500819504261017,
+      "learning_rate": 0.00014999252988468197,
+      "loss": 1.7241,
+      "step": 94
+    },
+    {
+      "epoch": 0.014570552147239263,
+      "grad_norm": 0.37793272733688354,
+      "learning_rate": 0.0001499920058312052,
+      "loss": 1.757,
+      "step": 95
+    },
+    {
+      "epoch": 0.014723926380368098,
+      "grad_norm": 0.3865213692188263,
+      "learning_rate": 0.00014999146401448716,
+      "loss": 1.7314,
+      "step": 96
+    },
+    {
+      "epoch": 0.014877300613496933,
+      "grad_norm": 0.3815392851829529,
+      "learning_rate": 0.00014999090443465623,
+      "loss": 1.7154,
+      "step": 97
+    },
+    {
+      "epoch": 0.015030674846625767,
+      "grad_norm": 0.36950916051864624,
+      "learning_rate": 0.000149990327091845,
+      "loss": 1.7868,
+      "step": 98
+    },
+    {
+      "epoch": 0.015184049079754602,
+      "grad_norm": 0.3632974624633789,
+      "learning_rate": 0.00014998973198619014,
+      "loss": 1.7254,
+      "step": 99
+    },
+    {
+      "epoch": 0.015337423312883436,
+      "grad_norm": 0.3812977969646454,
+      "learning_rate": 0.0001499891191178327,
+      "loss": 1.7361,
+      "step": 100
+    },
+    {
+      "epoch": 0.01549079754601227,
+      "grad_norm": 0.3810684382915497,
+      "learning_rate": 0.0001499884884869178,
+      "loss": 1.7317,
+      "step": 101
+    },
+    {
+      "epoch": 0.015644171779141104,
+      "grad_norm": 0.3631119132041931,
+      "learning_rate": 0.00014998784009359479,
+      "loss": 1.7379,
+      "step": 102
+    },
+    {
+      "epoch": 0.015797546012269938,
+      "grad_norm": 0.3660447299480438,
+      "learning_rate": 0.00014998717393801734,
+      "loss": 1.6887,
+      "step": 103
+    },
+    {
+      "epoch": 0.015950920245398775,
+      "grad_norm": 0.36023271083831787,
+      "learning_rate": 0.00014998649002034313,
+      "loss": 1.7343,
+      "step": 104
+    },
+    {
+      "epoch": 0.016104294478527608,
+      "grad_norm": 0.3579729199409485,
+      "learning_rate": 0.00014998578834073424,
+      "loss": 1.7795,
+      "step": 105
+    },
+    {
+      "epoch": 0.01625766871165644,
+      "grad_norm": 0.36055469512939453,
+      "learning_rate": 0.00014998506889935686,
+      "loss": 1.7447,
+      "step": 106
+    },
+    {
+      "epoch": 0.016411042944785275,
+      "grad_norm": 0.376589298248291,
+      "learning_rate": 0.0001499843316963814,
+      "loss": 1.7043,
+      "step": 107
+    },
+    {
+      "epoch": 0.016564417177914112,
+      "grad_norm": 0.4640180468559265,
+      "learning_rate": 0.00014998357673198247,
+      "loss": 1.6977,
+      "step": 108
+    },
+    {
+      "epoch": 0.016717791411042945,
+      "grad_norm": 0.3960256278514862,
+      "learning_rate": 0.0001499828040063389,
+      "loss": 1.68,
+      "step": 109
+    },
+    {
+      "epoch": 0.01687116564417178,
+      "grad_norm": 0.41461309790611267,
+      "learning_rate": 0.00014998201351963374,
+      "loss": 1.7206,
+      "step": 110
+    },
+    {
+      "epoch": 0.017024539877300612,
+      "grad_norm": 0.3750212490558624,
+      "learning_rate": 0.0001499812052720542,
+      "loss": 1.7451,
+      "step": 111
+    },
+    {
+      "epoch": 0.01717791411042945,
+      "grad_norm": 0.34901657700538635,
+      "learning_rate": 0.00014998037926379176,
+      "loss": 1.7414,
+      "step": 112
+    },
+    {
+      "epoch": 0.017331288343558283,
+      "grad_norm": 0.5079634189605713,
+      "learning_rate": 0.00014997953549504206,
+      "loss": 1.689,
+      "step": 113
+    },
+    {
+      "epoch": 0.017484662576687116,
+      "grad_norm": 0.4373054504394531,
+      "learning_rate": 0.00014997867396600496,
+      "loss": 1.6263,
+      "step": 114
+    },
+    {
+      "epoch": 0.01763803680981595,
+      "grad_norm": 0.37907862663269043,
+      "learning_rate": 0.00014997779467688452,
+      "loss": 1.7365,
+      "step": 115
+    },
+    {
+      "epoch": 0.017791411042944787,
+      "grad_norm": 0.37847113609313965,
+      "learning_rate": 0.00014997689762788905,
+      "loss": 1.6808,
+      "step": 116
+    },
+    {
+      "epoch": 0.01794478527607362,
+      "grad_norm": 0.4039837121963501,
+      "learning_rate": 0.00014997598281923103,
+      "loss": 1.6876,
+      "step": 117
+    },
+    {
+      "epoch": 0.018098159509202454,
+      "grad_norm": 0.3757103383541107,
+      "learning_rate": 0.0001499750502511271,
+      "loss": 1.6944,
+      "step": 118
+    },
+    {
+      "epoch": 0.018251533742331287,
+      "grad_norm": 0.3628157675266266,
+      "learning_rate": 0.0001499740999237982,
+      "loss": 1.7242,
+      "step": 119
+    },
+    {
+      "epoch": 0.018404907975460124,
+      "grad_norm": 0.36390653252601624,
+      "learning_rate": 0.00014997313183746943,
+      "loss": 1.7077,
+      "step": 120
+    },
+    {
+      "epoch": 0.018558282208588957,
+      "grad_norm": 0.4106927216053009,
+      "learning_rate": 0.00014997214599237006,
+      "loss": 1.7515,
+      "step": 121
+    },
+    {
+      "epoch": 0.01871165644171779,
+      "grad_norm": 0.39268365502357483,
+      "learning_rate": 0.00014997114238873366,
+      "loss": 1.7139,
+      "step": 122
+    },
+    {
+      "epoch": 0.018865030674846624,
+      "grad_norm": 0.3702599108219147,
+      "learning_rate": 0.0001499701210267979,
+      "loss": 1.6837,
+      "step": 123
+    },
+    {
+      "epoch": 0.01901840490797546,
+      "grad_norm": 0.4021584689617157,
+      "learning_rate": 0.00014996908190680478,
+      "loss": 1.6914,
+      "step": 124
+    },
+    {
+      "epoch": 0.019171779141104295,
+      "grad_norm": 0.371452271938324,
+      "learning_rate": 0.00014996802502900035,
+      "loss": 1.6912,
+      "step": 125
+    },
+    {
+      "epoch": 0.01932515337423313,
+      "grad_norm": 0.35980576276779175,
+      "learning_rate": 0.00014996695039363501,
+      "loss": 1.7333,
+      "step": 126
+    },
+    {
+      "epoch": 0.019478527607361962,
+      "grad_norm": 0.3630543351173401,
+      "learning_rate": 0.00014996585800096328,
+      "loss": 1.7041,
+      "step": 127
+    },
+    {
+      "epoch": 0.0196319018404908,
+      "grad_norm": 0.3905601501464844,
+      "learning_rate": 0.00014996474785124392,
+      "loss": 1.7007,
+      "step": 128
+    },
+    {
+      "epoch": 0.019785276073619632,
+      "grad_norm": 0.382161945104599,
+      "learning_rate": 0.0001499636199447399,
+      "loss": 1.5833,
+      "step": 129
+    },
+    {
+      "epoch": 0.019938650306748466,
+      "grad_norm": 0.392764687538147,
+      "learning_rate": 0.00014996247428171838,
+      "loss": 1.7209,
+      "step": 130
+    },
+    {
+      "epoch": 0.0200920245398773,
+      "grad_norm": 0.3893950581550598,
+      "learning_rate": 0.0001499613108624507,
+      "loss": 1.6965,
+      "step": 131
+    },
+    {
+      "epoch": 0.020245398773006136,
+      "grad_norm": 0.3623042106628418,
+      "learning_rate": 0.0001499601296872125,
+      "loss": 1.6457,
+      "step": 132
+    },
+    {
+      "epoch": 0.02039877300613497,
+      "grad_norm": 0.3802337348461151,
+      "learning_rate": 0.00014995893075628352,
+      "loss": 1.59,
+      "step": 133
+    },
+    {
+      "epoch": 0.020552147239263803,
+      "grad_norm": 0.42636531591415405,
+      "learning_rate": 0.00014995771406994777,
+      "loss": 1.7366,
+      "step": 134
+    },
+    {
+      "epoch": 0.020705521472392636,
+      "grad_norm": 0.35441824793815613,
+      "learning_rate": 0.00014995647962849341,
+      "loss": 1.7291,
+      "step": 135
+    },
+    {
+      "epoch": 0.020858895705521473,
+      "grad_norm": 0.36408284306526184,
+      "learning_rate": 0.00014995522743221286,
+      "loss": 1.6469,
+      "step": 136
+    },
+    {
+      "epoch": 0.021012269938650307,
+      "grad_norm": 0.38769856095314026,
+      "learning_rate": 0.00014995395748140274,
+      "loss": 1.6942,
+      "step": 137
+    },
+    {
+      "epoch": 0.02116564417177914,
+      "grad_norm": 0.38913822174072266,
+      "learning_rate": 0.00014995266977636384,
+      "loss": 1.7446,
+      "step": 138
+    },
+    {
+      "epoch": 0.021319018404907974,
+      "grad_norm": 0.4078815281391144,
+      "learning_rate": 0.0001499513643174012,
+      "loss": 1.7533,
+      "step": 139
+    },
+    {
+      "epoch": 0.02147239263803681,
+      "grad_norm": 0.38553088903427124,
+      "learning_rate": 0.000149950041104824,
+      "loss": 1.7261,
+      "step": 140
+    },
+    {
+      "epoch": 0.021625766871165644,
+      "grad_norm": 0.3599337935447693,
+      "learning_rate": 0.00014994870013894574,
+      "loss": 1.6358,
+      "step": 141
+    },
+    {
+      "epoch": 0.021779141104294478,
+      "grad_norm": 0.4201655685901642,
+      "learning_rate": 0.00014994734142008398,
+      "loss": 1.5583,
+      "step": 142
+    },
+    {
+      "epoch": 0.02193251533742331,
+      "grad_norm": 0.3919838070869446,
+      "learning_rate": 0.00014994596494856058,
+      "loss": 1.6705,
+      "step": 143
+    },
+    {
+      "epoch": 0.022085889570552148,
+      "grad_norm": 0.5253005027770996,
+      "learning_rate": 0.0001499445707247016,
+      "loss": 1.6681,
+      "step": 144
+    },
+    {
+      "epoch": 0.02223926380368098,
+      "grad_norm": 0.5169997811317444,
+      "learning_rate": 0.0001499431587488373,
+      "loss": 1.6749,
+      "step": 145
+    },
+    {
+      "epoch": 0.022392638036809815,
+      "grad_norm": 0.43623775243759155,
+      "learning_rate": 0.00014994172902130207,
+      "loss": 1.6605,
+      "step": 146
+    },
+    {
+      "epoch": 0.022546012269938652,
+      "grad_norm": 0.4605841338634491,
+      "learning_rate": 0.00014994028154243462,
+      "loss": 1.6781,
+      "step": 147
+    },
+    {
+      "epoch": 0.022699386503067485,
+      "grad_norm": 0.519533097743988,
+      "learning_rate": 0.00014993881631257777,
+      "loss": 1.6724,
+      "step": 148
+    },
+    {
+      "epoch": 0.02285276073619632,
+      "grad_norm": 0.4139060080051422,
+      "learning_rate": 0.00014993733333207865,
+      "loss": 1.6773,
+      "step": 149
+    },
+    {
+      "epoch": 0.023006134969325152,
+      "grad_norm": 0.3932259678840637,
+      "learning_rate": 0.0001499358326012885,
+      "loss": 1.5962,
+      "step": 150
+    },
+    {
+      "epoch": 0.02315950920245399,
+      "grad_norm": 0.8100685477256775,
+      "learning_rate": 0.0001499343141205628,
+      "loss": 1.6741,
+      "step": 151
+    },
+    {
+      "epoch": 0.023312883435582823,
+      "grad_norm": 0.6980057954788208,
+      "learning_rate": 0.00014993277789026118,
+      "loss": 1.7353,
+      "step": 152
+    },
+    {
+      "epoch": 0.023466257668711656,
+      "grad_norm": 0.536426842212677,
+      "learning_rate": 0.0001499312239107476,
+      "loss": 1.7074,
+      "step": 153
+    },
+    {
+      "epoch": 0.02361963190184049,
+      "grad_norm": 0.4194886088371277,
+      "learning_rate": 0.0001499296521823901,
+      "loss": 1.6419,
+      "step": 154
+    },
+    {
+      "epoch": 0.023773006134969327,
+      "grad_norm": 0.42030543088912964,
+      "learning_rate": 0.000149928062705561,
+      "loss": 1.7288,
+      "step": 155
+    },
+    {
+      "epoch": 0.02392638036809816,
+      "grad_norm": 0.42655694484710693,
+      "learning_rate": 0.00014992645548063678,
+      "loss": 1.7153,
+      "step": 156
+    },
+    {
+      "epoch": 0.024079754601226994,
+      "grad_norm": 0.6100133061408997,
+      "learning_rate": 0.00014992483050799813,
+      "loss": 1.6224,
+      "step": 157
+    },
+    {
+      "epoch": 0.024233128834355827,
+      "grad_norm": 0.390532910823822,
+      "learning_rate": 0.00014992318778803,
+      "loss": 1.7034,
+      "step": 158
+    },
+    {
+      "epoch": 0.024386503067484664,
+      "grad_norm": 0.41450461745262146,
+      "learning_rate": 0.00014992152732112144,
+      "loss": 1.6717,
+      "step": 159
+    },
+    {
+      "epoch": 0.024539877300613498,
+      "grad_norm": 0.3604941964149475,
+      "learning_rate": 0.0001499198491076658,
+      "loss": 1.6595,
+      "step": 160
+    },
+    {
+      "epoch": 0.02469325153374233,
+      "grad_norm": 0.3706038296222687,
+      "learning_rate": 0.0001499181531480606,
+      "loss": 1.716,
+      "step": 161
+    },
+    {
+      "epoch": 0.024846625766871164,
+      "grad_norm": 0.35165998339653015,
+      "learning_rate": 0.00014991643944270756,
+      "loss": 1.7243,
+      "step": 162
+    },
+    {
+      "epoch": 0.025,
+      "grad_norm": 0.35479745268821716,
+      "learning_rate": 0.00014991470799201255,
+      "loss": 1.6664,
+      "step": 163
+    },
+    {
+      "epoch": 0.025153374233128835,
+      "grad_norm": 0.34846070408821106,
+      "learning_rate": 0.00014991295879638578,
+      "loss": 1.7091,
+      "step": 164
+    },
+    {
+      "epoch": 0.02530674846625767,
+      "grad_norm": 0.3654356896877289,
+      "learning_rate": 0.0001499111918562415,
+      "loss": 1.6044,
+      "step": 165
+    },
+    {
+      "epoch": 0.025460122699386502,
+      "grad_norm": 0.37394317984580994,
+      "learning_rate": 0.00014990940717199828,
+      "loss": 1.5668,
+      "step": 166
+    },
+    {
+      "epoch": 0.02561349693251534,
+      "grad_norm": 0.41581806540489197,
+      "learning_rate": 0.00014990760474407888,
+      "loss": 1.6678,
+      "step": 167
+    },
+    {
+      "epoch": 0.025766871165644172,
+      "grad_norm": 0.38020819425582886,
+      "learning_rate": 0.00014990578457291018,
+      "loss": 1.5945,
+      "step": 168
+    },
+    {
+      "epoch": 0.025920245398773006,
+      "grad_norm": 0.4416181743144989,
+      "learning_rate": 0.00014990394665892337,
+      "loss": 1.5818,
+      "step": 169
+    },
+    {
+      "epoch": 0.02607361963190184,
+      "grad_norm": 0.42521026730537415,
+      "learning_rate": 0.00014990209100255376,
+      "loss": 1.6327,
+      "step": 170
+    },
+    {
+      "epoch": 0.026226993865030676,
+      "grad_norm": 0.3844924569129944,
+      "learning_rate": 0.00014990021760424095,
+      "loss": 1.6172,
+      "step": 171
+    },
+    {
+      "epoch": 0.02638036809815951,
+      "grad_norm": 0.43099164962768555,
+      "learning_rate": 0.0001498983264644286,
+      "loss": 1.6216,
+      "step": 172
+    },
+    {
+      "epoch": 0.026533742331288343,
+      "grad_norm": 0.38728269934654236,
+      "learning_rate": 0.00014989641758356476,
+      "loss": 1.6963,
+      "step": 173
+    },
+    {
+      "epoch": 0.026687116564417177,
+      "grad_norm": 0.3995940685272217,
+      "learning_rate": 0.0001498944909621015,
+      "loss": 1.6563,
+      "step": 174
+    },
+    {
+      "epoch": 0.026840490797546013,
+      "grad_norm": 0.40860000252723694,
+      "learning_rate": 0.00014989254660049525,
+      "loss": 1.5398,
+      "step": 175
+    },
+    {
+      "epoch": 0.026993865030674847,
+      "grad_norm": 0.5215296745300293,
+      "learning_rate": 0.0001498905844992065,
+      "loss": 1.6312,
+      "step": 176
+    },
+    {
+      "epoch": 0.02714723926380368,
+      "grad_norm": 0.5953673124313354,
+      "learning_rate": 0.00014988860465870008,
+      "loss": 1.6027,
+      "step": 177
+    },
+    {
+      "epoch": 0.027300613496932514,
+      "grad_norm": 0.5164670944213867,
+      "learning_rate": 0.00014988660707944485,
+      "loss": 1.6295,
+      "step": 178
+    },
+    {
+      "epoch": 0.02745398773006135,
+      "grad_norm": 0.4317152500152588,
+      "learning_rate": 0.00014988459176191405,
+      "loss": 1.6667,
+      "step": 179
+    },
+    {
+      "epoch": 0.027607361963190184,
+      "grad_norm": 0.4172368347644806,
+      "learning_rate": 0.00014988255870658506,
+      "loss": 1.6452,
+      "step": 180
+    },
+    {
+      "epoch": 0.027760736196319018,
+      "grad_norm": 0.38960379362106323,
+      "learning_rate": 0.00014988050791393942,
+      "loss": 1.6533,
+      "step": 181
+    },
+    {
+      "epoch": 0.02791411042944785,
+      "grad_norm": 0.38012033700942993,
+      "learning_rate": 0.00014987843938446289,
+      "loss": 1.7115,
+      "step": 182
+    },
+    {
+      "epoch": 0.028067484662576688,
+      "grad_norm": 0.40224993228912354,
+      "learning_rate": 0.00014987635311864542,
+      "loss": 1.7212,
+      "step": 183
+    },
+    {
+      "epoch": 0.02822085889570552,
+      "grad_norm": 0.661195695400238,
+      "learning_rate": 0.00014987424911698123,
+      "loss": 1.5703,
+      "step": 184
+    },
+    {
+      "epoch": 0.028374233128834355,
+      "grad_norm": 0.5118616223335266,
+      "learning_rate": 0.00014987212737996866,
+      "loss": 1.6586,
+      "step": 185
+    },
+    {
+      "epoch": 0.02852760736196319,
+      "grad_norm": 0.4334467947483063,
+      "learning_rate": 0.00014986998790811028,
+      "loss": 1.5921,
+      "step": 186
+    },
+    {
+      "epoch": 0.028680981595092025,
+      "grad_norm": 0.6972788572311401,
+      "learning_rate": 0.00014986783070191292,
+      "loss": 1.6917,
+      "step": 187
+    },
+    {
+      "epoch": 0.02883435582822086,
+      "grad_norm": 0.5296037793159485,
+      "learning_rate": 0.00014986565576188745,
+      "loss": 1.6756,
+      "step": 188
+    },
+    {
+      "epoch": 0.028987730061349692,
+      "grad_norm": 0.36645233631134033,
+      "learning_rate": 0.00014986346308854914,
+      "loss": 1.7073,
+      "step": 189
+    },
+    {
+      "epoch": 0.029141104294478526,
+      "grad_norm": 0.3867458701133728,
+      "learning_rate": 0.0001498612526824173,
+      "loss": 1.6963,
+      "step": 190
+    },
+    {
+      "epoch": 0.029294478527607363,
+      "grad_norm": 0.3974872827529907,
+      "learning_rate": 0.00014985902454401555,
+      "loss": 1.6667,
+      "step": 191
+    },
+    {
+      "epoch": 0.029447852760736196,
+      "grad_norm": 0.36978384852409363,
+      "learning_rate": 0.00014985677867387163,
+      "loss": 1.6434,
+      "step": 192
+    },
+    {
+      "epoch": 0.02960122699386503,
+      "grad_norm": 0.4090369939804077,
+      "learning_rate": 0.00014985451507251756,
+      "loss": 1.6591,
+      "step": 193
+    },
+    {
+      "epoch": 0.029754601226993867,
+      "grad_norm": 0.4237601161003113,
+      "learning_rate": 0.00014985223374048946,
+      "loss": 1.6849,
+      "step": 194
+    },
+    {
+      "epoch": 0.0299079754601227,
+      "grad_norm": 0.4409900903701782,
+      "learning_rate": 0.00014984993467832776,
+      "loss": 1.5557,
+      "step": 195
+    },
+    {
+      "epoch": 0.030061349693251534,
+      "grad_norm": 0.47145578265190125,
+      "learning_rate": 0.00014984761788657702,
+      "loss": 1.6174,
+      "step": 196
+    },
+    {
+      "epoch": 0.030214723926380367,
+      "grad_norm": 0.40429362654685974,
+      "learning_rate": 0.00014984528336578598,
+      "loss": 1.5526,
+      "step": 197
+    },
+    {
+      "epoch": 0.030368098159509204,
+      "grad_norm": 0.4035743772983551,
+      "learning_rate": 0.00014984293111650765,
+      "loss": 1.5603,
+      "step": 198
+    },
+    {
+      "epoch": 0.030521472392638038,
+      "grad_norm": 0.5827813148498535,
+      "learning_rate": 0.00014984056113929919,
+      "loss": 1.5509,
+      "step": 199
+    },
+    {
+      "epoch": 0.03067484662576687,
+      "grad_norm": 0.46248817443847656,
+      "learning_rate": 0.00014983817343472197,
+      "loss": 1.649,
+      "step": 200
+    },
+    {
+      "epoch": 0.030828220858895704,
+      "grad_norm": 0.4427696764469147,
+      "learning_rate": 0.0001498357680033416,
+      "loss": 1.5689,
+      "step": 201
+    },
+    {
+      "epoch": 0.03098159509202454,
+      "grad_norm": 0.4999493658542633,
+      "learning_rate": 0.0001498333448457278,
+      "loss": 1.5961,
+      "step": 202
+    },
+    {
+      "epoch": 0.031134969325153375,
+      "grad_norm": 0.4135405421257019,
+      "learning_rate": 0.00014983090396245455,
+      "loss": 1.5909,
+      "step": 203
+    },
+    {
+      "epoch": 0.03128834355828221,
+      "grad_norm": 0.46542811393737793,
+      "learning_rate": 0.00014982844535410002,
+      "loss": 1.6512,
+      "step": 204
+    },
+    {
+      "epoch": 0.03144171779141104,
+      "grad_norm": 0.4624110460281372,
+      "learning_rate": 0.00014982596902124662,
+      "loss": 1.5765,
+      "step": 205
+    },
+    {
+      "epoch": 0.031595092024539875,
+      "grad_norm": 0.4685991704463959,
+      "learning_rate": 0.00014982347496448085,
+      "loss": 1.6481,
+      "step": 206
+    },
+    {
+      "epoch": 0.03174846625766871,
+      "grad_norm": 0.4442425072193146,
+      "learning_rate": 0.00014982096318439352,
+      "loss": 1.581,
+      "step": 207
+    },
+    {
+      "epoch": 0.03190184049079755,
+      "grad_norm": 0.38967493176460266,
+      "learning_rate": 0.00014981843368157957,
+      "loss": 1.5544,
+      "step": 208
+    },
+    {
+      "epoch": 0.03205521472392638,
+      "grad_norm": 0.7504474520683289,
+      "learning_rate": 0.00014981588645663816,
+      "loss": 1.4903,
+      "step": 209
+    },
+    {
+      "epoch": 0.032208588957055216,
+      "grad_norm": 0.4460054337978363,
+      "learning_rate": 0.00014981332151017265,
+      "loss": 1.5704,
+      "step": 210
+    },
+    {
+      "epoch": 0.03236196319018405,
+      "grad_norm": 0.3914797902107239,
+      "learning_rate": 0.00014981073884279058,
+      "loss": 1.6078,
+      "step": 211
+    },
+    {
+      "epoch": 0.03251533742331288,
+      "grad_norm": 0.3677014112472534,
+      "learning_rate": 0.00014980813845510375,
+      "loss": 1.6922,
+      "step": 212
+    },
+    {
+      "epoch": 0.03266871165644172,
+      "grad_norm": 0.392487496137619,
+      "learning_rate": 0.00014980552034772807,
+      "loss": 1.6631,
+      "step": 213
+    },
+    {
+      "epoch": 0.03282208588957055,
+      "grad_norm": 0.3803916275501251,
+      "learning_rate": 0.00014980288452128368,
+      "loss": 1.5205,
+      "step": 214
+    },
+    {
+      "epoch": 0.032975460122699383,
+      "grad_norm": 0.3823797404766083,
+      "learning_rate": 0.00014980023097639495,
+      "loss": 1.5371,
+      "step": 215
+    },
+    {
+      "epoch": 0.033128834355828224,
+      "grad_norm": 0.6218409538269043,
+      "learning_rate": 0.00014979755971369042,
+      "loss": 1.5377,
+      "step": 216
+    },
+    {
+      "epoch": 0.03328220858895706,
+      "grad_norm": 0.4620639383792877,
+      "learning_rate": 0.0001497948707338028,
+      "loss": 1.4432,
+      "step": 217
+    },
+    {
+      "epoch": 0.03343558282208589,
+      "grad_norm": 0.743587076663971,
+      "learning_rate": 0.00014979216403736907,
+      "loss": 1.5582,
+      "step": 218
+    },
+    {
+      "epoch": 0.033588957055214724,
+      "grad_norm": 0.7575177550315857,
+      "learning_rate": 0.0001497894396250303,
+      "loss": 1.556,
+      "step": 219
+    },
+    {
+      "epoch": 0.03374233128834356,
+      "grad_norm": 0.4221208095550537,
+      "learning_rate": 0.00014978669749743186,
+      "loss": 1.573,
+      "step": 220
+    },
+    {
+      "epoch": 0.03389570552147239,
+      "grad_norm": 0.4367414712905884,
+      "learning_rate": 0.0001497839376552233,
+      "loss": 1.4037,
+      "step": 221
+    },
+    {
+      "epoch": 0.034049079754601225,
+      "grad_norm": 0.5025057196617126,
+      "learning_rate": 0.00014978116009905826,
+      "loss": 1.5192,
+      "step": 222
+    },
+    {
+      "epoch": 0.03420245398773006,
+      "grad_norm": 0.4812071621417999,
+      "learning_rate": 0.00014977836482959472,
+      "loss": 1.5068,
+      "step": 223
+    },
+    {
+      "epoch": 0.0343558282208589,
+      "grad_norm": 0.46394768357276917,
+      "learning_rate": 0.00014977555184749478,
+      "loss": 1.5647,
+      "step": 224
+    },
+    {
+      "epoch": 0.03450920245398773,
+      "grad_norm": 0.4634012281894684,
+      "learning_rate": 0.0001497727211534247,
+      "loss": 1.5623,
+      "step": 225
+    },
+    {
+      "epoch": 0.034662576687116566,
+      "grad_norm": 0.4229913055896759,
+      "learning_rate": 0.00014976987274805507,
+      "loss": 1.5392,
+      "step": 226
+    },
+    {
+      "epoch": 0.0348159509202454,
+      "grad_norm": 0.6625610589981079,
+      "learning_rate": 0.0001497670066320605,
+      "loss": 1.5756,
+      "step": 227
+    },
+    {
+      "epoch": 0.03496932515337423,
+      "grad_norm": 0.6085181832313538,
+      "learning_rate": 0.00014976412280611994,
+      "loss": 1.53,
+      "step": 228
+    },
+    {
+      "epoch": 0.035122699386503066,
+      "grad_norm": 0.428016722202301,
+      "learning_rate": 0.00014976122127091643,
+      "loss": 1.5722,
+      "step": 229
+    },
+    {
+      "epoch": 0.0352760736196319,
+      "grad_norm": 0.4814985394477844,
+      "learning_rate": 0.00014975830202713732,
+      "loss": 1.4893,
+      "step": 230
+    },
+    {
+      "epoch": 0.03542944785276073,
+      "grad_norm": 0.5289542078971863,
+      "learning_rate": 0.00014975536507547403,
+      "loss": 1.6689,
+      "step": 231
+    },
+    {
+      "epoch": 0.03558282208588957,
+      "grad_norm": 0.46005043387413025,
+      "learning_rate": 0.00014975241041662225,
+      "loss": 1.5475,
+      "step": 232
+    },
+    {
+      "epoch": 0.03573619631901841,
+      "grad_norm": 0.48661136627197266,
+      "learning_rate": 0.00014974943805128182,
+      "loss": 1.5182,
+      "step": 233
+    },
+    {
+      "epoch": 0.03588957055214724,
+      "grad_norm": 0.44411274790763855,
+      "learning_rate": 0.00014974644798015685,
+      "loss": 1.5769,
+      "step": 234
+    },
+    {
+      "epoch": 0.036042944785276074,
+      "grad_norm": 0.46200031042099,
+      "learning_rate": 0.00014974344020395557,
+      "loss": 1.6059,
+      "step": 235
+    },
+    {
+      "epoch": 0.03619631901840491,
+      "grad_norm": 0.4868883490562439,
+      "learning_rate": 0.0001497404147233904,
+      "loss": 1.4562,
+      "step": 236
+    },
+    {
+      "epoch": 0.03634969325153374,
+      "grad_norm": 0.56839519739151,
+      "learning_rate": 0.00014973737153917802,
+      "loss": 1.55,
+      "step": 237
+    },
+    {
+      "epoch": 0.036503067484662574,
+      "grad_norm": 0.5236260294914246,
+      "learning_rate": 0.00014973431065203925,
+      "loss": 1.507,
+      "step": 238
+    },
+    {
+      "epoch": 0.03665644171779141,
+      "grad_norm": 0.43564942479133606,
+      "learning_rate": 0.00014973123206269911,
+      "loss": 1.5122,
+      "step": 239
+    },
+    {
+      "epoch": 0.03680981595092025,
+      "grad_norm": 0.44424116611480713,
+      "learning_rate": 0.00014972813577188683,
+      "loss": 1.5795,
+      "step": 240
+    },
+    {
+      "epoch": 0.03696319018404908,
+      "grad_norm": 0.4695936441421509,
+      "learning_rate": 0.00014972502178033583,
+      "loss": 1.5248,
+      "step": 241
+    },
+    {
+      "epoch": 0.037116564417177915,
+      "grad_norm": 0.7540334463119507,
+      "learning_rate": 0.00014972189008878367,
+      "loss": 1.5527,
+      "step": 242
+    },
+    {
+      "epoch": 0.03726993865030675,
+      "grad_norm": 0.610228955745697,
+      "learning_rate": 0.00014971874069797224,
+      "loss": 1.513,
+      "step": 243
+    },
+    {
+      "epoch": 0.03742331288343558,
+      "grad_norm": 0.42700299620628357,
+      "learning_rate": 0.00014971557360864742,
+      "loss": 1.5311,
+      "step": 244
+    },
+    {
+      "epoch": 0.037576687116564415,
+      "grad_norm": 0.5032939314842224,
+      "learning_rate": 0.0001497123888215595,
+      "loss": 1.6114,
+      "step": 245
+    },
+    {
+      "epoch": 0.03773006134969325,
+      "grad_norm": 0.4937465488910675,
+      "learning_rate": 0.00014970918633746278,
+      "loss": 1.5462,
+      "step": 246
+    },
+    {
+      "epoch": 0.03788343558282209,
+      "grad_norm": 0.4553655683994293,
+      "learning_rate": 0.00014970596615711588,
+      "loss": 1.5025,
+      "step": 247
+    },
+    {
+      "epoch": 0.03803680981595092,
+      "grad_norm": 0.4388618469238281,
+      "learning_rate": 0.0001497027282812815,
+      "loss": 1.4359,
+      "step": 248
+    },
+    {
+      "epoch": 0.038190184049079756,
+      "grad_norm": 0.5530393719673157,
+      "learning_rate": 0.00014969947271072665,
+      "loss": 1.586,
+      "step": 249
+    },
+    {
+      "epoch": 0.03834355828220859,
+      "grad_norm": 0.4664992690086365,
+      "learning_rate": 0.00014969619944622242,
+      "loss": 1.5207,
+      "step": 250
+    },
+    {
+      "epoch": 0.03849693251533742,
+      "grad_norm": 0.40434589982032776,
+      "learning_rate": 0.0001496929084885442,
+      "loss": 1.553,
+      "step": 251
+    },
+    {
+      "epoch": 0.03865030674846626,
+      "grad_norm": 0.461453914642334,
+      "learning_rate": 0.00014968959983847148,
+      "loss": 1.5142,
+      "step": 252
+    },
+    {
+      "epoch": 0.03880368098159509,
+      "grad_norm": 0.4253198504447937,
+      "learning_rate": 0.00014968627349678796,
+      "loss": 1.5201,
+      "step": 253
+    },
+    {
+      "epoch": 0.038957055214723924,
+      "grad_norm": 0.5482823252677917,
+      "learning_rate": 0.00014968292946428158,
+      "loss": 1.5218,
+      "step": 254
+    },
+    {
+      "epoch": 0.039110429447852764,
+      "grad_norm": 0.4378732740879059,
+      "learning_rate": 0.0001496795677417444,
+      "loss": 1.5511,
+      "step": 255
+    },
+    {
+      "epoch": 0.0392638036809816,
+      "grad_norm": 0.3832899034023285,
+      "learning_rate": 0.00014967618832997275,
+      "loss": 1.5805,
+      "step": 256
+    },
+    {
+      "epoch": 0.03941717791411043,
+      "grad_norm": 0.4855468273162842,
+      "learning_rate": 0.00014967279122976706,
+      "loss": 1.5317,
+      "step": 257
+    },
+    {
+      "epoch": 0.039570552147239264,
+      "grad_norm": 0.7556911110877991,
+      "learning_rate": 0.00014966937644193202,
+      "loss": 1.4778,
+      "step": 258
+    },
+    {
+      "epoch": 0.0397239263803681,
+      "grad_norm": 0.4173446595668793,
+      "learning_rate": 0.0001496659439672765,
+      "loss": 1.5291,
+      "step": 259
+    },
+    {
+      "epoch": 0.03987730061349693,
+      "grad_norm": 0.5004670023918152,
+      "learning_rate": 0.00014966249380661352,
+      "loss": 1.6987,
+      "step": 260
+    },
+    {
+      "epoch": 0.040030674846625765,
+      "grad_norm": 0.6741068959236145,
+      "learning_rate": 0.00014965902596076033,
+      "loss": 1.5106,
+      "step": 261
+    },
+    {
+      "epoch": 0.0401840490797546,
+      "grad_norm": 0.48477065563201904,
+      "learning_rate": 0.00014965554043053832,
+      "loss": 1.5403,
+      "step": 262
+    },
+    {
+      "epoch": 0.04033742331288344,
+      "grad_norm": 0.5601456165313721,
+      "learning_rate": 0.00014965203721677316,
+      "loss": 1.5555,
+      "step": 263
+    },
+    {
+      "epoch": 0.04049079754601227,
+      "grad_norm": 0.5726425051689148,
+      "learning_rate": 0.0001496485163202946,
+      "loss": 1.4952,
+      "step": 264
+    },
+    {
+      "epoch": 0.040644171779141106,
+      "grad_norm": 0.46574845910072327,
+      "learning_rate": 0.00014964497774193664,
+      "loss": 1.4647,
+      "step": 265
+    },
+    {
+      "epoch": 0.04079754601226994,
+      "grad_norm": 0.4819343090057373,
+      "learning_rate": 0.00014964142148253748,
+      "loss": 1.5013,
+      "step": 266
+    },
+    {
+      "epoch": 0.04095092024539877,
+      "grad_norm": 0.56939297914505,
+      "learning_rate": 0.00014963784754293947,
+      "loss": 1.404,
+      "step": 267
+    },
+    {
+      "epoch": 0.041104294478527606,
+      "grad_norm": 0.8476461172103882,
+      "learning_rate": 0.00014963425592398917,
+      "loss": 1.5369,
+      "step": 268
+    },
+    {
+      "epoch": 0.04125766871165644,
+      "grad_norm": 0.5947362780570984,
+      "learning_rate": 0.0001496306466265373,
+      "loss": 1.4772,
+      "step": 269
+    },
+    {
+      "epoch": 0.04141104294478527,
+      "grad_norm": 0.43048569560050964,
+      "learning_rate": 0.0001496270196514388,
+      "loss": 1.5772,
+      "step": 270
+    },
+    {
+      "epoch": 0.04156441717791411,
+      "grad_norm": 0.44426417350769043,
+      "learning_rate": 0.00014962337499955283,
+      "loss": 1.5896,
+      "step": 271
+    },
+    {
+      "epoch": 0.04171779141104295,
+      "grad_norm": 0.48519203066825867,
+      "learning_rate": 0.00014961971267174262,
+      "loss": 1.5028,
+      "step": 272
+    },
+    {
+      "epoch": 0.04187116564417178,
+      "grad_norm": 0.4280100464820862,
+      "learning_rate": 0.0001496160326688757,
+      "loss": 1.501,
+      "step": 273
+    },
+    {
+      "epoch": 0.042024539877300614,
+      "grad_norm": 0.6907528042793274,
+      "learning_rate": 0.00014961233499182375,
+      "loss": 1.6432,
+      "step": 274
+    },
+    {
+      "epoch": 0.04217791411042945,
+      "grad_norm": 0.7158175110816956,
+      "learning_rate": 0.00014960861964146262,
+      "loss": 1.4861,
+      "step": 275
+    },
+    {
+      "epoch": 0.04233128834355828,
+      "grad_norm": 0.4761661887168884,
+      "learning_rate": 0.00014960488661867237,
+      "loss": 1.4113,
+      "step": 276
+    },
+    {
+      "epoch": 0.042484662576687114,
+      "grad_norm": 0.8205477595329285,
+      "learning_rate": 0.00014960113592433724,
+      "loss": 1.4628,
+      "step": 277
+    },
+    {
+      "epoch": 0.04263803680981595,
+      "grad_norm": 0.5250906348228455,
+      "learning_rate": 0.0001495973675593456,
+      "loss": 1.5287,
+      "step": 278
+    },
+    {
+      "epoch": 0.04279141104294479,
+      "grad_norm": 0.44841963052749634,
+      "learning_rate": 0.00014959358152459012,
+      "loss": 1.5294,
+      "step": 279
+    },
+    {
+      "epoch": 0.04294478527607362,
+      "grad_norm": 0.42856255173683167,
+      "learning_rate": 0.00014958977782096757,
+      "loss": 1.4963,
+      "step": 280
+    },
+    {
+      "epoch": 0.043098159509202455,
+      "grad_norm": 0.48935380578041077,
+      "learning_rate": 0.00014958595644937894,
+      "loss": 1.6323,
+      "step": 281
+    },
+    {
+      "epoch": 0.04325153374233129,
+      "grad_norm": 0.5439465641975403,
+      "learning_rate": 0.00014958211741072932,
+      "loss": 1.6016,
+      "step": 282
+    },
+    {
+      "epoch": 0.04340490797546012,
+      "grad_norm": 0.48032259941101074,
+      "learning_rate": 0.00014957826070592816,
+      "loss": 1.4287,
+      "step": 283
+    },
+    {
+      "epoch": 0.043558282208588955,
+      "grad_norm": 0.9836577773094177,
+      "learning_rate": 0.00014957438633588895,
+      "loss": 1.5377,
+      "step": 284
+    },
+    {
+      "epoch": 0.04371165644171779,
+      "grad_norm": 0.5396369099617004,
+      "learning_rate": 0.00014957049430152942,
+      "loss": 1.5563,
+      "step": 285
+    },
+    {
+      "epoch": 0.04386503067484662,
+      "grad_norm": 0.4108019173145294,
+      "learning_rate": 0.00014956658460377144,
+      "loss": 1.5602,
+      "step": 286
+    },
+    {
+      "epoch": 0.04401840490797546,
+      "grad_norm": 0.4396299421787262,
+      "learning_rate": 0.00014956265724354106,
+      "loss": 1.4588,
+      "step": 287
+    },
+    {
+      "epoch": 0.044171779141104296,
+      "grad_norm": 0.42162081599235535,
+      "learning_rate": 0.00014955871222176864,
+      "loss": 1.4254,
+      "step": 288
+    },
+    {
+      "epoch": 0.04432515337423313,
+      "grad_norm": 0.4663374125957489,
+      "learning_rate": 0.0001495547495393886,
+      "loss": 1.5008,
+      "step": 289
+    },
+    {
+      "epoch": 0.04447852760736196,
+      "grad_norm": 0.43410372734069824,
+      "learning_rate": 0.00014955076919733953,
+      "loss": 1.5396,
+      "step": 290
+    },
+    {
+      "epoch": 0.0446319018404908,
+      "grad_norm": 0.4772189259529114,
+      "learning_rate": 0.0001495467711965643,
+      "loss": 1.4385,
+      "step": 291
+    },
+    {
+      "epoch": 0.04478527607361963,
+      "grad_norm": 0.6041788458824158,
+      "learning_rate": 0.00014954275553800988,
+      "loss": 1.4246,
+      "step": 292
+    },
+    {
+      "epoch": 0.044938650306748464,
+      "grad_norm": 0.4341111183166504,
+      "learning_rate": 0.00014953872222262746,
+      "loss": 1.5266,
+      "step": 293
+    },
+    {
+      "epoch": 0.045092024539877304,
+      "grad_norm": 0.4903193712234497,
+      "learning_rate": 0.00014953467125137243,
+      "loss": 1.4907,
+      "step": 294
+    },
+    {
+      "epoch": 0.04524539877300614,
+      "grad_norm": 0.5634788274765015,
+      "learning_rate": 0.0001495306026252043,
+      "loss": 1.4847,
+      "step": 295
+    },
+    {
+      "epoch": 0.04539877300613497,
+      "grad_norm": 0.4479643404483795,
+      "learning_rate": 0.0001495265163450868,
+      "loss": 1.4735,
+      "step": 296
+    },
+    {
+      "epoch": 0.045552147239263804,
+      "grad_norm": 0.4891916811466217,
+      "learning_rate": 0.00014952241241198788,
+      "loss": 1.5701,
+      "step": 297
+    },
+    {
+      "epoch": 0.04570552147239264,
+      "grad_norm": 0.599760115146637,
+      "learning_rate": 0.0001495182908268796,
+      "loss": 1.4248,
+      "step": 298
+    },
+    {
+      "epoch": 0.04585889570552147,
+      "grad_norm": 0.7414944171905518,
+      "learning_rate": 0.00014951415159073823,
+      "loss": 1.4404,
+      "step": 299
+    },
+    {
+      "epoch": 0.046012269938650305,
+      "grad_norm": 0.5190250873565674,
+      "learning_rate": 0.00014950999470454425,
+      "loss": 1.5104,
+      "step": 300
+    },
+    {
+      "epoch": 0.04616564417177914,
+      "grad_norm": 0.5141550898551941,
+      "learning_rate": 0.00014950582016928226,
+      "loss": 1.4577,
+      "step": 301
+    },
+    {
+      "epoch": 0.04631901840490798,
+      "grad_norm": 0.614512026309967,
+      "learning_rate": 0.0001495016279859411,
+      "loss": 1.4424,
+      "step": 302
+    },
+    {
+      "epoch": 0.04647239263803681,
+      "grad_norm": 0.47473815083503723,
+      "learning_rate": 0.00014949741815551378,
+      "loss": 1.424,
+      "step": 303
+    },
+    {
+      "epoch": 0.046625766871165646,
+      "grad_norm": 0.44575339555740356,
+      "learning_rate": 0.00014949319067899744,
+      "loss": 1.454,
+      "step": 304
+    },
+    {
+      "epoch": 0.04677914110429448,
+      "grad_norm": 0.5251998901367188,
+      "learning_rate": 0.00014948894555739345,
+      "loss": 1.5122,
+      "step": 305
+    },
+    {
+      "epoch": 0.04693251533742331,
+      "grad_norm": 0.521515429019928,
+      "learning_rate": 0.0001494846827917074,
+      "loss": 1.4802,
+      "step": 306
+    },
+    {
+      "epoch": 0.047085889570552146,
+      "grad_norm": 0.5664398074150085,
+      "learning_rate": 0.00014948040238294886,
+      "loss": 1.6552,
+      "step": 307
+    },
+    {
+      "epoch": 0.04723926380368098,
+      "grad_norm": 0.455660343170166,
+      "learning_rate": 0.00014947610433213189,
+      "loss": 1.4143,
+      "step": 308
+    },
+    {
+      "epoch": 0.04739263803680981,
+      "grad_norm": 0.45331138372421265,
+      "learning_rate": 0.00014947178864027448,
+      "loss": 1.3783,
+      "step": 309
+    },
+    {
+      "epoch": 0.04754601226993865,
+      "grad_norm": 0.5683020949363708,
+      "learning_rate": 0.00014946745530839886,
+      "loss": 1.4912,
+      "step": 310
+    },
+    {
+      "epoch": 0.04769938650306749,
+      "grad_norm": 0.4074256420135498,
+      "learning_rate": 0.00014946310433753155,
+      "loss": 1.5017,
+      "step": 311
+    },
+    {
+      "epoch": 0.04785276073619632,
+      "grad_norm": 0.513526201248169,
+      "learning_rate": 0.00014945873572870305,
+      "loss": 1.4865,
+      "step": 312
+    },
+    {
+      "epoch": 0.048006134969325154,
+      "grad_norm": 0.5456559062004089,
+      "learning_rate": 0.00014945434948294823,
+      "loss": 1.3709,
+      "step": 313
+    },
+    {
+      "epoch": 0.04815950920245399,
+      "grad_norm": 0.431872695684433,
+      "learning_rate": 0.000149449945601306,
+      "loss": 1.5086,
+      "step": 314
+    },
+    {
+      "epoch": 0.04831288343558282,
+      "grad_norm": 0.4263322353363037,
+      "learning_rate": 0.00014944552408481952,
+      "loss": 1.3941,
+      "step": 315
+    },
+    {
+      "epoch": 0.048466257668711654,
+      "grad_norm": 1.0746424198150635,
+      "learning_rate": 0.0001494410849345361,
+      "loss": 1.6212,
+      "step": 316
+    },
+    {
+      "epoch": 0.04861963190184049,
+      "grad_norm": 1.0165568590164185,
+      "learning_rate": 0.00014943662815150726,
+      "loss": 1.3994,
+      "step": 317
+    },
+    {
+      "epoch": 0.04877300613496933,
+      "grad_norm": 0.47195306420326233,
+      "learning_rate": 0.00014943215373678864,
+      "loss": 1.55,
+      "step": 318
+    },
+    {
+      "epoch": 0.04892638036809816,
+      "grad_norm": 0.5479450821876526,
+      "learning_rate": 0.0001494276616914401,
+      "loss": 1.4918,
+      "step": 319
+    },
+    {
+      "epoch": 0.049079754601226995,
+      "grad_norm": 0.44177329540252686,
+      "learning_rate": 0.00014942315201652568,
+      "loss": 1.4015,
+      "step": 320
+    },
+    {
+      "epoch": 0.04923312883435583,
+      "grad_norm": 0.46865057945251465,
+      "learning_rate": 0.00014941862471311355,
+      "loss": 1.448,
+      "step": 321
+    },
+    {
+      "epoch": 0.04938650306748466,
+      "grad_norm": 0.4543299376964569,
+      "learning_rate": 0.00014941407978227608,
+      "loss": 1.3791,
+      "step": 322
+    },
+    {
+      "epoch": 0.049539877300613495,
+      "grad_norm": 0.531876802444458,
+      "learning_rate": 0.00014940951722508988,
+      "loss": 1.4917,
+      "step": 323
+    },
+    {
+      "epoch": 0.04969325153374233,
+      "grad_norm": 0.4502822458744049,
+      "learning_rate": 0.00014940493704263563,
+      "loss": 1.5408,
+      "step": 324
+    },
+    {
+      "epoch": 0.04984662576687116,
+      "grad_norm": 0.45577719807624817,
+      "learning_rate": 0.00014940033923599823,
+      "loss": 1.3612,
+      "step": 325
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.687548041343689,
+      "learning_rate": 0.00014939572380626677,
+      "loss": 1.525,
+      "step": 326
+    },
+    {
+      "epoch": 0.050153374233128836,
+      "grad_norm": 0.46472668647766113,
+      "learning_rate": 0.00014939109075453447,
+      "loss": 1.5113,
+      "step": 327
+    },
+    {
+      "epoch": 0.05030674846625767,
+      "grad_norm": 0.4952874481678009,
+      "learning_rate": 0.0001493864400818988,
+      "loss": 1.4621,
+      "step": 328
+    },
+    {
+      "epoch": 0.0504601226993865,
+      "grad_norm": 0.59206223487854,
+      "learning_rate": 0.00014938177178946133,
+      "loss": 1.3832,
+      "step": 329
+    },
+    {
+      "epoch": 0.05061349693251534,
+      "grad_norm": 0.46549341082572937,
+      "learning_rate": 0.00014937708587832782,
+      "loss": 1.4584,
+      "step": 330
+    },
+    {
+      "epoch": 0.05076687116564417,
+      "grad_norm": 0.43143153190612793,
+      "learning_rate": 0.00014937238234960824,
+      "loss": 1.4774,
+      "step": 331
+    },
+    {
+      "epoch": 0.050920245398773004,
+      "grad_norm": 0.46019047498703003,
+      "learning_rate": 0.0001493676612044167,
+      "loss": 1.3498,
+      "step": 332
+    },
+    {
+      "epoch": 0.05107361963190184,
+      "grad_norm": 0.6220625042915344,
+      "learning_rate": 0.0001493629224438715,
+      "loss": 1.4629,
+      "step": 333
+    },
+    {
+      "epoch": 0.05122699386503068,
+      "grad_norm": 0.5862975120544434,
+      "learning_rate": 0.00014935816606909507,
+      "loss": 1.3758,
+      "step": 334
+    },
+    {
+      "epoch": 0.05138036809815951,
+      "grad_norm": 0.4843556582927704,
+      "learning_rate": 0.00014935339208121409,
+      "loss": 1.4275,
+      "step": 335
+    },
+    {
+      "epoch": 0.051533742331288344,
+      "grad_norm": 0.5658594369888306,
+      "learning_rate": 0.00014934860048135932,
+      "loss": 1.4144,
+      "step": 336
+    },
+    {
+      "epoch": 0.05168711656441718,
+      "grad_norm": 0.47703906893730164,
+      "learning_rate": 0.00014934379127066574,
+      "loss": 1.4879,
+      "step": 337
+    },
+    {
+      "epoch": 0.05184049079754601,
+      "grad_norm": 0.48222243785858154,
+      "learning_rate": 0.00014933896445027255,
+      "loss": 1.3547,
+      "step": 338
+    },
+    {
+      "epoch": 0.051993865030674845,
+      "grad_norm": 0.7616364359855652,
+      "learning_rate": 0.00014933412002132306,
+      "loss": 1.3807,
+      "step": 339
+    },
+    {
+      "epoch": 0.05214723926380368,
+      "grad_norm": 0.5141359567642212,
+      "learning_rate": 0.00014932925798496474,
+      "loss": 1.361,
+      "step": 340
+    },
+    {
+      "epoch": 0.05230061349693252,
+      "grad_norm": 0.42366892099380493,
+      "learning_rate": 0.00014932437834234923,
+      "loss": 1.4609,
+      "step": 341
+    },
+    {
+      "epoch": 0.05245398773006135,
+      "grad_norm": 0.3981532156467438,
+      "learning_rate": 0.00014931948109463245,
+      "loss": 1.4818,
+      "step": 342
+    },
+    {
+      "epoch": 0.052607361963190186,
+      "grad_norm": 0.4604555070400238,
+      "learning_rate": 0.0001493145662429743,
+      "loss": 1.438,
+      "step": 343
+    },
+    {
+      "epoch": 0.05276073619631902,
+      "grad_norm": 0.4341568052768707,
+      "learning_rate": 0.00014930963378853901,
+      "loss": 1.4034,
+      "step": 344
+    },
+    {
+      "epoch": 0.05291411042944785,
+      "grad_norm": 0.44069844484329224,
+      "learning_rate": 0.00014930468373249492,
+      "loss": 1.4653,
+      "step": 345
+    },
+    {
+      "epoch": 0.053067484662576686,
+      "grad_norm": 0.49977535009384155,
+      "learning_rate": 0.00014929971607601452,
+      "loss": 1.43,
+      "step": 346
+    },
+    {
+      "epoch": 0.05322085889570552,
+      "grad_norm": 0.4593197703361511,
+      "learning_rate": 0.00014929473082027453,
+      "loss": 1.4534,
+      "step": 347
+    },
+    {
+      "epoch": 0.05337423312883435,
+      "grad_norm": 0.4495786428451538,
+      "learning_rate": 0.00014928972796645574,
+      "loss": 1.4036,
+      "step": 348
+    },
+    {
+      "epoch": 0.05352760736196319,
+      "grad_norm": 0.9325106143951416,
+      "learning_rate": 0.00014928470751574325,
+      "loss": 1.4296,
+      "step": 349
+    },
+    {
+      "epoch": 0.05368098159509203,
+      "grad_norm": 0.5847094655036926,
+      "learning_rate": 0.00014927966946932616,
+      "loss": 1.5048,
+      "step": 350
+    },
+    {
+      "epoch": 0.05383435582822086,
+      "grad_norm": 0.7088392972946167,
+      "learning_rate": 0.00014927461382839788,
+      "loss": 1.4356,
+      "step": 351
+    },
+    {
+      "epoch": 0.053987730061349694,
+      "grad_norm": 0.527802586555481,
+      "learning_rate": 0.00014926954059415594,
+      "loss": 1.274,
+      "step": 352
+    },
+    {
+      "epoch": 0.05414110429447853,
+      "grad_norm": 0.8103647232055664,
+      "learning_rate": 0.00014926444976780197,
+      "loss": 1.5647,
+      "step": 353
+    },
+    {
+      "epoch": 0.05429447852760736,
+      "grad_norm": 0.570330023765564,
+      "learning_rate": 0.00014925934135054188,
+      "loss": 1.5279,
+      "step": 354
+    },
+    {
+      "epoch": 0.054447852760736194,
+      "grad_norm": 0.4824731647968292,
+      "learning_rate": 0.00014925421534358566,
+      "loss": 1.4582,
+      "step": 355
+    },
+    {
+      "epoch": 0.05460122699386503,
+      "grad_norm": 0.4886772334575653,
+      "learning_rate": 0.00014924907174814752,
+      "loss": 1.3263,
+      "step": 356
+    },
+    {
+      "epoch": 0.05475460122699387,
+      "grad_norm": 0.5513357520103455,
+      "learning_rate": 0.00014924391056544585,
+      "loss": 1.3749,
+      "step": 357
+    },
+    {
+      "epoch": 0.0549079754601227,
+      "grad_norm": 0.49871954321861267,
+      "learning_rate": 0.0001492387317967031,
+      "loss": 1.4499,
+      "step": 358
+    },
+    {
+      "epoch": 0.055061349693251535,
+      "grad_norm": 0.4645892381668091,
+      "learning_rate": 0.00014923353544314598,
+      "loss": 1.3598,
+      "step": 359
+    },
+    {
+      "epoch": 0.05521472392638037,
+      "grad_norm": 0.43729010224342346,
+      "learning_rate": 0.00014922832150600536,
+      "loss": 1.3919,
+      "step": 360
+    },
+    {
+      "epoch": 0.0553680981595092,
+      "grad_norm": 0.43087756633758545,
+      "learning_rate": 0.00014922308998651624,
+      "loss": 1.4825,
+      "step": 361
+    },
+    {
+      "epoch": 0.055521472392638035,
+      "grad_norm": 0.4052712023258209,
+      "learning_rate": 0.00014921784088591783,
+      "loss": 1.3899,
+      "step": 362
+    },
+    {
+      "epoch": 0.05567484662576687,
+      "grad_norm": 0.42732927203178406,
+      "learning_rate": 0.00014921257420545344,
+      "loss": 1.4339,
+      "step": 363
+    },
+    {
+      "epoch": 0.0558282208588957,
+      "grad_norm": 0.4733315110206604,
+      "learning_rate": 0.0001492072899463706,
+      "loss": 1.3713,
+      "step": 364
+    },
+    {
+      "epoch": 0.05598159509202454,
+      "grad_norm": 0.46981754899024963,
+      "learning_rate": 0.000149201988109921,
+      "loss": 1.3866,
+      "step": 365
+    },
+    {
+      "epoch": 0.056134969325153376,
+      "grad_norm": 0.4517139196395874,
+      "learning_rate": 0.00014919666869736044,
+      "loss": 1.3888,
+      "step": 366
+    },
+    {
+      "epoch": 0.05628834355828221,
+      "grad_norm": 0.4770098030567169,
+      "learning_rate": 0.0001491913317099489,
+      "loss": 1.3575,
+      "step": 367
+    },
+    {
+      "epoch": 0.05644171779141104,
+      "grad_norm": 0.5802912712097168,
+      "learning_rate": 0.00014918597714895065,
+      "loss": 1.4062,
+      "step": 368
+    },
+    {
+      "epoch": 0.05659509202453988,
+      "grad_norm": 0.45984718203544617,
+      "learning_rate": 0.00014918060501563392,
+      "loss": 1.3626,
+      "step": 369
+    },
+    {
+      "epoch": 0.05674846625766871,
+      "grad_norm": 0.3906403183937073,
+      "learning_rate": 0.00014917521531127126,
+      "loss": 1.4365,
+      "step": 370
+    },
+    {
+      "epoch": 0.056901840490797544,
+      "grad_norm": 0.497295618057251,
+      "learning_rate": 0.00014916980803713928,
+      "loss": 1.5084,
+      "step": 371
+    },
+    {
+      "epoch": 0.05705521472392638,
+      "grad_norm": 0.4369620680809021,
+      "learning_rate": 0.00014916438319451878,
+      "loss": 1.3618,
+      "step": 372
+    },
+    {
+      "epoch": 0.05720858895705522,
+      "grad_norm": 0.4122920632362366,
+      "learning_rate": 0.0001491589407846948,
+      "loss": 1.372,
+      "step": 373
+    },
+    {
+      "epoch": 0.05736196319018405,
+      "grad_norm": 0.5999494791030884,
+      "learning_rate": 0.00014915348080895638,
+      "loss": 1.3864,
+      "step": 374
+    },
+    {
+      "epoch": 0.057515337423312884,
+      "grad_norm": 0.4002017378807068,
+      "learning_rate": 0.00014914800326859692,
+      "loss": 1.4932,
+      "step": 375
+    },
+    {
+      "epoch": 0.05766871165644172,
+      "grad_norm": 0.5213642120361328,
+      "learning_rate": 0.0001491425081649138,
+      "loss": 1.4196,
+      "step": 376
+    },
+    {
+      "epoch": 0.05782208588957055,
+      "grad_norm": 0.8480618000030518,
+      "learning_rate": 0.00014913699549920866,
+      "loss": 1.4089,
+      "step": 377
+    },
+    {
+      "epoch": 0.057975460122699385,
+      "grad_norm": 0.5111736059188843,
+      "learning_rate": 0.00014913146527278728,
+      "loss": 1.3274,
+      "step": 378
+    },
+    {
+      "epoch": 0.05812883435582822,
+      "grad_norm": 0.3971315920352936,
+      "learning_rate": 0.0001491259174869596,
+      "loss": 1.359,
+      "step": 379
+    },
+    {
+      "epoch": 0.05828220858895705,
+      "grad_norm": 0.41635820269584656,
+      "learning_rate": 0.0001491203521430397,
+      "loss": 1.4776,
+      "step": 380
+    },
+    {
+      "epoch": 0.05843558282208589,
+      "grad_norm": 0.45930641889572144,
+      "learning_rate": 0.00014911476924234587,
+      "loss": 1.2534,
+      "step": 381
+    },
+    {
+      "epoch": 0.058588957055214726,
+      "grad_norm": 0.568717896938324,
+      "learning_rate": 0.00014910916878620048,
+      "loss": 1.3414,
+      "step": 382
+    },
+    {
+      "epoch": 0.05874233128834356,
+      "grad_norm": 0.46514275670051575,
+      "learning_rate": 0.0001491035507759301,
+      "loss": 1.3384,
+      "step": 383
+    },
+    {
+      "epoch": 0.05889570552147239,
+      "grad_norm": 0.43338680267333984,
+      "learning_rate": 0.00014909791521286553,
+      "loss": 1.4168,
+      "step": 384
+    },
+    {
+      "epoch": 0.059049079754601226,
+      "grad_norm": 0.5989061594009399,
+      "learning_rate": 0.00014909226209834157,
+      "loss": 1.4465,
+      "step": 385
+    },
+    {
+      "epoch": 0.05920245398773006,
+      "grad_norm": 0.5000714063644409,
+      "learning_rate": 0.0001490865914336973,
+      "loss": 1.4102,
+      "step": 386
+    },
+    {
+      "epoch": 0.05935582822085889,
+      "grad_norm": 0.44436636567115784,
+      "learning_rate": 0.00014908090322027597,
+      "loss": 1.4225,
+      "step": 387
+    },
+    {
+      "epoch": 0.05950920245398773,
+      "grad_norm": 0.8168815970420837,
+      "learning_rate": 0.00014907519745942482,
+      "loss": 1.4337,
+      "step": 388
+    },
+    {
+      "epoch": 0.05966257668711657,
+      "grad_norm": 0.6770294904708862,
+      "learning_rate": 0.00014906947415249547,
+      "loss": 1.3675,
+      "step": 389
+    },
+    {
+      "epoch": 0.0598159509202454,
+      "grad_norm": 0.478729784488678,
+      "learning_rate": 0.00014906373330084358,
+      "loss": 1.388,
+      "step": 390
+    },
+    {
+      "epoch": 0.059969325153374234,
+      "grad_norm": 0.4507676959037781,
+      "learning_rate": 0.00014905797490582894,
+      "loss": 1.2495,
+      "step": 391
+    },
+    {
+      "epoch": 0.06012269938650307,
+      "grad_norm": 0.9043616652488708,
+      "learning_rate": 0.00014905219896881554,
+      "loss": 1.4431,
+      "step": 392
+    },
+    {
+      "epoch": 0.0602760736196319,
+      "grad_norm": 0.6604290008544922,
+      "learning_rate": 0.0001490464054911715,
+      "loss": 1.3674,
+      "step": 393
+    },
+    {
+      "epoch": 0.060429447852760734,
+      "grad_norm": 0.5075806379318237,
+      "learning_rate": 0.00014904059447426918,
+      "loss": 1.5379,
+      "step": 394
+    },
+    {
+      "epoch": 0.06058282208588957,
+      "grad_norm": 0.4346505403518677,
+      "learning_rate": 0.00014903476591948496,
+      "loss": 1.4447,
+      "step": 395
+    },
+    {
+      "epoch": 0.06073619631901841,
+      "grad_norm": 0.5195268392562866,
+      "learning_rate": 0.00014902891982819944,
+      "loss": 1.3203,
+      "step": 396
+    },
+    {
+      "epoch": 0.06088957055214724,
+      "grad_norm": 0.48815247416496277,
+      "learning_rate": 0.00014902305620179742,
+      "loss": 1.2059,
+      "step": 397
+    },
+    {
+      "epoch": 0.061042944785276075,
+      "grad_norm": 0.4953227639198303,
+      "learning_rate": 0.0001490171750416678,
+      "loss": 1.4056,
+      "step": 398
+    },
+    {
+      "epoch": 0.06119631901840491,
+      "grad_norm": 0.4652651250362396,
+      "learning_rate": 0.0001490112763492036,
+      "loss": 1.3771,
+      "step": 399
+    },
+    {
+      "epoch": 0.06134969325153374,
+      "grad_norm": 0.5088167190551758,
+      "learning_rate": 0.00014900536012580208,
+      "loss": 1.4686,
+      "step": 400
+    },
+    {
+      "epoch": 0.061503067484662576,
+      "grad_norm": 0.4817758798599243,
+      "learning_rate": 0.00014899942637286456,
+      "loss": 1.3932,
+      "step": 401
+    },
+    {
+      "epoch": 0.06165644171779141,
+      "grad_norm": 0.45939183235168457,
+      "learning_rate": 0.00014899347509179662,
+      "loss": 1.4312,
+      "step": 402
+    },
+    {
+      "epoch": 0.06180981595092024,
+      "grad_norm": 0.5755530595779419,
+      "learning_rate": 0.00014898750628400788,
+      "loss": 1.3378,
+      "step": 403
+    },
+    {
+      "epoch": 0.06196319018404908,
+      "grad_norm": 0.5162791609764099,
+      "learning_rate": 0.0001489815199509122,
+      "loss": 1.3722,
+      "step": 404
+    },
+    {
+      "epoch": 0.062116564417177916,
+      "grad_norm": 0.5068424940109253,
+      "learning_rate": 0.00014897551609392752,
+      "loss": 1.2644,
+      "step": 405
+    },
+    {
+      "epoch": 0.06226993865030675,
+      "grad_norm": 0.4198472499847412,
+      "learning_rate": 0.00014896949471447597,
+      "loss": 1.3214,
+      "step": 406
+    },
+    {
+      "epoch": 0.06242331288343558,
+      "grad_norm": 0.4970983862876892,
+      "learning_rate": 0.00014896345581398384,
+      "loss": 1.2571,
+      "step": 407
+    },
+    {
+      "epoch": 0.06257668711656442,
+      "grad_norm": 0.4446052014827728,
+      "learning_rate": 0.00014895739939388154,
+      "loss": 1.3199,
+      "step": 408
+    },
+    {
+      "epoch": 0.06273006134969325,
+      "grad_norm": 0.6020923852920532,
+      "learning_rate": 0.0001489513254556037,
+      "loss": 1.2303,
+      "step": 409
+    },
+    {
+      "epoch": 0.06288343558282208,
+      "grad_norm": 0.8033818006515503,
+      "learning_rate": 0.00014894523400058894,
+      "loss": 1.4693,
+      "step": 410
+    },
+    {
+      "epoch": 0.06303680981595092,
+      "grad_norm": 0.8705591559410095,
+      "learning_rate": 0.0001489391250302802,
+      "loss": 1.2397,
+      "step": 411
+    },
+    {
+      "epoch": 0.06319018404907975,
+      "grad_norm": 0.48758605122566223,
+      "learning_rate": 0.0001489329985461245,
+      "loss": 1.3679,
+      "step": 412
+    },
+    {
+      "epoch": 0.06334355828220858,
+      "grad_norm": 0.6478848457336426,
+      "learning_rate": 0.00014892685454957302,
+      "loss": 1.4673,
+      "step": 413
+    },
+    {
+      "epoch": 0.06349693251533742,
+      "grad_norm": 0.39195746183395386,
+      "learning_rate": 0.00014892069304208106,
+      "loss": 1.5083,
+      "step": 414
+    },
+    {
+      "epoch": 0.06365030674846625,
+      "grad_norm": 0.5908502340316772,
+      "learning_rate": 0.00014891451402510806,
+      "loss": 1.4938,
+      "step": 415
+    },
+    {
+      "epoch": 0.0638036809815951,
+      "grad_norm": 0.4813465178012848,
+      "learning_rate": 0.0001489083175001177,
+      "loss": 1.1437,
+      "step": 416
+    },
+    {
+      "epoch": 0.06395705521472393,
+      "grad_norm": 0.863193690776825,
+      "learning_rate": 0.00014890210346857766,
+      "loss": 1.4059,
+      "step": 417
+    },
+    {
+      "epoch": 0.06411042944785277,
+      "grad_norm": 0.515499472618103,
+      "learning_rate": 0.00014889587193195995,
+      "loss": 1.475,
+      "step": 418
+    },
+    {
+      "epoch": 0.0642638036809816,
+      "grad_norm": 0.5976134538650513,
+      "learning_rate": 0.00014888962289174052,
+      "loss": 1.3275,
+      "step": 419
+    },
+    {
+      "epoch": 0.06441717791411043,
+      "grad_norm": 0.5400510430335999,
+      "learning_rate": 0.00014888335634939966,
+      "loss": 1.3615,
+      "step": 420
+    },
+    {
+      "epoch": 0.06457055214723927,
+      "grad_norm": 0.5568822622299194,
+      "learning_rate": 0.00014887707230642165,
+      "loss": 1.2207,
+      "step": 421
+    },
+    {
+      "epoch": 0.0647239263803681,
+      "grad_norm": 0.502244234085083,
+      "learning_rate": 0.00014887077076429503,
+      "loss": 1.4529,
+      "step": 422
+    },
+    {
+      "epoch": 0.06487730061349693,
+      "grad_norm": 0.5405155420303345,
+      "learning_rate": 0.00014886445172451236,
+      "loss": 1.2438,
+      "step": 423
+    },
+    {
+      "epoch": 0.06503067484662577,
+      "grad_norm": 0.4825628399848938,
+      "learning_rate": 0.00014885811518857052,
+      "loss": 1.3799,
+      "step": 424
+    },
+    {
+      "epoch": 0.0651840490797546,
+      "grad_norm": 0.591012179851532,
+      "learning_rate": 0.0001488517611579704,
+      "loss": 1.3854,
+      "step": 425
+    },
+    {
+      "epoch": 0.06533742331288343,
+      "grad_norm": 0.54218590259552,
+      "learning_rate": 0.00014884538963421705,
+      "loss": 1.4183,
+      "step": 426
+    },
+    {
+      "epoch": 0.06549079754601227,
+      "grad_norm": 0.4336499571800232,
+      "learning_rate": 0.00014883900061881967,
+      "loss": 1.3236,
+      "step": 427
+    },
+    {
+      "epoch": 0.0656441717791411,
+      "grad_norm": 0.5193835496902466,
+      "learning_rate": 0.00014883259411329166,
+      "loss": 1.3897,
+      "step": 428
+    },
+    {
+      "epoch": 0.06579754601226993,
+      "grad_norm": 0.41640183329582214,
+      "learning_rate": 0.00014882617011915047,
+      "loss": 1.3772,
+      "step": 429
+    },
+    {
+      "epoch": 0.06595092024539877,
+      "grad_norm": 0.55613774061203,
+      "learning_rate": 0.00014881972863791778,
+      "loss": 1.3518,
+      "step": 430
+    },
+    {
+      "epoch": 0.0661042944785276,
+      "grad_norm": 0.6740829944610596,
+      "learning_rate": 0.0001488132696711194,
+      "loss": 1.2597,
+      "step": 431
+    },
+    {
+      "epoch": 0.06625766871165645,
+      "grad_norm": 0.4811161160469055,
+      "learning_rate": 0.0001488067932202852,
+      "loss": 1.4136,
+      "step": 432
+    },
+    {
+      "epoch": 0.06641104294478528,
+      "grad_norm": 0.4066157937049866,
+      "learning_rate": 0.00014880029928694924,
+      "loss": 1.2396,
+      "step": 433
+    },
+    {
+      "epoch": 0.06656441717791411,
+      "grad_norm": 0.4148232340812683,
+      "learning_rate": 0.00014879378787264975,
+      "loss": 1.4289,
+      "step": 434
+    },
+    {
+      "epoch": 0.06671779141104295,
+      "grad_norm": 0.42814043164253235,
+      "learning_rate": 0.00014878725897892912,
+      "loss": 1.3465,
+      "step": 435
+    },
+    {
+      "epoch": 0.06687116564417178,
+      "grad_norm": 0.3809296190738678,
+      "learning_rate": 0.0001487807126073338,
+      "loss": 1.3647,
+      "step": 436
+    },
+    {
+      "epoch": 0.06702453987730062,
+      "grad_norm": 0.43945470452308655,
+      "learning_rate": 0.0001487741487594144,
+      "loss": 1.3475,
+      "step": 437
+    },
+    {
+      "epoch": 0.06717791411042945,
+      "grad_norm": 0.49180132150650024,
+      "learning_rate": 0.0001487675674367257,
+      "loss": 1.3264,
+      "step": 438
+    },
+    {
+      "epoch": 0.06733128834355828,
+      "grad_norm": 0.5954174995422363,
+      "learning_rate": 0.00014876096864082665,
+      "loss": 1.2671,
+      "step": 439
+    },
+    {
+      "epoch": 0.06748466257668712,
+      "grad_norm": 0.4337195158004761,
+      "learning_rate": 0.00014875435237328023,
+      "loss": 1.2867,
+      "step": 440
+    },
+    {
+      "epoch": 0.06763803680981595,
+      "grad_norm": 0.9139065146446228,
+      "learning_rate": 0.0001487477186356537,
+      "loss": 1.3478,
+      "step": 441
+    },
+    {
+      "epoch": 0.06779141104294478,
+      "grad_norm": 0.6302591562271118,
+      "learning_rate": 0.0001487410674295183,
+      "loss": 1.3726,
+      "step": 442
+    },
+    {
+      "epoch": 0.06794478527607362,
+      "grad_norm": 0.6063013672828674,
+      "learning_rate": 0.00014873439875644955,
+      "loss": 1.3317,
+      "step": 443
+    },
+    {
+      "epoch": 0.06809815950920245,
+      "grad_norm": 0.5359678268432617,
+      "learning_rate": 0.00014872771261802703,
+      "loss": 1.3585,
+      "step": 444
+    },
+    {
+      "epoch": 0.06825153374233128,
+      "grad_norm": 0.6703312397003174,
+      "learning_rate": 0.00014872100901583448,
+      "loss": 1.3542,
+      "step": 445
+    },
+    {
+      "epoch": 0.06840490797546012,
+      "grad_norm": 0.566304624080658,
+      "learning_rate": 0.00014871428795145975,
+      "loss": 1.3121,
+      "step": 446
+    },
+    {
+      "epoch": 0.06855828220858896,
+      "grad_norm": 0.48193395137786865,
+      "learning_rate": 0.0001487075494264949,
+      "loss": 1.3726,
+      "step": 447
+    },
+    {
+      "epoch": 0.0687116564417178,
+      "grad_norm": 0.7194489240646362,
+      "learning_rate": 0.00014870079344253602,
+      "loss": 1.433,
+      "step": 448
+    },
+    {
+      "epoch": 0.06886503067484663,
+      "grad_norm": 0.45869988203048706,
+      "learning_rate": 0.0001486940200011834,
+      "loss": 1.3654,
+      "step": 449
+    },
+    {
+      "epoch": 0.06901840490797546,
+      "grad_norm": 0.5410028100013733,
+      "learning_rate": 0.00014868722910404148,
+      "loss": 1.2767,
+      "step": 450
+    },
+    {
+      "epoch": 0.0691717791411043,
+      "grad_norm": 0.45854243636131287,
+      "learning_rate": 0.0001486804207527188,
+      "loss": 1.4058,
+      "step": 451
+    },
+    {
+      "epoch": 0.06932515337423313,
+      "grad_norm": 0.5670607089996338,
+      "learning_rate": 0.00014867359494882804,
+      "loss": 1.2185,
+      "step": 452
+    },
+    {
+      "epoch": 0.06947852760736196,
+      "grad_norm": 0.43948128819465637,
+      "learning_rate": 0.000148666751693986,
+      "loss": 1.4028,
+      "step": 453
+    },
+    {
+      "epoch": 0.0696319018404908,
+      "grad_norm": 0.5671201944351196,
+      "learning_rate": 0.00014865989098981365,
+      "loss": 1.2917,
+      "step": 454
+    },
+    {
+      "epoch": 0.06978527607361963,
+      "grad_norm": 0.559069037437439,
+      "learning_rate": 0.0001486530128379361,
+      "loss": 1.2963,
+      "step": 455
+    },
+    {
+      "epoch": 0.06993865030674846,
+      "grad_norm": 0.47423461079597473,
+      "learning_rate": 0.0001486461172399825,
+      "loss": 1.3214,
+      "step": 456
+    },
+    {
+      "epoch": 0.0700920245398773,
+      "grad_norm": 0.446857213973999,
+      "learning_rate": 0.00014863920419758625,
+      "loss": 1.2282,
+      "step": 457
+    },
+    {
+      "epoch": 0.07024539877300613,
+      "grad_norm": 0.4252924919128418,
+      "learning_rate": 0.00014863227371238484,
+      "loss": 1.1997,
+      "step": 458
+    },
+    {
+      "epoch": 0.07039877300613497,
+      "grad_norm": 0.4934619963169098,
+      "learning_rate": 0.00014862532578601985,
+      "loss": 1.3417,
+      "step": 459
+    },
+    {
+      "epoch": 0.0705521472392638,
+      "grad_norm": 0.4634302854537964,
+      "learning_rate": 0.00014861836042013703,
+      "loss": 1.4027,
+      "step": 460
+    },
+    {
+      "epoch": 0.07070552147239263,
+      "grad_norm": 0.4769023656845093,
+      "learning_rate": 0.00014861137761638628,
+      "loss": 1.4032,
+      "step": 461
+    },
+    {
+      "epoch": 0.07085889570552147,
+      "grad_norm": 0.5140354633331299,
+      "learning_rate": 0.00014860437737642155,
+      "loss": 1.3437,
+      "step": 462
+    },
+    {
+      "epoch": 0.07101226993865031,
+      "grad_norm": 0.5393059849739075,
+      "learning_rate": 0.00014859735970190106,
+      "loss": 1.3602,
+      "step": 463
+    },
+    {
+      "epoch": 0.07116564417177915,
+      "grad_norm": 0.4592728912830353,
+      "learning_rate": 0.00014859032459448702,
+      "loss": 1.2348,
+      "step": 464
+    },
+    {
+      "epoch": 0.07131901840490798,
+      "grad_norm": 0.5363672971725464,
+      "learning_rate": 0.00014858327205584583,
+      "loss": 1.2783,
+      "step": 465
+    },
+    {
+      "epoch": 0.07147239263803681,
+      "grad_norm": 0.6304312944412231,
+      "learning_rate": 0.00014857620208764803,
+      "loss": 1.36,
+      "step": 466
+    },
+    {
+      "epoch": 0.07162576687116565,
+      "grad_norm": 0.42689624428749084,
+      "learning_rate": 0.00014856911469156823,
+      "loss": 1.4174,
+      "step": 467
+    },
+    {
+      "epoch": 0.07177914110429448,
+      "grad_norm": 0.6914393305778503,
+      "learning_rate": 0.00014856200986928527,
+      "loss": 1.5282,
+      "step": 468
+    },
+    {
+      "epoch": 0.07193251533742331,
+      "grad_norm": 0.5081053972244263,
+      "learning_rate": 0.00014855488762248207,
+      "loss": 1.0983,
+      "step": 469
+    },
+    {
+      "epoch": 0.07208588957055215,
+      "grad_norm": 0.7178828120231628,
+      "learning_rate": 0.0001485477479528456,
+      "loss": 1.3618,
+      "step": 470
+    },
+    {
+      "epoch": 0.07223926380368098,
+      "grad_norm": 0.5675197243690491,
+      "learning_rate": 0.00014854059086206702,
+      "loss": 1.4688,
+      "step": 471
+    },
+    {
+      "epoch": 0.07239263803680981,
+      "grad_norm": 0.4940665662288666,
+      "learning_rate": 0.00014853341635184168,
+      "loss": 1.2941,
+      "step": 472
+    },
+    {
+      "epoch": 0.07254601226993865,
+      "grad_norm": 0.5538465976715088,
+      "learning_rate": 0.00014852622442386895,
+      "loss": 1.2817,
+      "step": 473
+    },
+    {
+      "epoch": 0.07269938650306748,
+      "grad_norm": 0.44795095920562744,
+      "learning_rate": 0.00014851901507985238,
+      "loss": 1.5361,
+      "step": 474
+    },
+    {
+      "epoch": 0.07285276073619631,
+      "grad_norm": 0.4201502203941345,
+      "learning_rate": 0.00014851178832149967,
+      "loss": 1.3162,
+      "step": 475
+    },
+    {
+      "epoch": 0.07300613496932515,
+      "grad_norm": 0.430527925491333,
+      "learning_rate": 0.00014850454415052256,
+      "loss": 1.4798,
+      "step": 476
+    },
+    {
+      "epoch": 0.07315950920245398,
+      "grad_norm": 0.5481934547424316,
+      "learning_rate": 0.00014849728256863698,
+      "loss": 1.3888,
+      "step": 477
+    },
+    {
+      "epoch": 0.07331288343558282,
+      "grad_norm": 0.5167529582977295,
+      "learning_rate": 0.000148490003577563,
+      "loss": 1.2306,
+      "step": 478
+    },
+    {
+      "epoch": 0.07346625766871166,
+      "grad_norm": 0.431482195854187,
+      "learning_rate": 0.00014848270717902478,
+      "loss": 1.3137,
+      "step": 479
+    },
+    {
+      "epoch": 0.0736196319018405,
+      "grad_norm": 0.4432452619075775,
+      "learning_rate": 0.00014847539337475054,
+      "loss": 1.3593,
+      "step": 480
+    },
+    {
+      "epoch": 0.07377300613496933,
+      "grad_norm": 0.4434761703014374,
+      "learning_rate": 0.00014846806216647278,
+      "loss": 1.4222,
+      "step": 481
+    },
+    {
+      "epoch": 0.07392638036809816,
+      "grad_norm": 0.7436431050300598,
+      "learning_rate": 0.00014846071355592797,
+      "loss": 1.175,
+      "step": 482
+    },
+    {
+      "epoch": 0.074079754601227,
+      "grad_norm": 1.075213074684143,
+      "learning_rate": 0.00014845334754485677,
+      "loss": 1.3525,
+      "step": 483
+    },
+    {
+      "epoch": 0.07423312883435583,
+      "grad_norm": 0.48838797211647034,
+      "learning_rate": 0.000148445964135004,
+      "loss": 1.43,
+      "step": 484
+    },
+    {
+      "epoch": 0.07438650306748466,
+      "grad_norm": 0.5812395811080933,
+      "learning_rate": 0.0001484385633281185,
+      "loss": 1.3449,
+      "step": 485
+    },
+    {
+      "epoch": 0.0745398773006135,
+      "grad_norm": 0.5197191834449768,
+      "learning_rate": 0.00014843114512595336,
+      "loss": 1.3225,
+      "step": 486
+    },
+    {
+      "epoch": 0.07469325153374233,
+      "grad_norm": 0.9942704439163208,
+      "learning_rate": 0.00014842370953026564,
+      "loss": 1.4337,
+      "step": 487
+    },
+    {
+      "epoch": 0.07484662576687116,
+      "grad_norm": 0.5735070705413818,
+      "learning_rate": 0.00014841625654281662,
+      "loss": 1.3229,
+      "step": 488
+    },
+    {
+      "epoch": 0.075,
+      "grad_norm": 0.7372777462005615,
+      "learning_rate": 0.0001484087861653717,
+      "loss": 1.1848,
+      "step": 489
+    },
+    {
+      "epoch": 0.07515337423312883,
+      "grad_norm": 0.7764583230018616,
+      "learning_rate": 0.00014840129839970037,
+      "loss": 1.3465,
+      "step": 490
+    },
+    {
+      "epoch": 0.07530674846625766,
+      "grad_norm": 0.49510499835014343,
+      "learning_rate": 0.00014839379324757624,
+      "loss": 1.3071,
+      "step": 491
+    },
+    {
+      "epoch": 0.0754601226993865,
+      "grad_norm": 0.5933728814125061,
+      "learning_rate": 0.00014838627071077704,
+      "loss": 1.2909,
+      "step": 492
+    },
+    {
+      "epoch": 0.07561349693251533,
+      "grad_norm": 0.4268637001514435,
+      "learning_rate": 0.00014837873079108464,
+      "loss": 1.4213,
+      "step": 493
+    },
+    {
+      "epoch": 0.07576687116564418,
+      "grad_norm": 0.721684455871582,
+      "learning_rate": 0.00014837117349028497,
+      "loss": 1.34,
+      "step": 494
+    },
+    {
+      "epoch": 0.07592024539877301,
+      "grad_norm": 0.45970723032951355,
+      "learning_rate": 0.00014836359881016816,
+      "loss": 1.3035,
+      "step": 495
+    },
+    {
+      "epoch": 0.07607361963190185,
+      "grad_norm": 0.5905528664588928,
+      "learning_rate": 0.0001483560067525284,
+      "loss": 1.3341,
+      "step": 496
+    },
+    {
+      "epoch": 0.07622699386503068,
+      "grad_norm": 0.6083369255065918,
+      "learning_rate": 0.000148348397319164,
+      "loss": 1.5115,
+      "step": 497
+    },
+    {
+      "epoch": 0.07638036809815951,
+      "grad_norm": 0.41297653317451477,
+      "learning_rate": 0.0001483407705118774,
+      "loss": 1.3217,
+      "step": 498
+    },
+    {
+      "epoch": 0.07653374233128835,
+      "grad_norm": 0.3917851448059082,
+      "learning_rate": 0.00014833312633247514,
+      "loss": 1.467,
+      "step": 499
+    },
+    {
+      "epoch": 0.07668711656441718,
+      "grad_norm": 0.5250905156135559,
+      "learning_rate": 0.00014832546478276792,
+      "loss": 1.3991,
+      "step": 500
+    },
+    {
+      "epoch": 0.07684049079754601,
+      "grad_norm": 0.5031157732009888,
+      "learning_rate": 0.00014831778586457046,
+      "loss": 1.2316,
+      "step": 501
+    },
+    {
+      "epoch": 0.07699386503067485,
+      "grad_norm": 0.4872812330722809,
+      "learning_rate": 0.0001483100895797017,
+      "loss": 1.2758,
+      "step": 502
+    },
+    {
+      "epoch": 0.07714723926380368,
+      "grad_norm": 0.5064236521720886,
+      "learning_rate": 0.00014830237592998463,
+      "loss": 1.3706,
+      "step": 503
+    },
+    {
+      "epoch": 0.07730061349693251,
+      "grad_norm": 0.41143012046813965,
+      "learning_rate": 0.00014829464491724638,
+      "loss": 1.3348,
+      "step": 504
+    },
+    {
+      "epoch": 0.07745398773006135,
+      "grad_norm": 0.5962451696395874,
+      "learning_rate": 0.00014828689654331818,
+      "loss": 1.4348,
+      "step": 505
+    },
+    {
+      "epoch": 0.07760736196319018,
+      "grad_norm": 0.4838153123855591,
+      "learning_rate": 0.00014827913081003538,
+      "loss": 1.3359,
+      "step": 506
+    },
+    {
+      "epoch": 0.07776073619631901,
+      "grad_norm": 0.5272953510284424,
+      "learning_rate": 0.0001482713477192374,
+      "loss": 1.3956,
+      "step": 507
+    },
+    {
+      "epoch": 0.07791411042944785,
+      "grad_norm": 0.42692282795906067,
+      "learning_rate": 0.00014826354727276786,
+      "loss": 1.2617,
+      "step": 508
+    },
+    {
+      "epoch": 0.07806748466257668,
+      "grad_norm": 0.47148358821868896,
+      "learning_rate": 0.00014825572947247442,
+      "loss": 1.3607,
+      "step": 509
+    },
+    {
+      "epoch": 0.07822085889570553,
+      "grad_norm": 0.42702433466911316,
+      "learning_rate": 0.00014824789432020888,
+      "loss": 1.4379,
+      "step": 510
+    },
+    {
+      "epoch": 0.07837423312883436,
+      "grad_norm": 0.5548335313796997,
+      "learning_rate": 0.00014824004181782712,
+      "loss": 1.3075,
+      "step": 511
+    },
+    {
+      "epoch": 0.0785276073619632,
+      "grad_norm": 0.47418245673179626,
+      "learning_rate": 0.00014823217196718915,
+      "loss": 1.3315,
+      "step": 512
+    },
+    {
+      "epoch": 0.07868098159509203,
+      "grad_norm": 0.4181951880455017,
+      "learning_rate": 0.0001482242847701591,
+      "loss": 1.2132,
+      "step": 513
+    },
+    {
+      "epoch": 0.07883435582822086,
+      "grad_norm": 0.4804813861846924,
+      "learning_rate": 0.0001482163802286052,
+      "loss": 1.3431,
+      "step": 514
+    },
+    {
+      "epoch": 0.0789877300613497,
+      "grad_norm": 0.39072558283805847,
+      "learning_rate": 0.0001482084583443998,
+      "loss": 1.4361,
+      "step": 515
+    },
+    {
+      "epoch": 0.07914110429447853,
+      "grad_norm": 0.45596057176589966,
+      "learning_rate": 0.0001482005191194193,
+      "loss": 1.335,
+      "step": 516
+    },
+    {
+      "epoch": 0.07929447852760736,
+      "grad_norm": 0.4300730526447296,
+      "learning_rate": 0.0001481925625555443,
+      "loss": 1.1945,
+      "step": 517
+    },
+    {
+      "epoch": 0.0794478527607362,
+      "grad_norm": 0.6591091752052307,
+      "learning_rate": 0.00014818458865465946,
+      "loss": 1.3654,
+      "step": 518
+    },
+    {
+      "epoch": 0.07960122699386503,
+      "grad_norm": 0.42098766565322876,
+      "learning_rate": 0.0001481765974186535,
+      "loss": 1.2897,
+      "step": 519
+    },
+    {
+      "epoch": 0.07975460122699386,
+      "grad_norm": 0.5882658362388611,
+      "learning_rate": 0.00014816858884941932,
+      "loss": 1.2867,
+      "step": 520
+    },
+    {
+      "epoch": 0.0799079754601227,
+      "grad_norm": 0.5603997707366943,
+      "learning_rate": 0.0001481605629488539,
+      "loss": 1.2631,
+      "step": 521
+    },
+    {
+      "epoch": 0.08006134969325153,
+      "grad_norm": 0.39910754561424255,
+      "learning_rate": 0.0001481525197188583,
+      "loss": 1.2609,
+      "step": 522
+    },
+    {
+      "epoch": 0.08021472392638036,
+      "grad_norm": 0.5855087637901306,
+      "learning_rate": 0.00014814445916133773,
+      "loss": 1.2172,
+      "step": 523
+    },
+    {
+      "epoch": 0.0803680981595092,
+      "grad_norm": 0.4452013373374939,
+      "learning_rate": 0.00014813638127820152,
+      "loss": 1.3106,
+      "step": 524
+    },
+    {
+      "epoch": 0.08052147239263803,
+      "grad_norm": 0.7003870606422424,
+      "learning_rate": 0.00014812828607136297,
+      "loss": 1.2133,
+      "step": 525
+    },
+    {
+      "epoch": 0.08067484662576688,
+      "grad_norm": 0.40929195284843445,
+      "learning_rate": 0.00014812017354273965,
+      "loss": 1.3314,
+      "step": 526
+    },
+    {
+      "epoch": 0.08082822085889571,
+      "grad_norm": 0.4900018572807312,
+      "learning_rate": 0.00014811204369425315,
+      "loss": 1.3926,
+      "step": 527
+    },
+    {
+      "epoch": 0.08098159509202454,
+      "grad_norm": 0.38799721002578735,
+      "learning_rate": 0.00014810389652782915,
+      "loss": 1.2919,
+      "step": 528
+    },
+    {
+      "epoch": 0.08113496932515338,
+      "grad_norm": 0.4534439146518707,
+      "learning_rate": 0.0001480957320453975,
+      "loss": 1.3401,
+      "step": 529
+    },
+    {
+      "epoch": 0.08128834355828221,
+      "grad_norm": 0.42503777146339417,
+      "learning_rate": 0.00014808755024889206,
+      "loss": 1.0604,
+      "step": 530
+    },
+    {
+      "epoch": 0.08144171779141104,
+      "grad_norm": 0.4545483887195587,
+      "learning_rate": 0.00014807935114025087,
+      "loss": 1.2343,
+      "step": 531
+    },
+    {
+      "epoch": 0.08159509202453988,
+      "grad_norm": 0.37469080090522766,
+      "learning_rate": 0.00014807113472141607,
+      "loss": 1.2469,
+      "step": 532
+    },
+    {
+      "epoch": 0.08174846625766871,
+      "grad_norm": 0.5288442373275757,
+      "learning_rate": 0.0001480629009943338,
+      "loss": 1.3496,
+      "step": 533
+    },
+    {
+      "epoch": 0.08190184049079755,
+      "grad_norm": 0.5469565987586975,
+      "learning_rate": 0.0001480546499609544,
+      "loss": 1.3514,
+      "step": 534
+    },
+    {
+      "epoch": 0.08205521472392638,
+      "grad_norm": 0.4116729497909546,
+      "learning_rate": 0.0001480463816232323,
+      "loss": 1.2688,
+      "step": 535
+    },
+    {
+      "epoch": 0.08220858895705521,
+      "grad_norm": 0.5259972214698792,
+      "learning_rate": 0.00014803809598312598,
+      "loss": 1.1901,
+      "step": 536
+    },
+    {
+      "epoch": 0.08236196319018405,
+      "grad_norm": 0.8216963410377502,
+      "learning_rate": 0.00014802979304259808,
+      "loss": 1.4001,
+      "step": 537
+    },
+    {
+      "epoch": 0.08251533742331288,
+      "grad_norm": 0.5650008320808411,
+      "learning_rate": 0.00014802147280361527,
+      "loss": 1.4349,
+      "step": 538
+    },
+    {
+      "epoch": 0.08266871165644171,
+      "grad_norm": 0.5415148735046387,
+      "learning_rate": 0.00014801313526814836,
+      "loss": 1.3737,
+      "step": 539
+    },
+    {
+      "epoch": 0.08282208588957055,
+      "grad_norm": 0.5397143959999084,
+      "learning_rate": 0.00014800478043817226,
+      "loss": 1.2631,
+      "step": 540
+    },
+    {
+      "epoch": 0.0829754601226994,
+      "grad_norm": 0.7763659358024597,
+      "learning_rate": 0.00014799640831566596,
+      "loss": 1.2372,
+      "step": 541
+    },
+    {
+      "epoch": 0.08312883435582823,
+      "grad_norm": 0.8116266131401062,
+      "learning_rate": 0.00014798801890261257,
+      "loss": 1.2506,
+      "step": 542
+    },
+    {
+      "epoch": 0.08328220858895706,
+      "grad_norm": 0.35098856687545776,
+      "learning_rate": 0.00014797961220099922,
+      "loss": 1.3039,
+      "step": 543
+    },
+    {
+      "epoch": 0.0834355828220859,
+      "grad_norm": 0.570016622543335,
+      "learning_rate": 0.0001479711882128173,
+      "loss": 1.2277,
+      "step": 544
+    },
+    {
+      "epoch": 0.08358895705521473,
+      "grad_norm": 0.4362933933734894,
+      "learning_rate": 0.00014796274694006207,
+      "loss": 1.3024,
+      "step": 545
+    },
+    {
+      "epoch": 0.08374233128834356,
+      "grad_norm": 0.6373287439346313,
+      "learning_rate": 0.00014795428838473306,
+      "loss": 1.2644,
+      "step": 546
+    },
+    {
+      "epoch": 0.0838957055214724,
+      "grad_norm": 0.44138941168785095,
+      "learning_rate": 0.00014794581254883382,
+      "loss": 1.2249,
+      "step": 547
+    },
+    {
+      "epoch": 0.08404907975460123,
+      "grad_norm": 0.6656278967857361,
+      "learning_rate": 0.00014793731943437205,
+      "loss": 1.3615,
+      "step": 548
+    },
+    {
+      "epoch": 0.08420245398773006,
+      "grad_norm": 0.42734357714653015,
+      "learning_rate": 0.00014792880904335943,
+      "loss": 1.2156,
+      "step": 549
+    },
+    {
+      "epoch": 0.0843558282208589,
+      "grad_norm": 0.6081749796867371,
+      "learning_rate": 0.00014792028137781187,
+      "loss": 1.4341,
+      "step": 550
+    },
+    {
+      "epoch": 0.08450920245398773,
+      "grad_norm": 0.39509207010269165,
+      "learning_rate": 0.00014791173643974924,
+      "loss": 1.3268,
+      "step": 551
+    },
+    {
+      "epoch": 0.08466257668711656,
+      "grad_norm": 0.775597095489502,
+      "learning_rate": 0.00014790317423119562,
+      "loss": 1.231,
+      "step": 552
+    },
+    {
+      "epoch": 0.0848159509202454,
+      "grad_norm": 0.404861181974411,
+      "learning_rate": 0.0001478945947541791,
+      "loss": 1.4337,
+      "step": 553
+    },
+    {
+      "epoch": 0.08496932515337423,
+      "grad_norm": 0.8275524377822876,
+      "learning_rate": 0.00014788599801073192,
+      "loss": 1.4491,
+      "step": 554
+    },
+    {
+      "epoch": 0.08512269938650306,
+      "grad_norm": 0.4180584251880646,
+      "learning_rate": 0.0001478773840028903,
+      "loss": 1.2126,
+      "step": 555
+    },
+    {
+      "epoch": 0.0852760736196319,
+      "grad_norm": 0.599632203578949,
+      "learning_rate": 0.00014786875273269473,
+      "loss": 1.3695,
+      "step": 556
+    },
+    {
+      "epoch": 0.08542944785276074,
+      "grad_norm": 0.5414772033691406,
+      "learning_rate": 0.0001478601042021896,
+      "loss": 1.2723,
+      "step": 557
+    },
+    {
+      "epoch": 0.08558282208588958,
+      "grad_norm": 0.5061226487159729,
+      "learning_rate": 0.00014785143841342352,
+      "loss": 1.3186,
+      "step": 558
+    },
+    {
+      "epoch": 0.08573619631901841,
+      "grad_norm": 0.41651973128318787,
+      "learning_rate": 0.00014784275536844916,
+      "loss": 1.4098,
+      "step": 559
+    },
+    {
+      "epoch": 0.08588957055214724,
+      "grad_norm": 0.5278149247169495,
+      "learning_rate": 0.00014783405506932321,
+      "loss": 1.3238,
+      "step": 560
+    },
+    {
+      "epoch": 0.08604294478527608,
+      "grad_norm": 0.4451618492603302,
+      "learning_rate": 0.0001478253375181065,
+      "loss": 1.152,
+      "step": 561
+    },
+    {
+      "epoch": 0.08619631901840491,
+      "grad_norm": 0.4073576331138611,
+      "learning_rate": 0.000147816602716864,
+      "loss": 1.3643,
+      "step": 562
+    },
+    {
+      "epoch": 0.08634969325153374,
+      "grad_norm": 0.4320569932460785,
+      "learning_rate": 0.00014780785066766464,
+      "loss": 1.3134,
+      "step": 563
+    },
+    {
+      "epoch": 0.08650306748466258,
+      "grad_norm": 0.42690572142601013,
+      "learning_rate": 0.00014779908137258152,
+      "loss": 1.2915,
+      "step": 564
+    },
+    {
+      "epoch": 0.08665644171779141,
+      "grad_norm": 0.5024705529212952,
+      "learning_rate": 0.00014779029483369185,
+      "loss": 1.2805,
+      "step": 565
+    },
+    {
+      "epoch": 0.08680981595092024,
+      "grad_norm": 0.6136749386787415,
+      "learning_rate": 0.00014778149105307685,
+      "loss": 1.2383,
+      "step": 566
+    },
+    {
+      "epoch": 0.08696319018404908,
+      "grad_norm": 0.47031715512275696,
+      "learning_rate": 0.00014777267003282188,
+      "loss": 1.3345,
+      "step": 567
+    },
+    {
+      "epoch": 0.08711656441717791,
+      "grad_norm": 0.3894667327404022,
+      "learning_rate": 0.00014776383177501632,
+      "loss": 1.1828,
+      "step": 568
+    },
+    {
+      "epoch": 0.08726993865030674,
+      "grad_norm": 0.48029863834381104,
+      "learning_rate": 0.0001477549762817537,
+      "loss": 1.3297,
+      "step": 569
+    },
+    {
+      "epoch": 0.08742331288343558,
+      "grad_norm": 0.5909090638160706,
+      "learning_rate": 0.00014774610355513165,
+      "loss": 1.3705,
+      "step": 570
+    },
+    {
+      "epoch": 0.08757668711656441,
+      "grad_norm": 0.510029673576355,
+      "learning_rate": 0.00014773721359725176,
+      "loss": 1.0939,
+      "step": 571
+    },
+    {
+      "epoch": 0.08773006134969324,
+      "grad_norm": 0.45577895641326904,
+      "learning_rate": 0.0001477283064102198,
+      "loss": 1.2735,
+      "step": 572
+    },
+    {
+      "epoch": 0.08788343558282209,
+      "grad_norm": 0.4269903898239136,
+      "learning_rate": 0.0001477193819961457,
+      "loss": 1.2794,
+      "step": 573
+    },
+    {
+      "epoch": 0.08803680981595093,
+      "grad_norm": 0.43689975142478943,
+      "learning_rate": 0.0001477104403571432,
+      "loss": 1.2906,
+      "step": 574
+    },
+    {
+      "epoch": 0.08819018404907976,
+      "grad_norm": 0.4200446903705597,
+      "learning_rate": 0.00014770148149533042,
+      "loss": 1.222,
+      "step": 575
+    },
+    {
+      "epoch": 0.08834355828220859,
+      "grad_norm": 0.4607657790184021,
+      "learning_rate": 0.00014769250541282942,
+      "loss": 1.3368,
+      "step": 576
+    },
+    {
+      "epoch": 0.08849693251533743,
+      "grad_norm": 0.3637036085128784,
+      "learning_rate": 0.00014768351211176628,
+      "loss": 1.3924,
+      "step": 577
+    },
+    {
+      "epoch": 0.08865030674846626,
+      "grad_norm": 0.4128216803073883,
+      "learning_rate": 0.0001476745015942713,
+      "loss": 1.195,
+      "step": 578
+    },
+    {
+      "epoch": 0.08880368098159509,
+      "grad_norm": 0.4523659348487854,
+      "learning_rate": 0.0001476654738624788,
+      "loss": 1.1272,
+      "step": 579
+    },
+    {
+      "epoch": 0.08895705521472393,
+      "grad_norm": 0.4197641909122467,
+      "learning_rate": 0.00014765642891852708,
+      "loss": 1.2259,
+      "step": 580
+    },
+    {
+      "epoch": 0.08911042944785276,
+      "grad_norm": 0.39657992124557495,
+      "learning_rate": 0.00014764736676455868,
+      "loss": 1.373,
+      "step": 581
+    },
+    {
+      "epoch": 0.0892638036809816,
+      "grad_norm": 0.41460514068603516,
+      "learning_rate": 0.0001476382874027201,
+      "loss": 1.3429,
+      "step": 582
+    },
+    {
+      "epoch": 0.08941717791411043,
+      "grad_norm": 0.4582538604736328,
+      "learning_rate": 0.00014762919083516194,
+      "loss": 1.2286,
+      "step": 583
+    },
+    {
+      "epoch": 0.08957055214723926,
+      "grad_norm": 0.36807987093925476,
+      "learning_rate": 0.00014762007706403896,
+      "loss": 1.3341,
+      "step": 584
+    },
+    {
+      "epoch": 0.0897239263803681,
+      "grad_norm": 0.38717028498649597,
+      "learning_rate": 0.00014761094609150985,
+      "loss": 1.1607,
+      "step": 585
+    },
+    {
+      "epoch": 0.08987730061349693,
+      "grad_norm": 0.7174474000930786,
+      "learning_rate": 0.00014760179791973747,
+      "loss": 1.1726,
+      "step": 586
+    },
+    {
+      "epoch": 0.09003067484662576,
+      "grad_norm": 0.671405553817749,
+      "learning_rate": 0.00014759263255088878,
+      "loss": 1.2684,
+      "step": 587
+    },
+    {
+      "epoch": 0.09018404907975461,
+      "grad_norm": 0.7479199171066284,
+      "learning_rate": 0.0001475834499871347,
+      "loss": 1.366,
+      "step": 588
+    },
+    {
+      "epoch": 0.09033742331288344,
+      "grad_norm": 0.7044978737831116,
+      "learning_rate": 0.0001475742502306503,
+      "loss": 1.2193,
+      "step": 589
+    },
+    {
+      "epoch": 0.09049079754601227,
+      "grad_norm": 0.6694434285163879,
+      "learning_rate": 0.00014756503328361474,
+      "loss": 1.204,
+      "step": 590
+    },
+    {
+      "epoch": 0.09064417177914111,
+      "grad_norm": 0.6148474216461182,
+      "learning_rate": 0.00014755579914821123,
+      "loss": 1.4125,
+      "step": 591
+    },
+    {
+      "epoch": 0.09079754601226994,
+      "grad_norm": 1.0183055400848389,
+      "learning_rate": 0.00014754654782662703,
+      "loss": 1.2666,
+      "step": 592
+    },
+    {
+      "epoch": 0.09095092024539878,
+      "grad_norm": 0.501554012298584,
+      "learning_rate": 0.0001475372793210535,
+      "loss": 1.2915,
+      "step": 593
+    },
+    {
+      "epoch": 0.09110429447852761,
+      "grad_norm": 0.698258101940155,
+      "learning_rate": 0.000147527993633686,
+      "loss": 1.3616,
+      "step": 594
+    },
+    {
+      "epoch": 0.09125766871165644,
+      "grad_norm": 0.4906356632709503,
+      "learning_rate": 0.00014751869076672411,
+      "loss": 1.222,
+      "step": 595
+    },
+    {
+      "epoch": 0.09141104294478528,
+      "grad_norm": 0.47147291898727417,
+      "learning_rate": 0.0001475093707223713,
+      "loss": 1.276,
+      "step": 596
+    },
+    {
+      "epoch": 0.09156441717791411,
+      "grad_norm": 1.2358698844909668,
+      "learning_rate": 0.00014750003350283522,
+      "loss": 1.2026,
+      "step": 597
+    },
+    {
+      "epoch": 0.09171779141104294,
+      "grad_norm": 0.6165987253189087,
+      "learning_rate": 0.0001474906791103276,
+      "loss": 1.3868,
+      "step": 598
+    },
+    {
+      "epoch": 0.09187116564417178,
+      "grad_norm": 0.6322820782661438,
+      "learning_rate": 0.00014748130754706418,
+      "loss": 1.3276,
+      "step": 599
+    },
+    {
+      "epoch": 0.09202453987730061,
+      "grad_norm": 0.48793137073516846,
+      "learning_rate": 0.00014747191881526477,
+      "loss": 1.2175,
+      "step": 600
+    },
+    {
+      "epoch": 0.09217791411042944,
+      "grad_norm": 1.005854845046997,
+      "learning_rate": 0.00014746251291715325,
+      "loss": 1.4285,
+      "step": 601
+    },
+    {
+      "epoch": 0.09233128834355828,
+      "grad_norm": 0.5546073317527771,
+      "learning_rate": 0.00014745308985495765,
+      "loss": 1.2079,
+      "step": 602
+    },
+    {
+      "epoch": 0.09248466257668711,
+      "grad_norm": 0.8411008715629578,
+      "learning_rate": 0.00014744364963090992,
+      "loss": 1.3627,
+      "step": 603
+    },
+    {
+      "epoch": 0.09263803680981596,
+      "grad_norm": 0.5839352011680603,
+      "learning_rate": 0.00014743419224724622,
+      "loss": 1.3488,
+      "step": 604
+    },
+    {
+      "epoch": 0.09279141104294479,
+      "grad_norm": 1.1033393144607544,
+      "learning_rate": 0.00014742471770620664,
+      "loss": 1.3911,
+      "step": 605
+    },
+    {
+      "epoch": 0.09294478527607362,
+      "grad_norm": 0.49690014123916626,
+      "learning_rate": 0.00014741522601003542,
+      "loss": 1.3681,
+      "step": 606
+    },
+    {
+      "epoch": 0.09309815950920246,
+      "grad_norm": 0.5824717879295349,
+      "learning_rate": 0.00014740571716098089,
+      "loss": 1.2094,
+      "step": 607
+    },
+    {
+      "epoch": 0.09325153374233129,
+      "grad_norm": 0.4469147026538849,
+      "learning_rate": 0.0001473961911612953,
+      "loss": 1.4052,
+      "step": 608
+    },
+    {
+      "epoch": 0.09340490797546012,
+      "grad_norm": 0.5548704266548157,
+      "learning_rate": 0.0001473866480132352,
+      "loss": 1.3637,
+      "step": 609
+    },
+    {
+      "epoch": 0.09355828220858896,
+      "grad_norm": 0.5111848711967468,
+      "learning_rate": 0.00014737708771906093,
+      "loss": 1.3237,
+      "step": 610
+    },
+    {
+      "epoch": 0.09371165644171779,
+      "grad_norm": 0.5298082828521729,
+      "learning_rate": 0.00014736751028103707,
+      "loss": 1.2761,
+      "step": 611
+    },
+    {
+      "epoch": 0.09386503067484663,
+      "grad_norm": 0.5042627453804016,
+      "learning_rate": 0.00014735791570143222,
+      "loss": 1.3804,
+      "step": 612
+    },
+    {
+      "epoch": 0.09401840490797546,
+      "grad_norm": 0.7127280831336975,
+      "learning_rate": 0.00014734830398251903,
+      "loss": 1.1849,
+      "step": 613
+    },
+    {
+      "epoch": 0.09417177914110429,
+      "grad_norm": 0.5852988958358765,
+      "learning_rate": 0.00014733867512657422,
+      "loss": 1.1847,
+      "step": 614
+    },
+    {
+      "epoch": 0.09432515337423313,
+      "grad_norm": 0.5466170310974121,
+      "learning_rate": 0.00014732902913587855,
+      "loss": 1.4012,
+      "step": 615
+    },
+    {
+      "epoch": 0.09447852760736196,
+      "grad_norm": 0.4419586658477783,
+      "learning_rate": 0.00014731936601271686,
+      "loss": 1.1321,
+      "step": 616
+    },
+    {
+      "epoch": 0.09463190184049079,
+      "grad_norm": 0.4462755024433136,
+      "learning_rate": 0.000147309685759378,
+      "loss": 1.3095,
+      "step": 617
+    },
+    {
+      "epoch": 0.09478527607361963,
+      "grad_norm": 0.6070041060447693,
+      "learning_rate": 0.000147299988378155,
+      "loss": 1.2674,
+      "step": 618
+    },
+    {
+      "epoch": 0.09493865030674846,
+      "grad_norm": 0.6949111223220825,
+      "learning_rate": 0.0001472902738713448,
+      "loss": 1.3123,
+      "step": 619
+    },
+    {
+      "epoch": 0.0950920245398773,
+      "grad_norm": 0.40998876094818115,
+      "learning_rate": 0.00014728054224124844,
+      "loss": 1.2802,
+      "step": 620
+    },
+    {
+      "epoch": 0.09524539877300614,
+      "grad_norm": 0.4410196840763092,
+      "learning_rate": 0.0001472707934901711,
+      "loss": 1.3769,
+      "step": 621
+    },
+    {
+      "epoch": 0.09539877300613497,
+      "grad_norm": 0.5631739497184753,
+      "learning_rate": 0.0001472610276204219,
+      "loss": 1.2597,
+      "step": 622
+    },
+    {
+      "epoch": 0.09555214723926381,
+      "grad_norm": 0.5435857176780701,
+      "learning_rate": 0.00014725124463431413,
+      "loss": 1.1757,
+      "step": 623
+    },
+    {
+      "epoch": 0.09570552147239264,
+      "grad_norm": 0.7644972801208496,
+      "learning_rate": 0.00014724144453416498,
+      "loss": 1.2717,
+      "step": 624
+    },
+    {
+      "epoch": 0.09585889570552147,
+      "grad_norm": 0.6015825867652893,
+      "learning_rate": 0.00014723162732229586,
+      "loss": 1.2463,
+      "step": 625
+    },
+    {
+      "epoch": 0.09601226993865031,
+      "grad_norm": 0.3898197412490845,
+      "learning_rate": 0.00014722179300103214,
+      "loss": 1.4581,
+      "step": 626
+    },
+    {
+      "epoch": 0.09616564417177914,
+      "grad_norm": 0.5000936388969421,
+      "learning_rate": 0.00014721194157270325,
+      "loss": 1.2033,
+      "step": 627
+    },
+    {
+      "epoch": 0.09631901840490797,
+      "grad_norm": 0.44766244292259216,
+      "learning_rate": 0.00014720207303964267,
+      "loss": 1.3494,
+      "step": 628
+    },
+    {
+      "epoch": 0.09647239263803681,
+      "grad_norm": 0.41217562556266785,
+      "learning_rate": 0.000147192187404188,
+      "loss": 1.3168,
+      "step": 629
+    },
+    {
+      "epoch": 0.09662576687116564,
+      "grad_norm": 0.6604576706886292,
+      "learning_rate": 0.00014718228466868076,
+      "loss": 1.2897,
+      "step": 630
+    },
+    {
+      "epoch": 0.09677914110429447,
+      "grad_norm": 0.6487378478050232,
+      "learning_rate": 0.00014717236483546663,
+      "loss": 1.2109,
+      "step": 631
+    },
+    {
+      "epoch": 0.09693251533742331,
+      "grad_norm": 0.5256462693214417,
+      "learning_rate": 0.00014716242790689533,
+      "loss": 1.2156,
+      "step": 632
+    },
+    {
+      "epoch": 0.09708588957055214,
+      "grad_norm": 0.9279944896697998,
+      "learning_rate": 0.00014715247388532055,
+      "loss": 1.2752,
+      "step": 633
+    },
+    {
+      "epoch": 0.09723926380368098,
+      "grad_norm": 0.4033508002758026,
+      "learning_rate": 0.00014714250277310013,
+      "loss": 1.2557,
+      "step": 634
+    },
+    {
+      "epoch": 0.09739263803680982,
+      "grad_norm": 0.5507687330245972,
+      "learning_rate": 0.00014713251457259592,
+      "loss": 1.4868,
+      "step": 635
+    },
+    {
+      "epoch": 0.09754601226993866,
+      "grad_norm": 0.4332635700702667,
+      "learning_rate": 0.00014712250928617374,
+      "loss": 1.2415,
+      "step": 636
+    },
+    {
+      "epoch": 0.09769938650306749,
+      "grad_norm": 0.5033321976661682,
+      "learning_rate": 0.00014711248691620362,
+      "loss": 1.2995,
+      "step": 637
+    },
+    {
+      "epoch": 0.09785276073619632,
+      "grad_norm": 0.6211531162261963,
+      "learning_rate": 0.00014710244746505947,
+      "loss": 1.3006,
+      "step": 638
+    },
+    {
+      "epoch": 0.09800613496932516,
+      "grad_norm": 0.3979441523551941,
+      "learning_rate": 0.00014709239093511935,
+      "loss": 1.2704,
+      "step": 639
+    },
+    {
+      "epoch": 0.09815950920245399,
+      "grad_norm": 0.7468326091766357,
+      "learning_rate": 0.00014708231732876535,
+      "loss": 1.1697,
+      "step": 640
+    },
+    {
+      "epoch": 0.09831288343558282,
+      "grad_norm": 0.5905004143714905,
+      "learning_rate": 0.00014707222664838356,
+      "loss": 1.283,
+      "step": 641
+    },
+    {
+      "epoch": 0.09846625766871166,
+      "grad_norm": 0.6057485342025757,
+      "learning_rate": 0.00014706211889636416,
+      "loss": 1.3074,
+      "step": 642
+    },
+    {
+      "epoch": 0.09861963190184049,
+      "grad_norm": 0.48701801896095276,
+      "learning_rate": 0.00014705199407510133,
+      "loss": 1.2235,
+      "step": 643
+    },
+    {
+      "epoch": 0.09877300613496932,
+      "grad_norm": 0.5471395254135132,
+      "learning_rate": 0.00014704185218699335,
+      "loss": 1.2812,
+      "step": 644
+    },
+    {
+      "epoch": 0.09892638036809816,
+      "grad_norm": 0.48325681686401367,
+      "learning_rate": 0.00014703169323444253,
+      "loss": 1.1527,
+      "step": 645
+    },
+    {
+      "epoch": 0.09907975460122699,
+      "grad_norm": 0.601097047328949,
+      "learning_rate": 0.00014702151721985516,
+      "loss": 1.1992,
+      "step": 646
+    },
+    {
+      "epoch": 0.09923312883435582,
+      "grad_norm": 0.42635834217071533,
+      "learning_rate": 0.00014701132414564168,
+      "loss": 1.1609,
+      "step": 647
+    },
+    {
+      "epoch": 0.09938650306748466,
+      "grad_norm": 0.42430904507637024,
+      "learning_rate": 0.00014700111401421643,
+      "loss": 1.1376,
+      "step": 648
+    },
+    {
+      "epoch": 0.09953987730061349,
+      "grad_norm": 0.3631018102169037,
+      "learning_rate": 0.00014699088682799792,
+      "loss": 1.4393,
+      "step": 649
+    },
+    {
+      "epoch": 0.09969325153374232,
+      "grad_norm": 0.44148290157318115,
+      "learning_rate": 0.0001469806425894086,
+      "loss": 1.2814,
+      "step": 650
+    },
+    {
+      "epoch": 0.09984662576687117,
+      "grad_norm": 0.3927432894706726,
+      "learning_rate": 0.00014697038130087508,
+      "loss": 1.2619,
+      "step": 651
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4822012186050415,
+      "learning_rate": 0.0001469601029648279,
+      "loss": 1.21,
+      "step": 652
+    },
+    {
+      "epoch": 0.10015337423312884,
+      "grad_norm": 0.5334418416023254,
+      "learning_rate": 0.00014694980758370163,
+      "loss": 1.3901,
+      "step": 653
+    },
+    {
+      "epoch": 0.10030674846625767,
+      "grad_norm": 0.6232284903526306,
+      "learning_rate": 0.000146939495159935,
+      "loss": 1.4457,
+      "step": 654
+    },
+    {
+      "epoch": 0.1004601226993865,
+      "grad_norm": 0.3717365860939026,
+      "learning_rate": 0.00014692916569597063,
+      "loss": 1.27,
+      "step": 655
+    },
+    {
+      "epoch": 0.10061349693251534,
+      "grad_norm": 0.5195326805114746,
+      "learning_rate": 0.0001469188191942553,
+      "loss": 1.3111,
+      "step": 656
+    },
+    {
+      "epoch": 0.10076687116564417,
+      "grad_norm": 0.360432505607605,
+      "learning_rate": 0.00014690845565723973,
+      "loss": 1.1315,
+      "step": 657
+    },
+    {
+      "epoch": 0.100920245398773,
+      "grad_norm": 0.5961891412734985,
+      "learning_rate": 0.00014689807508737874,
+      "loss": 1.2611,
+      "step": 658
+    },
+    {
+      "epoch": 0.10107361963190184,
+      "grad_norm": 0.5239169001579285,
+      "learning_rate": 0.00014688767748713112,
+      "loss": 1.1239,
+      "step": 659
+    },
+    {
+      "epoch": 0.10122699386503067,
+      "grad_norm": 0.38524919748306274,
+      "learning_rate": 0.0001468772628589598,
+      "loss": 1.1712,
+      "step": 660
+    },
+    {
+      "epoch": 0.1013803680981595,
+      "grad_norm": 0.3785124719142914,
+      "learning_rate": 0.00014686683120533161,
+      "loss": 1.3438,
+      "step": 661
+    },
+    {
+      "epoch": 0.10153374233128834,
+      "grad_norm": 0.4589962959289551,
+      "learning_rate": 0.0001468563825287175,
+      "loss": 1.1021,
+      "step": 662
+    },
+    {
+      "epoch": 0.10168711656441717,
+      "grad_norm": 0.443163126707077,
+      "learning_rate": 0.00014684591683159244,
+      "loss": 1.182,
+      "step": 663
+    },
+    {
+      "epoch": 0.10184049079754601,
+      "grad_norm": 1.1890807151794434,
+      "learning_rate": 0.00014683543411643548,
+      "loss": 1.3975,
+      "step": 664
+    },
+    {
+      "epoch": 0.10199386503067484,
+      "grad_norm": 0.44043096899986267,
+      "learning_rate": 0.00014682493438572953,
+      "loss": 1.2248,
+      "step": 665
+    },
+    {
+      "epoch": 0.10214723926380367,
+      "grad_norm": 0.8207892179489136,
+      "learning_rate": 0.00014681441764196173,
+      "loss": 1.1933,
+      "step": 666
+    },
+    {
+      "epoch": 0.10230061349693252,
+      "grad_norm": 0.4834432303905487,
+      "learning_rate": 0.00014680388388762312,
+      "loss": 1.2117,
+      "step": 667
+    },
+    {
+      "epoch": 0.10245398773006135,
+      "grad_norm": 0.8632691502571106,
+      "learning_rate": 0.00014679333312520886,
+      "loss": 1.3375,
+      "step": 668
+    },
+    {
+      "epoch": 0.10260736196319019,
+      "grad_norm": 0.48036423325538635,
+      "learning_rate": 0.00014678276535721804,
+      "loss": 1.2649,
+      "step": 669
+    },
+    {
+      "epoch": 0.10276073619631902,
+      "grad_norm": 0.5681169033050537,
+      "learning_rate": 0.00014677218058615384,
+      "loss": 1.312,
+      "step": 670
+    },
+    {
+      "epoch": 0.10291411042944786,
+      "grad_norm": 0.5166031122207642,
+      "learning_rate": 0.00014676157881452352,
+      "loss": 1.3123,
+      "step": 671
+    },
+    {
+      "epoch": 0.10306748466257669,
+      "grad_norm": 0.45652541518211365,
+      "learning_rate": 0.00014675096004483824,
+      "loss": 1.2639,
+      "step": 672
+    },
+    {
+      "epoch": 0.10322085889570552,
+      "grad_norm": 0.5452796220779419,
+      "learning_rate": 0.00014674032427961327,
+      "loss": 1.2562,
+      "step": 673
+    },
+    {
+      "epoch": 0.10337423312883436,
+      "grad_norm": 0.8694941997528076,
+      "learning_rate": 0.00014672967152136787,
+      "loss": 1.2638,
+      "step": 674
+    },
+    {
+      "epoch": 0.10352760736196319,
+      "grad_norm": 0.45909711718559265,
+      "learning_rate": 0.0001467190017726254,
+      "loss": 1.2926,
+      "step": 675
+    },
+    {
+      "epoch": 0.10368098159509202,
+      "grad_norm": 0.7866394519805908,
+      "learning_rate": 0.00014670831503591313,
+      "loss": 1.1234,
+      "step": 676
+    },
+    {
+      "epoch": 0.10383435582822086,
+      "grad_norm": 0.5738254189491272,
+      "learning_rate": 0.00014669761131376244,
+      "loss": 1.2861,
+      "step": 677
+    },
+    {
+      "epoch": 0.10398773006134969,
+      "grad_norm": 0.6348480582237244,
+      "learning_rate": 0.0001466868906087087,
+      "loss": 1.2894,
+      "step": 678
+    },
+    {
+      "epoch": 0.10414110429447852,
+      "grad_norm": 0.40673431754112244,
+      "learning_rate": 0.0001466761529232913,
+      "loss": 1.3662,
+      "step": 679
+    },
+    {
+      "epoch": 0.10429447852760736,
+      "grad_norm": 0.6973959803581238,
+      "learning_rate": 0.00014666539826005363,
+      "loss": 1.2144,
+      "step": 680
+    },
+    {
+      "epoch": 0.10444785276073619,
+      "grad_norm": 0.5669949054718018,
+      "learning_rate": 0.00014665462662154317,
+      "loss": 1.2531,
+      "step": 681
+    },
+    {
+      "epoch": 0.10460122699386504,
+      "grad_norm": 1.042041301727295,
+      "learning_rate": 0.00014664383801031138,
+      "loss": 1.2741,
+      "step": 682
+    },
+    {
+      "epoch": 0.10475460122699387,
+      "grad_norm": 0.4904883801937103,
+      "learning_rate": 0.00014663303242891374,
+      "loss": 1.248,
+      "step": 683
+    },
+    {
+      "epoch": 0.1049079754601227,
+      "grad_norm": 0.5515105128288269,
+      "learning_rate": 0.00014662220987990976,
+      "loss": 1.1735,
+      "step": 684
+    },
+    {
+      "epoch": 0.10506134969325154,
+      "grad_norm": 0.45439085364341736,
+      "learning_rate": 0.00014661137036586292,
+      "loss": 1.2504,
+      "step": 685
+    },
+    {
+      "epoch": 0.10521472392638037,
+      "grad_norm": 0.5754830241203308,
+      "learning_rate": 0.0001466005138893408,
+      "loss": 1.0789,
+      "step": 686
+    },
+    {
+      "epoch": 0.1053680981595092,
+      "grad_norm": 0.7441419363021851,
+      "learning_rate": 0.00014658964045291493,
+      "loss": 1.2921,
+      "step": 687
+    },
+    {
+      "epoch": 0.10552147239263804,
+      "grad_norm": 0.4807514250278473,
+      "learning_rate": 0.0001465787500591609,
+      "loss": 1.475,
+      "step": 688
+    },
+    {
+      "epoch": 0.10567484662576687,
+      "grad_norm": 0.562820315361023,
+      "learning_rate": 0.0001465678427106583,
+      "loss": 1.3039,
+      "step": 689
+    },
+    {
+      "epoch": 0.1058282208588957,
+      "grad_norm": 0.7970765233039856,
+      "learning_rate": 0.00014655691840999074,
+      "loss": 1.2619,
+      "step": 690
+    },
+    {
+      "epoch": 0.10598159509202454,
+      "grad_norm": 0.6146219968795776,
+      "learning_rate": 0.0001465459771597458,
+      "loss": 1.1856,
+      "step": 691
+    },
+    {
+      "epoch": 0.10613496932515337,
+      "grad_norm": 0.9856784343719482,
+      "learning_rate": 0.0001465350189625152,
+      "loss": 1.2256,
+      "step": 692
+    },
+    {
+      "epoch": 0.1062883435582822,
+      "grad_norm": 0.7528601288795471,
+      "learning_rate": 0.0001465240438208945,
+      "loss": 1.2629,
+      "step": 693
+    },
+    {
+      "epoch": 0.10644171779141104,
+      "grad_norm": 0.4854978919029236,
+      "learning_rate": 0.00014651305173748346,
+      "loss": 1.2831,
+      "step": 694
+    },
+    {
+      "epoch": 0.10659509202453987,
+      "grad_norm": 0.4876590967178345,
+      "learning_rate": 0.00014650204271488565,
+      "loss": 1.2088,
+      "step": 695
+    },
+    {
+      "epoch": 0.1067484662576687,
+      "grad_norm": 0.5084669589996338,
+      "learning_rate": 0.00014649101675570886,
+      "loss": 1.3002,
+      "step": 696
+    },
+    {
+      "epoch": 0.10690184049079754,
+      "grad_norm": 0.41993334889411926,
+      "learning_rate": 0.00014647997386256473,
+      "loss": 1.3686,
+      "step": 697
+    },
+    {
+      "epoch": 0.10705521472392639,
+      "grad_norm": 0.5362027287483215,
+      "learning_rate": 0.00014646891403806902,
+      "loss": 1.153,
+      "step": 698
+    },
+    {
+      "epoch": 0.10720858895705522,
+      "grad_norm": 0.5166748762130737,
+      "learning_rate": 0.0001464578372848414,
+      "loss": 1.0834,
+      "step": 699
+    },
+    {
+      "epoch": 0.10736196319018405,
+      "grad_norm": 0.5641416311264038,
+      "learning_rate": 0.00014644674360550565,
+      "loss": 1.3014,
+      "step": 700
+    },
+    {
+      "epoch": 0.10751533742331289,
+      "grad_norm": 0.5209351181983948,
+      "learning_rate": 0.00014643563300268948,
+      "loss": 1.2206,
+      "step": 701
+    },
+    {
+      "epoch": 0.10766871165644172,
+      "grad_norm": 0.5761834383010864,
+      "learning_rate": 0.00014642450547902464,
+      "loss": 1.2136,
+      "step": 702
+    },
+    {
+      "epoch": 0.10782208588957055,
+      "grad_norm": 0.6117256283760071,
+      "learning_rate": 0.00014641336103714692,
+      "loss": 1.4002,
+      "step": 703
+    },
+    {
+      "epoch": 0.10797546012269939,
+      "grad_norm": 0.41749241948127747,
+      "learning_rate": 0.00014640219967969608,
+      "loss": 1.3204,
+      "step": 704
+    },
+    {
+      "epoch": 0.10812883435582822,
+      "grad_norm": 0.48151376843452454,
+      "learning_rate": 0.0001463910214093159,
+      "loss": 1.1271,
+      "step": 705
+    },
+    {
+      "epoch": 0.10828220858895705,
+      "grad_norm": 0.4332837760448456,
+      "learning_rate": 0.0001463798262286541,
+      "loss": 1.1958,
+      "step": 706
+    },
+    {
+      "epoch": 0.10843558282208589,
+      "grad_norm": 0.40836796164512634,
+      "learning_rate": 0.0001463686141403625,
+      "loss": 1.3194,
+      "step": 707
+    },
+    {
+      "epoch": 0.10858895705521472,
+      "grad_norm": 0.5874041318893433,
+      "learning_rate": 0.00014635738514709695,
+      "loss": 1.0538,
+      "step": 708
+    },
+    {
+      "epoch": 0.10874233128834355,
+      "grad_norm": 0.5597360134124756,
+      "learning_rate": 0.00014634613925151716,
+      "loss": 1.2784,
+      "step": 709
+    },
+    {
+      "epoch": 0.10889570552147239,
+      "grad_norm": 0.7818507552146912,
+      "learning_rate": 0.00014633487645628697,
+      "loss": 1.3195,
+      "step": 710
+    },
+    {
+      "epoch": 0.10904907975460122,
+      "grad_norm": 0.5778970122337341,
+      "learning_rate": 0.00014632359676407418,
+      "loss": 1.302,
+      "step": 711
+    },
+    {
+      "epoch": 0.10920245398773006,
+      "grad_norm": 0.49210768938064575,
+      "learning_rate": 0.00014631230017755057,
+      "loss": 1.404,
+      "step": 712
+    },
+    {
+      "epoch": 0.10935582822085889,
+      "grad_norm": 0.4152347445487976,
+      "learning_rate": 0.00014630098669939198,
+      "loss": 1.2711,
+      "step": 713
+    },
+    {
+      "epoch": 0.10950920245398774,
+      "grad_norm": 0.46252959966659546,
+      "learning_rate": 0.00014628965633227818,
+      "loss": 1.2581,
+      "step": 714
+    },
+    {
+      "epoch": 0.10966257668711657,
+      "grad_norm": 0.705820620059967,
+      "learning_rate": 0.000146278309078893,
+      "loss": 1.2448,
+      "step": 715
+    },
+    {
+      "epoch": 0.1098159509202454,
+      "grad_norm": 0.4767616391181946,
+      "learning_rate": 0.00014626694494192425,
+      "loss": 1.2694,
+      "step": 716
+    },
+    {
+      "epoch": 0.10996932515337424,
+      "grad_norm": 0.6891629695892334,
+      "learning_rate": 0.00014625556392406372,
+      "loss": 1.1287,
+      "step": 717
+    },
+    {
+      "epoch": 0.11012269938650307,
+      "grad_norm": 0.4378080666065216,
+      "learning_rate": 0.0001462441660280072,
+      "loss": 1.2197,
+      "step": 718
+    },
+    {
+      "epoch": 0.1102760736196319,
+      "grad_norm": 1.2264528274536133,
+      "learning_rate": 0.00014623275125645455,
+      "loss": 1.3721,
+      "step": 719
+    },
+    {
+      "epoch": 0.11042944785276074,
+      "grad_norm": 0.3677716553211212,
+      "learning_rate": 0.00014622131961210952,
+      "loss": 1.2189,
+      "step": 720
+    },
+    {
+      "epoch": 0.11058282208588957,
+      "grad_norm": 0.9777968525886536,
+      "learning_rate": 0.0001462098710976799,
+      "loss": 1.1686,
+      "step": 721
+    },
+    {
+      "epoch": 0.1107361963190184,
+      "grad_norm": 0.41792306303977966,
+      "learning_rate": 0.00014619840571587753,
+      "loss": 1.2064,
+      "step": 722
+    },
+    {
+      "epoch": 0.11088957055214724,
+      "grad_norm": 0.5822682976722717,
+      "learning_rate": 0.00014618692346941812,
+      "loss": 1.2519,
+      "step": 723
+    },
+    {
+      "epoch": 0.11104294478527607,
+      "grad_norm": 0.7767218947410583,
+      "learning_rate": 0.00014617542436102154,
+      "loss": 1.3621,
+      "step": 724
+    },
+    {
+      "epoch": 0.1111963190184049,
+      "grad_norm": 0.5194628834724426,
+      "learning_rate": 0.0001461639083934115,
+      "loss": 1.2282,
+      "step": 725
+    },
+    {
+      "epoch": 0.11134969325153374,
+      "grad_norm": 0.5655984282493591,
+      "learning_rate": 0.0001461523755693158,
+      "loss": 1.3131,
+      "step": 726
+    },
+    {
+      "epoch": 0.11150306748466257,
+      "grad_norm": 0.4200071692466736,
+      "learning_rate": 0.00014614082589146623,
+      "loss": 1.3897,
+      "step": 727
+    },
+    {
+      "epoch": 0.1116564417177914,
+      "grad_norm": 0.7865282893180847,
+      "learning_rate": 0.00014612925936259848,
+      "loss": 1.2533,
+      "step": 728
+    },
+    {
+      "epoch": 0.11180981595092025,
+      "grad_norm": 0.43898725509643555,
+      "learning_rate": 0.00014611767598545232,
+      "loss": 1.2558,
+      "step": 729
+    },
+    {
+      "epoch": 0.11196319018404909,
+      "grad_norm": 0.46632713079452515,
+      "learning_rate": 0.0001461060757627715,
+      "loss": 1.1979,
+      "step": 730
+    },
+    {
+      "epoch": 0.11211656441717792,
+      "grad_norm": 0.4613082706928253,
+      "learning_rate": 0.00014609445869730376,
+      "loss": 1.1672,
+      "step": 731
+    },
+    {
+      "epoch": 0.11226993865030675,
+      "grad_norm": 0.4198003113269806,
+      "learning_rate": 0.0001460828247918008,
+      "loss": 1.1799,
+      "step": 732
+    },
+    {
+      "epoch": 0.11242331288343559,
+      "grad_norm": 0.5567918419837952,
+      "learning_rate": 0.0001460711740490183,
+      "loss": 1.1242,
+      "step": 733
+    },
+    {
+      "epoch": 0.11257668711656442,
+      "grad_norm": 0.5446353554725647,
+      "learning_rate": 0.000146059506471716,
+      "loss": 1.1769,
+      "step": 734
+    },
+    {
+      "epoch": 0.11273006134969325,
+      "grad_norm": 0.6399310231208801,
+      "learning_rate": 0.00014604782206265754,
+      "loss": 1.2978,
+      "step": 735
+    },
+    {
+      "epoch": 0.11288343558282209,
+      "grad_norm": 0.4838942587375641,
+      "learning_rate": 0.00014603612082461062,
+      "loss": 1.27,
+      "step": 736
+    },
+    {
+      "epoch": 0.11303680981595092,
+      "grad_norm": 0.7107148170471191,
+      "learning_rate": 0.00014602440276034685,
+      "loss": 1.2664,
+      "step": 737
+    },
+    {
+      "epoch": 0.11319018404907975,
+      "grad_norm": 0.4061047434806824,
+      "learning_rate": 0.00014601266787264193,
+      "loss": 1.0861,
+      "step": 738
+    },
+    {
+      "epoch": 0.11334355828220859,
+      "grad_norm": 0.5581299662590027,
+      "learning_rate": 0.00014600091616427546,
+      "loss": 1.0668,
+      "step": 739
+    },
+    {
+      "epoch": 0.11349693251533742,
+      "grad_norm": 0.6086634993553162,
+      "learning_rate": 0.00014598914763803102,
+      "loss": 1.1926,
+      "step": 740
+    },
+    {
+      "epoch": 0.11365030674846625,
+      "grad_norm": 0.6133199334144592,
+      "learning_rate": 0.00014597736229669627,
+      "loss": 1.2121,
+      "step": 741
+    },
+    {
+      "epoch": 0.11380368098159509,
+      "grad_norm": 0.5800121426582336,
+      "learning_rate": 0.00014596556014306267,
+      "loss": 1.2368,
+      "step": 742
+    },
+    {
+      "epoch": 0.11395705521472392,
+      "grad_norm": 0.4823463261127472,
+      "learning_rate": 0.0001459537411799259,
+      "loss": 1.1943,
+      "step": 743
+    },
+    {
+      "epoch": 0.11411042944785275,
+      "grad_norm": 5.4602766036987305,
+      "learning_rate": 0.00014594190541008544,
+      "loss": 1.2646,
+      "step": 744
+    },
+    {
+      "epoch": 0.1142638036809816,
+      "grad_norm": 1.228232502937317,
+      "learning_rate": 0.00014593005283634484,
+      "loss": 1.2086,
+      "step": 745
+    },
+    {
+      "epoch": 0.11441717791411044,
+      "grad_norm": 0.3682299554347992,
+      "learning_rate": 0.00014591818346151153,
+      "loss": 1.3065,
+      "step": 746
+    },
+    {
+      "epoch": 0.11457055214723927,
+      "grad_norm": 0.8390713930130005,
+      "learning_rate": 0.0001459062972883971,
+      "loss": 1.2582,
+      "step": 747
+    },
+    {
+      "epoch": 0.1147239263803681,
+      "grad_norm": 0.48809686303138733,
+      "learning_rate": 0.0001458943943198169,
+      "loss": 1.3068,
+      "step": 748
+    },
+    {
+      "epoch": 0.11487730061349694,
+      "grad_norm": 0.8421530723571777,
+      "learning_rate": 0.00014588247455859046,
+      "loss": 1.2769,
+      "step": 749
+    },
+    {
+      "epoch": 0.11503067484662577,
+      "grad_norm": 0.49530625343322754,
+      "learning_rate": 0.00014587053800754114,
+      "loss": 1.3534,
+      "step": 750
+    },
+    {
+      "epoch": 0.1151840490797546,
+      "grad_norm": 0.7441544532775879,
+      "learning_rate": 0.00014585858466949634,
+      "loss": 1.1932,
+      "step": 751
+    },
+    {
+      "epoch": 0.11533742331288344,
+      "grad_norm": 0.42309680581092834,
+      "learning_rate": 0.00014584661454728745,
+      "loss": 1.1997,
+      "step": 752
+    },
+    {
+      "epoch": 0.11549079754601227,
+      "grad_norm": 0.4171423614025116,
+      "learning_rate": 0.0001458346276437498,
+      "loss": 1.2274,
+      "step": 753
+    },
+    {
+      "epoch": 0.1156441717791411,
+      "grad_norm": 0.5527334809303284,
+      "learning_rate": 0.00014582262396172272,
+      "loss": 1.3397,
+      "step": 754
+    },
+    {
+      "epoch": 0.11579754601226994,
+      "grad_norm": 0.39605894684791565,
+      "learning_rate": 0.00014581060350404948,
+      "loss": 1.2608,
+      "step": 755
+    },
+    {
+      "epoch": 0.11595092024539877,
+      "grad_norm": 0.43258658051490784,
+      "learning_rate": 0.00014579856627357736,
+      "loss": 1.1513,
+      "step": 756
+    },
+    {
+      "epoch": 0.1161042944785276,
+      "grad_norm": 0.5120593905448914,
+      "learning_rate": 0.0001457865122731576,
+      "loss": 1.2616,
+      "step": 757
+    },
+    {
+      "epoch": 0.11625766871165644,
+      "grad_norm": 0.4305022954940796,
+      "learning_rate": 0.00014577444150564543,
+      "loss": 1.2102,
+      "step": 758
+    },
+    {
+      "epoch": 0.11641104294478527,
+      "grad_norm": 0.385321706533432,
+      "learning_rate": 0.0001457623539739,
+      "loss": 1.152,
+      "step": 759
+    },
+    {
+      "epoch": 0.1165644171779141,
+      "grad_norm": 0.4039337635040283,
+      "learning_rate": 0.00014575024968078448,
+      "loss": 1.2138,
+      "step": 760
+    },
+    {
+      "epoch": 0.11671779141104295,
+      "grad_norm": 0.5123145580291748,
+      "learning_rate": 0.00014573812862916604,
+      "loss": 1.3049,
+      "step": 761
+    },
+    {
+      "epoch": 0.11687116564417178,
+      "grad_norm": 0.54781574010849,
+      "learning_rate": 0.00014572599082191568,
+      "loss": 1.2106,
+      "step": 762
+    },
+    {
+      "epoch": 0.11702453987730062,
+      "grad_norm": 0.3811926245689392,
+      "learning_rate": 0.00014571383626190852,
+      "loss": 1.2806,
+      "step": 763
+    },
+    {
+      "epoch": 0.11717791411042945,
+      "grad_norm": 0.5609819889068604,
+      "learning_rate": 0.00014570166495202363,
+      "loss": 1.2614,
+      "step": 764
+    },
+    {
+      "epoch": 0.11733128834355828,
+      "grad_norm": 0.42993399500846863,
+      "learning_rate": 0.00014568947689514396,
+      "loss": 1.1975,
+      "step": 765
+    },
+    {
+      "epoch": 0.11748466257668712,
+      "grad_norm": 0.539181113243103,
+      "learning_rate": 0.00014567727209415646,
+      "loss": 1.2873,
+      "step": 766
+    },
+    {
+      "epoch": 0.11763803680981595,
+      "grad_norm": 0.36532437801361084,
+      "learning_rate": 0.00014566505055195212,
+      "loss": 1.2508,
+      "step": 767
+    },
+    {
+      "epoch": 0.11779141104294479,
+      "grad_norm": 0.4971954822540283,
+      "learning_rate": 0.00014565281227142576,
+      "loss": 1.1503,
+      "step": 768
+    },
+    {
+      "epoch": 0.11794478527607362,
+      "grad_norm": 1.3910174369812012,
+      "learning_rate": 0.00014564055725547632,
+      "loss": 1.3375,
+      "step": 769
+    },
+    {
+      "epoch": 0.11809815950920245,
+      "grad_norm": 0.5616905689239502,
+      "learning_rate": 0.0001456282855070066,
+      "loss": 1.2785,
+      "step": 770
+    },
+    {
+      "epoch": 0.11825153374233129,
+      "grad_norm": 0.4290854036808014,
+      "learning_rate": 0.00014561599702892338,
+      "loss": 1.2828,
+      "step": 771
+    },
+    {
+      "epoch": 0.11840490797546012,
+      "grad_norm": 0.3684541583061218,
+      "learning_rate": 0.0001456036918241374,
+      "loss": 1.2097,
+      "step": 772
+    },
+    {
+      "epoch": 0.11855828220858895,
+      "grad_norm": 0.46067917346954346,
+      "learning_rate": 0.00014559136989556343,
+      "loss": 1.2501,
+      "step": 773
+    },
+    {
+      "epoch": 0.11871165644171779,
+      "grad_norm": 0.3659669756889343,
+      "learning_rate": 0.00014557903124612007,
+      "loss": 1.229,
+      "step": 774
+    },
+    {
+      "epoch": 0.11886503067484662,
+      "grad_norm": 0.39746764302253723,
+      "learning_rate": 0.00014556667587873004,
+      "loss": 1.1884,
+      "step": 775
+    },
+    {
+      "epoch": 0.11901840490797547,
+      "grad_norm": 0.5905351042747498,
+      "learning_rate": 0.00014555430379631986,
+      "loss": 1.2238,
+      "step": 776
+    },
+    {
+      "epoch": 0.1191717791411043,
+      "grad_norm": 0.4590882658958435,
+      "learning_rate": 0.00014554191500182012,
+      "loss": 1.0162,
+      "step": 777
+    },
+    {
+      "epoch": 0.11932515337423313,
+      "grad_norm": 0.7793764472007751,
+      "learning_rate": 0.00014552950949816532,
+      "loss": 1.3413,
+      "step": 778
+    },
+    {
+      "epoch": 0.11947852760736197,
+      "grad_norm": 0.4517294466495514,
+      "learning_rate": 0.00014551708728829398,
+      "loss": 1.2668,
+      "step": 779
+    },
+    {
+      "epoch": 0.1196319018404908,
+      "grad_norm": 0.7772916555404663,
+      "learning_rate": 0.0001455046483751485,
+      "loss": 1.2702,
+      "step": 780
+    },
+    {
+      "epoch": 0.11978527607361963,
+      "grad_norm": 0.45899680256843567,
+      "learning_rate": 0.00014549219276167524,
+      "loss": 1.1286,
+      "step": 781
+    },
+    {
+      "epoch": 0.11993865030674847,
+      "grad_norm": 1.403138518333435,
+      "learning_rate": 0.00014547972045082458,
+      "loss": 1.473,
+      "step": 782
+    },
+    {
+      "epoch": 0.1200920245398773,
+      "grad_norm": 0.6466749310493469,
+      "learning_rate": 0.0001454672314455508,
+      "loss": 1.2062,
+      "step": 783
+    },
+    {
+      "epoch": 0.12024539877300613,
+      "grad_norm": 0.6214339733123779,
+      "learning_rate": 0.00014545472574881214,
+      "loss": 1.2797,
+      "step": 784
+    },
+    {
+      "epoch": 0.12039877300613497,
+      "grad_norm": 0.39480194449424744,
+      "learning_rate": 0.00014544220336357082,
+      "loss": 1.2681,
+      "step": 785
+    },
+    {
+      "epoch": 0.1205521472392638,
+      "grad_norm": 0.43037939071655273,
+      "learning_rate": 0.00014542966429279298,
+      "loss": 1.2533,
+      "step": 786
+    },
+    {
+      "epoch": 0.12070552147239264,
+      "grad_norm": 0.5847401022911072,
+      "learning_rate": 0.0001454171085394488,
+      "loss": 1.3295,
+      "step": 787
+    },
+    {
+      "epoch": 0.12085889570552147,
+      "grad_norm": 0.5810413360595703,
+      "learning_rate": 0.00014540453610651226,
+      "loss": 1.2127,
+      "step": 788
+    },
+    {
+      "epoch": 0.1210122699386503,
+      "grad_norm": 0.4295422434806824,
+      "learning_rate": 0.00014539194699696138,
+      "loss": 1.3024,
+      "step": 789
+    },
+    {
+      "epoch": 0.12116564417177914,
+      "grad_norm": 0.5489316582679749,
+      "learning_rate": 0.00014537934121377818,
+      "loss": 1.1798,
+      "step": 790
+    },
+    {
+      "epoch": 0.12131901840490797,
+      "grad_norm": 0.42049461603164673,
+      "learning_rate": 0.00014536671875994852,
+      "loss": 1.2132,
+      "step": 791
+    },
+    {
+      "epoch": 0.12147239263803682,
+      "grad_norm": 0.3933230936527252,
+      "learning_rate": 0.0001453540796384623,
+      "loss": 1.1911,
+      "step": 792
+    },
+    {
+      "epoch": 0.12162576687116565,
+      "grad_norm": 0.41901588439941406,
+      "learning_rate": 0.00014534142385231328,
+      "loss": 1.2757,
+      "step": 793
+    },
+    {
+      "epoch": 0.12177914110429448,
+      "grad_norm": 0.36867618560791016,
+      "learning_rate": 0.00014532875140449927,
+      "loss": 1.3445,
+      "step": 794
+    },
+    {
+      "epoch": 0.12193251533742332,
+      "grad_norm": 0.4413376748561859,
+      "learning_rate": 0.00014531606229802193,
+      "loss": 1.1883,
+      "step": 795
+    },
+    {
+      "epoch": 0.12208588957055215,
+      "grad_norm": 0.4728180766105652,
+      "learning_rate": 0.00014530335653588692,
+      "loss": 1.1968,
+      "step": 796
+    },
+    {
+      "epoch": 0.12223926380368098,
+      "grad_norm": 0.5243786573410034,
+      "learning_rate": 0.00014529063412110387,
+      "loss": 1.1579,
+      "step": 797
+    },
+    {
+      "epoch": 0.12239263803680982,
+      "grad_norm": 0.7133045196533203,
+      "learning_rate": 0.00014527789505668626,
+      "loss": 1.2474,
+      "step": 798
+    },
+    {
+      "epoch": 0.12254601226993865,
+      "grad_norm": 0.6316635608673096,
+      "learning_rate": 0.00014526513934565163,
+      "loss": 1.1294,
+      "step": 799
+    },
+    {
+      "epoch": 0.12269938650306748,
+      "grad_norm": 0.6349585056304932,
+      "learning_rate": 0.00014525236699102135,
+      "loss": 1.2842,
+      "step": 800
+    },
+    {
+      "epoch": 0.12285276073619632,
+      "grad_norm": 0.5772204399108887,
+      "learning_rate": 0.00014523957799582085,
+      "loss": 1.2504,
+      "step": 801
+    },
+    {
+      "epoch": 0.12300613496932515,
+      "grad_norm": 0.7428522706031799,
+      "learning_rate": 0.00014522677236307937,
+      "loss": 1.2806,
+      "step": 802
+    },
+    {
+      "epoch": 0.12315950920245398,
+      "grad_norm": 0.49810150265693665,
+      "learning_rate": 0.00014521395009583024,
+      "loss": 1.0786,
+      "step": 803
+    },
+    {
+      "epoch": 0.12331288343558282,
+      "grad_norm": 0.5507228374481201,
+      "learning_rate": 0.00014520111119711057,
+      "loss": 1.1571,
+      "step": 804
+    },
+    {
+      "epoch": 0.12346625766871165,
+      "grad_norm": 0.49051791429519653,
+      "learning_rate": 0.00014518825566996153,
+      "loss": 1.1739,
+      "step": 805
+    },
+    {
+      "epoch": 0.12361963190184048,
+      "grad_norm": 0.8210214376449585,
+      "learning_rate": 0.0001451753835174282,
+      "loss": 1.298,
+      "step": 806
+    },
+    {
+      "epoch": 0.12377300613496932,
+      "grad_norm": 0.5052042603492737,
+      "learning_rate": 0.0001451624947425596,
+      "loss": 1.264,
+      "step": 807
+    },
+    {
+      "epoch": 0.12392638036809817,
+      "grad_norm": 0.7050299048423767,
+      "learning_rate": 0.0001451495893484086,
+      "loss": 1.2675,
+      "step": 808
+    },
+    {
+      "epoch": 0.124079754601227,
+      "grad_norm": 0.5001578330993652,
+      "learning_rate": 0.00014513666733803217,
+      "loss": 1.2113,
+      "step": 809
+    },
+    {
+      "epoch": 0.12423312883435583,
+      "grad_norm": 0.40423932671546936,
+      "learning_rate": 0.00014512372871449105,
+      "loss": 1.347,
+      "step": 810
+    },
+    {
+      "epoch": 0.12438650306748467,
+      "grad_norm": 0.3877594470977783,
+      "learning_rate": 0.00014511077348085003,
+      "loss": 1.2561,
+      "step": 811
+    },
+    {
+      "epoch": 0.1245398773006135,
+      "grad_norm": 0.5787234902381897,
+      "learning_rate": 0.0001450978016401778,
+      "loss": 1.2541,
+      "step": 812
+    },
+    {
+      "epoch": 0.12469325153374233,
+      "grad_norm": 0.42153164744377136,
+      "learning_rate": 0.000145084813195547,
+      "loss": 1.3494,
+      "step": 813
+    },
+    {
+      "epoch": 0.12484662576687117,
+      "grad_norm": 0.47907838225364685,
+      "learning_rate": 0.00014507180815003415,
+      "loss": 1.2537,
+      "step": 814
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 0.37564322352409363,
+      "learning_rate": 0.00014505878650671972,
+      "loss": 1.1189,
+      "step": 815
+    },
+    {
+      "epoch": 0.12515337423312883,
+      "grad_norm": 0.4874243438243866,
+      "learning_rate": 0.00014504574826868817,
+      "loss": 1.1676,
+      "step": 816
+    },
+    {
+      "epoch": 0.12530674846625767,
+      "grad_norm": 0.45957306027412415,
+      "learning_rate": 0.00014503269343902781,
+      "loss": 1.1126,
+      "step": 817
+    },
+    {
+      "epoch": 0.1254601226993865,
+      "grad_norm": 0.5088058710098267,
+      "learning_rate": 0.00014501962202083095,
+      "loss": 1.0584,
+      "step": 818
+    },
+    {
+      "epoch": 0.12561349693251533,
+      "grad_norm": 0.7506797313690186,
+      "learning_rate": 0.00014500653401719376,
+      "loss": 1.2008,
+      "step": 819
+    },
+    {
+      "epoch": 0.12576687116564417,
+      "grad_norm": 0.6399716734886169,
+      "learning_rate": 0.00014499342943121642,
+      "loss": 1.2454,
+      "step": 820
+    },
+    {
+      "epoch": 0.125920245398773,
+      "grad_norm": 0.5990877151489258,
+      "learning_rate": 0.00014498030826600293,
+      "loss": 1.322,
+      "step": 821
+    },
+    {
+      "epoch": 0.12607361963190183,
+      "grad_norm": 0.4886583387851715,
+      "learning_rate": 0.00014496717052466134,
+      "loss": 1.1405,
+      "step": 822
+    },
+    {
+      "epoch": 0.12622699386503067,
+      "grad_norm": 0.5315207242965698,
+      "learning_rate": 0.00014495401621030356,
+      "loss": 1.317,
+      "step": 823
+    },
+    {
+      "epoch": 0.1263803680981595,
+      "grad_norm": 0.3816460967063904,
+      "learning_rate": 0.00014494084532604541,
+      "loss": 1.2151,
+      "step": 824
+    },
+    {
+      "epoch": 0.12653374233128833,
+      "grad_norm": 0.46838194131851196,
+      "learning_rate": 0.00014492765787500666,
+      "loss": 1.2977,
+      "step": 825
+    },
+    {
+      "epoch": 0.12668711656441717,
+      "grad_norm": 0.442996621131897,
+      "learning_rate": 0.00014491445386031102,
+      "loss": 1.2534,
+      "step": 826
+    },
+    {
+      "epoch": 0.126840490797546,
+      "grad_norm": 0.8242411613464355,
+      "learning_rate": 0.00014490123328508606,
+      "loss": 1.2357,
+      "step": 827
+    },
+    {
+      "epoch": 0.12699386503067484,
+      "grad_norm": 0.4229716360569,
+      "learning_rate": 0.00014488799615246337,
+      "loss": 1.1904,
+      "step": 828
+    },
+    {
+      "epoch": 0.12714723926380367,
+      "grad_norm": 0.43018126487731934,
+      "learning_rate": 0.0001448747424655784,
+      "loss": 1.1557,
+      "step": 829
+    },
+    {
+      "epoch": 0.1273006134969325,
+      "grad_norm": 0.42852962017059326,
+      "learning_rate": 0.0001448614722275705,
+      "loss": 1.2274,
+      "step": 830
+    },
+    {
+      "epoch": 0.12745398773006136,
+      "grad_norm": 0.4913936257362366,
+      "learning_rate": 0.00014484818544158298,
+      "loss": 1.2248,
+      "step": 831
+    },
+    {
+      "epoch": 0.1276073619631902,
+      "grad_norm": 0.36642903089523315,
+      "learning_rate": 0.00014483488211076309,
+      "loss": 1.1951,
+      "step": 832
+    },
+    {
+      "epoch": 0.12776073619631903,
+      "grad_norm": 0.6291371583938599,
+      "learning_rate": 0.00014482156223826192,
+      "loss": 1.2671,
+      "step": 833
+    },
+    {
+      "epoch": 0.12791411042944786,
+      "grad_norm": 0.44361504912376404,
+      "learning_rate": 0.00014480822582723458,
+      "loss": 1.2102,
+      "step": 834
+    },
+    {
+      "epoch": 0.1280674846625767,
+      "grad_norm": 0.49332115054130554,
+      "learning_rate": 0.00014479487288084002,
+      "loss": 1.2882,
+      "step": 835
+    },
+    {
+      "epoch": 0.12822085889570553,
+      "grad_norm": 0.3480657935142517,
+      "learning_rate": 0.00014478150340224114,
+      "loss": 1.2912,
+      "step": 836
+    },
+    {
+      "epoch": 0.12837423312883436,
+      "grad_norm": 0.34533295035362244,
+      "learning_rate": 0.00014476811739460473,
+      "loss": 1.2638,
+      "step": 837
+    },
+    {
+      "epoch": 0.1285276073619632,
+      "grad_norm": 0.4844033122062683,
+      "learning_rate": 0.00014475471486110155,
+      "loss": 1.2285,
+      "step": 838
+    },
+    {
+      "epoch": 0.12868098159509203,
+      "grad_norm": 0.40068188309669495,
+      "learning_rate": 0.00014474129580490617,
+      "loss": 1.3102,
+      "step": 839
+    },
+    {
+      "epoch": 0.12883435582822086,
+      "grad_norm": 0.38730692863464355,
+      "learning_rate": 0.0001447278602291972,
+      "loss": 1.2211,
+      "step": 840
+    },
+    {
+      "epoch": 0.1289877300613497,
+      "grad_norm": 1.1225190162658691,
+      "learning_rate": 0.0001447144081371571,
+      "loss": 1.2554,
+      "step": 841
+    },
+    {
+      "epoch": 0.12914110429447853,
+      "grad_norm": 0.41648203134536743,
+      "learning_rate": 0.00014470093953197223,
+      "loss": 1.2186,
+      "step": 842
+    },
+    {
+      "epoch": 0.12929447852760736,
+      "grad_norm": 0.6380814909934998,
+      "learning_rate": 0.0001446874544168329,
+      "loss": 1.1093,
+      "step": 843
+    },
+    {
+      "epoch": 0.1294478527607362,
+      "grad_norm": 0.3950760066509247,
+      "learning_rate": 0.00014467395279493326,
+      "loss": 1.2497,
+      "step": 844
+    },
+    {
+      "epoch": 0.12960122699386503,
+      "grad_norm": 0.487688809633255,
+      "learning_rate": 0.00014466043466947146,
+      "loss": 1.2004,
+      "step": 845
+    },
+    {
+      "epoch": 0.12975460122699387,
+      "grad_norm": 0.5548173785209656,
+      "learning_rate": 0.00014464690004364948,
+      "loss": 1.3388,
+      "step": 846
+    },
+    {
+      "epoch": 0.1299079754601227,
+      "grad_norm": 0.537651777267456,
+      "learning_rate": 0.00014463334892067333,
+      "loss": 1.3584,
+      "step": 847
+    },
+    {
+      "epoch": 0.13006134969325153,
+      "grad_norm": 0.7923974990844727,
+      "learning_rate": 0.00014461978130375273,
+      "loss": 1.1207,
+      "step": 848
+    },
+    {
+      "epoch": 0.13021472392638037,
+      "grad_norm": 0.45239830017089844,
+      "learning_rate": 0.0001446061971961015,
+      "loss": 1.2447,
+      "step": 849
+    },
+    {
+      "epoch": 0.1303680981595092,
+      "grad_norm": 0.5661341547966003,
+      "learning_rate": 0.00014459259660093727,
+      "loss": 1.2048,
+      "step": 850
+    },
+    {
+      "epoch": 0.13052147239263803,
+      "grad_norm": 0.4260557293891907,
+      "learning_rate": 0.0001445789795214816,
+      "loss": 1.2673,
+      "step": 851
+    },
+    {
+      "epoch": 0.13067484662576687,
+      "grad_norm": 0.5774537920951843,
+      "learning_rate": 0.0001445653459609599,
+      "loss": 1.3267,
+      "step": 852
+    },
+    {
+      "epoch": 0.1308282208588957,
+      "grad_norm": 0.5803173184394836,
+      "learning_rate": 0.00014455169592260156,
+      "loss": 1.1861,
+      "step": 853
+    },
+    {
+      "epoch": 0.13098159509202453,
+      "grad_norm": 0.580595850944519,
+      "learning_rate": 0.00014453802940963986,
+      "loss": 1.2039,
+      "step": 854
+    },
+    {
+      "epoch": 0.13113496932515337,
+      "grad_norm": 0.8031509518623352,
+      "learning_rate": 0.00014452434642531198,
+      "loss": 1.2794,
+      "step": 855
+    },
+    {
+      "epoch": 0.1312883435582822,
+      "grad_norm": 0.34702932834625244,
+      "learning_rate": 0.00014451064697285892,
+      "loss": 1.1968,
+      "step": 856
+    },
+    {
+      "epoch": 0.13144171779141103,
+      "grad_norm": 0.6688801050186157,
+      "learning_rate": 0.0001444969310555257,
+      "loss": 1.3055,
+      "step": 857
+    },
+    {
+      "epoch": 0.13159509202453987,
+      "grad_norm": 0.3785598874092102,
+      "learning_rate": 0.0001444831986765612,
+      "loss": 1.2446,
+      "step": 858
+    },
+    {
+      "epoch": 0.1317484662576687,
+      "grad_norm": 0.418057382106781,
+      "learning_rate": 0.00014446944983921814,
+      "loss": 1.1057,
+      "step": 859
+    },
+    {
+      "epoch": 0.13190184049079753,
+      "grad_norm": 0.4399944245815277,
+      "learning_rate": 0.0001444556845467532,
+      "loss": 1.1676,
+      "step": 860
+    },
+    {
+      "epoch": 0.13205521472392637,
+      "grad_norm": 0.44490736722946167,
+      "learning_rate": 0.00014444190280242695,
+      "loss": 1.319,
+      "step": 861
+    },
+    {
+      "epoch": 0.1322085889570552,
+      "grad_norm": 0.5250846743583679,
+      "learning_rate": 0.0001444281046095039,
+      "loss": 1.3803,
+      "step": 862
+    },
+    {
+      "epoch": 0.13236196319018406,
+      "grad_norm": 0.43551379442214966,
+      "learning_rate": 0.0001444142899712523,
+      "loss": 1.3715,
+      "step": 863
+    },
+    {
+      "epoch": 0.1325153374233129,
+      "grad_norm": 0.544020414352417,
+      "learning_rate": 0.00014440045889094452,
+      "loss": 1.2505,
+      "step": 864
+    },
+    {
+      "epoch": 0.13266871165644173,
+      "grad_norm": 0.3474222719669342,
+      "learning_rate": 0.00014438661137185663,
+      "loss": 1.1919,
+      "step": 865
+    },
+    {
+      "epoch": 0.13282208588957056,
+      "grad_norm": 0.4889809489250183,
+      "learning_rate": 0.00014437274741726865,
+      "loss": 1.1646,
+      "step": 866
+    },
+    {
+      "epoch": 0.1329754601226994,
+      "grad_norm": 0.3717634975910187,
+      "learning_rate": 0.0001443588670304646,
+      "loss": 1.2763,
+      "step": 867
+    },
+    {
+      "epoch": 0.13312883435582823,
+      "grad_norm": 0.38745248317718506,
+      "learning_rate": 0.00014434497021473226,
+      "loss": 1.3548,
+      "step": 868
+    },
+    {
+      "epoch": 0.13328220858895706,
+      "grad_norm": 0.38336706161499023,
+      "learning_rate": 0.00014433105697336332,
+      "loss": 1.0662,
+      "step": 869
+    },
+    {
+      "epoch": 0.1334355828220859,
+      "grad_norm": 0.525192141532898,
+      "learning_rate": 0.00014431712730965344,
+      "loss": 1.2185,
+      "step": 870
+    },
+    {
+      "epoch": 0.13358895705521473,
+      "grad_norm": 0.596077024936676,
+      "learning_rate": 0.0001443031812269021,
+      "loss": 1.1821,
+      "step": 871
+    },
+    {
+      "epoch": 0.13374233128834356,
+      "grad_norm": 0.344664603471756,
+      "learning_rate": 0.00014428921872841263,
+      "loss": 1.2208,
+      "step": 872
+    },
+    {
+      "epoch": 0.1338957055214724,
+      "grad_norm": 0.4343579113483429,
+      "learning_rate": 0.0001442752398174924,
+      "loss": 1.2767,
+      "step": 873
+    },
+    {
+      "epoch": 0.13404907975460123,
+      "grad_norm": 0.610854983329773,
+      "learning_rate": 0.00014426124449745252,
+      "loss": 1.2115,
+      "step": 874
+    },
+    {
+      "epoch": 0.13420245398773006,
+      "grad_norm": 0.3365573287010193,
+      "learning_rate": 0.00014424723277160805,
+      "loss": 1.3046,
+      "step": 875
+    },
+    {
+      "epoch": 0.1343558282208589,
+      "grad_norm": 0.6626908779144287,
+      "learning_rate": 0.00014423320464327792,
+      "loss": 1.1658,
+      "step": 876
+    },
+    {
+      "epoch": 0.13450920245398773,
+      "grad_norm": 0.6231021881103516,
+      "learning_rate": 0.00014421916011578498,
+      "loss": 1.0494,
+      "step": 877
+    },
+    {
+      "epoch": 0.13466257668711656,
+      "grad_norm": 0.4749593436717987,
+      "learning_rate": 0.00014420509919245588,
+      "loss": 1.3911,
+      "step": 878
+    },
+    {
+      "epoch": 0.1348159509202454,
+      "grad_norm": 0.41368749737739563,
+      "learning_rate": 0.00014419102187662126,
+      "loss": 1.0664,
+      "step": 879
+    },
+    {
+      "epoch": 0.13496932515337423,
+      "grad_norm": 0.4697296917438507,
+      "learning_rate": 0.00014417692817161556,
+      "loss": 1.2727,
+      "step": 880
+    },
+    {
+      "epoch": 0.13512269938650306,
+      "grad_norm": 0.45082950592041016,
+      "learning_rate": 0.00014416281808077717,
+      "loss": 1.1305,
+      "step": 881
+    },
+    {
+      "epoch": 0.1352760736196319,
+      "grad_norm": 0.620059609413147,
+      "learning_rate": 0.0001441486916074483,
+      "loss": 1.218,
+      "step": 882
+    },
+    {
+      "epoch": 0.13542944785276073,
+      "grad_norm": 0.6860255599021912,
+      "learning_rate": 0.00014413454875497505,
+      "loss": 1.2412,
+      "step": 883
+    },
+    {
+      "epoch": 0.13558282208588956,
+      "grad_norm": 0.49157240986824036,
+      "learning_rate": 0.00014412038952670743,
+      "loss": 1.434,
+      "step": 884
+    },
+    {
+      "epoch": 0.1357361963190184,
+      "grad_norm": 0.7819257378578186,
+      "learning_rate": 0.00014410621392599933,
+      "loss": 1.1724,
+      "step": 885
+    },
+    {
+      "epoch": 0.13588957055214723,
+      "grad_norm": 0.4717603027820587,
+      "learning_rate": 0.00014409202195620851,
+      "loss": 1.2844,
+      "step": 886
+    },
+    {
+      "epoch": 0.13604294478527607,
+      "grad_norm": 0.45863500237464905,
+      "learning_rate": 0.00014407781362069658,
+      "loss": 1.1241,
+      "step": 887
+    },
+    {
+      "epoch": 0.1361963190184049,
+      "grad_norm": 0.5072398781776428,
+      "learning_rate": 0.00014406358892282903,
+      "loss": 1.102,
+      "step": 888
+    },
+    {
+      "epoch": 0.13634969325153373,
+      "grad_norm": 0.3757067322731018,
+      "learning_rate": 0.00014404934786597527,
+      "loss": 1.2218,
+      "step": 889
+    },
+    {
+      "epoch": 0.13650306748466257,
+      "grad_norm": 0.3540547788143158,
+      "learning_rate": 0.00014403509045350857,
+      "loss": 1.3729,
+      "step": 890
+    },
+    {
+      "epoch": 0.1366564417177914,
+      "grad_norm": 0.42970430850982666,
+      "learning_rate": 0.00014402081668880599,
+      "loss": 1.158,
+      "step": 891
+    },
+    {
+      "epoch": 0.13680981595092023,
+      "grad_norm": 0.3858935236930847,
+      "learning_rate": 0.00014400652657524862,
+      "loss": 1.2434,
+      "step": 892
+    },
+    {
+      "epoch": 0.13696319018404907,
+      "grad_norm": 0.490856796503067,
+      "learning_rate": 0.00014399222011622132,
+      "loss": 1.2032,
+      "step": 893
+    },
+    {
+      "epoch": 0.13711656441717793,
+      "grad_norm": 0.5603492259979248,
+      "learning_rate": 0.00014397789731511278,
+      "loss": 1.2931,
+      "step": 894
+    },
+    {
+      "epoch": 0.13726993865030676,
+      "grad_norm": 0.5156564712524414,
+      "learning_rate": 0.0001439635581753157,
+      "loss": 1.1331,
+      "step": 895
+    },
+    {
+      "epoch": 0.1374233128834356,
+      "grad_norm": 0.4871954321861267,
+      "learning_rate": 0.0001439492027002265,
+      "loss": 1.2994,
+      "step": 896
+    },
+    {
+      "epoch": 0.13757668711656443,
+      "grad_norm": 0.6525068879127502,
+      "learning_rate": 0.00014393483089324559,
+      "loss": 1.273,
+      "step": 897
+    },
+    {
+      "epoch": 0.13773006134969326,
+      "grad_norm": 0.4161410331726074,
+      "learning_rate": 0.00014392044275777716,
+      "loss": 1.1801,
+      "step": 898
+    },
+    {
+      "epoch": 0.1378834355828221,
+      "grad_norm": 0.5837424397468567,
+      "learning_rate": 0.00014390603829722933,
+      "loss": 1.306,
+      "step": 899
+    },
+    {
+      "epoch": 0.13803680981595093,
+      "grad_norm": 0.3595598340034485,
+      "learning_rate": 0.00014389161751501405,
+      "loss": 1.3637,
+      "step": 900
+    },
+    {
+      "epoch": 0.13819018404907976,
+      "grad_norm": 0.3748437762260437,
+      "learning_rate": 0.00014387718041454717,
+      "loss": 1.321,
+      "step": 901
+    },
+    {
+      "epoch": 0.1383435582822086,
+      "grad_norm": 0.5539718866348267,
+      "learning_rate": 0.00014386272699924837,
+      "loss": 1.2127,
+      "step": 902
+    },
+    {
+      "epoch": 0.13849693251533743,
+      "grad_norm": 0.32123151421546936,
+      "learning_rate": 0.0001438482572725412,
+      "loss": 1.2725,
+      "step": 903
+    },
+    {
+      "epoch": 0.13865030674846626,
+      "grad_norm": 0.9548490643501282,
+      "learning_rate": 0.0001438337712378531,
+      "loss": 1.125,
+      "step": 904
+    },
+    {
+      "epoch": 0.1388036809815951,
+      "grad_norm": 0.6954799294471741,
+      "learning_rate": 0.00014381926889861534,
+      "loss": 1.2717,
+      "step": 905
+    },
+    {
+      "epoch": 0.13895705521472393,
+      "grad_norm": 0.3539749085903168,
+      "learning_rate": 0.00014380475025826308,
+      "loss": 1.313,
+      "step": 906
+    },
+    {
+      "epoch": 0.13911042944785276,
+      "grad_norm": 0.40778326988220215,
+      "learning_rate": 0.00014379021532023528,
+      "loss": 1.2701,
+      "step": 907
+    },
+    {
+      "epoch": 0.1392638036809816,
+      "grad_norm": 0.43370234966278076,
+      "learning_rate": 0.0001437756640879749,
+      "loss": 1.1356,
+      "step": 908
+    },
+    {
+      "epoch": 0.13941717791411043,
+      "grad_norm": 0.44014111161231995,
+      "learning_rate": 0.00014376109656492856,
+      "loss": 1.235,
+      "step": 909
+    },
+    {
+      "epoch": 0.13957055214723926,
+      "grad_norm": 0.46440231800079346,
+      "learning_rate": 0.00014374651275454693,
+      "loss": 1.2131,
+      "step": 910
+    },
+    {
+      "epoch": 0.1397239263803681,
+      "grad_norm": 0.42208975553512573,
+      "learning_rate": 0.00014373191266028442,
+      "loss": 1.1337,
+      "step": 911
+    },
+    {
+      "epoch": 0.13987730061349693,
+      "grad_norm": 0.5172114968299866,
+      "learning_rate": 0.00014371729628559935,
+      "loss": 1.3029,
+      "step": 912
+    },
+    {
+      "epoch": 0.14003067484662576,
+      "grad_norm": 0.39711248874664307,
+      "learning_rate": 0.00014370266363395386,
+      "loss": 1.2104,
+      "step": 913
+    },
+    {
+      "epoch": 0.1401840490797546,
+      "grad_norm": 0.36572620272636414,
+      "learning_rate": 0.00014368801470881397,
+      "loss": 1.2979,
+      "step": 914
+    },
+    {
+      "epoch": 0.14033742331288343,
+      "grad_norm": 0.39108961820602417,
+      "learning_rate": 0.00014367334951364953,
+      "loss": 1.2472,
+      "step": 915
+    },
+    {
+      "epoch": 0.14049079754601226,
+      "grad_norm": 0.36105772852897644,
+      "learning_rate": 0.00014365866805193431,
+      "loss": 1.129,
+      "step": 916
+    },
+    {
+      "epoch": 0.1406441717791411,
+      "grad_norm": 0.45772993564605713,
+      "learning_rate": 0.00014364397032714583,
+      "loss": 1.1116,
+      "step": 917
+    },
+    {
+      "epoch": 0.14079754601226993,
+      "grad_norm": 0.40730756521224976,
+      "learning_rate": 0.00014362925634276555,
+      "loss": 1.1973,
+      "step": 918
+    },
+    {
+      "epoch": 0.14095092024539876,
+      "grad_norm": 0.3718886375427246,
+      "learning_rate": 0.00014361452610227876,
+      "loss": 1.1804,
+      "step": 919
+    },
+    {
+      "epoch": 0.1411042944785276,
+      "grad_norm": 0.4633801579475403,
+      "learning_rate": 0.00014359977960917456,
+      "loss": 1.208,
+      "step": 920
+    },
+    {
+      "epoch": 0.14125766871165643,
+      "grad_norm": 0.7726576328277588,
+      "learning_rate": 0.00014358501686694597,
+      "loss": 1.2718,
+      "step": 921
+    },
+    {
+      "epoch": 0.14141104294478526,
+      "grad_norm": 0.36928802728652954,
+      "learning_rate": 0.00014357023787908975,
+      "loss": 1.3078,
+      "step": 922
+    },
+    {
+      "epoch": 0.1415644171779141,
+      "grad_norm": 0.3427782654762268,
+      "learning_rate": 0.00014355544264910662,
+      "loss": 1.2032,
+      "step": 923
+    },
+    {
+      "epoch": 0.14171779141104293,
+      "grad_norm": 0.4516969919204712,
+      "learning_rate": 0.00014354063118050112,
+      "loss": 1.2516,
+      "step": 924
+    },
+    {
+      "epoch": 0.1418711656441718,
+      "grad_norm": 0.3878684341907501,
+      "learning_rate": 0.0001435258034767816,
+      "loss": 1.2009,
+      "step": 925
+    },
+    {
+      "epoch": 0.14202453987730063,
+      "grad_norm": 0.30950307846069336,
+      "learning_rate": 0.00014351095954146024,
+      "loss": 1.1527,
+      "step": 926
+    },
+    {
+      "epoch": 0.14217791411042946,
+      "grad_norm": 0.3476774990558624,
+      "learning_rate": 0.00014349609937805316,
+      "loss": 1.2711,
+      "step": 927
+    },
+    {
+      "epoch": 0.1423312883435583,
+      "grad_norm": 0.3822958171367645,
+      "learning_rate": 0.00014348122299008025,
+      "loss": 1.3124,
+      "step": 928
+    },
+    {
+      "epoch": 0.14248466257668713,
+      "grad_norm": 0.4613816738128662,
+      "learning_rate": 0.00014346633038106522,
+      "loss": 1.2648,
+      "step": 929
+    },
+    {
+      "epoch": 0.14263803680981596,
+      "grad_norm": 0.3477330505847931,
+      "learning_rate": 0.00014345142155453572,
+      "loss": 1.2962,
+      "step": 930
+    },
+    {
+      "epoch": 0.1427914110429448,
+      "grad_norm": 0.5870327353477478,
+      "learning_rate": 0.0001434364965140231,
+      "loss": 1.1153,
+      "step": 931
+    },
+    {
+      "epoch": 0.14294478527607363,
+      "grad_norm": 0.34226319193840027,
+      "learning_rate": 0.0001434215552630627,
+      "loss": 1.188,
+      "step": 932
+    },
+    {
+      "epoch": 0.14309815950920246,
+      "grad_norm": 1.0119986534118652,
+      "learning_rate": 0.00014340659780519361,
+      "loss": 1.3391,
+      "step": 933
+    },
+    {
+      "epoch": 0.1432515337423313,
+      "grad_norm": 0.532867968082428,
+      "learning_rate": 0.00014339162414395878,
+      "loss": 1.2327,
+      "step": 934
+    },
+    {
+      "epoch": 0.14340490797546013,
+      "grad_norm": 0.40939241647720337,
+      "learning_rate": 0.000143376634282905,
+      "loss": 1.2547,
+      "step": 935
+    },
+    {
+      "epoch": 0.14355828220858896,
+      "grad_norm": 0.510563850402832,
+      "learning_rate": 0.00014336162822558288,
+      "loss": 1.1174,
+      "step": 936
+    },
+    {
+      "epoch": 0.1437116564417178,
+      "grad_norm": 0.5246591567993164,
+      "learning_rate": 0.0001433466059755469,
+      "loss": 1.1455,
+      "step": 937
+    },
+    {
+      "epoch": 0.14386503067484663,
+      "grad_norm": 0.9741495251655579,
+      "learning_rate": 0.00014333156753635533,
+      "loss": 1.2051,
+      "step": 938
+    },
+    {
+      "epoch": 0.14401840490797546,
+      "grad_norm": 0.7041084170341492,
+      "learning_rate": 0.00014331651291157036,
+      "loss": 1.2647,
+      "step": 939
+    },
+    {
+      "epoch": 0.1441717791411043,
+      "grad_norm": 0.6096442937850952,
+      "learning_rate": 0.0001433014421047579,
+      "loss": 1.277,
+      "step": 940
+    },
+    {
+      "epoch": 0.14432515337423313,
+      "grad_norm": 0.3860096037387848,
+      "learning_rate": 0.00014328635511948778,
+      "loss": 1.3253,
+      "step": 941
+    },
+    {
+      "epoch": 0.14447852760736196,
+      "grad_norm": 0.7076362371444702,
+      "learning_rate": 0.0001432712519593336,
+      "loss": 1.1189,
+      "step": 942
+    },
+    {
+      "epoch": 0.1446319018404908,
+      "grad_norm": 0.35366472601890564,
+      "learning_rate": 0.00014325613262787285,
+      "loss": 1.1259,
+      "step": 943
+    },
+    {
+      "epoch": 0.14478527607361963,
+      "grad_norm": 0.5243180990219116,
+      "learning_rate": 0.00014324099712868682,
+      "loss": 1.1909,
+      "step": 944
+    },
+    {
+      "epoch": 0.14493865030674846,
+      "grad_norm": 0.3988249897956848,
+      "learning_rate": 0.00014322584546536062,
+      "loss": 1.0622,
+      "step": 945
+    },
+    {
+      "epoch": 0.1450920245398773,
+      "grad_norm": 0.6021813750267029,
+      "learning_rate": 0.0001432106776414832,
+      "loss": 1.1815,
+      "step": 946
+    },
+    {
+      "epoch": 0.14524539877300613,
+      "grad_norm": 0.5586457848548889,
+      "learning_rate": 0.00014319549366064737,
+      "loss": 1.1022,
+      "step": 947
+    },
+    {
+      "epoch": 0.14539877300613496,
+      "grad_norm": 0.4719068109989166,
+      "learning_rate": 0.0001431802935264497,
+      "loss": 1.3703,
+      "step": 948
+    },
+    {
+      "epoch": 0.1455521472392638,
+      "grad_norm": 0.38756102323532104,
+      "learning_rate": 0.00014316507724249062,
+      "loss": 1.3275,
+      "step": 949
+    },
+    {
+      "epoch": 0.14570552147239263,
+      "grad_norm": 1.094046950340271,
+      "learning_rate": 0.0001431498448123744,
+      "loss": 1.1686,
+      "step": 950
+    },
+    {
+      "epoch": 0.14585889570552146,
+      "grad_norm": 0.7981559038162231,
+      "learning_rate": 0.00014313459623970914,
+      "loss": 1.1055,
+      "step": 951
+    },
+    {
+      "epoch": 0.1460122699386503,
+      "grad_norm": 0.5953325033187866,
+      "learning_rate": 0.00014311933152810674,
+      "loss": 1.183,
+      "step": 952
+    },
+    {
+      "epoch": 0.14616564417177913,
+      "grad_norm": 0.37586501240730286,
+      "learning_rate": 0.0001431040506811829,
+      "loss": 1.2268,
+      "step": 953
+    },
+    {
+      "epoch": 0.14631901840490796,
+      "grad_norm": 0.4386526644229889,
+      "learning_rate": 0.0001430887537025572,
+      "loss": 1.1997,
+      "step": 954
+    },
+    {
+      "epoch": 0.1464723926380368,
+      "grad_norm": 0.7090454697608948,
+      "learning_rate": 0.00014307344059585303,
+      "loss": 1.2371,
+      "step": 955
+    },
+    {
+      "epoch": 0.14662576687116563,
+      "grad_norm": 0.5226252675056458,
+      "learning_rate": 0.00014305811136469752,
+      "loss": 1.2074,
+      "step": 956
+    },
+    {
+      "epoch": 0.1467791411042945,
+      "grad_norm": 0.41718220710754395,
+      "learning_rate": 0.00014304276601272172,
+      "loss": 1.2457,
+      "step": 957
+    },
+    {
+      "epoch": 0.14693251533742333,
+      "grad_norm": 0.41600221395492554,
+      "learning_rate": 0.00014302740454356046,
+      "loss": 1.3714,
+      "step": 958
+    },
+    {
+      "epoch": 0.14708588957055216,
+      "grad_norm": 0.36179399490356445,
+      "learning_rate": 0.0001430120269608524,
+      "loss": 1.1942,
+      "step": 959
+    },
+    {
+      "epoch": 0.147239263803681,
+      "grad_norm": 0.3597133159637451,
+      "learning_rate": 0.00014299663326823998,
+      "loss": 1.2886,
+      "step": 960
+    },
+    {
+      "epoch": 0.14739263803680983,
+      "grad_norm": 0.4911383390426636,
+      "learning_rate": 0.00014298122346936948,
+      "loss": 1.2644,
+      "step": 961
+    },
+    {
+      "epoch": 0.14754601226993866,
+      "grad_norm": 0.5117217898368835,
+      "learning_rate": 0.00014296579756789101,
+      "loss": 1.1415,
+      "step": 962
+    },
+    {
+      "epoch": 0.1476993865030675,
+      "grad_norm": 0.6255547404289246,
+      "learning_rate": 0.00014295035556745848,
+      "loss": 1.1405,
+      "step": 963
+    },
+    {
+      "epoch": 0.14785276073619633,
+      "grad_norm": 0.4516870975494385,
+      "learning_rate": 0.00014293489747172958,
+      "loss": 1.2808,
+      "step": 964
+    },
+    {
+      "epoch": 0.14800613496932516,
+      "grad_norm": 0.46257010102272034,
+      "learning_rate": 0.0001429194232843659,
+      "loss": 1.1703,
+      "step": 965
+    },
+    {
+      "epoch": 0.148159509202454,
+      "grad_norm": 0.49126505851745605,
+      "learning_rate": 0.00014290393300903277,
+      "loss": 1.238,
+      "step": 966
+    },
+    {
+      "epoch": 0.14831288343558283,
+      "grad_norm": 0.4858570992946625,
+      "learning_rate": 0.0001428884266493993,
+      "loss": 1.2217,
+      "step": 967
+    },
+    {
+      "epoch": 0.14846625766871166,
+      "grad_norm": 0.3443072736263275,
+      "learning_rate": 0.00014287290420913852,
+      "loss": 1.1962,
+      "step": 968
+    },
+    {
+      "epoch": 0.1486196319018405,
+      "grad_norm": 0.4481574296951294,
+      "learning_rate": 0.00014285736569192716,
+      "loss": 1.1887,
+      "step": 969
+    },
+    {
+      "epoch": 0.14877300613496933,
+      "grad_norm": 0.5995777249336243,
+      "learning_rate": 0.00014284181110144585,
+      "loss": 1.1076,
+      "step": 970
+    },
+    {
+      "epoch": 0.14892638036809816,
+      "grad_norm": 0.41445115208625793,
+      "learning_rate": 0.00014282624044137892,
+      "loss": 1.2423,
+      "step": 971
+    },
+    {
+      "epoch": 0.149079754601227,
+      "grad_norm": 0.35934779047966003,
+      "learning_rate": 0.00014281065371541464,
+      "loss": 1.1612,
+      "step": 972
+    },
+    {
+      "epoch": 0.14923312883435583,
+      "grad_norm": 0.3953668177127838,
+      "learning_rate": 0.00014279505092724498,
+      "loss": 1.296,
+      "step": 973
+    },
+    {
+      "epoch": 0.14938650306748466,
+      "grad_norm": 0.6208137273788452,
+      "learning_rate": 0.00014277943208056574,
+      "loss": 1.2458,
+      "step": 974
+    },
+    {
+      "epoch": 0.1495398773006135,
+      "grad_norm": 0.9362353086471558,
+      "learning_rate": 0.0001427637971790765,
+      "loss": 1.1835,
+      "step": 975
+    },
+    {
+      "epoch": 0.14969325153374233,
+      "grad_norm": 0.5192467570304871,
+      "learning_rate": 0.0001427481462264808,
+      "loss": 1.1989,
+      "step": 976
+    },
+    {
+      "epoch": 0.14984662576687116,
+      "grad_norm": 0.3463788628578186,
+      "learning_rate": 0.00014273247922648569,
+      "loss": 1.4032,
+      "step": 977
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5507043600082397,
+      "learning_rate": 0.0001427167961828023,
+      "loss": 1.202,
+      "step": 978
+    },
+    {
+      "epoch": 0.15015337423312883,
+      "grad_norm": 0.42106783390045166,
+      "learning_rate": 0.00014270109709914544,
+      "loss": 1.2367,
+      "step": 979
+    },
+    {
+      "epoch": 0.15030674846625766,
+      "grad_norm": 0.43204399943351746,
+      "learning_rate": 0.0001426853819792337,
+      "loss": 1.072,
+      "step": 980
+    },
+    {
+      "epoch": 0.1504601226993865,
+      "grad_norm": 0.6941382884979248,
+      "learning_rate": 0.0001426696508267895,
+      "loss": 1.2252,
+      "step": 981
+    },
+    {
+      "epoch": 0.15061349693251533,
+      "grad_norm": 0.5600700378417969,
+      "learning_rate": 0.00014265390364553906,
+      "loss": 1.1897,
+      "step": 982
+    },
+    {
+      "epoch": 0.15076687116564416,
+      "grad_norm": 0.6339186429977417,
+      "learning_rate": 0.00014263814043921237,
+      "loss": 1.2274,
+      "step": 983
+    },
+    {
+      "epoch": 0.150920245398773,
+      "grad_norm": 0.48370078206062317,
+      "learning_rate": 0.00014262236121154325,
+      "loss": 1.2382,
+      "step": 984
+    },
+    {
+      "epoch": 0.15107361963190183,
+      "grad_norm": 0.46227484941482544,
+      "learning_rate": 0.00014260656596626933,
+      "loss": 1.1778,
+      "step": 985
+    },
+    {
+      "epoch": 0.15122699386503066,
+      "grad_norm": 0.3565542697906494,
+      "learning_rate": 0.00014259075470713197,
+      "loss": 1.0487,
+      "step": 986
+    },
+    {
+      "epoch": 0.1513803680981595,
+      "grad_norm": 0.4410135746002197,
+      "learning_rate": 0.00014257492743787638,
+      "loss": 1.2718,
+      "step": 987
+    },
+    {
+      "epoch": 0.15153374233128836,
+      "grad_norm": 0.44414249062538147,
+      "learning_rate": 0.00014255908416225152,
+      "loss": 1.0313,
+      "step": 988
+    },
+    {
+      "epoch": 0.1516871165644172,
+      "grad_norm": 0.7337961792945862,
+      "learning_rate": 0.00014254322488401017,
+      "loss": 1.2147,
+      "step": 989
+    },
+    {
+      "epoch": 0.15184049079754602,
+      "grad_norm": 0.3991456925868988,
+      "learning_rate": 0.00014252734960690892,
+      "loss": 1.1964,
+      "step": 990
+    },
+    {
+      "epoch": 0.15199386503067486,
+      "grad_norm": 0.39302685856819153,
+      "learning_rate": 0.00014251145833470808,
+      "loss": 1.2971,
+      "step": 991
+    },
+    {
+      "epoch": 0.1521472392638037,
+      "grad_norm": 0.34671080112457275,
+      "learning_rate": 0.00014249555107117182,
+      "loss": 1.1228,
+      "step": 992
+    },
+    {
+      "epoch": 0.15230061349693252,
+      "grad_norm": 0.3572523593902588,
+      "learning_rate": 0.00014247962782006803,
+      "loss": 1.2334,
+      "step": 993
+    },
+    {
+      "epoch": 0.15245398773006136,
+      "grad_norm": 0.6943639516830444,
+      "learning_rate": 0.0001424636885851685,
+      "loss": 1.1416,
+      "step": 994
+    },
+    {
+      "epoch": 0.1526073619631902,
+      "grad_norm": 0.3824521005153656,
+      "learning_rate": 0.00014244773337024864,
+      "loss": 1.2347,
+      "step": 995
+    },
+    {
+      "epoch": 0.15276073619631902,
+      "grad_norm": 0.5617176294326782,
+      "learning_rate": 0.0001424317621790878,
+      "loss": 1.2015,
+      "step": 996
+    },
+    {
+      "epoch": 0.15291411042944786,
+      "grad_norm": 0.5470498204231262,
+      "learning_rate": 0.00014241577501546905,
+      "loss": 1.211,
+      "step": 997
+    },
+    {
+      "epoch": 0.1530674846625767,
+      "grad_norm": 0.4709789752960205,
+      "learning_rate": 0.00014239977188317923,
+      "loss": 1.2266,
+      "step": 998
+    },
+    {
+      "epoch": 0.15322085889570553,
+      "grad_norm": 0.4241445064544678,
+      "learning_rate": 0.00014238375278600898,
+      "loss": 1.3449,
+      "step": 999
+    },
+    {
+      "epoch": 0.15337423312883436,
+      "grad_norm": 0.3999594748020172,
+      "learning_rate": 0.0001423677177277527,
+      "loss": 1.3412,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1535276073619632,
+      "grad_norm": 0.5106109976768494,
+      "learning_rate": 0.00014235166671220861,
+      "loss": 1.303,
+      "step": 1001
+    },
+    {
+      "epoch": 0.15368098159509203,
+      "grad_norm": 0.37052586674690247,
+      "learning_rate": 0.00014233559974317867,
+      "loss": 1.3139,
+      "step": 1002
+    },
+    {
+      "epoch": 0.15383435582822086,
+      "grad_norm": 0.37761712074279785,
+      "learning_rate": 0.0001423195168244687,
+      "loss": 1.1879,
+      "step": 1003
+    },
+    {
+      "epoch": 0.1539877300613497,
+      "grad_norm": 0.40408360958099365,
+      "learning_rate": 0.00014230341795988814,
+      "loss": 1.2416,
+      "step": 1004
+    },
+    {
+      "epoch": 0.15414110429447853,
+      "grad_norm": 0.6532825231552124,
+      "learning_rate": 0.0001422873031532504,
+      "loss": 1.2043,
+      "step": 1005
+    },
+    {
+      "epoch": 0.15429447852760736,
+      "grad_norm": 0.6255736351013184,
+      "learning_rate": 0.00014227117240837248,
+      "loss": 1.2098,
+      "step": 1006
+    },
+    {
+      "epoch": 0.1544478527607362,
+      "grad_norm": 0.35969769954681396,
+      "learning_rate": 0.00014225502572907535,
+      "loss": 1.1321,
+      "step": 1007
+    },
+    {
+      "epoch": 0.15460122699386503,
+      "grad_norm": 0.3736885190010071,
+      "learning_rate": 0.0001422388631191835,
+      "loss": 1.2306,
+      "step": 1008
+    },
+    {
+      "epoch": 0.15475460122699386,
+      "grad_norm": 0.6000912189483643,
+      "learning_rate": 0.0001422226845825255,
+      "loss": 1.265,
+      "step": 1009
+    },
+    {
+      "epoch": 0.1549079754601227,
+      "grad_norm": 0.3804703950881958,
+      "learning_rate": 0.00014220649012293345,
+      "loss": 1.1478,
+      "step": 1010
+    },
+    {
+      "epoch": 0.15506134969325153,
+      "grad_norm": 0.9573713541030884,
+      "learning_rate": 0.0001421902797442433,
+      "loss": 1.1614,
+      "step": 1011
+    },
+    {
+      "epoch": 0.15521472392638036,
+      "grad_norm": 0.6572113037109375,
+      "learning_rate": 0.00014217405345029485,
+      "loss": 1.3525,
+      "step": 1012
+    },
+    {
+      "epoch": 0.1553680981595092,
+      "grad_norm": 0.49434101581573486,
+      "learning_rate": 0.0001421578112449315,
+      "loss": 1.1055,
+      "step": 1013
+    },
+    {
+      "epoch": 0.15552147239263803,
+      "grad_norm": 0.9106312394142151,
+      "learning_rate": 0.00014214155313200055,
+      "loss": 1.2168,
+      "step": 1014
+    },
+    {
+      "epoch": 0.15567484662576686,
+      "grad_norm": 0.6043204069137573,
+      "learning_rate": 0.0001421252791153531,
+      "loss": 1.2059,
+      "step": 1015
+    },
+    {
+      "epoch": 0.1558282208588957,
+      "grad_norm": 0.4197806417942047,
+      "learning_rate": 0.00014210898919884385,
+      "loss": 1.2486,
+      "step": 1016
+    },
+    {
+      "epoch": 0.15598159509202453,
+      "grad_norm": 0.3580225706100464,
+      "learning_rate": 0.00014209268338633143,
+      "loss": 1.1212,
+      "step": 1017
+    },
+    {
+      "epoch": 0.15613496932515336,
+      "grad_norm": 0.5046915411949158,
+      "learning_rate": 0.00014207636168167816,
+      "loss": 1.0991,
+      "step": 1018
+    },
+    {
+      "epoch": 0.15628834355828222,
+      "grad_norm": 0.5020412802696228,
+      "learning_rate": 0.00014206002408875015,
+      "loss": 1.1254,
+      "step": 1019
+    },
+    {
+      "epoch": 0.15644171779141106,
+      "grad_norm": 0.35541245341300964,
+      "learning_rate": 0.0001420436706114172,
+      "loss": 1.0501,
+      "step": 1020
+    },
+    {
+      "epoch": 0.1565950920245399,
+      "grad_norm": 0.42204344272613525,
+      "learning_rate": 0.00014202730125355306,
+      "loss": 1.2032,
+      "step": 1021
+    },
+    {
+      "epoch": 0.15674846625766872,
+      "grad_norm": 0.3421989977359772,
+      "learning_rate": 0.00014201091601903496,
+      "loss": 1.1772,
+      "step": 1022
+    },
+    {
+      "epoch": 0.15690184049079756,
+      "grad_norm": 0.5651016235351562,
+      "learning_rate": 0.00014199451491174415,
+      "loss": 1.2838,
+      "step": 1023
+    },
+    {
+      "epoch": 0.1570552147239264,
+      "grad_norm": 0.4249191880226135,
+      "learning_rate": 0.00014197809793556552,
+      "loss": 1.2018,
+      "step": 1024
+    },
+    {
+      "epoch": 0.15720858895705522,
+      "grad_norm": 0.689967155456543,
+      "learning_rate": 0.00014196166509438768,
+      "loss": 1.2413,
+      "step": 1025
+    },
+    {
+      "epoch": 0.15736196319018406,
+      "grad_norm": 1.0080453157424927,
+      "learning_rate": 0.0001419452163921031,
+      "loss": 1.0306,
+      "step": 1026
+    },
+    {
+      "epoch": 0.1575153374233129,
+      "grad_norm": 0.5568225979804993,
+      "learning_rate": 0.00014192875183260796,
+      "loss": 1.2598,
+      "step": 1027
+    },
+    {
+      "epoch": 0.15766871165644172,
+      "grad_norm": 0.5192050933837891,
+      "learning_rate": 0.0001419122714198022,
+      "loss": 1.1952,
+      "step": 1028
+    },
+    {
+      "epoch": 0.15782208588957056,
+      "grad_norm": 0.359229177236557,
+      "learning_rate": 0.00014189577515758947,
+      "loss": 1.1636,
+      "step": 1029
+    },
+    {
+      "epoch": 0.1579754601226994,
+      "grad_norm": 0.3988804817199707,
+      "learning_rate": 0.00014187926304987726,
+      "loss": 1.2089,
+      "step": 1030
+    },
+    {
+      "epoch": 0.15812883435582822,
+      "grad_norm": 0.5517054200172424,
+      "learning_rate": 0.00014186273510057671,
+      "loss": 1.2341,
+      "step": 1031
+    },
+    {
+      "epoch": 0.15828220858895706,
+      "grad_norm": 0.33341965079307556,
+      "learning_rate": 0.00014184619131360284,
+      "loss": 1.222,
+      "step": 1032
+    },
+    {
+      "epoch": 0.1584355828220859,
+      "grad_norm": 0.3521129786968231,
+      "learning_rate": 0.00014182963169287428,
+      "loss": 1.1458,
+      "step": 1033
+    },
+    {
+      "epoch": 0.15858895705521472,
+      "grad_norm": 0.3736705183982849,
+      "learning_rate": 0.00014181305624231356,
+      "loss": 1.2368,
+      "step": 1034
+    },
+    {
+      "epoch": 0.15874233128834356,
+      "grad_norm": 0.47151219844818115,
+      "learning_rate": 0.00014179646496584683,
+      "loss": 1.0864,
+      "step": 1035
+    },
+    {
+      "epoch": 0.1588957055214724,
+      "grad_norm": 0.6190378665924072,
+      "learning_rate": 0.00014177985786740404,
+      "loss": 1.2726,
+      "step": 1036
+    },
+    {
+      "epoch": 0.15904907975460122,
+      "grad_norm": 0.6650306582450867,
+      "learning_rate": 0.0001417632349509189,
+      "loss": 1.163,
+      "step": 1037
+    },
+    {
+      "epoch": 0.15920245398773006,
+      "grad_norm": 0.5691866874694824,
+      "learning_rate": 0.00014174659622032885,
+      "loss": 1.2476,
+      "step": 1038
+    },
+    {
+      "epoch": 0.1593558282208589,
+      "grad_norm": 0.45317548513412476,
+      "learning_rate": 0.00014172994167957508,
+      "loss": 1.0492,
+      "step": 1039
+    },
+    {
+      "epoch": 0.15950920245398773,
+      "grad_norm": 0.38135525584220886,
+      "learning_rate": 0.00014171327133260256,
+      "loss": 1.1273,
+      "step": 1040
+    },
+    {
+      "epoch": 0.15966257668711656,
+      "grad_norm": 0.7529379725456238,
+      "learning_rate": 0.0001416965851833599,
+      "loss": 1.1118,
+      "step": 1041
+    },
+    {
+      "epoch": 0.1598159509202454,
+      "grad_norm": 0.3577733635902405,
+      "learning_rate": 0.0001416798832357996,
+      "loss": 1.0521,
+      "step": 1042
+    },
+    {
+      "epoch": 0.15996932515337423,
+      "grad_norm": 0.3320363461971283,
+      "learning_rate": 0.00014166316549387776,
+      "loss": 1.2731,
+      "step": 1043
+    },
+    {
+      "epoch": 0.16012269938650306,
+      "grad_norm": 0.628596305847168,
+      "learning_rate": 0.0001416464319615543,
+      "loss": 1.3022,
+      "step": 1044
+    },
+    {
+      "epoch": 0.1602760736196319,
+      "grad_norm": 0.45977655053138733,
+      "learning_rate": 0.0001416296826427929,
+      "loss": 1.253,
+      "step": 1045
+    },
+    {
+      "epoch": 0.16042944785276073,
+      "grad_norm": 1.080838918685913,
+      "learning_rate": 0.00014161291754156093,
+      "loss": 1.2492,
+      "step": 1046
+    },
+    {
+      "epoch": 0.16058282208588956,
+      "grad_norm": 0.33007872104644775,
+      "learning_rate": 0.0001415961366618295,
+      "loss": 1.1931,
+      "step": 1047
+    },
+    {
+      "epoch": 0.1607361963190184,
+      "grad_norm": 0.49386709928512573,
+      "learning_rate": 0.00014157934000757347,
+      "loss": 1.2044,
+      "step": 1048
+    },
+    {
+      "epoch": 0.16088957055214723,
+      "grad_norm": 0.39435461163520813,
+      "learning_rate": 0.00014156252758277144,
+      "loss": 1.0522,
+      "step": 1049
+    },
+    {
+      "epoch": 0.16104294478527606,
+      "grad_norm": 0.3926445245742798,
+      "learning_rate": 0.00014154569939140577,
+      "loss": 1.2831,
+      "step": 1050
+    },
+    {
+      "epoch": 0.16119631901840492,
+      "grad_norm": 0.4545574486255646,
+      "learning_rate": 0.00014152885543746252,
+      "loss": 1.3292,
+      "step": 1051
+    },
+    {
+      "epoch": 0.16134969325153375,
+      "grad_norm": 0.9473252892494202,
+      "learning_rate": 0.00014151199572493144,
+      "loss": 1.2123,
+      "step": 1052
+    },
+    {
+      "epoch": 0.1615030674846626,
+      "grad_norm": 0.942137598991394,
+      "learning_rate": 0.00014149512025780613,
+      "loss": 1.2737,
+      "step": 1053
+    },
+    {
+      "epoch": 0.16165644171779142,
+      "grad_norm": 0.8104510307312012,
+      "learning_rate": 0.0001414782290400838,
+      "loss": 1.2318,
+      "step": 1054
+    },
+    {
+      "epoch": 0.16180981595092025,
+      "grad_norm": 0.46195563673973083,
+      "learning_rate": 0.0001414613220757655,
+      "loss": 1.3264,
+      "step": 1055
+    },
+    {
+      "epoch": 0.1619631901840491,
+      "grad_norm": 0.8912544250488281,
+      "learning_rate": 0.00014144439936885592,
+      "loss": 1.21,
+      "step": 1056
+    },
+    {
+      "epoch": 0.16211656441717792,
+      "grad_norm": 0.41986313462257385,
+      "learning_rate": 0.00014142746092336356,
+      "loss": 1.2952,
+      "step": 1057
+    },
+    {
+      "epoch": 0.16226993865030676,
+      "grad_norm": 0.6209827661514282,
+      "learning_rate": 0.00014141050674330052,
+      "loss": 1.2212,
+      "step": 1058
+    },
+    {
+      "epoch": 0.1624233128834356,
+      "grad_norm": 0.4614139497280121,
+      "learning_rate": 0.0001413935368326828,
+      "loss": 1.2106,
+      "step": 1059
+    },
+    {
+      "epoch": 0.16257668711656442,
+      "grad_norm": 0.7930997610092163,
+      "learning_rate": 0.00014137655119553,
+      "loss": 1.2733,
+      "step": 1060
+    },
+    {
+      "epoch": 0.16273006134969326,
+      "grad_norm": 0.5114085674285889,
+      "learning_rate": 0.00014135954983586545,
+      "loss": 1.1603,
+      "step": 1061
+    },
+    {
+      "epoch": 0.1628834355828221,
+      "grad_norm": 0.7346270680427551,
+      "learning_rate": 0.0001413425327577163,
+      "loss": 1.3065,
+      "step": 1062
+    },
+    {
+      "epoch": 0.16303680981595092,
+      "grad_norm": 0.4948417544364929,
+      "learning_rate": 0.0001413254999651133,
+      "loss": 1.1991,
+      "step": 1063
+    },
+    {
+      "epoch": 0.16319018404907976,
+      "grad_norm": 0.6928455233573914,
+      "learning_rate": 0.000141308451462091,
+      "loss": 1.3423,
+      "step": 1064
+    },
+    {
+      "epoch": 0.1633435582822086,
+      "grad_norm": 0.6951407790184021,
+      "learning_rate": 0.0001412913872526877,
+      "loss": 1.142,
+      "step": 1065
+    },
+    {
+      "epoch": 0.16349693251533742,
+      "grad_norm": 0.35444721579551697,
+      "learning_rate": 0.0001412743073409453,
+      "loss": 1.1685,
+      "step": 1066
+    },
+    {
+      "epoch": 0.16365030674846626,
+      "grad_norm": 0.48605695366859436,
+      "learning_rate": 0.00014125721173090954,
+      "loss": 1.1573,
+      "step": 1067
+    },
+    {
+      "epoch": 0.1638036809815951,
+      "grad_norm": 0.5005863904953003,
+      "learning_rate": 0.0001412401004266298,
+      "loss": 1.3086,
+      "step": 1068
+    },
+    {
+      "epoch": 0.16395705521472392,
+      "grad_norm": 0.5854507684707642,
+      "learning_rate": 0.00014122297343215924,
+      "loss": 1.369,
+      "step": 1069
+    },
+    {
+      "epoch": 0.16411042944785276,
+      "grad_norm": 0.4311809539794922,
+      "learning_rate": 0.00014120583075155467,
+      "loss": 1.1221,
+      "step": 1070
+    },
+    {
+      "epoch": 0.1642638036809816,
+      "grad_norm": 0.43862247467041016,
+      "learning_rate": 0.0001411886723888767,
+      "loss": 1.3067,
+      "step": 1071
+    },
+    {
+      "epoch": 0.16441717791411042,
+      "grad_norm": 0.780173122882843,
+      "learning_rate": 0.00014117149834818956,
+      "loss": 1.3662,
+      "step": 1072
+    },
+    {
+      "epoch": 0.16457055214723926,
+      "grad_norm": 0.5280118584632874,
+      "learning_rate": 0.00014115430863356125,
+      "loss": 1.2046,
+      "step": 1073
+    },
+    {
+      "epoch": 0.1647239263803681,
+      "grad_norm": 0.5641762018203735,
+      "learning_rate": 0.0001411371032490635,
+      "loss": 1.0654,
+      "step": 1074
+    },
+    {
+      "epoch": 0.16487730061349692,
+      "grad_norm": 0.4292205274105072,
+      "learning_rate": 0.00014111988219877165,
+      "loss": 1.3925,
+      "step": 1075
+    },
+    {
+      "epoch": 0.16503067484662576,
+      "grad_norm": 0.8196471929550171,
+      "learning_rate": 0.00014110264548676491,
+      "loss": 1.2257,
+      "step": 1076
+    },
+    {
+      "epoch": 0.1651840490797546,
+      "grad_norm": 0.413025826215744,
+      "learning_rate": 0.00014108539311712606,
+      "loss": 1.0374,
+      "step": 1077
+    },
+    {
+      "epoch": 0.16533742331288342,
+      "grad_norm": 0.5046497583389282,
+      "learning_rate": 0.00014106812509394166,
+      "loss": 1.2055,
+      "step": 1078
+    },
+    {
+      "epoch": 0.16549079754601226,
+      "grad_norm": 0.5826169848442078,
+      "learning_rate": 0.00014105084142130195,
+      "loss": 1.2691,
+      "step": 1079
+    },
+    {
+      "epoch": 0.1656441717791411,
+      "grad_norm": 0.4098675847053528,
+      "learning_rate": 0.0001410335421033009,
+      "loss": 1.2874,
+      "step": 1080
+    },
+    {
+      "epoch": 0.16579754601226993,
+      "grad_norm": 0.4397736191749573,
+      "learning_rate": 0.00014101622714403616,
+      "loss": 1.1513,
+      "step": 1081
+    },
+    {
+      "epoch": 0.1659509202453988,
+      "grad_norm": 0.4411703944206238,
+      "learning_rate": 0.0001409988965476091,
+      "loss": 1.3157,
+      "step": 1082
+    },
+    {
+      "epoch": 0.16610429447852762,
+      "grad_norm": 0.4149448871612549,
+      "learning_rate": 0.00014098155031812482,
+      "loss": 1.151,
+      "step": 1083
+    },
+    {
+      "epoch": 0.16625766871165645,
+      "grad_norm": 0.46042555570602417,
+      "learning_rate": 0.00014096418845969207,
+      "loss": 1.0167,
+      "step": 1084
+    },
+    {
+      "epoch": 0.1664110429447853,
+      "grad_norm": 0.39350542426109314,
+      "learning_rate": 0.00014094681097642333,
+      "loss": 1.1416,
+      "step": 1085
+    },
+    {
+      "epoch": 0.16656441717791412,
+      "grad_norm": 0.49666649103164673,
+      "learning_rate": 0.00014092941787243477,
+      "loss": 1.3553,
+      "step": 1086
+    },
+    {
+      "epoch": 0.16671779141104295,
+      "grad_norm": 0.4671136140823364,
+      "learning_rate": 0.00014091200915184625,
+      "loss": 1.1979,
+      "step": 1087
+    },
+    {
+      "epoch": 0.1668711656441718,
+      "grad_norm": 0.7664749622344971,
+      "learning_rate": 0.00014089458481878142,
+      "loss": 1.2663,
+      "step": 1088
+    },
+    {
+      "epoch": 0.16702453987730062,
+      "grad_norm": 0.6404178142547607,
+      "learning_rate": 0.00014087714487736748,
+      "loss": 1.238,
+      "step": 1089
+    },
+    {
+      "epoch": 0.16717791411042945,
+      "grad_norm": 0.4325612187385559,
+      "learning_rate": 0.00014085968933173542,
+      "loss": 1.0772,
+      "step": 1090
+    },
+    {
+      "epoch": 0.1673312883435583,
+      "grad_norm": 0.40936538577079773,
+      "learning_rate": 0.00014084221818601991,
+      "loss": 1.2392,
+      "step": 1091
+    },
+    {
+      "epoch": 0.16748466257668712,
+      "grad_norm": 0.33305689692497253,
+      "learning_rate": 0.00014082473144435935,
+      "loss": 1.2234,
+      "step": 1092
+    },
+    {
+      "epoch": 0.16763803680981595,
+      "grad_norm": 0.4741230309009552,
+      "learning_rate": 0.00014080722911089574,
+      "loss": 1.2985,
+      "step": 1093
+    },
+    {
+      "epoch": 0.1677914110429448,
+      "grad_norm": 0.3069056570529938,
+      "learning_rate": 0.00014078971118977487,
+      "loss": 1.1945,
+      "step": 1094
+    },
+    {
+      "epoch": 0.16794478527607362,
+      "grad_norm": 0.6780207753181458,
+      "learning_rate": 0.00014077217768514614,
+      "loss": 1.1194,
+      "step": 1095
+    },
+    {
+      "epoch": 0.16809815950920245,
+      "grad_norm": 0.36013585329055786,
+      "learning_rate": 0.00014075462860116275,
+      "loss": 1.3243,
+      "step": 1096
+    },
+    {
+      "epoch": 0.1682515337423313,
+      "grad_norm": 0.6774555444717407,
+      "learning_rate": 0.0001407370639419815,
+      "loss": 1.131,
+      "step": 1097
+    },
+    {
+      "epoch": 0.16840490797546012,
+      "grad_norm": 0.4291854500770569,
+      "learning_rate": 0.00014071948371176287,
+      "loss": 1.2137,
+      "step": 1098
+    },
+    {
+      "epoch": 0.16855828220858896,
+      "grad_norm": 0.5698674321174622,
+      "learning_rate": 0.00014070188791467112,
+      "loss": 1.2161,
+      "step": 1099
+    },
+    {
+      "epoch": 0.1687116564417178,
+      "grad_norm": 0.8331619501113892,
+      "learning_rate": 0.00014068427655487405,
+      "loss": 1.2187,
+      "step": 1100
+    },
+    {
+      "epoch": 0.16886503067484662,
+      "grad_norm": 0.6767890453338623,
+      "learning_rate": 0.00014066664963654333,
+      "loss": 1.1341,
+      "step": 1101
+    },
+    {
+      "epoch": 0.16901840490797546,
+      "grad_norm": 0.5038207769393921,
+      "learning_rate": 0.00014064900716385416,
+      "loss": 1.3785,
+      "step": 1102
+    },
+    {
+      "epoch": 0.1691717791411043,
+      "grad_norm": 0.5180702805519104,
+      "learning_rate": 0.00014063134914098552,
+      "loss": 1.3337,
+      "step": 1103
+    },
+    {
+      "epoch": 0.16932515337423312,
+      "grad_norm": 0.4176473319530487,
+      "learning_rate": 0.00014061367557212005,
+      "loss": 1.2921,
+      "step": 1104
+    },
+    {
+      "epoch": 0.16947852760736196,
+      "grad_norm": 0.474020779132843,
+      "learning_rate": 0.000140595986461444,
+      "loss": 1.227,
+      "step": 1105
+    },
+    {
+      "epoch": 0.1696319018404908,
+      "grad_norm": 1.223692774772644,
+      "learning_rate": 0.00014057828181314742,
+      "loss": 1.0083,
+      "step": 1106
+    },
+    {
+      "epoch": 0.16978527607361962,
+      "grad_norm": 0.36724853515625,
+      "learning_rate": 0.00014056056163142395,
+      "loss": 1.1995,
+      "step": 1107
+    },
+    {
+      "epoch": 0.16993865030674846,
+      "grad_norm": 0.5425660014152527,
+      "learning_rate": 0.00014054282592047097,
+      "loss": 1.1451,
+      "step": 1108
+    },
+    {
+      "epoch": 0.1700920245398773,
+      "grad_norm": 0.46019020676612854,
+      "learning_rate": 0.00014052507468448948,
+      "loss": 1.2576,
+      "step": 1109
+    },
+    {
+      "epoch": 0.17024539877300612,
+      "grad_norm": 0.7236688137054443,
+      "learning_rate": 0.00014050730792768422,
+      "loss": 1.0581,
+      "step": 1110
+    },
+    {
+      "epoch": 0.17039877300613496,
+      "grad_norm": 0.3565762937068939,
+      "learning_rate": 0.00014048952565426354,
+      "loss": 1.3184,
+      "step": 1111
+    },
+    {
+      "epoch": 0.1705521472392638,
+      "grad_norm": 0.3493216037750244,
+      "learning_rate": 0.00014047172786843954,
+      "loss": 1.2908,
+      "step": 1112
+    },
+    {
+      "epoch": 0.17070552147239265,
+      "grad_norm": 0.43589818477630615,
+      "learning_rate": 0.00014045391457442793,
+      "loss": 1.2051,
+      "step": 1113
+    },
+    {
+      "epoch": 0.17085889570552149,
+      "grad_norm": 0.4687861502170563,
+      "learning_rate": 0.0001404360857764481,
+      "loss": 1.2212,
+      "step": 1114
+    },
+    {
+      "epoch": 0.17101226993865032,
+      "grad_norm": 0.3462468981742859,
+      "learning_rate": 0.00014041824147872318,
+      "loss": 1.2132,
+      "step": 1115
+    },
+    {
+      "epoch": 0.17116564417177915,
+      "grad_norm": 0.45358753204345703,
+      "learning_rate": 0.00014040038168547985,
+      "loss": 1.2403,
+      "step": 1116
+    },
+    {
+      "epoch": 0.17131901840490799,
+      "grad_norm": 0.40706667304039,
+      "learning_rate": 0.0001403825064009486,
+      "loss": 1.0256,
+      "step": 1117
+    },
+    {
+      "epoch": 0.17147239263803682,
+      "grad_norm": 0.3800080418586731,
+      "learning_rate": 0.0001403646156293635,
+      "loss": 1.1051,
+      "step": 1118
+    },
+    {
+      "epoch": 0.17162576687116565,
+      "grad_norm": 0.285297691822052,
+      "learning_rate": 0.00014034670937496227,
+      "loss": 1.1294,
+      "step": 1119
+    },
+    {
+      "epoch": 0.17177914110429449,
+      "grad_norm": 0.39089110493659973,
+      "learning_rate": 0.0001403287876419864,
+      "loss": 1.1408,
+      "step": 1120
+    },
+    {
+      "epoch": 0.17193251533742332,
+      "grad_norm": 0.3382580280303955,
+      "learning_rate": 0.00014031085043468095,
+      "loss": 1.2036,
+      "step": 1121
+    },
+    {
+      "epoch": 0.17208588957055215,
+      "grad_norm": 0.3455091118812561,
+      "learning_rate": 0.00014029289775729467,
+      "loss": 1.3114,
+      "step": 1122
+    },
+    {
+      "epoch": 0.172239263803681,
+      "grad_norm": 0.3015062212944031,
+      "learning_rate": 0.00014027492961408,
+      "loss": 1.1774,
+      "step": 1123
+    },
+    {
+      "epoch": 0.17239263803680982,
+      "grad_norm": 0.2915518283843994,
+      "learning_rate": 0.00014025694600929304,
+      "loss": 1.2392,
+      "step": 1124
+    },
+    {
+      "epoch": 0.17254601226993865,
+      "grad_norm": 1.1231883764266968,
+      "learning_rate": 0.0001402389469471935,
+      "loss": 1.2121,
+      "step": 1125
+    },
+    {
+      "epoch": 0.1726993865030675,
+      "grad_norm": 0.338023841381073,
+      "learning_rate": 0.00014022093243204483,
+      "loss": 1.1788,
+      "step": 1126
+    },
+    {
+      "epoch": 0.17285276073619632,
+      "grad_norm": 1.0729116201400757,
+      "learning_rate": 0.00014020290246811407,
+      "loss": 1.1659,
+      "step": 1127
+    },
+    {
+      "epoch": 0.17300613496932515,
+      "grad_norm": 0.42871764302253723,
+      "learning_rate": 0.00014018485705967197,
+      "loss": 1.1538,
+      "step": 1128
+    },
+    {
+      "epoch": 0.173159509202454,
+      "grad_norm": 1.10825514793396,
+      "learning_rate": 0.0001401667962109929,
+      "loss": 1.1146,
+      "step": 1129
+    },
+    {
+      "epoch": 0.17331288343558282,
+      "grad_norm": 0.3630538880825043,
+      "learning_rate": 0.00014014871992635493,
+      "loss": 1.1985,
+      "step": 1130
+    },
+    {
+      "epoch": 0.17346625766871165,
+      "grad_norm": 0.6109446287155151,
+      "learning_rate": 0.00014013062821003974,
+      "loss": 1.2265,
+      "step": 1131
+    },
+    {
+      "epoch": 0.1736196319018405,
+      "grad_norm": 0.46161046624183655,
+      "learning_rate": 0.00014011252106633268,
+      "loss": 1.315,
+      "step": 1132
+    },
+    {
+      "epoch": 0.17377300613496932,
+      "grad_norm": 0.6320381760597229,
+      "learning_rate": 0.00014009439849952277,
+      "loss": 1.1824,
+      "step": 1133
+    },
+    {
+      "epoch": 0.17392638036809815,
+      "grad_norm": 0.38861578702926636,
+      "learning_rate": 0.0001400762605139027,
+      "loss": 1.1749,
+      "step": 1134
+    },
+    {
+      "epoch": 0.174079754601227,
+      "grad_norm": 0.40446123480796814,
+      "learning_rate": 0.00014005810711376876,
+      "loss": 1.2824,
+      "step": 1135
+    },
+    {
+      "epoch": 0.17423312883435582,
+      "grad_norm": 0.45437321066856384,
+      "learning_rate": 0.00014003993830342094,
+      "loss": 1.1112,
+      "step": 1136
+    },
+    {
+      "epoch": 0.17438650306748466,
+      "grad_norm": 0.43737006187438965,
+      "learning_rate": 0.00014002175408716282,
+      "loss": 1.2579,
+      "step": 1137
+    },
+    {
+      "epoch": 0.1745398773006135,
+      "grad_norm": 0.38767343759536743,
+      "learning_rate": 0.0001400035544693017,
+      "loss": 1.1077,
+      "step": 1138
+    },
+    {
+      "epoch": 0.17469325153374232,
+      "grad_norm": 0.47691109776496887,
+      "learning_rate": 0.0001399853394541485,
+      "loss": 1.1527,
+      "step": 1139
+    },
+    {
+      "epoch": 0.17484662576687116,
+      "grad_norm": 0.5723860263824463,
+      "learning_rate": 0.00013996710904601775,
+      "loss": 1.1509,
+      "step": 1140
+    },
+    {
+      "epoch": 0.175,
+      "grad_norm": 0.4794568419456482,
+      "learning_rate": 0.00013994886324922767,
+      "loss": 1.2126,
+      "step": 1141
+    },
+    {
+      "epoch": 0.17515337423312882,
+      "grad_norm": 0.2785758078098297,
+      "learning_rate": 0.00013993060206810015,
+      "loss": 1.2065,
+      "step": 1142
+    },
+    {
+      "epoch": 0.17530674846625766,
+      "grad_norm": 0.3598911762237549,
+      "learning_rate": 0.00013991232550696064,
+      "loss": 1.0883,
+      "step": 1143
+    },
+    {
+      "epoch": 0.1754601226993865,
+      "grad_norm": 0.41302305459976196,
+      "learning_rate": 0.00013989403357013829,
+      "loss": 1.1004,
+      "step": 1144
+    },
+    {
+      "epoch": 0.17561349693251535,
+      "grad_norm": 0.36111176013946533,
+      "learning_rate": 0.00013987572626196593,
+      "loss": 1.3043,
+      "step": 1145
+    },
+    {
+      "epoch": 0.17576687116564418,
+      "grad_norm": 0.3471079170703888,
+      "learning_rate": 0.00013985740358677992,
+      "loss": 1.2128,
+      "step": 1146
+    },
+    {
+      "epoch": 0.17592024539877302,
+      "grad_norm": 0.40614938735961914,
+      "learning_rate": 0.00013983906554892035,
+      "loss": 1.2175,
+      "step": 1147
+    },
+    {
+      "epoch": 0.17607361963190185,
+      "grad_norm": 0.2792016863822937,
+      "learning_rate": 0.00013982071215273094,
+      "loss": 1.2038,
+      "step": 1148
+    },
+    {
+      "epoch": 0.17622699386503068,
+      "grad_norm": 0.3880993723869324,
+      "learning_rate": 0.00013980234340255898,
+      "loss": 1.1737,
+      "step": 1149
+    },
+    {
+      "epoch": 0.17638036809815952,
+      "grad_norm": 0.3723905384540558,
+      "learning_rate": 0.00013978395930275552,
+      "loss": 1.2944,
+      "step": 1150
+    },
+    {
+      "epoch": 0.17653374233128835,
+      "grad_norm": 0.348897248506546,
+      "learning_rate": 0.0001397655598576751,
+      "loss": 1.1985,
+      "step": 1151
+    },
+    {
+      "epoch": 0.17668711656441718,
+      "grad_norm": 0.3374224305152893,
+      "learning_rate": 0.00013974714507167604,
+      "loss": 1.1832,
+      "step": 1152
+    },
+    {
+      "epoch": 0.17684049079754602,
+      "grad_norm": 0.33870846033096313,
+      "learning_rate": 0.00013972871494912018,
+      "loss": 1.1839,
+      "step": 1153
+    },
+    {
+      "epoch": 0.17699386503067485,
+      "grad_norm": 0.46305862069129944,
+      "learning_rate": 0.00013971026949437303,
+      "loss": 1.1613,
+      "step": 1154
+    },
+    {
+      "epoch": 0.17714723926380369,
+      "grad_norm": 0.29042232036590576,
+      "learning_rate": 0.00013969180871180374,
+      "loss": 1.2321,
+      "step": 1155
+    },
+    {
+      "epoch": 0.17730061349693252,
+      "grad_norm": 0.4682149589061737,
+      "learning_rate": 0.00013967333260578513,
+      "loss": 1.2178,
+      "step": 1156
+    },
+    {
+      "epoch": 0.17745398773006135,
+      "grad_norm": 0.28799858689308167,
+      "learning_rate": 0.00013965484118069353,
+      "loss": 1.3305,
+      "step": 1157
+    },
+    {
+      "epoch": 0.17760736196319019,
+      "grad_norm": 0.35807713866233826,
+      "learning_rate": 0.00013963633444090902,
+      "loss": 1.2473,
+      "step": 1158
+    },
+    {
+      "epoch": 0.17776073619631902,
+      "grad_norm": 0.44646841287612915,
+      "learning_rate": 0.0001396178123908153,
+      "loss": 1.1894,
+      "step": 1159
+    },
+    {
+      "epoch": 0.17791411042944785,
+      "grad_norm": 0.822752058506012,
+      "learning_rate": 0.00013959927503479958,
+      "loss": 1.107,
+      "step": 1160
+    },
+    {
+      "epoch": 0.17806748466257669,
+      "grad_norm": 0.42735758423805237,
+      "learning_rate": 0.00013958072237725284,
+      "loss": 1.2264,
+      "step": 1161
+    },
+    {
+      "epoch": 0.17822085889570552,
+      "grad_norm": 0.4227905571460724,
+      "learning_rate": 0.00013956215442256957,
+      "loss": 1.1186,
+      "step": 1162
+    },
+    {
+      "epoch": 0.17837423312883435,
+      "grad_norm": 0.3355790078639984,
+      "learning_rate": 0.00013954357117514798,
+      "loss": 1.1262,
+      "step": 1163
+    },
+    {
+      "epoch": 0.1785276073619632,
+      "grad_norm": 0.3445092737674713,
+      "learning_rate": 0.00013952497263938982,
+      "loss": 1.2103,
+      "step": 1164
+    },
+    {
+      "epoch": 0.17868098159509202,
+      "grad_norm": 0.31808537244796753,
+      "learning_rate": 0.0001395063588197005,
+      "loss": 1.3181,
+      "step": 1165
+    },
+    {
+      "epoch": 0.17883435582822085,
+      "grad_norm": 0.43470337986946106,
+      "learning_rate": 0.00013948772972048907,
+      "loss": 1.1336,
+      "step": 1166
+    },
+    {
+      "epoch": 0.1789877300613497,
+      "grad_norm": 0.5171157717704773,
+      "learning_rate": 0.00013946908534616815,
+      "loss": 1.1797,
+      "step": 1167
+    },
+    {
+      "epoch": 0.17914110429447852,
+      "grad_norm": 0.3581295907497406,
+      "learning_rate": 0.00013945042570115404,
+      "loss": 1.2939,
+      "step": 1168
+    },
+    {
+      "epoch": 0.17929447852760735,
+      "grad_norm": 0.38410869240760803,
+      "learning_rate": 0.00013943175078986656,
+      "loss": 1.048,
+      "step": 1169
+    },
+    {
+      "epoch": 0.1794478527607362,
+      "grad_norm": 0.34006235003471375,
+      "learning_rate": 0.0001394130606167293,
+      "loss": 1.1793,
+      "step": 1170
+    },
+    {
+      "epoch": 0.17960122699386502,
+      "grad_norm": 0.9034362435340881,
+      "learning_rate": 0.00013939435518616925,
+      "loss": 1.2384,
+      "step": 1171
+    },
+    {
+      "epoch": 0.17975460122699385,
+      "grad_norm": 0.7272456884384155,
+      "learning_rate": 0.00013937563450261726,
+      "loss": 1.2275,
+      "step": 1172
+    },
+    {
+      "epoch": 0.1799079754601227,
+      "grad_norm": 0.4246615171432495,
+      "learning_rate": 0.0001393568985705076,
+      "loss": 1.0974,
+      "step": 1173
+    },
+    {
+      "epoch": 0.18006134969325152,
+      "grad_norm": 0.3137120008468628,
+      "learning_rate": 0.00013933814739427823,
+      "loss": 1.2109,
+      "step": 1174
+    },
+    {
+      "epoch": 0.18021472392638035,
+      "grad_norm": 0.3872048854827881,
+      "learning_rate": 0.0001393193809783707,
+      "loss": 1.1944,
+      "step": 1175
+    },
+    {
+      "epoch": 0.18036809815950922,
+      "grad_norm": 0.5548997521400452,
+      "learning_rate": 0.0001393005993272302,
+      "loss": 1.1502,
+      "step": 1176
+    },
+    {
+      "epoch": 0.18052147239263805,
+      "grad_norm": 0.4862046241760254,
+      "learning_rate": 0.00013928180244530554,
+      "loss": 1.2354,
+      "step": 1177
+    },
+    {
+      "epoch": 0.18067484662576688,
+      "grad_norm": 0.5642412304878235,
+      "learning_rate": 0.0001392629903370491,
+      "loss": 0.98,
+      "step": 1178
+    },
+    {
+      "epoch": 0.18082822085889572,
+      "grad_norm": 0.3447502851486206,
+      "learning_rate": 0.00013924416300691679,
+      "loss": 1.2749,
+      "step": 1179
+    },
+    {
+      "epoch": 0.18098159509202455,
+      "grad_norm": 0.5684177875518799,
+      "learning_rate": 0.0001392253204593683,
+      "loss": 1.1431,
+      "step": 1180
+    },
+    {
+      "epoch": 0.18113496932515338,
+      "grad_norm": 0.3061545491218567,
+      "learning_rate": 0.00013920646269886683,
+      "loss": 1.2517,
+      "step": 1181
+    },
+    {
+      "epoch": 0.18128834355828222,
+      "grad_norm": 0.3448631465435028,
+      "learning_rate": 0.00013918758972987917,
+      "loss": 1.2522,
+      "step": 1182
+    },
+    {
+      "epoch": 0.18144171779141105,
+      "grad_norm": 0.41935625672340393,
+      "learning_rate": 0.0001391687015568757,
+      "loss": 1.2797,
+      "step": 1183
+    },
+    {
+      "epoch": 0.18159509202453988,
+      "grad_norm": 0.48955971002578735,
+      "learning_rate": 0.00013914979818433049,
+      "loss": 1.2117,
+      "step": 1184
+    },
+    {
+      "epoch": 0.18174846625766872,
+      "grad_norm": 0.3896166682243347,
+      "learning_rate": 0.00013913087961672113,
+      "loss": 1.2183,
+      "step": 1185
+    },
+    {
+      "epoch": 0.18190184049079755,
+      "grad_norm": 0.3411356508731842,
+      "learning_rate": 0.0001391119458585288,
+      "loss": 1.0944,
+      "step": 1186
+    },
+    {
+      "epoch": 0.18205521472392638,
+      "grad_norm": 0.7475496530532837,
+      "learning_rate": 0.00013909299691423834,
+      "loss": 1.221,
+      "step": 1187
+    },
+    {
+      "epoch": 0.18220858895705522,
+      "grad_norm": 0.3311949372291565,
+      "learning_rate": 0.00013907403278833817,
+      "loss": 1.2711,
+      "step": 1188
+    },
+    {
+      "epoch": 0.18236196319018405,
+      "grad_norm": 0.7380959391593933,
+      "learning_rate": 0.00013905505348532027,
+      "loss": 1.2634,
+      "step": 1189
+    },
+    {
+      "epoch": 0.18251533742331288,
+      "grad_norm": 0.37988942861557007,
+      "learning_rate": 0.00013903605900968024,
+      "loss": 1.1057,
+      "step": 1190
+    },
+    {
+      "epoch": 0.18266871165644172,
+      "grad_norm": 0.6455687284469604,
+      "learning_rate": 0.0001390170493659173,
+      "loss": 1.3444,
+      "step": 1191
+    },
+    {
+      "epoch": 0.18282208588957055,
+      "grad_norm": 0.49459758400917053,
+      "learning_rate": 0.0001389980245585342,
+      "loss": 1.2441,
+      "step": 1192
+    },
+    {
+      "epoch": 0.18297546012269938,
+      "grad_norm": 0.3220534920692444,
+      "learning_rate": 0.00013897898459203735,
+      "loss": 1.1988,
+      "step": 1193
+    },
+    {
+      "epoch": 0.18312883435582822,
+      "grad_norm": 0.5718726515769958,
+      "learning_rate": 0.00013895992947093668,
+      "loss": 1.2769,
+      "step": 1194
+    },
+    {
+      "epoch": 0.18328220858895705,
+      "grad_norm": 0.3755846917629242,
+      "learning_rate": 0.00013894085919974577,
+      "loss": 1.1118,
+      "step": 1195
+    },
+    {
+      "epoch": 0.18343558282208589,
+      "grad_norm": 0.33112457394599915,
+      "learning_rate": 0.00013892177378298177,
+      "loss": 1.1665,
+      "step": 1196
+    },
+    {
+      "epoch": 0.18358895705521472,
+      "grad_norm": 0.9967913031578064,
+      "learning_rate": 0.00013890267322516539,
+      "loss": 1.2407,
+      "step": 1197
+    },
+    {
+      "epoch": 0.18374233128834355,
+      "grad_norm": 0.6321672797203064,
+      "learning_rate": 0.00013888355753082098,
+      "loss": 1.064,
+      "step": 1198
+    },
+    {
+      "epoch": 0.18389570552147239,
+      "grad_norm": 0.39529067277908325,
+      "learning_rate": 0.00013886442670447643,
+      "loss": 1.2367,
+      "step": 1199
+    },
+    {
+      "epoch": 0.18404907975460122,
+      "grad_norm": 0.42543596029281616,
+      "learning_rate": 0.0001388452807506632,
+      "loss": 1.1401,
+      "step": 1200
+    },
+    {
+      "epoch": 0.18420245398773005,
+      "grad_norm": 0.3527466654777527,
+      "learning_rate": 0.00013882611967391644,
+      "loss": 1.2991,
+      "step": 1201
+    },
+    {
+      "epoch": 0.18435582822085889,
+      "grad_norm": 0.6947577595710754,
+      "learning_rate": 0.00013880694347877473,
+      "loss": 1.2093,
+      "step": 1202
+    },
+    {
+      "epoch": 0.18450920245398772,
+      "grad_norm": 0.49983134865760803,
+      "learning_rate": 0.00013878775216978034,
+      "loss": 1.3414,
+      "step": 1203
+    },
+    {
+      "epoch": 0.18466257668711655,
+      "grad_norm": 0.3950105607509613,
+      "learning_rate": 0.00013876854575147908,
+      "loss": 1.1633,
+      "step": 1204
+    },
+    {
+      "epoch": 0.1848159509202454,
+      "grad_norm": 0.8230614066123962,
+      "learning_rate": 0.00013874932422842034,
+      "loss": 1.257,
+      "step": 1205
+    },
+    {
+      "epoch": 0.18496932515337422,
+      "grad_norm": 0.5518431663513184,
+      "learning_rate": 0.0001387300876051571,
+      "loss": 1.2219,
+      "step": 1206
+    },
+    {
+      "epoch": 0.18512269938650308,
+      "grad_norm": 0.6637043952941895,
+      "learning_rate": 0.00013871083588624589,
+      "loss": 1.2435,
+      "step": 1207
+    },
+    {
+      "epoch": 0.18527607361963191,
+      "grad_norm": 0.446323961019516,
+      "learning_rate": 0.00013869156907624688,
+      "loss": 1.1988,
+      "step": 1208
+    },
+    {
+      "epoch": 0.18542944785276075,
+      "grad_norm": 0.6545414924621582,
+      "learning_rate": 0.00013867228717972372,
+      "loss": 1.2749,
+      "step": 1209
+    },
+    {
+      "epoch": 0.18558282208588958,
+      "grad_norm": 0.3705739676952362,
+      "learning_rate": 0.0001386529902012437,
+      "loss": 1.2053,
+      "step": 1210
+    },
+    {
+      "epoch": 0.18573619631901842,
+      "grad_norm": 0.4742700755596161,
+      "learning_rate": 0.00013863367814537767,
+      "loss": 1.1561,
+      "step": 1211
+    },
+    {
+      "epoch": 0.18588957055214725,
+      "grad_norm": 0.5390597581863403,
+      "learning_rate": 0.00013861435101670007,
+      "loss": 1.1399,
+      "step": 1212
+    },
+    {
+      "epoch": 0.18604294478527608,
+      "grad_norm": 0.3651621639728546,
+      "learning_rate": 0.00013859500881978885,
+      "loss": 1.3425,
+      "step": 1213
+    },
+    {
+      "epoch": 0.18619631901840492,
+      "grad_norm": 0.8817558884620667,
+      "learning_rate": 0.0001385756515592256,
+      "loss": 1.1946,
+      "step": 1214
+    },
+    {
+      "epoch": 0.18634969325153375,
+      "grad_norm": 0.9217113852500916,
+      "learning_rate": 0.00013855627923959536,
+      "loss": 1.2152,
+      "step": 1215
+    },
+    {
+      "epoch": 0.18650306748466258,
+      "grad_norm": 1.4506409168243408,
+      "learning_rate": 0.00013853689186548696,
+      "loss": 1.1558,
+      "step": 1216
+    },
+    {
+      "epoch": 0.18665644171779142,
+      "grad_norm": 0.3862997591495514,
+      "learning_rate": 0.00013851748944149255,
+      "loss": 1.1886,
+      "step": 1217
+    },
+    {
+      "epoch": 0.18680981595092025,
+      "grad_norm": 0.8543208837509155,
+      "learning_rate": 0.00013849807197220798,
+      "loss": 1.213,
+      "step": 1218
+    },
+    {
+      "epoch": 0.18696319018404908,
+      "grad_norm": 0.4121759235858917,
+      "learning_rate": 0.00013847863946223263,
+      "loss": 1.2262,
+      "step": 1219
+    },
+    {
+      "epoch": 0.18711656441717792,
+      "grad_norm": 0.6287878751754761,
+      "learning_rate": 0.00013845919191616952,
+      "loss": 1.0132,
+      "step": 1220
+    },
+    {
+      "epoch": 0.18726993865030675,
+      "grad_norm": 0.4476358890533447,
+      "learning_rate": 0.00013843972933862505,
+      "loss": 1.1477,
+      "step": 1221
+    },
+    {
+      "epoch": 0.18742331288343558,
+      "grad_norm": 0.4244545102119446,
+      "learning_rate": 0.00013842025173420937,
+      "loss": 1.1685,
+      "step": 1222
+    },
+    {
+      "epoch": 0.18757668711656442,
+      "grad_norm": 1.0576450824737549,
+      "learning_rate": 0.00013840075910753606,
+      "loss": 1.1994,
+      "step": 1223
+    },
+    {
+      "epoch": 0.18773006134969325,
+      "grad_norm": 0.3933957517147064,
+      "learning_rate": 0.00013838125146322234,
+      "loss": 1.1066,
+      "step": 1224
+    },
+    {
+      "epoch": 0.18788343558282208,
+      "grad_norm": 0.4830858111381531,
+      "learning_rate": 0.00013836172880588894,
+      "loss": 1.2163,
+      "step": 1225
+    },
+    {
+      "epoch": 0.18803680981595092,
+      "grad_norm": 0.44613009691238403,
+      "learning_rate": 0.00013834219114016022,
+      "loss": 1.1515,
+      "step": 1226
+    },
+    {
+      "epoch": 0.18819018404907975,
+      "grad_norm": 0.37876129150390625,
+      "learning_rate": 0.00013832263847066392,
+      "loss": 1.2274,
+      "step": 1227
+    },
+    {
+      "epoch": 0.18834355828220858,
+      "grad_norm": 0.42368945479393005,
+      "learning_rate": 0.00013830307080203152,
+      "loss": 1.2584,
+      "step": 1228
+    },
+    {
+      "epoch": 0.18849693251533742,
+      "grad_norm": 0.513860821723938,
+      "learning_rate": 0.000138283488138898,
+      "loss": 1.1592,
+      "step": 1229
+    },
+    {
+      "epoch": 0.18865030674846625,
+      "grad_norm": 0.7934274077415466,
+      "learning_rate": 0.00013826389048590188,
+      "loss": 1.124,
+      "step": 1230
+    },
+    {
+      "epoch": 0.18880368098159508,
+      "grad_norm": 0.3297276496887207,
+      "learning_rate": 0.00013824427784768517,
+      "loss": 1.1501,
+      "step": 1231
+    },
+    {
+      "epoch": 0.18895705521472392,
+      "grad_norm": 0.5297401547431946,
+      "learning_rate": 0.00013822465022889354,
+      "loss": 1.1504,
+      "step": 1232
+    },
+    {
+      "epoch": 0.18911042944785275,
+      "grad_norm": 0.39665183424949646,
+      "learning_rate": 0.0001382050076341761,
+      "loss": 1.1019,
+      "step": 1233
+    },
+    {
+      "epoch": 0.18926380368098158,
+      "grad_norm": 1.1127557754516602,
+      "learning_rate": 0.0001381853500681856,
+      "loss": 1.3707,
+      "step": 1234
+    },
+    {
+      "epoch": 0.18941717791411042,
+      "grad_norm": 0.5515221357345581,
+      "learning_rate": 0.0001381656775355783,
+      "loss": 1.2661,
+      "step": 1235
+    },
+    {
+      "epoch": 0.18957055214723925,
+      "grad_norm": 0.6015550494194031,
+      "learning_rate": 0.000138145990041014,
+      "loss": 1.3419,
+      "step": 1236
+    },
+    {
+      "epoch": 0.18972392638036809,
+      "grad_norm": 0.5050998330116272,
+      "learning_rate": 0.00013812628758915602,
+      "loss": 1.1718,
+      "step": 1237
+    },
+    {
+      "epoch": 0.18987730061349692,
+      "grad_norm": 0.5648759603500366,
+      "learning_rate": 0.0001381065701846713,
+      "loss": 1.1269,
+      "step": 1238
+    },
+    {
+      "epoch": 0.19003067484662578,
+      "grad_norm": 0.6239713430404663,
+      "learning_rate": 0.0001380868378322302,
+      "loss": 1.1843,
+      "step": 1239
+    },
+    {
+      "epoch": 0.1901840490797546,
+      "grad_norm": 0.5946897864341736,
+      "learning_rate": 0.00013806709053650675,
+      "loss": 1.0943,
+      "step": 1240
+    },
+    {
+      "epoch": 0.19033742331288345,
+      "grad_norm": 0.31699755787849426,
+      "learning_rate": 0.00013804732830217844,
+      "loss": 1.1347,
+      "step": 1241
+    },
+    {
+      "epoch": 0.19049079754601228,
+      "grad_norm": 0.5166706442832947,
+      "learning_rate": 0.00013802755113392632,
+      "loss": 1.1328,
+      "step": 1242
+    },
+    {
+      "epoch": 0.19064417177914111,
+      "grad_norm": 0.4165552854537964,
+      "learning_rate": 0.000138007759036435,
+      "loss": 1.095,
+      "step": 1243
+    },
+    {
+      "epoch": 0.19079754601226995,
+      "grad_norm": 0.4528188109397888,
+      "learning_rate": 0.00013798795201439254,
+      "loss": 1.1563,
+      "step": 1244
+    },
+    {
+      "epoch": 0.19095092024539878,
+      "grad_norm": 0.6442364454269409,
+      "learning_rate": 0.00013796813007249068,
+      "loss": 1.1603,
+      "step": 1245
+    },
+    {
+      "epoch": 0.19110429447852761,
+      "grad_norm": 0.5871281623840332,
+      "learning_rate": 0.00013794829321542458,
+      "loss": 1.1422,
+      "step": 1246
+    },
+    {
+      "epoch": 0.19125766871165645,
+      "grad_norm": 0.8799555897712708,
+      "learning_rate": 0.00013792844144789294,
+      "loss": 1.128,
+      "step": 1247
+    },
+    {
+      "epoch": 0.19141104294478528,
+      "grad_norm": 0.33640292286872864,
+      "learning_rate": 0.00013790857477459807,
+      "loss": 1.1633,
+      "step": 1248
+    },
+    {
+      "epoch": 0.19156441717791411,
+      "grad_norm": 0.48547548055648804,
+      "learning_rate": 0.00013788869320024568,
+      "loss": 1.1563,
+      "step": 1249
+    },
+    {
+      "epoch": 0.19171779141104295,
+      "grad_norm": 0.3200492560863495,
+      "learning_rate": 0.0001378687967295452,
+      "loss": 1.2557,
+      "step": 1250
+    },
+    {
+      "epoch": 0.19187116564417178,
+      "grad_norm": 0.4115026891231537,
+      "learning_rate": 0.00013784888536720936,
+      "loss": 1.1856,
+      "step": 1251
+    },
+    {
+      "epoch": 0.19202453987730062,
+      "grad_norm": 0.5192675590515137,
+      "learning_rate": 0.0001378289591179546,
+      "loss": 1.2002,
+      "step": 1252
+    },
+    {
+      "epoch": 0.19217791411042945,
+      "grad_norm": 0.3569779694080353,
+      "learning_rate": 0.00013780901798650082,
+      "loss": 1.3022,
+      "step": 1253
+    },
+    {
+      "epoch": 0.19233128834355828,
+      "grad_norm": 0.48523423075675964,
+      "learning_rate": 0.0001377890619775714,
+      "loss": 1.2156,
+      "step": 1254
+    },
+    {
+      "epoch": 0.19248466257668712,
+      "grad_norm": 0.8600319623947144,
+      "learning_rate": 0.00013776909109589335,
+      "loss": 1.118,
+      "step": 1255
+    },
+    {
+      "epoch": 0.19263803680981595,
+      "grad_norm": 0.5117350220680237,
+      "learning_rate": 0.0001377491053461971,
+      "loss": 1.2446,
+      "step": 1256
+    },
+    {
+      "epoch": 0.19279141104294478,
+      "grad_norm": 0.47303178906440735,
+      "learning_rate": 0.00013772910473321666,
+      "loss": 1.1024,
+      "step": 1257
+    },
+    {
+      "epoch": 0.19294478527607362,
+      "grad_norm": 0.9025350213050842,
+      "learning_rate": 0.00013770908926168953,
+      "loss": 1.0749,
+      "step": 1258
+    },
+    {
+      "epoch": 0.19309815950920245,
+      "grad_norm": 0.3786260187625885,
+      "learning_rate": 0.00013768905893635677,
+      "loss": 1.2443,
+      "step": 1259
+    },
+    {
+      "epoch": 0.19325153374233128,
+      "grad_norm": 0.6722860932350159,
+      "learning_rate": 0.00013766901376196294,
+      "loss": 1.0533,
+      "step": 1260
+    },
+    {
+      "epoch": 0.19340490797546012,
+      "grad_norm": 0.3648119568824768,
+      "learning_rate": 0.00013764895374325604,
+      "loss": 1.1055,
+      "step": 1261
+    },
+    {
+      "epoch": 0.19355828220858895,
+      "grad_norm": 0.48103785514831543,
+      "learning_rate": 0.00013762887888498773,
+      "loss": 1.2403,
+      "step": 1262
+    },
+    {
+      "epoch": 0.19371165644171778,
+      "grad_norm": 0.7607157230377197,
+      "learning_rate": 0.0001376087891919131,
+      "loss": 1.2835,
+      "step": 1263
+    },
+    {
+      "epoch": 0.19386503067484662,
+      "grad_norm": 0.5584524273872375,
+      "learning_rate": 0.0001375886846687907,
+      "loss": 1.2877,
+      "step": 1264
+    },
+    {
+      "epoch": 0.19401840490797545,
+      "grad_norm": 0.4431840777397156,
+      "learning_rate": 0.00013756856532038276,
+      "loss": 1.2236,
+      "step": 1265
+    },
+    {
+      "epoch": 0.19417177914110428,
+      "grad_norm": 0.3900325298309326,
+      "learning_rate": 0.00013754843115145482,
+      "loss": 1.1681,
+      "step": 1266
+    },
+    {
+      "epoch": 0.19432515337423312,
+      "grad_norm": 0.645054280757904,
+      "learning_rate": 0.0001375282821667761,
+      "loss": 1.2195,
+      "step": 1267
+    },
+    {
+      "epoch": 0.19447852760736195,
+      "grad_norm": 0.6786963939666748,
+      "learning_rate": 0.00013750811837111922,
+      "loss": 1.3302,
+      "step": 1268
+    },
+    {
+      "epoch": 0.19463190184049078,
+      "grad_norm": 0.69007408618927,
+      "learning_rate": 0.0001374879397692604,
+      "loss": 1.2055,
+      "step": 1269
+    },
+    {
+      "epoch": 0.19478527607361965,
+      "grad_norm": 0.5566274523735046,
+      "learning_rate": 0.00013746774636597922,
+      "loss": 1.1858,
+      "step": 1270
+    },
+    {
+      "epoch": 0.19493865030674848,
+      "grad_norm": 0.5410163402557373,
+      "learning_rate": 0.00013744753816605896,
+      "loss": 1.2334,
+      "step": 1271
+    },
+    {
+      "epoch": 0.1950920245398773,
+      "grad_norm": 0.49454814195632935,
+      "learning_rate": 0.00013742731517428622,
+      "loss": 1.2265,
+      "step": 1272
+    },
+    {
+      "epoch": 0.19524539877300615,
+      "grad_norm": 0.5418010354042053,
+      "learning_rate": 0.00013740707739545128,
+      "loss": 1.1269,
+      "step": 1273
+    },
+    {
+      "epoch": 0.19539877300613498,
+      "grad_norm": 0.45860517024993896,
+      "learning_rate": 0.00013738682483434774,
+      "loss": 1.1674,
+      "step": 1274
+    },
+    {
+      "epoch": 0.1955521472392638,
+      "grad_norm": 0.3171859681606293,
+      "learning_rate": 0.00013736655749577284,
+      "loss": 1.2418,
+      "step": 1275
+    },
+    {
+      "epoch": 0.19570552147239265,
+      "grad_norm": 0.515257716178894,
+      "learning_rate": 0.0001373462753845273,
+      "loss": 1.2762,
+      "step": 1276
+    },
+    {
+      "epoch": 0.19585889570552148,
+      "grad_norm": 0.4884953498840332,
+      "learning_rate": 0.00013732597850541525,
+      "loss": 1.2762,
+      "step": 1277
+    },
+    {
+      "epoch": 0.1960122699386503,
+      "grad_norm": 0.6861414909362793,
+      "learning_rate": 0.00013730566686324442,
+      "loss": 1.2185,
+      "step": 1278
+    },
+    {
+      "epoch": 0.19616564417177915,
+      "grad_norm": 0.8693483471870422,
+      "learning_rate": 0.000137285340462826,
+      "loss": 1.2394,
+      "step": 1279
+    },
+    {
+      "epoch": 0.19631901840490798,
+      "grad_norm": 0.29843106865882874,
+      "learning_rate": 0.00013726499930897466,
+      "loss": 1.1557,
+      "step": 1280
+    },
+    {
+      "epoch": 0.1964723926380368,
+      "grad_norm": 0.42790350317955017,
+      "learning_rate": 0.00013724464340650856,
+      "loss": 1.2384,
+      "step": 1281
+    },
+    {
+      "epoch": 0.19662576687116565,
+      "grad_norm": 0.562095046043396,
+      "learning_rate": 0.00013722427276024942,
+      "loss": 1.1222,
+      "step": 1282
+    },
+    {
+      "epoch": 0.19677914110429448,
+      "grad_norm": 0.35019150376319885,
+      "learning_rate": 0.00013720388737502236,
+      "loss": 1.0555,
+      "step": 1283
+    },
+    {
+      "epoch": 0.19693251533742331,
+      "grad_norm": 0.42238637804985046,
+      "learning_rate": 0.00013718348725565602,
+      "loss": 1.106,
+      "step": 1284
+    },
+    {
+      "epoch": 0.19708588957055215,
+      "grad_norm": 1.0306576490402222,
+      "learning_rate": 0.0001371630724069826,
+      "loss": 1.253,
+      "step": 1285
+    },
+    {
+      "epoch": 0.19723926380368098,
+      "grad_norm": 0.6462970972061157,
+      "learning_rate": 0.0001371426428338377,
+      "loss": 1.3147,
+      "step": 1286
+    },
+    {
+      "epoch": 0.19739263803680981,
+      "grad_norm": 0.8770545721054077,
+      "learning_rate": 0.00013712219854106048,
+      "loss": 1.3998,
+      "step": 1287
+    },
+    {
+      "epoch": 0.19754601226993865,
+      "grad_norm": 0.5981436371803284,
+      "learning_rate": 0.0001371017395334935,
+      "loss": 1.3105,
+      "step": 1288
+    },
+    {
+      "epoch": 0.19769938650306748,
+      "grad_norm": 0.38686010241508484,
+      "learning_rate": 0.00013708126581598288,
+      "loss": 1.3159,
+      "step": 1289
+    },
+    {
+      "epoch": 0.19785276073619631,
+      "grad_norm": 0.5505843758583069,
+      "learning_rate": 0.00013706077739337818,
+      "loss": 1.1252,
+      "step": 1290
+    },
+    {
+      "epoch": 0.19800613496932515,
+      "grad_norm": 0.5607936382293701,
+      "learning_rate": 0.00013704027427053248,
+      "loss": 1.0365,
+      "step": 1291
+    },
+    {
+      "epoch": 0.19815950920245398,
+      "grad_norm": 0.42114922404289246,
+      "learning_rate": 0.0001370197564523023,
+      "loss": 1.1657,
+      "step": 1292
+    },
+    {
+      "epoch": 0.19831288343558282,
+      "grad_norm": 0.3713596761226654,
+      "learning_rate": 0.00013699922394354771,
+      "loss": 1.194,
+      "step": 1293
+    },
+    {
+      "epoch": 0.19846625766871165,
+      "grad_norm": 0.4431571066379547,
+      "learning_rate": 0.00013697867674913213,
+      "loss": 1.154,
+      "step": 1294
+    },
+    {
+      "epoch": 0.19861963190184048,
+      "grad_norm": 0.4971466064453125,
+      "learning_rate": 0.00013695811487392264,
+      "loss": 1.2412,
+      "step": 1295
+    },
+    {
+      "epoch": 0.19877300613496932,
+      "grad_norm": 0.5079706907272339,
+      "learning_rate": 0.00013693753832278965,
+      "loss": 1.3054,
+      "step": 1296
+    },
+    {
+      "epoch": 0.19892638036809815,
+      "grad_norm": 1.080243468284607,
+      "learning_rate": 0.00013691694710060712,
+      "loss": 1.1833,
+      "step": 1297
+    },
+    {
+      "epoch": 0.19907975460122698,
+      "grad_norm": 0.3502369225025177,
+      "learning_rate": 0.0001368963412122524,
+      "loss": 1.2352,
+      "step": 1298
+    },
+    {
+      "epoch": 0.19923312883435582,
+      "grad_norm": 0.40748098492622375,
+      "learning_rate": 0.00013687572066260642,
+      "loss": 1.2405,
+      "step": 1299
+    },
+    {
+      "epoch": 0.19938650306748465,
+      "grad_norm": 0.4777142107486725,
+      "learning_rate": 0.00013685508545655355,
+      "loss": 1.2633,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1995398773006135,
+      "grad_norm": 0.3972133696079254,
+      "learning_rate": 0.00013683443559898162,
+      "loss": 1.1439,
+      "step": 1301
+    },
+    {
+      "epoch": 0.19969325153374234,
+      "grad_norm": 0.5087442398071289,
+      "learning_rate": 0.0001368137710947819,
+      "loss": 1.212,
+      "step": 1302
+    },
+    {
+      "epoch": 0.19984662576687118,
+      "grad_norm": 0.5820403099060059,
+      "learning_rate": 0.00013679309194884916,
+      "loss": 1.2869,
+      "step": 1303
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.49128150939941406,
+      "learning_rate": 0.00013677239816608166,
+      "loss": 1.3123,
+      "step": 1304
+    },
+    {
+      "epoch": 0.20015337423312884,
+      "grad_norm": 0.34268200397491455,
+      "learning_rate": 0.0001367516897513811,
+      "loss": 1.2795,
+      "step": 1305
+    },
+    {
+      "epoch": 0.20030674846625768,
+      "grad_norm": 0.3634507358074188,
+      "learning_rate": 0.00013673096670965265,
+      "loss": 1.2271,
+      "step": 1306
+    },
+    {
+      "epoch": 0.2004601226993865,
+      "grad_norm": 1.0192762613296509,
+      "learning_rate": 0.00013671022904580493,
+      "loss": 1.1235,
+      "step": 1307
+    },
+    {
+      "epoch": 0.20061349693251534,
+      "grad_norm": 0.8439657688140869,
+      "learning_rate": 0.00013668947676475005,
+      "loss": 1.27,
+      "step": 1308
+    },
+    {
+      "epoch": 0.20076687116564418,
+      "grad_norm": 0.4166775047779083,
+      "learning_rate": 0.00013666870987140358,
+      "loss": 1.0864,
+      "step": 1309
+    },
+    {
+      "epoch": 0.200920245398773,
+      "grad_norm": 0.5215516686439514,
+      "learning_rate": 0.00013664792837068452,
+      "loss": 1.3433,
+      "step": 1310
+    },
+    {
+      "epoch": 0.20107361963190185,
+      "grad_norm": 0.3629275858402252,
+      "learning_rate": 0.00013662713226751539,
+      "loss": 1.0567,
+      "step": 1311
+    },
+    {
+      "epoch": 0.20122699386503068,
+      "grad_norm": 0.37714487314224243,
+      "learning_rate": 0.00013660632156682208,
+      "loss": 1.1539,
+      "step": 1312
+    },
+    {
+      "epoch": 0.2013803680981595,
+      "grad_norm": 0.3101533055305481,
+      "learning_rate": 0.00013658549627353403,
+      "loss": 1.2772,
+      "step": 1313
+    },
+    {
+      "epoch": 0.20153374233128835,
+      "grad_norm": 0.4658983647823334,
+      "learning_rate": 0.00013656465639258407,
+      "loss": 1.1911,
+      "step": 1314
+    },
+    {
+      "epoch": 0.20168711656441718,
+      "grad_norm": 0.7514652609825134,
+      "learning_rate": 0.00013654380192890854,
+      "loss": 1.2584,
+      "step": 1315
+    },
+    {
+      "epoch": 0.201840490797546,
+      "grad_norm": 1.0433018207550049,
+      "learning_rate": 0.00013652293288744717,
+      "loss": 1.0716,
+      "step": 1316
+    },
+    {
+      "epoch": 0.20199386503067485,
+      "grad_norm": 1.2453595399856567,
+      "learning_rate": 0.0001365020492731432,
+      "loss": 1.1912,
+      "step": 1317
+    },
+    {
+      "epoch": 0.20214723926380368,
+      "grad_norm": 0.32366400957107544,
+      "learning_rate": 0.0001364811510909433,
+      "loss": 1.1208,
+      "step": 1318
+    },
+    {
+      "epoch": 0.2023006134969325,
+      "grad_norm": 0.5037803053855896,
+      "learning_rate": 0.00013646023834579756,
+      "loss": 1.2274,
+      "step": 1319
+    },
+    {
+      "epoch": 0.20245398773006135,
+      "grad_norm": 0.4481677711009979,
+      "learning_rate": 0.0001364393110426596,
+      "loss": 1.3169,
+      "step": 1320
+    },
+    {
+      "epoch": 0.20260736196319018,
+      "grad_norm": 0.6590549349784851,
+      "learning_rate": 0.0001364183691864864,
+      "loss": 1.1005,
+      "step": 1321
+    },
+    {
+      "epoch": 0.202760736196319,
+      "grad_norm": 0.40658706426620483,
+      "learning_rate": 0.00013639741278223842,
+      "loss": 1.1529,
+      "step": 1322
+    },
+    {
+      "epoch": 0.20291411042944785,
+      "grad_norm": 0.6158697605133057,
+      "learning_rate": 0.00013637644183487962,
+      "loss": 1.2685,
+      "step": 1323
+    },
+    {
+      "epoch": 0.20306748466257668,
+      "grad_norm": 0.4483517110347748,
+      "learning_rate": 0.0001363554563493773,
+      "loss": 1.2524,
+      "step": 1324
+    },
+    {
+      "epoch": 0.20322085889570551,
+      "grad_norm": 0.4719666838645935,
+      "learning_rate": 0.00013633445633070228,
+      "loss": 1.318,
+      "step": 1325
+    },
+    {
+      "epoch": 0.20337423312883435,
+      "grad_norm": 0.5082921981811523,
+      "learning_rate": 0.0001363134417838288,
+      "loss": 1.2067,
+      "step": 1326
+    },
+    {
+      "epoch": 0.20352760736196318,
+      "grad_norm": 0.3916122615337372,
+      "learning_rate": 0.0001362924127137346,
+      "loss": 1.2325,
+      "step": 1327
+    },
+    {
+      "epoch": 0.20368098159509201,
+      "grad_norm": 0.4416019022464752,
+      "learning_rate": 0.00013627136912540072,
+      "loss": 1.2206,
+      "step": 1328
+    },
+    {
+      "epoch": 0.20383435582822085,
+      "grad_norm": 0.7656328678131104,
+      "learning_rate": 0.00013625031102381175,
+      "loss": 1.341,
+      "step": 1329
+    },
+    {
+      "epoch": 0.20398773006134968,
+      "grad_norm": 0.5802013278007507,
+      "learning_rate": 0.0001362292384139557,
+      "loss": 1.2024,
+      "step": 1330
+    },
+    {
+      "epoch": 0.20414110429447851,
+      "grad_norm": 0.5355190634727478,
+      "learning_rate": 0.000136208151300824,
+      "loss": 1.2854,
+      "step": 1331
+    },
+    {
+      "epoch": 0.20429447852760735,
+      "grad_norm": 0.3389717638492584,
+      "learning_rate": 0.00013618704968941154,
+      "loss": 1.1814,
+      "step": 1332
+    },
+    {
+      "epoch": 0.2044478527607362,
+      "grad_norm": 0.6282284259796143,
+      "learning_rate": 0.00013616593358471658,
+      "loss": 1.1303,
+      "step": 1333
+    },
+    {
+      "epoch": 0.20460122699386504,
+      "grad_norm": 0.34664472937583923,
+      "learning_rate": 0.0001361448029917409,
+      "loss": 1.2253,
+      "step": 1334
+    },
+    {
+      "epoch": 0.20475460122699388,
+      "grad_norm": 0.7212837934494019,
+      "learning_rate": 0.00013612365791548968,
+      "loss": 1.2718,
+      "step": 1335
+    },
+    {
+      "epoch": 0.2049079754601227,
+      "grad_norm": 0.4610905945301056,
+      "learning_rate": 0.0001361024983609715,
+      "loss": 1.2865,
+      "step": 1336
+    },
+    {
+      "epoch": 0.20506134969325154,
+      "grad_norm": 0.4226832985877991,
+      "learning_rate": 0.00013608132433319837,
+      "loss": 1.1657,
+      "step": 1337
+    },
+    {
+      "epoch": 0.20521472392638038,
+      "grad_norm": 0.48476704955101013,
+      "learning_rate": 0.0001360601358371858,
+      "loss": 1.2226,
+      "step": 1338
+    },
+    {
+      "epoch": 0.2053680981595092,
+      "grad_norm": 0.33844107389450073,
+      "learning_rate": 0.00013603893287795262,
+      "loss": 0.9664,
+      "step": 1339
+    },
+    {
+      "epoch": 0.20552147239263804,
+      "grad_norm": 0.404498815536499,
+      "learning_rate": 0.0001360177154605212,
+      "loss": 1.3033,
+      "step": 1340
+    },
+    {
+      "epoch": 0.20567484662576688,
+      "grad_norm": 0.3838912844657898,
+      "learning_rate": 0.00013599648358991723,
+      "loss": 1.1681,
+      "step": 1341
+    },
+    {
+      "epoch": 0.2058282208588957,
+      "grad_norm": 0.4045764207839966,
+      "learning_rate": 0.0001359752372711699,
+      "loss": 1.2145,
+      "step": 1342
+    },
+    {
+      "epoch": 0.20598159509202454,
+      "grad_norm": 0.6098929643630981,
+      "learning_rate": 0.0001359539765093118,
+      "loss": 1.1654,
+      "step": 1343
+    },
+    {
+      "epoch": 0.20613496932515338,
+      "grad_norm": 0.5765594244003296,
+      "learning_rate": 0.00013593270130937888,
+      "loss": 1.0986,
+      "step": 1344
+    },
+    {
+      "epoch": 0.2062883435582822,
+      "grad_norm": 0.3323592245578766,
+      "learning_rate": 0.00013591141167641063,
+      "loss": 1.2961,
+      "step": 1345
+    },
+    {
+      "epoch": 0.20644171779141104,
+      "grad_norm": 0.35089361667633057,
+      "learning_rate": 0.00013589010761544986,
+      "loss": 1.2042,
+      "step": 1346
+    },
+    {
+      "epoch": 0.20659509202453988,
+      "grad_norm": 0.38721007108688354,
+      "learning_rate": 0.0001358687891315428,
+      "loss": 1.0989,
+      "step": 1347
+    },
+    {
+      "epoch": 0.2067484662576687,
+      "grad_norm": 0.48348769545555115,
+      "learning_rate": 0.0001358474562297392,
+      "loss": 1.0817,
+      "step": 1348
+    },
+    {
+      "epoch": 0.20690184049079755,
+      "grad_norm": 0.5801207423210144,
+      "learning_rate": 0.0001358261089150921,
+      "loss": 1.0924,
+      "step": 1349
+    },
+    {
+      "epoch": 0.20705521472392638,
+      "grad_norm": 0.3601516783237457,
+      "learning_rate": 0.00013580474719265804,
+      "loss": 1.2056,
+      "step": 1350
+    },
+    {
+      "epoch": 0.2072085889570552,
+      "grad_norm": 0.3875848650932312,
+      "learning_rate": 0.00013578337106749688,
+      "loss": 1.2329,
+      "step": 1351
+    },
+    {
+      "epoch": 0.20736196319018405,
+      "grad_norm": 0.5556846857070923,
+      "learning_rate": 0.000135761980544672,
+      "loss": 1.1435,
+      "step": 1352
+    },
+    {
+      "epoch": 0.20751533742331288,
+      "grad_norm": 0.31709370017051697,
+      "learning_rate": 0.0001357405756292501,
+      "loss": 1.2578,
+      "step": 1353
+    },
+    {
+      "epoch": 0.2076687116564417,
+      "grad_norm": 0.4118722975254059,
+      "learning_rate": 0.0001357191563263014,
+      "loss": 1.1059,
+      "step": 1354
+    },
+    {
+      "epoch": 0.20782208588957055,
+      "grad_norm": 0.7188805341720581,
+      "learning_rate": 0.00013569772264089938,
+      "loss": 1.1804,
+      "step": 1355
+    },
+    {
+      "epoch": 0.20797546012269938,
+      "grad_norm": 0.32123687863349915,
+      "learning_rate": 0.00013567627457812106,
+      "loss": 1.3689,
+      "step": 1356
+    },
+    {
+      "epoch": 0.2081288343558282,
+      "grad_norm": 0.6456804871559143,
+      "learning_rate": 0.00013565481214304677,
+      "loss": 1.2107,
+      "step": 1357
+    },
+    {
+      "epoch": 0.20828220858895705,
+      "grad_norm": 0.46022534370422363,
+      "learning_rate": 0.00013563333534076028,
+      "loss": 1.2527,
+      "step": 1358
+    },
+    {
+      "epoch": 0.20843558282208588,
+      "grad_norm": 0.37033626437187195,
+      "learning_rate": 0.00013561184417634883,
+      "loss": 1.163,
+      "step": 1359
+    },
+    {
+      "epoch": 0.2085889570552147,
+      "grad_norm": 0.5817553400993347,
+      "learning_rate": 0.00013559033865490294,
+      "loss": 1.2054,
+      "step": 1360
+    },
+    {
+      "epoch": 0.20874233128834355,
+      "grad_norm": 0.4064948856830597,
+      "learning_rate": 0.0001355688187815166,
+      "loss": 1.1956,
+      "step": 1361
+    },
+    {
+      "epoch": 0.20889570552147238,
+      "grad_norm": 0.2974809408187866,
+      "learning_rate": 0.0001355472845612872,
+      "loss": 1.2305,
+      "step": 1362
+    },
+    {
+      "epoch": 0.2090490797546012,
+      "grad_norm": 0.368295818567276,
+      "learning_rate": 0.00013552573599931553,
+      "loss": 1.1908,
+      "step": 1363
+    },
+    {
+      "epoch": 0.20920245398773007,
+      "grad_norm": 0.4414481520652771,
+      "learning_rate": 0.00013550417310070572,
+      "loss": 1.108,
+      "step": 1364
+    },
+    {
+      "epoch": 0.2093558282208589,
+      "grad_norm": 0.38083013892173767,
+      "learning_rate": 0.0001354825958705654,
+      "loss": 1.1611,
+      "step": 1365
+    },
+    {
+      "epoch": 0.20950920245398774,
+      "grad_norm": 0.4108550548553467,
+      "learning_rate": 0.00013546100431400547,
+      "loss": 1.2776,
+      "step": 1366
+    },
+    {
+      "epoch": 0.20966257668711658,
+      "grad_norm": 0.8463228940963745,
+      "learning_rate": 0.00013543939843614037,
+      "loss": 1.1382,
+      "step": 1367
+    },
+    {
+      "epoch": 0.2098159509202454,
+      "grad_norm": 0.4300602376461029,
+      "learning_rate": 0.00013541777824208778,
+      "loss": 1.2387,
+      "step": 1368
+    },
+    {
+      "epoch": 0.20996932515337424,
+      "grad_norm": 1.0507773160934448,
+      "learning_rate": 0.00013539614373696884,
+      "loss": 1.2212,
+      "step": 1369
+    },
+    {
+      "epoch": 0.21012269938650308,
+      "grad_norm": 0.4803310036659241,
+      "learning_rate": 0.00013537449492590815,
+      "loss": 1.1533,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2102760736196319,
+      "grad_norm": 0.6644026637077332,
+      "learning_rate": 0.00013535283181403357,
+      "loss": 1.09,
+      "step": 1371
+    },
+    {
+      "epoch": 0.21042944785276074,
+      "grad_norm": 0.6922277808189392,
+      "learning_rate": 0.0001353311544064764,
+      "loss": 1.3702,
+      "step": 1372
+    },
+    {
+      "epoch": 0.21058282208588958,
+      "grad_norm": 0.3923341929912567,
+      "learning_rate": 0.00013530946270837142,
+      "loss": 1.1836,
+      "step": 1373
+    },
+    {
+      "epoch": 0.2107361963190184,
+      "grad_norm": 0.5082353353500366,
+      "learning_rate": 0.00013528775672485663,
+      "loss": 1.1769,
+      "step": 1374
+    },
+    {
+      "epoch": 0.21088957055214724,
+      "grad_norm": 0.5632988810539246,
+      "learning_rate": 0.0001352660364610735,
+      "loss": 1.182,
+      "step": 1375
+    },
+    {
+      "epoch": 0.21104294478527608,
+      "grad_norm": 0.7537752389907837,
+      "learning_rate": 0.00013524430192216687,
+      "loss": 1.3453,
+      "step": 1376
+    },
+    {
+      "epoch": 0.2111963190184049,
+      "grad_norm": 0.3823428153991699,
+      "learning_rate": 0.00013522255311328502,
+      "loss": 1.1569,
+      "step": 1377
+    },
+    {
+      "epoch": 0.21134969325153374,
+      "grad_norm": 0.4108635187149048,
+      "learning_rate": 0.0001352007900395795,
+      "loss": 1.1861,
+      "step": 1378
+    },
+    {
+      "epoch": 0.21150306748466258,
+      "grad_norm": 0.35405609011650085,
+      "learning_rate": 0.00013517901270620534,
+      "loss": 1.2978,
+      "step": 1379
+    },
+    {
+      "epoch": 0.2116564417177914,
+      "grad_norm": 0.392356812953949,
+      "learning_rate": 0.00013515722111832087,
+      "loss": 1.3144,
+      "step": 1380
+    },
+    {
+      "epoch": 0.21180981595092024,
+      "grad_norm": 0.359629362821579,
+      "learning_rate": 0.00013513541528108785,
+      "loss": 1.2117,
+      "step": 1381
+    },
+    {
+      "epoch": 0.21196319018404908,
+      "grad_norm": 0.4284852147102356,
+      "learning_rate": 0.0001351135951996714,
+      "loss": 1.1966,
+      "step": 1382
+    },
+    {
+      "epoch": 0.2121165644171779,
+      "grad_norm": 0.4751740097999573,
+      "learning_rate": 0.00013509176087923996,
+      "loss": 1.1421,
+      "step": 1383
+    },
+    {
+      "epoch": 0.21226993865030674,
+      "grad_norm": 0.3155258893966675,
+      "learning_rate": 0.00013506991232496547,
+      "loss": 1.0915,
+      "step": 1384
+    },
+    {
+      "epoch": 0.21242331288343558,
+      "grad_norm": 0.42437633872032166,
+      "learning_rate": 0.0001350480495420231,
+      "loss": 1.175,
+      "step": 1385
+    },
+    {
+      "epoch": 0.2125766871165644,
+      "grad_norm": 0.29920029640197754,
+      "learning_rate": 0.00013502617253559145,
+      "loss": 1.179,
+      "step": 1386
+    },
+    {
+      "epoch": 0.21273006134969324,
+      "grad_norm": 0.2559267580509186,
+      "learning_rate": 0.00013500428131085255,
+      "loss": 1.0262,
+      "step": 1387
+    },
+    {
+      "epoch": 0.21288343558282208,
+      "grad_norm": 0.4405561685562134,
+      "learning_rate": 0.0001349823758729917,
+      "loss": 1.1768,
+      "step": 1388
+    },
+    {
+      "epoch": 0.2130368098159509,
+      "grad_norm": 0.4656549096107483,
+      "learning_rate": 0.00013496045622719764,
+      "loss": 1.1524,
+      "step": 1389
+    },
+    {
+      "epoch": 0.21319018404907975,
+      "grad_norm": 0.2569158971309662,
+      "learning_rate": 0.00013493852237866238,
+      "loss": 1.289,
+      "step": 1390
+    },
+    {
+      "epoch": 0.21334355828220858,
+      "grad_norm": 0.3134523332118988,
+      "learning_rate": 0.00013491657433258145,
+      "loss": 1.2267,
+      "step": 1391
+    },
+    {
+      "epoch": 0.2134969325153374,
+      "grad_norm": 0.3231922686100006,
+      "learning_rate": 0.00013489461209415356,
+      "loss": 1.2234,
+      "step": 1392
+    },
+    {
+      "epoch": 0.21365030674846625,
+      "grad_norm": 0.299771249294281,
+      "learning_rate": 0.00013487263566858092,
+      "loss": 1.2479,
+      "step": 1393
+    },
+    {
+      "epoch": 0.21380368098159508,
+      "grad_norm": 0.36011672019958496,
+      "learning_rate": 0.000134850645061069,
+      "loss": 1.0598,
+      "step": 1394
+    },
+    {
+      "epoch": 0.21395705521472394,
+      "grad_norm": 0.3833593726158142,
+      "learning_rate": 0.0001348286402768268,
+      "loss": 1.1616,
+      "step": 1395
+    },
+    {
+      "epoch": 0.21411042944785277,
+      "grad_norm": 0.3183632493019104,
+      "learning_rate": 0.00013480662132106644,
+      "loss": 1.1091,
+      "step": 1396
+    },
+    {
+      "epoch": 0.2142638036809816,
+      "grad_norm": 0.3841853439807892,
+      "learning_rate": 0.00013478458819900354,
+      "loss": 1.2246,
+      "step": 1397
+    },
+    {
+      "epoch": 0.21441717791411044,
+      "grad_norm": 0.45916858315467834,
+      "learning_rate": 0.00013476254091585708,
+      "loss": 0.9817,
+      "step": 1398
+    },
+    {
+      "epoch": 0.21457055214723927,
+      "grad_norm": 0.2923356890678406,
+      "learning_rate": 0.00013474047947684937,
+      "loss": 1.2363,
+      "step": 1399
+    },
+    {
+      "epoch": 0.2147239263803681,
+      "grad_norm": 0.4143669307231903,
+      "learning_rate": 0.00013471840388720605,
+      "loss": 1.1851,
+      "step": 1400
+    },
+    {
+      "epoch": 0.21487730061349694,
+      "grad_norm": 1.1738274097442627,
+      "learning_rate": 0.00013469631415215612,
+      "loss": 1.1964,
+      "step": 1401
+    },
+    {
+      "epoch": 0.21503067484662577,
+      "grad_norm": 0.33784547448158264,
+      "learning_rate": 0.00013467421027693198,
+      "loss": 1.2173,
+      "step": 1402
+    },
+    {
+      "epoch": 0.2151840490797546,
+      "grad_norm": 0.804108738899231,
+      "learning_rate": 0.0001346520922667693,
+      "loss": 1.1994,
+      "step": 1403
+    },
+    {
+      "epoch": 0.21533742331288344,
+      "grad_norm": 0.2841300964355469,
+      "learning_rate": 0.00013462996012690717,
+      "loss": 1.1488,
+      "step": 1404
+    },
+    {
+      "epoch": 0.21549079754601227,
+      "grad_norm": 0.3315947651863098,
+      "learning_rate": 0.000134607813862588,
+      "loss": 1.1136,
+      "step": 1405
+    },
+    {
+      "epoch": 0.2156441717791411,
+      "grad_norm": 0.4034654200077057,
+      "learning_rate": 0.0001345856534790575,
+      "loss": 1.2185,
+      "step": 1406
+    },
+    {
+      "epoch": 0.21579754601226994,
+      "grad_norm": 0.4774540364742279,
+      "learning_rate": 0.00013456347898156478,
+      "loss": 1.2288,
+      "step": 1407
+    },
+    {
+      "epoch": 0.21595092024539878,
+      "grad_norm": 0.41431012749671936,
+      "learning_rate": 0.0001345412903753623,
+      "loss": 1.1504,
+      "step": 1408
+    },
+    {
+      "epoch": 0.2161042944785276,
+      "grad_norm": 0.5912563800811768,
+      "learning_rate": 0.00013451908766570582,
+      "loss": 1.3026,
+      "step": 1409
+    },
+    {
+      "epoch": 0.21625766871165644,
+      "grad_norm": 0.3930104374885559,
+      "learning_rate": 0.0001344968708578545,
+      "loss": 1.1719,
+      "step": 1410
+    },
+    {
+      "epoch": 0.21641104294478528,
+      "grad_norm": 0.30039188265800476,
+      "learning_rate": 0.00013447463995707078,
+      "loss": 1.2414,
+      "step": 1411
+    },
+    {
+      "epoch": 0.2165644171779141,
+      "grad_norm": 0.40156009793281555,
+      "learning_rate": 0.00013445239496862044,
+      "loss": 1.0976,
+      "step": 1412
+    },
+    {
+      "epoch": 0.21671779141104294,
+      "grad_norm": 0.31799373030662537,
+      "learning_rate": 0.0001344301358977726,
+      "loss": 1.1012,
+      "step": 1413
+    },
+    {
+      "epoch": 0.21687116564417178,
+      "grad_norm": 0.6262158751487732,
+      "learning_rate": 0.00013440786274979979,
+      "loss": 1.1293,
+      "step": 1414
+    },
+    {
+      "epoch": 0.2170245398773006,
+      "grad_norm": 0.28730446100234985,
+      "learning_rate": 0.00013438557552997778,
+      "loss": 1.1288,
+      "step": 1415
+    },
+    {
+      "epoch": 0.21717791411042944,
+      "grad_norm": 0.36861464381217957,
+      "learning_rate": 0.00013436327424358574,
+      "loss": 1.2756,
+      "step": 1416
+    },
+    {
+      "epoch": 0.21733128834355828,
+      "grad_norm": 0.8517671227455139,
+      "learning_rate": 0.00013434095889590608,
+      "loss": 1.2472,
+      "step": 1417
+    },
+    {
+      "epoch": 0.2174846625766871,
+      "grad_norm": 0.3622899353504181,
+      "learning_rate": 0.00013431862949222467,
+      "loss": 1.1706,
+      "step": 1418
+    },
+    {
+      "epoch": 0.21763803680981594,
+      "grad_norm": 1.1624623537063599,
+      "learning_rate": 0.0001342962860378306,
+      "loss": 1.3068,
+      "step": 1419
+    },
+    {
+      "epoch": 0.21779141104294478,
+      "grad_norm": 0.35962969064712524,
+      "learning_rate": 0.00013427392853801638,
+      "loss": 1.2028,
+      "step": 1420
+    },
+    {
+      "epoch": 0.2179447852760736,
+      "grad_norm": 0.3189540505409241,
+      "learning_rate": 0.0001342515569980777,
+      "loss": 1.1895,
+      "step": 1421
+    },
+    {
+      "epoch": 0.21809815950920244,
+      "grad_norm": 0.9896800518035889,
+      "learning_rate": 0.00013422917142331375,
+      "loss": 1.162,
+      "step": 1422
+    },
+    {
+      "epoch": 0.21825153374233128,
+      "grad_norm": 0.4820398688316345,
+      "learning_rate": 0.00013420677181902698,
+      "loss": 1.4153,
+      "step": 1423
+    },
+    {
+      "epoch": 0.2184049079754601,
+      "grad_norm": 1.0959323644638062,
+      "learning_rate": 0.0001341843581905231,
+      "loss": 1.1714,
+      "step": 1424
+    },
+    {
+      "epoch": 0.21855828220858894,
+      "grad_norm": 0.6305782794952393,
+      "learning_rate": 0.0001341619305431112,
+      "loss": 1.3218,
+      "step": 1425
+    },
+    {
+      "epoch": 0.21871165644171778,
+      "grad_norm": 0.4958251714706421,
+      "learning_rate": 0.0001341394888821037,
+      "loss": 1.2132,
+      "step": 1426
+    },
+    {
+      "epoch": 0.21886503067484664,
+      "grad_norm": 0.5175288319587708,
+      "learning_rate": 0.00013411703321281632,
+      "loss": 1.2698,
+      "step": 1427
+    },
+    {
+      "epoch": 0.21901840490797547,
+      "grad_norm": 1.0391154289245605,
+      "learning_rate": 0.00013409456354056807,
+      "loss": 1.1706,
+      "step": 1428
+    },
+    {
+      "epoch": 0.2191717791411043,
+      "grad_norm": 0.8039506673812866,
+      "learning_rate": 0.00013407207987068133,
+      "loss": 1.2311,
+      "step": 1429
+    },
+    {
+      "epoch": 0.21932515337423314,
+      "grad_norm": 0.47560667991638184,
+      "learning_rate": 0.0001340495822084818,
+      "loss": 1.2608,
+      "step": 1430
+    },
+    {
+      "epoch": 0.21947852760736197,
+      "grad_norm": 0.48472705483436584,
+      "learning_rate": 0.00013402707055929842,
+      "loss": 1.0575,
+      "step": 1431
+    },
+    {
+      "epoch": 0.2196319018404908,
+      "grad_norm": 0.3514997363090515,
+      "learning_rate": 0.0001340045449284635,
+      "loss": 1.0869,
+      "step": 1432
+    },
+    {
+      "epoch": 0.21978527607361964,
+      "grad_norm": 0.37568989396095276,
+      "learning_rate": 0.0001339820053213127,
+      "loss": 1.1604,
+      "step": 1433
+    },
+    {
+      "epoch": 0.21993865030674847,
+      "grad_norm": 0.4217923879623413,
+      "learning_rate": 0.0001339594517431849,
+      "loss": 1.2172,
+      "step": 1434
+    },
+    {
+      "epoch": 0.2200920245398773,
+      "grad_norm": 0.34138113260269165,
+      "learning_rate": 0.00013393688419942228,
+      "loss": 1.1591,
+      "step": 1435
+    },
+    {
+      "epoch": 0.22024539877300614,
+      "grad_norm": 0.4642728567123413,
+      "learning_rate": 0.00013391430269537046,
+      "loss": 1.2713,
+      "step": 1436
+    },
+    {
+      "epoch": 0.22039877300613497,
+      "grad_norm": 0.2767432928085327,
+      "learning_rate": 0.00013389170723637826,
+      "loss": 1.299,
+      "step": 1437
+    },
+    {
+      "epoch": 0.2205521472392638,
+      "grad_norm": 0.4519413709640503,
+      "learning_rate": 0.00013386909782779787,
+      "loss": 1.2298,
+      "step": 1438
+    },
+    {
+      "epoch": 0.22070552147239264,
+      "grad_norm": 0.5791484713554382,
+      "learning_rate": 0.00013384647447498466,
+      "loss": 1.3106,
+      "step": 1439
+    },
+    {
+      "epoch": 0.22085889570552147,
+      "grad_norm": 0.5004804730415344,
+      "learning_rate": 0.00013382383718329744,
+      "loss": 1.3025,
+      "step": 1440
+    },
+    {
+      "epoch": 0.2210122699386503,
+      "grad_norm": 0.7393519282341003,
+      "learning_rate": 0.0001338011859580983,
+      "loss": 1.1441,
+      "step": 1441
+    },
+    {
+      "epoch": 0.22116564417177914,
+      "grad_norm": 0.3756941258907318,
+      "learning_rate": 0.00013377852080475255,
+      "loss": 1.214,
+      "step": 1442
+    },
+    {
+      "epoch": 0.22131901840490797,
+      "grad_norm": 0.4095316231250763,
+      "learning_rate": 0.00013375584172862885,
+      "loss": 1.0953,
+      "step": 1443
+    },
+    {
+      "epoch": 0.2214723926380368,
+      "grad_norm": 0.6681280732154846,
+      "learning_rate": 0.0001337331487350992,
+      "loss": 1.0174,
+      "step": 1444
+    },
+    {
+      "epoch": 0.22162576687116564,
+      "grad_norm": 0.6622613072395325,
+      "learning_rate": 0.0001337104418295388,
+      "loss": 1.1628,
+      "step": 1445
+    },
+    {
+      "epoch": 0.22177914110429447,
+      "grad_norm": 0.5640711188316345,
+      "learning_rate": 0.00013368772101732626,
+      "loss": 1.2165,
+      "step": 1446
+    },
+    {
+      "epoch": 0.2219325153374233,
+      "grad_norm": 0.4797951579093933,
+      "learning_rate": 0.0001336649863038434,
+      "loss": 1.0518,
+      "step": 1447
+    },
+    {
+      "epoch": 0.22208588957055214,
+      "grad_norm": 0.2803958058357239,
+      "learning_rate": 0.0001336422376944753,
+      "loss": 1.073,
+      "step": 1448
+    },
+    {
+      "epoch": 0.22223926380368098,
+      "grad_norm": 0.5000907182693481,
+      "learning_rate": 0.0001336194751946105,
+      "loss": 1.1679,
+      "step": 1449
+    },
+    {
+      "epoch": 0.2223926380368098,
+      "grad_norm": 0.2899826169013977,
+      "learning_rate": 0.00013359669880964062,
+      "loss": 1.2139,
+      "step": 1450
+    },
+    {
+      "epoch": 0.22254601226993864,
+      "grad_norm": 0.30941277742385864,
+      "learning_rate": 0.0001335739085449607,
+      "loss": 1.2292,
+      "step": 1451
+    },
+    {
+      "epoch": 0.22269938650306748,
+      "grad_norm": 0.8701250553131104,
+      "learning_rate": 0.00013355110440596907,
+      "loss": 1.1538,
+      "step": 1452
+    },
+    {
+      "epoch": 0.2228527607361963,
+      "grad_norm": 0.387128621339798,
+      "learning_rate": 0.00013352828639806726,
+      "loss": 1.082,
+      "step": 1453
+    },
+    {
+      "epoch": 0.22300613496932514,
+      "grad_norm": 0.8385788202285767,
+      "learning_rate": 0.00013350545452666014,
+      "loss": 1.1946,
+      "step": 1454
+    },
+    {
+      "epoch": 0.22315950920245398,
+      "grad_norm": 0.43165335059165955,
+      "learning_rate": 0.0001334826087971559,
+      "loss": 1.1783,
+      "step": 1455
+    },
+    {
+      "epoch": 0.2233128834355828,
+      "grad_norm": 0.3087812066078186,
+      "learning_rate": 0.00013345974921496597,
+      "loss": 1.1694,
+      "step": 1456
+    },
+    {
+      "epoch": 0.22346625766871164,
+      "grad_norm": 0.4324992001056671,
+      "learning_rate": 0.00013343687578550503,
+      "loss": 1.1421,
+      "step": 1457
+    },
+    {
+      "epoch": 0.2236196319018405,
+      "grad_norm": 0.3609766960144043,
+      "learning_rate": 0.00013341398851419108,
+      "loss": 1.3317,
+      "step": 1458
+    },
+    {
+      "epoch": 0.22377300613496934,
+      "grad_norm": 0.41812819242477417,
+      "learning_rate": 0.0001333910874064454,
+      "loss": 1.1672,
+      "step": 1459
+    },
+    {
+      "epoch": 0.22392638036809817,
+      "grad_norm": 0.2980424761772156,
+      "learning_rate": 0.00013336817246769255,
+      "loss": 1.2138,
+      "step": 1460
+    },
+    {
+      "epoch": 0.224079754601227,
+      "grad_norm": 0.354069322347641,
+      "learning_rate": 0.00013334524370336035,
+      "loss": 1.1567,
+      "step": 1461
+    },
+    {
+      "epoch": 0.22423312883435584,
+      "grad_norm": 0.2948873043060303,
+      "learning_rate": 0.0001333223011188799,
+      "loss": 1.2039,
+      "step": 1462
+    },
+    {
+      "epoch": 0.22438650306748467,
+      "grad_norm": 0.27754291892051697,
+      "learning_rate": 0.00013329934471968555,
+      "loss": 1.0982,
+      "step": 1463
+    },
+    {
+      "epoch": 0.2245398773006135,
+      "grad_norm": 0.3554466664791107,
+      "learning_rate": 0.00013327637451121497,
+      "loss": 1.1143,
+      "step": 1464
+    },
+    {
+      "epoch": 0.22469325153374234,
+      "grad_norm": 0.31265246868133545,
+      "learning_rate": 0.00013325339049890912,
+      "loss": 1.1255,
+      "step": 1465
+    },
+    {
+      "epoch": 0.22484662576687117,
+      "grad_norm": 0.6797874569892883,
+      "learning_rate": 0.00013323039268821213,
+      "loss": 1.1882,
+      "step": 1466
+    },
+    {
+      "epoch": 0.225,
+      "grad_norm": 0.40149444341659546,
+      "learning_rate": 0.00013320738108457147,
+      "loss": 1.1169,
+      "step": 1467
+    },
+    {
+      "epoch": 0.22515337423312884,
+      "grad_norm": 0.5934581756591797,
+      "learning_rate": 0.00013318435569343784,
+      "loss": 1.0942,
+      "step": 1468
+    },
+    {
+      "epoch": 0.22530674846625767,
+      "grad_norm": 0.3019876480102539,
+      "learning_rate": 0.0001331613165202653,
+      "loss": 1.2499,
+      "step": 1469
+    },
+    {
+      "epoch": 0.2254601226993865,
+      "grad_norm": 0.2999240458011627,
+      "learning_rate": 0.00013313826357051102,
+      "loss": 1.113,
+      "step": 1470
+    },
+    {
+      "epoch": 0.22561349693251534,
+      "grad_norm": 0.33106759190559387,
+      "learning_rate": 0.00013311519684963558,
+      "loss": 1.0747,
+      "step": 1471
+    },
+    {
+      "epoch": 0.22576687116564417,
+      "grad_norm": 0.275715708732605,
+      "learning_rate": 0.0001330921163631027,
+      "loss": 1.2722,
+      "step": 1472
+    },
+    {
+      "epoch": 0.225920245398773,
+      "grad_norm": 0.34964901208877563,
+      "learning_rate": 0.00013306902211637948,
+      "loss": 1.2345,
+      "step": 1473
+    },
+    {
+      "epoch": 0.22607361963190184,
+      "grad_norm": 0.3151015639305115,
+      "learning_rate": 0.0001330459141149362,
+      "loss": 1.3989,
+      "step": 1474
+    },
+    {
+      "epoch": 0.22622699386503067,
+      "grad_norm": 0.423977792263031,
+      "learning_rate": 0.0001330227923642464,
+      "loss": 1.0759,
+      "step": 1475
+    },
+    {
+      "epoch": 0.2263803680981595,
+      "grad_norm": 0.2691545784473419,
+      "learning_rate": 0.0001329996568697869,
+      "loss": 1.2185,
+      "step": 1476
+    },
+    {
+      "epoch": 0.22653374233128834,
+      "grad_norm": 0.3952186107635498,
+      "learning_rate": 0.00013297650763703777,
+      "loss": 1.2272,
+      "step": 1477
+    },
+    {
+      "epoch": 0.22668711656441717,
+      "grad_norm": 0.3136208653450012,
+      "learning_rate": 0.00013295334467148235,
+      "loss": 1.1848,
+      "step": 1478
+    },
+    {
+      "epoch": 0.226840490797546,
+      "grad_norm": 0.5769270658493042,
+      "learning_rate": 0.0001329301679786072,
+      "loss": 0.9864,
+      "step": 1479
+    },
+    {
+      "epoch": 0.22699386503067484,
+      "grad_norm": 0.27135899662971497,
+      "learning_rate": 0.00013290697756390213,
+      "loss": 1.251,
+      "step": 1480
+    },
+    {
+      "epoch": 0.22714723926380367,
+      "grad_norm": 0.2560366094112396,
+      "learning_rate": 0.00013288377343286025,
+      "loss": 1.1393,
+      "step": 1481
+    },
+    {
+      "epoch": 0.2273006134969325,
+      "grad_norm": 0.31022390723228455,
+      "learning_rate": 0.00013286055559097787,
+      "loss": 1.1551,
+      "step": 1482
+    },
+    {
+      "epoch": 0.22745398773006134,
+      "grad_norm": 0.7119417190551758,
+      "learning_rate": 0.00013283732404375455,
+      "loss": 1.2879,
+      "step": 1483
+    },
+    {
+      "epoch": 0.22760736196319017,
+      "grad_norm": 0.9106839299201965,
+      "learning_rate": 0.00013281407879669313,
+      "loss": 1.2188,
+      "step": 1484
+    },
+    {
+      "epoch": 0.227760736196319,
+      "grad_norm": 0.5690062642097473,
+      "learning_rate": 0.0001327908198552997,
+      "loss": 1.3168,
+      "step": 1485
+    },
+    {
+      "epoch": 0.22791411042944784,
+      "grad_norm": 0.6921231746673584,
+      "learning_rate": 0.00013276754722508355,
+      "loss": 1.1125,
+      "step": 1486
+    },
+    {
+      "epoch": 0.22806748466257667,
+      "grad_norm": 0.3539598882198334,
+      "learning_rate": 0.0001327442609115572,
+      "loss": 1.1065,
+      "step": 1487
+    },
+    {
+      "epoch": 0.2282208588957055,
+      "grad_norm": 0.43658408522605896,
+      "learning_rate": 0.00013272096092023646,
+      "loss": 1.1278,
+      "step": 1488
+    },
+    {
+      "epoch": 0.22837423312883437,
+      "grad_norm": 0.44805747270584106,
+      "learning_rate": 0.0001326976472566404,
+      "loss": 1.2302,
+      "step": 1489
+    },
+    {
+      "epoch": 0.2285276073619632,
+      "grad_norm": 0.38446635007858276,
+      "learning_rate": 0.00013267431992629127,
+      "loss": 1.1511,
+      "step": 1490
+    },
+    {
+      "epoch": 0.22868098159509204,
+      "grad_norm": 0.6325763463973999,
+      "learning_rate": 0.00013265097893471457,
+      "loss": 1.1937,
+      "step": 1491
+    },
+    {
+      "epoch": 0.22883435582822087,
+      "grad_norm": 0.45965588092803955,
+      "learning_rate": 0.00013262762428743903,
+      "loss": 1.2391,
+      "step": 1492
+    },
+    {
+      "epoch": 0.2289877300613497,
+      "grad_norm": 0.43657946586608887,
+      "learning_rate": 0.00013260425598999668,
+      "loss": 1.0032,
+      "step": 1493
+    },
+    {
+      "epoch": 0.22914110429447854,
+      "grad_norm": 0.33290451765060425,
+      "learning_rate": 0.00013258087404792264,
+      "loss": 1.2743,
+      "step": 1494
+    },
+    {
+      "epoch": 0.22929447852760737,
+      "grad_norm": 0.3993147611618042,
+      "learning_rate": 0.00013255747846675547,
+      "loss": 1.1266,
+      "step": 1495
+    },
+    {
+      "epoch": 0.2294478527607362,
+      "grad_norm": 0.42214447259902954,
+      "learning_rate": 0.0001325340692520368,
+      "loss": 1.155,
+      "step": 1496
+    },
+    {
+      "epoch": 0.22960122699386504,
+      "grad_norm": 0.2955145835876465,
+      "learning_rate": 0.00013251064640931146,
+      "loss": 1.1394,
+      "step": 1497
+    },
+    {
+      "epoch": 0.22975460122699387,
+      "grad_norm": 0.4689525067806244,
+      "learning_rate": 0.0001324872099441277,
+      "loss": 1.3291,
+      "step": 1498
+    },
+    {
+      "epoch": 0.2299079754601227,
+      "grad_norm": 0.45544981956481934,
+      "learning_rate": 0.00013246375986203683,
+      "loss": 0.9955,
+      "step": 1499
+    },
+    {
+      "epoch": 0.23006134969325154,
+      "grad_norm": 0.3708842694759369,
+      "learning_rate": 0.00013244029616859342,
+      "loss": 1.2085,
+      "step": 1500
+    },
+    {
+      "epoch": 0.23021472392638037,
+      "grad_norm": 0.37807542085647583,
+      "learning_rate": 0.00013241681886935527,
+      "loss": 1.2724,
+      "step": 1501
+    },
+    {
+      "epoch": 0.2303680981595092,
+      "grad_norm": 0.6577174067497253,
+      "learning_rate": 0.00013239332796988347,
+      "loss": 0.9357,
+      "step": 1502
+    },
+    {
+      "epoch": 0.23052147239263804,
+      "grad_norm": 0.3179166316986084,
+      "learning_rate": 0.00013236982347574223,
+      "loss": 1.3045,
+      "step": 1503
+    },
+    {
+      "epoch": 0.23067484662576687,
+      "grad_norm": 0.7733244895935059,
+      "learning_rate": 0.000132346305392499,
+      "loss": 1.1542,
+      "step": 1504
+    },
+    {
+      "epoch": 0.2308282208588957,
+      "grad_norm": 0.7481251358985901,
+      "learning_rate": 0.0001323227737257245,
+      "loss": 1.1197,
+      "step": 1505
+    },
+    {
+      "epoch": 0.23098159509202454,
+      "grad_norm": 0.4447888135910034,
+      "learning_rate": 0.0001322992284809927,
+      "loss": 1.2777,
+      "step": 1506
+    },
+    {
+      "epoch": 0.23113496932515337,
+      "grad_norm": 1.3066809177398682,
+      "learning_rate": 0.0001322756696638806,
+      "loss": 1.2807,
+      "step": 1507
+    },
+    {
+      "epoch": 0.2312883435582822,
+      "grad_norm": 0.28385233879089355,
+      "learning_rate": 0.00013225209727996866,
+      "loss": 1.1472,
+      "step": 1508
+    },
+    {
+      "epoch": 0.23144171779141104,
+      "grad_norm": 0.4637300968170166,
+      "learning_rate": 0.00013222851133484038,
+      "loss": 1.2579,
+      "step": 1509
+    },
+    {
+      "epoch": 0.23159509202453987,
+      "grad_norm": 0.4624904990196228,
+      "learning_rate": 0.0001322049118340825,
+      "loss": 1.1,
+      "step": 1510
+    },
+    {
+      "epoch": 0.2317484662576687,
+      "grad_norm": 0.6591308116912842,
+      "learning_rate": 0.00013218129878328505,
+      "loss": 1.2031,
+      "step": 1511
+    },
+    {
+      "epoch": 0.23190184049079754,
+      "grad_norm": 0.535393238067627,
+      "learning_rate": 0.00013215767218804117,
+      "loss": 1.134,
+      "step": 1512
+    },
+    {
+      "epoch": 0.23205521472392637,
+      "grad_norm": 0.5642839074134827,
+      "learning_rate": 0.0001321340320539473,
+      "loss": 1.0489,
+      "step": 1513
+    },
+    {
+      "epoch": 0.2322085889570552,
+      "grad_norm": 0.29670730233192444,
+      "learning_rate": 0.000132110378386603,
+      "loss": 1.1733,
+      "step": 1514
+    },
+    {
+      "epoch": 0.23236196319018404,
+      "grad_norm": 0.5632471442222595,
+      "learning_rate": 0.00013208671119161115,
+      "loss": 1.261,
+      "step": 1515
+    },
+    {
+      "epoch": 0.23251533742331287,
+      "grad_norm": 0.5878940224647522,
+      "learning_rate": 0.00013206303047457767,
+      "loss": 1.2192,
+      "step": 1516
+    },
+    {
+      "epoch": 0.2326687116564417,
+      "grad_norm": 0.41816994547843933,
+      "learning_rate": 0.0001320393362411118,
+      "loss": 1.0962,
+      "step": 1517
+    },
+    {
+      "epoch": 0.23282208588957054,
+      "grad_norm": 0.32931211590766907,
+      "learning_rate": 0.000132015628496826,
+      "loss": 1.1743,
+      "step": 1518
+    },
+    {
+      "epoch": 0.23297546012269937,
+      "grad_norm": 0.4399535059928894,
+      "learning_rate": 0.00013199190724733588,
+      "loss": 1.1551,
+      "step": 1519
+    },
+    {
+      "epoch": 0.2331288343558282,
+      "grad_norm": 0.3388459384441376,
+      "learning_rate": 0.00013196817249826022,
+      "loss": 1.2779,
+      "step": 1520
+    },
+    {
+      "epoch": 0.23328220858895707,
+      "grad_norm": 0.4899957478046417,
+      "learning_rate": 0.00013194442425522107,
+      "loss": 1.1607,
+      "step": 1521
+    },
+    {
+      "epoch": 0.2334355828220859,
+      "grad_norm": 0.6026613712310791,
+      "learning_rate": 0.00013192066252384356,
+      "loss": 1.1878,
+      "step": 1522
+    },
+    {
+      "epoch": 0.23358895705521474,
+      "grad_norm": 0.27890220284461975,
+      "learning_rate": 0.00013189688730975621,
+      "loss": 1.1949,
+      "step": 1523
+    },
+    {
+      "epoch": 0.23374233128834357,
+      "grad_norm": 0.4225197434425354,
+      "learning_rate": 0.00013187309861859055,
+      "loss": 1.2629,
+      "step": 1524
+    },
+    {
+      "epoch": 0.2338957055214724,
+      "grad_norm": 0.40191560983657837,
+      "learning_rate": 0.0001318492964559814,
+      "loss": 1.0638,
+      "step": 1525
+    },
+    {
+      "epoch": 0.23404907975460124,
+      "grad_norm": 0.4727327525615692,
+      "learning_rate": 0.0001318254808275667,
+      "loss": 1.1454,
+      "step": 1526
+    },
+    {
+      "epoch": 0.23420245398773007,
+      "grad_norm": 0.33942678570747375,
+      "learning_rate": 0.00013180165173898764,
+      "loss": 1.1462,
+      "step": 1527
+    },
+    {
+      "epoch": 0.2343558282208589,
+      "grad_norm": 1.219256043434143,
+      "learning_rate": 0.0001317778091958886,
+      "loss": 1.2231,
+      "step": 1528
+    },
+    {
+      "epoch": 0.23450920245398774,
+      "grad_norm": 0.35999107360839844,
+      "learning_rate": 0.0001317539532039171,
+      "loss": 1.1842,
+      "step": 1529
+    },
+    {
+      "epoch": 0.23466257668711657,
+      "grad_norm": 0.5462523102760315,
+      "learning_rate": 0.0001317300837687239,
+      "loss": 1.1497,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2348159509202454,
+      "grad_norm": 0.31782227754592896,
+      "learning_rate": 0.0001317062008959629,
+      "loss": 1.2165,
+      "step": 1531
+    },
+    {
+      "epoch": 0.23496932515337424,
+      "grad_norm": 0.35214725136756897,
+      "learning_rate": 0.0001316823045912912,
+      "loss": 1.1966,
+      "step": 1532
+    },
+    {
+      "epoch": 0.23512269938650307,
+      "grad_norm": 0.26995301246643066,
+      "learning_rate": 0.0001316583948603691,
+      "loss": 1.25,
+      "step": 1533
+    },
+    {
+      "epoch": 0.2352760736196319,
+      "grad_norm": 0.38915473222732544,
+      "learning_rate": 0.00013163447170886,
+      "loss": 1.1589,
+      "step": 1534
+    },
+    {
+      "epoch": 0.23542944785276074,
+      "grad_norm": 0.40822306275367737,
+      "learning_rate": 0.00013161053514243064,
+      "loss": 1.1088,
+      "step": 1535
+    },
+    {
+      "epoch": 0.23558282208588957,
+      "grad_norm": 0.3806876242160797,
+      "learning_rate": 0.00013158658516675075,
+      "loss": 1.3371,
+      "step": 1536
+    },
+    {
+      "epoch": 0.2357361963190184,
+      "grad_norm": 0.27456697821617126,
+      "learning_rate": 0.00013156262178749337,
+      "loss": 1.1186,
+      "step": 1537
+    },
+    {
+      "epoch": 0.23588957055214724,
+      "grad_norm": 0.29795610904693604,
+      "learning_rate": 0.0001315386450103346,
+      "loss": 1.1643,
+      "step": 1538
+    },
+    {
+      "epoch": 0.23604294478527607,
+      "grad_norm": 0.4605606198310852,
+      "learning_rate": 0.0001315146548409539,
+      "loss": 1.2991,
+      "step": 1539
+    },
+    {
+      "epoch": 0.2361963190184049,
+      "grad_norm": 0.2858566641807556,
+      "learning_rate": 0.0001314906512850337,
+      "loss": 1.082,
+      "step": 1540
+    },
+    {
+      "epoch": 0.23634969325153374,
+      "grad_norm": 0.28060099482536316,
+      "learning_rate": 0.0001314666343482597,
+      "loss": 1.1567,
+      "step": 1541
+    },
+    {
+      "epoch": 0.23650306748466257,
+      "grad_norm": 0.26639530062675476,
+      "learning_rate": 0.00013144260403632078,
+      "loss": 1.1533,
+      "step": 1542
+    },
+    {
+      "epoch": 0.2366564417177914,
+      "grad_norm": 0.3050497770309448,
+      "learning_rate": 0.00013141856035490894,
+      "loss": 1.272,
+      "step": 1543
+    },
+    {
+      "epoch": 0.23680981595092024,
+      "grad_norm": 0.34869149327278137,
+      "learning_rate": 0.00013139450330971932,
+      "loss": 1.0787,
+      "step": 1544
+    },
+    {
+      "epoch": 0.23696319018404907,
+      "grad_norm": 0.49356991052627563,
+      "learning_rate": 0.0001313704329064504,
+      "loss": 1.2094,
+      "step": 1545
+    },
+    {
+      "epoch": 0.2371165644171779,
+      "grad_norm": 0.288102924823761,
+      "learning_rate": 0.0001313463491508036,
+      "loss": 1.0986,
+      "step": 1546
+    },
+    {
+      "epoch": 0.23726993865030674,
+      "grad_norm": 0.3910624384880066,
+      "learning_rate": 0.00013132225204848363,
+      "loss": 1.1423,
+      "step": 1547
+    },
+    {
+      "epoch": 0.23742331288343557,
+      "grad_norm": 0.3096100389957428,
+      "learning_rate": 0.00013129814160519834,
+      "loss": 1.271,
+      "step": 1548
+    },
+    {
+      "epoch": 0.2375766871165644,
+      "grad_norm": 0.2893604338169098,
+      "learning_rate": 0.00013127401782665873,
+      "loss": 1.1672,
+      "step": 1549
+    },
+    {
+      "epoch": 0.23773006134969324,
+      "grad_norm": 0.2662157416343689,
+      "learning_rate": 0.00013124988071857896,
+      "loss": 1.1243,
+      "step": 1550
+    },
+    {
+      "epoch": 0.23788343558282207,
+      "grad_norm": 0.6219305396080017,
+      "learning_rate": 0.00013122573028667635,
+      "loss": 1.1435,
+      "step": 1551
+    },
+    {
+      "epoch": 0.23803680981595093,
+      "grad_norm": 0.262065589427948,
+      "learning_rate": 0.00013120156653667137,
+      "loss": 1.1038,
+      "step": 1552
+    },
+    {
+      "epoch": 0.23819018404907977,
+      "grad_norm": 0.3350154757499695,
+      "learning_rate": 0.00013117738947428765,
+      "loss": 1.2244,
+      "step": 1553
+    },
+    {
+      "epoch": 0.2383435582822086,
+      "grad_norm": 0.30722594261169434,
+      "learning_rate": 0.00013115319910525202,
+      "loss": 1.0959,
+      "step": 1554
+    },
+    {
+      "epoch": 0.23849693251533743,
+      "grad_norm": 0.3229091465473175,
+      "learning_rate": 0.00013112899543529436,
+      "loss": 1.1479,
+      "step": 1555
+    },
+    {
+      "epoch": 0.23865030674846627,
+      "grad_norm": 0.5248367190361023,
+      "learning_rate": 0.00013110477847014774,
+      "loss": 1.2398,
+      "step": 1556
+    },
+    {
+      "epoch": 0.2388036809815951,
+      "grad_norm": 0.2705022692680359,
+      "learning_rate": 0.00013108054821554848,
+      "loss": 1.3052,
+      "step": 1557
+    },
+    {
+      "epoch": 0.23895705521472393,
+      "grad_norm": 0.3054780960083008,
+      "learning_rate": 0.00013105630467723586,
+      "loss": 1.107,
+      "step": 1558
+    },
+    {
+      "epoch": 0.23911042944785277,
+      "grad_norm": 0.5173863172531128,
+      "learning_rate": 0.00013103204786095248,
+      "loss": 1.118,
+      "step": 1559
+    },
+    {
+      "epoch": 0.2392638036809816,
+      "grad_norm": 0.39121127128601074,
+      "learning_rate": 0.00013100777777244398,
+      "loss": 1.2098,
+      "step": 1560
+    },
+    {
+      "epoch": 0.23941717791411044,
+      "grad_norm": 0.38801413774490356,
+      "learning_rate": 0.00013098349441745922,
+      "loss": 1.0835,
+      "step": 1561
+    },
+    {
+      "epoch": 0.23957055214723927,
+      "grad_norm": 0.7164775729179382,
+      "learning_rate": 0.00013095919780175012,
+      "loss": 1.1154,
+      "step": 1562
+    },
+    {
+      "epoch": 0.2397239263803681,
+      "grad_norm": 0.43947407603263855,
+      "learning_rate": 0.0001309348879310718,
+      "loss": 1.2327,
+      "step": 1563
+    },
+    {
+      "epoch": 0.23987730061349694,
+      "grad_norm": 0.5260670185089111,
+      "learning_rate": 0.0001309105648111825,
+      "loss": 1.2511,
+      "step": 1564
+    },
+    {
+      "epoch": 0.24003067484662577,
+      "grad_norm": 0.37852105498313904,
+      "learning_rate": 0.0001308862284478436,
+      "loss": 1.271,
+      "step": 1565
+    },
+    {
+      "epoch": 0.2401840490797546,
+      "grad_norm": 0.29975423216819763,
+      "learning_rate": 0.0001308618788468196,
+      "loss": 1.0846,
+      "step": 1566
+    },
+    {
+      "epoch": 0.24033742331288344,
+      "grad_norm": 0.3109259307384491,
+      "learning_rate": 0.0001308375160138782,
+      "loss": 1.0987,
+      "step": 1567
+    },
+    {
+      "epoch": 0.24049079754601227,
+      "grad_norm": 0.4331613779067993,
+      "learning_rate": 0.0001308131399547901,
+      "loss": 1.2136,
+      "step": 1568
+    },
+    {
+      "epoch": 0.2406441717791411,
+      "grad_norm": 0.3757682740688324,
+      "learning_rate": 0.00013078875067532933,
+      "loss": 1.2074,
+      "step": 1569
+    },
+    {
+      "epoch": 0.24079754601226994,
+      "grad_norm": 0.5531485676765442,
+      "learning_rate": 0.00013076434818127286,
+      "loss": 1.1896,
+      "step": 1570
+    },
+    {
+      "epoch": 0.24095092024539877,
+      "grad_norm": 0.4385623335838318,
+      "learning_rate": 0.0001307399324784009,
+      "loss": 1.29,
+      "step": 1571
+    },
+    {
+      "epoch": 0.2411042944785276,
+      "grad_norm": 0.33539628982543945,
+      "learning_rate": 0.00013071550357249677,
+      "loss": 1.1937,
+      "step": 1572
+    },
+    {
+      "epoch": 0.24125766871165644,
+      "grad_norm": 0.34692442417144775,
+      "learning_rate": 0.0001306910614693469,
+      "loss": 1.1421,
+      "step": 1573
+    },
+    {
+      "epoch": 0.24141104294478527,
+      "grad_norm": 0.32472455501556396,
+      "learning_rate": 0.0001306666061747408,
+      "loss": 1.2103,
+      "step": 1574
+    },
+    {
+      "epoch": 0.2415644171779141,
+      "grad_norm": 0.5431707501411438,
+      "learning_rate": 0.00013064213769447126,
+      "loss": 1.3049,
+      "step": 1575
+    },
+    {
+      "epoch": 0.24171779141104294,
+      "grad_norm": 0.5633724927902222,
+      "learning_rate": 0.000130617656034334,
+      "loss": 1.1674,
+      "step": 1576
+    },
+    {
+      "epoch": 0.24187116564417177,
+      "grad_norm": 0.566586434841156,
+      "learning_rate": 0.000130593161200128,
+      "loss": 1.1754,
+      "step": 1577
+    },
+    {
+      "epoch": 0.2420245398773006,
+      "grad_norm": 0.2965088188648224,
+      "learning_rate": 0.00013056865319765532,
+      "loss": 1.1865,
+      "step": 1578
+    },
+    {
+      "epoch": 0.24217791411042944,
+      "grad_norm": 0.34909582138061523,
+      "learning_rate": 0.0001305441320327211,
+      "loss": 1.1709,
+      "step": 1579
+    },
+    {
+      "epoch": 0.24233128834355827,
+      "grad_norm": 0.2724282741546631,
+      "learning_rate": 0.00013051959771113365,
+      "loss": 1.2471,
+      "step": 1580
+    },
+    {
+      "epoch": 0.2424846625766871,
+      "grad_norm": 0.4572576582431793,
+      "learning_rate": 0.0001304950502387044,
+      "loss": 1.0691,
+      "step": 1581
+    },
+    {
+      "epoch": 0.24263803680981594,
+      "grad_norm": 0.3091803193092346,
+      "learning_rate": 0.0001304704896212478,
+      "loss": 1.1134,
+      "step": 1582
+    },
+    {
+      "epoch": 0.2427914110429448,
+      "grad_norm": 0.6603121757507324,
+      "learning_rate": 0.00013044591586458158,
+      "loss": 1.2978,
+      "step": 1583
+    },
+    {
+      "epoch": 0.24294478527607363,
+      "grad_norm": 0.3515135645866394,
+      "learning_rate": 0.00013042132897452642,
+      "loss": 1.1155,
+      "step": 1584
+    },
+    {
+      "epoch": 0.24309815950920247,
+      "grad_norm": 0.29700207710266113,
+      "learning_rate": 0.0001303967289569062,
+      "loss": 0.9949,
+      "step": 1585
+    },
+    {
+      "epoch": 0.2432515337423313,
+      "grad_norm": 0.3264654874801636,
+      "learning_rate": 0.00013037211581754787,
+      "loss": 1.0008,
+      "step": 1586
+    },
+    {
+      "epoch": 0.24340490797546013,
+      "grad_norm": 0.26765918731689453,
+      "learning_rate": 0.00013034748956228156,
+      "loss": 1.2055,
+      "step": 1587
+    },
+    {
+      "epoch": 0.24355828220858897,
+      "grad_norm": 0.3094084560871124,
+      "learning_rate": 0.0001303228501969404,
+      "loss": 1.1765,
+      "step": 1588
+    },
+    {
+      "epoch": 0.2437116564417178,
+      "grad_norm": 0.37831974029541016,
+      "learning_rate": 0.00013029819772736065,
+      "loss": 1.2314,
+      "step": 1589
+    },
+    {
+      "epoch": 0.24386503067484663,
+      "grad_norm": 0.2780807614326477,
+      "learning_rate": 0.0001302735321593818,
+      "loss": 0.9494,
+      "step": 1590
+    },
+    {
+      "epoch": 0.24401840490797547,
+      "grad_norm": 0.31754422187805176,
+      "learning_rate": 0.0001302488534988463,
+      "loss": 1.1048,
+      "step": 1591
+    },
+    {
+      "epoch": 0.2441717791411043,
+      "grad_norm": 0.3702795207500458,
+      "learning_rate": 0.00013022416175159974,
+      "loss": 1.1608,
+      "step": 1592
+    },
+    {
+      "epoch": 0.24432515337423313,
+      "grad_norm": 0.4959076941013336,
+      "learning_rate": 0.0001301994569234908,
+      "loss": 1.0805,
+      "step": 1593
+    },
+    {
+      "epoch": 0.24447852760736197,
+      "grad_norm": 0.24731668829917908,
+      "learning_rate": 0.0001301747390203713,
+      "loss": 1.0746,
+      "step": 1594
+    },
+    {
+      "epoch": 0.2446319018404908,
+      "grad_norm": 0.2618066668510437,
+      "learning_rate": 0.0001301500080480961,
+      "loss": 1.1575,
+      "step": 1595
+    },
+    {
+      "epoch": 0.24478527607361963,
+      "grad_norm": 0.5735589265823364,
+      "learning_rate": 0.00013012526401252325,
+      "loss": 1.188,
+      "step": 1596
+    },
+    {
+      "epoch": 0.24493865030674847,
+      "grad_norm": 0.3973346948623657,
+      "learning_rate": 0.00013010050691951373,
+      "loss": 1.0854,
+      "step": 1597
+    },
+    {
+      "epoch": 0.2450920245398773,
+      "grad_norm": 0.39300745725631714,
+      "learning_rate": 0.00013007573677493184,
+      "loss": 1.1993,
+      "step": 1598
+    },
+    {
+      "epoch": 0.24524539877300613,
+      "grad_norm": 0.25208231806755066,
+      "learning_rate": 0.00013005095358464478,
+      "loss": 1.1769,
+      "step": 1599
+    },
+    {
+      "epoch": 0.24539877300613497,
+      "grad_norm": 0.2967680096626282,
+      "learning_rate": 0.00013002615735452288,
+      "loss": 1.1462,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2455521472392638,
+      "grad_norm": 0.3026011288166046,
+      "learning_rate": 0.00013000134809043964,
+      "loss": 1.1678,
+      "step": 1601
+    },
+    {
+      "epoch": 0.24570552147239264,
+      "grad_norm": 0.48279646039009094,
+      "learning_rate": 0.00012997652579827153,
+      "loss": 1.14,
+      "step": 1602
+    },
+    {
+      "epoch": 0.24585889570552147,
+      "grad_norm": 0.33869418501853943,
+      "learning_rate": 0.00012995169048389822,
+      "loss": 1.1898,
+      "step": 1603
+    },
+    {
+      "epoch": 0.2460122699386503,
+      "grad_norm": 0.39897239208221436,
+      "learning_rate": 0.00012992684215320241,
+      "loss": 1.1408,
+      "step": 1604
+    },
+    {
+      "epoch": 0.24616564417177914,
+      "grad_norm": 0.5536998510360718,
+      "learning_rate": 0.00012990198081206988,
+      "loss": 1.1201,
+      "step": 1605
+    },
+    {
+      "epoch": 0.24631901840490797,
+      "grad_norm": 0.2741105556488037,
+      "learning_rate": 0.00012987710646638947,
+      "loss": 1.2297,
+      "step": 1606
+    },
+    {
+      "epoch": 0.2464723926380368,
+      "grad_norm": 0.33702632784843445,
+      "learning_rate": 0.00012985221912205315,
+      "loss": 1.1969,
+      "step": 1607
+    },
+    {
+      "epoch": 0.24662576687116564,
+      "grad_norm": 0.6848388910293579,
+      "learning_rate": 0.00012982731878495595,
+      "loss": 1.3391,
+      "step": 1608
+    },
+    {
+      "epoch": 0.24677914110429447,
+      "grad_norm": 0.38302531838417053,
+      "learning_rate": 0.00012980240546099598,
+      "loss": 1.0269,
+      "step": 1609
+    },
+    {
+      "epoch": 0.2469325153374233,
+      "grad_norm": 0.41164326667785645,
+      "learning_rate": 0.0001297774791560744,
+      "loss": 1.1699,
+      "step": 1610
+    },
+    {
+      "epoch": 0.24708588957055214,
+      "grad_norm": 0.24806590378284454,
+      "learning_rate": 0.0001297525398760955,
+      "loss": 1.0524,
+      "step": 1611
+    },
+    {
+      "epoch": 0.24723926380368097,
+      "grad_norm": 0.2660265266895294,
+      "learning_rate": 0.00012972758762696656,
+      "loss": 1.3099,
+      "step": 1612
+    },
+    {
+      "epoch": 0.2473926380368098,
+      "grad_norm": 0.3011568486690521,
+      "learning_rate": 0.000129702622414598,
+      "loss": 1.1521,
+      "step": 1613
+    },
+    {
+      "epoch": 0.24754601226993864,
+      "grad_norm": 0.3706291913986206,
+      "learning_rate": 0.00012967764424490333,
+      "loss": 1.1881,
+      "step": 1614
+    },
+    {
+      "epoch": 0.2476993865030675,
+      "grad_norm": 0.3098074495792389,
+      "learning_rate": 0.00012965265312379903,
+      "loss": 1.1578,
+      "step": 1615
+    },
+    {
+      "epoch": 0.24785276073619633,
+      "grad_norm": 0.32589712738990784,
+      "learning_rate": 0.00012962764905720476,
+      "loss": 1.2844,
+      "step": 1616
+    },
+    {
+      "epoch": 0.24800613496932516,
+      "grad_norm": 0.7810177803039551,
+      "learning_rate": 0.00012960263205104316,
+      "loss": 1.1977,
+      "step": 1617
+    },
+    {
+      "epoch": 0.248159509202454,
+      "grad_norm": 0.23413997888565063,
+      "learning_rate": 0.00012957760211123999,
+      "loss": 1.1948,
+      "step": 1618
+    },
+    {
+      "epoch": 0.24831288343558283,
+      "grad_norm": 0.37678784132003784,
+      "learning_rate": 0.00012955255924372404,
+      "loss": 1.1039,
+      "step": 1619
+    },
+    {
+      "epoch": 0.24846625766871167,
+      "grad_norm": 0.353674978017807,
+      "learning_rate": 0.00012952750345442718,
+      "loss": 1.1921,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2486196319018405,
+      "grad_norm": 0.6872379779815674,
+      "learning_rate": 0.00012950243474928432,
+      "loss": 1.0687,
+      "step": 1621
+    },
+    {
+      "epoch": 0.24877300613496933,
+      "grad_norm": 0.24359232187271118,
+      "learning_rate": 0.0001294773531342335,
+      "loss": 1.0969,
+      "step": 1622
+    },
+    {
+      "epoch": 0.24892638036809817,
+      "grad_norm": 0.2645212411880493,
+      "learning_rate": 0.0001294522586152157,
+      "loss": 1.1016,
+      "step": 1623
+    },
+    {
+      "epoch": 0.249079754601227,
+      "grad_norm": 0.3842701315879822,
+      "learning_rate": 0.00012942715119817504,
+      "loss": 1.1293,
+      "step": 1624
+    },
+    {
+      "epoch": 0.24923312883435583,
+      "grad_norm": 0.7154290676116943,
+      "learning_rate": 0.0001294020308890587,
+      "loss": 1.1423,
+      "step": 1625
+    },
+    {
+      "epoch": 0.24938650306748467,
+      "grad_norm": 0.28489992022514343,
+      "learning_rate": 0.00012937689769381688,
+      "loss": 1.1743,
+      "step": 1626
+    },
+    {
+      "epoch": 0.2495398773006135,
+      "grad_norm": 0.4794822335243225,
+      "learning_rate": 0.0001293517516184028,
+      "loss": 1.3101,
+      "step": 1627
+    },
+    {
+      "epoch": 0.24969325153374233,
+      "grad_norm": 0.322126567363739,
+      "learning_rate": 0.00012932659266877284,
+      "loss": 1.086,
+      "step": 1628
+    },
+    {
+      "epoch": 0.24984662576687117,
+      "grad_norm": 0.3989594578742981,
+      "learning_rate": 0.00012930142085088633,
+      "loss": 1.3027,
+      "step": 1629
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3991124927997589,
+      "learning_rate": 0.00012927623617070566,
+      "loss": 1.1214,
+      "step": 1630
+    },
+    {
+      "epoch": 0.25015337423312883,
+      "grad_norm": 0.24785183370113373,
+      "learning_rate": 0.00012925103863419628,
+      "loss": 1.1813,
+      "step": 1631
+    },
+    {
+      "epoch": 0.25030674846625767,
+      "grad_norm": 0.6339516043663025,
+      "learning_rate": 0.00012922582824732674,
+      "loss": 1.0664,
+      "step": 1632
+    },
+    {
+      "epoch": 0.2504601226993865,
+      "grad_norm": 0.31884288787841797,
+      "learning_rate": 0.00012920060501606857,
+      "loss": 1.1399,
+      "step": 1633
+    },
+    {
+      "epoch": 0.25061349693251533,
+      "grad_norm": 0.368842750787735,
+      "learning_rate": 0.00012917536894639634,
+      "loss": 1.1254,
+      "step": 1634
+    },
+    {
+      "epoch": 0.25076687116564417,
+      "grad_norm": 0.6924677491188049,
+      "learning_rate": 0.00012915012004428766,
+      "loss": 1.1477,
+      "step": 1635
+    },
+    {
+      "epoch": 0.250920245398773,
+      "grad_norm": 0.41479969024658203,
+      "learning_rate": 0.00012912485831572326,
+      "loss": 1.1946,
+      "step": 1636
+    },
+    {
+      "epoch": 0.25107361963190183,
+      "grad_norm": 0.616326093673706,
+      "learning_rate": 0.0001290995837666868,
+      "loss": 1.2471,
+      "step": 1637
+    },
+    {
+      "epoch": 0.25122699386503067,
+      "grad_norm": 0.34676626324653625,
+      "learning_rate": 0.00012907429640316502,
+      "loss": 1.1719,
+      "step": 1638
+    },
+    {
+      "epoch": 0.2513803680981595,
+      "grad_norm": 0.3227722644805908,
+      "learning_rate": 0.00012904899623114775,
+      "loss": 1.0074,
+      "step": 1639
+    },
+    {
+      "epoch": 0.25153374233128833,
+      "grad_norm": 0.27663588523864746,
+      "learning_rate": 0.00012902368325662776,
+      "loss": 1.0776,
+      "step": 1640
+    },
+    {
+      "epoch": 0.25168711656441717,
+      "grad_norm": 0.28498637676239014,
+      "learning_rate": 0.00012899835748560092,
+      "loss": 1.1391,
+      "step": 1641
+    },
+    {
+      "epoch": 0.251840490797546,
+      "grad_norm": 0.3311587870121002,
+      "learning_rate": 0.00012897301892406604,
+      "loss": 1.1629,
+      "step": 1642
+    },
+    {
+      "epoch": 0.25199386503067484,
+      "grad_norm": 0.2729542553424835,
+      "learning_rate": 0.0001289476675780251,
+      "loss": 1.0783,
+      "step": 1643
+    },
+    {
+      "epoch": 0.25214723926380367,
+      "grad_norm": 0.38988688588142395,
+      "learning_rate": 0.00012892230345348303,
+      "loss": 1.0579,
+      "step": 1644
+    },
+    {
+      "epoch": 0.2523006134969325,
+      "grad_norm": 0.6141564249992371,
+      "learning_rate": 0.00012889692655644776,
+      "loss": 1.1624,
+      "step": 1645
+    },
+    {
+      "epoch": 0.25245398773006134,
+      "grad_norm": 0.9049823880195618,
+      "learning_rate": 0.00012887153689293028,
+      "loss": 1.2854,
+      "step": 1646
+    },
+    {
+      "epoch": 0.25260736196319017,
+      "grad_norm": 0.44914108514785767,
+      "learning_rate": 0.0001288461344689446,
+      "loss": 1.2272,
+      "step": 1647
+    },
+    {
+      "epoch": 0.252760736196319,
+      "grad_norm": 0.31848281621932983,
+      "learning_rate": 0.00012882071929050777,
+      "loss": 1.2427,
+      "step": 1648
+    },
+    {
+      "epoch": 0.25291411042944784,
+      "grad_norm": 0.4284230172634125,
+      "learning_rate": 0.00012879529136363983,
+      "loss": 1.1106,
+      "step": 1649
+    },
+    {
+      "epoch": 0.25306748466257667,
+      "grad_norm": 0.3254087269306183,
+      "learning_rate": 0.0001287698506943639,
+      "loss": 1.2928,
+      "step": 1650
+    },
+    {
+      "epoch": 0.2532208588957055,
+      "grad_norm": 0.3734246790409088,
+      "learning_rate": 0.000128744397288706,
+      "loss": 0.9772,
+      "step": 1651
+    },
+    {
+      "epoch": 0.25337423312883434,
+      "grad_norm": 0.27902284264564514,
+      "learning_rate": 0.00012871893115269528,
+      "loss": 1.3885,
+      "step": 1652
+    },
+    {
+      "epoch": 0.25352760736196317,
+      "grad_norm": 0.4739260673522949,
+      "learning_rate": 0.00012869345229236386,
+      "loss": 1.0924,
+      "step": 1653
+    },
+    {
+      "epoch": 0.253680981595092,
+      "grad_norm": 0.27953019738197327,
+      "learning_rate": 0.00012866796071374687,
+      "loss": 1.0516,
+      "step": 1654
+    },
+    {
+      "epoch": 0.25383435582822084,
+      "grad_norm": 0.9085614681243896,
+      "learning_rate": 0.00012864245642288245,
+      "loss": 1.2831,
+      "step": 1655
+    },
+    {
+      "epoch": 0.25398773006134967,
+      "grad_norm": 0.26268115639686584,
+      "learning_rate": 0.0001286169394258118,
+      "loss": 1.1576,
+      "step": 1656
+    },
+    {
+      "epoch": 0.2541411042944785,
+      "grad_norm": 0.3797646760940552,
+      "learning_rate": 0.00012859140972857906,
+      "loss": 1.1062,
+      "step": 1657
+    },
+    {
+      "epoch": 0.25429447852760734,
+      "grad_norm": 0.40319445729255676,
+      "learning_rate": 0.00012856586733723141,
+      "loss": 1.2496,
+      "step": 1658
+    },
+    {
+      "epoch": 0.25444785276073617,
+      "grad_norm": 0.5084361433982849,
+      "learning_rate": 0.0001285403122578191,
+      "loss": 1.3044,
+      "step": 1659
+    },
+    {
+      "epoch": 0.254601226993865,
+      "grad_norm": 0.31476977467536926,
+      "learning_rate": 0.0001285147444963952,
+      "loss": 1.2459,
+      "step": 1660
+    },
+    {
+      "epoch": 0.25475460122699384,
+      "grad_norm": 0.6129370927810669,
+      "learning_rate": 0.00012848916405901602,
+      "loss": 1.1319,
+      "step": 1661
+    },
+    {
+      "epoch": 0.2549079754601227,
+      "grad_norm": 0.3505834639072418,
+      "learning_rate": 0.0001284635709517407,
+      "loss": 1.2903,
+      "step": 1662
+    },
+    {
+      "epoch": 0.25506134969325156,
+      "grad_norm": 0.5543643832206726,
+      "learning_rate": 0.00012843796518063145,
+      "loss": 1.1427,
+      "step": 1663
+    },
+    {
+      "epoch": 0.2552147239263804,
+      "grad_norm": 0.2996883988380432,
+      "learning_rate": 0.00012841234675175348,
+      "loss": 1.134,
+      "step": 1664
+    },
+    {
+      "epoch": 0.2553680981595092,
+      "grad_norm": 0.2726325988769531,
+      "learning_rate": 0.000128386715671175,
+      "loss": 1.1246,
+      "step": 1665
+    },
+    {
+      "epoch": 0.25552147239263806,
+      "grad_norm": 0.33595702052116394,
+      "learning_rate": 0.00012836107194496717,
+      "loss": 1.214,
+      "step": 1666
+    },
+    {
+      "epoch": 0.2556748466257669,
+      "grad_norm": 0.33279335498809814,
+      "learning_rate": 0.0001283354155792042,
+      "loss": 1.2262,
+      "step": 1667
+    },
+    {
+      "epoch": 0.2558282208588957,
+      "grad_norm": 0.551201343536377,
+      "learning_rate": 0.00012830974657996325,
+      "loss": 1.0922,
+      "step": 1668
+    },
+    {
+      "epoch": 0.25598159509202456,
+      "grad_norm": 0.26764634251594543,
+      "learning_rate": 0.00012828406495332456,
+      "loss": 1.0642,
+      "step": 1669
+    },
+    {
+      "epoch": 0.2561349693251534,
+      "grad_norm": 0.32928338646888733,
+      "learning_rate": 0.00012825837070537123,
+      "loss": 1.0973,
+      "step": 1670
+    },
+    {
+      "epoch": 0.25628834355828223,
+      "grad_norm": 0.26610153913497925,
+      "learning_rate": 0.00012823266384218946,
+      "loss": 1.2582,
+      "step": 1671
+    },
+    {
+      "epoch": 0.25644171779141106,
+      "grad_norm": 0.28347620368003845,
+      "learning_rate": 0.00012820694436986834,
+      "loss": 1.2093,
+      "step": 1672
+    },
+    {
+      "epoch": 0.2565950920245399,
+      "grad_norm": 0.2946341335773468,
+      "learning_rate": 0.0001281812122945001,
+      "loss": 1.0805,
+      "step": 1673
+    },
+    {
+      "epoch": 0.25674846625766873,
+      "grad_norm": 0.3004854917526245,
+      "learning_rate": 0.00012815546762217972,
+      "loss": 1.2067,
+      "step": 1674
+    },
+    {
+      "epoch": 0.25690184049079756,
+      "grad_norm": 0.33331528306007385,
+      "learning_rate": 0.00012812971035900542,
+      "loss": 1.067,
+      "step": 1675
+    },
+    {
+      "epoch": 0.2570552147239264,
+      "grad_norm": 0.25203633308410645,
+      "learning_rate": 0.00012810394051107824,
+      "loss": 1.0184,
+      "step": 1676
+    },
+    {
+      "epoch": 0.25720858895705523,
+      "grad_norm": 0.3199441134929657,
+      "learning_rate": 0.00012807815808450225,
+      "loss": 1.114,
+      "step": 1677
+    },
+    {
+      "epoch": 0.25736196319018406,
+      "grad_norm": 0.2580183148384094,
+      "learning_rate": 0.00012805236308538448,
+      "loss": 1.1274,
+      "step": 1678
+    },
+    {
+      "epoch": 0.2575153374233129,
+      "grad_norm": 0.24814966320991516,
+      "learning_rate": 0.00012802655551983496,
+      "loss": 1.1974,
+      "step": 1679
+    },
+    {
+      "epoch": 0.25766871165644173,
+      "grad_norm": 0.2282392531633377,
+      "learning_rate": 0.0001280007353939667,
+      "loss": 0.9954,
+      "step": 1680
+    },
+    {
+      "epoch": 0.25782208588957056,
+      "grad_norm": 0.5551994442939758,
+      "learning_rate": 0.00012797490271389565,
+      "loss": 1.2324,
+      "step": 1681
+    },
+    {
+      "epoch": 0.2579754601226994,
+      "grad_norm": 0.267819881439209,
+      "learning_rate": 0.00012794905748574075,
+      "loss": 1.0386,
+      "step": 1682
+    },
+    {
+      "epoch": 0.25812883435582823,
+      "grad_norm": 0.29086530208587646,
+      "learning_rate": 0.00012792319971562398,
+      "loss": 1.1319,
+      "step": 1683
+    },
+    {
+      "epoch": 0.25828220858895706,
+      "grad_norm": 0.336639940738678,
+      "learning_rate": 0.00012789732940967017,
+      "loss": 1.1122,
+      "step": 1684
+    },
+    {
+      "epoch": 0.2584355828220859,
+      "grad_norm": 0.28534552454948425,
+      "learning_rate": 0.0001278714465740072,
+      "loss": 1.3037,
+      "step": 1685
+    },
+    {
+      "epoch": 0.25858895705521473,
+      "grad_norm": 0.27174675464630127,
+      "learning_rate": 0.0001278455512147659,
+      "loss": 1.3008,
+      "step": 1686
+    },
+    {
+      "epoch": 0.25874233128834356,
+      "grad_norm": 0.26546740531921387,
+      "learning_rate": 0.00012781964333808006,
+      "loss": 1.1149,
+      "step": 1687
+    },
+    {
+      "epoch": 0.2588957055214724,
+      "grad_norm": 0.30852973461151123,
+      "learning_rate": 0.00012779372295008645,
+      "loss": 1.1905,
+      "step": 1688
+    },
+    {
+      "epoch": 0.25904907975460123,
+      "grad_norm": 0.47795557975769043,
+      "learning_rate": 0.00012776779005692476,
+      "loss": 0.9764,
+      "step": 1689
+    },
+    {
+      "epoch": 0.25920245398773006,
+      "grad_norm": 0.24956677854061127,
+      "learning_rate": 0.0001277418446647377,
+      "loss": 1.1117,
+      "step": 1690
+    },
+    {
+      "epoch": 0.2593558282208589,
+      "grad_norm": 0.31286460161209106,
+      "learning_rate": 0.00012771588677967091,
+      "loss": 1.0498,
+      "step": 1691
+    },
+    {
+      "epoch": 0.25950920245398773,
+      "grad_norm": 0.25914907455444336,
+      "learning_rate": 0.000127689916407873,
+      "loss": 1.2762,
+      "step": 1692
+    },
+    {
+      "epoch": 0.25966257668711656,
+      "grad_norm": 0.40418484807014465,
+      "learning_rate": 0.00012766393355549555,
+      "loss": 1.153,
+      "step": 1693
+    },
+    {
+      "epoch": 0.2598159509202454,
+      "grad_norm": 0.28620991110801697,
+      "learning_rate": 0.00012763793822869303,
+      "loss": 1.0541,
+      "step": 1694
+    },
+    {
+      "epoch": 0.25996932515337423,
+      "grad_norm": 0.46040502190589905,
+      "learning_rate": 0.00012761193043362294,
+      "loss": 1.2471,
+      "step": 1695
+    },
+    {
+      "epoch": 0.26012269938650306,
+      "grad_norm": 0.4214593768119812,
+      "learning_rate": 0.00012758591017644573,
+      "loss": 1.0255,
+      "step": 1696
+    },
+    {
+      "epoch": 0.2602760736196319,
+      "grad_norm": 0.38870853185653687,
+      "learning_rate": 0.00012755987746332475,
+      "loss": 1.1442,
+      "step": 1697
+    },
+    {
+      "epoch": 0.26042944785276073,
+      "grad_norm": 0.3891431987285614,
+      "learning_rate": 0.00012753383230042632,
+      "loss": 1.2204,
+      "step": 1698
+    },
+    {
+      "epoch": 0.26058282208588956,
+      "grad_norm": 0.28943106532096863,
+      "learning_rate": 0.00012750777469391973,
+      "loss": 1.2372,
+      "step": 1699
+    },
+    {
+      "epoch": 0.2607361963190184,
+      "grad_norm": 0.5560736060142517,
+      "learning_rate": 0.0001274817046499772,
+      "loss": 1.2105,
+      "step": 1700
+    },
+    {
+      "epoch": 0.26088957055214723,
+      "grad_norm": 0.2501446008682251,
+      "learning_rate": 0.00012745562217477395,
+      "loss": 0.9993,
+      "step": 1701
+    },
+    {
+      "epoch": 0.26104294478527607,
+      "grad_norm": 0.2808799147605896,
+      "learning_rate": 0.00012742952727448802,
+      "loss": 1.2523,
+      "step": 1702
+    },
+    {
+      "epoch": 0.2611963190184049,
+      "grad_norm": 0.21843339502811432,
+      "learning_rate": 0.0001274034199553005,
+      "loss": 1.1569,
+      "step": 1703
+    },
+    {
+      "epoch": 0.26134969325153373,
+      "grad_norm": 0.2325458824634552,
+      "learning_rate": 0.0001273773002233954,
+      "loss": 1.1998,
+      "step": 1704
+    },
+    {
+      "epoch": 0.26150306748466257,
+      "grad_norm": 0.23753198981285095,
+      "learning_rate": 0.00012735116808495965,
+      "loss": 1.1851,
+      "step": 1705
+    },
+    {
+      "epoch": 0.2616564417177914,
+      "grad_norm": 0.30201247334480286,
+      "learning_rate": 0.00012732502354618317,
+      "loss": 1.166,
+      "step": 1706
+    },
+    {
+      "epoch": 0.26180981595092023,
+      "grad_norm": 0.240064337849617,
+      "learning_rate": 0.0001272988666132587,
+      "loss": 1.0564,
+      "step": 1707
+    },
+    {
+      "epoch": 0.26196319018404907,
+      "grad_norm": 0.2625720798969269,
+      "learning_rate": 0.00012727269729238203,
+      "loss": 1.0722,
+      "step": 1708
+    },
+    {
+      "epoch": 0.2621165644171779,
+      "grad_norm": 0.22432206571102142,
+      "learning_rate": 0.00012724651558975185,
+      "loss": 1.1209,
+      "step": 1709
+    },
+    {
+      "epoch": 0.26226993865030673,
+      "grad_norm": 0.24269315600395203,
+      "learning_rate": 0.0001272203215115698,
+      "loss": 1.1437,
+      "step": 1710
+    },
+    {
+      "epoch": 0.26242331288343557,
+      "grad_norm": 0.23188909888267517,
+      "learning_rate": 0.0001271941150640404,
+      "loss": 1.1045,
+      "step": 1711
+    },
+    {
+      "epoch": 0.2625766871165644,
+      "grad_norm": 0.5129312872886658,
+      "learning_rate": 0.00012716789625337113,
+      "loss": 1.1127,
+      "step": 1712
+    },
+    {
+      "epoch": 0.26273006134969323,
+      "grad_norm": 0.3686440885066986,
+      "learning_rate": 0.0001271416650857724,
+      "loss": 1.1245,
+      "step": 1713
+    },
+    {
+      "epoch": 0.26288343558282207,
+      "grad_norm": 0.27216291427612305,
+      "learning_rate": 0.00012711542156745758,
+      "loss": 1.1647,
+      "step": 1714
+    },
+    {
+      "epoch": 0.2630368098159509,
+      "grad_norm": 0.3804262578487396,
+      "learning_rate": 0.00012708916570464292,
+      "loss": 1.0962,
+      "step": 1715
+    },
+    {
+      "epoch": 0.26319018404907973,
+      "grad_norm": 0.23700635135173798,
+      "learning_rate": 0.00012706289750354758,
+      "loss": 1.0641,
+      "step": 1716
+    },
+    {
+      "epoch": 0.26334355828220857,
+      "grad_norm": 0.27520328760147095,
+      "learning_rate": 0.00012703661697039366,
+      "loss": 1.2409,
+      "step": 1717
+    },
+    {
+      "epoch": 0.2634969325153374,
+      "grad_norm": 0.27076905965805054,
+      "learning_rate": 0.00012701032411140624,
+      "loss": 1.2397,
+      "step": 1718
+    },
+    {
+      "epoch": 0.26365030674846623,
+      "grad_norm": 0.2264924794435501,
+      "learning_rate": 0.00012698401893281326,
+      "loss": 1.1398,
+      "step": 1719
+    },
+    {
+      "epoch": 0.26380368098159507,
+      "grad_norm": 0.2567868232727051,
+      "learning_rate": 0.00012695770144084554,
+      "loss": 1.2208,
+      "step": 1720
+    },
+    {
+      "epoch": 0.2639570552147239,
+      "grad_norm": 0.2233199030160904,
+      "learning_rate": 0.00012693137164173688,
+      "loss": 1.2547,
+      "step": 1721
+    },
+    {
+      "epoch": 0.26411042944785273,
+      "grad_norm": 0.33531323075294495,
+      "learning_rate": 0.00012690502954172404,
+      "loss": 1.1577,
+      "step": 1722
+    },
+    {
+      "epoch": 0.26426380368098157,
+      "grad_norm": 0.26825258135795593,
+      "learning_rate": 0.00012687867514704653,
+      "loss": 1.0822,
+      "step": 1723
+    },
+    {
+      "epoch": 0.2644171779141104,
+      "grad_norm": 0.27282729744911194,
+      "learning_rate": 0.00012685230846394698,
+      "loss": 1.1196,
+      "step": 1724
+    },
+    {
+      "epoch": 0.2645705521472393,
+      "grad_norm": 0.24176284670829773,
+      "learning_rate": 0.00012682592949867075,
+      "loss": 1.1136,
+      "step": 1725
+    },
+    {
+      "epoch": 0.2647239263803681,
+      "grad_norm": 0.2544703483581543,
+      "learning_rate": 0.00012679953825746622,
+      "loss": 1.1217,
+      "step": 1726
+    },
+    {
+      "epoch": 0.26487730061349696,
+      "grad_norm": 0.629332959651947,
+      "learning_rate": 0.0001267731347465846,
+      "loss": 1.2309,
+      "step": 1727
+    },
+    {
+      "epoch": 0.2650306748466258,
+      "grad_norm": 0.275493860244751,
+      "learning_rate": 0.0001267467189722801,
+      "loss": 1.1885,
+      "step": 1728
+    },
+    {
+      "epoch": 0.2651840490797546,
+      "grad_norm": 0.358589231967926,
+      "learning_rate": 0.00012672029094080978,
+      "loss": 1.2346,
+      "step": 1729
+    },
+    {
+      "epoch": 0.26533742331288346,
+      "grad_norm": 0.6857182383537292,
+      "learning_rate": 0.0001266938506584336,
+      "loss": 1.1533,
+      "step": 1730
+    },
+    {
+      "epoch": 0.2654907975460123,
+      "grad_norm": 0.34166422486305237,
+      "learning_rate": 0.0001266673981314144,
+      "loss": 1.1481,
+      "step": 1731
+    },
+    {
+      "epoch": 0.2656441717791411,
+      "grad_norm": 0.4112060070037842,
+      "learning_rate": 0.00012664093336601796,
+      "loss": 1.1367,
+      "step": 1732
+    },
+    {
+      "epoch": 0.26579754601226996,
+      "grad_norm": 0.7333977818489075,
+      "learning_rate": 0.000126614456368513,
+      "loss": 1.2046,
+      "step": 1733
+    },
+    {
+      "epoch": 0.2659509202453988,
+      "grad_norm": 0.38565728068351746,
+      "learning_rate": 0.00012658796714517102,
+      "loss": 1.106,
+      "step": 1734
+    },
+    {
+      "epoch": 0.2661042944785276,
+      "grad_norm": 0.35075485706329346,
+      "learning_rate": 0.00012656146570226651,
+      "loss": 1.2096,
+      "step": 1735
+    },
+    {
+      "epoch": 0.26625766871165646,
+      "grad_norm": 0.6002593636512756,
+      "learning_rate": 0.00012653495204607685,
+      "loss": 1.3417,
+      "step": 1736
+    },
+    {
+      "epoch": 0.2664110429447853,
+      "grad_norm": 0.4263046681880951,
+      "learning_rate": 0.00012650842618288223,
+      "loss": 1.158,
+      "step": 1737
+    },
+    {
+      "epoch": 0.2665644171779141,
+      "grad_norm": 0.6455430388450623,
+      "learning_rate": 0.00012648188811896583,
+      "loss": 1.1279,
+      "step": 1738
+    },
+    {
+      "epoch": 0.26671779141104296,
+      "grad_norm": 0.557421863079071,
+      "learning_rate": 0.00012645533786061368,
+      "loss": 1.1416,
+      "step": 1739
+    },
+    {
+      "epoch": 0.2668711656441718,
+      "grad_norm": 0.7362133860588074,
+      "learning_rate": 0.0001264287754141147,
+      "loss": 1.1648,
+      "step": 1740
+    },
+    {
+      "epoch": 0.2670245398773006,
+      "grad_norm": 0.4018508791923523,
+      "learning_rate": 0.00012640220078576064,
+      "loss": 1.2045,
+      "step": 1741
+    },
+    {
+      "epoch": 0.26717791411042946,
+      "grad_norm": 0.5262569189071655,
+      "learning_rate": 0.0001263756139818463,
+      "loss": 1.2193,
+      "step": 1742
+    },
+    {
+      "epoch": 0.2673312883435583,
+      "grad_norm": 0.5104323029518127,
+      "learning_rate": 0.00012634901500866919,
+      "loss": 1.196,
+      "step": 1743
+    },
+    {
+      "epoch": 0.2674846625766871,
+      "grad_norm": 0.2668098509311676,
+      "learning_rate": 0.00012632240387252974,
+      "loss": 1.2029,
+      "step": 1744
+    },
+    {
+      "epoch": 0.26763803680981596,
+      "grad_norm": 0.3146851658821106,
+      "learning_rate": 0.00012629578057973134,
+      "loss": 1.1464,
+      "step": 1745
+    },
+    {
+      "epoch": 0.2677914110429448,
+      "grad_norm": 0.3695160448551178,
+      "learning_rate": 0.0001262691451365802,
+      "loss": 1.0813,
+      "step": 1746
+    },
+    {
+      "epoch": 0.2679447852760736,
+      "grad_norm": 0.3052177429199219,
+      "learning_rate": 0.0001262424975493854,
+      "loss": 1.1196,
+      "step": 1747
+    },
+    {
+      "epoch": 0.26809815950920246,
+      "grad_norm": 0.26070263981819153,
+      "learning_rate": 0.0001262158378244589,
+      "loss": 1.1139,
+      "step": 1748
+    },
+    {
+      "epoch": 0.2682515337423313,
+      "grad_norm": 0.46527954936027527,
+      "learning_rate": 0.0001261891659681156,
+      "loss": 1.1175,
+      "step": 1749
+    },
+    {
+      "epoch": 0.2684049079754601,
+      "grad_norm": 0.2961979806423187,
+      "learning_rate": 0.00012616248198667315,
+      "loss": 1.2349,
+      "step": 1750
+    },
+    {
+      "epoch": 0.26855828220858896,
+      "grad_norm": 0.3458435535430908,
+      "learning_rate": 0.00012613578588645223,
+      "loss": 1.159,
+      "step": 1751
+    },
+    {
+      "epoch": 0.2687116564417178,
+      "grad_norm": 0.45342785120010376,
+      "learning_rate": 0.00012610907767377622,
+      "loss": 1.0924,
+      "step": 1752
+    },
+    {
+      "epoch": 0.26886503067484663,
+      "grad_norm": 0.2586349546909332,
+      "learning_rate": 0.0001260823573549715,
+      "loss": 1.1619,
+      "step": 1753
+    },
+    {
+      "epoch": 0.26901840490797546,
+      "grad_norm": 0.33087363839149475,
+      "learning_rate": 0.00012605562493636725,
+      "loss": 1.0953,
+      "step": 1754
+    },
+    {
+      "epoch": 0.2691717791411043,
+      "grad_norm": 0.2637923061847687,
+      "learning_rate": 0.0001260288804242956,
+      "loss": 1.1439,
+      "step": 1755
+    },
+    {
+      "epoch": 0.26932515337423313,
+      "grad_norm": 0.2820630371570587,
+      "learning_rate": 0.00012600212382509135,
+      "loss": 1.2832,
+      "step": 1756
+    },
+    {
+      "epoch": 0.26947852760736196,
+      "grad_norm": 0.33386531472206116,
+      "learning_rate": 0.00012597535514509244,
+      "loss": 1.2436,
+      "step": 1757
+    },
+    {
+      "epoch": 0.2696319018404908,
+      "grad_norm": 0.25223755836486816,
+      "learning_rate": 0.00012594857439063946,
+      "loss": 1.2886,
+      "step": 1758
+    },
+    {
+      "epoch": 0.26978527607361963,
+      "grad_norm": 0.31392180919647217,
+      "learning_rate": 0.00012592178156807589,
+      "loss": 1.181,
+      "step": 1759
+    },
+    {
+      "epoch": 0.26993865030674846,
+      "grad_norm": 0.27364081144332886,
+      "learning_rate": 0.00012589497668374816,
+      "loss": 1.2832,
+      "step": 1760
+    },
+    {
+      "epoch": 0.2700920245398773,
+      "grad_norm": 0.3053947389125824,
+      "learning_rate": 0.0001258681597440055,
+      "loss": 1.2305,
+      "step": 1761
+    },
+    {
+      "epoch": 0.27024539877300613,
+      "grad_norm": 1.3622281551361084,
+      "learning_rate": 0.00012584133075519997,
+      "loss": 1.2534,
+      "step": 1762
+    },
+    {
+      "epoch": 0.27039877300613496,
+      "grad_norm": 0.34615224599838257,
+      "learning_rate": 0.0001258144897236865,
+      "loss": 1.1349,
+      "step": 1763
+    },
+    {
+      "epoch": 0.2705521472392638,
+      "grad_norm": 0.3894805610179901,
+      "learning_rate": 0.00012578763665582293,
+      "loss": 1.1989,
+      "step": 1764
+    },
+    {
+      "epoch": 0.27070552147239263,
+      "grad_norm": 1.4577516317367554,
+      "learning_rate": 0.00012576077155796987,
+      "loss": 1.0441,
+      "step": 1765
+    },
+    {
+      "epoch": 0.27085889570552146,
+      "grad_norm": 0.2927125096321106,
+      "learning_rate": 0.00012573389443649085,
+      "loss": 1.2088,
+      "step": 1766
+    },
+    {
+      "epoch": 0.2710122699386503,
+      "grad_norm": 0.33564674854278564,
+      "learning_rate": 0.00012570700529775214,
+      "loss": 1.1487,
+      "step": 1767
+    },
+    {
+      "epoch": 0.27116564417177913,
+      "grad_norm": 0.4361121356487274,
+      "learning_rate": 0.00012568010414812302,
+      "loss": 1.3016,
+      "step": 1768
+    },
+    {
+      "epoch": 0.27131901840490796,
+      "grad_norm": 0.32853755354881287,
+      "learning_rate": 0.00012565319099397545,
+      "loss": 1.1567,
+      "step": 1769
+    },
+    {
+      "epoch": 0.2714723926380368,
+      "grad_norm": 0.36490413546562195,
+      "learning_rate": 0.00012562626584168438,
+      "loss": 1.189,
+      "step": 1770
+    },
+    {
+      "epoch": 0.27162576687116563,
+      "grad_norm": 1.008055329322815,
+      "learning_rate": 0.00012559932869762747,
+      "loss": 1.1768,
+      "step": 1771
+    },
+    {
+      "epoch": 0.27177914110429446,
+      "grad_norm": 0.4403283894062042,
+      "learning_rate": 0.00012557237956818527,
+      "loss": 1.3133,
+      "step": 1772
+    },
+    {
+      "epoch": 0.2719325153374233,
+      "grad_norm": 0.35770583152770996,
+      "learning_rate": 0.00012554541845974124,
+      "loss": 1.1145,
+      "step": 1773
+    },
+    {
+      "epoch": 0.27208588957055213,
+      "grad_norm": 0.35985687375068665,
+      "learning_rate": 0.00012551844537868158,
+      "loss": 1.1072,
+      "step": 1774
+    },
+    {
+      "epoch": 0.27223926380368096,
+      "grad_norm": 0.4859484136104584,
+      "learning_rate": 0.00012549146033139533,
+      "loss": 1.1982,
+      "step": 1775
+    },
+    {
+      "epoch": 0.2723926380368098,
+      "grad_norm": 0.3751656115055084,
+      "learning_rate": 0.00012546446332427447,
+      "loss": 1.073,
+      "step": 1776
+    },
+    {
+      "epoch": 0.27254601226993863,
+      "grad_norm": 0.42645615339279175,
+      "learning_rate": 0.0001254374543637137,
+      "loss": 1.3011,
+      "step": 1777
+    },
+    {
+      "epoch": 0.27269938650306746,
+      "grad_norm": 0.2741505801677704,
+      "learning_rate": 0.00012541043345611056,
+      "loss": 1.1538,
+      "step": 1778
+    },
+    {
+      "epoch": 0.2728527607361963,
+      "grad_norm": 0.281800240278244,
+      "learning_rate": 0.00012538340060786553,
+      "loss": 1.2188,
+      "step": 1779
+    },
+    {
+      "epoch": 0.27300613496932513,
+      "grad_norm": 0.31412461400032043,
+      "learning_rate": 0.00012535635582538177,
+      "loss": 1.1067,
+      "step": 1780
+    },
+    {
+      "epoch": 0.27315950920245397,
+      "grad_norm": 0.32550278306007385,
+      "learning_rate": 0.00012532929911506535,
+      "loss": 1.1193,
+      "step": 1781
+    },
+    {
+      "epoch": 0.2733128834355828,
+      "grad_norm": 0.26710861921310425,
+      "learning_rate": 0.00012530223048332518,
+      "loss": 1.2483,
+      "step": 1782
+    },
+    {
+      "epoch": 0.27346625766871163,
+      "grad_norm": 0.4228115975856781,
+      "learning_rate": 0.00012527514993657295,
+      "loss": 1.1246,
+      "step": 1783
+    },
+    {
+      "epoch": 0.27361963190184047,
+      "grad_norm": 0.45894473791122437,
+      "learning_rate": 0.00012524805748122315,
+      "loss": 1.1547,
+      "step": 1784
+    },
+    {
+      "epoch": 0.2737730061349693,
+      "grad_norm": 0.9309108257293701,
+      "learning_rate": 0.00012522095312369317,
+      "loss": 1.2007,
+      "step": 1785
+    },
+    {
+      "epoch": 0.27392638036809813,
+      "grad_norm": 0.29931697249412537,
+      "learning_rate": 0.0001251938368704032,
+      "loss": 1.1235,
+      "step": 1786
+    },
+    {
+      "epoch": 0.274079754601227,
+      "grad_norm": 0.35071128606796265,
+      "learning_rate": 0.00012516670872777616,
+      "loss": 1.3176,
+      "step": 1787
+    },
+    {
+      "epoch": 0.27423312883435585,
+      "grad_norm": 0.6034968495368958,
+      "learning_rate": 0.00012513956870223789,
+      "loss": 1.3146,
+      "step": 1788
+    },
+    {
+      "epoch": 0.2743865030674847,
+      "grad_norm": 0.27235767245292664,
+      "learning_rate": 0.000125112416800217,
+      "loss": 1.1766,
+      "step": 1789
+    },
+    {
+      "epoch": 0.2745398773006135,
+      "grad_norm": 0.27549147605895996,
+      "learning_rate": 0.00012508525302814494,
+      "loss": 0.9949,
+      "step": 1790
+    },
+    {
+      "epoch": 0.27469325153374236,
+      "grad_norm": 0.9262846112251282,
+      "learning_rate": 0.00012505807739245594,
+      "loss": 1.1585,
+      "step": 1791
+    },
+    {
+      "epoch": 0.2748466257668712,
+      "grad_norm": 0.47159233689308167,
+      "learning_rate": 0.00012503088989958703,
+      "loss": 1.024,
+      "step": 1792
+    },
+    {
+      "epoch": 0.275,
+      "grad_norm": 0.36469513177871704,
+      "learning_rate": 0.00012500369055597806,
+      "loss": 1.3261,
+      "step": 1793
+    },
+    {
+      "epoch": 0.27515337423312886,
+      "grad_norm": 0.31499096751213074,
+      "learning_rate": 0.00012497647936807177,
+      "loss": 1.1726,
+      "step": 1794
+    },
+    {
+      "epoch": 0.2753067484662577,
+      "grad_norm": 0.33638596534729004,
+      "learning_rate": 0.00012494925634231352,
+      "loss": 1.16,
+      "step": 1795
+    },
+    {
+      "epoch": 0.2754601226993865,
+      "grad_norm": 0.5230518579483032,
+      "learning_rate": 0.00012492202148515168,
+      "loss": 1.171,
+      "step": 1796
+    },
+    {
+      "epoch": 0.27561349693251536,
+      "grad_norm": 0.25849413871765137,
+      "learning_rate": 0.0001248947748030373,
+      "loss": 1.1703,
+      "step": 1797
+    },
+    {
+      "epoch": 0.2757668711656442,
+      "grad_norm": 0.2648497521877289,
+      "learning_rate": 0.00012486751630242426,
+      "loss": 1.1261,
+      "step": 1798
+    },
+    {
+      "epoch": 0.275920245398773,
+      "grad_norm": 0.33965373039245605,
+      "learning_rate": 0.00012484024598976922,
+      "loss": 0.9817,
+      "step": 1799
+    },
+    {
+      "epoch": 0.27607361963190186,
+      "grad_norm": 0.3139823079109192,
+      "learning_rate": 0.00012481296387153166,
+      "loss": 1.07,
+      "step": 1800
+    },
+    {
+      "epoch": 0.2762269938650307,
+      "grad_norm": 0.2756589353084564,
+      "learning_rate": 0.0001247856699541739,
+      "loss": 1.0735,
+      "step": 1801
+    },
+    {
+      "epoch": 0.2763803680981595,
+      "grad_norm": 0.2767266035079956,
+      "learning_rate": 0.00012475836424416097,
+      "loss": 1.2717,
+      "step": 1802
+    },
+    {
+      "epoch": 0.27653374233128836,
+      "grad_norm": 1.0867480039596558,
+      "learning_rate": 0.00012473104674796072,
+      "loss": 1.2385,
+      "step": 1803
+    },
+    {
+      "epoch": 0.2766871165644172,
+      "grad_norm": 0.2601228356361389,
+      "learning_rate": 0.00012470371747204384,
+      "loss": 1.1975,
+      "step": 1804
+    },
+    {
+      "epoch": 0.276840490797546,
+      "grad_norm": 0.29320228099823,
+      "learning_rate": 0.00012467637642288375,
+      "loss": 0.9783,
+      "step": 1805
+    },
+    {
+      "epoch": 0.27699386503067486,
+      "grad_norm": 0.2687963843345642,
+      "learning_rate": 0.00012464902360695667,
+      "loss": 1.2444,
+      "step": 1806
+    },
+    {
+      "epoch": 0.2771472392638037,
+      "grad_norm": 0.8859814405441284,
+      "learning_rate": 0.00012462165903074163,
+      "loss": 1.1625,
+      "step": 1807
+    },
+    {
+      "epoch": 0.2773006134969325,
+      "grad_norm": 0.377537339925766,
+      "learning_rate": 0.00012459428270072045,
+      "loss": 1.0382,
+      "step": 1808
+    },
+    {
+      "epoch": 0.27745398773006136,
+      "grad_norm": 0.31914326548576355,
+      "learning_rate": 0.0001245668946233777,
+      "loss": 1.2286,
+      "step": 1809
+    },
+    {
+      "epoch": 0.2776073619631902,
+      "grad_norm": 0.4909219741821289,
+      "learning_rate": 0.0001245394948052008,
+      "loss": 1.1942,
+      "step": 1810
+    },
+    {
+      "epoch": 0.277760736196319,
+      "grad_norm": 0.8578031659126282,
+      "learning_rate": 0.00012451208325267983,
+      "loss": 1.0689,
+      "step": 1811
+    },
+    {
+      "epoch": 0.27791411042944786,
+      "grad_norm": 0.35371869802474976,
+      "learning_rate": 0.00012448465997230777,
+      "loss": 1.1361,
+      "step": 1812
+    },
+    {
+      "epoch": 0.2780674846625767,
+      "grad_norm": 0.49104273319244385,
+      "learning_rate": 0.0001244572249705803,
+      "loss": 1.2209,
+      "step": 1813
+    },
+    {
+      "epoch": 0.2782208588957055,
+      "grad_norm": 0.4191496968269348,
+      "learning_rate": 0.00012442977825399593,
+      "loss": 1.2952,
+      "step": 1814
+    },
+    {
+      "epoch": 0.27837423312883436,
+      "grad_norm": 0.35341188311576843,
+      "learning_rate": 0.00012440231982905592,
+      "loss": 1.1338,
+      "step": 1815
+    },
+    {
+      "epoch": 0.2785276073619632,
+      "grad_norm": 0.27151957154273987,
+      "learning_rate": 0.0001243748497022643,
+      "loss": 1.1465,
+      "step": 1816
+    },
+    {
+      "epoch": 0.278680981595092,
+      "grad_norm": 0.2607406973838806,
+      "learning_rate": 0.00012434736788012786,
+      "loss": 1.3473,
+      "step": 1817
+    },
+    {
+      "epoch": 0.27883435582822086,
+      "grad_norm": 0.26552867889404297,
+      "learning_rate": 0.00012431987436915622,
+      "loss": 1.2787,
+      "step": 1818
+    },
+    {
+      "epoch": 0.2789877300613497,
+      "grad_norm": 0.34142789244651794,
+      "learning_rate": 0.0001242923691758617,
+      "loss": 1.2142,
+      "step": 1819
+    },
+    {
+      "epoch": 0.2791411042944785,
+      "grad_norm": 0.5956395864486694,
+      "learning_rate": 0.0001242648523067594,
+      "loss": 1.1795,
+      "step": 1820
+    },
+    {
+      "epoch": 0.27929447852760736,
+      "grad_norm": 0.30675971508026123,
+      "learning_rate": 0.0001242373237683672,
+      "loss": 1.2895,
+      "step": 1821
+    },
+    {
+      "epoch": 0.2794478527607362,
+      "grad_norm": 0.3154914081096649,
+      "learning_rate": 0.00012420978356720574,
+      "loss": 1.1518,
+      "step": 1822
+    },
+    {
+      "epoch": 0.279601226993865,
+      "grad_norm": 0.2820705771446228,
+      "learning_rate": 0.00012418223170979848,
+      "loss": 1.2031,
+      "step": 1823
+    },
+    {
+      "epoch": 0.27975460122699386,
+      "grad_norm": 0.2699308395385742,
+      "learning_rate": 0.00012415466820267152,
+      "loss": 1.2484,
+      "step": 1824
+    },
+    {
+      "epoch": 0.2799079754601227,
+      "grad_norm": 0.2572730481624603,
+      "learning_rate": 0.00012412709305235383,
+      "loss": 1.1528,
+      "step": 1825
+    },
+    {
+      "epoch": 0.2800613496932515,
+      "grad_norm": 0.3126010596752167,
+      "learning_rate": 0.00012409950626537705,
+      "loss": 1.2159,
+      "step": 1826
+    },
+    {
+      "epoch": 0.28021472392638036,
+      "grad_norm": 0.5773135423660278,
+      "learning_rate": 0.00012407190784827565,
+      "loss": 1.0507,
+      "step": 1827
+    },
+    {
+      "epoch": 0.2803680981595092,
+      "grad_norm": 0.3662653863430023,
+      "learning_rate": 0.00012404429780758683,
+      "loss": 1.1938,
+      "step": 1828
+    },
+    {
+      "epoch": 0.280521472392638,
+      "grad_norm": 0.32324546575546265,
+      "learning_rate": 0.00012401667614985053,
+      "loss": 1.1721,
+      "step": 1829
+    },
+    {
+      "epoch": 0.28067484662576686,
+      "grad_norm": 0.32635340094566345,
+      "learning_rate": 0.00012398904288160942,
+      "loss": 1.2459,
+      "step": 1830
+    },
+    {
+      "epoch": 0.2808282208588957,
+      "grad_norm": 0.4830225110054016,
+      "learning_rate": 0.000123961398009409,
+      "loss": 0.9969,
+      "step": 1831
+    },
+    {
+      "epoch": 0.2809815950920245,
+      "grad_norm": 0.3115139603614807,
+      "learning_rate": 0.00012393374153979743,
+      "loss": 1.2145,
+      "step": 1832
+    },
+    {
+      "epoch": 0.28113496932515336,
+      "grad_norm": 0.3283381760120392,
+      "learning_rate": 0.00012390607347932567,
+      "loss": 0.98,
+      "step": 1833
+    },
+    {
+      "epoch": 0.2812883435582822,
+      "grad_norm": 0.34774792194366455,
+      "learning_rate": 0.00012387839383454742,
+      "loss": 1.1315,
+      "step": 1834
+    },
+    {
+      "epoch": 0.28144171779141103,
+      "grad_norm": 0.3749578893184662,
+      "learning_rate": 0.0001238507026120191,
+      "loss": 1.2148,
+      "step": 1835
+    },
+    {
+      "epoch": 0.28159509202453986,
+      "grad_norm": 0.3452282249927521,
+      "learning_rate": 0.0001238229998182999,
+      "loss": 1.2336,
+      "step": 1836
+    },
+    {
+      "epoch": 0.2817484662576687,
+      "grad_norm": 0.27290740609169006,
+      "learning_rate": 0.00012379528545995171,
+      "loss": 1.1656,
+      "step": 1837
+    },
+    {
+      "epoch": 0.28190184049079753,
+      "grad_norm": 1.1912314891815186,
+      "learning_rate": 0.00012376755954353924,
+      "loss": 1.2628,
+      "step": 1838
+    },
+    {
+      "epoch": 0.28205521472392636,
+      "grad_norm": 0.5588482022285461,
+      "learning_rate": 0.0001237398220756298,
+      "loss": 1.0995,
+      "step": 1839
+    },
+    {
+      "epoch": 0.2822085889570552,
+      "grad_norm": 0.7528854012489319,
+      "learning_rate": 0.0001237120730627936,
+      "loss": 1.0785,
+      "step": 1840
+    },
+    {
+      "epoch": 0.28236196319018403,
+      "grad_norm": 0.3596801459789276,
+      "learning_rate": 0.0001236843125116035,
+      "loss": 1.0764,
+      "step": 1841
+    },
+    {
+      "epoch": 0.28251533742331286,
+      "grad_norm": 0.2877620458602905,
+      "learning_rate": 0.00012365654042863507,
+      "loss": 1.2853,
+      "step": 1842
+    },
+    {
+      "epoch": 0.2826687116564417,
+      "grad_norm": 0.40135982632637024,
+      "learning_rate": 0.00012362875682046668,
+      "loss": 1.1589,
+      "step": 1843
+    },
+    {
+      "epoch": 0.28282208588957053,
+      "grad_norm": 0.2544049918651581,
+      "learning_rate": 0.00012360096169367933,
+      "loss": 1.2857,
+      "step": 1844
+    },
+    {
+      "epoch": 0.28297546012269936,
+      "grad_norm": 0.4005971848964691,
+      "learning_rate": 0.00012357315505485684,
+      "loss": 1.1052,
+      "step": 1845
+    },
+    {
+      "epoch": 0.2831288343558282,
+      "grad_norm": 0.27655598521232605,
+      "learning_rate": 0.00012354533691058574,
+      "loss": 1.2972,
+      "step": 1846
+    },
+    {
+      "epoch": 0.28328220858895703,
+      "grad_norm": 0.4048658013343811,
+      "learning_rate": 0.00012351750726745522,
+      "loss": 1.0443,
+      "step": 1847
+    },
+    {
+      "epoch": 0.28343558282208586,
+      "grad_norm": 0.261182963848114,
+      "learning_rate": 0.00012348966613205732,
+      "loss": 1.1491,
+      "step": 1848
+    },
+    {
+      "epoch": 0.2835889570552147,
+      "grad_norm": 0.2675437033176422,
+      "learning_rate": 0.00012346181351098669,
+      "loss": 1.1461,
+      "step": 1849
+    },
+    {
+      "epoch": 0.2837423312883436,
+      "grad_norm": 0.30994632840156555,
+      "learning_rate": 0.0001234339494108407,
+      "loss": 1.2354,
+      "step": 1850
+    },
+    {
+      "epoch": 0.2838957055214724,
+      "grad_norm": 0.23087720572948456,
+      "learning_rate": 0.00012340607383821956,
+      "loss": 1.2197,
+      "step": 1851
+    },
+    {
+      "epoch": 0.28404907975460125,
+      "grad_norm": 0.2573588490486145,
+      "learning_rate": 0.00012337818679972605,
+      "loss": 1.2168,
+      "step": 1852
+    },
+    {
+      "epoch": 0.2842024539877301,
+      "grad_norm": 0.26689615845680237,
+      "learning_rate": 0.00012335028830196573,
+      "loss": 1.2338,
+      "step": 1853
+    },
+    {
+      "epoch": 0.2843558282208589,
+      "grad_norm": 0.2517935633659363,
+      "learning_rate": 0.0001233223783515469,
+      "loss": 1.1624,
+      "step": 1854
+    },
+    {
+      "epoch": 0.28450920245398775,
+      "grad_norm": 0.2692297697067261,
+      "learning_rate": 0.00012329445695508057,
+      "loss": 1.0912,
+      "step": 1855
+    },
+    {
+      "epoch": 0.2846625766871166,
+      "grad_norm": 0.24611712992191315,
+      "learning_rate": 0.00012326652411918037,
+      "loss": 1.1313,
+      "step": 1856
+    },
+    {
+      "epoch": 0.2848159509202454,
+      "grad_norm": 0.3156646490097046,
+      "learning_rate": 0.00012323857985046277,
+      "loss": 1.3675,
+      "step": 1857
+    },
+    {
+      "epoch": 0.28496932515337425,
+      "grad_norm": 0.21803468465805054,
+      "learning_rate": 0.00012321062415554683,
+      "loss": 1.2124,
+      "step": 1858
+    },
+    {
+      "epoch": 0.2851226993865031,
+      "grad_norm": 0.22948789596557617,
+      "learning_rate": 0.00012318265704105443,
+      "loss": 1.1742,
+      "step": 1859
+    },
+    {
+      "epoch": 0.2852760736196319,
+      "grad_norm": 0.32487693428993225,
+      "learning_rate": 0.00012315467851361006,
+      "loss": 1.236,
+      "step": 1860
+    },
+    {
+      "epoch": 0.28542944785276075,
+      "grad_norm": 0.6182332634925842,
+      "learning_rate": 0.00012312668857984095,
+      "loss": 1.2134,
+      "step": 1861
+    },
+    {
+      "epoch": 0.2855828220858896,
+      "grad_norm": 0.7873092889785767,
+      "learning_rate": 0.00012309868724637705,
+      "loss": 1.238,
+      "step": 1862
+    },
+    {
+      "epoch": 0.2857361963190184,
+      "grad_norm": 0.23776070773601532,
+      "learning_rate": 0.00012307067451985102,
+      "loss": 1.2245,
+      "step": 1863
+    },
+    {
+      "epoch": 0.28588957055214725,
+      "grad_norm": 0.6169922351837158,
+      "learning_rate": 0.00012304265040689812,
+      "loss": 1.1486,
+      "step": 1864
+    },
+    {
+      "epoch": 0.2860429447852761,
+      "grad_norm": 0.4745227098464966,
+      "learning_rate": 0.00012301461491415644,
+      "loss": 1.2083,
+      "step": 1865
+    },
+    {
+      "epoch": 0.2861963190184049,
+      "grad_norm": 0.2495294064283371,
+      "learning_rate": 0.00012298656804826666,
+      "loss": 1.0265,
+      "step": 1866
+    },
+    {
+      "epoch": 0.28634969325153375,
+      "grad_norm": 0.6950542330741882,
+      "learning_rate": 0.00012295850981587224,
+      "loss": 1.0784,
+      "step": 1867
+    },
+    {
+      "epoch": 0.2865030674846626,
+      "grad_norm": 0.3061631917953491,
+      "learning_rate": 0.00012293044022361927,
+      "loss": 1.1581,
+      "step": 1868
+    },
+    {
+      "epoch": 0.2866564417177914,
+      "grad_norm": 0.4083435833454132,
+      "learning_rate": 0.00012290235927815655,
+      "loss": 1.0888,
+      "step": 1869
+    },
+    {
+      "epoch": 0.28680981595092025,
+      "grad_norm": 0.4837428629398346,
+      "learning_rate": 0.00012287426698613554,
+      "loss": 1.129,
+      "step": 1870
+    },
+    {
+      "epoch": 0.2869631901840491,
+      "grad_norm": 0.5466214418411255,
+      "learning_rate": 0.00012284616335421047,
+      "loss": 1.1787,
+      "step": 1871
+    },
+    {
+      "epoch": 0.2871165644171779,
+      "grad_norm": 0.38480639457702637,
+      "learning_rate": 0.0001228180483890382,
+      "loss": 1.2003,
+      "step": 1872
+    },
+    {
+      "epoch": 0.28726993865030676,
+      "grad_norm": 0.38878071308135986,
+      "learning_rate": 0.00012278992209727825,
+      "loss": 1.2383,
+      "step": 1873
+    },
+    {
+      "epoch": 0.2874233128834356,
+      "grad_norm": 0.2481946349143982,
+      "learning_rate": 0.00012276178448559287,
+      "loss": 1.2758,
+      "step": 1874
+    },
+    {
+      "epoch": 0.2875766871165644,
+      "grad_norm": 0.2779454290866852,
+      "learning_rate": 0.00012273363556064695,
+      "loss": 1.2941,
+      "step": 1875
+    },
+    {
+      "epoch": 0.28773006134969326,
+      "grad_norm": 0.3368411958217621,
+      "learning_rate": 0.00012270547532910813,
+      "loss": 1.1099,
+      "step": 1876
+    },
+    {
+      "epoch": 0.2878834355828221,
+      "grad_norm": 0.5561997890472412,
+      "learning_rate": 0.00012267730379764662,
+      "loss": 1.0678,
+      "step": 1877
+    },
+    {
+      "epoch": 0.2880368098159509,
+      "grad_norm": 0.2926631569862366,
+      "learning_rate": 0.0001226491209729354,
+      "loss": 1.2304,
+      "step": 1878
+    },
+    {
+      "epoch": 0.28819018404907976,
+      "grad_norm": 0.9711184501647949,
+      "learning_rate": 0.0001226209268616501,
+      "loss": 1.2118,
+      "step": 1879
+    },
+    {
+      "epoch": 0.2883435582822086,
+      "grad_norm": 0.32794806361198425,
+      "learning_rate": 0.00012259272147046901,
+      "loss": 1.2235,
+      "step": 1880
+    },
+    {
+      "epoch": 0.2884969325153374,
+      "grad_norm": 0.3814946115016937,
+      "learning_rate": 0.00012256450480607308,
+      "loss": 1.2535,
+      "step": 1881
+    },
+    {
+      "epoch": 0.28865030674846626,
+      "grad_norm": 0.42511552572250366,
+      "learning_rate": 0.00012253627687514596,
+      "loss": 1.1321,
+      "step": 1882
+    },
+    {
+      "epoch": 0.2888036809815951,
+      "grad_norm": 0.2878671884536743,
+      "learning_rate": 0.00012250803768437397,
+      "loss": 1.1067,
+      "step": 1883
+    },
+    {
+      "epoch": 0.2889570552147239,
+      "grad_norm": 0.29115909337997437,
+      "learning_rate": 0.00012247978724044605,
+      "loss": 1.1533,
+      "step": 1884
+    },
+    {
+      "epoch": 0.28911042944785276,
+      "grad_norm": 0.29047396779060364,
+      "learning_rate": 0.0001224515255500539,
+      "loss": 1.0623,
+      "step": 1885
+    },
+    {
+      "epoch": 0.2892638036809816,
+      "grad_norm": 0.35951316356658936,
+      "learning_rate": 0.00012242325261989175,
+      "loss": 1.3013,
+      "step": 1886
+    },
+    {
+      "epoch": 0.2894171779141104,
+      "grad_norm": 0.7687309980392456,
+      "learning_rate": 0.00012239496845665662,
+      "loss": 1.1613,
+      "step": 1887
+    },
+    {
+      "epoch": 0.28957055214723926,
+      "grad_norm": 0.33891361951828003,
+      "learning_rate": 0.00012236667306704811,
+      "loss": 1.2351,
+      "step": 1888
+    },
+    {
+      "epoch": 0.2897239263803681,
+      "grad_norm": 0.5337998867034912,
+      "learning_rate": 0.0001223383664577685,
+      "loss": 1.2827,
+      "step": 1889
+    },
+    {
+      "epoch": 0.2898773006134969,
+      "grad_norm": 0.28730982542037964,
+      "learning_rate": 0.00012231004863552276,
+      "loss": 1.3112,
+      "step": 1890
+    },
+    {
+      "epoch": 0.29003067484662576,
+      "grad_norm": 0.25794607400894165,
+      "learning_rate": 0.00012228171960701847,
+      "loss": 1.3741,
+      "step": 1891
+    },
+    {
+      "epoch": 0.2901840490797546,
+      "grad_norm": 0.5016089677810669,
+      "learning_rate": 0.0001222533793789659,
+      "loss": 1.118,
+      "step": 1892
+    },
+    {
+      "epoch": 0.2903374233128834,
+      "grad_norm": 0.27106767892837524,
+      "learning_rate": 0.0001222250279580779,
+      "loss": 1.0012,
+      "step": 1893
+    },
+    {
+      "epoch": 0.29049079754601226,
+      "grad_norm": 0.2565261423587799,
+      "learning_rate": 0.0001221966653510701,
+      "loss": 1.2484,
+      "step": 1894
+    },
+    {
+      "epoch": 0.2906441717791411,
+      "grad_norm": 0.5944375395774841,
+      "learning_rate": 0.00012216829156466068,
+      "loss": 1.2169,
+      "step": 1895
+    },
+    {
+      "epoch": 0.2907975460122699,
+      "grad_norm": 0.30955949425697327,
+      "learning_rate": 0.00012213990660557051,
+      "loss": 1.2254,
+      "step": 1896
+    },
+    {
+      "epoch": 0.29095092024539876,
+      "grad_norm": 0.33194753527641296,
+      "learning_rate": 0.00012211151048052303,
+      "loss": 1.2406,
+      "step": 1897
+    },
+    {
+      "epoch": 0.2911042944785276,
+      "grad_norm": 0.3325173556804657,
+      "learning_rate": 0.00012208310319624448,
+      "loss": 1.1345,
+      "step": 1898
+    },
+    {
+      "epoch": 0.2912576687116564,
+      "grad_norm": 0.4195181727409363,
+      "learning_rate": 0.00012205468475946356,
+      "loss": 1.0114,
+      "step": 1899
+    },
+    {
+      "epoch": 0.29141104294478526,
+      "grad_norm": 0.23835013806819916,
+      "learning_rate": 0.00012202625517691176,
+      "loss": 1.182,
+      "step": 1900
+    },
+    {
+      "epoch": 0.2915644171779141,
+      "grad_norm": 0.3492606580257416,
+      "learning_rate": 0.00012199781445532313,
+      "loss": 1.0664,
+      "step": 1901
+    },
+    {
+      "epoch": 0.2917177914110429,
+      "grad_norm": 0.516598105430603,
+      "learning_rate": 0.00012196936260143439,
+      "loss": 1.274,
+      "step": 1902
+    },
+    {
+      "epoch": 0.29187116564417176,
+      "grad_norm": 0.23151113092899323,
+      "learning_rate": 0.00012194089962198487,
+      "loss": 1.1429,
+      "step": 1903
+    },
+    {
+      "epoch": 0.2920245398773006,
+      "grad_norm": 0.31203627586364746,
+      "learning_rate": 0.00012191242552371656,
+      "loss": 1.2591,
+      "step": 1904
+    },
+    {
+      "epoch": 0.2921779141104294,
+      "grad_norm": 0.2378513514995575,
+      "learning_rate": 0.0001218839403133741,
+      "loss": 1.1271,
+      "step": 1905
+    },
+    {
+      "epoch": 0.29233128834355826,
+      "grad_norm": 0.2664697766304016,
+      "learning_rate": 0.00012185544399770471,
+      "loss": 1.0674,
+      "step": 1906
+    },
+    {
+      "epoch": 0.2924846625766871,
+      "grad_norm": 0.2708868086338043,
+      "learning_rate": 0.00012182693658345827,
+      "loss": 1.1596,
+      "step": 1907
+    },
+    {
+      "epoch": 0.2926380368098159,
+      "grad_norm": 0.36455556750297546,
+      "learning_rate": 0.0001217984180773873,
+      "loss": 1.0423,
+      "step": 1908
+    },
+    {
+      "epoch": 0.29279141104294476,
+      "grad_norm": 0.2276659607887268,
+      "learning_rate": 0.00012176988848624692,
+      "loss": 1.1322,
+      "step": 1909
+    },
+    {
+      "epoch": 0.2929447852760736,
+      "grad_norm": 0.26716727018356323,
+      "learning_rate": 0.00012174134781679491,
+      "loss": 1.2832,
+      "step": 1910
+    },
+    {
+      "epoch": 0.2930981595092024,
+      "grad_norm": 0.24976113438606262,
+      "learning_rate": 0.00012171279607579163,
+      "loss": 1.1849,
+      "step": 1911
+    },
+    {
+      "epoch": 0.29325153374233126,
+      "grad_norm": 0.2293761521577835,
+      "learning_rate": 0.00012168423327000011,
+      "loss": 1.1575,
+      "step": 1912
+    },
+    {
+      "epoch": 0.29340490797546015,
+      "grad_norm": 0.23514318466186523,
+      "learning_rate": 0.00012165565940618597,
+      "loss": 1.0303,
+      "step": 1913
+    },
+    {
+      "epoch": 0.293558282208589,
+      "grad_norm": 0.2838674783706665,
+      "learning_rate": 0.00012162707449111744,
+      "loss": 1.2303,
+      "step": 1914
+    },
+    {
+      "epoch": 0.2937116564417178,
+      "grad_norm": 0.2449359893798828,
+      "learning_rate": 0.00012159847853156542,
+      "loss": 1.2085,
+      "step": 1915
+    },
+    {
+      "epoch": 0.29386503067484665,
+      "grad_norm": 0.3092651665210724,
+      "learning_rate": 0.00012156987153430335,
+      "loss": 1.166,
+      "step": 1916
+    },
+    {
+      "epoch": 0.2940184049079755,
+      "grad_norm": 0.29424700140953064,
+      "learning_rate": 0.00012154125350610738,
+      "loss": 1.1887,
+      "step": 1917
+    },
+    {
+      "epoch": 0.2941717791411043,
+      "grad_norm": 0.24386832118034363,
+      "learning_rate": 0.00012151262445375614,
+      "loss": 1.0145,
+      "step": 1918
+    },
+    {
+      "epoch": 0.29432515337423315,
+      "grad_norm": 0.22343406081199646,
+      "learning_rate": 0.00012148398438403102,
+      "loss": 1.1853,
+      "step": 1919
+    },
+    {
+      "epoch": 0.294478527607362,
+      "grad_norm": 0.3495458960533142,
+      "learning_rate": 0.00012145533330371593,
+      "loss": 1.2336,
+      "step": 1920
+    },
+    {
+      "epoch": 0.2946319018404908,
+      "grad_norm": 0.2311302274465561,
+      "learning_rate": 0.00012142667121959739,
+      "loss": 1.1073,
+      "step": 1921
+    },
+    {
+      "epoch": 0.29478527607361965,
+      "grad_norm": 0.22799399495124817,
+      "learning_rate": 0.00012139799813846456,
+      "loss": 1.098,
+      "step": 1922
+    },
+    {
+      "epoch": 0.2949386503067485,
+      "grad_norm": 0.2798739969730377,
+      "learning_rate": 0.00012136931406710916,
+      "loss": 1.1033,
+      "step": 1923
+    },
+    {
+      "epoch": 0.2950920245398773,
+      "grad_norm": 0.29123950004577637,
+      "learning_rate": 0.00012134061901232558,
+      "loss": 1.1612,
+      "step": 1924
+    },
+    {
+      "epoch": 0.29524539877300615,
+      "grad_norm": 0.22587506473064423,
+      "learning_rate": 0.00012131191298091077,
+      "loss": 1.1468,
+      "step": 1925
+    },
+    {
+      "epoch": 0.295398773006135,
+      "grad_norm": 0.22429734468460083,
+      "learning_rate": 0.00012128319597966427,
+      "loss": 1.1867,
+      "step": 1926
+    },
+    {
+      "epoch": 0.2955521472392638,
+      "grad_norm": 0.32708224654197693,
+      "learning_rate": 0.00012125446801538823,
+      "loss": 1.085,
+      "step": 1927
+    },
+    {
+      "epoch": 0.29570552147239265,
+      "grad_norm": 0.22231736779212952,
+      "learning_rate": 0.0001212257290948874,
+      "loss": 1.0899,
+      "step": 1928
+    },
+    {
+      "epoch": 0.2958588957055215,
+      "grad_norm": 0.36758795380592346,
+      "learning_rate": 0.00012119697922496915,
+      "loss": 1.1989,
+      "step": 1929
+    },
+    {
+      "epoch": 0.2960122699386503,
+      "grad_norm": 0.2168715000152588,
+      "learning_rate": 0.00012116821841244337,
+      "loss": 1.1625,
+      "step": 1930
+    },
+    {
+      "epoch": 0.29616564417177915,
+      "grad_norm": 0.21805234253406525,
+      "learning_rate": 0.00012113944666412262,
+      "loss": 1.0619,
+      "step": 1931
+    },
+    {
+      "epoch": 0.296319018404908,
+      "grad_norm": 0.2775552570819855,
+      "learning_rate": 0.00012111066398682204,
+      "loss": 1.0164,
+      "step": 1932
+    },
+    {
+      "epoch": 0.2964723926380368,
+      "grad_norm": 0.2534048557281494,
+      "learning_rate": 0.00012108187038735927,
+      "loss": 1.1576,
+      "step": 1933
+    },
+    {
+      "epoch": 0.29662576687116565,
+      "grad_norm": 0.21596451103687286,
+      "learning_rate": 0.00012105306587255468,
+      "loss": 1.1061,
+      "step": 1934
+    },
+    {
+      "epoch": 0.2967791411042945,
+      "grad_norm": 0.19572919607162476,
+      "learning_rate": 0.0001210242504492311,
+      "loss": 1.0353,
+      "step": 1935
+    },
+    {
+      "epoch": 0.2969325153374233,
+      "grad_norm": 0.3992101848125458,
+      "learning_rate": 0.00012099542412421402,
+      "loss": 1.1887,
+      "step": 1936
+    },
+    {
+      "epoch": 0.29708588957055215,
+      "grad_norm": 0.197307750582695,
+      "learning_rate": 0.0001209665869043315,
+      "loss": 1.0208,
+      "step": 1937
+    },
+    {
+      "epoch": 0.297239263803681,
+      "grad_norm": 0.2667563855648041,
+      "learning_rate": 0.00012093773879641411,
+      "loss": 1.0348,
+      "step": 1938
+    },
+    {
+      "epoch": 0.2973926380368098,
+      "grad_norm": 0.21156510710716248,
+      "learning_rate": 0.00012090887980729513,
+      "loss": 1.0911,
+      "step": 1939
+    },
+    {
+      "epoch": 0.29754601226993865,
+      "grad_norm": 0.23344705998897552,
+      "learning_rate": 0.00012088000994381031,
+      "loss": 1.067,
+      "step": 1940
+    },
+    {
+      "epoch": 0.2976993865030675,
+      "grad_norm": 0.22214551270008087,
+      "learning_rate": 0.000120851129212798,
+      "loss": 1.0977,
+      "step": 1941
+    },
+    {
+      "epoch": 0.2978527607361963,
+      "grad_norm": 0.2295672446489334,
+      "learning_rate": 0.00012082223762109915,
+      "loss": 1.1701,
+      "step": 1942
+    },
+    {
+      "epoch": 0.29800613496932515,
+      "grad_norm": 0.29851287603378296,
+      "learning_rate": 0.00012079333517555727,
+      "loss": 1.101,
+      "step": 1943
+    },
+    {
+      "epoch": 0.298159509202454,
+      "grad_norm": 0.2696934938430786,
+      "learning_rate": 0.00012076442188301843,
+      "loss": 1.0175,
+      "step": 1944
+    },
+    {
+      "epoch": 0.2983128834355828,
+      "grad_norm": 0.2519541084766388,
+      "learning_rate": 0.00012073549775033127,
+      "loss": 1.2449,
+      "step": 1945
+    },
+    {
+      "epoch": 0.29846625766871165,
+      "grad_norm": 0.21956078708171844,
+      "learning_rate": 0.00012070656278434702,
+      "loss": 1.1375,
+      "step": 1946
+    },
+    {
+      "epoch": 0.2986196319018405,
+      "grad_norm": 0.46382272243499756,
+      "learning_rate": 0.00012067761699191944,
+      "loss": 1.216,
+      "step": 1947
+    },
+    {
+      "epoch": 0.2987730061349693,
+      "grad_norm": 0.21049021184444427,
+      "learning_rate": 0.00012064866037990491,
+      "loss": 1.1382,
+      "step": 1948
+    },
+    {
+      "epoch": 0.29892638036809815,
+      "grad_norm": 0.2210572063922882,
+      "learning_rate": 0.0001206196929551623,
+      "loss": 1.1333,
+      "step": 1949
+    },
+    {
+      "epoch": 0.299079754601227,
+      "grad_norm": 0.46571096777915955,
+      "learning_rate": 0.00012059071472455311,
+      "loss": 1.0526,
+      "step": 1950
+    },
+    {
+      "epoch": 0.2992331288343558,
+      "grad_norm": 0.2259930670261383,
+      "learning_rate": 0.00012056172569494134,
+      "loss": 1.202,
+      "step": 1951
+    },
+    {
+      "epoch": 0.29938650306748466,
+      "grad_norm": 0.21275845170021057,
+      "learning_rate": 0.0001205327258731936,
+      "loss": 1.0014,
+      "step": 1952
+    },
+    {
+      "epoch": 0.2995398773006135,
+      "grad_norm": 0.22605876624584198,
+      "learning_rate": 0.00012050371526617904,
+      "loss": 1.2406,
+      "step": 1953
+    },
+    {
+      "epoch": 0.2996932515337423,
+      "grad_norm": 0.20811183750629425,
+      "learning_rate": 0.00012047469388076933,
+      "loss": 1.0232,
+      "step": 1954
+    },
+    {
+      "epoch": 0.29984662576687116,
+      "grad_norm": 0.21276848018169403,
+      "learning_rate": 0.00012044566172383873,
+      "loss": 1.1063,
+      "step": 1955
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.30129796266555786,
+      "learning_rate": 0.00012041661880226408,
+      "loss": 1.1942,
+      "step": 1956
+    },
+    {
+      "epoch": 0.3001533742331288,
+      "grad_norm": 0.2115873098373413,
+      "learning_rate": 0.00012038756512292465,
+      "loss": 0.9852,
+      "step": 1957
+    },
+    {
+      "epoch": 0.30030674846625766,
+      "grad_norm": 0.22482295334339142,
+      "learning_rate": 0.00012035850069270243,
+      "loss": 1.1148,
+      "step": 1958
+    },
+    {
+      "epoch": 0.3004601226993865,
+      "grad_norm": 0.2308046966791153,
+      "learning_rate": 0.0001203294255184818,
+      "loss": 1.1481,
+      "step": 1959
+    },
+    {
+      "epoch": 0.3006134969325153,
+      "grad_norm": 0.2622653841972351,
+      "learning_rate": 0.0001203003396071498,
+      "loss": 1.1392,
+      "step": 1960
+    },
+    {
+      "epoch": 0.30076687116564416,
+      "grad_norm": 0.22390864789485931,
+      "learning_rate": 0.00012027124296559595,
+      "loss": 1.2039,
+      "step": 1961
+    },
+    {
+      "epoch": 0.300920245398773,
+      "grad_norm": 0.2333284467458725,
+      "learning_rate": 0.00012024213560071229,
+      "loss": 1.1257,
+      "step": 1962
+    },
+    {
+      "epoch": 0.3010736196319018,
+      "grad_norm": 0.46139562129974365,
+      "learning_rate": 0.0001202130175193935,
+      "loss": 1.2704,
+      "step": 1963
+    },
+    {
+      "epoch": 0.30122699386503066,
+      "grad_norm": 0.22276277840137482,
+      "learning_rate": 0.00012018388872853667,
+      "loss": 1.0238,
+      "step": 1964
+    },
+    {
+      "epoch": 0.3013803680981595,
+      "grad_norm": 0.3860635459423065,
+      "learning_rate": 0.00012015474923504155,
+      "loss": 1.0799,
+      "step": 1965
+    },
+    {
+      "epoch": 0.3015337423312883,
+      "grad_norm": 0.22241513431072235,
+      "learning_rate": 0.00012012559904581032,
+      "loss": 1.1198,
+      "step": 1966
+    },
+    {
+      "epoch": 0.30168711656441716,
+      "grad_norm": 0.2257394790649414,
+      "learning_rate": 0.00012009643816774778,
+      "loss": 1.2048,
+      "step": 1967
+    },
+    {
+      "epoch": 0.301840490797546,
+      "grad_norm": 0.22904177010059357,
+      "learning_rate": 0.0001200672666077612,
+      "loss": 1.1626,
+      "step": 1968
+    },
+    {
+      "epoch": 0.3019938650306748,
+      "grad_norm": 0.4125727117061615,
+      "learning_rate": 0.00012003808437276038,
+      "loss": 1.1725,
+      "step": 1969
+    },
+    {
+      "epoch": 0.30214723926380366,
+      "grad_norm": 0.7140176296234131,
+      "learning_rate": 0.00012000889146965774,
+      "loss": 1.1459,
+      "step": 1970
+    },
+    {
+      "epoch": 0.3023006134969325,
+      "grad_norm": 0.24838979542255402,
+      "learning_rate": 0.00011997968790536811,
+      "loss": 1.1136,
+      "step": 1971
+    },
+    {
+      "epoch": 0.3024539877300613,
+      "grad_norm": 1.2875839471817017,
+      "learning_rate": 0.00011995047368680888,
+      "loss": 1.2019,
+      "step": 1972
+    },
+    {
+      "epoch": 0.30260736196319016,
+      "grad_norm": 0.36224353313446045,
+      "learning_rate": 0.0001199212488209,
+      "loss": 1.2049,
+      "step": 1973
+    },
+    {
+      "epoch": 0.302760736196319,
+      "grad_norm": 0.7017869353294373,
+      "learning_rate": 0.00011989201331456394,
+      "loss": 1.121,
+      "step": 1974
+    },
+    {
+      "epoch": 0.3029141104294479,
+      "grad_norm": 0.2917768359184265,
+      "learning_rate": 0.00011986276717472564,
+      "loss": 1.1939,
+      "step": 1975
+    },
+    {
+      "epoch": 0.3030674846625767,
+      "grad_norm": 0.3192695379257202,
+      "learning_rate": 0.00011983351040831258,
+      "loss": 1.1632,
+      "step": 1976
+    },
+    {
+      "epoch": 0.30322085889570555,
+      "grad_norm": 0.44383761286735535,
+      "learning_rate": 0.00011980424302225478,
+      "loss": 1.0418,
+      "step": 1977
+    },
+    {
+      "epoch": 0.3033742331288344,
+      "grad_norm": 0.47158437967300415,
+      "learning_rate": 0.00011977496502348477,
+      "loss": 1.1743,
+      "step": 1978
+    },
+    {
+      "epoch": 0.3035276073619632,
+      "grad_norm": 0.298312783241272,
+      "learning_rate": 0.00011974567641893756,
+      "loss": 1.0975,
+      "step": 1979
+    },
+    {
+      "epoch": 0.30368098159509205,
+      "grad_norm": 0.24666467308998108,
+      "learning_rate": 0.00011971637721555072,
+      "loss": 1.0884,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3038343558282209,
+      "grad_norm": 0.4058324694633484,
+      "learning_rate": 0.00011968706742026431,
+      "loss": 1.1262,
+      "step": 1981
+    },
+    {
+      "epoch": 0.3039877300613497,
+      "grad_norm": 0.4799357056617737,
+      "learning_rate": 0.00011965774704002086,
+      "loss": 1.1408,
+      "step": 1982
+    },
+    {
+      "epoch": 0.30414110429447855,
+      "grad_norm": 0.28924715518951416,
+      "learning_rate": 0.00011962841608176548,
+      "loss": 1.2202,
+      "step": 1983
+    },
+    {
+      "epoch": 0.3042944785276074,
+      "grad_norm": 0.39448630809783936,
+      "learning_rate": 0.00011959907455244575,
+      "loss": 1.3787,
+      "step": 1984
+    },
+    {
+      "epoch": 0.3044478527607362,
+      "grad_norm": 0.29516351222991943,
+      "learning_rate": 0.00011956972245901172,
+      "loss": 1.1461,
+      "step": 1985
+    },
+    {
+      "epoch": 0.30460122699386505,
+      "grad_norm": 0.25890031456947327,
+      "learning_rate": 0.00011954035980841601,
+      "loss": 1.0936,
+      "step": 1986
+    },
+    {
+      "epoch": 0.3047546012269939,
+      "grad_norm": 0.2846240997314453,
+      "learning_rate": 0.00011951098660761367,
+      "loss": 1.0365,
+      "step": 1987
+    },
+    {
+      "epoch": 0.3049079754601227,
+      "grad_norm": 0.3203687071800232,
+      "learning_rate": 0.00011948160286356234,
+      "loss": 1.2144,
+      "step": 1988
+    },
+    {
+      "epoch": 0.30506134969325155,
+      "grad_norm": 0.4400498867034912,
+      "learning_rate": 0.00011945220858322203,
+      "loss": 1.1224,
+      "step": 1989
+    },
+    {
+      "epoch": 0.3052147239263804,
+      "grad_norm": 0.2816298007965088,
+      "learning_rate": 0.00011942280377355539,
+      "loss": 1.2439,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3053680981595092,
+      "grad_norm": 0.3046875298023224,
+      "learning_rate": 0.00011939338844152744,
+      "loss": 1.1623,
+      "step": 1991
+    },
+    {
+      "epoch": 0.30552147239263805,
+      "grad_norm": 0.27858686447143555,
+      "learning_rate": 0.00011936396259410577,
+      "loss": 1.1022,
+      "step": 1992
+    },
+    {
+      "epoch": 0.3056748466257669,
+      "grad_norm": 0.5457581877708435,
+      "learning_rate": 0.00011933452623826043,
+      "loss": 1.1842,
+      "step": 1993
+    },
+    {
+      "epoch": 0.3058282208588957,
+      "grad_norm": 0.24072328209877014,
+      "learning_rate": 0.00011930507938096396,
+      "loss": 1.2368,
+      "step": 1994
+    },
+    {
+      "epoch": 0.30598159509202455,
+      "grad_norm": 0.2542443573474884,
+      "learning_rate": 0.00011927562202919139,
+      "loss": 1.106,
+      "step": 1995
+    },
+    {
+      "epoch": 0.3061349693251534,
+      "grad_norm": 0.4752184748649597,
+      "learning_rate": 0.00011924615418992025,
+      "loss": 1.0968,
+      "step": 1996
+    },
+    {
+      "epoch": 0.3062883435582822,
+      "grad_norm": 0.33898496627807617,
+      "learning_rate": 0.00011921667587013054,
+      "loss": 1.0413,
+      "step": 1997
+    },
+    {
+      "epoch": 0.30644171779141105,
+      "grad_norm": 0.2699562907218933,
+      "learning_rate": 0.00011918718707680471,
+      "loss": 1.2121,
+      "step": 1998
+    },
+    {
+      "epoch": 0.3065950920245399,
+      "grad_norm": 0.23603251576423645,
+      "learning_rate": 0.00011915768781692779,
+      "loss": 0.9902,
+      "step": 1999
+    },
+    {
+      "epoch": 0.3067484662576687,
+      "grad_norm": 0.2548014521598816,
+      "learning_rate": 0.00011912817809748717,
+      "loss": 1.1544,
+      "step": 2000
+    },
+    {
+      "epoch": 0.30690184049079755,
+      "grad_norm": 0.2765936255455017,
+      "learning_rate": 0.0001190986579254728,
+      "loss": 1.0595,
+      "step": 2001
+    },
+    {
+      "epoch": 0.3070552147239264,
+      "grad_norm": 0.22175633907318115,
+      "learning_rate": 0.00011906912730787707,
+      "loss": 1.0768,
+      "step": 2002
+    },
+    {
+      "epoch": 0.3072085889570552,
+      "grad_norm": 0.36023789644241333,
+      "learning_rate": 0.00011903958625169484,
+      "loss": 1.0889,
+      "step": 2003
+    },
+    {
+      "epoch": 0.30736196319018405,
+      "grad_norm": 0.24995064735412598,
+      "learning_rate": 0.00011901003476392346,
+      "loss": 1.2765,
+      "step": 2004
+    },
+    {
+      "epoch": 0.3075153374233129,
+      "grad_norm": 0.22217421233654022,
+      "learning_rate": 0.00011898047285156274,
+      "loss": 1.2017,
+      "step": 2005
+    },
+    {
+      "epoch": 0.3076687116564417,
+      "grad_norm": 0.24546882510185242,
+      "learning_rate": 0.000118950900521615,
+      "loss": 1.2107,
+      "step": 2006
+    },
+    {
+      "epoch": 0.30782208588957055,
+      "grad_norm": 0.27184733748435974,
+      "learning_rate": 0.00011892131778108497,
+      "loss": 1.2278,
+      "step": 2007
+    },
+    {
+      "epoch": 0.3079754601226994,
+      "grad_norm": 0.35606515407562256,
+      "learning_rate": 0.00011889172463697984,
+      "loss": 1.075,
+      "step": 2008
+    },
+    {
+      "epoch": 0.3081288343558282,
+      "grad_norm": 0.641994297504425,
+      "learning_rate": 0.00011886212109630933,
+      "loss": 1.2851,
+      "step": 2009
+    },
+    {
+      "epoch": 0.30828220858895705,
+      "grad_norm": 0.2205173522233963,
+      "learning_rate": 0.00011883250716608557,
+      "loss": 1.0885,
+      "step": 2010
+    },
+    {
+      "epoch": 0.3084355828220859,
+      "grad_norm": 0.45107245445251465,
+      "learning_rate": 0.0001188028828533232,
+      "loss": 1.0494,
+      "step": 2011
+    },
+    {
+      "epoch": 0.3085889570552147,
+      "grad_norm": 0.4053805470466614,
+      "learning_rate": 0.0001187732481650392,
+      "loss": 1.0679,
+      "step": 2012
+    },
+    {
+      "epoch": 0.30874233128834355,
+      "grad_norm": 0.3488190174102783,
+      "learning_rate": 0.0001187436031082532,
+      "loss": 1.1976,
+      "step": 2013
+    },
+    {
+      "epoch": 0.3088957055214724,
+      "grad_norm": 0.43984925746917725,
+      "learning_rate": 0.0001187139476899871,
+      "loss": 1.0511,
+      "step": 2014
+    },
+    {
+      "epoch": 0.3090490797546012,
+      "grad_norm": 0.47614163160324097,
+      "learning_rate": 0.00011868428191726536,
+      "loss": 1.1279,
+      "step": 2015
+    },
+    {
+      "epoch": 0.30920245398773005,
+      "grad_norm": 0.2610242962837219,
+      "learning_rate": 0.00011865460579711487,
+      "loss": 1.1606,
+      "step": 2016
+    },
+    {
+      "epoch": 0.3093558282208589,
+      "grad_norm": 0.29505789279937744,
+      "learning_rate": 0.00011862491933656497,
+      "loss": 1.0803,
+      "step": 2017
+    },
+    {
+      "epoch": 0.3095092024539877,
+      "grad_norm": 0.4754345715045929,
+      "learning_rate": 0.00011859522254264742,
+      "loss": 1.2349,
+      "step": 2018
+    },
+    {
+      "epoch": 0.30966257668711655,
+      "grad_norm": 0.3153233230113983,
+      "learning_rate": 0.0001185655154223965,
+      "loss": 1.0223,
+      "step": 2019
+    },
+    {
+      "epoch": 0.3098159509202454,
+      "grad_norm": 0.3752325773239136,
+      "learning_rate": 0.00011853579798284886,
+      "loss": 1.2841,
+      "step": 2020
+    },
+    {
+      "epoch": 0.3099693251533742,
+      "grad_norm": 0.3091168701648712,
+      "learning_rate": 0.00011850607023104362,
+      "loss": 1.075,
+      "step": 2021
+    },
+    {
+      "epoch": 0.31012269938650305,
+      "grad_norm": 0.47162890434265137,
+      "learning_rate": 0.00011847633217402236,
+      "loss": 1.1458,
+      "step": 2022
+    },
+    {
+      "epoch": 0.3102760736196319,
+      "grad_norm": 0.2502751052379608,
+      "learning_rate": 0.00011844658381882905,
+      "loss": 1.1262,
+      "step": 2023
+    },
+    {
+      "epoch": 0.3104294478527607,
+      "grad_norm": 0.29777321219444275,
+      "learning_rate": 0.00011841682517251021,
+      "loss": 1.1611,
+      "step": 2024
+    },
+    {
+      "epoch": 0.31058282208588955,
+      "grad_norm": 0.2561899721622467,
+      "learning_rate": 0.00011838705624211469,
+      "loss": 1.1419,
+      "step": 2025
+    },
+    {
+      "epoch": 0.3107361963190184,
+      "grad_norm": 0.2598625719547272,
+      "learning_rate": 0.00011835727703469377,
+      "loss": 1.1426,
+      "step": 2026
+    },
+    {
+      "epoch": 0.3108895705521472,
+      "grad_norm": 0.23384545743465424,
+      "learning_rate": 0.00011832748755730126,
+      "loss": 1.1286,
+      "step": 2027
+    },
+    {
+      "epoch": 0.31104294478527605,
+      "grad_norm": 0.27308690547943115,
+      "learning_rate": 0.00011829768781699333,
+      "loss": 1.123,
+      "step": 2028
+    },
+    {
+      "epoch": 0.3111963190184049,
+      "grad_norm": 0.28117576241493225,
+      "learning_rate": 0.0001182678778208286,
+      "loss": 1.0987,
+      "step": 2029
+    },
+    {
+      "epoch": 0.3113496932515337,
+      "grad_norm": 0.47760146856307983,
+      "learning_rate": 0.0001182380575758681,
+      "loss": 0.9773,
+      "step": 2030
+    },
+    {
+      "epoch": 0.31150306748466255,
+      "grad_norm": 0.2887839376926422,
+      "learning_rate": 0.00011820822708917532,
+      "loss": 1.2621,
+      "step": 2031
+    },
+    {
+      "epoch": 0.3116564417177914,
+      "grad_norm": 0.259050190448761,
+      "learning_rate": 0.00011817838636781619,
+      "loss": 1.1619,
+      "step": 2032
+    },
+    {
+      "epoch": 0.3118098159509202,
+      "grad_norm": 0.26510950922966003,
+      "learning_rate": 0.00011814853541885897,
+      "loss": 1.1575,
+      "step": 2033
+    },
+    {
+      "epoch": 0.31196319018404906,
+      "grad_norm": 0.2392127960920334,
+      "learning_rate": 0.0001181186742493745,
+      "loss": 1.1729,
+      "step": 2034
+    },
+    {
+      "epoch": 0.3121165644171779,
+      "grad_norm": 0.21716976165771484,
+      "learning_rate": 0.00011808880286643585,
+      "loss": 1.112,
+      "step": 2035
+    },
+    {
+      "epoch": 0.3122699386503067,
+      "grad_norm": 0.47390568256378174,
+      "learning_rate": 0.00011805892127711865,
+      "loss": 1.0973,
+      "step": 2036
+    },
+    {
+      "epoch": 0.31242331288343556,
+      "grad_norm": 0.2280590683221817,
+      "learning_rate": 0.00011802902948850095,
+      "loss": 1.0546,
+      "step": 2037
+    },
+    {
+      "epoch": 0.31257668711656444,
+      "grad_norm": 0.38014882802963257,
+      "learning_rate": 0.0001179991275076631,
+      "loss": 1.2802,
+      "step": 2038
+    },
+    {
+      "epoch": 0.3127300613496933,
+      "grad_norm": 0.22329308092594147,
+      "learning_rate": 0.000117969215341688,
+      "loss": 1.226,
+      "step": 2039
+    },
+    {
+      "epoch": 0.3128834355828221,
+      "grad_norm": 0.2228647917509079,
+      "learning_rate": 0.00011793929299766084,
+      "loss": 1.0913,
+      "step": 2040
+    },
+    {
+      "epoch": 0.31303680981595094,
+      "grad_norm": 0.21949118375778198,
+      "learning_rate": 0.00011790936048266931,
+      "loss": 1.0809,
+      "step": 2041
+    },
+    {
+      "epoch": 0.3131901840490798,
+      "grad_norm": 0.4141925275325775,
+      "learning_rate": 0.00011787941780380348,
+      "loss": 0.9828,
+      "step": 2042
+    },
+    {
+      "epoch": 0.3133435582822086,
+      "grad_norm": 0.21674688160419464,
+      "learning_rate": 0.00011784946496815583,
+      "loss": 1.138,
+      "step": 2043
+    },
+    {
+      "epoch": 0.31349693251533745,
+      "grad_norm": 0.23845529556274414,
+      "learning_rate": 0.00011781950198282123,
+      "loss": 1.2699,
+      "step": 2044
+    },
+    {
+      "epoch": 0.3136503067484663,
+      "grad_norm": 0.4109309911727905,
+      "learning_rate": 0.00011778952885489697,
+      "loss": 0.9424,
+      "step": 2045
+    },
+    {
+      "epoch": 0.3138036809815951,
+      "grad_norm": 0.2484615594148636,
+      "learning_rate": 0.00011775954559148273,
+      "loss": 1.2782,
+      "step": 2046
+    },
+    {
+      "epoch": 0.31395705521472395,
+      "grad_norm": 0.2723847031593323,
+      "learning_rate": 0.00011772955219968062,
+      "loss": 1.0509,
+      "step": 2047
+    },
+    {
+      "epoch": 0.3141104294478528,
+      "grad_norm": 0.22099396586418152,
+      "learning_rate": 0.00011769954868659514,
+      "loss": 1.0876,
+      "step": 2048
+    },
+    {
+      "epoch": 0.3142638036809816,
+      "grad_norm": 0.6517888903617859,
+      "learning_rate": 0.00011766953505933312,
+      "loss": 1.1069,
+      "step": 2049
+    },
+    {
+      "epoch": 0.31441717791411045,
+      "grad_norm": 0.23072803020477295,
+      "learning_rate": 0.0001176395113250039,
+      "loss": 1.0347,
+      "step": 2050
+    },
+    {
+      "epoch": 0.3145705521472393,
+      "grad_norm": 0.3843873143196106,
+      "learning_rate": 0.0001176094774907191,
+      "loss": 1.1818,
+      "step": 2051
+    },
+    {
+      "epoch": 0.3147239263803681,
+      "grad_norm": 1.0863109827041626,
+      "learning_rate": 0.00011757943356359284,
+      "loss": 1.1498,
+      "step": 2052
+    },
+    {
+      "epoch": 0.31487730061349695,
+      "grad_norm": 0.2807559669017792,
+      "learning_rate": 0.00011754937955074157,
+      "loss": 1.2862,
+      "step": 2053
+    },
+    {
+      "epoch": 0.3150306748466258,
+      "grad_norm": 0.3272719979286194,
+      "learning_rate": 0.00011751931545928409,
+      "loss": 1.2568,
+      "step": 2054
+    },
+    {
+      "epoch": 0.3151840490797546,
+      "grad_norm": 0.5127188563346863,
+      "learning_rate": 0.00011748924129634171,
+      "loss": 1.2351,
+      "step": 2055
+    },
+    {
+      "epoch": 0.31533742331288345,
+      "grad_norm": 0.29457730054855347,
+      "learning_rate": 0.00011745915706903801,
+      "loss": 1.1442,
+      "step": 2056
+    },
+    {
+      "epoch": 0.3154907975460123,
+      "grad_norm": 0.3086037337779999,
+      "learning_rate": 0.00011742906278449897,
+      "loss": 1.1998,
+      "step": 2057
+    },
+    {
+      "epoch": 0.3156441717791411,
+      "grad_norm": 0.2668285667896271,
+      "learning_rate": 0.00011739895844985302,
+      "loss": 1.154,
+      "step": 2058
+    },
+    {
+      "epoch": 0.31579754601226995,
+      "grad_norm": 0.8680452704429626,
+      "learning_rate": 0.00011736884407223091,
+      "loss": 1.1477,
+      "step": 2059
+    },
+    {
+      "epoch": 0.3159509202453988,
+      "grad_norm": 0.3466576337814331,
+      "learning_rate": 0.0001173387196587658,
+      "loss": 1.2203,
+      "step": 2060
+    },
+    {
+      "epoch": 0.3161042944785276,
+      "grad_norm": 0.26049551367759705,
+      "learning_rate": 0.0001173085852165932,
+      "loss": 1.2139,
+      "step": 2061
+    },
+    {
+      "epoch": 0.31625766871165645,
+      "grad_norm": 0.5072387456893921,
+      "learning_rate": 0.000117278440752851,
+      "loss": 1.248,
+      "step": 2062
+    },
+    {
+      "epoch": 0.3164110429447853,
+      "grad_norm": 0.3260441720485687,
+      "learning_rate": 0.00011724828627467948,
+      "loss": 1.1912,
+      "step": 2063
+    },
+    {
+      "epoch": 0.3165644171779141,
+      "grad_norm": 0.25832754373550415,
+      "learning_rate": 0.0001172181217892213,
+      "loss": 1.1708,
+      "step": 2064
+    },
+    {
+      "epoch": 0.31671779141104295,
+      "grad_norm": 0.4657942056655884,
+      "learning_rate": 0.00011718794730362147,
+      "loss": 1.2258,
+      "step": 2065
+    },
+    {
+      "epoch": 0.3168711656441718,
+      "grad_norm": 0.3225032687187195,
+      "learning_rate": 0.00011715776282502735,
+      "loss": 1.0978,
+      "step": 2066
+    },
+    {
+      "epoch": 0.3170245398773006,
+      "grad_norm": 0.27882614731788635,
+      "learning_rate": 0.00011712756836058871,
+      "loss": 1.1092,
+      "step": 2067
+    },
+    {
+      "epoch": 0.31717791411042945,
+      "grad_norm": 0.43778833746910095,
+      "learning_rate": 0.00011709736391745769,
+      "loss": 1.0947,
+      "step": 2068
+    },
+    {
+      "epoch": 0.3173312883435583,
+      "grad_norm": 0.29696327447891235,
+      "learning_rate": 0.00011706714950278871,
+      "loss": 1.1603,
+      "step": 2069
+    },
+    {
+      "epoch": 0.3174846625766871,
+      "grad_norm": 0.5270256996154785,
+      "learning_rate": 0.00011703692512373868,
+      "loss": 0.9688,
+      "step": 2070
+    },
+    {
+      "epoch": 0.31763803680981595,
+      "grad_norm": 0.26528674364089966,
+      "learning_rate": 0.00011700669078746676,
+      "loss": 1.198,
+      "step": 2071
+    },
+    {
+      "epoch": 0.3177914110429448,
+      "grad_norm": 0.2558128833770752,
+      "learning_rate": 0.0001169764465011345,
+      "loss": 1.1946,
+      "step": 2072
+    },
+    {
+      "epoch": 0.3179447852760736,
+      "grad_norm": 0.24115176498889923,
+      "learning_rate": 0.00011694619227190585,
+      "loss": 1.2108,
+      "step": 2073
+    },
+    {
+      "epoch": 0.31809815950920245,
+      "grad_norm": 0.3094039857387543,
+      "learning_rate": 0.00011691592810694706,
+      "loss": 1.1924,
+      "step": 2074
+    },
+    {
+      "epoch": 0.3182515337423313,
+      "grad_norm": 0.23247317969799042,
+      "learning_rate": 0.00011688565401342679,
+      "loss": 1.1715,
+      "step": 2075
+    },
+    {
+      "epoch": 0.3184049079754601,
+      "grad_norm": 0.2718925178050995,
+      "learning_rate": 0.000116855369998516,
+      "loss": 1.0364,
+      "step": 2076
+    },
+    {
+      "epoch": 0.31855828220858895,
+      "grad_norm": 0.24640357494354248,
+      "learning_rate": 0.000116825076069388,
+      "loss": 1.2613,
+      "step": 2077
+    },
+    {
+      "epoch": 0.3187116564417178,
+      "grad_norm": 0.9361047148704529,
+      "learning_rate": 0.00011679477223321849,
+      "loss": 1.2511,
+      "step": 2078
+    },
+    {
+      "epoch": 0.3188650306748466,
+      "grad_norm": 0.23513992130756378,
+      "learning_rate": 0.00011676445849718544,
+      "loss": 1.1903,
+      "step": 2079
+    },
+    {
+      "epoch": 0.31901840490797545,
+      "grad_norm": 0.6282753348350525,
+      "learning_rate": 0.00011673413486846931,
+      "loss": 1.1084,
+      "step": 2080
+    },
+    {
+      "epoch": 0.3191717791411043,
+      "grad_norm": 0.24602270126342773,
+      "learning_rate": 0.00011670380135425274,
+      "loss": 1.2141,
+      "step": 2081
+    },
+    {
+      "epoch": 0.3193251533742331,
+      "grad_norm": 0.24885410070419312,
+      "learning_rate": 0.00011667345796172082,
+      "loss": 1.0674,
+      "step": 2082
+    },
+    {
+      "epoch": 0.31947852760736195,
+      "grad_norm": 0.3875042796134949,
+      "learning_rate": 0.0001166431046980609,
+      "loss": 1.1293,
+      "step": 2083
+    },
+    {
+      "epoch": 0.3196319018404908,
+      "grad_norm": 0.24105136096477509,
+      "learning_rate": 0.00011661274157046277,
+      "loss": 1.1099,
+      "step": 2084
+    },
+    {
+      "epoch": 0.3197852760736196,
+      "grad_norm": 0.2315361499786377,
+      "learning_rate": 0.00011658236858611844,
+      "loss": 1.2054,
+      "step": 2085
+    },
+    {
+      "epoch": 0.31993865030674845,
+      "grad_norm": 0.5660565495491028,
+      "learning_rate": 0.00011655198575222236,
+      "loss": 1.1904,
+      "step": 2086
+    },
+    {
+      "epoch": 0.3200920245398773,
+      "grad_norm": 0.30285975337028503,
+      "learning_rate": 0.00011652159307597123,
+      "loss": 1.0915,
+      "step": 2087
+    },
+    {
+      "epoch": 0.3202453987730061,
+      "grad_norm": 0.24959087371826172,
+      "learning_rate": 0.00011649119056456415,
+      "loss": 1.0665,
+      "step": 2088
+    },
+    {
+      "epoch": 0.32039877300613495,
+      "grad_norm": 0.33122485876083374,
+      "learning_rate": 0.00011646077822520248,
+      "loss": 1.1477,
+      "step": 2089
+    },
+    {
+      "epoch": 0.3205521472392638,
+      "grad_norm": 0.47002488374710083,
+      "learning_rate": 0.00011643035606508996,
+      "loss": 1.1345,
+      "step": 2090
+    },
+    {
+      "epoch": 0.3207055214723926,
+      "grad_norm": 0.3842456340789795,
+      "learning_rate": 0.00011639992409143264,
+      "loss": 1.1241,
+      "step": 2091
+    },
+    {
+      "epoch": 0.32085889570552145,
+      "grad_norm": 0.5561758279800415,
+      "learning_rate": 0.00011636948231143891,
+      "loss": 1.1482,
+      "step": 2092
+    },
+    {
+      "epoch": 0.3210122699386503,
+      "grad_norm": 0.3623267412185669,
+      "learning_rate": 0.00011633903073231941,
+      "loss": 1.1437,
+      "step": 2093
+    },
+    {
+      "epoch": 0.3211656441717791,
+      "grad_norm": 0.5673458576202393,
+      "learning_rate": 0.00011630856936128725,
+      "loss": 1.1359,
+      "step": 2094
+    },
+    {
+      "epoch": 0.32131901840490795,
+      "grad_norm": 0.2157488465309143,
+      "learning_rate": 0.00011627809820555771,
+      "loss": 1.1207,
+      "step": 2095
+    },
+    {
+      "epoch": 0.3214723926380368,
+      "grad_norm": 0.255827397108078,
+      "learning_rate": 0.00011624761727234844,
+      "loss": 1.2211,
+      "step": 2096
+    },
+    {
+      "epoch": 0.3216257668711656,
+      "grad_norm": 0.514578104019165,
+      "learning_rate": 0.00011621712656887944,
+      "loss": 1.3417,
+      "step": 2097
+    },
+    {
+      "epoch": 0.32177914110429445,
+      "grad_norm": 0.3177904188632965,
+      "learning_rate": 0.00011618662610237299,
+      "loss": 1.1345,
+      "step": 2098
+    },
+    {
+      "epoch": 0.3219325153374233,
+      "grad_norm": 0.41988497972488403,
+      "learning_rate": 0.00011615611588005368,
+      "loss": 1.0715,
+      "step": 2099
+    },
+    {
+      "epoch": 0.3220858895705521,
+      "grad_norm": 0.2319345474243164,
+      "learning_rate": 0.00011612559590914842,
+      "loss": 1.2045,
+      "step": 2100
+    },
+    {
+      "epoch": 0.322239263803681,
+      "grad_norm": 0.22138464450836182,
+      "learning_rate": 0.00011609506619688641,
+      "loss": 1.1096,
+      "step": 2101
+    },
+    {
+      "epoch": 0.32239263803680984,
+      "grad_norm": 0.3099164366722107,
+      "learning_rate": 0.00011606452675049922,
+      "loss": 1.1083,
+      "step": 2102
+    },
+    {
+      "epoch": 0.3225460122699387,
+      "grad_norm": 0.3520711064338684,
+      "learning_rate": 0.00011603397757722066,
+      "loss": 1.2004,
+      "step": 2103
+    },
+    {
+      "epoch": 0.3226993865030675,
+      "grad_norm": 0.22952105104923248,
+      "learning_rate": 0.00011600341868428686,
+      "loss": 1.0185,
+      "step": 2104
+    },
+    {
+      "epoch": 0.32285276073619634,
+      "grad_norm": 0.2458094358444214,
+      "learning_rate": 0.00011597285007893626,
+      "loss": 1.2534,
+      "step": 2105
+    },
+    {
+      "epoch": 0.3230061349693252,
+      "grad_norm": 0.3337112069129944,
+      "learning_rate": 0.0001159422717684096,
+      "loss": 1.0903,
+      "step": 2106
+    },
+    {
+      "epoch": 0.323159509202454,
+      "grad_norm": 0.22169286012649536,
+      "learning_rate": 0.00011591168375994991,
+      "loss": 1.0131,
+      "step": 2107
+    },
+    {
+      "epoch": 0.32331288343558284,
+      "grad_norm": 0.3341331481933594,
+      "learning_rate": 0.00011588108606080254,
+      "loss": 1.2118,
+      "step": 2108
+    },
+    {
+      "epoch": 0.3234662576687117,
+      "grad_norm": 0.7077994346618652,
+      "learning_rate": 0.00011585047867821512,
+      "loss": 1.1148,
+      "step": 2109
+    },
+    {
+      "epoch": 0.3236196319018405,
+      "grad_norm": 0.3344380259513855,
+      "learning_rate": 0.00011581986161943758,
+      "loss": 1.2024,
+      "step": 2110
+    },
+    {
+      "epoch": 0.32377300613496934,
+      "grad_norm": 0.6363966464996338,
+      "learning_rate": 0.00011578923489172209,
+      "loss": 1.0977,
+      "step": 2111
+    },
+    {
+      "epoch": 0.3239263803680982,
+      "grad_norm": 0.25915586948394775,
+      "learning_rate": 0.00011575859850232321,
+      "loss": 1.2349,
+      "step": 2112
+    },
+    {
+      "epoch": 0.324079754601227,
+      "grad_norm": 0.24833527207374573,
+      "learning_rate": 0.0001157279524584977,
+      "loss": 1.1616,
+      "step": 2113
+    },
+    {
+      "epoch": 0.32423312883435584,
+      "grad_norm": 0.3449287712574005,
+      "learning_rate": 0.00011569729676750465,
+      "loss": 1.0735,
+      "step": 2114
+    },
+    {
+      "epoch": 0.3243865030674847,
+      "grad_norm": 0.5721024870872498,
+      "learning_rate": 0.00011566663143660545,
+      "loss": 1.2685,
+      "step": 2115
+    },
+    {
+      "epoch": 0.3245398773006135,
+      "grad_norm": 0.48847925662994385,
+      "learning_rate": 0.00011563595647306372,
+      "loss": 1.2095,
+      "step": 2116
+    },
+    {
+      "epoch": 0.32469325153374234,
+      "grad_norm": 0.3152576982975006,
+      "learning_rate": 0.0001156052718841454,
+      "loss": 1.2176,
+      "step": 2117
+    },
+    {
+      "epoch": 0.3248466257668712,
+      "grad_norm": 0.2739075720310211,
+      "learning_rate": 0.00011557457767711871,
+      "loss": 1.1618,
+      "step": 2118
+    },
+    {
+      "epoch": 0.325,
+      "grad_norm": 0.29229143261909485,
+      "learning_rate": 0.00011554387385925414,
+      "loss": 1.1215,
+      "step": 2119
+    },
+    {
+      "epoch": 0.32515337423312884,
+      "grad_norm": 0.281669944524765,
+      "learning_rate": 0.0001155131604378244,
+      "loss": 1.2123,
+      "step": 2120
+    },
+    {
+      "epoch": 0.3253067484662577,
+      "grad_norm": 0.31511685252189636,
+      "learning_rate": 0.00011548243742010464,
+      "loss": 1.1002,
+      "step": 2121
+    },
+    {
+      "epoch": 0.3254601226993865,
+      "grad_norm": 0.26750531792640686,
+      "learning_rate": 0.0001154517048133721,
+      "loss": 1.138,
+      "step": 2122
+    },
+    {
+      "epoch": 0.32561349693251534,
+      "grad_norm": 0.4569969177246094,
+      "learning_rate": 0.00011542096262490636,
+      "loss": 1.1399,
+      "step": 2123
+    },
+    {
+      "epoch": 0.3257668711656442,
+      "grad_norm": 0.28135594725608826,
+      "learning_rate": 0.0001153902108619893,
+      "loss": 1.1618,
+      "step": 2124
+    },
+    {
+      "epoch": 0.325920245398773,
+      "grad_norm": 0.4052767753601074,
+      "learning_rate": 0.00011535944953190505,
+      "loss": 1.1619,
+      "step": 2125
+    },
+    {
+      "epoch": 0.32607361963190185,
+      "grad_norm": 0.3280336558818817,
+      "learning_rate": 0.00011532867864193997,
+      "loss": 1.1352,
+      "step": 2126
+    },
+    {
+      "epoch": 0.3262269938650307,
+      "grad_norm": 0.2339053750038147,
+      "learning_rate": 0.00011529789819938275,
+      "loss": 1.1858,
+      "step": 2127
+    },
+    {
+      "epoch": 0.3263803680981595,
+      "grad_norm": 0.2548159658908844,
+      "learning_rate": 0.00011526710821152427,
+      "loss": 0.9973,
+      "step": 2128
+    },
+    {
+      "epoch": 0.32653374233128835,
+      "grad_norm": 0.2190786749124527,
+      "learning_rate": 0.00011523630868565773,
+      "loss": 1.1591,
+      "step": 2129
+    },
+    {
+      "epoch": 0.3266871165644172,
+      "grad_norm": 0.36930644512176514,
+      "learning_rate": 0.00011520549962907857,
+      "loss": 1.1956,
+      "step": 2130
+    },
+    {
+      "epoch": 0.326840490797546,
+      "grad_norm": 0.23340868949890137,
+      "learning_rate": 0.00011517468104908444,
+      "loss": 1.2265,
+      "step": 2131
+    },
+    {
+      "epoch": 0.32699386503067485,
+      "grad_norm": 0.23325209319591522,
+      "learning_rate": 0.00011514385295297534,
+      "loss": 1.0311,
+      "step": 2132
+    },
+    {
+      "epoch": 0.3271472392638037,
+      "grad_norm": 0.3745498061180115,
+      "learning_rate": 0.00011511301534805348,
+      "loss": 1.1044,
+      "step": 2133
+    },
+    {
+      "epoch": 0.3273006134969325,
+      "grad_norm": 0.2236793339252472,
+      "learning_rate": 0.00011508216824162326,
+      "loss": 1.1407,
+      "step": 2134
+    },
+    {
+      "epoch": 0.32745398773006135,
+      "grad_norm": 0.21163193881511688,
+      "learning_rate": 0.00011505131164099143,
+      "loss": 1.0913,
+      "step": 2135
+    },
+    {
+      "epoch": 0.3276073619631902,
+      "grad_norm": 0.22671017050743103,
+      "learning_rate": 0.00011502044555346693,
+      "loss": 1.2528,
+      "step": 2136
+    },
+    {
+      "epoch": 0.327760736196319,
+      "grad_norm": 0.22990642488002777,
+      "learning_rate": 0.00011498956998636099,
+      "loss": 1.0766,
+      "step": 2137
+    },
+    {
+      "epoch": 0.32791411042944785,
+      "grad_norm": 0.2587161362171173,
+      "learning_rate": 0.00011495868494698702,
+      "loss": 1.1742,
+      "step": 2138
+    },
+    {
+      "epoch": 0.3280674846625767,
+      "grad_norm": 0.38465067744255066,
+      "learning_rate": 0.00011492779044266073,
+      "loss": 1.3242,
+      "step": 2139
+    },
+    {
+      "epoch": 0.3282208588957055,
+      "grad_norm": 0.21239735186100006,
+      "learning_rate": 0.00011489688648070005,
+      "loss": 1.1808,
+      "step": 2140
+    },
+    {
+      "epoch": 0.32837423312883435,
+      "grad_norm": 0.2297668159008026,
+      "learning_rate": 0.00011486597306842514,
+      "loss": 1.0858,
+      "step": 2141
+    },
+    {
+      "epoch": 0.3285276073619632,
+      "grad_norm": 0.20530425012111664,
+      "learning_rate": 0.00011483505021315847,
+      "loss": 1.0975,
+      "step": 2142
+    },
+    {
+      "epoch": 0.328680981595092,
+      "grad_norm": 0.21028612554073334,
+      "learning_rate": 0.00011480411792222462,
+      "loss": 1.1322,
+      "step": 2143
+    },
+    {
+      "epoch": 0.32883435582822085,
+      "grad_norm": 0.22614608705043793,
+      "learning_rate": 0.00011477317620295049,
+      "loss": 1.176,
+      "step": 2144
+    },
+    {
+      "epoch": 0.3289877300613497,
+      "grad_norm": 0.44087469577789307,
+      "learning_rate": 0.00011474222506266522,
+      "loss": 1.3011,
+      "step": 2145
+    },
+    {
+      "epoch": 0.3291411042944785,
+      "grad_norm": 0.4684230387210846,
+      "learning_rate": 0.00011471126450870016,
+      "loss": 0.9884,
+      "step": 2146
+    },
+    {
+      "epoch": 0.32929447852760735,
+      "grad_norm": 0.21588344871997833,
+      "learning_rate": 0.00011468029454838886,
+      "loss": 1.1513,
+      "step": 2147
+    },
+    {
+      "epoch": 0.3294478527607362,
+      "grad_norm": 0.5984287261962891,
+      "learning_rate": 0.00011464931518906715,
+      "loss": 1.0594,
+      "step": 2148
+    },
+    {
+      "epoch": 0.329601226993865,
+      "grad_norm": 0.22537560760974884,
+      "learning_rate": 0.00011461832643807307,
+      "loss": 1.1534,
+      "step": 2149
+    },
+    {
+      "epoch": 0.32975460122699385,
+      "grad_norm": 0.2393924444913864,
+      "learning_rate": 0.00011458732830274686,
+      "loss": 1.241,
+      "step": 2150
+    },
+    {
+      "epoch": 0.3299079754601227,
+      "grad_norm": 0.21872982382774353,
+      "learning_rate": 0.00011455632079043101,
+      "loss": 1.0362,
+      "step": 2151
+    },
+    {
+      "epoch": 0.3300613496932515,
+      "grad_norm": 0.46357035636901855,
+      "learning_rate": 0.00011452530390847019,
+      "loss": 1.101,
+      "step": 2152
+    },
+    {
+      "epoch": 0.33021472392638035,
+      "grad_norm": 0.2308637946844101,
+      "learning_rate": 0.00011449427766421138,
+      "loss": 1.2581,
+      "step": 2153
+    },
+    {
+      "epoch": 0.3303680981595092,
+      "grad_norm": 0.5972962379455566,
+      "learning_rate": 0.00011446324206500367,
+      "loss": 1.1919,
+      "step": 2154
+    },
+    {
+      "epoch": 0.330521472392638,
+      "grad_norm": 0.3233962655067444,
+      "learning_rate": 0.00011443219711819842,
+      "loss": 1.1572,
+      "step": 2155
+    },
+    {
+      "epoch": 0.33067484662576685,
+      "grad_norm": 0.33283883333206177,
+      "learning_rate": 0.00011440114283114923,
+      "loss": 1.1164,
+      "step": 2156
+    },
+    {
+      "epoch": 0.3308282208588957,
+      "grad_norm": 0.2424173653125763,
+      "learning_rate": 0.00011437007921121187,
+      "loss": 1.256,
+      "step": 2157
+    },
+    {
+      "epoch": 0.3309815950920245,
+      "grad_norm": 0.237359881401062,
+      "learning_rate": 0.0001143390062657443,
+      "loss": 1.1099,
+      "step": 2158
+    },
+    {
+      "epoch": 0.33113496932515335,
+      "grad_norm": 0.32807326316833496,
+      "learning_rate": 0.00011430792400210676,
+      "loss": 1.1418,
+      "step": 2159
+    },
+    {
+      "epoch": 0.3312883435582822,
+      "grad_norm": 0.34481582045555115,
+      "learning_rate": 0.00011427683242766165,
+      "loss": 1.1779,
+      "step": 2160
+    },
+    {
+      "epoch": 0.331441717791411,
+      "grad_norm": 0.23089967668056488,
+      "learning_rate": 0.00011424573154977356,
+      "loss": 1.273,
+      "step": 2161
+    },
+    {
+      "epoch": 0.33159509202453985,
+      "grad_norm": 0.2335558980703354,
+      "learning_rate": 0.00011421462137580933,
+      "loss": 1.1382,
+      "step": 2162
+    },
+    {
+      "epoch": 0.33174846625766874,
+      "grad_norm": 0.22664405405521393,
+      "learning_rate": 0.00011418350191313798,
+      "loss": 1.198,
+      "step": 2163
+    },
+    {
+      "epoch": 0.3319018404907976,
+      "grad_norm": 0.38447779417037964,
+      "learning_rate": 0.00011415237316913073,
+      "loss": 1.1906,
+      "step": 2164
+    },
+    {
+      "epoch": 0.3320552147239264,
+      "grad_norm": 0.22901588678359985,
+      "learning_rate": 0.00011412123515116098,
+      "loss": 1.2217,
+      "step": 2165
+    },
+    {
+      "epoch": 0.33220858895705524,
+      "grad_norm": 0.22537654638290405,
+      "learning_rate": 0.00011409008786660436,
+      "loss": 1.0888,
+      "step": 2166
+    },
+    {
+      "epoch": 0.3323619631901841,
+      "grad_norm": 0.22759351134300232,
+      "learning_rate": 0.00011405893132283866,
+      "loss": 1.1583,
+      "step": 2167
+    },
+    {
+      "epoch": 0.3325153374233129,
+      "grad_norm": 0.3474869728088379,
+      "learning_rate": 0.0001140277655272439,
+      "loss": 1.2018,
+      "step": 2168
+    },
+    {
+      "epoch": 0.33266871165644174,
+      "grad_norm": 0.24630072712898254,
+      "learning_rate": 0.00011399659048720226,
+      "loss": 1.2015,
+      "step": 2169
+    },
+    {
+      "epoch": 0.3328220858895706,
+      "grad_norm": 0.5960997343063354,
+      "learning_rate": 0.00011396540621009815,
+      "loss": 1.2146,
+      "step": 2170
+    },
+    {
+      "epoch": 0.3329754601226994,
+      "grad_norm": 0.5164998769760132,
+      "learning_rate": 0.00011393421270331814,
+      "loss": 1.3234,
+      "step": 2171
+    },
+    {
+      "epoch": 0.33312883435582824,
+      "grad_norm": 0.2485114485025406,
+      "learning_rate": 0.00011390300997425098,
+      "loss": 1.0826,
+      "step": 2172
+    },
+    {
+      "epoch": 0.3332822085889571,
+      "grad_norm": 0.2219211459159851,
+      "learning_rate": 0.00011387179803028756,
+      "loss": 1.1627,
+      "step": 2173
+    },
+    {
+      "epoch": 0.3334355828220859,
+      "grad_norm": 0.5440203547477722,
+      "learning_rate": 0.00011384057687882109,
+      "loss": 1.2414,
+      "step": 2174
+    },
+    {
+      "epoch": 0.33358895705521474,
+      "grad_norm": 0.23794159293174744,
+      "learning_rate": 0.00011380934652724683,
+      "loss": 1.2738,
+      "step": 2175
+    },
+    {
+      "epoch": 0.3337423312883436,
+      "grad_norm": 0.3872488737106323,
+      "learning_rate": 0.00011377810698296228,
+      "loss": 1.0579,
+      "step": 2176
+    },
+    {
+      "epoch": 0.3338957055214724,
+      "grad_norm": 0.2342870831489563,
+      "learning_rate": 0.00011374685825336706,
+      "loss": 1.1089,
+      "step": 2177
+    },
+    {
+      "epoch": 0.33404907975460124,
+      "grad_norm": 0.3174893856048584,
+      "learning_rate": 0.0001137156003458631,
+      "loss": 1.2354,
+      "step": 2178
+    },
+    {
+      "epoch": 0.3342024539877301,
+      "grad_norm": 0.3542334735393524,
+      "learning_rate": 0.00011368433326785433,
+      "loss": 1.0854,
+      "step": 2179
+    },
+    {
+      "epoch": 0.3343558282208589,
+      "grad_norm": 0.2220126837491989,
+      "learning_rate": 0.00011365305702674696,
+      "loss": 1.1673,
+      "step": 2180
+    },
+    {
+      "epoch": 0.33450920245398774,
+      "grad_norm": 0.27878862619400024,
+      "learning_rate": 0.00011362177162994934,
+      "loss": 1.2068,
+      "step": 2181
+    },
+    {
+      "epoch": 0.3346625766871166,
+      "grad_norm": 0.36227118968963623,
+      "learning_rate": 0.00011359047708487202,
+      "loss": 1.2089,
+      "step": 2182
+    },
+    {
+      "epoch": 0.3348159509202454,
+      "grad_norm": 0.8196086287498474,
+      "learning_rate": 0.00011355917339892765,
+      "loss": 1.1882,
+      "step": 2183
+    },
+    {
+      "epoch": 0.33496932515337424,
+      "grad_norm": 0.278204083442688,
+      "learning_rate": 0.00011352786057953112,
+      "loss": 1.2216,
+      "step": 2184
+    },
+    {
+      "epoch": 0.3351226993865031,
+      "grad_norm": 0.23634231090545654,
+      "learning_rate": 0.00011349653863409948,
+      "loss": 1.1781,
+      "step": 2185
+    },
+    {
+      "epoch": 0.3352760736196319,
+      "grad_norm": 0.2887524962425232,
+      "learning_rate": 0.00011346520757005183,
+      "loss": 1.2033,
+      "step": 2186
+    },
+    {
+      "epoch": 0.33542944785276074,
+      "grad_norm": 0.3041462004184723,
+      "learning_rate": 0.00011343386739480955,
+      "loss": 1.241,
+      "step": 2187
+    },
+    {
+      "epoch": 0.3355828220858896,
+      "grad_norm": 0.24550284445285797,
+      "learning_rate": 0.00011340251811579614,
+      "loss": 1.1406,
+      "step": 2188
+    },
+    {
+      "epoch": 0.3357361963190184,
+      "grad_norm": 0.3352280557155609,
+      "learning_rate": 0.00011337115974043725,
+      "loss": 1.262,
+      "step": 2189
+    },
+    {
+      "epoch": 0.33588957055214724,
+      "grad_norm": 0.545216977596283,
+      "learning_rate": 0.0001133397922761607,
+      "loss": 1.1195,
+      "step": 2190
+    },
+    {
+      "epoch": 0.3360429447852761,
+      "grad_norm": 0.23915894329547882,
+      "learning_rate": 0.00011330841573039643,
+      "loss": 1.1759,
+      "step": 2191
+    },
+    {
+      "epoch": 0.3361963190184049,
+      "grad_norm": 0.2881857752799988,
+      "learning_rate": 0.0001132770301105766,
+      "loss": 1.2719,
+      "step": 2192
+    },
+    {
+      "epoch": 0.33634969325153374,
+      "grad_norm": 0.2310599535703659,
+      "learning_rate": 0.00011324563542413545,
+      "loss": 1.1154,
+      "step": 2193
+    },
+    {
+      "epoch": 0.3365030674846626,
+      "grad_norm": 0.3102778196334839,
+      "learning_rate": 0.00011321423167850936,
+      "loss": 1.1184,
+      "step": 2194
+    },
+    {
+      "epoch": 0.3366564417177914,
+      "grad_norm": 0.22077269852161407,
+      "learning_rate": 0.00011318281888113693,
+      "loss": 1.099,
+      "step": 2195
+    },
+    {
+      "epoch": 0.33680981595092024,
+      "grad_norm": 0.2338707596063614,
+      "learning_rate": 0.00011315139703945883,
+      "loss": 1.3089,
+      "step": 2196
+    },
+    {
+      "epoch": 0.3369631901840491,
+      "grad_norm": 0.21680043637752533,
+      "learning_rate": 0.00011311996616091795,
+      "loss": 1.1574,
+      "step": 2197
+    },
+    {
+      "epoch": 0.3371165644171779,
+      "grad_norm": 0.2374686300754547,
+      "learning_rate": 0.00011308852625295922,
+      "loss": 1.0571,
+      "step": 2198
+    },
+    {
+      "epoch": 0.33726993865030674,
+      "grad_norm": 0.22704748809337616,
+      "learning_rate": 0.00011305707732302979,
+      "loss": 1.214,
+      "step": 2199
+    },
+    {
+      "epoch": 0.3374233128834356,
+      "grad_norm": 0.8966697454452515,
+      "learning_rate": 0.00011302561937857888,
+      "loss": 1.2058,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3375766871165644,
+      "grad_norm": 0.5003755688667297,
+      "learning_rate": 0.00011299415242705798,
+      "loss": 1.2876,
+      "step": 2201
+    },
+    {
+      "epoch": 0.33773006134969324,
+      "grad_norm": 0.256887286901474,
+      "learning_rate": 0.00011296267647592052,
+      "loss": 1.3315,
+      "step": 2202
+    },
+    {
+      "epoch": 0.3378834355828221,
+      "grad_norm": 0.27039989829063416,
+      "learning_rate": 0.00011293119153262221,
+      "loss": 1.203,
+      "step": 2203
+    },
+    {
+      "epoch": 0.3380368098159509,
+      "grad_norm": 0.36862775683403015,
+      "learning_rate": 0.0001128996976046208,
+      "loss": 1.0979,
+      "step": 2204
+    },
+    {
+      "epoch": 0.33819018404907975,
+      "grad_norm": 0.26202648878097534,
+      "learning_rate": 0.0001128681946993763,
+      "loss": 1.2078,
+      "step": 2205
+    },
+    {
+      "epoch": 0.3383435582822086,
+      "grad_norm": 0.3117836117744446,
+      "learning_rate": 0.00011283668282435065,
+      "loss": 1.2517,
+      "step": 2206
+    },
+    {
+      "epoch": 0.3384969325153374,
+      "grad_norm": 0.3130289614200592,
+      "learning_rate": 0.00011280516198700806,
+      "loss": 1.1709,
+      "step": 2207
+    },
+    {
+      "epoch": 0.33865030674846625,
+      "grad_norm": 0.42023149132728577,
+      "learning_rate": 0.00011277363219481485,
+      "loss": 1.1569,
+      "step": 2208
+    },
+    {
+      "epoch": 0.3388036809815951,
+      "grad_norm": 0.2341425120830536,
+      "learning_rate": 0.00011274209345523936,
+      "loss": 1.0775,
+      "step": 2209
+    },
+    {
+      "epoch": 0.3389570552147239,
+      "grad_norm": 0.2323361039161682,
+      "learning_rate": 0.0001127105457757522,
+      "loss": 1.1239,
+      "step": 2210
+    },
+    {
+      "epoch": 0.33911042944785275,
+      "grad_norm": 0.3473592698574066,
+      "learning_rate": 0.00011267898916382598,
+      "loss": 1.2045,
+      "step": 2211
+    },
+    {
+      "epoch": 0.3392638036809816,
+      "grad_norm": 0.6648304462432861,
+      "learning_rate": 0.00011264742362693549,
+      "loss": 1.0885,
+      "step": 2212
+    },
+    {
+      "epoch": 0.3394171779141104,
+      "grad_norm": 0.35814911127090454,
+      "learning_rate": 0.00011261584917255758,
+      "loss": 1.126,
+      "step": 2213
+    },
+    {
+      "epoch": 0.33957055214723925,
+      "grad_norm": 0.5801516771316528,
+      "learning_rate": 0.00011258426580817128,
+      "loss": 1.2135,
+      "step": 2214
+    },
+    {
+      "epoch": 0.3397239263803681,
+      "grad_norm": 0.26346829533576965,
+      "learning_rate": 0.00011255267354125767,
+      "loss": 1.1377,
+      "step": 2215
+    },
+    {
+      "epoch": 0.3398773006134969,
+      "grad_norm": 0.24434280395507812,
+      "learning_rate": 0.00011252107237929994,
+      "loss": 1.3792,
+      "step": 2216
+    },
+    {
+      "epoch": 0.34003067484662575,
+      "grad_norm": 0.3159669041633606,
+      "learning_rate": 0.00011248946232978344,
+      "loss": 1.1062,
+      "step": 2217
+    },
+    {
+      "epoch": 0.3401840490797546,
+      "grad_norm": 0.24887773394584656,
+      "learning_rate": 0.00011245784340019559,
+      "loss": 1.3233,
+      "step": 2218
+    },
+    {
+      "epoch": 0.3403374233128834,
+      "grad_norm": 0.2450355738401413,
+      "learning_rate": 0.00011242621559802592,
+      "loss": 1.1101,
+      "step": 2219
+    },
+    {
+      "epoch": 0.34049079754601225,
+      "grad_norm": 0.23642706871032715,
+      "learning_rate": 0.00011239457893076603,
+      "loss": 1.1409,
+      "step": 2220
+    },
+    {
+      "epoch": 0.3406441717791411,
+      "grad_norm": 0.27718886733055115,
+      "learning_rate": 0.0001123629334059097,
+      "loss": 1.2034,
+      "step": 2221
+    },
+    {
+      "epoch": 0.3407975460122699,
+      "grad_norm": 0.2350548356771469,
+      "learning_rate": 0.00011233127903095271,
+      "loss": 1.1106,
+      "step": 2222
+    },
+    {
+      "epoch": 0.34095092024539875,
+      "grad_norm": 0.2343282848596573,
+      "learning_rate": 0.00011229961581339301,
+      "loss": 1.1099,
+      "step": 2223
+    },
+    {
+      "epoch": 0.3411042944785276,
+      "grad_norm": 0.31704166531562805,
+      "learning_rate": 0.00011226794376073059,
+      "loss": 1.096,
+      "step": 2224
+    },
+    {
+      "epoch": 0.3412576687116564,
+      "grad_norm": 0.5446470379829407,
+      "learning_rate": 0.0001122362628804676,
+      "loss": 0.9997,
+      "step": 2225
+    },
+    {
+      "epoch": 0.3414110429447853,
+      "grad_norm": 0.21466156840324402,
+      "learning_rate": 0.00011220457318010822,
+      "loss": 1.0805,
+      "step": 2226
+    },
+    {
+      "epoch": 0.34156441717791414,
+      "grad_norm": 0.23476915061473846,
+      "learning_rate": 0.00011217287466715874,
+      "loss": 1.2912,
+      "step": 2227
+    },
+    {
+      "epoch": 0.34171779141104297,
+      "grad_norm": 0.21756812930107117,
+      "learning_rate": 0.00011214116734912755,
+      "loss": 1.0257,
+      "step": 2228
+    },
+    {
+      "epoch": 0.3418711656441718,
+      "grad_norm": 0.30071771144866943,
+      "learning_rate": 0.0001121094512335251,
+      "loss": 1.2581,
+      "step": 2229
+    },
+    {
+      "epoch": 0.34202453987730064,
+      "grad_norm": 0.22593587636947632,
+      "learning_rate": 0.00011207772632786392,
+      "loss": 1.09,
+      "step": 2230
+    },
+    {
+      "epoch": 0.34217791411042947,
+      "grad_norm": 0.22183966636657715,
+      "learning_rate": 0.00011204599263965868,
+      "loss": 1.1097,
+      "step": 2231
+    },
+    {
+      "epoch": 0.3423312883435583,
+      "grad_norm": 0.23617132008075714,
+      "learning_rate": 0.00011201425017642605,
+      "loss": 1.1099,
+      "step": 2232
+    },
+    {
+      "epoch": 0.34248466257668714,
+      "grad_norm": 0.41551658511161804,
+      "learning_rate": 0.00011198249894568484,
+      "loss": 1.2472,
+      "step": 2233
+    },
+    {
+      "epoch": 0.34263803680981597,
+      "grad_norm": 0.2525964081287384,
+      "learning_rate": 0.00011195073895495594,
+      "loss": 1.0983,
+      "step": 2234
+    },
+    {
+      "epoch": 0.3427914110429448,
+      "grad_norm": 0.2557397186756134,
+      "learning_rate": 0.00011191897021176226,
+      "loss": 1.0918,
+      "step": 2235
+    },
+    {
+      "epoch": 0.34294478527607364,
+      "grad_norm": 0.3604578375816345,
+      "learning_rate": 0.00011188719272362881,
+      "loss": 1.1612,
+      "step": 2236
+    },
+    {
+      "epoch": 0.34309815950920247,
+      "grad_norm": 0.2730843126773834,
+      "learning_rate": 0.00011185540649808268,
+      "loss": 1.1469,
+      "step": 2237
+    },
+    {
+      "epoch": 0.3432515337423313,
+      "grad_norm": 0.2247643917798996,
+      "learning_rate": 0.00011182361154265301,
+      "loss": 1.055,
+      "step": 2238
+    },
+    {
+      "epoch": 0.34340490797546014,
+      "grad_norm": 0.194431334733963,
+      "learning_rate": 0.00011179180786487107,
+      "loss": 1.118,
+      "step": 2239
+    },
+    {
+      "epoch": 0.34355828220858897,
+      "grad_norm": 0.22817997634410858,
+      "learning_rate": 0.0001117599954722701,
+      "loss": 1.2065,
+      "step": 2240
+    },
+    {
+      "epoch": 0.3437116564417178,
+      "grad_norm": 0.23923316597938538,
+      "learning_rate": 0.00011172817437238548,
+      "loss": 1.1387,
+      "step": 2241
+    },
+    {
+      "epoch": 0.34386503067484664,
+      "grad_norm": 0.26103705167770386,
+      "learning_rate": 0.0001116963445727546,
+      "loss": 1.2836,
+      "step": 2242
+    },
+    {
+      "epoch": 0.3440184049079755,
+      "grad_norm": 0.2259214073419571,
+      "learning_rate": 0.00011166450608091695,
+      "loss": 1.1907,
+      "step": 2243
+    },
+    {
+      "epoch": 0.3441717791411043,
+      "grad_norm": 1.0400322675704956,
+      "learning_rate": 0.00011163265890441408,
+      "loss": 1.2747,
+      "step": 2244
+    },
+    {
+      "epoch": 0.34432515337423314,
+      "grad_norm": 0.6471238136291504,
+      "learning_rate": 0.00011160080305078958,
+      "loss": 1.1422,
+      "step": 2245
+    },
+    {
+      "epoch": 0.344478527607362,
+      "grad_norm": 0.3667556047439575,
+      "learning_rate": 0.00011156893852758906,
+      "loss": 1.0526,
+      "step": 2246
+    },
+    {
+      "epoch": 0.3446319018404908,
+      "grad_norm": 0.282048761844635,
+      "learning_rate": 0.00011153706534236025,
+      "loss": 1.1891,
+      "step": 2247
+    },
+    {
+      "epoch": 0.34478527607361964,
+      "grad_norm": 0.6529294848442078,
+      "learning_rate": 0.00011150518350265289,
+      "loss": 1.0471,
+      "step": 2248
+    },
+    {
+      "epoch": 0.3449386503067485,
+      "grad_norm": 0.9169056415557861,
+      "learning_rate": 0.00011147329301601882,
+      "loss": 0.9738,
+      "step": 2249
+    },
+    {
+      "epoch": 0.3450920245398773,
+      "grad_norm": 0.2730126976966858,
+      "learning_rate": 0.00011144139389001184,
+      "loss": 1.2938,
+      "step": 2250
+    },
+    {
+      "epoch": 0.34524539877300614,
+      "grad_norm": 0.27185800671577454,
+      "learning_rate": 0.00011140948613218788,
+      "loss": 1.1166,
+      "step": 2251
+    },
+    {
+      "epoch": 0.345398773006135,
+      "grad_norm": 0.2661115825176239,
+      "learning_rate": 0.00011137756975010484,
+      "loss": 1.1458,
+      "step": 2252
+    },
+    {
+      "epoch": 0.3455521472392638,
+      "grad_norm": 0.29870855808258057,
+      "learning_rate": 0.00011134564475132275,
+      "loss": 1.198,
+      "step": 2253
+    },
+    {
+      "epoch": 0.34570552147239264,
+      "grad_norm": 0.3498530089855194,
+      "learning_rate": 0.00011131371114340363,
+      "loss": 1.2221,
+      "step": 2254
+    },
+    {
+      "epoch": 0.3458588957055215,
+      "grad_norm": 0.23807935416698456,
+      "learning_rate": 0.0001112817689339115,
+      "loss": 1.2107,
+      "step": 2255
+    },
+    {
+      "epoch": 0.3460122699386503,
+      "grad_norm": 0.2705216407775879,
+      "learning_rate": 0.00011124981813041254,
+      "loss": 1.0816,
+      "step": 2256
+    },
+    {
+      "epoch": 0.34616564417177914,
+      "grad_norm": 0.4178018271923065,
+      "learning_rate": 0.00011121785874047484,
+      "loss": 1.181,
+      "step": 2257
+    },
+    {
+      "epoch": 0.346319018404908,
+      "grad_norm": 0.23916126787662506,
+      "learning_rate": 0.00011118589077166856,
+      "loss": 1.1529,
+      "step": 2258
+    },
+    {
+      "epoch": 0.3464723926380368,
+      "grad_norm": 0.32956087589263916,
+      "learning_rate": 0.00011115391423156591,
+      "loss": 1.1105,
+      "step": 2259
+    },
+    {
+      "epoch": 0.34662576687116564,
+      "grad_norm": 0.24128425121307373,
+      "learning_rate": 0.00011112192912774113,
+      "loss": 1.1768,
+      "step": 2260
+    },
+    {
+      "epoch": 0.3467791411042945,
+      "grad_norm": 0.27682942152023315,
+      "learning_rate": 0.0001110899354677705,
+      "loss": 1.0818,
+      "step": 2261
+    },
+    {
+      "epoch": 0.3469325153374233,
+      "grad_norm": 0.3265593647956848,
+      "learning_rate": 0.0001110579332592323,
+      "loss": 1.1068,
+      "step": 2262
+    },
+    {
+      "epoch": 0.34708588957055214,
+      "grad_norm": 0.26334449648857117,
+      "learning_rate": 0.00011102592250970681,
+      "loss": 1.1622,
+      "step": 2263
+    },
+    {
+      "epoch": 0.347239263803681,
+      "grad_norm": 0.2725241184234619,
+      "learning_rate": 0.00011099390322677641,
+      "loss": 1.0773,
+      "step": 2264
+    },
+    {
+      "epoch": 0.3473926380368098,
+      "grad_norm": 0.7363737225532532,
+      "learning_rate": 0.00011096187541802545,
+      "loss": 1.2028,
+      "step": 2265
+    },
+    {
+      "epoch": 0.34754601226993864,
+      "grad_norm": 0.2223881334066391,
+      "learning_rate": 0.00011092983909104031,
+      "loss": 0.9282,
+      "step": 2266
+    },
+    {
+      "epoch": 0.3476993865030675,
+      "grad_norm": 0.30277490615844727,
+      "learning_rate": 0.00011089779425340936,
+      "loss": 1.132,
+      "step": 2267
+    },
+    {
+      "epoch": 0.3478527607361963,
+      "grad_norm": 0.2252754420042038,
+      "learning_rate": 0.00011086574091272302,
+      "loss": 1.0972,
+      "step": 2268
+    },
+    {
+      "epoch": 0.34800613496932514,
+      "grad_norm": 0.2607453465461731,
+      "learning_rate": 0.00011083367907657373,
+      "loss": 1.1253,
+      "step": 2269
+    },
+    {
+      "epoch": 0.348159509202454,
+      "grad_norm": 0.5650967359542847,
+      "learning_rate": 0.00011080160875255594,
+      "loss": 1.0671,
+      "step": 2270
+    },
+    {
+      "epoch": 0.3483128834355828,
+      "grad_norm": 0.20605497062206268,
+      "learning_rate": 0.00011076952994826608,
+      "loss": 1.1471,
+      "step": 2271
+    },
+    {
+      "epoch": 0.34846625766871164,
+      "grad_norm": 0.21261581778526306,
+      "learning_rate": 0.00011073744267130259,
+      "loss": 1.0512,
+      "step": 2272
+    },
+    {
+      "epoch": 0.3486196319018405,
+      "grad_norm": 0.7565560936927795,
+      "learning_rate": 0.00011070534692926597,
+      "loss": 1.0996,
+      "step": 2273
+    },
+    {
+      "epoch": 0.3487730061349693,
+      "grad_norm": 0.3338976204395294,
+      "learning_rate": 0.00011067324272975863,
+      "loss": 1.1194,
+      "step": 2274
+    },
+    {
+      "epoch": 0.34892638036809814,
+      "grad_norm": 0.2903406322002411,
+      "learning_rate": 0.00011064113008038513,
+      "loss": 1.2109,
+      "step": 2275
+    },
+    {
+      "epoch": 0.349079754601227,
+      "grad_norm": 0.7075420022010803,
+      "learning_rate": 0.00011060900898875187,
+      "loss": 1.2057,
+      "step": 2276
+    },
+    {
+      "epoch": 0.3492331288343558,
+      "grad_norm": 0.9775124788284302,
+      "learning_rate": 0.00011057687946246736,
+      "loss": 1.2541,
+      "step": 2277
+    },
+    {
+      "epoch": 0.34938650306748464,
+      "grad_norm": 0.2291848510503769,
+      "learning_rate": 0.00011054474150914207,
+      "loss": 1.0569,
+      "step": 2278
+    },
+    {
+      "epoch": 0.3495398773006135,
+      "grad_norm": 0.4416504204273224,
+      "learning_rate": 0.00011051259513638846,
+      "loss": 1.1534,
+      "step": 2279
+    },
+    {
+      "epoch": 0.3496932515337423,
+      "grad_norm": 0.555528461933136,
+      "learning_rate": 0.00011048044035182099,
+      "loss": 1.1115,
+      "step": 2280
+    },
+    {
+      "epoch": 0.34984662576687114,
+      "grad_norm": 0.23718573153018951,
+      "learning_rate": 0.0001104482771630561,
+      "loss": 1.1522,
+      "step": 2281
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.23635195195674896,
+      "learning_rate": 0.00011041610557771224,
+      "loss": 1.1228,
+      "step": 2282
+    },
+    {
+      "epoch": 0.3501533742331288,
+      "grad_norm": 0.28144773840904236,
+      "learning_rate": 0.00011038392560340988,
+      "loss": 1.2034,
+      "step": 2283
+    },
+    {
+      "epoch": 0.35030674846625764,
+      "grad_norm": 0.2782566249370575,
+      "learning_rate": 0.00011035173724777142,
+      "loss": 1.1983,
+      "step": 2284
+    },
+    {
+      "epoch": 0.3504601226993865,
+      "grad_norm": 0.3447224795818329,
+      "learning_rate": 0.00011031954051842124,
+      "loss": 1.0516,
+      "step": 2285
+    },
+    {
+      "epoch": 0.3506134969325153,
+      "grad_norm": 0.25099799036979675,
+      "learning_rate": 0.00011028733542298577,
+      "loss": 1.0884,
+      "step": 2286
+    },
+    {
+      "epoch": 0.35076687116564415,
+      "grad_norm": 0.21801820397377014,
+      "learning_rate": 0.00011025512196909336,
+      "loss": 1.0611,
+      "step": 2287
+    },
+    {
+      "epoch": 0.350920245398773,
+      "grad_norm": 0.2553553879261017,
+      "learning_rate": 0.00011022290016437436,
+      "loss": 1.1408,
+      "step": 2288
+    },
+    {
+      "epoch": 0.35107361963190187,
+      "grad_norm": 0.22888065874576569,
+      "learning_rate": 0.00011019067001646111,
+      "loss": 1.0815,
+      "step": 2289
+    },
+    {
+      "epoch": 0.3512269938650307,
+      "grad_norm": 0.4823405146598816,
+      "learning_rate": 0.00011015843153298792,
+      "loss": 1.1311,
+      "step": 2290
+    },
+    {
+      "epoch": 0.35138036809815953,
+      "grad_norm": 0.21930362284183502,
+      "learning_rate": 0.00011012618472159105,
+      "loss": 1.2593,
+      "step": 2291
+    },
+    {
+      "epoch": 0.35153374233128837,
+      "grad_norm": 0.21783937513828278,
+      "learning_rate": 0.00011009392958990876,
+      "loss": 1.1709,
+      "step": 2292
+    },
+    {
+      "epoch": 0.3516871165644172,
+      "grad_norm": 0.33965542912483215,
+      "learning_rate": 0.00011006166614558128,
+      "loss": 1.1889,
+      "step": 2293
+    },
+    {
+      "epoch": 0.35184049079754603,
+      "grad_norm": 0.25767749547958374,
+      "learning_rate": 0.00011002939439625079,
+      "loss": 1.0906,
+      "step": 2294
+    },
+    {
+      "epoch": 0.35199386503067487,
+      "grad_norm": 0.24223719537258148,
+      "learning_rate": 0.00010999711434956146,
+      "loss": 1.1151,
+      "step": 2295
+    },
+    {
+      "epoch": 0.3521472392638037,
+      "grad_norm": 0.7662160992622375,
+      "learning_rate": 0.00010996482601315944,
+      "loss": 1.2423,
+      "step": 2296
+    },
+    {
+      "epoch": 0.35230061349693254,
+      "grad_norm": 0.2097785472869873,
+      "learning_rate": 0.00010993252939469277,
+      "loss": 1.0873,
+      "step": 2297
+    },
+    {
+      "epoch": 0.35245398773006137,
+      "grad_norm": 0.205241858959198,
+      "learning_rate": 0.00010990022450181152,
+      "loss": 1.0392,
+      "step": 2298
+    },
+    {
+      "epoch": 0.3526073619631902,
+      "grad_norm": 0.5106040835380554,
+      "learning_rate": 0.00010986791134216773,
+      "loss": 1.1468,
+      "step": 2299
+    },
+    {
+      "epoch": 0.35276073619631904,
+      "grad_norm": 0.2589695155620575,
+      "learning_rate": 0.0001098355899234153,
+      "loss": 1.1606,
+      "step": 2300
+    },
+    {
+      "epoch": 0.35291411042944787,
+      "grad_norm": 0.4777984619140625,
+      "learning_rate": 0.00010980326025321022,
+      "loss": 1.2022,
+      "step": 2301
+    },
+    {
+      "epoch": 0.3530674846625767,
+      "grad_norm": 1.2484053373336792,
+      "learning_rate": 0.00010977092233921035,
+      "loss": 1.1595,
+      "step": 2302
+    },
+    {
+      "epoch": 0.35322085889570554,
+      "grad_norm": 0.22485268115997314,
+      "learning_rate": 0.0001097385761890755,
+      "loss": 1.1195,
+      "step": 2303
+    },
+    {
+      "epoch": 0.35337423312883437,
+      "grad_norm": 0.24357753992080688,
+      "learning_rate": 0.00010970622181046748,
+      "loss": 1.1575,
+      "step": 2304
+    },
+    {
+      "epoch": 0.3535276073619632,
+      "grad_norm": 0.4232584238052368,
+      "learning_rate": 0.00010967385921105001,
+      "loss": 1.211,
+      "step": 2305
+    },
+    {
+      "epoch": 0.35368098159509204,
+      "grad_norm": 0.4186762273311615,
+      "learning_rate": 0.00010964148839848878,
+      "loss": 1.0164,
+      "step": 2306
+    },
+    {
+      "epoch": 0.35383435582822087,
+      "grad_norm": 0.6382477283477783,
+      "learning_rate": 0.00010960910938045139,
+      "loss": 1.2695,
+      "step": 2307
+    },
+    {
+      "epoch": 0.3539877300613497,
+      "grad_norm": 0.29954084753990173,
+      "learning_rate": 0.00010957672216460745,
+      "loss": 0.9354,
+      "step": 2308
+    },
+    {
+      "epoch": 0.35414110429447854,
+      "grad_norm": 0.2907813787460327,
+      "learning_rate": 0.00010954432675862844,
+      "loss": 1.1604,
+      "step": 2309
+    },
+    {
+      "epoch": 0.35429447852760737,
+      "grad_norm": 0.5316698551177979,
+      "learning_rate": 0.00010951192317018778,
+      "loss": 1.1549,
+      "step": 2310
+    },
+    {
+      "epoch": 0.3544478527607362,
+      "grad_norm": 0.2894304394721985,
+      "learning_rate": 0.00010947951140696093,
+      "loss": 1.0423,
+      "step": 2311
+    },
+    {
+      "epoch": 0.35460122699386504,
+      "grad_norm": 0.2792621850967407,
+      "learning_rate": 0.00010944709147662517,
+      "loss": 1.1995,
+      "step": 2312
+    },
+    {
+      "epoch": 0.35475460122699387,
+      "grad_norm": 0.32874566316604614,
+      "learning_rate": 0.00010941466338685977,
+      "loss": 1.2657,
+      "step": 2313
+    },
+    {
+      "epoch": 0.3549079754601227,
+      "grad_norm": 0.30107027292251587,
+      "learning_rate": 0.00010938222714534591,
+      "loss": 1.1186,
+      "step": 2314
+    },
+    {
+      "epoch": 0.35506134969325154,
+      "grad_norm": 0.2845155894756317,
+      "learning_rate": 0.00010934978275976674,
+      "loss": 1.0654,
+      "step": 2315
+    },
+    {
+      "epoch": 0.35521472392638037,
+      "grad_norm": 0.4581781029701233,
+      "learning_rate": 0.00010931733023780726,
+      "loss": 0.99,
+      "step": 2316
+    },
+    {
+      "epoch": 0.3553680981595092,
+      "grad_norm": 0.363207072019577,
+      "learning_rate": 0.00010928486958715449,
+      "loss": 1.1752,
+      "step": 2317
+    },
+    {
+      "epoch": 0.35552147239263804,
+      "grad_norm": 0.2895558178424835,
+      "learning_rate": 0.00010925240081549733,
+      "loss": 1.1412,
+      "step": 2318
+    },
+    {
+      "epoch": 0.35567484662576687,
+      "grad_norm": 0.23707033693790436,
+      "learning_rate": 0.00010921992393052661,
+      "loss": 0.9902,
+      "step": 2319
+    },
+    {
+      "epoch": 0.3558282208588957,
+      "grad_norm": 0.2475333958864212,
+      "learning_rate": 0.0001091874389399351,
+      "loss": 1.0924,
+      "step": 2320
+    },
+    {
+      "epoch": 0.35598159509202454,
+      "grad_norm": 0.25761526823043823,
+      "learning_rate": 0.0001091549458514174,
+      "loss": 1.1867,
+      "step": 2321
+    },
+    {
+      "epoch": 0.35613496932515337,
+      "grad_norm": 0.39130645990371704,
+      "learning_rate": 0.00010912244467267018,
+      "loss": 1.0915,
+      "step": 2322
+    },
+    {
+      "epoch": 0.3562883435582822,
+      "grad_norm": 0.24919815361499786,
+      "learning_rate": 0.00010908993541139187,
+      "loss": 1.227,
+      "step": 2323
+    },
+    {
+      "epoch": 0.35644171779141104,
+      "grad_norm": 0.8676552176475525,
+      "learning_rate": 0.00010905741807528294,
+      "loss": 1.2009,
+      "step": 2324
+    },
+    {
+      "epoch": 0.3565950920245399,
+      "grad_norm": 0.21798589825630188,
+      "learning_rate": 0.00010902489267204572,
+      "loss": 1.2042,
+      "step": 2325
+    },
+    {
+      "epoch": 0.3567484662576687,
+      "grad_norm": 0.2806626856327057,
+      "learning_rate": 0.00010899235920938442,
+      "loss": 1.0452,
+      "step": 2326
+    },
+    {
+      "epoch": 0.35690184049079754,
+      "grad_norm": 0.38935405015945435,
+      "learning_rate": 0.00010895981769500521,
+      "loss": 1.1202,
+      "step": 2327
+    },
+    {
+      "epoch": 0.3570552147239264,
+      "grad_norm": 0.2573637068271637,
+      "learning_rate": 0.00010892726813661618,
+      "loss": 1.1521,
+      "step": 2328
+    },
+    {
+      "epoch": 0.3572085889570552,
+      "grad_norm": 0.4005904793739319,
+      "learning_rate": 0.00010889471054192722,
+      "loss": 1.0597,
+      "step": 2329
+    },
+    {
+      "epoch": 0.35736196319018404,
+      "grad_norm": 0.24706833064556122,
+      "learning_rate": 0.00010886214491865027,
+      "loss": 1.2027,
+      "step": 2330
+    },
+    {
+      "epoch": 0.3575153374233129,
+      "grad_norm": 1.003148078918457,
+      "learning_rate": 0.00010882957127449906,
+      "loss": 1.2043,
+      "step": 2331
+    },
+    {
+      "epoch": 0.3576687116564417,
+      "grad_norm": 0.34944143891334534,
+      "learning_rate": 0.00010879698961718927,
+      "loss": 1.1197,
+      "step": 2332
+    },
+    {
+      "epoch": 0.35782208588957054,
+      "grad_norm": 0.38279879093170166,
+      "learning_rate": 0.00010876439995443847,
+      "loss": 1.0073,
+      "step": 2333
+    },
+    {
+      "epoch": 0.3579754601226994,
+      "grad_norm": 0.4149549603462219,
+      "learning_rate": 0.00010873180229396611,
+      "loss": 1.1555,
+      "step": 2334
+    },
+    {
+      "epoch": 0.3581288343558282,
+      "grad_norm": 0.7360826730728149,
+      "learning_rate": 0.00010869919664349356,
+      "loss": 1.0831,
+      "step": 2335
+    },
+    {
+      "epoch": 0.35828220858895704,
+      "grad_norm": 0.4775683283805847,
+      "learning_rate": 0.00010866658301074407,
+      "loss": 1.2336,
+      "step": 2336
+    },
+    {
+      "epoch": 0.3584355828220859,
+      "grad_norm": 0.35014936327934265,
+      "learning_rate": 0.00010863396140344277,
+      "loss": 1.1187,
+      "step": 2337
+    },
+    {
+      "epoch": 0.3585889570552147,
+      "grad_norm": 0.25096985697746277,
+      "learning_rate": 0.0001086013318293167,
+      "loss": 1.1579,
+      "step": 2338
+    },
+    {
+      "epoch": 0.35874233128834354,
+      "grad_norm": 0.27783241868019104,
+      "learning_rate": 0.00010856869429609477,
+      "loss": 1.0126,
+      "step": 2339
+    },
+    {
+      "epoch": 0.3588957055214724,
+      "grad_norm": 0.2635081708431244,
+      "learning_rate": 0.0001085360488115078,
+      "loss": 0.95,
+      "step": 2340
+    },
+    {
+      "epoch": 0.3590490797546012,
+      "grad_norm": 0.26063308119773865,
+      "learning_rate": 0.00010850339538328846,
+      "loss": 1.1315,
+      "step": 2341
+    },
+    {
+      "epoch": 0.35920245398773004,
+      "grad_norm": 0.40031543374061584,
+      "learning_rate": 0.00010847073401917133,
+      "loss": 1.0493,
+      "step": 2342
+    },
+    {
+      "epoch": 0.3593558282208589,
+      "grad_norm": 0.2540077269077301,
+      "learning_rate": 0.00010843806472689284,
+      "loss": 1.2074,
+      "step": 2343
+    },
+    {
+      "epoch": 0.3595092024539877,
+      "grad_norm": 0.27282342314720154,
+      "learning_rate": 0.00010840538751419131,
+      "loss": 1.2953,
+      "step": 2344
+    },
+    {
+      "epoch": 0.35966257668711654,
+      "grad_norm": 0.2451404184103012,
+      "learning_rate": 0.00010837270238880696,
+      "loss": 1.2988,
+      "step": 2345
+    },
+    {
+      "epoch": 0.3598159509202454,
+      "grad_norm": 0.22721104323863983,
+      "learning_rate": 0.00010834000935848187,
+      "loss": 1.2065,
+      "step": 2346
+    },
+    {
+      "epoch": 0.3599693251533742,
+      "grad_norm": 0.6175394058227539,
+      "learning_rate": 0.00010830730843095998,
+      "loss": 1.1169,
+      "step": 2347
+    },
+    {
+      "epoch": 0.36012269938650304,
+      "grad_norm": 0.22079049050807953,
+      "learning_rate": 0.00010827459961398708,
+      "loss": 1.0999,
+      "step": 2348
+    },
+    {
+      "epoch": 0.3602760736196319,
+      "grad_norm": 0.2357230931520462,
+      "learning_rate": 0.0001082418829153109,
+      "loss": 1.2432,
+      "step": 2349
+    },
+    {
+      "epoch": 0.3604294478527607,
+      "grad_norm": 0.4917770326137543,
+      "learning_rate": 0.00010820915834268097,
+      "loss": 1.065,
+      "step": 2350
+    },
+    {
+      "epoch": 0.3605828220858896,
+      "grad_norm": 0.2788459360599518,
+      "learning_rate": 0.00010817642590384873,
+      "loss": 1.2455,
+      "step": 2351
+    },
+    {
+      "epoch": 0.36073619631901843,
+      "grad_norm": 0.22051605582237244,
+      "learning_rate": 0.00010814368560656745,
+      "loss": 0.9501,
+      "step": 2352
+    },
+    {
+      "epoch": 0.36088957055214727,
+      "grad_norm": 0.2543274760246277,
+      "learning_rate": 0.00010811093745859226,
+      "loss": 1.0099,
+      "step": 2353
+    },
+    {
+      "epoch": 0.3610429447852761,
+      "grad_norm": 1.0231744050979614,
+      "learning_rate": 0.00010807818146768017,
+      "loss": 1.1137,
+      "step": 2354
+    },
+    {
+      "epoch": 0.36119631901840493,
+      "grad_norm": 0.23485730588436127,
+      "learning_rate": 0.0001080454176415901,
+      "loss": 1.0826,
+      "step": 2355
+    },
+    {
+      "epoch": 0.36134969325153377,
+      "grad_norm": 0.24545204639434814,
+      "learning_rate": 0.00010801264598808266,
+      "loss": 1.2597,
+      "step": 2356
+    },
+    {
+      "epoch": 0.3615030674846626,
+      "grad_norm": 0.5202247500419617,
+      "learning_rate": 0.0001079798665149205,
+      "loss": 1.0561,
+      "step": 2357
+    },
+    {
+      "epoch": 0.36165644171779143,
+      "grad_norm": 0.2407538741827011,
+      "learning_rate": 0.000107947079229868,
+      "loss": 1.2175,
+      "step": 2358
+    },
+    {
+      "epoch": 0.36180981595092027,
+      "grad_norm": 0.23571135103702545,
+      "learning_rate": 0.00010791428414069147,
+      "loss": 1.1579,
+      "step": 2359
+    },
+    {
+      "epoch": 0.3619631901840491,
+      "grad_norm": 0.2599712014198303,
+      "learning_rate": 0.000107881481255159,
+      "loss": 1.1878,
+      "step": 2360
+    },
+    {
+      "epoch": 0.36211656441717793,
+      "grad_norm": 0.24012036621570587,
+      "learning_rate": 0.00010784867058104058,
+      "loss": 1.1205,
+      "step": 2361
+    },
+    {
+      "epoch": 0.36226993865030677,
+      "grad_norm": 0.4671995937824249,
+      "learning_rate": 0.00010781585212610803,
+      "loss": 1.2162,
+      "step": 2362
+    },
+    {
+      "epoch": 0.3624233128834356,
+      "grad_norm": 0.5224483609199524,
+      "learning_rate": 0.000107783025898135,
+      "loss": 1.2441,
+      "step": 2363
+    },
+    {
+      "epoch": 0.36257668711656443,
+      "grad_norm": 0.21108107268810272,
+      "learning_rate": 0.00010775019190489697,
+      "loss": 1.0665,
+      "step": 2364
+    },
+    {
+      "epoch": 0.36273006134969327,
+      "grad_norm": 0.21228383481502533,
+      "learning_rate": 0.00010771735015417131,
+      "loss": 1.1701,
+      "step": 2365
+    },
+    {
+      "epoch": 0.3628834355828221,
+      "grad_norm": 0.23914079368114471,
+      "learning_rate": 0.00010768450065373714,
+      "loss": 1.2096,
+      "step": 2366
+    },
+    {
+      "epoch": 0.36303680981595093,
+      "grad_norm": 0.3105505406856537,
+      "learning_rate": 0.00010765164341137555,
+      "loss": 1.214,
+      "step": 2367
+    },
+    {
+      "epoch": 0.36319018404907977,
+      "grad_norm": 0.2572573721408844,
+      "learning_rate": 0.00010761877843486931,
+      "loss": 1.2505,
+      "step": 2368
+    },
+    {
+      "epoch": 0.3633435582822086,
+      "grad_norm": 0.4450485408306122,
+      "learning_rate": 0.00010758590573200313,
+      "loss": 1.1265,
+      "step": 2369
+    },
+    {
+      "epoch": 0.36349693251533743,
+      "grad_norm": 0.21891994774341583,
+      "learning_rate": 0.00010755302531056353,
+      "loss": 1.0264,
+      "step": 2370
+    },
+    {
+      "epoch": 0.36365030674846627,
+      "grad_norm": 0.25320807099342346,
+      "learning_rate": 0.00010752013717833882,
+      "loss": 1.1304,
+      "step": 2371
+    },
+    {
+      "epoch": 0.3638036809815951,
+      "grad_norm": 0.23188531398773193,
+      "learning_rate": 0.00010748724134311918,
+      "loss": 1.3276,
+      "step": 2372
+    },
+    {
+      "epoch": 0.36395705521472393,
+      "grad_norm": 0.20284503698349,
+      "learning_rate": 0.00010745433781269659,
+      "loss": 1.0206,
+      "step": 2373
+    },
+    {
+      "epoch": 0.36411042944785277,
+      "grad_norm": 0.2091296762228012,
+      "learning_rate": 0.00010742142659486483,
+      "loss": 1.0975,
+      "step": 2374
+    },
+    {
+      "epoch": 0.3642638036809816,
+      "grad_norm": 0.3131512403488159,
+      "learning_rate": 0.0001073885076974196,
+      "loss": 1.0975,
+      "step": 2375
+    },
+    {
+      "epoch": 0.36441717791411044,
+      "grad_norm": 0.21189506351947784,
+      "learning_rate": 0.00010735558112815827,
+      "loss": 0.9972,
+      "step": 2376
+    },
+    {
+      "epoch": 0.36457055214723927,
+      "grad_norm": 0.37394455075263977,
+      "learning_rate": 0.00010732264689488017,
+      "loss": 1.2058,
+      "step": 2377
+    },
+    {
+      "epoch": 0.3647239263803681,
+      "grad_norm": 0.21865034103393555,
+      "learning_rate": 0.00010728970500538633,
+      "loss": 1.2004,
+      "step": 2378
+    },
+    {
+      "epoch": 0.36487730061349694,
+      "grad_norm": 0.22142834961414337,
+      "learning_rate": 0.00010725675546747965,
+      "loss": 1.1127,
+      "step": 2379
+    },
+    {
+      "epoch": 0.36503067484662577,
+      "grad_norm": 0.2051433026790619,
+      "learning_rate": 0.00010722379828896488,
+      "loss": 0.9635,
+      "step": 2380
+    },
+    {
+      "epoch": 0.3651840490797546,
+      "grad_norm": 0.2567044198513031,
+      "learning_rate": 0.0001071908334776485,
+      "loss": 1.0328,
+      "step": 2381
+    },
+    {
+      "epoch": 0.36533742331288344,
+      "grad_norm": 0.9264582991600037,
+      "learning_rate": 0.00010715786104133883,
+      "loss": 1.1769,
+      "step": 2382
+    },
+    {
+      "epoch": 0.36549079754601227,
+      "grad_norm": 0.30112287402153015,
+      "learning_rate": 0.00010712488098784603,
+      "loss": 1.1825,
+      "step": 2383
+    },
+    {
+      "epoch": 0.3656441717791411,
+      "grad_norm": 0.3013554811477661,
+      "learning_rate": 0.00010709189332498203,
+      "loss": 1.1722,
+      "step": 2384
+    },
+    {
+      "epoch": 0.36579754601226994,
+      "grad_norm": 0.2167431116104126,
+      "learning_rate": 0.00010705889806056053,
+      "loss": 1.0135,
+      "step": 2385
+    },
+    {
+      "epoch": 0.36595092024539877,
+      "grad_norm": 0.35778719186782837,
+      "learning_rate": 0.0001070258952023971,
+      "loss": 1.2006,
+      "step": 2386
+    },
+    {
+      "epoch": 0.3661042944785276,
+      "grad_norm": 0.33251339197158813,
+      "learning_rate": 0.00010699288475830908,
+      "loss": 1.1228,
+      "step": 2387
+    },
+    {
+      "epoch": 0.36625766871165644,
+      "grad_norm": 0.2222868949174881,
+      "learning_rate": 0.00010695986673611557,
+      "loss": 1.2246,
+      "step": 2388
+    },
+    {
+      "epoch": 0.36641104294478527,
+      "grad_norm": 0.21383428573608398,
+      "learning_rate": 0.00010692684114363754,
+      "loss": 1.0055,
+      "step": 2389
+    },
+    {
+      "epoch": 0.3665644171779141,
+      "grad_norm": 0.2361694574356079,
+      "learning_rate": 0.00010689380798869767,
+      "loss": 1.2691,
+      "step": 2390
+    },
+    {
+      "epoch": 0.36671779141104294,
+      "grad_norm": 0.6031351089477539,
+      "learning_rate": 0.00010686076727912052,
+      "loss": 1.0159,
+      "step": 2391
+    },
+    {
+      "epoch": 0.36687116564417177,
+      "grad_norm": 0.24755312502384186,
+      "learning_rate": 0.00010682771902273234,
+      "loss": 1.0991,
+      "step": 2392
+    },
+    {
+      "epoch": 0.3670245398773006,
+      "grad_norm": 0.20430293679237366,
+      "learning_rate": 0.00010679466322736128,
+      "loss": 1.0981,
+      "step": 2393
+    },
+    {
+      "epoch": 0.36717791411042944,
+      "grad_norm": 0.5178117156028748,
+      "learning_rate": 0.00010676159990083715,
+      "loss": 1.0368,
+      "step": 2394
+    },
+    {
+      "epoch": 0.36733128834355827,
+      "grad_norm": 0.6060917377471924,
+      "learning_rate": 0.00010672852905099167,
+      "loss": 1.0908,
+      "step": 2395
+    },
+    {
+      "epoch": 0.3674846625766871,
+      "grad_norm": 0.27103808522224426,
+      "learning_rate": 0.00010669545068565822,
+      "loss": 1.175,
+      "step": 2396
+    },
+    {
+      "epoch": 0.36763803680981594,
+      "grad_norm": 0.2330983579158783,
+      "learning_rate": 0.00010666236481267209,
+      "loss": 1.09,
+      "step": 2397
+    },
+    {
+      "epoch": 0.36779141104294477,
+      "grad_norm": 0.23016589879989624,
+      "learning_rate": 0.00010662927143987023,
+      "loss": 1.1387,
+      "step": 2398
+    },
+    {
+      "epoch": 0.3679447852760736,
+      "grad_norm": 0.36958545446395874,
+      "learning_rate": 0.00010659617057509141,
+      "loss": 1.1632,
+      "step": 2399
+    },
+    {
+      "epoch": 0.36809815950920244,
+      "grad_norm": 0.2648904323577881,
+      "learning_rate": 0.00010656306222617623,
+      "loss": 1.0351,
+      "step": 2400
+    },
+    {
+      "epoch": 0.36825153374233127,
+      "grad_norm": 0.2524549663066864,
+      "learning_rate": 0.00010652994640096698,
+      "loss": 1.1966,
+      "step": 2401
+    },
+    {
+      "epoch": 0.3684049079754601,
+      "grad_norm": 0.26206475496292114,
+      "learning_rate": 0.00010649682310730774,
+      "loss": 1.1642,
+      "step": 2402
+    },
+    {
+      "epoch": 0.36855828220858894,
+      "grad_norm": 0.23221324384212494,
+      "learning_rate": 0.0001064636923530444,
+      "loss": 1.261,
+      "step": 2403
+    },
+    {
+      "epoch": 0.36871165644171777,
+      "grad_norm": 0.45974791049957275,
+      "learning_rate": 0.0001064305541460246,
+      "loss": 1.206,
+      "step": 2404
+    },
+    {
+      "epoch": 0.3688650306748466,
+      "grad_norm": 0.2335430085659027,
+      "learning_rate": 0.0001063974084940977,
+      "loss": 1.1444,
+      "step": 2405
+    },
+    {
+      "epoch": 0.36901840490797544,
+      "grad_norm": 0.2885693311691284,
+      "learning_rate": 0.0001063642554051149,
+      "loss": 1.1802,
+      "step": 2406
+    },
+    {
+      "epoch": 0.3691717791411043,
+      "grad_norm": 0.20670290291309357,
+      "learning_rate": 0.00010633109488692909,
+      "loss": 1.2268,
+      "step": 2407
+    },
+    {
+      "epoch": 0.3693251533742331,
+      "grad_norm": 0.25239109992980957,
+      "learning_rate": 0.00010629792694739496,
+      "loss": 1.1452,
+      "step": 2408
+    },
+    {
+      "epoch": 0.36947852760736194,
+      "grad_norm": 0.609539806842804,
+      "learning_rate": 0.00010626475159436895,
+      "loss": 1.0598,
+      "step": 2409
+    },
+    {
+      "epoch": 0.3696319018404908,
+      "grad_norm": 0.2023230642080307,
+      "learning_rate": 0.00010623156883570926,
+      "loss": 1.1704,
+      "step": 2410
+    },
+    {
+      "epoch": 0.3697852760736196,
+      "grad_norm": 0.3003746271133423,
+      "learning_rate": 0.00010619837867927585,
+      "loss": 1.0031,
+      "step": 2411
+    },
+    {
+      "epoch": 0.36993865030674844,
+      "grad_norm": 0.2194560468196869,
+      "learning_rate": 0.00010616518113293038,
+      "loss": 1.1061,
+      "step": 2412
+    },
+    {
+      "epoch": 0.3700920245398773,
+      "grad_norm": 0.3135942220687866,
+      "learning_rate": 0.00010613197620453635,
+      "loss": 1.1165,
+      "step": 2413
+    },
+    {
+      "epoch": 0.37024539877300616,
+      "grad_norm": 0.2114485800266266,
+      "learning_rate": 0.00010609876390195893,
+      "loss": 1.1853,
+      "step": 2414
+    },
+    {
+      "epoch": 0.370398773006135,
+      "grad_norm": 0.45604079961776733,
+      "learning_rate": 0.00010606554423306506,
+      "loss": 1.2511,
+      "step": 2415
+    },
+    {
+      "epoch": 0.37055214723926383,
+      "grad_norm": 0.2684125304222107,
+      "learning_rate": 0.00010603231720572347,
+      "loss": 1.1348,
+      "step": 2416
+    },
+    {
+      "epoch": 0.37070552147239266,
+      "grad_norm": 0.2191537469625473,
+      "learning_rate": 0.00010599908282780456,
+      "loss": 1.1392,
+      "step": 2417
+    },
+    {
+      "epoch": 0.3708588957055215,
+      "grad_norm": 1.2727528810501099,
+      "learning_rate": 0.00010596584110718053,
+      "loss": 1.0308,
+      "step": 2418
+    },
+    {
+      "epoch": 0.37101226993865033,
+      "grad_norm": 0.22624953091144562,
+      "learning_rate": 0.00010593259205172529,
+      "loss": 1.2544,
+      "step": 2419
+    },
+    {
+      "epoch": 0.37116564417177916,
+      "grad_norm": 0.2011311948299408,
+      "learning_rate": 0.00010589933566931447,
+      "loss": 1.196,
+      "step": 2420
+    },
+    {
+      "epoch": 0.371319018404908,
+      "grad_norm": 0.6138514876365662,
+      "learning_rate": 0.00010586607196782549,
+      "loss": 1.1113,
+      "step": 2421
+    },
+    {
+      "epoch": 0.37147239263803683,
+      "grad_norm": 0.23382601141929626,
+      "learning_rate": 0.00010583280095513746,
+      "loss": 1.1079,
+      "step": 2422
+    },
+    {
+      "epoch": 0.37162576687116566,
+      "grad_norm": 0.2276948243379593,
+      "learning_rate": 0.0001057995226391312,
+      "loss": 1.2196,
+      "step": 2423
+    },
+    {
+      "epoch": 0.3717791411042945,
+      "grad_norm": 0.2566922605037689,
+      "learning_rate": 0.00010576623702768935,
+      "loss": 1.1699,
+      "step": 2424
+    },
+    {
+      "epoch": 0.37193251533742333,
+      "grad_norm": 0.38378238677978516,
+      "learning_rate": 0.00010573294412869617,
+      "loss": 1.0831,
+      "step": 2425
+    },
+    {
+      "epoch": 0.37208588957055216,
+      "grad_norm": 0.3740137815475464,
+      "learning_rate": 0.00010569964395003776,
+      "loss": 1.231,
+      "step": 2426
+    },
+    {
+      "epoch": 0.372239263803681,
+      "grad_norm": 0.35522225499153137,
+      "learning_rate": 0.00010566633649960184,
+      "loss": 1.0241,
+      "step": 2427
+    },
+    {
+      "epoch": 0.37239263803680983,
+      "grad_norm": 0.39372220635414124,
+      "learning_rate": 0.00010563302178527788,
+      "loss": 1.1238,
+      "step": 2428
+    },
+    {
+      "epoch": 0.37254601226993866,
+      "grad_norm": 0.3515973687171936,
+      "learning_rate": 0.00010559969981495711,
+      "loss": 1.1394,
+      "step": 2429
+    },
+    {
+      "epoch": 0.3726993865030675,
+      "grad_norm": 0.23316068947315216,
+      "learning_rate": 0.00010556637059653245,
+      "loss": 1.1385,
+      "step": 2430
+    },
+    {
+      "epoch": 0.37285276073619633,
+      "grad_norm": 0.26693201065063477,
+      "learning_rate": 0.00010553303413789856,
+      "loss": 1.102,
+      "step": 2431
+    },
+    {
+      "epoch": 0.37300613496932516,
+      "grad_norm": 0.27642327547073364,
+      "learning_rate": 0.00010549969044695176,
+      "loss": 1.2536,
+      "step": 2432
+    },
+    {
+      "epoch": 0.373159509202454,
+      "grad_norm": 0.2356724590063095,
+      "learning_rate": 0.00010546633953159014,
+      "loss": 1.1728,
+      "step": 2433
+    },
+    {
+      "epoch": 0.37331288343558283,
+      "grad_norm": 0.20566748082637787,
+      "learning_rate": 0.00010543298139971347,
+      "loss": 1.2171,
+      "step": 2434
+    },
+    {
+      "epoch": 0.37346625766871167,
+      "grad_norm": 0.19501802325248718,
+      "learning_rate": 0.00010539961605922323,
+      "loss": 1.0791,
+      "step": 2435
+    },
+    {
+      "epoch": 0.3736196319018405,
+      "grad_norm": 0.3327271640300751,
+      "learning_rate": 0.00010536624351802265,
+      "loss": 1.0485,
+      "step": 2436
+    },
+    {
+      "epoch": 0.37377300613496933,
+      "grad_norm": 0.3412679433822632,
+      "learning_rate": 0.0001053328637840166,
+      "loss": 1.028,
+      "step": 2437
+    },
+    {
+      "epoch": 0.37392638036809817,
+      "grad_norm": 0.22047950327396393,
+      "learning_rate": 0.00010529947686511171,
+      "loss": 1.2375,
+      "step": 2438
+    },
+    {
+      "epoch": 0.374079754601227,
+      "grad_norm": 0.24975819885730743,
+      "learning_rate": 0.00010526608276921628,
+      "loss": 1.1618,
+      "step": 2439
+    },
+    {
+      "epoch": 0.37423312883435583,
+      "grad_norm": 0.23064488172531128,
+      "learning_rate": 0.00010523268150424031,
+      "loss": 1.1422,
+      "step": 2440
+    },
+    {
+      "epoch": 0.37438650306748467,
+      "grad_norm": 0.24046337604522705,
+      "learning_rate": 0.00010519927307809554,
+      "loss": 1.0981,
+      "step": 2441
+    },
+    {
+      "epoch": 0.3745398773006135,
+      "grad_norm": 0.37612104415893555,
+      "learning_rate": 0.00010516585749869533,
+      "loss": 1.1141,
+      "step": 2442
+    },
+    {
+      "epoch": 0.37469325153374233,
+      "grad_norm": 0.24391721189022064,
+      "learning_rate": 0.00010513243477395479,
+      "loss": 1.1246,
+      "step": 2443
+    },
+    {
+      "epoch": 0.37484662576687117,
+      "grad_norm": 0.6591930389404297,
+      "learning_rate": 0.0001050990049117907,
+      "loss": 1.1204,
+      "step": 2444
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 0.2129911631345749,
+      "learning_rate": 0.00010506556792012157,
+      "loss": 1.0779,
+      "step": 2445
+    },
+    {
+      "epoch": 0.37515337423312883,
+      "grad_norm": 0.3146311938762665,
+      "learning_rate": 0.00010503212380686753,
+      "loss": 1.1517,
+      "step": 2446
+    },
+    {
+      "epoch": 0.37530674846625767,
+      "grad_norm": 0.3550461232662201,
+      "learning_rate": 0.00010499867257995047,
+      "loss": 1.1322,
+      "step": 2447
+    },
+    {
+      "epoch": 0.3754601226993865,
+      "grad_norm": 0.21718236804008484,
+      "learning_rate": 0.00010496521424729393,
+      "loss": 1.022,
+      "step": 2448
+    },
+    {
+      "epoch": 0.37561349693251533,
+      "grad_norm": 0.23840586841106415,
+      "learning_rate": 0.00010493174881682311,
+      "loss": 1.0609,
+      "step": 2449
+    },
+    {
+      "epoch": 0.37576687116564417,
+      "grad_norm": 0.20586925745010376,
+      "learning_rate": 0.00010489827629646493,
+      "loss": 1.0704,
+      "step": 2450
+    },
+    {
+      "epoch": 0.375920245398773,
+      "grad_norm": 0.22544316947460175,
+      "learning_rate": 0.00010486479669414797,
+      "loss": 1.1531,
+      "step": 2451
+    },
+    {
+      "epoch": 0.37607361963190183,
+      "grad_norm": 0.20023788511753082,
+      "learning_rate": 0.00010483131001780247,
+      "loss": 1.1045,
+      "step": 2452
+    },
+    {
+      "epoch": 0.37622699386503067,
+      "grad_norm": 0.207672581076622,
+      "learning_rate": 0.00010479781627536042,
+      "loss": 1.2653,
+      "step": 2453
+    },
+    {
+      "epoch": 0.3763803680981595,
+      "grad_norm": 0.2023344337940216,
+      "learning_rate": 0.00010476431547475538,
+      "loss": 1.1009,
+      "step": 2454
+    },
+    {
+      "epoch": 0.37653374233128833,
+      "grad_norm": 0.2181950956583023,
+      "learning_rate": 0.00010473080762392266,
+      "loss": 1.0574,
+      "step": 2455
+    },
+    {
+      "epoch": 0.37668711656441717,
+      "grad_norm": 0.22382469475269318,
+      "learning_rate": 0.00010469729273079923,
+      "loss": 1.12,
+      "step": 2456
+    },
+    {
+      "epoch": 0.376840490797546,
+      "grad_norm": 0.208431214094162,
+      "learning_rate": 0.00010466377080332367,
+      "loss": 1.1686,
+      "step": 2457
+    },
+    {
+      "epoch": 0.37699386503067484,
+      "grad_norm": 0.23381951451301575,
+      "learning_rate": 0.00010463024184943629,
+      "loss": 1.1746,
+      "step": 2458
+    },
+    {
+      "epoch": 0.37714723926380367,
+      "grad_norm": 0.21349789202213287,
+      "learning_rate": 0.00010459670587707904,
+      "loss": 1.2023,
+      "step": 2459
+    },
+    {
+      "epoch": 0.3773006134969325,
+      "grad_norm": 0.2327684611082077,
+      "learning_rate": 0.00010456316289419552,
+      "loss": 1.0425,
+      "step": 2460
+    },
+    {
+      "epoch": 0.37745398773006134,
+      "grad_norm": 0.20930907130241394,
+      "learning_rate": 0.00010452961290873103,
+      "loss": 0.9639,
+      "step": 2461
+    },
+    {
+      "epoch": 0.37760736196319017,
+      "grad_norm": 0.3498322367668152,
+      "learning_rate": 0.0001044960559286325,
+      "loss": 1.1969,
+      "step": 2462
+    },
+    {
+      "epoch": 0.377760736196319,
+      "grad_norm": 0.2019176334142685,
+      "learning_rate": 0.00010446249196184851,
+      "loss": 1.0999,
+      "step": 2463
+    },
+    {
+      "epoch": 0.37791411042944784,
+      "grad_norm": 0.2592250406742096,
+      "learning_rate": 0.00010442892101632931,
+      "loss": 1.0881,
+      "step": 2464
+    },
+    {
+      "epoch": 0.37806748466257667,
+      "grad_norm": 0.20402561128139496,
+      "learning_rate": 0.00010439534310002678,
+      "loss": 1.0645,
+      "step": 2465
+    },
+    {
+      "epoch": 0.3782208588957055,
+      "grad_norm": 0.19624631106853485,
+      "learning_rate": 0.00010436175822089448,
+      "loss": 1.0158,
+      "step": 2466
+    },
+    {
+      "epoch": 0.37837423312883434,
+      "grad_norm": 0.3445425033569336,
+      "learning_rate": 0.00010432816638688763,
+      "loss": 1.1539,
+      "step": 2467
+    },
+    {
+      "epoch": 0.37852760736196317,
+      "grad_norm": 0.20075830817222595,
+      "learning_rate": 0.00010429456760596306,
+      "loss": 1.059,
+      "step": 2468
+    },
+    {
+      "epoch": 0.378680981595092,
+      "grad_norm": 0.1938285529613495,
+      "learning_rate": 0.00010426096188607923,
+      "loss": 1.0903,
+      "step": 2469
+    },
+    {
+      "epoch": 0.37883435582822084,
+      "grad_norm": 1.6145378351211548,
+      "learning_rate": 0.00010422734923519635,
+      "loss": 1.1232,
+      "step": 2470
+    },
+    {
+      "epoch": 0.37898773006134967,
+      "grad_norm": 0.19625777006149292,
+      "learning_rate": 0.00010419372966127613,
+      "loss": 1.098,
+      "step": 2471
+    },
+    {
+      "epoch": 0.3791411042944785,
+      "grad_norm": 0.3624955117702484,
+      "learning_rate": 0.00010416010317228202,
+      "loss": 1.0498,
+      "step": 2472
+    },
+    {
+      "epoch": 0.37929447852760734,
+      "grad_norm": 0.4879778027534485,
+      "learning_rate": 0.00010412646977617904,
+      "loss": 1.3205,
+      "step": 2473
+    },
+    {
+      "epoch": 0.37944785276073617,
+      "grad_norm": 0.38369467854499817,
+      "learning_rate": 0.00010409282948093394,
+      "loss": 1.1786,
+      "step": 2474
+    },
+    {
+      "epoch": 0.379601226993865,
+      "grad_norm": 0.2914874851703644,
+      "learning_rate": 0.000104059182294515,
+      "loss": 1.1625,
+      "step": 2475
+    },
+    {
+      "epoch": 0.37975460122699384,
+      "grad_norm": 0.2553163468837738,
+      "learning_rate": 0.00010402552822489217,
+      "loss": 1.1032,
+      "step": 2476
+    },
+    {
+      "epoch": 0.3799079754601227,
+      "grad_norm": 0.2143537551164627,
+      "learning_rate": 0.00010399186728003707,
+      "loss": 1.1607,
+      "step": 2477
+    },
+    {
+      "epoch": 0.38006134969325156,
+      "grad_norm": 0.21552975475788116,
+      "learning_rate": 0.0001039581994679229,
+      "loss": 1.1814,
+      "step": 2478
+    },
+    {
+      "epoch": 0.3802147239263804,
+      "grad_norm": 0.24259217083454132,
+      "learning_rate": 0.0001039245247965245,
+      "loss": 1.2295,
+      "step": 2479
+    },
+    {
+      "epoch": 0.3803680981595092,
+      "grad_norm": 0.26586195826530457,
+      "learning_rate": 0.00010389084327381835,
+      "loss": 1.0345,
+      "step": 2480
+    },
+    {
+      "epoch": 0.38052147239263806,
+      "grad_norm": 0.2664244771003723,
+      "learning_rate": 0.00010385715490778254,
+      "loss": 1.2816,
+      "step": 2481
+    },
+    {
+      "epoch": 0.3806748466257669,
+      "grad_norm": 0.1904495358467102,
+      "learning_rate": 0.00010382345970639677,
+      "loss": 1.0189,
+      "step": 2482
+    },
+    {
+      "epoch": 0.3808282208588957,
+      "grad_norm": 0.5123971700668335,
+      "learning_rate": 0.00010378975767764239,
+      "loss": 1.1675,
+      "step": 2483
+    },
+    {
+      "epoch": 0.38098159509202456,
+      "grad_norm": 0.29069262742996216,
+      "learning_rate": 0.0001037560488295023,
+      "loss": 1.1105,
+      "step": 2484
+    },
+    {
+      "epoch": 0.3811349693251534,
+      "grad_norm": 0.28372979164123535,
+      "learning_rate": 0.00010372233316996112,
+      "loss": 1.1118,
+      "step": 2485
+    },
+    {
+      "epoch": 0.38128834355828223,
+      "grad_norm": 0.20762629806995392,
+      "learning_rate": 0.000103688610707005,
+      "loss": 1.1775,
+      "step": 2486
+    },
+    {
+      "epoch": 0.38144171779141106,
+      "grad_norm": 0.22006504237651825,
+      "learning_rate": 0.00010365488144862171,
+      "loss": 1.0363,
+      "step": 2487
+    },
+    {
+      "epoch": 0.3815950920245399,
+      "grad_norm": 0.24446511268615723,
+      "learning_rate": 0.0001036211454028007,
+      "loss": 1.2037,
+      "step": 2488
+    },
+    {
+      "epoch": 0.38174846625766873,
+      "grad_norm": 0.24875061213970184,
+      "learning_rate": 0.00010358740257753292,
+      "loss": 1.0079,
+      "step": 2489
+    },
+    {
+      "epoch": 0.38190184049079756,
+      "grad_norm": 0.21152977645397186,
+      "learning_rate": 0.00010355365298081101,
+      "loss": 0.9341,
+      "step": 2490
+    },
+    {
+      "epoch": 0.3820552147239264,
+      "grad_norm": 0.34620460867881775,
+      "learning_rate": 0.00010351989662062922,
+      "loss": 1.0608,
+      "step": 2491
+    },
+    {
+      "epoch": 0.38220858895705523,
+      "grad_norm": 1.0460968017578125,
+      "learning_rate": 0.00010348613350498328,
+      "loss": 1.1823,
+      "step": 2492
+    },
+    {
+      "epoch": 0.38236196319018406,
+      "grad_norm": 0.24160589277744293,
+      "learning_rate": 0.00010345236364187067,
+      "loss": 1.2354,
+      "step": 2493
+    },
+    {
+      "epoch": 0.3825153374233129,
+      "grad_norm": 0.3539619445800781,
+      "learning_rate": 0.00010341858703929039,
+      "loss": 1.1566,
+      "step": 2494
+    },
+    {
+      "epoch": 0.38266871165644173,
+      "grad_norm": 0.20056772232055664,
+      "learning_rate": 0.00010338480370524305,
+      "loss": 1.1339,
+      "step": 2495
+    },
+    {
+      "epoch": 0.38282208588957056,
+      "grad_norm": 0.41644981503486633,
+      "learning_rate": 0.00010335101364773085,
+      "loss": 1.0038,
+      "step": 2496
+    },
+    {
+      "epoch": 0.3829754601226994,
+      "grad_norm": 0.2545158565044403,
+      "learning_rate": 0.0001033172168747576,
+      "loss": 1.1231,
+      "step": 2497
+    },
+    {
+      "epoch": 0.38312883435582823,
+      "grad_norm": 0.20312146842479706,
+      "learning_rate": 0.00010328341339432869,
+      "loss": 1.1533,
+      "step": 2498
+    },
+    {
+      "epoch": 0.38328220858895706,
+      "grad_norm": 0.20917072892189026,
+      "learning_rate": 0.00010324960321445111,
+      "loss": 1.1232,
+      "step": 2499
+    },
+    {
+      "epoch": 0.3834355828220859,
+      "grad_norm": 0.2931499779224396,
+      "learning_rate": 0.00010321578634313338,
+      "loss": 0.9588,
+      "step": 2500
+    },
+    {
+      "epoch": 0.38358895705521473,
+      "grad_norm": 0.21562887728214264,
+      "learning_rate": 0.00010318196278838572,
+      "loss": 1.0976,
+      "step": 2501
+    },
+    {
+      "epoch": 0.38374233128834356,
+      "grad_norm": 0.2033701092004776,
+      "learning_rate": 0.00010314813255821981,
+      "loss": 1.092,
+      "step": 2502
+    },
+    {
+      "epoch": 0.3838957055214724,
+      "grad_norm": 0.8089235424995422,
+      "learning_rate": 0.00010311429566064899,
+      "loss": 1.0993,
+      "step": 2503
+    },
+    {
+      "epoch": 0.38404907975460123,
+      "grad_norm": 0.22584927082061768,
+      "learning_rate": 0.00010308045210368811,
+      "loss": 1.1967,
+      "step": 2504
+    },
+    {
+      "epoch": 0.38420245398773006,
+      "grad_norm": 0.3506816327571869,
+      "learning_rate": 0.00010304660189535371,
+      "loss": 1.16,
+      "step": 2505
+    },
+    {
+      "epoch": 0.3843558282208589,
+      "grad_norm": 0.3503791391849518,
+      "learning_rate": 0.0001030127450436638,
+      "loss": 1.162,
+      "step": 2506
+    },
+    {
+      "epoch": 0.38450920245398773,
+      "grad_norm": 0.2595718801021576,
+      "learning_rate": 0.00010297888155663798,
+      "loss": 1.2325,
+      "step": 2507
+    },
+    {
+      "epoch": 0.38466257668711656,
+      "grad_norm": 0.25214067101478577,
+      "learning_rate": 0.00010294501144229747,
+      "loss": 1.1338,
+      "step": 2508
+    },
+    {
+      "epoch": 0.3848159509202454,
+      "grad_norm": 0.42905083298683167,
+      "learning_rate": 0.00010291113470866504,
+      "loss": 1.1716,
+      "step": 2509
+    },
+    {
+      "epoch": 0.38496932515337423,
+      "grad_norm": 0.4055924713611603,
+      "learning_rate": 0.000102877251363765,
+      "loss": 1.2032,
+      "step": 2510
+    },
+    {
+      "epoch": 0.38512269938650306,
+      "grad_norm": 0.22197411954402924,
+      "learning_rate": 0.00010284336141562324,
+      "loss": 1.132,
+      "step": 2511
+    },
+    {
+      "epoch": 0.3852760736196319,
+      "grad_norm": 0.7938892245292664,
+      "learning_rate": 0.00010280946487226727,
+      "loss": 1.0865,
+      "step": 2512
+    },
+    {
+      "epoch": 0.38542944785276073,
+      "grad_norm": 0.8538442254066467,
+      "learning_rate": 0.00010277556174172603,
+      "loss": 1.1154,
+      "step": 2513
+    },
+    {
+      "epoch": 0.38558282208588956,
+      "grad_norm": 0.2417086809873581,
+      "learning_rate": 0.00010274165203203017,
+      "loss": 1.1999,
+      "step": 2514
+    },
+    {
+      "epoch": 0.3857361963190184,
+      "grad_norm": 0.239262193441391,
+      "learning_rate": 0.0001027077357512118,
+      "loss": 0.9941,
+      "step": 2515
+    },
+    {
+      "epoch": 0.38588957055214723,
+      "grad_norm": 0.3122091293334961,
+      "learning_rate": 0.00010267381290730463,
+      "loss": 1.0188,
+      "step": 2516
+    },
+    {
+      "epoch": 0.38604294478527607,
+      "grad_norm": 0.22704201936721802,
+      "learning_rate": 0.0001026398835083439,
+      "loss": 1.2273,
+      "step": 2517
+    },
+    {
+      "epoch": 0.3861963190184049,
+      "grad_norm": 0.2519001066684723,
+      "learning_rate": 0.00010260594756236642,
+      "loss": 0.955,
+      "step": 2518
+    },
+    {
+      "epoch": 0.38634969325153373,
+      "grad_norm": 0.6293007731437683,
+      "learning_rate": 0.00010257200507741056,
+      "loss": 1.0494,
+      "step": 2519
+    },
+    {
+      "epoch": 0.38650306748466257,
+      "grad_norm": 0.24627865850925446,
+      "learning_rate": 0.0001025380560615162,
+      "loss": 1.0903,
+      "step": 2520
+    },
+    {
+      "epoch": 0.3866564417177914,
+      "grad_norm": 0.38064128160476685,
+      "learning_rate": 0.0001025041005227248,
+      "loss": 1.1386,
+      "step": 2521
+    },
+    {
+      "epoch": 0.38680981595092023,
+      "grad_norm": 0.25124678015708923,
+      "learning_rate": 0.00010247013846907937,
+      "loss": 1.2359,
+      "step": 2522
+    },
+    {
+      "epoch": 0.38696319018404907,
+      "grad_norm": 0.4061981737613678,
+      "learning_rate": 0.00010243616990862442,
+      "loss": 0.9908,
+      "step": 2523
+    },
+    {
+      "epoch": 0.3871165644171779,
+      "grad_norm": 0.24852539598941803,
+      "learning_rate": 0.00010240219484940606,
+      "loss": 1.1119,
+      "step": 2524
+    },
+    {
+      "epoch": 0.38726993865030673,
+      "grad_norm": 0.39001208543777466,
+      "learning_rate": 0.00010236821329947191,
+      "loss": 1.1683,
+      "step": 2525
+    },
+    {
+      "epoch": 0.38742331288343557,
+      "grad_norm": 0.2443695068359375,
+      "learning_rate": 0.00010233422526687112,
+      "loss": 1.1663,
+      "step": 2526
+    },
+    {
+      "epoch": 0.3875766871165644,
+      "grad_norm": 0.2508279085159302,
+      "learning_rate": 0.00010230023075965437,
+      "loss": 1.1062,
+      "step": 2527
+    },
+    {
+      "epoch": 0.38773006134969323,
+      "grad_norm": 0.42127323150634766,
+      "learning_rate": 0.00010226622978587388,
+      "loss": 1.0816,
+      "step": 2528
+    },
+    {
+      "epoch": 0.38788343558282207,
+      "grad_norm": 0.2057909220457077,
+      "learning_rate": 0.00010223222235358346,
+      "loss": 1.1239,
+      "step": 2529
+    },
+    {
+      "epoch": 0.3880368098159509,
+      "grad_norm": 0.2186688333749771,
+      "learning_rate": 0.00010219820847083832,
+      "loss": 1.2328,
+      "step": 2530
+    },
+    {
+      "epoch": 0.38819018404907973,
+      "grad_norm": 0.2101743370294571,
+      "learning_rate": 0.00010216418814569535,
+      "loss": 1.1342,
+      "step": 2531
+    },
+    {
+      "epoch": 0.38834355828220857,
+      "grad_norm": 0.2073497772216797,
+      "learning_rate": 0.00010213016138621282,
+      "loss": 1.0839,
+      "step": 2532
+    },
+    {
+      "epoch": 0.3884969325153374,
+      "grad_norm": 0.22619086503982544,
+      "learning_rate": 0.00010209612820045068,
+      "loss": 1.1438,
+      "step": 2533
+    },
+    {
+      "epoch": 0.38865030674846623,
+      "grad_norm": 0.198466494679451,
+      "learning_rate": 0.00010206208859647029,
+      "loss": 1.1573,
+      "step": 2534
+    },
+    {
+      "epoch": 0.38880368098159507,
+      "grad_norm": 0.2047959715127945,
+      "learning_rate": 0.0001020280425823345,
+      "loss": 1.1479,
+      "step": 2535
+    },
+    {
+      "epoch": 0.3889570552147239,
+      "grad_norm": 0.25286662578582764,
+      "learning_rate": 0.00010199399016610781,
+      "loss": 1.1716,
+      "step": 2536
+    },
+    {
+      "epoch": 0.38911042944785273,
+      "grad_norm": 0.3163773715496063,
+      "learning_rate": 0.00010195993135585613,
+      "loss": 1.1562,
+      "step": 2537
+    },
+    {
+      "epoch": 0.38926380368098157,
+      "grad_norm": 0.2863243818283081,
+      "learning_rate": 0.0001019258661596469,
+      "loss": 1.1632,
+      "step": 2538
+    },
+    {
+      "epoch": 0.3894171779141104,
+      "grad_norm": 0.5595969557762146,
+      "learning_rate": 0.00010189179458554912,
+      "loss": 1.0903,
+      "step": 2539
+    },
+    {
+      "epoch": 0.3895705521472393,
+      "grad_norm": 0.21528899669647217,
+      "learning_rate": 0.00010185771664163331,
+      "loss": 1.2339,
+      "step": 2540
+    },
+    {
+      "epoch": 0.3897239263803681,
+      "grad_norm": 0.21200229227542877,
+      "learning_rate": 0.00010182363233597139,
+      "loss": 1.0888,
+      "step": 2541
+    },
+    {
+      "epoch": 0.38987730061349696,
+      "grad_norm": 1.4303981065750122,
+      "learning_rate": 0.00010178954167663688,
+      "loss": 1.1556,
+      "step": 2542
+    },
+    {
+      "epoch": 0.3900306748466258,
+      "grad_norm": 0.20898601412773132,
+      "learning_rate": 0.00010175544467170476,
+      "loss": 1.1433,
+      "step": 2543
+    },
+    {
+      "epoch": 0.3901840490797546,
+      "grad_norm": 0.22791609168052673,
+      "learning_rate": 0.0001017213413292516,
+      "loss": 1.1695,
+      "step": 2544
+    },
+    {
+      "epoch": 0.39033742331288346,
+      "grad_norm": 0.5367591977119446,
+      "learning_rate": 0.00010168723165735534,
+      "loss": 1.1323,
+      "step": 2545
+    },
+    {
+      "epoch": 0.3904907975460123,
+      "grad_norm": 0.299679160118103,
+      "learning_rate": 0.00010165311566409553,
+      "loss": 1.1393,
+      "step": 2546
+    },
+    {
+      "epoch": 0.3906441717791411,
+      "grad_norm": 0.3442368507385254,
+      "learning_rate": 0.00010161899335755316,
+      "loss": 1.148,
+      "step": 2547
+    },
+    {
+      "epoch": 0.39079754601226996,
+      "grad_norm": 0.40139004588127136,
+      "learning_rate": 0.0001015848647458107,
+      "loss": 1.1334,
+      "step": 2548
+    },
+    {
+      "epoch": 0.3909509202453988,
+      "grad_norm": 0.40697404742240906,
+      "learning_rate": 0.00010155072983695217,
+      "loss": 1.1512,
+      "step": 2549
+    },
+    {
+      "epoch": 0.3911042944785276,
+      "grad_norm": 0.3360229432582855,
+      "learning_rate": 0.00010151658863906307,
+      "loss": 1.2077,
+      "step": 2550
+    },
+    {
+      "epoch": 0.39125766871165646,
+      "grad_norm": 0.21205496788024902,
+      "learning_rate": 0.00010148244116023034,
+      "loss": 1.1698,
+      "step": 2551
+    },
+    {
+      "epoch": 0.3914110429447853,
+      "grad_norm": 0.18613575398921967,
+      "learning_rate": 0.00010144828740854242,
+      "loss": 0.8826,
+      "step": 2552
+    },
+    {
+      "epoch": 0.3915644171779141,
+      "grad_norm": 0.24769316613674164,
+      "learning_rate": 0.00010141412739208934,
+      "loss": 1.1587,
+      "step": 2553
+    },
+    {
+      "epoch": 0.39171779141104296,
+      "grad_norm": 0.2050507366657257,
+      "learning_rate": 0.00010137996111896245,
+      "loss": 1.1539,
+      "step": 2554
+    },
+    {
+      "epoch": 0.3918711656441718,
+      "grad_norm": 0.23550301790237427,
+      "learning_rate": 0.00010134578859725472,
+      "loss": 1.2832,
+      "step": 2555
+    },
+    {
+      "epoch": 0.3920245398773006,
+      "grad_norm": 0.22266478836536407,
+      "learning_rate": 0.00010131160983506048,
+      "loss": 1.1759,
+      "step": 2556
+    },
+    {
+      "epoch": 0.39217791411042946,
+      "grad_norm": 0.22980253398418427,
+      "learning_rate": 0.00010127742484047565,
+      "loss": 1.1146,
+      "step": 2557
+    },
+    {
+      "epoch": 0.3923312883435583,
+      "grad_norm": 0.2689424753189087,
+      "learning_rate": 0.00010124323362159756,
+      "loss": 1.2068,
+      "step": 2558
+    },
+    {
+      "epoch": 0.3924846625766871,
+      "grad_norm": 0.19971536099910736,
+      "learning_rate": 0.00010120903618652504,
+      "loss": 1.1767,
+      "step": 2559
+    },
+    {
+      "epoch": 0.39263803680981596,
+      "grad_norm": 0.21566547453403473,
+      "learning_rate": 0.00010117483254335837,
+      "loss": 1.155,
+      "step": 2560
+    },
+    {
+      "epoch": 0.3927914110429448,
+      "grad_norm": 0.26533767580986023,
+      "learning_rate": 0.00010114062270019934,
+      "loss": 1.1121,
+      "step": 2561
+    },
+    {
+      "epoch": 0.3929447852760736,
+      "grad_norm": 0.21605005860328674,
+      "learning_rate": 0.00010110640666515116,
+      "loss": 1.209,
+      "step": 2562
+    },
+    {
+      "epoch": 0.39309815950920246,
+      "grad_norm": 0.2533816695213318,
+      "learning_rate": 0.00010107218444631853,
+      "loss": 0.9826,
+      "step": 2563
+    },
+    {
+      "epoch": 0.3932515337423313,
+      "grad_norm": 0.6976297497749329,
+      "learning_rate": 0.00010103795605180763,
+      "loss": 1.2278,
+      "step": 2564
+    },
+    {
+      "epoch": 0.3934049079754601,
+      "grad_norm": 0.19673821330070496,
+      "learning_rate": 0.00010100372148972606,
+      "loss": 1.0275,
+      "step": 2565
+    },
+    {
+      "epoch": 0.39355828220858896,
+      "grad_norm": 0.28093674778938293,
+      "learning_rate": 0.00010096948076818294,
+      "loss": 1.1814,
+      "step": 2566
+    },
+    {
+      "epoch": 0.3937116564417178,
+      "grad_norm": 0.19864977896213531,
+      "learning_rate": 0.00010093523389528879,
+      "loss": 1.2092,
+      "step": 2567
+    },
+    {
+      "epoch": 0.39386503067484663,
+      "grad_norm": 0.2037159651517868,
+      "learning_rate": 0.00010090098087915565,
+      "loss": 1.3279,
+      "step": 2568
+    },
+    {
+      "epoch": 0.39401840490797546,
+      "grad_norm": 0.19845017790794373,
+      "learning_rate": 0.00010086672172789695,
+      "loss": 1.0575,
+      "step": 2569
+    },
+    {
+      "epoch": 0.3941717791411043,
+      "grad_norm": 0.3801080882549286,
+      "learning_rate": 0.00010083245644962758,
+      "loss": 1.1524,
+      "step": 2570
+    },
+    {
+      "epoch": 0.39432515337423313,
+      "grad_norm": 0.2034035474061966,
+      "learning_rate": 0.00010079818505246396,
+      "loss": 1.1915,
+      "step": 2571
+    },
+    {
+      "epoch": 0.39447852760736196,
+      "grad_norm": 0.2249310165643692,
+      "learning_rate": 0.00010076390754452386,
+      "loss": 1.0421,
+      "step": 2572
+    },
+    {
+      "epoch": 0.3946319018404908,
+      "grad_norm": 0.19774840772151947,
+      "learning_rate": 0.00010072962393392657,
+      "loss": 1.0348,
+      "step": 2573
+    },
+    {
+      "epoch": 0.39478527607361963,
+      "grad_norm": 0.20836445689201355,
+      "learning_rate": 0.00010069533422879276,
+      "loss": 1.1014,
+      "step": 2574
+    },
+    {
+      "epoch": 0.39493865030674846,
+      "grad_norm": 0.18617203831672668,
+      "learning_rate": 0.00010066103843724463,
+      "loss": 1.1041,
+      "step": 2575
+    },
+    {
+      "epoch": 0.3950920245398773,
+      "grad_norm": 0.20640195906162262,
+      "learning_rate": 0.00010062673656740574,
+      "loss": 1.2171,
+      "step": 2576
+    },
+    {
+      "epoch": 0.39524539877300613,
+      "grad_norm": 0.19124098122119904,
+      "learning_rate": 0.00010059242862740113,
+      "loss": 1.0887,
+      "step": 2577
+    },
+    {
+      "epoch": 0.39539877300613496,
+      "grad_norm": 0.7065445184707642,
+      "learning_rate": 0.00010055811462535726,
+      "loss": 1.1381,
+      "step": 2578
+    },
+    {
+      "epoch": 0.3955521472392638,
+      "grad_norm": 0.5006560683250427,
+      "learning_rate": 0.00010052379456940203,
+      "loss": 1.0244,
+      "step": 2579
+    },
+    {
+      "epoch": 0.39570552147239263,
+      "grad_norm": 0.27687016129493713,
+      "learning_rate": 0.00010048946846766477,
+      "loss": 1.1609,
+      "step": 2580
+    },
+    {
+      "epoch": 0.39585889570552146,
+      "grad_norm": 0.20640812814235687,
+      "learning_rate": 0.00010045513632827629,
+      "loss": 1.1969,
+      "step": 2581
+    },
+    {
+      "epoch": 0.3960122699386503,
+      "grad_norm": 0.20042988657951355,
+      "learning_rate": 0.00010042079815936878,
+      "loss": 0.9063,
+      "step": 2582
+    },
+    {
+      "epoch": 0.39616564417177913,
+      "grad_norm": 0.2937188446521759,
+      "learning_rate": 0.00010038645396907583,
+      "loss": 1.1693,
+      "step": 2583
+    },
+    {
+      "epoch": 0.39631901840490796,
+      "grad_norm": 0.2234763354063034,
+      "learning_rate": 0.00010035210376553255,
+      "loss": 1.151,
+      "step": 2584
+    },
+    {
+      "epoch": 0.3964723926380368,
+      "grad_norm": 0.3138175308704376,
+      "learning_rate": 0.00010031774755687538,
+      "loss": 1.2583,
+      "step": 2585
+    },
+    {
+      "epoch": 0.39662576687116563,
+      "grad_norm": 0.20175500214099884,
+      "learning_rate": 0.00010028338535124224,
+      "loss": 1.0824,
+      "step": 2586
+    },
+    {
+      "epoch": 0.39677914110429446,
+      "grad_norm": 0.19755424559116364,
+      "learning_rate": 0.00010024901715677246,
+      "loss": 1.1094,
+      "step": 2587
+    },
+    {
+      "epoch": 0.3969325153374233,
+      "grad_norm": 0.2038612961769104,
+      "learning_rate": 0.00010021464298160676,
+      "loss": 1.0412,
+      "step": 2588
+    },
+    {
+      "epoch": 0.39708588957055213,
+      "grad_norm": 0.2092968374490738,
+      "learning_rate": 0.00010018026283388732,
+      "loss": 1.1841,
+      "step": 2589
+    },
+    {
+      "epoch": 0.39723926380368096,
+      "grad_norm": 0.19677473604679108,
+      "learning_rate": 0.00010014587672175772,
+      "loss": 1.0564,
+      "step": 2590
+    },
+    {
+      "epoch": 0.3973926380368098,
+      "grad_norm": 0.24636343121528625,
+      "learning_rate": 0.00010011148465336291,
+      "loss": 1.1429,
+      "step": 2591
+    },
+    {
+      "epoch": 0.39754601226993863,
+      "grad_norm": 0.20779484510421753,
+      "learning_rate": 0.00010007708663684929,
+      "loss": 1.1533,
+      "step": 2592
+    },
+    {
+      "epoch": 0.39769938650306746,
+      "grad_norm": 0.20610220730304718,
+      "learning_rate": 0.00010004268268036469,
+      "loss": 1.2307,
+      "step": 2593
+    },
+    {
+      "epoch": 0.3978527607361963,
+      "grad_norm": 0.19073477387428284,
+      "learning_rate": 0.00010000827279205829,
+      "loss": 1.1753,
+      "step": 2594
+    },
+    {
+      "epoch": 0.39800613496932513,
+      "grad_norm": 0.3707866966724396,
+      "learning_rate": 9.997385698008076e-05,
+      "loss": 0.9588,
+      "step": 2595
+    },
+    {
+      "epoch": 0.39815950920245397,
+      "grad_norm": 0.19675074517726898,
+      "learning_rate": 9.993943525258408e-05,
+      "loss": 1.2104,
+      "step": 2596
+    },
+    {
+      "epoch": 0.3983128834355828,
+      "grad_norm": 0.2467436045408249,
+      "learning_rate": 9.990500761772169e-05,
+      "loss": 1.2352,
+      "step": 2597
+    },
+    {
+      "epoch": 0.39846625766871163,
+      "grad_norm": 0.2740110456943512,
+      "learning_rate": 9.987057408364841e-05,
+      "loss": 1.12,
+      "step": 2598
+    },
+    {
+      "epoch": 0.39861963190184047,
+      "grad_norm": 0.22025059163570404,
+      "learning_rate": 9.983613465852043e-05,
+      "loss": 1.1427,
+      "step": 2599
+    },
+    {
+      "epoch": 0.3987730061349693,
+      "grad_norm": 0.18945367634296417,
+      "learning_rate": 9.980168935049539e-05,
+      "loss": 1.1219,
+      "step": 2600
+    },
+    {
+      "epoch": 0.39892638036809813,
+      "grad_norm": 0.2149667590856552,
+      "learning_rate": 9.976723816773229e-05,
+      "loss": 1.1765,
+      "step": 2601
+    },
+    {
+      "epoch": 0.399079754601227,
+      "grad_norm": 0.24212020635604858,
+      "learning_rate": 9.973278111839153e-05,
+      "loss": 1.0609,
+      "step": 2602
+    },
+    {
+      "epoch": 0.39923312883435585,
+      "grad_norm": 0.19352123141288757,
+      "learning_rate": 9.969831821063492e-05,
+      "loss": 0.9852,
+      "step": 2603
+    },
+    {
+      "epoch": 0.3993865030674847,
+      "grad_norm": 0.5595475435256958,
+      "learning_rate": 9.966384945262559e-05,
+      "loss": 1.1124,
+      "step": 2604
+    },
+    {
+      "epoch": 0.3995398773006135,
+      "grad_norm": 0.26186299324035645,
+      "learning_rate": 9.962937485252812e-05,
+      "loss": 1.1314,
+      "step": 2605
+    },
+    {
+      "epoch": 0.39969325153374236,
+      "grad_norm": 0.4155636727809906,
+      "learning_rate": 9.959489441850848e-05,
+      "loss": 1.1608,
+      "step": 2606
+    },
+    {
+      "epoch": 0.3998466257668712,
+      "grad_norm": 0.19638502597808838,
+      "learning_rate": 9.956040815873397e-05,
+      "loss": 1.0406,
+      "step": 2607
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8691263198852539,
+      "learning_rate": 9.952591608137331e-05,
+      "loss": 1.0142,
+      "step": 2608
+    },
+    {
+      "epoch": 0.40015337423312886,
+      "grad_norm": 0.23327504098415375,
+      "learning_rate": 9.94914181945966e-05,
+      "loss": 1.1162,
+      "step": 2609
+    },
+    {
+      "epoch": 0.4003067484662577,
+      "grad_norm": 0.20776480436325073,
+      "learning_rate": 9.945691450657528e-05,
+      "loss": 1.2065,
+      "step": 2610
+    },
+    {
+      "epoch": 0.4004601226993865,
+      "grad_norm": 0.20773537456989288,
+      "learning_rate": 9.942240502548219e-05,
+      "loss": 1.2401,
+      "step": 2611
+    },
+    {
+      "epoch": 0.40061349693251536,
+      "grad_norm": 0.4705096185207367,
+      "learning_rate": 9.938788975949152e-05,
+      "loss": 1.0569,
+      "step": 2612
+    },
+    {
+      "epoch": 0.4007668711656442,
+      "grad_norm": 0.34275171160697937,
+      "learning_rate": 9.935336871677887e-05,
+      "loss": 1.0872,
+      "step": 2613
+    },
+    {
+      "epoch": 0.400920245398773,
+      "grad_norm": 0.46426495909690857,
+      "learning_rate": 9.931884190552118e-05,
+      "loss": 1.0618,
+      "step": 2614
+    },
+    {
+      "epoch": 0.40107361963190186,
+      "grad_norm": 0.6266953349113464,
+      "learning_rate": 9.928430933389677e-05,
+      "loss": 1.0914,
+      "step": 2615
+    },
+    {
+      "epoch": 0.4012269938650307,
+      "grad_norm": 0.29145199060440063,
+      "learning_rate": 9.924977101008531e-05,
+      "loss": 1.1689,
+      "step": 2616
+    },
+    {
+      "epoch": 0.4013803680981595,
+      "grad_norm": 0.31275057792663574,
+      "learning_rate": 9.921522694226784e-05,
+      "loss": 1.1071,
+      "step": 2617
+    },
+    {
+      "epoch": 0.40153374233128836,
+      "grad_norm": 0.8621084690093994,
+      "learning_rate": 9.918067713862677e-05,
+      "loss": 1.0824,
+      "step": 2618
+    },
+    {
+      "epoch": 0.4016871165644172,
+      "grad_norm": 1.0446206331253052,
+      "learning_rate": 9.914612160734585e-05,
+      "loss": 1.0638,
+      "step": 2619
+    },
+    {
+      "epoch": 0.401840490797546,
+      "grad_norm": 0.22402727603912354,
+      "learning_rate": 9.911156035661019e-05,
+      "loss": 1.1746,
+      "step": 2620
+    },
+    {
+      "epoch": 0.40199386503067486,
+      "grad_norm": 0.24227148294448853,
+      "learning_rate": 9.907699339460625e-05,
+      "loss": 1.0477,
+      "step": 2621
+    },
+    {
+      "epoch": 0.4021472392638037,
+      "grad_norm": 0.21767298877239227,
+      "learning_rate": 9.904242072952189e-05,
+      "loss": 1.0805,
+      "step": 2622
+    },
+    {
+      "epoch": 0.4023006134969325,
+      "grad_norm": 0.23689454793930054,
+      "learning_rate": 9.900784236954626e-05,
+      "loss": 1.3959,
+      "step": 2623
+    },
+    {
+      "epoch": 0.40245398773006136,
+      "grad_norm": 0.3619484007358551,
+      "learning_rate": 9.897325832286987e-05,
+      "loss": 1.139,
+      "step": 2624
+    },
+    {
+      "epoch": 0.4026073619631902,
+      "grad_norm": 0.7343416810035706,
+      "learning_rate": 9.893866859768464e-05,
+      "loss": 1.1842,
+      "step": 2625
+    },
+    {
+      "epoch": 0.402760736196319,
+      "grad_norm": 0.22161367535591125,
+      "learning_rate": 9.890407320218374e-05,
+      "loss": 1.1201,
+      "step": 2626
+    },
+    {
+      "epoch": 0.40291411042944786,
+      "grad_norm": 0.22009001672267914,
+      "learning_rate": 9.886947214456173e-05,
+      "loss": 1.0318,
+      "step": 2627
+    },
+    {
+      "epoch": 0.4030674846625767,
+      "grad_norm": 0.28609874844551086,
+      "learning_rate": 9.883486543301453e-05,
+      "loss": 1.2089,
+      "step": 2628
+    },
+    {
+      "epoch": 0.4032208588957055,
+      "grad_norm": 0.7825326919555664,
+      "learning_rate": 9.880025307573935e-05,
+      "loss": 1.0633,
+      "step": 2629
+    },
+    {
+      "epoch": 0.40337423312883436,
+      "grad_norm": 0.2800739109516144,
+      "learning_rate": 9.87656350809348e-05,
+      "loss": 1.0882,
+      "step": 2630
+    },
+    {
+      "epoch": 0.4035276073619632,
+      "grad_norm": 0.5831973552703857,
+      "learning_rate": 9.873101145680077e-05,
+      "loss": 1.092,
+      "step": 2631
+    },
+    {
+      "epoch": 0.403680981595092,
+      "grad_norm": 0.757469117641449,
+      "learning_rate": 9.869638221153853e-05,
+      "loss": 1.0652,
+      "step": 2632
+    },
+    {
+      "epoch": 0.40383435582822086,
+      "grad_norm": 0.24778321385383606,
+      "learning_rate": 9.866174735335062e-05,
+      "loss": 1.103,
+      "step": 2633
+    },
+    {
+      "epoch": 0.4039877300613497,
+      "grad_norm": 0.25633469223976135,
+      "learning_rate": 9.862710689044095e-05,
+      "loss": 0.9605,
+      "step": 2634
+    },
+    {
+      "epoch": 0.4041411042944785,
+      "grad_norm": 0.9311344623565674,
+      "learning_rate": 9.859246083101478e-05,
+      "loss": 1.2157,
+      "step": 2635
+    },
+    {
+      "epoch": 0.40429447852760736,
+      "grad_norm": 0.5161888003349304,
+      "learning_rate": 9.855780918327864e-05,
+      "loss": 1.2578,
+      "step": 2636
+    },
+    {
+      "epoch": 0.4044478527607362,
+      "grad_norm": 0.2273307889699936,
+      "learning_rate": 9.852315195544044e-05,
+      "loss": 1.2382,
+      "step": 2637
+    },
+    {
+      "epoch": 0.404601226993865,
+      "grad_norm": 0.3018631637096405,
+      "learning_rate": 9.848848915570936e-05,
+      "loss": 1.1181,
+      "step": 2638
+    },
+    {
+      "epoch": 0.40475460122699386,
+      "grad_norm": 0.29040977358818054,
+      "learning_rate": 9.845382079229593e-05,
+      "loss": 1.2763,
+      "step": 2639
+    },
+    {
+      "epoch": 0.4049079754601227,
+      "grad_norm": 0.9716011881828308,
+      "learning_rate": 9.841914687341204e-05,
+      "loss": 1.1678,
+      "step": 2640
+    },
+    {
+      "epoch": 0.4050613496932515,
+      "grad_norm": 0.38862425088882446,
+      "learning_rate": 9.838446740727076e-05,
+      "loss": 1.1453,
+      "step": 2641
+    },
+    {
+      "epoch": 0.40521472392638036,
+      "grad_norm": 0.28988996148109436,
+      "learning_rate": 9.834978240208663e-05,
+      "loss": 1.0668,
+      "step": 2642
+    },
+    {
+      "epoch": 0.4053680981595092,
+      "grad_norm": 0.8118105530738831,
+      "learning_rate": 9.831509186607538e-05,
+      "loss": 1.0611,
+      "step": 2643
+    },
+    {
+      "epoch": 0.405521472392638,
+      "grad_norm": 1.125091552734375,
+      "learning_rate": 9.828039580745416e-05,
+      "loss": 1.0507,
+      "step": 2644
+    },
+    {
+      "epoch": 0.40567484662576686,
+      "grad_norm": 0.30394065380096436,
+      "learning_rate": 9.824569423444133e-05,
+      "loss": 1.0697,
+      "step": 2645
+    },
+    {
+      "epoch": 0.4058282208588957,
+      "grad_norm": 0.4725097417831421,
+      "learning_rate": 9.821098715525664e-05,
+      "loss": 1.335,
+      "step": 2646
+    },
+    {
+      "epoch": 0.4059815950920245,
+      "grad_norm": 0.926630973815918,
+      "learning_rate": 9.817627457812105e-05,
+      "loss": 1.2961,
+      "step": 2647
+    },
+    {
+      "epoch": 0.40613496932515336,
+      "grad_norm": 0.44699627161026,
+      "learning_rate": 9.814155651125693e-05,
+      "loss": 1.1423,
+      "step": 2648
+    },
+    {
+      "epoch": 0.4062883435582822,
+      "grad_norm": 0.32848548889160156,
+      "learning_rate": 9.810683296288783e-05,
+      "loss": 1.2273,
+      "step": 2649
+    },
+    {
+      "epoch": 0.40644171779141103,
+      "grad_norm": 0.6397431492805481,
+      "learning_rate": 9.807210394123873e-05,
+      "loss": 0.9838,
+      "step": 2650
+    },
+    {
+      "epoch": 0.40659509202453986,
+      "grad_norm": 0.5640486478805542,
+      "learning_rate": 9.80373694545358e-05,
+      "loss": 1.1217,
+      "step": 2651
+    },
+    {
+      "epoch": 0.4067484662576687,
+      "grad_norm": 0.47879117727279663,
+      "learning_rate": 9.800262951100656e-05,
+      "loss": 1.1817,
+      "step": 2652
+    },
+    {
+      "epoch": 0.40690184049079753,
+      "grad_norm": 0.28983911871910095,
+      "learning_rate": 9.79678841188798e-05,
+      "loss": 1.0269,
+      "step": 2653
+    },
+    {
+      "epoch": 0.40705521472392636,
+      "grad_norm": 0.4302183985710144,
+      "learning_rate": 9.793313328638562e-05,
+      "loss": 1.3615,
+      "step": 2654
+    },
+    {
+      "epoch": 0.4072085889570552,
+      "grad_norm": 0.3655952513217926,
+      "learning_rate": 9.789837702175537e-05,
+      "loss": 1.1812,
+      "step": 2655
+    },
+    {
+      "epoch": 0.40736196319018403,
+      "grad_norm": 0.278461217880249,
+      "learning_rate": 9.786361533322174e-05,
+      "loss": 1.1903,
+      "step": 2656
+    },
+    {
+      "epoch": 0.40751533742331286,
+      "grad_norm": 0.28493303060531616,
+      "learning_rate": 9.782884822901867e-05,
+      "loss": 1.2105,
+      "step": 2657
+    },
+    {
+      "epoch": 0.4076687116564417,
+      "grad_norm": 0.2575989067554474,
+      "learning_rate": 9.779407571738137e-05,
+      "loss": 1.2099,
+      "step": 2658
+    },
+    {
+      "epoch": 0.40782208588957053,
+      "grad_norm": 0.49580103158950806,
+      "learning_rate": 9.775929780654639e-05,
+      "loss": 1.1204,
+      "step": 2659
+    },
+    {
+      "epoch": 0.40797546012269936,
+      "grad_norm": 0.4431646764278412,
+      "learning_rate": 9.772451450475148e-05,
+      "loss": 1.0727,
+      "step": 2660
+    },
+    {
+      "epoch": 0.4081288343558282,
+      "grad_norm": 0.23784826695919037,
+      "learning_rate": 9.768972582023576e-05,
+      "loss": 1.1403,
+      "step": 2661
+    },
+    {
+      "epoch": 0.40828220858895703,
+      "grad_norm": 0.23785635828971863,
+      "learning_rate": 9.765493176123952e-05,
+      "loss": 1.1384,
+      "step": 2662
+    },
+    {
+      "epoch": 0.40843558282208586,
+      "grad_norm": 0.6385679841041565,
+      "learning_rate": 9.762013233600438e-05,
+      "loss": 1.0173,
+      "step": 2663
+    },
+    {
+      "epoch": 0.4085889570552147,
+      "grad_norm": 0.3683789074420929,
+      "learning_rate": 9.758532755277326e-05,
+      "loss": 1.2007,
+      "step": 2664
+    },
+    {
+      "epoch": 0.4087423312883436,
+      "grad_norm": 0.21628734469413757,
+      "learning_rate": 9.755051741979028e-05,
+      "loss": 1.0864,
+      "step": 2665
+    },
+    {
+      "epoch": 0.4088957055214724,
+      "grad_norm": 0.23748043179512024,
+      "learning_rate": 9.75157019453009e-05,
+      "loss": 1.1508,
+      "step": 2666
+    },
+    {
+      "epoch": 0.40904907975460125,
+      "grad_norm": 0.4721824526786804,
+      "learning_rate": 9.748088113755177e-05,
+      "loss": 0.9744,
+      "step": 2667
+    },
+    {
+      "epoch": 0.4092024539877301,
+      "grad_norm": 0.2209247350692749,
+      "learning_rate": 9.744605500479084e-05,
+      "loss": 1.1864,
+      "step": 2668
+    },
+    {
+      "epoch": 0.4093558282208589,
+      "grad_norm": 0.24785536527633667,
+      "learning_rate": 9.741122355526738e-05,
+      "loss": 1.2385,
+      "step": 2669
+    },
+    {
+      "epoch": 0.40950920245398775,
+      "grad_norm": 0.3114257752895355,
+      "learning_rate": 9.737638679723179e-05,
+      "loss": 1.0727,
+      "step": 2670
+    },
+    {
+      "epoch": 0.4096625766871166,
+      "grad_norm": 0.35667404532432556,
+      "learning_rate": 9.734154473893586e-05,
+      "loss": 1.1172,
+      "step": 2671
+    },
+    {
+      "epoch": 0.4098159509202454,
+      "grad_norm": 0.24872253835201263,
+      "learning_rate": 9.730669738863253e-05,
+      "loss": 1.1625,
+      "step": 2672
+    },
+    {
+      "epoch": 0.40996932515337425,
+      "grad_norm": 0.32286447286605835,
+      "learning_rate": 9.727184475457606e-05,
+      "loss": 1.1924,
+      "step": 2673
+    },
+    {
+      "epoch": 0.4101226993865031,
+      "grad_norm": 0.24281348288059235,
+      "learning_rate": 9.723698684502193e-05,
+      "loss": 1.0652,
+      "step": 2674
+    },
+    {
+      "epoch": 0.4102760736196319,
+      "grad_norm": 0.20418302714824677,
+      "learning_rate": 9.720212366822691e-05,
+      "loss": 1.1745,
+      "step": 2675
+    },
+    {
+      "epoch": 0.41042944785276075,
+      "grad_norm": 0.4843696355819702,
+      "learning_rate": 9.716725523244893e-05,
+      "loss": 1.1579,
+      "step": 2676
+    },
+    {
+      "epoch": 0.4105828220858896,
+      "grad_norm": 0.3205069601535797,
+      "learning_rate": 9.713238154594727e-05,
+      "loss": 1.2118,
+      "step": 2677
+    },
+    {
+      "epoch": 0.4107361963190184,
+      "grad_norm": 0.26448124647140503,
+      "learning_rate": 9.709750261698239e-05,
+      "loss": 1.284,
+      "step": 2678
+    },
+    {
+      "epoch": 0.41088957055214725,
+      "grad_norm": 0.5180304646492004,
+      "learning_rate": 9.7062618453816e-05,
+      "loss": 1.0229,
+      "step": 2679
+    },
+    {
+      "epoch": 0.4110429447852761,
+      "grad_norm": 0.33064115047454834,
+      "learning_rate": 9.702772906471107e-05,
+      "loss": 1.2503,
+      "step": 2680
+    },
+    {
+      "epoch": 0.4111963190184049,
+      "grad_norm": 0.25297847390174866,
+      "learning_rate": 9.69928344579318e-05,
+      "loss": 1.1563,
+      "step": 2681
+    },
+    {
+      "epoch": 0.41134969325153375,
+      "grad_norm": 0.2722991406917572,
+      "learning_rate": 9.695793464174361e-05,
+      "loss": 1.1289,
+      "step": 2682
+    },
+    {
+      "epoch": 0.4115030674846626,
+      "grad_norm": 0.1985693871974945,
+      "learning_rate": 9.69230296244132e-05,
+      "loss": 0.9679,
+      "step": 2683
+    },
+    {
+      "epoch": 0.4116564417177914,
+      "grad_norm": 0.3466608226299286,
+      "learning_rate": 9.688811941420842e-05,
+      "loss": 1.0913,
+      "step": 2684
+    },
+    {
+      "epoch": 0.41180981595092025,
+      "grad_norm": 0.19432079792022705,
+      "learning_rate": 9.685320401939841e-05,
+      "loss": 1.1051,
+      "step": 2685
+    },
+    {
+      "epoch": 0.4119631901840491,
+      "grad_norm": 0.1949218213558197,
+      "learning_rate": 9.681828344825352e-05,
+      "loss": 0.9893,
+      "step": 2686
+    },
+    {
+      "epoch": 0.4121165644171779,
+      "grad_norm": 0.19530819356441498,
+      "learning_rate": 9.678335770904537e-05,
+      "loss": 1.1546,
+      "step": 2687
+    },
+    {
+      "epoch": 0.41226993865030676,
+      "grad_norm": 0.18814663589000702,
+      "learning_rate": 9.674842681004673e-05,
+      "loss": 0.9511,
+      "step": 2688
+    },
+    {
+      "epoch": 0.4124233128834356,
+      "grad_norm": 0.20040805637836456,
+      "learning_rate": 9.671349075953163e-05,
+      "loss": 1.1728,
+      "step": 2689
+    },
+    {
+      "epoch": 0.4125766871165644,
+      "grad_norm": 0.19699981808662415,
+      "learning_rate": 9.667854956577535e-05,
+      "loss": 1.1014,
+      "step": 2690
+    },
+    {
+      "epoch": 0.41273006134969326,
+      "grad_norm": 0.19791506230831146,
+      "learning_rate": 9.664360323705432e-05,
+      "loss": 1.2913,
+      "step": 2691
+    },
+    {
+      "epoch": 0.4128834355828221,
+      "grad_norm": 0.1973227560520172,
+      "learning_rate": 9.660865178164624e-05,
+      "loss": 1.1299,
+      "step": 2692
+    },
+    {
+      "epoch": 0.4130368098159509,
+      "grad_norm": 0.2056787759065628,
+      "learning_rate": 9.657369520783001e-05,
+      "loss": 1.1658,
+      "step": 2693
+    },
+    {
+      "epoch": 0.41319018404907976,
+      "grad_norm": 0.31302326917648315,
+      "learning_rate": 9.653873352388573e-05,
+      "loss": 1.0776,
+      "step": 2694
+    },
+    {
+      "epoch": 0.4133435582822086,
+      "grad_norm": 0.20679320394992828,
+      "learning_rate": 9.650376673809472e-05,
+      "loss": 1.1426,
+      "step": 2695
+    },
+    {
+      "epoch": 0.4134969325153374,
+      "grad_norm": 0.2226579636335373,
+      "learning_rate": 9.646879485873954e-05,
+      "loss": 1.0526,
+      "step": 2696
+    },
+    {
+      "epoch": 0.41365030674846626,
+      "grad_norm": 0.19804169237613678,
+      "learning_rate": 9.643381789410389e-05,
+      "loss": 1.1698,
+      "step": 2697
+    },
+    {
+      "epoch": 0.4138036809815951,
+      "grad_norm": 0.1731129139661789,
+      "learning_rate": 9.639883585247272e-05,
+      "loss": 0.9407,
+      "step": 2698
+    },
+    {
+      "epoch": 0.4139570552147239,
+      "grad_norm": 0.19783996045589447,
+      "learning_rate": 9.636384874213217e-05,
+      "loss": 1.1227,
+      "step": 2699
+    },
+    {
+      "epoch": 0.41411042944785276,
+      "grad_norm": 0.19401469826698303,
+      "learning_rate": 9.632885657136958e-05,
+      "loss": 1.1882,
+      "step": 2700
+    },
+    {
+      "epoch": 0.4142638036809816,
+      "grad_norm": 0.32239410281181335,
+      "learning_rate": 9.62938593484735e-05,
+      "loss": 1.194,
+      "step": 2701
+    },
+    {
+      "epoch": 0.4144171779141104,
+      "grad_norm": 0.19184055924415588,
+      "learning_rate": 9.625885708173367e-05,
+      "loss": 0.9864,
+      "step": 2702
+    },
+    {
+      "epoch": 0.41457055214723926,
+      "grad_norm": 0.18981805443763733,
+      "learning_rate": 9.622384977944102e-05,
+      "loss": 1.1118,
+      "step": 2703
+    },
+    {
+      "epoch": 0.4147239263803681,
+      "grad_norm": 0.17381292581558228,
+      "learning_rate": 9.618883744988768e-05,
+      "loss": 1.0123,
+      "step": 2704
+    },
+    {
+      "epoch": 0.4148773006134969,
+      "grad_norm": 0.1871747076511383,
+      "learning_rate": 9.615382010136697e-05,
+      "loss": 1.0126,
+      "step": 2705
+    },
+    {
+      "epoch": 0.41503067484662576,
+      "grad_norm": 0.216110497713089,
+      "learning_rate": 9.611879774217335e-05,
+      "loss": 1.1216,
+      "step": 2706
+    },
+    {
+      "epoch": 0.4151840490797546,
+      "grad_norm": 0.18437710404396057,
+      "learning_rate": 9.608377038060258e-05,
+      "loss": 1.0785,
+      "step": 2707
+    },
+    {
+      "epoch": 0.4153374233128834,
+      "grad_norm": 0.42110344767570496,
+      "learning_rate": 9.60487380249515e-05,
+      "loss": 1.1584,
+      "step": 2708
+    },
+    {
+      "epoch": 0.41549079754601226,
+      "grad_norm": 0.1828433722257614,
+      "learning_rate": 9.601370068351815e-05,
+      "loss": 1.0573,
+      "step": 2709
+    },
+    {
+      "epoch": 0.4156441717791411,
+      "grad_norm": 0.21547716856002808,
+      "learning_rate": 9.597865836460186e-05,
+      "loss": 1.3005,
+      "step": 2710
+    },
+    {
+      "epoch": 0.4157975460122699,
+      "grad_norm": 0.19449225068092346,
+      "learning_rate": 9.594361107650293e-05,
+      "loss": 1.1753,
+      "step": 2711
+    },
+    {
+      "epoch": 0.41595092024539876,
+      "grad_norm": 0.16765469312667847,
+      "learning_rate": 9.590855882752305e-05,
+      "loss": 0.9454,
+      "step": 2712
+    },
+    {
+      "epoch": 0.4161042944785276,
+      "grad_norm": 0.21245166659355164,
+      "learning_rate": 9.587350162596495e-05,
+      "loss": 1.1107,
+      "step": 2713
+    },
+    {
+      "epoch": 0.4162576687116564,
+      "grad_norm": 0.6117340922355652,
+      "learning_rate": 9.583843948013257e-05,
+      "loss": 1.2275,
+      "step": 2714
+    },
+    {
+      "epoch": 0.41641104294478526,
+      "grad_norm": 0.19648610055446625,
+      "learning_rate": 9.580337239833106e-05,
+      "loss": 1.1948,
+      "step": 2715
+    },
+    {
+      "epoch": 0.4165644171779141,
+      "grad_norm": 0.19113889336585999,
+      "learning_rate": 9.576830038886668e-05,
+      "loss": 1.2238,
+      "step": 2716
+    },
+    {
+      "epoch": 0.4167177914110429,
+      "grad_norm": 0.22827202081680298,
+      "learning_rate": 9.57332234600469e-05,
+      "loss": 1.1367,
+      "step": 2717
+    },
+    {
+      "epoch": 0.41687116564417176,
+      "grad_norm": 0.20057086646556854,
+      "learning_rate": 9.569814162018031e-05,
+      "loss": 1.1447,
+      "step": 2718
+    },
+    {
+      "epoch": 0.4170245398773006,
+      "grad_norm": 0.23576216399669647,
+      "learning_rate": 9.566305487757672e-05,
+      "loss": 1.1091,
+      "step": 2719
+    },
+    {
+      "epoch": 0.4171779141104294,
+      "grad_norm": 0.23772644996643066,
+      "learning_rate": 9.562796324054706e-05,
+      "loss": 1.2735,
+      "step": 2720
+    },
+    {
+      "epoch": 0.41733128834355826,
+      "grad_norm": 0.29214784502983093,
+      "learning_rate": 9.559286671740343e-05,
+      "loss": 1.1006,
+      "step": 2721
+    },
+    {
+      "epoch": 0.4174846625766871,
+      "grad_norm": 0.6258183121681213,
+      "learning_rate": 9.55577653164591e-05,
+      "loss": 0.9983,
+      "step": 2722
+    },
+    {
+      "epoch": 0.4176380368098159,
+      "grad_norm": 0.18970300257205963,
+      "learning_rate": 9.552265904602846e-05,
+      "loss": 1.095,
+      "step": 2723
+    },
+    {
+      "epoch": 0.41779141104294476,
+      "grad_norm": 0.44734400510787964,
+      "learning_rate": 9.548754791442712e-05,
+      "loss": 0.9528,
+      "step": 2724
+    },
+    {
+      "epoch": 0.4179447852760736,
+      "grad_norm": 0.19948838651180267,
+      "learning_rate": 9.545243192997178e-05,
+      "loss": 1.0614,
+      "step": 2725
+    },
+    {
+      "epoch": 0.4180981595092024,
+      "grad_norm": 0.20125146210193634,
+      "learning_rate": 9.54173111009803e-05,
+      "loss": 1.1629,
+      "step": 2726
+    },
+    {
+      "epoch": 0.41825153374233126,
+      "grad_norm": 0.6206796169281006,
+      "learning_rate": 9.53821854357717e-05,
+      "loss": 1.0741,
+      "step": 2727
+    },
+    {
+      "epoch": 0.41840490797546015,
+      "grad_norm": 0.4673416018486023,
+      "learning_rate": 9.534705494266613e-05,
+      "loss": 1.1735,
+      "step": 2728
+    },
+    {
+      "epoch": 0.418558282208589,
+      "grad_norm": 0.1932254135608673,
+      "learning_rate": 9.531191962998493e-05,
+      "loss": 1.1714,
+      "step": 2729
+    },
+    {
+      "epoch": 0.4187116564417178,
+      "grad_norm": 0.47512224316596985,
+      "learning_rate": 9.52767795060505e-05,
+      "loss": 1.0194,
+      "step": 2730
+    },
+    {
+      "epoch": 0.41886503067484665,
+      "grad_norm": 0.23819732666015625,
+      "learning_rate": 9.52416345791865e-05,
+      "loss": 1.0435,
+      "step": 2731
+    },
+    {
+      "epoch": 0.4190184049079755,
+      "grad_norm": 0.17585274577140808,
+      "learning_rate": 9.520648485771758e-05,
+      "loss": 1.0279,
+      "step": 2732
+    },
+    {
+      "epoch": 0.4191717791411043,
+      "grad_norm": 0.24209927022457123,
+      "learning_rate": 9.517133034996964e-05,
+      "loss": 1.0549,
+      "step": 2733
+    },
+    {
+      "epoch": 0.41932515337423315,
+      "grad_norm": 0.20605552196502686,
+      "learning_rate": 9.513617106426966e-05,
+      "loss": 1.109,
+      "step": 2734
+    },
+    {
+      "epoch": 0.419478527607362,
+      "grad_norm": 0.3058285117149353,
+      "learning_rate": 9.510100700894577e-05,
+      "loss": 1.2813,
+      "step": 2735
+    },
+    {
+      "epoch": 0.4196319018404908,
+      "grad_norm": 0.24217043817043304,
+      "learning_rate": 9.506583819232724e-05,
+      "loss": 0.9991,
+      "step": 2736
+    },
+    {
+      "epoch": 0.41978527607361965,
+      "grad_norm": 0.1910662204027176,
+      "learning_rate": 9.503066462274444e-05,
+      "loss": 1.176,
+      "step": 2737
+    },
+    {
+      "epoch": 0.4199386503067485,
+      "grad_norm": 0.2610887289047241,
+      "learning_rate": 9.499548630852887e-05,
+      "loss": 1.31,
+      "step": 2738
+    },
+    {
+      "epoch": 0.4200920245398773,
+      "grad_norm": 0.2453286051750183,
+      "learning_rate": 9.49603032580132e-05,
+      "loss": 1.1955,
+      "step": 2739
+    },
+    {
+      "epoch": 0.42024539877300615,
+      "grad_norm": 0.2025267779827118,
+      "learning_rate": 9.492511547953113e-05,
+      "loss": 1.16,
+      "step": 2740
+    },
+    {
+      "epoch": 0.420398773006135,
+      "grad_norm": 0.264873743057251,
+      "learning_rate": 9.488992298141759e-05,
+      "loss": 1.1892,
+      "step": 2741
+    },
+    {
+      "epoch": 0.4205521472392638,
+      "grad_norm": 0.23771484196186066,
+      "learning_rate": 9.485472577200852e-05,
+      "loss": 1.2044,
+      "step": 2742
+    },
+    {
+      "epoch": 0.42070552147239265,
+      "grad_norm": 0.1812715232372284,
+      "learning_rate": 9.481952385964107e-05,
+      "loss": 1.0518,
+      "step": 2743
+    },
+    {
+      "epoch": 0.4208588957055215,
+      "grad_norm": 0.19266551733016968,
+      "learning_rate": 9.478431725265343e-05,
+      "loss": 1.0554,
+      "step": 2744
+    },
+    {
+      "epoch": 0.4210122699386503,
+      "grad_norm": 0.20322883129119873,
+      "learning_rate": 9.474910595938498e-05,
+      "loss": 1.1825,
+      "step": 2745
+    },
+    {
+      "epoch": 0.42116564417177915,
+      "grad_norm": 0.2718859016895294,
+      "learning_rate": 9.47138899881761e-05,
+      "loss": 1.0109,
+      "step": 2746
+    },
+    {
+      "epoch": 0.421319018404908,
+      "grad_norm": 0.5000318884849548,
+      "learning_rate": 9.467866934736844e-05,
+      "loss": 1.0512,
+      "step": 2747
+    },
+    {
+      "epoch": 0.4214723926380368,
+      "grad_norm": 0.41442298889160156,
+      "learning_rate": 9.464344404530455e-05,
+      "loss": 1.0191,
+      "step": 2748
+    },
+    {
+      "epoch": 0.42162576687116565,
+      "grad_norm": 0.20330314338207245,
+      "learning_rate": 9.460821409032827e-05,
+      "loss": 1.1539,
+      "step": 2749
+    },
+    {
+      "epoch": 0.4217791411042945,
+      "grad_norm": 0.33926984667778015,
+      "learning_rate": 9.457297949078442e-05,
+      "loss": 1.0282,
+      "step": 2750
+    },
+    {
+      "epoch": 0.4219325153374233,
+      "grad_norm": 0.20329637825489044,
+      "learning_rate": 9.4537740255019e-05,
+      "loss": 1.0698,
+      "step": 2751
+    },
+    {
+      "epoch": 0.42208588957055215,
+      "grad_norm": 0.21812704205513,
+      "learning_rate": 9.450249639137902e-05,
+      "loss": 1.0958,
+      "step": 2752
+    },
+    {
+      "epoch": 0.422239263803681,
+      "grad_norm": 0.3627301752567291,
+      "learning_rate": 9.446724790821273e-05,
+      "loss": 1.0949,
+      "step": 2753
+    },
+    {
+      "epoch": 0.4223926380368098,
+      "grad_norm": 0.19875159859657288,
+      "learning_rate": 9.443199481386932e-05,
+      "loss": 1.2904,
+      "step": 2754
+    },
+    {
+      "epoch": 0.42254601226993865,
+      "grad_norm": 0.1881684958934784,
+      "learning_rate": 9.439673711669914e-05,
+      "loss": 1.1424,
+      "step": 2755
+    },
+    {
+      "epoch": 0.4226993865030675,
+      "grad_norm": 0.18532918393611908,
+      "learning_rate": 9.436147482505362e-05,
+      "loss": 1.0533,
+      "step": 2756
+    },
+    {
+      "epoch": 0.4228527607361963,
+      "grad_norm": 0.2116365283727646,
+      "learning_rate": 9.432620794728533e-05,
+      "loss": 1.0777,
+      "step": 2757
+    },
+    {
+      "epoch": 0.42300613496932515,
+      "grad_norm": 0.20873315632343292,
+      "learning_rate": 9.429093649174783e-05,
+      "loss": 1.0654,
+      "step": 2758
+    },
+    {
+      "epoch": 0.423159509202454,
+      "grad_norm": 0.2083897590637207,
+      "learning_rate": 9.425566046679584e-05,
+      "loss": 1.1838,
+      "step": 2759
+    },
+    {
+      "epoch": 0.4233128834355828,
+      "grad_norm": 0.18657729029655457,
+      "learning_rate": 9.422037988078516e-05,
+      "loss": 0.9882,
+      "step": 2760
+    },
+    {
+      "epoch": 0.42346625766871165,
+      "grad_norm": 0.4262439012527466,
+      "learning_rate": 9.418509474207258e-05,
+      "loss": 1.0631,
+      "step": 2761
+    },
+    {
+      "epoch": 0.4236196319018405,
+      "grad_norm": 0.18261700868606567,
+      "learning_rate": 9.41498050590161e-05,
+      "loss": 1.006,
+      "step": 2762
+    },
+    {
+      "epoch": 0.4237730061349693,
+      "grad_norm": 0.28129711747169495,
+      "learning_rate": 9.41145108399747e-05,
+      "loss": 1.1813,
+      "step": 2763
+    },
+    {
+      "epoch": 0.42392638036809815,
+      "grad_norm": 0.9670384526252747,
+      "learning_rate": 9.40792120933085e-05,
+      "loss": 1.0464,
+      "step": 2764
+    },
+    {
+      "epoch": 0.424079754601227,
+      "grad_norm": 0.2007206678390503,
+      "learning_rate": 9.404390882737862e-05,
+      "loss": 1.1538,
+      "step": 2765
+    },
+    {
+      "epoch": 0.4242331288343558,
+      "grad_norm": 0.47508248686790466,
+      "learning_rate": 9.400860105054732e-05,
+      "loss": 1.1609,
+      "step": 2766
+    },
+    {
+      "epoch": 0.42438650306748466,
+      "grad_norm": 0.2843158543109894,
+      "learning_rate": 9.397328877117788e-05,
+      "loss": 1.0944,
+      "step": 2767
+    },
+    {
+      "epoch": 0.4245398773006135,
+      "grad_norm": 0.19018416106700897,
+      "learning_rate": 9.39379719976347e-05,
+      "loss": 1.082,
+      "step": 2768
+    },
+    {
+      "epoch": 0.4246932515337423,
+      "grad_norm": 0.21966424584388733,
+      "learning_rate": 9.390265073828316e-05,
+      "loss": 1.2069,
+      "step": 2769
+    },
+    {
+      "epoch": 0.42484662576687116,
+      "grad_norm": 0.20814302563667297,
+      "learning_rate": 9.386732500148978e-05,
+      "loss": 1.0507,
+      "step": 2770
+    },
+    {
+      "epoch": 0.425,
+      "grad_norm": 0.18270622193813324,
+      "learning_rate": 9.38319947956221e-05,
+      "loss": 1.1415,
+      "step": 2771
+    },
+    {
+      "epoch": 0.4251533742331288,
+      "grad_norm": 0.19704443216323853,
+      "learning_rate": 9.379666012904875e-05,
+      "loss": 1.178,
+      "step": 2772
+    },
+    {
+      "epoch": 0.42530674846625766,
+      "grad_norm": 0.18709015846252441,
+      "learning_rate": 9.376132101013938e-05,
+      "loss": 1.0478,
+      "step": 2773
+    },
+    {
+      "epoch": 0.4254601226993865,
+      "grad_norm": 0.4005066752433777,
+      "learning_rate": 9.372597744726476e-05,
+      "loss": 1.0255,
+      "step": 2774
+    },
+    {
+      "epoch": 0.4256134969325153,
+      "grad_norm": 0.3168583810329437,
+      "learning_rate": 9.369062944879659e-05,
+      "loss": 1.2648,
+      "step": 2775
+    },
+    {
+      "epoch": 0.42576687116564416,
+      "grad_norm": 0.44313931465148926,
+      "learning_rate": 9.365527702310774e-05,
+      "loss": 1.1774,
+      "step": 2776
+    },
+    {
+      "epoch": 0.425920245398773,
+      "grad_norm": 0.19479180872440338,
+      "learning_rate": 9.361992017857208e-05,
+      "loss": 1.1078,
+      "step": 2777
+    },
+    {
+      "epoch": 0.4260736196319018,
+      "grad_norm": 0.20662395656108856,
+      "learning_rate": 9.358455892356454e-05,
+      "loss": 1.1841,
+      "step": 2778
+    },
+    {
+      "epoch": 0.42622699386503066,
+      "grad_norm": 0.20020242035388947,
+      "learning_rate": 9.354919326646106e-05,
+      "loss": 1.1759,
+      "step": 2779
+    },
+    {
+      "epoch": 0.4263803680981595,
+      "grad_norm": 0.24016380310058594,
+      "learning_rate": 9.351382321563867e-05,
+      "loss": 1.125,
+      "step": 2780
+    },
+    {
+      "epoch": 0.4265337423312883,
+      "grad_norm": 0.21590808033943176,
+      "learning_rate": 9.34784487794754e-05,
+      "loss": 0.9091,
+      "step": 2781
+    },
+    {
+      "epoch": 0.42668711656441716,
+      "grad_norm": 0.2576092481613159,
+      "learning_rate": 9.344306996635039e-05,
+      "loss": 1.2231,
+      "step": 2782
+    },
+    {
+      "epoch": 0.426840490797546,
+      "grad_norm": 0.19620594382286072,
+      "learning_rate": 9.340768678464372e-05,
+      "loss": 1.0787,
+      "step": 2783
+    },
+    {
+      "epoch": 0.4269938650306748,
+      "grad_norm": 0.19355285167694092,
+      "learning_rate": 9.337229924273653e-05,
+      "loss": 1.223,
+      "step": 2784
+    },
+    {
+      "epoch": 0.42714723926380366,
+      "grad_norm": 0.2132529616355896,
+      "learning_rate": 9.333690734901106e-05,
+      "loss": 1.2311,
+      "step": 2785
+    },
+    {
+      "epoch": 0.4273006134969325,
+      "grad_norm": 0.20811593532562256,
+      "learning_rate": 9.330151111185051e-05,
+      "loss": 1.1749,
+      "step": 2786
+    },
+    {
+      "epoch": 0.4274539877300613,
+      "grad_norm": 0.1844632625579834,
+      "learning_rate": 9.326611053963913e-05,
+      "loss": 1.1393,
+      "step": 2787
+    },
+    {
+      "epoch": 0.42760736196319016,
+      "grad_norm": 0.19996188580989838,
+      "learning_rate": 9.32307056407622e-05,
+      "loss": 1.1522,
+      "step": 2788
+    },
+    {
+      "epoch": 0.427760736196319,
+      "grad_norm": 0.6936095952987671,
+      "learning_rate": 9.319529642360607e-05,
+      "loss": 1.0872,
+      "step": 2789
+    },
+    {
+      "epoch": 0.4279141104294479,
+      "grad_norm": 0.5277085304260254,
+      "learning_rate": 9.315988289655798e-05,
+      "loss": 1.1202,
+      "step": 2790
+    },
+    {
+      "epoch": 0.4280674846625767,
+      "grad_norm": 0.1880098134279251,
+      "learning_rate": 9.312446506800635e-05,
+      "loss": 1.1166,
+      "step": 2791
+    },
+    {
+      "epoch": 0.42822085889570555,
+      "grad_norm": 0.1962825357913971,
+      "learning_rate": 9.308904294634051e-05,
+      "loss": 0.9921,
+      "step": 2792
+    },
+    {
+      "epoch": 0.4283742331288344,
+      "grad_norm": 0.2003127485513687,
+      "learning_rate": 9.305361653995085e-05,
+      "loss": 1.0845,
+      "step": 2793
+    },
+    {
+      "epoch": 0.4285276073619632,
+      "grad_norm": 0.2780044674873352,
+      "learning_rate": 9.301818585722875e-05,
+      "loss": 1.1549,
+      "step": 2794
+    },
+    {
+      "epoch": 0.42868098159509205,
+      "grad_norm": 0.18174302577972412,
+      "learning_rate": 9.298275090656669e-05,
+      "loss": 1.1146,
+      "step": 2795
+    },
+    {
+      "epoch": 0.4288343558282209,
+      "grad_norm": 0.18864718079566956,
+      "learning_rate": 9.294731169635802e-05,
+      "loss": 1.2258,
+      "step": 2796
+    },
+    {
+      "epoch": 0.4289877300613497,
+      "grad_norm": 0.18561610579490662,
+      "learning_rate": 9.291186823499721e-05,
+      "loss": 0.9715,
+      "step": 2797
+    },
+    {
+      "epoch": 0.42914110429447855,
+      "grad_norm": 0.22671082615852356,
+      "learning_rate": 9.287642053087967e-05,
+      "loss": 1.0509,
+      "step": 2798
+    },
+    {
+      "epoch": 0.4292944785276074,
+      "grad_norm": 0.1851448267698288,
+      "learning_rate": 9.284096859240188e-05,
+      "loss": 1.1111,
+      "step": 2799
+    },
+    {
+      "epoch": 0.4294478527607362,
+      "grad_norm": 0.19739854335784912,
+      "learning_rate": 9.280551242796126e-05,
+      "loss": 1.0127,
+      "step": 2800
+    },
+    {
+      "epoch": 0.42960122699386505,
+      "grad_norm": 0.22910991311073303,
+      "learning_rate": 9.277005204595624e-05,
+      "loss": 1.0425,
+      "step": 2801
+    },
+    {
+      "epoch": 0.4297546012269939,
+      "grad_norm": 0.20392802357673645,
+      "learning_rate": 9.273458745478631e-05,
+      "loss": 1.111,
+      "step": 2802
+    },
+    {
+      "epoch": 0.4299079754601227,
+      "grad_norm": 0.21286192536354065,
+      "learning_rate": 9.269911866285191e-05,
+      "loss": 1.023,
+      "step": 2803
+    },
+    {
+      "epoch": 0.43006134969325155,
+      "grad_norm": 0.19045209884643555,
+      "learning_rate": 9.266364567855444e-05,
+      "loss": 1.1425,
+      "step": 2804
+    },
+    {
+      "epoch": 0.4302147239263804,
+      "grad_norm": 0.1793903112411499,
+      "learning_rate": 9.262816851029636e-05,
+      "loss": 0.9858,
+      "step": 2805
+    },
+    {
+      "epoch": 0.4303680981595092,
+      "grad_norm": 0.36526554822921753,
+      "learning_rate": 9.259268716648106e-05,
+      "loss": 1.0823,
+      "step": 2806
+    },
+    {
+      "epoch": 0.43052147239263805,
+      "grad_norm": 0.18784278631210327,
+      "learning_rate": 9.2557201655513e-05,
+      "loss": 1.138,
+      "step": 2807
+    },
+    {
+      "epoch": 0.4306748466257669,
+      "grad_norm": 0.19819802045822144,
+      "learning_rate": 9.252171198579754e-05,
+      "loss": 1.0932,
+      "step": 2808
+    },
+    {
+      "epoch": 0.4308282208588957,
+      "grad_norm": 0.1907021701335907,
+      "learning_rate": 9.248621816574108e-05,
+      "loss": 1.0027,
+      "step": 2809
+    },
+    {
+      "epoch": 0.43098159509202455,
+      "grad_norm": 0.18676075339317322,
+      "learning_rate": 9.245072020375098e-05,
+      "loss": 1.1844,
+      "step": 2810
+    },
+    {
+      "epoch": 0.4311349693251534,
+      "grad_norm": 0.19258108735084534,
+      "learning_rate": 9.241521810823562e-05,
+      "loss": 1.1624,
+      "step": 2811
+    },
+    {
+      "epoch": 0.4312883435582822,
+      "grad_norm": 0.22481131553649902,
+      "learning_rate": 9.237971188760426e-05,
+      "loss": 1.0987,
+      "step": 2812
+    },
+    {
+      "epoch": 0.43144171779141105,
+      "grad_norm": 0.1947210729122162,
+      "learning_rate": 9.234420155026725e-05,
+      "loss": 1.0949,
+      "step": 2813
+    },
+    {
+      "epoch": 0.4315950920245399,
+      "grad_norm": 0.253302663564682,
+      "learning_rate": 9.230868710463588e-05,
+      "loss": 1.1369,
+      "step": 2814
+    },
+    {
+      "epoch": 0.4317484662576687,
+      "grad_norm": 0.17204664647579193,
+      "learning_rate": 9.227316855912235e-05,
+      "loss": 0.9645,
+      "step": 2815
+    },
+    {
+      "epoch": 0.43190184049079755,
+      "grad_norm": 0.19119073450565338,
+      "learning_rate": 9.223764592213994e-05,
+      "loss": 1.1253,
+      "step": 2816
+    },
+    {
+      "epoch": 0.4320552147239264,
+      "grad_norm": 0.17278462648391724,
+      "learning_rate": 9.220211920210283e-05,
+      "loss": 0.9432,
+      "step": 2817
+    },
+    {
+      "epoch": 0.4322085889570552,
+      "grad_norm": 0.17700186371803284,
+      "learning_rate": 9.216658840742616e-05,
+      "loss": 1.0365,
+      "step": 2818
+    },
+    {
+      "epoch": 0.43236196319018405,
+      "grad_norm": 0.2905201017856598,
+      "learning_rate": 9.213105354652606e-05,
+      "loss": 1.2241,
+      "step": 2819
+    },
+    {
+      "epoch": 0.4325153374233129,
+      "grad_norm": 0.1890917867422104,
+      "learning_rate": 9.209551462781964e-05,
+      "loss": 1.2126,
+      "step": 2820
+    },
+    {
+      "epoch": 0.4326687116564417,
+      "grad_norm": 0.20080871880054474,
+      "learning_rate": 9.20599716597249e-05,
+      "loss": 1.1168,
+      "step": 2821
+    },
+    {
+      "epoch": 0.43282208588957055,
+      "grad_norm": 0.18008360266685486,
+      "learning_rate": 9.202442465066089e-05,
+      "loss": 1.1454,
+      "step": 2822
+    },
+    {
+      "epoch": 0.4329754601226994,
+      "grad_norm": 0.19770358502864838,
+      "learning_rate": 9.198887360904757e-05,
+      "loss": 1.1502,
+      "step": 2823
+    },
+    {
+      "epoch": 0.4331288343558282,
+      "grad_norm": 0.18325120210647583,
+      "learning_rate": 9.195331854330587e-05,
+      "loss": 1.0452,
+      "step": 2824
+    },
+    {
+      "epoch": 0.43328220858895705,
+      "grad_norm": 0.19642315804958344,
+      "learning_rate": 9.191775946185764e-05,
+      "loss": 1.0814,
+      "step": 2825
+    },
+    {
+      "epoch": 0.4334355828220859,
+      "grad_norm": 0.1912439912557602,
+      "learning_rate": 9.18821963731257e-05,
+      "loss": 1.1381,
+      "step": 2826
+    },
+    {
+      "epoch": 0.4335889570552147,
+      "grad_norm": 0.17682555317878723,
+      "learning_rate": 9.184662928553384e-05,
+      "loss": 1.0662,
+      "step": 2827
+    },
+    {
+      "epoch": 0.43374233128834355,
+      "grad_norm": 0.18681925535202026,
+      "learning_rate": 9.18110582075068e-05,
+      "loss": 1.164,
+      "step": 2828
+    },
+    {
+      "epoch": 0.4338957055214724,
+      "grad_norm": 0.18759578466415405,
+      "learning_rate": 9.177548314747019e-05,
+      "loss": 1.1587,
+      "step": 2829
+    },
+    {
+      "epoch": 0.4340490797546012,
+      "grad_norm": 0.1786603033542633,
+      "learning_rate": 9.173990411385068e-05,
+      "loss": 1.2151,
+      "step": 2830
+    },
+    {
+      "epoch": 0.43420245398773005,
+      "grad_norm": 0.23540641367435455,
+      "learning_rate": 9.170432111507579e-05,
+      "loss": 1.16,
+      "step": 2831
+    },
+    {
+      "epoch": 0.4343558282208589,
+      "grad_norm": 0.20033970475196838,
+      "learning_rate": 9.166873415957404e-05,
+      "loss": 1.1776,
+      "step": 2832
+    },
+    {
+      "epoch": 0.4345092024539877,
+      "grad_norm": 0.18616433441638947,
+      "learning_rate": 9.163314325577482e-05,
+      "loss": 1.0353,
+      "step": 2833
+    },
+    {
+      "epoch": 0.43466257668711655,
+      "grad_norm": 0.18619492650032043,
+      "learning_rate": 9.15975484121085e-05,
+      "loss": 1.1504,
+      "step": 2834
+    },
+    {
+      "epoch": 0.4348159509202454,
+      "grad_norm": 0.19051499664783478,
+      "learning_rate": 9.156194963700639e-05,
+      "loss": 1.2128,
+      "step": 2835
+    },
+    {
+      "epoch": 0.4349693251533742,
+      "grad_norm": 0.1820095032453537,
+      "learning_rate": 9.152634693890068e-05,
+      "loss": 1.2605,
+      "step": 2836
+    },
+    {
+      "epoch": 0.43512269938650305,
+      "grad_norm": 0.19009649753570557,
+      "learning_rate": 9.149074032622458e-05,
+      "loss": 1.0221,
+      "step": 2837
+    },
+    {
+      "epoch": 0.4352760736196319,
+      "grad_norm": 0.20088344812393188,
+      "learning_rate": 9.145512980741219e-05,
+      "loss": 1.1934,
+      "step": 2838
+    },
+    {
+      "epoch": 0.4354294478527607,
+      "grad_norm": 0.18488138914108276,
+      "learning_rate": 9.141951539089843e-05,
+      "loss": 1.1069,
+      "step": 2839
+    },
+    {
+      "epoch": 0.43558282208588955,
+      "grad_norm": 0.17815108597278595,
+      "learning_rate": 9.13838970851193e-05,
+      "loss": 1.0892,
+      "step": 2840
+    },
+    {
+      "epoch": 0.4357361963190184,
+      "grad_norm": 0.17443816363811493,
+      "learning_rate": 9.134827489851164e-05,
+      "loss": 1.0868,
+      "step": 2841
+    },
+    {
+      "epoch": 0.4358895705521472,
+      "grad_norm": 0.18889440596103668,
+      "learning_rate": 9.131264883951319e-05,
+      "loss": 1.0406,
+      "step": 2842
+    },
+    {
+      "epoch": 0.43604294478527605,
+      "grad_norm": 0.17918244004249573,
+      "learning_rate": 9.12770189165627e-05,
+      "loss": 1.1236,
+      "step": 2843
+    },
+    {
+      "epoch": 0.4361963190184049,
+      "grad_norm": 0.16756457090377808,
+      "learning_rate": 9.124138513809972e-05,
+      "loss": 1.0087,
+      "step": 2844
+    },
+    {
+      "epoch": 0.4363496932515337,
+      "grad_norm": 0.18427762389183044,
+      "learning_rate": 9.120574751256479e-05,
+      "loss": 1.1161,
+      "step": 2845
+    },
+    {
+      "epoch": 0.43650306748466255,
+      "grad_norm": 0.1851513385772705,
+      "learning_rate": 9.117010604839937e-05,
+      "loss": 1.2121,
+      "step": 2846
+    },
+    {
+      "epoch": 0.4366564417177914,
+      "grad_norm": 0.18116386234760284,
+      "learning_rate": 9.113446075404574e-05,
+      "loss": 1.22,
+      "step": 2847
+    },
+    {
+      "epoch": 0.4368098159509202,
+      "grad_norm": 0.1848672330379486,
+      "learning_rate": 9.109881163794719e-05,
+      "loss": 1.1184,
+      "step": 2848
+    },
+    {
+      "epoch": 0.43696319018404906,
+      "grad_norm": 0.1753009706735611,
+      "learning_rate": 9.106315870854784e-05,
+      "loss": 1.1319,
+      "step": 2849
+    },
+    {
+      "epoch": 0.4371165644171779,
+      "grad_norm": 0.18137356638908386,
+      "learning_rate": 9.102750197429279e-05,
+      "loss": 1.1233,
+      "step": 2850
+    },
+    {
+      "epoch": 0.4372699386503067,
+      "grad_norm": 0.17534518241882324,
+      "learning_rate": 9.099184144362795e-05,
+      "loss": 1.1799,
+      "step": 2851
+    },
+    {
+      "epoch": 0.43742331288343556,
+      "grad_norm": 0.16961492598056793,
+      "learning_rate": 9.095617712500019e-05,
+      "loss": 1.0423,
+      "step": 2852
+    },
+    {
+      "epoch": 0.43757668711656444,
+      "grad_norm": 0.18844301998615265,
+      "learning_rate": 9.092050902685729e-05,
+      "loss": 1.228,
+      "step": 2853
+    },
+    {
+      "epoch": 0.4377300613496933,
+      "grad_norm": 0.21648100018501282,
+      "learning_rate": 9.088483715764784e-05,
+      "loss": 1.0992,
+      "step": 2854
+    },
+    {
+      "epoch": 0.4378834355828221,
+      "grad_norm": 0.1775430142879486,
+      "learning_rate": 9.084916152582143e-05,
+      "loss": 1.0835,
+      "step": 2855
+    },
+    {
+      "epoch": 0.43803680981595094,
+      "grad_norm": 0.1744329035282135,
+      "learning_rate": 9.081348213982847e-05,
+      "loss": 1.0427,
+      "step": 2856
+    },
+    {
+      "epoch": 0.4381901840490798,
+      "grad_norm": 0.17121067643165588,
+      "learning_rate": 9.077779900812029e-05,
+      "loss": 0.9422,
+      "step": 2857
+    },
+    {
+      "epoch": 0.4383435582822086,
+      "grad_norm": 0.1786738485097885,
+      "learning_rate": 9.074211213914907e-05,
+      "loss": 1.1348,
+      "step": 2858
+    },
+    {
+      "epoch": 0.43849693251533745,
+      "grad_norm": 0.1809164434671402,
+      "learning_rate": 9.070642154136798e-05,
+      "loss": 1.1191,
+      "step": 2859
+    },
+    {
+      "epoch": 0.4386503067484663,
+      "grad_norm": 0.1744803935289383,
+      "learning_rate": 9.067072722323092e-05,
+      "loss": 1.1945,
+      "step": 2860
+    },
+    {
+      "epoch": 0.4388036809815951,
+      "grad_norm": 0.18012723326683044,
+      "learning_rate": 9.063502919319276e-05,
+      "loss": 1.1767,
+      "step": 2861
+    },
+    {
+      "epoch": 0.43895705521472395,
+      "grad_norm": 0.16661274433135986,
+      "learning_rate": 9.059932745970927e-05,
+      "loss": 1.0772,
+      "step": 2862
+    },
+    {
+      "epoch": 0.4391104294478528,
+      "grad_norm": 0.16723419725894928,
+      "learning_rate": 9.056362203123705e-05,
+      "loss": 1.0335,
+      "step": 2863
+    },
+    {
+      "epoch": 0.4392638036809816,
+      "grad_norm": 0.7214125990867615,
+      "learning_rate": 9.052791291623357e-05,
+      "loss": 1.2568,
+      "step": 2864
+    },
+    {
+      "epoch": 0.43941717791411045,
+      "grad_norm": 0.1671818494796753,
+      "learning_rate": 9.049220012315721e-05,
+      "loss": 1.0484,
+      "step": 2865
+    },
+    {
+      "epoch": 0.4395705521472393,
+      "grad_norm": 0.20968560874462128,
+      "learning_rate": 9.045648366046722e-05,
+      "loss": 0.9994,
+      "step": 2866
+    },
+    {
+      "epoch": 0.4397239263803681,
+      "grad_norm": 0.35248351097106934,
+      "learning_rate": 9.04207635366237e-05,
+      "loss": 1.168,
+      "step": 2867
+    },
+    {
+      "epoch": 0.43987730061349695,
+      "grad_norm": 0.2248559445142746,
+      "learning_rate": 9.038503976008758e-05,
+      "loss": 1.0914,
+      "step": 2868
+    },
+    {
+      "epoch": 0.4400306748466258,
+      "grad_norm": 0.2192009538412094,
+      "learning_rate": 9.034931233932075e-05,
+      "loss": 1.1314,
+      "step": 2869
+    },
+    {
+      "epoch": 0.4401840490797546,
+      "grad_norm": 0.21001996099948883,
+      "learning_rate": 9.031358128278588e-05,
+      "loss": 1.2913,
+      "step": 2870
+    },
+    {
+      "epoch": 0.44033742331288345,
+      "grad_norm": 0.2003633826971054,
+      "learning_rate": 9.027784659894652e-05,
+      "loss": 1.154,
+      "step": 2871
+    },
+    {
+      "epoch": 0.4404907975460123,
+      "grad_norm": 0.22259283065795898,
+      "learning_rate": 9.024210829626712e-05,
+      "loss": 1.0245,
+      "step": 2872
+    },
+    {
+      "epoch": 0.4406441717791411,
+      "grad_norm": 0.1818084716796875,
+      "learning_rate": 9.020636638321292e-05,
+      "loss": 1.0608,
+      "step": 2873
+    },
+    {
+      "epoch": 0.44079754601226995,
+      "grad_norm": 0.17804022133350372,
+      "learning_rate": 9.01706208682501e-05,
+      "loss": 1.0541,
+      "step": 2874
+    },
+    {
+      "epoch": 0.4409509202453988,
+      "grad_norm": 0.2011212557554245,
+      "learning_rate": 9.013487175984561e-05,
+      "loss": 1.1428,
+      "step": 2875
+    },
+    {
+      "epoch": 0.4411042944785276,
+      "grad_norm": 0.1793699562549591,
+      "learning_rate": 9.009911906646729e-05,
+      "loss": 1.0997,
+      "step": 2876
+    },
+    {
+      "epoch": 0.44125766871165645,
+      "grad_norm": 0.3107013702392578,
+      "learning_rate": 9.006336279658383e-05,
+      "loss": 1.1244,
+      "step": 2877
+    },
+    {
+      "epoch": 0.4414110429447853,
+      "grad_norm": 0.1796242594718933,
+      "learning_rate": 9.002760295866477e-05,
+      "loss": 1.1024,
+      "step": 2878
+    },
+    {
+      "epoch": 0.4415644171779141,
+      "grad_norm": 0.1730370670557022,
+      "learning_rate": 8.999183956118045e-05,
+      "loss": 0.9896,
+      "step": 2879
+    },
+    {
+      "epoch": 0.44171779141104295,
+      "grad_norm": 0.18172144889831543,
+      "learning_rate": 8.995607261260216e-05,
+      "loss": 1.0293,
+      "step": 2880
+    },
+    {
+      "epoch": 0.4418711656441718,
+      "grad_norm": 0.18349625170230865,
+      "learning_rate": 8.992030212140193e-05,
+      "loss": 1.1076,
+      "step": 2881
+    },
+    {
+      "epoch": 0.4420245398773006,
+      "grad_norm": 0.18350404500961304,
+      "learning_rate": 8.988452809605263e-05,
+      "loss": 1.0183,
+      "step": 2882
+    },
+    {
+      "epoch": 0.44217791411042945,
+      "grad_norm": 0.5083114504814148,
+      "learning_rate": 8.984875054502803e-05,
+      "loss": 1.0508,
+      "step": 2883
+    },
+    {
+      "epoch": 0.4423312883435583,
+      "grad_norm": 0.20792530477046967,
+      "learning_rate": 8.981296947680271e-05,
+      "loss": 1.212,
+      "step": 2884
+    },
+    {
+      "epoch": 0.4424846625766871,
+      "grad_norm": 0.22570449113845825,
+      "learning_rate": 8.977718489985205e-05,
+      "loss": 1.0796,
+      "step": 2885
+    },
+    {
+      "epoch": 0.44263803680981595,
+      "grad_norm": 0.20488135516643524,
+      "learning_rate": 8.974139682265231e-05,
+      "loss": 1.2089,
+      "step": 2886
+    },
+    {
+      "epoch": 0.4427914110429448,
+      "grad_norm": 0.18447823822498322,
+      "learning_rate": 8.970560525368054e-05,
+      "loss": 1.2665,
+      "step": 2887
+    },
+    {
+      "epoch": 0.4429447852760736,
+      "grad_norm": 0.19167743623256683,
+      "learning_rate": 8.966981020141468e-05,
+      "loss": 1.1568,
+      "step": 2888
+    },
+    {
+      "epoch": 0.44309815950920245,
+      "grad_norm": 0.3244103491306305,
+      "learning_rate": 8.963401167433338e-05,
+      "loss": 1.208,
+      "step": 2889
+    },
+    {
+      "epoch": 0.4432515337423313,
+      "grad_norm": 0.6652431488037109,
+      "learning_rate": 8.959820968091622e-05,
+      "loss": 1.2313,
+      "step": 2890
+    },
+    {
+      "epoch": 0.4434049079754601,
+      "grad_norm": 0.18742717802524567,
+      "learning_rate": 8.956240422964354e-05,
+      "loss": 1.0503,
+      "step": 2891
+    },
+    {
+      "epoch": 0.44355828220858895,
+      "grad_norm": 0.2239595353603363,
+      "learning_rate": 8.952659532899656e-05,
+      "loss": 1.1535,
+      "step": 2892
+    },
+    {
+      "epoch": 0.4437116564417178,
+      "grad_norm": 0.19431546330451965,
+      "learning_rate": 8.949078298745725e-05,
+      "loss": 1.1189,
+      "step": 2893
+    },
+    {
+      "epoch": 0.4438650306748466,
+      "grad_norm": 0.2843902111053467,
+      "learning_rate": 8.945496721350845e-05,
+      "loss": 1.1999,
+      "step": 2894
+    },
+    {
+      "epoch": 0.44401840490797545,
+      "grad_norm": 0.18931303918361664,
+      "learning_rate": 8.941914801563375e-05,
+      "loss": 1.0446,
+      "step": 2895
+    },
+    {
+      "epoch": 0.4441717791411043,
+      "grad_norm": 0.1776357889175415,
+      "learning_rate": 8.938332540231762e-05,
+      "loss": 1.0937,
+      "step": 2896
+    },
+    {
+      "epoch": 0.4443251533742331,
+      "grad_norm": 0.19808872044086456,
+      "learning_rate": 8.934749938204529e-05,
+      "loss": 0.9871,
+      "step": 2897
+    },
+    {
+      "epoch": 0.44447852760736195,
+      "grad_norm": 0.19629482924938202,
+      "learning_rate": 8.931166996330282e-05,
+      "loss": 1.0669,
+      "step": 2898
+    },
+    {
+      "epoch": 0.4446319018404908,
+      "grad_norm": 0.40424779057502747,
+      "learning_rate": 8.927583715457708e-05,
+      "loss": 1.0967,
+      "step": 2899
+    },
+    {
+      "epoch": 0.4447852760736196,
+      "grad_norm": 0.1839829981327057,
+      "learning_rate": 8.92400009643557e-05,
+      "loss": 1.1081,
+      "step": 2900
+    },
+    {
+      "epoch": 0.44493865030674845,
+      "grad_norm": 0.2133030891418457,
+      "learning_rate": 8.920416140112719e-05,
+      "loss": 1.1046,
+      "step": 2901
+    },
+    {
+      "epoch": 0.4450920245398773,
+      "grad_norm": 0.1961429864168167,
+      "learning_rate": 8.916831847338079e-05,
+      "loss": 1.2604,
+      "step": 2902
+    },
+    {
+      "epoch": 0.4452453987730061,
+      "grad_norm": 0.30197247862815857,
+      "learning_rate": 8.913247218960655e-05,
+      "loss": 1.1713,
+      "step": 2903
+    },
+    {
+      "epoch": 0.44539877300613495,
+      "grad_norm": 0.20241853594779968,
+      "learning_rate": 8.909662255829532e-05,
+      "loss": 1.1671,
+      "step": 2904
+    },
+    {
+      "epoch": 0.4455521472392638,
+      "grad_norm": 0.18751366436481476,
+      "learning_rate": 8.906076958793876e-05,
+      "loss": 1.1431,
+      "step": 2905
+    },
+    {
+      "epoch": 0.4457055214723926,
+      "grad_norm": 0.18559636175632477,
+      "learning_rate": 8.902491328702933e-05,
+      "loss": 1.1466,
+      "step": 2906
+    },
+    {
+      "epoch": 0.44585889570552145,
+      "grad_norm": 0.1762618124485016,
+      "learning_rate": 8.89890536640602e-05,
+      "loss": 1.0696,
+      "step": 2907
+    },
+    {
+      "epoch": 0.4460122699386503,
+      "grad_norm": 0.19974097609519958,
+      "learning_rate": 8.895319072752545e-05,
+      "loss": 1.0935,
+      "step": 2908
+    },
+    {
+      "epoch": 0.4461656441717791,
+      "grad_norm": 0.17399561405181885,
+      "learning_rate": 8.891732448591984e-05,
+      "loss": 1.0709,
+      "step": 2909
+    },
+    {
+      "epoch": 0.44631901840490795,
+      "grad_norm": 0.18627117574214935,
+      "learning_rate": 8.888145494773895e-05,
+      "loss": 1.1562,
+      "step": 2910
+    },
+    {
+      "epoch": 0.4464723926380368,
+      "grad_norm": 0.20805209875106812,
+      "learning_rate": 8.884558212147915e-05,
+      "loss": 1.2176,
+      "step": 2911
+    },
+    {
+      "epoch": 0.4466257668711656,
+      "grad_norm": 0.18888361752033234,
+      "learning_rate": 8.880970601563758e-05,
+      "loss": 0.9561,
+      "step": 2912
+    },
+    {
+      "epoch": 0.44677914110429445,
+      "grad_norm": 0.19582141935825348,
+      "learning_rate": 8.877382663871216e-05,
+      "loss": 1.1938,
+      "step": 2913
+    },
+    {
+      "epoch": 0.4469325153374233,
+      "grad_norm": 0.20500609278678894,
+      "learning_rate": 8.873794399920158e-05,
+      "loss": 1.1892,
+      "step": 2914
+    },
+    {
+      "epoch": 0.4470858895705521,
+      "grad_norm": 0.17617574334144592,
+      "learning_rate": 8.870205810560532e-05,
+      "loss": 1.0534,
+      "step": 2915
+    },
+    {
+      "epoch": 0.447239263803681,
+      "grad_norm": 0.18535910546779633,
+      "learning_rate": 8.86661689664236e-05,
+      "loss": 1.1228,
+      "step": 2916
+    },
+    {
+      "epoch": 0.44739263803680984,
+      "grad_norm": 0.1852298378944397,
+      "learning_rate": 8.863027659015743e-05,
+      "loss": 1.0502,
+      "step": 2917
+    },
+    {
+      "epoch": 0.4475460122699387,
+      "grad_norm": 0.20653344690799713,
+      "learning_rate": 8.859438098530857e-05,
+      "loss": 1.17,
+      "step": 2918
+    },
+    {
+      "epoch": 0.4476993865030675,
+      "grad_norm": 0.1811387836933136,
+      "learning_rate": 8.855848216037958e-05,
+      "loss": 1.1578,
+      "step": 2919
+    },
+    {
+      "epoch": 0.44785276073619634,
+      "grad_norm": 0.17355945706367493,
+      "learning_rate": 8.852258012387373e-05,
+      "loss": 1.0851,
+      "step": 2920
+    },
+    {
+      "epoch": 0.4480061349693252,
+      "grad_norm": 0.1685906946659088,
+      "learning_rate": 8.848667488429512e-05,
+      "loss": 1.0068,
+      "step": 2921
+    },
+    {
+      "epoch": 0.448159509202454,
+      "grad_norm": 0.3505186140537262,
+      "learning_rate": 8.845076645014853e-05,
+      "loss": 1.1433,
+      "step": 2922
+    },
+    {
+      "epoch": 0.44831288343558284,
+      "grad_norm": 0.17260311543941498,
+      "learning_rate": 8.841485482993958e-05,
+      "loss": 1.0905,
+      "step": 2923
+    },
+    {
+      "epoch": 0.4484662576687117,
+      "grad_norm": 0.15919803082942963,
+      "learning_rate": 8.837894003217454e-05,
+      "loss": 0.9049,
+      "step": 2924
+    },
+    {
+      "epoch": 0.4486196319018405,
+      "grad_norm": 0.21413205564022064,
+      "learning_rate": 8.834302206536054e-05,
+      "loss": 1.0494,
+      "step": 2925
+    },
+    {
+      "epoch": 0.44877300613496934,
+      "grad_norm": 0.15951237082481384,
+      "learning_rate": 8.830710093800541e-05,
+      "loss": 0.9383,
+      "step": 2926
+    },
+    {
+      "epoch": 0.4489263803680982,
+      "grad_norm": 0.17997069656848907,
+      "learning_rate": 8.82711766586177e-05,
+      "loss": 1.1346,
+      "step": 2927
+    },
+    {
+      "epoch": 0.449079754601227,
+      "grad_norm": 0.19919730722904205,
+      "learning_rate": 8.823524923570679e-05,
+      "loss": 1.2961,
+      "step": 2928
+    },
+    {
+      "epoch": 0.44923312883435584,
+      "grad_norm": 0.1739238053560257,
+      "learning_rate": 8.81993186777827e-05,
+      "loss": 1.104,
+      "step": 2929
+    },
+    {
+      "epoch": 0.4493865030674847,
+      "grad_norm": 0.18657776713371277,
+      "learning_rate": 8.816338499335629e-05,
+      "loss": 0.9607,
+      "step": 2930
+    },
+    {
+      "epoch": 0.4495398773006135,
+      "grad_norm": 0.1863935887813568,
+      "learning_rate": 8.812744819093909e-05,
+      "loss": 1.0867,
+      "step": 2931
+    },
+    {
+      "epoch": 0.44969325153374234,
+      "grad_norm": 0.3487640917301178,
+      "learning_rate": 8.809150827904341e-05,
+      "loss": 1.2363,
+      "step": 2932
+    },
+    {
+      "epoch": 0.4498466257668712,
+      "grad_norm": 0.17084182798862457,
+      "learning_rate": 8.805556526618225e-05,
+      "loss": 1.0834,
+      "step": 2933
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.18539951741695404,
+      "learning_rate": 8.801961916086942e-05,
+      "loss": 1.1437,
+      "step": 2934
+    },
+    {
+      "epoch": 0.45015337423312884,
+      "grad_norm": 0.1797419935464859,
+      "learning_rate": 8.79836699716194e-05,
+      "loss": 1.0443,
+      "step": 2935
+    },
+    {
+      "epoch": 0.4503067484662577,
+      "grad_norm": 0.3089222013950348,
+      "learning_rate": 8.79477177069474e-05,
+      "loss": 0.9502,
+      "step": 2936
+    },
+    {
+      "epoch": 0.4504601226993865,
+      "grad_norm": 0.32173117995262146,
+      "learning_rate": 8.791176237536941e-05,
+      "loss": 1.1091,
+      "step": 2937
+    },
+    {
+      "epoch": 0.45061349693251534,
+      "grad_norm": 0.17575564980506897,
+      "learning_rate": 8.78758039854021e-05,
+      "loss": 1.0957,
+      "step": 2938
+    },
+    {
+      "epoch": 0.4507668711656442,
+      "grad_norm": 0.19442306458950043,
+      "learning_rate": 8.783984254556288e-05,
+      "loss": 1.2121,
+      "step": 2939
+    },
+    {
+      "epoch": 0.450920245398773,
+      "grad_norm": 0.20062631368637085,
+      "learning_rate": 8.780387806436988e-05,
+      "loss": 1.2142,
+      "step": 2940
+    },
+    {
+      "epoch": 0.45107361963190185,
+      "grad_norm": 0.21505619585514069,
+      "learning_rate": 8.776791055034194e-05,
+      "loss": 1.1804,
+      "step": 2941
+    },
+    {
+      "epoch": 0.4512269938650307,
+      "grad_norm": 0.1857578456401825,
+      "learning_rate": 8.773194001199865e-05,
+      "loss": 1.157,
+      "step": 2942
+    },
+    {
+      "epoch": 0.4513803680981595,
+      "grad_norm": 0.18144625425338745,
+      "learning_rate": 8.769596645786027e-05,
+      "loss": 1.0331,
+      "step": 2943
+    },
+    {
+      "epoch": 0.45153374233128835,
+      "grad_norm": 0.17125898599624634,
+      "learning_rate": 8.765998989644783e-05,
+      "loss": 1.1195,
+      "step": 2944
+    },
+    {
+      "epoch": 0.4516871165644172,
+      "grad_norm": 0.18194173276424408,
+      "learning_rate": 8.762401033628304e-05,
+      "loss": 1.1256,
+      "step": 2945
+    },
+    {
+      "epoch": 0.451840490797546,
+      "grad_norm": 0.173585444688797,
+      "learning_rate": 8.758802778588831e-05,
+      "loss": 1.1469,
+      "step": 2946
+    },
+    {
+      "epoch": 0.45199386503067485,
+      "grad_norm": 0.19690896570682526,
+      "learning_rate": 8.755204225378678e-05,
+      "loss": 1.0918,
+      "step": 2947
+    },
+    {
+      "epoch": 0.4521472392638037,
+      "grad_norm": 0.18373191356658936,
+      "learning_rate": 8.751605374850227e-05,
+      "loss": 1.1688,
+      "step": 2948
+    },
+    {
+      "epoch": 0.4523006134969325,
+      "grad_norm": 0.18684855103492737,
+      "learning_rate": 8.748006227855933e-05,
+      "loss": 1.0545,
+      "step": 2949
+    },
+    {
+      "epoch": 0.45245398773006135,
+      "grad_norm": 0.22068950533866882,
+      "learning_rate": 8.74440678524832e-05,
+      "loss": 1.1362,
+      "step": 2950
+    },
+    {
+      "epoch": 0.4526073619631902,
+      "grad_norm": 0.35876670479774475,
+      "learning_rate": 8.740807047879988e-05,
+      "loss": 1.1032,
+      "step": 2951
+    },
+    {
+      "epoch": 0.452760736196319,
+      "grad_norm": 0.18497459590435028,
+      "learning_rate": 8.737207016603593e-05,
+      "loss": 1.1833,
+      "step": 2952
+    },
+    {
+      "epoch": 0.45291411042944785,
+      "grad_norm": 0.18193531036376953,
+      "learning_rate": 8.733606692271874e-05,
+      "loss": 1.1583,
+      "step": 2953
+    },
+    {
+      "epoch": 0.4530674846625767,
+      "grad_norm": 0.20839650928974152,
+      "learning_rate": 8.730006075737631e-05,
+      "loss": 1.0603,
+      "step": 2954
+    },
+    {
+      "epoch": 0.4532208588957055,
+      "grad_norm": 0.19117304682731628,
+      "learning_rate": 8.72640516785374e-05,
+      "loss": 1.1713,
+      "step": 2955
+    },
+    {
+      "epoch": 0.45337423312883435,
+      "grad_norm": 0.19122177362442017,
+      "learning_rate": 8.72280396947314e-05,
+      "loss": 1.1261,
+      "step": 2956
+    },
+    {
+      "epoch": 0.4535276073619632,
+      "grad_norm": 0.16023336350917816,
+      "learning_rate": 8.719202481448842e-05,
+      "loss": 0.9697,
+      "step": 2957
+    },
+    {
+      "epoch": 0.453680981595092,
+      "grad_norm": 0.18539200723171234,
+      "learning_rate": 8.715600704633925e-05,
+      "loss": 1.133,
+      "step": 2958
+    },
+    {
+      "epoch": 0.45383435582822085,
+      "grad_norm": 0.28000709414482117,
+      "learning_rate": 8.711998639881536e-05,
+      "loss": 1.0926,
+      "step": 2959
+    },
+    {
+      "epoch": 0.4539877300613497,
+      "grad_norm": 0.1677703857421875,
+      "learning_rate": 8.708396288044894e-05,
+      "loss": 1.0452,
+      "step": 2960
+    },
+    {
+      "epoch": 0.4541411042944785,
+      "grad_norm": 0.1856243908405304,
+      "learning_rate": 8.704793649977277e-05,
+      "loss": 1.1985,
+      "step": 2961
+    },
+    {
+      "epoch": 0.45429447852760735,
+      "grad_norm": 0.18393976986408234,
+      "learning_rate": 8.701190726532038e-05,
+      "loss": 1.1156,
+      "step": 2962
+    },
+    {
+      "epoch": 0.4544478527607362,
+      "grad_norm": 0.449712336063385,
+      "learning_rate": 8.697587518562597e-05,
+      "loss": 1.1456,
+      "step": 2963
+    },
+    {
+      "epoch": 0.454601226993865,
+      "grad_norm": 0.21187464892864227,
+      "learning_rate": 8.69398402692244e-05,
+      "loss": 1.0275,
+      "step": 2964
+    },
+    {
+      "epoch": 0.45475460122699385,
+      "grad_norm": 0.18304894864559174,
+      "learning_rate": 8.690380252465122e-05,
+      "loss": 1.2587,
+      "step": 2965
+    },
+    {
+      "epoch": 0.4549079754601227,
+      "grad_norm": 0.17827458679676056,
+      "learning_rate": 8.686776196044262e-05,
+      "loss": 1.0401,
+      "step": 2966
+    },
+    {
+      "epoch": 0.4550613496932515,
+      "grad_norm": 0.18347550928592682,
+      "learning_rate": 8.683171858513547e-05,
+      "loss": 1.1513,
+      "step": 2967
+    },
+    {
+      "epoch": 0.45521472392638035,
+      "grad_norm": 0.1998286098241806,
+      "learning_rate": 8.679567240726733e-05,
+      "loss": 1.1868,
+      "step": 2968
+    },
+    {
+      "epoch": 0.4553680981595092,
+      "grad_norm": 0.18630172312259674,
+      "learning_rate": 8.675962343537636e-05,
+      "loss": 1.211,
+      "step": 2969
+    },
+    {
+      "epoch": 0.455521472392638,
+      "grad_norm": 0.1745014488697052,
+      "learning_rate": 8.672357167800146e-05,
+      "loss": 1.1008,
+      "step": 2970
+    },
+    {
+      "epoch": 0.45567484662576685,
+      "grad_norm": 0.17459824681282043,
+      "learning_rate": 8.668751714368214e-05,
+      "loss": 1.1818,
+      "step": 2971
+    },
+    {
+      "epoch": 0.4558282208588957,
+      "grad_norm": 0.18815895915031433,
+      "learning_rate": 8.665145984095859e-05,
+      "loss": 1.0574,
+      "step": 2972
+    },
+    {
+      "epoch": 0.4559815950920245,
+      "grad_norm": 0.18239791691303253,
+      "learning_rate": 8.661539977837167e-05,
+      "loss": 1.009,
+      "step": 2973
+    },
+    {
+      "epoch": 0.45613496932515335,
+      "grad_norm": 0.1724919080734253,
+      "learning_rate": 8.657933696446283e-05,
+      "loss": 1.1219,
+      "step": 2974
+    },
+    {
+      "epoch": 0.4562883435582822,
+      "grad_norm": 0.18050749599933624,
+      "learning_rate": 8.654327140777423e-05,
+      "loss": 1.0086,
+      "step": 2975
+    },
+    {
+      "epoch": 0.456441717791411,
+      "grad_norm": 0.1557498574256897,
+      "learning_rate": 8.650720311684866e-05,
+      "loss": 0.8415,
+      "step": 2976
+    },
+    {
+      "epoch": 0.45659509202453985,
+      "grad_norm": 0.1679290235042572,
+      "learning_rate": 8.647113210022955e-05,
+      "loss": 1.093,
+      "step": 2977
+    },
+    {
+      "epoch": 0.45674846625766874,
+      "grad_norm": 0.20219172537326813,
+      "learning_rate": 8.643505836646103e-05,
+      "loss": 1.0162,
+      "step": 2978
+    },
+    {
+      "epoch": 0.4569018404907976,
+      "grad_norm": 0.19633986055850983,
+      "learning_rate": 8.639898192408778e-05,
+      "loss": 1.1133,
+      "step": 2979
+    },
+    {
+      "epoch": 0.4570552147239264,
+      "grad_norm": 0.19668221473693848,
+      "learning_rate": 8.63629027816552e-05,
+      "loss": 1.2484,
+      "step": 2980
+    },
+    {
+      "epoch": 0.45720858895705524,
+      "grad_norm": 0.17567576467990875,
+      "learning_rate": 8.63268209477093e-05,
+      "loss": 1.2012,
+      "step": 2981
+    },
+    {
+      "epoch": 0.4573619631901841,
+      "grad_norm": 0.16535621881484985,
+      "learning_rate": 8.62907364307967e-05,
+      "loss": 1.0171,
+      "step": 2982
+    },
+    {
+      "epoch": 0.4575153374233129,
+      "grad_norm": 0.1767207384109497,
+      "learning_rate": 8.625464923946474e-05,
+      "loss": 1.1234,
+      "step": 2983
+    },
+    {
+      "epoch": 0.45766871165644174,
+      "grad_norm": 0.18185818195343018,
+      "learning_rate": 8.621855938226128e-05,
+      "loss": 1.1557,
+      "step": 2984
+    },
+    {
+      "epoch": 0.4578220858895706,
+      "grad_norm": 0.18163810670375824,
+      "learning_rate": 8.618246686773486e-05,
+      "loss": 1.1578,
+      "step": 2985
+    },
+    {
+      "epoch": 0.4579754601226994,
+      "grad_norm": 0.19656196236610413,
+      "learning_rate": 8.614637170443474e-05,
+      "loss": 1.2519,
+      "step": 2986
+    },
+    {
+      "epoch": 0.45812883435582824,
+      "grad_norm": 0.16383744776248932,
+      "learning_rate": 8.611027390091067e-05,
+      "loss": 0.9921,
+      "step": 2987
+    },
+    {
+      "epoch": 0.4582822085889571,
+      "grad_norm": 0.18539997935295105,
+      "learning_rate": 8.607417346571307e-05,
+      "loss": 1.1284,
+      "step": 2988
+    },
+    {
+      "epoch": 0.4584355828220859,
+      "grad_norm": 0.3120240867137909,
+      "learning_rate": 8.603807040739302e-05,
+      "loss": 1.1798,
+      "step": 2989
+    },
+    {
+      "epoch": 0.45858895705521474,
+      "grad_norm": 0.17893950641155243,
+      "learning_rate": 8.600196473450218e-05,
+      "loss": 1.148,
+      "step": 2990
+    },
+    {
+      "epoch": 0.4587423312883436,
+      "grad_norm": 0.19057869911193848,
+      "learning_rate": 8.596585645559287e-05,
+      "loss": 1.2022,
+      "step": 2991
+    },
+    {
+      "epoch": 0.4588957055214724,
+      "grad_norm": 0.18307508528232574,
+      "learning_rate": 8.592974557921798e-05,
+      "loss": 1.0807,
+      "step": 2992
+    },
+    {
+      "epoch": 0.45904907975460124,
+      "grad_norm": 0.16110311448574066,
+      "learning_rate": 8.589363211393105e-05,
+      "loss": 0.9842,
+      "step": 2993
+    },
+    {
+      "epoch": 0.4592024539877301,
+      "grad_norm": 0.16933010518550873,
+      "learning_rate": 8.58575160682862e-05,
+      "loss": 1.1023,
+      "step": 2994
+    },
+    {
+      "epoch": 0.4593558282208589,
+      "grad_norm": 0.1768992692232132,
+      "learning_rate": 8.582139745083821e-05,
+      "loss": 1.1502,
+      "step": 2995
+    },
+    {
+      "epoch": 0.45950920245398774,
+      "grad_norm": 0.18022505939006805,
+      "learning_rate": 8.578527627014243e-05,
+      "loss": 1.1628,
+      "step": 2996
+    },
+    {
+      "epoch": 0.4596625766871166,
+      "grad_norm": 0.17020092904567719,
+      "learning_rate": 8.574915253475481e-05,
+      "loss": 1.0793,
+      "step": 2997
+    },
+    {
+      "epoch": 0.4598159509202454,
+      "grad_norm": 0.1807938516139984,
+      "learning_rate": 8.571302625323195e-05,
+      "loss": 1.2805,
+      "step": 2998
+    },
+    {
+      "epoch": 0.45996932515337424,
+      "grad_norm": 0.17949186265468597,
+      "learning_rate": 8.567689743413103e-05,
+      "loss": 1.0989,
+      "step": 2999
+    },
+    {
+      "epoch": 0.4601226993865031,
+      "grad_norm": 0.1762734055519104,
+      "learning_rate": 8.56407660860098e-05,
+      "loss": 1.0,
+      "step": 3000
+    },
+    {
+      "epoch": 0.4602760736196319,
+      "grad_norm": 0.20224101841449738,
+      "learning_rate": 8.560463221742666e-05,
+      "loss": 1.1494,
+      "step": 3001
+    },
+    {
+      "epoch": 0.46042944785276074,
+      "grad_norm": 0.18240110576152802,
+      "learning_rate": 8.556849583694059e-05,
+      "loss": 1.2105,
+      "step": 3002
+    },
+    {
+      "epoch": 0.4605828220858896,
+      "grad_norm": 0.2938978970050812,
+      "learning_rate": 8.553235695311113e-05,
+      "loss": 1.0838,
+      "step": 3003
+    },
+    {
+      "epoch": 0.4607361963190184,
+      "grad_norm": 0.1756850928068161,
+      "learning_rate": 8.549621557449846e-05,
+      "loss": 1.0785,
+      "step": 3004
+    },
+    {
+      "epoch": 0.46088957055214724,
+      "grad_norm": 0.17004607617855072,
+      "learning_rate": 8.546007170966333e-05,
+      "loss": 1.0648,
+      "step": 3005
+    },
+    {
+      "epoch": 0.4610429447852761,
+      "grad_norm": 0.19428950548171997,
+      "learning_rate": 8.542392536716706e-05,
+      "loss": 0.9693,
+      "step": 3006
+    },
+    {
+      "epoch": 0.4611963190184049,
+      "grad_norm": 0.18621695041656494,
+      "learning_rate": 8.538777655557162e-05,
+      "loss": 1.0715,
+      "step": 3007
+    },
+    {
+      "epoch": 0.46134969325153374,
+      "grad_norm": 0.1671433448791504,
+      "learning_rate": 8.535162528343952e-05,
+      "loss": 1.0347,
+      "step": 3008
+    },
+    {
+      "epoch": 0.4615030674846626,
+      "grad_norm": 0.16483332216739655,
+      "learning_rate": 8.531547155933382e-05,
+      "loss": 1.0166,
+      "step": 3009
+    },
+    {
+      "epoch": 0.4616564417177914,
+      "grad_norm": 0.18975330889225006,
+      "learning_rate": 8.527931539181823e-05,
+      "loss": 1.1851,
+      "step": 3010
+    },
+    {
+      "epoch": 0.46180981595092024,
+      "grad_norm": 0.1956935077905655,
+      "learning_rate": 8.524315678945698e-05,
+      "loss": 0.9155,
+      "step": 3011
+    },
+    {
+      "epoch": 0.4619631901840491,
+      "grad_norm": 0.16841883957386017,
+      "learning_rate": 8.520699576081493e-05,
+      "loss": 1.0776,
+      "step": 3012
+    },
+    {
+      "epoch": 0.4621165644171779,
+      "grad_norm": 0.1870298534631729,
+      "learning_rate": 8.517083231445745e-05,
+      "loss": 1.2156,
+      "step": 3013
+    },
+    {
+      "epoch": 0.46226993865030674,
+      "grad_norm": 0.2105487734079361,
+      "learning_rate": 8.513466645895056e-05,
+      "loss": 1.1189,
+      "step": 3014
+    },
+    {
+      "epoch": 0.4624233128834356,
+      "grad_norm": 0.18547849357128143,
+      "learning_rate": 8.509849820286079e-05,
+      "loss": 1.2112,
+      "step": 3015
+    },
+    {
+      "epoch": 0.4625766871165644,
+      "grad_norm": 0.1823100745677948,
+      "learning_rate": 8.506232755475529e-05,
+      "loss": 1.0526,
+      "step": 3016
+    },
+    {
+      "epoch": 0.46273006134969324,
+      "grad_norm": 0.21993638575077057,
+      "learning_rate": 8.50261545232017e-05,
+      "loss": 1.0893,
+      "step": 3017
+    },
+    {
+      "epoch": 0.4628834355828221,
+      "grad_norm": 0.1928110271692276,
+      "learning_rate": 8.498997911676828e-05,
+      "loss": 1.133,
+      "step": 3018
+    },
+    {
+      "epoch": 0.4630368098159509,
+      "grad_norm": 0.1868249475955963,
+      "learning_rate": 8.495380134402385e-05,
+      "loss": 1.085,
+      "step": 3019
+    },
+    {
+      "epoch": 0.46319018404907975,
+      "grad_norm": 0.16866308450698853,
+      "learning_rate": 8.491762121353779e-05,
+      "loss": 1.0259,
+      "step": 3020
+    },
+    {
+      "epoch": 0.4633435582822086,
+      "grad_norm": 0.18015320599079132,
+      "learning_rate": 8.488143873388004e-05,
+      "loss": 1.1152,
+      "step": 3021
+    },
+    {
+      "epoch": 0.4634969325153374,
+      "grad_norm": 0.20193496346473694,
+      "learning_rate": 8.484525391362106e-05,
+      "loss": 1.2561,
+      "step": 3022
+    },
+    {
+      "epoch": 0.46365030674846625,
+      "grad_norm": 0.16302141547203064,
+      "learning_rate": 8.480906676133192e-05,
+      "loss": 1.0805,
+      "step": 3023
+    },
+    {
+      "epoch": 0.4638036809815951,
+      "grad_norm": 0.1741969734430313,
+      "learning_rate": 8.47728772855842e-05,
+      "loss": 1.1854,
+      "step": 3024
+    },
+    {
+      "epoch": 0.4639570552147239,
+      "grad_norm": 0.1711106300354004,
+      "learning_rate": 8.473668549495007e-05,
+      "loss": 1.0678,
+      "step": 3025
+    },
+    {
+      "epoch": 0.46411042944785275,
+      "grad_norm": 0.17004378139972687,
+      "learning_rate": 8.470049139800218e-05,
+      "loss": 0.9649,
+      "step": 3026
+    },
+    {
+      "epoch": 0.4642638036809816,
+      "grad_norm": 0.18481206893920898,
+      "learning_rate": 8.46642950033138e-05,
+      "loss": 1.1826,
+      "step": 3027
+    },
+    {
+      "epoch": 0.4644171779141104,
+      "grad_norm": 0.18992342054843903,
+      "learning_rate": 8.462809631945868e-05,
+      "loss": 1.2474,
+      "step": 3028
+    },
+    {
+      "epoch": 0.46457055214723925,
+      "grad_norm": 0.17712606489658356,
+      "learning_rate": 8.459189535501119e-05,
+      "loss": 1.085,
+      "step": 3029
+    },
+    {
+      "epoch": 0.4647239263803681,
+      "grad_norm": 0.16024668514728546,
+      "learning_rate": 8.455569211854621e-05,
+      "loss": 0.9886,
+      "step": 3030
+    },
+    {
+      "epoch": 0.4648773006134969,
+      "grad_norm": 0.18597331643104553,
+      "learning_rate": 8.451948661863911e-05,
+      "loss": 1.1841,
+      "step": 3031
+    },
+    {
+      "epoch": 0.46503067484662575,
+      "grad_norm": 0.20810918509960175,
+      "learning_rate": 8.448327886386582e-05,
+      "loss": 1.1854,
+      "step": 3032
+    },
+    {
+      "epoch": 0.4651840490797546,
+      "grad_norm": 0.1767541468143463,
+      "learning_rate": 8.444706886280283e-05,
+      "loss": 1.0262,
+      "step": 3033
+    },
+    {
+      "epoch": 0.4653374233128834,
+      "grad_norm": 0.17253388464450836,
+      "learning_rate": 8.441085662402716e-05,
+      "loss": 1.0553,
+      "step": 3034
+    },
+    {
+      "epoch": 0.46549079754601225,
+      "grad_norm": 0.17705030739307404,
+      "learning_rate": 8.437464215611633e-05,
+      "loss": 1.1892,
+      "step": 3035
+    },
+    {
+      "epoch": 0.4656441717791411,
+      "grad_norm": 0.20617061853408813,
+      "learning_rate": 8.433842546764844e-05,
+      "loss": 1.0752,
+      "step": 3036
+    },
+    {
+      "epoch": 0.4657975460122699,
+      "grad_norm": 0.1704021543264389,
+      "learning_rate": 8.430220656720203e-05,
+      "loss": 1.1436,
+      "step": 3037
+    },
+    {
+      "epoch": 0.46595092024539875,
+      "grad_norm": 0.16896378993988037,
+      "learning_rate": 8.426598546335624e-05,
+      "loss": 0.9426,
+      "step": 3038
+    },
+    {
+      "epoch": 0.4661042944785276,
+      "grad_norm": 0.16879817843437195,
+      "learning_rate": 8.422976216469072e-05,
+      "loss": 1.0647,
+      "step": 3039
+    },
+    {
+      "epoch": 0.4662576687116564,
+      "grad_norm": 0.16654084622859955,
+      "learning_rate": 8.41935366797856e-05,
+      "loss": 0.991,
+      "step": 3040
+    },
+    {
+      "epoch": 0.4664110429447853,
+      "grad_norm": 0.18649433553218842,
+      "learning_rate": 8.415730901722158e-05,
+      "loss": 1.114,
+      "step": 3041
+    },
+    {
+      "epoch": 0.46656441717791414,
+      "grad_norm": 0.16858479380607605,
+      "learning_rate": 8.412107918557982e-05,
+      "loss": 0.9944,
+      "step": 3042
+    },
+    {
+      "epoch": 0.46671779141104297,
+      "grad_norm": 0.17839005589485168,
+      "learning_rate": 8.408484719344206e-05,
+      "loss": 1.1056,
+      "step": 3043
+    },
+    {
+      "epoch": 0.4668711656441718,
+      "grad_norm": 0.16194123029708862,
+      "learning_rate": 8.40486130493905e-05,
+      "loss": 0.891,
+      "step": 3044
+    },
+    {
+      "epoch": 0.46702453987730064,
+      "grad_norm": 0.18518096208572388,
+      "learning_rate": 8.401237676200787e-05,
+      "loss": 1.1395,
+      "step": 3045
+    },
+    {
+      "epoch": 0.46717791411042947,
+      "grad_norm": 0.2089030146598816,
+      "learning_rate": 8.397613833987739e-05,
+      "loss": 1.0855,
+      "step": 3046
+    },
+    {
+      "epoch": 0.4673312883435583,
+      "grad_norm": 0.18343348801136017,
+      "learning_rate": 8.393989779158283e-05,
+      "loss": 1.1633,
+      "step": 3047
+    },
+    {
+      "epoch": 0.46748466257668714,
+      "grad_norm": 0.1784767210483551,
+      "learning_rate": 8.39036551257084e-05,
+      "loss": 1.1007,
+      "step": 3048
+    },
+    {
+      "epoch": 0.46763803680981597,
+      "grad_norm": 0.1746671050786972,
+      "learning_rate": 8.386741035083884e-05,
+      "loss": 1.1146,
+      "step": 3049
+    },
+    {
+      "epoch": 0.4677914110429448,
+      "grad_norm": 0.18811342120170593,
+      "learning_rate": 8.383116347555941e-05,
+      "loss": 1.186,
+      "step": 3050
+    },
+    {
+      "epoch": 0.46794478527607364,
+      "grad_norm": 0.17835330963134766,
+      "learning_rate": 8.379491450845589e-05,
+      "loss": 1.2942,
+      "step": 3051
+    },
+    {
+      "epoch": 0.46809815950920247,
+      "grad_norm": 0.16628144681453705,
+      "learning_rate": 8.375866345811446e-05,
+      "loss": 0.9986,
+      "step": 3052
+    },
+    {
+      "epoch": 0.4682515337423313,
+      "grad_norm": 0.5324901342391968,
+      "learning_rate": 8.372241033312187e-05,
+      "loss": 1.0462,
+      "step": 3053
+    },
+    {
+      "epoch": 0.46840490797546014,
+      "grad_norm": 0.16728751361370087,
+      "learning_rate": 8.368615514206536e-05,
+      "loss": 1.0945,
+      "step": 3054
+    },
+    {
+      "epoch": 0.46855828220858897,
+      "grad_norm": 0.18563146889209747,
+      "learning_rate": 8.36498978935326e-05,
+      "loss": 1.0774,
+      "step": 3055
+    },
+    {
+      "epoch": 0.4687116564417178,
+      "grad_norm": 0.16791577637195587,
+      "learning_rate": 8.361363859611182e-05,
+      "loss": 0.9779,
+      "step": 3056
+    },
+    {
+      "epoch": 0.46886503067484664,
+      "grad_norm": 0.1934942752122879,
+      "learning_rate": 8.35773772583917e-05,
+      "loss": 1.1539,
+      "step": 3057
+    },
+    {
+      "epoch": 0.4690184049079755,
+      "grad_norm": 0.17296479642391205,
+      "learning_rate": 8.35411138889614e-05,
+      "loss": 1.0257,
+      "step": 3058
+    },
+    {
+      "epoch": 0.4691717791411043,
+      "grad_norm": 0.22492633759975433,
+      "learning_rate": 8.35048484964106e-05,
+      "loss": 1.1086,
+      "step": 3059
+    },
+    {
+      "epoch": 0.46932515337423314,
+      "grad_norm": 0.338558167219162,
+      "learning_rate": 8.346858108932937e-05,
+      "loss": 0.9795,
+      "step": 3060
+    },
+    {
+      "epoch": 0.469478527607362,
+      "grad_norm": 0.166054829955101,
+      "learning_rate": 8.343231167630836e-05,
+      "loss": 0.9647,
+      "step": 3061
+    },
+    {
+      "epoch": 0.4696319018404908,
+      "grad_norm": 0.1918661892414093,
+      "learning_rate": 8.339604026593863e-05,
+      "loss": 1.0582,
+      "step": 3062
+    },
+    {
+      "epoch": 0.46978527607361964,
+      "grad_norm": 0.16336724162101746,
+      "learning_rate": 8.335976686681174e-05,
+      "loss": 1.03,
+      "step": 3063
+    },
+    {
+      "epoch": 0.4699386503067485,
+      "grad_norm": 0.1722850203514099,
+      "learning_rate": 8.332349148751971e-05,
+      "loss": 1.0777,
+      "step": 3064
+    },
+    {
+      "epoch": 0.4700920245398773,
+      "grad_norm": 0.2173023372888565,
+      "learning_rate": 8.328721413665504e-05,
+      "loss": 1.0763,
+      "step": 3065
+    },
+    {
+      "epoch": 0.47024539877300614,
+      "grad_norm": 0.19207273423671722,
+      "learning_rate": 8.32509348228107e-05,
+      "loss": 1.0086,
+      "step": 3066
+    },
+    {
+      "epoch": 0.470398773006135,
+      "grad_norm": 0.21819952130317688,
+      "learning_rate": 8.32146535545801e-05,
+      "loss": 0.9746,
+      "step": 3067
+    },
+    {
+      "epoch": 0.4705521472392638,
+      "grad_norm": 0.18284624814987183,
+      "learning_rate": 8.317837034055711e-05,
+      "loss": 1.0247,
+      "step": 3068
+    },
+    {
+      "epoch": 0.47070552147239264,
+      "grad_norm": 0.1747242659330368,
+      "learning_rate": 8.314208518933613e-05,
+      "loss": 0.996,
+      "step": 3069
+    },
+    {
+      "epoch": 0.4708588957055215,
+      "grad_norm": 0.18892984092235565,
+      "learning_rate": 8.31057981095119e-05,
+      "loss": 1.184,
+      "step": 3070
+    },
+    {
+      "epoch": 0.4710122699386503,
+      "grad_norm": 0.1812051236629486,
+      "learning_rate": 8.306950910967975e-05,
+      "loss": 0.9514,
+      "step": 3071
+    },
+    {
+      "epoch": 0.47116564417177914,
+      "grad_norm": 0.1883111298084259,
+      "learning_rate": 8.303321819843538e-05,
+      "loss": 1.1619,
+      "step": 3072
+    },
+    {
+      "epoch": 0.471319018404908,
+      "grad_norm": 0.17553798854351044,
+      "learning_rate": 8.299692538437495e-05,
+      "loss": 1.0719,
+      "step": 3073
+    },
+    {
+      "epoch": 0.4714723926380368,
+      "grad_norm": 0.17909272015094757,
+      "learning_rate": 8.296063067609507e-05,
+      "loss": 1.1599,
+      "step": 3074
+    },
+    {
+      "epoch": 0.47162576687116564,
+      "grad_norm": 0.1829393059015274,
+      "learning_rate": 8.292433408219284e-05,
+      "loss": 1.1215,
+      "step": 3075
+    },
+    {
+      "epoch": 0.4717791411042945,
+      "grad_norm": 0.18627776205539703,
+      "learning_rate": 8.288803561126578e-05,
+      "loss": 1.115,
+      "step": 3076
+    },
+    {
+      "epoch": 0.4719325153374233,
+      "grad_norm": 0.17852477729320526,
+      "learning_rate": 8.285173527191182e-05,
+      "loss": 1.1606,
+      "step": 3077
+    },
+    {
+      "epoch": 0.47208588957055214,
+      "grad_norm": 0.21346773207187653,
+      "learning_rate": 8.281543307272941e-05,
+      "loss": 1.2458,
+      "step": 3078
+    },
+    {
+      "epoch": 0.472239263803681,
+      "grad_norm": 0.17781712114810944,
+      "learning_rate": 8.277912902231735e-05,
+      "loss": 1.1521,
+      "step": 3079
+    },
+    {
+      "epoch": 0.4723926380368098,
+      "grad_norm": 0.19087755680084229,
+      "learning_rate": 8.274282312927499e-05,
+      "loss": 1.094,
+      "step": 3080
+    },
+    {
+      "epoch": 0.47254601226993864,
+      "grad_norm": 0.24578963220119476,
+      "learning_rate": 8.270651540220199e-05,
+      "loss": 1.1622,
+      "step": 3081
+    },
+    {
+      "epoch": 0.4726993865030675,
+      "grad_norm": 0.16819509863853455,
+      "learning_rate": 8.26702058496985e-05,
+      "loss": 1.0807,
+      "step": 3082
+    },
+    {
+      "epoch": 0.4728527607361963,
+      "grad_norm": 0.1703372448682785,
+      "learning_rate": 8.263389448036515e-05,
+      "loss": 1.0917,
+      "step": 3083
+    },
+    {
+      "epoch": 0.47300613496932514,
+      "grad_norm": 0.19573110342025757,
+      "learning_rate": 8.259758130280294e-05,
+      "loss": 1.1446,
+      "step": 3084
+    },
+    {
+      "epoch": 0.473159509202454,
+      "grad_norm": 0.17236454784870148,
+      "learning_rate": 8.256126632561331e-05,
+      "loss": 1.0177,
+      "step": 3085
+    },
+    {
+      "epoch": 0.4733128834355828,
+      "grad_norm": 0.1987570822238922,
+      "learning_rate": 8.252494955739815e-05,
+      "loss": 1.1595,
+      "step": 3086
+    },
+    {
+      "epoch": 0.47346625766871164,
+      "grad_norm": 0.17334100604057312,
+      "learning_rate": 8.248863100675974e-05,
+      "loss": 1.068,
+      "step": 3087
+    },
+    {
+      "epoch": 0.4736196319018405,
+      "grad_norm": 0.17860974371433258,
+      "learning_rate": 8.245231068230084e-05,
+      "loss": 0.9787,
+      "step": 3088
+    },
+    {
+      "epoch": 0.4737730061349693,
+      "grad_norm": 0.18481960892677307,
+      "learning_rate": 8.241598859262453e-05,
+      "loss": 1.2666,
+      "step": 3089
+    },
+    {
+      "epoch": 0.47392638036809814,
+      "grad_norm": 0.17158453166484833,
+      "learning_rate": 8.237966474633439e-05,
+      "loss": 1.0576,
+      "step": 3090
+    },
+    {
+      "epoch": 0.474079754601227,
+      "grad_norm": 0.20149801671504974,
+      "learning_rate": 8.234333915203437e-05,
+      "loss": 1.1363,
+      "step": 3091
+    },
+    {
+      "epoch": 0.4742331288343558,
+      "grad_norm": 0.24340111017227173,
+      "learning_rate": 8.230701181832892e-05,
+      "loss": 1.2164,
+      "step": 3092
+    },
+    {
+      "epoch": 0.47438650306748464,
+      "grad_norm": 0.16910584270954132,
+      "learning_rate": 8.227068275382281e-05,
+      "loss": 1.0331,
+      "step": 3093
+    },
+    {
+      "epoch": 0.4745398773006135,
+      "grad_norm": 0.17552785575389862,
+      "learning_rate": 8.223435196712123e-05,
+      "loss": 1.1093,
+      "step": 3094
+    },
+    {
+      "epoch": 0.4746932515337423,
+      "grad_norm": 0.16980332136154175,
+      "learning_rate": 8.219801946682982e-05,
+      "loss": 0.9721,
+      "step": 3095
+    },
+    {
+      "epoch": 0.47484662576687114,
+      "grad_norm": 0.17738831043243408,
+      "learning_rate": 8.216168526155459e-05,
+      "loss": 1.1428,
+      "step": 3096
+    },
+    {
+      "epoch": 0.475,
+      "grad_norm": 0.190529927611351,
+      "learning_rate": 8.212534935990199e-05,
+      "loss": 1.0808,
+      "step": 3097
+    },
+    {
+      "epoch": 0.4751533742331288,
+      "grad_norm": 0.174271821975708,
+      "learning_rate": 8.208901177047882e-05,
+      "loss": 1.1631,
+      "step": 3098
+    },
+    {
+      "epoch": 0.47530674846625764,
+      "grad_norm": 0.18442140519618988,
+      "learning_rate": 8.20526725018923e-05,
+      "loss": 1.139,
+      "step": 3099
+    },
+    {
+      "epoch": 0.4754601226993865,
+      "grad_norm": 0.18024253845214844,
+      "learning_rate": 8.201633156275011e-05,
+      "loss": 1.2234,
+      "step": 3100
+    },
+    {
+      "epoch": 0.4756134969325153,
+      "grad_norm": 0.16328118741512299,
+      "learning_rate": 8.197998896166025e-05,
+      "loss": 1.0217,
+      "step": 3101
+    },
+    {
+      "epoch": 0.47576687116564415,
+      "grad_norm": 0.1783020943403244,
+      "learning_rate": 8.19436447072311e-05,
+      "loss": 1.1312,
+      "step": 3102
+    },
+    {
+      "epoch": 0.475920245398773,
+      "grad_norm": 0.18994952738285065,
+      "learning_rate": 8.19072988080715e-05,
+      "loss": 1.1307,
+      "step": 3103
+    },
+    {
+      "epoch": 0.47607361963190187,
+      "grad_norm": 0.17654873430728912,
+      "learning_rate": 8.187095127279064e-05,
+      "loss": 1.0364,
+      "step": 3104
+    },
+    {
+      "epoch": 0.4762269938650307,
+      "grad_norm": 0.1869022101163864,
+      "learning_rate": 8.183460210999811e-05,
+      "loss": 1.2554,
+      "step": 3105
+    },
+    {
+      "epoch": 0.47638036809815953,
+      "grad_norm": 0.16928358376026154,
+      "learning_rate": 8.179825132830386e-05,
+      "loss": 1.107,
+      "step": 3106
+    },
+    {
+      "epoch": 0.47653374233128837,
+      "grad_norm": 0.1715088188648224,
+      "learning_rate": 8.176189893631829e-05,
+      "loss": 1.0061,
+      "step": 3107
+    },
+    {
+      "epoch": 0.4766871165644172,
+      "grad_norm": 0.1659255176782608,
+      "learning_rate": 8.172554494265209e-05,
+      "loss": 1.0616,
+      "step": 3108
+    },
+    {
+      "epoch": 0.47684049079754603,
+      "grad_norm": 0.17763087153434753,
+      "learning_rate": 8.168918935591641e-05,
+      "loss": 1.1732,
+      "step": 3109
+    },
+    {
+      "epoch": 0.47699386503067487,
+      "grad_norm": 0.17108909785747528,
+      "learning_rate": 8.16528321847227e-05,
+      "loss": 1.0494,
+      "step": 3110
+    },
+    {
+      "epoch": 0.4771472392638037,
+      "grad_norm": 0.17605121433734894,
+      "learning_rate": 8.161647343768286e-05,
+      "loss": 1.1444,
+      "step": 3111
+    },
+    {
+      "epoch": 0.47730061349693254,
+      "grad_norm": 0.17387795448303223,
+      "learning_rate": 8.15801131234091e-05,
+      "loss": 1.0872,
+      "step": 3112
+    },
+    {
+      "epoch": 0.47745398773006137,
+      "grad_norm": 0.17761188745498657,
+      "learning_rate": 8.154375125051407e-05,
+      "loss": 1.1266,
+      "step": 3113
+    },
+    {
+      "epoch": 0.4776073619631902,
+      "grad_norm": 0.18502120673656464,
+      "learning_rate": 8.150738782761073e-05,
+      "loss": 1.1781,
+      "step": 3114
+    },
+    {
+      "epoch": 0.47776073619631904,
+      "grad_norm": 0.17503711581230164,
+      "learning_rate": 8.147102286331245e-05,
+      "loss": 1.1766,
+      "step": 3115
+    },
+    {
+      "epoch": 0.47791411042944787,
+      "grad_norm": 0.16709618270397186,
+      "learning_rate": 8.143465636623291e-05,
+      "loss": 1.0638,
+      "step": 3116
+    },
+    {
+      "epoch": 0.4780674846625767,
+      "grad_norm": 0.16848856210708618,
+      "learning_rate": 8.139828834498619e-05,
+      "loss": 1.1262,
+      "step": 3117
+    },
+    {
+      "epoch": 0.47822085889570554,
+      "grad_norm": 0.16712258756160736,
+      "learning_rate": 8.136191880818674e-05,
+      "loss": 1.0431,
+      "step": 3118
+    },
+    {
+      "epoch": 0.47837423312883437,
+      "grad_norm": 0.17360533773899078,
+      "learning_rate": 8.132554776444936e-05,
+      "loss": 1.1064,
+      "step": 3119
+    },
+    {
+      "epoch": 0.4785276073619632,
+      "grad_norm": 0.17327716946601868,
+      "learning_rate": 8.128917522238921e-05,
+      "loss": 1.194,
+      "step": 3120
+    },
+    {
+      "epoch": 0.47868098159509204,
+      "grad_norm": 0.1902066171169281,
+      "learning_rate": 8.125280119062178e-05,
+      "loss": 1.0612,
+      "step": 3121
+    },
+    {
+      "epoch": 0.47883435582822087,
+      "grad_norm": 0.18563537299633026,
+      "learning_rate": 8.121642567776293e-05,
+      "loss": 1.1069,
+      "step": 3122
+    },
+    {
+      "epoch": 0.4789877300613497,
+      "grad_norm": 0.17914380133152008,
+      "learning_rate": 8.118004869242888e-05,
+      "loss": 1.0861,
+      "step": 3123
+    },
+    {
+      "epoch": 0.47914110429447854,
+      "grad_norm": 1.1535758972167969,
+      "learning_rate": 8.11436702432362e-05,
+      "loss": 1.0134,
+      "step": 3124
+    },
+    {
+      "epoch": 0.47929447852760737,
+      "grad_norm": 0.18209898471832275,
+      "learning_rate": 8.110729033880178e-05,
+      "loss": 1.1392,
+      "step": 3125
+    },
+    {
+      "epoch": 0.4794478527607362,
+      "grad_norm": 0.18520505726337433,
+      "learning_rate": 8.107090898774287e-05,
+      "loss": 1.2533,
+      "step": 3126
+    },
+    {
+      "epoch": 0.47960122699386504,
+      "grad_norm": 0.17671112716197968,
+      "learning_rate": 8.103452619867709e-05,
+      "loss": 1.1051,
+      "step": 3127
+    },
+    {
+      "epoch": 0.47975460122699387,
+      "grad_norm": 0.17208343744277954,
+      "learning_rate": 8.099814198022235e-05,
+      "loss": 1.1103,
+      "step": 3128
+    },
+    {
+      "epoch": 0.4799079754601227,
+      "grad_norm": 0.17011429369449615,
+      "learning_rate": 8.096175634099693e-05,
+      "loss": 1.0037,
+      "step": 3129
+    },
+    {
+      "epoch": 0.48006134969325154,
+      "grad_norm": 0.20530948042869568,
+      "learning_rate": 8.092536928961946e-05,
+      "loss": 1.0943,
+      "step": 3130
+    },
+    {
+      "epoch": 0.48021472392638037,
+      "grad_norm": 0.17680084705352783,
+      "learning_rate": 8.088898083470885e-05,
+      "loss": 1.1166,
+      "step": 3131
+    },
+    {
+      "epoch": 0.4803680981595092,
+      "grad_norm": 0.17588894069194794,
+      "learning_rate": 8.08525909848844e-05,
+      "loss": 1.0911,
+      "step": 3132
+    },
+    {
+      "epoch": 0.48052147239263804,
+      "grad_norm": 0.16407588124275208,
+      "learning_rate": 8.081619974876573e-05,
+      "loss": 1.0449,
+      "step": 3133
+    },
+    {
+      "epoch": 0.48067484662576687,
+      "grad_norm": 0.17733675241470337,
+      "learning_rate": 8.07798071349727e-05,
+      "loss": 1.1223,
+      "step": 3134
+    },
+    {
+      "epoch": 0.4808282208588957,
+      "grad_norm": 0.18496139347553253,
+      "learning_rate": 8.07434131521257e-05,
+      "loss": 1.1786,
+      "step": 3135
+    },
+    {
+      "epoch": 0.48098159509202454,
+      "grad_norm": 0.15911750495433807,
+      "learning_rate": 8.070701780884523e-05,
+      "loss": 0.9004,
+      "step": 3136
+    },
+    {
+      "epoch": 0.48113496932515337,
+      "grad_norm": 0.16960233449935913,
+      "learning_rate": 8.067062111375221e-05,
+      "loss": 1.1182,
+      "step": 3137
+    },
+    {
+      "epoch": 0.4812883435582822,
+      "grad_norm": 0.17896592617034912,
+      "learning_rate": 8.063422307546788e-05,
+      "loss": 1.1013,
+      "step": 3138
+    },
+    {
+      "epoch": 0.48144171779141104,
+      "grad_norm": 0.19473567605018616,
+      "learning_rate": 8.05978237026138e-05,
+      "loss": 1.1893,
+      "step": 3139
+    },
+    {
+      "epoch": 0.4815950920245399,
+      "grad_norm": 0.18751852214336395,
+      "learning_rate": 8.056142300381178e-05,
+      "loss": 1.2495,
+      "step": 3140
+    },
+    {
+      "epoch": 0.4817484662576687,
+      "grad_norm": 0.1763666272163391,
+      "learning_rate": 8.052502098768408e-05,
+      "loss": 1.0928,
+      "step": 3141
+    },
+    {
+      "epoch": 0.48190184049079754,
+      "grad_norm": 0.19210687279701233,
+      "learning_rate": 8.048861766285312e-05,
+      "loss": 1.1886,
+      "step": 3142
+    },
+    {
+      "epoch": 0.4820552147239264,
+      "grad_norm": 0.16984228789806366,
+      "learning_rate": 8.045221303794172e-05,
+      "loss": 1.1314,
+      "step": 3143
+    },
+    {
+      "epoch": 0.4822085889570552,
+      "grad_norm": 0.1710156351327896,
+      "learning_rate": 8.041580712157302e-05,
+      "loss": 1.0276,
+      "step": 3144
+    },
+    {
+      "epoch": 0.48236196319018404,
+      "grad_norm": 0.17372262477874756,
+      "learning_rate": 8.037939992237039e-05,
+      "loss": 1.0906,
+      "step": 3145
+    },
+    {
+      "epoch": 0.4825153374233129,
+      "grad_norm": 0.16878043115139008,
+      "learning_rate": 8.034299144895757e-05,
+      "loss": 0.9893,
+      "step": 3146
+    },
+    {
+      "epoch": 0.4826687116564417,
+      "grad_norm": 0.16948290169239044,
+      "learning_rate": 8.030658170995856e-05,
+      "loss": 1.0466,
+      "step": 3147
+    },
+    {
+      "epoch": 0.48282208588957054,
+      "grad_norm": 0.1822916865348816,
+      "learning_rate": 8.02701707139977e-05,
+      "loss": 0.9936,
+      "step": 3148
+    },
+    {
+      "epoch": 0.4829754601226994,
+      "grad_norm": 0.1771332323551178,
+      "learning_rate": 8.02337584696996e-05,
+      "loss": 1.1677,
+      "step": 3149
+    },
+    {
+      "epoch": 0.4831288343558282,
+      "grad_norm": 0.18053914606571198,
+      "learning_rate": 8.019734498568916e-05,
+      "loss": 1.1576,
+      "step": 3150
+    },
+    {
+      "epoch": 0.48328220858895704,
+      "grad_norm": 0.16492819786071777,
+      "learning_rate": 8.016093027059163e-05,
+      "loss": 1.1178,
+      "step": 3151
+    },
+    {
+      "epoch": 0.4834355828220859,
+      "grad_norm": 0.16707400977611542,
+      "learning_rate": 8.012451433303246e-05,
+      "loss": 1.0999,
+      "step": 3152
+    },
+    {
+      "epoch": 0.4835889570552147,
+      "grad_norm": 0.20346611738204956,
+      "learning_rate": 8.008809718163744e-05,
+      "loss": 1.2379,
+      "step": 3153
+    },
+    {
+      "epoch": 0.48374233128834354,
+      "grad_norm": 0.21714836359024048,
+      "learning_rate": 8.005167882503267e-05,
+      "loss": 1.1069,
+      "step": 3154
+    },
+    {
+      "epoch": 0.4838957055214724,
+      "grad_norm": 0.1746944934129715,
+      "learning_rate": 8.001525927184444e-05,
+      "loss": 1.0944,
+      "step": 3155
+    },
+    {
+      "epoch": 0.4840490797546012,
+      "grad_norm": 0.17370717227458954,
+      "learning_rate": 7.99788385306995e-05,
+      "loss": 1.1364,
+      "step": 3156
+    },
+    {
+      "epoch": 0.48420245398773004,
+      "grad_norm": 0.16743828356266022,
+      "learning_rate": 7.994241661022474e-05,
+      "loss": 0.9586,
+      "step": 3157
+    },
+    {
+      "epoch": 0.4843558282208589,
+      "grad_norm": 0.16328927874565125,
+      "learning_rate": 7.990599351904733e-05,
+      "loss": 0.9557,
+      "step": 3158
+    },
+    {
+      "epoch": 0.4845092024539877,
+      "grad_norm": 0.1832338273525238,
+      "learning_rate": 7.986956926579479e-05,
+      "loss": 1.1184,
+      "step": 3159
+    },
+    {
+      "epoch": 0.48466257668711654,
+      "grad_norm": 0.1732587367296219,
+      "learning_rate": 7.983314385909484e-05,
+      "loss": 1.147,
+      "step": 3160
+    },
+    {
+      "epoch": 0.4848159509202454,
+      "grad_norm": 0.1909385770559311,
+      "learning_rate": 7.979671730757553e-05,
+      "loss": 1.144,
+      "step": 3161
+    },
+    {
+      "epoch": 0.4849693251533742,
+      "grad_norm": 0.17659664154052734,
+      "learning_rate": 7.976028961986514e-05,
+      "loss": 1.1034,
+      "step": 3162
+    },
+    {
+      "epoch": 0.48512269938650304,
+      "grad_norm": 0.17157600820064545,
+      "learning_rate": 7.972386080459229e-05,
+      "loss": 1.0311,
+      "step": 3163
+    },
+    {
+      "epoch": 0.4852760736196319,
+      "grad_norm": 0.16316038370132446,
+      "learning_rate": 7.968743087038578e-05,
+      "loss": 1.1053,
+      "step": 3164
+    },
+    {
+      "epoch": 0.4854294478527607,
+      "grad_norm": 0.1701359748840332,
+      "learning_rate": 7.965099982587472e-05,
+      "loss": 1.1024,
+      "step": 3165
+    },
+    {
+      "epoch": 0.4855828220858896,
+      "grad_norm": 0.17056812345981598,
+      "learning_rate": 7.961456767968847e-05,
+      "loss": 1.071,
+      "step": 3166
+    },
+    {
+      "epoch": 0.48573619631901843,
+      "grad_norm": 0.1784372627735138,
+      "learning_rate": 7.957813444045668e-05,
+      "loss": 0.9077,
+      "step": 3167
+    },
+    {
+      "epoch": 0.48588957055214727,
+      "grad_norm": 0.1823170930147171,
+      "learning_rate": 7.954170011680919e-05,
+      "loss": 1.1131,
+      "step": 3168
+    },
+    {
+      "epoch": 0.4860429447852761,
+      "grad_norm": 0.17063477635383606,
+      "learning_rate": 7.950526471737619e-05,
+      "loss": 1.1521,
+      "step": 3169
+    },
+    {
+      "epoch": 0.48619631901840493,
+      "grad_norm": 0.17004498839378357,
+      "learning_rate": 7.946882825078804e-05,
+      "loss": 1.0658,
+      "step": 3170
+    },
+    {
+      "epoch": 0.48634969325153377,
+      "grad_norm": 0.17229071259498596,
+      "learning_rate": 7.943239072567541e-05,
+      "loss": 1.0372,
+      "step": 3171
+    },
+    {
+      "epoch": 0.4865030674846626,
+      "grad_norm": 0.23226846754550934,
+      "learning_rate": 7.939595215066922e-05,
+      "loss": 1.094,
+      "step": 3172
+    },
+    {
+      "epoch": 0.48665644171779143,
+      "grad_norm": 0.17724066972732544,
+      "learning_rate": 7.935951253440059e-05,
+      "loss": 1.1402,
+      "step": 3173
+    },
+    {
+      "epoch": 0.48680981595092027,
+      "grad_norm": 0.17798961699008942,
+      "learning_rate": 7.93230718855009e-05,
+      "loss": 1.241,
+      "step": 3174
+    },
+    {
+      "epoch": 0.4869631901840491,
+      "grad_norm": 0.1703213006258011,
+      "learning_rate": 7.928663021260183e-05,
+      "loss": 1.1453,
+      "step": 3175
+    },
+    {
+      "epoch": 0.48711656441717793,
+      "grad_norm": 0.1689971685409546,
+      "learning_rate": 7.925018752433524e-05,
+      "loss": 1.0272,
+      "step": 3176
+    },
+    {
+      "epoch": 0.48726993865030677,
+      "grad_norm": 0.17413723468780518,
+      "learning_rate": 7.921374382933325e-05,
+      "loss": 1.152,
+      "step": 3177
+    },
+    {
+      "epoch": 0.4874233128834356,
+      "grad_norm": 0.18641696870326996,
+      "learning_rate": 7.917729913622826e-05,
+      "loss": 1.1363,
+      "step": 3178
+    },
+    {
+      "epoch": 0.48757668711656443,
+      "grad_norm": 0.17912258207798004,
+      "learning_rate": 7.914085345365282e-05,
+      "loss": 1.0814,
+      "step": 3179
+    },
+    {
+      "epoch": 0.48773006134969327,
+      "grad_norm": 0.17489096522331238,
+      "learning_rate": 7.910440679023979e-05,
+      "loss": 1.1411,
+      "step": 3180
+    },
+    {
+      "epoch": 0.4878834355828221,
+      "grad_norm": 0.17071355879306793,
+      "learning_rate": 7.906795915462223e-05,
+      "loss": 1.1833,
+      "step": 3181
+    },
+    {
+      "epoch": 0.48803680981595093,
+      "grad_norm": 0.17924107611179352,
+      "learning_rate": 7.903151055543343e-05,
+      "loss": 1.186,
+      "step": 3182
+    },
+    {
+      "epoch": 0.48819018404907977,
+      "grad_norm": 0.17922861874103546,
+      "learning_rate": 7.899506100130692e-05,
+      "loss": 1.1522,
+      "step": 3183
+    },
+    {
+      "epoch": 0.4883435582822086,
+      "grad_norm": 0.16666735708713531,
+      "learning_rate": 7.895861050087644e-05,
+      "loss": 1.0251,
+      "step": 3184
+    },
+    {
+      "epoch": 0.48849693251533743,
+      "grad_norm": 0.19411693513393402,
+      "learning_rate": 7.892215906277597e-05,
+      "loss": 1.3706,
+      "step": 3185
+    },
+    {
+      "epoch": 0.48865030674846627,
+      "grad_norm": 0.16878367960453033,
+      "learning_rate": 7.888570669563974e-05,
+      "loss": 1.0628,
+      "step": 3186
+    },
+    {
+      "epoch": 0.4888036809815951,
+      "grad_norm": 0.17006945610046387,
+      "learning_rate": 7.884925340810211e-05,
+      "loss": 1.0605,
+      "step": 3187
+    },
+    {
+      "epoch": 0.48895705521472393,
+      "grad_norm": 0.17750848829746246,
+      "learning_rate": 7.881279920879772e-05,
+      "loss": 1.0482,
+      "step": 3188
+    },
+    {
+      "epoch": 0.48911042944785277,
+      "grad_norm": 0.17672303318977356,
+      "learning_rate": 7.877634410636147e-05,
+      "loss": 1.1136,
+      "step": 3189
+    },
+    {
+      "epoch": 0.4892638036809816,
+      "grad_norm": 0.17415787279605865,
+      "learning_rate": 7.873988810942837e-05,
+      "loss": 1.091,
+      "step": 3190
+    },
+    {
+      "epoch": 0.48941717791411044,
+      "grad_norm": 0.17538082599639893,
+      "learning_rate": 7.870343122663374e-05,
+      "loss": 1.0779,
+      "step": 3191
+    },
+    {
+      "epoch": 0.48957055214723927,
+      "grad_norm": 0.18596848845481873,
+      "learning_rate": 7.866697346661302e-05,
+      "loss": 1.0631,
+      "step": 3192
+    },
+    {
+      "epoch": 0.4897239263803681,
+      "grad_norm": 0.15459267795085907,
+      "learning_rate": 7.863051483800194e-05,
+      "loss": 1.0085,
+      "step": 3193
+    },
+    {
+      "epoch": 0.48987730061349694,
+      "grad_norm": 0.16093023121356964,
+      "learning_rate": 7.85940553494364e-05,
+      "loss": 0.9356,
+      "step": 3194
+    },
+    {
+      "epoch": 0.49003067484662577,
+      "grad_norm": 0.18677198886871338,
+      "learning_rate": 7.855759500955248e-05,
+      "loss": 1.2157,
+      "step": 3195
+    },
+    {
+      "epoch": 0.4901840490797546,
+      "grad_norm": 0.16449683904647827,
+      "learning_rate": 7.852113382698648e-05,
+      "loss": 1.0357,
+      "step": 3196
+    },
+    {
+      "epoch": 0.49033742331288344,
+      "grad_norm": 0.1645248681306839,
+      "learning_rate": 7.848467181037491e-05,
+      "loss": 1.0716,
+      "step": 3197
+    },
+    {
+      "epoch": 0.49049079754601227,
+      "grad_norm": 0.16969524323940277,
+      "learning_rate": 7.844820896835451e-05,
+      "loss": 1.0795,
+      "step": 3198
+    },
+    {
+      "epoch": 0.4906441717791411,
+      "grad_norm": 0.17600451409816742,
+      "learning_rate": 7.841174530956215e-05,
+      "loss": 1.1436,
+      "step": 3199
+    },
+    {
+      "epoch": 0.49079754601226994,
+      "grad_norm": 0.1766415536403656,
+      "learning_rate": 7.837528084263492e-05,
+      "loss": 1.2938,
+      "step": 3200
+    },
+    {
+      "epoch": 0.49095092024539877,
+      "grad_norm": 0.17288640141487122,
+      "learning_rate": 7.833881557621008e-05,
+      "loss": 1.165,
+      "step": 3201
+    },
+    {
+      "epoch": 0.4911042944785276,
+      "grad_norm": 0.1794213503599167,
+      "learning_rate": 7.830234951892512e-05,
+      "loss": 1.1371,
+      "step": 3202
+    },
+    {
+      "epoch": 0.49125766871165644,
+      "grad_norm": 0.1798541396856308,
+      "learning_rate": 7.826588267941773e-05,
+      "loss": 1.1478,
+      "step": 3203
+    },
+    {
+      "epoch": 0.49141104294478527,
+      "grad_norm": 0.17735347151756287,
+      "learning_rate": 7.822941506632571e-05,
+      "loss": 1.1274,
+      "step": 3204
+    },
+    {
+      "epoch": 0.4915644171779141,
+      "grad_norm": 0.18796475231647491,
+      "learning_rate": 7.819294668828712e-05,
+      "loss": 1.2433,
+      "step": 3205
+    },
+    {
+      "epoch": 0.49171779141104294,
+      "grad_norm": 0.1599234938621521,
+      "learning_rate": 7.815647755394013e-05,
+      "loss": 1.0207,
+      "step": 3206
+    },
+    {
+      "epoch": 0.49187116564417177,
+      "grad_norm": 0.1761898547410965,
+      "learning_rate": 7.812000767192316e-05,
+      "loss": 1.1414,
+      "step": 3207
+    },
+    {
+      "epoch": 0.4920245398773006,
+      "grad_norm": 0.1805347353219986,
+      "learning_rate": 7.808353705087478e-05,
+      "loss": 1.1467,
+      "step": 3208
+    },
+    {
+      "epoch": 0.49217791411042944,
+      "grad_norm": 0.1647939682006836,
+      "learning_rate": 7.804706569943369e-05,
+      "loss": 1.0659,
+      "step": 3209
+    },
+    {
+      "epoch": 0.49233128834355827,
+      "grad_norm": 0.16765812039375305,
+      "learning_rate": 7.801059362623884e-05,
+      "loss": 1.0166,
+      "step": 3210
+    },
+    {
+      "epoch": 0.4924846625766871,
+      "grad_norm": 0.162796750664711,
+      "learning_rate": 7.79741208399293e-05,
+      "loss": 1.0175,
+      "step": 3211
+    },
+    {
+      "epoch": 0.49263803680981594,
+      "grad_norm": 0.17611801624298096,
+      "learning_rate": 7.793764734914434e-05,
+      "loss": 1.1671,
+      "step": 3212
+    },
+    {
+      "epoch": 0.49279141104294477,
+      "grad_norm": 0.16977733373641968,
+      "learning_rate": 7.790117316252335e-05,
+      "loss": 1.0492,
+      "step": 3213
+    },
+    {
+      "epoch": 0.4929447852760736,
+      "grad_norm": 0.17329084873199463,
+      "learning_rate": 7.786469828870592e-05,
+      "loss": 1.0741,
+      "step": 3214
+    },
+    {
+      "epoch": 0.49309815950920244,
+      "grad_norm": 0.17608624696731567,
+      "learning_rate": 7.782822273633185e-05,
+      "loss": 1.0954,
+      "step": 3215
+    },
+    {
+      "epoch": 0.49325153374233127,
+      "grad_norm": 0.15726163983345032,
+      "learning_rate": 7.779174651404096e-05,
+      "loss": 1.0706,
+      "step": 3216
+    },
+    {
+      "epoch": 0.4934049079754601,
+      "grad_norm": 0.17672860622406006,
+      "learning_rate": 7.775526963047337e-05,
+      "loss": 1.1014,
+      "step": 3217
+    },
+    {
+      "epoch": 0.49355828220858894,
+      "grad_norm": 0.16979670524597168,
+      "learning_rate": 7.77187920942693e-05,
+      "loss": 1.069,
+      "step": 3218
+    },
+    {
+      "epoch": 0.49371165644171777,
+      "grad_norm": 0.18491843342781067,
+      "learning_rate": 7.768231391406912e-05,
+      "loss": 1.2059,
+      "step": 3219
+    },
+    {
+      "epoch": 0.4938650306748466,
+      "grad_norm": 0.16660651564598083,
+      "learning_rate": 7.764583509851336e-05,
+      "loss": 1.0641,
+      "step": 3220
+    },
+    {
+      "epoch": 0.49401840490797544,
+      "grad_norm": 0.18447935581207275,
+      "learning_rate": 7.760935565624272e-05,
+      "loss": 1.2624,
+      "step": 3221
+    },
+    {
+      "epoch": 0.4941717791411043,
+      "grad_norm": 0.17194898426532745,
+      "learning_rate": 7.757287559589801e-05,
+      "loss": 1.0158,
+      "step": 3222
+    },
+    {
+      "epoch": 0.4943251533742331,
+      "grad_norm": 0.1694585680961609,
+      "learning_rate": 7.75363949261202e-05,
+      "loss": 1.1476,
+      "step": 3223
+    },
+    {
+      "epoch": 0.49447852760736194,
+      "grad_norm": 0.166235089302063,
+      "learning_rate": 7.749991365555042e-05,
+      "loss": 0.9871,
+      "step": 3224
+    },
+    {
+      "epoch": 0.4946319018404908,
+      "grad_norm": 0.18250764906406403,
+      "learning_rate": 7.746343179282995e-05,
+      "loss": 1.1897,
+      "step": 3225
+    },
+    {
+      "epoch": 0.4947852760736196,
+      "grad_norm": 0.17327867448329926,
+      "learning_rate": 7.742694934660017e-05,
+      "loss": 1.107,
+      "step": 3226
+    },
+    {
+      "epoch": 0.49493865030674844,
+      "grad_norm": 0.16831481456756592,
+      "learning_rate": 7.73904663255026e-05,
+      "loss": 1.0182,
+      "step": 3227
+    },
+    {
+      "epoch": 0.4950920245398773,
+      "grad_norm": 0.16892239451408386,
+      "learning_rate": 7.735398273817898e-05,
+      "loss": 1.1793,
+      "step": 3228
+    },
+    {
+      "epoch": 0.49524539877300616,
+      "grad_norm": 0.1727597862482071,
+      "learning_rate": 7.731749859327107e-05,
+      "loss": 1.0974,
+      "step": 3229
+    },
+    {
+      "epoch": 0.495398773006135,
+      "grad_norm": 0.1689879149198532,
+      "learning_rate": 7.728101389942083e-05,
+      "loss": 1.1625,
+      "step": 3230
+    },
+    {
+      "epoch": 0.49555214723926383,
+      "grad_norm": 0.17369799315929413,
+      "learning_rate": 7.724452866527031e-05,
+      "loss": 1.1128,
+      "step": 3231
+    },
+    {
+      "epoch": 0.49570552147239266,
+      "grad_norm": 0.15727372467517853,
+      "learning_rate": 7.720804289946177e-05,
+      "loss": 1.0495,
+      "step": 3232
+    },
+    {
+      "epoch": 0.4958588957055215,
+      "grad_norm": 0.16576316952705383,
+      "learning_rate": 7.717155661063747e-05,
+      "loss": 1.0861,
+      "step": 3233
+    },
+    {
+      "epoch": 0.49601226993865033,
+      "grad_norm": 0.1608206331729889,
+      "learning_rate": 7.71350698074399e-05,
+      "loss": 1.1234,
+      "step": 3234
+    },
+    {
+      "epoch": 0.49616564417177916,
+      "grad_norm": 0.1656356006860733,
+      "learning_rate": 7.709858249851162e-05,
+      "loss": 1.1736,
+      "step": 3235
+    },
+    {
+      "epoch": 0.496319018404908,
+      "grad_norm": 0.17607444524765015,
+      "learning_rate": 7.706209469249534e-05,
+      "loss": 1.1569,
+      "step": 3236
+    },
+    {
+      "epoch": 0.49647239263803683,
+      "grad_norm": 0.16946889460086823,
+      "learning_rate": 7.702560639803388e-05,
+      "loss": 1.1759,
+      "step": 3237
+    },
+    {
+      "epoch": 0.49662576687116566,
+      "grad_norm": 0.16052377223968506,
+      "learning_rate": 7.69891176237701e-05,
+      "loss": 1.0359,
+      "step": 3238
+    },
+    {
+      "epoch": 0.4967791411042945,
+      "grad_norm": 0.16981737315654755,
+      "learning_rate": 7.69526283783471e-05,
+      "loss": 1.1444,
+      "step": 3239
+    },
+    {
+      "epoch": 0.49693251533742333,
+      "grad_norm": 0.16778415441513062,
+      "learning_rate": 7.6916138670408e-05,
+      "loss": 1.1345,
+      "step": 3240
+    },
+    {
+      "epoch": 0.49708588957055216,
+      "grad_norm": 0.17207223176956177,
+      "learning_rate": 7.687964850859609e-05,
+      "loss": 1.0892,
+      "step": 3241
+    },
+    {
+      "epoch": 0.497239263803681,
+      "grad_norm": 0.15521806478500366,
+      "learning_rate": 7.684315790155473e-05,
+      "loss": 1.0055,
+      "step": 3242
+    },
+    {
+      "epoch": 0.49739263803680983,
+      "grad_norm": 0.16693201661109924,
+      "learning_rate": 7.68066668579274e-05,
+      "loss": 1.0076,
+      "step": 3243
+    },
+    {
+      "epoch": 0.49754601226993866,
+      "grad_norm": 0.15961594879627228,
+      "learning_rate": 7.677017538635766e-05,
+      "loss": 1.0537,
+      "step": 3244
+    },
+    {
+      "epoch": 0.4976993865030675,
+      "grad_norm": 0.1675877869129181,
+      "learning_rate": 7.673368349548918e-05,
+      "loss": 1.0674,
+      "step": 3245
+    },
+    {
+      "epoch": 0.49785276073619633,
+      "grad_norm": 0.16247627139091492,
+      "learning_rate": 7.669719119396576e-05,
+      "loss": 1.0068,
+      "step": 3246
+    },
+    {
+      "epoch": 0.49800613496932516,
+      "grad_norm": 0.16573938727378845,
+      "learning_rate": 7.666069849043127e-05,
+      "loss": 0.9903,
+      "step": 3247
+    },
+    {
+      "epoch": 0.498159509202454,
+      "grad_norm": 0.16628819704055786,
+      "learning_rate": 7.662420539352968e-05,
+      "loss": 1.1759,
+      "step": 3248
+    },
+    {
+      "epoch": 0.49831288343558283,
+      "grad_norm": 0.17175593972206116,
+      "learning_rate": 7.658771191190504e-05,
+      "loss": 1.1611,
+      "step": 3249
+    },
+    {
+      "epoch": 0.49846625766871167,
+      "grad_norm": 0.18383678793907166,
+      "learning_rate": 7.655121805420156e-05,
+      "loss": 1.153,
+      "step": 3250
+    },
+    {
+      "epoch": 0.4986196319018405,
+      "grad_norm": 0.15648098289966583,
+      "learning_rate": 7.651472382906341e-05,
+      "loss": 1.1162,
+      "step": 3251
+    },
+    {
+      "epoch": 0.49877300613496933,
+      "grad_norm": 0.1833643913269043,
+      "learning_rate": 7.647822924513496e-05,
+      "loss": 1.2226,
+      "step": 3252
+    },
+    {
+      "epoch": 0.49892638036809817,
+      "grad_norm": 0.16378627717494965,
+      "learning_rate": 7.644173431106062e-05,
+      "loss": 1.0631,
+      "step": 3253
+    },
+    {
+      "epoch": 0.499079754601227,
+      "grad_norm": 0.16075760126113892,
+      "learning_rate": 7.640523903548487e-05,
+      "loss": 1.0301,
+      "step": 3254
+    },
+    {
+      "epoch": 0.49923312883435583,
+      "grad_norm": 0.16514970362186432,
+      "learning_rate": 7.636874342705233e-05,
+      "loss": 1.0322,
+      "step": 3255
+    },
+    {
+      "epoch": 0.49938650306748467,
+      "grad_norm": 0.1729796677827835,
+      "learning_rate": 7.633224749440763e-05,
+      "loss": 1.1298,
+      "step": 3256
+    },
+    {
+      "epoch": 0.4995398773006135,
+      "grad_norm": 0.18403829634189606,
+      "learning_rate": 7.629575124619551e-05,
+      "loss": 1.2231,
+      "step": 3257
+    },
+    {
+      "epoch": 0.49969325153374233,
+      "grad_norm": 0.17214560508728027,
+      "learning_rate": 7.62592546910608e-05,
+      "loss": 1.1318,
+      "step": 3258
+    },
+    {
+      "epoch": 0.49984662576687117,
+      "grad_norm": 0.16402646899223328,
+      "learning_rate": 7.622275783764835e-05,
+      "loss": 1.1676,
+      "step": 3259
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1661108136177063,
+      "learning_rate": 7.618626069460314e-05,
+      "loss": 1.0362,
+      "step": 3260
+    },
+    {
+      "epoch": 0.5001533742331289,
+      "grad_norm": 0.17139609158039093,
+      "learning_rate": 7.614976327057014e-05,
+      "loss": 1.022,
+      "step": 3261
+    },
+    {
+      "epoch": 0.5003067484662577,
+      "grad_norm": 0.18168310821056366,
+      "learning_rate": 7.611326557419452e-05,
+      "loss": 1.2017,
+      "step": 3262
+    },
+    {
+      "epoch": 0.5004601226993866,
+      "grad_norm": 0.16155001521110535,
+      "learning_rate": 7.607676761412138e-05,
+      "loss": 1.0398,
+      "step": 3263
+    },
+    {
+      "epoch": 0.5006134969325153,
+      "grad_norm": 0.1535506546497345,
+      "learning_rate": 7.604026939899597e-05,
+      "loss": 0.9427,
+      "step": 3264
+    },
+    {
+      "epoch": 0.5007668711656442,
+      "grad_norm": 0.17772312462329865,
+      "learning_rate": 7.600377093746354e-05,
+      "loss": 1.0729,
+      "step": 3265
+    },
+    {
+      "epoch": 0.500920245398773,
+      "grad_norm": 0.1521425098180771,
+      "learning_rate": 7.596727223816943e-05,
+      "loss": 0.9689,
+      "step": 3266
+    },
+    {
+      "epoch": 0.5010736196319019,
+      "grad_norm": 0.16941644251346588,
+      "learning_rate": 7.5930773309759e-05,
+      "loss": 1.1076,
+      "step": 3267
+    },
+    {
+      "epoch": 0.5012269938650307,
+      "grad_norm": 0.15884579718112946,
+      "learning_rate": 7.589427416087776e-05,
+      "loss": 1.014,
+      "step": 3268
+    },
+    {
+      "epoch": 0.5013803680981596,
+      "grad_norm": 0.1640690118074417,
+      "learning_rate": 7.585777480017115e-05,
+      "loss": 0.981,
+      "step": 3269
+    },
+    {
+      "epoch": 0.5015337423312883,
+      "grad_norm": 0.17912231385707855,
+      "learning_rate": 7.582127523628475e-05,
+      "loss": 1.1525,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5016871165644172,
+      "grad_norm": 0.1692584604024887,
+      "learning_rate": 7.578477547786415e-05,
+      "loss": 1.0981,
+      "step": 3271
+    },
+    {
+      "epoch": 0.501840490797546,
+      "grad_norm": 0.16946624219417572,
+      "learning_rate": 7.5748275533555e-05,
+      "loss": 0.9906,
+      "step": 3272
+    },
+    {
+      "epoch": 0.5019938650306749,
+      "grad_norm": 0.1930263787508011,
+      "learning_rate": 7.571177541200297e-05,
+      "loss": 1.2428,
+      "step": 3273
+    },
+    {
+      "epoch": 0.5021472392638037,
+      "grad_norm": 0.5781621932983398,
+      "learning_rate": 7.567527512185376e-05,
+      "loss": 1.0704,
+      "step": 3274
+    },
+    {
+      "epoch": 0.5023006134969326,
+      "grad_norm": 0.15949833393096924,
+      "learning_rate": 7.56387746717532e-05,
+      "loss": 1.0455,
+      "step": 3275
+    },
+    {
+      "epoch": 0.5024539877300613,
+      "grad_norm": 0.16776955127716064,
+      "learning_rate": 7.560227407034706e-05,
+      "loss": 1.1101,
+      "step": 3276
+    },
+    {
+      "epoch": 0.5026073619631902,
+      "grad_norm": 0.16978415846824646,
+      "learning_rate": 7.556577332628117e-05,
+      "loss": 1.202,
+      "step": 3277
+    },
+    {
+      "epoch": 0.502760736196319,
+      "grad_norm": 0.1664252132177353,
+      "learning_rate": 7.552927244820144e-05,
+      "loss": 1.084,
+      "step": 3278
+    },
+    {
+      "epoch": 0.5029141104294479,
+      "grad_norm": 0.16852708160877228,
+      "learning_rate": 7.549277144475377e-05,
+      "loss": 1.0929,
+      "step": 3279
+    },
+    {
+      "epoch": 0.5030674846625767,
+      "grad_norm": 0.16112615168094635,
+      "learning_rate": 7.545627032458406e-05,
+      "loss": 0.9176,
+      "step": 3280
+    },
+    {
+      "epoch": 0.5032208588957056,
+      "grad_norm": 0.17106615006923676,
+      "learning_rate": 7.541976909633831e-05,
+      "loss": 1.0451,
+      "step": 3281
+    },
+    {
+      "epoch": 0.5033742331288343,
+      "grad_norm": 0.16641750931739807,
+      "learning_rate": 7.538326776866246e-05,
+      "loss": 1.1354,
+      "step": 3282
+    },
+    {
+      "epoch": 0.5035276073619632,
+      "grad_norm": 0.17311017215251923,
+      "learning_rate": 7.534676635020261e-05,
+      "loss": 1.1285,
+      "step": 3283
+    },
+    {
+      "epoch": 0.503680981595092,
+      "grad_norm": 0.1794460117816925,
+      "learning_rate": 7.531026484960473e-05,
+      "loss": 1.1689,
+      "step": 3284
+    },
+    {
+      "epoch": 0.5038343558282209,
+      "grad_norm": 0.16847433149814606,
+      "learning_rate": 7.52737632755149e-05,
+      "loss": 1.0892,
+      "step": 3285
+    },
+    {
+      "epoch": 0.5039877300613497,
+      "grad_norm": 0.16328932344913483,
+      "learning_rate": 7.523726163657918e-05,
+      "loss": 1.01,
+      "step": 3286
+    },
+    {
+      "epoch": 0.5041411042944786,
+      "grad_norm": 0.17427417635917664,
+      "learning_rate": 7.520075994144366e-05,
+      "loss": 1.0674,
+      "step": 3287
+    },
+    {
+      "epoch": 0.5042944785276073,
+      "grad_norm": 0.1827038675546646,
+      "learning_rate": 7.516425819875445e-05,
+      "loss": 1.0847,
+      "step": 3288
+    },
+    {
+      "epoch": 0.5044478527607362,
+      "grad_norm": 0.16056060791015625,
+      "learning_rate": 7.512775641715763e-05,
+      "loss": 1.021,
+      "step": 3289
+    },
+    {
+      "epoch": 0.504601226993865,
+      "grad_norm": 0.1623144894838333,
+      "learning_rate": 7.509125460529937e-05,
+      "loss": 0.9382,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5047546012269939,
+      "grad_norm": 0.1742529571056366,
+      "learning_rate": 7.505475277182575e-05,
+      "loss": 1.1389,
+      "step": 3291
+    },
+    {
+      "epoch": 0.5049079754601227,
+      "grad_norm": 0.171508327126503,
+      "learning_rate": 7.501825092538294e-05,
+      "loss": 1.1376,
+      "step": 3292
+    },
+    {
+      "epoch": 0.5050613496932516,
+      "grad_norm": 0.15497560799121857,
+      "learning_rate": 7.498174907461707e-05,
+      "loss": 1.0091,
+      "step": 3293
+    },
+    {
+      "epoch": 0.5052147239263803,
+      "grad_norm": 0.1726723462343216,
+      "learning_rate": 7.494524722817425e-05,
+      "loss": 1.1508,
+      "step": 3294
+    },
+    {
+      "epoch": 0.5053680981595092,
+      "grad_norm": 0.18096669018268585,
+      "learning_rate": 7.490874539470064e-05,
+      "loss": 1.1903,
+      "step": 3295
+    },
+    {
+      "epoch": 0.505521472392638,
+      "grad_norm": 0.16929739713668823,
+      "learning_rate": 7.487224358284236e-05,
+      "loss": 1.114,
+      "step": 3296
+    },
+    {
+      "epoch": 0.5056748466257669,
+      "grad_norm": 0.16578556597232819,
+      "learning_rate": 7.483574180124555e-05,
+      "loss": 1.1363,
+      "step": 3297
+    },
+    {
+      "epoch": 0.5058282208588957,
+      "grad_norm": 0.17365536093711853,
+      "learning_rate": 7.479924005855634e-05,
+      "loss": 1.154,
+      "step": 3298
+    },
+    {
+      "epoch": 0.5059815950920246,
+      "grad_norm": 0.17543357610702515,
+      "learning_rate": 7.47627383634208e-05,
+      "loss": 1.1927,
+      "step": 3299
+    },
+    {
+      "epoch": 0.5061349693251533,
+      "grad_norm": 0.18644626438617706,
+      "learning_rate": 7.472623672448508e-05,
+      "loss": 1.2407,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5062883435582822,
+      "grad_norm": 0.1847156137228012,
+      "learning_rate": 7.468973515039527e-05,
+      "loss": 1.2195,
+      "step": 3301
+    },
+    {
+      "epoch": 0.506441717791411,
+      "grad_norm": 0.1822473555803299,
+      "learning_rate": 7.465323364979739e-05,
+      "loss": 1.2005,
+      "step": 3302
+    },
+    {
+      "epoch": 0.5065950920245399,
+      "grad_norm": 0.16739699244499207,
+      "learning_rate": 7.461673223133753e-05,
+      "loss": 1.0774,
+      "step": 3303
+    },
+    {
+      "epoch": 0.5067484662576687,
+      "grad_norm": 0.1651996225118637,
+      "learning_rate": 7.45802309036617e-05,
+      "loss": 1.0564,
+      "step": 3304
+    },
+    {
+      "epoch": 0.5069018404907976,
+      "grad_norm": 0.16678839921951294,
+      "learning_rate": 7.454372967541595e-05,
+      "loss": 1.0953,
+      "step": 3305
+    },
+    {
+      "epoch": 0.5070552147239263,
+      "grad_norm": 0.1851450800895691,
+      "learning_rate": 7.450722855524625e-05,
+      "loss": 1.2649,
+      "step": 3306
+    },
+    {
+      "epoch": 0.5072085889570552,
+      "grad_norm": 0.16653668880462646,
+      "learning_rate": 7.447072755179856e-05,
+      "loss": 1.0246,
+      "step": 3307
+    },
+    {
+      "epoch": 0.507361963190184,
+      "grad_norm": 0.1728491634130478,
+      "learning_rate": 7.443422667371883e-05,
+      "loss": 1.1285,
+      "step": 3308
+    },
+    {
+      "epoch": 0.5075153374233129,
+      "grad_norm": 0.17294463515281677,
+      "learning_rate": 7.439772592965295e-05,
+      "loss": 1.1186,
+      "step": 3309
+    },
+    {
+      "epoch": 0.5076687116564417,
+      "grad_norm": 0.17820493876934052,
+      "learning_rate": 7.436122532824681e-05,
+      "loss": 1.2123,
+      "step": 3310
+    },
+    {
+      "epoch": 0.5078220858895706,
+      "grad_norm": 0.16092544794082642,
+      "learning_rate": 7.432472487814624e-05,
+      "loss": 1.142,
+      "step": 3311
+    },
+    {
+      "epoch": 0.5079754601226993,
+      "grad_norm": 0.16658955812454224,
+      "learning_rate": 7.428822458799705e-05,
+      "loss": 1.187,
+      "step": 3312
+    },
+    {
+      "epoch": 0.5081288343558282,
+      "grad_norm": 0.1674429476261139,
+      "learning_rate": 7.425172446644501e-05,
+      "loss": 1.0706,
+      "step": 3313
+    },
+    {
+      "epoch": 0.508282208588957,
+      "grad_norm": 0.16181570291519165,
+      "learning_rate": 7.421522452213583e-05,
+      "loss": 1.0777,
+      "step": 3314
+    },
+    {
+      "epoch": 0.5084355828220859,
+      "grad_norm": 0.1672963798046112,
+      "learning_rate": 7.417872476371522e-05,
+      "loss": 1.0598,
+      "step": 3315
+    },
+    {
+      "epoch": 0.5085889570552147,
+      "grad_norm": 0.1826997697353363,
+      "learning_rate": 7.414222519982883e-05,
+      "loss": 1.1544,
+      "step": 3316
+    },
+    {
+      "epoch": 0.5087423312883436,
+      "grad_norm": 0.15632064640522003,
+      "learning_rate": 7.410572583912223e-05,
+      "loss": 1.0242,
+      "step": 3317
+    },
+    {
+      "epoch": 0.5088957055214723,
+      "grad_norm": 0.15974071621894836,
+      "learning_rate": 7.406922669024097e-05,
+      "loss": 0.9935,
+      "step": 3318
+    },
+    {
+      "epoch": 0.5090490797546012,
+      "grad_norm": 0.1714462786912918,
+      "learning_rate": 7.403272776183056e-05,
+      "loss": 1.1252,
+      "step": 3319
+    },
+    {
+      "epoch": 0.50920245398773,
+      "grad_norm": 0.15460112690925598,
+      "learning_rate": 7.399622906253646e-05,
+      "loss": 0.9754,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5093558282208589,
+      "grad_norm": 0.18914638459682465,
+      "learning_rate": 7.395973060100403e-05,
+      "loss": 1.1346,
+      "step": 3321
+    },
+    {
+      "epoch": 0.5095092024539877,
+      "grad_norm": 0.171733096241951,
+      "learning_rate": 7.39232323858786e-05,
+      "loss": 1.1685,
+      "step": 3322
+    },
+    {
+      "epoch": 0.5096625766871166,
+      "grad_norm": 0.1573234647512436,
+      "learning_rate": 7.388673442580547e-05,
+      "loss": 1.0581,
+      "step": 3323
+    },
+    {
+      "epoch": 0.5098159509202455,
+      "grad_norm": 0.14696155488491058,
+      "learning_rate": 7.385023672942983e-05,
+      "loss": 0.9276,
+      "step": 3324
+    },
+    {
+      "epoch": 0.5099693251533742,
+      "grad_norm": 0.16925278306007385,
+      "learning_rate": 7.381373930539687e-05,
+      "loss": 1.0954,
+      "step": 3325
+    },
+    {
+      "epoch": 0.5101226993865031,
+      "grad_norm": 0.1775648593902588,
+      "learning_rate": 7.377724216235165e-05,
+      "loss": 1.1765,
+      "step": 3326
+    },
+    {
+      "epoch": 0.5102760736196319,
+      "grad_norm": 0.16561901569366455,
+      "learning_rate": 7.37407453089392e-05,
+      "loss": 1.19,
+      "step": 3327
+    },
+    {
+      "epoch": 0.5104294478527608,
+      "grad_norm": 0.16938982903957367,
+      "learning_rate": 7.370424875380449e-05,
+      "loss": 1.0261,
+      "step": 3328
+    },
+    {
+      "epoch": 0.5105828220858896,
+      "grad_norm": 0.18938085436820984,
+      "learning_rate": 7.366775250559237e-05,
+      "loss": 1.1437,
+      "step": 3329
+    },
+    {
+      "epoch": 0.5107361963190185,
+      "grad_norm": 0.14740580320358276,
+      "learning_rate": 7.363125657294766e-05,
+      "loss": 0.9507,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5108895705521472,
+      "grad_norm": 0.17781108617782593,
+      "learning_rate": 7.359476096451512e-05,
+      "loss": 1.1892,
+      "step": 3331
+    },
+    {
+      "epoch": 0.5110429447852761,
+      "grad_norm": 0.1698606312274933,
+      "learning_rate": 7.355826568893939e-05,
+      "loss": 1.0676,
+      "step": 3332
+    },
+    {
+      "epoch": 0.5111963190184049,
+      "grad_norm": 0.16987648606300354,
+      "learning_rate": 7.352177075486503e-05,
+      "loss": 1.079,
+      "step": 3333
+    },
+    {
+      "epoch": 0.5113496932515338,
+      "grad_norm": 0.18848556280136108,
+      "learning_rate": 7.348527617093659e-05,
+      "loss": 1.0484,
+      "step": 3334
+    },
+    {
+      "epoch": 0.5115030674846626,
+      "grad_norm": 0.16819991171360016,
+      "learning_rate": 7.344878194579844e-05,
+      "loss": 1.0728,
+      "step": 3335
+    },
+    {
+      "epoch": 0.5116564417177915,
+      "grad_norm": 0.19539354741573334,
+      "learning_rate": 7.341228808809495e-05,
+      "loss": 1.1907,
+      "step": 3336
+    },
+    {
+      "epoch": 0.5118098159509202,
+      "grad_norm": 0.17517228424549103,
+      "learning_rate": 7.337579460647033e-05,
+      "loss": 1.1445,
+      "step": 3337
+    },
+    {
+      "epoch": 0.5119631901840491,
+      "grad_norm": 0.15810920298099518,
+      "learning_rate": 7.333930150956873e-05,
+      "loss": 1.0636,
+      "step": 3338
+    },
+    {
+      "epoch": 0.5121165644171779,
+      "grad_norm": 0.15890701115131378,
+      "learning_rate": 7.330280880603424e-05,
+      "loss": 1.1514,
+      "step": 3339
+    },
+    {
+      "epoch": 0.5122699386503068,
+      "grad_norm": 0.16299547255039215,
+      "learning_rate": 7.326631650451082e-05,
+      "loss": 0.9918,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5124233128834356,
+      "grad_norm": 0.18357041478157043,
+      "learning_rate": 7.322982461364234e-05,
+      "loss": 1.19,
+      "step": 3341
+    },
+    {
+      "epoch": 0.5125766871165645,
+      "grad_norm": 0.16792072355747223,
+      "learning_rate": 7.31933331420726e-05,
+      "loss": 1.105,
+      "step": 3342
+    },
+    {
+      "epoch": 0.5127300613496932,
+      "grad_norm": 0.18591275811195374,
+      "learning_rate": 7.315684209844527e-05,
+      "loss": 1.049,
+      "step": 3343
+    },
+    {
+      "epoch": 0.5128834355828221,
+      "grad_norm": 0.1530311405658722,
+      "learning_rate": 7.312035149140391e-05,
+      "loss": 1.0848,
+      "step": 3344
+    },
+    {
+      "epoch": 0.5130368098159509,
+      "grad_norm": 0.15444731712341309,
+      "learning_rate": 7.3083861329592e-05,
+      "loss": 1.0115,
+      "step": 3345
+    },
+    {
+      "epoch": 0.5131901840490798,
+      "grad_norm": 0.1659417450428009,
+      "learning_rate": 7.304737162165292e-05,
+      "loss": 1.1261,
+      "step": 3346
+    },
+    {
+      "epoch": 0.5133435582822086,
+      "grad_norm": 0.21587207913398743,
+      "learning_rate": 7.301088237622991e-05,
+      "loss": 1.106,
+      "step": 3347
+    },
+    {
+      "epoch": 0.5134969325153375,
+      "grad_norm": 0.15876300632953644,
+      "learning_rate": 7.297439360196615e-05,
+      "loss": 1.0222,
+      "step": 3348
+    },
+    {
+      "epoch": 0.5136503067484662,
+      "grad_norm": 0.1618252545595169,
+      "learning_rate": 7.293790530750466e-05,
+      "loss": 1.1003,
+      "step": 3349
+    },
+    {
+      "epoch": 0.5138036809815951,
+      "grad_norm": 0.16400085389614105,
+      "learning_rate": 7.290141750148836e-05,
+      "loss": 1.006,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5139570552147239,
+      "grad_norm": 0.16508464515209198,
+      "learning_rate": 7.28649301925601e-05,
+      "loss": 1.1666,
+      "step": 3351
+    },
+    {
+      "epoch": 0.5141104294478528,
+      "grad_norm": 0.15854419767856598,
+      "learning_rate": 7.282844338936254e-05,
+      "loss": 1.0311,
+      "step": 3352
+    },
+    {
+      "epoch": 0.5142638036809816,
+      "grad_norm": 0.18131181597709656,
+      "learning_rate": 7.279195710053826e-05,
+      "loss": 1.1388,
+      "step": 3353
+    },
+    {
+      "epoch": 0.5144171779141105,
+      "grad_norm": 0.16736222803592682,
+      "learning_rate": 7.275547133472969e-05,
+      "loss": 1.1701,
+      "step": 3354
+    },
+    {
+      "epoch": 0.5145705521472392,
+      "grad_norm": 0.20031829178333282,
+      "learning_rate": 7.271898610057918e-05,
+      "loss": 1.2481,
+      "step": 3355
+    },
+    {
+      "epoch": 0.5147239263803681,
+      "grad_norm": 0.15900033712387085,
+      "learning_rate": 7.268250140672894e-05,
+      "loss": 0.9427,
+      "step": 3356
+    },
+    {
+      "epoch": 0.5148773006134969,
+      "grad_norm": 0.15247607231140137,
+      "learning_rate": 7.264601726182101e-05,
+      "loss": 0.9729,
+      "step": 3357
+    },
+    {
+      "epoch": 0.5150306748466258,
+      "grad_norm": 0.17016887664794922,
+      "learning_rate": 7.260953367449737e-05,
+      "loss": 1.1398,
+      "step": 3358
+    },
+    {
+      "epoch": 0.5151840490797546,
+      "grad_norm": 0.1677098423242569,
+      "learning_rate": 7.257305065339982e-05,
+      "loss": 1.1482,
+      "step": 3359
+    },
+    {
+      "epoch": 0.5153374233128835,
+      "grad_norm": 0.1724252700805664,
+      "learning_rate": 7.253656820717003e-05,
+      "loss": 1.029,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5154907975460122,
+      "grad_norm": 0.17439420521259308,
+      "learning_rate": 7.250008634444955e-05,
+      "loss": 1.2223,
+      "step": 3361
+    },
+    {
+      "epoch": 0.5156441717791411,
+      "grad_norm": 0.1711495816707611,
+      "learning_rate": 7.246360507387978e-05,
+      "loss": 1.1061,
+      "step": 3362
+    },
+    {
+      "epoch": 0.5157975460122699,
+      "grad_norm": 0.1570705622434616,
+      "learning_rate": 7.242712440410199e-05,
+      "loss": 1.0757,
+      "step": 3363
+    },
+    {
+      "epoch": 0.5159509202453988,
+      "grad_norm": 0.18052349984645844,
+      "learning_rate": 7.239064434375727e-05,
+      "loss": 1.2864,
+      "step": 3364
+    },
+    {
+      "epoch": 0.5161042944785276,
+      "grad_norm": 0.1629783660173416,
+      "learning_rate": 7.235416490148662e-05,
+      "loss": 1.0572,
+      "step": 3365
+    },
+    {
+      "epoch": 0.5162576687116565,
+      "grad_norm": 0.15918542444705963,
+      "learning_rate": 7.231768608593087e-05,
+      "loss": 0.9947,
+      "step": 3366
+    },
+    {
+      "epoch": 0.5164110429447852,
+      "grad_norm": 0.15513356029987335,
+      "learning_rate": 7.228120790573068e-05,
+      "loss": 1.0189,
+      "step": 3367
+    },
+    {
+      "epoch": 0.5165644171779141,
+      "grad_norm": 0.5872828364372253,
+      "learning_rate": 7.224473036952661e-05,
+      "loss": 1.0059,
+      "step": 3368
+    },
+    {
+      "epoch": 0.5167177914110429,
+      "grad_norm": 0.15995055437088013,
+      "learning_rate": 7.220825348595903e-05,
+      "loss": 1.0144,
+      "step": 3369
+    },
+    {
+      "epoch": 0.5168711656441718,
+      "grad_norm": 0.18242594599723816,
+      "learning_rate": 7.217177726366815e-05,
+      "loss": 1.1858,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5170245398773006,
+      "grad_norm": 0.1740311235189438,
+      "learning_rate": 7.213530171129406e-05,
+      "loss": 1.1259,
+      "step": 3371
+    },
+    {
+      "epoch": 0.5171779141104295,
+      "grad_norm": 0.15719301998615265,
+      "learning_rate": 7.209882683747665e-05,
+      "loss": 1.0525,
+      "step": 3372
+    },
+    {
+      "epoch": 0.5173312883435582,
+      "grad_norm": 0.338139146566391,
+      "learning_rate": 7.206235265085566e-05,
+      "loss": 1.1068,
+      "step": 3373
+    },
+    {
+      "epoch": 0.5174846625766871,
+      "grad_norm": 0.20080961287021637,
+      "learning_rate": 7.202587916007069e-05,
+      "loss": 1.0599,
+      "step": 3374
+    },
+    {
+      "epoch": 0.5176380368098159,
+      "grad_norm": 0.18345503509044647,
+      "learning_rate": 7.198940637376114e-05,
+      "loss": 1.1531,
+      "step": 3375
+    },
+    {
+      "epoch": 0.5177914110429448,
+      "grad_norm": 0.1622041016817093,
+      "learning_rate": 7.19529343005663e-05,
+      "loss": 1.1054,
+      "step": 3376
+    },
+    {
+      "epoch": 0.5179447852760736,
+      "grad_norm": 0.15915414690971375,
+      "learning_rate": 7.191646294912522e-05,
+      "loss": 1.0629,
+      "step": 3377
+    },
+    {
+      "epoch": 0.5180981595092025,
+      "grad_norm": 0.17462489008903503,
+      "learning_rate": 7.187999232807684e-05,
+      "loss": 1.1846,
+      "step": 3378
+    },
+    {
+      "epoch": 0.5182515337423312,
+      "grad_norm": 0.199808731675148,
+      "learning_rate": 7.184352244605987e-05,
+      "loss": 1.2231,
+      "step": 3379
+    },
+    {
+      "epoch": 0.5184049079754601,
+      "grad_norm": 0.1728401780128479,
+      "learning_rate": 7.180705331171289e-05,
+      "loss": 1.108,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5185582822085889,
+      "grad_norm": 0.1687219738960266,
+      "learning_rate": 7.177058493367428e-05,
+      "loss": 1.1215,
+      "step": 3381
+    },
+    {
+      "epoch": 0.5187116564417178,
+      "grad_norm": 0.16782769560813904,
+      "learning_rate": 7.173411732058226e-05,
+      "loss": 1.0587,
+      "step": 3382
+    },
+    {
+      "epoch": 0.5188650306748466,
+      "grad_norm": 0.1627640426158905,
+      "learning_rate": 7.169765048107486e-05,
+      "loss": 0.9914,
+      "step": 3383
+    },
+    {
+      "epoch": 0.5190184049079755,
+      "grad_norm": 0.229523167014122,
+      "learning_rate": 7.166118442378991e-05,
+      "loss": 0.9573,
+      "step": 3384
+    },
+    {
+      "epoch": 0.5191717791411042,
+      "grad_norm": 0.17556867003440857,
+      "learning_rate": 7.162471915736508e-05,
+      "loss": 1.0932,
+      "step": 3385
+    },
+    {
+      "epoch": 0.5193251533742331,
+      "grad_norm": 0.15965168178081512,
+      "learning_rate": 7.158825469043785e-05,
+      "loss": 1.0119,
+      "step": 3386
+    },
+    {
+      "epoch": 0.519478527607362,
+      "grad_norm": 0.16492964327335358,
+      "learning_rate": 7.155179103164548e-05,
+      "loss": 0.962,
+      "step": 3387
+    },
+    {
+      "epoch": 0.5196319018404908,
+      "grad_norm": 0.16900041699409485,
+      "learning_rate": 7.151532818962508e-05,
+      "loss": 1.0769,
+      "step": 3388
+    },
+    {
+      "epoch": 0.5197852760736197,
+      "grad_norm": 0.1769964098930359,
+      "learning_rate": 7.147886617301352e-05,
+      "loss": 1.1318,
+      "step": 3389
+    },
+    {
+      "epoch": 0.5199386503067485,
+      "grad_norm": 0.1757764369249344,
+      "learning_rate": 7.144240499044753e-05,
+      "loss": 1.1541,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5200920245398774,
+      "grad_norm": 0.2492876499891281,
+      "learning_rate": 7.140594465056361e-05,
+      "loss": 1.0117,
+      "step": 3391
+    },
+    {
+      "epoch": 0.5202453987730061,
+      "grad_norm": 0.17647626996040344,
+      "learning_rate": 7.136948516199806e-05,
+      "loss": 1.0768,
+      "step": 3392
+    },
+    {
+      "epoch": 0.520398773006135,
+      "grad_norm": 0.17370787262916565,
+      "learning_rate": 7.133302653338697e-05,
+      "loss": 1.1389,
+      "step": 3393
+    },
+    {
+      "epoch": 0.5205521472392638,
+      "grad_norm": 0.2141655683517456,
+      "learning_rate": 7.129656877336629e-05,
+      "loss": 1.2193,
+      "step": 3394
+    },
+    {
+      "epoch": 0.5207055214723927,
+      "grad_norm": 0.1689879596233368,
+      "learning_rate": 7.126011189057164e-05,
+      "loss": 1.0308,
+      "step": 3395
+    },
+    {
+      "epoch": 0.5208588957055215,
+      "grad_norm": 0.1751512885093689,
+      "learning_rate": 7.122365589363855e-05,
+      "loss": 1.1565,
+      "step": 3396
+    },
+    {
+      "epoch": 0.5210122699386504,
+      "grad_norm": 0.16323870420455933,
+      "learning_rate": 7.118720079120228e-05,
+      "loss": 1.0029,
+      "step": 3397
+    },
+    {
+      "epoch": 0.5211656441717791,
+      "grad_norm": 0.16858617961406708,
+      "learning_rate": 7.11507465918979e-05,
+      "loss": 1.0366,
+      "step": 3398
+    },
+    {
+      "epoch": 0.521319018404908,
+      "grad_norm": 0.1831703931093216,
+      "learning_rate": 7.111429330436026e-05,
+      "loss": 1.0412,
+      "step": 3399
+    },
+    {
+      "epoch": 0.5214723926380368,
+      "grad_norm": 0.16090618073940277,
+      "learning_rate": 7.1077840937224e-05,
+      "loss": 1.1178,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5216257668711657,
+      "grad_norm": 0.1841476857662201,
+      "learning_rate": 7.104138949912355e-05,
+      "loss": 1.1417,
+      "step": 3401
+    },
+    {
+      "epoch": 0.5217791411042945,
+      "grad_norm": 0.16766014695167542,
+      "learning_rate": 7.100493899869306e-05,
+      "loss": 1.0335,
+      "step": 3402
+    },
+    {
+      "epoch": 0.5219325153374234,
+      "grad_norm": 0.15937942266464233,
+      "learning_rate": 7.096848944456656e-05,
+      "loss": 1.0217,
+      "step": 3403
+    },
+    {
+      "epoch": 0.5220858895705521,
+      "grad_norm": 0.17221161723136902,
+      "learning_rate": 7.093204084537776e-05,
+      "loss": 1.0344,
+      "step": 3404
+    },
+    {
+      "epoch": 0.522239263803681,
+      "grad_norm": 0.1644064486026764,
+      "learning_rate": 7.089559320976018e-05,
+      "loss": 1.1833,
+      "step": 3405
+    },
+    {
+      "epoch": 0.5223926380368098,
+      "grad_norm": 0.16269896924495697,
+      "learning_rate": 7.085914654634717e-05,
+      "loss": 1.099,
+      "step": 3406
+    },
+    {
+      "epoch": 0.5225460122699387,
+      "grad_norm": 0.17388860881328583,
+      "learning_rate": 7.082270086377173e-05,
+      "loss": 1.197,
+      "step": 3407
+    },
+    {
+      "epoch": 0.5226993865030675,
+      "grad_norm": 0.24951070547103882,
+      "learning_rate": 7.078625617066673e-05,
+      "loss": 1.0176,
+      "step": 3408
+    },
+    {
+      "epoch": 0.5228527607361964,
+      "grad_norm": 0.17695893347263336,
+      "learning_rate": 7.074981247566475e-05,
+      "loss": 1.1411,
+      "step": 3409
+    },
+    {
+      "epoch": 0.5230061349693251,
+      "grad_norm": 0.15939389169216156,
+      "learning_rate": 7.071336978739817e-05,
+      "loss": 0.9847,
+      "step": 3410
+    },
+    {
+      "epoch": 0.523159509202454,
+      "grad_norm": 0.17157989740371704,
+      "learning_rate": 7.067692811449909e-05,
+      "loss": 1.1198,
+      "step": 3411
+    },
+    {
+      "epoch": 0.5233128834355828,
+      "grad_norm": 0.2771384119987488,
+      "learning_rate": 7.064048746559941e-05,
+      "loss": 1.1152,
+      "step": 3412
+    },
+    {
+      "epoch": 0.5234662576687117,
+      "grad_norm": 0.17456112802028656,
+      "learning_rate": 7.060404784933078e-05,
+      "loss": 1.1985,
+      "step": 3413
+    },
+    {
+      "epoch": 0.5236196319018405,
+      "grad_norm": 0.17151398956775665,
+      "learning_rate": 7.056760927432459e-05,
+      "loss": 1.1115,
+      "step": 3414
+    },
+    {
+      "epoch": 0.5237730061349694,
+      "grad_norm": 0.15843623876571655,
+      "learning_rate": 7.053117174921196e-05,
+      "loss": 1.0209,
+      "step": 3415
+    },
+    {
+      "epoch": 0.5239263803680981,
+      "grad_norm": 0.17492695152759552,
+      "learning_rate": 7.049473528262382e-05,
+      "loss": 1.0814,
+      "step": 3416
+    },
+    {
+      "epoch": 0.524079754601227,
+      "grad_norm": 0.15875624120235443,
+      "learning_rate": 7.04582998831908e-05,
+      "loss": 1.0029,
+      "step": 3417
+    },
+    {
+      "epoch": 0.5242331288343558,
+      "grad_norm": 0.15954287350177765,
+      "learning_rate": 7.042186555954333e-05,
+      "loss": 0.9981,
+      "step": 3418
+    },
+    {
+      "epoch": 0.5243865030674847,
+      "grad_norm": 0.15742646157741547,
+      "learning_rate": 7.038543232031152e-05,
+      "loss": 0.9213,
+      "step": 3419
+    },
+    {
+      "epoch": 0.5245398773006135,
+      "grad_norm": 0.16170817613601685,
+      "learning_rate": 7.034900017412527e-05,
+      "loss": 1.0148,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5246932515337424,
+      "grad_norm": 0.19887501001358032,
+      "learning_rate": 7.031256912961423e-05,
+      "loss": 1.0069,
+      "step": 3421
+    },
+    {
+      "epoch": 0.5248466257668711,
+      "grad_norm": 0.1697273999452591,
+      "learning_rate": 7.027613919540771e-05,
+      "loss": 1.0508,
+      "step": 3422
+    },
+    {
+      "epoch": 0.525,
+      "grad_norm": 0.17574365437030792,
+      "learning_rate": 7.023971038013485e-05,
+      "loss": 1.0174,
+      "step": 3423
+    },
+    {
+      "epoch": 0.5251533742331288,
+      "grad_norm": 0.16753290593624115,
+      "learning_rate": 7.020328269242447e-05,
+      "loss": 1.1575,
+      "step": 3424
+    },
+    {
+      "epoch": 0.5253067484662577,
+      "grad_norm": 0.17729055881500244,
+      "learning_rate": 7.016685614090516e-05,
+      "loss": 1.0505,
+      "step": 3425
+    },
+    {
+      "epoch": 0.5254601226993865,
+      "grad_norm": 0.17626497149467468,
+      "learning_rate": 7.013043073420522e-05,
+      "loss": 1.104,
+      "step": 3426
+    },
+    {
+      "epoch": 0.5256134969325154,
+      "grad_norm": 0.1746252030134201,
+      "learning_rate": 7.009400648095267e-05,
+      "loss": 1.1235,
+      "step": 3427
+    },
+    {
+      "epoch": 0.5257668711656441,
+      "grad_norm": 0.16394571959972382,
+      "learning_rate": 7.005758338977525e-05,
+      "loss": 1.0793,
+      "step": 3428
+    },
+    {
+      "epoch": 0.525920245398773,
+      "grad_norm": 0.15105700492858887,
+      "learning_rate": 7.00211614693005e-05,
+      "loss": 1.0419,
+      "step": 3429
+    },
+    {
+      "epoch": 0.5260736196319018,
+      "grad_norm": 0.1823078989982605,
+      "learning_rate": 6.998474072815554e-05,
+      "loss": 1.1561,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5262269938650307,
+      "grad_norm": 0.16824288666248322,
+      "learning_rate": 6.994832117496736e-05,
+      "loss": 1.1946,
+      "step": 3431
+    },
+    {
+      "epoch": 0.5263803680981595,
+      "grad_norm": 0.1831098049879074,
+      "learning_rate": 6.991190281836258e-05,
+      "loss": 1.0008,
+      "step": 3432
+    },
+    {
+      "epoch": 0.5265337423312884,
+      "grad_norm": 0.17319327592849731,
+      "learning_rate": 6.987548566696756e-05,
+      "loss": 1.1401,
+      "step": 3433
+    },
+    {
+      "epoch": 0.5266871165644171,
+      "grad_norm": 0.16220615804195404,
+      "learning_rate": 6.983906972940838e-05,
+      "loss": 0.9706,
+      "step": 3434
+    },
+    {
+      "epoch": 0.526840490797546,
+      "grad_norm": 0.18086391687393188,
+      "learning_rate": 6.980265501431082e-05,
+      "loss": 1.1595,
+      "step": 3435
+    },
+    {
+      "epoch": 0.5269938650306748,
+      "grad_norm": 0.15880908071994781,
+      "learning_rate": 6.976624153030041e-05,
+      "loss": 1.0337,
+      "step": 3436
+    },
+    {
+      "epoch": 0.5271472392638037,
+      "grad_norm": 0.16082361340522766,
+      "learning_rate": 6.97298292860023e-05,
+      "loss": 1.1384,
+      "step": 3437
+    },
+    {
+      "epoch": 0.5273006134969325,
+      "grad_norm": 0.18312779068946838,
+      "learning_rate": 6.969341829004145e-05,
+      "loss": 1.1518,
+      "step": 3438
+    },
+    {
+      "epoch": 0.5274539877300614,
+      "grad_norm": 0.16452737152576447,
+      "learning_rate": 6.965700855104244e-05,
+      "loss": 1.0587,
+      "step": 3439
+    },
+    {
+      "epoch": 0.5276073619631901,
+      "grad_norm": 0.18254221975803375,
+      "learning_rate": 6.962060007762961e-05,
+      "loss": 1.1599,
+      "step": 3440
+    },
+    {
+      "epoch": 0.527760736196319,
+      "grad_norm": 0.1626570224761963,
+      "learning_rate": 6.958419287842699e-05,
+      "loss": 1.0899,
+      "step": 3441
+    },
+    {
+      "epoch": 0.5279141104294478,
+      "grad_norm": 0.1655672788619995,
+      "learning_rate": 6.954778696205825e-05,
+      "loss": 1.0362,
+      "step": 3442
+    },
+    {
+      "epoch": 0.5280674846625767,
+      "grad_norm": 0.18394744396209717,
+      "learning_rate": 6.951138233714687e-05,
+      "loss": 1.1193,
+      "step": 3443
+    },
+    {
+      "epoch": 0.5282208588957055,
+      "grad_norm": 0.15760815143585205,
+      "learning_rate": 6.947497901231591e-05,
+      "loss": 1.1311,
+      "step": 3444
+    },
+    {
+      "epoch": 0.5283742331288344,
+      "grad_norm": 0.17035575211048126,
+      "learning_rate": 6.943857699618818e-05,
+      "loss": 1.0485,
+      "step": 3445
+    },
+    {
+      "epoch": 0.5285276073619631,
+      "grad_norm": 0.1665990799665451,
+      "learning_rate": 6.940217629738619e-05,
+      "loss": 1.0393,
+      "step": 3446
+    },
+    {
+      "epoch": 0.528680981595092,
+      "grad_norm": 0.16503994166851044,
+      "learning_rate": 6.93657769245321e-05,
+      "loss": 1.1645,
+      "step": 3447
+    },
+    {
+      "epoch": 0.5288343558282208,
+      "grad_norm": 0.16222040355205536,
+      "learning_rate": 6.932937888624779e-05,
+      "loss": 1.0243,
+      "step": 3448
+    },
+    {
+      "epoch": 0.5289877300613497,
+      "grad_norm": 0.16192078590393066,
+      "learning_rate": 6.929298219115477e-05,
+      "loss": 1.048,
+      "step": 3449
+    },
+    {
+      "epoch": 0.5291411042944786,
+      "grad_norm": 0.1603655219078064,
+      "learning_rate": 6.92565868478743e-05,
+      "loss": 1.1407,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5292944785276074,
+      "grad_norm": 0.15752007067203522,
+      "learning_rate": 6.922019286502727e-05,
+      "loss": 0.9969,
+      "step": 3451
+    },
+    {
+      "epoch": 0.5294478527607362,
+      "grad_norm": 0.1572183221578598,
+      "learning_rate": 6.918380025123429e-05,
+      "loss": 0.9787,
+      "step": 3452
+    },
+    {
+      "epoch": 0.529601226993865,
+      "grad_norm": 0.16903875768184662,
+      "learning_rate": 6.91474090151156e-05,
+      "loss": 1.115,
+      "step": 3453
+    },
+    {
+      "epoch": 0.5297546012269939,
+      "grad_norm": 0.16609913110733032,
+      "learning_rate": 6.911101916529115e-05,
+      "loss": 1.173,
+      "step": 3454
+    },
+    {
+      "epoch": 0.5299079754601227,
+      "grad_norm": 0.16663315892219543,
+      "learning_rate": 6.907463071038054e-05,
+      "loss": 1.0857,
+      "step": 3455
+    },
+    {
+      "epoch": 0.5300613496932516,
+      "grad_norm": 0.1632128804922104,
+      "learning_rate": 6.903824365900307e-05,
+      "loss": 1.0939,
+      "step": 3456
+    },
+    {
+      "epoch": 0.5302147239263804,
+      "grad_norm": 0.17721736431121826,
+      "learning_rate": 6.900185801977765e-05,
+      "loss": 1.0965,
+      "step": 3457
+    },
+    {
+      "epoch": 0.5303680981595092,
+      "grad_norm": 0.15368637442588806,
+      "learning_rate": 6.896547380132291e-05,
+      "loss": 0.9053,
+      "step": 3458
+    },
+    {
+      "epoch": 0.530521472392638,
+      "grad_norm": 0.16886506974697113,
+      "learning_rate": 6.892909101225711e-05,
+      "loss": 1.1413,
+      "step": 3459
+    },
+    {
+      "epoch": 0.5306748466257669,
+      "grad_norm": 0.17498710751533508,
+      "learning_rate": 6.889270966119822e-05,
+      "loss": 1.0903,
+      "step": 3460
+    },
+    {
+      "epoch": 0.5308282208588957,
+      "grad_norm": 0.15783564746379852,
+      "learning_rate": 6.88563297567638e-05,
+      "loss": 1.0724,
+      "step": 3461
+    },
+    {
+      "epoch": 0.5309815950920246,
+      "grad_norm": 0.15758667886257172,
+      "learning_rate": 6.881995130757111e-05,
+      "loss": 1.0231,
+      "step": 3462
+    },
+    {
+      "epoch": 0.5311349693251534,
+      "grad_norm": 0.16468128561973572,
+      "learning_rate": 6.878357432223706e-05,
+      "loss": 1.1191,
+      "step": 3463
+    },
+    {
+      "epoch": 0.5312883435582823,
+      "grad_norm": 0.17636190354824066,
+      "learning_rate": 6.874719880937822e-05,
+      "loss": 1.0369,
+      "step": 3464
+    },
+    {
+      "epoch": 0.531441717791411,
+      "grad_norm": 0.16678442060947418,
+      "learning_rate": 6.871082477761079e-05,
+      "loss": 1.0501,
+      "step": 3465
+    },
+    {
+      "epoch": 0.5315950920245399,
+      "grad_norm": 0.16989123821258545,
+      "learning_rate": 6.867445223555063e-05,
+      "loss": 1.1394,
+      "step": 3466
+    },
+    {
+      "epoch": 0.5317484662576687,
+      "grad_norm": 0.1685888022184372,
+      "learning_rate": 6.863808119181324e-05,
+      "loss": 1.0585,
+      "step": 3467
+    },
+    {
+      "epoch": 0.5319018404907976,
+      "grad_norm": 0.17636343836784363,
+      "learning_rate": 6.860171165501381e-05,
+      "loss": 1.0593,
+      "step": 3468
+    },
+    {
+      "epoch": 0.5320552147239264,
+      "grad_norm": 0.18702596426010132,
+      "learning_rate": 6.856534363376709e-05,
+      "loss": 1.1949,
+      "step": 3469
+    },
+    {
+      "epoch": 0.5322085889570553,
+      "grad_norm": 0.15549130737781525,
+      "learning_rate": 6.852897713668755e-05,
+      "loss": 0.981,
+      "step": 3470
+    },
+    {
+      "epoch": 0.532361963190184,
+      "grad_norm": 0.17029328644275665,
+      "learning_rate": 6.849261217238927e-05,
+      "loss": 1.111,
+      "step": 3471
+    },
+    {
+      "epoch": 0.5325153374233129,
+      "grad_norm": 0.17223390936851501,
+      "learning_rate": 6.845624874948593e-05,
+      "loss": 1.0727,
+      "step": 3472
+    },
+    {
+      "epoch": 0.5326687116564417,
+      "grad_norm": 0.15325692296028137,
+      "learning_rate": 6.84198868765909e-05,
+      "loss": 1.0581,
+      "step": 3473
+    },
+    {
+      "epoch": 0.5328220858895706,
+      "grad_norm": 0.1582840383052826,
+      "learning_rate": 6.838352656231714e-05,
+      "loss": 1.0649,
+      "step": 3474
+    },
+    {
+      "epoch": 0.5329754601226994,
+      "grad_norm": 0.1748669147491455,
+      "learning_rate": 6.83471678152773e-05,
+      "loss": 1.1645,
+      "step": 3475
+    },
+    {
+      "epoch": 0.5331288343558283,
+      "grad_norm": 0.17215079069137573,
+      "learning_rate": 6.83108106440836e-05,
+      "loss": 1.2311,
+      "step": 3476
+    },
+    {
+      "epoch": 0.533282208588957,
+      "grad_norm": 0.16793349385261536,
+      "learning_rate": 6.827445505734791e-05,
+      "loss": 1.0624,
+      "step": 3477
+    },
+    {
+      "epoch": 0.5334355828220859,
+      "grad_norm": 0.1609848141670227,
+      "learning_rate": 6.823810106368171e-05,
+      "loss": 1.0986,
+      "step": 3478
+    },
+    {
+      "epoch": 0.5335889570552147,
+      "grad_norm": 0.15882481634616852,
+      "learning_rate": 6.820174867169614e-05,
+      "loss": 1.0889,
+      "step": 3479
+    },
+    {
+      "epoch": 0.5337423312883436,
+      "grad_norm": 0.1547410786151886,
+      "learning_rate": 6.816539789000191e-05,
+      "loss": 1.0467,
+      "step": 3480
+    },
+    {
+      "epoch": 0.5338957055214724,
+      "grad_norm": 0.16782033443450928,
+      "learning_rate": 6.812904872720937e-05,
+      "loss": 1.212,
+      "step": 3481
+    },
+    {
+      "epoch": 0.5340490797546013,
+      "grad_norm": 0.1640184223651886,
+      "learning_rate": 6.809270119192852e-05,
+      "loss": 1.0551,
+      "step": 3482
+    },
+    {
+      "epoch": 0.53420245398773,
+      "grad_norm": 0.16266794502735138,
+      "learning_rate": 6.805635529276892e-05,
+      "loss": 1.0371,
+      "step": 3483
+    },
+    {
+      "epoch": 0.5343558282208589,
+      "grad_norm": 0.16706693172454834,
+      "learning_rate": 6.802001103833976e-05,
+      "loss": 1.03,
+      "step": 3484
+    },
+    {
+      "epoch": 0.5345092024539877,
+      "grad_norm": 0.16543152928352356,
+      "learning_rate": 6.798366843724988e-05,
+      "loss": 1.031,
+      "step": 3485
+    },
+    {
+      "epoch": 0.5346625766871166,
+      "grad_norm": 0.15794984996318817,
+      "learning_rate": 6.794732749810767e-05,
+      "loss": 1.0326,
+      "step": 3486
+    },
+    {
+      "epoch": 0.5348159509202454,
+      "grad_norm": 0.16290351748466492,
+      "learning_rate": 6.791098822952117e-05,
+      "loss": 1.0323,
+      "step": 3487
+    },
+    {
+      "epoch": 0.5349693251533743,
+      "grad_norm": 0.17673058807849884,
+      "learning_rate": 6.7874650640098e-05,
+      "loss": 1.1378,
+      "step": 3488
+    },
+    {
+      "epoch": 0.535122699386503,
+      "grad_norm": 0.1633237898349762,
+      "learning_rate": 6.783831473844539e-05,
+      "loss": 1.044,
+      "step": 3489
+    },
+    {
+      "epoch": 0.5352760736196319,
+      "grad_norm": 0.16477954387664795,
+      "learning_rate": 6.780198053317016e-05,
+      "loss": 1.1629,
+      "step": 3490
+    },
+    {
+      "epoch": 0.5354294478527607,
+      "grad_norm": 0.17096509039402008,
+      "learning_rate": 6.776564803287875e-05,
+      "loss": 1.1412,
+      "step": 3491
+    },
+    {
+      "epoch": 0.5355828220858896,
+      "grad_norm": 0.16369840502738953,
+      "learning_rate": 6.772931724617718e-05,
+      "loss": 1.06,
+      "step": 3492
+    },
+    {
+      "epoch": 0.5357361963190184,
+      "grad_norm": 0.16654792428016663,
+      "learning_rate": 6.769298818167107e-05,
+      "loss": 0.9584,
+      "step": 3493
+    },
+    {
+      "epoch": 0.5358895705521473,
+      "grad_norm": 0.15233644843101501,
+      "learning_rate": 6.76566608479656e-05,
+      "loss": 0.9548,
+      "step": 3494
+    },
+    {
+      "epoch": 0.536042944785276,
+      "grad_norm": 0.14970964193344116,
+      "learning_rate": 6.762033525366561e-05,
+      "loss": 0.9452,
+      "step": 3495
+    },
+    {
+      "epoch": 0.5361963190184049,
+      "grad_norm": 0.16172778606414795,
+      "learning_rate": 6.758401140737547e-05,
+      "loss": 1.0894,
+      "step": 3496
+    },
+    {
+      "epoch": 0.5363496932515337,
+      "grad_norm": 0.1653529405593872,
+      "learning_rate": 6.754768931769916e-05,
+      "loss": 1.1743,
+      "step": 3497
+    },
+    {
+      "epoch": 0.5365030674846626,
+      "grad_norm": 0.1605312079191208,
+      "learning_rate": 6.751136899324024e-05,
+      "loss": 1.0979,
+      "step": 3498
+    },
+    {
+      "epoch": 0.5366564417177914,
+      "grad_norm": 0.16550888121128082,
+      "learning_rate": 6.747505044260184e-05,
+      "loss": 1.1481,
+      "step": 3499
+    },
+    {
+      "epoch": 0.5368098159509203,
+      "grad_norm": 0.16474755108356476,
+      "learning_rate": 6.743873367438668e-05,
+      "loss": 1.1058,
+      "step": 3500
+    },
+    {
+      "epoch": 0.536963190184049,
+      "grad_norm": 0.16065819561481476,
+      "learning_rate": 6.740241869719706e-05,
+      "loss": 1.1008,
+      "step": 3501
+    },
+    {
+      "epoch": 0.5371165644171779,
+      "grad_norm": 0.15991096198558807,
+      "learning_rate": 6.736610551963485e-05,
+      "loss": 1.04,
+      "step": 3502
+    },
+    {
+      "epoch": 0.5372699386503067,
+      "grad_norm": 0.17472554743289948,
+      "learning_rate": 6.732979415030149e-05,
+      "loss": 1.1611,
+      "step": 3503
+    },
+    {
+      "epoch": 0.5374233128834356,
+      "grad_norm": 0.15896394848823547,
+      "learning_rate": 6.729348459779801e-05,
+      "loss": 1.0338,
+      "step": 3504
+    },
+    {
+      "epoch": 0.5375766871165644,
+      "grad_norm": 0.16643494367599487,
+      "learning_rate": 6.725717687072501e-05,
+      "loss": 1.1047,
+      "step": 3505
+    },
+    {
+      "epoch": 0.5377300613496933,
+      "grad_norm": 0.162796750664711,
+      "learning_rate": 6.722087097768263e-05,
+      "loss": 1.016,
+      "step": 3506
+    },
+    {
+      "epoch": 0.537883435582822,
+      "grad_norm": 0.17863915860652924,
+      "learning_rate": 6.718456692727059e-05,
+      "loss": 1.1395,
+      "step": 3507
+    },
+    {
+      "epoch": 0.5380368098159509,
+      "grad_norm": 0.16205811500549316,
+      "learning_rate": 6.714826472808817e-05,
+      "loss": 1.0084,
+      "step": 3508
+    },
+    {
+      "epoch": 0.5381901840490797,
+      "grad_norm": 0.16102614998817444,
+      "learning_rate": 6.711196438873422e-05,
+      "loss": 1.0441,
+      "step": 3509
+    },
+    {
+      "epoch": 0.5383435582822086,
+      "grad_norm": 0.14174002408981323,
+      "learning_rate": 6.707566591780716e-05,
+      "loss": 0.858,
+      "step": 3510
+    },
+    {
+      "epoch": 0.5384969325153375,
+      "grad_norm": 0.16296502947807312,
+      "learning_rate": 6.703936932390492e-05,
+      "loss": 1.0924,
+      "step": 3511
+    },
+    {
+      "epoch": 0.5386503067484663,
+      "grad_norm": 0.16159355640411377,
+      "learning_rate": 6.700307461562506e-05,
+      "loss": 1.1361,
+      "step": 3512
+    },
+    {
+      "epoch": 0.5388036809815951,
+      "grad_norm": 0.15749932825565338,
+      "learning_rate": 6.696678180156462e-05,
+      "loss": 1.0097,
+      "step": 3513
+    },
+    {
+      "epoch": 0.5389570552147239,
+      "grad_norm": 0.14929069578647614,
+      "learning_rate": 6.693049089032025e-05,
+      "loss": 1.0118,
+      "step": 3514
+    },
+    {
+      "epoch": 0.5391104294478528,
+      "grad_norm": 0.1631220132112503,
+      "learning_rate": 6.68942018904881e-05,
+      "loss": 1.0304,
+      "step": 3515
+    },
+    {
+      "epoch": 0.5392638036809816,
+      "grad_norm": 0.1777171492576599,
+      "learning_rate": 6.68579148106639e-05,
+      "loss": 1.2407,
+      "step": 3516
+    },
+    {
+      "epoch": 0.5394171779141105,
+      "grad_norm": 0.15380235016345978,
+      "learning_rate": 6.682162965944289e-05,
+      "loss": 1.0422,
+      "step": 3517
+    },
+    {
+      "epoch": 0.5395705521472393,
+      "grad_norm": 0.16666916012763977,
+      "learning_rate": 6.678534644541992e-05,
+      "loss": 1.1132,
+      "step": 3518
+    },
+    {
+      "epoch": 0.5397239263803681,
+      "grad_norm": 0.1599230170249939,
+      "learning_rate": 6.674906517718931e-05,
+      "loss": 1.0177,
+      "step": 3519
+    },
+    {
+      "epoch": 0.5398773006134969,
+      "grad_norm": 0.16284504532814026,
+      "learning_rate": 6.671278586334495e-05,
+      "loss": 1.0765,
+      "step": 3520
+    },
+    {
+      "epoch": 0.5400306748466258,
+      "grad_norm": 0.18423151969909668,
+      "learning_rate": 6.66765085124803e-05,
+      "loss": 1.1657,
+      "step": 3521
+    },
+    {
+      "epoch": 0.5401840490797546,
+      "grad_norm": 0.1703919619321823,
+      "learning_rate": 6.664023313318828e-05,
+      "loss": 1.1036,
+      "step": 3522
+    },
+    {
+      "epoch": 0.5403374233128835,
+      "grad_norm": 0.15698012709617615,
+      "learning_rate": 6.660395973406139e-05,
+      "loss": 1.0663,
+      "step": 3523
+    },
+    {
+      "epoch": 0.5404907975460123,
+      "grad_norm": 0.1785694807767868,
+      "learning_rate": 6.656768832369165e-05,
+      "loss": 1.2472,
+      "step": 3524
+    },
+    {
+      "epoch": 0.5406441717791411,
+      "grad_norm": 0.1650916188955307,
+      "learning_rate": 6.653141891067064e-05,
+      "loss": 1.0912,
+      "step": 3525
+    },
+    {
+      "epoch": 0.5407975460122699,
+      "grad_norm": 0.15214000642299652,
+      "learning_rate": 6.649515150358942e-05,
+      "loss": 1.0231,
+      "step": 3526
+    },
+    {
+      "epoch": 0.5409509202453988,
+      "grad_norm": 0.16546806693077087,
+      "learning_rate": 6.645888611103858e-05,
+      "loss": 1.0798,
+      "step": 3527
+    },
+    {
+      "epoch": 0.5411042944785276,
+      "grad_norm": 0.1601184457540512,
+      "learning_rate": 6.642262274160829e-05,
+      "loss": 1.0624,
+      "step": 3528
+    },
+    {
+      "epoch": 0.5412576687116565,
+      "grad_norm": 0.15619443356990814,
+      "learning_rate": 6.638636140388816e-05,
+      "loss": 1.1092,
+      "step": 3529
+    },
+    {
+      "epoch": 0.5414110429447853,
+      "grad_norm": 0.16686996817588806,
+      "learning_rate": 6.635010210646738e-05,
+      "loss": 1.1123,
+      "step": 3530
+    },
+    {
+      "epoch": 0.5415644171779141,
+      "grad_norm": 0.1585853099822998,
+      "learning_rate": 6.631384485793463e-05,
+      "loss": 0.9105,
+      "step": 3531
+    },
+    {
+      "epoch": 0.5417177914110429,
+      "grad_norm": 0.17483842372894287,
+      "learning_rate": 6.62775896668781e-05,
+      "loss": 1.1741,
+      "step": 3532
+    },
+    {
+      "epoch": 0.5418711656441718,
+      "grad_norm": 0.1699761003255844,
+      "learning_rate": 6.624133654188554e-05,
+      "loss": 1.1626,
+      "step": 3533
+    },
+    {
+      "epoch": 0.5420245398773006,
+      "grad_norm": 0.16776113212108612,
+      "learning_rate": 6.620508549154411e-05,
+      "loss": 1.0304,
+      "step": 3534
+    },
+    {
+      "epoch": 0.5421779141104295,
+      "grad_norm": 0.18671639263629913,
+      "learning_rate": 6.616883652444056e-05,
+      "loss": 1.2019,
+      "step": 3535
+    },
+    {
+      "epoch": 0.5423312883435583,
+      "grad_norm": 0.15667033195495605,
+      "learning_rate": 6.613258964916115e-05,
+      "loss": 0.9835,
+      "step": 3536
+    },
+    {
+      "epoch": 0.5424846625766871,
+      "grad_norm": 0.15698322653770447,
+      "learning_rate": 6.609634487429161e-05,
+      "loss": 1.1038,
+      "step": 3537
+    },
+    {
+      "epoch": 0.5426380368098159,
+      "grad_norm": 0.1714984029531479,
+      "learning_rate": 6.606010220841717e-05,
+      "loss": 1.1052,
+      "step": 3538
+    },
+    {
+      "epoch": 0.5427914110429448,
+      "grad_norm": 0.16964519023895264,
+      "learning_rate": 6.60238616601226e-05,
+      "loss": 1.0817,
+      "step": 3539
+    },
+    {
+      "epoch": 0.5429447852760736,
+      "grad_norm": 0.17209139466285706,
+      "learning_rate": 6.598762323799212e-05,
+      "loss": 1.1148,
+      "step": 3540
+    },
+    {
+      "epoch": 0.5430981595092025,
+      "grad_norm": 0.1616567224264145,
+      "learning_rate": 6.59513869506095e-05,
+      "loss": 0.9749,
+      "step": 3541
+    },
+    {
+      "epoch": 0.5432515337423313,
+      "grad_norm": 0.15292274951934814,
+      "learning_rate": 6.591515280655793e-05,
+      "loss": 0.9576,
+      "step": 3542
+    },
+    {
+      "epoch": 0.5434049079754601,
+      "grad_norm": 0.15400253236293793,
+      "learning_rate": 6.587892081442018e-05,
+      "loss": 1.0078,
+      "step": 3543
+    },
+    {
+      "epoch": 0.5435582822085889,
+      "grad_norm": 0.17184637486934662,
+      "learning_rate": 6.584269098277843e-05,
+      "loss": 1.0732,
+      "step": 3544
+    },
+    {
+      "epoch": 0.5437116564417178,
+      "grad_norm": 0.14827391505241394,
+      "learning_rate": 6.58064633202144e-05,
+      "loss": 0.9646,
+      "step": 3545
+    },
+    {
+      "epoch": 0.5438650306748466,
+      "grad_norm": 0.1494305580854416,
+      "learning_rate": 6.577023783530928e-05,
+      "loss": 1.0255,
+      "step": 3546
+    },
+    {
+      "epoch": 0.5440184049079755,
+      "grad_norm": 0.17969827353954315,
+      "learning_rate": 6.573401453664374e-05,
+      "loss": 1.1071,
+      "step": 3547
+    },
+    {
+      "epoch": 0.5441717791411043,
+      "grad_norm": 0.22177502512931824,
+      "learning_rate": 6.569779343279796e-05,
+      "loss": 1.0414,
+      "step": 3548
+    },
+    {
+      "epoch": 0.5443251533742332,
+      "grad_norm": 0.16644063591957092,
+      "learning_rate": 6.566157453235158e-05,
+      "loss": 1.057,
+      "step": 3549
+    },
+    {
+      "epoch": 0.5444785276073619,
+      "grad_norm": 0.14325271546840668,
+      "learning_rate": 6.562535784388366e-05,
+      "loss": 0.917,
+      "step": 3550
+    },
+    {
+      "epoch": 0.5446319018404908,
+      "grad_norm": 0.15619418025016785,
+      "learning_rate": 6.558914337597284e-05,
+      "loss": 1.0226,
+      "step": 3551
+    },
+    {
+      "epoch": 0.5447852760736196,
+      "grad_norm": 0.16915465891361237,
+      "learning_rate": 6.555293113719716e-05,
+      "loss": 1.175,
+      "step": 3552
+    },
+    {
+      "epoch": 0.5449386503067485,
+      "grad_norm": 0.16401277482509613,
+      "learning_rate": 6.551672113613419e-05,
+      "loss": 1.0918,
+      "step": 3553
+    },
+    {
+      "epoch": 0.5450920245398773,
+      "grad_norm": 0.1795947253704071,
+      "learning_rate": 6.54805133813609e-05,
+      "loss": 1.0462,
+      "step": 3554
+    },
+    {
+      "epoch": 0.5452453987730062,
+      "grad_norm": 0.177162766456604,
+      "learning_rate": 6.544430788145378e-05,
+      "loss": 1.1969,
+      "step": 3555
+    },
+    {
+      "epoch": 0.5453987730061349,
+      "grad_norm": 0.15788620710372925,
+      "learning_rate": 6.54081046449888e-05,
+      "loss": 1.083,
+      "step": 3556
+    },
+    {
+      "epoch": 0.5455521472392638,
+      "grad_norm": 0.1626262068748474,
+      "learning_rate": 6.537190368054132e-05,
+      "loss": 1.091,
+      "step": 3557
+    },
+    {
+      "epoch": 0.5457055214723926,
+      "grad_norm": 0.1693645864725113,
+      "learning_rate": 6.533570499668622e-05,
+      "loss": 1.0893,
+      "step": 3558
+    },
+    {
+      "epoch": 0.5458588957055215,
+      "grad_norm": 0.17898011207580566,
+      "learning_rate": 6.529950860199784e-05,
+      "loss": 1.1941,
+      "step": 3559
+    },
+    {
+      "epoch": 0.5460122699386503,
+      "grad_norm": 0.14966538548469543,
+      "learning_rate": 6.526331450504995e-05,
+      "loss": 1.026,
+      "step": 3560
+    },
+    {
+      "epoch": 0.5461656441717792,
+      "grad_norm": 0.1687460094690323,
+      "learning_rate": 6.52271227144158e-05,
+      "loss": 1.0129,
+      "step": 3561
+    },
+    {
+      "epoch": 0.5463190184049079,
+      "grad_norm": 0.16827961802482605,
+      "learning_rate": 6.519093323866808e-05,
+      "loss": 1.1101,
+      "step": 3562
+    },
+    {
+      "epoch": 0.5464723926380368,
+      "grad_norm": 0.15890853106975555,
+      "learning_rate": 6.515474608637894e-05,
+      "loss": 1.1117,
+      "step": 3563
+    },
+    {
+      "epoch": 0.5466257668711656,
+      "grad_norm": 0.17539769411087036,
+      "learning_rate": 6.511856126611997e-05,
+      "loss": 1.1592,
+      "step": 3564
+    },
+    {
+      "epoch": 0.5467791411042945,
+      "grad_norm": 0.16487926244735718,
+      "learning_rate": 6.508237878646221e-05,
+      "loss": 1.1283,
+      "step": 3565
+    },
+    {
+      "epoch": 0.5469325153374233,
+      "grad_norm": 0.174005389213562,
+      "learning_rate": 6.504619865597616e-05,
+      "loss": 1.1068,
+      "step": 3566
+    },
+    {
+      "epoch": 0.5470858895705522,
+      "grad_norm": 0.16081731021404266,
+      "learning_rate": 6.501002088323172e-05,
+      "loss": 1.0651,
+      "step": 3567
+    },
+    {
+      "epoch": 0.5472392638036809,
+      "grad_norm": 0.14418929815292358,
+      "learning_rate": 6.497384547679832e-05,
+      "loss": 0.9053,
+      "step": 3568
+    },
+    {
+      "epoch": 0.5473926380368098,
+      "grad_norm": 0.17355632781982422,
+      "learning_rate": 6.49376724452447e-05,
+      "loss": 1.2306,
+      "step": 3569
+    },
+    {
+      "epoch": 0.5475460122699386,
+      "grad_norm": 0.17231911420822144,
+      "learning_rate": 6.490150179713918e-05,
+      "loss": 1.1102,
+      "step": 3570
+    },
+    {
+      "epoch": 0.5476993865030675,
+      "grad_norm": 0.1623077243566513,
+      "learning_rate": 6.48653335410494e-05,
+      "loss": 1.0583,
+      "step": 3571
+    },
+    {
+      "epoch": 0.5478527607361963,
+      "grad_norm": 0.17018643021583557,
+      "learning_rate": 6.482916768554252e-05,
+      "loss": 1.069,
+      "step": 3572
+    },
+    {
+      "epoch": 0.5480061349693252,
+      "grad_norm": 0.16610197722911835,
+      "learning_rate": 6.479300423918505e-05,
+      "loss": 1.0736,
+      "step": 3573
+    },
+    {
+      "epoch": 0.548159509202454,
+      "grad_norm": 0.19330105185508728,
+      "learning_rate": 6.4756843210543e-05,
+      "loss": 1.1969,
+      "step": 3574
+    },
+    {
+      "epoch": 0.5483128834355828,
+      "grad_norm": 0.1642577350139618,
+      "learning_rate": 6.472068460818175e-05,
+      "loss": 1.0355,
+      "step": 3575
+    },
+    {
+      "epoch": 0.5484662576687117,
+      "grad_norm": 0.15902185440063477,
+      "learning_rate": 6.468452844066616e-05,
+      "loss": 1.0931,
+      "step": 3576
+    },
+    {
+      "epoch": 0.5486196319018405,
+      "grad_norm": 0.16471059620380402,
+      "learning_rate": 6.464837471656048e-05,
+      "loss": 1.1087,
+      "step": 3577
+    },
+    {
+      "epoch": 0.5487730061349694,
+      "grad_norm": 0.1739111840724945,
+      "learning_rate": 6.461222344442836e-05,
+      "loss": 1.0987,
+      "step": 3578
+    },
+    {
+      "epoch": 0.5489263803680982,
+      "grad_norm": 0.1611456722021103,
+      "learning_rate": 6.457607463283293e-05,
+      "loss": 1.1191,
+      "step": 3579
+    },
+    {
+      "epoch": 0.549079754601227,
+      "grad_norm": 0.1770811378955841,
+      "learning_rate": 6.453992829033668e-05,
+      "loss": 1.0983,
+      "step": 3580
+    },
+    {
+      "epoch": 0.5492331288343558,
+      "grad_norm": 0.15897606313228607,
+      "learning_rate": 6.450378442550154e-05,
+      "loss": 0.9574,
+      "step": 3581
+    },
+    {
+      "epoch": 0.5493865030674847,
+      "grad_norm": 0.16113418340682983,
+      "learning_rate": 6.446764304688887e-05,
+      "loss": 1.0034,
+      "step": 3582
+    },
+    {
+      "epoch": 0.5495398773006135,
+      "grad_norm": 0.16184616088867188,
+      "learning_rate": 6.44315041630594e-05,
+      "loss": 1.0776,
+      "step": 3583
+    },
+    {
+      "epoch": 0.5496932515337424,
+      "grad_norm": 0.17578361928462982,
+      "learning_rate": 6.439536778257334e-05,
+      "loss": 1.227,
+      "step": 3584
+    },
+    {
+      "epoch": 0.5498466257668712,
+      "grad_norm": 0.15214155614376068,
+      "learning_rate": 6.435923391399019e-05,
+      "loss": 1.0176,
+      "step": 3585
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.16106829047203064,
+      "learning_rate": 6.432310256586897e-05,
+      "loss": 1.122,
+      "step": 3586
+    },
+    {
+      "epoch": 0.5501533742331288,
+      "grad_norm": 0.14416632056236267,
+      "learning_rate": 6.428697374676803e-05,
+      "loss": 0.8759,
+      "step": 3587
+    },
+    {
+      "epoch": 0.5503067484662577,
+      "grad_norm": 0.1838780641555786,
+      "learning_rate": 6.425084746524517e-05,
+      "loss": 1.2815,
+      "step": 3588
+    },
+    {
+      "epoch": 0.5504601226993865,
+      "grad_norm": 0.15805886685848236,
+      "learning_rate": 6.421472372985757e-05,
+      "loss": 1.1003,
+      "step": 3589
+    },
+    {
+      "epoch": 0.5506134969325154,
+      "grad_norm": 0.16670551896095276,
+      "learning_rate": 6.417860254916178e-05,
+      "loss": 1.2004,
+      "step": 3590
+    },
+    {
+      "epoch": 0.5507668711656442,
+      "grad_norm": 0.16837374866008759,
+      "learning_rate": 6.41424839317138e-05,
+      "loss": 1.1314,
+      "step": 3591
+    },
+    {
+      "epoch": 0.550920245398773,
+      "grad_norm": 0.15947063267230988,
+      "learning_rate": 6.410636788606897e-05,
+      "loss": 1.1178,
+      "step": 3592
+    },
+    {
+      "epoch": 0.5510736196319018,
+      "grad_norm": 0.16819867491722107,
+      "learning_rate": 6.407025442078202e-05,
+      "loss": 1.2335,
+      "step": 3593
+    },
+    {
+      "epoch": 0.5512269938650307,
+      "grad_norm": 0.15961135923862457,
+      "learning_rate": 6.403414354440712e-05,
+      "loss": 1.0643,
+      "step": 3594
+    },
+    {
+      "epoch": 0.5513803680981595,
+      "grad_norm": 0.15976320207118988,
+      "learning_rate": 6.39980352654978e-05,
+      "loss": 0.9984,
+      "step": 3595
+    },
+    {
+      "epoch": 0.5515337423312884,
+      "grad_norm": 0.17756947875022888,
+      "learning_rate": 6.396192959260697e-05,
+      "loss": 1.1814,
+      "step": 3596
+    },
+    {
+      "epoch": 0.5516871165644172,
+      "grad_norm": 0.17268332839012146,
+      "learning_rate": 6.392582653428693e-05,
+      "loss": 1.135,
+      "step": 3597
+    },
+    {
+      "epoch": 0.551840490797546,
+      "grad_norm": 0.17706866562366486,
+      "learning_rate": 6.388972609908934e-05,
+      "loss": 1.2045,
+      "step": 3598
+    },
+    {
+      "epoch": 0.5519938650306748,
+      "grad_norm": 0.15976423025131226,
+      "learning_rate": 6.385362829556527e-05,
+      "loss": 0.9897,
+      "step": 3599
+    },
+    {
+      "epoch": 0.5521472392638037,
+      "grad_norm": 0.1786051094532013,
+      "learning_rate": 6.381753313226513e-05,
+      "loss": 1.2498,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5523006134969325,
+      "grad_norm": 0.16335231065750122,
+      "learning_rate": 6.378144061773874e-05,
+      "loss": 1.0571,
+      "step": 3601
+    },
+    {
+      "epoch": 0.5524539877300614,
+      "grad_norm": 0.16432592272758484,
+      "learning_rate": 6.374535076053529e-05,
+      "loss": 1.0935,
+      "step": 3602
+    },
+    {
+      "epoch": 0.5526073619631902,
+      "grad_norm": 0.15722216665744781,
+      "learning_rate": 6.37092635692033e-05,
+      "loss": 1.0413,
+      "step": 3603
+    },
+    {
+      "epoch": 0.552760736196319,
+      "grad_norm": 0.18273420631885529,
+      "learning_rate": 6.36731790522907e-05,
+      "loss": 1.1641,
+      "step": 3604
+    },
+    {
+      "epoch": 0.5529141104294478,
+      "grad_norm": 0.16466134786605835,
+      "learning_rate": 6.36370972183448e-05,
+      "loss": 1.0815,
+      "step": 3605
+    },
+    {
+      "epoch": 0.5530674846625767,
+      "grad_norm": 0.1588336080312729,
+      "learning_rate": 6.360101807591222e-05,
+      "loss": 1.0205,
+      "step": 3606
+    },
+    {
+      "epoch": 0.5532208588957055,
+      "grad_norm": 0.17187876999378204,
+      "learning_rate": 6.3564941633539e-05,
+      "loss": 0.9826,
+      "step": 3607
+    },
+    {
+      "epoch": 0.5533742331288344,
+      "grad_norm": 0.15956705808639526,
+      "learning_rate": 6.352886789977045e-05,
+      "loss": 1.0969,
+      "step": 3608
+    },
+    {
+      "epoch": 0.5535276073619632,
+      "grad_norm": 0.1537466198205948,
+      "learning_rate": 6.349279688315135e-05,
+      "loss": 1.007,
+      "step": 3609
+    },
+    {
+      "epoch": 0.553680981595092,
+      "grad_norm": 0.1723203808069229,
+      "learning_rate": 6.345672859222579e-05,
+      "loss": 1.047,
+      "step": 3610
+    },
+    {
+      "epoch": 0.5538343558282208,
+      "grad_norm": 0.1780226081609726,
+      "learning_rate": 6.342066303553719e-05,
+      "loss": 1.1274,
+      "step": 3611
+    },
+    {
+      "epoch": 0.5539877300613497,
+      "grad_norm": 0.16775117814540863,
+      "learning_rate": 6.338460022162833e-05,
+      "loss": 1.102,
+      "step": 3612
+    },
+    {
+      "epoch": 0.5541411042944785,
+      "grad_norm": 0.18992358446121216,
+      "learning_rate": 6.334854015904137e-05,
+      "loss": 1.0894,
+      "step": 3613
+    },
+    {
+      "epoch": 0.5542944785276074,
+      "grad_norm": 0.14848209917545319,
+      "learning_rate": 6.331248285631783e-05,
+      "loss": 0.9672,
+      "step": 3614
+    },
+    {
+      "epoch": 0.5544478527607362,
+      "grad_norm": 0.15796998143196106,
+      "learning_rate": 6.327642832199852e-05,
+      "loss": 1.1186,
+      "step": 3615
+    },
+    {
+      "epoch": 0.554601226993865,
+      "grad_norm": 0.16426272690296173,
+      "learning_rate": 6.324037656462362e-05,
+      "loss": 1.0862,
+      "step": 3616
+    },
+    {
+      "epoch": 0.5547546012269938,
+      "grad_norm": 0.17439760267734528,
+      "learning_rate": 6.320432759273266e-05,
+      "loss": 1.1806,
+      "step": 3617
+    },
+    {
+      "epoch": 0.5549079754601227,
+      "grad_norm": 0.1612442582845688,
+      "learning_rate": 6.31682814148645e-05,
+      "loss": 0.9689,
+      "step": 3618
+    },
+    {
+      "epoch": 0.5550613496932515,
+      "grad_norm": 0.16931049525737762,
+      "learning_rate": 6.313223803955738e-05,
+      "loss": 1.2426,
+      "step": 3619
+    },
+    {
+      "epoch": 0.5552147239263804,
+      "grad_norm": 0.1663127839565277,
+      "learning_rate": 6.309619747534877e-05,
+      "loss": 1.0877,
+      "step": 3620
+    },
+    {
+      "epoch": 0.5553680981595092,
+      "grad_norm": 0.16798798739910126,
+      "learning_rate": 6.306015973077559e-05,
+      "loss": 1.0618,
+      "step": 3621
+    },
+    {
+      "epoch": 0.555521472392638,
+      "grad_norm": 0.16924402117729187,
+      "learning_rate": 6.302412481437402e-05,
+      "loss": 1.017,
+      "step": 3622
+    },
+    {
+      "epoch": 0.5556748466257668,
+      "grad_norm": 0.16516603529453278,
+      "learning_rate": 6.298809273467961e-05,
+      "loss": 1.0906,
+      "step": 3623
+    },
+    {
+      "epoch": 0.5558282208588957,
+      "grad_norm": 0.16970479488372803,
+      "learning_rate": 6.295206350022723e-05,
+      "loss": 1.0129,
+      "step": 3624
+    },
+    {
+      "epoch": 0.5559815950920245,
+      "grad_norm": 0.1709173321723938,
+      "learning_rate": 6.291603711955106e-05,
+      "loss": 1.0391,
+      "step": 3625
+    },
+    {
+      "epoch": 0.5561349693251534,
+      "grad_norm": 0.19226308166980743,
+      "learning_rate": 6.288001360118463e-05,
+      "loss": 1.1811,
+      "step": 3626
+    },
+    {
+      "epoch": 0.5562883435582822,
+      "grad_norm": 0.1552196890115738,
+      "learning_rate": 6.284399295366075e-05,
+      "loss": 1.0962,
+      "step": 3627
+    },
+    {
+      "epoch": 0.556441717791411,
+      "grad_norm": 0.18280701339244843,
+      "learning_rate": 6.280797518551158e-05,
+      "loss": 1.1456,
+      "step": 3628
+    },
+    {
+      "epoch": 0.5565950920245398,
+      "grad_norm": 0.15923529863357544,
+      "learning_rate": 6.27719603052686e-05,
+      "loss": 1.0155,
+      "step": 3629
+    },
+    {
+      "epoch": 0.5567484662576687,
+      "grad_norm": 0.1712086796760559,
+      "learning_rate": 6.273594832146259e-05,
+      "loss": 1.1521,
+      "step": 3630
+    },
+    {
+      "epoch": 0.5569018404907975,
+      "grad_norm": 0.15881407260894775,
+      "learning_rate": 6.269993924262368e-05,
+      "loss": 1.1254,
+      "step": 3631
+    },
+    {
+      "epoch": 0.5570552147239264,
+      "grad_norm": 0.16038641333580017,
+      "learning_rate": 6.266393307728126e-05,
+      "loss": 1.1195,
+      "step": 3632
+    },
+    {
+      "epoch": 0.5572085889570552,
+      "grad_norm": 0.17119790613651276,
+      "learning_rate": 6.262792983396405e-05,
+      "loss": 1.1385,
+      "step": 3633
+    },
+    {
+      "epoch": 0.557361963190184,
+      "grad_norm": 0.15316584706306458,
+      "learning_rate": 6.259192952120013e-05,
+      "loss": 0.9676,
+      "step": 3634
+    },
+    {
+      "epoch": 0.5575153374233128,
+      "grad_norm": 0.17309799790382385,
+      "learning_rate": 6.255593214751678e-05,
+      "loss": 1.1542,
+      "step": 3635
+    },
+    {
+      "epoch": 0.5576687116564417,
+      "grad_norm": 0.16148614883422852,
+      "learning_rate": 6.251993772144066e-05,
+      "loss": 1.0684,
+      "step": 3636
+    },
+    {
+      "epoch": 0.5578220858895706,
+      "grad_norm": 0.16583286225795746,
+      "learning_rate": 6.248394625149774e-05,
+      "loss": 1.0991,
+      "step": 3637
+    },
+    {
+      "epoch": 0.5579754601226994,
+      "grad_norm": 0.1631803810596466,
+      "learning_rate": 6.244795774621322e-05,
+      "loss": 1.1257,
+      "step": 3638
+    },
+    {
+      "epoch": 0.5581288343558283,
+      "grad_norm": 0.1540846824645996,
+      "learning_rate": 6.241197221411169e-05,
+      "loss": 0.9971,
+      "step": 3639
+    },
+    {
+      "epoch": 0.558282208588957,
+      "grad_norm": 0.1661279797554016,
+      "learning_rate": 6.237598966371694e-05,
+      "loss": 1.1528,
+      "step": 3640
+    },
+    {
+      "epoch": 0.5584355828220859,
+      "grad_norm": 0.15717022120952606,
+      "learning_rate": 6.234001010355215e-05,
+      "loss": 1.0187,
+      "step": 3641
+    },
+    {
+      "epoch": 0.5585889570552147,
+      "grad_norm": 0.16186368465423584,
+      "learning_rate": 6.230403354213973e-05,
+      "loss": 1.0722,
+      "step": 3642
+    },
+    {
+      "epoch": 0.5587423312883436,
+      "grad_norm": 0.16851234436035156,
+      "learning_rate": 6.226805998800136e-05,
+      "loss": 1.1287,
+      "step": 3643
+    },
+    {
+      "epoch": 0.5588957055214724,
+      "grad_norm": 0.15491002798080444,
+      "learning_rate": 6.223208944965807e-05,
+      "loss": 1.0362,
+      "step": 3644
+    },
+    {
+      "epoch": 0.5590490797546013,
+      "grad_norm": 0.15516987442970276,
+      "learning_rate": 6.219612193563014e-05,
+      "loss": 0.9986,
+      "step": 3645
+    },
+    {
+      "epoch": 0.55920245398773,
+      "grad_norm": 0.16270320117473602,
+      "learning_rate": 6.216015745443713e-05,
+      "loss": 1.0222,
+      "step": 3646
+    },
+    {
+      "epoch": 0.5593558282208589,
+      "grad_norm": 0.16396905481815338,
+      "learning_rate": 6.212419601459791e-05,
+      "loss": 1.074,
+      "step": 3647
+    },
+    {
+      "epoch": 0.5595092024539877,
+      "grad_norm": 0.16450683772563934,
+      "learning_rate": 6.208823762463059e-05,
+      "loss": 1.0386,
+      "step": 3648
+    },
+    {
+      "epoch": 0.5596625766871166,
+      "grad_norm": 0.16309912502765656,
+      "learning_rate": 6.205228229305262e-05,
+      "loss": 1.1212,
+      "step": 3649
+    },
+    {
+      "epoch": 0.5598159509202454,
+      "grad_norm": 0.1526799499988556,
+      "learning_rate": 6.201633002838062e-05,
+      "loss": 1.0182,
+      "step": 3650
+    },
+    {
+      "epoch": 0.5599693251533743,
+      "grad_norm": 0.1496737152338028,
+      "learning_rate": 6.198038083913059e-05,
+      "loss": 1.0285,
+      "step": 3651
+    },
+    {
+      "epoch": 0.560122699386503,
+      "grad_norm": 0.15974776446819305,
+      "learning_rate": 6.194443473381775e-05,
+      "loss": 1.0043,
+      "step": 3652
+    },
+    {
+      "epoch": 0.5602760736196319,
+      "grad_norm": 0.15510636568069458,
+      "learning_rate": 6.190849172095661e-05,
+      "loss": 1.0186,
+      "step": 3653
+    },
+    {
+      "epoch": 0.5604294478527607,
+      "grad_norm": 0.16349685192108154,
+      "learning_rate": 6.18725518090609e-05,
+      "loss": 1.0637,
+      "step": 3654
+    },
+    {
+      "epoch": 0.5605828220858896,
+      "grad_norm": 0.16266795992851257,
+      "learning_rate": 6.18366150066437e-05,
+      "loss": 1.0579,
+      "step": 3655
+    },
+    {
+      "epoch": 0.5607361963190184,
+      "grad_norm": 0.1754283308982849,
+      "learning_rate": 6.180068132221728e-05,
+      "loss": 1.189,
+      "step": 3656
+    },
+    {
+      "epoch": 0.5608895705521473,
+      "grad_norm": 0.16057947278022766,
+      "learning_rate": 6.17647507642932e-05,
+      "loss": 1.1077,
+      "step": 3657
+    },
+    {
+      "epoch": 0.561042944785276,
+      "grad_norm": 0.1557060331106186,
+      "learning_rate": 6.172882334138227e-05,
+      "loss": 1.0166,
+      "step": 3658
+    },
+    {
+      "epoch": 0.5611963190184049,
+      "grad_norm": 0.15987737476825714,
+      "learning_rate": 6.169289906199456e-05,
+      "loss": 1.0649,
+      "step": 3659
+    },
+    {
+      "epoch": 0.5613496932515337,
+      "grad_norm": 0.17160408198833466,
+      "learning_rate": 6.165697793463943e-05,
+      "loss": 1.1151,
+      "step": 3660
+    },
+    {
+      "epoch": 0.5615030674846626,
+      "grad_norm": 0.13948149979114532,
+      "learning_rate": 6.162105996782545e-05,
+      "loss": 0.8711,
+      "step": 3661
+    },
+    {
+      "epoch": 0.5616564417177914,
+      "grad_norm": 0.16325750946998596,
+      "learning_rate": 6.158514517006042e-05,
+      "loss": 1.122,
+      "step": 3662
+    },
+    {
+      "epoch": 0.5618098159509203,
+      "grad_norm": 0.16990768909454346,
+      "learning_rate": 6.154923354985146e-05,
+      "loss": 1.1758,
+      "step": 3663
+    },
+    {
+      "epoch": 0.561963190184049,
+      "grad_norm": 0.13591697812080383,
+      "learning_rate": 6.151332511570487e-05,
+      "loss": 0.7704,
+      "step": 3664
+    },
+    {
+      "epoch": 0.5621165644171779,
+      "grad_norm": 0.16672992706298828,
+      "learning_rate": 6.147741987612625e-05,
+      "loss": 1.1041,
+      "step": 3665
+    },
+    {
+      "epoch": 0.5622699386503067,
+      "grad_norm": 0.16135475039482117,
+      "learning_rate": 6.144151783962041e-05,
+      "loss": 1.0689,
+      "step": 3666
+    },
+    {
+      "epoch": 0.5624233128834356,
+      "grad_norm": 0.16738636791706085,
+      "learning_rate": 6.140561901469142e-05,
+      "loss": 1.0712,
+      "step": 3667
+    },
+    {
+      "epoch": 0.5625766871165644,
+      "grad_norm": 0.15032322704792023,
+      "learning_rate": 6.136972340984257e-05,
+      "loss": 0.986,
+      "step": 3668
+    },
+    {
+      "epoch": 0.5627300613496933,
+      "grad_norm": 0.1742803305387497,
+      "learning_rate": 6.13338310335764e-05,
+      "loss": 1.1984,
+      "step": 3669
+    },
+    {
+      "epoch": 0.5628834355828221,
+      "grad_norm": 0.16349761188030243,
+      "learning_rate": 6.129794189439468e-05,
+      "loss": 1.135,
+      "step": 3670
+    },
+    {
+      "epoch": 0.563036809815951,
+      "grad_norm": 0.1564071774482727,
+      "learning_rate": 6.126205600079841e-05,
+      "loss": 1.0187,
+      "step": 3671
+    },
+    {
+      "epoch": 0.5631901840490797,
+      "grad_norm": 0.15495312213897705,
+      "learning_rate": 6.122617336128784e-05,
+      "loss": 0.9119,
+      "step": 3672
+    },
+    {
+      "epoch": 0.5633435582822086,
+      "grad_norm": 0.16303059458732605,
+      "learning_rate": 6.119029398436241e-05,
+      "loss": 1.031,
+      "step": 3673
+    },
+    {
+      "epoch": 0.5634969325153374,
+      "grad_norm": 0.16641974449157715,
+      "learning_rate": 6.115441787852084e-05,
+      "loss": 1.0443,
+      "step": 3674
+    },
+    {
+      "epoch": 0.5636503067484663,
+      "grad_norm": 0.1529271900653839,
+      "learning_rate": 6.111854505226105e-05,
+      "loss": 1.0383,
+      "step": 3675
+    },
+    {
+      "epoch": 0.5638036809815951,
+      "grad_norm": 0.16431909799575806,
+      "learning_rate": 6.108267551408015e-05,
+      "loss": 1.1202,
+      "step": 3676
+    },
+    {
+      "epoch": 0.563957055214724,
+      "grad_norm": 0.16649247705936432,
+      "learning_rate": 6.104680927247455e-05,
+      "loss": 1.0111,
+      "step": 3677
+    },
+    {
+      "epoch": 0.5641104294478527,
+      "grad_norm": 0.14434537291526794,
+      "learning_rate": 6.101094633593978e-05,
+      "loss": 0.8835,
+      "step": 3678
+    },
+    {
+      "epoch": 0.5642638036809816,
+      "grad_norm": 0.1479896456003189,
+      "learning_rate": 6.097508671297067e-05,
+      "loss": 0.9742,
+      "step": 3679
+    },
+    {
+      "epoch": 0.5644171779141104,
+      "grad_norm": 0.16065913438796997,
+      "learning_rate": 6.093923041206123e-05,
+      "loss": 1.0835,
+      "step": 3680
+    },
+    {
+      "epoch": 0.5645705521472393,
+      "grad_norm": 0.15337218344211578,
+      "learning_rate": 6.090337744170468e-05,
+      "loss": 1.0626,
+      "step": 3681
+    },
+    {
+      "epoch": 0.5647239263803681,
+      "grad_norm": 0.1466309279203415,
+      "learning_rate": 6.0867527810393446e-05,
+      "loss": 0.8486,
+      "step": 3682
+    },
+    {
+      "epoch": 0.564877300613497,
+      "grad_norm": 0.1558169424533844,
+      "learning_rate": 6.083168152661921e-05,
+      "loss": 0.9429,
+      "step": 3683
+    },
+    {
+      "epoch": 0.5650306748466257,
+      "grad_norm": 0.1686062216758728,
+      "learning_rate": 6.0795838598872814e-05,
+      "loss": 1.0674,
+      "step": 3684
+    },
+    {
+      "epoch": 0.5651840490797546,
+      "grad_norm": 0.15899786353111267,
+      "learning_rate": 6.07599990356443e-05,
+      "loss": 1.0562,
+      "step": 3685
+    },
+    {
+      "epoch": 0.5653374233128834,
+      "grad_norm": 0.15008363127708435,
+      "learning_rate": 6.0724162845422934e-05,
+      "loss": 0.9703,
+      "step": 3686
+    },
+    {
+      "epoch": 0.5654907975460123,
+      "grad_norm": 0.18717341125011444,
+      "learning_rate": 6.068833003669718e-05,
+      "loss": 1.3273,
+      "step": 3687
+    },
+    {
+      "epoch": 0.5656441717791411,
+      "grad_norm": 0.14495129883289337,
+      "learning_rate": 6.065250061795471e-05,
+      "loss": 1.0114,
+      "step": 3688
+    },
+    {
+      "epoch": 0.56579754601227,
+      "grad_norm": 0.15488483011722565,
+      "learning_rate": 6.0616674597682386e-05,
+      "loss": 0.9788,
+      "step": 3689
+    },
+    {
+      "epoch": 0.5659509202453987,
+      "grad_norm": 0.1543501764535904,
+      "learning_rate": 6.0580851984366255e-05,
+      "loss": 0.9394,
+      "step": 3690
+    },
+    {
+      "epoch": 0.5661042944785276,
+      "grad_norm": 0.1525564342737198,
+      "learning_rate": 6.054503278649156e-05,
+      "loss": 1.0311,
+      "step": 3691
+    },
+    {
+      "epoch": 0.5662576687116564,
+      "grad_norm": 0.16630974411964417,
+      "learning_rate": 6.050921701254276e-05,
+      "loss": 1.1929,
+      "step": 3692
+    },
+    {
+      "epoch": 0.5664110429447853,
+      "grad_norm": 0.15991638600826263,
+      "learning_rate": 6.0473404671003455e-05,
+      "loss": 1.12,
+      "step": 3693
+    },
+    {
+      "epoch": 0.5665644171779141,
+      "grad_norm": 0.1749444156885147,
+      "learning_rate": 6.043759577035647e-05,
+      "loss": 1.156,
+      "step": 3694
+    },
+    {
+      "epoch": 0.566717791411043,
+      "grad_norm": 0.16473659873008728,
+      "learning_rate": 6.0401790319083804e-05,
+      "loss": 1.0381,
+      "step": 3695
+    },
+    {
+      "epoch": 0.5668711656441717,
+      "grad_norm": 0.16702233254909515,
+      "learning_rate": 6.036598832566662e-05,
+      "loss": 1.144,
+      "step": 3696
+    },
+    {
+      "epoch": 0.5670245398773006,
+      "grad_norm": 0.1547979861497879,
+      "learning_rate": 6.033018979858533e-05,
+      "loss": 0.997,
+      "step": 3697
+    },
+    {
+      "epoch": 0.5671779141104294,
+      "grad_norm": 0.15923479199409485,
+      "learning_rate": 6.0294394746319436e-05,
+      "loss": 0.9908,
+      "step": 3698
+    },
+    {
+      "epoch": 0.5673312883435583,
+      "grad_norm": 0.16092449426651,
+      "learning_rate": 6.025860317734767e-05,
+      "loss": 1.058,
+      "step": 3699
+    },
+    {
+      "epoch": 0.5674846625766872,
+      "grad_norm": 2.213817834854126,
+      "learning_rate": 6.0222815100147934e-05,
+      "loss": 1.1134,
+      "step": 3700
+    },
+    {
+      "epoch": 0.567638036809816,
+      "grad_norm": 0.16098032891750336,
+      "learning_rate": 6.0187030523197276e-05,
+      "loss": 1.1485,
+      "step": 3701
+    },
+    {
+      "epoch": 0.5677914110429448,
+      "grad_norm": 0.16785117983818054,
+      "learning_rate": 6.0151249454971945e-05,
+      "loss": 1.1331,
+      "step": 3702
+    },
+    {
+      "epoch": 0.5679447852760736,
+      "grad_norm": 0.1522005945444107,
+      "learning_rate": 6.0115471903947345e-05,
+      "loss": 1.0063,
+      "step": 3703
+    },
+    {
+      "epoch": 0.5680981595092025,
+      "grad_norm": 0.16794054210186005,
+      "learning_rate": 6.0079697878598076e-05,
+      "loss": 1.1142,
+      "step": 3704
+    },
+    {
+      "epoch": 0.5682515337423313,
+      "grad_norm": 0.1679549217224121,
+      "learning_rate": 6.004392738739782e-05,
+      "loss": 1.0184,
+      "step": 3705
+    },
+    {
+      "epoch": 0.5684049079754602,
+      "grad_norm": 0.16524876654148102,
+      "learning_rate": 6.000816043881952e-05,
+      "loss": 1.1299,
+      "step": 3706
+    },
+    {
+      "epoch": 0.568558282208589,
+      "grad_norm": 0.1753331571817398,
+      "learning_rate": 5.9972397041335235e-05,
+      "loss": 1.1238,
+      "step": 3707
+    },
+    {
+      "epoch": 0.5687116564417178,
+      "grad_norm": 0.161179780960083,
+      "learning_rate": 5.993663720341616e-05,
+      "loss": 1.0895,
+      "step": 3708
+    },
+    {
+      "epoch": 0.5688650306748466,
+      "grad_norm": 0.17398221790790558,
+      "learning_rate": 5.9900880933532704e-05,
+      "loss": 1.1325,
+      "step": 3709
+    },
+    {
+      "epoch": 0.5690184049079755,
+      "grad_norm": 0.1657515913248062,
+      "learning_rate": 5.9865128240154387e-05,
+      "loss": 1.1578,
+      "step": 3710
+    },
+    {
+      "epoch": 0.5691717791411043,
+      "grad_norm": 0.17717880010604858,
+      "learning_rate": 5.9829379131749886e-05,
+      "loss": 1.0906,
+      "step": 3711
+    },
+    {
+      "epoch": 0.5693251533742332,
+      "grad_norm": 0.16475020349025726,
+      "learning_rate": 5.979363361678707e-05,
+      "loss": 1.0895,
+      "step": 3712
+    },
+    {
+      "epoch": 0.569478527607362,
+      "grad_norm": 0.1666116714477539,
+      "learning_rate": 5.975789170373289e-05,
+      "loss": 1.1187,
+      "step": 3713
+    },
+    {
+      "epoch": 0.5696319018404908,
+      "grad_norm": 0.16622595489025116,
+      "learning_rate": 5.972215340105348e-05,
+      "loss": 1.1597,
+      "step": 3714
+    },
+    {
+      "epoch": 0.5697852760736196,
+      "grad_norm": 0.16865120828151703,
+      "learning_rate": 5.9686418717214124e-05,
+      "loss": 1.2047,
+      "step": 3715
+    },
+    {
+      "epoch": 0.5699386503067485,
+      "grad_norm": 0.15252836048603058,
+      "learning_rate": 5.965068766067925e-05,
+      "loss": 1.1437,
+      "step": 3716
+    },
+    {
+      "epoch": 0.5700920245398773,
+      "grad_norm": 0.1539345234632492,
+      "learning_rate": 5.96149602399124e-05,
+      "loss": 1.0059,
+      "step": 3717
+    },
+    {
+      "epoch": 0.5702453987730062,
+      "grad_norm": 0.17709121108055115,
+      "learning_rate": 5.957923646337629e-05,
+      "loss": 1.1365,
+      "step": 3718
+    },
+    {
+      "epoch": 0.570398773006135,
+      "grad_norm": 0.166261687874794,
+      "learning_rate": 5.954351633953277e-05,
+      "loss": 1.0245,
+      "step": 3719
+    },
+    {
+      "epoch": 0.5705521472392638,
+      "grad_norm": 0.15864305198192596,
+      "learning_rate": 5.950779987684278e-05,
+      "loss": 1.0232,
+      "step": 3720
+    },
+    {
+      "epoch": 0.5707055214723926,
+      "grad_norm": 0.1587650328874588,
+      "learning_rate": 5.947208708376643e-05,
+      "loss": 1.1716,
+      "step": 3721
+    },
+    {
+      "epoch": 0.5708588957055215,
+      "grad_norm": 0.16697479784488678,
+      "learning_rate": 5.9436377968762955e-05,
+      "loss": 1.0894,
+      "step": 3722
+    },
+    {
+      "epoch": 0.5710122699386503,
+      "grad_norm": 0.17073458433151245,
+      "learning_rate": 5.9400672540290725e-05,
+      "loss": 1.0737,
+      "step": 3723
+    },
+    {
+      "epoch": 0.5711656441717792,
+      "grad_norm": 0.16165028512477875,
+      "learning_rate": 5.936497080680723e-05,
+      "loss": 1.1298,
+      "step": 3724
+    },
+    {
+      "epoch": 0.571319018404908,
+      "grad_norm": 0.17224034667015076,
+      "learning_rate": 5.932927277676909e-05,
+      "loss": 1.1848,
+      "step": 3725
+    },
+    {
+      "epoch": 0.5714723926380368,
+      "grad_norm": 0.16054213047027588,
+      "learning_rate": 5.9293578458632024e-05,
+      "loss": 0.9988,
+      "step": 3726
+    },
+    {
+      "epoch": 0.5716257668711656,
+      "grad_norm": 0.16606584191322327,
+      "learning_rate": 5.925788786085092e-05,
+      "loss": 1.1091,
+      "step": 3727
+    },
+    {
+      "epoch": 0.5717791411042945,
+      "grad_norm": 0.16085775196552277,
+      "learning_rate": 5.922220099187972e-05,
+      "loss": 1.0793,
+      "step": 3728
+    },
+    {
+      "epoch": 0.5719325153374233,
+      "grad_norm": 0.15575139224529266,
+      "learning_rate": 5.918651786017154e-05,
+      "loss": 0.9803,
+      "step": 3729
+    },
+    {
+      "epoch": 0.5720858895705522,
+      "grad_norm": 0.1702861487865448,
+      "learning_rate": 5.915083847417858e-05,
+      "loss": 1.1131,
+      "step": 3730
+    },
+    {
+      "epoch": 0.572239263803681,
+      "grad_norm": 0.16304805874824524,
+      "learning_rate": 5.9115162842352164e-05,
+      "loss": 1.115,
+      "step": 3731
+    },
+    {
+      "epoch": 0.5723926380368098,
+      "grad_norm": 0.16027241945266724,
+      "learning_rate": 5.9079490973142724e-05,
+      "loss": 1.0237,
+      "step": 3732
+    },
+    {
+      "epoch": 0.5725460122699386,
+      "grad_norm": 0.1514197438955307,
+      "learning_rate": 5.9043822874999806e-05,
+      "loss": 0.9972,
+      "step": 3733
+    },
+    {
+      "epoch": 0.5726993865030675,
+      "grad_norm": 0.1438630074262619,
+      "learning_rate": 5.900815855637206e-05,
+      "loss": 1.0292,
+      "step": 3734
+    },
+    {
+      "epoch": 0.5728527607361963,
+      "grad_norm": 0.1669236123561859,
+      "learning_rate": 5.8972498025707226e-05,
+      "loss": 1.1201,
+      "step": 3735
+    },
+    {
+      "epoch": 0.5730061349693252,
+      "grad_norm": 0.15993891656398773,
+      "learning_rate": 5.893684129145216e-05,
+      "loss": 0.9938,
+      "step": 3736
+    },
+    {
+      "epoch": 0.573159509202454,
+      "grad_norm": 0.1659070998430252,
+      "learning_rate": 5.8901188362052814e-05,
+      "loss": 1.1891,
+      "step": 3737
+    },
+    {
+      "epoch": 0.5733128834355828,
+      "grad_norm": 0.14042247831821442,
+      "learning_rate": 5.8865539245954266e-05,
+      "loss": 0.8713,
+      "step": 3738
+    },
+    {
+      "epoch": 0.5734662576687116,
+      "grad_norm": 0.1598520427942276,
+      "learning_rate": 5.882989395160063e-05,
+      "loss": 1.039,
+      "step": 3739
+    },
+    {
+      "epoch": 0.5736196319018405,
+      "grad_norm": 0.4619072675704956,
+      "learning_rate": 5.879425248743518e-05,
+      "loss": 1.0751,
+      "step": 3740
+    },
+    {
+      "epoch": 0.5737730061349693,
+      "grad_norm": 0.16262522339820862,
+      "learning_rate": 5.875861486190027e-05,
+      "loss": 1.1282,
+      "step": 3741
+    },
+    {
+      "epoch": 0.5739263803680982,
+      "grad_norm": 0.16728949546813965,
+      "learning_rate": 5.8722981083437294e-05,
+      "loss": 1.2112,
+      "step": 3742
+    },
+    {
+      "epoch": 0.574079754601227,
+      "grad_norm": 0.16615043580532074,
+      "learning_rate": 5.8687351160486786e-05,
+      "loss": 1.0357,
+      "step": 3743
+    },
+    {
+      "epoch": 0.5742331288343558,
+      "grad_norm": 0.1537671685218811,
+      "learning_rate": 5.865172510148836e-05,
+      "loss": 1.0719,
+      "step": 3744
+    },
+    {
+      "epoch": 0.5743865030674846,
+      "grad_norm": 0.16694007813930511,
+      "learning_rate": 5.8616102914880694e-05,
+      "loss": 1.0849,
+      "step": 3745
+    },
+    {
+      "epoch": 0.5745398773006135,
+      "grad_norm": 0.16307784616947174,
+      "learning_rate": 5.858048460910157e-05,
+      "loss": 1.1671,
+      "step": 3746
+    },
+    {
+      "epoch": 0.5746932515337423,
+      "grad_norm": 0.18139229714870453,
+      "learning_rate": 5.854487019258782e-05,
+      "loss": 1.1338,
+      "step": 3747
+    },
+    {
+      "epoch": 0.5748466257668712,
+      "grad_norm": 0.15427757799625397,
+      "learning_rate": 5.8509259673775394e-05,
+      "loss": 1.003,
+      "step": 3748
+    },
+    {
+      "epoch": 0.575,
+      "grad_norm": 0.16606976091861725,
+      "learning_rate": 5.84736530610993e-05,
+      "loss": 1.0565,
+      "step": 3749
+    },
+    {
+      "epoch": 0.5751533742331288,
+      "grad_norm": 0.45251569151878357,
+      "learning_rate": 5.8438050362993613e-05,
+      "loss": 1.0769,
+      "step": 3750
+    },
+    {
+      "epoch": 0.5753067484662576,
+      "grad_norm": 0.166673943400383,
+      "learning_rate": 5.84024515878915e-05,
+      "loss": 1.1558,
+      "step": 3751
+    },
+    {
+      "epoch": 0.5754601226993865,
+      "grad_norm": 0.17040245234966278,
+      "learning_rate": 5.836685674422518e-05,
+      "loss": 1.0937,
+      "step": 3752
+    },
+    {
+      "epoch": 0.5756134969325153,
+      "grad_norm": 0.157133549451828,
+      "learning_rate": 5.8331265840425954e-05,
+      "loss": 0.9912,
+      "step": 3753
+    },
+    {
+      "epoch": 0.5757668711656442,
+      "grad_norm": 0.2430478185415268,
+      "learning_rate": 5.8295678884924205e-05,
+      "loss": 1.1314,
+      "step": 3754
+    },
+    {
+      "epoch": 0.575920245398773,
+      "grad_norm": 0.1656164526939392,
+      "learning_rate": 5.826009588614931e-05,
+      "loss": 0.9865,
+      "step": 3755
+    },
+    {
+      "epoch": 0.5760736196319018,
+      "grad_norm": 0.1569017916917801,
+      "learning_rate": 5.822451685252979e-05,
+      "loss": 1.0115,
+      "step": 3756
+    },
+    {
+      "epoch": 0.5762269938650306,
+      "grad_norm": 0.1571589857339859,
+      "learning_rate": 5.818894179249321e-05,
+      "loss": 1.0591,
+      "step": 3757
+    },
+    {
+      "epoch": 0.5763803680981595,
+      "grad_norm": 0.1796732395887375,
+      "learning_rate": 5.8153370714466154e-05,
+      "loss": 1.242,
+      "step": 3758
+    },
+    {
+      "epoch": 0.5765337423312883,
+      "grad_norm": 0.1790691465139389,
+      "learning_rate": 5.81178036268743e-05,
+      "loss": 1.1818,
+      "step": 3759
+    },
+    {
+      "epoch": 0.5766871165644172,
+      "grad_norm": 0.160001739859581,
+      "learning_rate": 5.808224053814236e-05,
+      "loss": 1.0112,
+      "step": 3760
+    },
+    {
+      "epoch": 0.576840490797546,
+      "grad_norm": 0.1620199829339981,
+      "learning_rate": 5.8046681456694126e-05,
+      "loss": 1.0118,
+      "step": 3761
+    },
+    {
+      "epoch": 0.5769938650306748,
+      "grad_norm": 0.16317425668239594,
+      "learning_rate": 5.801112639095242e-05,
+      "loss": 1.2331,
+      "step": 3762
+    },
+    {
+      "epoch": 0.5771472392638037,
+      "grad_norm": 0.1599256694316864,
+      "learning_rate": 5.79755753493391e-05,
+      "loss": 1.0202,
+      "step": 3763
+    },
+    {
+      "epoch": 0.5773006134969325,
+      "grad_norm": 0.16799144446849823,
+      "learning_rate": 5.794002834027509e-05,
+      "loss": 1.2237,
+      "step": 3764
+    },
+    {
+      "epoch": 0.5774539877300614,
+      "grad_norm": 0.16814011335372925,
+      "learning_rate": 5.790448537218037e-05,
+      "loss": 1.1824,
+      "step": 3765
+    },
+    {
+      "epoch": 0.5776073619631902,
+      "grad_norm": 0.15885727107524872,
+      "learning_rate": 5.786894645347393e-05,
+      "loss": 1.0217,
+      "step": 3766
+    },
+    {
+      "epoch": 0.5777607361963191,
+      "grad_norm": 0.1591104418039322,
+      "learning_rate": 5.783341159257383e-05,
+      "loss": 1.1333,
+      "step": 3767
+    },
+    {
+      "epoch": 0.5779141104294478,
+      "grad_norm": 0.14788179099559784,
+      "learning_rate": 5.779788079789716e-05,
+      "loss": 1.0102,
+      "step": 3768
+    },
+    {
+      "epoch": 0.5780674846625767,
+      "grad_norm": 0.16045820713043213,
+      "learning_rate": 5.7762354077860057e-05,
+      "loss": 1.1485,
+      "step": 3769
+    },
+    {
+      "epoch": 0.5782208588957055,
+      "grad_norm": 0.16793256998062134,
+      "learning_rate": 5.772683144087764e-05,
+      "loss": 1.1251,
+      "step": 3770
+    },
+    {
+      "epoch": 0.5783742331288344,
+      "grad_norm": 8.444486618041992,
+      "learning_rate": 5.769131289536414e-05,
+      "loss": 1.017,
+      "step": 3771
+    },
+    {
+      "epoch": 0.5785276073619632,
+      "grad_norm": 0.16264086961746216,
+      "learning_rate": 5.765579844973275e-05,
+      "loss": 0.9076,
+      "step": 3772
+    },
+    {
+      "epoch": 0.5786809815950921,
+      "grad_norm": 0.16539834439754486,
+      "learning_rate": 5.762028811239575e-05,
+      "loss": 1.1419,
+      "step": 3773
+    },
+    {
+      "epoch": 0.5788343558282208,
+      "grad_norm": 0.16343916952610016,
+      "learning_rate": 5.758478189176441e-05,
+      "loss": 0.9748,
+      "step": 3774
+    },
+    {
+      "epoch": 0.5789877300613497,
+      "grad_norm": 0.14553923904895782,
+      "learning_rate": 5.7549279796249016e-05,
+      "loss": 0.9243,
+      "step": 3775
+    },
+    {
+      "epoch": 0.5791411042944785,
+      "grad_norm": 0.16341689229011536,
+      "learning_rate": 5.7513781834258924e-05,
+      "loss": 1.0952,
+      "step": 3776
+    },
+    {
+      "epoch": 0.5792944785276074,
+      "grad_norm": 0.1455143839120865,
+      "learning_rate": 5.747828801420248e-05,
+      "loss": 1.043,
+      "step": 3777
+    },
+    {
+      "epoch": 0.5794478527607362,
+      "grad_norm": 0.15395206212997437,
+      "learning_rate": 5.744279834448702e-05,
+      "loss": 1.0099,
+      "step": 3778
+    },
+    {
+      "epoch": 0.5796012269938651,
+      "grad_norm": 0.16091716289520264,
+      "learning_rate": 5.740731283351895e-05,
+      "loss": 1.1757,
+      "step": 3779
+    },
+    {
+      "epoch": 0.5797546012269938,
+      "grad_norm": 0.16973790526390076,
+      "learning_rate": 5.7371831489703665e-05,
+      "loss": 1.0972,
+      "step": 3780
+    },
+    {
+      "epoch": 0.5799079754601227,
+      "grad_norm": 0.1664891093969345,
+      "learning_rate": 5.733635432144556e-05,
+      "loss": 1.1008,
+      "step": 3781
+    },
+    {
+      "epoch": 0.5800613496932515,
+      "grad_norm": 0.14447227120399475,
+      "learning_rate": 5.730088133714809e-05,
+      "loss": 0.9337,
+      "step": 3782
+    },
+    {
+      "epoch": 0.5802147239263804,
+      "grad_norm": 0.1648135483264923,
+      "learning_rate": 5.726541254521366e-05,
+      "loss": 1.2272,
+      "step": 3783
+    },
+    {
+      "epoch": 0.5803680981595092,
+      "grad_norm": 0.4565576910972595,
+      "learning_rate": 5.722994795404372e-05,
+      "loss": 1.081,
+      "step": 3784
+    },
+    {
+      "epoch": 0.5805214723926381,
+      "grad_norm": 0.16919881105422974,
+      "learning_rate": 5.719448757203873e-05,
+      "loss": 1.0864,
+      "step": 3785
+    },
+    {
+      "epoch": 0.5806748466257668,
+      "grad_norm": 0.16920538246631622,
+      "learning_rate": 5.71590314075981e-05,
+      "loss": 1.163,
+      "step": 3786
+    },
+    {
+      "epoch": 0.5808282208588957,
+      "grad_norm": 0.16719959676265717,
+      "learning_rate": 5.71235794691203e-05,
+      "loss": 1.0602,
+      "step": 3787
+    },
+    {
+      "epoch": 0.5809815950920245,
+      "grad_norm": 0.15594466030597687,
+      "learning_rate": 5.708813176500277e-05,
+      "loss": 1.0472,
+      "step": 3788
+    },
+    {
+      "epoch": 0.5811349693251534,
+      "grad_norm": 0.15677832067012787,
+      "learning_rate": 5.705268830364197e-05,
+      "loss": 1.054,
+      "step": 3789
+    },
+    {
+      "epoch": 0.5812883435582822,
+      "grad_norm": 0.1587744653224945,
+      "learning_rate": 5.7017249093433305e-05,
+      "loss": 1.013,
+      "step": 3790
+    },
+    {
+      "epoch": 0.5814417177914111,
+      "grad_norm": 0.16133522987365723,
+      "learning_rate": 5.698181414277123e-05,
+      "loss": 1.1522,
+      "step": 3791
+    },
+    {
+      "epoch": 0.5815950920245399,
+      "grad_norm": 0.15945760905742645,
+      "learning_rate": 5.6946383460049155e-05,
+      "loss": 1.1355,
+      "step": 3792
+    },
+    {
+      "epoch": 0.5817484662576687,
+      "grad_norm": 0.21529027819633484,
+      "learning_rate": 5.69109570536595e-05,
+      "loss": 1.0161,
+      "step": 3793
+    },
+    {
+      "epoch": 0.5819018404907975,
+      "grad_norm": 0.16811174154281616,
+      "learning_rate": 5.687553493199365e-05,
+      "loss": 1.1488,
+      "step": 3794
+    },
+    {
+      "epoch": 0.5820552147239264,
+      "grad_norm": 0.15788760781288147,
+      "learning_rate": 5.684011710344201e-05,
+      "loss": 0.9934,
+      "step": 3795
+    },
+    {
+      "epoch": 0.5822085889570552,
+      "grad_norm": 0.15974211692810059,
+      "learning_rate": 5.680470357639394e-05,
+      "loss": 0.9175,
+      "step": 3796
+    },
+    {
+      "epoch": 0.5823619631901841,
+      "grad_norm": 0.16767363250255585,
+      "learning_rate": 5.676929435923779e-05,
+      "loss": 1.205,
+      "step": 3797
+    },
+    {
+      "epoch": 0.5825153374233129,
+      "grad_norm": 0.16370126605033875,
+      "learning_rate": 5.673388946036087e-05,
+      "loss": 1.2066,
+      "step": 3798
+    },
+    {
+      "epoch": 0.5826687116564417,
+      "grad_norm": 0.16948892176151276,
+      "learning_rate": 5.66984888881495e-05,
+      "loss": 1.1438,
+      "step": 3799
+    },
+    {
+      "epoch": 0.5828220858895705,
+      "grad_norm": 0.1655070036649704,
+      "learning_rate": 5.666309265098894e-05,
+      "loss": 1.1633,
+      "step": 3800
+    },
+    {
+      "epoch": 0.5829754601226994,
+      "grad_norm": 0.14617659151554108,
+      "learning_rate": 5.662770075726347e-05,
+      "loss": 0.979,
+      "step": 3801
+    },
+    {
+      "epoch": 0.5831288343558282,
+      "grad_norm": 0.1556042730808258,
+      "learning_rate": 5.659231321535629e-05,
+      "loss": 1.0743,
+      "step": 3802
+    },
+    {
+      "epoch": 0.5832822085889571,
+      "grad_norm": 0.1647399216890335,
+      "learning_rate": 5.6556930033649606e-05,
+      "loss": 1.067,
+      "step": 3803
+    },
+    {
+      "epoch": 0.5834355828220859,
+      "grad_norm": 0.17167364060878754,
+      "learning_rate": 5.652155122052458e-05,
+      "loss": 1.0288,
+      "step": 3804
+    },
+    {
+      "epoch": 0.5835889570552147,
+      "grad_norm": 0.17160798609256744,
+      "learning_rate": 5.648617678436133e-05,
+      "loss": 1.1534,
+      "step": 3805
+    },
+    {
+      "epoch": 0.5837423312883435,
+      "grad_norm": 0.1520906239748001,
+      "learning_rate": 5.645080673353894e-05,
+      "loss": 0.9682,
+      "step": 3806
+    },
+    {
+      "epoch": 0.5838957055214724,
+      "grad_norm": 0.15805594623088837,
+      "learning_rate": 5.641544107643547e-05,
+      "loss": 1.1279,
+      "step": 3807
+    },
+    {
+      "epoch": 0.5840490797546012,
+      "grad_norm": 0.15761947631835938,
+      "learning_rate": 5.6380079821427915e-05,
+      "loss": 1.0508,
+      "step": 3808
+    },
+    {
+      "epoch": 0.5842024539877301,
+      "grad_norm": 0.16327354311943054,
+      "learning_rate": 5.6344722976892256e-05,
+      "loss": 1.0185,
+      "step": 3809
+    },
+    {
+      "epoch": 0.5843558282208589,
+      "grad_norm": 0.1776847392320633,
+      "learning_rate": 5.63093705512034e-05,
+      "loss": 1.2319,
+      "step": 3810
+    },
+    {
+      "epoch": 0.5845092024539877,
+      "grad_norm": 0.154203861951828,
+      "learning_rate": 5.627402255273525e-05,
+      "loss": 1.0538,
+      "step": 3811
+    },
+    {
+      "epoch": 0.5846625766871165,
+      "grad_norm": 0.1553698480129242,
+      "learning_rate": 5.623867898986061e-05,
+      "loss": 0.9997,
+      "step": 3812
+    },
+    {
+      "epoch": 0.5848159509202454,
+      "grad_norm": 0.177860826253891,
+      "learning_rate": 5.620333987095125e-05,
+      "loss": 1.1588,
+      "step": 3813
+    },
+    {
+      "epoch": 0.5849693251533742,
+      "grad_norm": 0.163923442363739,
+      "learning_rate": 5.6168005204377905e-05,
+      "loss": 1.0532,
+      "step": 3814
+    },
+    {
+      "epoch": 0.5851226993865031,
+      "grad_norm": 0.16161547601222992,
+      "learning_rate": 5.613267499851022e-05,
+      "loss": 1.0491,
+      "step": 3815
+    },
+    {
+      "epoch": 0.5852760736196319,
+      "grad_norm": 0.15714149177074432,
+      "learning_rate": 5.6097349261716846e-05,
+      "loss": 1.0396,
+      "step": 3816
+    },
+    {
+      "epoch": 0.5854294478527607,
+      "grad_norm": 0.16964268684387207,
+      "learning_rate": 5.606202800236531e-05,
+      "loss": 1.2907,
+      "step": 3817
+    },
+    {
+      "epoch": 0.5855828220858895,
+      "grad_norm": 0.1599869430065155,
+      "learning_rate": 5.6026711228822116e-05,
+      "loss": 1.0963,
+      "step": 3818
+    },
+    {
+      "epoch": 0.5857361963190184,
+      "grad_norm": 0.15354548394680023,
+      "learning_rate": 5.5991398949452677e-05,
+      "loss": 1.1379,
+      "step": 3819
+    },
+    {
+      "epoch": 0.5858895705521472,
+      "grad_norm": 0.15147291123867035,
+      "learning_rate": 5.5956091172621384e-05,
+      "loss": 0.9499,
+      "step": 3820
+    },
+    {
+      "epoch": 0.5860429447852761,
+      "grad_norm": 0.1538514345884323,
+      "learning_rate": 5.5920787906691524e-05,
+      "loss": 0.9173,
+      "step": 3821
+    },
+    {
+      "epoch": 0.5861963190184049,
+      "grad_norm": 0.16150332987308502,
+      "learning_rate": 5.58854891600253e-05,
+      "loss": 1.0462,
+      "step": 3822
+    },
+    {
+      "epoch": 0.5863496932515337,
+      "grad_norm": 0.1741340160369873,
+      "learning_rate": 5.585019494098391e-05,
+      "loss": 1.1343,
+      "step": 3823
+    },
+    {
+      "epoch": 0.5865030674846625,
+      "grad_norm": 0.15470580756664276,
+      "learning_rate": 5.5814905257927414e-05,
+      "loss": 1.1074,
+      "step": 3824
+    },
+    {
+      "epoch": 0.5866564417177914,
+      "grad_norm": 0.1521306186914444,
+      "learning_rate": 5.577962011921485e-05,
+      "loss": 0.9939,
+      "step": 3825
+    },
+    {
+      "epoch": 0.5868098159509203,
+      "grad_norm": 0.17932985723018646,
+      "learning_rate": 5.5744339533204144e-05,
+      "loss": 1.2615,
+      "step": 3826
+    },
+    {
+      "epoch": 0.5869631901840491,
+      "grad_norm": 1.9783152341842651,
+      "learning_rate": 5.5709063508252156e-05,
+      "loss": 1.1468,
+      "step": 3827
+    },
+    {
+      "epoch": 0.587116564417178,
+      "grad_norm": 0.15449807047843933,
+      "learning_rate": 5.567379205271466e-05,
+      "loss": 1.0375,
+      "step": 3828
+    },
+    {
+      "epoch": 0.5872699386503067,
+      "grad_norm": 0.1449441760778427,
+      "learning_rate": 5.563852517494635e-05,
+      "loss": 0.8949,
+      "step": 3829
+    },
+    {
+      "epoch": 0.5874233128834356,
+      "grad_norm": 0.16825325787067413,
+      "learning_rate": 5.5603262883300844e-05,
+      "loss": 1.1385,
+      "step": 3830
+    },
+    {
+      "epoch": 0.5875766871165644,
+      "grad_norm": 0.17017288506031036,
+      "learning_rate": 5.5568005186130666e-05,
+      "loss": 1.1843,
+      "step": 3831
+    },
+    {
+      "epoch": 0.5877300613496933,
+      "grad_norm": 0.17107781767845154,
+      "learning_rate": 5.553275209178726e-05,
+      "loss": 1.1653,
+      "step": 3832
+    },
+    {
+      "epoch": 0.5878834355828221,
+      "grad_norm": 0.15557049214839935,
+      "learning_rate": 5.5497503608620945e-05,
+      "loss": 1.046,
+      "step": 3833
+    },
+    {
+      "epoch": 0.588036809815951,
+      "grad_norm": 0.16129997372627258,
+      "learning_rate": 5.5462259744981e-05,
+      "loss": 1.1346,
+      "step": 3834
+    },
+    {
+      "epoch": 0.5881901840490797,
+      "grad_norm": 0.15197572112083435,
+      "learning_rate": 5.542702050921557e-05,
+      "loss": 1.0636,
+      "step": 3835
+    },
+    {
+      "epoch": 0.5883435582822086,
+      "grad_norm": 0.17665384709835052,
+      "learning_rate": 5.539178590967173e-05,
+      "loss": 1.1765,
+      "step": 3836
+    },
+    {
+      "epoch": 0.5884969325153374,
+      "grad_norm": 0.1638471782207489,
+      "learning_rate": 5.535655595469544e-05,
+      "loss": 1.0707,
+      "step": 3837
+    },
+    {
+      "epoch": 0.5886503067484663,
+      "grad_norm": 0.16900871694087982,
+      "learning_rate": 5.5321330652631565e-05,
+      "loss": 1.1038,
+      "step": 3838
+    },
+    {
+      "epoch": 0.5888036809815951,
+      "grad_norm": 0.1483481228351593,
+      "learning_rate": 5.528611001182388e-05,
+      "loss": 0.9633,
+      "step": 3839
+    },
+    {
+      "epoch": 0.588957055214724,
+      "grad_norm": 0.14949189126491547,
+      "learning_rate": 5.525089404061503e-05,
+      "loss": 1.0002,
+      "step": 3840
+    },
+    {
+      "epoch": 0.5891104294478527,
+      "grad_norm": 0.1645728051662445,
+      "learning_rate": 5.5215682747346566e-05,
+      "loss": 0.9739,
+      "step": 3841
+    },
+    {
+      "epoch": 0.5892638036809816,
+      "grad_norm": 0.17217572033405304,
+      "learning_rate": 5.5180476140358935e-05,
+      "loss": 1.2277,
+      "step": 3842
+    },
+    {
+      "epoch": 0.5894171779141104,
+      "grad_norm": 0.1647120714187622,
+      "learning_rate": 5.514527422799148e-05,
+      "loss": 1.2107,
+      "step": 3843
+    },
+    {
+      "epoch": 0.5895705521472393,
+      "grad_norm": 0.15840600430965424,
+      "learning_rate": 5.5110077018582415e-05,
+      "loss": 1.0176,
+      "step": 3844
+    },
+    {
+      "epoch": 0.5897239263803681,
+      "grad_norm": 0.16716374456882477,
+      "learning_rate": 5.5074884520468854e-05,
+      "loss": 1.1475,
+      "step": 3845
+    },
+    {
+      "epoch": 0.589877300613497,
+      "grad_norm": 0.16216494143009186,
+      "learning_rate": 5.503969674198679e-05,
+      "loss": 1.036,
+      "step": 3846
+    },
+    {
+      "epoch": 0.5900306748466257,
+      "grad_norm": 0.16636112332344055,
+      "learning_rate": 5.500451369147112e-05,
+      "loss": 1.1013,
+      "step": 3847
+    },
+    {
+      "epoch": 0.5901840490797546,
+      "grad_norm": 0.14755800366401672,
+      "learning_rate": 5.496933537725556e-05,
+      "loss": 0.9735,
+      "step": 3848
+    },
+    {
+      "epoch": 0.5903374233128834,
+      "grad_norm": 0.15014997124671936,
+      "learning_rate": 5.493416180767274e-05,
+      "loss": 0.9539,
+      "step": 3849
+    },
+    {
+      "epoch": 0.5904907975460123,
+      "grad_norm": 0.16543090343475342,
+      "learning_rate": 5.489899299105421e-05,
+      "loss": 1.2358,
+      "step": 3850
+    },
+    {
+      "epoch": 0.5906441717791411,
+      "grad_norm": 0.1841915398836136,
+      "learning_rate": 5.4863828935730326e-05,
+      "loss": 1.1879,
+      "step": 3851
+    },
+    {
+      "epoch": 0.59079754601227,
+      "grad_norm": 0.16208407282829285,
+      "learning_rate": 5.482866965003035e-05,
+      "loss": 1.031,
+      "step": 3852
+    },
+    {
+      "epoch": 0.5909509202453987,
+      "grad_norm": 0.16590836644172668,
+      "learning_rate": 5.479351514228242e-05,
+      "loss": 1.0973,
+      "step": 3853
+    },
+    {
+      "epoch": 0.5911042944785276,
+      "grad_norm": 0.16360202431678772,
+      "learning_rate": 5.47583654208135e-05,
+      "loss": 1.0022,
+      "step": 3854
+    },
+    {
+      "epoch": 0.5912576687116564,
+      "grad_norm": 0.16325893998146057,
+      "learning_rate": 5.47232204939495e-05,
+      "loss": 1.145,
+      "step": 3855
+    },
+    {
+      "epoch": 0.5914110429447853,
+      "grad_norm": 0.1610279530286789,
+      "learning_rate": 5.468808037001509e-05,
+      "loss": 1.0247,
+      "step": 3856
+    },
+    {
+      "epoch": 0.5915644171779141,
+      "grad_norm": 0.1664309799671173,
+      "learning_rate": 5.4652945057333874e-05,
+      "loss": 1.1044,
+      "step": 3857
+    },
+    {
+      "epoch": 0.591717791411043,
+      "grad_norm": 0.15159845352172852,
+      "learning_rate": 5.461781456422833e-05,
+      "loss": 1.0616,
+      "step": 3858
+    },
+    {
+      "epoch": 0.5918711656441717,
+      "grad_norm": 0.16440284252166748,
+      "learning_rate": 5.458268889901972e-05,
+      "loss": 1.0987,
+      "step": 3859
+    },
+    {
+      "epoch": 0.5920245398773006,
+      "grad_norm": 0.1563626378774643,
+      "learning_rate": 5.454756807002823e-05,
+      "loss": 1.0876,
+      "step": 3860
+    },
+    {
+      "epoch": 0.5921779141104294,
+      "grad_norm": 0.15570829808712006,
+      "learning_rate": 5.451245208557288e-05,
+      "loss": 1.1205,
+      "step": 3861
+    },
+    {
+      "epoch": 0.5923312883435583,
+      "grad_norm": 0.21625575423240662,
+      "learning_rate": 5.4477340953971545e-05,
+      "loss": 1.0517,
+      "step": 3862
+    },
+    {
+      "epoch": 0.5924846625766871,
+      "grad_norm": 0.17258931696414948,
+      "learning_rate": 5.4442234683540916e-05,
+      "loss": 1.1213,
+      "step": 3863
+    },
+    {
+      "epoch": 0.592638036809816,
+      "grad_norm": 0.16419757902622223,
+      "learning_rate": 5.4407133282596585e-05,
+      "loss": 1.1825,
+      "step": 3864
+    },
+    {
+      "epoch": 0.5927914110429447,
+      "grad_norm": 0.14991186559200287,
+      "learning_rate": 5.437203675945296e-05,
+      "loss": 1.118,
+      "step": 3865
+    },
+    {
+      "epoch": 0.5929447852760736,
+      "grad_norm": 0.15634749829769135,
+      "learning_rate": 5.433694512242325e-05,
+      "loss": 1.0059,
+      "step": 3866
+    },
+    {
+      "epoch": 0.5930981595092024,
+      "grad_norm": 0.17230486869812012,
+      "learning_rate": 5.4301858379819685e-05,
+      "loss": 1.2294,
+      "step": 3867
+    },
+    {
+      "epoch": 0.5932515337423313,
+      "grad_norm": 0.1619168221950531,
+      "learning_rate": 5.426677653995309e-05,
+      "loss": 1.1195,
+      "step": 3868
+    },
+    {
+      "epoch": 0.5934049079754601,
+      "grad_norm": 0.16316215693950653,
+      "learning_rate": 5.42316996111333e-05,
+      "loss": 1.1496,
+      "step": 3869
+    },
+    {
+      "epoch": 0.593558282208589,
+      "grad_norm": 0.16251909732818604,
+      "learning_rate": 5.419662760166892e-05,
+      "loss": 1.1611,
+      "step": 3870
+    },
+    {
+      "epoch": 0.5937116564417177,
+      "grad_norm": 0.14678719639778137,
+      "learning_rate": 5.4161560519867404e-05,
+      "loss": 1.0091,
+      "step": 3871
+    },
+    {
+      "epoch": 0.5938650306748466,
+      "grad_norm": 0.14761289954185486,
+      "learning_rate": 5.412649837403504e-05,
+      "loss": 1.042,
+      "step": 3872
+    },
+    {
+      "epoch": 0.5940184049079754,
+      "grad_norm": 0.15718349814414978,
+      "learning_rate": 5.409144117247693e-05,
+      "loss": 1.0396,
+      "step": 3873
+    },
+    {
+      "epoch": 0.5941717791411043,
+      "grad_norm": 0.1931951940059662,
+      "learning_rate": 5.405638892349705e-05,
+      "loss": 1.0601,
+      "step": 3874
+    },
+    {
+      "epoch": 0.5943251533742331,
+      "grad_norm": 0.16978251934051514,
+      "learning_rate": 5.402134163539815e-05,
+      "loss": 1.2238,
+      "step": 3875
+    },
+    {
+      "epoch": 0.594478527607362,
+      "grad_norm": 0.1526729315519333,
+      "learning_rate": 5.398629931648182e-05,
+      "loss": 1.0369,
+      "step": 3876
+    },
+    {
+      "epoch": 0.5946319018404908,
+      "grad_norm": 0.160925030708313,
+      "learning_rate": 5.395126197504849e-05,
+      "loss": 0.9857,
+      "step": 3877
+    },
+    {
+      "epoch": 0.5947852760736196,
+      "grad_norm": 0.16828401386737823,
+      "learning_rate": 5.391622961939742e-05,
+      "loss": 1.123,
+      "step": 3878
+    },
+    {
+      "epoch": 0.5949386503067484,
+      "grad_norm": 0.1618143767118454,
+      "learning_rate": 5.388120225782663e-05,
+      "loss": 1.0022,
+      "step": 3879
+    },
+    {
+      "epoch": 0.5950920245398773,
+      "grad_norm": 0.16502051055431366,
+      "learning_rate": 5.384617989863304e-05,
+      "loss": 1.0956,
+      "step": 3880
+    },
+    {
+      "epoch": 0.5952453987730061,
+      "grad_norm": 0.16438570618629456,
+      "learning_rate": 5.381116255011231e-05,
+      "loss": 1.1095,
+      "step": 3881
+    },
+    {
+      "epoch": 0.595398773006135,
+      "grad_norm": 0.15074849128723145,
+      "learning_rate": 5.377615022055898e-05,
+      "loss": 0.9567,
+      "step": 3882
+    },
+    {
+      "epoch": 0.5955521472392638,
+      "grad_norm": 0.1564747840166092,
+      "learning_rate": 5.3741142918266325e-05,
+      "loss": 0.9772,
+      "step": 3883
+    },
+    {
+      "epoch": 0.5957055214723926,
+      "grad_norm": 0.16980160772800446,
+      "learning_rate": 5.3706140651526495e-05,
+      "loss": 1.0787,
+      "step": 3884
+    },
+    {
+      "epoch": 0.5958588957055214,
+      "grad_norm": 0.14337258040905,
+      "learning_rate": 5.367114342863042e-05,
+      "loss": 0.8426,
+      "step": 3885
+    },
+    {
+      "epoch": 0.5960122699386503,
+      "grad_norm": 0.16472147405147552,
+      "learning_rate": 5.363615125786783e-05,
+      "loss": 1.082,
+      "step": 3886
+    },
+    {
+      "epoch": 0.5961656441717792,
+      "grad_norm": 0.16490456461906433,
+      "learning_rate": 5.360116414752728e-05,
+      "loss": 1.0267,
+      "step": 3887
+    },
+    {
+      "epoch": 0.596319018404908,
+      "grad_norm": 0.16613732278347015,
+      "learning_rate": 5.35661821058961e-05,
+      "loss": 1.2242,
+      "step": 3888
+    },
+    {
+      "epoch": 0.5964723926380369,
+      "grad_norm": 0.1617889106273651,
+      "learning_rate": 5.3531205141260444e-05,
+      "loss": 1.1046,
+      "step": 3889
+    },
+    {
+      "epoch": 0.5966257668711656,
+      "grad_norm": 0.14433616399765015,
+      "learning_rate": 5.3496233261905265e-05,
+      "loss": 1.0032,
+      "step": 3890
+    },
+    {
+      "epoch": 0.5967791411042945,
+      "grad_norm": 0.16089080274105072,
+      "learning_rate": 5.346126647611427e-05,
+      "loss": 0.9897,
+      "step": 3891
+    },
+    {
+      "epoch": 0.5969325153374233,
+      "grad_norm": 0.1409217268228531,
+      "learning_rate": 5.342630479216999e-05,
+      "loss": 0.8838,
+      "step": 3892
+    },
+    {
+      "epoch": 0.5970858895705522,
+      "grad_norm": 0.17226989567279816,
+      "learning_rate": 5.339134821835375e-05,
+      "loss": 1.1574,
+      "step": 3893
+    },
+    {
+      "epoch": 0.597239263803681,
+      "grad_norm": 0.1657126247882843,
+      "learning_rate": 5.3356396762945675e-05,
+      "loss": 1.0196,
+      "step": 3894
+    },
+    {
+      "epoch": 0.5973926380368099,
+      "grad_norm": 0.16235093772411346,
+      "learning_rate": 5.3321450434224655e-05,
+      "loss": 1.1243,
+      "step": 3895
+    },
+    {
+      "epoch": 0.5975460122699386,
+      "grad_norm": 0.16567087173461914,
+      "learning_rate": 5.3286509240468354e-05,
+      "loss": 1.024,
+      "step": 3896
+    },
+    {
+      "epoch": 0.5976993865030675,
+      "grad_norm": 0.1603376269340515,
+      "learning_rate": 5.325157318995328e-05,
+      "loss": 1.073,
+      "step": 3897
+    },
+    {
+      "epoch": 0.5978527607361963,
+      "grad_norm": 0.1636907011270523,
+      "learning_rate": 5.321664229095464e-05,
+      "loss": 1.1164,
+      "step": 3898
+    },
+    {
+      "epoch": 0.5980061349693252,
+      "grad_norm": 0.15526147186756134,
+      "learning_rate": 5.318171655174648e-05,
+      "loss": 0.988,
+      "step": 3899
+    },
+    {
+      "epoch": 0.598159509202454,
+      "grad_norm": 0.15556731820106506,
+      "learning_rate": 5.314679598060161e-05,
+      "loss": 1.0174,
+      "step": 3900
+    },
+    {
+      "epoch": 0.5983128834355829,
+      "grad_norm": 0.15255306661128998,
+      "learning_rate": 5.31118805857916e-05,
+      "loss": 0.9747,
+      "step": 3901
+    },
+    {
+      "epoch": 0.5984662576687116,
+      "grad_norm": 0.16824842989444733,
+      "learning_rate": 5.307697037558681e-05,
+      "loss": 1.1487,
+      "step": 3902
+    },
+    {
+      "epoch": 0.5986196319018405,
+      "grad_norm": 0.16928155720233917,
+      "learning_rate": 5.3042065358256376e-05,
+      "loss": 1.1244,
+      "step": 3903
+    },
+    {
+      "epoch": 0.5987730061349693,
+      "grad_norm": 0.16556628048419952,
+      "learning_rate": 5.300716554206819e-05,
+      "loss": 1.1662,
+      "step": 3904
+    },
+    {
+      "epoch": 0.5989263803680982,
+      "grad_norm": 0.14854003489017487,
+      "learning_rate": 5.2972270935288934e-05,
+      "loss": 1.011,
+      "step": 3905
+    },
+    {
+      "epoch": 0.599079754601227,
+      "grad_norm": 0.15456773340702057,
+      "learning_rate": 5.2937381546184e-05,
+      "loss": 0.9828,
+      "step": 3906
+    },
+    {
+      "epoch": 0.5992331288343559,
+      "grad_norm": 0.15888185799121857,
+      "learning_rate": 5.2902497383017615e-05,
+      "loss": 1.0441,
+      "step": 3907
+    },
+    {
+      "epoch": 0.5993865030674846,
+      "grad_norm": 0.15722249448299408,
+      "learning_rate": 5.2867618454052734e-05,
+      "loss": 1.0794,
+      "step": 3908
+    },
+    {
+      "epoch": 0.5995398773006135,
+      "grad_norm": 0.16912032663822174,
+      "learning_rate": 5.2832744767551055e-05,
+      "loss": 1.2357,
+      "step": 3909
+    },
+    {
+      "epoch": 0.5996932515337423,
+      "grad_norm": 0.15952743589878082,
+      "learning_rate": 5.279787633177309e-05,
+      "loss": 1.1209,
+      "step": 3910
+    },
+    {
+      "epoch": 0.5998466257668712,
+      "grad_norm": 0.15343554317951202,
+      "learning_rate": 5.2763013154978046e-05,
+      "loss": 1.0733,
+      "step": 3911
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.15807461738586426,
+      "learning_rate": 5.272815524542392e-05,
+      "loss": 1.0142,
+      "step": 3912
+    },
+    {
+      "epoch": 0.6001533742331289,
+      "grad_norm": 0.18484333157539368,
+      "learning_rate": 5.2693302611367454e-05,
+      "loss": 1.201,
+      "step": 3913
+    },
+    {
+      "epoch": 0.6003067484662576,
+      "grad_norm": 0.16329152882099152,
+      "learning_rate": 5.265845526106413e-05,
+      "loss": 1.0749,
+      "step": 3914
+    },
+    {
+      "epoch": 0.6004601226993865,
+      "grad_norm": 0.1550145000219345,
+      "learning_rate": 5.262361320276818e-05,
+      "loss": 1.1418,
+      "step": 3915
+    },
+    {
+      "epoch": 0.6006134969325153,
+      "grad_norm": 0.15800510346889496,
+      "learning_rate": 5.258877644473261e-05,
+      "loss": 1.1271,
+      "step": 3916
+    },
+    {
+      "epoch": 0.6007668711656442,
+      "grad_norm": 0.17786899209022522,
+      "learning_rate": 5.255394499520914e-05,
+      "loss": 1.0702,
+      "step": 3917
+    },
+    {
+      "epoch": 0.600920245398773,
+      "grad_norm": 0.1625654399394989,
+      "learning_rate": 5.2519118862448235e-05,
+      "loss": 1.0658,
+      "step": 3918
+    },
+    {
+      "epoch": 0.6010736196319019,
+      "grad_norm": 0.1635540872812271,
+      "learning_rate": 5.248429805469911e-05,
+      "loss": 1.1104,
+      "step": 3919
+    },
+    {
+      "epoch": 0.6012269938650306,
+      "grad_norm": 0.1569156050682068,
+      "learning_rate": 5.244948258020972e-05,
+      "loss": 1.1067,
+      "step": 3920
+    },
+    {
+      "epoch": 0.6013803680981595,
+      "grad_norm": 0.17162126302719116,
+      "learning_rate": 5.2414672447226743e-05,
+      "loss": 1.0462,
+      "step": 3921
+    },
+    {
+      "epoch": 0.6015337423312883,
+      "grad_norm": 0.17604993283748627,
+      "learning_rate": 5.237986766399562e-05,
+      "loss": 1.2196,
+      "step": 3922
+    },
+    {
+      "epoch": 0.6016871165644172,
+      "grad_norm": 0.16010592877864838,
+      "learning_rate": 5.234506823876048e-05,
+      "loss": 1.1462,
+      "step": 3923
+    },
+    {
+      "epoch": 0.601840490797546,
+      "grad_norm": 0.15078873932361603,
+      "learning_rate": 5.231027417976424e-05,
+      "loss": 1.0048,
+      "step": 3924
+    },
+    {
+      "epoch": 0.6019938650306749,
+      "grad_norm": 0.15796580910682678,
+      "learning_rate": 5.2275485495248506e-05,
+      "loss": 1.0576,
+      "step": 3925
+    },
+    {
+      "epoch": 0.6021472392638036,
+      "grad_norm": 0.14666616916656494,
+      "learning_rate": 5.2240702193453606e-05,
+      "loss": 0.9346,
+      "step": 3926
+    },
+    {
+      "epoch": 0.6023006134969325,
+      "grad_norm": 0.1644924134016037,
+      "learning_rate": 5.2205924282618614e-05,
+      "loss": 1.0413,
+      "step": 3927
+    },
+    {
+      "epoch": 0.6024539877300613,
+      "grad_norm": 0.15695805847644806,
+      "learning_rate": 5.2171151770981334e-05,
+      "loss": 1.0314,
+      "step": 3928
+    },
+    {
+      "epoch": 0.6026073619631902,
+      "grad_norm": 0.15922202169895172,
+      "learning_rate": 5.213638466677825e-05,
+      "loss": 1.0853,
+      "step": 3929
+    },
+    {
+      "epoch": 0.602760736196319,
+      "grad_norm": 0.15381553769111633,
+      "learning_rate": 5.2101622978244615e-05,
+      "loss": 1.0583,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6029141104294479,
+      "grad_norm": 0.1669517159461975,
+      "learning_rate": 5.206686671361438e-05,
+      "loss": 1.1435,
+      "step": 3931
+    },
+    {
+      "epoch": 0.6030674846625766,
+      "grad_norm": 0.16659021377563477,
+      "learning_rate": 5.203211588112021e-05,
+      "loss": 1.0567,
+      "step": 3932
+    },
+    {
+      "epoch": 0.6032208588957055,
+      "grad_norm": 0.16408196091651917,
+      "learning_rate": 5.1997370488993444e-05,
+      "loss": 1.0686,
+      "step": 3933
+    },
+    {
+      "epoch": 0.6033742331288343,
+      "grad_norm": 0.16988317668437958,
+      "learning_rate": 5.19626305454642e-05,
+      "loss": 1.1044,
+      "step": 3934
+    },
+    {
+      "epoch": 0.6035276073619632,
+      "grad_norm": 0.15843534469604492,
+      "learning_rate": 5.1927896058761275e-05,
+      "loss": 1.1137,
+      "step": 3935
+    },
+    {
+      "epoch": 0.603680981595092,
+      "grad_norm": 0.1483002007007599,
+      "learning_rate": 5.189316703711216e-05,
+      "loss": 0.9451,
+      "step": 3936
+    },
+    {
+      "epoch": 0.6038343558282209,
+      "grad_norm": 0.15308158099651337,
+      "learning_rate": 5.185844348874308e-05,
+      "loss": 1.0152,
+      "step": 3937
+    },
+    {
+      "epoch": 0.6039877300613496,
+      "grad_norm": 0.15336956083774567,
+      "learning_rate": 5.182372542187895e-05,
+      "loss": 0.9422,
+      "step": 3938
+    },
+    {
+      "epoch": 0.6041411042944785,
+      "grad_norm": 0.14870087802410126,
+      "learning_rate": 5.1789012844743364e-05,
+      "loss": 0.9347,
+      "step": 3939
+    },
+    {
+      "epoch": 0.6042944785276073,
+      "grad_norm": 0.16334392130374908,
+      "learning_rate": 5.175430576555867e-05,
+      "loss": 0.9412,
+      "step": 3940
+    },
+    {
+      "epoch": 0.6044478527607362,
+      "grad_norm": 0.1572083979845047,
+      "learning_rate": 5.171960419254585e-05,
+      "loss": 1.0153,
+      "step": 3941
+    },
+    {
+      "epoch": 0.604601226993865,
+      "grad_norm": 0.16756299138069153,
+      "learning_rate": 5.1684908133924615e-05,
+      "loss": 1.1173,
+      "step": 3942
+    },
+    {
+      "epoch": 0.6047546012269939,
+      "grad_norm": 0.1607980728149414,
+      "learning_rate": 5.165021759791339e-05,
+      "loss": 1.1499,
+      "step": 3943
+    },
+    {
+      "epoch": 0.6049079754601226,
+      "grad_norm": 0.16711091995239258,
+      "learning_rate": 5.161553259272924e-05,
+      "loss": 1.1258,
+      "step": 3944
+    },
+    {
+      "epoch": 0.6050613496932515,
+      "grad_norm": 0.1664416790008545,
+      "learning_rate": 5.158085312658799e-05,
+      "loss": 1.1524,
+      "step": 3945
+    },
+    {
+      "epoch": 0.6052147239263803,
+      "grad_norm": 0.1568172425031662,
+      "learning_rate": 5.154617920770405e-05,
+      "loss": 0.9856,
+      "step": 3946
+    },
+    {
+      "epoch": 0.6053680981595092,
+      "grad_norm": 0.15207922458648682,
+      "learning_rate": 5.151151084429064e-05,
+      "loss": 1.0283,
+      "step": 3947
+    },
+    {
+      "epoch": 0.605521472392638,
+      "grad_norm": 0.16035543382167816,
+      "learning_rate": 5.147684804455957e-05,
+      "loss": 1.1184,
+      "step": 3948
+    },
+    {
+      "epoch": 0.6056748466257669,
+      "grad_norm": 0.1567773073911667,
+      "learning_rate": 5.1442190816721365e-05,
+      "loss": 1.0653,
+      "step": 3949
+    },
+    {
+      "epoch": 0.6058282208588958,
+      "grad_norm": 0.1557001918554306,
+      "learning_rate": 5.140753916898523e-05,
+      "loss": 1.0889,
+      "step": 3950
+    },
+    {
+      "epoch": 0.6059815950920245,
+      "grad_norm": 0.19200055301189423,
+      "learning_rate": 5.137289310955902e-05,
+      "loss": 0.9004,
+      "step": 3951
+    },
+    {
+      "epoch": 0.6061349693251534,
+      "grad_norm": 0.15841273963451385,
+      "learning_rate": 5.133825264664937e-05,
+      "loss": 1.0863,
+      "step": 3952
+    },
+    {
+      "epoch": 0.6062883435582822,
+      "grad_norm": 0.17449960112571716,
+      "learning_rate": 5.130361778846147e-05,
+      "loss": 1.2163,
+      "step": 3953
+    },
+    {
+      "epoch": 0.6064417177914111,
+      "grad_norm": 0.188052698969841,
+      "learning_rate": 5.1268988543199214e-05,
+      "loss": 1.2692,
+      "step": 3954
+    },
+    {
+      "epoch": 0.6065950920245399,
+      "grad_norm": 0.16686880588531494,
+      "learning_rate": 5.123436491906517e-05,
+      "loss": 1.0827,
+      "step": 3955
+    },
+    {
+      "epoch": 0.6067484662576688,
+      "grad_norm": 0.16472604870796204,
+      "learning_rate": 5.1199746924260624e-05,
+      "loss": 1.0449,
+      "step": 3956
+    },
+    {
+      "epoch": 0.6069018404907975,
+      "grad_norm": 0.16700950264930725,
+      "learning_rate": 5.116513456698546e-05,
+      "loss": 1.1144,
+      "step": 3957
+    },
+    {
+      "epoch": 0.6070552147239264,
+      "grad_norm": 0.16153235733509064,
+      "learning_rate": 5.113052785543825e-05,
+      "loss": 1.0628,
+      "step": 3958
+    },
+    {
+      "epoch": 0.6072085889570552,
+      "grad_norm": 0.17771539092063904,
+      "learning_rate": 5.109592679781624e-05,
+      "loss": 1.1716,
+      "step": 3959
+    },
+    {
+      "epoch": 0.6073619631901841,
+      "grad_norm": 0.1562386453151703,
+      "learning_rate": 5.106133140231536e-05,
+      "loss": 1.1057,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6075153374233129,
+      "grad_norm": 0.1586359143257141,
+      "learning_rate": 5.102674167713011e-05,
+      "loss": 1.0788,
+      "step": 3961
+    },
+    {
+      "epoch": 0.6076687116564418,
+      "grad_norm": 0.16150328516960144,
+      "learning_rate": 5.099215763045373e-05,
+      "loss": 0.9741,
+      "step": 3962
+    },
+    {
+      "epoch": 0.6078220858895705,
+      "grad_norm": 0.1536196619272232,
+      "learning_rate": 5.09575792704781e-05,
+      "loss": 1.0256,
+      "step": 3963
+    },
+    {
+      "epoch": 0.6079754601226994,
+      "grad_norm": 0.1649237722158432,
+      "learning_rate": 5.0923006605393736e-05,
+      "loss": 1.2044,
+      "step": 3964
+    },
+    {
+      "epoch": 0.6081288343558282,
+      "grad_norm": 0.1627538949251175,
+      "learning_rate": 5.0888439643389813e-05,
+      "loss": 1.0911,
+      "step": 3965
+    },
+    {
+      "epoch": 0.6082822085889571,
+      "grad_norm": 0.16669350862503052,
+      "learning_rate": 5.085387839265415e-05,
+      "loss": 1.1227,
+      "step": 3966
+    },
+    {
+      "epoch": 0.6084355828220859,
+      "grad_norm": 0.17664358019828796,
+      "learning_rate": 5.081932286137324e-05,
+      "loss": 1.167,
+      "step": 3967
+    },
+    {
+      "epoch": 0.6085889570552148,
+      "grad_norm": 0.1761554777622223,
+      "learning_rate": 5.078477305773216e-05,
+      "loss": 1.1255,
+      "step": 3968
+    },
+    {
+      "epoch": 0.6087423312883435,
+      "grad_norm": 0.15784402191638947,
+      "learning_rate": 5.075022898991469e-05,
+      "loss": 1.1363,
+      "step": 3969
+    },
+    {
+      "epoch": 0.6088957055214724,
+      "grad_norm": 0.16434703767299652,
+      "learning_rate": 5.071569066610323e-05,
+      "loss": 1.1364,
+      "step": 3970
+    },
+    {
+      "epoch": 0.6090490797546012,
+      "grad_norm": 0.17500193417072296,
+      "learning_rate": 5.068115809447882e-05,
+      "loss": 1.1496,
+      "step": 3971
+    },
+    {
+      "epoch": 0.6092024539877301,
+      "grad_norm": 0.1555159091949463,
+      "learning_rate": 5.0646631283221135e-05,
+      "loss": 1.0611,
+      "step": 3972
+    },
+    {
+      "epoch": 0.6093558282208589,
+      "grad_norm": 0.17021189630031586,
+      "learning_rate": 5.061211024050848e-05,
+      "loss": 1.0498,
+      "step": 3973
+    },
+    {
+      "epoch": 0.6095092024539878,
+      "grad_norm": 0.15629823505878448,
+      "learning_rate": 5.057759497451782e-05,
+      "loss": 1.1078,
+      "step": 3974
+    },
+    {
+      "epoch": 0.6096625766871165,
+      "grad_norm": 0.17778685688972473,
+      "learning_rate": 5.054308549342473e-05,
+      "loss": 1.2515,
+      "step": 3975
+    },
+    {
+      "epoch": 0.6098159509202454,
+      "grad_norm": 0.1601831614971161,
+      "learning_rate": 5.05085818054034e-05,
+      "loss": 1.0474,
+      "step": 3976
+    },
+    {
+      "epoch": 0.6099693251533742,
+      "grad_norm": 0.1767289936542511,
+      "learning_rate": 5.047408391862667e-05,
+      "loss": 1.2759,
+      "step": 3977
+    },
+    {
+      "epoch": 0.6101226993865031,
+      "grad_norm": 0.17520849406719208,
+      "learning_rate": 5.0439591841266014e-05,
+      "loss": 1.3079,
+      "step": 3978
+    },
+    {
+      "epoch": 0.6102760736196319,
+      "grad_norm": 0.1650875061750412,
+      "learning_rate": 5.04051055814915e-05,
+      "loss": 1.053,
+      "step": 3979
+    },
+    {
+      "epoch": 0.6104294478527608,
+      "grad_norm": 0.1695176661014557,
+      "learning_rate": 5.037062514747187e-05,
+      "loss": 1.1412,
+      "step": 3980
+    },
+    {
+      "epoch": 0.6105828220858895,
+      "grad_norm": 0.17090082168579102,
+      "learning_rate": 5.033615054737441e-05,
+      "loss": 1.1365,
+      "step": 3981
+    },
+    {
+      "epoch": 0.6107361963190184,
+      "grad_norm": 0.1812809556722641,
+      "learning_rate": 5.030168178936509e-05,
+      "loss": 1.2529,
+      "step": 3982
+    },
+    {
+      "epoch": 0.6108895705521472,
+      "grad_norm": 0.16145825386047363,
+      "learning_rate": 5.0267218881608474e-05,
+      "loss": 1.1255,
+      "step": 3983
+    },
+    {
+      "epoch": 0.6110429447852761,
+      "grad_norm": 0.15409447252750397,
+      "learning_rate": 5.023276183226772e-05,
+      "loss": 1.0657,
+      "step": 3984
+    },
+    {
+      "epoch": 0.6111963190184049,
+      "grad_norm": 0.16209319233894348,
+      "learning_rate": 5.019831064950462e-05,
+      "loss": 1.1021,
+      "step": 3985
+    },
+    {
+      "epoch": 0.6113496932515338,
+      "grad_norm": 0.17209653556346893,
+      "learning_rate": 5.016386534147958e-05,
+      "loss": 1.2067,
+      "step": 3986
+    },
+    {
+      "epoch": 0.6115030674846625,
+      "grad_norm": 0.14058496057987213,
+      "learning_rate": 5.01294259163516e-05,
+      "loss": 0.9096,
+      "step": 3987
+    },
+    {
+      "epoch": 0.6116564417177914,
+      "grad_norm": 0.15729482471942902,
+      "learning_rate": 5.00949923822783e-05,
+      "loss": 1.0609,
+      "step": 3988
+    },
+    {
+      "epoch": 0.6118098159509202,
+      "grad_norm": 0.18334858119487762,
+      "learning_rate": 5.006056474741591e-05,
+      "loss": 1.1611,
+      "step": 3989
+    },
+    {
+      "epoch": 0.6119631901840491,
+      "grad_norm": 0.15676042437553406,
+      "learning_rate": 5.002614301991924e-05,
+      "loss": 1.0163,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6121165644171779,
+      "grad_norm": 0.1645875722169876,
+      "learning_rate": 4.999172720794171e-05,
+      "loss": 1.2024,
+      "step": 3991
+    },
+    {
+      "epoch": 0.6122699386503068,
+      "grad_norm": 0.15078315138816833,
+      "learning_rate": 4.9957317319635326e-05,
+      "loss": 1.0906,
+      "step": 3992
+    },
+    {
+      "epoch": 0.6124233128834355,
+      "grad_norm": 0.15845845639705658,
+      "learning_rate": 4.9922913363150724e-05,
+      "loss": 1.0822,
+      "step": 3993
+    },
+    {
+      "epoch": 0.6125766871165644,
+      "grad_norm": 0.15892338752746582,
+      "learning_rate": 4.988851534663708e-05,
+      "loss": 1.1006,
+      "step": 3994
+    },
+    {
+      "epoch": 0.6127300613496932,
+      "grad_norm": 0.14927585422992706,
+      "learning_rate": 4.9854123278242284e-05,
+      "loss": 0.9534,
+      "step": 3995
+    },
+    {
+      "epoch": 0.6128834355828221,
+      "grad_norm": 0.16095024347305298,
+      "learning_rate": 4.981973716611267e-05,
+      "loss": 1.0322,
+      "step": 3996
+    },
+    {
+      "epoch": 0.6130368098159509,
+      "grad_norm": 0.16583456099033356,
+      "learning_rate": 4.9785357018393214e-05,
+      "loss": 1.0776,
+      "step": 3997
+    },
+    {
+      "epoch": 0.6131901840490798,
+      "grad_norm": 0.21744617819786072,
+      "learning_rate": 4.9750982843227515e-05,
+      "loss": 1.2103,
+      "step": 3998
+    },
+    {
+      "epoch": 0.6133435582822085,
+      "grad_norm": 0.186132550239563,
+      "learning_rate": 4.971661464875773e-05,
+      "loss": 1.1169,
+      "step": 3999
+    },
+    {
+      "epoch": 0.6134969325153374,
+      "grad_norm": 0.17573612928390503,
+      "learning_rate": 4.9682252443124594e-05,
+      "loss": 1.1795,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6136503067484662,
+      "grad_norm": 0.17564579844474792,
+      "learning_rate": 4.964789623446743e-05,
+      "loss": 1.0563,
+      "step": 4001
+    },
+    {
+      "epoch": 0.6138036809815951,
+      "grad_norm": 0.1464127004146576,
+      "learning_rate": 4.961354603092416e-05,
+      "loss": 0.947,
+      "step": 4002
+    },
+    {
+      "epoch": 0.6139570552147239,
+      "grad_norm": 0.27968648076057434,
+      "learning_rate": 4.9579201840631225e-05,
+      "loss": 1.0274,
+      "step": 4003
+    },
+    {
+      "epoch": 0.6141104294478528,
+      "grad_norm": 0.14721627533435822,
+      "learning_rate": 4.95448636717237e-05,
+      "loss": 0.9574,
+      "step": 4004
+    },
+    {
+      "epoch": 0.6142638036809815,
+      "grad_norm": 0.14948388934135437,
+      "learning_rate": 4.951053153233522e-05,
+      "loss": 0.9658,
+      "step": 4005
+    },
+    {
+      "epoch": 0.6144171779141104,
+      "grad_norm": 0.14715680480003357,
+      "learning_rate": 4.947620543059798e-05,
+      "loss": 0.9191,
+      "step": 4006
+    },
+    {
+      "epoch": 0.6145705521472392,
+      "grad_norm": 0.1585419923067093,
+      "learning_rate": 4.9441885374642744e-05,
+      "loss": 1.0514,
+      "step": 4007
+    },
+    {
+      "epoch": 0.6147239263803681,
+      "grad_norm": 0.15230122208595276,
+      "learning_rate": 4.940757137259886e-05,
+      "loss": 0.93,
+      "step": 4008
+    },
+    {
+      "epoch": 0.6148773006134969,
+      "grad_norm": 0.16954533755779266,
+      "learning_rate": 4.9373263432594243e-05,
+      "loss": 1.1106,
+      "step": 4009
+    },
+    {
+      "epoch": 0.6150306748466258,
+      "grad_norm": 0.16469143331050873,
+      "learning_rate": 4.9338961562755364e-05,
+      "loss": 1.1349,
+      "step": 4010
+    },
+    {
+      "epoch": 0.6151840490797545,
+      "grad_norm": 0.1664043515920639,
+      "learning_rate": 4.930466577120722e-05,
+      "loss": 1.1117,
+      "step": 4011
+    },
+    {
+      "epoch": 0.6153374233128834,
+      "grad_norm": 0.1761433631181717,
+      "learning_rate": 4.927037606607343e-05,
+      "loss": 1.1181,
+      "step": 4012
+    },
+    {
+      "epoch": 0.6154907975460123,
+      "grad_norm": 0.15655045211315155,
+      "learning_rate": 4.923609245547613e-05,
+      "loss": 1.1492,
+      "step": 4013
+    },
+    {
+      "epoch": 0.6156441717791411,
+      "grad_norm": 0.15494222939014435,
+      "learning_rate": 4.9201814947536034e-05,
+      "loss": 0.9704,
+      "step": 4014
+    },
+    {
+      "epoch": 0.61579754601227,
+      "grad_norm": 0.1563873291015625,
+      "learning_rate": 4.916754355037241e-05,
+      "loss": 1.0633,
+      "step": 4015
+    },
+    {
+      "epoch": 0.6159509202453988,
+      "grad_norm": 0.15292173624038696,
+      "learning_rate": 4.913327827210306e-05,
+      "loss": 0.9953,
+      "step": 4016
+    },
+    {
+      "epoch": 0.6161042944785277,
+      "grad_norm": 0.16080833971500397,
+      "learning_rate": 4.909901912084435e-05,
+      "loss": 1.0883,
+      "step": 4017
+    },
+    {
+      "epoch": 0.6162576687116564,
+      "grad_norm": 0.14609433710575104,
+      "learning_rate": 4.90647661047112e-05,
+      "loss": 0.9561,
+      "step": 4018
+    },
+    {
+      "epoch": 0.6164110429447853,
+      "grad_norm": 0.16182322800159454,
+      "learning_rate": 4.9030519231817064e-05,
+      "loss": 1.0709,
+      "step": 4019
+    },
+    {
+      "epoch": 0.6165644171779141,
+      "grad_norm": 0.15286783874034882,
+      "learning_rate": 4.8996278510273935e-05,
+      "loss": 1.016,
+      "step": 4020
+    },
+    {
+      "epoch": 0.616717791411043,
+      "grad_norm": 0.18100109696388245,
+      "learning_rate": 4.896204394819238e-05,
+      "loss": 1.145,
+      "step": 4021
+    },
+    {
+      "epoch": 0.6168711656441718,
+      "grad_norm": 0.16162200272083282,
+      "learning_rate": 4.8927815553681474e-05,
+      "loss": 1.1183,
+      "step": 4022
+    },
+    {
+      "epoch": 0.6170245398773007,
+      "grad_norm": 0.15942735970020294,
+      "learning_rate": 4.8893593334848844e-05,
+      "loss": 1.0626,
+      "step": 4023
+    },
+    {
+      "epoch": 0.6171779141104294,
+      "grad_norm": 0.14344464242458344,
+      "learning_rate": 4.8859377299800656e-05,
+      "loss": 0.9194,
+      "step": 4024
+    },
+    {
+      "epoch": 0.6173312883435583,
+      "grad_norm": 0.1562381088733673,
+      "learning_rate": 4.882516745664163e-05,
+      "loss": 1.0231,
+      "step": 4025
+    },
+    {
+      "epoch": 0.6174846625766871,
+      "grad_norm": 0.1552782505750656,
+      "learning_rate": 4.8790963813474965e-05,
+      "loss": 1.04,
+      "step": 4026
+    },
+    {
+      "epoch": 0.617638036809816,
+      "grad_norm": 0.16723258793354034,
+      "learning_rate": 4.875676637840245e-05,
+      "loss": 1.0919,
+      "step": 4027
+    },
+    {
+      "epoch": 0.6177914110429448,
+      "grad_norm": 0.1605692356824875,
+      "learning_rate": 4.872257515952436e-05,
+      "loss": 1.0577,
+      "step": 4028
+    },
+    {
+      "epoch": 0.6179447852760737,
+      "grad_norm": 0.16720004379749298,
+      "learning_rate": 4.868839016493953e-05,
+      "loss": 1.1027,
+      "step": 4029
+    },
+    {
+      "epoch": 0.6180981595092024,
+      "grad_norm": 0.15089797973632812,
+      "learning_rate": 4.8654211402745304e-05,
+      "loss": 1.0746,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6182515337423313,
+      "grad_norm": 0.16284462809562683,
+      "learning_rate": 4.862003888103755e-05,
+      "loss": 1.2014,
+      "step": 4031
+    },
+    {
+      "epoch": 0.6184049079754601,
+      "grad_norm": 0.15542402863502502,
+      "learning_rate": 4.858587260791067e-05,
+      "loss": 0.9561,
+      "step": 4032
+    },
+    {
+      "epoch": 0.618558282208589,
+      "grad_norm": 0.18438683450222015,
+      "learning_rate": 4.855171259145758e-05,
+      "loss": 1.2668,
+      "step": 4033
+    },
+    {
+      "epoch": 0.6187116564417178,
+      "grad_norm": 0.1589433252811432,
+      "learning_rate": 4.851755883976969e-05,
+      "loss": 1.0552,
+      "step": 4034
+    },
+    {
+      "epoch": 0.6188650306748467,
+      "grad_norm": 0.17488019168376923,
+      "learning_rate": 4.8483411360936955e-05,
+      "loss": 1.0755,
+      "step": 4035
+    },
+    {
+      "epoch": 0.6190184049079754,
+      "grad_norm": 0.15297681093215942,
+      "learning_rate": 4.84492701630478e-05,
+      "loss": 1.0552,
+      "step": 4036
+    },
+    {
+      "epoch": 0.6191717791411043,
+      "grad_norm": 0.15701426565647125,
+      "learning_rate": 4.8415135254189294e-05,
+      "loss": 1.1159,
+      "step": 4037
+    },
+    {
+      "epoch": 0.6193251533742331,
+      "grad_norm": 0.15415164828300476,
+      "learning_rate": 4.838100664244684e-05,
+      "loss": 1.0994,
+      "step": 4038
+    },
+    {
+      "epoch": 0.619478527607362,
+      "grad_norm": 0.16221213340759277,
+      "learning_rate": 4.834688433590446e-05,
+      "loss": 1.0954,
+      "step": 4039
+    },
+    {
+      "epoch": 0.6196319018404908,
+      "grad_norm": 0.15984806418418884,
+      "learning_rate": 4.831276834264464e-05,
+      "loss": 1.028,
+      "step": 4040
+    },
+    {
+      "epoch": 0.6197852760736197,
+      "grad_norm": 0.155140683054924,
+      "learning_rate": 4.827865867074839e-05,
+      "loss": 1.065,
+      "step": 4041
+    },
+    {
+      "epoch": 0.6199386503067484,
+      "grad_norm": 0.1536492556333542,
+      "learning_rate": 4.8244555328295204e-05,
+      "loss": 1.079,
+      "step": 4042
+    },
+    {
+      "epoch": 0.6200920245398773,
+      "grad_norm": 0.15318217873573303,
+      "learning_rate": 4.821045832336311e-05,
+      "loss": 1.0579,
+      "step": 4043
+    },
+    {
+      "epoch": 0.6202453987730061,
+      "grad_norm": 0.1541278213262558,
+      "learning_rate": 4.81763676640286e-05,
+      "loss": 1.123,
+      "step": 4044
+    },
+    {
+      "epoch": 0.620398773006135,
+      "grad_norm": 0.16251613199710846,
+      "learning_rate": 4.814228335836669e-05,
+      "loss": 1.1904,
+      "step": 4045
+    },
+    {
+      "epoch": 0.6205521472392638,
+      "grad_norm": 0.14456701278686523,
+      "learning_rate": 4.8108205414450844e-05,
+      "loss": 0.9077,
+      "step": 4046
+    },
+    {
+      "epoch": 0.6207055214723927,
+      "grad_norm": 0.16376003623008728,
+      "learning_rate": 4.8074133840353074e-05,
+      "loss": 1.1498,
+      "step": 4047
+    },
+    {
+      "epoch": 0.6208588957055214,
+      "grad_norm": 0.1603405475616455,
+      "learning_rate": 4.804006864414387e-05,
+      "loss": 1.0863,
+      "step": 4048
+    },
+    {
+      "epoch": 0.6210122699386503,
+      "grad_norm": 0.18010610342025757,
+      "learning_rate": 4.800600983389218e-05,
+      "loss": 1.1585,
+      "step": 4049
+    },
+    {
+      "epoch": 0.6211656441717791,
+      "grad_norm": 0.15335537493228912,
+      "learning_rate": 4.797195741766548e-05,
+      "loss": 1.0299,
+      "step": 4050
+    },
+    {
+      "epoch": 0.621319018404908,
+      "grad_norm": 0.16920021176338196,
+      "learning_rate": 4.793791140352971e-05,
+      "loss": 1.2071,
+      "step": 4051
+    },
+    {
+      "epoch": 0.6214723926380368,
+      "grad_norm": 0.15667133033275604,
+      "learning_rate": 4.790387179954931e-05,
+      "loss": 1.0431,
+      "step": 4052
+    },
+    {
+      "epoch": 0.6216257668711657,
+      "grad_norm": 0.1535995751619339,
+      "learning_rate": 4.7869838613787156e-05,
+      "loss": 0.9509,
+      "step": 4053
+    },
+    {
+      "epoch": 0.6217791411042944,
+      "grad_norm": 0.17284314334392548,
+      "learning_rate": 4.7835811854304654e-05,
+      "loss": 1.0406,
+      "step": 4054
+    },
+    {
+      "epoch": 0.6219325153374233,
+      "grad_norm": 0.16701477766036987,
+      "learning_rate": 4.7801791529161685e-05,
+      "loss": 1.1658,
+      "step": 4055
+    },
+    {
+      "epoch": 0.6220858895705521,
+      "grad_norm": 0.1513189971446991,
+      "learning_rate": 4.776777764641655e-05,
+      "loss": 1.0124,
+      "step": 4056
+    },
+    {
+      "epoch": 0.622239263803681,
+      "grad_norm": 0.15735474228858948,
+      "learning_rate": 4.773377021412612e-05,
+      "loss": 1.0393,
+      "step": 4057
+    },
+    {
+      "epoch": 0.6223926380368098,
+      "grad_norm": 0.16500148177146912,
+      "learning_rate": 4.7699769240345635e-05,
+      "loss": 0.9996,
+      "step": 4058
+    },
+    {
+      "epoch": 0.6225460122699387,
+      "grad_norm": 0.16476504504680634,
+      "learning_rate": 4.766577473312888e-05,
+      "loss": 1.1328,
+      "step": 4059
+    },
+    {
+      "epoch": 0.6226993865030674,
+      "grad_norm": 0.17266681790351868,
+      "learning_rate": 4.763178670052809e-05,
+      "loss": 1.2245,
+      "step": 4060
+    },
+    {
+      "epoch": 0.6228527607361963,
+      "grad_norm": 0.1665343940258026,
+      "learning_rate": 4.759780515059393e-05,
+      "loss": 1.1321,
+      "step": 4061
+    },
+    {
+      "epoch": 0.6230061349693251,
+      "grad_norm": 0.15755121409893036,
+      "learning_rate": 4.756383009137558e-05,
+      "loss": 1.0759,
+      "step": 4062
+    },
+    {
+      "epoch": 0.623159509202454,
+      "grad_norm": 0.16117440164089203,
+      "learning_rate": 4.752986153092064e-05,
+      "loss": 1.1967,
+      "step": 4063
+    },
+    {
+      "epoch": 0.6233128834355828,
+      "grad_norm": 0.16191531717777252,
+      "learning_rate": 4.749589947727519e-05,
+      "loss": 1.1186,
+      "step": 4064
+    },
+    {
+      "epoch": 0.6234662576687117,
+      "grad_norm": 0.1579698622226715,
+      "learning_rate": 4.74619439384838e-05,
+      "loss": 1.0539,
+      "step": 4065
+    },
+    {
+      "epoch": 0.6236196319018404,
+      "grad_norm": 0.16369062662124634,
+      "learning_rate": 4.742799492258945e-05,
+      "loss": 1.046,
+      "step": 4066
+    },
+    {
+      "epoch": 0.6237730061349693,
+      "grad_norm": 0.15067362785339355,
+      "learning_rate": 4.739405243763357e-05,
+      "loss": 0.9068,
+      "step": 4067
+    },
+    {
+      "epoch": 0.6239263803680981,
+      "grad_norm": 0.1682489514350891,
+      "learning_rate": 4.736011649165611e-05,
+      "loss": 0.9886,
+      "step": 4068
+    },
+    {
+      "epoch": 0.624079754601227,
+      "grad_norm": 0.17243598401546478,
+      "learning_rate": 4.732618709269539e-05,
+      "loss": 1.1519,
+      "step": 4069
+    },
+    {
+      "epoch": 0.6242331288343558,
+      "grad_norm": 0.1424737423658371,
+      "learning_rate": 4.7292264248788216e-05,
+      "loss": 0.9831,
+      "step": 4070
+    },
+    {
+      "epoch": 0.6243865030674847,
+      "grad_norm": 0.15912824869155884,
+      "learning_rate": 4.725834796796985e-05,
+      "loss": 1.1292,
+      "step": 4071
+    },
+    {
+      "epoch": 0.6245398773006134,
+      "grad_norm": 0.16757068037986755,
+      "learning_rate": 4.722443825827397e-05,
+      "loss": 1.1645,
+      "step": 4072
+    },
+    {
+      "epoch": 0.6246932515337423,
+      "grad_norm": 0.1542944461107254,
+      "learning_rate": 4.7190535127732755e-05,
+      "loss": 1.0959,
+      "step": 4073
+    },
+    {
+      "epoch": 0.6248466257668711,
+      "grad_norm": 0.16689881682395935,
+      "learning_rate": 4.715663858437675e-05,
+      "loss": 1.1283,
+      "step": 4074
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 0.15449704229831696,
+      "learning_rate": 4.712274863623502e-05,
+      "loss": 1.0443,
+      "step": 4075
+    },
+    {
+      "epoch": 0.6251533742331289,
+      "grad_norm": 0.15791812539100647,
+      "learning_rate": 4.7088865291334976e-05,
+      "loss": 1.0395,
+      "step": 4076
+    },
+    {
+      "epoch": 0.6253067484662577,
+      "grad_norm": 0.172937273979187,
+      "learning_rate": 4.7054988557702535e-05,
+      "loss": 1.1377,
+      "step": 4077
+    },
+    {
+      "epoch": 0.6254601226993866,
+      "grad_norm": 0.14146703481674194,
+      "learning_rate": 4.702111844336203e-05,
+      "loss": 0.9545,
+      "step": 4078
+    },
+    {
+      "epoch": 0.6256134969325153,
+      "grad_norm": 0.18029752373695374,
+      "learning_rate": 4.698725495633619e-05,
+      "loss": 1.1637,
+      "step": 4079
+    },
+    {
+      "epoch": 0.6257668711656442,
+      "grad_norm": 0.15237957239151,
+      "learning_rate": 4.6953398104646284e-05,
+      "loss": 1.0998,
+      "step": 4080
+    },
+    {
+      "epoch": 0.625920245398773,
+      "grad_norm": 0.1501048505306244,
+      "learning_rate": 4.691954789631187e-05,
+      "loss": 1.0484,
+      "step": 4081
+    },
+    {
+      "epoch": 0.6260736196319019,
+      "grad_norm": 0.16016629338264465,
+      "learning_rate": 4.6885704339351006e-05,
+      "loss": 1.1515,
+      "step": 4082
+    },
+    {
+      "epoch": 0.6262269938650307,
+      "grad_norm": 0.15225543081760406,
+      "learning_rate": 4.685186744178018e-05,
+      "loss": 0.9792,
+      "step": 4083
+    },
+    {
+      "epoch": 0.6263803680981596,
+      "grad_norm": 0.16258083283901215,
+      "learning_rate": 4.681803721161426e-05,
+      "loss": 1.1732,
+      "step": 4084
+    },
+    {
+      "epoch": 0.6265337423312883,
+      "grad_norm": 0.152770534157753,
+      "learning_rate": 4.6784213656866584e-05,
+      "loss": 0.9272,
+      "step": 4085
+    },
+    {
+      "epoch": 0.6266871165644172,
+      "grad_norm": 0.13663353025913239,
+      "learning_rate": 4.675039678554888e-05,
+      "loss": 0.9512,
+      "step": 4086
+    },
+    {
+      "epoch": 0.626840490797546,
+      "grad_norm": 0.149224653840065,
+      "learning_rate": 4.671658660567129e-05,
+      "loss": 1.0243,
+      "step": 4087
+    },
+    {
+      "epoch": 0.6269938650306749,
+      "grad_norm": 0.1681971698999405,
+      "learning_rate": 4.6682783125242386e-05,
+      "loss": 1.1371,
+      "step": 4088
+    },
+    {
+      "epoch": 0.6271472392638037,
+      "grad_norm": 0.1784522831439972,
+      "learning_rate": 4.6648986352269146e-05,
+      "loss": 1.2133,
+      "step": 4089
+    },
+    {
+      "epoch": 0.6273006134969326,
+      "grad_norm": 0.1719326227903366,
+      "learning_rate": 4.661519629475696e-05,
+      "loss": 1.1845,
+      "step": 4090
+    },
+    {
+      "epoch": 0.6274539877300613,
+      "grad_norm": 0.15536406636238098,
+      "learning_rate": 4.6581412960709614e-05,
+      "loss": 1.106,
+      "step": 4091
+    },
+    {
+      "epoch": 0.6276073619631902,
+      "grad_norm": 0.1535634547472,
+      "learning_rate": 4.654763635812933e-05,
+      "loss": 1.0273,
+      "step": 4092
+    },
+    {
+      "epoch": 0.627760736196319,
+      "grad_norm": 0.17215478420257568,
+      "learning_rate": 4.651386649501671e-05,
+      "loss": 1.1612,
+      "step": 4093
+    },
+    {
+      "epoch": 0.6279141104294479,
+      "grad_norm": 0.1650753915309906,
+      "learning_rate": 4.648010337937079e-05,
+      "loss": 1.1381,
+      "step": 4094
+    },
+    {
+      "epoch": 0.6280674846625767,
+      "grad_norm": 0.16678518056869507,
+      "learning_rate": 4.6446347019188974e-05,
+      "loss": 1.1453,
+      "step": 4095
+    },
+    {
+      "epoch": 0.6282208588957056,
+      "grad_norm": 0.1460491567850113,
+      "learning_rate": 4.6412597422467076e-05,
+      "loss": 0.9835,
+      "step": 4096
+    },
+    {
+      "epoch": 0.6283742331288343,
+      "grad_norm": 0.15596385300159454,
+      "learning_rate": 4.63788545971993e-05,
+      "loss": 1.0116,
+      "step": 4097
+    },
+    {
+      "epoch": 0.6285276073619632,
+      "grad_norm": 0.15481920540332794,
+      "learning_rate": 4.6345118551378266e-05,
+      "loss": 1.0789,
+      "step": 4098
+    },
+    {
+      "epoch": 0.628680981595092,
+      "grad_norm": 0.16684682667255402,
+      "learning_rate": 4.631138929299501e-05,
+      "loss": 1.0918,
+      "step": 4099
+    },
+    {
+      "epoch": 0.6288343558282209,
+      "grad_norm": 0.15650717914104462,
+      "learning_rate": 4.627766683003888e-05,
+      "loss": 0.9821,
+      "step": 4100
+    },
+    {
+      "epoch": 0.6289877300613497,
+      "grad_norm": 0.16026943922042847,
+      "learning_rate": 4.624395117049768e-05,
+      "loss": 1.058,
+      "step": 4101
+    },
+    {
+      "epoch": 0.6291411042944786,
+      "grad_norm": 0.15863636136054993,
+      "learning_rate": 4.6210242322357614e-05,
+      "loss": 1.1623,
+      "step": 4102
+    },
+    {
+      "epoch": 0.6292944785276073,
+      "grad_norm": 0.15459735691547394,
+      "learning_rate": 4.6176540293603226e-05,
+      "loss": 1.0578,
+      "step": 4103
+    },
+    {
+      "epoch": 0.6294478527607362,
+      "grad_norm": 0.15733712911605835,
+      "learning_rate": 4.6142845092217464e-05,
+      "loss": 1.1368,
+      "step": 4104
+    },
+    {
+      "epoch": 0.629601226993865,
+      "grad_norm": 0.16239690780639648,
+      "learning_rate": 4.610915672618164e-05,
+      "loss": 1.0797,
+      "step": 4105
+    },
+    {
+      "epoch": 0.6297546012269939,
+      "grad_norm": 0.16430510580539703,
+      "learning_rate": 4.607547520347549e-05,
+      "loss": 1.1253,
+      "step": 4106
+    },
+    {
+      "epoch": 0.6299079754601227,
+      "grad_norm": 0.15031488239765167,
+      "learning_rate": 4.604180053207709e-05,
+      "loss": 0.9486,
+      "step": 4107
+    },
+    {
+      "epoch": 0.6300613496932516,
+      "grad_norm": 0.1635013371706009,
+      "learning_rate": 4.600813271996292e-05,
+      "loss": 1.1191,
+      "step": 4108
+    },
+    {
+      "epoch": 0.6302147239263803,
+      "grad_norm": 0.1573510617017746,
+      "learning_rate": 4.597447177510782e-05,
+      "loss": 1.084,
+      "step": 4109
+    },
+    {
+      "epoch": 0.6303680981595092,
+      "grad_norm": 0.1738271266222,
+      "learning_rate": 4.594081770548502e-05,
+      "loss": 1.125,
+      "step": 4110
+    },
+    {
+      "epoch": 0.630521472392638,
+      "grad_norm": 0.1455516219139099,
+      "learning_rate": 4.590717051906608e-05,
+      "loss": 1.0339,
+      "step": 4111
+    },
+    {
+      "epoch": 0.6306748466257669,
+      "grad_norm": 0.152591273188591,
+      "learning_rate": 4.587353022382096e-05,
+      "loss": 1.0189,
+      "step": 4112
+    },
+    {
+      "epoch": 0.6308282208588957,
+      "grad_norm": 0.16439639031887054,
+      "learning_rate": 4.583989682771799e-05,
+      "loss": 1.0496,
+      "step": 4113
+    },
+    {
+      "epoch": 0.6309815950920246,
+      "grad_norm": 0.1565714329481125,
+      "learning_rate": 4.5806270338723875e-05,
+      "loss": 1.0755,
+      "step": 4114
+    },
+    {
+      "epoch": 0.6311349693251533,
+      "grad_norm": 0.15730218589305878,
+      "learning_rate": 4.577265076480366e-05,
+      "loss": 1.079,
+      "step": 4115
+    },
+    {
+      "epoch": 0.6312883435582822,
+      "grad_norm": 0.15289506316184998,
+      "learning_rate": 4.573903811392075e-05,
+      "loss": 1.0176,
+      "step": 4116
+    },
+    {
+      "epoch": 0.631441717791411,
+      "grad_norm": 0.1596888154745102,
+      "learning_rate": 4.570543239403695e-05,
+      "loss": 1.03,
+      "step": 4117
+    },
+    {
+      "epoch": 0.6315950920245399,
+      "grad_norm": 0.160442516207695,
+      "learning_rate": 4.5671833613112394e-05,
+      "loss": 1.0324,
+      "step": 4118
+    },
+    {
+      "epoch": 0.6317484662576687,
+      "grad_norm": 0.16553239524364471,
+      "learning_rate": 4.5638241779105525e-05,
+      "loss": 1.1494,
+      "step": 4119
+    },
+    {
+      "epoch": 0.6319018404907976,
+      "grad_norm": 0.1660393923521042,
+      "learning_rate": 4.560465689997323e-05,
+      "loss": 1.1543,
+      "step": 4120
+    },
+    {
+      "epoch": 0.6320552147239263,
+      "grad_norm": 0.16864372789859772,
+      "learning_rate": 4.5571078983670686e-05,
+      "loss": 1.0829,
+      "step": 4121
+    },
+    {
+      "epoch": 0.6322085889570552,
+      "grad_norm": 0.15463875234127045,
+      "learning_rate": 4.5537508038151486e-05,
+      "loss": 1.0161,
+      "step": 4122
+    },
+    {
+      "epoch": 0.632361963190184,
+      "grad_norm": 0.16050134599208832,
+      "learning_rate": 4.550394407136748e-05,
+      "loss": 1.0974,
+      "step": 4123
+    },
+    {
+      "epoch": 0.6325153374233129,
+      "grad_norm": 0.15535379946231842,
+      "learning_rate": 4.547038709126894e-05,
+      "loss": 0.9756,
+      "step": 4124
+    },
+    {
+      "epoch": 0.6326687116564417,
+      "grad_norm": 0.16100555658340454,
+      "learning_rate": 4.543683710580445e-05,
+      "loss": 1.0536,
+      "step": 4125
+    },
+    {
+      "epoch": 0.6328220858895706,
+      "grad_norm": 0.1669153869152069,
+      "learning_rate": 4.540329412292095e-05,
+      "loss": 1.0401,
+      "step": 4126
+    },
+    {
+      "epoch": 0.6329754601226993,
+      "grad_norm": 0.16109785437583923,
+      "learning_rate": 4.536975815056369e-05,
+      "loss": 1.0691,
+      "step": 4127
+    },
+    {
+      "epoch": 0.6331288343558282,
+      "grad_norm": 0.158873051404953,
+      "learning_rate": 4.533622919667631e-05,
+      "loss": 1.1209,
+      "step": 4128
+    },
+    {
+      "epoch": 0.633282208588957,
+      "grad_norm": 0.16331413388252258,
+      "learning_rate": 4.5302707269200754e-05,
+      "loss": 1.0944,
+      "step": 4129
+    },
+    {
+      "epoch": 0.6334355828220859,
+      "grad_norm": 0.16411657631397247,
+      "learning_rate": 4.526919237607732e-05,
+      "loss": 1.1862,
+      "step": 4130
+    },
+    {
+      "epoch": 0.6335889570552147,
+      "grad_norm": 0.1519576907157898,
+      "learning_rate": 4.5235684525244616e-05,
+      "loss": 1.0226,
+      "step": 4131
+    },
+    {
+      "epoch": 0.6337423312883436,
+      "grad_norm": 0.15682350099086761,
+      "learning_rate": 4.520218372463958e-05,
+      "loss": 1.086,
+      "step": 4132
+    },
+    {
+      "epoch": 0.6338957055214723,
+      "grad_norm": 0.16704848408699036,
+      "learning_rate": 4.5168689982197515e-05,
+      "loss": 1.0989,
+      "step": 4133
+    },
+    {
+      "epoch": 0.6340490797546012,
+      "grad_norm": 0.15215148031711578,
+      "learning_rate": 4.513520330585203e-05,
+      "loss": 0.9666,
+      "step": 4134
+    },
+    {
+      "epoch": 0.63420245398773,
+      "grad_norm": 0.15420414507389069,
+      "learning_rate": 4.510172370353506e-05,
+      "loss": 1.0903,
+      "step": 4135
+    },
+    {
+      "epoch": 0.6343558282208589,
+      "grad_norm": 0.1553548127412796,
+      "learning_rate": 4.506825118317688e-05,
+      "loss": 1.0257,
+      "step": 4136
+    },
+    {
+      "epoch": 0.6345092024539877,
+      "grad_norm": 0.15454690158367157,
+      "learning_rate": 4.503478575270606e-05,
+      "loss": 1.1121,
+      "step": 4137
+    },
+    {
+      "epoch": 0.6346625766871166,
+      "grad_norm": 0.15265466272830963,
+      "learning_rate": 4.500132742004951e-05,
+      "loss": 0.9745,
+      "step": 4138
+    },
+    {
+      "epoch": 0.6348159509202455,
+      "grad_norm": 0.15836651623249054,
+      "learning_rate": 4.496787619313246e-05,
+      "loss": 1.0973,
+      "step": 4139
+    },
+    {
+      "epoch": 0.6349693251533742,
+      "grad_norm": 0.14983439445495605,
+      "learning_rate": 4.493443207987843e-05,
+      "loss": 0.9484,
+      "step": 4140
+    },
+    {
+      "epoch": 0.6351226993865031,
+      "grad_norm": 0.14671963453292847,
+      "learning_rate": 4.4900995088209295e-05,
+      "loss": 1.0285,
+      "step": 4141
+    },
+    {
+      "epoch": 0.6352760736196319,
+      "grad_norm": 0.1596812605857849,
+      "learning_rate": 4.486756522604522e-05,
+      "loss": 1.0111,
+      "step": 4142
+    },
+    {
+      "epoch": 0.6354294478527608,
+      "grad_norm": 0.15917697548866272,
+      "learning_rate": 4.483414250130467e-05,
+      "loss": 1.0833,
+      "step": 4143
+    },
+    {
+      "epoch": 0.6355828220858896,
+      "grad_norm": 0.15177205204963684,
+      "learning_rate": 4.480072692190446e-05,
+      "loss": 1.0599,
+      "step": 4144
+    },
+    {
+      "epoch": 0.6357361963190185,
+      "grad_norm": 0.1574317067861557,
+      "learning_rate": 4.476731849575967e-05,
+      "loss": 1.021,
+      "step": 4145
+    },
+    {
+      "epoch": 0.6358895705521472,
+      "grad_norm": 0.17041276395320892,
+      "learning_rate": 4.473391723078371e-05,
+      "loss": 1.0847,
+      "step": 4146
+    },
+    {
+      "epoch": 0.6360429447852761,
+      "grad_norm": 0.1526332050561905,
+      "learning_rate": 4.4700523134888275e-05,
+      "loss": 1.0224,
+      "step": 4147
+    },
+    {
+      "epoch": 0.6361963190184049,
+      "grad_norm": 0.15788044035434723,
+      "learning_rate": 4.466713621598339e-05,
+      "loss": 1.1805,
+      "step": 4148
+    },
+    {
+      "epoch": 0.6363496932515338,
+      "grad_norm": 0.1644408255815506,
+      "learning_rate": 4.4633756481977345e-05,
+      "loss": 1.2121,
+      "step": 4149
+    },
+    {
+      "epoch": 0.6365030674846626,
+      "grad_norm": 0.16078224778175354,
+      "learning_rate": 4.460038394077677e-05,
+      "loss": 1.2027,
+      "step": 4150
+    },
+    {
+      "epoch": 0.6366564417177915,
+      "grad_norm": 0.16066177189350128,
+      "learning_rate": 4.456701860028654e-05,
+      "loss": 1.1459,
+      "step": 4151
+    },
+    {
+      "epoch": 0.6368098159509202,
+      "grad_norm": 0.1663476824760437,
+      "learning_rate": 4.453366046840987e-05,
+      "loss": 1.1895,
+      "step": 4152
+    },
+    {
+      "epoch": 0.6369631901840491,
+      "grad_norm": 0.14706870913505554,
+      "learning_rate": 4.450030955304825e-05,
+      "loss": 1.0196,
+      "step": 4153
+    },
+    {
+      "epoch": 0.6371165644171779,
+      "grad_norm": 0.14665493369102478,
+      "learning_rate": 4.446696586210146e-05,
+      "loss": 1.0551,
+      "step": 4154
+    },
+    {
+      "epoch": 0.6372699386503068,
+      "grad_norm": 0.16520848870277405,
+      "learning_rate": 4.443362940346755e-05,
+      "loss": 1.1173,
+      "step": 4155
+    },
+    {
+      "epoch": 0.6374233128834356,
+      "grad_norm": 0.14953812956809998,
+      "learning_rate": 4.4400300185042895e-05,
+      "loss": 0.9995,
+      "step": 4156
+    },
+    {
+      "epoch": 0.6375766871165645,
+      "grad_norm": 0.1596766710281372,
+      "learning_rate": 4.4366978214722124e-05,
+      "loss": 1.1019,
+      "step": 4157
+    },
+    {
+      "epoch": 0.6377300613496932,
+      "grad_norm": 0.16075439751148224,
+      "learning_rate": 4.433366350039817e-05,
+      "loss": 1.0688,
+      "step": 4158
+    },
+    {
+      "epoch": 0.6378834355828221,
+      "grad_norm": 0.15425916016101837,
+      "learning_rate": 4.4300356049962245e-05,
+      "loss": 0.9789,
+      "step": 4159
+    },
+    {
+      "epoch": 0.6380368098159509,
+      "grad_norm": 0.16283060610294342,
+      "learning_rate": 4.4267055871303814e-05,
+      "loss": 1.1635,
+      "step": 4160
+    },
+    {
+      "epoch": 0.6381901840490798,
+      "grad_norm": 0.16086675226688385,
+      "learning_rate": 4.423376297231067e-05,
+      "loss": 1.1332,
+      "step": 4161
+    },
+    {
+      "epoch": 0.6383435582822086,
+      "grad_norm": 0.15543517470359802,
+      "learning_rate": 4.420047736086881e-05,
+      "loss": 1.0689,
+      "step": 4162
+    },
+    {
+      "epoch": 0.6384969325153375,
+      "grad_norm": 0.15878212451934814,
+      "learning_rate": 4.416719904486257e-05,
+      "loss": 1.0846,
+      "step": 4163
+    },
+    {
+      "epoch": 0.6386503067484662,
+      "grad_norm": 0.16383205354213715,
+      "learning_rate": 4.413392803217449e-05,
+      "loss": 1.1632,
+      "step": 4164
+    },
+    {
+      "epoch": 0.6388036809815951,
+      "grad_norm": 0.16529908776283264,
+      "learning_rate": 4.4100664330685516e-05,
+      "loss": 1.1208,
+      "step": 4165
+    },
+    {
+      "epoch": 0.6389570552147239,
+      "grad_norm": 0.1418522447347641,
+      "learning_rate": 4.4067407948274706e-05,
+      "loss": 1.0142,
+      "step": 4166
+    },
+    {
+      "epoch": 0.6391104294478528,
+      "grad_norm": 0.16291998326778412,
+      "learning_rate": 4.403415889281945e-05,
+      "loss": 1.0666,
+      "step": 4167
+    },
+    {
+      "epoch": 0.6392638036809816,
+      "grad_norm": 0.14886784553527832,
+      "learning_rate": 4.400091717219542e-05,
+      "loss": 0.9651,
+      "step": 4168
+    },
+    {
+      "epoch": 0.6394171779141105,
+      "grad_norm": 0.1514214426279068,
+      "learning_rate": 4.3967682794276517e-05,
+      "loss": 1.0276,
+      "step": 4169
+    },
+    {
+      "epoch": 0.6395705521472392,
+      "grad_norm": 0.1500834971666336,
+      "learning_rate": 4.393445576693492e-05,
+      "loss": 0.9107,
+      "step": 4170
+    },
+    {
+      "epoch": 0.6397239263803681,
+      "grad_norm": 0.15735629200935364,
+      "learning_rate": 4.3901236098041066e-05,
+      "loss": 1.0975,
+      "step": 4171
+    },
+    {
+      "epoch": 0.6398773006134969,
+      "grad_norm": 0.17131754755973816,
+      "learning_rate": 4.386802379546364e-05,
+      "loss": 1.1847,
+      "step": 4172
+    },
+    {
+      "epoch": 0.6400306748466258,
+      "grad_norm": 0.162659153342247,
+      "learning_rate": 4.383481886706961e-05,
+      "loss": 1.0305,
+      "step": 4173
+    },
+    {
+      "epoch": 0.6401840490797546,
+      "grad_norm": 0.18608862161636353,
+      "learning_rate": 4.380162132072416e-05,
+      "loss": 1.1943,
+      "step": 4174
+    },
+    {
+      "epoch": 0.6403374233128835,
+      "grad_norm": 0.15512801706790924,
+      "learning_rate": 4.376843116429071e-05,
+      "loss": 1.1157,
+      "step": 4175
+    },
+    {
+      "epoch": 0.6404907975460122,
+      "grad_norm": 0.15910673141479492,
+      "learning_rate": 4.373524840563104e-05,
+      "loss": 1.0512,
+      "step": 4176
+    },
+    {
+      "epoch": 0.6406441717791411,
+      "grad_norm": 0.16498352587223053,
+      "learning_rate": 4.370207305260505e-05,
+      "loss": 1.17,
+      "step": 4177
+    },
+    {
+      "epoch": 0.6407975460122699,
+      "grad_norm": 0.16312827169895172,
+      "learning_rate": 4.3668905113070914e-05,
+      "loss": 1.0509,
+      "step": 4178
+    },
+    {
+      "epoch": 0.6409509202453988,
+      "grad_norm": 0.1684739738702774,
+      "learning_rate": 4.363574459488511e-05,
+      "loss": 1.1786,
+      "step": 4179
+    },
+    {
+      "epoch": 0.6411042944785276,
+      "grad_norm": 0.17141911387443542,
+      "learning_rate": 4.360259150590229e-05,
+      "loss": 0.9502,
+      "step": 4180
+    },
+    {
+      "epoch": 0.6412576687116565,
+      "grad_norm": 0.15395741164684296,
+      "learning_rate": 4.356944585397541e-05,
+      "loss": 1.0475,
+      "step": 4181
+    },
+    {
+      "epoch": 0.6414110429447852,
+      "grad_norm": 0.15900783240795135,
+      "learning_rate": 4.3536307646955584e-05,
+      "loss": 1.0124,
+      "step": 4182
+    },
+    {
+      "epoch": 0.6415644171779141,
+      "grad_norm": 0.1542302370071411,
+      "learning_rate": 4.3503176892692267e-05,
+      "loss": 0.9905,
+      "step": 4183
+    },
+    {
+      "epoch": 0.6417177914110429,
+      "grad_norm": 0.15114253759384155,
+      "learning_rate": 4.347005359903305e-05,
+      "loss": 1.0504,
+      "step": 4184
+    },
+    {
+      "epoch": 0.6418711656441718,
+      "grad_norm": 0.17450426518917084,
+      "learning_rate": 4.343693777382378e-05,
+      "loss": 1.0949,
+      "step": 4185
+    },
+    {
+      "epoch": 0.6420245398773006,
+      "grad_norm": 0.14286664128303528,
+      "learning_rate": 4.340382942490859e-05,
+      "loss": 0.903,
+      "step": 4186
+    },
+    {
+      "epoch": 0.6421779141104295,
+      "grad_norm": 0.15033487975597382,
+      "learning_rate": 4.3370728560129774e-05,
+      "loss": 0.9431,
+      "step": 4187
+    },
+    {
+      "epoch": 0.6423312883435582,
+      "grad_norm": 0.15217600762844086,
+      "learning_rate": 4.3337635187327915e-05,
+      "loss": 0.9507,
+      "step": 4188
+    },
+    {
+      "epoch": 0.6424846625766871,
+      "grad_norm": 0.17148920893669128,
+      "learning_rate": 4.330454931434175e-05,
+      "loss": 1.212,
+      "step": 4189
+    },
+    {
+      "epoch": 0.6426380368098159,
+      "grad_norm": 0.16978555917739868,
+      "learning_rate": 4.327147094900833e-05,
+      "loss": 1.0771,
+      "step": 4190
+    },
+    {
+      "epoch": 0.6427914110429448,
+      "grad_norm": 0.15417224168777466,
+      "learning_rate": 4.323840009916282e-05,
+      "loss": 1.0578,
+      "step": 4191
+    },
+    {
+      "epoch": 0.6429447852760736,
+      "grad_norm": 0.15253567695617676,
+      "learning_rate": 4.320533677263872e-05,
+      "loss": 1.011,
+      "step": 4192
+    },
+    {
+      "epoch": 0.6430981595092025,
+      "grad_norm": 0.16444191336631775,
+      "learning_rate": 4.3172280977267656e-05,
+      "loss": 1.1529,
+      "step": 4193
+    },
+    {
+      "epoch": 0.6432515337423312,
+      "grad_norm": 0.154244065284729,
+      "learning_rate": 4.313923272087947e-05,
+      "loss": 1.1063,
+      "step": 4194
+    },
+    {
+      "epoch": 0.6434049079754601,
+      "grad_norm": 0.16112159192562103,
+      "learning_rate": 4.310619201130233e-05,
+      "loss": 1.0848,
+      "step": 4195
+    },
+    {
+      "epoch": 0.6435582822085889,
+      "grad_norm": 0.16567301750183105,
+      "learning_rate": 4.307315885636245e-05,
+      "loss": 1.0994,
+      "step": 4196
+    },
+    {
+      "epoch": 0.6437116564417178,
+      "grad_norm": 0.16046570241451263,
+      "learning_rate": 4.304013326388443e-05,
+      "loss": 1.0894,
+      "step": 4197
+    },
+    {
+      "epoch": 0.6438650306748466,
+      "grad_norm": 0.1569291353225708,
+      "learning_rate": 4.3007115241690915e-05,
+      "loss": 0.9391,
+      "step": 4198
+    },
+    {
+      "epoch": 0.6440184049079755,
+      "grad_norm": 0.1487690508365631,
+      "learning_rate": 4.29741047976029e-05,
+      "loss": 0.9553,
+      "step": 4199
+    },
+    {
+      "epoch": 0.6441717791411042,
+      "grad_norm": 0.1519821733236313,
+      "learning_rate": 4.2941101939439496e-05,
+      "loss": 0.99,
+      "step": 4200
+    },
+    {
+      "epoch": 0.6443251533742331,
+      "grad_norm": 0.15131717920303345,
+      "learning_rate": 4.2908106675017984e-05,
+      "loss": 1.0378,
+      "step": 4201
+    },
+    {
+      "epoch": 0.644478527607362,
+      "grad_norm": 0.15518660843372345,
+      "learning_rate": 4.287511901215398e-05,
+      "loss": 1.0629,
+      "step": 4202
+    },
+    {
+      "epoch": 0.6446319018404908,
+      "grad_norm": 0.15553152561187744,
+      "learning_rate": 4.2842138958661154e-05,
+      "loss": 1.0342,
+      "step": 4203
+    },
+    {
+      "epoch": 0.6447852760736197,
+      "grad_norm": 0.14753149449825287,
+      "learning_rate": 4.2809166522351514e-05,
+      "loss": 0.9711,
+      "step": 4204
+    },
+    {
+      "epoch": 0.6449386503067485,
+      "grad_norm": 0.1678260713815689,
+      "learning_rate": 4.2776201711035114e-05,
+      "loss": 1.1588,
+      "step": 4205
+    },
+    {
+      "epoch": 0.6450920245398774,
+      "grad_norm": 0.17150460183620453,
+      "learning_rate": 4.274324453252032e-05,
+      "loss": 1.1584,
+      "step": 4206
+    },
+    {
+      "epoch": 0.6452453987730061,
+      "grad_norm": 0.16658475995063782,
+      "learning_rate": 4.271029499461367e-05,
+      "loss": 1.174,
+      "step": 4207
+    },
+    {
+      "epoch": 0.645398773006135,
+      "grad_norm": 0.15804807841777802,
+      "learning_rate": 4.2677353105119815e-05,
+      "loss": 1.0558,
+      "step": 4208
+    },
+    {
+      "epoch": 0.6455521472392638,
+      "grad_norm": 0.16138483583927155,
+      "learning_rate": 4.264441887184172e-05,
+      "loss": 1.0411,
+      "step": 4209
+    },
+    {
+      "epoch": 0.6457055214723927,
+      "grad_norm": 0.1656871736049652,
+      "learning_rate": 4.261149230258038e-05,
+      "loss": 1.0578,
+      "step": 4210
+    },
+    {
+      "epoch": 0.6458588957055215,
+      "grad_norm": 0.15863549709320068,
+      "learning_rate": 4.257857340513513e-05,
+      "loss": 1.0479,
+      "step": 4211
+    },
+    {
+      "epoch": 0.6460122699386504,
+      "grad_norm": 0.1695554256439209,
+      "learning_rate": 4.254566218730342e-05,
+      "loss": 1.202,
+      "step": 4212
+    },
+    {
+      "epoch": 0.6461656441717791,
+      "grad_norm": 0.1522192507982254,
+      "learning_rate": 4.251275865688079e-05,
+      "loss": 1.0136,
+      "step": 4213
+    },
+    {
+      "epoch": 0.646319018404908,
+      "grad_norm": 0.15386785566806793,
+      "learning_rate": 4.247986282166117e-05,
+      "loss": 0.9935,
+      "step": 4214
+    },
+    {
+      "epoch": 0.6464723926380368,
+      "grad_norm": 0.15300044417381287,
+      "learning_rate": 4.244697468943645e-05,
+      "loss": 0.9853,
+      "step": 4215
+    },
+    {
+      "epoch": 0.6466257668711657,
+      "grad_norm": 0.16664716601371765,
+      "learning_rate": 4.2414094267996855e-05,
+      "loss": 1.1869,
+      "step": 4216
+    },
+    {
+      "epoch": 0.6467791411042945,
+      "grad_norm": 0.1618921011686325,
+      "learning_rate": 4.2381221565130676e-05,
+      "loss": 1.0361,
+      "step": 4217
+    },
+    {
+      "epoch": 0.6469325153374234,
+      "grad_norm": 0.15727147459983826,
+      "learning_rate": 4.234835658862446e-05,
+      "loss": 0.9909,
+      "step": 4218
+    },
+    {
+      "epoch": 0.6470858895705521,
+      "grad_norm": 0.16167579591274261,
+      "learning_rate": 4.2315499346262856e-05,
+      "loss": 1.1718,
+      "step": 4219
+    },
+    {
+      "epoch": 0.647239263803681,
+      "grad_norm": 0.15506502985954285,
+      "learning_rate": 4.22826498458287e-05,
+      "loss": 1.0832,
+      "step": 4220
+    },
+    {
+      "epoch": 0.6473926380368098,
+      "grad_norm": 0.1561659425497055,
+      "learning_rate": 4.2249808095103035e-05,
+      "loss": 1.0453,
+      "step": 4221
+    },
+    {
+      "epoch": 0.6475460122699387,
+      "grad_norm": 0.17632365226745605,
+      "learning_rate": 4.2216974101865e-05,
+      "loss": 1.2768,
+      "step": 4222
+    },
+    {
+      "epoch": 0.6476993865030675,
+      "grad_norm": 0.15960724651813507,
+      "learning_rate": 4.218414787389197e-05,
+      "loss": 1.0897,
+      "step": 4223
+    },
+    {
+      "epoch": 0.6478527607361964,
+      "grad_norm": 0.16434144973754883,
+      "learning_rate": 4.21513294189594e-05,
+      "loss": 1.0486,
+      "step": 4224
+    },
+    {
+      "epoch": 0.6480061349693251,
+      "grad_norm": 0.16351114213466644,
+      "learning_rate": 4.211851874484099e-05,
+      "loss": 1.0691,
+      "step": 4225
+    },
+    {
+      "epoch": 0.648159509202454,
+      "grad_norm": 0.16473975777626038,
+      "learning_rate": 4.2085715859308525e-05,
+      "loss": 1.117,
+      "step": 4226
+    },
+    {
+      "epoch": 0.6483128834355828,
+      "grad_norm": 0.17202623188495636,
+      "learning_rate": 4.2052920770132e-05,
+      "loss": 1.104,
+      "step": 4227
+    },
+    {
+      "epoch": 0.6484662576687117,
+      "grad_norm": 0.16176828742027283,
+      "learning_rate": 4.2020133485079524e-05,
+      "loss": 1.1317,
+      "step": 4228
+    },
+    {
+      "epoch": 0.6486196319018405,
+      "grad_norm": 0.1725386381149292,
+      "learning_rate": 4.198735401191734e-05,
+      "loss": 1.1975,
+      "step": 4229
+    },
+    {
+      "epoch": 0.6487730061349694,
+      "grad_norm": 0.16195085644721985,
+      "learning_rate": 4.195458235840993e-05,
+      "loss": 1.1099,
+      "step": 4230
+    },
+    {
+      "epoch": 0.6489263803680981,
+      "grad_norm": 0.148160919547081,
+      "learning_rate": 4.1921818532319805e-05,
+      "loss": 0.9687,
+      "step": 4231
+    },
+    {
+      "epoch": 0.649079754601227,
+      "grad_norm": 0.15499632060527802,
+      "learning_rate": 4.188906254140774e-05,
+      "loss": 1.0629,
+      "step": 4232
+    },
+    {
+      "epoch": 0.6492331288343558,
+      "grad_norm": 0.159387469291687,
+      "learning_rate": 4.185631439343253e-05,
+      "loss": 1.0292,
+      "step": 4233
+    },
+    {
+      "epoch": 0.6493865030674847,
+      "grad_norm": 0.1712397187948227,
+      "learning_rate": 4.1823574096151256e-05,
+      "loss": 1.2267,
+      "step": 4234
+    },
+    {
+      "epoch": 0.6495398773006135,
+      "grad_norm": 0.1575055867433548,
+      "learning_rate": 4.179084165731903e-05,
+      "loss": 1.0308,
+      "step": 4235
+    },
+    {
+      "epoch": 0.6496932515337424,
+      "grad_norm": 0.16220234334468842,
+      "learning_rate": 4.1758117084689094e-05,
+      "loss": 1.0874,
+      "step": 4236
+    },
+    {
+      "epoch": 0.6498466257668711,
+      "grad_norm": 0.15492837131023407,
+      "learning_rate": 4.172540038601292e-05,
+      "loss": 1.1408,
+      "step": 4237
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.16737353801727295,
+      "learning_rate": 4.169269156904003e-05,
+      "loss": 1.123,
+      "step": 4238
+    },
+    {
+      "epoch": 0.6501533742331288,
+      "grad_norm": 0.1745455116033554,
+      "learning_rate": 4.1659990641518146e-05,
+      "loss": 1.1289,
+      "step": 4239
+    },
+    {
+      "epoch": 0.6503067484662577,
+      "grad_norm": 0.2124985158443451,
+      "learning_rate": 4.162729761119303e-05,
+      "loss": 1.1226,
+      "step": 4240
+    },
+    {
+      "epoch": 0.6504601226993865,
+      "grad_norm": 0.16254720091819763,
+      "learning_rate": 4.1594612485808696e-05,
+      "loss": 1.1532,
+      "step": 4241
+    },
+    {
+      "epoch": 0.6506134969325154,
+      "grad_norm": 0.1683128923177719,
+      "learning_rate": 4.156193527310719e-05,
+      "loss": 1.218,
+      "step": 4242
+    },
+    {
+      "epoch": 0.6507668711656441,
+      "grad_norm": 0.15618206560611725,
+      "learning_rate": 4.1529265980828686e-05,
+      "loss": 0.986,
+      "step": 4243
+    },
+    {
+      "epoch": 0.650920245398773,
+      "grad_norm": 0.15879414975643158,
+      "learning_rate": 4.149660461671156e-05,
+      "loss": 1.0529,
+      "step": 4244
+    },
+    {
+      "epoch": 0.6510736196319018,
+      "grad_norm": 0.167622372508049,
+      "learning_rate": 4.14639511884922e-05,
+      "loss": 1.2177,
+      "step": 4245
+    },
+    {
+      "epoch": 0.6512269938650307,
+      "grad_norm": 0.17096753418445587,
+      "learning_rate": 4.143130570390524e-05,
+      "loss": 1.0928,
+      "step": 4246
+    },
+    {
+      "epoch": 0.6513803680981595,
+      "grad_norm": 0.15859466791152954,
+      "learning_rate": 4.1398668170683305e-05,
+      "loss": 1.0916,
+      "step": 4247
+    },
+    {
+      "epoch": 0.6515337423312884,
+      "grad_norm": 0.1679517775774002,
+      "learning_rate": 4.136603859655724e-05,
+      "loss": 1.1035,
+      "step": 4248
+    },
+    {
+      "epoch": 0.6516871165644171,
+      "grad_norm": 0.15547356009483337,
+      "learning_rate": 4.1333416989255935e-05,
+      "loss": 1.0068,
+      "step": 4249
+    },
+    {
+      "epoch": 0.651840490797546,
+      "grad_norm": 0.16352684795856476,
+      "learning_rate": 4.130080335650643e-05,
+      "loss": 1.125,
+      "step": 4250
+    },
+    {
+      "epoch": 0.6519938650306748,
+      "grad_norm": 0.16490840911865234,
+      "learning_rate": 4.126819770603389e-05,
+      "loss": 1.1556,
+      "step": 4251
+    },
+    {
+      "epoch": 0.6521472392638037,
+      "grad_norm": 0.15922421216964722,
+      "learning_rate": 4.1235600045561505e-05,
+      "loss": 1.0697,
+      "step": 4252
+    },
+    {
+      "epoch": 0.6523006134969325,
+      "grad_norm": 0.15966758131980896,
+      "learning_rate": 4.120301038281072e-05,
+      "loss": 1.134,
+      "step": 4253
+    },
+    {
+      "epoch": 0.6524539877300614,
+      "grad_norm": 0.16938716173171997,
+      "learning_rate": 4.117042872550094e-05,
+      "loss": 1.2135,
+      "step": 4254
+    },
+    {
+      "epoch": 0.6526073619631901,
+      "grad_norm": 0.173355370759964,
+      "learning_rate": 4.113785508134972e-05,
+      "loss": 1.1215,
+      "step": 4255
+    },
+    {
+      "epoch": 0.652760736196319,
+      "grad_norm": 0.15840348601341248,
+      "learning_rate": 4.1105289458072766e-05,
+      "loss": 1.1247,
+      "step": 4256
+    },
+    {
+      "epoch": 0.6529141104294478,
+      "grad_norm": 0.15972819924354553,
+      "learning_rate": 4.107273186338381e-05,
+      "loss": 1.0886,
+      "step": 4257
+    },
+    {
+      "epoch": 0.6530674846625767,
+      "grad_norm": 0.1588718146085739,
+      "learning_rate": 4.104018230499477e-05,
+      "loss": 1.1035,
+      "step": 4258
+    },
+    {
+      "epoch": 0.6532208588957055,
+      "grad_norm": 0.16256824135780334,
+      "learning_rate": 4.100764079061556e-05,
+      "loss": 1.146,
+      "step": 4259
+    },
+    {
+      "epoch": 0.6533742331288344,
+      "grad_norm": 0.15615764260292053,
+      "learning_rate": 4.097510732795427e-05,
+      "loss": 1.0462,
+      "step": 4260
+    },
+    {
+      "epoch": 0.6535276073619631,
+      "grad_norm": 0.15832509100437164,
+      "learning_rate": 4.094258192471703e-05,
+      "loss": 1.0838,
+      "step": 4261
+    },
+    {
+      "epoch": 0.653680981595092,
+      "grad_norm": 0.1528276652097702,
+      "learning_rate": 4.0910064588608116e-05,
+      "loss": 0.8615,
+      "step": 4262
+    },
+    {
+      "epoch": 0.6538343558282208,
+      "grad_norm": 0.1610703468322754,
+      "learning_rate": 4.0877555327329845e-05,
+      "loss": 1.0235,
+      "step": 4263
+    },
+    {
+      "epoch": 0.6539877300613497,
+      "grad_norm": 0.1466621607542038,
+      "learning_rate": 4.0845054148582576e-05,
+      "loss": 0.8838,
+      "step": 4264
+    },
+    {
+      "epoch": 0.6541411042944786,
+      "grad_norm": 0.15962162613868713,
+      "learning_rate": 4.081256106006491e-05,
+      "loss": 1.154,
+      "step": 4265
+    },
+    {
+      "epoch": 0.6542944785276074,
+      "grad_norm": 0.16617649793624878,
+      "learning_rate": 4.078007606947336e-05,
+      "loss": 1.0569,
+      "step": 4266
+    },
+    {
+      "epoch": 0.6544478527607362,
+      "grad_norm": 0.1464460790157318,
+      "learning_rate": 4.074759918450265e-05,
+      "loss": 0.9629,
+      "step": 4267
+    },
+    {
+      "epoch": 0.654601226993865,
+      "grad_norm": 0.15803605318069458,
+      "learning_rate": 4.071513041284548e-05,
+      "loss": 0.9705,
+      "step": 4268
+    },
+    {
+      "epoch": 0.6547546012269939,
+      "grad_norm": 0.16167838871479034,
+      "learning_rate": 4.068266976219273e-05,
+      "loss": 0.8872,
+      "step": 4269
+    },
+    {
+      "epoch": 0.6549079754601227,
+      "grad_norm": 0.15657852590084076,
+      "learning_rate": 4.065021724023329e-05,
+      "loss": 1.1141,
+      "step": 4270
+    },
+    {
+      "epoch": 0.6550613496932516,
+      "grad_norm": 0.15178610384464264,
+      "learning_rate": 4.061777285465408e-05,
+      "loss": 1.0657,
+      "step": 4271
+    },
+    {
+      "epoch": 0.6552147239263804,
+      "grad_norm": 0.1721733957529068,
+      "learning_rate": 4.058533661314023e-05,
+      "loss": 1.1317,
+      "step": 4272
+    },
+    {
+      "epoch": 0.6553680981595092,
+      "grad_norm": 0.1599392145872116,
+      "learning_rate": 4.055290852337481e-05,
+      "loss": 1.1002,
+      "step": 4273
+    },
+    {
+      "epoch": 0.655521472392638,
+      "grad_norm": 0.17357072234153748,
+      "learning_rate": 4.0520488593039066e-05,
+      "loss": 1.133,
+      "step": 4274
+    },
+    {
+      "epoch": 0.6556748466257669,
+      "grad_norm": 0.17255747318267822,
+      "learning_rate": 4.048807682981219e-05,
+      "loss": 1.1451,
+      "step": 4275
+    },
+    {
+      "epoch": 0.6558282208588957,
+      "grad_norm": 0.1667288839817047,
+      "learning_rate": 4.045567324137157e-05,
+      "loss": 1.1631,
+      "step": 4276
+    },
+    {
+      "epoch": 0.6559815950920246,
+      "grad_norm": 0.1570214182138443,
+      "learning_rate": 4.0423277835392573e-05,
+      "loss": 1.028,
+      "step": 4277
+    },
+    {
+      "epoch": 0.6561349693251534,
+      "grad_norm": 0.15232422947883606,
+      "learning_rate": 4.0390890619548594e-05,
+      "loss": 1.0082,
+      "step": 4278
+    },
+    {
+      "epoch": 0.6562883435582823,
+      "grad_norm": 0.1588517129421234,
+      "learning_rate": 4.035851160151123e-05,
+      "loss": 1.1022,
+      "step": 4279
+    },
+    {
+      "epoch": 0.656441717791411,
+      "grad_norm": 0.1822834014892578,
+      "learning_rate": 4.0326140788949976e-05,
+      "loss": 1.2082,
+      "step": 4280
+    },
+    {
+      "epoch": 0.6565950920245399,
+      "grad_norm": 0.14410391449928284,
+      "learning_rate": 4.029377818953252e-05,
+      "loss": 1.1044,
+      "step": 4281
+    },
+    {
+      "epoch": 0.6567484662576687,
+      "grad_norm": 0.1590224653482437,
+      "learning_rate": 4.0261423810924485e-05,
+      "loss": 1.1572,
+      "step": 4282
+    },
+    {
+      "epoch": 0.6569018404907976,
+      "grad_norm": 0.16456656157970428,
+      "learning_rate": 4.022907766078966e-05,
+      "loss": 1.0476,
+      "step": 4283
+    },
+    {
+      "epoch": 0.6570552147239264,
+      "grad_norm": 0.1609247922897339,
+      "learning_rate": 4.019673974678977e-05,
+      "loss": 1.1903,
+      "step": 4284
+    },
+    {
+      "epoch": 0.6572085889570553,
+      "grad_norm": 0.1492173671722412,
+      "learning_rate": 4.01644100765847e-05,
+      "loss": 1.0303,
+      "step": 4285
+    },
+    {
+      "epoch": 0.657361963190184,
+      "grad_norm": 0.1563778668642044,
+      "learning_rate": 4.0132088657832304e-05,
+      "loss": 1.0147,
+      "step": 4286
+    },
+    {
+      "epoch": 0.6575153374233129,
+      "grad_norm": 0.15758267045021057,
+      "learning_rate": 4.0099775498188476e-05,
+      "loss": 1.1012,
+      "step": 4287
+    },
+    {
+      "epoch": 0.6576687116564417,
+      "grad_norm": 0.15548963844776154,
+      "learning_rate": 4.006747060530725e-05,
+      "loss": 1.1262,
+      "step": 4288
+    },
+    {
+      "epoch": 0.6578220858895706,
+      "grad_norm": 0.1497388780117035,
+      "learning_rate": 4.0035173986840565e-05,
+      "loss": 1.0407,
+      "step": 4289
+    },
+    {
+      "epoch": 0.6579754601226994,
+      "grad_norm": 0.16109821200370789,
+      "learning_rate": 4.000288565043854e-05,
+      "loss": 1.0807,
+      "step": 4290
+    },
+    {
+      "epoch": 0.6581288343558283,
+      "grad_norm": 0.35926222801208496,
+      "learning_rate": 3.9970605603749195e-05,
+      "loss": 1.1688,
+      "step": 4291
+    },
+    {
+      "epoch": 0.658282208588957,
+      "grad_norm": 0.15786033868789673,
+      "learning_rate": 3.99383338544187e-05,
+      "loss": 1.0672,
+      "step": 4292
+    },
+    {
+      "epoch": 0.6584355828220859,
+      "grad_norm": 0.15514659881591797,
+      "learning_rate": 3.990607041009123e-05,
+      "loss": 0.9375,
+      "step": 4293
+    },
+    {
+      "epoch": 0.6585889570552147,
+      "grad_norm": 0.17768561840057373,
+      "learning_rate": 3.987381527840893e-05,
+      "loss": 1.2664,
+      "step": 4294
+    },
+    {
+      "epoch": 0.6587423312883436,
+      "grad_norm": 0.1659986525774002,
+      "learning_rate": 3.984156846701208e-05,
+      "loss": 1.0329,
+      "step": 4295
+    },
+    {
+      "epoch": 0.6588957055214724,
+      "grad_norm": 0.1473858803510666,
+      "learning_rate": 3.9809329983538864e-05,
+      "loss": 0.9432,
+      "step": 4296
+    },
+    {
+      "epoch": 0.6590490797546013,
+      "grad_norm": 0.1542430967092514,
+      "learning_rate": 3.977709983562562e-05,
+      "loss": 1.022,
+      "step": 4297
+    },
+    {
+      "epoch": 0.65920245398773,
+      "grad_norm": 0.1622529774904251,
+      "learning_rate": 3.974487803090664e-05,
+      "loss": 1.067,
+      "step": 4298
+    },
+    {
+      "epoch": 0.6593558282208589,
+      "grad_norm": 0.21890589594841003,
+      "learning_rate": 3.971266457701421e-05,
+      "loss": 1.1779,
+      "step": 4299
+    },
+    {
+      "epoch": 0.6595092024539877,
+      "grad_norm": 0.16704028844833374,
+      "learning_rate": 3.968045948157875e-05,
+      "loss": 1.2092,
+      "step": 4300
+    },
+    {
+      "epoch": 0.6596625766871166,
+      "grad_norm": 0.15885594487190247,
+      "learning_rate": 3.964826275222857e-05,
+      "loss": 1.0782,
+      "step": 4301
+    },
+    {
+      "epoch": 0.6598159509202454,
+      "grad_norm": 0.15500549972057343,
+      "learning_rate": 3.9616074396590107e-05,
+      "loss": 1.0719,
+      "step": 4302
+    },
+    {
+      "epoch": 0.6599693251533743,
+      "grad_norm": 0.17408056557178497,
+      "learning_rate": 3.958389442228773e-05,
+      "loss": 1.2737,
+      "step": 4303
+    },
+    {
+      "epoch": 0.660122699386503,
+      "grad_norm": 0.15466268360614777,
+      "learning_rate": 3.95517228369439e-05,
+      "loss": 1.0336,
+      "step": 4304
+    },
+    {
+      "epoch": 0.6602760736196319,
+      "grad_norm": 0.15790748596191406,
+      "learning_rate": 3.9519559648179035e-05,
+      "loss": 1.0443,
+      "step": 4305
+    },
+    {
+      "epoch": 0.6604294478527607,
+      "grad_norm": 0.15243235230445862,
+      "learning_rate": 3.9487404863611545e-05,
+      "loss": 1.0657,
+      "step": 4306
+    },
+    {
+      "epoch": 0.6605828220858896,
+      "grad_norm": 0.15943321585655212,
+      "learning_rate": 3.945525849085793e-05,
+      "loss": 1.0978,
+      "step": 4307
+    },
+    {
+      "epoch": 0.6607361963190184,
+      "grad_norm": 0.14388729631900787,
+      "learning_rate": 3.942312053753262e-05,
+      "loss": 0.9368,
+      "step": 4308
+    },
+    {
+      "epoch": 0.6608895705521473,
+      "grad_norm": 0.15687178075313568,
+      "learning_rate": 3.939099101124812e-05,
+      "loss": 1.0967,
+      "step": 4309
+    },
+    {
+      "epoch": 0.661042944785276,
+      "grad_norm": 0.16107988357543945,
+      "learning_rate": 3.935886991961486e-05,
+      "loss": 1.1217,
+      "step": 4310
+    },
+    {
+      "epoch": 0.6611963190184049,
+      "grad_norm": 0.153348907828331,
+      "learning_rate": 3.932675727024135e-05,
+      "loss": 1.0304,
+      "step": 4311
+    },
+    {
+      "epoch": 0.6613496932515337,
+      "grad_norm": 0.16674523055553436,
+      "learning_rate": 3.929465307073405e-05,
+      "loss": 1.2408,
+      "step": 4312
+    },
+    {
+      "epoch": 0.6615030674846626,
+      "grad_norm": 0.1527973860502243,
+      "learning_rate": 3.92625573286974e-05,
+      "loss": 1.0612,
+      "step": 4313
+    },
+    {
+      "epoch": 0.6616564417177914,
+      "grad_norm": 0.15312857925891876,
+      "learning_rate": 3.923047005173393e-05,
+      "loss": 0.9311,
+      "step": 4314
+    },
+    {
+      "epoch": 0.6618098159509203,
+      "grad_norm": 0.14754623174667358,
+      "learning_rate": 3.9198391247444036e-05,
+      "loss": 0.9376,
+      "step": 4315
+    },
+    {
+      "epoch": 0.661963190184049,
+      "grad_norm": 0.14537326991558075,
+      "learning_rate": 3.916632092342625e-05,
+      "loss": 0.9762,
+      "step": 4316
+    },
+    {
+      "epoch": 0.6621165644171779,
+      "grad_norm": 0.16158458590507507,
+      "learning_rate": 3.913425908727696e-05,
+      "loss": 1.0793,
+      "step": 4317
+    },
+    {
+      "epoch": 0.6622699386503067,
+      "grad_norm": 0.15586626529693604,
+      "learning_rate": 3.910220574659065e-05,
+      "loss": 1.1053,
+      "step": 4318
+    },
+    {
+      "epoch": 0.6624233128834356,
+      "grad_norm": 0.15231674909591675,
+      "learning_rate": 3.907016090895968e-05,
+      "loss": 1.0225,
+      "step": 4319
+    },
+    {
+      "epoch": 0.6625766871165644,
+      "grad_norm": 0.18038927018642426,
+      "learning_rate": 3.903812458197454e-05,
+      "loss": 1.2159,
+      "step": 4320
+    },
+    {
+      "epoch": 0.6627300613496933,
+      "grad_norm": 0.1457068920135498,
+      "learning_rate": 3.900609677322359e-05,
+      "loss": 0.946,
+      "step": 4321
+    },
+    {
+      "epoch": 0.662883435582822,
+      "grad_norm": 0.4744737148284912,
+      "learning_rate": 3.897407749029318e-05,
+      "loss": 1.1802,
+      "step": 4322
+    },
+    {
+      "epoch": 0.6630368098159509,
+      "grad_norm": 0.15916495025157928,
+      "learning_rate": 3.894206674076772e-05,
+      "loss": 1.1307,
+      "step": 4323
+    },
+    {
+      "epoch": 0.6631901840490797,
+      "grad_norm": 0.16184011101722717,
+      "learning_rate": 3.8910064532229493e-05,
+      "loss": 1.0744,
+      "step": 4324
+    },
+    {
+      "epoch": 0.6633435582822086,
+      "grad_norm": 0.15676051378250122,
+      "learning_rate": 3.887807087225888e-05,
+      "loss": 1.0594,
+      "step": 4325
+    },
+    {
+      "epoch": 0.6634969325153375,
+      "grad_norm": 0.15250150859355927,
+      "learning_rate": 3.884608576843409e-05,
+      "loss": 1.0212,
+      "step": 4326
+    },
+    {
+      "epoch": 0.6636503067484663,
+      "grad_norm": 0.1587228924036026,
+      "learning_rate": 3.8814109228331456e-05,
+      "loss": 1.0615,
+      "step": 4327
+    },
+    {
+      "epoch": 0.6638036809815951,
+      "grad_norm": 0.15888157486915588,
+      "learning_rate": 3.87821412595252e-05,
+      "loss": 1.116,
+      "step": 4328
+    },
+    {
+      "epoch": 0.6639570552147239,
+      "grad_norm": 0.1581210196018219,
+      "learning_rate": 3.875018186958746e-05,
+      "loss": 0.9375,
+      "step": 4329
+    },
+    {
+      "epoch": 0.6641104294478528,
+      "grad_norm": 0.14942006766796112,
+      "learning_rate": 3.8718231066088495e-05,
+      "loss": 1.024,
+      "step": 4330
+    },
+    {
+      "epoch": 0.6642638036809816,
+      "grad_norm": 0.15661093592643738,
+      "learning_rate": 3.8686288856596366e-05,
+      "loss": 1.0741,
+      "step": 4331
+    },
+    {
+      "epoch": 0.6644171779141105,
+      "grad_norm": 0.16886527836322784,
+      "learning_rate": 3.8654355248677256e-05,
+      "loss": 1.0404,
+      "step": 4332
+    },
+    {
+      "epoch": 0.6645705521472393,
+      "grad_norm": 0.15575972199440002,
+      "learning_rate": 3.862243024989515e-05,
+      "loss": 1.0263,
+      "step": 4333
+    },
+    {
+      "epoch": 0.6647239263803681,
+      "grad_norm": 0.16187970340251923,
+      "learning_rate": 3.8590513867812114e-05,
+      "loss": 1.0447,
+      "step": 4334
+    },
+    {
+      "epoch": 0.6648773006134969,
+      "grad_norm": 0.1495710164308548,
+      "learning_rate": 3.8558606109988147e-05,
+      "loss": 1.0292,
+      "step": 4335
+    },
+    {
+      "epoch": 0.6650306748466258,
+      "grad_norm": 0.1653447449207306,
+      "learning_rate": 3.852670698398117e-05,
+      "loss": 1.1253,
+      "step": 4336
+    },
+    {
+      "epoch": 0.6651840490797546,
+      "grad_norm": 0.16111351549625397,
+      "learning_rate": 3.8494816497347084e-05,
+      "loss": 1.1428,
+      "step": 4337
+    },
+    {
+      "epoch": 0.6653374233128835,
+      "grad_norm": 0.1632365882396698,
+      "learning_rate": 3.846293465763972e-05,
+      "loss": 1.1258,
+      "step": 4338
+    },
+    {
+      "epoch": 0.6654907975460123,
+      "grad_norm": 0.16432106494903564,
+      "learning_rate": 3.843106147241093e-05,
+      "loss": 1.1849,
+      "step": 4339
+    },
+    {
+      "epoch": 0.6656441717791411,
+      "grad_norm": 0.16820837557315826,
+      "learning_rate": 3.839919694921043e-05,
+      "loss": 1.1626,
+      "step": 4340
+    },
+    {
+      "epoch": 0.6657975460122699,
+      "grad_norm": 0.16986069083213806,
+      "learning_rate": 3.836734109558589e-05,
+      "loss": 1.2683,
+      "step": 4341
+    },
+    {
+      "epoch": 0.6659509202453988,
+      "grad_norm": 0.15822920203208923,
+      "learning_rate": 3.833549391908304e-05,
+      "loss": 1.0884,
+      "step": 4342
+    },
+    {
+      "epoch": 0.6661042944785276,
+      "grad_norm": 0.17211344838142395,
+      "learning_rate": 3.830365542724538e-05,
+      "loss": 1.1418,
+      "step": 4343
+    },
+    {
+      "epoch": 0.6662576687116565,
+      "grad_norm": 0.1580507457256317,
+      "learning_rate": 3.827182562761453e-05,
+      "loss": 1.0075,
+      "step": 4344
+    },
+    {
+      "epoch": 0.6664110429447853,
+      "grad_norm": 0.17606805264949799,
+      "learning_rate": 3.824000452772988e-05,
+      "loss": 1.097,
+      "step": 4345
+    },
+    {
+      "epoch": 0.6665644171779141,
+      "grad_norm": 0.1615937203168869,
+      "learning_rate": 3.820819213512893e-05,
+      "loss": 1.0575,
+      "step": 4346
+    },
+    {
+      "epoch": 0.6667177914110429,
+      "grad_norm": 0.1570366770029068,
+      "learning_rate": 3.8176388457346986e-05,
+      "loss": 1.0985,
+      "step": 4347
+    },
+    {
+      "epoch": 0.6668711656441718,
+      "grad_norm": 0.16763897240161896,
+      "learning_rate": 3.814459350191732e-05,
+      "loss": 1.0896,
+      "step": 4348
+    },
+    {
+      "epoch": 0.6670245398773006,
+      "grad_norm": 0.17233878374099731,
+      "learning_rate": 3.811280727637121e-05,
+      "loss": 1.1271,
+      "step": 4349
+    },
+    {
+      "epoch": 0.6671779141104295,
+      "grad_norm": 0.1638813316822052,
+      "learning_rate": 3.808102978823773e-05,
+      "loss": 1.0935,
+      "step": 4350
+    },
+    {
+      "epoch": 0.6673312883435583,
+      "grad_norm": 0.15781360864639282,
+      "learning_rate": 3.804926104504406e-05,
+      "loss": 1.092,
+      "step": 4351
+    },
+    {
+      "epoch": 0.6674846625766871,
+      "grad_norm": 0.15230731666088104,
+      "learning_rate": 3.801750105431512e-05,
+      "loss": 1.0363,
+      "step": 4352
+    },
+    {
+      "epoch": 0.6676380368098159,
+      "grad_norm": 0.16590295732021332,
+      "learning_rate": 3.798574982357394e-05,
+      "loss": 1.0912,
+      "step": 4353
+    },
+    {
+      "epoch": 0.6677914110429448,
+      "grad_norm": 0.15731804072856903,
+      "learning_rate": 3.795400736034131e-05,
+      "loss": 1.1195,
+      "step": 4354
+    },
+    {
+      "epoch": 0.6679447852760736,
+      "grad_norm": 0.1641869992017746,
+      "learning_rate": 3.7922273672136074e-05,
+      "loss": 1.1859,
+      "step": 4355
+    },
+    {
+      "epoch": 0.6680981595092025,
+      "grad_norm": 0.1613423377275467,
+      "learning_rate": 3.7890548766474914e-05,
+      "loss": 0.9921,
+      "step": 4356
+    },
+    {
+      "epoch": 0.6682515337423313,
+      "grad_norm": 0.1788252741098404,
+      "learning_rate": 3.785883265087245e-05,
+      "loss": 1.1437,
+      "step": 4357
+    },
+    {
+      "epoch": 0.6684049079754601,
+      "grad_norm": 0.14580023288726807,
+      "learning_rate": 3.7827125332841255e-05,
+      "loss": 0.9247,
+      "step": 4358
+    },
+    {
+      "epoch": 0.6685582822085889,
+      "grad_norm": 0.14230574667453766,
+      "learning_rate": 3.779542681989176e-05,
+      "loss": 0.9267,
+      "step": 4359
+    },
+    {
+      "epoch": 0.6687116564417178,
+      "grad_norm": 0.15303421020507812,
+      "learning_rate": 3.776373711953238e-05,
+      "loss": 1.0503,
+      "step": 4360
+    },
+    {
+      "epoch": 0.6688650306748466,
+      "grad_norm": 0.15893405675888062,
+      "learning_rate": 3.7732056239269376e-05,
+      "loss": 1.0835,
+      "step": 4361
+    },
+    {
+      "epoch": 0.6690184049079755,
+      "grad_norm": 0.15672078728675842,
+      "learning_rate": 3.770038418660699e-05,
+      "loss": 1.1123,
+      "step": 4362
+    },
+    {
+      "epoch": 0.6691717791411043,
+      "grad_norm": 0.1615694910287857,
+      "learning_rate": 3.766872096904729e-05,
+      "loss": 1.0545,
+      "step": 4363
+    },
+    {
+      "epoch": 0.6693251533742332,
+      "grad_norm": 0.14451193809509277,
+      "learning_rate": 3.7637066594090294e-05,
+      "loss": 0.9259,
+      "step": 4364
+    },
+    {
+      "epoch": 0.6694785276073619,
+      "grad_norm": 0.16119402647018433,
+      "learning_rate": 3.760542106923397e-05,
+      "loss": 1.0178,
+      "step": 4365
+    },
+    {
+      "epoch": 0.6696319018404908,
+      "grad_norm": 0.15624383091926575,
+      "learning_rate": 3.757378440197408e-05,
+      "loss": 1.047,
+      "step": 4366
+    },
+    {
+      "epoch": 0.6697852760736196,
+      "grad_norm": 0.16492727398872375,
+      "learning_rate": 3.7542156599804414e-05,
+      "loss": 1.0323,
+      "step": 4367
+    },
+    {
+      "epoch": 0.6699386503067485,
+      "grad_norm": 0.15404000878334045,
+      "learning_rate": 3.751053767021655e-05,
+      "loss": 1.0463,
+      "step": 4368
+    },
+    {
+      "epoch": 0.6700920245398773,
+      "grad_norm": 0.17345097661018372,
+      "learning_rate": 3.747892762070007e-05,
+      "loss": 1.254,
+      "step": 4369
+    },
+    {
+      "epoch": 0.6702453987730062,
+      "grad_norm": 0.1649002879858017,
+      "learning_rate": 3.7447326458742366e-05,
+      "loss": 1.0444,
+      "step": 4370
+    },
+    {
+      "epoch": 0.6703987730061349,
+      "grad_norm": 0.16894447803497314,
+      "learning_rate": 3.741573419182873e-05,
+      "loss": 1.2118,
+      "step": 4371
+    },
+    {
+      "epoch": 0.6705521472392638,
+      "grad_norm": 0.14763273298740387,
+      "learning_rate": 3.738415082744243e-05,
+      "loss": 0.9587,
+      "step": 4372
+    },
+    {
+      "epoch": 0.6707055214723926,
+      "grad_norm": 0.16178584098815918,
+      "learning_rate": 3.735257637306451e-05,
+      "loss": 1.1382,
+      "step": 4373
+    },
+    {
+      "epoch": 0.6708588957055215,
+      "grad_norm": 0.16589082777500153,
+      "learning_rate": 3.7321010836174023e-05,
+      "loss": 1.1781,
+      "step": 4374
+    },
+    {
+      "epoch": 0.6710122699386503,
+      "grad_norm": 0.1366875171661377,
+      "learning_rate": 3.7289454224247795e-05,
+      "loss": 0.885,
+      "step": 4375
+    },
+    {
+      "epoch": 0.6711656441717792,
+      "grad_norm": 0.1637764573097229,
+      "learning_rate": 3.725790654476061e-05,
+      "loss": 1.0593,
+      "step": 4376
+    },
+    {
+      "epoch": 0.6713190184049079,
+      "grad_norm": 0.16613249480724335,
+      "learning_rate": 3.722636780518516e-05,
+      "loss": 1.0903,
+      "step": 4377
+    },
+    {
+      "epoch": 0.6714723926380368,
+      "grad_norm": 0.1621791273355484,
+      "learning_rate": 3.7194838012991916e-05,
+      "loss": 1.1343,
+      "step": 4378
+    },
+    {
+      "epoch": 0.6716257668711656,
+      "grad_norm": 0.17521212995052338,
+      "learning_rate": 3.716331717564935e-05,
+      "loss": 1.0783,
+      "step": 4379
+    },
+    {
+      "epoch": 0.6717791411042945,
+      "grad_norm": 0.15942871570587158,
+      "learning_rate": 3.7131805300623685e-05,
+      "loss": 1.0605,
+      "step": 4380
+    },
+    {
+      "epoch": 0.6719325153374233,
+      "grad_norm": 0.1482733190059662,
+      "learning_rate": 3.7100302395379164e-05,
+      "loss": 0.9488,
+      "step": 4381
+    },
+    {
+      "epoch": 0.6720858895705522,
+      "grad_norm": 0.15223097801208496,
+      "learning_rate": 3.706880846737779e-05,
+      "loss": 0.9971,
+      "step": 4382
+    },
+    {
+      "epoch": 0.6722392638036809,
+      "grad_norm": 0.17094838619232178,
+      "learning_rate": 3.7037323524079466e-05,
+      "loss": 1.0845,
+      "step": 4383
+    },
+    {
+      "epoch": 0.6723926380368098,
+      "grad_norm": 0.1568271666765213,
+      "learning_rate": 3.700584757294202e-05,
+      "loss": 1.1109,
+      "step": 4384
+    },
+    {
+      "epoch": 0.6725460122699386,
+      "grad_norm": 0.15975546836853027,
+      "learning_rate": 3.697438062142108e-05,
+      "loss": 1.0672,
+      "step": 4385
+    },
+    {
+      "epoch": 0.6726993865030675,
+      "grad_norm": 0.14256882667541504,
+      "learning_rate": 3.6942922676970214e-05,
+      "loss": 0.8424,
+      "step": 4386
+    },
+    {
+      "epoch": 0.6728527607361963,
+      "grad_norm": 0.1582542061805725,
+      "learning_rate": 3.691147374704076e-05,
+      "loss": 1.0353,
+      "step": 4387
+    },
+    {
+      "epoch": 0.6730061349693252,
+      "grad_norm": 0.1701362282037735,
+      "learning_rate": 3.688003383908205e-05,
+      "loss": 1.159,
+      "step": 4388
+    },
+    {
+      "epoch": 0.673159509202454,
+      "grad_norm": 0.15822738409042358,
+      "learning_rate": 3.684860296054113e-05,
+      "loss": 1.171,
+      "step": 4389
+    },
+    {
+      "epoch": 0.6733128834355828,
+      "grad_norm": 0.1528329998254776,
+      "learning_rate": 3.681718111886306e-05,
+      "loss": 0.9974,
+      "step": 4390
+    },
+    {
+      "epoch": 0.6734662576687117,
+      "grad_norm": 0.16526983678340912,
+      "learning_rate": 3.678576832149064e-05,
+      "loss": 1.074,
+      "step": 4391
+    },
+    {
+      "epoch": 0.6736196319018405,
+      "grad_norm": 0.15962718427181244,
+      "learning_rate": 3.675436457586455e-05,
+      "loss": 1.1113,
+      "step": 4392
+    },
+    {
+      "epoch": 0.6737730061349694,
+      "grad_norm": 0.15394537150859833,
+      "learning_rate": 3.67229698894234e-05,
+      "loss": 1.0663,
+      "step": 4393
+    },
+    {
+      "epoch": 0.6739263803680982,
+      "grad_norm": 0.15282468497753143,
+      "learning_rate": 3.669158426960354e-05,
+      "loss": 1.0331,
+      "step": 4394
+    },
+    {
+      "epoch": 0.674079754601227,
+      "grad_norm": 0.15937189757823944,
+      "learning_rate": 3.66602077238393e-05,
+      "loss": 1.1543,
+      "step": 4395
+    },
+    {
+      "epoch": 0.6742331288343558,
+      "grad_norm": 0.14116787910461426,
+      "learning_rate": 3.662884025956273e-05,
+      "loss": 0.883,
+      "step": 4396
+    },
+    {
+      "epoch": 0.6743865030674847,
+      "grad_norm": 0.15346594154834747,
+      "learning_rate": 3.659748188420387e-05,
+      "loss": 1.0636,
+      "step": 4397
+    },
+    {
+      "epoch": 0.6745398773006135,
+      "grad_norm": 0.15183106064796448,
+      "learning_rate": 3.656613260519047e-05,
+      "loss": 0.9783,
+      "step": 4398
+    },
+    {
+      "epoch": 0.6746932515337424,
+      "grad_norm": 0.15648363530635834,
+      "learning_rate": 3.653479242994817e-05,
+      "loss": 1.113,
+      "step": 4399
+    },
+    {
+      "epoch": 0.6748466257668712,
+      "grad_norm": 0.16409996151924133,
+      "learning_rate": 3.6503461365900534e-05,
+      "loss": 1.0144,
+      "step": 4400
+    },
+    {
+      "epoch": 0.675,
+      "grad_norm": 0.14533723890781403,
+      "learning_rate": 3.647213942046884e-05,
+      "loss": 0.94,
+      "step": 4401
+    },
+    {
+      "epoch": 0.6751533742331288,
+      "grad_norm": 0.16307547688484192,
+      "learning_rate": 3.644082660107233e-05,
+      "loss": 1.1121,
+      "step": 4402
+    },
+    {
+      "epoch": 0.6753067484662577,
+      "grad_norm": 0.15523040294647217,
+      "learning_rate": 3.640952291512797e-05,
+      "loss": 1.1116,
+      "step": 4403
+    },
+    {
+      "epoch": 0.6754601226993865,
+      "grad_norm": 0.15901292860507965,
+      "learning_rate": 3.637822837005065e-05,
+      "loss": 1.0271,
+      "step": 4404
+    },
+    {
+      "epoch": 0.6756134969325154,
+      "grad_norm": 0.15642738342285156,
+      "learning_rate": 3.634694297325305e-05,
+      "loss": 1.1257,
+      "step": 4405
+    },
+    {
+      "epoch": 0.6757668711656442,
+      "grad_norm": 0.1544211357831955,
+      "learning_rate": 3.631566673214567e-05,
+      "loss": 1.0848,
+      "step": 4406
+    },
+    {
+      "epoch": 0.675920245398773,
+      "grad_norm": 0.1489080935716629,
+      "learning_rate": 3.628439965413692e-05,
+      "loss": 0.96,
+      "step": 4407
+    },
+    {
+      "epoch": 0.6760736196319018,
+      "grad_norm": 0.16093581914901733,
+      "learning_rate": 3.625314174663292e-05,
+      "loss": 1.019,
+      "step": 4408
+    },
+    {
+      "epoch": 0.6762269938650307,
+      "grad_norm": 0.18092598021030426,
+      "learning_rate": 3.6221893017037736e-05,
+      "loss": 1.1792,
+      "step": 4409
+    },
+    {
+      "epoch": 0.6763803680981595,
+      "grad_norm": 0.1508086770772934,
+      "learning_rate": 3.619065347275317e-05,
+      "loss": 0.922,
+      "step": 4410
+    },
+    {
+      "epoch": 0.6765337423312884,
+      "grad_norm": 0.16245847940444946,
+      "learning_rate": 3.615942312117891e-05,
+      "loss": 1.0056,
+      "step": 4411
+    },
+    {
+      "epoch": 0.6766871165644172,
+      "grad_norm": 0.1556919813156128,
+      "learning_rate": 3.612820196971242e-05,
+      "loss": 1.0328,
+      "step": 4412
+    },
+    {
+      "epoch": 0.676840490797546,
+      "grad_norm": 0.15338407456874847,
+      "learning_rate": 3.609699002574904e-05,
+      "loss": 1.0771,
+      "step": 4413
+    },
+    {
+      "epoch": 0.6769938650306748,
+      "grad_norm": 0.15298648178577423,
+      "learning_rate": 3.606578729668187e-05,
+      "loss": 1.0989,
+      "step": 4414
+    },
+    {
+      "epoch": 0.6771472392638037,
+      "grad_norm": 0.15607735514640808,
+      "learning_rate": 3.603459378990183e-05,
+      "loss": 1.0137,
+      "step": 4415
+    },
+    {
+      "epoch": 0.6773006134969325,
+      "grad_norm": 0.16778235137462616,
+      "learning_rate": 3.600340951279774e-05,
+      "loss": 1.1069,
+      "step": 4416
+    },
+    {
+      "epoch": 0.6774539877300614,
+      "grad_norm": 0.15295717120170593,
+      "learning_rate": 3.597223447275609e-05,
+      "loss": 1.0211,
+      "step": 4417
+    },
+    {
+      "epoch": 0.6776073619631902,
+      "grad_norm": 0.15631230175495148,
+      "learning_rate": 3.594106867716132e-05,
+      "loss": 1.0399,
+      "step": 4418
+    },
+    {
+      "epoch": 0.677760736196319,
+      "grad_norm": 0.15878616273403168,
+      "learning_rate": 3.590991213339564e-05,
+      "loss": 1.1702,
+      "step": 4419
+    },
+    {
+      "epoch": 0.6779141104294478,
+      "grad_norm": 0.15896347165107727,
+      "learning_rate": 3.5878764848839e-05,
+      "loss": 1.0841,
+      "step": 4420
+    },
+    {
+      "epoch": 0.6780674846625767,
+      "grad_norm": 0.15696169435977936,
+      "learning_rate": 3.584762683086926e-05,
+      "loss": 1.0865,
+      "step": 4421
+    },
+    {
+      "epoch": 0.6782208588957055,
+      "grad_norm": 0.14999616146087646,
+      "learning_rate": 3.581649808686199e-05,
+      "loss": 1.0101,
+      "step": 4422
+    },
+    {
+      "epoch": 0.6783742331288344,
+      "grad_norm": 0.14092575013637543,
+      "learning_rate": 3.5785378624190645e-05,
+      "loss": 0.8979,
+      "step": 4423
+    },
+    {
+      "epoch": 0.6785276073619632,
+      "grad_norm": 0.16287577152252197,
+      "learning_rate": 3.57542684502264e-05,
+      "loss": 1.0862,
+      "step": 4424
+    },
+    {
+      "epoch": 0.678680981595092,
+      "grad_norm": 0.15387265384197235,
+      "learning_rate": 3.572316757233833e-05,
+      "loss": 1.0003,
+      "step": 4425
+    },
+    {
+      "epoch": 0.6788343558282208,
+      "grad_norm": 0.15361180901527405,
+      "learning_rate": 3.569207599789323e-05,
+      "loss": 0.9866,
+      "step": 4426
+    },
+    {
+      "epoch": 0.6789877300613497,
+      "grad_norm": 0.15980538725852966,
+      "learning_rate": 3.566099373425568e-05,
+      "loss": 1.1696,
+      "step": 4427
+    },
+    {
+      "epoch": 0.6791411042944785,
+      "grad_norm": 0.1447245478630066,
+      "learning_rate": 3.562992078878813e-05,
+      "loss": 0.9018,
+      "step": 4428
+    },
+    {
+      "epoch": 0.6792944785276074,
+      "grad_norm": 0.15301081538200378,
+      "learning_rate": 3.5598857168850744e-05,
+      "loss": 1.0033,
+      "step": 4429
+    },
+    {
+      "epoch": 0.6794478527607362,
+      "grad_norm": 0.1619458943605423,
+      "learning_rate": 3.556780288180156e-05,
+      "loss": 1.0436,
+      "step": 4430
+    },
+    {
+      "epoch": 0.679601226993865,
+      "grad_norm": 0.1627204418182373,
+      "learning_rate": 3.553675793499631e-05,
+      "loss": 1.0715,
+      "step": 4431
+    },
+    {
+      "epoch": 0.6797546012269938,
+      "grad_norm": 0.15594378113746643,
+      "learning_rate": 3.5505722335788615e-05,
+      "loss": 1.0997,
+      "step": 4432
+    },
+    {
+      "epoch": 0.6799079754601227,
+      "grad_norm": 0.1646922081708908,
+      "learning_rate": 3.547469609152981e-05,
+      "loss": 1.0144,
+      "step": 4433
+    },
+    {
+      "epoch": 0.6800613496932515,
+      "grad_norm": 0.15654811263084412,
+      "learning_rate": 3.544367920956899e-05,
+      "loss": 1.017,
+      "step": 4434
+    },
+    {
+      "epoch": 0.6802147239263804,
+      "grad_norm": 0.14238181710243225,
+      "learning_rate": 3.541267169725314e-05,
+      "loss": 0.9931,
+      "step": 4435
+    },
+    {
+      "epoch": 0.6803680981595092,
+      "grad_norm": 0.14157864451408386,
+      "learning_rate": 3.5381673561926915e-05,
+      "loss": 0.9652,
+      "step": 4436
+    },
+    {
+      "epoch": 0.680521472392638,
+      "grad_norm": 0.15161658823490143,
+      "learning_rate": 3.535068481093283e-05,
+      "loss": 1.0131,
+      "step": 4437
+    },
+    {
+      "epoch": 0.6806748466257668,
+      "grad_norm": 0.15585848689079285,
+      "learning_rate": 3.531970545161111e-05,
+      "loss": 1.0327,
+      "step": 4438
+    },
+    {
+      "epoch": 0.6808282208588957,
+      "grad_norm": 0.15641841292381287,
+      "learning_rate": 3.5288735491299833e-05,
+      "loss": 1.1367,
+      "step": 4439
+    },
+    {
+      "epoch": 0.6809815950920245,
+      "grad_norm": 0.14978274703025818,
+      "learning_rate": 3.525777493733478e-05,
+      "loss": 0.9948,
+      "step": 4440
+    },
+    {
+      "epoch": 0.6811349693251534,
+      "grad_norm": 0.15914486348628998,
+      "learning_rate": 3.522682379704949e-05,
+      "loss": 1.102,
+      "step": 4441
+    },
+    {
+      "epoch": 0.6812883435582822,
+      "grad_norm": 0.1387794315814972,
+      "learning_rate": 3.519588207777539e-05,
+      "loss": 0.8648,
+      "step": 4442
+    },
+    {
+      "epoch": 0.681441717791411,
+      "grad_norm": 0.17018060386180878,
+      "learning_rate": 3.5164949786841526e-05,
+      "loss": 1.1236,
+      "step": 4443
+    },
+    {
+      "epoch": 0.6815950920245398,
+      "grad_norm": 0.1620759218931198,
+      "learning_rate": 3.513402693157485e-05,
+      "loss": 0.9986,
+      "step": 4444
+    },
+    {
+      "epoch": 0.6817484662576687,
+      "grad_norm": 0.16036604344844818,
+      "learning_rate": 3.510311351929993e-05,
+      "loss": 1.1102,
+      "step": 4445
+    },
+    {
+      "epoch": 0.6819018404907975,
+      "grad_norm": 0.16540485620498657,
+      "learning_rate": 3.507220955733927e-05,
+      "loss": 1.0851,
+      "step": 4446
+    },
+    {
+      "epoch": 0.6820552147239264,
+      "grad_norm": 0.1621365249156952,
+      "learning_rate": 3.504131505301296e-05,
+      "loss": 1.0719,
+      "step": 4447
+    },
+    {
+      "epoch": 0.6822085889570552,
+      "grad_norm": 0.15728580951690674,
+      "learning_rate": 3.501043001363901e-05,
+      "loss": 1.2175,
+      "step": 4448
+    },
+    {
+      "epoch": 0.682361963190184,
+      "grad_norm": 0.14690391719341278,
+      "learning_rate": 3.497955444653306e-05,
+      "loss": 1.009,
+      "step": 4449
+    },
+    {
+      "epoch": 0.6825153374233128,
+      "grad_norm": 0.14237406849861145,
+      "learning_rate": 3.494868835900856e-05,
+      "loss": 0.9339,
+      "step": 4450
+    },
+    {
+      "epoch": 0.6826687116564417,
+      "grad_norm": 0.13317032158374786,
+      "learning_rate": 3.4917831758376744e-05,
+      "loss": 0.8298,
+      "step": 4451
+    },
+    {
+      "epoch": 0.6828220858895706,
+      "grad_norm": 0.15216466784477234,
+      "learning_rate": 3.488698465194652e-05,
+      "loss": 1.111,
+      "step": 4452
+    },
+    {
+      "epoch": 0.6829754601226994,
+      "grad_norm": 0.17114469408988953,
+      "learning_rate": 3.485614704702465e-05,
+      "loss": 1.053,
+      "step": 4453
+    },
+    {
+      "epoch": 0.6831288343558283,
+      "grad_norm": 0.15447744727134705,
+      "learning_rate": 3.482531895091554e-05,
+      "loss": 1.0308,
+      "step": 4454
+    },
+    {
+      "epoch": 0.683282208588957,
+      "grad_norm": 0.1437186896800995,
+      "learning_rate": 3.479450037092144e-05,
+      "loss": 0.9274,
+      "step": 4455
+    },
+    {
+      "epoch": 0.6834355828220859,
+      "grad_norm": 0.16932901740074158,
+      "learning_rate": 3.4763691314342285e-05,
+      "loss": 1.105,
+      "step": 4456
+    },
+    {
+      "epoch": 0.6835889570552147,
+      "grad_norm": 0.16048173606395721,
+      "learning_rate": 3.473289178847572e-05,
+      "loss": 1.0219,
+      "step": 4457
+    },
+    {
+      "epoch": 0.6837423312883436,
+      "grad_norm": 0.15929318964481354,
+      "learning_rate": 3.470210180061726e-05,
+      "loss": 1.1404,
+      "step": 4458
+    },
+    {
+      "epoch": 0.6838957055214724,
+      "grad_norm": 0.15658126771450043,
+      "learning_rate": 3.467132135806002e-05,
+      "loss": 1.0771,
+      "step": 4459
+    },
+    {
+      "epoch": 0.6840490797546013,
+      "grad_norm": 0.1614944338798523,
+      "learning_rate": 3.464055046809496e-05,
+      "loss": 1.1445,
+      "step": 4460
+    },
+    {
+      "epoch": 0.68420245398773,
+      "grad_norm": 0.15099824965000153,
+      "learning_rate": 3.460978913801069e-05,
+      "loss": 1.0449,
+      "step": 4461
+    },
+    {
+      "epoch": 0.6843558282208589,
+      "grad_norm": 0.15100497007369995,
+      "learning_rate": 3.4579037375093616e-05,
+      "loss": 1.0768,
+      "step": 4462
+    },
+    {
+      "epoch": 0.6845092024539877,
+      "grad_norm": 0.15015797317028046,
+      "learning_rate": 3.45482951866279e-05,
+      "loss": 0.9712,
+      "step": 4463
+    },
+    {
+      "epoch": 0.6846625766871166,
+      "grad_norm": 0.15040357410907745,
+      "learning_rate": 3.4517562579895336e-05,
+      "loss": 1.0243,
+      "step": 4464
+    },
+    {
+      "epoch": 0.6848159509202454,
+      "grad_norm": 0.14931735396385193,
+      "learning_rate": 3.448683956217557e-05,
+      "loss": 1.038,
+      "step": 4465
+    },
+    {
+      "epoch": 0.6849693251533743,
+      "grad_norm": 0.15978765487670898,
+      "learning_rate": 3.445612614074585e-05,
+      "loss": 1.1732,
+      "step": 4466
+    },
+    {
+      "epoch": 0.685122699386503,
+      "grad_norm": 0.16654500365257263,
+      "learning_rate": 3.442542232288128e-05,
+      "loss": 1.1306,
+      "step": 4467
+    },
+    {
+      "epoch": 0.6852760736196319,
+      "grad_norm": 0.14673475921154022,
+      "learning_rate": 3.43947281158546e-05,
+      "loss": 1.0216,
+      "step": 4468
+    },
+    {
+      "epoch": 0.6854294478527607,
+      "grad_norm": 0.13886986672878265,
+      "learning_rate": 3.4364043526936265e-05,
+      "loss": 0.9947,
+      "step": 4469
+    },
+    {
+      "epoch": 0.6855828220858896,
+      "grad_norm": 0.1614229530096054,
+      "learning_rate": 3.433336856339455e-05,
+      "loss": 1.1129,
+      "step": 4470
+    },
+    {
+      "epoch": 0.6857361963190184,
+      "grad_norm": 0.15790164470672607,
+      "learning_rate": 3.4302703232495325e-05,
+      "loss": 1.0954,
+      "step": 4471
+    },
+    {
+      "epoch": 0.6858895705521473,
+      "grad_norm": 0.14804179966449738,
+      "learning_rate": 3.427204754150229e-05,
+      "loss": 1.0784,
+      "step": 4472
+    },
+    {
+      "epoch": 0.686042944785276,
+      "grad_norm": 0.20319785177707672,
+      "learning_rate": 3.424140149767678e-05,
+      "loss": 1.2192,
+      "step": 4473
+    },
+    {
+      "epoch": 0.6861963190184049,
+      "grad_norm": 0.15258800983428955,
+      "learning_rate": 3.42107651082779e-05,
+      "loss": 1.0075,
+      "step": 4474
+    },
+    {
+      "epoch": 0.6863496932515337,
+      "grad_norm": 0.164951354265213,
+      "learning_rate": 3.4180138380562436e-05,
+      "loss": 1.1474,
+      "step": 4475
+    },
+    {
+      "epoch": 0.6865030674846626,
+      "grad_norm": 0.15822334587574005,
+      "learning_rate": 3.414952132178487e-05,
+      "loss": 1.0622,
+      "step": 4476
+    },
+    {
+      "epoch": 0.6866564417177914,
+      "grad_norm": 0.15353894233703613,
+      "learning_rate": 3.4118913939197455e-05,
+      "loss": 1.0466,
+      "step": 4477
+    },
+    {
+      "epoch": 0.6868098159509203,
+      "grad_norm": 0.14440461993217468,
+      "learning_rate": 3.408831624005007e-05,
+      "loss": 0.9784,
+      "step": 4478
+    },
+    {
+      "epoch": 0.686963190184049,
+      "grad_norm": 0.1379581242799759,
+      "learning_rate": 3.40577282315904e-05,
+      "loss": 0.8763,
+      "step": 4479
+    },
+    {
+      "epoch": 0.6871165644171779,
+      "grad_norm": 0.1579783707857132,
+      "learning_rate": 3.402714992106372e-05,
+      "loss": 1.0595,
+      "step": 4480
+    },
+    {
+      "epoch": 0.6872699386503067,
+      "grad_norm": 0.15881548821926117,
+      "learning_rate": 3.399658131571314e-05,
+      "loss": 0.9989,
+      "step": 4481
+    },
+    {
+      "epoch": 0.6874233128834356,
+      "grad_norm": 0.1614791303873062,
+      "learning_rate": 3.396602242277935e-05,
+      "loss": 0.9857,
+      "step": 4482
+    },
+    {
+      "epoch": 0.6875766871165644,
+      "grad_norm": 0.1427900344133377,
+      "learning_rate": 3.393547324950077e-05,
+      "loss": 0.9545,
+      "step": 4483
+    },
+    {
+      "epoch": 0.6877300613496933,
+      "grad_norm": 0.1491352915763855,
+      "learning_rate": 3.390493380311359e-05,
+      "loss": 0.9799,
+      "step": 4484
+    },
+    {
+      "epoch": 0.6878834355828221,
+      "grad_norm": 0.1579512506723404,
+      "learning_rate": 3.387440409085158e-05,
+      "loss": 1.1441,
+      "step": 4485
+    },
+    {
+      "epoch": 0.688036809815951,
+      "grad_norm": 0.15027284622192383,
+      "learning_rate": 3.384388411994632e-05,
+      "loss": 0.9601,
+      "step": 4486
+    },
+    {
+      "epoch": 0.6881901840490797,
+      "grad_norm": 0.1577225923538208,
+      "learning_rate": 3.3813373897627e-05,
+      "loss": 1.0646,
+      "step": 4487
+    },
+    {
+      "epoch": 0.6883435582822086,
+      "grad_norm": 0.16313867270946503,
+      "learning_rate": 3.378287343112055e-05,
+      "loss": 1.1646,
+      "step": 4488
+    },
+    {
+      "epoch": 0.6884969325153374,
+      "grad_norm": 0.17674291133880615,
+      "learning_rate": 3.375238272765154e-05,
+      "loss": 1.2706,
+      "step": 4489
+    },
+    {
+      "epoch": 0.6886503067484663,
+      "grad_norm": 0.15048839151859283,
+      "learning_rate": 3.3721901794442294e-05,
+      "loss": 1.0547,
+      "step": 4490
+    },
+    {
+      "epoch": 0.6888036809815951,
+      "grad_norm": 0.15365725755691528,
+      "learning_rate": 3.369143063871275e-05,
+      "loss": 1.0995,
+      "step": 4491
+    },
+    {
+      "epoch": 0.688957055214724,
+      "grad_norm": 0.16561256349086761,
+      "learning_rate": 3.366096926768056e-05,
+      "loss": 1.1055,
+      "step": 4492
+    },
+    {
+      "epoch": 0.6891104294478527,
+      "grad_norm": 0.15923088788986206,
+      "learning_rate": 3.363051768856111e-05,
+      "loss": 1.1382,
+      "step": 4493
+    },
+    {
+      "epoch": 0.6892638036809816,
+      "grad_norm": 0.15853071212768555,
+      "learning_rate": 3.360007590856735e-05,
+      "loss": 1.0736,
+      "step": 4494
+    },
+    {
+      "epoch": 0.6894171779141104,
+      "grad_norm": 0.22687534987926483,
+      "learning_rate": 3.356964393491005e-05,
+      "loss": 1.0542,
+      "step": 4495
+    },
+    {
+      "epoch": 0.6895705521472393,
+      "grad_norm": 0.1664365977048874,
+      "learning_rate": 3.353922177479751e-05,
+      "loss": 1.1319,
+      "step": 4496
+    },
+    {
+      "epoch": 0.6897239263803681,
+      "grad_norm": 0.14696256816387177,
+      "learning_rate": 3.350880943543586e-05,
+      "loss": 1.048,
+      "step": 4497
+    },
+    {
+      "epoch": 0.689877300613497,
+      "grad_norm": 0.15227681398391724,
+      "learning_rate": 3.3478406924028775e-05,
+      "loss": 1.0225,
+      "step": 4498
+    },
+    {
+      "epoch": 0.6900306748466257,
+      "grad_norm": 0.16218845546245575,
+      "learning_rate": 3.3448014247777636e-05,
+      "loss": 1.0069,
+      "step": 4499
+    },
+    {
+      "epoch": 0.6901840490797546,
+      "grad_norm": 0.16484732925891876,
+      "learning_rate": 3.3417631413881563e-05,
+      "loss": 1.1109,
+      "step": 4500
+    },
+    {
+      "epoch": 0.6903374233128834,
+      "grad_norm": 0.14668594300746918,
+      "learning_rate": 3.3387258429537234e-05,
+      "loss": 0.9861,
+      "step": 4501
+    },
+    {
+      "epoch": 0.6904907975460123,
+      "grad_norm": 0.14890913665294647,
+      "learning_rate": 3.3356895301939096e-05,
+      "loss": 1.0858,
+      "step": 4502
+    },
+    {
+      "epoch": 0.6906441717791411,
+      "grad_norm": 0.31909292936325073,
+      "learning_rate": 3.332654203827918e-05,
+      "loss": 1.0747,
+      "step": 4503
+    },
+    {
+      "epoch": 0.69079754601227,
+      "grad_norm": 0.15669703483581543,
+      "learning_rate": 3.329619864574723e-05,
+      "loss": 1.1106,
+      "step": 4504
+    },
+    {
+      "epoch": 0.6909509202453987,
+      "grad_norm": 0.15857093036174774,
+      "learning_rate": 3.3265865131530685e-05,
+      "loss": 1.084,
+      "step": 4505
+    },
+    {
+      "epoch": 0.6911042944785276,
+      "grad_norm": 0.14514023065567017,
+      "learning_rate": 3.323554150281452e-05,
+      "loss": 0.9396,
+      "step": 4506
+    },
+    {
+      "epoch": 0.6912576687116564,
+      "grad_norm": 0.15473279356956482,
+      "learning_rate": 3.3205227766781514e-05,
+      "loss": 0.9898,
+      "step": 4507
+    },
+    {
+      "epoch": 0.6914110429447853,
+      "grad_norm": 0.161356583237648,
+      "learning_rate": 3.3174923930611975e-05,
+      "loss": 1.0523,
+      "step": 4508
+    },
+    {
+      "epoch": 0.6915644171779141,
+      "grad_norm": 0.16521796584129333,
+      "learning_rate": 3.314463000148399e-05,
+      "loss": 1.0627,
+      "step": 4509
+    },
+    {
+      "epoch": 0.691717791411043,
+      "grad_norm": 0.14920157194137573,
+      "learning_rate": 3.31143459865732e-05,
+      "loss": 1.0348,
+      "step": 4510
+    },
+    {
+      "epoch": 0.6918711656441717,
+      "grad_norm": 0.15608203411102295,
+      "learning_rate": 3.30840718930529e-05,
+      "loss": 1.0339,
+      "step": 4511
+    },
+    {
+      "epoch": 0.6920245398773006,
+      "grad_norm": 0.15225042402744293,
+      "learning_rate": 3.3053807728094144e-05,
+      "loss": 1.0539,
+      "step": 4512
+    },
+    {
+      "epoch": 0.6921779141104294,
+      "grad_norm": 0.16419251263141632,
+      "learning_rate": 3.3023553498865473e-05,
+      "loss": 1.0912,
+      "step": 4513
+    },
+    {
+      "epoch": 0.6923312883435583,
+      "grad_norm": 0.15743084251880646,
+      "learning_rate": 3.299330921253325e-05,
+      "loss": 1.1383,
+      "step": 4514
+    },
+    {
+      "epoch": 0.6924846625766872,
+      "grad_norm": 0.1692809760570526,
+      "learning_rate": 3.296307487626131e-05,
+      "loss": 1.055,
+      "step": 4515
+    },
+    {
+      "epoch": 0.692638036809816,
+      "grad_norm": 0.1548701524734497,
+      "learning_rate": 3.293285049721127e-05,
+      "loss": 1.0254,
+      "step": 4516
+    },
+    {
+      "epoch": 0.6927914110429448,
+      "grad_norm": 0.1448838710784912,
+      "learning_rate": 3.290263608254232e-05,
+      "loss": 0.9088,
+      "step": 4517
+    },
+    {
+      "epoch": 0.6929447852760736,
+      "grad_norm": 0.16154997050762177,
+      "learning_rate": 3.2872431639411266e-05,
+      "loss": 1.0547,
+      "step": 4518
+    },
+    {
+      "epoch": 0.6930981595092025,
+      "grad_norm": 0.15312322974205017,
+      "learning_rate": 3.284223717497265e-05,
+      "loss": 1.0946,
+      "step": 4519
+    },
+    {
+      "epoch": 0.6932515337423313,
+      "grad_norm": 0.14747284352779388,
+      "learning_rate": 3.281205269637852e-05,
+      "loss": 1.0071,
+      "step": 4520
+    },
+    {
+      "epoch": 0.6934049079754602,
+      "grad_norm": 0.1681765764951706,
+      "learning_rate": 3.278187821077869e-05,
+      "loss": 1.0567,
+      "step": 4521
+    },
+    {
+      "epoch": 0.693558282208589,
+      "grad_norm": 0.1455790102481842,
+      "learning_rate": 3.2751713725320495e-05,
+      "loss": 0.9824,
+      "step": 4522
+    },
+    {
+      "epoch": 0.6937116564417178,
+      "grad_norm": 0.14128927886486053,
+      "learning_rate": 3.2721559247149e-05,
+      "loss": 0.9215,
+      "step": 4523
+    },
+    {
+      "epoch": 0.6938650306748466,
+      "grad_norm": 0.14747948944568634,
+      "learning_rate": 3.269141478340678e-05,
+      "loss": 0.9796,
+      "step": 4524
+    },
+    {
+      "epoch": 0.6940184049079755,
+      "grad_norm": 0.15070101618766785,
+      "learning_rate": 3.2661280341234186e-05,
+      "loss": 1.044,
+      "step": 4525
+    },
+    {
+      "epoch": 0.6941717791411043,
+      "grad_norm": 0.1579710841178894,
+      "learning_rate": 3.2631155927769076e-05,
+      "loss": 1.0783,
+      "step": 4526
+    },
+    {
+      "epoch": 0.6943251533742332,
+      "grad_norm": 0.15029607713222504,
+      "learning_rate": 3.260104155014696e-05,
+      "loss": 1.1055,
+      "step": 4527
+    },
+    {
+      "epoch": 0.694478527607362,
+      "grad_norm": 0.15994788706302643,
+      "learning_rate": 3.257093721550102e-05,
+      "loss": 1.0742,
+      "step": 4528
+    },
+    {
+      "epoch": 0.6946319018404908,
+      "grad_norm": 0.16349707543849945,
+      "learning_rate": 3.254084293096198e-05,
+      "loss": 1.0527,
+      "step": 4529
+    },
+    {
+      "epoch": 0.6947852760736196,
+      "grad_norm": 0.1601266711950302,
+      "learning_rate": 3.251075870365829e-05,
+      "loss": 1.111,
+      "step": 4530
+    },
+    {
+      "epoch": 0.6949386503067485,
+      "grad_norm": 0.16360405087471008,
+      "learning_rate": 3.248068454071588e-05,
+      "loss": 1.0462,
+      "step": 4531
+    },
+    {
+      "epoch": 0.6950920245398773,
+      "grad_norm": 0.16377083957195282,
+      "learning_rate": 3.245062044925844e-05,
+      "loss": 1.095,
+      "step": 4532
+    },
+    {
+      "epoch": 0.6952453987730062,
+      "grad_norm": 0.15840598940849304,
+      "learning_rate": 3.242056643640717e-05,
+      "loss": 1.0607,
+      "step": 4533
+    },
+    {
+      "epoch": 0.695398773006135,
+      "grad_norm": 0.1580493152141571,
+      "learning_rate": 3.239052250928089e-05,
+      "loss": 1.1147,
+      "step": 4534
+    },
+    {
+      "epoch": 0.6955521472392638,
+      "grad_norm": 0.1524648517370224,
+      "learning_rate": 3.236048867499612e-05,
+      "loss": 1.0105,
+      "step": 4535
+    },
+    {
+      "epoch": 0.6957055214723926,
+      "grad_norm": 0.15492001175880432,
+      "learning_rate": 3.2330464940666876e-05,
+      "loss": 1.0072,
+      "step": 4536
+    },
+    {
+      "epoch": 0.6958588957055215,
+      "grad_norm": 0.16926118731498718,
+      "learning_rate": 3.230045131340489e-05,
+      "loss": 1.1446,
+      "step": 4537
+    },
+    {
+      "epoch": 0.6960122699386503,
+      "grad_norm": 0.17113718390464783,
+      "learning_rate": 3.227044780031937e-05,
+      "loss": 1.2053,
+      "step": 4538
+    },
+    {
+      "epoch": 0.6961656441717792,
+      "grad_norm": 0.158729687333107,
+      "learning_rate": 3.2240454408517267e-05,
+      "loss": 1.0551,
+      "step": 4539
+    },
+    {
+      "epoch": 0.696319018404908,
+      "grad_norm": 0.14953166246414185,
+      "learning_rate": 3.221047114510305e-05,
+      "loss": 0.9851,
+      "step": 4540
+    },
+    {
+      "epoch": 0.6964723926380368,
+      "grad_norm": 0.17053116858005524,
+      "learning_rate": 3.218049801717878e-05,
+      "loss": 1.2047,
+      "step": 4541
+    },
+    {
+      "epoch": 0.6966257668711656,
+      "grad_norm": 0.16179822385311127,
+      "learning_rate": 3.215053503184419e-05,
+      "loss": 1.0643,
+      "step": 4542
+    },
+    {
+      "epoch": 0.6967791411042945,
+      "grad_norm": 0.15844307839870453,
+      "learning_rate": 3.2120582196196517e-05,
+      "loss": 1.0148,
+      "step": 4543
+    },
+    {
+      "epoch": 0.6969325153374233,
+      "grad_norm": 0.15230192244052887,
+      "learning_rate": 3.20906395173307e-05,
+      "loss": 1.0255,
+      "step": 4544
+    },
+    {
+      "epoch": 0.6970858895705522,
+      "grad_norm": 0.16097067296504974,
+      "learning_rate": 3.206070700233916e-05,
+      "loss": 1.1546,
+      "step": 4545
+    },
+    {
+      "epoch": 0.697239263803681,
+      "grad_norm": 0.15703046321868896,
+      "learning_rate": 3.2030784658311995e-05,
+      "loss": 1.0859,
+      "step": 4546
+    },
+    {
+      "epoch": 0.6973926380368098,
+      "grad_norm": 0.15658466517925262,
+      "learning_rate": 3.200087249233689e-05,
+      "loss": 1.066,
+      "step": 4547
+    },
+    {
+      "epoch": 0.6975460122699386,
+      "grad_norm": 0.14781983196735382,
+      "learning_rate": 3.197097051149903e-05,
+      "loss": 0.8747,
+      "step": 4548
+    },
+    {
+      "epoch": 0.6976993865030675,
+      "grad_norm": 0.1500227004289627,
+      "learning_rate": 3.194107872288132e-05,
+      "loss": 0.9446,
+      "step": 4549
+    },
+    {
+      "epoch": 0.6978527607361963,
+      "grad_norm": 0.15879333019256592,
+      "learning_rate": 3.1911197133564125e-05,
+      "loss": 0.9927,
+      "step": 4550
+    },
+    {
+      "epoch": 0.6980061349693252,
+      "grad_norm": 0.14226990938186646,
+      "learning_rate": 3.1881325750625505e-05,
+      "loss": 0.9821,
+      "step": 4551
+    },
+    {
+      "epoch": 0.698159509202454,
+      "grad_norm": 0.17220234870910645,
+      "learning_rate": 3.185146458114101e-05,
+      "loss": 1.3277,
+      "step": 4552
+    },
+    {
+      "epoch": 0.6983128834355828,
+      "grad_norm": 0.1532985270023346,
+      "learning_rate": 3.1821613632183795e-05,
+      "loss": 0.954,
+      "step": 4553
+    },
+    {
+      "epoch": 0.6984662576687116,
+      "grad_norm": 0.15497009456157684,
+      "learning_rate": 3.1791772910824665e-05,
+      "loss": 1.0143,
+      "step": 4554
+    },
+    {
+      "epoch": 0.6986196319018405,
+      "grad_norm": 0.14855705201625824,
+      "learning_rate": 3.176194242413188e-05,
+      "loss": 0.9297,
+      "step": 4555
+    },
+    {
+      "epoch": 0.6987730061349693,
+      "grad_norm": 0.14983654022216797,
+      "learning_rate": 3.17321221791714e-05,
+      "loss": 1.0209,
+      "step": 4556
+    },
+    {
+      "epoch": 0.6989263803680982,
+      "grad_norm": 0.15649889409542084,
+      "learning_rate": 3.1702312183006655e-05,
+      "loss": 1.1013,
+      "step": 4557
+    },
+    {
+      "epoch": 0.699079754601227,
+      "grad_norm": 0.14873401820659637,
+      "learning_rate": 3.1672512442698726e-05,
+      "loss": 1.0751,
+      "step": 4558
+    },
+    {
+      "epoch": 0.6992331288343558,
+      "grad_norm": 0.15112358331680298,
+      "learning_rate": 3.1642722965306205e-05,
+      "loss": 1.0204,
+      "step": 4559
+    },
+    {
+      "epoch": 0.6993865030674846,
+      "grad_norm": 0.15726788341999054,
+      "learning_rate": 3.161294375788532e-05,
+      "loss": 1.0945,
+      "step": 4560
+    },
+    {
+      "epoch": 0.6995398773006135,
+      "grad_norm": 0.14182141423225403,
+      "learning_rate": 3.1583174827489785e-05,
+      "loss": 0.9292,
+      "step": 4561
+    },
+    {
+      "epoch": 0.6996932515337423,
+      "grad_norm": 0.1686347872018814,
+      "learning_rate": 3.155341618117092e-05,
+      "loss": 1.2383,
+      "step": 4562
+    },
+    {
+      "epoch": 0.6998466257668712,
+      "grad_norm": 0.16628539562225342,
+      "learning_rate": 3.1523667825977654e-05,
+      "loss": 1.0322,
+      "step": 4563
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.163345068693161,
+      "learning_rate": 3.149392976895637e-05,
+      "loss": 1.0649,
+      "step": 4564
+    },
+    {
+      "epoch": 0.7001533742331288,
+      "grad_norm": 0.14128173887729645,
+      "learning_rate": 3.1464202017151145e-05,
+      "loss": 0.9073,
+      "step": 4565
+    },
+    {
+      "epoch": 0.7003067484662576,
+      "grad_norm": 0.16104736924171448,
+      "learning_rate": 3.143448457760348e-05,
+      "loss": 1.0665,
+      "step": 4566
+    },
+    {
+      "epoch": 0.7004601226993865,
+      "grad_norm": 0.15005414187908173,
+      "learning_rate": 3.140477745735256e-05,
+      "loss": 1.0485,
+      "step": 4567
+    },
+    {
+      "epoch": 0.7006134969325153,
+      "grad_norm": 0.1540173590183258,
+      "learning_rate": 3.1375080663435035e-05,
+      "loss": 1.0039,
+      "step": 4568
+    },
+    {
+      "epoch": 0.7007668711656442,
+      "grad_norm": 0.15845166146755219,
+      "learning_rate": 3.1345394202885116e-05,
+      "loss": 1.0492,
+      "step": 4569
+    },
+    {
+      "epoch": 0.700920245398773,
+      "grad_norm": 0.15778520703315735,
+      "learning_rate": 3.1315718082734646e-05,
+      "loss": 0.9739,
+      "step": 4570
+    },
+    {
+      "epoch": 0.7010736196319018,
+      "grad_norm": 0.15788094699382782,
+      "learning_rate": 3.1286052310012894e-05,
+      "loss": 1.0355,
+      "step": 4571
+    },
+    {
+      "epoch": 0.7012269938650306,
+      "grad_norm": 0.1500125378370285,
+      "learning_rate": 3.125639689174681e-05,
+      "loss": 1.0316,
+      "step": 4572
+    },
+    {
+      "epoch": 0.7013803680981595,
+      "grad_norm": 0.1518303006887436,
+      "learning_rate": 3.122675183496077e-05,
+      "loss": 1.0596,
+      "step": 4573
+    },
+    {
+      "epoch": 0.7015337423312883,
+      "grad_norm": 0.13943350315093994,
+      "learning_rate": 3.119711714667681e-05,
+      "loss": 0.9514,
+      "step": 4574
+    },
+    {
+      "epoch": 0.7016871165644172,
+      "grad_norm": 0.16341152787208557,
+      "learning_rate": 3.116749283391443e-05,
+      "loss": 1.0353,
+      "step": 4575
+    },
+    {
+      "epoch": 0.701840490797546,
+      "grad_norm": 0.1640615016222,
+      "learning_rate": 3.113787890369066e-05,
+      "loss": 1.043,
+      "step": 4576
+    },
+    {
+      "epoch": 0.7019938650306748,
+      "grad_norm": 0.14637258648872375,
+      "learning_rate": 3.110827536302017e-05,
+      "loss": 0.9876,
+      "step": 4577
+    },
+    {
+      "epoch": 0.7021472392638037,
+      "grad_norm": 0.16557632386684418,
+      "learning_rate": 3.107868221891504e-05,
+      "loss": 1.0879,
+      "step": 4578
+    },
+    {
+      "epoch": 0.7023006134969325,
+      "grad_norm": 0.15702037513256073,
+      "learning_rate": 3.104909947838501e-05,
+      "loss": 1.1215,
+      "step": 4579
+    },
+    {
+      "epoch": 0.7024539877300614,
+      "grad_norm": 0.15598827600479126,
+      "learning_rate": 3.101952714843724e-05,
+      "loss": 1.0114,
+      "step": 4580
+    },
+    {
+      "epoch": 0.7026073619631902,
+      "grad_norm": 0.15218444168567657,
+      "learning_rate": 3.098996523607655e-05,
+      "loss": 0.9487,
+      "step": 4581
+    },
+    {
+      "epoch": 0.7027607361963191,
+      "grad_norm": 0.16236959397792816,
+      "learning_rate": 3.096041374830516e-05,
+      "loss": 1.0553,
+      "step": 4582
+    },
+    {
+      "epoch": 0.7029141104294478,
+      "grad_norm": 0.1628352701663971,
+      "learning_rate": 3.093087269212294e-05,
+      "loss": 1.1178,
+      "step": 4583
+    },
+    {
+      "epoch": 0.7030674846625767,
+      "grad_norm": 0.16686590015888214,
+      "learning_rate": 3.090134207452722e-05,
+      "loss": 1.1157,
+      "step": 4584
+    },
+    {
+      "epoch": 0.7032208588957055,
+      "grad_norm": 0.14539609849452972,
+      "learning_rate": 3.087182190251283e-05,
+      "loss": 0.9679,
+      "step": 4585
+    },
+    {
+      "epoch": 0.7033742331288344,
+      "grad_norm": 0.15306894481182098,
+      "learning_rate": 3.084231218307222e-05,
+      "loss": 0.9544,
+      "step": 4586
+    },
+    {
+      "epoch": 0.7035276073619632,
+      "grad_norm": 0.14341360330581665,
+      "learning_rate": 3.0812812923195274e-05,
+      "loss": 0.9818,
+      "step": 4587
+    },
+    {
+      "epoch": 0.7036809815950921,
+      "grad_norm": 0.16655504703521729,
+      "learning_rate": 3.078332412986944e-05,
+      "loss": 1.1512,
+      "step": 4588
+    },
+    {
+      "epoch": 0.7038343558282208,
+      "grad_norm": 0.16586560010910034,
+      "learning_rate": 3.0753845810079746e-05,
+      "loss": 1.0848,
+      "step": 4589
+    },
+    {
+      "epoch": 0.7039877300613497,
+      "grad_norm": 0.1697733998298645,
+      "learning_rate": 3.072437797080859e-05,
+      "loss": 1.2012,
+      "step": 4590
+    },
+    {
+      "epoch": 0.7041411042944785,
+      "grad_norm": 0.1661807745695114,
+      "learning_rate": 3.0694920619036036e-05,
+      "loss": 0.9674,
+      "step": 4591
+    },
+    {
+      "epoch": 0.7042944785276074,
+      "grad_norm": 0.16604669392108917,
+      "learning_rate": 3.066547376173955e-05,
+      "loss": 1.0435,
+      "step": 4592
+    },
+    {
+      "epoch": 0.7044478527607362,
+      "grad_norm": 0.1620442271232605,
+      "learning_rate": 3.0636037405894215e-05,
+      "loss": 1.1017,
+      "step": 4593
+    },
+    {
+      "epoch": 0.7046012269938651,
+      "grad_norm": 0.15877309441566467,
+      "learning_rate": 3.0606611558472535e-05,
+      "loss": 1.0654,
+      "step": 4594
+    },
+    {
+      "epoch": 0.7047546012269938,
+      "grad_norm": 0.1504252552986145,
+      "learning_rate": 3.05771962264446e-05,
+      "loss": 1.0069,
+      "step": 4595
+    },
+    {
+      "epoch": 0.7049079754601227,
+      "grad_norm": 0.1422571986913681,
+      "learning_rate": 3.054779141677796e-05,
+      "loss": 0.8488,
+      "step": 4596
+    },
+    {
+      "epoch": 0.7050613496932515,
+      "grad_norm": 0.17338404059410095,
+      "learning_rate": 3.0518397136437656e-05,
+      "loss": 1.0842,
+      "step": 4597
+    },
+    {
+      "epoch": 0.7052147239263804,
+      "grad_norm": 0.16005489230155945,
+      "learning_rate": 3.0489013392386323e-05,
+      "loss": 1.1629,
+      "step": 4598
+    },
+    {
+      "epoch": 0.7053680981595092,
+      "grad_norm": 0.15240567922592163,
+      "learning_rate": 3.0459640191583977e-05,
+      "loss": 1.0349,
+      "step": 4599
+    },
+    {
+      "epoch": 0.7055214723926381,
+      "grad_norm": 0.14655594527721405,
+      "learning_rate": 3.0430277540988277e-05,
+      "loss": 0.9056,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7056748466257668,
+      "grad_norm": 0.16694054007530212,
+      "learning_rate": 3.0400925447554236e-05,
+      "loss": 1.1583,
+      "step": 4601
+    },
+    {
+      "epoch": 0.7058282208588957,
+      "grad_norm": 0.1573842316865921,
+      "learning_rate": 3.0371583918234512e-05,
+      "loss": 0.9738,
+      "step": 4602
+    },
+    {
+      "epoch": 0.7059815950920245,
+      "grad_norm": 0.17175859212875366,
+      "learning_rate": 3.0342252959979137e-05,
+      "loss": 1.0574,
+      "step": 4603
+    },
+    {
+      "epoch": 0.7061349693251534,
+      "grad_norm": 0.16460727155208588,
+      "learning_rate": 3.031293257973569e-05,
+      "loss": 1.1182,
+      "step": 4604
+    },
+    {
+      "epoch": 0.7062883435582822,
+      "grad_norm": 0.14960789680480957,
+      "learning_rate": 3.0283622784449284e-05,
+      "loss": 0.9722,
+      "step": 4605
+    },
+    {
+      "epoch": 0.7064417177914111,
+      "grad_norm": 0.15768195688724518,
+      "learning_rate": 3.0254323581062426e-05,
+      "loss": 0.937,
+      "step": 4606
+    },
+    {
+      "epoch": 0.7065950920245399,
+      "grad_norm": 0.16398966312408447,
+      "learning_rate": 3.0225034976515234e-05,
+      "loss": 0.9347,
+      "step": 4607
+    },
+    {
+      "epoch": 0.7067484662576687,
+      "grad_norm": 0.14992304146289825,
+      "learning_rate": 3.0195756977745204e-05,
+      "loss": 0.9663,
+      "step": 4608
+    },
+    {
+      "epoch": 0.7069018404907975,
+      "grad_norm": 0.17738746106624603,
+      "learning_rate": 3.016648959168742e-05,
+      "loss": 1.1982,
+      "step": 4609
+    },
+    {
+      "epoch": 0.7070552147239264,
+      "grad_norm": 0.15887269377708435,
+      "learning_rate": 3.0137232825274373e-05,
+      "loss": 1.0997,
+      "step": 4610
+    },
+    {
+      "epoch": 0.7072085889570552,
+      "grad_norm": 0.15159371495246887,
+      "learning_rate": 3.0107986685436048e-05,
+      "loss": 0.9914,
+      "step": 4611
+    },
+    {
+      "epoch": 0.7073619631901841,
+      "grad_norm": 0.1594579666852951,
+      "learning_rate": 3.0078751179099984e-05,
+      "loss": 1.0763,
+      "step": 4612
+    },
+    {
+      "epoch": 0.7075153374233129,
+      "grad_norm": 0.1532166600227356,
+      "learning_rate": 3.00495263131911e-05,
+      "loss": 1.005,
+      "step": 4613
+    },
+    {
+      "epoch": 0.7076687116564417,
+      "grad_norm": 0.16191327571868896,
+      "learning_rate": 3.0020312094631894e-05,
+      "loss": 1.0989,
+      "step": 4614
+    },
+    {
+      "epoch": 0.7078220858895705,
+      "grad_norm": 0.1549447625875473,
+      "learning_rate": 2.9991108530342246e-05,
+      "loss": 1.018,
+      "step": 4615
+    },
+    {
+      "epoch": 0.7079754601226994,
+      "grad_norm": 0.16356804966926575,
+      "learning_rate": 2.9961915627239596e-05,
+      "loss": 1.1295,
+      "step": 4616
+    },
+    {
+      "epoch": 0.7081288343558282,
+      "grad_norm": 0.157591313123703,
+      "learning_rate": 2.9932733392238792e-05,
+      "loss": 1.0335,
+      "step": 4617
+    },
+    {
+      "epoch": 0.7082822085889571,
+      "grad_norm": 0.1550099402666092,
+      "learning_rate": 2.990356183225222e-05,
+      "loss": 0.9423,
+      "step": 4618
+    },
+    {
+      "epoch": 0.7084355828220859,
+      "grad_norm": 0.16547532379627228,
+      "learning_rate": 2.987440095418969e-05,
+      "loss": 1.1155,
+      "step": 4619
+    },
+    {
+      "epoch": 0.7085889570552147,
+      "grad_norm": 0.15183918178081512,
+      "learning_rate": 2.9845250764958457e-05,
+      "loss": 0.9642,
+      "step": 4620
+    },
+    {
+      "epoch": 0.7087423312883435,
+      "grad_norm": 0.15749786794185638,
+      "learning_rate": 2.9816111271463337e-05,
+      "loss": 1.1109,
+      "step": 4621
+    },
+    {
+      "epoch": 0.7088957055214724,
+      "grad_norm": 0.1556425392627716,
+      "learning_rate": 2.978698248060651e-05,
+      "loss": 0.9614,
+      "step": 4622
+    },
+    {
+      "epoch": 0.7090490797546012,
+      "grad_norm": 0.14377060532569885,
+      "learning_rate": 2.9757864399287713e-05,
+      "loss": 0.94,
+      "step": 4623
+    },
+    {
+      "epoch": 0.7092024539877301,
+      "grad_norm": 0.16300755739212036,
+      "learning_rate": 2.9728757034404054e-05,
+      "loss": 1.0629,
+      "step": 4624
+    },
+    {
+      "epoch": 0.7093558282208589,
+      "grad_norm": 0.15461722016334534,
+      "learning_rate": 2.9699660392850205e-05,
+      "loss": 1.0131,
+      "step": 4625
+    },
+    {
+      "epoch": 0.7095092024539877,
+      "grad_norm": 0.1388702243566513,
+      "learning_rate": 2.967057448151821e-05,
+      "loss": 0.9085,
+      "step": 4626
+    },
+    {
+      "epoch": 0.7096625766871165,
+      "grad_norm": 0.162318617105484,
+      "learning_rate": 2.964149930729757e-05,
+      "loss": 1.092,
+      "step": 4627
+    },
+    {
+      "epoch": 0.7098159509202454,
+      "grad_norm": 0.16493630409240723,
+      "learning_rate": 2.9612434877075347e-05,
+      "loss": 1.0623,
+      "step": 4628
+    },
+    {
+      "epoch": 0.7099693251533742,
+      "grad_norm": 0.14439798891544342,
+      "learning_rate": 2.9583381197735935e-05,
+      "loss": 1.0149,
+      "step": 4629
+    },
+    {
+      "epoch": 0.7101226993865031,
+      "grad_norm": 0.16714046895503998,
+      "learning_rate": 2.9554338276161265e-05,
+      "loss": 1.0763,
+      "step": 4630
+    },
+    {
+      "epoch": 0.7102760736196319,
+      "grad_norm": 0.15250417590141296,
+      "learning_rate": 2.9525306119230665e-05,
+      "loss": 1.1088,
+      "step": 4631
+    },
+    {
+      "epoch": 0.7104294478527607,
+      "grad_norm": 0.15543603897094727,
+      "learning_rate": 2.949628473382094e-05,
+      "loss": 0.9612,
+      "step": 4632
+    },
+    {
+      "epoch": 0.7105828220858895,
+      "grad_norm": 0.17160038650035858,
+      "learning_rate": 2.9467274126806382e-05,
+      "loss": 1.0943,
+      "step": 4633
+    },
+    {
+      "epoch": 0.7107361963190184,
+      "grad_norm": 0.16398188471794128,
+      "learning_rate": 2.9438274305058633e-05,
+      "loss": 1.0806,
+      "step": 4634
+    },
+    {
+      "epoch": 0.7108895705521472,
+      "grad_norm": 0.1650051772594452,
+      "learning_rate": 2.940928527544688e-05,
+      "loss": 1.185,
+      "step": 4635
+    },
+    {
+      "epoch": 0.7110429447852761,
+      "grad_norm": 0.14938130974769592,
+      "learning_rate": 2.9380307044837672e-05,
+      "loss": 0.8723,
+      "step": 4636
+    },
+    {
+      "epoch": 0.7111963190184049,
+      "grad_norm": 0.1491820216178894,
+      "learning_rate": 2.9351339620095082e-05,
+      "loss": 1.0868,
+      "step": 4637
+    },
+    {
+      "epoch": 0.7113496932515337,
+      "grad_norm": 0.15607237815856934,
+      "learning_rate": 2.9322383008080555e-05,
+      "loss": 0.9813,
+      "step": 4638
+    },
+    {
+      "epoch": 0.7115030674846625,
+      "grad_norm": 0.14669564366340637,
+      "learning_rate": 2.929343721565297e-05,
+      "loss": 0.9669,
+      "step": 4639
+    },
+    {
+      "epoch": 0.7116564417177914,
+      "grad_norm": 0.16853703558444977,
+      "learning_rate": 2.9264502249668727e-05,
+      "loss": 1.2099,
+      "step": 4640
+    },
+    {
+      "epoch": 0.7118098159509203,
+      "grad_norm": 0.15338140726089478,
+      "learning_rate": 2.923557811698156e-05,
+      "loss": 1.1283,
+      "step": 4641
+    },
+    {
+      "epoch": 0.7119631901840491,
+      "grad_norm": 0.16580119729042053,
+      "learning_rate": 2.9206664824442725e-05,
+      "loss": 1.1236,
+      "step": 4642
+    },
+    {
+      "epoch": 0.712116564417178,
+      "grad_norm": 0.1600951850414276,
+      "learning_rate": 2.917776237890083e-05,
+      "loss": 0.9711,
+      "step": 4643
+    },
+    {
+      "epoch": 0.7122699386503067,
+      "grad_norm": 0.15509898960590363,
+      "learning_rate": 2.914887078720199e-05,
+      "loss": 1.0014,
+      "step": 4644
+    },
+    {
+      "epoch": 0.7124233128834356,
+      "grad_norm": 0.15286383032798767,
+      "learning_rate": 2.9119990056189697e-05,
+      "loss": 1.0128,
+      "step": 4645
+    },
+    {
+      "epoch": 0.7125766871165644,
+      "grad_norm": 0.15845797955989838,
+      "learning_rate": 2.9091120192704862e-05,
+      "loss": 1.0806,
+      "step": 4646
+    },
+    {
+      "epoch": 0.7127300613496933,
+      "grad_norm": 0.15458694100379944,
+      "learning_rate": 2.9062261203585885e-05,
+      "loss": 1.0608,
+      "step": 4647
+    },
+    {
+      "epoch": 0.7128834355828221,
+      "grad_norm": 0.15030039846897125,
+      "learning_rate": 2.9033413095668507e-05,
+      "loss": 0.9521,
+      "step": 4648
+    },
+    {
+      "epoch": 0.713036809815951,
+      "grad_norm": 0.1745256781578064,
+      "learning_rate": 2.9004575875785977e-05,
+      "loss": 1.0956,
+      "step": 4649
+    },
+    {
+      "epoch": 0.7131901840490797,
+      "grad_norm": 0.1675093173980713,
+      "learning_rate": 2.897574955076889e-05,
+      "loss": 1.1822,
+      "step": 4650
+    },
+    {
+      "epoch": 0.7133435582822086,
+      "grad_norm": 0.15339823067188263,
+      "learning_rate": 2.8946934127445325e-05,
+      "loss": 1.0107,
+      "step": 4651
+    },
+    {
+      "epoch": 0.7134969325153374,
+      "grad_norm": 0.1569569855928421,
+      "learning_rate": 2.891812961264071e-05,
+      "loss": 1.0723,
+      "step": 4652
+    },
+    {
+      "epoch": 0.7136503067484663,
+      "grad_norm": 0.1537383645772934,
+      "learning_rate": 2.8889336013177972e-05,
+      "loss": 0.9921,
+      "step": 4653
+    },
+    {
+      "epoch": 0.7138036809815951,
+      "grad_norm": 0.15265832841396332,
+      "learning_rate": 2.8860553335877384e-05,
+      "loss": 1.0347,
+      "step": 4654
+    },
+    {
+      "epoch": 0.713957055214724,
+      "grad_norm": 0.1616775393486023,
+      "learning_rate": 2.8831781587556617e-05,
+      "loss": 0.9985,
+      "step": 4655
+    },
+    {
+      "epoch": 0.7141104294478527,
+      "grad_norm": 0.14239372313022614,
+      "learning_rate": 2.8803020775030864e-05,
+      "loss": 0.9582,
+      "step": 4656
+    },
+    {
+      "epoch": 0.7142638036809816,
+      "grad_norm": 0.14967161417007446,
+      "learning_rate": 2.8774270905112578e-05,
+      "loss": 1.0646,
+      "step": 4657
+    },
+    {
+      "epoch": 0.7144171779141104,
+      "grad_norm": 0.1485603153705597,
+      "learning_rate": 2.8745531984611768e-05,
+      "loss": 0.8819,
+      "step": 4658
+    },
+    {
+      "epoch": 0.7145705521472393,
+      "grad_norm": 0.16565236449241638,
+      "learning_rate": 2.8716804020335716e-05,
+      "loss": 1.1294,
+      "step": 4659
+    },
+    {
+      "epoch": 0.7147239263803681,
+      "grad_norm": 0.14740273356437683,
+      "learning_rate": 2.8688087019089224e-05,
+      "loss": 1.0034,
+      "step": 4660
+    },
+    {
+      "epoch": 0.714877300613497,
+      "grad_norm": 0.15110133588314056,
+      "learning_rate": 2.8659380987674424e-05,
+      "loss": 1.1242,
+      "step": 4661
+    },
+    {
+      "epoch": 0.7150306748466257,
+      "grad_norm": 0.1586732566356659,
+      "learning_rate": 2.863068593289083e-05,
+      "loss": 1.0138,
+      "step": 4662
+    },
+    {
+      "epoch": 0.7151840490797546,
+      "grad_norm": 0.1539641171693802,
+      "learning_rate": 2.8602001861535467e-05,
+      "loss": 1.0314,
+      "step": 4663
+    },
+    {
+      "epoch": 0.7153374233128834,
+      "grad_norm": 0.17047712206840515,
+      "learning_rate": 2.8573328780402615e-05,
+      "loss": 1.1775,
+      "step": 4664
+    },
+    {
+      "epoch": 0.7154907975460123,
+      "grad_norm": 0.16656771302223206,
+      "learning_rate": 2.854466669628409e-05,
+      "loss": 1.1354,
+      "step": 4665
+    },
+    {
+      "epoch": 0.7156441717791411,
+      "grad_norm": 0.16711367666721344,
+      "learning_rate": 2.851601561596897e-05,
+      "loss": 1.0827,
+      "step": 4666
+    },
+    {
+      "epoch": 0.71579754601227,
+      "grad_norm": 0.16540174186229706,
+      "learning_rate": 2.8487375546243856e-05,
+      "loss": 1.0492,
+      "step": 4667
+    },
+    {
+      "epoch": 0.7159509202453987,
+      "grad_norm": 0.17620410025119781,
+      "learning_rate": 2.8458746493892655e-05,
+      "loss": 1.1288,
+      "step": 4668
+    },
+    {
+      "epoch": 0.7161042944785276,
+      "grad_norm": 0.1757420301437378,
+      "learning_rate": 2.8430128465696645e-05,
+      "loss": 1.1393,
+      "step": 4669
+    },
+    {
+      "epoch": 0.7162576687116564,
+      "grad_norm": 0.15250858664512634,
+      "learning_rate": 2.84015214684346e-05,
+      "loss": 1.0417,
+      "step": 4670
+    },
+    {
+      "epoch": 0.7164110429447853,
+      "grad_norm": 0.15904060006141663,
+      "learning_rate": 2.837292550888255e-05,
+      "loss": 1.0007,
+      "step": 4671
+    },
+    {
+      "epoch": 0.7165644171779141,
+      "grad_norm": 0.15607178211212158,
+      "learning_rate": 2.8344340593814045e-05,
+      "loss": 1.1184,
+      "step": 4672
+    },
+    {
+      "epoch": 0.716717791411043,
+      "grad_norm": 0.1566692590713501,
+      "learning_rate": 2.8315766729999883e-05,
+      "loss": 1.1017,
+      "step": 4673
+    },
+    {
+      "epoch": 0.7168711656441717,
+      "grad_norm": 0.17039957642555237,
+      "learning_rate": 2.8287203924208346e-05,
+      "loss": 1.1381,
+      "step": 4674
+    },
+    {
+      "epoch": 0.7170245398773006,
+      "grad_norm": 0.15379323065280914,
+      "learning_rate": 2.8258652183205085e-05,
+      "loss": 1.0865,
+      "step": 4675
+    },
+    {
+      "epoch": 0.7171779141104294,
+      "grad_norm": 0.16658678650856018,
+      "learning_rate": 2.823011151375305e-05,
+      "loss": 1.1088,
+      "step": 4676
+    },
+    {
+      "epoch": 0.7173312883435583,
+      "grad_norm": 0.15444034337997437,
+      "learning_rate": 2.8201581922612683e-05,
+      "loss": 0.9977,
+      "step": 4677
+    },
+    {
+      "epoch": 0.7174846625766871,
+      "grad_norm": 0.1618213951587677,
+      "learning_rate": 2.8173063416541698e-05,
+      "loss": 1.116,
+      "step": 4678
+    },
+    {
+      "epoch": 0.717638036809816,
+      "grad_norm": 0.16961054503917694,
+      "learning_rate": 2.8144556002295278e-05,
+      "loss": 1.2497,
+      "step": 4679
+    },
+    {
+      "epoch": 0.7177914110429447,
+      "grad_norm": 0.14909717440605164,
+      "learning_rate": 2.8116059686625898e-05,
+      "loss": 1.0174,
+      "step": 4680
+    },
+    {
+      "epoch": 0.7179447852760736,
+      "grad_norm": 0.1683913618326187,
+      "learning_rate": 2.8087574476283408e-05,
+      "loss": 1.002,
+      "step": 4681
+    },
+    {
+      "epoch": 0.7180981595092024,
+      "grad_norm": 0.15785446763038635,
+      "learning_rate": 2.8059100378015125e-05,
+      "loss": 1.0657,
+      "step": 4682
+    },
+    {
+      "epoch": 0.7182515337423313,
+      "grad_norm": 0.15460793673992157,
+      "learning_rate": 2.8030637398565604e-05,
+      "loss": 0.9637,
+      "step": 4683
+    },
+    {
+      "epoch": 0.7184049079754601,
+      "grad_norm": 0.18119178712368011,
+      "learning_rate": 2.8002185544676873e-05,
+      "loss": 1.3114,
+      "step": 4684
+    },
+    {
+      "epoch": 0.718558282208589,
+      "grad_norm": 0.15982501208782196,
+      "learning_rate": 2.7973744823088224e-05,
+      "loss": 1.0519,
+      "step": 4685
+    },
+    {
+      "epoch": 0.7187116564417177,
+      "grad_norm": 0.15914812684059143,
+      "learning_rate": 2.7945315240536435e-05,
+      "loss": 1.1249,
+      "step": 4686
+    },
+    {
+      "epoch": 0.7188650306748466,
+      "grad_norm": 0.15595190227031708,
+      "learning_rate": 2.7916896803755516e-05,
+      "loss": 1.0544,
+      "step": 4687
+    },
+    {
+      "epoch": 0.7190184049079754,
+      "grad_norm": 0.14317531883716583,
+      "learning_rate": 2.7888489519476957e-05,
+      "loss": 1.0036,
+      "step": 4688
+    },
+    {
+      "epoch": 0.7191717791411043,
+      "grad_norm": 0.140059232711792,
+      "learning_rate": 2.786009339442951e-05,
+      "loss": 1.0313,
+      "step": 4689
+    },
+    {
+      "epoch": 0.7193251533742331,
+      "grad_norm": 0.16263364255428314,
+      "learning_rate": 2.7831708435339304e-05,
+      "loss": 0.9254,
+      "step": 4690
+    },
+    {
+      "epoch": 0.719478527607362,
+      "grad_norm": 0.15536749362945557,
+      "learning_rate": 2.7803334648929896e-05,
+      "loss": 1.0588,
+      "step": 4691
+    },
+    {
+      "epoch": 0.7196319018404908,
+      "grad_norm": 0.15920695662498474,
+      "learning_rate": 2.7774972041922074e-05,
+      "loss": 1.0398,
+      "step": 4692
+    },
+    {
+      "epoch": 0.7197852760736196,
+      "grad_norm": 0.1410697102546692,
+      "learning_rate": 2.774662062103412e-05,
+      "loss": 0.9409,
+      "step": 4693
+    },
+    {
+      "epoch": 0.7199386503067484,
+      "grad_norm": 0.15904748439788818,
+      "learning_rate": 2.771828039298152e-05,
+      "loss": 1.0557,
+      "step": 4694
+    },
+    {
+      "epoch": 0.7200920245398773,
+      "grad_norm": 0.17269840836524963,
+      "learning_rate": 2.7689951364477237e-05,
+      "loss": 1.1474,
+      "step": 4695
+    },
+    {
+      "epoch": 0.7202453987730061,
+      "grad_norm": 0.15708015859127045,
+      "learning_rate": 2.76616335422315e-05,
+      "loss": 1.0077,
+      "step": 4696
+    },
+    {
+      "epoch": 0.720398773006135,
+      "grad_norm": 0.14838846027851105,
+      "learning_rate": 2.7633326932951883e-05,
+      "loss": 0.9325,
+      "step": 4697
+    },
+    {
+      "epoch": 0.7205521472392638,
+      "grad_norm": 0.16584771871566772,
+      "learning_rate": 2.7605031543343384e-05,
+      "loss": 1.1997,
+      "step": 4698
+    },
+    {
+      "epoch": 0.7207055214723926,
+      "grad_norm": 0.1587873101234436,
+      "learning_rate": 2.7576747380108232e-05,
+      "loss": 1.0304,
+      "step": 4699
+    },
+    {
+      "epoch": 0.7208588957055214,
+      "grad_norm": 0.15955893695354462,
+      "learning_rate": 2.7548474449946105e-05,
+      "loss": 1.1264,
+      "step": 4700
+    },
+    {
+      "epoch": 0.7210122699386503,
+      "grad_norm": 0.15433920919895172,
+      "learning_rate": 2.7520212759553923e-05,
+      "loss": 0.953,
+      "step": 4701
+    },
+    {
+      "epoch": 0.7211656441717792,
+      "grad_norm": 0.15636499226093292,
+      "learning_rate": 2.7491962315626024e-05,
+      "loss": 0.9867,
+      "step": 4702
+    },
+    {
+      "epoch": 0.721319018404908,
+      "grad_norm": 0.15389743447303772,
+      "learning_rate": 2.7463723124854044e-05,
+      "loss": 1.1182,
+      "step": 4703
+    },
+    {
+      "epoch": 0.7214723926380369,
+      "grad_norm": 0.1464344710111618,
+      "learning_rate": 2.743549519392692e-05,
+      "loss": 0.9879,
+      "step": 4704
+    },
+    {
+      "epoch": 0.7216257668711656,
+      "grad_norm": 0.1604791134595871,
+      "learning_rate": 2.7407278529531003e-05,
+      "loss": 1.032,
+      "step": 4705
+    },
+    {
+      "epoch": 0.7217791411042945,
+      "grad_norm": 0.15361931920051575,
+      "learning_rate": 2.7379073138349892e-05,
+      "loss": 1.0277,
+      "step": 4706
+    },
+    {
+      "epoch": 0.7219325153374233,
+      "grad_norm": 0.14814847707748413,
+      "learning_rate": 2.7350879027064605e-05,
+      "loss": 1.0149,
+      "step": 4707
+    },
+    {
+      "epoch": 0.7220858895705522,
+      "grad_norm": 1.156808614730835,
+      "learning_rate": 2.7322696202353377e-05,
+      "loss": 1.0951,
+      "step": 4708
+    },
+    {
+      "epoch": 0.722239263803681,
+      "grad_norm": 0.1372937262058258,
+      "learning_rate": 2.7294524670891883e-05,
+      "loss": 0.9049,
+      "step": 4709
+    },
+    {
+      "epoch": 0.7223926380368099,
+      "grad_norm": 0.16232243180274963,
+      "learning_rate": 2.7266364439353027e-05,
+      "loss": 1.0486,
+      "step": 4710
+    },
+    {
+      "epoch": 0.7225460122699386,
+      "grad_norm": 0.14344483613967896,
+      "learning_rate": 2.7238215514407135e-05,
+      "loss": 0.9547,
+      "step": 4711
+    },
+    {
+      "epoch": 0.7226993865030675,
+      "grad_norm": 0.17259953916072845,
+      "learning_rate": 2.7210077902721754e-05,
+      "loss": 1.0166,
+      "step": 4712
+    },
+    {
+      "epoch": 0.7228527607361963,
+      "grad_norm": 0.15605898201465607,
+      "learning_rate": 2.7181951610961796e-05,
+      "loss": 1.0167,
+      "step": 4713
+    },
+    {
+      "epoch": 0.7230061349693252,
+      "grad_norm": 0.17079085111618042,
+      "learning_rate": 2.7153836645789523e-05,
+      "loss": 1.0639,
+      "step": 4714
+    },
+    {
+      "epoch": 0.723159509202454,
+      "grad_norm": 0.1507808417081833,
+      "learning_rate": 2.712573301386444e-05,
+      "loss": 1.0158,
+      "step": 4715
+    },
+    {
+      "epoch": 0.7233128834355829,
+      "grad_norm": 0.17366670072078705,
+      "learning_rate": 2.709764072184345e-05,
+      "loss": 1.133,
+      "step": 4716
+    },
+    {
+      "epoch": 0.7234662576687116,
+      "grad_norm": 0.16143646836280823,
+      "learning_rate": 2.706955977638073e-05,
+      "loss": 1.0965,
+      "step": 4717
+    },
+    {
+      "epoch": 0.7236196319018405,
+      "grad_norm": 0.15532474219799042,
+      "learning_rate": 2.704149018412774e-05,
+      "loss": 0.9415,
+      "step": 4718
+    },
+    {
+      "epoch": 0.7237730061349693,
+      "grad_norm": 0.1751648187637329,
+      "learning_rate": 2.701343195173332e-05,
+      "loss": 1.1135,
+      "step": 4719
+    },
+    {
+      "epoch": 0.7239263803680982,
+      "grad_norm": 0.15531589090824127,
+      "learning_rate": 2.6985385085843536e-05,
+      "loss": 1.0532,
+      "step": 4720
+    },
+    {
+      "epoch": 0.724079754601227,
+      "grad_norm": 0.3691503405570984,
+      "learning_rate": 2.6957349593101866e-05,
+      "loss": 1.1532,
+      "step": 4721
+    },
+    {
+      "epoch": 0.7242331288343559,
+      "grad_norm": 0.8284315466880798,
+      "learning_rate": 2.6929325480148963e-05,
+      "loss": 1.1645,
+      "step": 4722
+    },
+    {
+      "epoch": 0.7243865030674846,
+      "grad_norm": 0.15610206127166748,
+      "learning_rate": 2.6901312753622922e-05,
+      "loss": 1.0424,
+      "step": 4723
+    },
+    {
+      "epoch": 0.7245398773006135,
+      "grad_norm": 0.15863794088363647,
+      "learning_rate": 2.687331142015904e-05,
+      "loss": 0.896,
+      "step": 4724
+    },
+    {
+      "epoch": 0.7246932515337423,
+      "grad_norm": 0.1617928296327591,
+      "learning_rate": 2.6845321486389933e-05,
+      "loss": 1.1061,
+      "step": 4725
+    },
+    {
+      "epoch": 0.7248466257668712,
+      "grad_norm": 0.24165242910385132,
+      "learning_rate": 2.6817342958945573e-05,
+      "loss": 1.095,
+      "step": 4726
+    },
+    {
+      "epoch": 0.725,
+      "grad_norm": 0.15974821150302887,
+      "learning_rate": 2.678937584445315e-05,
+      "loss": 1.0397,
+      "step": 4727
+    },
+    {
+      "epoch": 0.7251533742331289,
+      "grad_norm": 0.16846869885921478,
+      "learning_rate": 2.6761420149537232e-05,
+      "loss": 1.2192,
+      "step": 4728
+    },
+    {
+      "epoch": 0.7253067484662576,
+      "grad_norm": 0.20532642304897308,
+      "learning_rate": 2.6733475880819603e-05,
+      "loss": 0.9969,
+      "step": 4729
+    },
+    {
+      "epoch": 0.7254601226993865,
+      "grad_norm": 0.16356846690177917,
+      "learning_rate": 2.670554304491943e-05,
+      "loss": 1.0854,
+      "step": 4730
+    },
+    {
+      "epoch": 0.7256134969325153,
+      "grad_norm": 0.20300541818141937,
+      "learning_rate": 2.6677621648453084e-05,
+      "loss": 0.8804,
+      "step": 4731
+    },
+    {
+      "epoch": 0.7257668711656442,
+      "grad_norm": 0.20216643810272217,
+      "learning_rate": 2.664971169803425e-05,
+      "loss": 1.1358,
+      "step": 4732
+    },
+    {
+      "epoch": 0.725920245398773,
+      "grad_norm": 0.19910526275634766,
+      "learning_rate": 2.6621813200273956e-05,
+      "loss": 1.2295,
+      "step": 4733
+    },
+    {
+      "epoch": 0.7260736196319019,
+      "grad_norm": 0.16115906834602356,
+      "learning_rate": 2.659392616178043e-05,
+      "loss": 1.0659,
+      "step": 4734
+    },
+    {
+      "epoch": 0.7262269938650306,
+      "grad_norm": 0.29153046011924744,
+      "learning_rate": 2.656605058915928e-05,
+      "loss": 1.0341,
+      "step": 4735
+    },
+    {
+      "epoch": 0.7263803680981595,
+      "grad_norm": 0.1410980522632599,
+      "learning_rate": 2.65381864890133e-05,
+      "loss": 0.934,
+      "step": 4736
+    },
+    {
+      "epoch": 0.7265337423312883,
+      "grad_norm": 0.16717395186424255,
+      "learning_rate": 2.6510333867942667e-05,
+      "loss": 1.1431,
+      "step": 4737
+    },
+    {
+      "epoch": 0.7266871165644172,
+      "grad_norm": 0.16844061017036438,
+      "learning_rate": 2.6482492732544773e-05,
+      "loss": 1.0307,
+      "step": 4738
+    },
+    {
+      "epoch": 0.726840490797546,
+      "grad_norm": 0.25759509205818176,
+      "learning_rate": 2.645466308941427e-05,
+      "loss": 1.1576,
+      "step": 4739
+    },
+    {
+      "epoch": 0.7269938650306749,
+      "grad_norm": 0.15752065181732178,
+      "learning_rate": 2.642684494514317e-05,
+      "loss": 1.0018,
+      "step": 4740
+    },
+    {
+      "epoch": 0.7271472392638036,
+      "grad_norm": 0.1678353101015091,
+      "learning_rate": 2.6399038306320673e-05,
+      "loss": 1.2616,
+      "step": 4741
+    },
+    {
+      "epoch": 0.7273006134969325,
+      "grad_norm": 0.15784068405628204,
+      "learning_rate": 2.6371243179533333e-05,
+      "loss": 1.0513,
+      "step": 4742
+    },
+    {
+      "epoch": 0.7274539877300613,
+      "grad_norm": 0.17724674940109253,
+      "learning_rate": 2.6343459571364903e-05,
+      "loss": 1.2682,
+      "step": 4743
+    },
+    {
+      "epoch": 0.7276073619631902,
+      "grad_norm": 0.14742454886436462,
+      "learning_rate": 2.6315687488396484e-05,
+      "loss": 1.0662,
+      "step": 4744
+    },
+    {
+      "epoch": 0.727760736196319,
+      "grad_norm": 0.15546353161334991,
+      "learning_rate": 2.6287926937206364e-05,
+      "loss": 0.9933,
+      "step": 4745
+    },
+    {
+      "epoch": 0.7279141104294479,
+      "grad_norm": 0.1575142741203308,
+      "learning_rate": 2.6260177924370183e-05,
+      "loss": 1.0659,
+      "step": 4746
+    },
+    {
+      "epoch": 0.7280674846625766,
+      "grad_norm": 0.1574835181236267,
+      "learning_rate": 2.623244045646079e-05,
+      "loss": 1.0293,
+      "step": 4747
+    },
+    {
+      "epoch": 0.7282208588957055,
+      "grad_norm": 0.14697183668613434,
+      "learning_rate": 2.620471454004829e-05,
+      "loss": 0.8934,
+      "step": 4748
+    },
+    {
+      "epoch": 0.7283742331288343,
+      "grad_norm": 0.16566424071788788,
+      "learning_rate": 2.617700018170012e-05,
+      "loss": 1.1299,
+      "step": 4749
+    },
+    {
+      "epoch": 0.7285276073619632,
+      "grad_norm": 0.15119284391403198,
+      "learning_rate": 2.61492973879809e-05,
+      "loss": 1.0001,
+      "step": 4750
+    },
+    {
+      "epoch": 0.728680981595092,
+      "grad_norm": 0.16706572473049164,
+      "learning_rate": 2.612160616545259e-05,
+      "loss": 1.2347,
+      "step": 4751
+    },
+    {
+      "epoch": 0.7288343558282209,
+      "grad_norm": 0.20365415513515472,
+      "learning_rate": 2.6093926520674323e-05,
+      "loss": 1.0685,
+      "step": 4752
+    },
+    {
+      "epoch": 0.7289877300613496,
+      "grad_norm": 0.16475756466388702,
+      "learning_rate": 2.606625846020257e-05,
+      "loss": 1.2086,
+      "step": 4753
+    },
+    {
+      "epoch": 0.7291411042944785,
+      "grad_norm": 0.15041141211986542,
+      "learning_rate": 2.6038601990591023e-05,
+      "loss": 0.9654,
+      "step": 4754
+    },
+    {
+      "epoch": 0.7292944785276073,
+      "grad_norm": 0.1642054319381714,
+      "learning_rate": 2.6010957118390576e-05,
+      "loss": 1.0919,
+      "step": 4755
+    },
+    {
+      "epoch": 0.7294478527607362,
+      "grad_norm": 0.16732393205165863,
+      "learning_rate": 2.5983323850149492e-05,
+      "loss": 1.1167,
+      "step": 4756
+    },
+    {
+      "epoch": 0.729601226993865,
+      "grad_norm": 0.16902528703212738,
+      "learning_rate": 2.5955702192413168e-05,
+      "loss": 1.1504,
+      "step": 4757
+    },
+    {
+      "epoch": 0.7297546012269939,
+      "grad_norm": 0.15650643408298492,
+      "learning_rate": 2.5928092151724328e-05,
+      "loss": 1.0616,
+      "step": 4758
+    },
+    {
+      "epoch": 0.7299079754601226,
+      "grad_norm": 0.16118840873241425,
+      "learning_rate": 2.590049373462294e-05,
+      "loss": 1.1333,
+      "step": 4759
+    },
+    {
+      "epoch": 0.7300613496932515,
+      "grad_norm": 0.1621835082769394,
+      "learning_rate": 2.587290694764616e-05,
+      "loss": 0.9979,
+      "step": 4760
+    },
+    {
+      "epoch": 0.7302147239263803,
+      "grad_norm": 0.15653188526630402,
+      "learning_rate": 2.584533179732846e-05,
+      "loss": 0.9584,
+      "step": 4761
+    },
+    {
+      "epoch": 0.7303680981595092,
+      "grad_norm": 0.16381248831748962,
+      "learning_rate": 2.58177682902015e-05,
+      "loss": 0.9963,
+      "step": 4762
+    },
+    {
+      "epoch": 0.730521472392638,
+      "grad_norm": 0.1561465859413147,
+      "learning_rate": 2.5790216432794232e-05,
+      "loss": 1.0858,
+      "step": 4763
+    },
+    {
+      "epoch": 0.7306748466257669,
+      "grad_norm": 0.1663474142551422,
+      "learning_rate": 2.5762676231632778e-05,
+      "loss": 1.1135,
+      "step": 4764
+    },
+    {
+      "epoch": 0.7308282208588958,
+      "grad_norm": 0.1732131987810135,
+      "learning_rate": 2.57351476932406e-05,
+      "loss": 0.9749,
+      "step": 4765
+    },
+    {
+      "epoch": 0.7309815950920245,
+      "grad_norm": 0.1532256305217743,
+      "learning_rate": 2.570763082413831e-05,
+      "loss": 1.0469,
+      "step": 4766
+    },
+    {
+      "epoch": 0.7311349693251534,
+      "grad_norm": 0.1534772664308548,
+      "learning_rate": 2.568012563084377e-05,
+      "loss": 1.0509,
+      "step": 4767
+    },
+    {
+      "epoch": 0.7312883435582822,
+      "grad_norm": 0.15682703256607056,
+      "learning_rate": 2.565263211987213e-05,
+      "loss": 0.961,
+      "step": 4768
+    },
+    {
+      "epoch": 0.7314417177914111,
+      "grad_norm": 0.14603842794895172,
+      "learning_rate": 2.562515029773569e-05,
+      "loss": 1.0245,
+      "step": 4769
+    },
+    {
+      "epoch": 0.7315950920245399,
+      "grad_norm": 0.1570816934108734,
+      "learning_rate": 2.5597680170944075e-05,
+      "loss": 1.0538,
+      "step": 4770
+    },
+    {
+      "epoch": 0.7317484662576688,
+      "grad_norm": 0.14805246889591217,
+      "learning_rate": 2.5570221746004047e-05,
+      "loss": 1.0325,
+      "step": 4771
+    },
+    {
+      "epoch": 0.7319018404907975,
+      "grad_norm": 0.1592307686805725,
+      "learning_rate": 2.5542775029419694e-05,
+      "loss": 1.0821,
+      "step": 4772
+    },
+    {
+      "epoch": 0.7320552147239264,
+      "grad_norm": 0.15847885608673096,
+      "learning_rate": 2.5515340027692247e-05,
+      "loss": 1.1604,
+      "step": 4773
+    },
+    {
+      "epoch": 0.7322085889570552,
+      "grad_norm": 0.1547393798828125,
+      "learning_rate": 2.5487916747320165e-05,
+      "loss": 1.0992,
+      "step": 4774
+    },
+    {
+      "epoch": 0.7323619631901841,
+      "grad_norm": 0.15948578715324402,
+      "learning_rate": 2.546050519479921e-05,
+      "loss": 1.0449,
+      "step": 4775
+    },
+    {
+      "epoch": 0.7325153374233129,
+      "grad_norm": 0.16526654362678528,
+      "learning_rate": 2.543310537662227e-05,
+      "loss": 1.0393,
+      "step": 4776
+    },
+    {
+      "epoch": 0.7326687116564418,
+      "grad_norm": 0.1558866798877716,
+      "learning_rate": 2.540571729927954e-05,
+      "loss": 1.0704,
+      "step": 4777
+    },
+    {
+      "epoch": 0.7328220858895705,
+      "grad_norm": 0.1630602329969406,
+      "learning_rate": 2.5378340969258348e-05,
+      "loss": 1.0471,
+      "step": 4778
+    },
+    {
+      "epoch": 0.7329754601226994,
+      "grad_norm": 0.16562478244304657,
+      "learning_rate": 2.5350976393043328e-05,
+      "loss": 1.192,
+      "step": 4779
+    },
+    {
+      "epoch": 0.7331288343558282,
+      "grad_norm": 0.16687901318073273,
+      "learning_rate": 2.5323623577116245e-05,
+      "loss": 1.1623,
+      "step": 4780
+    },
+    {
+      "epoch": 0.7332822085889571,
+      "grad_norm": 0.14213518798351288,
+      "learning_rate": 2.5296282527956156e-05,
+      "loss": 1.0279,
+      "step": 4781
+    },
+    {
+      "epoch": 0.7334355828220859,
+      "grad_norm": 0.16214075684547424,
+      "learning_rate": 2.5268953252039278e-05,
+      "loss": 1.0326,
+      "step": 4782
+    },
+    {
+      "epoch": 0.7335889570552148,
+      "grad_norm": 0.16316555440425873,
+      "learning_rate": 2.524163575583902e-05,
+      "loss": 1.0523,
+      "step": 4783
+    },
+    {
+      "epoch": 0.7337423312883435,
+      "grad_norm": 0.1682974100112915,
+      "learning_rate": 2.5214330045826097e-05,
+      "loss": 1.1228,
+      "step": 4784
+    },
+    {
+      "epoch": 0.7338957055214724,
+      "grad_norm": 0.14995364844799042,
+      "learning_rate": 2.518703612846831e-05,
+      "loss": 0.9693,
+      "step": 4785
+    },
+    {
+      "epoch": 0.7340490797546012,
+      "grad_norm": 0.16020090878009796,
+      "learning_rate": 2.515975401023078e-05,
+      "loss": 1.1772,
+      "step": 4786
+    },
+    {
+      "epoch": 0.7342024539877301,
+      "grad_norm": 0.1748143881559372,
+      "learning_rate": 2.5132483697575733e-05,
+      "loss": 1.184,
+      "step": 4787
+    },
+    {
+      "epoch": 0.7343558282208589,
+      "grad_norm": 0.16117775440216064,
+      "learning_rate": 2.51052251969627e-05,
+      "loss": 1.0452,
+      "step": 4788
+    },
+    {
+      "epoch": 0.7345092024539878,
+      "grad_norm": 0.15506485104560852,
+      "learning_rate": 2.507797851484833e-05,
+      "loss": 1.0223,
+      "step": 4789
+    },
+    {
+      "epoch": 0.7346625766871165,
+      "grad_norm": 0.15724465250968933,
+      "learning_rate": 2.5050743657686475e-05,
+      "loss": 1.0367,
+      "step": 4790
+    },
+    {
+      "epoch": 0.7348159509202454,
+      "grad_norm": 0.1641533225774765,
+      "learning_rate": 2.5023520631928263e-05,
+      "loss": 1.0293,
+      "step": 4791
+    },
+    {
+      "epoch": 0.7349693251533742,
+      "grad_norm": 0.15270555019378662,
+      "learning_rate": 2.499630944402193e-05,
+      "loss": 0.9773,
+      "step": 4792
+    },
+    {
+      "epoch": 0.7351226993865031,
+      "grad_norm": 0.1602126508951187,
+      "learning_rate": 2.4969110100412993e-05,
+      "loss": 1.027,
+      "step": 4793
+    },
+    {
+      "epoch": 0.7352760736196319,
+      "grad_norm": 0.1611841917037964,
+      "learning_rate": 2.494192260754406e-05,
+      "loss": 1.045,
+      "step": 4794
+    },
+    {
+      "epoch": 0.7354294478527608,
+      "grad_norm": 0.15729700028896332,
+      "learning_rate": 2.4914746971855056e-05,
+      "loss": 1.1265,
+      "step": 4795
+    },
+    {
+      "epoch": 0.7355828220858895,
+      "grad_norm": 0.1405768096446991,
+      "learning_rate": 2.4887583199783002e-05,
+      "loss": 0.9382,
+      "step": 4796
+    },
+    {
+      "epoch": 0.7357361963190184,
+      "grad_norm": 0.15090888738632202,
+      "learning_rate": 2.4860431297762105e-05,
+      "loss": 0.9995,
+      "step": 4797
+    },
+    {
+      "epoch": 0.7358895705521472,
+      "grad_norm": 0.14585581421852112,
+      "learning_rate": 2.4833291272223853e-05,
+      "loss": 1.1003,
+      "step": 4798
+    },
+    {
+      "epoch": 0.7360429447852761,
+      "grad_norm": 0.15128780901432037,
+      "learning_rate": 2.480616312959682e-05,
+      "loss": 1.0661,
+      "step": 4799
+    },
+    {
+      "epoch": 0.7361963190184049,
+      "grad_norm": 0.22327548265457153,
+      "learning_rate": 2.4779046876306836e-05,
+      "loss": 1.0061,
+      "step": 4800
+    },
+    {
+      "epoch": 0.7363496932515338,
+      "grad_norm": 0.15386351943016052,
+      "learning_rate": 2.4751942518776852e-05,
+      "loss": 0.987,
+      "step": 4801
+    },
+    {
+      "epoch": 0.7365030674846625,
+      "grad_norm": 0.17636144161224365,
+      "learning_rate": 2.4724850063427058e-05,
+      "loss": 1.0831,
+      "step": 4802
+    },
+    {
+      "epoch": 0.7366564417177914,
+      "grad_norm": 0.16791610419750214,
+      "learning_rate": 2.469776951667482e-05,
+      "loss": 1.135,
+      "step": 4803
+    },
+    {
+      "epoch": 0.7368098159509202,
+      "grad_norm": 0.14386366307735443,
+      "learning_rate": 2.4670700884934627e-05,
+      "loss": 0.9649,
+      "step": 4804
+    },
+    {
+      "epoch": 0.7369631901840491,
+      "grad_norm": 0.15070487558841705,
+      "learning_rate": 2.4643644174618223e-05,
+      "loss": 0.9552,
+      "step": 4805
+    },
+    {
+      "epoch": 0.7371165644171779,
+      "grad_norm": 0.1404603272676468,
+      "learning_rate": 2.461659939213445e-05,
+      "loss": 0.9428,
+      "step": 4806
+    },
+    {
+      "epoch": 0.7372699386503068,
+      "grad_norm": 0.16909964382648468,
+      "learning_rate": 2.458956654388941e-05,
+      "loss": 1.045,
+      "step": 4807
+    },
+    {
+      "epoch": 0.7374233128834355,
+      "grad_norm": 0.15138790011405945,
+      "learning_rate": 2.4562545636286305e-05,
+      "loss": 1.04,
+      "step": 4808
+    },
+    {
+      "epoch": 0.7375766871165644,
+      "grad_norm": 0.149057999253273,
+      "learning_rate": 2.4535536675725512e-05,
+      "loss": 1.0279,
+      "step": 4809
+    },
+    {
+      "epoch": 0.7377300613496932,
+      "grad_norm": 0.1636347770690918,
+      "learning_rate": 2.4508539668604655e-05,
+      "loss": 1.1264,
+      "step": 4810
+    },
+    {
+      "epoch": 0.7378834355828221,
+      "grad_norm": 0.1525610387325287,
+      "learning_rate": 2.448155462131842e-05,
+      "loss": 1.0937,
+      "step": 4811
+    },
+    {
+      "epoch": 0.7380368098159509,
+      "grad_norm": 0.166118785738945,
+      "learning_rate": 2.4454581540258762e-05,
+      "loss": 1.1194,
+      "step": 4812
+    },
+    {
+      "epoch": 0.7381901840490798,
+      "grad_norm": 0.15521717071533203,
+      "learning_rate": 2.44276204318147e-05,
+      "loss": 1.0046,
+      "step": 4813
+    },
+    {
+      "epoch": 0.7383435582822085,
+      "grad_norm": 0.15372465550899506,
+      "learning_rate": 2.440067130237254e-05,
+      "loss": 1.0754,
+      "step": 4814
+    },
+    {
+      "epoch": 0.7384969325153374,
+      "grad_norm": 0.1371060013771057,
+      "learning_rate": 2.4373734158315604e-05,
+      "loss": 1.0231,
+      "step": 4815
+    },
+    {
+      "epoch": 0.7386503067484662,
+      "grad_norm": 0.1594177633523941,
+      "learning_rate": 2.4346809006024525e-05,
+      "loss": 1.0856,
+      "step": 4816
+    },
+    {
+      "epoch": 0.7388036809815951,
+      "grad_norm": 0.15640725195407867,
+      "learning_rate": 2.4319895851876983e-05,
+      "loss": 1.0252,
+      "step": 4817
+    },
+    {
+      "epoch": 0.7389570552147239,
+      "grad_norm": 0.16525106132030487,
+      "learning_rate": 2.4292994702247837e-05,
+      "loss": 1.1253,
+      "step": 4818
+    },
+    {
+      "epoch": 0.7391104294478528,
+      "grad_norm": 0.15349161624908447,
+      "learning_rate": 2.4266105563509165e-05,
+      "loss": 1.0461,
+      "step": 4819
+    },
+    {
+      "epoch": 0.7392638036809815,
+      "grad_norm": 0.15077096223831177,
+      "learning_rate": 2.4239228442030113e-05,
+      "loss": 1.0089,
+      "step": 4820
+    },
+    {
+      "epoch": 0.7394171779141104,
+      "grad_norm": 0.1590133011341095,
+      "learning_rate": 2.4212363344177063e-05,
+      "loss": 1.0291,
+      "step": 4821
+    },
+    {
+      "epoch": 0.7395705521472392,
+      "grad_norm": 0.14957161247730255,
+      "learning_rate": 2.418551027631348e-05,
+      "loss": 1.0664,
+      "step": 4822
+    },
+    {
+      "epoch": 0.7397239263803681,
+      "grad_norm": 0.16263411939144135,
+      "learning_rate": 2.415866924480003e-05,
+      "loss": 1.1126,
+      "step": 4823
+    },
+    {
+      "epoch": 0.7398773006134969,
+      "grad_norm": 0.15682156383991241,
+      "learning_rate": 2.413184025599451e-05,
+      "loss": 1.1193,
+      "step": 4824
+    },
+    {
+      "epoch": 0.7400306748466258,
+      "grad_norm": 0.15181097388267517,
+      "learning_rate": 2.410502331625182e-05,
+      "loss": 0.9965,
+      "step": 4825
+    },
+    {
+      "epoch": 0.7401840490797545,
+      "grad_norm": 0.16809095442295074,
+      "learning_rate": 2.4078218431924104e-05,
+      "loss": 1.1922,
+      "step": 4826
+    },
+    {
+      "epoch": 0.7403374233128834,
+      "grad_norm": 0.15254008769989014,
+      "learning_rate": 2.405142560936054e-05,
+      "loss": 1.0273,
+      "step": 4827
+    },
+    {
+      "epoch": 0.7404907975460123,
+      "grad_norm": 0.15413230657577515,
+      "learning_rate": 2.4024644854907556e-05,
+      "loss": 1.0321,
+      "step": 4828
+    },
+    {
+      "epoch": 0.7406441717791411,
+      "grad_norm": 0.1574150025844574,
+      "learning_rate": 2.3997876174908608e-05,
+      "loss": 1.0174,
+      "step": 4829
+    },
+    {
+      "epoch": 0.74079754601227,
+      "grad_norm": 0.14614856243133545,
+      "learning_rate": 2.397111957570442e-05,
+      "loss": 0.965,
+      "step": 4830
+    },
+    {
+      "epoch": 0.7409509202453988,
+      "grad_norm": 0.1587124615907669,
+      "learning_rate": 2.3944375063632746e-05,
+      "loss": 1.1416,
+      "step": 4831
+    },
+    {
+      "epoch": 0.7411042944785277,
+      "grad_norm": 0.15838930010795593,
+      "learning_rate": 2.39176426450285e-05,
+      "loss": 1.0682,
+      "step": 4832
+    },
+    {
+      "epoch": 0.7412576687116564,
+      "grad_norm": 0.15453669428825378,
+      "learning_rate": 2.3890922326223795e-05,
+      "loss": 1.0078,
+      "step": 4833
+    },
+    {
+      "epoch": 0.7414110429447853,
+      "grad_norm": 0.1455966681241989,
+      "learning_rate": 2.3864214113547785e-05,
+      "loss": 0.9416,
+      "step": 4834
+    },
+    {
+      "epoch": 0.7415644171779141,
+      "grad_norm": 0.14974582195281982,
+      "learning_rate": 2.383751801332685e-05,
+      "loss": 1.1315,
+      "step": 4835
+    },
+    {
+      "epoch": 0.741717791411043,
+      "grad_norm": 0.1717061847448349,
+      "learning_rate": 2.3810834031884413e-05,
+      "loss": 1.0733,
+      "step": 4836
+    },
+    {
+      "epoch": 0.7418711656441718,
+      "grad_norm": 0.15484069287776947,
+      "learning_rate": 2.3784162175541108e-05,
+      "loss": 1.0273,
+      "step": 4837
+    },
+    {
+      "epoch": 0.7420245398773007,
+      "grad_norm": 0.16261836886405945,
+      "learning_rate": 2.3757502450614606e-05,
+      "loss": 1.1133,
+      "step": 4838
+    },
+    {
+      "epoch": 0.7421779141104294,
+      "grad_norm": 0.18576017022132874,
+      "learning_rate": 2.373085486341981e-05,
+      "loss": 1.1794,
+      "step": 4839
+    },
+    {
+      "epoch": 0.7423312883435583,
+      "grad_norm": 0.1535072922706604,
+      "learning_rate": 2.3704219420268676e-05,
+      "loss": 1.0181,
+      "step": 4840
+    },
+    {
+      "epoch": 0.7424846625766871,
+      "grad_norm": 0.16047599911689758,
+      "learning_rate": 2.3677596127470257e-05,
+      "loss": 1.0399,
+      "step": 4841
+    },
+    {
+      "epoch": 0.742638036809816,
+      "grad_norm": 0.15500353276729584,
+      "learning_rate": 2.365098499133083e-05,
+      "loss": 1.077,
+      "step": 4842
+    },
+    {
+      "epoch": 0.7427914110429448,
+      "grad_norm": 0.15435835719108582,
+      "learning_rate": 2.3624386018153697e-05,
+      "loss": 1.0707,
+      "step": 4843
+    },
+    {
+      "epoch": 0.7429447852760737,
+      "grad_norm": 0.15177366137504578,
+      "learning_rate": 2.3597799214239315e-05,
+      "loss": 1.0587,
+      "step": 4844
+    },
+    {
+      "epoch": 0.7430981595092024,
+      "grad_norm": 0.15499411523342133,
+      "learning_rate": 2.3571224585885305e-05,
+      "loss": 1.1211,
+      "step": 4845
+    },
+    {
+      "epoch": 0.7432515337423313,
+      "grad_norm": 0.16660846769809723,
+      "learning_rate": 2.35446621393863e-05,
+      "loss": 1.0711,
+      "step": 4846
+    },
+    {
+      "epoch": 0.7434049079754601,
+      "grad_norm": 0.15047027170658112,
+      "learning_rate": 2.3518111881034162e-05,
+      "loss": 1.0214,
+      "step": 4847
+    },
+    {
+      "epoch": 0.743558282208589,
+      "grad_norm": 0.1401759833097458,
+      "learning_rate": 2.3491573817117753e-05,
+      "loss": 0.9375,
+      "step": 4848
+    },
+    {
+      "epoch": 0.7437116564417178,
+      "grad_norm": 0.16084319353103638,
+      "learning_rate": 2.3465047953923154e-05,
+      "loss": 1.0624,
+      "step": 4849
+    },
+    {
+      "epoch": 0.7438650306748467,
+      "grad_norm": 0.1464039385318756,
+      "learning_rate": 2.343853429773346e-05,
+      "loss": 0.9943,
+      "step": 4850
+    },
+    {
+      "epoch": 0.7440184049079754,
+      "grad_norm": 0.1678074300289154,
+      "learning_rate": 2.3412032854828964e-05,
+      "loss": 1.1261,
+      "step": 4851
+    },
+    {
+      "epoch": 0.7441717791411043,
+      "grad_norm": 0.15005123615264893,
+      "learning_rate": 2.3385543631487e-05,
+      "loss": 1.0788,
+      "step": 4852
+    },
+    {
+      "epoch": 0.7443251533742331,
+      "grad_norm": 0.1486799716949463,
+      "learning_rate": 2.3359066633982008e-05,
+      "loss": 0.9958,
+      "step": 4853
+    },
+    {
+      "epoch": 0.744478527607362,
+      "grad_norm": 0.16775429248809814,
+      "learning_rate": 2.3332601868585592e-05,
+      "loss": 1.0494,
+      "step": 4854
+    },
+    {
+      "epoch": 0.7446319018404908,
+      "grad_norm": 0.18107891082763672,
+      "learning_rate": 2.330614934156639e-05,
+      "loss": 1.0259,
+      "step": 4855
+    },
+    {
+      "epoch": 0.7447852760736197,
+      "grad_norm": 0.16532278060913086,
+      "learning_rate": 2.3279709059190213e-05,
+      "loss": 1.0245,
+      "step": 4856
+    },
+    {
+      "epoch": 0.7449386503067484,
+      "grad_norm": 0.15282846987247467,
+      "learning_rate": 2.3253281027719866e-05,
+      "loss": 1.0112,
+      "step": 4857
+    },
+    {
+      "epoch": 0.7450920245398773,
+      "grad_norm": 0.17147549986839294,
+      "learning_rate": 2.3226865253415385e-05,
+      "loss": 1.1139,
+      "step": 4858
+    },
+    {
+      "epoch": 0.7452453987730061,
+      "grad_norm": 0.14917372167110443,
+      "learning_rate": 2.32004617425338e-05,
+      "loss": 0.9963,
+      "step": 4859
+    },
+    {
+      "epoch": 0.745398773006135,
+      "grad_norm": 0.16155073046684265,
+      "learning_rate": 2.3174070501329254e-05,
+      "loss": 1.0301,
+      "step": 4860
+    },
+    {
+      "epoch": 0.7455521472392638,
+      "grad_norm": 0.1456802636384964,
+      "learning_rate": 2.314769153605303e-05,
+      "loss": 1.0326,
+      "step": 4861
+    },
+    {
+      "epoch": 0.7457055214723927,
+      "grad_norm": 0.260094553232193,
+      "learning_rate": 2.3121324852953444e-05,
+      "loss": 0.9837,
+      "step": 4862
+    },
+    {
+      "epoch": 0.7458588957055214,
+      "grad_norm": 0.1607215404510498,
+      "learning_rate": 2.3094970458275968e-05,
+      "loss": 1.0145,
+      "step": 4863
+    },
+    {
+      "epoch": 0.7460122699386503,
+      "grad_norm": 0.14627034962177277,
+      "learning_rate": 2.306862835826309e-05,
+      "loss": 0.8817,
+      "step": 4864
+    },
+    {
+      "epoch": 0.7461656441717791,
+      "grad_norm": 0.16227278113365173,
+      "learning_rate": 2.304229855915446e-05,
+      "loss": 1.1197,
+      "step": 4865
+    },
+    {
+      "epoch": 0.746319018404908,
+      "grad_norm": 0.15668316185474396,
+      "learning_rate": 2.3015981067186758e-05,
+      "loss": 1.0746,
+      "step": 4866
+    },
+    {
+      "epoch": 0.7464723926380368,
+      "grad_norm": 0.15932784974575043,
+      "learning_rate": 2.2989675888593743e-05,
+      "loss": 1.06,
+      "step": 4867
+    },
+    {
+      "epoch": 0.7466257668711657,
+      "grad_norm": 0.1579982042312622,
+      "learning_rate": 2.2963383029606327e-05,
+      "loss": 0.9752,
+      "step": 4868
+    },
+    {
+      "epoch": 0.7467791411042944,
+      "grad_norm": 0.1602020412683487,
+      "learning_rate": 2.2937102496452417e-05,
+      "loss": 1.1031,
+      "step": 4869
+    },
+    {
+      "epoch": 0.7469325153374233,
+      "grad_norm": 0.5916392207145691,
+      "learning_rate": 2.291083429535709e-05,
+      "loss": 1.1857,
+      "step": 4870
+    },
+    {
+      "epoch": 0.7470858895705521,
+      "grad_norm": 0.15400254726409912,
+      "learning_rate": 2.28845784325424e-05,
+      "loss": 0.9533,
+      "step": 4871
+    },
+    {
+      "epoch": 0.747239263803681,
+      "grad_norm": 0.1468094438314438,
+      "learning_rate": 2.2858334914227584e-05,
+      "loss": 0.8516,
+      "step": 4872
+    },
+    {
+      "epoch": 0.7473926380368098,
+      "grad_norm": 0.15203820168972015,
+      "learning_rate": 2.2832103746628866e-05,
+      "loss": 1.0059,
+      "step": 4873
+    },
+    {
+      "epoch": 0.7475460122699387,
+      "grad_norm": 0.15552541613578796,
+      "learning_rate": 2.2805884935959614e-05,
+      "loss": 1.0827,
+      "step": 4874
+    },
+    {
+      "epoch": 0.7476993865030674,
+      "grad_norm": 0.1446714848279953,
+      "learning_rate": 2.277967848843022e-05,
+      "loss": 0.9944,
+      "step": 4875
+    },
+    {
+      "epoch": 0.7478527607361963,
+      "grad_norm": 0.1525045484304428,
+      "learning_rate": 2.2753484410248148e-05,
+      "loss": 1.0105,
+      "step": 4876
+    },
+    {
+      "epoch": 0.7480061349693251,
+      "grad_norm": 0.15250660479068756,
+      "learning_rate": 2.2727302707617986e-05,
+      "loss": 0.9892,
+      "step": 4877
+    },
+    {
+      "epoch": 0.748159509202454,
+      "grad_norm": 0.1646517664194107,
+      "learning_rate": 2.2701133386741314e-05,
+      "loss": 1.085,
+      "step": 4878
+    },
+    {
+      "epoch": 0.7483128834355828,
+      "grad_norm": 0.16727416217327118,
+      "learning_rate": 2.267497645381686e-05,
+      "loss": 1.0696,
+      "step": 4879
+    },
+    {
+      "epoch": 0.7484662576687117,
+      "grad_norm": 0.16217383742332458,
+      "learning_rate": 2.2648831915040335e-05,
+      "loss": 1.1332,
+      "step": 4880
+    },
+    {
+      "epoch": 0.7486196319018404,
+      "grad_norm": 0.190058171749115,
+      "learning_rate": 2.2622699776604597e-05,
+      "loss": 1.0497,
+      "step": 4881
+    },
+    {
+      "epoch": 0.7487730061349693,
+      "grad_norm": 0.16254018247127533,
+      "learning_rate": 2.2596580044699513e-05,
+      "loss": 1.0424,
+      "step": 4882
+    },
+    {
+      "epoch": 0.7489263803680981,
+      "grad_norm": 0.149506077170372,
+      "learning_rate": 2.257047272551198e-05,
+      "loss": 1.0137,
+      "step": 4883
+    },
+    {
+      "epoch": 0.749079754601227,
+      "grad_norm": 0.16062863171100616,
+      "learning_rate": 2.254437782522606e-05,
+      "loss": 1.0987,
+      "step": 4884
+    },
+    {
+      "epoch": 0.7492331288343558,
+      "grad_norm": 0.16236479580402374,
+      "learning_rate": 2.2518295350022776e-05,
+      "loss": 1.0891,
+      "step": 4885
+    },
+    {
+      "epoch": 0.7493865030674847,
+      "grad_norm": 0.15753857791423798,
+      "learning_rate": 2.249222530608024e-05,
+      "loss": 1.0901,
+      "step": 4886
+    },
+    {
+      "epoch": 0.7495398773006134,
+      "grad_norm": 0.15305666625499725,
+      "learning_rate": 2.2466167699573674e-05,
+      "loss": 1.1313,
+      "step": 4887
+    },
+    {
+      "epoch": 0.7496932515337423,
+      "grad_norm": 0.16430655121803284,
+      "learning_rate": 2.2440122536675233e-05,
+      "loss": 1.1949,
+      "step": 4888
+    },
+    {
+      "epoch": 0.7498466257668711,
+      "grad_norm": 0.150715172290802,
+      "learning_rate": 2.2414089823554257e-05,
+      "loss": 1.0207,
+      "step": 4889
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1544128656387329,
+      "learning_rate": 2.238806956637702e-05,
+      "loss": 1.0704,
+      "step": 4890
+    },
+    {
+      "epoch": 0.7501533742331289,
+      "grad_norm": 0.1439797431230545,
+      "learning_rate": 2.236206177130695e-05,
+      "loss": 0.9962,
+      "step": 4891
+    },
+    {
+      "epoch": 0.7503067484662577,
+      "grad_norm": 0.16671083867549896,
+      "learning_rate": 2.2336066444504434e-05,
+      "loss": 1.1686,
+      "step": 4892
+    },
+    {
+      "epoch": 0.7504601226993866,
+      "grad_norm": 0.17932899296283722,
+      "learning_rate": 2.2310083592126976e-05,
+      "loss": 1.1786,
+      "step": 4893
+    },
+    {
+      "epoch": 0.7506134969325153,
+      "grad_norm": 0.1546257585287094,
+      "learning_rate": 2.228411322032908e-05,
+      "loss": 1.063,
+      "step": 4894
+    },
+    {
+      "epoch": 0.7507668711656442,
+      "grad_norm": 0.15660794079303741,
+      "learning_rate": 2.2258155335262285e-05,
+      "loss": 1.0451,
+      "step": 4895
+    },
+    {
+      "epoch": 0.750920245398773,
+      "grad_norm": 0.1547050178050995,
+      "learning_rate": 2.2232209943075243e-05,
+      "loss": 1.0122,
+      "step": 4896
+    },
+    {
+      "epoch": 0.7510736196319019,
+      "grad_norm": 0.15579164028167725,
+      "learning_rate": 2.2206277049913553e-05,
+      "loss": 1.0668,
+      "step": 4897
+    },
+    {
+      "epoch": 0.7512269938650307,
+      "grad_norm": 0.15486817061901093,
+      "learning_rate": 2.2180356661919943e-05,
+      "loss": 1.0598,
+      "step": 4898
+    },
+    {
+      "epoch": 0.7513803680981596,
+      "grad_norm": 0.15795797109603882,
+      "learning_rate": 2.215444878523409e-05,
+      "loss": 1.0162,
+      "step": 4899
+    },
+    {
+      "epoch": 0.7515337423312883,
+      "grad_norm": 0.16649675369262695,
+      "learning_rate": 2.2128553425992796e-05,
+      "loss": 1.1029,
+      "step": 4900
+    },
+    {
+      "epoch": 0.7516871165644172,
+      "grad_norm": 0.16680076718330383,
+      "learning_rate": 2.210267059032984e-05,
+      "loss": 1.092,
+      "step": 4901
+    },
+    {
+      "epoch": 0.751840490797546,
+      "grad_norm": 0.14619426429271698,
+      "learning_rate": 2.2076800284376013e-05,
+      "loss": 1.0134,
+      "step": 4902
+    },
+    {
+      "epoch": 0.7519938650306749,
+      "grad_norm": 0.15435869991779327,
+      "learning_rate": 2.205094251425924e-05,
+      "loss": 1.0269,
+      "step": 4903
+    },
+    {
+      "epoch": 0.7521472392638037,
+      "grad_norm": 0.1537514179944992,
+      "learning_rate": 2.202509728610435e-05,
+      "loss": 0.9169,
+      "step": 4904
+    },
+    {
+      "epoch": 0.7523006134969326,
+      "grad_norm": 0.15396396815776825,
+      "learning_rate": 2.1999264606033317e-05,
+      "loss": 1.046,
+      "step": 4905
+    },
+    {
+      "epoch": 0.7524539877300613,
+      "grad_norm": 0.158710777759552,
+      "learning_rate": 2.197344448016503e-05,
+      "loss": 1.0185,
+      "step": 4906
+    },
+    {
+      "epoch": 0.7526073619631902,
+      "grad_norm": 0.13783185184001923,
+      "learning_rate": 2.1947636914615522e-05,
+      "loss": 0.8766,
+      "step": 4907
+    },
+    {
+      "epoch": 0.752760736196319,
+      "grad_norm": 0.14992964267730713,
+      "learning_rate": 2.192184191549774e-05,
+      "loss": 1.1123,
+      "step": 4908
+    },
+    {
+      "epoch": 0.7529141104294479,
+      "grad_norm": 0.15453597903251648,
+      "learning_rate": 2.1896059488921756e-05,
+      "loss": 1.1139,
+      "step": 4909
+    },
+    {
+      "epoch": 0.7530674846625767,
+      "grad_norm": 0.14551620185375214,
+      "learning_rate": 2.1870289640994585e-05,
+      "loss": 0.9348,
+      "step": 4910
+    },
+    {
+      "epoch": 0.7532208588957056,
+      "grad_norm": 0.14462552964687347,
+      "learning_rate": 2.184453237782026e-05,
+      "loss": 0.984,
+      "step": 4911
+    },
+    {
+      "epoch": 0.7533742331288343,
+      "grad_norm": 0.16536812484264374,
+      "learning_rate": 2.181878770549993e-05,
+      "loss": 1.1072,
+      "step": 4912
+    },
+    {
+      "epoch": 0.7535276073619632,
+      "grad_norm": 0.1672157496213913,
+      "learning_rate": 2.179305563013164e-05,
+      "loss": 1.0714,
+      "step": 4913
+    },
+    {
+      "epoch": 0.753680981595092,
+      "grad_norm": 0.14931009709835052,
+      "learning_rate": 2.176733615781056e-05,
+      "loss": 0.9671,
+      "step": 4914
+    },
+    {
+      "epoch": 0.7538343558282209,
+      "grad_norm": 0.1632610708475113,
+      "learning_rate": 2.1741629294628757e-05,
+      "loss": 1.0508,
+      "step": 4915
+    },
+    {
+      "epoch": 0.7539877300613497,
+      "grad_norm": 0.16455641388893127,
+      "learning_rate": 2.171593504667544e-05,
+      "loss": 0.9713,
+      "step": 4916
+    },
+    {
+      "epoch": 0.7541411042944786,
+      "grad_norm": 0.16682060062885284,
+      "learning_rate": 2.1690253420036747e-05,
+      "loss": 1.122,
+      "step": 4917
+    },
+    {
+      "epoch": 0.7542944785276073,
+      "grad_norm": 0.1546294242143631,
+      "learning_rate": 2.1664584420795803e-05,
+      "loss": 1.0123,
+      "step": 4918
+    },
+    {
+      "epoch": 0.7544478527607362,
+      "grad_norm": 0.1646372675895691,
+      "learning_rate": 2.163892805503284e-05,
+      "loss": 1.1757,
+      "step": 4919
+    },
+    {
+      "epoch": 0.754601226993865,
+      "grad_norm": 0.16828979551792145,
+      "learning_rate": 2.1613284328825005e-05,
+      "loss": 1.1215,
+      "step": 4920
+    },
+    {
+      "epoch": 0.7547546012269939,
+      "grad_norm": 0.16115215420722961,
+      "learning_rate": 2.158765324824651e-05,
+      "loss": 1.0769,
+      "step": 4921
+    },
+    {
+      "epoch": 0.7549079754601227,
+      "grad_norm": 0.15510740876197815,
+      "learning_rate": 2.156203481936853e-05,
+      "loss": 0.9424,
+      "step": 4922
+    },
+    {
+      "epoch": 0.7550613496932516,
+      "grad_norm": 0.2407289445400238,
+      "learning_rate": 2.15364290482593e-05,
+      "loss": 0.9703,
+      "step": 4923
+    },
+    {
+      "epoch": 0.7552147239263803,
+      "grad_norm": 0.1592092365026474,
+      "learning_rate": 2.1510835940983995e-05,
+      "loss": 1.0736,
+      "step": 4924
+    },
+    {
+      "epoch": 0.7553680981595092,
+      "grad_norm": 0.1519709676504135,
+      "learning_rate": 2.1485255503604787e-05,
+      "loss": 1.0151,
+      "step": 4925
+    },
+    {
+      "epoch": 0.755521472392638,
+      "grad_norm": 0.1511695683002472,
+      "learning_rate": 2.1459687742180933e-05,
+      "loss": 0.9414,
+      "step": 4926
+    },
+    {
+      "epoch": 0.7556748466257669,
+      "grad_norm": 0.1586926430463791,
+      "learning_rate": 2.1434132662768566e-05,
+      "loss": 1.1195,
+      "step": 4927
+    },
+    {
+      "epoch": 0.7558282208588957,
+      "grad_norm": 0.16157181560993195,
+      "learning_rate": 2.1408590271420927e-05,
+      "loss": 1.0976,
+      "step": 4928
+    },
+    {
+      "epoch": 0.7559815950920246,
+      "grad_norm": 0.14870795607566833,
+      "learning_rate": 2.1383060574188194e-05,
+      "loss": 1.0858,
+      "step": 4929
+    },
+    {
+      "epoch": 0.7561349693251533,
+      "grad_norm": 0.16789990663528442,
+      "learning_rate": 2.1357543577117524e-05,
+      "loss": 1.1662,
+      "step": 4930
+    },
+    {
+      "epoch": 0.7562883435582822,
+      "grad_norm": 0.1712149977684021,
+      "learning_rate": 2.1332039286253136e-05,
+      "loss": 1.1422,
+      "step": 4931
+    },
+    {
+      "epoch": 0.756441717791411,
+      "grad_norm": 0.15163518488407135,
+      "learning_rate": 2.1306547707636127e-05,
+      "loss": 1.0339,
+      "step": 4932
+    },
+    {
+      "epoch": 0.7565950920245399,
+      "grad_norm": 0.16050104796886444,
+      "learning_rate": 2.1281068847304717e-05,
+      "loss": 1.0836,
+      "step": 4933
+    },
+    {
+      "epoch": 0.7567484662576687,
+      "grad_norm": 0.1556115299463272,
+      "learning_rate": 2.125560271129398e-05,
+      "loss": 0.9204,
+      "step": 4934
+    },
+    {
+      "epoch": 0.7569018404907976,
+      "grad_norm": 0.1556495875120163,
+      "learning_rate": 2.1230149305636097e-05,
+      "loss": 1.0511,
+      "step": 4935
+    },
+    {
+      "epoch": 0.7570552147239263,
+      "grad_norm": 0.15599294006824493,
+      "learning_rate": 2.1204708636360145e-05,
+      "loss": 0.9883,
+      "step": 4936
+    },
+    {
+      "epoch": 0.7572085889570552,
+      "grad_norm": 0.16333626210689545,
+      "learning_rate": 2.117928070949221e-05,
+      "loss": 1.0817,
+      "step": 4937
+    },
+    {
+      "epoch": 0.757361963190184,
+      "grad_norm": 0.15060292184352875,
+      "learning_rate": 2.11538655310554e-05,
+      "loss": 1.052,
+      "step": 4938
+    },
+    {
+      "epoch": 0.7575153374233129,
+      "grad_norm": 0.15757085382938385,
+      "learning_rate": 2.1128463107069716e-05,
+      "loss": 1.0449,
+      "step": 4939
+    },
+    {
+      "epoch": 0.7576687116564417,
+      "grad_norm": 0.14835135638713837,
+      "learning_rate": 2.110307344355225e-05,
+      "loss": 0.9052,
+      "step": 4940
+    },
+    {
+      "epoch": 0.7578220858895706,
+      "grad_norm": 0.1570204347372055,
+      "learning_rate": 2.1077696546516963e-05,
+      "loss": 1.1189,
+      "step": 4941
+    },
+    {
+      "epoch": 0.7579754601226993,
+      "grad_norm": 0.15273284912109375,
+      "learning_rate": 2.1052332421974886e-05,
+      "loss": 1.0693,
+      "step": 4942
+    },
+    {
+      "epoch": 0.7581288343558282,
+      "grad_norm": 0.15271270275115967,
+      "learning_rate": 2.1026981075933937e-05,
+      "loss": 1.1169,
+      "step": 4943
+    },
+    {
+      "epoch": 0.758282208588957,
+      "grad_norm": 0.16320936381816864,
+      "learning_rate": 2.100164251439909e-05,
+      "loss": 1.1074,
+      "step": 4944
+    },
+    {
+      "epoch": 0.7584355828220859,
+      "grad_norm": 0.15577249228954315,
+      "learning_rate": 2.097631674337224e-05,
+      "loss": 1.0278,
+      "step": 4945
+    },
+    {
+      "epoch": 0.7585889570552147,
+      "grad_norm": 0.15728303790092468,
+      "learning_rate": 2.0951003768852224e-05,
+      "loss": 1.0823,
+      "step": 4946
+    },
+    {
+      "epoch": 0.7587423312883436,
+      "grad_norm": 0.16765731573104858,
+      "learning_rate": 2.0925703596834953e-05,
+      "loss": 1.1711,
+      "step": 4947
+    },
+    {
+      "epoch": 0.7588957055214723,
+      "grad_norm": 0.14541569352149963,
+      "learning_rate": 2.0900416233313177e-05,
+      "loss": 1.0452,
+      "step": 4948
+    },
+    {
+      "epoch": 0.7590490797546012,
+      "grad_norm": 0.1585385799407959,
+      "learning_rate": 2.0875141684276734e-05,
+      "loss": 1.112,
+      "step": 4949
+    },
+    {
+      "epoch": 0.75920245398773,
+      "grad_norm": 0.14387774467468262,
+      "learning_rate": 2.0849879955712307e-05,
+      "loss": 0.9376,
+      "step": 4950
+    },
+    {
+      "epoch": 0.7593558282208589,
+      "grad_norm": 0.16150708496570587,
+      "learning_rate": 2.082463105360366e-05,
+      "loss": 1.2394,
+      "step": 4951
+    },
+    {
+      "epoch": 0.7595092024539877,
+      "grad_norm": 0.16476021707057953,
+      "learning_rate": 2.0799394983931435e-05,
+      "loss": 1.0996,
+      "step": 4952
+    },
+    {
+      "epoch": 0.7596625766871166,
+      "grad_norm": 0.15049542486667633,
+      "learning_rate": 2.0774171752673242e-05,
+      "loss": 0.961,
+      "step": 4953
+    },
+    {
+      "epoch": 0.7598159509202455,
+      "grad_norm": 0.16384409368038177,
+      "learning_rate": 2.074896136580371e-05,
+      "loss": 1.1655,
+      "step": 4954
+    },
+    {
+      "epoch": 0.7599693251533742,
+      "grad_norm": 0.15569357573986053,
+      "learning_rate": 2.072376382929434e-05,
+      "loss": 1.0873,
+      "step": 4955
+    },
+    {
+      "epoch": 0.7601226993865031,
+      "grad_norm": 0.14492732286453247,
+      "learning_rate": 2.0698579149113677e-05,
+      "loss": 0.9559,
+      "step": 4956
+    },
+    {
+      "epoch": 0.7602760736196319,
+      "grad_norm": 0.1517828404903412,
+      "learning_rate": 2.067340733122714e-05,
+      "loss": 1.06,
+      "step": 4957
+    },
+    {
+      "epoch": 0.7604294478527608,
+      "grad_norm": 0.15721721947193146,
+      "learning_rate": 2.0648248381597173e-05,
+      "loss": 0.995,
+      "step": 4958
+    },
+    {
+      "epoch": 0.7605828220858896,
+      "grad_norm": 0.15350444614887238,
+      "learning_rate": 2.062310230618313e-05,
+      "loss": 1.0248,
+      "step": 4959
+    },
+    {
+      "epoch": 0.7607361963190185,
+      "grad_norm": 0.1499471366405487,
+      "learning_rate": 2.059796911094128e-05,
+      "loss": 0.8659,
+      "step": 4960
+    },
+    {
+      "epoch": 0.7608895705521472,
+      "grad_norm": 0.15131284296512604,
+      "learning_rate": 2.0572848801824944e-05,
+      "loss": 0.9473,
+      "step": 4961
+    },
+    {
+      "epoch": 0.7610429447852761,
+      "grad_norm": 0.1529626101255417,
+      "learning_rate": 2.0547741384784295e-05,
+      "loss": 1.0389,
+      "step": 4962
+    },
+    {
+      "epoch": 0.7611963190184049,
+      "grad_norm": 0.14026862382888794,
+      "learning_rate": 2.052264686576651e-05,
+      "loss": 0.9719,
+      "step": 4963
+    },
+    {
+      "epoch": 0.7613496932515338,
+      "grad_norm": 0.1663055270910263,
+      "learning_rate": 2.049756525071566e-05,
+      "loss": 1.194,
+      "step": 4964
+    },
+    {
+      "epoch": 0.7615030674846626,
+      "grad_norm": 0.1516270786523819,
+      "learning_rate": 2.047249654557283e-05,
+      "loss": 1.0337,
+      "step": 4965
+    },
+    {
+      "epoch": 0.7616564417177915,
+      "grad_norm": 0.16902519762516022,
+      "learning_rate": 2.044744075627595e-05,
+      "loss": 0.9519,
+      "step": 4966
+    },
+    {
+      "epoch": 0.7618098159509202,
+      "grad_norm": 0.15783967077732086,
+      "learning_rate": 2.0422397888760018e-05,
+      "loss": 1.0001,
+      "step": 4967
+    },
+    {
+      "epoch": 0.7619631901840491,
+      "grad_norm": 0.15725120902061462,
+      "learning_rate": 2.039736794895685e-05,
+      "loss": 1.084,
+      "step": 4968
+    },
+    {
+      "epoch": 0.7621165644171779,
+      "grad_norm": 0.1532997190952301,
+      "learning_rate": 2.037235094279524e-05,
+      "loss": 0.9764,
+      "step": 4969
+    },
+    {
+      "epoch": 0.7622699386503068,
+      "grad_norm": 0.15504780411720276,
+      "learning_rate": 2.0347346876200947e-05,
+      "loss": 0.9226,
+      "step": 4970
+    },
+    {
+      "epoch": 0.7624233128834356,
+      "grad_norm": 0.15956611931324005,
+      "learning_rate": 2.032235575509667e-05,
+      "loss": 1.0851,
+      "step": 4971
+    },
+    {
+      "epoch": 0.7625766871165645,
+      "grad_norm": 0.15383020043373108,
+      "learning_rate": 2.029737758540197e-05,
+      "loss": 1.1079,
+      "step": 4972
+    },
+    {
+      "epoch": 0.7627300613496932,
+      "grad_norm": 0.1591903418302536,
+      "learning_rate": 2.027241237303343e-05,
+      "loss": 0.9866,
+      "step": 4973
+    },
+    {
+      "epoch": 0.7628834355828221,
+      "grad_norm": 0.1489236056804657,
+      "learning_rate": 2.024746012390449e-05,
+      "loss": 1.0006,
+      "step": 4974
+    },
+    {
+      "epoch": 0.7630368098159509,
+      "grad_norm": 0.1624889075756073,
+      "learning_rate": 2.022252084392558e-05,
+      "loss": 1.0326,
+      "step": 4975
+    },
+    {
+      "epoch": 0.7631901840490798,
+      "grad_norm": 0.13921262323856354,
+      "learning_rate": 2.0197594539003996e-05,
+      "loss": 0.8781,
+      "step": 4976
+    },
+    {
+      "epoch": 0.7633435582822086,
+      "grad_norm": 0.16525699198246002,
+      "learning_rate": 2.017268121504403e-05,
+      "loss": 1.1188,
+      "step": 4977
+    },
+    {
+      "epoch": 0.7634969325153375,
+      "grad_norm": 0.16340181231498718,
+      "learning_rate": 2.014778087794682e-05,
+      "loss": 1.1694,
+      "step": 4978
+    },
+    {
+      "epoch": 0.7636503067484662,
+      "grad_norm": 0.1794653981924057,
+      "learning_rate": 2.0122893533610514e-05,
+      "loss": 1.096,
+      "step": 4979
+    },
+    {
+      "epoch": 0.7638036809815951,
+      "grad_norm": 0.1529432088136673,
+      "learning_rate": 2.0098019187930124e-05,
+      "loss": 1.0662,
+      "step": 4980
+    },
+    {
+      "epoch": 0.7639570552147239,
+      "grad_norm": 0.14733092486858368,
+      "learning_rate": 2.0073157846797562e-05,
+      "loss": 0.9495,
+      "step": 4981
+    },
+    {
+      "epoch": 0.7641104294478528,
+      "grad_norm": 0.1484023630619049,
+      "learning_rate": 2.0048309516101756e-05,
+      "loss": 0.9987,
+      "step": 4982
+    },
+    {
+      "epoch": 0.7642638036809816,
+      "grad_norm": 0.1503826528787613,
+      "learning_rate": 2.0023474201728446e-05,
+      "loss": 0.9495,
+      "step": 4983
+    },
+    {
+      "epoch": 0.7644171779141105,
+      "grad_norm": 0.1477353721857071,
+      "learning_rate": 1.9998651909560365e-05,
+      "loss": 0.9543,
+      "step": 4984
+    },
+    {
+      "epoch": 0.7645705521472392,
+      "grad_norm": 0.15062960982322693,
+      "learning_rate": 1.9973842645477094e-05,
+      "loss": 1.0401,
+      "step": 4985
+    },
+    {
+      "epoch": 0.7647239263803681,
+      "grad_norm": 0.13873188197612762,
+      "learning_rate": 1.9949046415355217e-05,
+      "loss": 0.9512,
+      "step": 4986
+    },
+    {
+      "epoch": 0.7648773006134969,
+      "grad_norm": 0.17755068838596344,
+      "learning_rate": 1.9924263225068154e-05,
+      "loss": 1.0735,
+      "step": 4987
+    },
+    {
+      "epoch": 0.7650306748466258,
+      "grad_norm": 0.15398763120174408,
+      "learning_rate": 1.9899493080486232e-05,
+      "loss": 1.1241,
+      "step": 4988
+    },
+    {
+      "epoch": 0.7651840490797546,
+      "grad_norm": 0.14475573599338531,
+      "learning_rate": 1.9874735987476763e-05,
+      "loss": 0.959,
+      "step": 4989
+    },
+    {
+      "epoch": 0.7653374233128835,
+      "grad_norm": 0.15250547230243683,
+      "learning_rate": 1.984999195190388e-05,
+      "loss": 0.9751,
+      "step": 4990
+    },
+    {
+      "epoch": 0.7654907975460122,
+      "grad_norm": 0.1546533703804016,
+      "learning_rate": 1.9825260979628703e-05,
+      "loss": 1.1167,
+      "step": 4991
+    },
+    {
+      "epoch": 0.7656441717791411,
+      "grad_norm": 0.15204840898513794,
+      "learning_rate": 1.980054307650919e-05,
+      "loss": 0.9877,
+      "step": 4992
+    },
+    {
+      "epoch": 0.7657975460122699,
+      "grad_norm": 0.1659376472234726,
+      "learning_rate": 1.977583824840026e-05,
+      "loss": 1.0141,
+      "step": 4993
+    },
+    {
+      "epoch": 0.7659509202453988,
+      "grad_norm": 0.14427024126052856,
+      "learning_rate": 1.97511465011537e-05,
+      "loss": 0.9532,
+      "step": 4994
+    },
+    {
+      "epoch": 0.7661042944785276,
+      "grad_norm": 0.14955876767635345,
+      "learning_rate": 1.9726467840618175e-05,
+      "loss": 0.9476,
+      "step": 4995
+    },
+    {
+      "epoch": 0.7662576687116565,
+      "grad_norm": 0.15106454491615295,
+      "learning_rate": 1.970180227263933e-05,
+      "loss": 1.0239,
+      "step": 4996
+    },
+    {
+      "epoch": 0.7664110429447852,
+      "grad_norm": 0.16141484677791595,
+      "learning_rate": 1.9677149803059617e-05,
+      "loss": 1.1808,
+      "step": 4997
+    },
+    {
+      "epoch": 0.7665644171779141,
+      "grad_norm": 0.14750875532627106,
+      "learning_rate": 1.9652510437718464e-05,
+      "loss": 0.9967,
+      "step": 4998
+    },
+    {
+      "epoch": 0.7667177914110429,
+      "grad_norm": 0.16225315630435944,
+      "learning_rate": 1.962788418245212e-05,
+      "loss": 1.1264,
+      "step": 4999
+    },
+    {
+      "epoch": 0.7668711656441718,
+      "grad_norm": 0.1554587036371231,
+      "learning_rate": 1.9603271043093817e-05,
+      "loss": 1.0628,
+      "step": 5000
+    },
+    {
+      "epoch": 0.7670245398773006,
+      "grad_norm": 0.15829472243785858,
+      "learning_rate": 1.957867102547358e-05,
+      "loss": 1.052,
+      "step": 5001
+    },
+    {
+      "epoch": 0.7671779141104295,
+      "grad_norm": 0.16758573055267334,
+      "learning_rate": 1.9554084135418426e-05,
+      "loss": 1.0469,
+      "step": 5002
+    },
+    {
+      "epoch": 0.7673312883435582,
+      "grad_norm": 0.15607556700706482,
+      "learning_rate": 1.95295103787522e-05,
+      "loss": 1.0082,
+      "step": 5003
+    },
+    {
+      "epoch": 0.7674846625766871,
+      "grad_norm": 0.16924582421779633,
+      "learning_rate": 1.950494976129561e-05,
+      "loss": 1.2042,
+      "step": 5004
+    },
+    {
+      "epoch": 0.7676380368098159,
+      "grad_norm": 0.15887904167175293,
+      "learning_rate": 1.9480402288866346e-05,
+      "loss": 1.0701,
+      "step": 5005
+    },
+    {
+      "epoch": 0.7677914110429448,
+      "grad_norm": 0.15378306806087494,
+      "learning_rate": 1.9455867967278887e-05,
+      "loss": 0.9929,
+      "step": 5006
+    },
+    {
+      "epoch": 0.7679447852760736,
+      "grad_norm": 0.15678109228610992,
+      "learning_rate": 1.943134680234469e-05,
+      "loss": 0.9038,
+      "step": 5007
+    },
+    {
+      "epoch": 0.7680981595092025,
+      "grad_norm": 0.1590825915336609,
+      "learning_rate": 1.9406838799871985e-05,
+      "loss": 1.0085,
+      "step": 5008
+    },
+    {
+      "epoch": 0.7682515337423312,
+      "grad_norm": 0.1817246824502945,
+      "learning_rate": 1.9382343965666e-05,
+      "loss": 1.0952,
+      "step": 5009
+    },
+    {
+      "epoch": 0.7684049079754601,
+      "grad_norm": 0.15530440211296082,
+      "learning_rate": 1.935786230552876e-05,
+      "loss": 1.0925,
+      "step": 5010
+    },
+    {
+      "epoch": 0.7685582822085889,
+      "grad_norm": 0.16113033890724182,
+      "learning_rate": 1.9333393825259184e-05,
+      "loss": 1.0266,
+      "step": 5011
+    },
+    {
+      "epoch": 0.7687116564417178,
+      "grad_norm": 0.1644577831029892,
+      "learning_rate": 1.9308938530653126e-05,
+      "loss": 0.987,
+      "step": 5012
+    },
+    {
+      "epoch": 0.7688650306748466,
+      "grad_norm": 0.15382815897464752,
+      "learning_rate": 1.9284496427503204e-05,
+      "loss": 1.0828,
+      "step": 5013
+    },
+    {
+      "epoch": 0.7690184049079755,
+      "grad_norm": 0.16416391730308533,
+      "learning_rate": 1.9260067521599072e-05,
+      "loss": 1.0468,
+      "step": 5014
+    },
+    {
+      "epoch": 0.7691717791411042,
+      "grad_norm": 0.14969447255134583,
+      "learning_rate": 1.923565181872712e-05,
+      "loss": 1.062,
+      "step": 5015
+    },
+    {
+      "epoch": 0.7693251533742331,
+      "grad_norm": 0.15253598988056183,
+      "learning_rate": 1.9211249324670647e-05,
+      "loss": 0.9418,
+      "step": 5016
+    },
+    {
+      "epoch": 0.769478527607362,
+      "grad_norm": 0.15605583786964417,
+      "learning_rate": 1.9186860045209863e-05,
+      "loss": 1.0051,
+      "step": 5017
+    },
+    {
+      "epoch": 0.7696319018404908,
+      "grad_norm": 0.16020622849464417,
+      "learning_rate": 1.9162483986121786e-05,
+      "loss": 1.1374,
+      "step": 5018
+    },
+    {
+      "epoch": 0.7697852760736197,
+      "grad_norm": 0.14846539497375488,
+      "learning_rate": 1.913812115318038e-05,
+      "loss": 0.9713,
+      "step": 5019
+    },
+    {
+      "epoch": 0.7699386503067485,
+      "grad_norm": 0.16731469333171844,
+      "learning_rate": 1.9113771552156378e-05,
+      "loss": 1.0056,
+      "step": 5020
+    },
+    {
+      "epoch": 0.7700920245398774,
+      "grad_norm": 0.15272608399391174,
+      "learning_rate": 1.9089435188817487e-05,
+      "loss": 0.9701,
+      "step": 5021
+    },
+    {
+      "epoch": 0.7702453987730061,
+      "grad_norm": 0.16156768798828125,
+      "learning_rate": 1.906511206892819e-05,
+      "loss": 1.1121,
+      "step": 5022
+    },
+    {
+      "epoch": 0.770398773006135,
+      "grad_norm": 0.17605283856391907,
+      "learning_rate": 1.904080219824986e-05,
+      "loss": 1.1701,
+      "step": 5023
+    },
+    {
+      "epoch": 0.7705521472392638,
+      "grad_norm": 0.15134312212467194,
+      "learning_rate": 1.9016505582540774e-05,
+      "loss": 0.9912,
+      "step": 5024
+    },
+    {
+      "epoch": 0.7707055214723927,
+      "grad_norm": 0.1574927121400833,
+      "learning_rate": 1.8992222227555992e-05,
+      "loss": 1.0123,
+      "step": 5025
+    },
+    {
+      "epoch": 0.7708588957055215,
+      "grad_norm": 0.15418072044849396,
+      "learning_rate": 1.8967952139047516e-05,
+      "loss": 1.0464,
+      "step": 5026
+    },
+    {
+      "epoch": 0.7710122699386504,
+      "grad_norm": 0.14263483881950378,
+      "learning_rate": 1.8943695322764123e-05,
+      "loss": 0.9191,
+      "step": 5027
+    },
+    {
+      "epoch": 0.7711656441717791,
+      "grad_norm": 0.1592971831560135,
+      "learning_rate": 1.891945178445153e-05,
+      "loss": 1.1011,
+      "step": 5028
+    },
+    {
+      "epoch": 0.771319018404908,
+      "grad_norm": 0.17282187938690186,
+      "learning_rate": 1.8895221529852254e-05,
+      "loss": 1.1938,
+      "step": 5029
+    },
+    {
+      "epoch": 0.7714723926380368,
+      "grad_norm": 0.1621592789888382,
+      "learning_rate": 1.8871004564705638e-05,
+      "loss": 1.076,
+      "step": 5030
+    },
+    {
+      "epoch": 0.7716257668711657,
+      "grad_norm": 0.16386282444000244,
+      "learning_rate": 1.884680089474798e-05,
+      "loss": 1.121,
+      "step": 5031
+    },
+    {
+      "epoch": 0.7717791411042945,
+      "grad_norm": 0.14998912811279297,
+      "learning_rate": 1.8822610525712316e-05,
+      "loss": 1.0567,
+      "step": 5032
+    },
+    {
+      "epoch": 0.7719325153374234,
+      "grad_norm": 0.1674668788909912,
+      "learning_rate": 1.879843346332861e-05,
+      "loss": 1.164,
+      "step": 5033
+    },
+    {
+      "epoch": 0.7720858895705521,
+      "grad_norm": 0.1555359959602356,
+      "learning_rate": 1.8774269713323632e-05,
+      "loss": 1.1797,
+      "step": 5034
+    },
+    {
+      "epoch": 0.772239263803681,
+      "grad_norm": 0.14323054254055023,
+      "learning_rate": 1.875011928142103e-05,
+      "loss": 0.9791,
+      "step": 5035
+    },
+    {
+      "epoch": 0.7723926380368098,
+      "grad_norm": 0.15449552237987518,
+      "learning_rate": 1.872598217334125e-05,
+      "loss": 1.0088,
+      "step": 5036
+    },
+    {
+      "epoch": 0.7725460122699387,
+      "grad_norm": 0.15940268337726593,
+      "learning_rate": 1.8701858394801643e-05,
+      "loss": 1.0892,
+      "step": 5037
+    },
+    {
+      "epoch": 0.7726993865030675,
+      "grad_norm": 0.1548135131597519,
+      "learning_rate": 1.8677747951516366e-05,
+      "loss": 0.9904,
+      "step": 5038
+    },
+    {
+      "epoch": 0.7728527607361964,
+      "grad_norm": 0.15277628600597382,
+      "learning_rate": 1.8653650849196392e-05,
+      "loss": 1.0432,
+      "step": 5039
+    },
+    {
+      "epoch": 0.7730061349693251,
+      "grad_norm": 0.14695514738559723,
+      "learning_rate": 1.86295670935496e-05,
+      "loss": 0.9965,
+      "step": 5040
+    },
+    {
+      "epoch": 0.773159509202454,
+      "grad_norm": 0.14430734515190125,
+      "learning_rate": 1.8605496690280643e-05,
+      "loss": 0.9969,
+      "step": 5041
+    },
+    {
+      "epoch": 0.7733128834355828,
+      "grad_norm": 0.15836571156978607,
+      "learning_rate": 1.858143964509108e-05,
+      "loss": 0.9988,
+      "step": 5042
+    },
+    {
+      "epoch": 0.7734662576687117,
+      "grad_norm": 0.17474053800106049,
+      "learning_rate": 1.855739596367922e-05,
+      "loss": 1.1584,
+      "step": 5043
+    },
+    {
+      "epoch": 0.7736196319018405,
+      "grad_norm": 0.17450954020023346,
+      "learning_rate": 1.8533365651740294e-05,
+      "loss": 1.0821,
+      "step": 5044
+    },
+    {
+      "epoch": 0.7737730061349694,
+      "grad_norm": 0.15599225461483002,
+      "learning_rate": 1.850934871496631e-05,
+      "loss": 1.0186,
+      "step": 5045
+    },
+    {
+      "epoch": 0.7739263803680981,
+      "grad_norm": 0.1646985113620758,
+      "learning_rate": 1.8485345159046096e-05,
+      "loss": 1.128,
+      "step": 5046
+    },
+    {
+      "epoch": 0.774079754601227,
+      "grad_norm": 0.16185422241687775,
+      "learning_rate": 1.8461354989665384e-05,
+      "loss": 1.0462,
+      "step": 5047
+    },
+    {
+      "epoch": 0.7742331288343558,
+      "grad_norm": 0.15563446283340454,
+      "learning_rate": 1.8437378212506642e-05,
+      "loss": 1.0142,
+      "step": 5048
+    },
+    {
+      "epoch": 0.7743865030674847,
+      "grad_norm": 0.1617828905582428,
+      "learning_rate": 1.841341483324926e-05,
+      "loss": 1.1841,
+      "step": 5049
+    },
+    {
+      "epoch": 0.7745398773006135,
+      "grad_norm": 0.15550780296325684,
+      "learning_rate": 1.8389464857569362e-05,
+      "loss": 1.0517,
+      "step": 5050
+    },
+    {
+      "epoch": 0.7746932515337424,
+      "grad_norm": 0.15321731567382812,
+      "learning_rate": 1.836552829113998e-05,
+      "loss": 0.989,
+      "step": 5051
+    },
+    {
+      "epoch": 0.7748466257668711,
+      "grad_norm": 0.1605292409658432,
+      "learning_rate": 1.8341605139630912e-05,
+      "loss": 1.0886,
+      "step": 5052
+    },
+    {
+      "epoch": 0.775,
+      "grad_norm": 0.16210776567459106,
+      "learning_rate": 1.8317695408708785e-05,
+      "loss": 1.0444,
+      "step": 5053
+    },
+    {
+      "epoch": 0.7751533742331288,
+      "grad_norm": 0.16230027377605438,
+      "learning_rate": 1.82937991040371e-05,
+      "loss": 1.1125,
+      "step": 5054
+    },
+    {
+      "epoch": 0.7753067484662577,
+      "grad_norm": 0.1594068557024002,
+      "learning_rate": 1.8269916231276068e-05,
+      "loss": 1.1326,
+      "step": 5055
+    },
+    {
+      "epoch": 0.7754601226993865,
+      "grad_norm": 0.1625036597251892,
+      "learning_rate": 1.824604679608288e-05,
+      "loss": 0.9647,
+      "step": 5056
+    },
+    {
+      "epoch": 0.7756134969325154,
+      "grad_norm": 0.14693734049797058,
+      "learning_rate": 1.8222190804111398e-05,
+      "loss": 1.0338,
+      "step": 5057
+    },
+    {
+      "epoch": 0.7757668711656441,
+      "grad_norm": 0.15117250382900238,
+      "learning_rate": 1.819834826101235e-05,
+      "loss": 1.059,
+      "step": 5058
+    },
+    {
+      "epoch": 0.775920245398773,
+      "grad_norm": 0.15295112133026123,
+      "learning_rate": 1.8174519172433308e-05,
+      "loss": 0.9907,
+      "step": 5059
+    },
+    {
+      "epoch": 0.7760736196319018,
+      "grad_norm": 0.16680429875850677,
+      "learning_rate": 1.81507035440186e-05,
+      "loss": 1.178,
+      "step": 5060
+    },
+    {
+      "epoch": 0.7762269938650307,
+      "grad_norm": 0.13648228347301483,
+      "learning_rate": 1.8126901381409446e-05,
+      "loss": 0.9046,
+      "step": 5061
+    },
+    {
+      "epoch": 0.7763803680981595,
+      "grad_norm": 0.1637558788061142,
+      "learning_rate": 1.8103112690243766e-05,
+      "loss": 1.1567,
+      "step": 5062
+    },
+    {
+      "epoch": 0.7765337423312884,
+      "grad_norm": 0.18085458874702454,
+      "learning_rate": 1.807933747615641e-05,
+      "loss": 1.1051,
+      "step": 5063
+    },
+    {
+      "epoch": 0.7766871165644171,
+      "grad_norm": 0.16263402998447418,
+      "learning_rate": 1.8055575744778946e-05,
+      "loss": 1.1113,
+      "step": 5064
+    },
+    {
+      "epoch": 0.776840490797546,
+      "grad_norm": 0.14077363908290863,
+      "learning_rate": 1.8031827501739766e-05,
+      "loss": 0.9538,
+      "step": 5065
+    },
+    {
+      "epoch": 0.7769938650306748,
+      "grad_norm": 0.14812524616718292,
+      "learning_rate": 1.8008092752664115e-05,
+      "loss": 0.9987,
+      "step": 5066
+    },
+    {
+      "epoch": 0.7771472392638037,
+      "grad_norm": 1.4554287195205688,
+      "learning_rate": 1.7984371503173957e-05,
+      "loss": 1.1069,
+      "step": 5067
+    },
+    {
+      "epoch": 0.7773006134969325,
+      "grad_norm": 0.14712680876255035,
+      "learning_rate": 1.7960663758888172e-05,
+      "loss": 1.0165,
+      "step": 5068
+    },
+    {
+      "epoch": 0.7774539877300614,
+      "grad_norm": 0.17158052325248718,
+      "learning_rate": 1.7936969525422316e-05,
+      "loss": 1.2228,
+      "step": 5069
+    },
+    {
+      "epoch": 0.7776073619631901,
+      "grad_norm": 0.15102095901966095,
+      "learning_rate": 1.7913288808388855e-05,
+      "loss": 1.0155,
+      "step": 5070
+    },
+    {
+      "epoch": 0.777760736196319,
+      "grad_norm": 0.1621231585741043,
+      "learning_rate": 1.7889621613396984e-05,
+      "loss": 1.0932,
+      "step": 5071
+    },
+    {
+      "epoch": 0.7779141104294478,
+      "grad_norm": 0.15455608069896698,
+      "learning_rate": 1.786596794605269e-05,
+      "loss": 1.0682,
+      "step": 5072
+    },
+    {
+      "epoch": 0.7780674846625767,
+      "grad_norm": 0.16021138429641724,
+      "learning_rate": 1.784232781195882e-05,
+      "loss": 1.0333,
+      "step": 5073
+    },
+    {
+      "epoch": 0.7782208588957055,
+      "grad_norm": 0.165344700217247,
+      "learning_rate": 1.7818701216714947e-05,
+      "loss": 1.072,
+      "step": 5074
+    },
+    {
+      "epoch": 0.7783742331288344,
+      "grad_norm": 0.14591671526432037,
+      "learning_rate": 1.7795088165917504e-05,
+      "loss": 0.9912,
+      "step": 5075
+    },
+    {
+      "epoch": 0.7785276073619631,
+      "grad_norm": 0.16486574709415436,
+      "learning_rate": 1.7771488665159627e-05,
+      "loss": 1.1083,
+      "step": 5076
+    },
+    {
+      "epoch": 0.778680981595092,
+      "grad_norm": 0.15570954978466034,
+      "learning_rate": 1.774790272003134e-05,
+      "loss": 1.064,
+      "step": 5077
+    },
+    {
+      "epoch": 0.7788343558282208,
+      "grad_norm": 0.15804103016853333,
+      "learning_rate": 1.772433033611937e-05,
+      "loss": 1.0913,
+      "step": 5078
+    },
+    {
+      "epoch": 0.7789877300613497,
+      "grad_norm": 0.14358454942703247,
+      "learning_rate": 1.770077151900731e-05,
+      "loss": 1.0314,
+      "step": 5079
+    },
+    {
+      "epoch": 0.7791411042944786,
+      "grad_norm": 0.1525515913963318,
+      "learning_rate": 1.7677226274275487e-05,
+      "loss": 0.9818,
+      "step": 5080
+    },
+    {
+      "epoch": 0.7792944785276074,
+      "grad_norm": 0.17292633652687073,
+      "learning_rate": 1.7653694607501e-05,
+      "loss": 1.2522,
+      "step": 5081
+    },
+    {
+      "epoch": 0.7794478527607362,
+      "grad_norm": 0.15385732054710388,
+      "learning_rate": 1.7630176524257792e-05,
+      "loss": 1.0428,
+      "step": 5082
+    },
+    {
+      "epoch": 0.779601226993865,
+      "grad_norm": 0.16042907536029816,
+      "learning_rate": 1.7606672030116536e-05,
+      "loss": 1.156,
+      "step": 5083
+    },
+    {
+      "epoch": 0.7797546012269939,
+      "grad_norm": 0.15725523233413696,
+      "learning_rate": 1.758318113064473e-05,
+      "loss": 1.182,
+      "step": 5084
+    },
+    {
+      "epoch": 0.7799079754601227,
+      "grad_norm": 0.15973664820194244,
+      "learning_rate": 1.755970383140658e-05,
+      "loss": 1.0829,
+      "step": 5085
+    },
+    {
+      "epoch": 0.7800613496932516,
+      "grad_norm": 0.15760840475559235,
+      "learning_rate": 1.7536240137963177e-05,
+      "loss": 1.064,
+      "step": 5086
+    },
+    {
+      "epoch": 0.7802147239263804,
+      "grad_norm": 0.1493833363056183,
+      "learning_rate": 1.7512790055872304e-05,
+      "loss": 1.0209,
+      "step": 5087
+    },
+    {
+      "epoch": 0.7803680981595092,
+      "grad_norm": 0.1575067639350891,
+      "learning_rate": 1.748935359068852e-05,
+      "loss": 1.0711,
+      "step": 5088
+    },
+    {
+      "epoch": 0.780521472392638,
+      "grad_norm": 0.15666989982128143,
+      "learning_rate": 1.7465930747963232e-05,
+      "loss": 1.0101,
+      "step": 5089
+    },
+    {
+      "epoch": 0.7806748466257669,
+      "grad_norm": 0.14693285524845123,
+      "learning_rate": 1.7442521533244522e-05,
+      "loss": 1.0644,
+      "step": 5090
+    },
+    {
+      "epoch": 0.7808282208588957,
+      "grad_norm": 0.16005074977874756,
+      "learning_rate": 1.7419125952077348e-05,
+      "loss": 1.134,
+      "step": 5091
+    },
+    {
+      "epoch": 0.7809815950920246,
+      "grad_norm": 0.15427349507808685,
+      "learning_rate": 1.7395744010003327e-05,
+      "loss": 1.0041,
+      "step": 5092
+    },
+    {
+      "epoch": 0.7811349693251534,
+      "grad_norm": 0.14965461194515228,
+      "learning_rate": 1.737237571256097e-05,
+      "loss": 0.9517,
+      "step": 5093
+    },
+    {
+      "epoch": 0.7812883435582823,
+      "grad_norm": 0.14859610795974731,
+      "learning_rate": 1.7349021065285444e-05,
+      "loss": 1.099,
+      "step": 5094
+    },
+    {
+      "epoch": 0.781441717791411,
+      "grad_norm": 0.1567378044128418,
+      "learning_rate": 1.7325680073708727e-05,
+      "loss": 1.0644,
+      "step": 5095
+    },
+    {
+      "epoch": 0.7815950920245399,
+      "grad_norm": 0.1572311818599701,
+      "learning_rate": 1.730235274335959e-05,
+      "loss": 0.9216,
+      "step": 5096
+    },
+    {
+      "epoch": 0.7817484662576687,
+      "grad_norm": 0.13857290148735046,
+      "learning_rate": 1.7279039079763512e-05,
+      "loss": 0.9591,
+      "step": 5097
+    },
+    {
+      "epoch": 0.7819018404907976,
+      "grad_norm": 0.15886491537094116,
+      "learning_rate": 1.7255739088442784e-05,
+      "loss": 1.0585,
+      "step": 5098
+    },
+    {
+      "epoch": 0.7820552147239264,
+      "grad_norm": 0.16191424429416656,
+      "learning_rate": 1.7232452774916454e-05,
+      "loss": 1.1166,
+      "step": 5099
+    },
+    {
+      "epoch": 0.7822085889570553,
+      "grad_norm": 0.16922032833099365,
+      "learning_rate": 1.7209180144700278e-05,
+      "loss": 1.1615,
+      "step": 5100
+    },
+    {
+      "epoch": 0.782361963190184,
+      "grad_norm": 0.14725349843502045,
+      "learning_rate": 1.7185921203306842e-05,
+      "loss": 0.8973,
+      "step": 5101
+    },
+    {
+      "epoch": 0.7825153374233129,
+      "grad_norm": 0.15577849745750427,
+      "learning_rate": 1.7162675956245424e-05,
+      "loss": 0.9336,
+      "step": 5102
+    },
+    {
+      "epoch": 0.7826687116564417,
+      "grad_norm": 0.1478036493062973,
+      "learning_rate": 1.7139444409022135e-05,
+      "loss": 0.9178,
+      "step": 5103
+    },
+    {
+      "epoch": 0.7828220858895706,
+      "grad_norm": 0.1523665338754654,
+      "learning_rate": 1.7116226567139734e-05,
+      "loss": 0.9164,
+      "step": 5104
+    },
+    {
+      "epoch": 0.7829754601226994,
+      "grad_norm": 0.16216666996479034,
+      "learning_rate": 1.7093022436097865e-05,
+      "loss": 1.0783,
+      "step": 5105
+    },
+    {
+      "epoch": 0.7831288343558283,
+      "grad_norm": 0.2763657867908478,
+      "learning_rate": 1.706983202139281e-05,
+      "loss": 1.1548,
+      "step": 5106
+    },
+    {
+      "epoch": 0.783282208588957,
+      "grad_norm": 0.15616950392723083,
+      "learning_rate": 1.7046655328517637e-05,
+      "loss": 1.1506,
+      "step": 5107
+    },
+    {
+      "epoch": 0.7834355828220859,
+      "grad_norm": 0.1622968465089798,
+      "learning_rate": 1.702349236296222e-05,
+      "loss": 1.1079,
+      "step": 5108
+    },
+    {
+      "epoch": 0.7835889570552147,
+      "grad_norm": 0.15844421088695526,
+      "learning_rate": 1.7000343130213075e-05,
+      "loss": 1.0997,
+      "step": 5109
+    },
+    {
+      "epoch": 0.7837423312883436,
+      "grad_norm": 0.15583525598049164,
+      "learning_rate": 1.697720763575359e-05,
+      "loss": 0.9951,
+      "step": 5110
+    },
+    {
+      "epoch": 0.7838957055214724,
+      "grad_norm": 0.15496324002742767,
+      "learning_rate": 1.6954085885063774e-05,
+      "loss": 1.0652,
+      "step": 5111
+    },
+    {
+      "epoch": 0.7840490797546013,
+      "grad_norm": 0.15239845216274261,
+      "learning_rate": 1.6930977883620496e-05,
+      "loss": 1.0723,
+      "step": 5112
+    },
+    {
+      "epoch": 0.78420245398773,
+      "grad_norm": 0.16520878672599792,
+      "learning_rate": 1.690788363689726e-05,
+      "loss": 1.1269,
+      "step": 5113
+    },
+    {
+      "epoch": 0.7843558282208589,
+      "grad_norm": 0.1584097295999527,
+      "learning_rate": 1.688480315036442e-05,
+      "loss": 0.9337,
+      "step": 5114
+    },
+    {
+      "epoch": 0.7845092024539877,
+      "grad_norm": 0.1595587283372879,
+      "learning_rate": 1.6861736429488977e-05,
+      "loss": 1.142,
+      "step": 5115
+    },
+    {
+      "epoch": 0.7846625766871166,
+      "grad_norm": 0.15035440027713776,
+      "learning_rate": 1.6838683479734706e-05,
+      "loss": 1.0071,
+      "step": 5116
+    },
+    {
+      "epoch": 0.7848159509202454,
+      "grad_norm": 0.15220750868320465,
+      "learning_rate": 1.6815644306562148e-05,
+      "loss": 0.9644,
+      "step": 5117
+    },
+    {
+      "epoch": 0.7849693251533743,
+      "grad_norm": 0.14775633811950684,
+      "learning_rate": 1.679261891542853e-05,
+      "loss": 1.004,
+      "step": 5118
+    },
+    {
+      "epoch": 0.785122699386503,
+      "grad_norm": 0.15227387845516205,
+      "learning_rate": 1.6769607311787876e-05,
+      "loss": 1.0133,
+      "step": 5119
+    },
+    {
+      "epoch": 0.7852760736196319,
+      "grad_norm": 0.14613862335681915,
+      "learning_rate": 1.6746609501090866e-05,
+      "loss": 1.0526,
+      "step": 5120
+    },
+    {
+      "epoch": 0.7854294478527607,
+      "grad_norm": 0.1581389158964157,
+      "learning_rate": 1.6723625488785e-05,
+      "loss": 0.9815,
+      "step": 5121
+    },
+    {
+      "epoch": 0.7855828220858896,
+      "grad_norm": 0.22287459671497345,
+      "learning_rate": 1.6700655280314452e-05,
+      "loss": 0.9468,
+      "step": 5122
+    },
+    {
+      "epoch": 0.7857361963190184,
+      "grad_norm": 0.15954537689685822,
+      "learning_rate": 1.6677698881120107e-05,
+      "loss": 1.0815,
+      "step": 5123
+    },
+    {
+      "epoch": 0.7858895705521473,
+      "grad_norm": 0.14720569550991058,
+      "learning_rate": 1.6654756296639664e-05,
+      "loss": 1.0074,
+      "step": 5124
+    },
+    {
+      "epoch": 0.786042944785276,
+      "grad_norm": 0.16058772802352905,
+      "learning_rate": 1.6631827532307443e-05,
+      "loss": 1.1045,
+      "step": 5125
+    },
+    {
+      "epoch": 0.7861963190184049,
+      "grad_norm": 0.16769537329673767,
+      "learning_rate": 1.6608912593554596e-05,
+      "loss": 1.1004,
+      "step": 5126
+    },
+    {
+      "epoch": 0.7863496932515337,
+      "grad_norm": 0.16153274476528168,
+      "learning_rate": 1.6586011485808913e-05,
+      "loss": 1.1511,
+      "step": 5127
+    },
+    {
+      "epoch": 0.7865030674846626,
+      "grad_norm": 0.17218117415905,
+      "learning_rate": 1.6563124214494973e-05,
+      "loss": 1.2408,
+      "step": 5128
+    },
+    {
+      "epoch": 0.7866564417177914,
+      "grad_norm": 0.15736933052539825,
+      "learning_rate": 1.654025078503404e-05,
+      "loss": 1.0498,
+      "step": 5129
+    },
+    {
+      "epoch": 0.7868098159509203,
+      "grad_norm": 0.15207502245903015,
+      "learning_rate": 1.6517391202844076e-05,
+      "loss": 1.0709,
+      "step": 5130
+    },
+    {
+      "epoch": 0.786963190184049,
+      "grad_norm": 0.13978588581085205,
+      "learning_rate": 1.6494545473339845e-05,
+      "loss": 0.9723,
+      "step": 5131
+    },
+    {
+      "epoch": 0.7871165644171779,
+      "grad_norm": 0.16792739927768707,
+      "learning_rate": 1.6471713601932735e-05,
+      "loss": 1.0223,
+      "step": 5132
+    },
+    {
+      "epoch": 0.7872699386503067,
+      "grad_norm": 0.1657523810863495,
+      "learning_rate": 1.6448895594030942e-05,
+      "loss": 1.0715,
+      "step": 5133
+    },
+    {
+      "epoch": 0.7874233128834356,
+      "grad_norm": 0.15719912946224213,
+      "learning_rate": 1.642609145503928e-05,
+      "loss": 1.0279,
+      "step": 5134
+    },
+    {
+      "epoch": 0.7875766871165644,
+      "grad_norm": 0.18613675236701965,
+      "learning_rate": 1.6403301190359388e-05,
+      "loss": 1.3001,
+      "step": 5135
+    },
+    {
+      "epoch": 0.7877300613496933,
+      "grad_norm": 0.16254331171512604,
+      "learning_rate": 1.6380524805389507e-05,
+      "loss": 1.0781,
+      "step": 5136
+    },
+    {
+      "epoch": 0.787883435582822,
+      "grad_norm": 0.1525297611951828,
+      "learning_rate": 1.635776230552469e-05,
+      "loss": 1.0921,
+      "step": 5137
+    },
+    {
+      "epoch": 0.7880368098159509,
+      "grad_norm": 0.16951067745685577,
+      "learning_rate": 1.633501369615663e-05,
+      "loss": 1.2108,
+      "step": 5138
+    },
+    {
+      "epoch": 0.7881901840490797,
+      "grad_norm": 0.1716470569372177,
+      "learning_rate": 1.6312278982673743e-05,
+      "loss": 1.2493,
+      "step": 5139
+    },
+    {
+      "epoch": 0.7883435582822086,
+      "grad_norm": 0.15147896111011505,
+      "learning_rate": 1.6289558170461182e-05,
+      "loss": 0.9907,
+      "step": 5140
+    },
+    {
+      "epoch": 0.7884969325153375,
+      "grad_norm": 0.1399131417274475,
+      "learning_rate": 1.6266851264900807e-05,
+      "loss": 0.9374,
+      "step": 5141
+    },
+    {
+      "epoch": 0.7886503067484663,
+      "grad_norm": 0.1561935842037201,
+      "learning_rate": 1.6244158271371136e-05,
+      "loss": 0.9773,
+      "step": 5142
+    },
+    {
+      "epoch": 0.7888036809815951,
+      "grad_norm": 0.16166828572750092,
+      "learning_rate": 1.6221479195247457e-05,
+      "loss": 1.0683,
+      "step": 5143
+    },
+    {
+      "epoch": 0.7889570552147239,
+      "grad_norm": 0.16014066338539124,
+      "learning_rate": 1.6198814041901696e-05,
+      "loss": 1.0301,
+      "step": 5144
+    },
+    {
+      "epoch": 0.7891104294478528,
+      "grad_norm": 0.1555147022008896,
+      "learning_rate": 1.6176162816702543e-05,
+      "loss": 1.0997,
+      "step": 5145
+    },
+    {
+      "epoch": 0.7892638036809816,
+      "grad_norm": 0.15404841303825378,
+      "learning_rate": 1.6153525525015326e-05,
+      "loss": 1.0324,
+      "step": 5146
+    },
+    {
+      "epoch": 0.7894171779141105,
+      "grad_norm": 0.1580309271812439,
+      "learning_rate": 1.613090217220214e-05,
+      "loss": 1.0243,
+      "step": 5147
+    },
+    {
+      "epoch": 0.7895705521472393,
+      "grad_norm": 0.14723524451255798,
+      "learning_rate": 1.6108292763621704e-05,
+      "loss": 1.053,
+      "step": 5148
+    },
+    {
+      "epoch": 0.7897239263803681,
+      "grad_norm": 0.15492655336856842,
+      "learning_rate": 1.608569730462952e-05,
+      "loss": 1.0455,
+      "step": 5149
+    },
+    {
+      "epoch": 0.7898773006134969,
+      "grad_norm": 0.15425118803977966,
+      "learning_rate": 1.6063115800577716e-05,
+      "loss": 1.0152,
+      "step": 5150
+    },
+    {
+      "epoch": 0.7900306748466258,
+      "grad_norm": 0.16009299457073212,
+      "learning_rate": 1.604054825681512e-05,
+      "loss": 1.109,
+      "step": 5151
+    },
+    {
+      "epoch": 0.7901840490797546,
+      "grad_norm": 0.1592734009027481,
+      "learning_rate": 1.6017994678687306e-05,
+      "loss": 1.1335,
+      "step": 5152
+    },
+    {
+      "epoch": 0.7903374233128835,
+      "grad_norm": 0.14592015743255615,
+      "learning_rate": 1.5995455071536472e-05,
+      "loss": 0.9804,
+      "step": 5153
+    },
+    {
+      "epoch": 0.7904907975460123,
+      "grad_norm": 0.1394699662923813,
+      "learning_rate": 1.597292944070157e-05,
+      "loss": 0.8963,
+      "step": 5154
+    },
+    {
+      "epoch": 0.7906441717791411,
+      "grad_norm": 0.14961464703083038,
+      "learning_rate": 1.595041779151819e-05,
+      "loss": 0.9682,
+      "step": 5155
+    },
+    {
+      "epoch": 0.7907975460122699,
+      "grad_norm": 0.1468290090560913,
+      "learning_rate": 1.5927920129318653e-05,
+      "loss": 0.9318,
+      "step": 5156
+    },
+    {
+      "epoch": 0.7909509202453988,
+      "grad_norm": 0.15555760264396667,
+      "learning_rate": 1.5905436459431935e-05,
+      "loss": 0.956,
+      "step": 5157
+    },
+    {
+      "epoch": 0.7911042944785276,
+      "grad_norm": 0.15478423237800598,
+      "learning_rate": 1.5882966787183687e-05,
+      "loss": 1.125,
+      "step": 5158
+    },
+    {
+      "epoch": 0.7912576687116565,
+      "grad_norm": 0.16678215563297272,
+      "learning_rate": 1.586051111789631e-05,
+      "loss": 1.0976,
+      "step": 5159
+    },
+    {
+      "epoch": 0.7914110429447853,
+      "grad_norm": 0.1537007987499237,
+      "learning_rate": 1.5838069456888788e-05,
+      "loss": 1.0659,
+      "step": 5160
+    },
+    {
+      "epoch": 0.7915644171779141,
+      "grad_norm": 0.15301015973091125,
+      "learning_rate": 1.5815641809476903e-05,
+      "loss": 1.0286,
+      "step": 5161
+    },
+    {
+      "epoch": 0.7917177914110429,
+      "grad_norm": 0.15242184698581696,
+      "learning_rate": 1.579322818097301e-05,
+      "loss": 1.0119,
+      "step": 5162
+    },
+    {
+      "epoch": 0.7918711656441718,
+      "grad_norm": 0.15082843601703644,
+      "learning_rate": 1.5770828576686222e-05,
+      "loss": 0.9767,
+      "step": 5163
+    },
+    {
+      "epoch": 0.7920245398773006,
+      "grad_norm": 0.15858341753482819,
+      "learning_rate": 1.5748443001922288e-05,
+      "loss": 1.2019,
+      "step": 5164
+    },
+    {
+      "epoch": 0.7921779141104295,
+      "grad_norm": 0.14033429324626923,
+      "learning_rate": 1.5726071461983627e-05,
+      "loss": 0.9046,
+      "step": 5165
+    },
+    {
+      "epoch": 0.7923312883435583,
+      "grad_norm": 0.16146448254585266,
+      "learning_rate": 1.5703713962169383e-05,
+      "loss": 1.1806,
+      "step": 5166
+    },
+    {
+      "epoch": 0.7924846625766871,
+      "grad_norm": 0.1424746960401535,
+      "learning_rate": 1.5681370507775313e-05,
+      "loss": 0.8768,
+      "step": 5167
+    },
+    {
+      "epoch": 0.7926380368098159,
+      "grad_norm": 0.14937210083007812,
+      "learning_rate": 1.5659041104093904e-05,
+      "loss": 0.8852,
+      "step": 5168
+    },
+    {
+      "epoch": 0.7927914110429448,
+      "grad_norm": 0.1481458693742752,
+      "learning_rate": 1.563672575641425e-05,
+      "loss": 0.9042,
+      "step": 5169
+    },
+    {
+      "epoch": 0.7929447852760736,
+      "grad_norm": 0.1506829559803009,
+      "learning_rate": 1.56144244700222e-05,
+      "loss": 1.103,
+      "step": 5170
+    },
+    {
+      "epoch": 0.7930981595092025,
+      "grad_norm": 0.15223276615142822,
+      "learning_rate": 1.559213725020019e-05,
+      "loss": 1.0003,
+      "step": 5171
+    },
+    {
+      "epoch": 0.7932515337423313,
+      "grad_norm": 0.15568529069423676,
+      "learning_rate": 1.556986410222739e-05,
+      "loss": 1.0185,
+      "step": 5172
+    },
+    {
+      "epoch": 0.7934049079754601,
+      "grad_norm": 0.16619466245174408,
+      "learning_rate": 1.554760503137958e-05,
+      "loss": 1.0241,
+      "step": 5173
+    },
+    {
+      "epoch": 0.7935582822085889,
+      "grad_norm": 0.1519506275653839,
+      "learning_rate": 1.5525360042929225e-05,
+      "loss": 1.0002,
+      "step": 5174
+    },
+    {
+      "epoch": 0.7937116564417178,
+      "grad_norm": 0.14970724284648895,
+      "learning_rate": 1.5503129142145495e-05,
+      "loss": 0.9772,
+      "step": 5175
+    },
+    {
+      "epoch": 0.7938650306748466,
+      "grad_norm": 0.14958788454532623,
+      "learning_rate": 1.5480912334294155e-05,
+      "loss": 1.0465,
+      "step": 5176
+    },
+    {
+      "epoch": 0.7940184049079755,
+      "grad_norm": 0.1526130884885788,
+      "learning_rate": 1.54587096246377e-05,
+      "loss": 1.0014,
+      "step": 5177
+    },
+    {
+      "epoch": 0.7941717791411043,
+      "grad_norm": 0.15542428195476532,
+      "learning_rate": 1.5436521018435216e-05,
+      "loss": 1.0408,
+      "step": 5178
+    },
+    {
+      "epoch": 0.7943251533742332,
+      "grad_norm": 0.16321760416030884,
+      "learning_rate": 1.5414346520942517e-05,
+      "loss": 1.0908,
+      "step": 5179
+    },
+    {
+      "epoch": 0.7944785276073619,
+      "grad_norm": 0.14491243660449982,
+      "learning_rate": 1.5392186137412034e-05,
+      "loss": 1.0223,
+      "step": 5180
+    },
+    {
+      "epoch": 0.7946319018404908,
+      "grad_norm": 0.16331495344638824,
+      "learning_rate": 1.537003987309283e-05,
+      "loss": 1.0717,
+      "step": 5181
+    },
+    {
+      "epoch": 0.7947852760736196,
+      "grad_norm": 0.14930886030197144,
+      "learning_rate": 1.53479077332307e-05,
+      "loss": 1.0165,
+      "step": 5182
+    },
+    {
+      "epoch": 0.7949386503067485,
+      "grad_norm": 0.15739698708057404,
+      "learning_rate": 1.5325789723068e-05,
+      "loss": 1.1382,
+      "step": 5183
+    },
+    {
+      "epoch": 0.7950920245398773,
+      "grad_norm": 0.15759453177452087,
+      "learning_rate": 1.530368584784386e-05,
+      "loss": 0.9663,
+      "step": 5184
+    },
+    {
+      "epoch": 0.7952453987730062,
+      "grad_norm": 0.1484224796295166,
+      "learning_rate": 1.528159611279395e-05,
+      "loss": 1.0553,
+      "step": 5185
+    },
+    {
+      "epoch": 0.7953987730061349,
+      "grad_norm": 0.15801848471164703,
+      "learning_rate": 1.5259520523150618e-05,
+      "loss": 1.0308,
+      "step": 5186
+    },
+    {
+      "epoch": 0.7955521472392638,
+      "grad_norm": 0.14157868921756744,
+      "learning_rate": 1.5237459084142908e-05,
+      "loss": 0.986,
+      "step": 5187
+    },
+    {
+      "epoch": 0.7957055214723926,
+      "grad_norm": 0.16210196912288666,
+      "learning_rate": 1.521541180099644e-05,
+      "loss": 1.115,
+      "step": 5188
+    },
+    {
+      "epoch": 0.7958588957055215,
+      "grad_norm": 0.15573857724666595,
+      "learning_rate": 1.5193378678933562e-05,
+      "loss": 1.077,
+      "step": 5189
+    },
+    {
+      "epoch": 0.7960122699386503,
+      "grad_norm": 0.16179420053958893,
+      "learning_rate": 1.5171359723173191e-05,
+      "loss": 1.1538,
+      "step": 5190
+    },
+    {
+      "epoch": 0.7961656441717792,
+      "grad_norm": 0.15161646902561188,
+      "learning_rate": 1.514935493893096e-05,
+      "loss": 1.0463,
+      "step": 5191
+    },
+    {
+      "epoch": 0.7963190184049079,
+      "grad_norm": 0.16006891429424286,
+      "learning_rate": 1.5127364331419086e-05,
+      "loss": 1.1034,
+      "step": 5192
+    },
+    {
+      "epoch": 0.7964723926380368,
+      "grad_norm": 0.14698579907417297,
+      "learning_rate": 1.5105387905846428e-05,
+      "loss": 0.9548,
+      "step": 5193
+    },
+    {
+      "epoch": 0.7966257668711656,
+      "grad_norm": 0.14644916355609894,
+      "learning_rate": 1.5083425667418552e-05,
+      "loss": 1.0295,
+      "step": 5194
+    },
+    {
+      "epoch": 0.7967791411042945,
+      "grad_norm": 0.16617482900619507,
+      "learning_rate": 1.5061477621337583e-05,
+      "loss": 1.0807,
+      "step": 5195
+    },
+    {
+      "epoch": 0.7969325153374233,
+      "grad_norm": 0.15939976274967194,
+      "learning_rate": 1.5039543772802354e-05,
+      "loss": 1.0527,
+      "step": 5196
+    },
+    {
+      "epoch": 0.7970858895705522,
+      "grad_norm": 0.16029827296733856,
+      "learning_rate": 1.5017624127008266e-05,
+      "loss": 1.0886,
+      "step": 5197
+    },
+    {
+      "epoch": 0.7972392638036809,
+      "grad_norm": 0.14678840339183807,
+      "learning_rate": 1.4995718689147433e-05,
+      "loss": 1.0179,
+      "step": 5198
+    },
+    {
+      "epoch": 0.7973926380368098,
+      "grad_norm": 0.16032183170318604,
+      "learning_rate": 1.4973827464408535e-05,
+      "loss": 1.0591,
+      "step": 5199
+    },
+    {
+      "epoch": 0.7975460122699386,
+      "grad_norm": 0.14996986091136932,
+      "learning_rate": 1.4951950457976896e-05,
+      "loss": 1.0072,
+      "step": 5200
+    },
+    {
+      "epoch": 0.7976993865030675,
+      "grad_norm": 0.1632947027683258,
+      "learning_rate": 1.4930087675034544e-05,
+      "loss": 1.1892,
+      "step": 5201
+    },
+    {
+      "epoch": 0.7978527607361963,
+      "grad_norm": 0.1516823172569275,
+      "learning_rate": 1.4908239120760018e-05,
+      "loss": 0.9033,
+      "step": 5202
+    },
+    {
+      "epoch": 0.7980061349693252,
+      "grad_norm": 0.15614616870880127,
+      "learning_rate": 1.4886404800328606e-05,
+      "loss": 0.9904,
+      "step": 5203
+    },
+    {
+      "epoch": 0.798159509202454,
+      "grad_norm": 0.15119053423404694,
+      "learning_rate": 1.4864584718912134e-05,
+      "loss": 0.9219,
+      "step": 5204
+    },
+    {
+      "epoch": 0.7983128834355828,
+      "grad_norm": 0.15632766485214233,
+      "learning_rate": 1.4842778881679113e-05,
+      "loss": 1.186,
+      "step": 5205
+    },
+    {
+      "epoch": 0.7984662576687117,
+      "grad_norm": 0.14037279784679413,
+      "learning_rate": 1.4820987293794642e-05,
+      "loss": 0.9582,
+      "step": 5206
+    },
+    {
+      "epoch": 0.7986196319018405,
+      "grad_norm": 0.15237537026405334,
+      "learning_rate": 1.4799209960420478e-05,
+      "loss": 1.0297,
+      "step": 5207
+    },
+    {
+      "epoch": 0.7987730061349694,
+      "grad_norm": 0.1516215056180954,
+      "learning_rate": 1.4777446886714983e-05,
+      "loss": 1.0593,
+      "step": 5208
+    },
+    {
+      "epoch": 0.7989263803680982,
+      "grad_norm": 0.17925027012825012,
+      "learning_rate": 1.4755698077833112e-05,
+      "loss": 1.2906,
+      "step": 5209
+    },
+    {
+      "epoch": 0.799079754601227,
+      "grad_norm": 0.15453961491584778,
+      "learning_rate": 1.4733963538926516e-05,
+      "loss": 1.0231,
+      "step": 5210
+    },
+    {
+      "epoch": 0.7992331288343558,
+      "grad_norm": 0.1620304435491562,
+      "learning_rate": 1.471224327514338e-05,
+      "loss": 1.0623,
+      "step": 5211
+    },
+    {
+      "epoch": 0.7993865030674847,
+      "grad_norm": 0.1584639698266983,
+      "learning_rate": 1.4690537291628584e-05,
+      "loss": 1.0662,
+      "step": 5212
+    },
+    {
+      "epoch": 0.7995398773006135,
+      "grad_norm": 0.29372310638427734,
+      "learning_rate": 1.4668845593523571e-05,
+      "loss": 1.1575,
+      "step": 5213
+    },
+    {
+      "epoch": 0.7996932515337424,
+      "grad_norm": 0.31444892287254333,
+      "learning_rate": 1.4647168185966434e-05,
+      "loss": 1.0401,
+      "step": 5214
+    },
+    {
+      "epoch": 0.7998466257668712,
+      "grad_norm": 0.15988114476203918,
+      "learning_rate": 1.4625505074091862e-05,
+      "loss": 0.9943,
+      "step": 5215
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1502654105424881,
+      "learning_rate": 1.4603856263031147e-05,
+      "loss": 1.0083,
+      "step": 5216
+    },
+    {
+      "epoch": 0.8001533742331288,
+      "grad_norm": 0.15703654289245605,
+      "learning_rate": 1.4582221757912238e-05,
+      "loss": 1.0844,
+      "step": 5217
+    },
+    {
+      "epoch": 0.8003067484662577,
+      "grad_norm": 0.15125107765197754,
+      "learning_rate": 1.4560601563859637e-05,
+      "loss": 0.9469,
+      "step": 5218
+    },
+    {
+      "epoch": 0.8004601226993865,
+      "grad_norm": 0.16515254974365234,
+      "learning_rate": 1.453899568599452e-05,
+      "loss": 1.0313,
+      "step": 5219
+    },
+    {
+      "epoch": 0.8006134969325154,
+      "grad_norm": 0.1700449436903,
+      "learning_rate": 1.4517404129434595e-05,
+      "loss": 1.1376,
+      "step": 5220
+    },
+    {
+      "epoch": 0.8007668711656442,
+      "grad_norm": 0.16214244067668915,
+      "learning_rate": 1.4495826899294272e-05,
+      "loss": 1.0489,
+      "step": 5221
+    },
+    {
+      "epoch": 0.800920245398773,
+      "grad_norm": 0.15550541877746582,
+      "learning_rate": 1.4474264000684485e-05,
+      "loss": 1.0754,
+      "step": 5222
+    },
+    {
+      "epoch": 0.8010736196319018,
+      "grad_norm": 0.13802295923233032,
+      "learning_rate": 1.4452715438712793e-05,
+      "loss": 0.87,
+      "step": 5223
+    },
+    {
+      "epoch": 0.8012269938650307,
+      "grad_norm": 0.15525995194911957,
+      "learning_rate": 1.44311812184834e-05,
+      "loss": 1.005,
+      "step": 5224
+    },
+    {
+      "epoch": 0.8013803680981595,
+      "grad_norm": 0.16275157034397125,
+      "learning_rate": 1.440966134509704e-05,
+      "loss": 1.0978,
+      "step": 5225
+    },
+    {
+      "epoch": 0.8015337423312884,
+      "grad_norm": 0.15075835585594177,
+      "learning_rate": 1.4388155823651158e-05,
+      "loss": 1.0518,
+      "step": 5226
+    },
+    {
+      "epoch": 0.8016871165644172,
+      "grad_norm": 0.1627943366765976,
+      "learning_rate": 1.43666646592397e-05,
+      "loss": 1.1538,
+      "step": 5227
+    },
+    {
+      "epoch": 0.801840490797546,
+      "grad_norm": 0.15043292939662933,
+      "learning_rate": 1.434518785695322e-05,
+      "loss": 1.0413,
+      "step": 5228
+    },
+    {
+      "epoch": 0.8019938650306748,
+      "grad_norm": 0.15858949720859528,
+      "learning_rate": 1.4323725421878949e-05,
+      "loss": 1.044,
+      "step": 5229
+    },
+    {
+      "epoch": 0.8021472392638037,
+      "grad_norm": 0.14852404594421387,
+      "learning_rate": 1.4302277359100606e-05,
+      "loss": 0.9725,
+      "step": 5230
+    },
+    {
+      "epoch": 0.8023006134969325,
+      "grad_norm": 0.13998112082481384,
+      "learning_rate": 1.4280843673698603e-05,
+      "loss": 0.9076,
+      "step": 5231
+    },
+    {
+      "epoch": 0.8024539877300614,
+      "grad_norm": 0.15140080451965332,
+      "learning_rate": 1.425942437074987e-05,
+      "loss": 1.0267,
+      "step": 5232
+    },
+    {
+      "epoch": 0.8026073619631902,
+      "grad_norm": 0.15888454020023346,
+      "learning_rate": 1.4238019455328006e-05,
+      "loss": 1.1235,
+      "step": 5233
+    },
+    {
+      "epoch": 0.802760736196319,
+      "grad_norm": 0.17761437594890594,
+      "learning_rate": 1.4216628932503132e-05,
+      "loss": 1.1641,
+      "step": 5234
+    },
+    {
+      "epoch": 0.8029141104294478,
+      "grad_norm": 0.15858528017997742,
+      "learning_rate": 1.4195252807341969e-05,
+      "loss": 1.0281,
+      "step": 5235
+    },
+    {
+      "epoch": 0.8030674846625767,
+      "grad_norm": 0.1390526294708252,
+      "learning_rate": 1.4173891084907892e-05,
+      "loss": 0.8999,
+      "step": 5236
+    },
+    {
+      "epoch": 0.8032208588957055,
+      "grad_norm": 0.15516042709350586,
+      "learning_rate": 1.415254377026078e-05,
+      "loss": 1.0135,
+      "step": 5237
+    },
+    {
+      "epoch": 0.8033742331288344,
+      "grad_norm": 0.15697352588176727,
+      "learning_rate": 1.4131210868457177e-05,
+      "loss": 1.0928,
+      "step": 5238
+    },
+    {
+      "epoch": 0.8035276073619632,
+      "grad_norm": 0.15299899876117706,
+      "learning_rate": 1.4109892384550134e-05,
+      "loss": 1.1293,
+      "step": 5239
+    },
+    {
+      "epoch": 0.803680981595092,
+      "grad_norm": 0.1550423502922058,
+      "learning_rate": 1.4088588323589362e-05,
+      "loss": 1.0545,
+      "step": 5240
+    },
+    {
+      "epoch": 0.8038343558282208,
+      "grad_norm": 0.15697802603244781,
+      "learning_rate": 1.406729869062109e-05,
+      "loss": 1.0627,
+      "step": 5241
+    },
+    {
+      "epoch": 0.8039877300613497,
+      "grad_norm": 0.15049420297145844,
+      "learning_rate": 1.4046023490688203e-05,
+      "loss": 1.0206,
+      "step": 5242
+    },
+    {
+      "epoch": 0.8041411042944785,
+      "grad_norm": 0.1428128182888031,
+      "learning_rate": 1.4024762728830092e-05,
+      "loss": 0.9008,
+      "step": 5243
+    },
+    {
+      "epoch": 0.8042944785276074,
+      "grad_norm": 0.15246379375457764,
+      "learning_rate": 1.4003516410082754e-05,
+      "loss": 0.9917,
+      "step": 5244
+    },
+    {
+      "epoch": 0.8044478527607362,
+      "grad_norm": 0.14584305882453918,
+      "learning_rate": 1.39822845394788e-05,
+      "loss": 0.967,
+      "step": 5245
+    },
+    {
+      "epoch": 0.804601226993865,
+      "grad_norm": 0.15744082629680634,
+      "learning_rate": 1.3961067122047362e-05,
+      "loss": 1.1052,
+      "step": 5246
+    },
+    {
+      "epoch": 0.8047546012269938,
+      "grad_norm": 0.16700398921966553,
+      "learning_rate": 1.3939864162814205e-05,
+      "loss": 1.0352,
+      "step": 5247
+    },
+    {
+      "epoch": 0.8049079754601227,
+      "grad_norm": 0.1612730771303177,
+      "learning_rate": 1.391867566680161e-05,
+      "loss": 1.1313,
+      "step": 5248
+    },
+    {
+      "epoch": 0.8050613496932515,
+      "grad_norm": 0.1562173068523407,
+      "learning_rate": 1.3897501639028504e-05,
+      "loss": 1.0532,
+      "step": 5249
+    },
+    {
+      "epoch": 0.8052147239263804,
+      "grad_norm": 0.14953723549842834,
+      "learning_rate": 1.3876342084510328e-05,
+      "loss": 1.0713,
+      "step": 5250
+    },
+    {
+      "epoch": 0.8053680981595092,
+      "grad_norm": 0.14709927141666412,
+      "learning_rate": 1.3855197008259087e-05,
+      "loss": 1.0741,
+      "step": 5251
+    },
+    {
+      "epoch": 0.805521472392638,
+      "grad_norm": 0.13552725315093994,
+      "learning_rate": 1.3834066415283424e-05,
+      "loss": 0.9123,
+      "step": 5252
+    },
+    {
+      "epoch": 0.8056748466257668,
+      "grad_norm": 0.15881463885307312,
+      "learning_rate": 1.3812950310588469e-05,
+      "loss": 1.0229,
+      "step": 5253
+    },
+    {
+      "epoch": 0.8058282208588957,
+      "grad_norm": 0.161329984664917,
+      "learning_rate": 1.3791848699176006e-05,
+      "loss": 1.1261,
+      "step": 5254
+    },
+    {
+      "epoch": 0.8059815950920245,
+      "grad_norm": 0.15904633700847626,
+      "learning_rate": 1.3770761586044294e-05,
+      "loss": 1.0762,
+      "step": 5255
+    },
+    {
+      "epoch": 0.8061349693251534,
+      "grad_norm": 0.16097107529640198,
+      "learning_rate": 1.3749688976188254e-05,
+      "loss": 1.1544,
+      "step": 5256
+    },
+    {
+      "epoch": 0.8062883435582822,
+      "grad_norm": 0.15701910853385925,
+      "learning_rate": 1.37286308745993e-05,
+      "loss": 0.9984,
+      "step": 5257
+    },
+    {
+      "epoch": 0.806441717791411,
+      "grad_norm": 0.16883784532546997,
+      "learning_rate": 1.3707587286265403e-05,
+      "loss": 1.1758,
+      "step": 5258
+    },
+    {
+      "epoch": 0.8065950920245398,
+      "grad_norm": 0.1513289213180542,
+      "learning_rate": 1.3686558216171176e-05,
+      "loss": 1.0057,
+      "step": 5259
+    },
+    {
+      "epoch": 0.8067484662576687,
+      "grad_norm": 0.14872494339942932,
+      "learning_rate": 1.3665543669297707e-05,
+      "loss": 0.9506,
+      "step": 5260
+    },
+    {
+      "epoch": 0.8069018404907975,
+      "grad_norm": 0.16748589277267456,
+      "learning_rate": 1.3644543650622707e-05,
+      "loss": 1.0597,
+      "step": 5261
+    },
+    {
+      "epoch": 0.8070552147239264,
+      "grad_norm": 0.14692017436027527,
+      "learning_rate": 1.3623558165120382e-05,
+      "loss": 1.0051,
+      "step": 5262
+    },
+    {
+      "epoch": 0.8072085889570552,
+      "grad_norm": 0.14741258323192596,
+      "learning_rate": 1.3602587217761571e-05,
+      "loss": 1.0086,
+      "step": 5263
+    },
+    {
+      "epoch": 0.807361963190184,
+      "grad_norm": 0.1610267162322998,
+      "learning_rate": 1.35816308135136e-05,
+      "loss": 1.1514,
+      "step": 5264
+    },
+    {
+      "epoch": 0.8075153374233128,
+      "grad_norm": 0.1501212865114212,
+      "learning_rate": 1.3560688957340407e-05,
+      "loss": 0.9353,
+      "step": 5265
+    },
+    {
+      "epoch": 0.8076687116564417,
+      "grad_norm": 0.1556597352027893,
+      "learning_rate": 1.3539761654202441e-05,
+      "loss": 1.1024,
+      "step": 5266
+    },
+    {
+      "epoch": 0.8078220858895706,
+      "grad_norm": 0.16099700331687927,
+      "learning_rate": 1.3518848909056704e-05,
+      "loss": 1.0325,
+      "step": 5267
+    },
+    {
+      "epoch": 0.8079754601226994,
+      "grad_norm": 0.16193760931491852,
+      "learning_rate": 1.349795072685679e-05,
+      "loss": 1.0643,
+      "step": 5268
+    },
+    {
+      "epoch": 0.8081288343558283,
+      "grad_norm": 0.17420490086078644,
+      "learning_rate": 1.3477067112552825e-05,
+      "loss": 1.116,
+      "step": 5269
+    },
+    {
+      "epoch": 0.808282208588957,
+      "grad_norm": 0.1485450118780136,
+      "learning_rate": 1.3456198071091451e-05,
+      "loss": 1.0096,
+      "step": 5270
+    },
+    {
+      "epoch": 0.8084355828220859,
+      "grad_norm": 0.1621159315109253,
+      "learning_rate": 1.3435343607415913e-05,
+      "loss": 1.0876,
+      "step": 5271
+    },
+    {
+      "epoch": 0.8085889570552147,
+      "grad_norm": 0.15780602395534515,
+      "learning_rate": 1.3414503726465952e-05,
+      "loss": 0.9968,
+      "step": 5272
+    },
+    {
+      "epoch": 0.8087423312883436,
+      "grad_norm": 0.16445016860961914,
+      "learning_rate": 1.3393678433177904e-05,
+      "loss": 1.0775,
+      "step": 5273
+    },
+    {
+      "epoch": 0.8088957055214724,
+      "grad_norm": 0.16098670661449432,
+      "learning_rate": 1.3372867732484596e-05,
+      "loss": 1.0117,
+      "step": 5274
+    },
+    {
+      "epoch": 0.8090490797546013,
+      "grad_norm": 0.16139888763427734,
+      "learning_rate": 1.335207162931546e-05,
+      "loss": 1.04,
+      "step": 5275
+    },
+    {
+      "epoch": 0.80920245398773,
+      "grad_norm": 0.14465440809726715,
+      "learning_rate": 1.33312901285964e-05,
+      "loss": 0.9125,
+      "step": 5276
+    },
+    {
+      "epoch": 0.8093558282208589,
+      "grad_norm": 0.14953339099884033,
+      "learning_rate": 1.3310523235249942e-05,
+      "loss": 1.098,
+      "step": 5277
+    },
+    {
+      "epoch": 0.8095092024539877,
+      "grad_norm": 0.15358446538448334,
+      "learning_rate": 1.3289770954195077e-05,
+      "loss": 0.9693,
+      "step": 5278
+    },
+    {
+      "epoch": 0.8096625766871166,
+      "grad_norm": 0.1623840481042862,
+      "learning_rate": 1.326903329034735e-05,
+      "loss": 1.1366,
+      "step": 5279
+    },
+    {
+      "epoch": 0.8098159509202454,
+      "grad_norm": 0.14874358475208282,
+      "learning_rate": 1.3248310248618904e-05,
+      "loss": 1.0256,
+      "step": 5280
+    },
+    {
+      "epoch": 0.8099693251533743,
+      "grad_norm": 0.15049009025096893,
+      "learning_rate": 1.322760183391833e-05,
+      "loss": 1.0198,
+      "step": 5281
+    },
+    {
+      "epoch": 0.810122699386503,
+      "grad_norm": 0.14770236611366272,
+      "learning_rate": 1.3206908051150838e-05,
+      "loss": 1.1387,
+      "step": 5282
+    },
+    {
+      "epoch": 0.8102760736196319,
+      "grad_norm": 0.1521225869655609,
+      "learning_rate": 1.31862289052181e-05,
+      "loss": 0.962,
+      "step": 5283
+    },
+    {
+      "epoch": 0.8104294478527607,
+      "grad_norm": 0.1571078598499298,
+      "learning_rate": 1.3165564401018385e-05,
+      "loss": 1.0219,
+      "step": 5284
+    },
+    {
+      "epoch": 0.8105828220858896,
+      "grad_norm": 0.15944351255893707,
+      "learning_rate": 1.3144914543446449e-05,
+      "loss": 1.0361,
+      "step": 5285
+    },
+    {
+      "epoch": 0.8107361963190184,
+      "grad_norm": 0.1595488339662552,
+      "learning_rate": 1.3124279337393562e-05,
+      "loss": 1.1827,
+      "step": 5286
+    },
+    {
+      "epoch": 0.8108895705521473,
+      "grad_norm": 0.15332965552806854,
+      "learning_rate": 1.3103658787747603e-05,
+      "loss": 1.0915,
+      "step": 5287
+    },
+    {
+      "epoch": 0.811042944785276,
+      "grad_norm": 0.14701484143733978,
+      "learning_rate": 1.3083052899392884e-05,
+      "loss": 0.9746,
+      "step": 5288
+    },
+    {
+      "epoch": 0.8111963190184049,
+      "grad_norm": 0.16376611590385437,
+      "learning_rate": 1.3062461677210342e-05,
+      "loss": 1.1076,
+      "step": 5289
+    },
+    {
+      "epoch": 0.8113496932515337,
+      "grad_norm": 0.1663869023323059,
+      "learning_rate": 1.3041885126077334e-05,
+      "loss": 1.084,
+      "step": 5290
+    },
+    {
+      "epoch": 0.8115030674846626,
+      "grad_norm": 0.16493773460388184,
+      "learning_rate": 1.3021323250867835e-05,
+      "loss": 1.1284,
+      "step": 5291
+    },
+    {
+      "epoch": 0.8116564417177914,
+      "grad_norm": 0.14604440331459045,
+      "learning_rate": 1.30007760564523e-05,
+      "loss": 0.9793,
+      "step": 5292
+    },
+    {
+      "epoch": 0.8118098159509203,
+      "grad_norm": 0.1590249389410019,
+      "learning_rate": 1.2980243547697678e-05,
+      "loss": 1.1056,
+      "step": 5293
+    },
+    {
+      "epoch": 0.811963190184049,
+      "grad_norm": 0.1410522758960724,
+      "learning_rate": 1.2959725729467518e-05,
+      "loss": 0.9735,
+      "step": 5294
+    },
+    {
+      "epoch": 0.8121165644171779,
+      "grad_norm": 0.15747523307800293,
+      "learning_rate": 1.2939222606621802e-05,
+      "loss": 1.0132,
+      "step": 5295
+    },
+    {
+      "epoch": 0.8122699386503067,
+      "grad_norm": 0.16124756634235382,
+      "learning_rate": 1.2918734184017119e-05,
+      "loss": 1.0272,
+      "step": 5296
+    },
+    {
+      "epoch": 0.8124233128834356,
+      "grad_norm": 0.1455930769443512,
+      "learning_rate": 1.2898260466506478e-05,
+      "loss": 0.995,
+      "step": 5297
+    },
+    {
+      "epoch": 0.8125766871165644,
+      "grad_norm": 0.15270616114139557,
+      "learning_rate": 1.2877801458939507e-05,
+      "loss": 1.0467,
+      "step": 5298
+    },
+    {
+      "epoch": 0.8127300613496933,
+      "grad_norm": 0.159566268324852,
+      "learning_rate": 1.2857357166162266e-05,
+      "loss": 1.0038,
+      "step": 5299
+    },
+    {
+      "epoch": 0.8128834355828221,
+      "grad_norm": 0.15326109528541565,
+      "learning_rate": 1.2836927593017388e-05,
+      "loss": 1.0334,
+      "step": 5300
+    },
+    {
+      "epoch": 0.813036809815951,
+      "grad_norm": 0.160331130027771,
+      "learning_rate": 1.2816512744343976e-05,
+      "loss": 1.0437,
+      "step": 5301
+    },
+    {
+      "epoch": 0.8131901840490797,
+      "grad_norm": 0.15425312519073486,
+      "learning_rate": 1.2796112624977655e-05,
+      "loss": 1.0214,
+      "step": 5302
+    },
+    {
+      "epoch": 0.8133435582822086,
+      "grad_norm": 0.14734752476215363,
+      "learning_rate": 1.2775727239750601e-05,
+      "loss": 1.0059,
+      "step": 5303
+    },
+    {
+      "epoch": 0.8134969325153374,
+      "grad_norm": 0.1514655202627182,
+      "learning_rate": 1.2755356593491437e-05,
+      "loss": 0.9573,
+      "step": 5304
+    },
+    {
+      "epoch": 0.8136503067484663,
+      "grad_norm": 0.15275131165981293,
+      "learning_rate": 1.2735000691025358e-05,
+      "loss": 1.0741,
+      "step": 5305
+    },
+    {
+      "epoch": 0.8138036809815951,
+      "grad_norm": 0.14414991438388824,
+      "learning_rate": 1.2714659537173999e-05,
+      "loss": 1.0218,
+      "step": 5306
+    },
+    {
+      "epoch": 0.813957055214724,
+      "grad_norm": 0.18240796029567719,
+      "learning_rate": 1.2694333136755575e-05,
+      "loss": 1.1173,
+      "step": 5307
+    },
+    {
+      "epoch": 0.8141104294478527,
+      "grad_norm": 0.1595798134803772,
+      "learning_rate": 1.2674021494584757e-05,
+      "loss": 1.0182,
+      "step": 5308
+    },
+    {
+      "epoch": 0.8142638036809816,
+      "grad_norm": 0.156894251704216,
+      "learning_rate": 1.2653724615472708e-05,
+      "loss": 1.0704,
+      "step": 5309
+    },
+    {
+      "epoch": 0.8144171779141104,
+      "grad_norm": 0.1560276597738266,
+      "learning_rate": 1.2633442504227136e-05,
+      "loss": 1.1028,
+      "step": 5310
+    },
+    {
+      "epoch": 0.8145705521472393,
+      "grad_norm": 0.15613117814064026,
+      "learning_rate": 1.2613175165652236e-05,
+      "loss": 1.1903,
+      "step": 5311
+    },
+    {
+      "epoch": 0.8147239263803681,
+      "grad_norm": 0.15555442869663239,
+      "learning_rate": 1.2592922604548717e-05,
+      "loss": 1.2037,
+      "step": 5312
+    },
+    {
+      "epoch": 0.814877300613497,
+      "grad_norm": 0.15760821104049683,
+      "learning_rate": 1.2572684825713764e-05,
+      "loss": 1.0464,
+      "step": 5313
+    },
+    {
+      "epoch": 0.8150306748466257,
+      "grad_norm": 0.1494477093219757,
+      "learning_rate": 1.2552461833941035e-05,
+      "loss": 1.0493,
+      "step": 5314
+    },
+    {
+      "epoch": 0.8151840490797546,
+      "grad_norm": 0.16877663135528564,
+      "learning_rate": 1.2532253634020766e-05,
+      "loss": 1.1169,
+      "step": 5315
+    },
+    {
+      "epoch": 0.8153374233128834,
+      "grad_norm": 0.158935084939003,
+      "learning_rate": 1.2512060230739595e-05,
+      "loss": 1.0782,
+      "step": 5316
+    },
+    {
+      "epoch": 0.8154907975460123,
+      "grad_norm": 0.17175713181495667,
+      "learning_rate": 1.2491881628880755e-05,
+      "loss": 1.155,
+      "step": 5317
+    },
+    {
+      "epoch": 0.8156441717791411,
+      "grad_norm": 0.16893261671066284,
+      "learning_rate": 1.2471717833223877e-05,
+      "loss": 1.1791,
+      "step": 5318
+    },
+    {
+      "epoch": 0.81579754601227,
+      "grad_norm": 0.15607298910617828,
+      "learning_rate": 1.245156884854516e-05,
+      "loss": 0.9845,
+      "step": 5319
+    },
+    {
+      "epoch": 0.8159509202453987,
+      "grad_norm": 0.15970678627490997,
+      "learning_rate": 1.2431434679617245e-05,
+      "loss": 1.1317,
+      "step": 5320
+    },
+    {
+      "epoch": 0.8161042944785276,
+      "grad_norm": 0.1577739417552948,
+      "learning_rate": 1.2411315331209273e-05,
+      "loss": 1.1079,
+      "step": 5321
+    },
+    {
+      "epoch": 0.8162576687116564,
+      "grad_norm": 0.1593695729970932,
+      "learning_rate": 1.239121080808691e-05,
+      "loss": 1.0973,
+      "step": 5322
+    },
+    {
+      "epoch": 0.8164110429447853,
+      "grad_norm": 0.1524224877357483,
+      "learning_rate": 1.2371121115012244e-05,
+      "loss": 0.9465,
+      "step": 5323
+    },
+    {
+      "epoch": 0.8165644171779141,
+      "grad_norm": 0.14033262431621552,
+      "learning_rate": 1.2351046256743938e-05,
+      "loss": 0.9074,
+      "step": 5324
+    },
+    {
+      "epoch": 0.816717791411043,
+      "grad_norm": 0.15240827202796936,
+      "learning_rate": 1.2330986238037052e-05,
+      "loss": 0.9853,
+      "step": 5325
+    },
+    {
+      "epoch": 0.8168711656441717,
+      "grad_norm": 0.1561513990163803,
+      "learning_rate": 1.2310941063643204e-05,
+      "loss": 1.1086,
+      "step": 5326
+    },
+    {
+      "epoch": 0.8170245398773006,
+      "grad_norm": 0.16267681121826172,
+      "learning_rate": 1.2290910738310453e-05,
+      "loss": 1.0839,
+      "step": 5327
+    },
+    {
+      "epoch": 0.8171779141104294,
+      "grad_norm": 0.15258094668388367,
+      "learning_rate": 1.227089526678333e-05,
+      "loss": 0.9958,
+      "step": 5328
+    },
+    {
+      "epoch": 0.8173312883435583,
+      "grad_norm": 0.14883004128932953,
+      "learning_rate": 1.2250894653802897e-05,
+      "loss": 0.9418,
+      "step": 5329
+    },
+    {
+      "epoch": 0.8174846625766872,
+      "grad_norm": 0.16255828738212585,
+      "learning_rate": 1.2230908904106637e-05,
+      "loss": 1.172,
+      "step": 5330
+    },
+    {
+      "epoch": 0.817638036809816,
+      "grad_norm": 0.14039400219917297,
+      "learning_rate": 1.2210938022428593e-05,
+      "loss": 0.9733,
+      "step": 5331
+    },
+    {
+      "epoch": 0.8177914110429448,
+      "grad_norm": 0.15146596729755402,
+      "learning_rate": 1.219098201349918e-05,
+      "loss": 1.0581,
+      "step": 5332
+    },
+    {
+      "epoch": 0.8179447852760736,
+      "grad_norm": 0.1407800018787384,
+      "learning_rate": 1.2171040882045397e-05,
+      "loss": 0.8788,
+      "step": 5333
+    },
+    {
+      "epoch": 0.8180981595092025,
+      "grad_norm": 0.16536623239517212,
+      "learning_rate": 1.2151114632790635e-05,
+      "loss": 1.1233,
+      "step": 5334
+    },
+    {
+      "epoch": 0.8182515337423313,
+      "grad_norm": 0.15313588082790375,
+      "learning_rate": 1.213120327045482e-05,
+      "loss": 1.0381,
+      "step": 5335
+    },
+    {
+      "epoch": 0.8184049079754602,
+      "grad_norm": 0.15495413541793823,
+      "learning_rate": 1.2111306799754314e-05,
+      "loss": 0.9504,
+      "step": 5336
+    },
+    {
+      "epoch": 0.818558282208589,
+      "grad_norm": 0.16080059111118317,
+      "learning_rate": 1.2091425225401938e-05,
+      "loss": 1.1383,
+      "step": 5337
+    },
+    {
+      "epoch": 0.8187116564417178,
+      "grad_norm": 0.1662069708108902,
+      "learning_rate": 1.2071558552107058e-05,
+      "loss": 1.1434,
+      "step": 5338
+    },
+    {
+      "epoch": 0.8188650306748466,
+      "grad_norm": 0.15645921230316162,
+      "learning_rate": 1.2051706784575418e-05,
+      "loss": 1.0478,
+      "step": 5339
+    },
+    {
+      "epoch": 0.8190184049079755,
+      "grad_norm": 0.14407534897327423,
+      "learning_rate": 1.2031869927509316e-05,
+      "loss": 0.9933,
+      "step": 5340
+    },
+    {
+      "epoch": 0.8191717791411043,
+      "grad_norm": 0.15653568506240845,
+      "learning_rate": 1.2012047985607434e-05,
+      "loss": 1.0269,
+      "step": 5341
+    },
+    {
+      "epoch": 0.8193251533742332,
+      "grad_norm": 0.1629955768585205,
+      "learning_rate": 1.1992240963565009e-05,
+      "loss": 1.062,
+      "step": 5342
+    },
+    {
+      "epoch": 0.819478527607362,
+      "grad_norm": 0.15928030014038086,
+      "learning_rate": 1.1972448866073686e-05,
+      "loss": 0.9273,
+      "step": 5343
+    },
+    {
+      "epoch": 0.8196319018404908,
+      "grad_norm": 0.15566560626029968,
+      "learning_rate": 1.1952671697821558e-05,
+      "loss": 1.0325,
+      "step": 5344
+    },
+    {
+      "epoch": 0.8197852760736196,
+      "grad_norm": 0.15286238491535187,
+      "learning_rate": 1.1932909463493254e-05,
+      "loss": 1.0727,
+      "step": 5345
+    },
+    {
+      "epoch": 0.8199386503067485,
+      "grad_norm": 0.1596602201461792,
+      "learning_rate": 1.1913162167769794e-05,
+      "loss": 1.1049,
+      "step": 5346
+    },
+    {
+      "epoch": 0.8200920245398773,
+      "grad_norm": 0.17447683215141296,
+      "learning_rate": 1.1893429815328712e-05,
+      "loss": 1.1787,
+      "step": 5347
+    },
+    {
+      "epoch": 0.8202453987730062,
+      "grad_norm": 0.14300666749477386,
+      "learning_rate": 1.187371241084396e-05,
+      "loss": 0.9534,
+      "step": 5348
+    },
+    {
+      "epoch": 0.820398773006135,
+      "grad_norm": 0.14928869903087616,
+      "learning_rate": 1.1854009958985997e-05,
+      "loss": 0.9111,
+      "step": 5349
+    },
+    {
+      "epoch": 0.8205521472392638,
+      "grad_norm": 0.1570281833410263,
+      "learning_rate": 1.1834322464421695e-05,
+      "loss": 1.0005,
+      "step": 5350
+    },
+    {
+      "epoch": 0.8207055214723926,
+      "grad_norm": 0.15513072907924652,
+      "learning_rate": 1.1814649931814382e-05,
+      "loss": 1.0368,
+      "step": 5351
+    },
+    {
+      "epoch": 0.8208588957055215,
+      "grad_norm": 0.14527440071105957,
+      "learning_rate": 1.1794992365823899e-05,
+      "loss": 0.9238,
+      "step": 5352
+    },
+    {
+      "epoch": 0.8210122699386503,
+      "grad_norm": 0.15251928567886353,
+      "learning_rate": 1.1775349771106447e-05,
+      "loss": 1.0097,
+      "step": 5353
+    },
+    {
+      "epoch": 0.8211656441717792,
+      "grad_norm": 0.16155360639095306,
+      "learning_rate": 1.1755722152314808e-05,
+      "loss": 1.0715,
+      "step": 5354
+    },
+    {
+      "epoch": 0.821319018404908,
+      "grad_norm": 0.1554698795080185,
+      "learning_rate": 1.173610951409811e-05,
+      "loss": 0.957,
+      "step": 5355
+    },
+    {
+      "epoch": 0.8214723926380368,
+      "grad_norm": 0.1610928326845169,
+      "learning_rate": 1.171651186110196e-05,
+      "loss": 1.1889,
+      "step": 5356
+    },
+    {
+      "epoch": 0.8216257668711656,
+      "grad_norm": 0.1535339206457138,
+      "learning_rate": 1.1696929197968451e-05,
+      "loss": 1.1017,
+      "step": 5357
+    },
+    {
+      "epoch": 0.8217791411042945,
+      "grad_norm": 0.14928407967090607,
+      "learning_rate": 1.167736152933606e-05,
+      "loss": 0.931,
+      "step": 5358
+    },
+    {
+      "epoch": 0.8219325153374233,
+      "grad_norm": 0.14681820571422577,
+      "learning_rate": 1.1657808859839793e-05,
+      "loss": 0.9467,
+      "step": 5359
+    },
+    {
+      "epoch": 0.8220858895705522,
+      "grad_norm": 0.15944994986057281,
+      "learning_rate": 1.1638271194111021e-05,
+      "loss": 1.0426,
+      "step": 5360
+    },
+    {
+      "epoch": 0.822239263803681,
+      "grad_norm": 0.15610475838184357,
+      "learning_rate": 1.1618748536777642e-05,
+      "loss": 1.0359,
+      "step": 5361
+    },
+    {
+      "epoch": 0.8223926380368098,
+      "grad_norm": 0.15307720005512238,
+      "learning_rate": 1.1599240892463932e-05,
+      "loss": 0.9854,
+      "step": 5362
+    },
+    {
+      "epoch": 0.8225460122699386,
+      "grad_norm": 0.16392160952091217,
+      "learning_rate": 1.157974826579062e-05,
+      "loss": 0.9909,
+      "step": 5363
+    },
+    {
+      "epoch": 0.8226993865030675,
+      "grad_norm": 0.15698425471782684,
+      "learning_rate": 1.1560270661374938e-05,
+      "loss": 1.0639,
+      "step": 5364
+    },
+    {
+      "epoch": 0.8228527607361963,
+      "grad_norm": 0.1556168645620346,
+      "learning_rate": 1.1540808083830478e-05,
+      "loss": 1.1097,
+      "step": 5365
+    },
+    {
+      "epoch": 0.8230061349693252,
+      "grad_norm": 0.1478423923254013,
+      "learning_rate": 1.1521360537767335e-05,
+      "loss": 0.9448,
+      "step": 5366
+    },
+    {
+      "epoch": 0.823159509202454,
+      "grad_norm": 0.15289640426635742,
+      "learning_rate": 1.1501928027792e-05,
+      "loss": 1.0302,
+      "step": 5367
+    },
+    {
+      "epoch": 0.8233128834355828,
+      "grad_norm": 0.16254179179668427,
+      "learning_rate": 1.1482510558507452e-05,
+      "loss": 1.0762,
+      "step": 5368
+    },
+    {
+      "epoch": 0.8234662576687116,
+      "grad_norm": 0.15861298143863678,
+      "learning_rate": 1.1463108134513032e-05,
+      "loss": 1.0909,
+      "step": 5369
+    },
+    {
+      "epoch": 0.8236196319018405,
+      "grad_norm": 0.16268287599086761,
+      "learning_rate": 1.1443720760404609e-05,
+      "loss": 1.129,
+      "step": 5370
+    },
+    {
+      "epoch": 0.8237730061349693,
+      "grad_norm": 0.1502092480659485,
+      "learning_rate": 1.1424348440774426e-05,
+      "loss": 1.0284,
+      "step": 5371
+    },
+    {
+      "epoch": 0.8239263803680982,
+      "grad_norm": 0.1534148007631302,
+      "learning_rate": 1.1404991180211142e-05,
+      "loss": 1.0682,
+      "step": 5372
+    },
+    {
+      "epoch": 0.824079754601227,
+      "grad_norm": 0.14727599918842316,
+      "learning_rate": 1.1385648983299928e-05,
+      "loss": 0.8992,
+      "step": 5373
+    },
+    {
+      "epoch": 0.8242331288343558,
+      "grad_norm": 0.1499035358428955,
+      "learning_rate": 1.1366321854622307e-05,
+      "loss": 0.997,
+      "step": 5374
+    },
+    {
+      "epoch": 0.8243865030674846,
+      "grad_norm": 0.1579156219959259,
+      "learning_rate": 1.1347009798756294e-05,
+      "loss": 1.1139,
+      "step": 5375
+    },
+    {
+      "epoch": 0.8245398773006135,
+      "grad_norm": 0.15651346743106842,
+      "learning_rate": 1.1327712820276269e-05,
+      "loss": 1.0446,
+      "step": 5376
+    },
+    {
+      "epoch": 0.8246932515337423,
+      "grad_norm": 0.1591351181268692,
+      "learning_rate": 1.1308430923753128e-05,
+      "loss": 1.1218,
+      "step": 5377
+    },
+    {
+      "epoch": 0.8248466257668712,
+      "grad_norm": 0.15971088409423828,
+      "learning_rate": 1.128916411375411e-05,
+      "loss": 1.0571,
+      "step": 5378
+    },
+    {
+      "epoch": 0.825,
+      "grad_norm": 0.1466526836156845,
+      "learning_rate": 1.1269912394842904e-05,
+      "loss": 1.0306,
+      "step": 5379
+    },
+    {
+      "epoch": 0.8251533742331288,
+      "grad_norm": 0.16279444098472595,
+      "learning_rate": 1.1250675771579663e-05,
+      "loss": 1.1755,
+      "step": 5380
+    },
+    {
+      "epoch": 0.8253067484662576,
+      "grad_norm": 0.14202232658863068,
+      "learning_rate": 1.1231454248520915e-05,
+      "loss": 0.9239,
+      "step": 5381
+    },
+    {
+      "epoch": 0.8254601226993865,
+      "grad_norm": 0.15178915858268738,
+      "learning_rate": 1.1212247830219665e-05,
+      "loss": 1.1071,
+      "step": 5382
+    },
+    {
+      "epoch": 0.8256134969325153,
+      "grad_norm": 0.16872212290763855,
+      "learning_rate": 1.119305652122526e-05,
+      "loss": 1.1143,
+      "step": 5383
+    },
+    {
+      "epoch": 0.8257668711656442,
+      "grad_norm": 0.15460734069347382,
+      "learning_rate": 1.117388032608356e-05,
+      "loss": 0.9947,
+      "step": 5384
+    },
+    {
+      "epoch": 0.825920245398773,
+      "grad_norm": 0.14666278660297394,
+      "learning_rate": 1.1154719249336786e-05,
+      "loss": 0.9532,
+      "step": 5385
+    },
+    {
+      "epoch": 0.8260736196319018,
+      "grad_norm": 0.16167226433753967,
+      "learning_rate": 1.1135573295523575e-05,
+      "loss": 1.1997,
+      "step": 5386
+    },
+    {
+      "epoch": 0.8262269938650306,
+      "grad_norm": 0.15145562589168549,
+      "learning_rate": 1.1116442469179026e-05,
+      "loss": 1.0318,
+      "step": 5387
+    },
+    {
+      "epoch": 0.8263803680981595,
+      "grad_norm": 0.14120714366436005,
+      "learning_rate": 1.1097326774834597e-05,
+      "loss": 0.8702,
+      "step": 5388
+    },
+    {
+      "epoch": 0.8265337423312883,
+      "grad_norm": 0.1572810262441635,
+      "learning_rate": 1.1078226217018241e-05,
+      "loss": 1.1223,
+      "step": 5389
+    },
+    {
+      "epoch": 0.8266871165644172,
+      "grad_norm": 0.1676148623228073,
+      "learning_rate": 1.1059140800254221e-05,
+      "loss": 1.0928,
+      "step": 5390
+    },
+    {
+      "epoch": 0.826840490797546,
+      "grad_norm": 0.14842696487903595,
+      "learning_rate": 1.1040070529063322e-05,
+      "loss": 0.9069,
+      "step": 5391
+    },
+    {
+      "epoch": 0.8269938650306748,
+      "grad_norm": 0.15916170179843903,
+      "learning_rate": 1.1021015407962648e-05,
+      "loss": 1.1787,
+      "step": 5392
+    },
+    {
+      "epoch": 0.8271472392638037,
+      "grad_norm": 0.1565488874912262,
+      "learning_rate": 1.1001975441465794e-05,
+      "loss": 1.0442,
+      "step": 5393
+    },
+    {
+      "epoch": 0.8273006134969325,
+      "grad_norm": 0.14617577195167542,
+      "learning_rate": 1.0982950634082704e-05,
+      "loss": 0.9133,
+      "step": 5394
+    },
+    {
+      "epoch": 0.8274539877300614,
+      "grad_norm": 0.15405066311359406,
+      "learning_rate": 1.0963940990319734e-05,
+      "loss": 1.0256,
+      "step": 5395
+    },
+    {
+      "epoch": 0.8276073619631902,
+      "grad_norm": 0.1501809060573578,
+      "learning_rate": 1.094494651467972e-05,
+      "loss": 1.0371,
+      "step": 5396
+    },
+    {
+      "epoch": 0.8277607361963191,
+      "grad_norm": 0.1448415219783783,
+      "learning_rate": 1.0925967211661832e-05,
+      "loss": 0.8907,
+      "step": 5397
+    },
+    {
+      "epoch": 0.8279141104294478,
+      "grad_norm": 0.149242103099823,
+      "learning_rate": 1.0907003085761648e-05,
+      "loss": 1.0231,
+      "step": 5398
+    },
+    {
+      "epoch": 0.8280674846625767,
+      "grad_norm": 0.14549896121025085,
+      "learning_rate": 1.0888054141471208e-05,
+      "loss": 0.9985,
+      "step": 5399
+    },
+    {
+      "epoch": 0.8282208588957055,
+      "grad_norm": 0.15291832387447357,
+      "learning_rate": 1.0869120383278882e-05,
+      "loss": 1.0563,
+      "step": 5400
+    },
+    {
+      "epoch": 0.8283742331288344,
+      "grad_norm": 0.1565527617931366,
+      "learning_rate": 1.0850201815669512e-05,
+      "loss": 1.0296,
+      "step": 5401
+    },
+    {
+      "epoch": 0.8285276073619632,
+      "grad_norm": 0.15605196356773376,
+      "learning_rate": 1.0831298443124284e-05,
+      "loss": 1.0531,
+      "step": 5402
+    },
+    {
+      "epoch": 0.8286809815950921,
+      "grad_norm": 0.14492963254451752,
+      "learning_rate": 1.081241027012083e-05,
+      "loss": 0.8806,
+      "step": 5403
+    },
+    {
+      "epoch": 0.8288343558282208,
+      "grad_norm": 0.1563064455986023,
+      "learning_rate": 1.0793537301133149e-05,
+      "loss": 0.9778,
+      "step": 5404
+    },
+    {
+      "epoch": 0.8289877300613497,
+      "grad_norm": 0.14753411710262299,
+      "learning_rate": 1.0774679540631678e-05,
+      "loss": 0.9439,
+      "step": 5405
+    },
+    {
+      "epoch": 0.8291411042944785,
+      "grad_norm": 0.1735677272081375,
+      "learning_rate": 1.0755836993083198e-05,
+      "loss": 1.2204,
+      "step": 5406
+    },
+    {
+      "epoch": 0.8292944785276074,
+      "grad_norm": 0.16039517521858215,
+      "learning_rate": 1.073700966295091e-05,
+      "loss": 1.0421,
+      "step": 5407
+    },
+    {
+      "epoch": 0.8294478527607362,
+      "grad_norm": 0.151750847697258,
+      "learning_rate": 1.071819755469444e-05,
+      "loss": 1.0374,
+      "step": 5408
+    },
+    {
+      "epoch": 0.8296012269938651,
+      "grad_norm": 0.1563665121793747,
+      "learning_rate": 1.0699400672769758e-05,
+      "loss": 1.0709,
+      "step": 5409
+    },
+    {
+      "epoch": 0.8297546012269938,
+      "grad_norm": 0.17858736217021942,
+      "learning_rate": 1.0680619021629285e-05,
+      "loss": 1.1362,
+      "step": 5410
+    },
+    {
+      "epoch": 0.8299079754601227,
+      "grad_norm": 0.15624620020389557,
+      "learning_rate": 1.0661852605721762e-05,
+      "loss": 1.0454,
+      "step": 5411
+    },
+    {
+      "epoch": 0.8300613496932515,
+      "grad_norm": 0.15973436832427979,
+      "learning_rate": 1.06431014294924e-05,
+      "loss": 1.2041,
+      "step": 5412
+    },
+    {
+      "epoch": 0.8302147239263804,
+      "grad_norm": 0.1578376144170761,
+      "learning_rate": 1.062436549738274e-05,
+      "loss": 1.0536,
+      "step": 5413
+    },
+    {
+      "epoch": 0.8303680981595092,
+      "grad_norm": 0.15964920818805695,
+      "learning_rate": 1.0605644813830719e-05,
+      "loss": 1.1883,
+      "step": 5414
+    },
+    {
+      "epoch": 0.8305214723926381,
+      "grad_norm": 0.1575678139925003,
+      "learning_rate": 1.058693938327071e-05,
+      "loss": 1.0525,
+      "step": 5415
+    },
+    {
+      "epoch": 0.8306748466257668,
+      "grad_norm": 0.15767648816108704,
+      "learning_rate": 1.0568249210133408e-05,
+      "loss": 1.0589,
+      "step": 5416
+    },
+    {
+      "epoch": 0.8308282208588957,
+      "grad_norm": 0.15135683119297028,
+      "learning_rate": 1.0549574298845958e-05,
+      "loss": 1.0056,
+      "step": 5417
+    },
+    {
+      "epoch": 0.8309815950920245,
+      "grad_norm": 0.15195679664611816,
+      "learning_rate": 1.0530914653831821e-05,
+      "loss": 1.0734,
+      "step": 5418
+    },
+    {
+      "epoch": 0.8311349693251534,
+      "grad_norm": 0.15424586832523346,
+      "learning_rate": 1.0512270279510917e-05,
+      "loss": 1.0901,
+      "step": 5419
+    },
+    {
+      "epoch": 0.8312883435582822,
+      "grad_norm": 0.16089563071727753,
+      "learning_rate": 1.0493641180299492e-05,
+      "loss": 1.128,
+      "step": 5420
+    },
+    {
+      "epoch": 0.8314417177914111,
+      "grad_norm": 0.1455945074558258,
+      "learning_rate": 1.0475027360610172e-05,
+      "loss": 0.9857,
+      "step": 5421
+    },
+    {
+      "epoch": 0.8315950920245399,
+      "grad_norm": 0.16361774504184723,
+      "learning_rate": 1.0456428824852023e-05,
+      "loss": 1.237,
+      "step": 5422
+    },
+    {
+      "epoch": 0.8317484662576687,
+      "grad_norm": 0.13592419028282166,
+      "learning_rate": 1.0437845577430412e-05,
+      "loss": 0.8144,
+      "step": 5423
+    },
+    {
+      "epoch": 0.8319018404907975,
+      "grad_norm": 0.1656583845615387,
+      "learning_rate": 1.0419277622747168e-05,
+      "loss": 1.1126,
+      "step": 5424
+    },
+    {
+      "epoch": 0.8320552147239264,
+      "grad_norm": 0.1475682407617569,
+      "learning_rate": 1.0400724965200407e-05,
+      "loss": 1.0777,
+      "step": 5425
+    },
+    {
+      "epoch": 0.8322085889570552,
+      "grad_norm": 0.1565970778465271,
+      "learning_rate": 1.0382187609184706e-05,
+      "loss": 0.967,
+      "step": 5426
+    },
+    {
+      "epoch": 0.8323619631901841,
+      "grad_norm": 0.1557929962873459,
+      "learning_rate": 1.036366555909095e-05,
+      "loss": 1.0501,
+      "step": 5427
+    },
+    {
+      "epoch": 0.8325153374233129,
+      "grad_norm": 0.1611061841249466,
+      "learning_rate": 1.0345158819306464e-05,
+      "loss": 1.1004,
+      "step": 5428
+    },
+    {
+      "epoch": 0.8326687116564417,
+      "grad_norm": 0.16039223968982697,
+      "learning_rate": 1.0326667394214885e-05,
+      "loss": 1.1663,
+      "step": 5429
+    },
+    {
+      "epoch": 0.8328220858895705,
+      "grad_norm": 0.15693561732769012,
+      "learning_rate": 1.0308191288196241e-05,
+      "loss": 0.9653,
+      "step": 5430
+    },
+    {
+      "epoch": 0.8329754601226994,
+      "grad_norm": 0.16268521547317505,
+      "learning_rate": 1.0289730505626973e-05,
+      "loss": 1.0749,
+      "step": 5431
+    },
+    {
+      "epoch": 0.8331288343558282,
+      "grad_norm": 0.16158628463745117,
+      "learning_rate": 1.0271285050879816e-05,
+      "loss": 1.0539,
+      "step": 5432
+    },
+    {
+      "epoch": 0.8332822085889571,
+      "grad_norm": 0.14369353652000427,
+      "learning_rate": 1.025285492832395e-05,
+      "loss": 1.0592,
+      "step": 5433
+    },
+    {
+      "epoch": 0.8334355828220859,
+      "grad_norm": 0.15930315852165222,
+      "learning_rate": 1.0234440142324868e-05,
+      "loss": 1.0764,
+      "step": 5434
+    },
+    {
+      "epoch": 0.8335889570552147,
+      "grad_norm": 0.13980022072792053,
+      "learning_rate": 1.0216040697244479e-05,
+      "loss": 0.9094,
+      "step": 5435
+    },
+    {
+      "epoch": 0.8337423312883435,
+      "grad_norm": 0.13985928893089294,
+      "learning_rate": 1.0197656597441015e-05,
+      "loss": 0.9207,
+      "step": 5436
+    },
+    {
+      "epoch": 0.8338957055214724,
+      "grad_norm": 0.15020383894443512,
+      "learning_rate": 1.0179287847269067e-05,
+      "loss": 1.039,
+      "step": 5437
+    },
+    {
+      "epoch": 0.8340490797546012,
+      "grad_norm": 0.1661488562822342,
+      "learning_rate": 1.0160934451079636e-05,
+      "loss": 1.1147,
+      "step": 5438
+    },
+    {
+      "epoch": 0.8342024539877301,
+      "grad_norm": 0.15292736887931824,
+      "learning_rate": 1.0142596413220064e-05,
+      "loss": 1.0031,
+      "step": 5439
+    },
+    {
+      "epoch": 0.8343558282208589,
+      "grad_norm": 0.16275793313980103,
+      "learning_rate": 1.0124273738034066e-05,
+      "loss": 1.1238,
+      "step": 5440
+    },
+    {
+      "epoch": 0.8345092024539877,
+      "grad_norm": 0.15573781728744507,
+      "learning_rate": 1.0105966429861689e-05,
+      "loss": 1.0636,
+      "step": 5441
+    },
+    {
+      "epoch": 0.8346625766871165,
+      "grad_norm": 0.15652349591255188,
+      "learning_rate": 1.0087674493039346e-05,
+      "loss": 0.9849,
+      "step": 5442
+    },
+    {
+      "epoch": 0.8348159509202454,
+      "grad_norm": 0.15644009411334991,
+      "learning_rate": 1.0069397931899847e-05,
+      "loss": 1.0479,
+      "step": 5443
+    },
+    {
+      "epoch": 0.8349693251533742,
+      "grad_norm": 0.1419977843761444,
+      "learning_rate": 1.0051136750772304e-05,
+      "loss": 1.0003,
+      "step": 5444
+    },
+    {
+      "epoch": 0.8351226993865031,
+      "grad_norm": 0.155643492937088,
+      "learning_rate": 1.0032890953982247e-05,
+      "loss": 1.033,
+      "step": 5445
+    },
+    {
+      "epoch": 0.8352760736196319,
+      "grad_norm": 0.1505500227212906,
+      "learning_rate": 1.0014660545851489e-05,
+      "loss": 1.0388,
+      "step": 5446
+    },
+    {
+      "epoch": 0.8354294478527607,
+      "grad_norm": 0.15696249902248383,
+      "learning_rate": 9.996445530698288e-06,
+      "loss": 1.0551,
+      "step": 5447
+    },
+    {
+      "epoch": 0.8355828220858895,
+      "grad_norm": 0.1496182233095169,
+      "learning_rate": 9.978245912837172e-06,
+      "loss": 0.9693,
+      "step": 5448
+    },
+    {
+      "epoch": 0.8357361963190184,
+      "grad_norm": 0.15796005725860596,
+      "learning_rate": 9.960061696579055e-06,
+      "loss": 1.0909,
+      "step": 5449
+    },
+    {
+      "epoch": 0.8358895705521472,
+      "grad_norm": 0.15684518218040466,
+      "learning_rate": 9.941892886231228e-06,
+      "loss": 0.99,
+      "step": 5450
+    },
+    {
+      "epoch": 0.8360429447852761,
+      "grad_norm": 0.16117775440216064,
+      "learning_rate": 9.923739486097282e-06,
+      "loss": 1.0561,
+      "step": 5451
+    },
+    {
+      "epoch": 0.8361963190184049,
+      "grad_norm": 0.16109292209148407,
+      "learning_rate": 9.905601500477218e-06,
+      "loss": 1.0392,
+      "step": 5452
+    },
+    {
+      "epoch": 0.8363496932515337,
+      "grad_norm": 0.1526142805814743,
+      "learning_rate": 9.887478933667314e-06,
+      "loss": 1.04,
+      "step": 5453
+    },
+    {
+      "epoch": 0.8365030674846625,
+      "grad_norm": 0.15561571717262268,
+      "learning_rate": 9.869371789960274e-06,
+      "loss": 1.1004,
+      "step": 5454
+    },
+    {
+      "epoch": 0.8366564417177914,
+      "grad_norm": 0.16695371270179749,
+      "learning_rate": 9.851280073645087e-06,
+      "loss": 1.2127,
+      "step": 5455
+    },
+    {
+      "epoch": 0.8368098159509203,
+      "grad_norm": 0.14573006331920624,
+      "learning_rate": 9.833203789007097e-06,
+      "loss": 1.012,
+      "step": 5456
+    },
+    {
+      "epoch": 0.8369631901840491,
+      "grad_norm": 0.1553337723016739,
+      "learning_rate": 9.815142940328033e-06,
+      "loss": 1.0225,
+      "step": 5457
+    },
+    {
+      "epoch": 0.837116564417178,
+      "grad_norm": 0.15045616030693054,
+      "learning_rate": 9.797097531885925e-06,
+      "loss": 0.9824,
+      "step": 5458
+    },
+    {
+      "epoch": 0.8372699386503067,
+      "grad_norm": 0.14976078271865845,
+      "learning_rate": 9.77906756795517e-06,
+      "loss": 0.9884,
+      "step": 5459
+    },
+    {
+      "epoch": 0.8374233128834356,
+      "grad_norm": 0.15263327956199646,
+      "learning_rate": 9.761053052806479e-06,
+      "loss": 1.1564,
+      "step": 5460
+    },
+    {
+      "epoch": 0.8375766871165644,
+      "grad_norm": 0.16418568789958954,
+      "learning_rate": 9.743053990706958e-06,
+      "loss": 1.129,
+      "step": 5461
+    },
+    {
+      "epoch": 0.8377300613496933,
+      "grad_norm": 0.15491777658462524,
+      "learning_rate": 9.725070385919978e-06,
+      "loss": 1.0963,
+      "step": 5462
+    },
+    {
+      "epoch": 0.8378834355828221,
+      "grad_norm": 0.15767401456832886,
+      "learning_rate": 9.70710224270532e-06,
+      "loss": 1.0656,
+      "step": 5463
+    },
+    {
+      "epoch": 0.838036809815951,
+      "grad_norm": 0.1785578578710556,
+      "learning_rate": 9.689149565319055e-06,
+      "loss": 1.0527,
+      "step": 5464
+    },
+    {
+      "epoch": 0.8381901840490797,
+      "grad_norm": 0.1663561314344406,
+      "learning_rate": 9.671212358013589e-06,
+      "loss": 1.0316,
+      "step": 5465
+    },
+    {
+      "epoch": 0.8383435582822086,
+      "grad_norm": 0.14359532296657562,
+      "learning_rate": 9.653290625037723e-06,
+      "loss": 0.9915,
+      "step": 5466
+    },
+    {
+      "epoch": 0.8384969325153374,
+      "grad_norm": 0.1665794402360916,
+      "learning_rate": 9.635384370636507e-06,
+      "loss": 1.1015,
+      "step": 5467
+    },
+    {
+      "epoch": 0.8386503067484663,
+      "grad_norm": 0.14983418583869934,
+      "learning_rate": 9.6174935990514e-06,
+      "loss": 0.9878,
+      "step": 5468
+    },
+    {
+      "epoch": 0.8388036809815951,
+      "grad_norm": 0.15333761274814606,
+      "learning_rate": 9.599618314520136e-06,
+      "loss": 1.0718,
+      "step": 5469
+    },
+    {
+      "epoch": 0.838957055214724,
+      "grad_norm": 0.14969536662101746,
+      "learning_rate": 9.58175852127683e-06,
+      "loss": 1.0221,
+      "step": 5470
+    },
+    {
+      "epoch": 0.8391104294478527,
+      "grad_norm": 0.1728050857782364,
+      "learning_rate": 9.563914223551902e-06,
+      "loss": 1.107,
+      "step": 5471
+    },
+    {
+      "epoch": 0.8392638036809816,
+      "grad_norm": 0.15288516879081726,
+      "learning_rate": 9.546085425572069e-06,
+      "loss": 1.0769,
+      "step": 5472
+    },
+    {
+      "epoch": 0.8394171779141104,
+      "grad_norm": 0.15772855281829834,
+      "learning_rate": 9.528272131560463e-06,
+      "loss": 1.1074,
+      "step": 5473
+    },
+    {
+      "epoch": 0.8395705521472393,
+      "grad_norm": 0.15546779334545135,
+      "learning_rate": 9.51047434573644e-06,
+      "loss": 1.0464,
+      "step": 5474
+    },
+    {
+      "epoch": 0.8397239263803681,
+      "grad_norm": 0.1523028463125229,
+      "learning_rate": 9.49269207231578e-06,
+      "loss": 1.0048,
+      "step": 5475
+    },
+    {
+      "epoch": 0.839877300613497,
+      "grad_norm": 0.15174996852874756,
+      "learning_rate": 9.474925315510504e-06,
+      "loss": 0.9464,
+      "step": 5476
+    },
+    {
+      "epoch": 0.8400306748466257,
+      "grad_norm": 0.16688114404678345,
+      "learning_rate": 9.457174079529028e-06,
+      "loss": 1.1008,
+      "step": 5477
+    },
+    {
+      "epoch": 0.8401840490797546,
+      "grad_norm": 0.1573776751756668,
+      "learning_rate": 9.43943836857605e-06,
+      "loss": 1.0946,
+      "step": 5478
+    },
+    {
+      "epoch": 0.8403374233128834,
+      "grad_norm": 0.15243613719940186,
+      "learning_rate": 9.42171818685258e-06,
+      "loss": 0.8958,
+      "step": 5479
+    },
+    {
+      "epoch": 0.8404907975460123,
+      "grad_norm": 0.14717568457126617,
+      "learning_rate": 9.404013538555983e-06,
+      "loss": 1.0124,
+      "step": 5480
+    },
+    {
+      "epoch": 0.8406441717791411,
+      "grad_norm": 0.153679758310318,
+      "learning_rate": 9.386324427879941e-06,
+      "loss": 1.0615,
+      "step": 5481
+    },
+    {
+      "epoch": 0.84079754601227,
+      "grad_norm": 0.15122535824775696,
+      "learning_rate": 9.368650859014454e-06,
+      "loss": 0.9698,
+      "step": 5482
+    },
+    {
+      "epoch": 0.8409509202453987,
+      "grad_norm": 0.14592500030994415,
+      "learning_rate": 9.350992836145823e-06,
+      "loss": 0.9023,
+      "step": 5483
+    },
+    {
+      "epoch": 0.8411042944785276,
+      "grad_norm": 0.15335550904273987,
+      "learning_rate": 9.333350363456655e-06,
+      "loss": 1.0786,
+      "step": 5484
+    },
+    {
+      "epoch": 0.8412576687116564,
+      "grad_norm": 0.15528474748134613,
+      "learning_rate": 9.315723445125937e-06,
+      "loss": 1.1317,
+      "step": 5485
+    },
+    {
+      "epoch": 0.8414110429447853,
+      "grad_norm": 0.16234458982944489,
+      "learning_rate": 9.298112085328887e-06,
+      "loss": 1.0617,
+      "step": 5486
+    },
+    {
+      "epoch": 0.8415644171779141,
+      "grad_norm": 0.15061132609844208,
+      "learning_rate": 9.280516288237113e-06,
+      "loss": 1.0355,
+      "step": 5487
+    },
+    {
+      "epoch": 0.841717791411043,
+      "grad_norm": 0.14800190925598145,
+      "learning_rate": 9.262936058018488e-06,
+      "loss": 1.0341,
+      "step": 5488
+    },
+    {
+      "epoch": 0.8418711656441717,
+      "grad_norm": 0.1601727157831192,
+      "learning_rate": 9.245371398837224e-06,
+      "loss": 1.0408,
+      "step": 5489
+    },
+    {
+      "epoch": 0.8420245398773006,
+      "grad_norm": 0.15349048376083374,
+      "learning_rate": 9.227822314853839e-06,
+      "loss": 0.9977,
+      "step": 5490
+    },
+    {
+      "epoch": 0.8421779141104294,
+      "grad_norm": 0.15835987031459808,
+      "learning_rate": 9.210288810225128e-06,
+      "loss": 1.1202,
+      "step": 5491
+    },
+    {
+      "epoch": 0.8423312883435583,
+      "grad_norm": 0.15628068149089813,
+      "learning_rate": 9.19277088910426e-06,
+      "loss": 1.1293,
+      "step": 5492
+    },
+    {
+      "epoch": 0.8424846625766871,
+      "grad_norm": 0.15434089303016663,
+      "learning_rate": 9.175268555640655e-06,
+      "loss": 1.0622,
+      "step": 5493
+    },
+    {
+      "epoch": 0.842638036809816,
+      "grad_norm": 0.1638839691877365,
+      "learning_rate": 9.157781813980084e-06,
+      "loss": 1.1796,
+      "step": 5494
+    },
+    {
+      "epoch": 0.8427914110429447,
+      "grad_norm": 0.1602957844734192,
+      "learning_rate": 9.140310668264572e-06,
+      "loss": 0.9633,
+      "step": 5495
+    },
+    {
+      "epoch": 0.8429447852760736,
+      "grad_norm": 0.1535712629556656,
+      "learning_rate": 9.122855122632528e-06,
+      "loss": 1.057,
+      "step": 5496
+    },
+    {
+      "epoch": 0.8430981595092024,
+      "grad_norm": 0.21786567568778992,
+      "learning_rate": 9.105415181218576e-06,
+      "loss": 1.1051,
+      "step": 5497
+    },
+    {
+      "epoch": 0.8432515337423313,
+      "grad_norm": 0.1574670970439911,
+      "learning_rate": 9.08799084815372e-06,
+      "loss": 1.0735,
+      "step": 5498
+    },
+    {
+      "epoch": 0.8434049079754601,
+      "grad_norm": 0.15386682748794556,
+      "learning_rate": 9.070582127565235e-06,
+      "loss": 1.0478,
+      "step": 5499
+    },
+    {
+      "epoch": 0.843558282208589,
+      "grad_norm": 0.1460115909576416,
+      "learning_rate": 9.053189023576662e-06,
+      "loss": 1.0405,
+      "step": 5500
+    },
+    {
+      "epoch": 0.8437116564417177,
+      "grad_norm": 0.15707892179489136,
+      "learning_rate": 9.035811540307917e-06,
+      "loss": 1.0194,
+      "step": 5501
+    },
+    {
+      "epoch": 0.8438650306748466,
+      "grad_norm": 0.1551826447248459,
+      "learning_rate": 9.018449681875154e-06,
+      "loss": 1.0742,
+      "step": 5502
+    },
+    {
+      "epoch": 0.8440184049079754,
+      "grad_norm": 0.1523311734199524,
+      "learning_rate": 9.001103452390867e-06,
+      "loss": 1.1318,
+      "step": 5503
+    },
+    {
+      "epoch": 0.8441717791411043,
+      "grad_norm": 0.15955227613449097,
+      "learning_rate": 8.983772855963814e-06,
+      "loss": 1.1181,
+      "step": 5504
+    },
+    {
+      "epoch": 0.8443251533742331,
+      "grad_norm": 0.15758192539215088,
+      "learning_rate": 8.966457896699092e-06,
+      "loss": 0.9505,
+      "step": 5505
+    },
+    {
+      "epoch": 0.844478527607362,
+      "grad_norm": 0.15061746537685394,
+      "learning_rate": 8.949158578698045e-06,
+      "loss": 1.0582,
+      "step": 5506
+    },
+    {
+      "epoch": 0.8446319018404908,
+      "grad_norm": 0.17009229958057404,
+      "learning_rate": 8.931874906058339e-06,
+      "loss": 1.2163,
+      "step": 5507
+    },
+    {
+      "epoch": 0.8447852760736196,
+      "grad_norm": 0.14818435907363892,
+      "learning_rate": 8.914606882873948e-06,
+      "loss": 0.9994,
+      "step": 5508
+    },
+    {
+      "epoch": 0.8449386503067484,
+      "grad_norm": 0.16651393473148346,
+      "learning_rate": 8.89735451323509e-06,
+      "loss": 1.1373,
+      "step": 5509
+    },
+    {
+      "epoch": 0.8450920245398773,
+      "grad_norm": 0.16477398574352264,
+      "learning_rate": 8.880117801228343e-06,
+      "loss": 1.058,
+      "step": 5510
+    },
+    {
+      "epoch": 0.8452453987730061,
+      "grad_norm": 0.15565432608127594,
+      "learning_rate": 8.86289675093651e-06,
+      "loss": 0.9197,
+      "step": 5511
+    },
+    {
+      "epoch": 0.845398773006135,
+      "grad_norm": 0.1460820883512497,
+      "learning_rate": 8.845691366438748e-06,
+      "loss": 1.0058,
+      "step": 5512
+    },
+    {
+      "epoch": 0.8455521472392638,
+      "grad_norm": 0.15284599363803864,
+      "learning_rate": 8.828501651810447e-06,
+      "loss": 1.1113,
+      "step": 5513
+    },
+    {
+      "epoch": 0.8457055214723926,
+      "grad_norm": 0.14608140289783478,
+      "learning_rate": 8.811327611123298e-06,
+      "loss": 1.0163,
+      "step": 5514
+    },
+    {
+      "epoch": 0.8458588957055214,
+      "grad_norm": 0.15169358253479004,
+      "learning_rate": 8.794169248445316e-06,
+      "loss": 0.9491,
+      "step": 5515
+    },
+    {
+      "epoch": 0.8460122699386503,
+      "grad_norm": 0.16342386603355408,
+      "learning_rate": 8.777026567840756e-06,
+      "loss": 1.1229,
+      "step": 5516
+    },
+    {
+      "epoch": 0.8461656441717792,
+      "grad_norm": 0.16035307943820953,
+      "learning_rate": 8.759899573370196e-06,
+      "loss": 1.105,
+      "step": 5517
+    },
+    {
+      "epoch": 0.846319018404908,
+      "grad_norm": 0.17152419686317444,
+      "learning_rate": 8.742788269090459e-06,
+      "loss": 1.0927,
+      "step": 5518
+    },
+    {
+      "epoch": 0.8464723926380369,
+      "grad_norm": 0.1613141894340515,
+      "learning_rate": 8.725692659054701e-06,
+      "loss": 1.1092,
+      "step": 5519
+    },
+    {
+      "epoch": 0.8466257668711656,
+      "grad_norm": 0.15692268311977386,
+      "learning_rate": 8.7086127473123e-06,
+      "loss": 1.1002,
+      "step": 5520
+    },
+    {
+      "epoch": 0.8467791411042945,
+      "grad_norm": 0.1704825758934021,
+      "learning_rate": 8.691548537908986e-06,
+      "loss": 1.1639,
+      "step": 5521
+    },
+    {
+      "epoch": 0.8469325153374233,
+      "grad_norm": 0.15934710204601288,
+      "learning_rate": 8.674500034886708e-06,
+      "loss": 1.1319,
+      "step": 5522
+    },
+    {
+      "epoch": 0.8470858895705522,
+      "grad_norm": 0.1421310305595398,
+      "learning_rate": 8.657467242283698e-06,
+      "loss": 0.8685,
+      "step": 5523
+    },
+    {
+      "epoch": 0.847239263803681,
+      "grad_norm": 0.14724200963974,
+      "learning_rate": 8.640450164134532e-06,
+      "loss": 1.0654,
+      "step": 5524
+    },
+    {
+      "epoch": 0.8473926380368099,
+      "grad_norm": 0.1422833800315857,
+      "learning_rate": 8.623448804470007e-06,
+      "loss": 0.9433,
+      "step": 5525
+    },
+    {
+      "epoch": 0.8475460122699386,
+      "grad_norm": 0.14209480583667755,
+      "learning_rate": 8.60646316731719e-06,
+      "loss": 0.9804,
+      "step": 5526
+    },
+    {
+      "epoch": 0.8476993865030675,
+      "grad_norm": 0.15032771229743958,
+      "learning_rate": 8.589493256699464e-06,
+      "loss": 1.059,
+      "step": 5527
+    },
+    {
+      "epoch": 0.8478527607361963,
+      "grad_norm": 0.16555677354335785,
+      "learning_rate": 8.572539076636439e-06,
+      "loss": 1.0559,
+      "step": 5528
+    },
+    {
+      "epoch": 0.8480061349693252,
+      "grad_norm": 0.16199877858161926,
+      "learning_rate": 8.555600631144061e-06,
+      "loss": 1.0561,
+      "step": 5529
+    },
+    {
+      "epoch": 0.848159509202454,
+      "grad_norm": 0.15513648092746735,
+      "learning_rate": 8.538677924234477e-06,
+      "loss": 1.0377,
+      "step": 5530
+    },
+    {
+      "epoch": 0.8483128834355829,
+      "grad_norm": 0.16461271047592163,
+      "learning_rate": 8.521770959916183e-06,
+      "loss": 1.0966,
+      "step": 5531
+    },
+    {
+      "epoch": 0.8484662576687116,
+      "grad_norm": 0.14730577170848846,
+      "learning_rate": 8.504879742193856e-06,
+      "loss": 1.0149,
+      "step": 5532
+    },
+    {
+      "epoch": 0.8486196319018405,
+      "grad_norm": 0.15010027587413788,
+      "learning_rate": 8.488004275068549e-06,
+      "loss": 1.0445,
+      "step": 5533
+    },
+    {
+      "epoch": 0.8487730061349693,
+      "grad_norm": 0.14839251339435577,
+      "learning_rate": 8.471144562537489e-06,
+      "loss": 0.9397,
+      "step": 5534
+    },
+    {
+      "epoch": 0.8489263803680982,
+      "grad_norm": 0.16183850169181824,
+      "learning_rate": 8.45430060859421e-06,
+      "loss": 1.0417,
+      "step": 5535
+    },
+    {
+      "epoch": 0.849079754601227,
+      "grad_norm": 0.15728166699409485,
+      "learning_rate": 8.437472417228535e-06,
+      "loss": 0.8805,
+      "step": 5536
+    },
+    {
+      "epoch": 0.8492331288343559,
+      "grad_norm": 0.14943775534629822,
+      "learning_rate": 8.420659992426507e-06,
+      "loss": 0.8952,
+      "step": 5537
+    },
+    {
+      "epoch": 0.8493865030674846,
+      "grad_norm": 0.14686070382595062,
+      "learning_rate": 8.403863338170497e-06,
+      "loss": 1.0823,
+      "step": 5538
+    },
+    {
+      "epoch": 0.8495398773006135,
+      "grad_norm": 0.13618247210979462,
+      "learning_rate": 8.387082458439048e-06,
+      "loss": 0.8082,
+      "step": 5539
+    },
+    {
+      "epoch": 0.8496932515337423,
+      "grad_norm": 0.16547273099422455,
+      "learning_rate": 8.370317357207074e-06,
+      "loss": 1.1187,
+      "step": 5540
+    },
+    {
+      "epoch": 0.8498466257668712,
+      "grad_norm": 0.15273410081863403,
+      "learning_rate": 8.353568038445679e-06,
+      "loss": 0.9614,
+      "step": 5541
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.14407038688659668,
+      "learning_rate": 8.336834506122233e-06,
+      "loss": 0.9804,
+      "step": 5542
+    },
+    {
+      "epoch": 0.8501533742331289,
+      "grad_norm": 0.1478063464164734,
+      "learning_rate": 8.320116764200398e-06,
+      "loss": 0.9807,
+      "step": 5543
+    },
+    {
+      "epoch": 0.8503067484662576,
+      "grad_norm": 0.17456699907779694,
+      "learning_rate": 8.303414816640074e-06,
+      "loss": 1.2742,
+      "step": 5544
+    },
+    {
+      "epoch": 0.8504601226993865,
+      "grad_norm": 0.14349602162837982,
+      "learning_rate": 8.286728667397439e-06,
+      "loss": 1.0168,
+      "step": 5545
+    },
+    {
+      "epoch": 0.8506134969325153,
+      "grad_norm": 0.17023174464702606,
+      "learning_rate": 8.270058320424894e-06,
+      "loss": 1.1934,
+      "step": 5546
+    },
+    {
+      "epoch": 0.8507668711656442,
+      "grad_norm": 0.1487690657377243,
+      "learning_rate": 8.253403779671142e-06,
+      "loss": 0.9798,
+      "step": 5547
+    },
+    {
+      "epoch": 0.850920245398773,
+      "grad_norm": 0.13945037126541138,
+      "learning_rate": 8.2367650490811e-06,
+      "loss": 0.9017,
+      "step": 5548
+    },
+    {
+      "epoch": 0.8510736196319019,
+      "grad_norm": 0.15389849245548248,
+      "learning_rate": 8.220142132595953e-06,
+      "loss": 1.0629,
+      "step": 5549
+    },
+    {
+      "epoch": 0.8512269938650306,
+      "grad_norm": 0.14960312843322754,
+      "learning_rate": 8.203535034153172e-06,
+      "loss": 1.0538,
+      "step": 5550
+    },
+    {
+      "epoch": 0.8513803680981595,
+      "grad_norm": 0.16190311312675476,
+      "learning_rate": 8.186943757686428e-06,
+      "loss": 1.0976,
+      "step": 5551
+    },
+    {
+      "epoch": 0.8515337423312883,
+      "grad_norm": 0.16071808338165283,
+      "learning_rate": 8.170368307125696e-06,
+      "loss": 1.1812,
+      "step": 5552
+    },
+    {
+      "epoch": 0.8516871165644172,
+      "grad_norm": 0.15176630020141602,
+      "learning_rate": 8.153808686397159e-06,
+      "loss": 0.8509,
+      "step": 5553
+    },
+    {
+      "epoch": 0.851840490797546,
+      "grad_norm": 0.16415463387966156,
+      "learning_rate": 8.137264899423279e-06,
+      "loss": 1.0035,
+      "step": 5554
+    },
+    {
+      "epoch": 0.8519938650306749,
+      "grad_norm": 0.15081705152988434,
+      "learning_rate": 8.120736950122739e-06,
+      "loss": 1.0753,
+      "step": 5555
+    },
+    {
+      "epoch": 0.8521472392638036,
+      "grad_norm": 0.15228867530822754,
+      "learning_rate": 8.104224842410523e-06,
+      "loss": 0.8719,
+      "step": 5556
+    },
+    {
+      "epoch": 0.8523006134969325,
+      "grad_norm": 0.15205180644989014,
+      "learning_rate": 8.087728580197803e-06,
+      "loss": 1.0093,
+      "step": 5557
+    },
+    {
+      "epoch": 0.8524539877300613,
+      "grad_norm": 0.1625048965215683,
+      "learning_rate": 8.07124816739202e-06,
+      "loss": 1.0412,
+      "step": 5558
+    },
+    {
+      "epoch": 0.8526073619631902,
+      "grad_norm": 0.15340091288089752,
+      "learning_rate": 8.054783607896884e-06,
+      "loss": 1.0491,
+      "step": 5559
+    },
+    {
+      "epoch": 0.852760736196319,
+      "grad_norm": 0.15574215352535248,
+      "learning_rate": 8.03833490561231e-06,
+      "loss": 1.0913,
+      "step": 5560
+    },
+    {
+      "epoch": 0.8529141104294479,
+      "grad_norm": 0.1605466902256012,
+      "learning_rate": 8.021902064434488e-06,
+      "loss": 1.0424,
+      "step": 5561
+    },
+    {
+      "epoch": 0.8530674846625766,
+      "grad_norm": 0.1534646898508072,
+      "learning_rate": 8.005485088255831e-06,
+      "loss": 1.0458,
+      "step": 5562
+    },
+    {
+      "epoch": 0.8532208588957055,
+      "grad_norm": 0.13794498145580292,
+      "learning_rate": 7.989083980965027e-06,
+      "loss": 0.8199,
+      "step": 5563
+    },
+    {
+      "epoch": 0.8533742331288343,
+      "grad_norm": 0.1555621325969696,
+      "learning_rate": 7.972698746446962e-06,
+      "loss": 0.9251,
+      "step": 5564
+    },
+    {
+      "epoch": 0.8535276073619632,
+      "grad_norm": 0.16353701055049896,
+      "learning_rate": 7.956329388582754e-06,
+      "loss": 1.0245,
+      "step": 5565
+    },
+    {
+      "epoch": 0.853680981595092,
+      "grad_norm": 0.16461163759231567,
+      "learning_rate": 7.939975911249846e-06,
+      "loss": 1.1402,
+      "step": 5566
+    },
+    {
+      "epoch": 0.8538343558282209,
+      "grad_norm": 0.15421150624752045,
+      "learning_rate": 7.923638318321813e-06,
+      "loss": 1.0235,
+      "step": 5567
+    },
+    {
+      "epoch": 0.8539877300613496,
+      "grad_norm": 0.17076607048511505,
+      "learning_rate": 7.907316613668548e-06,
+      "loss": 1.1922,
+      "step": 5568
+    },
+    {
+      "epoch": 0.8541411042944785,
+      "grad_norm": 0.15002666413784027,
+      "learning_rate": 7.891010801156147e-06,
+      "loss": 0.9919,
+      "step": 5569
+    },
+    {
+      "epoch": 0.8542944785276073,
+      "grad_norm": 0.1636301428079605,
+      "learning_rate": 7.874720884646901e-06,
+      "loss": 1.0625,
+      "step": 5570
+    },
+    {
+      "epoch": 0.8544478527607362,
+      "grad_norm": 0.16075609624385834,
+      "learning_rate": 7.858446867999432e-06,
+      "loss": 1.1612,
+      "step": 5571
+    },
+    {
+      "epoch": 0.854601226993865,
+      "grad_norm": 0.16149021685123444,
+      "learning_rate": 7.842188755068491e-06,
+      "loss": 1.1817,
+      "step": 5572
+    },
+    {
+      "epoch": 0.8547546012269939,
+      "grad_norm": 0.16690771281719208,
+      "learning_rate": 7.825946549705162e-06,
+      "loss": 1.1528,
+      "step": 5573
+    },
+    {
+      "epoch": 0.8549079754601226,
+      "grad_norm": 0.15787088871002197,
+      "learning_rate": 7.80972025575667e-06,
+      "loss": 0.9686,
+      "step": 5574
+    },
+    {
+      "epoch": 0.8550613496932515,
+      "grad_norm": 0.15853403508663177,
+      "learning_rate": 7.793509877066537e-06,
+      "loss": 1.0622,
+      "step": 5575
+    },
+    {
+      "epoch": 0.8552147239263803,
+      "grad_norm": 0.15507937967777252,
+      "learning_rate": 7.777315417474486e-06,
+      "loss": 1.0879,
+      "step": 5576
+    },
+    {
+      "epoch": 0.8553680981595092,
+      "grad_norm": 0.1576450616121292,
+      "learning_rate": 7.761136880816458e-06,
+      "loss": 1.0394,
+      "step": 5577
+    },
+    {
+      "epoch": 0.855521472392638,
+      "grad_norm": 0.15685135126113892,
+      "learning_rate": 7.744974270924659e-06,
+      "loss": 1.1461,
+      "step": 5578
+    },
+    {
+      "epoch": 0.8556748466257669,
+      "grad_norm": 0.1656665802001953,
+      "learning_rate": 7.728827591627479e-06,
+      "loss": 1.1988,
+      "step": 5579
+    },
+    {
+      "epoch": 0.8558282208588958,
+      "grad_norm": 0.16449320316314697,
+      "learning_rate": 7.712696846749592e-06,
+      "loss": 1.0602,
+      "step": 5580
+    },
+    {
+      "epoch": 0.8559815950920245,
+      "grad_norm": 0.14780840277671814,
+      "learning_rate": 7.696582040111824e-06,
+      "loss": 1.1092,
+      "step": 5581
+    },
+    {
+      "epoch": 0.8561349693251534,
+      "grad_norm": 0.14205093681812286,
+      "learning_rate": 7.680483175531296e-06,
+      "loss": 1.0279,
+      "step": 5582
+    },
+    {
+      "epoch": 0.8562883435582822,
+      "grad_norm": 0.1475415676832199,
+      "learning_rate": 7.66440025682131e-06,
+      "loss": 0.9431,
+      "step": 5583
+    },
+    {
+      "epoch": 0.8564417177914111,
+      "grad_norm": 0.1445731818675995,
+      "learning_rate": 7.648333287791375e-06,
+      "loss": 1.0162,
+      "step": 5584
+    },
+    {
+      "epoch": 0.8565950920245399,
+      "grad_norm": 0.1516113579273224,
+      "learning_rate": 7.632282272247298e-06,
+      "loss": 0.9852,
+      "step": 5585
+    },
+    {
+      "epoch": 0.8567484662576688,
+      "grad_norm": 0.1591026484966278,
+      "learning_rate": 7.616247213991019e-06,
+      "loss": 1.1477,
+      "step": 5586
+    },
+    {
+      "epoch": 0.8569018404907975,
+      "grad_norm": 0.1445356011390686,
+      "learning_rate": 7.6002281168207694e-06,
+      "loss": 0.978,
+      "step": 5587
+    },
+    {
+      "epoch": 0.8570552147239264,
+      "grad_norm": 0.167802095413208,
+      "learning_rate": 7.584224984530929e-06,
+      "loss": 1.2659,
+      "step": 5588
+    },
+    {
+      "epoch": 0.8572085889570552,
+      "grad_norm": 0.16223134100437164,
+      "learning_rate": 7.568237820912182e-06,
+      "loss": 1.1093,
+      "step": 5589
+    },
+    {
+      "epoch": 0.8573619631901841,
+      "grad_norm": 0.1420927196741104,
+      "learning_rate": 7.552266629751342e-06,
+      "loss": 0.9792,
+      "step": 5590
+    },
+    {
+      "epoch": 0.8575153374233129,
+      "grad_norm": 0.15963011980056763,
+      "learning_rate": 7.536311414831505e-06,
+      "loss": 1.1316,
+      "step": 5591
+    },
+    {
+      "epoch": 0.8576687116564418,
+      "grad_norm": 0.14417342841625214,
+      "learning_rate": 7.520372179931961e-06,
+      "loss": 0.8868,
+      "step": 5592
+    },
+    {
+      "epoch": 0.8578220858895705,
+      "grad_norm": 0.14535951614379883,
+      "learning_rate": 7.504448928828177e-06,
+      "loss": 0.923,
+      "step": 5593
+    },
+    {
+      "epoch": 0.8579754601226994,
+      "grad_norm": 0.15454821288585663,
+      "learning_rate": 7.4885416652919225e-06,
+      "loss": 1.0719,
+      "step": 5594
+    },
+    {
+      "epoch": 0.8581288343558282,
+      "grad_norm": 0.16672104597091675,
+      "learning_rate": 7.472650393091073e-06,
+      "loss": 1.1107,
+      "step": 5595
+    },
+    {
+      "epoch": 0.8582822085889571,
+      "grad_norm": 0.15474580228328705,
+      "learning_rate": 7.4567751159898105e-06,
+      "loss": 0.9601,
+      "step": 5596
+    },
+    {
+      "epoch": 0.8584355828220859,
+      "grad_norm": 0.1462346464395523,
+      "learning_rate": 7.4409158377484594e-06,
+      "loss": 1.0073,
+      "step": 5597
+    },
+    {
+      "epoch": 0.8585889570552148,
+      "grad_norm": 0.18017476797103882,
+      "learning_rate": 7.425072562123607e-06,
+      "loss": 0.9418,
+      "step": 5598
+    },
+    {
+      "epoch": 0.8587423312883435,
+      "grad_norm": 0.15724287927150726,
+      "learning_rate": 7.409245292868019e-06,
+      "loss": 1.0899,
+      "step": 5599
+    },
+    {
+      "epoch": 0.8588957055214724,
+      "grad_norm": 0.15466810762882233,
+      "learning_rate": 7.393434033730653e-06,
+      "loss": 1.0441,
+      "step": 5600
+    },
+    {
+      "epoch": 0.8590490797546012,
+      "grad_norm": 0.1493416279554367,
+      "learning_rate": 7.37763878845673e-06,
+      "loss": 1.0341,
+      "step": 5601
+    },
+    {
+      "epoch": 0.8592024539877301,
+      "grad_norm": 0.15843386948108673,
+      "learning_rate": 7.361859560787617e-06,
+      "loss": 1.0498,
+      "step": 5602
+    },
+    {
+      "epoch": 0.8593558282208589,
+      "grad_norm": 0.15831826627254486,
+      "learning_rate": 7.346096354460948e-06,
+      "loss": 1.1462,
+      "step": 5603
+    },
+    {
+      "epoch": 0.8595092024539878,
+      "grad_norm": 0.14763209223747253,
+      "learning_rate": 7.330349173210495e-06,
+      "loss": 1.0299,
+      "step": 5604
+    },
+    {
+      "epoch": 0.8596625766871165,
+      "grad_norm": 0.17075896263122559,
+      "learning_rate": 7.314618020766291e-06,
+      "loss": 1.0404,
+      "step": 5605
+    },
+    {
+      "epoch": 0.8598159509202454,
+      "grad_norm": 0.13892310857772827,
+      "learning_rate": 7.298902900854556e-06,
+      "loss": 0.962,
+      "step": 5606
+    },
+    {
+      "epoch": 0.8599693251533742,
+      "grad_norm": 0.16215568780899048,
+      "learning_rate": 7.283203817197661e-06,
+      "loss": 1.0646,
+      "step": 5607
+    },
+    {
+      "epoch": 0.8601226993865031,
+      "grad_norm": 0.14909866452217102,
+      "learning_rate": 7.267520773514282e-06,
+      "loss": 0.9813,
+      "step": 5608
+    },
+    {
+      "epoch": 0.8602760736196319,
+      "grad_norm": 0.14133580029010773,
+      "learning_rate": 7.251853773519198e-06,
+      "loss": 0.8685,
+      "step": 5609
+    },
+    {
+      "epoch": 0.8604294478527608,
+      "grad_norm": 0.15181945264339447,
+      "learning_rate": 7.236202820923456e-06,
+      "loss": 0.9715,
+      "step": 5610
+    },
+    {
+      "epoch": 0.8605828220858895,
+      "grad_norm": 0.13881054520606995,
+      "learning_rate": 7.220567919434267e-06,
+      "loss": 1.0265,
+      "step": 5611
+    },
+    {
+      "epoch": 0.8607361963190184,
+      "grad_norm": 0.14711779356002808,
+      "learning_rate": 7.204949072755018e-06,
+      "loss": 1.0864,
+      "step": 5612
+    },
+    {
+      "epoch": 0.8608895705521472,
+      "grad_norm": 0.16237761080265045,
+      "learning_rate": 7.189346284585349e-06,
+      "loss": 1.0683,
+      "step": 5613
+    },
+    {
+      "epoch": 0.8610429447852761,
+      "grad_norm": 0.15689440071582794,
+      "learning_rate": 7.173759558621054e-06,
+      "loss": 1.0003,
+      "step": 5614
+    },
+    {
+      "epoch": 0.8611963190184049,
+      "grad_norm": 0.15908172726631165,
+      "learning_rate": 7.158188898554154e-06,
+      "loss": 1.0826,
+      "step": 5615
+    },
+    {
+      "epoch": 0.8613496932515338,
+      "grad_norm": 0.17141461372375488,
+      "learning_rate": 7.142634308072826e-06,
+      "loss": 1.091,
+      "step": 5616
+    },
+    {
+      "epoch": 0.8615030674846625,
+      "grad_norm": 0.1638488620519638,
+      "learning_rate": 7.127095790861478e-06,
+      "loss": 1.0644,
+      "step": 5617
+    },
+    {
+      "epoch": 0.8616564417177914,
+      "grad_norm": 0.15371571481227875,
+      "learning_rate": 7.111573350600694e-06,
+      "loss": 1.0816,
+      "step": 5618
+    },
+    {
+      "epoch": 0.8618098159509202,
+      "grad_norm": 0.15320442616939545,
+      "learning_rate": 7.096066990967231e-06,
+      "loss": 1.0868,
+      "step": 5619
+    },
+    {
+      "epoch": 0.8619631901840491,
+      "grad_norm": 0.153359517455101,
+      "learning_rate": 7.080576715634092e-06,
+      "loss": 1.0091,
+      "step": 5620
+    },
+    {
+      "epoch": 0.8621165644171779,
+      "grad_norm": 0.14629200100898743,
+      "learning_rate": 7.065102528270394e-06,
+      "loss": 0.9277,
+      "step": 5621
+    },
+    {
+      "epoch": 0.8622699386503068,
+      "grad_norm": 0.17239609360694885,
+      "learning_rate": 7.0496444325415205e-06,
+      "loss": 0.9807,
+      "step": 5622
+    },
+    {
+      "epoch": 0.8624233128834355,
+      "grad_norm": 0.15601126849651337,
+      "learning_rate": 7.0342024321089766e-06,
+      "loss": 1.0955,
+      "step": 5623
+    },
+    {
+      "epoch": 0.8625766871165644,
+      "grad_norm": 0.1645570546388626,
+      "learning_rate": 7.01877653063051e-06,
+      "loss": 1.0895,
+      "step": 5624
+    },
+    {
+      "epoch": 0.8627300613496932,
+      "grad_norm": 0.15011246502399445,
+      "learning_rate": 7.003366731760007e-06,
+      "loss": 1.0955,
+      "step": 5625
+    },
+    {
+      "epoch": 0.8628834355828221,
+      "grad_norm": 0.1649065911769867,
+      "learning_rate": 6.98797303914759e-06,
+      "loss": 1.1323,
+      "step": 5626
+    },
+    {
+      "epoch": 0.8630368098159509,
+      "grad_norm": 0.16816462576389313,
+      "learning_rate": 6.972595456439525e-06,
+      "loss": 1.1375,
+      "step": 5627
+    },
+    {
+      "epoch": 0.8631901840490798,
+      "grad_norm": 0.14642401039600372,
+      "learning_rate": 6.957233987278257e-06,
+      "loss": 0.9752,
+      "step": 5628
+    },
+    {
+      "epoch": 0.8633435582822085,
+      "grad_norm": 0.15308232605457306,
+      "learning_rate": 6.941888635302475e-06,
+      "loss": 1.0233,
+      "step": 5629
+    },
+    {
+      "epoch": 0.8634969325153374,
+      "grad_norm": 0.1534113883972168,
+      "learning_rate": 6.926559404146964e-06,
+      "loss": 1.0332,
+      "step": 5630
+    },
+    {
+      "epoch": 0.8636503067484662,
+      "grad_norm": 0.14583458006381989,
+      "learning_rate": 6.911246297442777e-06,
+      "loss": 1.0237,
+      "step": 5631
+    },
+    {
+      "epoch": 0.8638036809815951,
+      "grad_norm": 0.16419419646263123,
+      "learning_rate": 6.895949318817071e-06,
+      "loss": 1.1747,
+      "step": 5632
+    },
+    {
+      "epoch": 0.8639570552147239,
+      "grad_norm": 0.16018326580524445,
+      "learning_rate": 6.880668471893255e-06,
+      "loss": 1.097,
+      "step": 5633
+    },
+    {
+      "epoch": 0.8641104294478528,
+      "grad_norm": 0.14795129001140594,
+      "learning_rate": 6.865403760290852e-06,
+      "loss": 0.9339,
+      "step": 5634
+    },
+    {
+      "epoch": 0.8642638036809815,
+      "grad_norm": 0.16700689494609833,
+      "learning_rate": 6.850155187625586e-06,
+      "loss": 1.0755,
+      "step": 5635
+    },
+    {
+      "epoch": 0.8644171779141104,
+      "grad_norm": 0.15911705791950226,
+      "learning_rate": 6.834922757509384e-06,
+      "loss": 1.0772,
+      "step": 5636
+    },
+    {
+      "epoch": 0.8645705521472392,
+      "grad_norm": 0.163739874958992,
+      "learning_rate": 6.81970647355031e-06,
+      "loss": 1.1749,
+      "step": 5637
+    },
+    {
+      "epoch": 0.8647239263803681,
+      "grad_norm": 0.16091439127922058,
+      "learning_rate": 6.804506339352647e-06,
+      "loss": 1.0715,
+      "step": 5638
+    },
+    {
+      "epoch": 0.8648773006134969,
+      "grad_norm": 0.1558120995759964,
+      "learning_rate": 6.789322358516791e-06,
+      "loss": 0.9177,
+      "step": 5639
+    },
+    {
+      "epoch": 0.8650306748466258,
+      "grad_norm": 0.15639573335647583,
+      "learning_rate": 6.774154534639381e-06,
+      "loss": 0.9874,
+      "step": 5640
+    },
+    {
+      "epoch": 0.8651840490797545,
+      "grad_norm": 0.1402474045753479,
+      "learning_rate": 6.759002871313185e-06,
+      "loss": 0.9932,
+      "step": 5641
+    },
+    {
+      "epoch": 0.8653374233128834,
+      "grad_norm": 0.15420833230018616,
+      "learning_rate": 6.74386737212714e-06,
+      "loss": 1.0172,
+      "step": 5642
+    },
+    {
+      "epoch": 0.8654907975460123,
+      "grad_norm": 0.1446029543876648,
+      "learning_rate": 6.728748040666396e-06,
+      "loss": 1.0345,
+      "step": 5643
+    },
+    {
+      "epoch": 0.8656441717791411,
+      "grad_norm": 0.15381528437137604,
+      "learning_rate": 6.71364488051222e-06,
+      "loss": 1.0182,
+      "step": 5644
+    },
+    {
+      "epoch": 0.86579754601227,
+      "grad_norm": 0.1527973860502243,
+      "learning_rate": 6.698557895242093e-06,
+      "loss": 1.0749,
+      "step": 5645
+    },
+    {
+      "epoch": 0.8659509202453988,
+      "grad_norm": 0.16014711558818817,
+      "learning_rate": 6.683487088429631e-06,
+      "loss": 1.1464,
+      "step": 5646
+    },
+    {
+      "epoch": 0.8661042944785277,
+      "grad_norm": 0.13359317183494568,
+      "learning_rate": 6.668432463644646e-06,
+      "loss": 0.9052,
+      "step": 5647
+    },
+    {
+      "epoch": 0.8662576687116564,
+      "grad_norm": 0.15322217345237732,
+      "learning_rate": 6.653394024453102e-06,
+      "loss": 0.9752,
+      "step": 5648
+    },
+    {
+      "epoch": 0.8664110429447853,
+      "grad_norm": 0.15306906402111053,
+      "learning_rate": 6.638371774417115e-06,
+      "loss": 0.9996,
+      "step": 5649
+    },
+    {
+      "epoch": 0.8665644171779141,
+      "grad_norm": 0.15372596681118011,
+      "learning_rate": 6.623365717094989e-06,
+      "loss": 1.1019,
+      "step": 5650
+    },
+    {
+      "epoch": 0.866717791411043,
+      "grad_norm": 0.15685302019119263,
+      "learning_rate": 6.608375856041198e-06,
+      "loss": 1.2344,
+      "step": 5651
+    },
+    {
+      "epoch": 0.8668711656441718,
+      "grad_norm": 0.15203475952148438,
+      "learning_rate": 6.593402194806369e-06,
+      "loss": 1.187,
+      "step": 5652
+    },
+    {
+      "epoch": 0.8670245398773007,
+      "grad_norm": 0.15334521234035492,
+      "learning_rate": 6.57844473693728e-06,
+      "loss": 1.0855,
+      "step": 5653
+    },
+    {
+      "epoch": 0.8671779141104294,
+      "grad_norm": 0.15749193727970123,
+      "learning_rate": 6.5635034859768735e-06,
+      "loss": 1.0365,
+      "step": 5654
+    },
+    {
+      "epoch": 0.8673312883435583,
+      "grad_norm": 0.15271078050136566,
+      "learning_rate": 6.548578445464284e-06,
+      "loss": 1.0393,
+      "step": 5655
+    },
+    {
+      "epoch": 0.8674846625766871,
+      "grad_norm": 0.16541656851768494,
+      "learning_rate": 6.53366961893475e-06,
+      "loss": 1.1696,
+      "step": 5656
+    },
+    {
+      "epoch": 0.867638036809816,
+      "grad_norm": 0.15660600364208221,
+      "learning_rate": 6.51877700991974e-06,
+      "loss": 1.0349,
+      "step": 5657
+    },
+    {
+      "epoch": 0.8677914110429448,
+      "grad_norm": 0.14697322249412537,
+      "learning_rate": 6.503900621946806e-06,
+      "loss": 0.964,
+      "step": 5658
+    },
+    {
+      "epoch": 0.8679447852760737,
+      "grad_norm": 0.16233357787132263,
+      "learning_rate": 6.4890404585397295e-06,
+      "loss": 1.0062,
+      "step": 5659
+    },
+    {
+      "epoch": 0.8680981595092024,
+      "grad_norm": 0.16188350319862366,
+      "learning_rate": 6.474196523218395e-06,
+      "loss": 1.1038,
+      "step": 5660
+    },
+    {
+      "epoch": 0.8682515337423313,
+      "grad_norm": 0.14782607555389404,
+      "learning_rate": 6.459368819498856e-06,
+      "loss": 1.0121,
+      "step": 5661
+    },
+    {
+      "epoch": 0.8684049079754601,
+      "grad_norm": 0.14797140657901764,
+      "learning_rate": 6.4445573508933535e-06,
+      "loss": 1.1289,
+      "step": 5662
+    },
+    {
+      "epoch": 0.868558282208589,
+      "grad_norm": 0.1595693826675415,
+      "learning_rate": 6.429762120910231e-06,
+      "loss": 1.0741,
+      "step": 5663
+    },
+    {
+      "epoch": 0.8687116564417178,
+      "grad_norm": 0.14981243014335632,
+      "learning_rate": 6.414983133054027e-06,
+      "loss": 0.957,
+      "step": 5664
+    },
+    {
+      "epoch": 0.8688650306748467,
+      "grad_norm": 0.14334695041179657,
+      "learning_rate": 6.400220390825409e-06,
+      "loss": 0.8746,
+      "step": 5665
+    },
+    {
+      "epoch": 0.8690184049079754,
+      "grad_norm": 0.16850592195987701,
+      "learning_rate": 6.385473897721219e-06,
+      "loss": 1.0275,
+      "step": 5666
+    },
+    {
+      "epoch": 0.8691717791411043,
+      "grad_norm": 0.15561118721961975,
+      "learning_rate": 6.370743657234415e-06,
+      "loss": 1.1177,
+      "step": 5667
+    },
+    {
+      "epoch": 0.8693251533742331,
+      "grad_norm": 0.16423290967941284,
+      "learning_rate": 6.356029672854146e-06,
+      "loss": 1.0098,
+      "step": 5668
+    },
+    {
+      "epoch": 0.869478527607362,
+      "grad_norm": 0.15144723653793335,
+      "learning_rate": 6.3413319480656925e-06,
+      "loss": 1.0568,
+      "step": 5669
+    },
+    {
+      "epoch": 0.8696319018404908,
+      "grad_norm": 0.16183042526245117,
+      "learning_rate": 6.3266504863504515e-06,
+      "loss": 0.9813,
+      "step": 5670
+    },
+    {
+      "epoch": 0.8697852760736197,
+      "grad_norm": 0.15830768644809723,
+      "learning_rate": 6.311985291186034e-06,
+      "loss": 1.0551,
+      "step": 5671
+    },
+    {
+      "epoch": 0.8699386503067484,
+      "grad_norm": 0.15651634335517883,
+      "learning_rate": 6.297336366046135e-06,
+      "loss": 0.9198,
+      "step": 5672
+    },
+    {
+      "epoch": 0.8700920245398773,
+      "grad_norm": 0.14621621370315552,
+      "learning_rate": 6.282703714400656e-06,
+      "loss": 1.1084,
+      "step": 5673
+    },
+    {
+      "epoch": 0.8702453987730061,
+      "grad_norm": 0.16033896803855896,
+      "learning_rate": 6.268087339715572e-06,
+      "loss": 1.0637,
+      "step": 5674
+    },
+    {
+      "epoch": 0.870398773006135,
+      "grad_norm": 0.15340100228786469,
+      "learning_rate": 6.253487245453065e-06,
+      "loss": 0.9933,
+      "step": 5675
+    },
+    {
+      "epoch": 0.8705521472392638,
+      "grad_norm": 0.15566231310367584,
+      "learning_rate": 6.238903435071443e-06,
+      "loss": 1.0902,
+      "step": 5676
+    },
+    {
+      "epoch": 0.8707055214723927,
+      "grad_norm": 0.1556725800037384,
+      "learning_rate": 6.224335912025119e-06,
+      "loss": 0.9899,
+      "step": 5677
+    },
+    {
+      "epoch": 0.8708588957055214,
+      "grad_norm": 0.16780412197113037,
+      "learning_rate": 6.209784679764715e-06,
+      "loss": 1.1845,
+      "step": 5678
+    },
+    {
+      "epoch": 0.8710122699386503,
+      "grad_norm": 0.15001992881298065,
+      "learning_rate": 6.1952497417369325e-06,
+      "loss": 0.9619,
+      "step": 5679
+    },
+    {
+      "epoch": 0.8711656441717791,
+      "grad_norm": 0.1404801905155182,
+      "learning_rate": 6.180731101384659e-06,
+      "loss": 1.0205,
+      "step": 5680
+    },
+    {
+      "epoch": 0.871319018404908,
+      "grad_norm": 0.15820451080799103,
+      "learning_rate": 6.166228762146885e-06,
+      "loss": 1.1673,
+      "step": 5681
+    },
+    {
+      "epoch": 0.8714723926380368,
+      "grad_norm": 0.15235202014446259,
+      "learning_rate": 6.1517427274587806e-06,
+      "loss": 0.9634,
+      "step": 5682
+    },
+    {
+      "epoch": 0.8716257668711657,
+      "grad_norm": 0.1363549381494522,
+      "learning_rate": 6.137273000751625e-06,
+      "loss": 0.8654,
+      "step": 5683
+    },
+    {
+      "epoch": 0.8717791411042944,
+      "grad_norm": 0.15408770740032196,
+      "learning_rate": 6.12281958545281e-06,
+      "loss": 1.0704,
+      "step": 5684
+    },
+    {
+      "epoch": 0.8719325153374233,
+      "grad_norm": 0.15513281524181366,
+      "learning_rate": 6.10838248498593e-06,
+      "loss": 0.9873,
+      "step": 5685
+    },
+    {
+      "epoch": 0.8720858895705521,
+      "grad_norm": 0.1462486982345581,
+      "learning_rate": 6.09396170277065e-06,
+      "loss": 0.9507,
+      "step": 5686
+    },
+    {
+      "epoch": 0.872239263803681,
+      "grad_norm": 0.15660686790943146,
+      "learning_rate": 6.079557242222838e-06,
+      "loss": 1.1085,
+      "step": 5687
+    },
+    {
+      "epoch": 0.8723926380368098,
+      "grad_norm": 0.14948897063732147,
+      "learning_rate": 6.065169106754406e-06,
+      "loss": 0.8636,
+      "step": 5688
+    },
+    {
+      "epoch": 0.8725460122699387,
+      "grad_norm": 0.15034958720207214,
+      "learning_rate": 6.050797299773494e-06,
+      "loss": 0.9627,
+      "step": 5689
+    },
+    {
+      "epoch": 0.8726993865030674,
+      "grad_norm": 0.1461433619260788,
+      "learning_rate": 6.036441824684304e-06,
+      "loss": 0.8449,
+      "step": 5690
+    },
+    {
+      "epoch": 0.8728527607361963,
+      "grad_norm": 0.15167222917079926,
+      "learning_rate": 6.022102684887209e-06,
+      "loss": 0.9852,
+      "step": 5691
+    },
+    {
+      "epoch": 0.8730061349693251,
+      "grad_norm": 0.1508154571056366,
+      "learning_rate": 6.007779883778682e-06,
+      "loss": 1.0917,
+      "step": 5692
+    },
+    {
+      "epoch": 0.873159509202454,
+      "grad_norm": 0.14213061332702637,
+      "learning_rate": 5.9934734247513526e-06,
+      "loss": 0.9715,
+      "step": 5693
+    },
+    {
+      "epoch": 0.8733128834355828,
+      "grad_norm": 0.16305845975875854,
+      "learning_rate": 5.979183311193983e-06,
+      "loss": 1.1349,
+      "step": 5694
+    },
+    {
+      "epoch": 0.8734662576687117,
+      "grad_norm": 0.15600937604904175,
+      "learning_rate": 5.964909546491442e-06,
+      "loss": 1.0485,
+      "step": 5695
+    },
+    {
+      "epoch": 0.8736196319018404,
+      "grad_norm": 0.1664569228887558,
+      "learning_rate": 5.950652134024708e-06,
+      "loss": 1.1154,
+      "step": 5696
+    },
+    {
+      "epoch": 0.8737730061349693,
+      "grad_norm": 0.162469744682312,
+      "learning_rate": 5.936411077170955e-06,
+      "loss": 1.0988,
+      "step": 5697
+    },
+    {
+      "epoch": 0.8739263803680981,
+      "grad_norm": 0.17145057022571564,
+      "learning_rate": 5.922186379303409e-06,
+      "loss": 1.1491,
+      "step": 5698
+    },
+    {
+      "epoch": 0.874079754601227,
+      "grad_norm": 0.16432908177375793,
+      "learning_rate": 5.907978043791475e-06,
+      "loss": 1.0766,
+      "step": 5699
+    },
+    {
+      "epoch": 0.8742331288343558,
+      "grad_norm": 0.16513079404830933,
+      "learning_rate": 5.8937860740006364e-06,
+      "loss": 1.1198,
+      "step": 5700
+    },
+    {
+      "epoch": 0.8743865030674847,
+      "grad_norm": 0.15564443171024323,
+      "learning_rate": 5.8796104732925455e-06,
+      "loss": 1.0197,
+      "step": 5701
+    },
+    {
+      "epoch": 0.8745398773006134,
+      "grad_norm": 0.1358862817287445,
+      "learning_rate": 5.865451245024941e-06,
+      "loss": 0.8408,
+      "step": 5702
+    },
+    {
+      "epoch": 0.8746932515337423,
+      "grad_norm": 0.16488254070281982,
+      "learning_rate": 5.851308392551698e-06,
+      "loss": 1.0556,
+      "step": 5703
+    },
+    {
+      "epoch": 0.8748466257668711,
+      "grad_norm": 0.1505875140428543,
+      "learning_rate": 5.837181919222831e-06,
+      "loss": 1.0722,
+      "step": 5704
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 0.1623571664094925,
+      "learning_rate": 5.823071828384418e-06,
+      "loss": 0.9879,
+      "step": 5705
+    },
+    {
+      "epoch": 0.8751533742331289,
+      "grad_norm": 0.1633947342634201,
+      "learning_rate": 5.80897812337873e-06,
+      "loss": 1.0784,
+      "step": 5706
+    },
+    {
+      "epoch": 0.8753067484662577,
+      "grad_norm": 0.1618642508983612,
+      "learning_rate": 5.794900807544104e-06,
+      "loss": 1.0561,
+      "step": 5707
+    },
+    {
+      "epoch": 0.8754601226993866,
+      "grad_norm": 0.16289368271827698,
+      "learning_rate": 5.780839884215022e-06,
+      "loss": 1.1208,
+      "step": 5708
+    },
+    {
+      "epoch": 0.8756134969325153,
+      "grad_norm": 0.1565026044845581,
+      "learning_rate": 5.766795356722053e-06,
+      "loss": 1.0547,
+      "step": 5709
+    },
+    {
+      "epoch": 0.8757668711656442,
+      "grad_norm": 0.16009585559368134,
+      "learning_rate": 5.752767228391939e-06,
+      "loss": 1.1596,
+      "step": 5710
+    },
+    {
+      "epoch": 0.875920245398773,
+      "grad_norm": 0.16512751579284668,
+      "learning_rate": 5.738755502547468e-06,
+      "loss": 1.0945,
+      "step": 5711
+    },
+    {
+      "epoch": 0.8760736196319019,
+      "grad_norm": 0.15330770611763,
+      "learning_rate": 5.7247601825075875e-06,
+      "loss": 0.9821,
+      "step": 5712
+    },
+    {
+      "epoch": 0.8762269938650307,
+      "grad_norm": 0.15823820233345032,
+      "learning_rate": 5.710781271587353e-06,
+      "loss": 1.0677,
+      "step": 5713
+    },
+    {
+      "epoch": 0.8763803680981596,
+      "grad_norm": 0.16428913176059723,
+      "learning_rate": 5.696818773097908e-06,
+      "loss": 0.9931,
+      "step": 5714
+    },
+    {
+      "epoch": 0.8765337423312883,
+      "grad_norm": 0.13639238476753235,
+      "learning_rate": 5.682872690346557e-06,
+      "loss": 0.8857,
+      "step": 5715
+    },
+    {
+      "epoch": 0.8766871165644172,
+      "grad_norm": 0.1518758237361908,
+      "learning_rate": 5.668943026636666e-06,
+      "loss": 0.9901,
+      "step": 5716
+    },
+    {
+      "epoch": 0.876840490797546,
+      "grad_norm": 0.1445581167936325,
+      "learning_rate": 5.6550297852677454e-06,
+      "loss": 0.8888,
+      "step": 5717
+    },
+    {
+      "epoch": 0.8769938650306749,
+      "grad_norm": 0.15615792572498322,
+      "learning_rate": 5.6411329695354e-06,
+      "loss": 1.0564,
+      "step": 5718
+    },
+    {
+      "epoch": 0.8771472392638037,
+      "grad_norm": 0.15769603848457336,
+      "learning_rate": 5.627252582731331e-06,
+      "loss": 0.9592,
+      "step": 5719
+    },
+    {
+      "epoch": 0.8773006134969326,
+      "grad_norm": 0.1581096202135086,
+      "learning_rate": 5.613388628143383e-06,
+      "loss": 1.0643,
+      "step": 5720
+    },
+    {
+      "epoch": 0.8774539877300613,
+      "grad_norm": 0.16767553985118866,
+      "learning_rate": 5.59954110905548e-06,
+      "loss": 1.2048,
+      "step": 5721
+    },
+    {
+      "epoch": 0.8776073619631902,
+      "grad_norm": 0.7973523139953613,
+      "learning_rate": 5.585710028747681e-06,
+      "loss": 1.0782,
+      "step": 5722
+    },
+    {
+      "epoch": 0.877760736196319,
+      "grad_norm": 0.16758692264556885,
+      "learning_rate": 5.571895390496106e-06,
+      "loss": 1.0519,
+      "step": 5723
+    },
+    {
+      "epoch": 0.8779141104294479,
+      "grad_norm": 0.17085489630699158,
+      "learning_rate": 5.558097197573031e-06,
+      "loss": 1.1559,
+      "step": 5724
+    },
+    {
+      "epoch": 0.8780674846625767,
+      "grad_norm": 0.17001433670520782,
+      "learning_rate": 5.5443154532467895e-06,
+      "loss": 1.1367,
+      "step": 5725
+    },
+    {
+      "epoch": 0.8782208588957056,
+      "grad_norm": 0.16413988173007965,
+      "learning_rate": 5.530550160781871e-06,
+      "loss": 1.0246,
+      "step": 5726
+    },
+    {
+      "epoch": 0.8783742331288343,
+      "grad_norm": 0.15367798507213593,
+      "learning_rate": 5.516801323438819e-06,
+      "loss": 0.97,
+      "step": 5727
+    },
+    {
+      "epoch": 0.8785276073619632,
+      "grad_norm": 0.1664074808359146,
+      "learning_rate": 5.503068944474284e-06,
+      "loss": 1.1553,
+      "step": 5728
+    },
+    {
+      "epoch": 0.878680981595092,
+      "grad_norm": 0.17225711047649384,
+      "learning_rate": 5.4893530271410745e-06,
+      "loss": 1.0901,
+      "step": 5729
+    },
+    {
+      "epoch": 0.8788343558282209,
+      "grad_norm": 0.16833461821079254,
+      "learning_rate": 5.475653574688024e-06,
+      "loss": 1.0858,
+      "step": 5730
+    },
+    {
+      "epoch": 0.8789877300613497,
+      "grad_norm": 0.155145525932312,
+      "learning_rate": 5.461970590360121e-06,
+      "loss": 1.0006,
+      "step": 5731
+    },
+    {
+      "epoch": 0.8791411042944786,
+      "grad_norm": 0.1464105099439621,
+      "learning_rate": 5.448304077398413e-06,
+      "loss": 0.9649,
+      "step": 5732
+    },
+    {
+      "epoch": 0.8792944785276073,
+      "grad_norm": 0.14439015090465546,
+      "learning_rate": 5.434654039040096e-06,
+      "loss": 0.9465,
+      "step": 5733
+    },
+    {
+      "epoch": 0.8794478527607362,
+      "grad_norm": 0.14978152513504028,
+      "learning_rate": 5.421020478518415e-06,
+      "loss": 0.9886,
+      "step": 5734
+    },
+    {
+      "epoch": 0.879601226993865,
+      "grad_norm": 0.1571166068315506,
+      "learning_rate": 5.407403399062704e-06,
+      "loss": 1.025,
+      "step": 5735
+    },
+    {
+      "epoch": 0.8797546012269939,
+      "grad_norm": 0.15805941820144653,
+      "learning_rate": 5.393802803898476e-06,
+      "loss": 1.0725,
+      "step": 5736
+    },
+    {
+      "epoch": 0.8799079754601227,
+      "grad_norm": 0.1546279937028885,
+      "learning_rate": 5.380218696247244e-06,
+      "loss": 1.0364,
+      "step": 5737
+    },
+    {
+      "epoch": 0.8800613496932516,
+      "grad_norm": 0.1745101362466812,
+      "learning_rate": 5.3666510793266675e-06,
+      "loss": 1.1193,
+      "step": 5738
+    },
+    {
+      "epoch": 0.8802147239263803,
+      "grad_norm": 0.16241243481636047,
+      "learning_rate": 5.353099956350493e-06,
+      "loss": 1.1098,
+      "step": 5739
+    },
+    {
+      "epoch": 0.8803680981595092,
+      "grad_norm": 0.15269169211387634,
+      "learning_rate": 5.339565330528528e-06,
+      "loss": 1.0289,
+      "step": 5740
+    },
+    {
+      "epoch": 0.880521472392638,
+      "grad_norm": 0.16068150103092194,
+      "learning_rate": 5.326047205066733e-06,
+      "loss": 1.0483,
+      "step": 5741
+    },
+    {
+      "epoch": 0.8806748466257669,
+      "grad_norm": 0.1566779464483261,
+      "learning_rate": 5.312545583167097e-06,
+      "loss": 1.0589,
+      "step": 5742
+    },
+    {
+      "epoch": 0.8808282208588957,
+      "grad_norm": 0.15958382189273834,
+      "learning_rate": 5.299060468027752e-06,
+      "loss": 1.0622,
+      "step": 5743
+    },
+    {
+      "epoch": 0.8809815950920246,
+      "grad_norm": 0.1557781994342804,
+      "learning_rate": 5.285591862842878e-06,
+      "loss": 1.0503,
+      "step": 5744
+    },
+    {
+      "epoch": 0.8811349693251533,
+      "grad_norm": 0.14341381192207336,
+      "learning_rate": 5.272139770802772e-06,
+      "loss": 0.9614,
+      "step": 5745
+    },
+    {
+      "epoch": 0.8812883435582822,
+      "grad_norm": 0.15170325338840485,
+      "learning_rate": 5.2587041950938184e-06,
+      "loss": 1.0991,
+      "step": 5746
+    },
+    {
+      "epoch": 0.881441717791411,
+      "grad_norm": 0.161540225148201,
+      "learning_rate": 5.2452851388984475e-06,
+      "loss": 1.0669,
+      "step": 5747
+    },
+    {
+      "epoch": 0.8815950920245399,
+      "grad_norm": 0.1541643887758255,
+      "learning_rate": 5.231882605395257e-06,
+      "loss": 1.0886,
+      "step": 5748
+    },
+    {
+      "epoch": 0.8817484662576687,
+      "grad_norm": 0.1515563577413559,
+      "learning_rate": 5.218496597758842e-06,
+      "loss": 1.0312,
+      "step": 5749
+    },
+    {
+      "epoch": 0.8819018404907976,
+      "grad_norm": 0.15532483160495758,
+      "learning_rate": 5.205127119159966e-06,
+      "loss": 1.0314,
+      "step": 5750
+    },
+    {
+      "epoch": 0.8820552147239263,
+      "grad_norm": 0.16019129753112793,
+      "learning_rate": 5.191774172765395e-06,
+      "loss": 1.0723,
+      "step": 5751
+    },
+    {
+      "epoch": 0.8822085889570552,
+      "grad_norm": 0.1569075584411621,
+      "learning_rate": 5.178437761738058e-06,
+      "loss": 1.0574,
+      "step": 5752
+    },
+    {
+      "epoch": 0.882361963190184,
+      "grad_norm": 0.15603147447109222,
+      "learning_rate": 5.165117889236911e-06,
+      "loss": 1.0456,
+      "step": 5753
+    },
+    {
+      "epoch": 0.8825153374233129,
+      "grad_norm": 0.17533469200134277,
+      "learning_rate": 5.151814558417005e-06,
+      "loss": 1.0988,
+      "step": 5754
+    },
+    {
+      "epoch": 0.8826687116564417,
+      "grad_norm": 0.1591043621301651,
+      "learning_rate": 5.1385277724295045e-06,
+      "loss": 1.0531,
+      "step": 5755
+    },
+    {
+      "epoch": 0.8828220858895706,
+      "grad_norm": 0.15566477179527283,
+      "learning_rate": 5.125257534421598e-06,
+      "loss": 0.9488,
+      "step": 5756
+    },
+    {
+      "epoch": 0.8829754601226993,
+      "grad_norm": 0.1703893095254898,
+      "learning_rate": 5.112003847536622e-06,
+      "loss": 1.0458,
+      "step": 5757
+    },
+    {
+      "epoch": 0.8831288343558282,
+      "grad_norm": 0.1579372137784958,
+      "learning_rate": 5.0987667149139235e-06,
+      "loss": 1.0637,
+      "step": 5758
+    },
+    {
+      "epoch": 0.883282208588957,
+      "grad_norm": 0.15190039575099945,
+      "learning_rate": 5.085546139688987e-06,
+      "loss": 1.0566,
+      "step": 5759
+    },
+    {
+      "epoch": 0.8834355828220859,
+      "grad_norm": 0.16215085983276367,
+      "learning_rate": 5.072342124993331e-06,
+      "loss": 1.0758,
+      "step": 5760
+    },
+    {
+      "epoch": 0.8835889570552147,
+      "grad_norm": 0.15225616097450256,
+      "learning_rate": 5.059154673954588e-06,
+      "loss": 1.1094,
+      "step": 5761
+    },
+    {
+      "epoch": 0.8837423312883436,
+      "grad_norm": 0.1529587060213089,
+      "learning_rate": 5.045983789696434e-06,
+      "loss": 1.0643,
+      "step": 5762
+    },
+    {
+      "epoch": 0.8838957055214723,
+      "grad_norm": 0.14848493039608002,
+      "learning_rate": 5.03282947533864e-06,
+      "loss": 1.0722,
+      "step": 5763
+    },
+    {
+      "epoch": 0.8840490797546012,
+      "grad_norm": 0.16094045341014862,
+      "learning_rate": 5.019691733997055e-06,
+      "loss": 1.0302,
+      "step": 5764
+    },
+    {
+      "epoch": 0.88420245398773,
+      "grad_norm": 0.14938275516033173,
+      "learning_rate": 5.006570568783583e-06,
+      "loss": 1.0623,
+      "step": 5765
+    },
+    {
+      "epoch": 0.8843558282208589,
+      "grad_norm": 0.14906343817710876,
+      "learning_rate": 4.993465982806236e-06,
+      "loss": 0.9351,
+      "step": 5766
+    },
+    {
+      "epoch": 0.8845092024539877,
+      "grad_norm": 0.16712233424186707,
+      "learning_rate": 4.980377979169048e-06,
+      "loss": 1.1797,
+      "step": 5767
+    },
+    {
+      "epoch": 0.8846625766871166,
+      "grad_norm": 0.1587630659341812,
+      "learning_rate": 4.967306560972181e-06,
+      "loss": 1.1218,
+      "step": 5768
+    },
+    {
+      "epoch": 0.8848159509202455,
+      "grad_norm": 0.1588369607925415,
+      "learning_rate": 4.954251731311834e-06,
+      "loss": 1.0354,
+      "step": 5769
+    },
+    {
+      "epoch": 0.8849693251533742,
+      "grad_norm": 0.15524156391620636,
+      "learning_rate": 4.941213493280266e-06,
+      "loss": 1.0246,
+      "step": 5770
+    },
+    {
+      "epoch": 0.8851226993865031,
+      "grad_norm": 0.14653874933719635,
+      "learning_rate": 4.9281918499658565e-06,
+      "loss": 0.9317,
+      "step": 5771
+    },
+    {
+      "epoch": 0.8852760736196319,
+      "grad_norm": 0.1676556020975113,
+      "learning_rate": 4.915186804452989e-06,
+      "loss": 1.1347,
+      "step": 5772
+    },
+    {
+      "epoch": 0.8854294478527608,
+      "grad_norm": 0.16220028698444366,
+      "learning_rate": 4.902198359822173e-06,
+      "loss": 1.0661,
+      "step": 5773
+    },
+    {
+      "epoch": 0.8855828220858896,
+      "grad_norm": 0.16912487149238586,
+      "learning_rate": 4.889226519149942e-06,
+      "loss": 1.0859,
+      "step": 5774
+    },
+    {
+      "epoch": 0.8857361963190185,
+      "grad_norm": 0.1584475189447403,
+      "learning_rate": 4.8762712855089424e-06,
+      "loss": 1.095,
+      "step": 5775
+    },
+    {
+      "epoch": 0.8858895705521472,
+      "grad_norm": 0.15351292490959167,
+      "learning_rate": 4.863332661967847e-06,
+      "loss": 0.9522,
+      "step": 5776
+    },
+    {
+      "epoch": 0.8860429447852761,
+      "grad_norm": 0.15598662197589874,
+      "learning_rate": 4.85041065159138e-06,
+      "loss": 0.9009,
+      "step": 5777
+    },
+    {
+      "epoch": 0.8861963190184049,
+      "grad_norm": 0.1495562344789505,
+      "learning_rate": 4.837505257440399e-06,
+      "loss": 1.0732,
+      "step": 5778
+    },
+    {
+      "epoch": 0.8863496932515338,
+      "grad_norm": 0.15162207186222076,
+      "learning_rate": 4.824616482571769e-06,
+      "loss": 0.9774,
+      "step": 5779
+    },
+    {
+      "epoch": 0.8865030674846626,
+      "grad_norm": 0.15782980620861053,
+      "learning_rate": 4.811744330038447e-06,
+      "loss": 1.0007,
+      "step": 5780
+    },
+    {
+      "epoch": 0.8866564417177915,
+      "grad_norm": 0.15851199626922607,
+      "learning_rate": 4.79888880288942e-06,
+      "loss": 1.0535,
+      "step": 5781
+    },
+    {
+      "epoch": 0.8868098159509202,
+      "grad_norm": 0.16146107017993927,
+      "learning_rate": 4.786049904169758e-06,
+      "loss": 1.0891,
+      "step": 5782
+    },
+    {
+      "epoch": 0.8869631901840491,
+      "grad_norm": 0.15411558747291565,
+      "learning_rate": 4.773227636920606e-06,
+      "loss": 1.0889,
+      "step": 5783
+    },
+    {
+      "epoch": 0.8871165644171779,
+      "grad_norm": 0.15200673043727875,
+      "learning_rate": 4.760422004179143e-06,
+      "loss": 1.0334,
+      "step": 5784
+    },
+    {
+      "epoch": 0.8872699386503068,
+      "grad_norm": 0.1477597951889038,
+      "learning_rate": 4.747633008978633e-06,
+      "loss": 1.0307,
+      "step": 5785
+    },
+    {
+      "epoch": 0.8874233128834356,
+      "grad_norm": 0.14918993413448334,
+      "learning_rate": 4.73486065434836e-06,
+      "loss": 0.9772,
+      "step": 5786
+    },
+    {
+      "epoch": 0.8875766871165645,
+      "grad_norm": 0.13921032845973969,
+      "learning_rate": 4.722104943313723e-06,
+      "loss": 0.8773,
+      "step": 5787
+    },
+    {
+      "epoch": 0.8877300613496932,
+      "grad_norm": 0.1427036076784134,
+      "learning_rate": 4.709365878896129e-06,
+      "loss": 0.8596,
+      "step": 5788
+    },
+    {
+      "epoch": 0.8878834355828221,
+      "grad_norm": 0.15141703188419342,
+      "learning_rate": 4.696643464113065e-06,
+      "loss": 0.9604,
+      "step": 5789
+    },
+    {
+      "epoch": 0.8880368098159509,
+      "grad_norm": 0.1629810333251953,
+      "learning_rate": 4.6839377019780685e-06,
+      "loss": 0.9963,
+      "step": 5790
+    },
+    {
+      "epoch": 0.8881901840490798,
+      "grad_norm": 0.14559035003185272,
+      "learning_rate": 4.6712485955007265e-06,
+      "loss": 1.0083,
+      "step": 5791
+    },
+    {
+      "epoch": 0.8883435582822086,
+      "grad_norm": 0.14738984405994415,
+      "learning_rate": 4.658576147686707e-06,
+      "loss": 0.9518,
+      "step": 5792
+    },
+    {
+      "epoch": 0.8884969325153375,
+      "grad_norm": 0.1554996222257614,
+      "learning_rate": 4.645920361537694e-06,
+      "loss": 0.9658,
+      "step": 5793
+    },
+    {
+      "epoch": 0.8886503067484662,
+      "grad_norm": 0.13928905129432678,
+      "learning_rate": 4.633281240051465e-06,
+      "loss": 0.9639,
+      "step": 5794
+    },
+    {
+      "epoch": 0.8888036809815951,
+      "grad_norm": 0.15555289387702942,
+      "learning_rate": 4.620658786221798e-06,
+      "loss": 1.1447,
+      "step": 5795
+    },
+    {
+      "epoch": 0.8889570552147239,
+      "grad_norm": 0.1626276820898056,
+      "learning_rate": 4.608053003038595e-06,
+      "loss": 1.0469,
+      "step": 5796
+    },
+    {
+      "epoch": 0.8891104294478528,
+      "grad_norm": 0.15400700271129608,
+      "learning_rate": 4.5954638934877416e-06,
+      "loss": 1.0704,
+      "step": 5797
+    },
+    {
+      "epoch": 0.8892638036809816,
+      "grad_norm": 0.14584322273731232,
+      "learning_rate": 4.582891460551194e-06,
+      "loss": 0.9734,
+      "step": 5798
+    },
+    {
+      "epoch": 0.8894171779141105,
+      "grad_norm": 0.15082399547100067,
+      "learning_rate": 4.570335707206987e-06,
+      "loss": 0.9688,
+      "step": 5799
+    },
+    {
+      "epoch": 0.8895705521472392,
+      "grad_norm": 0.15909363329410553,
+      "learning_rate": 4.557796636429168e-06,
+      "loss": 1.1012,
+      "step": 5800
+    },
+    {
+      "epoch": 0.8897239263803681,
+      "grad_norm": 0.1538352519273758,
+      "learning_rate": 4.5452742511878586e-06,
+      "loss": 1.0706,
+      "step": 5801
+    },
+    {
+      "epoch": 0.8898773006134969,
+      "grad_norm": 0.153822124004364,
+      "learning_rate": 4.532768554449204e-06,
+      "loss": 1.0448,
+      "step": 5802
+    },
+    {
+      "epoch": 0.8900306748466258,
+      "grad_norm": 0.15416836738586426,
+      "learning_rate": 4.520279549175418e-06,
+      "loss": 1.0258,
+      "step": 5803
+    },
+    {
+      "epoch": 0.8901840490797546,
+      "grad_norm": 0.16300542652606964,
+      "learning_rate": 4.50780723832476e-06,
+      "loss": 1.1219,
+      "step": 5804
+    },
+    {
+      "epoch": 0.8903374233128835,
+      "grad_norm": 0.14935365319252014,
+      "learning_rate": 4.4953516248514985e-06,
+      "loss": 1.0,
+      "step": 5805
+    },
+    {
+      "epoch": 0.8904907975460122,
+      "grad_norm": 0.15455570816993713,
+      "learning_rate": 4.482912711706008e-06,
+      "loss": 0.9948,
+      "step": 5806
+    },
+    {
+      "epoch": 0.8906441717791411,
+      "grad_norm": 0.14951899647712708,
+      "learning_rate": 4.470490501834648e-06,
+      "loss": 1.033,
+      "step": 5807
+    },
+    {
+      "epoch": 0.8907975460122699,
+      "grad_norm": 0.15210136771202087,
+      "learning_rate": 4.458084998179873e-06,
+      "loss": 1.105,
+      "step": 5808
+    },
+    {
+      "epoch": 0.8909509202453988,
+      "grad_norm": 0.15395419299602509,
+      "learning_rate": 4.445696203680141e-06,
+      "loss": 1.03,
+      "step": 5809
+    },
+    {
+      "epoch": 0.8911042944785276,
+      "grad_norm": 0.1721864938735962,
+      "learning_rate": 4.433324121269971e-06,
+      "loss": 1.2383,
+      "step": 5810
+    },
+    {
+      "epoch": 0.8912576687116565,
+      "grad_norm": 0.1492663323879242,
+      "learning_rate": 4.420968753879919e-06,
+      "loss": 0.9525,
+      "step": 5811
+    },
+    {
+      "epoch": 0.8914110429447852,
+      "grad_norm": 0.14481961727142334,
+      "learning_rate": 4.40863010443657e-06,
+      "loss": 0.8686,
+      "step": 5812
+    },
+    {
+      "epoch": 0.8915644171779141,
+      "grad_norm": 0.16896508634090424,
+      "learning_rate": 4.396308175862584e-06,
+      "loss": 1.253,
+      "step": 5813
+    },
+    {
+      "epoch": 0.8917177914110429,
+      "grad_norm": 0.15678295493125916,
+      "learning_rate": 4.384002971076611e-06,
+      "loss": 1.119,
+      "step": 5814
+    },
+    {
+      "epoch": 0.8918711656441718,
+      "grad_norm": 0.15469986200332642,
+      "learning_rate": 4.371714492993395e-06,
+      "loss": 1.0475,
+      "step": 5815
+    },
+    {
+      "epoch": 0.8920245398773006,
+      "grad_norm": 0.16731977462768555,
+      "learning_rate": 4.3594427445236554e-06,
+      "loss": 1.2292,
+      "step": 5816
+    },
+    {
+      "epoch": 0.8921779141104295,
+      "grad_norm": 0.16725195944309235,
+      "learning_rate": 4.34718772857422e-06,
+      "loss": 1.1589,
+      "step": 5817
+    },
+    {
+      "epoch": 0.8923312883435582,
+      "grad_norm": 0.16074301302433014,
+      "learning_rate": 4.334949448047881e-06,
+      "loss": 1.1353,
+      "step": 5818
+    },
+    {
+      "epoch": 0.8924846625766871,
+      "grad_norm": 0.14353086054325104,
+      "learning_rate": 4.322727905843529e-06,
+      "loss": 0.9318,
+      "step": 5819
+    },
+    {
+      "epoch": 0.8926380368098159,
+      "grad_norm": 0.1503216028213501,
+      "learning_rate": 4.31052310485604e-06,
+      "loss": 0.9801,
+      "step": 5820
+    },
+    {
+      "epoch": 0.8927914110429448,
+      "grad_norm": 0.15206632018089294,
+      "learning_rate": 4.298335047976351e-06,
+      "loss": 1.086,
+      "step": 5821
+    },
+    {
+      "epoch": 0.8929447852760736,
+      "grad_norm": 0.15070699155330658,
+      "learning_rate": 4.286163738091444e-06,
+      "loss": 0.9361,
+      "step": 5822
+    },
+    {
+      "epoch": 0.8930981595092025,
+      "grad_norm": 0.1715269535779953,
+      "learning_rate": 4.274009178084315e-06,
+      "loss": 1.1858,
+      "step": 5823
+    },
+    {
+      "epoch": 0.8932515337423312,
+      "grad_norm": 0.1521557718515396,
+      "learning_rate": 4.2618713708339665e-06,
+      "loss": 0.9443,
+      "step": 5824
+    },
+    {
+      "epoch": 0.8934049079754601,
+      "grad_norm": 0.16819889843463898,
+      "learning_rate": 4.2497503192155e-06,
+      "loss": 1.1089,
+      "step": 5825
+    },
+    {
+      "epoch": 0.8935582822085889,
+      "grad_norm": 0.16168436408042908,
+      "learning_rate": 4.237646026099986e-06,
+      "loss": 1.0841,
+      "step": 5826
+    },
+    {
+      "epoch": 0.8937116564417178,
+      "grad_norm": 0.1455746442079544,
+      "learning_rate": 4.2255584943545725e-06,
+      "loss": 0.9565,
+      "step": 5827
+    },
+    {
+      "epoch": 0.8938650306748466,
+      "grad_norm": 0.16026215255260468,
+      "learning_rate": 4.213487726842385e-06,
+      "loss": 1.0176,
+      "step": 5828
+    },
+    {
+      "epoch": 0.8940184049079755,
+      "grad_norm": 0.15701237320899963,
+      "learning_rate": 4.201433726422629e-06,
+      "loss": 1.0712,
+      "step": 5829
+    },
+    {
+      "epoch": 0.8941717791411042,
+      "grad_norm": 0.1638382077217102,
+      "learning_rate": 4.189396495950504e-06,
+      "loss": 1.1331,
+      "step": 5830
+    },
+    {
+      "epoch": 0.8943251533742331,
+      "grad_norm": 0.1494309902191162,
+      "learning_rate": 4.177376038277269e-06,
+      "loss": 1.0727,
+      "step": 5831
+    },
+    {
+      "epoch": 0.894478527607362,
+      "grad_norm": 0.15829096734523773,
+      "learning_rate": 4.165372356250182e-06,
+      "loss": 1.0153,
+      "step": 5832
+    },
+    {
+      "epoch": 0.8946319018404908,
+      "grad_norm": 0.1568623036146164,
+      "learning_rate": 4.153385452712527e-06,
+      "loss": 0.9755,
+      "step": 5833
+    },
+    {
+      "epoch": 0.8947852760736197,
+      "grad_norm": 0.14713525772094727,
+      "learning_rate": 4.141415330503639e-06,
+      "loss": 1.0057,
+      "step": 5834
+    },
+    {
+      "epoch": 0.8949386503067485,
+      "grad_norm": 0.15302518010139465,
+      "learning_rate": 4.129461992458835e-06,
+      "loss": 1.0619,
+      "step": 5835
+    },
+    {
+      "epoch": 0.8950920245398774,
+      "grad_norm": 0.1585974395275116,
+      "learning_rate": 4.117525441409525e-06,
+      "loss": 0.9611,
+      "step": 5836
+    },
+    {
+      "epoch": 0.8952453987730061,
+      "grad_norm": 0.15477091073989868,
+      "learning_rate": 4.105605680183064e-06,
+      "loss": 1.0025,
+      "step": 5837
+    },
+    {
+      "epoch": 0.895398773006135,
+      "grad_norm": 0.1568540632724762,
+      "learning_rate": 4.0937027116029e-06,
+      "loss": 1.0473,
+      "step": 5838
+    },
+    {
+      "epoch": 0.8955521472392638,
+      "grad_norm": 0.16229557991027832,
+      "learning_rate": 4.081816538488447e-06,
+      "loss": 1.05,
+      "step": 5839
+    },
+    {
+      "epoch": 0.8957055214723927,
+      "grad_norm": 0.14899510145187378,
+      "learning_rate": 4.06994716365516e-06,
+      "loss": 0.9567,
+      "step": 5840
+    },
+    {
+      "epoch": 0.8958588957055215,
+      "grad_norm": 0.1565985083580017,
+      "learning_rate": 4.058094589914548e-06,
+      "loss": 1.1574,
+      "step": 5841
+    },
+    {
+      "epoch": 0.8960122699386504,
+      "grad_norm": 0.14344468712806702,
+      "learning_rate": 4.046258820074083e-06,
+      "loss": 0.9636,
+      "step": 5842
+    },
+    {
+      "epoch": 0.8961656441717791,
+      "grad_norm": 0.14630700647830963,
+      "learning_rate": 4.0344398569373136e-06,
+      "loss": 0.9156,
+      "step": 5843
+    },
+    {
+      "epoch": 0.896319018404908,
+      "grad_norm": 0.1487908810377121,
+      "learning_rate": 4.022637703303741e-06,
+      "loss": 0.9635,
+      "step": 5844
+    },
+    {
+      "epoch": 0.8964723926380368,
+      "grad_norm": 0.3965422809123993,
+      "learning_rate": 4.010852361968964e-06,
+      "loss": 1.0041,
+      "step": 5845
+    },
+    {
+      "epoch": 0.8966257668711657,
+      "grad_norm": 0.1606234759092331,
+      "learning_rate": 3.99908383572454e-06,
+      "loss": 1.0103,
+      "step": 5846
+    },
+    {
+      "epoch": 0.8967791411042945,
+      "grad_norm": 0.15086081624031067,
+      "learning_rate": 3.987332127358056e-06,
+      "loss": 0.973,
+      "step": 5847
+    },
+    {
+      "epoch": 0.8969325153374234,
+      "grad_norm": 0.15696370601654053,
+      "learning_rate": 3.975597239653128e-06,
+      "loss": 1.007,
+      "step": 5848
+    },
+    {
+      "epoch": 0.8970858895705521,
+      "grad_norm": 0.15061095356941223,
+      "learning_rate": 3.963879175389373e-06,
+      "loss": 0.9255,
+      "step": 5849
+    },
+    {
+      "epoch": 0.897239263803681,
+      "grad_norm": 0.15637679398059845,
+      "learning_rate": 3.9521779373424545e-06,
+      "loss": 1.1085,
+      "step": 5850
+    },
+    {
+      "epoch": 0.8973926380368098,
+      "grad_norm": 0.17591843008995056,
+      "learning_rate": 3.940493528283989e-06,
+      "loss": 1.0427,
+      "step": 5851
+    },
+    {
+      "epoch": 0.8975460122699387,
+      "grad_norm": 0.14846207201480865,
+      "learning_rate": 3.928825950981687e-06,
+      "loss": 1.0188,
+      "step": 5852
+    },
+    {
+      "epoch": 0.8976993865030675,
+      "grad_norm": 0.17448090016841888,
+      "learning_rate": 3.9171752081991885e-06,
+      "loss": 1.1658,
+      "step": 5853
+    },
+    {
+      "epoch": 0.8978527607361964,
+      "grad_norm": 0.16077740490436554,
+      "learning_rate": 3.9055413026962274e-06,
+      "loss": 1.1646,
+      "step": 5854
+    },
+    {
+      "epoch": 0.8980061349693251,
+      "grad_norm": 0.15605364739894867,
+      "learning_rate": 3.893924237228483e-06,
+      "loss": 1.0694,
+      "step": 5855
+    },
+    {
+      "epoch": 0.898159509202454,
+      "grad_norm": 0.14850561320781708,
+      "learning_rate": 3.882324014547661e-06,
+      "loss": 1.0302,
+      "step": 5856
+    },
+    {
+      "epoch": 0.8983128834355828,
+      "grad_norm": 0.1555531620979309,
+      "learning_rate": 3.870740637401523e-06,
+      "loss": 1.1157,
+      "step": 5857
+    },
+    {
+      "epoch": 0.8984662576687117,
+      "grad_norm": 0.15149959921836853,
+      "learning_rate": 3.859174108533775e-06,
+      "loss": 1.0786,
+      "step": 5858
+    },
+    {
+      "epoch": 0.8986196319018405,
+      "grad_norm": 0.15650232136249542,
+      "learning_rate": 3.847624430684182e-06,
+      "loss": 1.1115,
+      "step": 5859
+    },
+    {
+      "epoch": 0.8987730061349694,
+      "grad_norm": 0.1665438413619995,
+      "learning_rate": 3.836091606588482e-06,
+      "loss": 1.1733,
+      "step": 5860
+    },
+    {
+      "epoch": 0.8989263803680981,
+      "grad_norm": 0.14378252625465393,
+      "learning_rate": 3.824575638978463e-06,
+      "loss": 0.8902,
+      "step": 5861
+    },
+    {
+      "epoch": 0.899079754601227,
+      "grad_norm": 0.1499444991350174,
+      "learning_rate": 3.8130765305818627e-06,
+      "loss": 1.025,
+      "step": 5862
+    },
+    {
+      "epoch": 0.8992331288343558,
+      "grad_norm": 0.14060218632221222,
+      "learning_rate": 3.8015942841224752e-06,
+      "loss": 0.9097,
+      "step": 5863
+    },
+    {
+      "epoch": 0.8993865030674847,
+      "grad_norm": 0.1635594367980957,
+      "learning_rate": 3.7901289023200916e-06,
+      "loss": 0.9023,
+      "step": 5864
+    },
+    {
+      "epoch": 0.8995398773006135,
+      "grad_norm": 0.1538282334804535,
+      "learning_rate": 3.778680387890473e-06,
+      "loss": 1.0476,
+      "step": 5865
+    },
+    {
+      "epoch": 0.8996932515337424,
+      "grad_norm": 0.1815045326948166,
+      "learning_rate": 3.7672487435454422e-06,
+      "loss": 1.0681,
+      "step": 5866
+    },
+    {
+      "epoch": 0.8998466257668711,
+      "grad_norm": 0.15962480008602142,
+      "learning_rate": 3.755833971992775e-06,
+      "loss": 1.0002,
+      "step": 5867
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.15540069341659546,
+      "learning_rate": 3.7444360759362664e-06,
+      "loss": 1.0349,
+      "step": 5868
+    },
+    {
+      "epoch": 0.9001533742331288,
+      "grad_norm": 0.1594664305448532,
+      "learning_rate": 3.73305505807574e-06,
+      "loss": 1.0915,
+      "step": 5869
+    },
+    {
+      "epoch": 0.9003067484662577,
+      "grad_norm": 0.15836378931999207,
+      "learning_rate": 3.721690921106982e-06,
+      "loss": 1.0358,
+      "step": 5870
+    },
+    {
+      "epoch": 0.9004601226993865,
+      "grad_norm": 0.14288723468780518,
+      "learning_rate": 3.7103436677218046e-06,
+      "loss": 0.9651,
+      "step": 5871
+    },
+    {
+      "epoch": 0.9006134969325154,
+      "grad_norm": 0.16262847185134888,
+      "learning_rate": 3.6990133006080088e-06,
+      "loss": 1.0831,
+      "step": 5872
+    },
+    {
+      "epoch": 0.9007668711656441,
+      "grad_norm": 0.14909861981868744,
+      "learning_rate": 3.6876998224494135e-06,
+      "loss": 0.9655,
+      "step": 5873
+    },
+    {
+      "epoch": 0.900920245398773,
+      "grad_norm": 0.1461794376373291,
+      "learning_rate": 3.6764032359258177e-06,
+      "loss": 0.9567,
+      "step": 5874
+    },
+    {
+      "epoch": 0.9010736196319018,
+      "grad_norm": 0.1529189497232437,
+      "learning_rate": 3.665123543713014e-06,
+      "loss": 1.1142,
+      "step": 5875
+    },
+    {
+      "epoch": 0.9012269938650307,
+      "grad_norm": 0.16322194039821625,
+      "learning_rate": 3.6538607484828314e-06,
+      "loss": 1.1146,
+      "step": 5876
+    },
+    {
+      "epoch": 0.9013803680981595,
+      "grad_norm": 0.16598811745643616,
+      "learning_rate": 3.642614852903045e-06,
+      "loss": 1.0124,
+      "step": 5877
+    },
+    {
+      "epoch": 0.9015337423312884,
+      "grad_norm": 0.1473853886127472,
+      "learning_rate": 3.631385859637473e-06,
+      "loss": 1.0012,
+      "step": 5878
+    },
+    {
+      "epoch": 0.9016871165644171,
+      "grad_norm": 0.15713989734649658,
+      "learning_rate": 3.620173771345897e-06,
+      "loss": 1.003,
+      "step": 5879
+    },
+    {
+      "epoch": 0.901840490797546,
+      "grad_norm": 0.1592036336660385,
+      "learning_rate": 3.608978590684117e-06,
+      "loss": 1.0668,
+      "step": 5880
+    },
+    {
+      "epoch": 0.9019938650306748,
+      "grad_norm": 0.13889774680137634,
+      "learning_rate": 3.59780032030392e-06,
+      "loss": 0.9934,
+      "step": 5881
+    },
+    {
+      "epoch": 0.9021472392638037,
+      "grad_norm": 0.1553330272436142,
+      "learning_rate": 3.586638962853064e-06,
+      "loss": 1.003,
+      "step": 5882
+    },
+    {
+      "epoch": 0.9023006134969325,
+      "grad_norm": 0.16046211123466492,
+      "learning_rate": 3.5754945209753414e-06,
+      "loss": 1.1233,
+      "step": 5883
+    },
+    {
+      "epoch": 0.9024539877300614,
+      "grad_norm": 0.1493244767189026,
+      "learning_rate": 3.5643669973105165e-06,
+      "loss": 0.9846,
+      "step": 5884
+    },
+    {
+      "epoch": 0.9026073619631901,
+      "grad_norm": 0.14437395334243774,
+      "learning_rate": 3.553256394494347e-06,
+      "loss": 0.9533,
+      "step": 5885
+    },
+    {
+      "epoch": 0.902760736196319,
+      "grad_norm": 0.1498003751039505,
+      "learning_rate": 3.5421627151585863e-06,
+      "loss": 1.0989,
+      "step": 5886
+    },
+    {
+      "epoch": 0.9029141104294478,
+      "grad_norm": 0.17019787430763245,
+      "learning_rate": 3.531085961930974e-06,
+      "loss": 1.0652,
+      "step": 5887
+    },
+    {
+      "epoch": 0.9030674846625767,
+      "grad_norm": 0.16410811245441437,
+      "learning_rate": 3.5200261374352445e-06,
+      "loss": 1.0572,
+      "step": 5888
+    },
+    {
+      "epoch": 0.9032208588957055,
+      "grad_norm": 0.15384036302566528,
+      "learning_rate": 3.5089832442911277e-06,
+      "loss": 1.0955,
+      "step": 5889
+    },
+    {
+      "epoch": 0.9033742331288344,
+      "grad_norm": 0.15721964836120605,
+      "learning_rate": 3.4979572851143306e-06,
+      "loss": 1.1283,
+      "step": 5890
+    },
+    {
+      "epoch": 0.9035276073619631,
+      "grad_norm": 0.15675444900989532,
+      "learning_rate": 3.4869482625165478e-06,
+      "loss": 1.1537,
+      "step": 5891
+    },
+    {
+      "epoch": 0.903680981595092,
+      "grad_norm": 0.16809481382369995,
+      "learning_rate": 3.475956179105485e-06,
+      "loss": 1.112,
+      "step": 5892
+    },
+    {
+      "epoch": 0.9038343558282208,
+      "grad_norm": 0.16824904084205627,
+      "learning_rate": 3.4649810374848014e-06,
+      "loss": 1.1536,
+      "step": 5893
+    },
+    {
+      "epoch": 0.9039877300613497,
+      "grad_norm": 0.14723548293113708,
+      "learning_rate": 3.4540228402541837e-06,
+      "loss": 0.9469,
+      "step": 5894
+    },
+    {
+      "epoch": 0.9041411042944786,
+      "grad_norm": 0.15264569222927094,
+      "learning_rate": 3.443081590009264e-06,
+      "loss": 1.0964,
+      "step": 5895
+    },
+    {
+      "epoch": 0.9042944785276074,
+      "grad_norm": 0.1618390679359436,
+      "learning_rate": 3.4321572893417027e-06,
+      "loss": 1.1499,
+      "step": 5896
+    },
+    {
+      "epoch": 0.9044478527607362,
+      "grad_norm": 0.15304644405841827,
+      "learning_rate": 3.4212499408391046e-06,
+      "loss": 0.9298,
+      "step": 5897
+    },
+    {
+      "epoch": 0.904601226993865,
+      "grad_norm": 0.15273432433605194,
+      "learning_rate": 3.4103595470850704e-06,
+      "loss": 0.9512,
+      "step": 5898
+    },
+    {
+      "epoch": 0.9047546012269939,
+      "grad_norm": 0.15641191601753235,
+      "learning_rate": 3.399486110659211e-06,
+      "loss": 1.0136,
+      "step": 5899
+    },
+    {
+      "epoch": 0.9049079754601227,
+      "grad_norm": 0.15808595716953278,
+      "learning_rate": 3.3886296341370743e-06,
+      "loss": 1.0161,
+      "step": 5900
+    },
+    {
+      "epoch": 0.9050613496932516,
+      "grad_norm": 0.1602485179901123,
+      "learning_rate": 3.3777901200902457e-06,
+      "loss": 1.1336,
+      "step": 5901
+    },
+    {
+      "epoch": 0.9052147239263804,
+      "grad_norm": 0.15380018949508667,
+      "learning_rate": 3.3669675710862376e-06,
+      "loss": 1.0342,
+      "step": 5902
+    },
+    {
+      "epoch": 0.9053680981595092,
+      "grad_norm": 0.15821324288845062,
+      "learning_rate": 3.3561619896885995e-06,
+      "loss": 0.9891,
+      "step": 5903
+    },
+    {
+      "epoch": 0.905521472392638,
+      "grad_norm": 0.13085947930812836,
+      "learning_rate": 3.345373378456817e-06,
+      "loss": 0.8515,
+      "step": 5904
+    },
+    {
+      "epoch": 0.9056748466257669,
+      "grad_norm": 0.15754137933254242,
+      "learning_rate": 3.3346017399463546e-06,
+      "loss": 0.9601,
+      "step": 5905
+    },
+    {
+      "epoch": 0.9058282208588957,
+      "grad_norm": 0.16358177363872528,
+      "learning_rate": 3.3238470767087048e-06,
+      "loss": 1.0531,
+      "step": 5906
+    },
+    {
+      "epoch": 0.9059815950920246,
+      "grad_norm": 0.15506300330162048,
+      "learning_rate": 3.3131093912912882e-06,
+      "loss": 0.9502,
+      "step": 5907
+    },
+    {
+      "epoch": 0.9061349693251534,
+      "grad_norm": 0.13719794154167175,
+      "learning_rate": 3.3023886862375366e-06,
+      "loss": 1.0142,
+      "step": 5908
+    },
+    {
+      "epoch": 0.9062883435582823,
+      "grad_norm": 0.1445118486881256,
+      "learning_rate": 3.291684964086852e-06,
+      "loss": 1.0045,
+      "step": 5909
+    },
+    {
+      "epoch": 0.906441717791411,
+      "grad_norm": 0.16929525136947632,
+      "learning_rate": 3.2809982273745733e-06,
+      "loss": 1.2061,
+      "step": 5910
+    },
+    {
+      "epoch": 0.9065950920245399,
+      "grad_norm": 0.1613595336675644,
+      "learning_rate": 3.270328478632092e-06,
+      "loss": 0.9884,
+      "step": 5911
+    },
+    {
+      "epoch": 0.9067484662576687,
+      "grad_norm": 0.15186302363872528,
+      "learning_rate": 3.259675720386712e-06,
+      "loss": 0.9898,
+      "step": 5912
+    },
+    {
+      "epoch": 0.9069018404907976,
+      "grad_norm": 0.1548752337694168,
+      "learning_rate": 3.2490399551617472e-06,
+      "loss": 0.9042,
+      "step": 5913
+    },
+    {
+      "epoch": 0.9070552147239264,
+      "grad_norm": 0.14568808674812317,
+      "learning_rate": 3.238421185476467e-06,
+      "loss": 0.975,
+      "step": 5914
+    },
+    {
+      "epoch": 0.9072085889570553,
+      "grad_norm": 0.15052668750286102,
+      "learning_rate": 3.2278194138461332e-06,
+      "loss": 1.0741,
+      "step": 5915
+    },
+    {
+      "epoch": 0.907361963190184,
+      "grad_norm": 0.1563505381345749,
+      "learning_rate": 3.217234642781963e-06,
+      "loss": 1.0613,
+      "step": 5916
+    },
+    {
+      "epoch": 0.9075153374233129,
+      "grad_norm": 0.16215071082115173,
+      "learning_rate": 3.2066668747911415e-06,
+      "loss": 1.0272,
+      "step": 5917
+    },
+    {
+      "epoch": 0.9076687116564417,
+      "grad_norm": 0.14061282575130463,
+      "learning_rate": 3.1961161123768756e-06,
+      "loss": 1.0403,
+      "step": 5918
+    },
+    {
+      "epoch": 0.9078220858895706,
+      "grad_norm": 0.141673743724823,
+      "learning_rate": 3.185582358038266e-06,
+      "loss": 0.9632,
+      "step": 5919
+    },
+    {
+      "epoch": 0.9079754601226994,
+      "grad_norm": 0.15755723416805267,
+      "learning_rate": 3.175065614270467e-06,
+      "loss": 1.0118,
+      "step": 5920
+    },
+    {
+      "epoch": 0.9081288343558283,
+      "grad_norm": 0.14841370284557343,
+      "learning_rate": 3.1645658835645265e-06,
+      "loss": 0.997,
+      "step": 5921
+    },
+    {
+      "epoch": 0.908282208588957,
+      "grad_norm": 0.15082977712154388,
+      "learning_rate": 3.154083168407531e-06,
+      "loss": 0.9458,
+      "step": 5922
+    },
+    {
+      "epoch": 0.9084355828220859,
+      "grad_norm": 0.14099958539009094,
+      "learning_rate": 3.143617471282478e-06,
+      "loss": 0.865,
+      "step": 5923
+    },
+    {
+      "epoch": 0.9085889570552147,
+      "grad_norm": 0.15425287187099457,
+      "learning_rate": 3.133168794668392e-06,
+      "loss": 1.0727,
+      "step": 5924
+    },
+    {
+      "epoch": 0.9087423312883436,
+      "grad_norm": 0.20319843292236328,
+      "learning_rate": 3.12273714104021e-06,
+      "loss": 0.9368,
+      "step": 5925
+    },
+    {
+      "epoch": 0.9088957055214724,
+      "grad_norm": 0.15830613672733307,
+      "learning_rate": 3.1123225128688646e-06,
+      "loss": 1.1197,
+      "step": 5926
+    },
+    {
+      "epoch": 0.9090490797546013,
+      "grad_norm": 0.146449014544487,
+      "learning_rate": 3.101924912621265e-06,
+      "loss": 0.947,
+      "step": 5927
+    },
+    {
+      "epoch": 0.90920245398773,
+      "grad_norm": 0.14848999679088593,
+      "learning_rate": 3.0915443427602575e-06,
+      "loss": 1.0776,
+      "step": 5928
+    },
+    {
+      "epoch": 0.9093558282208589,
+      "grad_norm": 0.14669252932071686,
+      "learning_rate": 3.081180805744693e-06,
+      "loss": 1.1116,
+      "step": 5929
+    },
+    {
+      "epoch": 0.9095092024539877,
+      "grad_norm": 0.1479872465133667,
+      "learning_rate": 3.0708343040293487e-06,
+      "loss": 0.8965,
+      "step": 5930
+    },
+    {
+      "epoch": 0.9096625766871166,
+      "grad_norm": 0.15174394845962524,
+      "learning_rate": 3.060504840064998e-06,
+      "loss": 0.9947,
+      "step": 5931
+    },
+    {
+      "epoch": 0.9098159509202454,
+      "grad_norm": 0.14934507012367249,
+      "learning_rate": 3.050192416298358e-06,
+      "loss": 0.9694,
+      "step": 5932
+    },
+    {
+      "epoch": 0.9099693251533743,
+      "grad_norm": 0.15751032531261444,
+      "learning_rate": 3.039897035172101e-06,
+      "loss": 1.0155,
+      "step": 5933
+    },
+    {
+      "epoch": 0.910122699386503,
+      "grad_norm": 0.1531027853488922,
+      "learning_rate": 3.0296186991249085e-06,
+      "loss": 1.0645,
+      "step": 5934
+    },
+    {
+      "epoch": 0.9102760736196319,
+      "grad_norm": 0.1473522186279297,
+      "learning_rate": 3.019357410591375e-06,
+      "loss": 0.9386,
+      "step": 5935
+    },
+    {
+      "epoch": 0.9104294478527607,
+      "grad_norm": 0.14811834692955017,
+      "learning_rate": 3.0091131720020893e-06,
+      "loss": 1.0192,
+      "step": 5936
+    },
+    {
+      "epoch": 0.9105828220858896,
+      "grad_norm": 0.1611083745956421,
+      "learning_rate": 2.998885985783561e-06,
+      "loss": 1.1526,
+      "step": 5937
+    },
+    {
+      "epoch": 0.9107361963190184,
+      "grad_norm": 0.14472997188568115,
+      "learning_rate": 2.9886758543583273e-06,
+      "loss": 1.0901,
+      "step": 5938
+    },
+    {
+      "epoch": 0.9108895705521473,
+      "grad_norm": 0.14949196577072144,
+      "learning_rate": 2.9784827801448195e-06,
+      "loss": 0.9781,
+      "step": 5939
+    },
+    {
+      "epoch": 0.911042944785276,
+      "grad_norm": 0.15692375600337982,
+      "learning_rate": 2.9683067655574573e-06,
+      "loss": 0.9942,
+      "step": 5940
+    },
+    {
+      "epoch": 0.9111963190184049,
+      "grad_norm": 0.15228824317455292,
+      "learning_rate": 2.9581478130066283e-06,
+      "loss": 0.9545,
+      "step": 5941
+    },
+    {
+      "epoch": 0.9113496932515337,
+      "grad_norm": 0.16351252794265747,
+      "learning_rate": 2.9480059248986497e-06,
+      "loss": 1.1221,
+      "step": 5942
+    },
+    {
+      "epoch": 0.9115030674846626,
+      "grad_norm": 0.14677830040454865,
+      "learning_rate": 2.937881103635842e-06,
+      "loss": 0.8996,
+      "step": 5943
+    },
+    {
+      "epoch": 0.9116564417177914,
+      "grad_norm": 0.14071190357208252,
+      "learning_rate": 2.927773351616436e-06,
+      "loss": 0.9643,
+      "step": 5944
+    },
+    {
+      "epoch": 0.9118098159509203,
+      "grad_norm": 0.15667179226875305,
+      "learning_rate": 2.9176826712346416e-06,
+      "loss": 1.0751,
+      "step": 5945
+    },
+    {
+      "epoch": 0.911963190184049,
+      "grad_norm": 0.14417657256126404,
+      "learning_rate": 2.907609064880631e-06,
+      "loss": 0.9586,
+      "step": 5946
+    },
+    {
+      "epoch": 0.9121165644171779,
+      "grad_norm": 0.15129834413528442,
+      "learning_rate": 2.8975525349405115e-06,
+      "loss": 1.0922,
+      "step": 5947
+    },
+    {
+      "epoch": 0.9122699386503067,
+      "grad_norm": 0.1334395706653595,
+      "learning_rate": 2.8875130837963694e-06,
+      "loss": 0.8772,
+      "step": 5948
+    },
+    {
+      "epoch": 0.9124233128834356,
+      "grad_norm": 0.1654721051454544,
+      "learning_rate": 2.8774907138262278e-06,
+      "loss": 1.1046,
+      "step": 5949
+    },
+    {
+      "epoch": 0.9125766871165644,
+      "grad_norm": 0.148402601480484,
+      "learning_rate": 2.8674854274040786e-06,
+      "loss": 1.0423,
+      "step": 5950
+    },
+    {
+      "epoch": 0.9127300613496933,
+      "grad_norm": 0.15731282532215118,
+      "learning_rate": 2.857497226899852e-06,
+      "loss": 0.968,
+      "step": 5951
+    },
+    {
+      "epoch": 0.912883435582822,
+      "grad_norm": 0.14478661119937897,
+      "learning_rate": 2.8475261146794296e-06,
+      "loss": 0.9084,
+      "step": 5952
+    },
+    {
+      "epoch": 0.9130368098159509,
+      "grad_norm": 0.16299277544021606,
+      "learning_rate": 2.837572093104673e-06,
+      "loss": 1.0424,
+      "step": 5953
+    },
+    {
+      "epoch": 0.9131901840490797,
+      "grad_norm": 0.18644511699676514,
+      "learning_rate": 2.827635164533354e-06,
+      "loss": 0.9397,
+      "step": 5954
+    },
+    {
+      "epoch": 0.9133435582822086,
+      "grad_norm": 0.1548665463924408,
+      "learning_rate": 2.8177153313192314e-06,
+      "loss": 1.0351,
+      "step": 5955
+    },
+    {
+      "epoch": 0.9134969325153375,
+      "grad_norm": 0.15024320781230927,
+      "learning_rate": 2.8078125958119926e-06,
+      "loss": 0.9973,
+      "step": 5956
+    },
+    {
+      "epoch": 0.9136503067484663,
+      "grad_norm": 0.16121520102024078,
+      "learning_rate": 2.7979269603573025e-06,
+      "loss": 1.1814,
+      "step": 5957
+    },
+    {
+      "epoch": 0.9138036809815951,
+      "grad_norm": 0.16208361089229584,
+      "learning_rate": 2.7880584272967297e-06,
+      "loss": 1.1673,
+      "step": 5958
+    },
+    {
+      "epoch": 0.9139570552147239,
+      "grad_norm": 0.1654001772403717,
+      "learning_rate": 2.7782069989678386e-06,
+      "loss": 1.0492,
+      "step": 5959
+    },
+    {
+      "epoch": 0.9141104294478528,
+      "grad_norm": 0.1618158519268036,
+      "learning_rate": 2.7683726777041198e-06,
+      "loss": 1.083,
+      "step": 5960
+    },
+    {
+      "epoch": 0.9142638036809816,
+      "grad_norm": 0.1610296070575714,
+      "learning_rate": 2.758555465834994e-06,
+      "loss": 1.1342,
+      "step": 5961
+    },
+    {
+      "epoch": 0.9144171779141105,
+      "grad_norm": 0.15311822295188904,
+      "learning_rate": 2.7487553656858756e-06,
+      "loss": 1.0793,
+      "step": 5962
+    },
+    {
+      "epoch": 0.9145705521472393,
+      "grad_norm": 0.1411464661359787,
+      "learning_rate": 2.738972379578075e-06,
+      "loss": 0.9226,
+      "step": 5963
+    },
+    {
+      "epoch": 0.9147239263803681,
+      "grad_norm": 0.1546267718076706,
+      "learning_rate": 2.729206509828896e-06,
+      "loss": 0.9979,
+      "step": 5964
+    },
+    {
+      "epoch": 0.9148773006134969,
+      "grad_norm": 0.13861538469791412,
+      "learning_rate": 2.7194577587515476e-06,
+      "loss": 0.9078,
+      "step": 5965
+    },
+    {
+      "epoch": 0.9150306748466258,
+      "grad_norm": 0.15345698595046997,
+      "learning_rate": 2.709726128655215e-06,
+      "loss": 1.031,
+      "step": 5966
+    },
+    {
+      "epoch": 0.9151840490797546,
+      "grad_norm": 0.15485289692878723,
+      "learning_rate": 2.7000116218450053e-06,
+      "loss": 1.0403,
+      "step": 5967
+    },
+    {
+      "epoch": 0.9153374233128835,
+      "grad_norm": 0.1624671369791031,
+      "learning_rate": 2.690314240621977e-06,
+      "loss": 1.1418,
+      "step": 5968
+    },
+    {
+      "epoch": 0.9154907975460123,
+      "grad_norm": 0.15345712006092072,
+      "learning_rate": 2.6806339872831427e-06,
+      "loss": 1.1075,
+      "step": 5969
+    },
+    {
+      "epoch": 0.9156441717791411,
+      "grad_norm": 0.1597621738910675,
+      "learning_rate": 2.670970864121444e-06,
+      "loss": 1.0089,
+      "step": 5970
+    },
+    {
+      "epoch": 0.9157975460122699,
+      "grad_norm": 0.15370814502239227,
+      "learning_rate": 2.661324873425774e-06,
+      "loss": 1.0888,
+      "step": 5971
+    },
+    {
+      "epoch": 0.9159509202453988,
+      "grad_norm": 0.16114428639411926,
+      "learning_rate": 2.6516960174809476e-06,
+      "loss": 1.0786,
+      "step": 5972
+    },
+    {
+      "epoch": 0.9161042944785276,
+      "grad_norm": 0.15520064532756805,
+      "learning_rate": 2.642084298567765e-06,
+      "loss": 0.9852,
+      "step": 5973
+    },
+    {
+      "epoch": 0.9162576687116565,
+      "grad_norm": 0.15656356513500214,
+      "learning_rate": 2.6324897189629214e-06,
+      "loss": 1.1043,
+      "step": 5974
+    },
+    {
+      "epoch": 0.9164110429447853,
+      "grad_norm": 0.16386081278324127,
+      "learning_rate": 2.6229122809390656e-06,
+      "loss": 1.1595,
+      "step": 5975
+    },
+    {
+      "epoch": 0.9165644171779141,
+      "grad_norm": 0.1570153683423996,
+      "learning_rate": 2.6133519867648084e-06,
+      "loss": 1.0936,
+      "step": 5976
+    },
+    {
+      "epoch": 0.9167177914110429,
+      "grad_norm": 0.1613694429397583,
+      "learning_rate": 2.603808838704663e-06,
+      "loss": 1.0804,
+      "step": 5977
+    },
+    {
+      "epoch": 0.9168711656441718,
+      "grad_norm": 0.15751634538173676,
+      "learning_rate": 2.594282839019113e-06,
+      "loss": 1.0133,
+      "step": 5978
+    },
+    {
+      "epoch": 0.9170245398773006,
+      "grad_norm": 0.15686161816120148,
+      "learning_rate": 2.5847739899645615e-06,
+      "loss": 0.968,
+      "step": 5979
+    },
+    {
+      "epoch": 0.9171779141104295,
+      "grad_norm": 0.16074183583259583,
+      "learning_rate": 2.575282293793357e-06,
+      "loss": 1.1434,
+      "step": 5980
+    },
+    {
+      "epoch": 0.9173312883435583,
+      "grad_norm": 0.14406302571296692,
+      "learning_rate": 2.565807752753776e-06,
+      "loss": 0.9303,
+      "step": 5981
+    },
+    {
+      "epoch": 0.9174846625766871,
+      "grad_norm": 0.14101062715053558,
+      "learning_rate": 2.556350369090057e-06,
+      "loss": 0.8883,
+      "step": 5982
+    },
+    {
+      "epoch": 0.9176380368098159,
+      "grad_norm": 0.14877983927726746,
+      "learning_rate": 2.5469101450423495e-06,
+      "loss": 1.0479,
+      "step": 5983
+    },
+    {
+      "epoch": 0.9177914110429448,
+      "grad_norm": 0.16489146649837494,
+      "learning_rate": 2.537487082846723e-06,
+      "loss": 1.1032,
+      "step": 5984
+    },
+    {
+      "epoch": 0.9179447852760736,
+      "grad_norm": 0.1745263636112213,
+      "learning_rate": 2.5280811847352258e-06,
+      "loss": 1.0492,
+      "step": 5985
+    },
+    {
+      "epoch": 0.9180981595092025,
+      "grad_norm": 0.14576703310012817,
+      "learning_rate": 2.518692452935808e-06,
+      "loss": 0.9331,
+      "step": 5986
+    },
+    {
+      "epoch": 0.9182515337423313,
+      "grad_norm": 0.14987510442733765,
+      "learning_rate": 2.5093208896723838e-06,
+      "loss": 1.0433,
+      "step": 5987
+    },
+    {
+      "epoch": 0.9184049079754601,
+      "grad_norm": 0.15420925617218018,
+      "learning_rate": 2.4999664971647514e-06,
+      "loss": 1.073,
+      "step": 5988
+    },
+    {
+      "epoch": 0.9185582822085889,
+      "grad_norm": 0.15947505831718445,
+      "learning_rate": 2.4906292776286972e-06,
+      "loss": 1.0447,
+      "step": 5989
+    },
+    {
+      "epoch": 0.9187116564417178,
+      "grad_norm": 0.15079933404922485,
+      "learning_rate": 2.4813092332758856e-06,
+      "loss": 0.9633,
+      "step": 5990
+    },
+    {
+      "epoch": 0.9188650306748466,
+      "grad_norm": 0.1595383733510971,
+      "learning_rate": 2.4720063663139665e-06,
+      "loss": 1.1418,
+      "step": 5991
+    },
+    {
+      "epoch": 0.9190184049079755,
+      "grad_norm": 0.1588064432144165,
+      "learning_rate": 2.462720678946495e-06,
+      "loss": 1.0394,
+      "step": 5992
+    },
+    {
+      "epoch": 0.9191717791411043,
+      "grad_norm": 0.15718252956867218,
+      "learning_rate": 2.453452173372944e-06,
+      "loss": 1.0526,
+      "step": 5993
+    },
+    {
+      "epoch": 0.9193251533742332,
+      "grad_norm": 0.16322046518325806,
+      "learning_rate": 2.4442008517887495e-06,
+      "loss": 1.0436,
+      "step": 5994
+    },
+    {
+      "epoch": 0.9194785276073619,
+      "grad_norm": 0.13943615555763245,
+      "learning_rate": 2.434966716385234e-06,
+      "loss": 0.9696,
+      "step": 5995
+    },
+    {
+      "epoch": 0.9196319018404908,
+      "grad_norm": 0.14939971268177032,
+      "learning_rate": 2.42574976934968e-06,
+      "loss": 0.9913,
+      "step": 5996
+    },
+    {
+      "epoch": 0.9197852760736196,
+      "grad_norm": 0.1777634173631668,
+      "learning_rate": 2.416550012865309e-06,
+      "loss": 1.1785,
+      "step": 5997
+    },
+    {
+      "epoch": 0.9199386503067485,
+      "grad_norm": 0.15695764124393463,
+      "learning_rate": 2.4073674491112277e-06,
+      "loss": 1.0515,
+      "step": 5998
+    },
+    {
+      "epoch": 0.9200920245398773,
+      "grad_norm": 0.15057872235774994,
+      "learning_rate": 2.3982020802625126e-06,
+      "loss": 0.9945,
+      "step": 5999
+    },
+    {
+      "epoch": 0.9202453987730062,
+      "grad_norm": 0.15419985353946686,
+      "learning_rate": 2.3890539084901435e-06,
+      "loss": 1.125,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9203987730061349,
+      "grad_norm": 0.15763753652572632,
+      "learning_rate": 2.3799229359610374e-06,
+      "loss": 1.0258,
+      "step": 6001
+    },
+    {
+      "epoch": 0.9205521472392638,
+      "grad_norm": 0.16248351335525513,
+      "learning_rate": 2.3708091648380388e-06,
+      "loss": 1.0578,
+      "step": 6002
+    },
+    {
+      "epoch": 0.9207055214723926,
+      "grad_norm": 0.1544421911239624,
+      "learning_rate": 2.3617125972798966e-06,
+      "loss": 0.981,
+      "step": 6003
+    },
+    {
+      "epoch": 0.9208588957055215,
+      "grad_norm": 0.15288609266281128,
+      "learning_rate": 2.352633235441312e-06,
+      "loss": 1.0246,
+      "step": 6004
+    },
+    {
+      "epoch": 0.9210122699386503,
+      "grad_norm": 0.15894842147827148,
+      "learning_rate": 2.3435710814728976e-06,
+      "loss": 1.0737,
+      "step": 6005
+    },
+    {
+      "epoch": 0.9211656441717792,
+      "grad_norm": 0.14844535291194916,
+      "learning_rate": 2.3345261375211947e-06,
+      "loss": 0.9812,
+      "step": 6006
+    },
+    {
+      "epoch": 0.9213190184049079,
+      "grad_norm": 0.14181934297084808,
+      "learning_rate": 2.3254984057286568e-06,
+      "loss": 0.9238,
+      "step": 6007
+    },
+    {
+      "epoch": 0.9214723926380368,
+      "grad_norm": 0.14965277910232544,
+      "learning_rate": 2.3164878882336895e-06,
+      "loss": 1.0039,
+      "step": 6008
+    },
+    {
+      "epoch": 0.9216257668711656,
+      "grad_norm": 0.15251348912715912,
+      "learning_rate": 2.3074945871705775e-06,
+      "loss": 0.9755,
+      "step": 6009
+    },
+    {
+      "epoch": 0.9217791411042945,
+      "grad_norm": 0.1450740247964859,
+      "learning_rate": 2.2985185046695583e-06,
+      "loss": 0.9138,
+      "step": 6010
+    },
+    {
+      "epoch": 0.9219325153374233,
+      "grad_norm": 0.15668326616287231,
+      "learning_rate": 2.289559642856781e-06,
+      "loss": 1.1096,
+      "step": 6011
+    },
+    {
+      "epoch": 0.9220858895705522,
+      "grad_norm": 0.13782154023647308,
+      "learning_rate": 2.280618003854315e-06,
+      "loss": 0.8646,
+      "step": 6012
+    },
+    {
+      "epoch": 0.9222392638036809,
+      "grad_norm": 0.1517089158296585,
+      "learning_rate": 2.2716935897801658e-06,
+      "loss": 1.0513,
+      "step": 6013
+    },
+    {
+      "epoch": 0.9223926380368098,
+      "grad_norm": 0.15845313668251038,
+      "learning_rate": 2.2627864027482344e-06,
+      "loss": 1.0676,
+      "step": 6014
+    },
+    {
+      "epoch": 0.9225460122699386,
+      "grad_norm": 0.16631728410720825,
+      "learning_rate": 2.2538964448683582e-06,
+      "loss": 1.2606,
+      "step": 6015
+    },
+    {
+      "epoch": 0.9226993865030675,
+      "grad_norm": 0.15559019148349762,
+      "learning_rate": 2.2450237182462693e-06,
+      "loss": 0.9687,
+      "step": 6016
+    },
+    {
+      "epoch": 0.9228527607361963,
+      "grad_norm": 0.16060636937618256,
+      "learning_rate": 2.2361682249836697e-06,
+      "loss": 1.0063,
+      "step": 6017
+    },
+    {
+      "epoch": 0.9230061349693252,
+      "grad_norm": 0.16213560104370117,
+      "learning_rate": 2.2273299671781236e-06,
+      "loss": 1.174,
+      "step": 6018
+    },
+    {
+      "epoch": 0.923159509202454,
+      "grad_norm": 0.15483447909355164,
+      "learning_rate": 2.2185089469231394e-06,
+      "loss": 0.9648,
+      "step": 6019
+    },
+    {
+      "epoch": 0.9233128834355828,
+      "grad_norm": 0.1654367744922638,
+      "learning_rate": 2.209705166308137e-06,
+      "loss": 1.092,
+      "step": 6020
+    },
+    {
+      "epoch": 0.9234662576687117,
+      "grad_norm": 0.1600506752729416,
+      "learning_rate": 2.2009186274184574e-06,
+      "loss": 1.104,
+      "step": 6021
+    },
+    {
+      "epoch": 0.9236196319018405,
+      "grad_norm": 0.15514373779296875,
+      "learning_rate": 2.1921493323353606e-06,
+      "loss": 1.0367,
+      "step": 6022
+    },
+    {
+      "epoch": 0.9237730061349694,
+      "grad_norm": 0.1545235514640808,
+      "learning_rate": 2.1833972831360014e-06,
+      "loss": 0.9718,
+      "step": 6023
+    },
+    {
+      "epoch": 0.9239263803680982,
+      "grad_norm": 0.158521369099617,
+      "learning_rate": 2.17466248189348e-06,
+      "loss": 1.0528,
+      "step": 6024
+    },
+    {
+      "epoch": 0.924079754601227,
+      "grad_norm": 0.15031425654888153,
+      "learning_rate": 2.165944930676791e-06,
+      "loss": 1.0011,
+      "step": 6025
+    },
+    {
+      "epoch": 0.9242331288343558,
+      "grad_norm": 0.15098656713962555,
+      "learning_rate": 2.1572446315508337e-06,
+      "loss": 1.0236,
+      "step": 6026
+    },
+    {
+      "epoch": 0.9243865030674847,
+      "grad_norm": 0.16604359447956085,
+      "learning_rate": 2.1485615865764583e-06,
+      "loss": 1.0771,
+      "step": 6027
+    },
+    {
+      "epoch": 0.9245398773006135,
+      "grad_norm": 0.1492164134979248,
+      "learning_rate": 2.139895797810379e-06,
+      "loss": 1.0128,
+      "step": 6028
+    },
+    {
+      "epoch": 0.9246932515337424,
+      "grad_norm": 0.15764212608337402,
+      "learning_rate": 2.1312472673052695e-06,
+      "loss": 0.9605,
+      "step": 6029
+    },
+    {
+      "epoch": 0.9248466257668712,
+      "grad_norm": 0.1459902822971344,
+      "learning_rate": 2.1226159971096752e-06,
+      "loss": 0.9957,
+      "step": 6030
+    },
+    {
+      "epoch": 0.925,
+      "grad_norm": 0.1460924744606018,
+      "learning_rate": 2.114001989268094e-06,
+      "loss": 1.0306,
+      "step": 6031
+    },
+    {
+      "epoch": 0.9251533742331288,
+      "grad_norm": 0.16127847135066986,
+      "learning_rate": 2.1054052458208847e-06,
+      "loss": 1.0694,
+      "step": 6032
+    },
+    {
+      "epoch": 0.9253067484662577,
+      "grad_norm": 0.16614152491092682,
+      "learning_rate": 2.0968257688043695e-06,
+      "loss": 1.0007,
+      "step": 6033
+    },
+    {
+      "epoch": 0.9254601226993865,
+      "grad_norm": 0.14390528202056885,
+      "learning_rate": 2.088263560250747e-06,
+      "loss": 0.9165,
+      "step": 6034
+    },
+    {
+      "epoch": 0.9256134969325154,
+      "grad_norm": 0.14491645991802216,
+      "learning_rate": 2.0797186221881283e-06,
+      "loss": 0.9078,
+      "step": 6035
+    },
+    {
+      "epoch": 0.9257668711656442,
+      "grad_norm": 0.14940379559993744,
+      "learning_rate": 2.0711909566405534e-06,
+      "loss": 1.0186,
+      "step": 6036
+    },
+    {
+      "epoch": 0.925920245398773,
+      "grad_norm": 0.17976130545139313,
+      "learning_rate": 2.0626805656279473e-06,
+      "loss": 1.0838,
+      "step": 6037
+    },
+    {
+      "epoch": 0.9260736196319018,
+      "grad_norm": 0.16202814877033234,
+      "learning_rate": 2.0541874511661477e-06,
+      "loss": 1.0383,
+      "step": 6038
+    },
+    {
+      "epoch": 0.9262269938650307,
+      "grad_norm": 0.1504490226507187,
+      "learning_rate": 2.045711615266929e-06,
+      "loss": 1.042,
+      "step": 6039
+    },
+    {
+      "epoch": 0.9263803680981595,
+      "grad_norm": 0.14487427473068237,
+      "learning_rate": 2.0372530599379184e-06,
+      "loss": 1.005,
+      "step": 6040
+    },
+    {
+      "epoch": 0.9265337423312884,
+      "grad_norm": 0.1654735952615738,
+      "learning_rate": 2.0288117871827046e-06,
+      "loss": 1.0482,
+      "step": 6041
+    },
+    {
+      "epoch": 0.9266871165644172,
+      "grad_norm": 0.14739736914634705,
+      "learning_rate": 2.0203877990007387e-06,
+      "loss": 0.8633,
+      "step": 6042
+    },
+    {
+      "epoch": 0.926840490797546,
+      "grad_norm": 0.15377013385295868,
+      "learning_rate": 2.0119810973874247e-06,
+      "loss": 0.9768,
+      "step": 6043
+    },
+    {
+      "epoch": 0.9269938650306748,
+      "grad_norm": 0.16317322850227356,
+      "learning_rate": 2.003591684334027e-06,
+      "loss": 0.8969,
+      "step": 6044
+    },
+    {
+      "epoch": 0.9271472392638037,
+      "grad_norm": 0.1598971039056778,
+      "learning_rate": 1.995219561827724e-06,
+      "loss": 1.1218,
+      "step": 6045
+    },
+    {
+      "epoch": 0.9273006134969325,
+      "grad_norm": 0.13765203952789307,
+      "learning_rate": 1.9868647318516366e-06,
+      "loss": 0.9032,
+      "step": 6046
+    },
+    {
+      "epoch": 0.9274539877300614,
+      "grad_norm": 0.1665959358215332,
+      "learning_rate": 1.9785271963847242e-06,
+      "loss": 1.1409,
+      "step": 6047
+    },
+    {
+      "epoch": 0.9276073619631902,
+      "grad_norm": 0.149894580245018,
+      "learning_rate": 1.9702069574019234e-06,
+      "loss": 0.929,
+      "step": 6048
+    },
+    {
+      "epoch": 0.927760736196319,
+      "grad_norm": 0.16531307995319366,
+      "learning_rate": 1.961904016874008e-06,
+      "loss": 1.0814,
+      "step": 6049
+    },
+    {
+      "epoch": 0.9279141104294478,
+      "grad_norm": 0.1490601897239685,
+      "learning_rate": 1.953618376767696e-06,
+      "loss": 0.9924,
+      "step": 6050
+    },
+    {
+      "epoch": 0.9280674846625767,
+      "grad_norm": 0.1491236835718155,
+      "learning_rate": 1.9453500390455845e-06,
+      "loss": 0.9463,
+      "step": 6051
+    },
+    {
+      "epoch": 0.9282208588957055,
+      "grad_norm": 0.14991188049316406,
+      "learning_rate": 1.9370990056661983e-06,
+      "loss": 0.9918,
+      "step": 6052
+    },
+    {
+      "epoch": 0.9283742331288344,
+      "grad_norm": 0.15511952340602875,
+      "learning_rate": 1.928865278583941e-06,
+      "loss": 1.0165,
+      "step": 6053
+    },
+    {
+      "epoch": 0.9285276073619632,
+      "grad_norm": 0.16552039980888367,
+      "learning_rate": 1.920648859749102e-06,
+      "loss": 1.0665,
+      "step": 6054
+    },
+    {
+      "epoch": 0.928680981595092,
+      "grad_norm": 0.15505027770996094,
+      "learning_rate": 1.912449751107925e-06,
+      "loss": 1.0323,
+      "step": 6055
+    },
+    {
+      "epoch": 0.9288343558282208,
+      "grad_norm": 0.14960920810699463,
+      "learning_rate": 1.9042679546024898e-06,
+      "loss": 1.0023,
+      "step": 6056
+    },
+    {
+      "epoch": 0.9289877300613497,
+      "grad_norm": 0.1532852202653885,
+      "learning_rate": 1.8961034721708373e-06,
+      "loss": 0.9761,
+      "step": 6057
+    },
+    {
+      "epoch": 0.9291411042944785,
+      "grad_norm": 0.15590858459472656,
+      "learning_rate": 1.887956305746846e-06,
+      "loss": 0.9829,
+      "step": 6058
+    },
+    {
+      "epoch": 0.9292944785276074,
+      "grad_norm": 0.1557212769985199,
+      "learning_rate": 1.8798264572603471e-06,
+      "loss": 1.0558,
+      "step": 6059
+    },
+    {
+      "epoch": 0.9294478527607362,
+      "grad_norm": 0.16249842941761017,
+      "learning_rate": 1.8717139286370252e-06,
+      "loss": 1.0766,
+      "step": 6060
+    },
+    {
+      "epoch": 0.929601226993865,
+      "grad_norm": 0.16333213448524475,
+      "learning_rate": 1.8636187217984932e-06,
+      "loss": 1.115,
+      "step": 6061
+    },
+    {
+      "epoch": 0.9297546012269938,
+      "grad_norm": 0.16872717440128326,
+      "learning_rate": 1.8555408386622506e-06,
+      "loss": 1.1622,
+      "step": 6062
+    },
+    {
+      "epoch": 0.9299079754601227,
+      "grad_norm": 0.15045209228992462,
+      "learning_rate": 1.8474802811416832e-06,
+      "loss": 1.0142,
+      "step": 6063
+    },
+    {
+      "epoch": 0.9300613496932515,
+      "grad_norm": 0.1419679969549179,
+      "learning_rate": 1.839437051146106e-06,
+      "loss": 1.0027,
+      "step": 6064
+    },
+    {
+      "epoch": 0.9302147239263804,
+      "grad_norm": 0.15909957885742188,
+      "learning_rate": 1.8314111505806779e-06,
+      "loss": 1.0888,
+      "step": 6065
+    },
+    {
+      "epoch": 0.9303680981595092,
+      "grad_norm": 0.1646888107061386,
+      "learning_rate": 1.8234025813464948e-06,
+      "loss": 1.1956,
+      "step": 6066
+    },
+    {
+      "epoch": 0.930521472392638,
+      "grad_norm": 0.16548165678977966,
+      "learning_rate": 1.8154113453405477e-06,
+      "loss": 1.0769,
+      "step": 6067
+    },
+    {
+      "epoch": 0.9306748466257668,
+      "grad_norm": 0.1517190784215927,
+      "learning_rate": 1.807437444455681e-06,
+      "loss": 1.0269,
+      "step": 6068
+    },
+    {
+      "epoch": 0.9308282208588957,
+      "grad_norm": 0.15104077756404877,
+      "learning_rate": 1.7994808805806838e-06,
+      "loss": 0.992,
+      "step": 6069
+    },
+    {
+      "epoch": 0.9309815950920245,
+      "grad_norm": 0.14755171537399292,
+      "learning_rate": 1.7915416556001982e-06,
+      "loss": 0.9792,
+      "step": 6070
+    },
+    {
+      "epoch": 0.9311349693251534,
+      "grad_norm": 0.15834470093250275,
+      "learning_rate": 1.783619771394787e-06,
+      "loss": 1.0987,
+      "step": 6071
+    },
+    {
+      "epoch": 0.9312883435582822,
+      "grad_norm": 0.16670343279838562,
+      "learning_rate": 1.7757152298408906e-06,
+      "loss": 1.1185,
+      "step": 6072
+    },
+    {
+      "epoch": 0.931441717791411,
+      "grad_norm": 0.15444029867649078,
+      "learning_rate": 1.7678280328108446e-06,
+      "loss": 1.0773,
+      "step": 6073
+    },
+    {
+      "epoch": 0.9315950920245398,
+      "grad_norm": 0.15091416239738464,
+      "learning_rate": 1.7599581821728799e-06,
+      "loss": 0.8889,
+      "step": 6074
+    },
+    {
+      "epoch": 0.9317484662576687,
+      "grad_norm": 0.1676899641752243,
+      "learning_rate": 1.7521056797911132e-06,
+      "loss": 1.0525,
+      "step": 6075
+    },
+    {
+      "epoch": 0.9319018404907975,
+      "grad_norm": 0.15001904964447021,
+      "learning_rate": 1.7442705275255652e-06,
+      "loss": 0.9581,
+      "step": 6076
+    },
+    {
+      "epoch": 0.9320552147239264,
+      "grad_norm": 0.14972887933254242,
+      "learning_rate": 1.7364527272321183e-06,
+      "loss": 1.0626,
+      "step": 6077
+    },
+    {
+      "epoch": 0.9322085889570552,
+      "grad_norm": 0.16126218438148499,
+      "learning_rate": 1.7286522807625825e-06,
+      "loss": 1.0168,
+      "step": 6078
+    },
+    {
+      "epoch": 0.932361963190184,
+      "grad_norm": 0.1527792066335678,
+      "learning_rate": 1.7208691899646215e-06,
+      "loss": 1.046,
+      "step": 6079
+    },
+    {
+      "epoch": 0.9325153374233128,
+      "grad_norm": 0.15487521886825562,
+      "learning_rate": 1.713103456681811e-06,
+      "loss": 1.0587,
+      "step": 6080
+    },
+    {
+      "epoch": 0.9326687116564417,
+      "grad_norm": 0.16363193094730377,
+      "learning_rate": 1.7053550827536121e-06,
+      "loss": 1.0914,
+      "step": 6081
+    },
+    {
+      "epoch": 0.9328220858895706,
+      "grad_norm": 0.1659364104270935,
+      "learning_rate": 1.6976240700153577e-06,
+      "loss": 1.0872,
+      "step": 6082
+    },
+    {
+      "epoch": 0.9329754601226994,
+      "grad_norm": 0.15130175650119781,
+      "learning_rate": 1.6899104202982905e-06,
+      "loss": 1.0846,
+      "step": 6083
+    },
+    {
+      "epoch": 0.9331288343558283,
+      "grad_norm": 0.17045357823371887,
+      "learning_rate": 1.6822141354295326e-06,
+      "loss": 1.1904,
+      "step": 6084
+    },
+    {
+      "epoch": 0.933282208588957,
+      "grad_norm": 0.16648514568805695,
+      "learning_rate": 1.6745352172320842e-06,
+      "loss": 1.0242,
+      "step": 6085
+    },
+    {
+      "epoch": 0.9334355828220859,
+      "grad_norm": 0.15923425555229187,
+      "learning_rate": 1.66687366752484e-06,
+      "loss": 1.1128,
+      "step": 6086
+    },
+    {
+      "epoch": 0.9335889570552147,
+      "grad_norm": 0.14567922055721283,
+      "learning_rate": 1.6592294881225904e-06,
+      "loss": 0.9625,
+      "step": 6087
+    },
+    {
+      "epoch": 0.9337423312883436,
+      "grad_norm": 0.158275306224823,
+      "learning_rate": 1.6516026808359868e-06,
+      "loss": 1.1451,
+      "step": 6088
+    },
+    {
+      "epoch": 0.9338957055214724,
+      "grad_norm": 0.15314657986164093,
+      "learning_rate": 1.643993247471584e-06,
+      "loss": 1.0271,
+      "step": 6089
+    },
+    {
+      "epoch": 0.9340490797546013,
+      "grad_norm": 0.14617325365543365,
+      "learning_rate": 1.6364011898318235e-06,
+      "loss": 0.9584,
+      "step": 6090
+    },
+    {
+      "epoch": 0.93420245398773,
+      "grad_norm": 0.16759973764419556,
+      "learning_rate": 1.6288265097150088e-06,
+      "loss": 1.0986,
+      "step": 6091
+    },
+    {
+      "epoch": 0.9343558282208589,
+      "grad_norm": 0.1560433954000473,
+      "learning_rate": 1.6212692089153546e-06,
+      "loss": 1.1404,
+      "step": 6092
+    },
+    {
+      "epoch": 0.9345092024539877,
+      "grad_norm": 0.1574346125125885,
+      "learning_rate": 1.6137292892229454e-06,
+      "loss": 1.044,
+      "step": 6093
+    },
+    {
+      "epoch": 0.9346625766871166,
+      "grad_norm": 0.15473192930221558,
+      "learning_rate": 1.6062067524237527e-06,
+      "loss": 1.0416,
+      "step": 6094
+    },
+    {
+      "epoch": 0.9348159509202454,
+      "grad_norm": 0.15070603787899017,
+      "learning_rate": 1.5987016002996177e-06,
+      "loss": 0.8873,
+      "step": 6095
+    },
+    {
+      "epoch": 0.9349693251533743,
+      "grad_norm": 0.14598338305950165,
+      "learning_rate": 1.5912138346282855e-06,
+      "loss": 1.0156,
+      "step": 6096
+    },
+    {
+      "epoch": 0.935122699386503,
+      "grad_norm": 0.14883489906787872,
+      "learning_rate": 1.58374345718337e-06,
+      "loss": 1.0309,
+      "step": 6097
+    },
+    {
+      "epoch": 0.9352760736196319,
+      "grad_norm": 0.15587270259857178,
+      "learning_rate": 1.5762904697343565e-06,
+      "loss": 1.114,
+      "step": 6098
+    },
+    {
+      "epoch": 0.9354294478527607,
+      "grad_norm": 0.15864789485931396,
+      "learning_rate": 1.568854874046649e-06,
+      "loss": 1.0453,
+      "step": 6099
+    },
+    {
+      "epoch": 0.9355828220858896,
+      "grad_norm": 0.16430217027664185,
+      "learning_rate": 1.561436671881472e-06,
+      "loss": 1.0888,
+      "step": 6100
+    },
+    {
+      "epoch": 0.9357361963190184,
+      "grad_norm": 0.1593499779701233,
+      "learning_rate": 1.5540358649959866e-06,
+      "loss": 1.0436,
+      "step": 6101
+    },
+    {
+      "epoch": 0.9358895705521473,
+      "grad_norm": 0.15240147709846497,
+      "learning_rate": 1.5466524551432074e-06,
+      "loss": 1.0395,
+      "step": 6102
+    },
+    {
+      "epoch": 0.936042944785276,
+      "grad_norm": 0.153757244348526,
+      "learning_rate": 1.539286444072027e-06,
+      "loss": 0.9875,
+      "step": 6103
+    },
+    {
+      "epoch": 0.9361963190184049,
+      "grad_norm": 0.16155293583869934,
+      "learning_rate": 1.5319378335272164e-06,
+      "loss": 0.9903,
+      "step": 6104
+    },
+    {
+      "epoch": 0.9363496932515337,
+      "grad_norm": 0.1442556232213974,
+      "learning_rate": 1.5246066252494417e-06,
+      "loss": 0.895,
+      "step": 6105
+    },
+    {
+      "epoch": 0.9365030674846626,
+      "grad_norm": 0.15271756052970886,
+      "learning_rate": 1.51729282097523e-06,
+      "loss": 0.9458,
+      "step": 6106
+    },
+    {
+      "epoch": 0.9366564417177914,
+      "grad_norm": 0.15482698380947113,
+      "learning_rate": 1.5099964224369793e-06,
+      "loss": 0.985,
+      "step": 6107
+    },
+    {
+      "epoch": 0.9368098159509203,
+      "grad_norm": 0.15694041550159454,
+      "learning_rate": 1.5027174313629986e-06,
+      "loss": 1.1045,
+      "step": 6108
+    },
+    {
+      "epoch": 0.936963190184049,
+      "grad_norm": 0.16060326993465424,
+      "learning_rate": 1.4954558494774254e-06,
+      "loss": 1.1184,
+      "step": 6109
+    },
+    {
+      "epoch": 0.9371165644171779,
+      "grad_norm": 0.1512058973312378,
+      "learning_rate": 1.4882116785003256e-06,
+      "loss": 0.9969,
+      "step": 6110
+    },
+    {
+      "epoch": 0.9372699386503067,
+      "grad_norm": 0.15925554931163788,
+      "learning_rate": 1.4809849201476014e-06,
+      "loss": 0.9594,
+      "step": 6111
+    },
+    {
+      "epoch": 0.9374233128834356,
+      "grad_norm": 0.1470421552658081,
+      "learning_rate": 1.473775576131042e-06,
+      "loss": 0.9752,
+      "step": 6112
+    },
+    {
+      "epoch": 0.9375766871165644,
+      "grad_norm": 0.15337421000003815,
+      "learning_rate": 1.4665836481583232e-06,
+      "loss": 1.0102,
+      "step": 6113
+    },
+    {
+      "epoch": 0.9377300613496933,
+      "grad_norm": 0.15003955364227295,
+      "learning_rate": 1.4594091379329737e-06,
+      "loss": 1.0241,
+      "step": 6114
+    },
+    {
+      "epoch": 0.9378834355828221,
+      "grad_norm": 0.16519220173358917,
+      "learning_rate": 1.4522520471544176e-06,
+      "loss": 1.1703,
+      "step": 6115
+    },
+    {
+      "epoch": 0.938036809815951,
+      "grad_norm": 0.14600934088230133,
+      "learning_rate": 1.445112377517932e-06,
+      "loss": 0.9768,
+      "step": 6116
+    },
+    {
+      "epoch": 0.9381901840490797,
+      "grad_norm": 0.14443185925483704,
+      "learning_rate": 1.4379901307146974e-06,
+      "loss": 0.9295,
+      "step": 6117
+    },
+    {
+      "epoch": 0.9383435582822086,
+      "grad_norm": 0.15800036489963531,
+      "learning_rate": 1.4308853084317395e-06,
+      "loss": 1.078,
+      "step": 6118
+    },
+    {
+      "epoch": 0.9384969325153374,
+      "grad_norm": 0.14208444952964783,
+      "learning_rate": 1.4237979123519622e-06,
+      "loss": 0.9614,
+      "step": 6119
+    },
+    {
+      "epoch": 0.9386503067484663,
+      "grad_norm": 0.149053692817688,
+      "learning_rate": 1.4167279441541556e-06,
+      "loss": 0.9924,
+      "step": 6120
+    },
+    {
+      "epoch": 0.9388036809815951,
+      "grad_norm": 0.16677020490169525,
+      "learning_rate": 1.4096754055129721e-06,
+      "loss": 1.0519,
+      "step": 6121
+    },
+    {
+      "epoch": 0.938957055214724,
+      "grad_norm": 0.14929889142513275,
+      "learning_rate": 1.4026402980989254e-06,
+      "loss": 1.0953,
+      "step": 6122
+    },
+    {
+      "epoch": 0.9391104294478527,
+      "grad_norm": 0.15210068225860596,
+      "learning_rate": 1.395622623578424e-06,
+      "loss": 0.9699,
+      "step": 6123
+    },
+    {
+      "epoch": 0.9392638036809816,
+      "grad_norm": 0.152539923787117,
+      "learning_rate": 1.3886223836137217e-06,
+      "loss": 0.9682,
+      "step": 6124
+    },
+    {
+      "epoch": 0.9394171779141104,
+      "grad_norm": 0.15124522149562836,
+      "learning_rate": 1.381639579862967e-06,
+      "loss": 1.0939,
+      "step": 6125
+    },
+    {
+      "epoch": 0.9395705521472393,
+      "grad_norm": 0.16662761569023132,
+      "learning_rate": 1.3746742139801453e-06,
+      "loss": 1.2025,
+      "step": 6126
+    },
+    {
+      "epoch": 0.9397239263803681,
+      "grad_norm": 0.14770177006721497,
+      "learning_rate": 1.3677262876151617e-06,
+      "loss": 0.9399,
+      "step": 6127
+    },
+    {
+      "epoch": 0.939877300613497,
+      "grad_norm": 0.14835749566555023,
+      "learning_rate": 1.360795802413725e-06,
+      "loss": 1.0649,
+      "step": 6128
+    },
+    {
+      "epoch": 0.9400306748466257,
+      "grad_norm": 0.1583913415670395,
+      "learning_rate": 1.3538827600174885e-06,
+      "loss": 1.0987,
+      "step": 6129
+    },
+    {
+      "epoch": 0.9401840490797546,
+      "grad_norm": 0.14769384264945984,
+      "learning_rate": 1.3469871620639006e-06,
+      "loss": 0.8373,
+      "step": 6130
+    },
+    {
+      "epoch": 0.9403374233128834,
+      "grad_norm": 0.1480213701725006,
+      "learning_rate": 1.3401090101863298e-06,
+      "loss": 1.0706,
+      "step": 6131
+    },
+    {
+      "epoch": 0.9404907975460123,
+      "grad_norm": 0.15560485422611237,
+      "learning_rate": 1.3332483060139814e-06,
+      "loss": 1.035,
+      "step": 6132
+    },
+    {
+      "epoch": 0.9406441717791411,
+      "grad_norm": 0.1476944088935852,
+      "learning_rate": 1.326405051171947e-06,
+      "loss": 0.9989,
+      "step": 6133
+    },
+    {
+      "epoch": 0.94079754601227,
+      "grad_norm": 0.15529006719589233,
+      "learning_rate": 1.31957924728118e-06,
+      "loss": 1.0234,
+      "step": 6134
+    },
+    {
+      "epoch": 0.9409509202453987,
+      "grad_norm": 0.15410137176513672,
+      "learning_rate": 1.3127708959584954e-06,
+      "loss": 1.0806,
+      "step": 6135
+    },
+    {
+      "epoch": 0.9411042944785276,
+      "grad_norm": 0.18097318708896637,
+      "learning_rate": 1.305979998816578e-06,
+      "loss": 1.2503,
+      "step": 6136
+    },
+    {
+      "epoch": 0.9412576687116564,
+      "grad_norm": 0.15022549033164978,
+      "learning_rate": 1.2992065574639747e-06,
+      "loss": 1.044,
+      "step": 6137
+    },
+    {
+      "epoch": 0.9414110429447853,
+      "grad_norm": 0.1544906497001648,
+      "learning_rate": 1.2924505735050939e-06,
+      "loss": 0.9616,
+      "step": 6138
+    },
+    {
+      "epoch": 0.9415644171779141,
+      "grad_norm": 0.1527995765209198,
+      "learning_rate": 1.28571204854023e-06,
+      "loss": 1.0364,
+      "step": 6139
+    },
+    {
+      "epoch": 0.941717791411043,
+      "grad_norm": 0.15569648146629333,
+      "learning_rate": 1.2789909841655067e-06,
+      "loss": 1.1301,
+      "step": 6140
+    },
+    {
+      "epoch": 0.9418711656441717,
+      "grad_norm": 0.15247412025928497,
+      "learning_rate": 1.2722873819729585e-06,
+      "loss": 0.962,
+      "step": 6141
+    },
+    {
+      "epoch": 0.9420245398773006,
+      "grad_norm": 0.14042535424232483,
+      "learning_rate": 1.2656012435504403e-06,
+      "loss": 1.0243,
+      "step": 6142
+    },
+    {
+      "epoch": 0.9421779141104294,
+      "grad_norm": 0.15317124128341675,
+      "learning_rate": 1.2589325704816938e-06,
+      "loss": 1.1308,
+      "step": 6143
+    },
+    {
+      "epoch": 0.9423312883435583,
+      "grad_norm": 0.15457367897033691,
+      "learning_rate": 1.2522813643463054e-06,
+      "loss": 1.0817,
+      "step": 6144
+    },
+    {
+      "epoch": 0.9424846625766872,
+      "grad_norm": 0.14899657666683197,
+      "learning_rate": 1.245647626719748e-06,
+      "loss": 1.0028,
+      "step": 6145
+    },
+    {
+      "epoch": 0.942638036809816,
+      "grad_norm": 0.1514570713043213,
+      "learning_rate": 1.239031359173348e-06,
+      "loss": 1.0547,
+      "step": 6146
+    },
+    {
+      "epoch": 0.9427914110429448,
+      "grad_norm": 0.145532488822937,
+      "learning_rate": 1.232432563274277e-06,
+      "loss": 1.0445,
+      "step": 6147
+    },
+    {
+      "epoch": 0.9429447852760736,
+      "grad_norm": 0.15610258281230927,
+      "learning_rate": 1.225851240585593e-06,
+      "loss": 1.0004,
+      "step": 6148
+    },
+    {
+      "epoch": 0.9430981595092025,
+      "grad_norm": 0.1502203643321991,
+      "learning_rate": 1.219287392666199e-06,
+      "loss": 1.0415,
+      "step": 6149
+    },
+    {
+      "epoch": 0.9432515337423313,
+      "grad_norm": 0.1378299444913864,
+      "learning_rate": 1.2127410210708677e-06,
+      "loss": 0.9891,
+      "step": 6150
+    },
+    {
+      "epoch": 0.9434049079754602,
+      "grad_norm": 0.16068792343139648,
+      "learning_rate": 1.2062121273502172e-06,
+      "loss": 1.1572,
+      "step": 6151
+    },
+    {
+      "epoch": 0.943558282208589,
+      "grad_norm": 0.15070787072181702,
+      "learning_rate": 1.1997007130507524e-06,
+      "loss": 1.0538,
+      "step": 6152
+    },
+    {
+      "epoch": 0.9437116564417178,
+      "grad_norm": 0.15376684069633484,
+      "learning_rate": 1.1932067797148143e-06,
+      "loss": 0.9054,
+      "step": 6153
+    },
+    {
+      "epoch": 0.9438650306748466,
+      "grad_norm": 0.164872944355011,
+      "learning_rate": 1.1867303288805974e-06,
+      "loss": 1.1722,
+      "step": 6154
+    },
+    {
+      "epoch": 0.9440184049079755,
+      "grad_norm": 0.16540972888469696,
+      "learning_rate": 1.1802713620821997e-06,
+      "loss": 1.096,
+      "step": 6155
+    },
+    {
+      "epoch": 0.9441717791411043,
+      "grad_norm": 0.15572814643383026,
+      "learning_rate": 1.173829880849514e-06,
+      "loss": 1.0549,
+      "step": 6156
+    },
+    {
+      "epoch": 0.9443251533742332,
+      "grad_norm": 0.15611490607261658,
+      "learning_rate": 1.1674058867083446e-06,
+      "loss": 1.0919,
+      "step": 6157
+    },
+    {
+      "epoch": 0.944478527607362,
+      "grad_norm": 0.14988277852535248,
+      "learning_rate": 1.1609993811803248e-06,
+      "loss": 0.9818,
+      "step": 6158
+    },
+    {
+      "epoch": 0.9446319018404908,
+      "grad_norm": 0.166375070810318,
+      "learning_rate": 1.154610365782957e-06,
+      "loss": 1.2246,
+      "step": 6159
+    },
+    {
+      "epoch": 0.9447852760736196,
+      "grad_norm": 0.15963999927043915,
+      "learning_rate": 1.1482388420295974e-06,
+      "loss": 1.0987,
+      "step": 6160
+    },
+    {
+      "epoch": 0.9449386503067485,
+      "grad_norm": 0.14946727454662323,
+      "learning_rate": 1.1418848114294553e-06,
+      "loss": 1.0431,
+      "step": 6161
+    },
+    {
+      "epoch": 0.9450920245398773,
+      "grad_norm": 0.16641531884670258,
+      "learning_rate": 1.1355482754876105e-06,
+      "loss": 1.1287,
+      "step": 6162
+    },
+    {
+      "epoch": 0.9452453987730062,
+      "grad_norm": 0.14432361721992493,
+      "learning_rate": 1.1292292357049705e-06,
+      "loss": 0.9005,
+      "step": 6163
+    },
+    {
+      "epoch": 0.945398773006135,
+      "grad_norm": 0.15901923179626465,
+      "learning_rate": 1.1229276935783376e-06,
+      "loss": 1.113,
+      "step": 6164
+    },
+    {
+      "epoch": 0.9455521472392638,
+      "grad_norm": 0.16408410668373108,
+      "learning_rate": 1.1166436506003352e-06,
+      "loss": 1.0024,
+      "step": 6165
+    },
+    {
+      "epoch": 0.9457055214723926,
+      "grad_norm": 0.16166144609451294,
+      "learning_rate": 1.1103771082594553e-06,
+      "loss": 0.9765,
+      "step": 6166
+    },
+    {
+      "epoch": 0.9458588957055215,
+      "grad_norm": 0.15553545951843262,
+      "learning_rate": 1.1041280680400444e-06,
+      "loss": 1.0502,
+      "step": 6167
+    },
+    {
+      "epoch": 0.9460122699386503,
+      "grad_norm": 0.15464439988136292,
+      "learning_rate": 1.0978965314223098e-06,
+      "loss": 0.9802,
+      "step": 6168
+    },
+    {
+      "epoch": 0.9461656441717792,
+      "grad_norm": 0.1528966724872589,
+      "learning_rate": 1.091682499882296e-06,
+      "loss": 1.0722,
+      "step": 6169
+    },
+    {
+      "epoch": 0.946319018404908,
+      "grad_norm": 0.16366887092590332,
+      "learning_rate": 1.085485974891917e-06,
+      "loss": 1.0961,
+      "step": 6170
+    },
+    {
+      "epoch": 0.9464723926380368,
+      "grad_norm": 0.15847529470920563,
+      "learning_rate": 1.0793069579189407e-06,
+      "loss": 1.129,
+      "step": 6171
+    },
+    {
+      "epoch": 0.9466257668711656,
+      "grad_norm": 0.15042929351329803,
+      "learning_rate": 1.0731454504269792e-06,
+      "loss": 1.1319,
+      "step": 6172
+    },
+    {
+      "epoch": 0.9467791411042945,
+      "grad_norm": 0.15789996087551117,
+      "learning_rate": 1.0670014538754822e-06,
+      "loss": 1.0875,
+      "step": 6173
+    },
+    {
+      "epoch": 0.9469325153374233,
+      "grad_norm": 0.16227252781391144,
+      "learning_rate": 1.0608749697197855e-06,
+      "loss": 1.026,
+      "step": 6174
+    },
+    {
+      "epoch": 0.9470858895705522,
+      "grad_norm": 0.1586507260799408,
+      "learning_rate": 1.054765999411053e-06,
+      "loss": 1.1428,
+      "step": 6175
+    },
+    {
+      "epoch": 0.947239263803681,
+      "grad_norm": 0.14976057410240173,
+      "learning_rate": 1.0486745443963106e-06,
+      "loss": 0.9792,
+      "step": 6176
+    },
+    {
+      "epoch": 0.9473926380368098,
+      "grad_norm": 0.15220226347446442,
+      "learning_rate": 1.0426006061184295e-06,
+      "loss": 1.0381,
+      "step": 6177
+    },
+    {
+      "epoch": 0.9475460122699386,
+      "grad_norm": 0.15265688300132751,
+      "learning_rate": 1.0365441860161416e-06,
+      "loss": 0.9792,
+      "step": 6178
+    },
+    {
+      "epoch": 0.9476993865030675,
+      "grad_norm": 0.15446622669696808,
+      "learning_rate": 1.0305052855240082e-06,
+      "loss": 1.1104,
+      "step": 6179
+    },
+    {
+      "epoch": 0.9478527607361963,
+      "grad_norm": 0.16018728911876678,
+      "learning_rate": 1.024483906072468e-06,
+      "loss": 1.0909,
+      "step": 6180
+    },
+    {
+      "epoch": 0.9480061349693252,
+      "grad_norm": 0.16581259667873383,
+      "learning_rate": 1.0184800490877886e-06,
+      "loss": 1.12,
+      "step": 6181
+    },
+    {
+      "epoch": 0.948159509202454,
+      "grad_norm": 0.14986376464366913,
+      "learning_rate": 1.012493715992091e-06,
+      "loss": 1.0818,
+      "step": 6182
+    },
+    {
+      "epoch": 0.9483128834355828,
+      "grad_norm": 0.1399035006761551,
+      "learning_rate": 1.0065249082033656e-06,
+      "loss": 0.8962,
+      "step": 6183
+    },
+    {
+      "epoch": 0.9484662576687116,
+      "grad_norm": 0.155815988779068,
+      "learning_rate": 1.0005736271354064e-06,
+      "loss": 1.0033,
+      "step": 6184
+    },
+    {
+      "epoch": 0.9486196319018405,
+      "grad_norm": 0.15815867483615875,
+      "learning_rate": 9.946398741979111e-07,
+      "loss": 1.1155,
+      "step": 6185
+    },
+    {
+      "epoch": 0.9487730061349693,
+      "grad_norm": 0.1630033403635025,
+      "learning_rate": 9.887236507963804e-07,
+      "loss": 1.1196,
+      "step": 6186
+    },
+    {
+      "epoch": 0.9489263803680982,
+      "grad_norm": 0.14352092146873474,
+      "learning_rate": 9.828249583321928e-07,
+      "loss": 0.984,
+      "step": 6187
+    },
+    {
+      "epoch": 0.949079754601227,
+      "grad_norm": 0.1608503758907318,
+      "learning_rate": 9.769437982025647e-07,
+      "loss": 1.1244,
+      "step": 6188
+    },
+    {
+      "epoch": 0.9492331288343558,
+      "grad_norm": 0.1624704897403717,
+      "learning_rate": 9.710801718005396e-07,
+      "loss": 1.0527,
+      "step": 6189
+    },
+    {
+      "epoch": 0.9493865030674846,
+      "grad_norm": 0.16882145404815674,
+      "learning_rate": 9.6523408051504e-07,
+      "loss": 1.1595,
+      "step": 6190
+    },
+    {
+      "epoch": 0.9495398773006135,
+      "grad_norm": 0.15219949185848236,
+      "learning_rate": 9.594055257308163e-07,
+      "loss": 1.0084,
+      "step": 6191
+    },
+    {
+      "epoch": 0.9496932515337423,
+      "grad_norm": 0.1459079533815384,
+      "learning_rate": 9.535945088284813e-07,
+      "loss": 0.9088,
+      "step": 6192
+    },
+    {
+      "epoch": 0.9498466257668712,
+      "grad_norm": 0.15652957558631897,
+      "learning_rate": 9.478010311844586e-07,
+      "loss": 1.0889,
+      "step": 6193
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.15421977639198303,
+      "learning_rate": 9.420250941710588e-07,
+      "loss": 1.2261,
+      "step": 6194
+    },
+    {
+      "epoch": 0.9501533742331288,
+      "grad_norm": 0.1574552208185196,
+      "learning_rate": 9.362666991564127e-07,
+      "loss": 0.9744,
+      "step": 6195
+    },
+    {
+      "epoch": 0.9503067484662576,
+      "grad_norm": 0.14896561205387115,
+      "learning_rate": 9.305258475044958e-07,
+      "loss": 1.0564,
+      "step": 6196
+    },
+    {
+      "epoch": 0.9504601226993865,
+      "grad_norm": 0.16929253935813904,
+      "learning_rate": 9.248025405751536e-07,
+      "loss": 1.2021,
+      "step": 6197
+    },
+    {
+      "epoch": 0.9506134969325153,
+      "grad_norm": 0.168552964925766,
+      "learning_rate": 9.190967797240351e-07,
+      "loss": 1.1562,
+      "step": 6198
+    },
+    {
+      "epoch": 0.9507668711656442,
+      "grad_norm": 0.15728260576725006,
+      "learning_rate": 9.134085663026757e-07,
+      "loss": 1.0295,
+      "step": 6199
+    },
+    {
+      "epoch": 0.950920245398773,
+      "grad_norm": 0.1616741269826889,
+      "learning_rate": 9.077379016584147e-07,
+      "loss": 1.142,
+      "step": 6200
+    },
+    {
+      "epoch": 0.9510736196319018,
+      "grad_norm": 0.1499522179365158,
+      "learning_rate": 9.020847871344605e-07,
+      "loss": 1.0143,
+      "step": 6201
+    },
+    {
+      "epoch": 0.9512269938650306,
+      "grad_norm": 0.15546908974647522,
+      "learning_rate": 8.964492240698672e-07,
+      "loss": 1.0511,
+      "step": 6202
+    },
+    {
+      "epoch": 0.9513803680981595,
+      "grad_norm": 0.15357044339179993,
+      "learning_rate": 8.908312137995088e-07,
+      "loss": 0.9903,
+      "step": 6203
+    },
+    {
+      "epoch": 0.9515337423312883,
+      "grad_norm": 0.17025436460971832,
+      "learning_rate": 8.852307576541207e-07,
+      "loss": 1.2418,
+      "step": 6204
+    },
+    {
+      "epoch": 0.9516871165644172,
+      "grad_norm": 0.1394793838262558,
+      "learning_rate": 8.79647856960275e-07,
+      "loss": 1.0177,
+      "step": 6205
+    },
+    {
+      "epoch": 0.951840490797546,
+      "grad_norm": 0.15590570867061615,
+      "learning_rate": 8.740825130403889e-07,
+      "loss": 0.9619,
+      "step": 6206
+    },
+    {
+      "epoch": 0.9519938650306748,
+      "grad_norm": 0.1594354212284088,
+      "learning_rate": 8.68534727212708e-07,
+      "loss": 1.026,
+      "step": 6207
+    },
+    {
+      "epoch": 0.9521472392638037,
+      "grad_norm": 0.15581004321575165,
+      "learning_rate": 8.630045007913311e-07,
+      "loss": 1.1197,
+      "step": 6208
+    },
+    {
+      "epoch": 0.9523006134969325,
+      "grad_norm": 0.1702142357826233,
+      "learning_rate": 8.574918350861937e-07,
+      "loss": 1.1807,
+      "step": 6209
+    },
+    {
+      "epoch": 0.9524539877300614,
+      "grad_norm": 0.15108083188533783,
+      "learning_rate": 8.519967314030763e-07,
+      "loss": 0.9955,
+      "step": 6210
+    },
+    {
+      "epoch": 0.9526073619631902,
+      "grad_norm": 0.14110465347766876,
+      "learning_rate": 8.465191910435958e-07,
+      "loss": 0.998,
+      "step": 6211
+    },
+    {
+      "epoch": 0.9527607361963191,
+      "grad_norm": 0.14918231964111328,
+      "learning_rate": 8.410592153051982e-07,
+      "loss": 0.904,
+      "step": 6212
+    },
+    {
+      "epoch": 0.9529141104294478,
+      "grad_norm": 0.1651293784379959,
+      "learning_rate": 8.356168054811902e-07,
+      "loss": 1.1682,
+      "step": 6213
+    },
+    {
+      "epoch": 0.9530674846625767,
+      "grad_norm": 0.14413373172283173,
+      "learning_rate": 8.301919628607073e-07,
+      "loss": 0.9177,
+      "step": 6214
+    },
+    {
+      "epoch": 0.9532208588957055,
+      "grad_norm": 0.1580570638179779,
+      "learning_rate": 8.247846887287218e-07,
+      "loss": 1.0159,
+      "step": 6215
+    },
+    {
+      "epoch": 0.9533742331288344,
+      "grad_norm": 0.1526566743850708,
+      "learning_rate": 8.193949843660425e-07,
+      "loss": 0.9157,
+      "step": 6216
+    },
+    {
+      "epoch": 0.9535276073619632,
+      "grad_norm": 0.14709758758544922,
+      "learning_rate": 8.14022851049323e-07,
+      "loss": 0.9691,
+      "step": 6217
+    },
+    {
+      "epoch": 0.9536809815950921,
+      "grad_norm": 0.16941197216510773,
+      "learning_rate": 8.086682900510538e-07,
+      "loss": 1.1798,
+      "step": 6218
+    },
+    {
+      "epoch": 0.9538343558282208,
+      "grad_norm": 0.14724084734916687,
+      "learning_rate": 8.033313026395539e-07,
+      "loss": 1.031,
+      "step": 6219
+    },
+    {
+      "epoch": 0.9539877300613497,
+      "grad_norm": 0.1633569896221161,
+      "learning_rate": 7.980118900790034e-07,
+      "loss": 1.1786,
+      "step": 6220
+    },
+    {
+      "epoch": 0.9541411042944785,
+      "grad_norm": 0.14959578216075897,
+      "learning_rate": 7.927100536293779e-07,
+      "loss": 1.0273,
+      "step": 6221
+    },
+    {
+      "epoch": 0.9542944785276074,
+      "grad_norm": 0.1504783183336258,
+      "learning_rate": 7.874257945465478e-07,
+      "loss": 1.0277,
+      "step": 6222
+    },
+    {
+      "epoch": 0.9544478527607362,
+      "grad_norm": 0.15919551253318787,
+      "learning_rate": 7.82159114082162e-07,
+      "loss": 1.112,
+      "step": 6223
+    },
+    {
+      "epoch": 0.9546012269938651,
+      "grad_norm": 0.15358926355838776,
+      "learning_rate": 7.769100134837392e-07,
+      "loss": 0.8952,
+      "step": 6224
+    },
+    {
+      "epoch": 0.9547546012269938,
+      "grad_norm": 0.14281798899173737,
+      "learning_rate": 7.716784939946347e-07,
+      "loss": 0.9863,
+      "step": 6225
+    },
+    {
+      "epoch": 0.9549079754601227,
+      "grad_norm": 0.14543578028678894,
+      "learning_rate": 7.664645568540074e-07,
+      "loss": 0.9258,
+      "step": 6226
+    },
+    {
+      "epoch": 0.9550613496932515,
+      "grad_norm": 0.1470409333705902,
+      "learning_rate": 7.612682032969025e-07,
+      "loss": 0.9116,
+      "step": 6227
+    },
+    {
+      "epoch": 0.9552147239263804,
+      "grad_norm": 0.18287567794322968,
+      "learning_rate": 7.560894345541524e-07,
+      "loss": 1.0938,
+      "step": 6228
+    },
+    {
+      "epoch": 0.9553680981595092,
+      "grad_norm": 0.16119332611560822,
+      "learning_rate": 7.509282518524506e-07,
+      "loss": 0.9863,
+      "step": 6229
+    },
+    {
+      "epoch": 0.9555214723926381,
+      "grad_norm": 0.15084701776504517,
+      "learning_rate": 7.457846564143194e-07,
+      "loss": 0.939,
+      "step": 6230
+    },
+    {
+      "epoch": 0.9556748466257668,
+      "grad_norm": 0.152459517121315,
+      "learning_rate": 7.406586494581091e-07,
+      "loss": 0.975,
+      "step": 6231
+    },
+    {
+      "epoch": 0.9558282208588957,
+      "grad_norm": 0.15429134666919708,
+      "learning_rate": 7.355502321980234e-07,
+      "loss": 1.0374,
+      "step": 6232
+    },
+    {
+      "epoch": 0.9559815950920245,
+      "grad_norm": 0.15740075707435608,
+      "learning_rate": 7.304594058440611e-07,
+      "loss": 1.0781,
+      "step": 6233
+    },
+    {
+      "epoch": 0.9561349693251534,
+      "grad_norm": 0.15175439417362213,
+      "learning_rate": 7.253861716021075e-07,
+      "loss": 0.9435,
+      "step": 6234
+    },
+    {
+      "epoch": 0.9562883435582822,
+      "grad_norm": 0.15691116452217102,
+      "learning_rate": 7.203305306738266e-07,
+      "loss": 1.0851,
+      "step": 6235
+    },
+    {
+      "epoch": 0.9564417177914111,
+      "grad_norm": 0.16317540407180786,
+      "learning_rate": 7.15292484256752e-07,
+      "loss": 1.1384,
+      "step": 6236
+    },
+    {
+      "epoch": 0.9565950920245399,
+      "grad_norm": 0.1567709594964981,
+      "learning_rate": 7.102720335442458e-07,
+      "loss": 1.1234,
+      "step": 6237
+    },
+    {
+      "epoch": 0.9567484662576687,
+      "grad_norm": 0.1542871743440628,
+      "learning_rate": 7.052691797254734e-07,
+      "loss": 0.9812,
+      "step": 6238
+    },
+    {
+      "epoch": 0.9569018404907975,
+      "grad_norm": 0.1617075651884079,
+      "learning_rate": 7.002839239854702e-07,
+      "loss": 1.1155,
+      "step": 6239
+    },
+    {
+      "epoch": 0.9570552147239264,
+      "grad_norm": 0.15022258460521698,
+      "learning_rate": 6.953162675050749e-07,
+      "loss": 1.0748,
+      "step": 6240
+    },
+    {
+      "epoch": 0.9572085889570552,
+      "grad_norm": 0.13938675820827484,
+      "learning_rate": 6.903662114609882e-07,
+      "loss": 0.8692,
+      "step": 6241
+    },
+    {
+      "epoch": 0.9573619631901841,
+      "grad_norm": 0.15437892079353333,
+      "learning_rate": 6.854337570256968e-07,
+      "loss": 0.9516,
+      "step": 6242
+    },
+    {
+      "epoch": 0.9575153374233129,
+      "grad_norm": 0.15136757493019104,
+      "learning_rate": 6.80518905367558e-07,
+      "loss": 1.0935,
+      "step": 6243
+    },
+    {
+      "epoch": 0.9576687116564417,
+      "grad_norm": 0.1706974059343338,
+      "learning_rate": 6.756216576507406e-07,
+      "loss": 1.2285,
+      "step": 6244
+    },
+    {
+      "epoch": 0.9578220858895705,
+      "grad_norm": 0.16213113069534302,
+      "learning_rate": 6.7074201503525e-07,
+      "loss": 0.9961,
+      "step": 6245
+    },
+    {
+      "epoch": 0.9579754601226994,
+      "grad_norm": 0.16852495074272156,
+      "learning_rate": 6.658799786769198e-07,
+      "loss": 1.0756,
+      "step": 6246
+    },
+    {
+      "epoch": 0.9581288343558282,
+      "grad_norm": 0.15730877220630646,
+      "learning_rate": 6.610355497274206e-07,
+      "loss": 1.0055,
+      "step": 6247
+    },
+    {
+      "epoch": 0.9582822085889571,
+      "grad_norm": 0.1704762876033783,
+      "learning_rate": 6.562087293342344e-07,
+      "loss": 1.1969,
+      "step": 6248
+    },
+    {
+      "epoch": 0.9584355828220859,
+      "grad_norm": 0.16665096580982208,
+      "learning_rate": 6.513995186406801e-07,
+      "loss": 1.0818,
+      "step": 6249
+    },
+    {
+      "epoch": 0.9585889570552147,
+      "grad_norm": 0.16329354047775269,
+      "learning_rate": 6.46607918785913e-07,
+      "loss": 1.0539,
+      "step": 6250
+    },
+    {
+      "epoch": 0.9587423312883435,
+      "grad_norm": 0.15198612213134766,
+      "learning_rate": 6.418339309049253e-07,
+      "loss": 1.0139,
+      "step": 6251
+    },
+    {
+      "epoch": 0.9588957055214724,
+      "grad_norm": 0.15468253195285797,
+      "learning_rate": 6.370775561284958e-07,
+      "loss": 1.0256,
+      "step": 6252
+    },
+    {
+      "epoch": 0.9590490797546012,
+      "grad_norm": 0.16485030949115753,
+      "learning_rate": 6.323387955832815e-07,
+      "loss": 1.1208,
+      "step": 6253
+    },
+    {
+      "epoch": 0.9592024539877301,
+      "grad_norm": 0.14800235629081726,
+      "learning_rate": 6.276176503917429e-07,
+      "loss": 1.0767,
+      "step": 6254
+    },
+    {
+      "epoch": 0.9593558282208589,
+      "grad_norm": 0.1444309800863266,
+      "learning_rate": 6.229141216721606e-07,
+      "loss": 0.9025,
+      "step": 6255
+    },
+    {
+      "epoch": 0.9595092024539877,
+      "grad_norm": 0.1661992073059082,
+      "learning_rate": 6.182282105386599e-07,
+      "loss": 1.1986,
+      "step": 6256
+    },
+    {
+      "epoch": 0.9596625766871165,
+      "grad_norm": 0.1498861461877823,
+      "learning_rate": 6.135599181011864e-07,
+      "loss": 1.0552,
+      "step": 6257
+    },
+    {
+      "epoch": 0.9598159509202454,
+      "grad_norm": 0.14205828309059143,
+      "learning_rate": 6.089092454655137e-07,
+      "loss": 1.0133,
+      "step": 6258
+    },
+    {
+      "epoch": 0.9599693251533742,
+      "grad_norm": 0.1576204001903534,
+      "learning_rate": 6.042761937332274e-07,
+      "loss": 1.1132,
+      "step": 6259
+    },
+    {
+      "epoch": 0.9601226993865031,
+      "grad_norm": 0.14713168144226074,
+      "learning_rate": 5.996607640017582e-07,
+      "loss": 1.0705,
+      "step": 6260
+    },
+    {
+      "epoch": 0.9602760736196319,
+      "grad_norm": 0.14991633594036102,
+      "learning_rate": 5.950629573643562e-07,
+      "loss": 0.9593,
+      "step": 6261
+    },
+    {
+      "epoch": 0.9604294478527607,
+      "grad_norm": 0.15325503051280975,
+      "learning_rate": 5.904827749101005e-07,
+      "loss": 1.0037,
+      "step": 6262
+    },
+    {
+      "epoch": 0.9605828220858895,
+      "grad_norm": 0.14655537903308868,
+      "learning_rate": 5.859202177238814e-07,
+      "loss": 0.9462,
+      "step": 6263
+    },
+    {
+      "epoch": 0.9607361963190184,
+      "grad_norm": 0.17026782035827637,
+      "learning_rate": 5.813752868864347e-07,
+      "loss": 1.071,
+      "step": 6264
+    },
+    {
+      "epoch": 0.9608895705521472,
+      "grad_norm": 0.15463854372501373,
+      "learning_rate": 5.768479834743156e-07,
+      "loss": 1.0214,
+      "step": 6265
+    },
+    {
+      "epoch": 0.9610429447852761,
+      "grad_norm": 0.1529400646686554,
+      "learning_rate": 5.723383085598831e-07,
+      "loss": 1.0113,
+      "step": 6266
+    },
+    {
+      "epoch": 0.9611963190184049,
+      "grad_norm": 0.1647195667028427,
+      "learning_rate": 5.678462632113495e-07,
+      "loss": 1.118,
+      "step": 6267
+    },
+    {
+      "epoch": 0.9613496932515337,
+      "grad_norm": 0.14383739233016968,
+      "learning_rate": 5.633718484927302e-07,
+      "loss": 0.9921,
+      "step": 6268
+    },
+    {
+      "epoch": 0.9615030674846625,
+      "grad_norm": 0.14452727138996124,
+      "learning_rate": 5.589150654638857e-07,
+      "loss": 0.9807,
+      "step": 6269
+    },
+    {
+      "epoch": 0.9616564417177914,
+      "grad_norm": 0.15858685970306396,
+      "learning_rate": 5.544759151804717e-07,
+      "loss": 1.0883,
+      "step": 6270
+    },
+    {
+      "epoch": 0.9618098159509203,
+      "grad_norm": 0.15765677392482758,
+      "learning_rate": 5.500543986939887e-07,
+      "loss": 1.0354,
+      "step": 6271
+    },
+    {
+      "epoch": 0.9619631901840491,
+      "grad_norm": 0.1515926718711853,
+      "learning_rate": 5.456505170517656e-07,
+      "loss": 0.9809,
+      "step": 6272
+    },
+    {
+      "epoch": 0.962116564417178,
+      "grad_norm": 0.1667451411485672,
+      "learning_rate": 5.412642712969261e-07,
+      "loss": 1.0895,
+      "step": 6273
+    },
+    {
+      "epoch": 0.9622699386503067,
+      "grad_norm": 0.14338566362857819,
+      "learning_rate": 5.368956624684478e-07,
+      "loss": 0.9473,
+      "step": 6274
+    },
+    {
+      "epoch": 0.9624233128834356,
+      "grad_norm": 0.1562606394290924,
+      "learning_rate": 5.325446916011028e-07,
+      "loss": 0.9471,
+      "step": 6275
+    },
+    {
+      "epoch": 0.9625766871165644,
+      "grad_norm": 0.14219145476818085,
+      "learning_rate": 5.282113597255083e-07,
+      "loss": 1.0144,
+      "step": 6276
+    },
+    {
+      "epoch": 0.9627300613496933,
+      "grad_norm": 0.15869639813899994,
+      "learning_rate": 5.238956678680933e-07,
+      "loss": 1.0277,
+      "step": 6277
+    },
+    {
+      "epoch": 0.9628834355828221,
+      "grad_norm": 0.15016284584999084,
+      "learning_rate": 5.195976170511068e-07,
+      "loss": 1.0194,
+      "step": 6278
+    },
+    {
+      "epoch": 0.963036809815951,
+      "grad_norm": 0.1383758932352066,
+      "learning_rate": 5.153172082926177e-07,
+      "loss": 0.833,
+      "step": 6279
+    },
+    {
+      "epoch": 0.9631901840490797,
+      "grad_norm": 0.15268322825431824,
+      "learning_rate": 5.110544426065316e-07,
+      "loss": 1.041,
+      "step": 6280
+    },
+    {
+      "epoch": 0.9633435582822086,
+      "grad_norm": 0.16171841323375702,
+      "learning_rate": 5.068093210025575e-07,
+      "loss": 1.0688,
+      "step": 6281
+    },
+    {
+      "epoch": 0.9634969325153374,
+      "grad_norm": 0.14937566220760345,
+      "learning_rate": 5.025818444862245e-07,
+      "loss": 1.1285,
+      "step": 6282
+    },
+    {
+      "epoch": 0.9636503067484663,
+      "grad_norm": 0.1552608460187912,
+      "learning_rate": 4.983720140588898e-07,
+      "loss": 0.9196,
+      "step": 6283
+    },
+    {
+      "epoch": 0.9638036809815951,
+      "grad_norm": 0.16508065164089203,
+      "learning_rate": 4.941798307177392e-07,
+      "loss": 1.1415,
+      "step": 6284
+    },
+    {
+      "epoch": 0.963957055214724,
+      "grad_norm": 0.14980602264404297,
+      "learning_rate": 4.900052954557615e-07,
+      "loss": 0.9355,
+      "step": 6285
+    },
+    {
+      "epoch": 0.9641104294478527,
+      "grad_norm": 0.1537647545337677,
+      "learning_rate": 4.858484092617659e-07,
+      "loss": 1.0365,
+      "step": 6286
+    },
+    {
+      "epoch": 0.9642638036809816,
+      "grad_norm": 0.1670050024986267,
+      "learning_rate": 4.817091731203981e-07,
+      "loss": 1.1124,
+      "step": 6287
+    },
+    {
+      "epoch": 0.9644171779141104,
+      "grad_norm": 0.19447560608386993,
+      "learning_rate": 4.775875880121155e-07,
+      "loss": 0.9659,
+      "step": 6288
+    },
+    {
+      "epoch": 0.9645705521472393,
+      "grad_norm": 0.1493082046508789,
+      "learning_rate": 4.7348365491317875e-07,
+      "loss": 0.9516,
+      "step": 6289
+    },
+    {
+      "epoch": 0.9647239263803681,
+      "grad_norm": 0.1454736590385437,
+      "learning_rate": 4.693973747956853e-07,
+      "loss": 0.9002,
+      "step": 6290
+    },
+    {
+      "epoch": 0.964877300613497,
+      "grad_norm": 0.15198495984077454,
+      "learning_rate": 4.6532874862755253e-07,
+      "loss": 1.0307,
+      "step": 6291
+    },
+    {
+      "epoch": 0.9650306748466257,
+      "grad_norm": 0.15156108140945435,
+      "learning_rate": 4.612777773725096e-07,
+      "loss": 0.9288,
+      "step": 6292
+    },
+    {
+      "epoch": 0.9651840490797546,
+      "grad_norm": 0.15597659349441528,
+      "learning_rate": 4.5724446199009725e-07,
+      "loss": 1.0268,
+      "step": 6293
+    },
+    {
+      "epoch": 0.9653374233128834,
+      "grad_norm": 0.15130852162837982,
+      "learning_rate": 4.5322880343567626e-07,
+      "loss": 0.983,
+      "step": 6294
+    },
+    {
+      "epoch": 0.9654907975460123,
+      "grad_norm": 0.14995364844799042,
+      "learning_rate": 4.492308026604441e-07,
+      "loss": 1.0072,
+      "step": 6295
+    },
+    {
+      "epoch": 0.9656441717791411,
+      "grad_norm": 0.15639200806617737,
+      "learning_rate": 4.452504606113849e-07,
+      "loss": 1.0934,
+      "step": 6296
+    },
+    {
+      "epoch": 0.96579754601227,
+      "grad_norm": 0.1561667025089264,
+      "learning_rate": 4.4128777823132786e-07,
+      "loss": 1.0801,
+      "step": 6297
+    },
+    {
+      "epoch": 0.9659509202453987,
+      "grad_norm": 0.14021354913711548,
+      "learning_rate": 4.3734275645890547e-07,
+      "loss": 0.8554,
+      "step": 6298
+    },
+    {
+      "epoch": 0.9661042944785276,
+      "grad_norm": 0.1616344302892685,
+      "learning_rate": 4.3341539622856194e-07,
+      "loss": 1.0752,
+      "step": 6299
+    },
+    {
+      "epoch": 0.9662576687116564,
+      "grad_norm": 0.13583676517009735,
+      "learning_rate": 4.2950569847057805e-07,
+      "loss": 0.8658,
+      "step": 6300
+    },
+    {
+      "epoch": 0.9664110429447853,
+      "grad_norm": 0.160183846950531,
+      "learning_rate": 4.2561366411102973e-07,
+      "loss": 1.0511,
+      "step": 6301
+    },
+    {
+      "epoch": 0.9665644171779141,
+      "grad_norm": 0.15687982738018036,
+      "learning_rate": 4.217392940718128e-07,
+      "loss": 1.0212,
+      "step": 6302
+    },
+    {
+      "epoch": 0.966717791411043,
+      "grad_norm": 0.1574207842350006,
+      "learning_rate": 4.1788258927065165e-07,
+      "loss": 1.0424,
+      "step": 6303
+    },
+    {
+      "epoch": 0.9668711656441717,
+      "grad_norm": 0.16007095575332642,
+      "learning_rate": 4.1404355062107376e-07,
+      "loss": 1.0816,
+      "step": 6304
+    },
+    {
+      "epoch": 0.9670245398773006,
+      "grad_norm": 0.15194804966449738,
+      "learning_rate": 4.102221790324184e-07,
+      "loss": 1.0276,
+      "step": 6305
+    },
+    {
+      "epoch": 0.9671779141104294,
+      "grad_norm": 0.14737242460250854,
+      "learning_rate": 4.064184754098698e-07,
+      "loss": 1.023,
+      "step": 6306
+    },
+    {
+      "epoch": 0.9673312883435583,
+      "grad_norm": 0.14269758760929108,
+      "learning_rate": 4.026324406543907e-07,
+      "loss": 1.0786,
+      "step": 6307
+    },
+    {
+      "epoch": 0.9674846625766871,
+      "grad_norm": 0.14846524596214294,
+      "learning_rate": 3.988640756627637e-07,
+      "loss": 1.0193,
+      "step": 6308
+    },
+    {
+      "epoch": 0.967638036809816,
+      "grad_norm": 0.1721796989440918,
+      "learning_rate": 3.951133813276164e-07,
+      "loss": 1.1234,
+      "step": 6309
+    },
+    {
+      "epoch": 0.9677914110429447,
+      "grad_norm": 0.1522304117679596,
+      "learning_rate": 3.913803585373632e-07,
+      "loss": 1.0622,
+      "step": 6310
+    },
+    {
+      "epoch": 0.9679447852760736,
+      "grad_norm": 0.15822382271289825,
+      "learning_rate": 3.8766500817623005e-07,
+      "loss": 1.1457,
+      "step": 6311
+    },
+    {
+      "epoch": 0.9680981595092024,
+      "grad_norm": 0.15415138006210327,
+      "learning_rate": 3.8396733112427135e-07,
+      "loss": 1.0103,
+      "step": 6312
+    },
+    {
+      "epoch": 0.9682515337423313,
+      "grad_norm": 0.15216194093227386,
+      "learning_rate": 3.802873282573615e-07,
+      "loss": 1.0087,
+      "step": 6313
+    },
+    {
+      "epoch": 0.9684049079754601,
+      "grad_norm": 0.15336623787879944,
+      "learning_rate": 3.766250004471616e-07,
+      "loss": 1.0803,
+      "step": 6314
+    },
+    {
+      "epoch": 0.968558282208589,
+      "grad_norm": 0.1601058840751648,
+      "learning_rate": 3.729803485611693e-07,
+      "loss": 1.0862,
+      "step": 6315
+    },
+    {
+      "epoch": 0.9687116564417177,
+      "grad_norm": 0.14905840158462524,
+      "learning_rate": 3.693533734626858e-07,
+      "loss": 0.9749,
+      "step": 6316
+    },
+    {
+      "epoch": 0.9688650306748466,
+      "grad_norm": 0.14991194009780884,
+      "learning_rate": 3.6574407601082376e-07,
+      "loss": 1.0814,
+      "step": 6317
+    },
+    {
+      "epoch": 0.9690184049079754,
+      "grad_norm": 0.15434548258781433,
+      "learning_rate": 3.6215245706051607e-07,
+      "loss": 1.0674,
+      "step": 6318
+    },
+    {
+      "epoch": 0.9691717791411043,
+      "grad_norm": 0.1576557457447052,
+      "learning_rate": 3.585785174625072e-07,
+      "loss": 1.0664,
+      "step": 6319
+    },
+    {
+      "epoch": 0.9693251533742331,
+      "grad_norm": 0.1617213785648346,
+      "learning_rate": 3.5502225806334493e-07,
+      "loss": 1.143,
+      "step": 6320
+    },
+    {
+      "epoch": 0.969478527607362,
+      "grad_norm": 0.1510230451822281,
+      "learning_rate": 3.5148367970539715e-07,
+      "loss": 0.9718,
+      "step": 6321
+    },
+    {
+      "epoch": 0.9696319018404908,
+      "grad_norm": 0.16138926148414612,
+      "learning_rate": 3.4796278322684335e-07,
+      "loss": 1.2128,
+      "step": 6322
+    },
+    {
+      "epoch": 0.9697852760736196,
+      "grad_norm": 0.16342595219612122,
+      "learning_rate": 3.4445956946166644e-07,
+      "loss": 0.9849,
+      "step": 6323
+    },
+    {
+      "epoch": 0.9699386503067484,
+      "grad_norm": 0.1558510661125183,
+      "learning_rate": 3.4097403923966925e-07,
+      "loss": 0.9907,
+      "step": 6324
+    },
+    {
+      "epoch": 0.9700920245398773,
+      "grad_norm": 0.15122804045677185,
+      "learning_rate": 3.375061933864748e-07,
+      "loss": 1.0724,
+      "step": 6325
+    },
+    {
+      "epoch": 0.9702453987730061,
+      "grad_norm": 0.15866290032863617,
+      "learning_rate": 3.340560327234926e-07,
+      "loss": 1.0027,
+      "step": 6326
+    },
+    {
+      "epoch": 0.970398773006135,
+      "grad_norm": 0.15495622158050537,
+      "learning_rate": 3.306235580679606e-07,
+      "loss": 1.1513,
+      "step": 6327
+    },
+    {
+      "epoch": 0.9705521472392638,
+      "grad_norm": 0.16197198629379272,
+      "learning_rate": 3.2720877023292846e-07,
+      "loss": 1.1438,
+      "step": 6328
+    },
+    {
+      "epoch": 0.9707055214723926,
+      "grad_norm": 0.1490265130996704,
+      "learning_rate": 3.2381167002724085e-07,
+      "loss": 1.036,
+      "step": 6329
+    },
+    {
+      "epoch": 0.9708588957055214,
+      "grad_norm": 0.14421436190605164,
+      "learning_rate": 3.204322582555874e-07,
+      "loss": 1.0487,
+      "step": 6330
+    },
+    {
+      "epoch": 0.9710122699386503,
+      "grad_norm": 0.1506464034318924,
+      "learning_rate": 3.170705357184111e-07,
+      "loss": 0.9426,
+      "step": 6331
+    },
+    {
+      "epoch": 0.9711656441717792,
+      "grad_norm": 0.1419663280248642,
+      "learning_rate": 3.1372650321202506e-07,
+      "loss": 0.9,
+      "step": 6332
+    },
+    {
+      "epoch": 0.971319018404908,
+      "grad_norm": 0.14421413838863373,
+      "learning_rate": 3.104001615285123e-07,
+      "loss": 0.8921,
+      "step": 6333
+    },
+    {
+      "epoch": 0.9714723926380369,
+      "grad_norm": 0.15304036438465118,
+      "learning_rate": 3.0709151145578425e-07,
+      "loss": 1.0546,
+      "step": 6334
+    },
+    {
+      "epoch": 0.9716257668711656,
+      "grad_norm": 0.15888525545597076,
+      "learning_rate": 3.0380055377755565e-07,
+      "loss": 1.0411,
+      "step": 6335
+    },
+    {
+      "epoch": 0.9717791411042945,
+      "grad_norm": 0.1402287632226944,
+      "learning_rate": 3.005272892733362e-07,
+      "loss": 0.9304,
+      "step": 6336
+    },
+    {
+      "epoch": 0.9719325153374233,
+      "grad_norm": 0.1588635891675949,
+      "learning_rate": 2.972717187184809e-07,
+      "loss": 1.1176,
+      "step": 6337
+    },
+    {
+      "epoch": 0.9720858895705522,
+      "grad_norm": 0.1643708199262619,
+      "learning_rate": 2.9403384288411436e-07,
+      "loss": 1.1976,
+      "step": 6338
+    },
+    {
+      "epoch": 0.972239263803681,
+      "grad_norm": 0.16195148229599,
+      "learning_rate": 2.908136625372065e-07,
+      "loss": 1.0937,
+      "step": 6339
+    },
+    {
+      "epoch": 0.9723926380368099,
+      "grad_norm": 0.15032362937927246,
+      "learning_rate": 2.8761117844048885e-07,
+      "loss": 1.0437,
+      "step": 6340
+    },
+    {
+      "epoch": 0.9725460122699386,
+      "grad_norm": 0.15755504369735718,
+      "learning_rate": 2.8442639135255464e-07,
+      "loss": 1.0282,
+      "step": 6341
+    },
+    {
+      "epoch": 0.9726993865030675,
+      "grad_norm": 0.1447463184595108,
+      "learning_rate": 2.8125930202776704e-07,
+      "loss": 0.9824,
+      "step": 6342
+    },
+    {
+      "epoch": 0.9728527607361963,
+      "grad_norm": 0.15043294429779053,
+      "learning_rate": 2.7810991121630933e-07,
+      "loss": 1.0456,
+      "step": 6343
+    },
+    {
+      "epoch": 0.9730061349693252,
+      "grad_norm": 0.1753837913274765,
+      "learning_rate": 2.749782196641764e-07,
+      "loss": 1.112,
+      "step": 6344
+    },
+    {
+      "epoch": 0.973159509202454,
+      "grad_norm": 0.15246742963790894,
+      "learning_rate": 2.7186422811316666e-07,
+      "loss": 1.0215,
+      "step": 6345
+    },
+    {
+      "epoch": 0.9733128834355829,
+      "grad_norm": 0.173289492726326,
+      "learning_rate": 2.6876793730088164e-07,
+      "loss": 1.0712,
+      "step": 6346
+    },
+    {
+      "epoch": 0.9734662576687116,
+      "grad_norm": 0.15385347604751587,
+      "learning_rate": 2.6568934796074305e-07,
+      "loss": 1.0246,
+      "step": 6347
+    },
+    {
+      "epoch": 0.9736196319018405,
+      "grad_norm": 0.15060555934906006,
+      "learning_rate": 2.6262846082196754e-07,
+      "loss": 0.9054,
+      "step": 6348
+    },
+    {
+      "epoch": 0.9737730061349693,
+      "grad_norm": 0.14608028531074524,
+      "learning_rate": 2.595852766095835e-07,
+      "loss": 1.0357,
+      "step": 6349
+    },
+    {
+      "epoch": 0.9739263803680982,
+      "grad_norm": 0.1558111011981964,
+      "learning_rate": 2.5655979604442266e-07,
+      "loss": 0.988,
+      "step": 6350
+    },
+    {
+      "epoch": 0.974079754601227,
+      "grad_norm": 0.17531879246234894,
+      "learning_rate": 2.5355201984313676e-07,
+      "loss": 1.1512,
+      "step": 6351
+    },
+    {
+      "epoch": 0.9742331288343559,
+      "grad_norm": 0.14728225767612457,
+      "learning_rate": 2.5056194871815595e-07,
+      "loss": 1.0147,
+      "step": 6352
+    },
+    {
+      "epoch": 0.9743865030674846,
+      "grad_norm": 0.15108294785022736,
+      "learning_rate": 2.475895833777469e-07,
+      "loss": 0.9682,
+      "step": 6353
+    },
+    {
+      "epoch": 0.9745398773006135,
+      "grad_norm": 0.15737317502498627,
+      "learning_rate": 2.446349245259549e-07,
+      "loss": 1.115,
+      "step": 6354
+    },
+    {
+      "epoch": 0.9746932515337423,
+      "grad_norm": 0.154462993144989,
+      "learning_rate": 2.4169797286267e-07,
+      "loss": 1.0476,
+      "step": 6355
+    },
+    {
+      "epoch": 0.9748466257668712,
+      "grad_norm": 0.14202247560024261,
+      "learning_rate": 2.387787290835358e-07,
+      "loss": 0.9451,
+      "step": 6356
+    },
+    {
+      "epoch": 0.975,
+      "grad_norm": 0.15064360201358795,
+      "learning_rate": 2.358771938800491e-07,
+      "loss": 1.0297,
+      "step": 6357
+    },
+    {
+      "epoch": 0.9751533742331289,
+      "grad_norm": 0.1533876210451126,
+      "learning_rate": 2.329933679394852e-07,
+      "loss": 0.9818,
+      "step": 6358
+    },
+    {
+      "epoch": 0.9753067484662576,
+      "grad_norm": 0.17058703303337097,
+      "learning_rate": 2.3012725194493097e-07,
+      "loss": 1.1752,
+      "step": 6359
+    },
+    {
+      "epoch": 0.9754601226993865,
+      "grad_norm": 0.14673443138599396,
+      "learning_rate": 2.2727884657527672e-07,
+      "loss": 1.1008,
+      "step": 6360
+    },
+    {
+      "epoch": 0.9756134969325153,
+      "grad_norm": 0.14420919120311737,
+      "learning_rate": 2.2444815250521608e-07,
+      "loss": 1.0036,
+      "step": 6361
+    },
+    {
+      "epoch": 0.9757668711656442,
+      "grad_norm": 0.1582859754562378,
+      "learning_rate": 2.21635170405271e-07,
+      "loss": 1.0495,
+      "step": 6362
+    },
+    {
+      "epoch": 0.975920245398773,
+      "grad_norm": 0.15785516798496246,
+      "learning_rate": 2.1883990094172522e-07,
+      "loss": 1.1099,
+      "step": 6363
+    },
+    {
+      "epoch": 0.9760736196319019,
+      "grad_norm": 0.14717921614646912,
+      "learning_rate": 2.1606234477670737e-07,
+      "loss": 0.9821,
+      "step": 6364
+    },
+    {
+      "epoch": 0.9762269938650306,
+      "grad_norm": 0.15523101389408112,
+      "learning_rate": 2.133025025681162e-07,
+      "loss": 1.0932,
+      "step": 6365
+    },
+    {
+      "epoch": 0.9763803680981595,
+      "grad_norm": 0.15908773243427277,
+      "learning_rate": 2.1056037496968714e-07,
+      "loss": 1.0595,
+      "step": 6366
+    },
+    {
+      "epoch": 0.9765337423312883,
+      "grad_norm": 0.14644497632980347,
+      "learning_rate": 2.0783596263093393e-07,
+      "loss": 1.0236,
+      "step": 6367
+    },
+    {
+      "epoch": 0.9766871165644172,
+      "grad_norm": 0.16142764687538147,
+      "learning_rate": 2.0512926619718207e-07,
+      "loss": 1.0339,
+      "step": 6368
+    },
+    {
+      "epoch": 0.976840490797546,
+      "grad_norm": 0.15123902261257172,
+      "learning_rate": 2.0244028630956866e-07,
+      "loss": 1.0255,
+      "step": 6369
+    },
+    {
+      "epoch": 0.9769938650306749,
+      "grad_norm": 0.16239862143993378,
+      "learning_rate": 1.9976902360503423e-07,
+      "loss": 1.0735,
+      "step": 6370
+    },
+    {
+      "epoch": 0.9771472392638036,
+      "grad_norm": 0.15784195065498352,
+      "learning_rate": 1.9711547871630595e-07,
+      "loss": 1.0777,
+      "step": 6371
+    },
+    {
+      "epoch": 0.9773006134969325,
+      "grad_norm": 0.1560586839914322,
+      "learning_rate": 1.9447965227192274e-07,
+      "loss": 1.0446,
+      "step": 6372
+    },
+    {
+      "epoch": 0.9774539877300613,
+      "grad_norm": 0.1510692685842514,
+      "learning_rate": 1.9186154489623517e-07,
+      "loss": 0.9718,
+      "step": 6373
+    },
+    {
+      "epoch": 0.9776073619631902,
+      "grad_norm": 0.14936463534832,
+      "learning_rate": 1.8926115720938882e-07,
+      "loss": 1.0384,
+      "step": 6374
+    },
+    {
+      "epoch": 0.977760736196319,
+      "grad_norm": 0.15005002915859222,
+      "learning_rate": 1.8667848982734102e-07,
+      "loss": 0.9734,
+      "step": 6375
+    },
+    {
+      "epoch": 0.9779141104294479,
+      "grad_norm": 0.16044670343399048,
+      "learning_rate": 1.841135433618357e-07,
+      "loss": 1.056,
+      "step": 6376
+    },
+    {
+      "epoch": 0.9780674846625766,
+      "grad_norm": 0.16425248980522156,
+      "learning_rate": 1.8156631842042858e-07,
+      "loss": 1.144,
+      "step": 6377
+    },
+    {
+      "epoch": 0.9782208588957055,
+      "grad_norm": 0.17038680613040924,
+      "learning_rate": 1.7903681560647033e-07,
+      "loss": 1.1314,
+      "step": 6378
+    },
+    {
+      "epoch": 0.9783742331288343,
+      "grad_norm": 0.14835049211978912,
+      "learning_rate": 1.7652503551913167e-07,
+      "loss": 0.9761,
+      "step": 6379
+    },
+    {
+      "epoch": 0.9785276073619632,
+      "grad_norm": 0.16391988098621368,
+      "learning_rate": 1.7403097875337835e-07,
+      "loss": 1.1231,
+      "step": 6380
+    },
+    {
+      "epoch": 0.978680981595092,
+      "grad_norm": 0.15794773399829865,
+      "learning_rate": 1.715546458999545e-07,
+      "loss": 1.0791,
+      "step": 6381
+    },
+    {
+      "epoch": 0.9788343558282209,
+      "grad_norm": 0.14556477963924408,
+      "learning_rate": 1.6909603754544088e-07,
+      "loss": 1.0298,
+      "step": 6382
+    },
+    {
+      "epoch": 0.9789877300613496,
+      "grad_norm": 0.1521809846162796,
+      "learning_rate": 1.66655154272205e-07,
+      "loss": 0.995,
+      "step": 6383
+    },
+    {
+      "epoch": 0.9791411042944785,
+      "grad_norm": 0.12611958384513855,
+      "learning_rate": 1.6423199665840104e-07,
+      "loss": 0.8333,
+      "step": 6384
+    },
+    {
+      "epoch": 0.9792944785276073,
+      "grad_norm": 0.15944786369800568,
+      "learning_rate": 1.6182656527801153e-07,
+      "loss": 0.9604,
+      "step": 6385
+    },
+    {
+      "epoch": 0.9794478527607362,
+      "grad_norm": 0.14800360798835754,
+      "learning_rate": 1.594388607008057e-07,
+      "loss": 0.9468,
+      "step": 6386
+    },
+    {
+      "epoch": 0.979601226993865,
+      "grad_norm": 0.16233693063259125,
+      "learning_rate": 1.5706888349234781e-07,
+      "loss": 1.0635,
+      "step": 6387
+    },
+    {
+      "epoch": 0.9797546012269939,
+      "grad_norm": 0.15099622309207916,
+      "learning_rate": 1.5471663421401382e-07,
+      "loss": 1.0333,
+      "step": 6388
+    },
+    {
+      "epoch": 0.9799079754601226,
+      "grad_norm": 0.1576409786939621,
+      "learning_rate": 1.52382113422983e-07,
+      "loss": 1.0169,
+      "step": 6389
+    },
+    {
+      "epoch": 0.9800613496932515,
+      "grad_norm": 0.15477386116981506,
+      "learning_rate": 1.5006532167222974e-07,
+      "loss": 1.0614,
+      "step": 6390
+    },
+    {
+      "epoch": 0.9802147239263803,
+      "grad_norm": 0.15720243752002716,
+      "learning_rate": 1.4776625951052334e-07,
+      "loss": 1.0989,
+      "step": 6391
+    },
+    {
+      "epoch": 0.9803680981595092,
+      "grad_norm": 0.15166939795017242,
+      "learning_rate": 1.4548492748243657e-07,
+      "loss": 1.0632,
+      "step": 6392
+    },
+    {
+      "epoch": 0.980521472392638,
+      "grad_norm": 0.15280641615390778,
+      "learning_rate": 1.4322132612836213e-07,
+      "loss": 1.1159,
+      "step": 6393
+    },
+    {
+      "epoch": 0.9806748466257669,
+      "grad_norm": 0.15916390717029572,
+      "learning_rate": 1.4097545598444616e-07,
+      "loss": 1.1012,
+      "step": 6394
+    },
+    {
+      "epoch": 0.9808282208588958,
+      "grad_norm": 0.15256546437740326,
+      "learning_rate": 1.3874731758269641e-07,
+      "loss": 0.9601,
+      "step": 6395
+    },
+    {
+      "epoch": 0.9809815950920245,
+      "grad_norm": 0.15760859847068787,
+      "learning_rate": 1.3653691145086576e-07,
+      "loss": 1.0793,
+      "step": 6396
+    },
+    {
+      "epoch": 0.9811349693251534,
+      "grad_norm": 0.15323799848556519,
+      "learning_rate": 1.3434423811253536e-07,
+      "loss": 0.9885,
+      "step": 6397
+    },
+    {
+      "epoch": 0.9812883435582822,
+      "grad_norm": 0.1493401676416397,
+      "learning_rate": 1.3216929808708975e-07,
+      "loss": 0.9418,
+      "step": 6398
+    },
+    {
+      "epoch": 0.9814417177914111,
+      "grad_norm": 0.15353742241859436,
+      "learning_rate": 1.3001209188969185e-07,
+      "loss": 1.0551,
+      "step": 6399
+    },
+    {
+      "epoch": 0.9815950920245399,
+      "grad_norm": 0.1597486138343811,
+      "learning_rate": 1.278726200313246e-07,
+      "loss": 1.1874,
+      "step": 6400
+    },
+    {
+      "epoch": 0.9817484662576688,
+      "grad_norm": 0.15423130989074707,
+      "learning_rate": 1.2575088301875757e-07,
+      "loss": 0.9379,
+      "step": 6401
+    },
+    {
+      "epoch": 0.9819018404907975,
+      "grad_norm": 0.14323356747627258,
+      "learning_rate": 1.236468813545638e-07,
+      "loss": 0.8992,
+      "step": 6402
+    },
+    {
+      "epoch": 0.9820552147239264,
+      "grad_norm": 0.1522848755121231,
+      "learning_rate": 1.2156061553710295e-07,
+      "loss": 1.0561,
+      "step": 6403
+    },
+    {
+      "epoch": 0.9822085889570552,
+      "grad_norm": 0.1502063125371933,
+      "learning_rate": 1.1949208606057137e-07,
+      "loss": 1.0535,
+      "step": 6404
+    },
+    {
+      "epoch": 0.9823619631901841,
+      "grad_norm": 0.1553916484117508,
+      "learning_rate": 1.1744129341491881e-07,
+      "loss": 1.1382,
+      "step": 6405
+    },
+    {
+      "epoch": 0.9825153374233129,
+      "grad_norm": 0.13918881118297577,
+      "learning_rate": 1.1540823808591504e-07,
+      "loss": 0.9138,
+      "step": 6406
+    },
+    {
+      "epoch": 0.9826687116564418,
+      "grad_norm": 0.15551245212554932,
+      "learning_rate": 1.1339292055513315e-07,
+      "loss": 1.0515,
+      "step": 6407
+    },
+    {
+      "epoch": 0.9828220858895705,
+      "grad_norm": 0.1539306938648224,
+      "learning_rate": 1.11395341299933e-07,
+      "loss": 1.0501,
+      "step": 6408
+    },
+    {
+      "epoch": 0.9829754601226994,
+      "grad_norm": 0.14882515370845795,
+      "learning_rate": 1.0941550079348604e-07,
+      "loss": 0.8951,
+      "step": 6409
+    },
+    {
+      "epoch": 0.9831288343558282,
+      "grad_norm": 0.1460411250591278,
+      "learning_rate": 1.0745339950474219e-07,
+      "loss": 1.0043,
+      "step": 6410
+    },
+    {
+      "epoch": 0.9832822085889571,
+      "grad_norm": 0.15536056458950043,
+      "learning_rate": 1.0550903789847965e-07,
+      "loss": 1.047,
+      "step": 6411
+    },
+    {
+      "epoch": 0.9834355828220859,
+      "grad_norm": 0.15789516270160675,
+      "learning_rate": 1.0358241643523002e-07,
+      "loss": 1.0305,
+      "step": 6412
+    },
+    {
+      "epoch": 0.9835889570552148,
+      "grad_norm": 0.15599548816680908,
+      "learning_rate": 1.0167353557136992e-07,
+      "loss": 1.0352,
+      "step": 6413
+    },
+    {
+      "epoch": 0.9837423312883435,
+      "grad_norm": 0.15261557698249817,
+      "learning_rate": 9.978239575904601e-08,
+      "loss": 1.0187,
+      "step": 6414
+    },
+    {
+      "epoch": 0.9838957055214724,
+      "grad_norm": 0.15650393068790436,
+      "learning_rate": 9.790899744620828e-08,
+      "loss": 0.9941,
+      "step": 6415
+    },
+    {
+      "epoch": 0.9840490797546012,
+      "grad_norm": 0.15831558406352997,
+      "learning_rate": 9.605334107661844e-08,
+      "loss": 1.0229,
+      "step": 6416
+    },
+    {
+      "epoch": 0.9842024539877301,
+      "grad_norm": 0.15120169520378113,
+      "learning_rate": 9.42154270897999e-08,
+      "loss": 1.0608,
+      "step": 6417
+    },
+    {
+      "epoch": 0.9843558282208589,
+      "grad_norm": 0.16397884488105774,
+      "learning_rate": 9.239525592111274e-08,
+      "loss": 1.1255,
+      "step": 6418
+    },
+    {
+      "epoch": 0.9845092024539878,
+      "grad_norm": 0.15982234477996826,
+      "learning_rate": 9.059282800169542e-08,
+      "loss": 0.956,
+      "step": 6419
+    },
+    {
+      "epoch": 0.9846625766871165,
+      "grad_norm": 0.15858109295368195,
+      "learning_rate": 8.880814375848977e-08,
+      "loss": 1.0904,
+      "step": 6420
+    },
+    {
+      "epoch": 0.9848159509202454,
+      "grad_norm": 0.1438794583082199,
+      "learning_rate": 8.704120361422429e-08,
+      "loss": 1.0404,
+      "step": 6421
+    },
+    {
+      "epoch": 0.9849693251533742,
+      "grad_norm": 0.15911556780338287,
+      "learning_rate": 8.529200798743086e-08,
+      "loss": 1.0305,
+      "step": 6422
+    },
+    {
+      "epoch": 0.9851226993865031,
+      "grad_norm": 0.14902205765247345,
+      "learning_rate": 8.356055729243638e-08,
+      "loss": 1.0687,
+      "step": 6423
+    },
+    {
+      "epoch": 0.9852760736196319,
+      "grad_norm": 0.1473216712474823,
+      "learning_rate": 8.184685193937945e-08,
+      "loss": 1.0821,
+      "step": 6424
+    },
+    {
+      "epoch": 0.9854294478527608,
+      "grad_norm": 0.14040841162204742,
+      "learning_rate": 8.015089233417704e-08,
+      "loss": 0.9992,
+      "step": 6425
+    },
+    {
+      "epoch": 0.9855828220858895,
+      "grad_norm": 0.16098347306251526,
+      "learning_rate": 7.847267887854114e-08,
+      "loss": 1.0308,
+      "step": 6426
+    },
+    {
+      "epoch": 0.9857361963190184,
+      "grad_norm": 0.1579059362411499,
+      "learning_rate": 7.681221196999543e-08,
+      "loss": 1.0448,
+      "step": 6427
+    },
+    {
+      "epoch": 0.9858895705521472,
+      "grad_norm": 0.15049676597118378,
+      "learning_rate": 7.516949200185862e-08,
+      "loss": 1.0941,
+      "step": 6428
+    },
+    {
+      "epoch": 0.9860429447852761,
+      "grad_norm": 0.151799738407135,
+      "learning_rate": 7.354451936321948e-08,
+      "loss": 0.9863,
+      "step": 6429
+    },
+    {
+      "epoch": 0.9861963190184049,
+      "grad_norm": 0.15460079908370972,
+      "learning_rate": 7.19372944389951e-08,
+      "loss": 1.1674,
+      "step": 6430
+    },
+    {
+      "epoch": 0.9863496932515338,
+      "grad_norm": 0.15139129757881165,
+      "learning_rate": 7.034781760988928e-08,
+      "loss": 0.9814,
+      "step": 6431
+    },
+    {
+      "epoch": 0.9865030674846625,
+      "grad_norm": 0.16422967612743378,
+      "learning_rate": 6.877608925239252e-08,
+      "loss": 1.0997,
+      "step": 6432
+    },
+    {
+      "epoch": 0.9866564417177914,
+      "grad_norm": 0.16387169063091278,
+      "learning_rate": 6.722210973880703e-08,
+      "loss": 1.0776,
+      "step": 6433
+    },
+    {
+      "epoch": 0.9868098159509202,
+      "grad_norm": 0.15200990438461304,
+      "learning_rate": 6.568587943720505e-08,
+      "loss": 0.9692,
+      "step": 6434
+    },
+    {
+      "epoch": 0.9869631901840491,
+      "grad_norm": 0.16587693989276886,
+      "learning_rate": 6.416739871148713e-08,
+      "loss": 1.087,
+      "step": 6435
+    },
+    {
+      "epoch": 0.9871165644171779,
+      "grad_norm": 0.14037171006202698,
+      "learning_rate": 6.266666792132392e-08,
+      "loss": 1.0104,
+      "step": 6436
+    },
+    {
+      "epoch": 0.9872699386503068,
+      "grad_norm": 0.1541118621826172,
+      "learning_rate": 6.11836874221977e-08,
+      "loss": 1.0826,
+      "step": 6437
+    },
+    {
+      "epoch": 0.9874233128834355,
+      "grad_norm": 0.15324193239212036,
+      "learning_rate": 5.971845756537752e-08,
+      "loss": 0.9795,
+      "step": 6438
+    },
+    {
+      "epoch": 0.9875766871165644,
+      "grad_norm": 0.1533045768737793,
+      "learning_rate": 5.827097869792741e-08,
+      "loss": 0.9817,
+      "step": 6439
+    },
+    {
+      "epoch": 0.9877300613496932,
+      "grad_norm": 0.15164563059806824,
+      "learning_rate": 5.6841251162706436e-08,
+      "loss": 0.9989,
+      "step": 6440
+    },
+    {
+      "epoch": 0.9878834355828221,
+      "grad_norm": 0.14919531345367432,
+      "learning_rate": 5.5429275298385366e-08,
+      "loss": 1.0096,
+      "step": 6441
+    },
+    {
+      "epoch": 0.9880368098159509,
+      "grad_norm": 0.14437633752822876,
+      "learning_rate": 5.4035051439405005e-08,
+      "loss": 0.9451,
+      "step": 6442
+    },
+    {
+      "epoch": 0.9881901840490798,
+      "grad_norm": 0.13767346739768982,
+      "learning_rate": 5.2658579916009524e-08,
+      "loss": 0.9392,
+      "step": 6443
+    },
+    {
+      "epoch": 0.9883435582822085,
+      "grad_norm": 0.14806336164474487,
+      "learning_rate": 5.1299861054254765e-08,
+      "loss": 1.0698,
+      "step": 6444
+    },
+    {
+      "epoch": 0.9884969325153374,
+      "grad_norm": 0.15597228705883026,
+      "learning_rate": 4.995889517597495e-08,
+      "loss": 1.0292,
+      "step": 6445
+    },
+    {
+      "epoch": 0.9886503067484662,
+      "grad_norm": 0.15635894238948822,
+      "learning_rate": 4.863568259879103e-08,
+      "loss": 1.0505,
+      "step": 6446
+    },
+    {
+      "epoch": 0.9888036809815951,
+      "grad_norm": 0.13727818429470062,
+      "learning_rate": 4.7330223636143914e-08,
+      "loss": 0.9461,
+      "step": 6447
+    },
+    {
+      "epoch": 0.9889570552147239,
+      "grad_norm": 0.1666625291109085,
+      "learning_rate": 4.604251859725294e-08,
+      "loss": 1.2175,
+      "step": 6448
+    },
+    {
+      "epoch": 0.9891104294478528,
+      "grad_norm": 0.15742111206054688,
+      "learning_rate": 4.4772567787124124e-08,
+      "loss": 1.0393,
+      "step": 6449
+    },
+    {
+      "epoch": 0.9892638036809815,
+      "grad_norm": 0.15941481292247772,
+      "learning_rate": 4.35203715065835e-08,
+      "loss": 1.1553,
+      "step": 6450
+    },
+    {
+      "epoch": 0.9894171779141104,
+      "grad_norm": 0.15379458665847778,
+      "learning_rate": 4.2285930052227135e-08,
+      "loss": 1.0292,
+      "step": 6451
+    },
+    {
+      "epoch": 0.9895705521472392,
+      "grad_norm": 0.15909186005592346,
+      "learning_rate": 4.106924371646281e-08,
+      "loss": 1.0049,
+      "step": 6452
+    },
+    {
+      "epoch": 0.9897239263803681,
+      "grad_norm": 0.14969666302204132,
+      "learning_rate": 3.987031278747666e-08,
+      "loss": 1.0848,
+      "step": 6453
+    },
+    {
+      "epoch": 0.9898773006134969,
+      "grad_norm": 0.13908755779266357,
+      "learning_rate": 3.86891375492665e-08,
+      "loss": 1.0135,
+      "step": 6454
+    },
+    {
+      "epoch": 0.9900306748466258,
+      "grad_norm": 0.15390372276306152,
+      "learning_rate": 3.7525718281608555e-08,
+      "loss": 0.9825,
+      "step": 6455
+    },
+    {
+      "epoch": 0.9901840490797545,
+      "grad_norm": 0.1417882889509201,
+      "learning_rate": 3.6380055260082366e-08,
+      "loss": 1.0004,
+      "step": 6456
+    },
+    {
+      "epoch": 0.9903374233128834,
+      "grad_norm": 0.17176292836666107,
+      "learning_rate": 3.52521487560542e-08,
+      "loss": 1.0468,
+      "step": 6457
+    },
+    {
+      "epoch": 0.9904907975460123,
+      "grad_norm": 0.14699341356754303,
+      "learning_rate": 3.414199903670201e-08,
+      "loss": 0.9539,
+      "step": 6458
+    },
+    {
+      "epoch": 0.9906441717791411,
+      "grad_norm": 0.14349627494812012,
+      "learning_rate": 3.304960636497378e-08,
+      "loss": 0.9742,
+      "step": 6459
+    },
+    {
+      "epoch": 0.99079754601227,
+      "grad_norm": 0.1622297465801239,
+      "learning_rate": 3.197497099963753e-08,
+      "loss": 1.0612,
+      "step": 6460
+    },
+    {
+      "epoch": 0.9909509202453988,
+      "grad_norm": 0.1575959473848343,
+      "learning_rate": 3.091809319521466e-08,
+      "loss": 1.0488,
+      "step": 6461
+    },
+    {
+      "epoch": 0.9911042944785277,
+      "grad_norm": 0.1647721379995346,
+      "learning_rate": 2.9878973202071555e-08,
+      "loss": 1.0744,
+      "step": 6462
+    },
+    {
+      "epoch": 0.9912576687116564,
+      "grad_norm": 0.161659374833107,
+      "learning_rate": 2.8857611266336346e-08,
+      "loss": 1.121,
+      "step": 6463
+    },
+    {
+      "epoch": 0.9914110429447853,
+      "grad_norm": 0.16683809459209442,
+      "learning_rate": 2.7854007629923846e-08,
+      "loss": 1.1086,
+      "step": 6464
+    },
+    {
+      "epoch": 0.9915644171779141,
+      "grad_norm": 0.1680736541748047,
+      "learning_rate": 2.6868162530568893e-08,
+      "loss": 1.1752,
+      "step": 6465
+    },
+    {
+      "epoch": 0.991717791411043,
+      "grad_norm": 0.1789245903491974,
+      "learning_rate": 2.5900076201793018e-08,
+      "loss": 1.2867,
+      "step": 6466
+    },
+    {
+      "epoch": 0.9918711656441718,
+      "grad_norm": 0.15177828073501587,
+      "learning_rate": 2.4949748872887808e-08,
+      "loss": 1.0406,
+      "step": 6467
+    },
+    {
+      "epoch": 0.9920245398773007,
+      "grad_norm": 0.143916517496109,
+      "learning_rate": 2.4017180768964862e-08,
+      "loss": 0.974,
+      "step": 6468
+    },
+    {
+      "epoch": 0.9921779141104294,
+      "grad_norm": 0.15291012823581696,
+      "learning_rate": 2.31023721109308e-08,
+      "loss": 1.0172,
+      "step": 6469
+    },
+    {
+      "epoch": 0.9923312883435583,
+      "grad_norm": 0.15567804872989655,
+      "learning_rate": 2.2205323115453975e-08,
+      "loss": 1.0546,
+      "step": 6470
+    },
+    {
+      "epoch": 0.9924846625766871,
+      "grad_norm": 0.15163074433803558,
+      "learning_rate": 2.132603399503108e-08,
+      "loss": 1.0121,
+      "step": 6471
+    },
+    {
+      "epoch": 0.992638036809816,
+      "grad_norm": 0.15621015429496765,
+      "learning_rate": 2.0464504957937167e-08,
+      "loss": 1.1177,
+      "step": 6472
+    },
+    {
+      "epoch": 0.9927914110429448,
+      "grad_norm": 0.15744642913341522,
+      "learning_rate": 1.962073620823401e-08,
+      "loss": 1.0676,
+      "step": 6473
+    },
+    {
+      "epoch": 0.9929447852760737,
+      "grad_norm": 0.15837515890598297,
+      "learning_rate": 1.8794727945786735e-08,
+      "loss": 0.9707,
+      "step": 6474
+    },
+    {
+      "epoch": 0.9930981595092024,
+      "grad_norm": 0.1628391295671463,
+      "learning_rate": 1.7986480366255495e-08,
+      "loss": 1.1427,
+      "step": 6475
+    },
+    {
+      "epoch": 0.9932515337423313,
+      "grad_norm": 0.1521443873643875,
+      "learning_rate": 1.7195993661087148e-08,
+      "loss": 1.1063,
+      "step": 6476
+    },
+    {
+      "epoch": 0.9934049079754601,
+      "grad_norm": 0.150417298078537,
+      "learning_rate": 1.6423268017515255e-08,
+      "loss": 0.9948,
+      "step": 6477
+    },
+    {
+      "epoch": 0.993558282208589,
+      "grad_norm": 0.14210069179534912,
+      "learning_rate": 1.5668303618585067e-08,
+      "loss": 1.0359,
+      "step": 6478
+    },
+    {
+      "epoch": 0.9937116564417178,
+      "grad_norm": 0.16306671500205994,
+      "learning_rate": 1.4931100643120196e-08,
+      "loss": 1.0943,
+      "step": 6479
+    },
+    {
+      "epoch": 0.9938650306748467,
+      "grad_norm": 0.15590588748455048,
+      "learning_rate": 1.4211659265739306e-08,
+      "loss": 1.0373,
+      "step": 6480
+    },
+    {
+      "epoch": 0.9940184049079754,
+      "grad_norm": 0.13812555372714996,
+      "learning_rate": 1.3509979656856074e-08,
+      "loss": 0.9308,
+      "step": 6481
+    },
+    {
+      "epoch": 0.9941717791411043,
+      "grad_norm": 0.15514235198497772,
+      "learning_rate": 1.2826061982670887e-08,
+      "loss": 1.0311,
+      "step": 6482
+    },
+    {
+      "epoch": 0.9943251533742331,
+      "grad_norm": 0.14908231794834137,
+      "learning_rate": 1.2159906405187492e-08,
+      "loss": 1.1745,
+      "step": 6483
+    },
+    {
+      "epoch": 0.994478527607362,
+      "grad_norm": 0.14877817034721375,
+      "learning_rate": 1.151151308220466e-08,
+      "loss": 0.9873,
+      "step": 6484
+    },
+    {
+      "epoch": 0.9946319018404908,
+      "grad_norm": 0.1503046452999115,
+      "learning_rate": 1.0880882167291216e-08,
+      "loss": 1.0725,
+      "step": 6485
+    },
+    {
+      "epoch": 0.9947852760736197,
+      "grad_norm": 0.155674010515213,
+      "learning_rate": 1.0268013809835996e-08,
+      "loss": 1.1334,
+      "step": 6486
+    },
+    {
+      "epoch": 0.9949386503067484,
+      "grad_norm": 0.15383431315422058,
+      "learning_rate": 9.672908154997883e-09,
+      "loss": 1.0952,
+      "step": 6487
+    },
+    {
+      "epoch": 0.9950920245398773,
+      "grad_norm": 0.15209563076496124,
+      "learning_rate": 9.095565343747446e-09,
+      "loss": 0.9886,
+      "step": 6488
+    },
+    {
+      "epoch": 0.9952453987730061,
+      "grad_norm": 0.15805040299892426,
+      "learning_rate": 8.535985512833633e-09,
+      "loss": 1.0774,
+      "step": 6489
+    },
+    {
+      "epoch": 0.995398773006135,
+      "grad_norm": 0.15245789289474487,
+      "learning_rate": 7.994168794808741e-09,
+      "loss": 1.1335,
+      "step": 6490
+    },
+    {
+      "epoch": 0.9955521472392638,
+      "grad_norm": 0.13701754808425903,
+      "learning_rate": 7.470115318003456e-09,
+      "loss": 1.0112,
+      "step": 6491
+    },
+    {
+      "epoch": 0.9957055214723927,
+      "grad_norm": 0.14866489171981812,
+      "learning_rate": 6.963825206560136e-09,
+      "loss": 1.0683,
+      "step": 6492
+    },
+    {
+      "epoch": 0.9958588957055214,
+      "grad_norm": 0.16040024161338806,
+      "learning_rate": 6.475298580391197e-09,
+      "loss": 1.1295,
+      "step": 6493
+    },
+    {
+      "epoch": 0.9960122699386503,
+      "grad_norm": 0.15185928344726562,
+      "learning_rate": 6.004535555220735e-09,
+      "loss": 1.1435,
+      "step": 6494
+    },
+    {
+      "epoch": 0.9961656441717791,
+      "grad_norm": 0.15037021040916443,
+      "learning_rate": 5.551536242559551e-09,
+      "loss": 1.0516,
+      "step": 6495
+    },
+    {
+      "epoch": 0.996319018404908,
+      "grad_norm": 0.16256538033485413,
+      "learning_rate": 5.116300749705149e-09,
+      "loss": 1.1381,
+      "step": 6496
+    },
+    {
+      "epoch": 0.9964723926380368,
+      "grad_norm": 0.15257202088832855,
+      "learning_rate": 4.698829179750063e-09,
+      "loss": 1.0276,
+      "step": 6497
+    },
+    {
+      "epoch": 0.9966257668711657,
+      "grad_norm": 0.15939125418663025,
+      "learning_rate": 4.299121631581859e-09,
+      "loss": 1.0906,
+      "step": 6498
+    },
+    {
+      "epoch": 0.9967791411042944,
+      "grad_norm": 0.1508501023054123,
+      "learning_rate": 3.917178199883131e-09,
+      "loss": 0.9136,
+      "step": 6499
+    },
+    {
+      "epoch": 0.9969325153374233,
+      "grad_norm": 0.15422071516513824,
+      "learning_rate": 3.5529989751148513e-09,
+      "loss": 1.0444,
+      "step": 6500
+    },
+    {
+      "epoch": 0.9970858895705521,
+      "grad_norm": 0.13963113725185394,
+      "learning_rate": 3.2065840435413493e-09,
+      "loss": 0.9008,
+      "step": 6501
+    },
+    {
+      "epoch": 0.997239263803681,
+      "grad_norm": 0.15956629812717438,
+      "learning_rate": 2.877933487221984e-09,
+      "loss": 0.9707,
+      "step": 6502
+    },
+    {
+      "epoch": 0.9973926380368098,
+      "grad_norm": 0.14838074147701263,
+      "learning_rate": 2.5670473840028184e-09,
+      "loss": 1.0447,
+      "step": 6503
+    },
+    {
+      "epoch": 0.9975460122699387,
+      "grad_norm": 0.1461222618818283,
+      "learning_rate": 2.2739258075249457e-09,
+      "loss": 0.9995,
+      "step": 6504
+    },
+    {
+      "epoch": 0.9976993865030674,
+      "grad_norm": 0.14386232197284698,
+      "learning_rate": 1.9985688272161625e-09,
+      "loss": 1.0121,
+      "step": 6505
+    },
+    {
+      "epoch": 0.9978527607361963,
+      "grad_norm": 0.1496867835521698,
+      "learning_rate": 1.7409765082992965e-09,
+      "loss": 0.9665,
+      "step": 6506
+    },
+    {
+      "epoch": 0.9980061349693251,
+      "grad_norm": 0.16312821209430695,
+      "learning_rate": 1.501148911792205e-09,
+      "loss": 1.1706,
+      "step": 6507
+    },
+    {
+      "epoch": 0.998159509202454,
+      "grad_norm": 0.15520977973937988,
+      "learning_rate": 1.2790860944994485e-09,
+      "loss": 1.0135,
+      "step": 6508
+    },
+    {
+      "epoch": 0.9983128834355828,
+      "grad_norm": 0.1555289328098297,
+      "learning_rate": 1.0747881090289456e-09,
+      "loss": 1.09,
+      "step": 6509
+    },
+    {
+      "epoch": 0.9984662576687117,
+      "grad_norm": 0.16135857999324799,
+      "learning_rate": 8.882550037586644e-10,
+      "loss": 1.0734,
+      "step": 6510
+    },
+    {
+      "epoch": 0.9986196319018404,
+      "grad_norm": 0.15284059941768646,
+      "learning_rate": 7.194868228782569e-10,
+      "loss": 1.0075,
+      "step": 6511
+    },
+    {
+      "epoch": 0.9987730061349693,
+      "grad_norm": 0.1560993492603302,
+      "learning_rate": 5.684836063724052e-10,
+      "loss": 1.0873,
+      "step": 6512
+    },
+    {
+      "epoch": 0.9989263803680981,
+      "grad_norm": 0.16097715497016907,
+      "learning_rate": 4.3524538999584193e-10,
+      "loss": 0.9877,
+      "step": 6513
+    },
+    {
+      "epoch": 0.999079754601227,
+      "grad_norm": 0.16746404767036438,
+      "learning_rate": 3.197722053233098e-10,
+      "loss": 1.0645,
+      "step": 6514
+    },
+    {
+      "epoch": 0.9992331288343558,
+      "grad_norm": 0.15279144048690796,
+      "learning_rate": 2.2206407969127537e-10,
+      "loss": 1.0162,
+      "step": 6515
+    },
+    {
+      "epoch": 0.9993865030674847,
+      "grad_norm": 0.1594628244638443,
+      "learning_rate": 1.4212103624788863e-10,
+      "loss": 0.9967,
+      "step": 6516
+    },
+    {
+      "epoch": 0.9995398773006134,
+      "grad_norm": 0.1491973102092743,
+      "learning_rate": 7.994309393632992e-11,
+      "loss": 0.9595,
+      "step": 6517
+    },
+    {
+      "epoch": 0.9996932515337423,
+      "grad_norm": 0.16072843968868256,
+      "learning_rate": 3.553026747815657e-11,
+      "loss": 1.013,
+      "step": 6518
+    },
+    {
+      "epoch": 0.9998466257668711,
+      "grad_norm": 0.15656042098999023,
+      "learning_rate": 8.882567398282859e-12,
+      "loss": 1.137,
+      "step": 6519
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1527823656797409,
+      "learning_rate": 0.0,
+      "loss": 1.0393,
+      "step": 6520
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 6520,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 81,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.0719074667412324e+19,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}