diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12012 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9994885657923577,
+  "eval_steps": 500,
+  "global_step": 1710,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005844962373054724,
+      "grad_norm": 23.7740973198474,
+      "learning_rate": 0.0,
+      "loss": 2.0451,
+      "step": 1
+    },
+    {
+      "epoch": 0.0011689924746109447,
+      "grad_norm": 8.111811192948283,
+      "learning_rate": 2.3981246656813147e-06,
+      "loss": 1.8552,
+      "step": 2
+    },
+    {
+      "epoch": 0.001753488711916417,
+      "grad_norm": 7.595564061701498,
+      "learning_rate": 3.800937667159344e-06,
+      "loss": 1.8272,
+      "step": 3
+    },
+    {
+      "epoch": 0.0023379849492218894,
+      "grad_norm": 7.71208849713537,
+      "learning_rate": 4.7962493313626295e-06,
+      "loss": 1.8446,
+      "step": 4
+    },
+    {
+      "epoch": 0.002922481186527362,
+      "grad_norm": 5.621052504639032,
+      "learning_rate": 5.568273036287808e-06,
+      "loss": 1.6579,
+      "step": 5
+    },
+    {
+      "epoch": 0.003506977423832834,
+      "grad_norm": 5.304511042722482,
+      "learning_rate": 6.199062332840659e-06,
+      "loss": 1.6567,
+      "step": 6
+    },
+    {
+      "epoch": 0.004091473661138306,
+      "grad_norm": 4.770638284739429,
+      "learning_rate": 6.732387083908185e-06,
+      "loss": 1.6008,
+      "step": 7
+    },
+    {
+      "epoch": 0.004675969898443779,
+      "grad_norm": 3.5929321835671684,
+      "learning_rate": 7.194373997043944e-06,
+      "loss": 1.6222,
+      "step": 8
+    },
+    {
+      "epoch": 0.005260466135749251,
+      "grad_norm": 3.970194488842916,
+      "learning_rate": 7.601875334318688e-06,
+      "loss": 1.5604,
+      "step": 9
+    },
+    {
+      "epoch": 0.005844962373054724,
+      "grad_norm": 3.5395647274060247,
+      "learning_rate": 7.966397701969125e-06,
+      "loss": 1.5217,
+      "step": 10
+    },
+    {
+      "epoch": 0.006429458610360195,
+      "grad_norm": 3.1016474754545413,
+      "learning_rate": 8.29614829389194e-06,
+      "loss": 1.5498,
+      "step": 11
+    },
+    {
+      "epoch": 0.007013954847665668,
+      "grad_norm": 3.0411161195685033,
+      "learning_rate": 8.597186998521974e-06,
+      "loss": 1.5215,
+      "step": 12
+    },
+    {
+      "epoch": 0.00759845108497114,
+      "grad_norm": 2.812109479029755,
+      "learning_rate": 8.874115761940966e-06,
+      "loss": 1.4919,
+      "step": 13
+    },
+    {
+      "epoch": 0.008182947322276613,
+      "grad_norm": 2.763020165895636,
+      "learning_rate": 9.1305117495895e-06,
+      "loss": 1.47,
+      "step": 14
+    },
+    {
+      "epoch": 0.008767443559582085,
+      "grad_norm": 2.5985395686062556,
+      "learning_rate": 9.369210703447152e-06,
+      "loss": 1.5189,
+      "step": 15
+    },
+    {
+      "epoch": 0.009351939796887558,
+      "grad_norm": 2.520151401027329,
+      "learning_rate": 9.592498662725259e-06,
+      "loss": 1.4776,
+      "step": 16
+    },
+    {
+      "epoch": 0.00993643603419303,
+      "grad_norm": 2.896832582433134,
+      "learning_rate": 9.802245459658269e-06,
+      "loss": 1.5361,
+      "step": 17
+    },
+    {
+      "epoch": 0.010520932271498503,
+      "grad_norm": 2.7914944309399656,
+      "learning_rate": 1e-05,
+      "loss": 1.4885,
+      "step": 18
+    },
+    {
+      "epoch": 0.011105428508803975,
+      "grad_norm": 2.6244689732269513,
+      "learning_rate": 1e-05,
+      "loss": 1.508,
+      "step": 19
+    },
+    {
+      "epoch": 0.011689924746109448,
+      "grad_norm": 2.5058653849638373,
+      "learning_rate": 1e-05,
+      "loss": 1.5275,
+      "step": 20
+    },
+    {
+      "epoch": 0.01227442098341492,
+      "grad_norm": 2.4701313643863254,
+      "learning_rate": 1e-05,
+      "loss": 1.4783,
+      "step": 21
+    },
+    {
+      "epoch": 0.01285891722072039,
+      "grad_norm": 2.557971352548508,
+      "learning_rate": 1e-05,
+      "loss": 1.478,
+      "step": 22
+    },
+    {
+      "epoch": 0.013443413458025863,
+      "grad_norm": 2.4241266520637974,
+      "learning_rate": 1e-05,
+      "loss": 1.4778,
+      "step": 23
+    },
+    {
+      "epoch": 0.014027909695331336,
+      "grad_norm": 2.473737259302716,
+      "learning_rate": 1e-05,
+      "loss": 1.4128,
+      "step": 24
+    },
+    {
+      "epoch": 0.014612405932636808,
+      "grad_norm": 2.698834286084992,
+      "learning_rate": 1e-05,
+      "loss": 1.4475,
+      "step": 25
+    },
+    {
+      "epoch": 0.01519690216994228,
+      "grad_norm": 2.7355237090787528,
+      "learning_rate": 1e-05,
+      "loss": 1.4308,
+      "step": 26
+    },
+    {
+      "epoch": 0.015781398407247755,
+      "grad_norm": 2.303923770999781,
+      "learning_rate": 1e-05,
+      "loss": 1.4267,
+      "step": 27
+    },
+    {
+      "epoch": 0.016365894644553226,
+      "grad_norm": 2.4550349462011396,
+      "learning_rate": 1e-05,
+      "loss": 1.4518,
+      "step": 28
+    },
+    {
+      "epoch": 0.016950390881858696,
+      "grad_norm": 2.5045285374390063,
+      "learning_rate": 1e-05,
+      "loss": 1.4484,
+      "step": 29
+    },
+    {
+      "epoch": 0.01753488711916417,
+      "grad_norm": 2.8021521695039997,
+      "learning_rate": 1e-05,
+      "loss": 1.4414,
+      "step": 30
+    },
+    {
+      "epoch": 0.01811938335646964,
+      "grad_norm": 2.304080556351434,
+      "learning_rate": 1e-05,
+      "loss": 1.4689,
+      "step": 31
+    },
+    {
+      "epoch": 0.018703879593775116,
+      "grad_norm": 2.2155680599451926,
+      "learning_rate": 1e-05,
+      "loss": 1.4259,
+      "step": 32
+    },
+    {
+      "epoch": 0.019288375831080586,
+      "grad_norm": 2.302070142487707,
+      "learning_rate": 1e-05,
+      "loss": 1.4404,
+      "step": 33
+    },
+    {
+      "epoch": 0.01987287206838606,
+      "grad_norm": 2.164878610881341,
+      "learning_rate": 1e-05,
+      "loss": 1.3989,
+      "step": 34
+    },
+    {
+      "epoch": 0.02045736830569153,
+      "grad_norm": 2.3406139793911915,
+      "learning_rate": 1e-05,
+      "loss": 1.438,
+      "step": 35
+    },
+    {
+      "epoch": 0.021041864542997005,
+      "grad_norm": 2.6775932730477408,
+      "learning_rate": 1e-05,
+      "loss": 1.3935,
+      "step": 36
+    },
+    {
+      "epoch": 0.021626360780302476,
+      "grad_norm": 2.228054948968253,
+      "learning_rate": 1e-05,
+      "loss": 1.3937,
+      "step": 37
+    },
+    {
+      "epoch": 0.02221085701760795,
+      "grad_norm": 2.312757175969669,
+      "learning_rate": 1e-05,
+      "loss": 1.4674,
+      "step": 38
+    },
+    {
+      "epoch": 0.02279535325491342,
+      "grad_norm": 2.523198817093508,
+      "learning_rate": 1e-05,
+      "loss": 1.397,
+      "step": 39
+    },
+    {
+      "epoch": 0.023379849492218895,
+      "grad_norm": 2.3757182723661567,
+      "learning_rate": 1e-05,
+      "loss": 1.4444,
+      "step": 40
+    },
+    {
+      "epoch": 0.023964345729524366,
+      "grad_norm": 2.4209072920678696,
+      "learning_rate": 1e-05,
+      "loss": 1.4311,
+      "step": 41
+    },
+    {
+      "epoch": 0.02454884196682984,
+      "grad_norm": 2.1746341621396397,
+      "learning_rate": 1e-05,
+      "loss": 1.3669,
+      "step": 42
+    },
+    {
+      "epoch": 0.02513333820413531,
+      "grad_norm": 2.300885573649233,
+      "learning_rate": 1e-05,
+      "loss": 1.4056,
+      "step": 43
+    },
+    {
+      "epoch": 0.02571783444144078,
+      "grad_norm": 2.3347454470953544,
+      "learning_rate": 1e-05,
+      "loss": 1.3696,
+      "step": 44
+    },
+    {
+      "epoch": 0.026302330678746256,
+      "grad_norm": 2.4073307575101657,
+      "learning_rate": 1e-05,
+      "loss": 1.3902,
+      "step": 45
+    },
+    {
+      "epoch": 0.026886826916051727,
+      "grad_norm": 2.3457872598267455,
+      "learning_rate": 1e-05,
+      "loss": 1.4496,
+      "step": 46
+    },
+    {
+      "epoch": 0.0274713231533572,
+      "grad_norm": 2.4240246275756916,
+      "learning_rate": 1e-05,
+      "loss": 1.4061,
+      "step": 47
+    },
+    {
+      "epoch": 0.02805581939066267,
+      "grad_norm": 2.28796704606821,
+      "learning_rate": 1e-05,
+      "loss": 1.4441,
+      "step": 48
+    },
+    {
+      "epoch": 0.028640315627968146,
+      "grad_norm": 2.3136253604640835,
+      "learning_rate": 1e-05,
+      "loss": 1.4159,
+      "step": 49
+    },
+    {
+      "epoch": 0.029224811865273616,
+      "grad_norm": 2.355764233711023,
+      "learning_rate": 1e-05,
+      "loss": 1.422,
+      "step": 50
+    },
+    {
+      "epoch": 0.02980930810257909,
+      "grad_norm": 2.2695482451127127,
+      "learning_rate": 1e-05,
+      "loss": 1.3719,
+      "step": 51
+    },
+    {
+      "epoch": 0.03039380433988456,
+      "grad_norm": 2.2205678658996217,
+      "learning_rate": 1e-05,
+      "loss": 1.3568,
+      "step": 52
+    },
+    {
+      "epoch": 0.030978300577190036,
+      "grad_norm": 2.148255358980416,
+      "learning_rate": 1e-05,
+      "loss": 1.4291,
+      "step": 53
+    },
+    {
+      "epoch": 0.03156279681449551,
+      "grad_norm": 2.0891444448978884,
+      "learning_rate": 1e-05,
+      "loss": 1.4024,
+      "step": 54
+    },
+    {
+      "epoch": 0.03214729305180098,
+      "grad_norm": 2.4794390062716545,
+      "learning_rate": 1e-05,
+      "loss": 1.4038,
+      "step": 55
+    },
+    {
+      "epoch": 0.03273178928910645,
+      "grad_norm": 2.3532259079808497,
+      "learning_rate": 1e-05,
+      "loss": 1.3784,
+      "step": 56
+    },
+    {
+      "epoch": 0.03331628552641192,
+      "grad_norm": 2.2981333716894037,
+      "learning_rate": 1e-05,
+      "loss": 1.4521,
+      "step": 57
+    },
+    {
+      "epoch": 0.03390078176371739,
+      "grad_norm": 2.2719465649300488,
+      "learning_rate": 1e-05,
+      "loss": 1.3872,
+      "step": 58
+    },
+    {
+      "epoch": 0.03448527800102287,
+      "grad_norm": 2.420124373451866,
+      "learning_rate": 1e-05,
+      "loss": 1.3877,
+      "step": 59
+    },
+    {
+      "epoch": 0.03506977423832834,
+      "grad_norm": 2.195924747800388,
+      "learning_rate": 1e-05,
+      "loss": 1.3418,
+      "step": 60
+    },
+    {
+      "epoch": 0.03565427047563381,
+      "grad_norm": 2.4202788940806177,
+      "learning_rate": 1e-05,
+      "loss": 1.3734,
+      "step": 61
+    },
+    {
+      "epoch": 0.03623876671293928,
+      "grad_norm": 2.2971799784237765,
+      "learning_rate": 1e-05,
+      "loss": 1.3472,
+      "step": 62
+    },
+    {
+      "epoch": 0.03682326295024476,
+      "grad_norm": 2.3830429397766153,
+      "learning_rate": 1e-05,
+      "loss": 1.4057,
+      "step": 63
+    },
+    {
+      "epoch": 0.03740775918755023,
+      "grad_norm": 2.02362142213084,
+      "learning_rate": 1e-05,
+      "loss": 1.3606,
+      "step": 64
+    },
+    {
+      "epoch": 0.0379922554248557,
+      "grad_norm": 2.324428939912622,
+      "learning_rate": 1e-05,
+      "loss": 1.368,
+      "step": 65
+    },
+    {
+      "epoch": 0.03857675166216117,
+      "grad_norm": 2.165481963337798,
+      "learning_rate": 1e-05,
+      "loss": 1.3945,
+      "step": 66
+    },
+    {
+      "epoch": 0.03916124789946665,
+      "grad_norm": 2.3638274273306785,
+      "learning_rate": 1e-05,
+      "loss": 1.3959,
+      "step": 67
+    },
+    {
+      "epoch": 0.03974574413677212,
+      "grad_norm": 2.2515706398373365,
+      "learning_rate": 1e-05,
+      "loss": 1.3402,
+      "step": 68
+    },
+    {
+      "epoch": 0.04033024037407759,
+      "grad_norm": 2.397448109805578,
+      "learning_rate": 1e-05,
+      "loss": 1.3482,
+      "step": 69
+    },
+    {
+      "epoch": 0.04091473661138306,
+      "grad_norm": 2.2005145478627397,
+      "learning_rate": 1e-05,
+      "loss": 1.3918,
+      "step": 70
+    },
+    {
+      "epoch": 0.04149923284868853,
+      "grad_norm": 2.730097088023348,
+      "learning_rate": 1e-05,
+      "loss": 1.375,
+      "step": 71
+    },
+    {
+      "epoch": 0.04208372908599401,
+      "grad_norm": 2.488806372993402,
+      "learning_rate": 1e-05,
+      "loss": 1.3734,
+      "step": 72
+    },
+    {
+      "epoch": 0.04266822532329948,
+      "grad_norm": 2.2519515647095747,
+      "learning_rate": 1e-05,
+      "loss": 1.3537,
+      "step": 73
+    },
+    {
+      "epoch": 0.04325272156060495,
+      "grad_norm": 2.393911711717019,
+      "learning_rate": 1e-05,
+      "loss": 1.3532,
+      "step": 74
+    },
+    {
+      "epoch": 0.04383721779791042,
+      "grad_norm": 2.12534660610871,
+      "learning_rate": 1e-05,
+      "loss": 1.3452,
+      "step": 75
+    },
+    {
+      "epoch": 0.0444217140352159,
+      "grad_norm": 2.225566823943393,
+      "learning_rate": 1e-05,
+      "loss": 1.3725,
+      "step": 76
+    },
+    {
+      "epoch": 0.04500621027252137,
+      "grad_norm": 2.229904772635389,
+      "learning_rate": 1e-05,
+      "loss": 1.3728,
+      "step": 77
+    },
+    {
+      "epoch": 0.04559070650982684,
+      "grad_norm": 2.1188819570623836,
+      "learning_rate": 1e-05,
+      "loss": 1.3738,
+      "step": 78
+    },
+    {
+      "epoch": 0.04617520274713231,
+      "grad_norm": 2.3529140157580892,
+      "learning_rate": 1e-05,
+      "loss": 1.3712,
+      "step": 79
+    },
+    {
+      "epoch": 0.04675969898443779,
+      "grad_norm": 2.23057360847372,
+      "learning_rate": 1e-05,
+      "loss": 1.3478,
+      "step": 80
+    },
+    {
+      "epoch": 0.04734419522174326,
+      "grad_norm": 2.243731248856658,
+      "learning_rate": 1e-05,
+      "loss": 1.3625,
+      "step": 81
+    },
+    {
+      "epoch": 0.04792869145904873,
+      "grad_norm": 2.153823685489636,
+      "learning_rate": 1e-05,
+      "loss": 1.3865,
+      "step": 82
+    },
+    {
+      "epoch": 0.0485131876963542,
+      "grad_norm": 2.3048452926470304,
+      "learning_rate": 1e-05,
+      "loss": 1.368,
+      "step": 83
+    },
+    {
+      "epoch": 0.04909768393365968,
+      "grad_norm": 2.2966943042053867,
+      "learning_rate": 1e-05,
+      "loss": 1.3522,
+      "step": 84
+    },
+    {
+      "epoch": 0.04968218017096515,
+      "grad_norm": 2.297181116309717,
+      "learning_rate": 1e-05,
+      "loss": 1.3955,
+      "step": 85
+    },
+    {
+      "epoch": 0.05026667640827062,
+      "grad_norm": 2.3078952028672246,
+      "learning_rate": 1e-05,
+      "loss": 1.3519,
+      "step": 86
+    },
+    {
+      "epoch": 0.05085117264557609,
+      "grad_norm": 2.2052246308849512,
+      "learning_rate": 1e-05,
+      "loss": 1.3418,
+      "step": 87
+    },
+    {
+      "epoch": 0.05143566888288156,
+      "grad_norm": 2.380553700768679,
+      "learning_rate": 1e-05,
+      "loss": 1.3113,
+      "step": 88
+    },
+    {
+      "epoch": 0.05202016512018704,
+      "grad_norm": 2.158213542077448,
+      "learning_rate": 1e-05,
+      "loss": 1.3826,
+      "step": 89
+    },
+    {
+      "epoch": 0.05260466135749251,
+      "grad_norm": 2.332887107153097,
+      "learning_rate": 1e-05,
+      "loss": 1.3646,
+      "step": 90
+    },
+    {
+      "epoch": 0.05318915759479798,
+      "grad_norm": 2.2941867324079936,
+      "learning_rate": 1e-05,
+      "loss": 1.364,
+      "step": 91
+    },
+    {
+      "epoch": 0.05377365383210345,
+      "grad_norm": 2.2853421820948654,
+      "learning_rate": 1e-05,
+      "loss": 1.3084,
+      "step": 92
+    },
+    {
+      "epoch": 0.05435815006940893,
+      "grad_norm": 2.270012523147772,
+      "learning_rate": 1e-05,
+      "loss": 1.347,
+      "step": 93
+    },
+    {
+      "epoch": 0.0549426463067144,
+      "grad_norm": 2.1060325502380466,
+      "learning_rate": 1e-05,
+      "loss": 1.401,
+      "step": 94
+    },
+    {
+      "epoch": 0.05552714254401987,
+      "grad_norm": 2.0839420617041315,
+      "learning_rate": 1e-05,
+      "loss": 1.3086,
+      "step": 95
+    },
+    {
+      "epoch": 0.05611163878132534,
+      "grad_norm": 2.291873364212861,
+      "learning_rate": 1e-05,
+      "loss": 1.3623,
+      "step": 96
+    },
+    {
+      "epoch": 0.05669613501863082,
+      "grad_norm": 2.240541437346825,
+      "learning_rate": 1e-05,
+      "loss": 1.3443,
+      "step": 97
+    },
+    {
+      "epoch": 0.05728063125593629,
+      "grad_norm": 2.2497651676780683,
+      "learning_rate": 1e-05,
+      "loss": 1.2849,
+      "step": 98
+    },
+    {
+      "epoch": 0.05786512749324176,
+      "grad_norm": 2.216519606003062,
+      "learning_rate": 1e-05,
+      "loss": 1.3595,
+      "step": 99
+    },
+    {
+      "epoch": 0.05844962373054723,
+      "grad_norm": 2.380169543966151,
+      "learning_rate": 1e-05,
+      "loss": 1.3843,
+      "step": 100
+    },
+    {
+      "epoch": 0.059034119967852704,
+      "grad_norm": 2.1203068760596437,
+      "learning_rate": 1e-05,
+      "loss": 1.3523,
+      "step": 101
+    },
+    {
+      "epoch": 0.05961861620515818,
+      "grad_norm": 1.9902485946867732,
+      "learning_rate": 1e-05,
+      "loss": 1.2904,
+      "step": 102
+    },
+    {
+      "epoch": 0.06020311244246365,
+      "grad_norm": 2.4029040425666777,
+      "learning_rate": 1e-05,
+      "loss": 1.3207,
+      "step": 103
+    },
+    {
+      "epoch": 0.06078760867976912,
+      "grad_norm": 2.4016353721253005,
+      "learning_rate": 1e-05,
+      "loss": 1.372,
+      "step": 104
+    },
+    {
+      "epoch": 0.061372104917074594,
+      "grad_norm": 2.2988502794700496,
+      "learning_rate": 1e-05,
+      "loss": 1.374,
+      "step": 105
+    },
+    {
+      "epoch": 0.06195660115438007,
+      "grad_norm": 2.428658729123216,
+      "learning_rate": 1e-05,
+      "loss": 1.3907,
+      "step": 106
+    },
+    {
+      "epoch": 0.06254109739168554,
+      "grad_norm": 2.3306851878836574,
+      "learning_rate": 1e-05,
+      "loss": 1.3501,
+      "step": 107
+    },
+    {
+      "epoch": 0.06312559362899102,
+      "grad_norm": 2.088179008614745,
+      "learning_rate": 1e-05,
+      "loss": 1.3587,
+      "step": 108
+    },
+    {
+      "epoch": 0.06371008986629649,
+      "grad_norm": 2.238528300831707,
+      "learning_rate": 1e-05,
+      "loss": 1.3559,
+      "step": 109
+    },
+    {
+      "epoch": 0.06429458610360196,
+      "grad_norm": 2.252621566823,
+      "learning_rate": 1e-05,
+      "loss": 1.3371,
+      "step": 110
+    },
+    {
+      "epoch": 0.06487908234090743,
+      "grad_norm": 2.060890128934192,
+      "learning_rate": 1e-05,
+      "loss": 1.3577,
+      "step": 111
+    },
+    {
+      "epoch": 0.0654635785782129,
+      "grad_norm": 2.192347853943285,
+      "learning_rate": 1e-05,
+      "loss": 1.3253,
+      "step": 112
+    },
+    {
+      "epoch": 0.06604807481551837,
+      "grad_norm": 2.0975008524665157,
+      "learning_rate": 1e-05,
+      "loss": 1.3228,
+      "step": 113
+    },
+    {
+      "epoch": 0.06663257105282384,
+      "grad_norm": 2.201156153872674,
+      "learning_rate": 1e-05,
+      "loss": 1.3093,
+      "step": 114
+    },
+    {
+      "epoch": 0.06721706729012931,
+      "grad_norm": 2.1443901176039084,
+      "learning_rate": 1e-05,
+      "loss": 1.3173,
+      "step": 115
+    },
+    {
+      "epoch": 0.06780156352743479,
+      "grad_norm": 2.076611165629815,
+      "learning_rate": 1e-05,
+      "loss": 1.3233,
+      "step": 116
+    },
+    {
+      "epoch": 0.06838605976474027,
+      "grad_norm": 2.0026277783843702,
+      "learning_rate": 1e-05,
+      "loss": 1.32,
+      "step": 117
+    },
+    {
+      "epoch": 0.06897055600204574,
+      "grad_norm": 2.134214156986593,
+      "learning_rate": 1e-05,
+      "loss": 1.3501,
+      "step": 118
+    },
+    {
+      "epoch": 0.06955505223935121,
+      "grad_norm": 2.3475474986839937,
+      "learning_rate": 1e-05,
+      "loss": 1.3008,
+      "step": 119
+    },
+    {
+      "epoch": 0.07013954847665668,
+      "grad_norm": 2.1863564643825217,
+      "learning_rate": 1e-05,
+      "loss": 1.3426,
+      "step": 120
+    },
+    {
+      "epoch": 0.07072404471396215,
+      "grad_norm": 2.407215736841061,
+      "learning_rate": 1e-05,
+      "loss": 1.3788,
+      "step": 121
+    },
+    {
+      "epoch": 0.07130854095126762,
+      "grad_norm": 2.3356743704375025,
+      "learning_rate": 1e-05,
+      "loss": 1.3158,
+      "step": 122
+    },
+    {
+      "epoch": 0.0718930371885731,
+      "grad_norm": 2.3348410141799203,
+      "learning_rate": 1e-05,
+      "loss": 1.3353,
+      "step": 123
+    },
+    {
+      "epoch": 0.07247753342587857,
+      "grad_norm": 2.2031606137677726,
+      "learning_rate": 1e-05,
+      "loss": 1.3685,
+      "step": 124
+    },
+    {
+      "epoch": 0.07306202966318405,
+      "grad_norm": 2.275103439816515,
+      "learning_rate": 1e-05,
+      "loss": 1.3062,
+      "step": 125
+    },
+    {
+      "epoch": 0.07364652590048952,
+      "grad_norm": 2.270142777226694,
+      "learning_rate": 1e-05,
+      "loss": 1.3148,
+      "step": 126
+    },
+    {
+      "epoch": 0.07423102213779499,
+      "grad_norm": 2.0442662440469084,
+      "learning_rate": 1e-05,
+      "loss": 1.316,
+      "step": 127
+    },
+    {
+      "epoch": 0.07481551837510046,
+      "grad_norm": 2.1058393196442005,
+      "learning_rate": 1e-05,
+      "loss": 1.3867,
+      "step": 128
+    },
+    {
+      "epoch": 0.07540001461240593,
+      "grad_norm": 2.2238399861678673,
+      "learning_rate": 1e-05,
+      "loss": 1.3194,
+      "step": 129
+    },
+    {
+      "epoch": 0.0759845108497114,
+      "grad_norm": 2.1403781348557214,
+      "learning_rate": 1e-05,
+      "loss": 1.2953,
+      "step": 130
+    },
+    {
+      "epoch": 0.07656900708701687,
+      "grad_norm": 2.0320013498338945,
+      "learning_rate": 1e-05,
+      "loss": 1.2877,
+      "step": 131
+    },
+    {
+      "epoch": 0.07715350332432235,
+      "grad_norm": 2.2465142323703593,
+      "learning_rate": 1e-05,
+      "loss": 1.3335,
+      "step": 132
+    },
+    {
+      "epoch": 0.07773799956162782,
+      "grad_norm": 1.8243081896524245,
+      "learning_rate": 1e-05,
+      "loss": 1.3587,
+      "step": 133
+    },
+    {
+      "epoch": 0.0783224957989333,
+      "grad_norm": 2.1738592983277316,
+      "learning_rate": 1e-05,
+      "loss": 1.3479,
+      "step": 134
+    },
+    {
+      "epoch": 0.07890699203623877,
+      "grad_norm": 2.271141310601414,
+      "learning_rate": 1e-05,
+      "loss": 1.2924,
+      "step": 135
+    },
+    {
+      "epoch": 0.07949148827354424,
+      "grad_norm": 2.3726379102257775,
+      "learning_rate": 1e-05,
+      "loss": 1.2975,
+      "step": 136
+    },
+    {
+      "epoch": 0.08007598451084971,
+      "grad_norm": 2.3421456865467842,
+      "learning_rate": 1e-05,
+      "loss": 1.3107,
+      "step": 137
+    },
+    {
+      "epoch": 0.08066048074815518,
+      "grad_norm": 2.270157210933452,
+      "learning_rate": 1e-05,
+      "loss": 1.2654,
+      "step": 138
+    },
+    {
+      "epoch": 0.08124497698546065,
+      "grad_norm": 2.2053577995272438,
+      "learning_rate": 1e-05,
+      "loss": 1.3522,
+      "step": 139
+    },
+    {
+      "epoch": 0.08182947322276612,
+      "grad_norm": 2.1928783074501825,
+      "learning_rate": 1e-05,
+      "loss": 1.3506,
+      "step": 140
+    },
+    {
+      "epoch": 0.0824139694600716,
+      "grad_norm": 2.185243241174161,
+      "learning_rate": 1e-05,
+      "loss": 1.3676,
+      "step": 141
+    },
+    {
+      "epoch": 0.08299846569737707,
+      "grad_norm": 2.0600506779660956,
+      "learning_rate": 1e-05,
+      "loss": 1.3634,
+      "step": 142
+    },
+    {
+      "epoch": 0.08358296193468255,
+      "grad_norm": 2.363041051218333,
+      "learning_rate": 1e-05,
+      "loss": 1.3419,
+      "step": 143
+    },
+    {
+      "epoch": 0.08416745817198802,
+      "grad_norm": 2.27447605040156,
+      "learning_rate": 1e-05,
+      "loss": 1.3471,
+      "step": 144
+    },
+    {
+      "epoch": 0.08475195440929349,
+      "grad_norm": 2.192672877721629,
+      "learning_rate": 1e-05,
+      "loss": 1.3497,
+      "step": 145
+    },
+    {
+      "epoch": 0.08533645064659896,
+      "grad_norm": 2.2094767230107477,
+      "learning_rate": 1e-05,
+      "loss": 1.3093,
+      "step": 146
+    },
+    {
+      "epoch": 0.08592094688390443,
+      "grad_norm": 2.051452197789312,
+      "learning_rate": 1e-05,
+      "loss": 1.3008,
+      "step": 147
+    },
+    {
+      "epoch": 0.0865054431212099,
+      "grad_norm": 2.1397171339362804,
+      "learning_rate": 1e-05,
+      "loss": 1.3199,
+      "step": 148
+    },
+    {
+      "epoch": 0.08708993935851538,
+      "grad_norm": 2.02987740530301,
+      "learning_rate": 1e-05,
+      "loss": 1.3611,
+      "step": 149
+    },
+    {
+      "epoch": 0.08767443559582085,
+      "grad_norm": 2.012492058121203,
+      "learning_rate": 1e-05,
+      "loss": 1.3507,
+      "step": 150
+    },
+    {
+      "epoch": 0.08825893183312633,
+      "grad_norm": 2.0969626545224114,
+      "learning_rate": 1e-05,
+      "loss": 1.3191,
+      "step": 151
+    },
+    {
+      "epoch": 0.0888434280704318,
+      "grad_norm": 2.3349987429661874,
+      "learning_rate": 1e-05,
+      "loss": 1.2836,
+      "step": 152
+    },
+    {
+      "epoch": 0.08942792430773727,
+      "grad_norm": 2.2574582422779352,
+      "learning_rate": 1e-05,
+      "loss": 1.3638,
+      "step": 153
+    },
+    {
+      "epoch": 0.09001242054504274,
+      "grad_norm": 1.8540871623515924,
+      "learning_rate": 1e-05,
+      "loss": 1.358,
+      "step": 154
+    },
+    {
+      "epoch": 0.09059691678234821,
+      "grad_norm": 2.3299472188986576,
+      "learning_rate": 1e-05,
+      "loss": 1.2849,
+      "step": 155
+    },
+    {
+      "epoch": 0.09118141301965368,
+      "grad_norm": 2.2577992480346745,
+      "learning_rate": 1e-05,
+      "loss": 1.3415,
+      "step": 156
+    },
+    {
+      "epoch": 0.09176590925695916,
+      "grad_norm": 2.166013934172933,
+      "learning_rate": 1e-05,
+      "loss": 1.3207,
+      "step": 157
+    },
+    {
+      "epoch": 0.09235040549426463,
+      "grad_norm": 2.1505279549968974,
+      "learning_rate": 1e-05,
+      "loss": 1.3554,
+      "step": 158
+    },
+    {
+      "epoch": 0.0929349017315701,
+      "grad_norm": 1.9250598639069556,
+      "learning_rate": 1e-05,
+      "loss": 1.3003,
+      "step": 159
+    },
+    {
+      "epoch": 0.09351939796887558,
+      "grad_norm": 2.114862362689793,
+      "learning_rate": 1e-05,
+      "loss": 1.2872,
+      "step": 160
+    },
+    {
+      "epoch": 0.09410389420618105,
+      "grad_norm": 2.1864220348430985,
+      "learning_rate": 1e-05,
+      "loss": 1.2872,
+      "step": 161
+    },
+    {
+      "epoch": 0.09468839044348652,
+      "grad_norm": 2.0930467564564874,
+      "learning_rate": 1e-05,
+      "loss": 1.3388,
+      "step": 162
+    },
+    {
+      "epoch": 0.095272886680792,
+      "grad_norm": 2.0990204068816873,
+      "learning_rate": 1e-05,
+      "loss": 1.3085,
+      "step": 163
+    },
+    {
+      "epoch": 0.09585738291809746,
+      "grad_norm": 2.0944163595135596,
+      "learning_rate": 1e-05,
+      "loss": 1.3791,
+      "step": 164
+    },
+    {
+      "epoch": 0.09644187915540293,
+      "grad_norm": 2.260848568528631,
+      "learning_rate": 1e-05,
+      "loss": 1.3126,
+      "step": 165
+    },
+    {
+      "epoch": 0.0970263753927084,
+      "grad_norm": 2.0505781331907422,
+      "learning_rate": 1e-05,
+      "loss": 1.2999,
+      "step": 166
+    },
+    {
+      "epoch": 0.09761087163001388,
+      "grad_norm": 2.235737272001482,
+      "learning_rate": 1e-05,
+      "loss": 1.362,
+      "step": 167
+    },
+    {
+      "epoch": 0.09819536786731936,
+      "grad_norm": 2.1077529013831073,
+      "learning_rate": 1e-05,
+      "loss": 1.3265,
+      "step": 168
+    },
+    {
+      "epoch": 0.09877986410462483,
+      "grad_norm": 2.161630665030747,
+      "learning_rate": 1e-05,
+      "loss": 1.314,
+      "step": 169
+    },
+    {
+      "epoch": 0.0993643603419303,
+      "grad_norm": 2.173358238177831,
+      "learning_rate": 1e-05,
+      "loss": 1.2667,
+      "step": 170
+    },
+    {
+      "epoch": 0.09994885657923577,
+      "grad_norm": 2.1140520367127222,
+      "learning_rate": 1e-05,
+      "loss": 1.2893,
+      "step": 171
+    },
+    {
+      "epoch": 0.10053335281654124,
+      "grad_norm": 1.9864003484344632,
+      "learning_rate": 1e-05,
+      "loss": 1.3292,
+      "step": 172
+    },
+    {
+      "epoch": 0.10111784905384671,
+      "grad_norm": 2.3099149889217836,
+      "learning_rate": 1e-05,
+      "loss": 1.3764,
+      "step": 173
+    },
+    {
+      "epoch": 0.10170234529115219,
+      "grad_norm": 2.002925214700107,
+      "learning_rate": 1e-05,
+      "loss": 1.3629,
+      "step": 174
+    },
+    {
+      "epoch": 0.10228684152845766,
+      "grad_norm": 2.1630008207136004,
+      "learning_rate": 1e-05,
+      "loss": 1.3371,
+      "step": 175
+    },
+    {
+      "epoch": 0.10287133776576313,
+      "grad_norm": 2.1087481489379942,
+      "learning_rate": 1e-05,
+      "loss": 1.2549,
+      "step": 176
+    },
+    {
+      "epoch": 0.10345583400306861,
+      "grad_norm": 1.97791964843698,
+      "learning_rate": 1e-05,
+      "loss": 1.3614,
+      "step": 177
+    },
+    {
+      "epoch": 0.10404033024037408,
+      "grad_norm": 2.175752033732102,
+      "learning_rate": 1e-05,
+      "loss": 1.2943,
+      "step": 178
+    },
+    {
+      "epoch": 0.10462482647767955,
+      "grad_norm": 2.027633109237481,
+      "learning_rate": 1e-05,
+      "loss": 1.3434,
+      "step": 179
+    },
+    {
+      "epoch": 0.10520932271498502,
+      "grad_norm": 2.5874981916649444,
+      "learning_rate": 1e-05,
+      "loss": 1.2909,
+      "step": 180
+    },
+    {
+      "epoch": 0.1057938189522905,
+      "grad_norm": 2.5290062200000984,
+      "learning_rate": 1e-05,
+      "loss": 1.3149,
+      "step": 181
+    },
+    {
+      "epoch": 0.10637831518959596,
+      "grad_norm": 2.001761982207369,
+      "learning_rate": 1e-05,
+      "loss": 1.2997,
+      "step": 182
+    },
+    {
+      "epoch": 0.10696281142690144,
+      "grad_norm": 1.981546864360565,
+      "learning_rate": 1e-05,
+      "loss": 1.3082,
+      "step": 183
+    },
+    {
+      "epoch": 0.1075473076642069,
+      "grad_norm": 2.067415015285412,
+      "learning_rate": 1e-05,
+      "loss": 1.3135,
+      "step": 184
+    },
+    {
+      "epoch": 0.10813180390151238,
+      "grad_norm": 2.1658757525333834,
+      "learning_rate": 1e-05,
+      "loss": 1.2575,
+      "step": 185
+    },
+    {
+      "epoch": 0.10871630013881786,
+      "grad_norm": 2.2618429365830073,
+      "learning_rate": 1e-05,
+      "loss": 1.277,
+      "step": 186
+    },
+    {
+      "epoch": 0.10930079637612333,
+      "grad_norm": 2.2750216896364175,
+      "learning_rate": 1e-05,
+      "loss": 1.33,
+      "step": 187
+    },
+    {
+      "epoch": 0.1098852926134288,
+      "grad_norm": 2.2344266970445648,
+      "learning_rate": 1e-05,
+      "loss": 1.2694,
+      "step": 188
+    },
+    {
+      "epoch": 0.11046978885073427,
+      "grad_norm": 2.1897134580522377,
+      "learning_rate": 1e-05,
+      "loss": 1.2786,
+      "step": 189
+    },
+    {
+      "epoch": 0.11105428508803974,
+      "grad_norm": 2.2376157503379837,
+      "learning_rate": 1e-05,
+      "loss": 1.2915,
+      "step": 190
+    },
+    {
+      "epoch": 0.11163878132534522,
+      "grad_norm": 2.19924670747834,
+      "learning_rate": 1e-05,
+      "loss": 1.3046,
+      "step": 191
+    },
+    {
+      "epoch": 0.11222327756265069,
+      "grad_norm": 2.0337741567945744,
+      "learning_rate": 1e-05,
+      "loss": 1.3003,
+      "step": 192
+    },
+    {
+      "epoch": 0.11280777379995616,
+      "grad_norm": 2.2298783511600475,
+      "learning_rate": 1e-05,
+      "loss": 1.3274,
+      "step": 193
+    },
+    {
+      "epoch": 0.11339227003726164,
+      "grad_norm": 2.177582903353876,
+      "learning_rate": 1e-05,
+      "loss": 1.297,
+      "step": 194
+    },
+    {
+      "epoch": 0.11397676627456711,
+      "grad_norm": 2.056323574813421,
+      "learning_rate": 1e-05,
+      "loss": 1.3152,
+      "step": 195
+    },
+    {
+      "epoch": 0.11456126251187258,
+      "grad_norm": 1.9548263675911643,
+      "learning_rate": 1e-05,
+      "loss": 1.3326,
+      "step": 196
+    },
+    {
+      "epoch": 0.11514575874917805,
+      "grad_norm": 2.263843718400184,
+      "learning_rate": 1e-05,
+      "loss": 1.2846,
+      "step": 197
+    },
+    {
+      "epoch": 0.11573025498648352,
+      "grad_norm": 1.971698078110625,
+      "learning_rate": 1e-05,
+      "loss": 1.3539,
+      "step": 198
+    },
+    {
+      "epoch": 0.116314751223789,
+      "grad_norm": 2.029529010327694,
+      "learning_rate": 1e-05,
+      "loss": 1.2968,
+      "step": 199
+    },
+    {
+      "epoch": 0.11689924746109447,
+      "grad_norm": 2.153864654126516,
+      "learning_rate": 1e-05,
+      "loss": 1.3294,
+      "step": 200
+    },
+    {
+      "epoch": 0.11748374369839994,
+      "grad_norm": 2.1964807806319846,
+      "learning_rate": 1e-05,
+      "loss": 1.259,
+      "step": 201
+    },
+    {
+      "epoch": 0.11806823993570541,
+      "grad_norm": 2.205255963973796,
+      "learning_rate": 1e-05,
+      "loss": 1.3395,
+      "step": 202
+    },
+    {
+      "epoch": 0.11865273617301089,
+      "grad_norm": 2.163928534859531,
+      "learning_rate": 1e-05,
+      "loss": 1.2687,
+      "step": 203
+    },
+    {
+      "epoch": 0.11923723241031636,
+      "grad_norm": 2.0533916338901865,
+      "learning_rate": 1e-05,
+      "loss": 1.2629,
+      "step": 204
+    },
+    {
+      "epoch": 0.11982172864762183,
+      "grad_norm": 2.016182166231378,
+      "learning_rate": 1e-05,
+      "loss": 1.2681,
+      "step": 205
+    },
+    {
+      "epoch": 0.1204062248849273,
+      "grad_norm": 2.0676764788616677,
+      "learning_rate": 1e-05,
+      "loss": 1.3458,
+      "step": 206
+    },
+    {
+      "epoch": 0.12099072112223277,
+      "grad_norm": 2.156001247098662,
+      "learning_rate": 1e-05,
+      "loss": 1.3,
+      "step": 207
+    },
+    {
+      "epoch": 0.12157521735953825,
+      "grad_norm": 2.270476516964796,
+      "learning_rate": 1e-05,
+      "loss": 1.3185,
+      "step": 208
+    },
+    {
+      "epoch": 0.12215971359684372,
+      "grad_norm": 2.103472203913147,
+      "learning_rate": 1e-05,
+      "loss": 1.2969,
+      "step": 209
+    },
+    {
+      "epoch": 0.12274420983414919,
+      "grad_norm": 2.2992839732099077,
+      "learning_rate": 1e-05,
+      "loss": 1.2673,
+      "step": 210
+    },
+    {
+      "epoch": 0.12332870607145467,
+      "grad_norm": 2.1248865051758976,
+      "learning_rate": 1e-05,
+      "loss": 1.3376,
+      "step": 211
+    },
+    {
+      "epoch": 0.12391320230876014,
+      "grad_norm": 2.215882624620382,
+      "learning_rate": 1e-05,
+      "loss": 1.3223,
+      "step": 212
+    },
+    {
+      "epoch": 0.12449769854606561,
+      "grad_norm": 2.070644564669489,
+      "learning_rate": 1e-05,
+      "loss": 1.3029,
+      "step": 213
+    },
+    {
+      "epoch": 0.12508219478337107,
+      "grad_norm": 2.0020856392276043,
+      "learning_rate": 1e-05,
+      "loss": 1.2544,
+      "step": 214
+    },
+    {
+      "epoch": 0.12566669102067654,
+      "grad_norm": 1.9569023772133114,
+      "learning_rate": 1e-05,
+      "loss": 1.3029,
+      "step": 215
+    },
+    {
+      "epoch": 0.12625118725798204,
+      "grad_norm": 2.2032138308058897,
+      "learning_rate": 1e-05,
+      "loss": 1.3036,
+      "step": 216
+    },
+    {
+      "epoch": 0.1268356834952875,
+      "grad_norm": 2.121404104774535,
+      "learning_rate": 1e-05,
+      "loss": 1.3315,
+      "step": 217
+    },
+    {
+      "epoch": 0.12742017973259298,
+      "grad_norm": 2.140419406758026,
+      "learning_rate": 1e-05,
+      "loss": 1.2963,
+      "step": 218
+    },
+    {
+      "epoch": 0.12800467596989845,
+      "grad_norm": 1.8824960865185534,
+      "learning_rate": 1e-05,
+      "loss": 1.316,
+      "step": 219
+    },
+    {
+      "epoch": 0.12858917220720392,
+      "grad_norm": 2.080018005868092,
+      "learning_rate": 1e-05,
+      "loss": 1.2997,
+      "step": 220
+    },
+    {
+      "epoch": 0.1291736684445094,
+      "grad_norm": 2.1466392848171023,
+      "learning_rate": 1e-05,
+      "loss": 1.2868,
+      "step": 221
+    },
+    {
+      "epoch": 0.12975816468181486,
+      "grad_norm": 2.0862797448562085,
+      "learning_rate": 1e-05,
+      "loss": 1.2657,
+      "step": 222
+    },
+    {
+      "epoch": 0.13034266091912033,
+      "grad_norm": 2.1449219828372623,
+      "learning_rate": 1e-05,
+      "loss": 1.2777,
+      "step": 223
+    },
+    {
+      "epoch": 0.1309271571564258,
+      "grad_norm": 2.0652658121090104,
+      "learning_rate": 1e-05,
+      "loss": 1.3272,
+      "step": 224
+    },
+    {
+      "epoch": 0.13151165339373128,
+      "grad_norm": 2.0771354810742872,
+      "learning_rate": 1e-05,
+      "loss": 1.3401,
+      "step": 225
+    },
+    {
+      "epoch": 0.13209614963103675,
+      "grad_norm": 2.0985843070722208,
+      "learning_rate": 1e-05,
+      "loss": 1.2559,
+      "step": 226
+    },
+    {
+      "epoch": 0.13268064586834222,
+      "grad_norm": 1.8922450951001708,
+      "learning_rate": 1e-05,
+      "loss": 1.3097,
+      "step": 227
+    },
+    {
+      "epoch": 0.1332651421056477,
+      "grad_norm": 2.078948999532085,
+      "learning_rate": 1e-05,
+      "loss": 1.3198,
+      "step": 228
+    },
+    {
+      "epoch": 0.13384963834295316,
+      "grad_norm": 2.1813589274238137,
+      "learning_rate": 1e-05,
+      "loss": 1.2821,
+      "step": 229
+    },
+    {
+      "epoch": 0.13443413458025863,
+      "grad_norm": 2.0046343421340693,
+      "learning_rate": 1e-05,
+      "loss": 1.306,
+      "step": 230
+    },
+    {
+      "epoch": 0.1350186308175641,
+      "grad_norm": 2.095001502727408,
+      "learning_rate": 1e-05,
+      "loss": 1.2919,
+      "step": 231
+    },
+    {
+      "epoch": 0.13560312705486957,
+      "grad_norm": 2.104493257341596,
+      "learning_rate": 1e-05,
+      "loss": 1.3217,
+      "step": 232
+    },
+    {
+      "epoch": 0.13618762329217507,
+      "grad_norm": 2.0803554787442766,
+      "learning_rate": 1e-05,
+      "loss": 1.2623,
+      "step": 233
+    },
+    {
+      "epoch": 0.13677211952948054,
+      "grad_norm": 2.028048350479765,
+      "learning_rate": 1e-05,
+      "loss": 1.283,
+      "step": 234
+    },
+    {
+      "epoch": 0.137356615766786,
+      "grad_norm": 2.059413037500954,
+      "learning_rate": 1e-05,
+      "loss": 1.3056,
+      "step": 235
+    },
+    {
+      "epoch": 0.13794111200409148,
+      "grad_norm": 2.0771318610118517,
+      "learning_rate": 1e-05,
+      "loss": 1.2996,
+      "step": 236
+    },
+    {
+      "epoch": 0.13852560824139695,
+      "grad_norm": 1.998930537856237,
+      "learning_rate": 1e-05,
+      "loss": 1.3124,
+      "step": 237
+    },
+    {
+      "epoch": 0.13911010447870242,
+      "grad_norm": 2.2885728207311002,
+      "learning_rate": 1e-05,
+      "loss": 1.3086,
+      "step": 238
+    },
+    {
+      "epoch": 0.1396946007160079,
+      "grad_norm": 2.1461012449070567,
+      "learning_rate": 1e-05,
+      "loss": 1.2732,
+      "step": 239
+    },
+    {
+      "epoch": 0.14027909695331336,
+      "grad_norm": 2.2417496514614736,
+      "learning_rate": 1e-05,
+      "loss": 1.3047,
+      "step": 240
+    },
+    {
+      "epoch": 0.14086359319061884,
+      "grad_norm": 2.164361889084499,
+      "learning_rate": 1e-05,
+      "loss": 1.3599,
+      "step": 241
+    },
+    {
+      "epoch": 0.1414480894279243,
+      "grad_norm": 1.996733327286818,
+      "learning_rate": 1e-05,
+      "loss": 1.3554,
+      "step": 242
+    },
+    {
+      "epoch": 0.14203258566522978,
+      "grad_norm": 2.0937054039601697,
+      "learning_rate": 1e-05,
+      "loss": 1.3294,
+      "step": 243
+    },
+    {
+      "epoch": 0.14261708190253525,
+      "grad_norm": 2.2895289094807256,
+      "learning_rate": 1e-05,
+      "loss": 1.34,
+      "step": 244
+    },
+    {
+      "epoch": 0.14320157813984072,
+      "grad_norm": 2.0694679852919857,
+      "learning_rate": 1e-05,
+      "loss": 1.3242,
+      "step": 245
+    },
+    {
+      "epoch": 0.1437860743771462,
+      "grad_norm": 2.1308559789514643,
+      "learning_rate": 1e-05,
+      "loss": 1.3103,
+      "step": 246
+    },
+    {
+      "epoch": 0.14437057061445166,
+      "grad_norm": 2.0176838695090806,
+      "learning_rate": 1e-05,
+      "loss": 1.2454,
+      "step": 247
+    },
+    {
+      "epoch": 0.14495506685175713,
+      "grad_norm": 2.297377314107247,
+      "learning_rate": 1e-05,
+      "loss": 1.2538,
+      "step": 248
+    },
+    {
+      "epoch": 0.1455395630890626,
+      "grad_norm": 1.9717740351911044,
+      "learning_rate": 1e-05,
+      "loss": 1.2973,
+      "step": 249
+    },
+    {
+      "epoch": 0.1461240593263681,
+      "grad_norm": 2.1173719817533487,
+      "learning_rate": 1e-05,
+      "loss": 1.3117,
+      "step": 250
+    },
+    {
+      "epoch": 0.14670855556367357,
+      "grad_norm": 1.918210258882671,
+      "learning_rate": 1e-05,
+      "loss": 1.3121,
+      "step": 251
+    },
+    {
+      "epoch": 0.14729305180097904,
+      "grad_norm": 2.07771375689505,
+      "learning_rate": 1e-05,
+      "loss": 1.3457,
+      "step": 252
+    },
+    {
+      "epoch": 0.1478775480382845,
+      "grad_norm": 2.183106156298185,
+      "learning_rate": 1e-05,
+      "loss": 1.2785,
+      "step": 253
+    },
+    {
+      "epoch": 0.14846204427558998,
+      "grad_norm": 2.0979572799310584,
+      "learning_rate": 1e-05,
+      "loss": 1.2684,
+      "step": 254
+    },
+    {
+      "epoch": 0.14904654051289545,
+      "grad_norm": 1.875950072152609,
+      "learning_rate": 1e-05,
+      "loss": 1.2631,
+      "step": 255
+    },
+    {
+      "epoch": 0.14963103675020092,
+      "grad_norm": 1.9091779454335909,
+      "learning_rate": 1e-05,
+      "loss": 1.3176,
+      "step": 256
+    },
+    {
+      "epoch": 0.1502155329875064,
+      "grad_norm": 1.9012080498697779,
+      "learning_rate": 1e-05,
+      "loss": 1.3387,
+      "step": 257
+    },
+    {
+      "epoch": 0.15080002922481187,
+      "grad_norm": 1.8573972993864167,
+      "learning_rate": 1e-05,
+      "loss": 1.2556,
+      "step": 258
+    },
+    {
+      "epoch": 0.15138452546211734,
+      "grad_norm": 1.9117519630550235,
+      "learning_rate": 1e-05,
+      "loss": 1.333,
+      "step": 259
+    },
+    {
+      "epoch": 0.1519690216994228,
+      "grad_norm": 2.112926819463512,
+      "learning_rate": 1e-05,
+      "loss": 1.2873,
+      "step": 260
+    },
+    {
+      "epoch": 0.15255351793672828,
+      "grad_norm": 2.023374809075834,
+      "learning_rate": 1e-05,
+      "loss": 1.3103,
+      "step": 261
+    },
+    {
+      "epoch": 0.15313801417403375,
+      "grad_norm": 2.1571839863573485,
+      "learning_rate": 1e-05,
+      "loss": 1.2661,
+      "step": 262
+    },
+    {
+      "epoch": 0.15372251041133922,
+      "grad_norm": 2.110535027131159,
+      "learning_rate": 1e-05,
+      "loss": 1.3057,
+      "step": 263
+    },
+    {
+      "epoch": 0.1543070066486447,
+      "grad_norm": 1.9505908560683503,
+      "learning_rate": 1e-05,
+      "loss": 1.2962,
+      "step": 264
+    },
+    {
+      "epoch": 0.15489150288595016,
+      "grad_norm": 2.14607270813795,
+      "learning_rate": 1e-05,
+      "loss": 1.2749,
+      "step": 265
+    },
+    {
+      "epoch": 0.15547599912325563,
+      "grad_norm": 1.852818048391368,
+      "learning_rate": 1e-05,
+      "loss": 1.2708,
+      "step": 266
+    },
+    {
+      "epoch": 0.15606049536056113,
+      "grad_norm": 2.010132543935311,
+      "learning_rate": 1e-05,
+      "loss": 1.2629,
+      "step": 267
+    },
+    {
+      "epoch": 0.1566449915978666,
+      "grad_norm": 1.9388045239195144,
+      "learning_rate": 1e-05,
+      "loss": 1.2842,
+      "step": 268
+    },
+    {
+      "epoch": 0.15722948783517207,
+      "grad_norm": 2.0617913809095065,
+      "learning_rate": 1e-05,
+      "loss": 1.3141,
+      "step": 269
+    },
+    {
+      "epoch": 0.15781398407247754,
+      "grad_norm": 2.1328157596194717,
+      "learning_rate": 1e-05,
+      "loss": 1.3068,
+      "step": 270
+    },
+    {
+      "epoch": 0.158398480309783,
+      "grad_norm": 1.9008830047721657,
+      "learning_rate": 1e-05,
+      "loss": 1.296,
+      "step": 271
+    },
+    {
+      "epoch": 0.15898297654708848,
+      "grad_norm": 2.1932056950263767,
+      "learning_rate": 1e-05,
+      "loss": 1.3442,
+      "step": 272
+    },
+    {
+      "epoch": 0.15956747278439395,
+      "grad_norm": 2.2117955137023593,
+      "learning_rate": 1e-05,
+      "loss": 1.2583,
+      "step": 273
+    },
+    {
+      "epoch": 0.16015196902169943,
+      "grad_norm": 1.992350300523065,
+      "learning_rate": 1e-05,
+      "loss": 1.2789,
+      "step": 274
+    },
+    {
+      "epoch": 0.1607364652590049,
+      "grad_norm": 2.167435757776945,
+      "learning_rate": 1e-05,
+      "loss": 1.248,
+      "step": 275
+    },
+    {
+      "epoch": 0.16132096149631037,
+      "grad_norm": 2.0482148557762816,
+      "learning_rate": 1e-05,
+      "loss": 1.2559,
+      "step": 276
+    },
+    {
+      "epoch": 0.16190545773361584,
+      "grad_norm": 2.1404150776553514,
+      "learning_rate": 1e-05,
+      "loss": 1.3408,
+      "step": 277
+    },
+    {
+      "epoch": 0.1624899539709213,
+      "grad_norm": 2.0428884775998286,
+      "learning_rate": 1e-05,
+      "loss": 1.2172,
+      "step": 278
+    },
+    {
+      "epoch": 0.16307445020822678,
+      "grad_norm": 2.134511606691306,
+      "learning_rate": 1e-05,
+      "loss": 1.2897,
+      "step": 279
+    },
+    {
+      "epoch": 0.16365894644553225,
+      "grad_norm": 2.172976285915068,
+      "learning_rate": 1e-05,
+      "loss": 1.2745,
+      "step": 280
+    },
+    {
+      "epoch": 0.16424344268283772,
+      "grad_norm": 2.198698683685309,
+      "learning_rate": 1e-05,
+      "loss": 1.2666,
+      "step": 281
+    },
+    {
+      "epoch": 0.1648279389201432,
+      "grad_norm": 2.0467973941746993,
+      "learning_rate": 1e-05,
+      "loss": 1.2944,
+      "step": 282
+    },
+    {
+      "epoch": 0.16541243515744866,
+      "grad_norm": 2.1286781191694746,
+      "learning_rate": 1e-05,
+      "loss": 1.2562,
+      "step": 283
+    },
+    {
+      "epoch": 0.16599693139475413,
+      "grad_norm": 1.9860948094839945,
+      "learning_rate": 1e-05,
+      "loss": 1.3353,
+      "step": 284
+    },
+    {
+      "epoch": 0.16658142763205963,
+      "grad_norm": 2.004075553824083,
+      "learning_rate": 1e-05,
+      "loss": 1.3664,
+      "step": 285
+    },
+    {
+      "epoch": 0.1671659238693651,
+      "grad_norm": 2.0595719890349686,
+      "learning_rate": 1e-05,
+      "loss": 1.2888,
+      "step": 286
+    },
+    {
+      "epoch": 0.16775042010667057,
+      "grad_norm": 1.8514586942745248,
+      "learning_rate": 1e-05,
+      "loss": 1.269,
+      "step": 287
+    },
+    {
+      "epoch": 0.16833491634397604,
+      "grad_norm": 1.9949536608779335,
+      "learning_rate": 1e-05,
+      "loss": 1.2906,
+      "step": 288
+    },
+    {
+      "epoch": 0.16891941258128151,
+      "grad_norm": 2.0598594653759603,
+      "learning_rate": 1e-05,
+      "loss": 1.257,
+      "step": 289
+    },
+    {
+      "epoch": 0.16950390881858698,
+      "grad_norm": 2.215605143698448,
+      "learning_rate": 1e-05,
+      "loss": 1.3164,
+      "step": 290
+    },
+    {
+      "epoch": 0.17008840505589246,
+      "grad_norm": 2.2096681603364483,
+      "learning_rate": 1e-05,
+      "loss": 1.3056,
+      "step": 291
+    },
+    {
+      "epoch": 0.17067290129319793,
+      "grad_norm": 2.2717503144037017,
+      "learning_rate": 1e-05,
+      "loss": 1.2831,
+      "step": 292
+    },
+    {
+      "epoch": 0.1712573975305034,
+      "grad_norm": 2.0819768452392857,
+      "learning_rate": 1e-05,
+      "loss": 1.3022,
+      "step": 293
+    },
+    {
+      "epoch": 0.17184189376780887,
+      "grad_norm": 2.3287236530422093,
+      "learning_rate": 1e-05,
+      "loss": 1.3139,
+      "step": 294
+    },
+    {
+      "epoch": 0.17242639000511434,
+      "grad_norm": 1.9776424006738431,
+      "learning_rate": 1e-05,
+      "loss": 1.3087,
+      "step": 295
+    },
+    {
+      "epoch": 0.1730108862424198,
+      "grad_norm": 2.0730338260885275,
+      "learning_rate": 1e-05,
+      "loss": 1.3469,
+      "step": 296
+    },
+    {
+      "epoch": 0.17359538247972528,
+      "grad_norm": 2.2449714018481544,
+      "learning_rate": 1e-05,
+      "loss": 1.2786,
+      "step": 297
+    },
+    {
+      "epoch": 0.17417987871703075,
+      "grad_norm": 1.9752771565647496,
+      "learning_rate": 1e-05,
+      "loss": 1.3122,
+      "step": 298
+    },
+    {
+      "epoch": 0.17476437495433622,
+      "grad_norm": 1.954892855440776,
+      "learning_rate": 1e-05,
+      "loss": 1.2691,
+      "step": 299
+    },
+    {
+      "epoch": 0.1753488711916417,
+      "grad_norm": 2.4196921508603655,
+      "learning_rate": 1e-05,
+      "loss": 1.2391,
+      "step": 300
+    },
+    {
+      "epoch": 0.17593336742894716,
+      "grad_norm": 2.026464953290186,
+      "learning_rate": 1e-05,
+      "loss": 1.3033,
+      "step": 301
+    },
+    {
+      "epoch": 0.17651786366625266,
+      "grad_norm": 2.13499677161983,
+      "learning_rate": 1e-05,
+      "loss": 1.2644,
+      "step": 302
+    },
+    {
+      "epoch": 0.17710235990355813,
+      "grad_norm": 2.007415375163882,
+      "learning_rate": 1e-05,
+      "loss": 1.3007,
+      "step": 303
+    },
+    {
+      "epoch": 0.1776868561408636,
+      "grad_norm": 2.0595706705209667,
+      "learning_rate": 1e-05,
+      "loss": 1.2638,
+      "step": 304
+    },
+    {
+      "epoch": 0.17827135237816907,
+      "grad_norm": 1.985428011831755,
+      "learning_rate": 1e-05,
+      "loss": 1.2753,
+      "step": 305
+    },
+    {
+      "epoch": 0.17885584861547454,
+      "grad_norm": 2.056289427044583,
+      "learning_rate": 1e-05,
+      "loss": 1.353,
+      "step": 306
+    },
+    {
+      "epoch": 0.17944034485278001,
+      "grad_norm": 2.31446285021816,
+      "learning_rate": 1e-05,
+      "loss": 1.2455,
+      "step": 307
+    },
+    {
+      "epoch": 0.18002484109008549,
+      "grad_norm": 2.033869117352267,
+      "learning_rate": 1e-05,
+      "loss": 1.2344,
+      "step": 308
+    },
+    {
+      "epoch": 0.18060933732739096,
+      "grad_norm": 1.8466281173787475,
+      "learning_rate": 1e-05,
+      "loss": 1.2681,
+      "step": 309
+    },
+    {
+      "epoch": 0.18119383356469643,
+      "grad_norm": 2.2876665522996547,
+      "learning_rate": 1e-05,
+      "loss": 1.2806,
+      "step": 310
+    },
+    {
+      "epoch": 0.1817783298020019,
+      "grad_norm": 2.0548726219039737,
+      "learning_rate": 1e-05,
+      "loss": 1.2283,
+      "step": 311
+    },
+    {
+      "epoch": 0.18236282603930737,
+      "grad_norm": 2.0667147713529825,
+      "learning_rate": 1e-05,
+      "loss": 1.2615,
+      "step": 312
+    },
+    {
+      "epoch": 0.18294732227661284,
+      "grad_norm": 1.9119272528996851,
+      "learning_rate": 1e-05,
+      "loss": 1.2933,
+      "step": 313
+    },
+    {
+      "epoch": 0.1835318185139183,
+      "grad_norm": 2.2226129894086215,
+      "learning_rate": 1e-05,
+      "loss": 1.2946,
+      "step": 314
+    },
+    {
+      "epoch": 0.18411631475122378,
+      "grad_norm": 2.2391568493225122,
+      "learning_rate": 1e-05,
+      "loss": 1.2963,
+      "step": 315
+    },
+    {
+      "epoch": 0.18470081098852925,
+      "grad_norm": 2.0179793916620845,
+      "learning_rate": 1e-05,
+      "loss": 1.2967,
+      "step": 316
+    },
+    {
+      "epoch": 0.18528530722583472,
+      "grad_norm": 1.9953151540273535,
+      "learning_rate": 1e-05,
+      "loss": 1.3217,
+      "step": 317
+    },
+    {
+      "epoch": 0.1858698034631402,
+      "grad_norm": 1.9523149224113234,
+      "learning_rate": 1e-05,
+      "loss": 1.277,
+      "step": 318
+    },
+    {
+      "epoch": 0.1864542997004457,
+      "grad_norm": 1.8470657713109726,
+      "learning_rate": 1e-05,
+      "loss": 1.2765,
+      "step": 319
+    },
+    {
+      "epoch": 0.18703879593775116,
+      "grad_norm": 1.9859711129103998,
+      "learning_rate": 1e-05,
+      "loss": 1.2616,
+      "step": 320
+    },
+    {
+      "epoch": 0.18762329217505663,
+      "grad_norm": 2.0379471618886957,
+      "learning_rate": 1e-05,
+      "loss": 1.2893,
+      "step": 321
+    },
+    {
+      "epoch": 0.1882077884123621,
+      "grad_norm": 1.9571408264734231,
+      "learning_rate": 1e-05,
+      "loss": 1.3254,
+      "step": 322
+    },
+    {
+      "epoch": 0.18879228464966757,
+      "grad_norm": 1.919409788322633,
+      "learning_rate": 1e-05,
+      "loss": 1.2872,
+      "step": 323
+    },
+    {
+      "epoch": 0.18937678088697305,
+      "grad_norm": 2.0757881798601057,
+      "learning_rate": 1e-05,
+      "loss": 1.3301,
+      "step": 324
+    },
+    {
+      "epoch": 0.18996127712427852,
+      "grad_norm": 1.925703964787337,
+      "learning_rate": 1e-05,
+      "loss": 1.2498,
+      "step": 325
+    },
+    {
+      "epoch": 0.190545773361584,
+      "grad_norm": 2.1842423873606562,
+      "learning_rate": 1e-05,
+      "loss": 1.2579,
+      "step": 326
+    },
+    {
+      "epoch": 0.19113026959888946,
+      "grad_norm": 1.9753765446565863,
+      "learning_rate": 1e-05,
+      "loss": 1.2111,
+      "step": 327
+    },
+    {
+      "epoch": 0.19171476583619493,
+      "grad_norm": 1.8981455329887007,
+      "learning_rate": 1e-05,
+      "loss": 1.2676,
+      "step": 328
+    },
+    {
+      "epoch": 0.1922992620735004,
+      "grad_norm": 2.2080540817312238,
+      "learning_rate": 1e-05,
+      "loss": 1.2717,
+      "step": 329
+    },
+    {
+      "epoch": 0.19288375831080587,
+      "grad_norm": 1.921841933634693,
+      "learning_rate": 1e-05,
+      "loss": 1.3193,
+      "step": 330
+    },
+    {
+      "epoch": 0.19346825454811134,
+      "grad_norm": 1.8859608782972828,
+      "learning_rate": 1e-05,
+      "loss": 1.2786,
+      "step": 331
+    },
+    {
+      "epoch": 0.1940527507854168,
+      "grad_norm": 1.7995817724177323,
+      "learning_rate": 1e-05,
+      "loss": 1.2591,
+      "step": 332
+    },
+    {
+      "epoch": 0.19463724702272228,
+      "grad_norm": 2.0581026553637294,
+      "learning_rate": 1e-05,
+      "loss": 1.2504,
+      "step": 333
+    },
+    {
+      "epoch": 0.19522174326002775,
+      "grad_norm": 2.012028456291,
+      "learning_rate": 1e-05,
+      "loss": 1.31,
+      "step": 334
+    },
+    {
+      "epoch": 0.19580623949733322,
+      "grad_norm": 2.207222184244356,
+      "learning_rate": 1e-05,
+      "loss": 1.2821,
+      "step": 335
+    },
+    {
+      "epoch": 0.19639073573463872,
+      "grad_norm": 2.139114260824696,
+      "learning_rate": 1e-05,
+      "loss": 1.2813,
+      "step": 336
+    },
+    {
+      "epoch": 0.1969752319719442,
+      "grad_norm": 2.0805380652147623,
+      "learning_rate": 1e-05,
+      "loss": 1.3582,
+      "step": 337
+    },
+    {
+      "epoch": 0.19755972820924966,
+      "grad_norm": 2.1924555456453567,
+      "learning_rate": 1e-05,
+      "loss": 1.2807,
+      "step": 338
+    },
+    {
+      "epoch": 0.19814422444655513,
+      "grad_norm": 2.2098845540332235,
+      "learning_rate": 1e-05,
+      "loss": 1.2567,
+      "step": 339
+    },
+    {
+      "epoch": 0.1987287206838606,
+      "grad_norm": 1.8291165559520288,
+      "learning_rate": 1e-05,
+      "loss": 1.2661,
+      "step": 340
+    },
+    {
+      "epoch": 0.19931321692116608,
+      "grad_norm": 1.9152114586374016,
+      "learning_rate": 1e-05,
+      "loss": 1.3184,
+      "step": 341
+    },
+    {
+      "epoch": 0.19989771315847155,
+      "grad_norm": 2.2052228092474864,
+      "learning_rate": 1e-05,
+      "loss": 1.2461,
+      "step": 342
+    },
+    {
+      "epoch": 0.20048220939577702,
+      "grad_norm": 2.2140812469347777,
+      "learning_rate": 1e-05,
+      "loss": 1.3002,
+      "step": 343
+    },
+    {
+      "epoch": 0.2010667056330825,
+      "grad_norm": 2.115905050577708,
+      "learning_rate": 1e-05,
+      "loss": 1.3103,
+      "step": 344
+    },
+    {
+      "epoch": 0.20165120187038796,
+      "grad_norm": 1.9492696474081166,
+      "learning_rate": 1e-05,
+      "loss": 1.3085,
+      "step": 345
+    },
+    {
+      "epoch": 0.20223569810769343,
+      "grad_norm": 1.8825965180676585,
+      "learning_rate": 1e-05,
+      "loss": 1.3078,
+      "step": 346
+    },
+    {
+      "epoch": 0.2028201943449989,
+      "grad_norm": 1.984474697160862,
+      "learning_rate": 1e-05,
+      "loss": 1.2607,
+      "step": 347
+    },
+    {
+      "epoch": 0.20340469058230437,
+      "grad_norm": 2.087648640283265,
+      "learning_rate": 1e-05,
+      "loss": 1.2921,
+      "step": 348
+    },
+    {
+      "epoch": 0.20398918681960984,
+      "grad_norm": 2.1565029783230965,
+      "learning_rate": 1e-05,
+      "loss": 1.2842,
+      "step": 349
+    },
+    {
+      "epoch": 0.2045736830569153,
+      "grad_norm": 2.4351112445775858,
+      "learning_rate": 1e-05,
+      "loss": 1.2958,
+      "step": 350
+    },
+    {
+      "epoch": 0.20515817929422078,
+      "grad_norm": 1.994356006525233,
+      "learning_rate": 1e-05,
+      "loss": 1.3056,
+      "step": 351
+    },
+    {
+      "epoch": 0.20574267553152625,
+      "grad_norm": 1.933034432878314,
+      "learning_rate": 1e-05,
+      "loss": 1.3308,
+      "step": 352
+    },
+    {
+      "epoch": 0.20632717176883175,
+      "grad_norm": 1.8902905387324658,
+      "learning_rate": 1e-05,
+      "loss": 1.3146,
+      "step": 353
+    },
+    {
+      "epoch": 0.20691166800613722,
+      "grad_norm": 1.9758162128993135,
+      "learning_rate": 1e-05,
+      "loss": 1.2463,
+      "step": 354
+    },
+    {
+      "epoch": 0.2074961642434427,
+      "grad_norm": 2.08056590277236,
+      "learning_rate": 1e-05,
+      "loss": 1.2616,
+      "step": 355
+    },
+    {
+      "epoch": 0.20808066048074816,
+      "grad_norm": 1.9878170924817915,
+      "learning_rate": 1e-05,
+      "loss": 1.2753,
+      "step": 356
+    },
+    {
+      "epoch": 0.20866515671805363,
+      "grad_norm": 2.1153583311028092,
+      "learning_rate": 1e-05,
+      "loss": 1.2784,
+      "step": 357
+    },
+    {
+      "epoch": 0.2092496529553591,
+      "grad_norm": 2.0365294364738205,
+      "learning_rate": 1e-05,
+      "loss": 1.3311,
+      "step": 358
+    },
+    {
+      "epoch": 0.20983414919266458,
+      "grad_norm": 2.0534335645922477,
+      "learning_rate": 1e-05,
+      "loss": 1.2921,
+      "step": 359
+    },
+    {
+      "epoch": 0.21041864542997005,
+      "grad_norm": 1.8891405110667008,
+      "learning_rate": 1e-05,
+      "loss": 1.3246,
+      "step": 360
+    },
+    {
+      "epoch": 0.21100314166727552,
+      "grad_norm": 2.08292968392948,
+      "learning_rate": 1e-05,
+      "loss": 1.2914,
+      "step": 361
+    },
+    {
+      "epoch": 0.211587637904581,
+      "grad_norm": 2.111071838967622,
+      "learning_rate": 1e-05,
+      "loss": 1.2392,
+      "step": 362
+    },
+    {
+      "epoch": 0.21217213414188646,
+      "grad_norm": 1.9910737439984199,
+      "learning_rate": 1e-05,
+      "loss": 1.2455,
+      "step": 363
+    },
+    {
+      "epoch": 0.21275663037919193,
+      "grad_norm": 2.228520498969153,
+      "learning_rate": 1e-05,
+      "loss": 1.2891,
+      "step": 364
+    },
+    {
+      "epoch": 0.2133411266164974,
+      "grad_norm": 2.043395815900406,
+      "learning_rate": 1e-05,
+      "loss": 1.2434,
+      "step": 365
+    },
+    {
+      "epoch": 0.21392562285380287,
+      "grad_norm": 2.2263190272087447,
+      "learning_rate": 1e-05,
+      "loss": 1.2957,
+      "step": 366
+    },
+    {
+      "epoch": 0.21451011909110834,
+      "grad_norm": 2.0969013787955206,
+      "learning_rate": 1e-05,
+      "loss": 1.2463,
+      "step": 367
+    },
+    {
+      "epoch": 0.2150946153284138,
+      "grad_norm": 1.9660664020372078,
+      "learning_rate": 1e-05,
+      "loss": 1.2799,
+      "step": 368
+    },
+    {
+      "epoch": 0.21567911156571928,
+      "grad_norm": 2.297136404147855,
+      "learning_rate": 1e-05,
+      "loss": 1.2768,
+      "step": 369
+    },
+    {
+      "epoch": 0.21626360780302475,
+      "grad_norm": 1.9186326334460775,
+      "learning_rate": 1e-05,
+      "loss": 1.2547,
+      "step": 370
+    },
+    {
+      "epoch": 0.21684810404033025,
+      "grad_norm": 1.8616142758304943,
+      "learning_rate": 1e-05,
+      "loss": 1.2499,
+      "step": 371
+    },
+    {
+      "epoch": 0.21743260027763572,
+      "grad_norm": 2.0769318741971485,
+      "learning_rate": 1e-05,
+      "loss": 1.2995,
+      "step": 372
+    },
+    {
+      "epoch": 0.2180170965149412,
+      "grad_norm": 1.9683869316338465,
+      "learning_rate": 1e-05,
+      "loss": 1.2303,
+      "step": 373
+    },
+    {
+      "epoch": 0.21860159275224667,
+      "grad_norm": 2.054264077278263,
+      "learning_rate": 1e-05,
+      "loss": 1.2575,
+      "step": 374
+    },
+    {
+      "epoch": 0.21918608898955214,
+      "grad_norm": 2.1398278456501365,
+      "learning_rate": 1e-05,
+      "loss": 1.2954,
+      "step": 375
+    },
+    {
+      "epoch": 0.2197705852268576,
+      "grad_norm": 2.238546062591368,
+      "learning_rate": 1e-05,
+      "loss": 1.2068,
+      "step": 376
+    },
+    {
+      "epoch": 0.22035508146416308,
+      "grad_norm": 2.0102704415645194,
+      "learning_rate": 1e-05,
+      "loss": 1.2968,
+      "step": 377
+    },
+    {
+      "epoch": 0.22093957770146855,
+      "grad_norm": 1.9804921611582418,
+      "learning_rate": 1e-05,
+      "loss": 1.3142,
+      "step": 378
+    },
+    {
+      "epoch": 0.22152407393877402,
+      "grad_norm": 1.9490590434729507,
+      "learning_rate": 1e-05,
+      "loss": 1.2604,
+      "step": 379
+    },
+    {
+      "epoch": 0.2221085701760795,
+      "grad_norm": 1.9513669840817702,
+      "learning_rate": 1e-05,
+      "loss": 1.2565,
+      "step": 380
+    },
+    {
+      "epoch": 0.22269306641338496,
+      "grad_norm": 2.0244002086254675,
+      "learning_rate": 1e-05,
+      "loss": 1.2616,
+      "step": 381
+    },
+    {
+      "epoch": 0.22327756265069043,
+      "grad_norm": 2.06688414158411,
+      "learning_rate": 1e-05,
+      "loss": 1.2814,
+      "step": 382
+    },
+    {
+      "epoch": 0.2238620588879959,
+      "grad_norm": 2.1684604344369256,
+      "learning_rate": 1e-05,
+      "loss": 1.284,
+      "step": 383
+    },
+    {
+      "epoch": 0.22444655512530137,
+      "grad_norm": 2.0579742216342174,
+      "learning_rate": 1e-05,
+      "loss": 1.2947,
+      "step": 384
+    },
+    {
+      "epoch": 0.22503105136260684,
+      "grad_norm": 1.9464630794339237,
+      "learning_rate": 1e-05,
+      "loss": 1.2533,
+      "step": 385
+    },
+    {
+      "epoch": 0.22561554759991231,
+      "grad_norm": 2.1713018757216234,
+      "learning_rate": 1e-05,
+      "loss": 1.24,
+      "step": 386
+    },
+    {
+      "epoch": 0.22620004383721778,
+      "grad_norm": 2.0538820427593705,
+      "learning_rate": 1e-05,
+      "loss": 1.2479,
+      "step": 387
+    },
+    {
+      "epoch": 0.22678454007452328,
+      "grad_norm": 2.179331130312452,
+      "learning_rate": 1e-05,
+      "loss": 1.2836,
+      "step": 388
+    },
+    {
+      "epoch": 0.22736903631182875,
+      "grad_norm": 2.197905194592063,
+      "learning_rate": 1e-05,
+      "loss": 1.2976,
+      "step": 389
+    },
+    {
+      "epoch": 0.22795353254913422,
+      "grad_norm": 2.1111511303330532,
+      "learning_rate": 1e-05,
+      "loss": 1.2809,
+      "step": 390
+    },
+    {
+      "epoch": 0.2285380287864397,
+      "grad_norm": 1.9920839705873312,
+      "learning_rate": 1e-05,
+      "loss": 1.2979,
+      "step": 391
+    },
+    {
+      "epoch": 0.22912252502374517,
+      "grad_norm": 2.0189954862192265,
+      "learning_rate": 1e-05,
+      "loss": 1.2307,
+      "step": 392
+    },
+    {
+      "epoch": 0.22970702126105064,
+      "grad_norm": 1.9885440230575102,
+      "learning_rate": 1e-05,
+      "loss": 1.2523,
+      "step": 393
+    },
+    {
+      "epoch": 0.2302915174983561,
+      "grad_norm": 2.0073953676895266,
+      "learning_rate": 1e-05,
+      "loss": 1.2696,
+      "step": 394
+    },
+    {
+      "epoch": 0.23087601373566158,
+      "grad_norm": 2.036334006096198,
+      "learning_rate": 1e-05,
+      "loss": 1.2552,
+      "step": 395
+    },
+    {
+      "epoch": 0.23146050997296705,
+      "grad_norm": 2.012712813221627,
+      "learning_rate": 1e-05,
+      "loss": 1.3034,
+      "step": 396
+    },
+    {
+      "epoch": 0.23204500621027252,
+      "grad_norm": 1.9914633909911443,
+      "learning_rate": 1e-05,
+      "loss": 1.2938,
+      "step": 397
+    },
+    {
+      "epoch": 0.232629502447578,
+      "grad_norm": 1.9283481242705944,
+      "learning_rate": 1e-05,
+      "loss": 1.3061,
+      "step": 398
+    },
+    {
+      "epoch": 0.23321399868488346,
+      "grad_norm": 2.0744554488771776,
+      "learning_rate": 1e-05,
+      "loss": 1.2684,
+      "step": 399
+    },
+    {
+      "epoch": 0.23379849492218893,
+      "grad_norm": 1.8213307012212814,
+      "learning_rate": 1e-05,
+      "loss": 1.2857,
+      "step": 400
+    },
+    {
+      "epoch": 0.2343829911594944,
+      "grad_norm": 2.0618681191150707,
+      "learning_rate": 1e-05,
+      "loss": 1.2097,
+      "step": 401
+    },
+    {
+      "epoch": 0.23496748739679987,
+      "grad_norm": 1.895535147278404,
+      "learning_rate": 1e-05,
+      "loss": 1.2587,
+      "step": 402
+    },
+    {
+      "epoch": 0.23555198363410534,
+      "grad_norm": 2.0546144433585463,
+      "learning_rate": 1e-05,
+      "loss": 1.2965,
+      "step": 403
+    },
+    {
+      "epoch": 0.23613647987141081,
+      "grad_norm": 1.9829228718743115,
+      "learning_rate": 1e-05,
+      "loss": 1.2627,
+      "step": 404
+    },
+    {
+      "epoch": 0.2367209761087163,
+      "grad_norm": 1.9455349646719426,
+      "learning_rate": 1e-05,
+      "loss": 1.2513,
+      "step": 405
+    },
+    {
+      "epoch": 0.23730547234602178,
+      "grad_norm": 2.1985507126061528,
+      "learning_rate": 1e-05,
+      "loss": 1.3039,
+      "step": 406
+    },
+    {
+      "epoch": 0.23788996858332725,
+      "grad_norm": 2.0240541408011534,
+      "learning_rate": 1e-05,
+      "loss": 1.2073,
+      "step": 407
+    },
+    {
+      "epoch": 0.23847446482063273,
+      "grad_norm": 2.2850452248213626,
+      "learning_rate": 1e-05,
+      "loss": 1.2217,
+      "step": 408
+    },
+    {
+      "epoch": 0.2390589610579382,
+      "grad_norm": 2.0470248264690243,
+      "learning_rate": 1e-05,
+      "loss": 1.2782,
+      "step": 409
+    },
+    {
+      "epoch": 0.23964345729524367,
+      "grad_norm": 1.913586048505923,
+      "learning_rate": 1e-05,
+      "loss": 1.2309,
+      "step": 410
+    },
+    {
+      "epoch": 0.24022795353254914,
+      "grad_norm": 1.9415326554132903,
+      "learning_rate": 1e-05,
+      "loss": 1.2821,
+      "step": 411
+    },
+    {
+      "epoch": 0.2408124497698546,
+      "grad_norm": 1.9520810625580314,
+      "learning_rate": 1e-05,
+      "loss": 1.2836,
+      "step": 412
+    },
+    {
+      "epoch": 0.24139694600716008,
+      "grad_norm": 1.9084913161655561,
+      "learning_rate": 1e-05,
+      "loss": 1.2269,
+      "step": 413
+    },
+    {
+      "epoch": 0.24198144224446555,
+      "grad_norm": 1.9457021561848948,
+      "learning_rate": 1e-05,
+      "loss": 1.2778,
+      "step": 414
+    },
+    {
+      "epoch": 0.24256593848177102,
+      "grad_norm": 1.9719833949034837,
+      "learning_rate": 1e-05,
+      "loss": 1.3087,
+      "step": 415
+    },
+    {
+      "epoch": 0.2431504347190765,
+      "grad_norm": 1.8865829743307645,
+      "learning_rate": 1e-05,
+      "loss": 1.2393,
+      "step": 416
+    },
+    {
+      "epoch": 0.24373493095638196,
+      "grad_norm": 1.8659445456366994,
+      "learning_rate": 1e-05,
+      "loss": 1.2412,
+      "step": 417
+    },
+    {
+      "epoch": 0.24431942719368743,
+      "grad_norm": 2.0079858773334474,
+      "learning_rate": 1e-05,
+      "loss": 1.2751,
+      "step": 418
+    },
+    {
+      "epoch": 0.2449039234309929,
+      "grad_norm": 1.893226333509476,
+      "learning_rate": 1e-05,
+      "loss": 1.2746,
+      "step": 419
+    },
+    {
+      "epoch": 0.24548841966829837,
+      "grad_norm": 2.032300998657237,
+      "learning_rate": 1e-05,
+      "loss": 1.2492,
+      "step": 420
+    },
+    {
+      "epoch": 0.24607291590560385,
+      "grad_norm": 1.7984024897492445,
+      "learning_rate": 1e-05,
+      "loss": 1.2612,
+      "step": 421
+    },
+    {
+      "epoch": 0.24665741214290934,
+      "grad_norm": 1.8428718363302001,
+      "learning_rate": 1e-05,
+      "loss": 1.2295,
+      "step": 422
+    },
+    {
+      "epoch": 0.24724190838021481,
+      "grad_norm": 1.9551708182124667,
+      "learning_rate": 1e-05,
+      "loss": 1.3042,
+      "step": 423
+    },
+    {
+      "epoch": 0.24782640461752029,
+      "grad_norm": 1.8553715169861047,
+      "learning_rate": 1e-05,
+      "loss": 1.2578,
+      "step": 424
+    },
+    {
+      "epoch": 0.24841090085482576,
+      "grad_norm": 1.8778002899893769,
+      "learning_rate": 1e-05,
+      "loss": 1.2203,
+      "step": 425
+    },
+    {
+      "epoch": 0.24899539709213123,
+      "grad_norm": 2.0122592940870208,
+      "learning_rate": 1e-05,
+      "loss": 1.2632,
+      "step": 426
+    },
+    {
+      "epoch": 0.2495798933294367,
+      "grad_norm": 1.9508619528301605,
+      "learning_rate": 1e-05,
+      "loss": 1.2686,
+      "step": 427
+    },
+    {
+      "epoch": 0.25016438956674214,
+      "grad_norm": 1.9171303139217792,
+      "learning_rate": 1e-05,
+      "loss": 1.2859,
+      "step": 428
+    },
+    {
+      "epoch": 0.25074888580404764,
+      "grad_norm": 2.0796743479655304,
+      "learning_rate": 1e-05,
+      "loss": 1.2877,
+      "step": 429
+    },
+    {
+      "epoch": 0.2513333820413531,
+      "grad_norm": 1.9970633774362665,
+      "learning_rate": 1e-05,
+      "loss": 1.3429,
+      "step": 430
+    },
+    {
+      "epoch": 0.2519178782786586,
+      "grad_norm": 2.0237370451810412,
+      "learning_rate": 1e-05,
+      "loss": 1.2456,
+      "step": 431
+    },
+    {
+      "epoch": 0.2525023745159641,
+      "grad_norm": 1.782231348373727,
+      "learning_rate": 1e-05,
+      "loss": 1.2699,
+      "step": 432
+    },
+    {
+      "epoch": 0.2530868707532695,
+      "grad_norm": 1.9770592048391713,
+      "learning_rate": 1e-05,
+      "loss": 1.2692,
+      "step": 433
+    },
+    {
+      "epoch": 0.253671366990575,
+      "grad_norm": 1.9232213055455618,
+      "learning_rate": 1e-05,
+      "loss": 1.2175,
+      "step": 434
+    },
+    {
+      "epoch": 0.25425586322788046,
+      "grad_norm": 2.015182065860531,
+      "learning_rate": 1e-05,
+      "loss": 1.2575,
+      "step": 435
+    },
+    {
+      "epoch": 0.25484035946518596,
+      "grad_norm": 1.8988878798642703,
+      "learning_rate": 1e-05,
+      "loss": 1.2841,
+      "step": 436
+    },
+    {
+      "epoch": 0.2554248557024914,
+      "grad_norm": 1.83096229156437,
+      "learning_rate": 1e-05,
+      "loss": 1.2714,
+      "step": 437
+    },
+    {
+      "epoch": 0.2560093519397969,
+      "grad_norm": 1.9523729588439458,
+      "learning_rate": 1e-05,
+      "loss": 1.2962,
+      "step": 438
+    },
+    {
+      "epoch": 0.25659384817710235,
+      "grad_norm": 2.0055487821514504,
+      "learning_rate": 1e-05,
+      "loss": 1.268,
+      "step": 439
+    },
+    {
+      "epoch": 0.25717834441440784,
+      "grad_norm": 2.0171517260298675,
+      "learning_rate": 1e-05,
+      "loss": 1.2608,
+      "step": 440
+    },
+    {
+      "epoch": 0.2577628406517133,
+      "grad_norm": 1.9399443424995328,
+      "learning_rate": 1e-05,
+      "loss": 1.3065,
+      "step": 441
+    },
+    {
+      "epoch": 0.2583473368890188,
+      "grad_norm": 1.9094646103544568,
+      "learning_rate": 1e-05,
+      "loss": 1.2535,
+      "step": 442
+    },
+    {
+      "epoch": 0.25893183312632423,
+      "grad_norm": 1.9232648723886796,
+      "learning_rate": 1e-05,
+      "loss": 1.2795,
+      "step": 443
+    },
+    {
+      "epoch": 0.2595163293636297,
+      "grad_norm": 2.0164524014505014,
+      "learning_rate": 1e-05,
+      "loss": 1.292,
+      "step": 444
+    },
+    {
+      "epoch": 0.26010082560093517,
+      "grad_norm": 2.0422212943827445,
+      "learning_rate": 1e-05,
+      "loss": 1.3121,
+      "step": 445
+    },
+    {
+      "epoch": 0.26068532183824067,
+      "grad_norm": 1.9794099286280404,
+      "learning_rate": 1e-05,
+      "loss": 1.2811,
+      "step": 446
+    },
+    {
+      "epoch": 0.2612698180755461,
+      "grad_norm": 1.8898848543207019,
+      "learning_rate": 1e-05,
+      "loss": 1.2431,
+      "step": 447
+    },
+    {
+      "epoch": 0.2618543143128516,
+      "grad_norm": 1.8663354871992899,
+      "learning_rate": 1e-05,
+      "loss": 1.277,
+      "step": 448
+    },
+    {
+      "epoch": 0.2624388105501571,
+      "grad_norm": 1.9777811134806018,
+      "learning_rate": 1e-05,
+      "loss": 1.2657,
+      "step": 449
+    },
+    {
+      "epoch": 0.26302330678746255,
+      "grad_norm": 1.8474920763062657,
+      "learning_rate": 1e-05,
+      "loss": 1.2889,
+      "step": 450
+    },
+    {
+      "epoch": 0.26360780302476805,
+      "grad_norm": 1.8695697397280835,
+      "learning_rate": 1e-05,
+      "loss": 1.2789,
+      "step": 451
+    },
+    {
+      "epoch": 0.2641922992620735,
+      "grad_norm": 2.069943830401144,
+      "learning_rate": 1e-05,
+      "loss": 1.2484,
+      "step": 452
+    },
+    {
+      "epoch": 0.264776795499379,
+      "grad_norm": 1.8928662576583344,
+      "learning_rate": 1e-05,
+      "loss": 1.2797,
+      "step": 453
+    },
+    {
+      "epoch": 0.26536129173668443,
+      "grad_norm": 1.991156930208093,
+      "learning_rate": 1e-05,
+      "loss": 1.2885,
+      "step": 454
+    },
+    {
+      "epoch": 0.26594578797398993,
+      "grad_norm": 1.9521373925649954,
+      "learning_rate": 1e-05,
+      "loss": 1.2612,
+      "step": 455
+    },
+    {
+      "epoch": 0.2665302842112954,
+      "grad_norm": 1.7925080528082347,
+      "learning_rate": 1e-05,
+      "loss": 1.2458,
+      "step": 456
+    },
+    {
+      "epoch": 0.2671147804486009,
+      "grad_norm": 2.223116313592963,
+      "learning_rate": 1e-05,
+      "loss": 1.2149,
+      "step": 457
+    },
+    {
+      "epoch": 0.2676992766859063,
+      "grad_norm": 2.0557576131098156,
+      "learning_rate": 1e-05,
+      "loss": 1.2557,
+      "step": 458
+    },
+    {
+      "epoch": 0.2682837729232118,
+      "grad_norm": 1.8246288695511588,
+      "learning_rate": 1e-05,
+      "loss": 1.278,
+      "step": 459
+    },
+    {
+      "epoch": 0.26886826916051726,
+      "grad_norm": 1.9453123039840678,
+      "learning_rate": 1e-05,
+      "loss": 1.2413,
+      "step": 460
+    },
+    {
+      "epoch": 0.26945276539782276,
+      "grad_norm": 1.854761663761935,
+      "learning_rate": 1e-05,
+      "loss": 1.2417,
+      "step": 461
+    },
+    {
+      "epoch": 0.2700372616351282,
+      "grad_norm": 1.7711393885291036,
+      "learning_rate": 1e-05,
+      "loss": 1.2898,
+      "step": 462
+    },
+    {
+      "epoch": 0.2706217578724337,
+      "grad_norm": 1.8165175469341248,
+      "learning_rate": 1e-05,
+      "loss": 1.2333,
+      "step": 463
+    },
+    {
+      "epoch": 0.27120625410973914,
+      "grad_norm": 1.8595408015562858,
+      "learning_rate": 1e-05,
+      "loss": 1.2522,
+      "step": 464
+    },
+    {
+      "epoch": 0.27179075034704464,
+      "grad_norm": 2.0923665633915616,
+      "learning_rate": 1e-05,
+      "loss": 1.239,
+      "step": 465
+    },
+    {
+      "epoch": 0.27237524658435014,
+      "grad_norm": 1.9433875365616446,
+      "learning_rate": 1e-05,
+      "loss": 1.2332,
+      "step": 466
+    },
+    {
+      "epoch": 0.2729597428216556,
+      "grad_norm": 1.8314448792761906,
+      "learning_rate": 1e-05,
+      "loss": 1.3016,
+      "step": 467
+    },
+    {
+      "epoch": 0.2735442390589611,
+      "grad_norm": 1.9746385841822227,
+      "learning_rate": 1e-05,
+      "loss": 1.2744,
+      "step": 468
+    },
+    {
+      "epoch": 0.2741287352962665,
+      "grad_norm": 1.8726161889357786,
+      "learning_rate": 1e-05,
+      "loss": 1.2786,
+      "step": 469
+    },
+    {
+      "epoch": 0.274713231533572,
+      "grad_norm": 2.0052767371604174,
+      "learning_rate": 1e-05,
+      "loss": 1.247,
+      "step": 470
+    },
+    {
+      "epoch": 0.27529772777087747,
+      "grad_norm": 1.9739498042130639,
+      "learning_rate": 1e-05,
+      "loss": 1.3023,
+      "step": 471
+    },
+    {
+      "epoch": 0.27588222400818296,
+      "grad_norm": 1.8571472047384954,
+      "learning_rate": 1e-05,
+      "loss": 1.2527,
+      "step": 472
+    },
+    {
+      "epoch": 0.2764667202454884,
+      "grad_norm": 1.8360010732109298,
+      "learning_rate": 1e-05,
+      "loss": 1.2692,
+      "step": 473
+    },
+    {
+      "epoch": 0.2770512164827939,
+      "grad_norm": 1.8995971184798859,
+      "learning_rate": 1e-05,
+      "loss": 1.2676,
+      "step": 474
+    },
+    {
+      "epoch": 0.27763571272009935,
+      "grad_norm": 2.077193450911563,
+      "learning_rate": 1e-05,
+      "loss": 1.2609,
+      "step": 475
+    },
+    {
+      "epoch": 0.27822020895740485,
+      "grad_norm": 2.0223384427748736,
+      "learning_rate": 1e-05,
+      "loss": 1.2241,
+      "step": 476
+    },
+    {
+      "epoch": 0.2788047051947103,
+      "grad_norm": 1.8282567961312726,
+      "learning_rate": 1e-05,
+      "loss": 1.236,
+      "step": 477
+    },
+    {
+      "epoch": 0.2793892014320158,
+      "grad_norm": 1.8832335852109048,
+      "learning_rate": 1e-05,
+      "loss": 1.2893,
+      "step": 478
+    },
+    {
+      "epoch": 0.27997369766932123,
+      "grad_norm": 1.9435321659929443,
+      "learning_rate": 1e-05,
+      "loss": 1.2234,
+      "step": 479
+    },
+    {
+      "epoch": 0.28055819390662673,
+      "grad_norm": 2.0630835201075115,
+      "learning_rate": 1e-05,
+      "loss": 1.2853,
+      "step": 480
+    },
+    {
+      "epoch": 0.2811426901439322,
+      "grad_norm": 2.0618458259903836,
+      "learning_rate": 1e-05,
+      "loss": 1.2031,
+      "step": 481
+    },
+    {
+      "epoch": 0.28172718638123767,
+      "grad_norm": 2.0489792224654457,
+      "learning_rate": 1e-05,
+      "loss": 1.2252,
+      "step": 482
+    },
+    {
+      "epoch": 0.28231168261854317,
+      "grad_norm": 2.062675206114832,
+      "learning_rate": 1e-05,
+      "loss": 1.2708,
+      "step": 483
+    },
+    {
+      "epoch": 0.2828961788558486,
+      "grad_norm": 1.889236624676505,
+      "learning_rate": 1e-05,
+      "loss": 1.2739,
+      "step": 484
+    },
+    {
+      "epoch": 0.2834806750931541,
+      "grad_norm": 2.0980334767871844,
+      "learning_rate": 1e-05,
+      "loss": 1.2699,
+      "step": 485
+    },
+    {
+      "epoch": 0.28406517133045955,
+      "grad_norm": 1.9201594133952309,
+      "learning_rate": 1e-05,
+      "loss": 1.2509,
+      "step": 486
+    },
+    {
+      "epoch": 0.28464966756776505,
+      "grad_norm": 1.9885800432967933,
+      "learning_rate": 1e-05,
+      "loss": 1.2853,
+      "step": 487
+    },
+    {
+      "epoch": 0.2852341638050705,
+      "grad_norm": 2.0336361793687567,
+      "learning_rate": 1e-05,
+      "loss": 1.2325,
+      "step": 488
+    },
+    {
+      "epoch": 0.285818660042376,
+      "grad_norm": 2.2467936824654107,
+      "learning_rate": 1e-05,
+      "loss": 1.2232,
+      "step": 489
+    },
+    {
+      "epoch": 0.28640315627968144,
+      "grad_norm": 1.9238760641672366,
+      "learning_rate": 1e-05,
+      "loss": 1.2699,
+      "step": 490
+    },
+    {
+      "epoch": 0.28698765251698694,
+      "grad_norm": 2.1692661606279855,
+      "learning_rate": 1e-05,
+      "loss": 1.3068,
+      "step": 491
+    },
+    {
+      "epoch": 0.2875721487542924,
+      "grad_norm": 1.9036039324998029,
+      "learning_rate": 1e-05,
+      "loss": 1.2437,
+      "step": 492
+    },
+    {
+      "epoch": 0.2881566449915979,
+      "grad_norm": 2.070606101766456,
+      "learning_rate": 1e-05,
+      "loss": 1.2457,
+      "step": 493
+    },
+    {
+      "epoch": 0.2887411412289033,
+      "grad_norm": 2.084503197465521,
+      "learning_rate": 1e-05,
+      "loss": 1.2047,
+      "step": 494
+    },
+    {
+      "epoch": 0.2893256374662088,
+      "grad_norm": 2.014008508275812,
+      "learning_rate": 1e-05,
+      "loss": 1.3102,
+      "step": 495
+    },
+    {
+      "epoch": 0.28991013370351426,
+      "grad_norm": 2.0042395211948736,
+      "learning_rate": 1e-05,
+      "loss": 1.2138,
+      "step": 496
+    },
+    {
+      "epoch": 0.29049462994081976,
+      "grad_norm": 1.972108580277803,
+      "learning_rate": 1e-05,
+      "loss": 1.3321,
+      "step": 497
+    },
+    {
+      "epoch": 0.2910791261781252,
+      "grad_norm": 1.9704453566962155,
+      "learning_rate": 1e-05,
+      "loss": 1.2409,
+      "step": 498
+    },
+    {
+      "epoch": 0.2916636224154307,
+      "grad_norm": 1.926173287542309,
+      "learning_rate": 1e-05,
+      "loss": 1.2673,
+      "step": 499
+    },
+    {
+      "epoch": 0.2922481186527362,
+      "grad_norm": 1.9894370603245293,
+      "learning_rate": 1e-05,
+      "loss": 1.2442,
+      "step": 500
+    },
+    {
+      "epoch": 0.29283261489004164,
+      "grad_norm": 1.9794268348806914,
+      "learning_rate": 1e-05,
+      "loss": 1.233,
+      "step": 501
+    },
+    {
+      "epoch": 0.29341711112734714,
+      "grad_norm": 1.9353010658401677,
+      "learning_rate": 1e-05,
+      "loss": 1.2816,
+      "step": 502
+    },
+    {
+      "epoch": 0.2940016073646526,
+      "grad_norm": 1.914545695851921,
+      "learning_rate": 1e-05,
+      "loss": 1.2318,
+      "step": 503
+    },
+    {
+      "epoch": 0.2945861036019581,
+      "grad_norm": 2.1202774248722354,
+      "learning_rate": 1e-05,
+      "loss": 1.2547,
+      "step": 504
+    },
+    {
+      "epoch": 0.2951705998392635,
+      "grad_norm": 1.9341241307170964,
+      "learning_rate": 1e-05,
+      "loss": 1.2445,
+      "step": 505
+    },
+    {
+      "epoch": 0.295755096076569,
+      "grad_norm": 1.9500676638674974,
+      "learning_rate": 1e-05,
+      "loss": 1.2195,
+      "step": 506
+    },
+    {
+      "epoch": 0.29633959231387447,
+      "grad_norm": 1.7559688212929336,
+      "learning_rate": 1e-05,
+      "loss": 1.297,
+      "step": 507
+    },
+    {
+      "epoch": 0.29692408855117997,
+      "grad_norm": 1.979845407116721,
+      "learning_rate": 1e-05,
+      "loss": 1.2511,
+      "step": 508
+    },
+    {
+      "epoch": 0.2975085847884854,
+      "grad_norm": 2.0245553230685114,
+      "learning_rate": 1e-05,
+      "loss": 1.2148,
+      "step": 509
+    },
+    {
+      "epoch": 0.2980930810257909,
+      "grad_norm": 1.7991873769678588,
+      "learning_rate": 1e-05,
+      "loss": 1.2918,
+      "step": 510
+    },
+    {
+      "epoch": 0.29867757726309635,
+      "grad_norm": 2.105677693376913,
+      "learning_rate": 1e-05,
+      "loss": 1.2092,
+      "step": 511
+    },
+    {
+      "epoch": 0.29926207350040185,
+      "grad_norm": 1.899482771238321,
+      "learning_rate": 1e-05,
+      "loss": 1.1972,
+      "step": 512
+    },
+    {
+      "epoch": 0.2998465697377073,
+      "grad_norm": 1.8269247421773263,
+      "learning_rate": 1e-05,
+      "loss": 1.2455,
+      "step": 513
+    },
+    {
+      "epoch": 0.3004310659750128,
+      "grad_norm": 1.9068261940275615,
+      "learning_rate": 1e-05,
+      "loss": 1.2199,
+      "step": 514
+    },
+    {
+      "epoch": 0.30101556221231823,
+      "grad_norm": 2.0555955880956764,
+      "learning_rate": 1e-05,
+      "loss": 1.257,
+      "step": 515
+    },
+    {
+      "epoch": 0.30160005844962373,
+      "grad_norm": 1.8566481143205649,
+      "learning_rate": 1e-05,
+      "loss": 1.2455,
+      "step": 516
+    },
+    {
+      "epoch": 0.30218455468692923,
+      "grad_norm": 1.952433275651338,
+      "learning_rate": 1e-05,
+      "loss": 1.2574,
+      "step": 517
+    },
+    {
+      "epoch": 0.3027690509242347,
+      "grad_norm": 1.9070322616020552,
+      "learning_rate": 1e-05,
+      "loss": 1.2266,
+      "step": 518
+    },
+    {
+      "epoch": 0.30335354716154017,
+      "grad_norm": 1.966795260885517,
+      "learning_rate": 1e-05,
+      "loss": 1.2439,
+      "step": 519
+    },
+    {
+      "epoch": 0.3039380433988456,
+      "grad_norm": 2.0154060712910393,
+      "learning_rate": 1e-05,
+      "loss": 1.282,
+      "step": 520
+    },
+    {
+      "epoch": 0.3045225396361511,
+      "grad_norm": 1.7952089031079697,
+      "learning_rate": 1e-05,
+      "loss": 1.2603,
+      "step": 521
+    },
+    {
+      "epoch": 0.30510703587345656,
+      "grad_norm": 1.9003208293622356,
+      "learning_rate": 1e-05,
+      "loss": 1.2782,
+      "step": 522
+    },
+    {
+      "epoch": 0.30569153211076205,
+      "grad_norm": 2.0597781377425073,
+      "learning_rate": 1e-05,
+      "loss": 1.1891,
+      "step": 523
+    },
+    {
+      "epoch": 0.3062760283480675,
+      "grad_norm": 1.8736955365158388,
+      "learning_rate": 1e-05,
+      "loss": 1.2521,
+      "step": 524
+    },
+    {
+      "epoch": 0.306860524585373,
+      "grad_norm": 1.9623904564741377,
+      "learning_rate": 1e-05,
+      "loss": 1.203,
+      "step": 525
+    },
+    {
+      "epoch": 0.30744502082267844,
+      "grad_norm": 2.0646564714685485,
+      "learning_rate": 1e-05,
+      "loss": 1.2738,
+      "step": 526
+    },
+    {
+      "epoch": 0.30802951705998394,
+      "grad_norm": 2.0285820609344376,
+      "learning_rate": 1e-05,
+      "loss": 1.2552,
+      "step": 527
+    },
+    {
+      "epoch": 0.3086140132972894,
+      "grad_norm": 1.9832454467324185,
+      "learning_rate": 1e-05,
+      "loss": 1.246,
+      "step": 528
+    },
+    {
+      "epoch": 0.3091985095345949,
+      "grad_norm": 1.8715781900385027,
+      "learning_rate": 1e-05,
+      "loss": 1.2505,
+      "step": 529
+    },
+    {
+      "epoch": 0.3097830057719003,
+      "grad_norm": 1.7248201867950546,
+      "learning_rate": 1e-05,
+      "loss": 1.2813,
+      "step": 530
+    },
+    {
+      "epoch": 0.3103675020092058,
+      "grad_norm": 2.1774503366015705,
+      "learning_rate": 1e-05,
+      "loss": 1.2501,
+      "step": 531
+    },
+    {
+      "epoch": 0.31095199824651126,
+      "grad_norm": 1.6374986677065475,
+      "learning_rate": 1e-05,
+      "loss": 1.2235,
+      "step": 532
+    },
+    {
+      "epoch": 0.31153649448381676,
+      "grad_norm": 1.8078270313855158,
+      "learning_rate": 1e-05,
+      "loss": 1.2911,
+      "step": 533
+    },
+    {
+      "epoch": 0.31212099072112226,
+      "grad_norm": 1.8503067973680813,
+      "learning_rate": 1e-05,
+      "loss": 1.2988,
+      "step": 534
+    },
+    {
+      "epoch": 0.3127054869584277,
+      "grad_norm": 1.8673881847278284,
+      "learning_rate": 1e-05,
+      "loss": 1.2862,
+      "step": 535
+    },
+    {
+      "epoch": 0.3132899831957332,
+      "grad_norm": 2.0635374143443066,
+      "learning_rate": 1e-05,
+      "loss": 1.2236,
+      "step": 536
+    },
+    {
+      "epoch": 0.31387447943303864,
+      "grad_norm": 1.7846188398595475,
+      "learning_rate": 1e-05,
+      "loss": 1.2367,
+      "step": 537
+    },
+    {
+      "epoch": 0.31445897567034414,
+      "grad_norm": 2.052513916038752,
+      "learning_rate": 1e-05,
+      "loss": 1.2868,
+      "step": 538
+    },
+    {
+      "epoch": 0.3150434719076496,
+      "grad_norm": 2.2494257231090535,
+      "learning_rate": 1e-05,
+      "loss": 1.2447,
+      "step": 539
+    },
+    {
+      "epoch": 0.3156279681449551,
+      "grad_norm": 1.7724861670049066,
+      "learning_rate": 1e-05,
+      "loss": 1.2555,
+      "step": 540
+    },
+    {
+      "epoch": 0.3162124643822605,
+      "grad_norm": 1.85452203173062,
+      "learning_rate": 1e-05,
+      "loss": 1.2131,
+      "step": 541
+    },
+    {
+      "epoch": 0.316796960619566,
+      "grad_norm": 1.8554548900562873,
+      "learning_rate": 1e-05,
+      "loss": 1.2782,
+      "step": 542
+    },
+    {
+      "epoch": 0.31738145685687147,
+      "grad_norm": 1.8149611378838015,
+      "learning_rate": 1e-05,
+      "loss": 1.2833,
+      "step": 543
+    },
+    {
+      "epoch": 0.31796595309417697,
+      "grad_norm": 1.9438556911296117,
+      "learning_rate": 1e-05,
+      "loss": 1.2221,
+      "step": 544
+    },
+    {
+      "epoch": 0.3185504493314824,
+      "grad_norm": 1.9703703013468485,
+      "learning_rate": 1e-05,
+      "loss": 1.2215,
+      "step": 545
+    },
+    {
+      "epoch": 0.3191349455687879,
+      "grad_norm": 1.8511550402485115,
+      "learning_rate": 1e-05,
+      "loss": 1.2576,
+      "step": 546
+    },
+    {
+      "epoch": 0.31971944180609335,
+      "grad_norm": 1.8272508352603085,
+      "learning_rate": 1e-05,
+      "loss": 1.2829,
+      "step": 547
+    },
+    {
+      "epoch": 0.32030393804339885,
+      "grad_norm": 1.8647855433726377,
+      "learning_rate": 1e-05,
+      "loss": 1.2803,
+      "step": 548
+    },
+    {
+      "epoch": 0.3208884342807043,
+      "grad_norm": 1.8196437499067115,
+      "learning_rate": 1e-05,
+      "loss": 1.2253,
+      "step": 549
+    },
+    {
+      "epoch": 0.3214729305180098,
+      "grad_norm": 1.67133577148971,
+      "learning_rate": 1e-05,
+      "loss": 1.2582,
+      "step": 550
+    },
+    {
+      "epoch": 0.3220574267553153,
+      "grad_norm": 2.1283672358161256,
+      "learning_rate": 1e-05,
+      "loss": 1.2592,
+      "step": 551
+    },
+    {
+      "epoch": 0.32264192299262073,
+      "grad_norm": 1.9459849674703746,
+      "learning_rate": 1e-05,
+      "loss": 1.2232,
+      "step": 552
+    },
+    {
+      "epoch": 0.32322641922992623,
+      "grad_norm": 2.1879511044402813,
+      "learning_rate": 1e-05,
+      "loss": 1.2614,
+      "step": 553
+    },
+    {
+      "epoch": 0.3238109154672317,
+      "grad_norm": 1.8493420181116036,
+      "learning_rate": 1e-05,
+      "loss": 1.2556,
+      "step": 554
+    },
+    {
+      "epoch": 0.3243954117045372,
+      "grad_norm": 2.1480704594525286,
+      "learning_rate": 1e-05,
+      "loss": 1.267,
+      "step": 555
+    },
+    {
+      "epoch": 0.3249799079418426,
+      "grad_norm": 1.8580725187433735,
+      "learning_rate": 1e-05,
+      "loss": 1.221,
+      "step": 556
+    },
+    {
+      "epoch": 0.3255644041791481,
+      "grad_norm": 2.0477585819993225,
+      "learning_rate": 1e-05,
+      "loss": 1.2615,
+      "step": 557
+    },
+    {
+      "epoch": 0.32614890041645356,
+      "grad_norm": 1.852686392265377,
+      "learning_rate": 1e-05,
+      "loss": 1.2613,
+      "step": 558
+    },
+    {
+      "epoch": 0.32673339665375906,
+      "grad_norm": 2.036634465075735,
+      "learning_rate": 1e-05,
+      "loss": 1.222,
+      "step": 559
+    },
+    {
+      "epoch": 0.3273178928910645,
+      "grad_norm": 1.9616405764710099,
+      "learning_rate": 1e-05,
+      "loss": 1.2696,
+      "step": 560
+    },
+    {
+      "epoch": 0.32790238912837,
+      "grad_norm": 1.9208164756188018,
+      "learning_rate": 1e-05,
+      "loss": 1.2444,
+      "step": 561
+    },
+    {
+      "epoch": 0.32848688536567544,
+      "grad_norm": 2.0429631090545595,
+      "learning_rate": 1e-05,
+      "loss": 1.2119,
+      "step": 562
+    },
+    {
+      "epoch": 0.32907138160298094,
+      "grad_norm": 1.9774936950784288,
+      "learning_rate": 1e-05,
+      "loss": 1.2673,
+      "step": 563
+    },
+    {
+      "epoch": 0.3296558778402864,
+      "grad_norm": 1.9637164430108291,
+      "learning_rate": 1e-05,
+      "loss": 1.2103,
+      "step": 564
+    },
+    {
+      "epoch": 0.3302403740775919,
+      "grad_norm": 1.90292546463272,
+      "learning_rate": 1e-05,
+      "loss": 1.2455,
+      "step": 565
+    },
+    {
+      "epoch": 0.3308248703148973,
+      "grad_norm": 2.0448861243820593,
+      "learning_rate": 1e-05,
+      "loss": 1.2528,
+      "step": 566
+    },
+    {
+      "epoch": 0.3314093665522028,
+      "grad_norm": 2.084980771679745,
+      "learning_rate": 1e-05,
+      "loss": 1.2589,
+      "step": 567
+    },
+    {
+      "epoch": 0.33199386278950827,
+      "grad_norm": 1.8402586821932294,
+      "learning_rate": 1e-05,
+      "loss": 1.2308,
+      "step": 568
+    },
+    {
+      "epoch": 0.33257835902681376,
+      "grad_norm": 1.8115131155487174,
+      "learning_rate": 1e-05,
+      "loss": 1.242,
+      "step": 569
+    },
+    {
+      "epoch": 0.33316285526411926,
+      "grad_norm": 1.9252622302745714,
+      "learning_rate": 1e-05,
+      "loss": 1.2363,
+      "step": 570
+    },
+    {
+      "epoch": 0.3337473515014247,
+      "grad_norm": 1.9341689022335478,
+      "learning_rate": 1e-05,
+      "loss": 1.2711,
+      "step": 571
+    },
+    {
+      "epoch": 0.3343318477387302,
+      "grad_norm": 1.9277839274697275,
+      "learning_rate": 1e-05,
+      "loss": 1.2357,
+      "step": 572
+    },
+    {
+      "epoch": 0.33491634397603565,
+      "grad_norm": 2.1365428366788684,
+      "learning_rate": 1e-05,
+      "loss": 1.2788,
+      "step": 573
+    },
+    {
+      "epoch": 0.33550084021334114,
+      "grad_norm": 1.9346383807683418,
+      "learning_rate": 1e-05,
+      "loss": 1.2122,
+      "step": 574
+    },
+    {
+      "epoch": 0.3360853364506466,
+      "grad_norm": 1.7859049797549464,
+      "learning_rate": 1e-05,
+      "loss": 1.1696,
+      "step": 575
+    },
+    {
+      "epoch": 0.3366698326879521,
+      "grad_norm": 1.9628058683077474,
+      "learning_rate": 1e-05,
+      "loss": 1.2241,
+      "step": 576
+    },
+    {
+      "epoch": 0.33725432892525753,
+      "grad_norm": 1.7304013879722078,
+      "learning_rate": 1e-05,
+      "loss": 1.1807,
+      "step": 577
+    },
+    {
+      "epoch": 0.33783882516256303,
+      "grad_norm": 1.8593766026185374,
+      "learning_rate": 1e-05,
+      "loss": 1.2591,
+      "step": 578
+    },
+    {
+      "epoch": 0.33842332139986847,
+      "grad_norm": 1.848039026704815,
+      "learning_rate": 1e-05,
+      "loss": 1.27,
+      "step": 579
+    },
+    {
+      "epoch": 0.33900781763717397,
+      "grad_norm": 1.9567046647088875,
+      "learning_rate": 1e-05,
+      "loss": 1.2624,
+      "step": 580
+    },
+    {
+      "epoch": 0.3395923138744794,
+      "grad_norm": 1.8731645048142613,
+      "learning_rate": 1e-05,
+      "loss": 1.2572,
+      "step": 581
+    },
+    {
+      "epoch": 0.3401768101117849,
+      "grad_norm": 1.8658894879383807,
+      "learning_rate": 1e-05,
+      "loss": 1.2092,
+      "step": 582
+    },
+    {
+      "epoch": 0.34076130634909035,
+      "grad_norm": 1.904220741019776,
+      "learning_rate": 1e-05,
+      "loss": 1.2926,
+      "step": 583
+    },
+    {
+      "epoch": 0.34134580258639585,
+      "grad_norm": 1.9494681134023866,
+      "learning_rate": 1e-05,
+      "loss": 1.2507,
+      "step": 584
+    },
+    {
+      "epoch": 0.3419302988237013,
+      "grad_norm": 1.999051358227611,
+      "learning_rate": 1e-05,
+      "loss": 1.2348,
+      "step": 585
+    },
+    {
+      "epoch": 0.3425147950610068,
+      "grad_norm": 1.8712211397530882,
+      "learning_rate": 1e-05,
+      "loss": 1.2005,
+      "step": 586
+    },
+    {
+      "epoch": 0.3430992912983123,
+      "grad_norm": 1.83955577946858,
+      "learning_rate": 1e-05,
+      "loss": 1.2272,
+      "step": 587
+    },
+    {
+      "epoch": 0.34368378753561774,
+      "grad_norm": 1.8632067546990136,
+      "learning_rate": 1e-05,
+      "loss": 1.1916,
+      "step": 588
+    },
+    {
+      "epoch": 0.34426828377292323,
+      "grad_norm": 1.8094789272128537,
+      "learning_rate": 1e-05,
+      "loss": 1.2355,
+      "step": 589
+    },
+    {
+      "epoch": 0.3448527800102287,
+      "grad_norm": 2.0484844483634825,
+      "learning_rate": 1e-05,
+      "loss": 1.2329,
+      "step": 590
+    },
+    {
+      "epoch": 0.3454372762475342,
+      "grad_norm": 1.8887328816602642,
+      "learning_rate": 1e-05,
+      "loss": 1.2076,
+      "step": 591
+    },
+    {
+      "epoch": 0.3460217724848396,
+      "grad_norm": 1.8463674764905278,
+      "learning_rate": 1e-05,
+      "loss": 1.26,
+      "step": 592
+    },
+    {
+      "epoch": 0.3466062687221451,
+      "grad_norm": 1.9572576857124795,
+      "learning_rate": 1e-05,
+      "loss": 1.2321,
+      "step": 593
+    },
+    {
+      "epoch": 0.34719076495945056,
+      "grad_norm": 2.33474599900278,
+      "learning_rate": 1e-05,
+      "loss": 1.2827,
+      "step": 594
+    },
+    {
+      "epoch": 0.34777526119675606,
+      "grad_norm": 2.017451280397668,
+      "learning_rate": 1e-05,
+      "loss": 1.2727,
+      "step": 595
+    },
+    {
+      "epoch": 0.3483597574340615,
+      "grad_norm": 2.3156806026305463,
+      "learning_rate": 1e-05,
+      "loss": 1.289,
+      "step": 596
+    },
+    {
+      "epoch": 0.348944253671367,
+      "grad_norm": 1.7860172449167409,
+      "learning_rate": 1e-05,
+      "loss": 1.1828,
+      "step": 597
+    },
+    {
+      "epoch": 0.34952874990867244,
+      "grad_norm": 1.9989557211266393,
+      "learning_rate": 1e-05,
+      "loss": 1.2986,
+      "step": 598
+    },
+    {
+      "epoch": 0.35011324614597794,
+      "grad_norm": 1.9560719511658293,
+      "learning_rate": 1e-05,
+      "loss": 1.2188,
+      "step": 599
+    },
+    {
+      "epoch": 0.3506977423832834,
+      "grad_norm": 2.1483096009605847,
+      "learning_rate": 1e-05,
+      "loss": 1.252,
+      "step": 600
+    },
+    {
+      "epoch": 0.3512822386205889,
+      "grad_norm": 1.8322821525960806,
+      "learning_rate": 1e-05,
+      "loss": 1.2487,
+      "step": 601
+    },
+    {
+      "epoch": 0.3518667348578943,
+      "grad_norm": 1.8750170985170502,
+      "learning_rate": 1e-05,
+      "loss": 1.2433,
+      "step": 602
+    },
+    {
+      "epoch": 0.3524512310951998,
+      "grad_norm": 2.0599922608331713,
+      "learning_rate": 1e-05,
+      "loss": 1.2501,
+      "step": 603
+    },
+    {
+      "epoch": 0.3530357273325053,
+      "grad_norm": 1.8568137370280302,
+      "learning_rate": 1e-05,
+      "loss": 1.2859,
+      "step": 604
+    },
+    {
+      "epoch": 0.35362022356981077,
+      "grad_norm": 1.8445086983117804,
+      "learning_rate": 1e-05,
+      "loss": 1.2592,
+      "step": 605
+    },
+    {
+      "epoch": 0.35420471980711626,
+      "grad_norm": 1.9998948752127053,
+      "learning_rate": 1e-05,
+      "loss": 1.2081,
+      "step": 606
+    },
+    {
+      "epoch": 0.3547892160444217,
+      "grad_norm": 1.8839796842198089,
+      "learning_rate": 1e-05,
+      "loss": 1.2129,
+      "step": 607
+    },
+    {
+      "epoch": 0.3553737122817272,
+      "grad_norm": 2.069231651147054,
+      "learning_rate": 1e-05,
+      "loss": 1.202,
+      "step": 608
+    },
+    {
+      "epoch": 0.35595820851903265,
+      "grad_norm": 1.9309936023941956,
+      "learning_rate": 1e-05,
+      "loss": 1.2529,
+      "step": 609
+    },
+    {
+      "epoch": 0.35654270475633815,
+      "grad_norm": 1.970897764804915,
+      "learning_rate": 1e-05,
+      "loss": 1.2353,
+      "step": 610
+    },
+    {
+      "epoch": 0.3571272009936436,
+      "grad_norm": 2.064391214192704,
+      "learning_rate": 1e-05,
+      "loss": 1.2695,
+      "step": 611
+    },
+    {
+      "epoch": 0.3577116972309491,
+      "grad_norm": 1.897097220716556,
+      "learning_rate": 1e-05,
+      "loss": 1.244,
+      "step": 612
+    },
+    {
+      "epoch": 0.35829619346825453,
+      "grad_norm": 2.1506093740921224,
+      "learning_rate": 1e-05,
+      "loss": 1.2813,
+      "step": 613
+    },
+    {
+      "epoch": 0.35888068970556003,
+      "grad_norm": 1.7933959805512387,
+      "learning_rate": 1e-05,
+      "loss": 1.2495,
+      "step": 614
+    },
+    {
+      "epoch": 0.3594651859428655,
+      "grad_norm": 1.8506002206188736,
+      "learning_rate": 1e-05,
+      "loss": 1.2357,
+      "step": 615
+    },
+    {
+      "epoch": 0.36004968218017097,
+      "grad_norm": 1.8338184361768353,
+      "learning_rate": 1e-05,
+      "loss": 1.2585,
+      "step": 616
+    },
+    {
+      "epoch": 0.3606341784174764,
+      "grad_norm": 2.1256225198615546,
+      "learning_rate": 1e-05,
+      "loss": 1.2652,
+      "step": 617
+    },
+    {
+      "epoch": 0.3612186746547819,
+      "grad_norm": 2.1569808824514864,
+      "learning_rate": 1e-05,
+      "loss": 1.2535,
+      "step": 618
+    },
+    {
+      "epoch": 0.36180317089208736,
+      "grad_norm": 2.052715309819321,
+      "learning_rate": 1e-05,
+      "loss": 1.2193,
+      "step": 619
+    },
+    {
+      "epoch": 0.36238766712939285,
+      "grad_norm": 2.0674149440240255,
+      "learning_rate": 1e-05,
+      "loss": 1.2417,
+      "step": 620
+    },
+    {
+      "epoch": 0.36297216336669835,
+      "grad_norm": 2.033537707483929,
+      "learning_rate": 1e-05,
+      "loss": 1.2631,
+      "step": 621
+    },
+    {
+      "epoch": 0.3635566596040038,
+      "grad_norm": 1.8277766264437378,
+      "learning_rate": 1e-05,
+      "loss": 1.269,
+      "step": 622
+    },
+    {
+      "epoch": 0.3641411558413093,
+      "grad_norm": 1.7504356853457703,
+      "learning_rate": 1e-05,
+      "loss": 1.3176,
+      "step": 623
+    },
+    {
+      "epoch": 0.36472565207861474,
+      "grad_norm": 1.8441024614111936,
+      "learning_rate": 1e-05,
+      "loss": 1.1915,
+      "step": 624
+    },
+    {
+      "epoch": 0.36531014831592024,
+      "grad_norm": 2.0175578935416225,
+      "learning_rate": 1e-05,
+      "loss": 1.2858,
+      "step": 625
+    },
+    {
+      "epoch": 0.3658946445532257,
+      "grad_norm": 1.8522996013379147,
+      "learning_rate": 1e-05,
+      "loss": 1.3031,
+      "step": 626
+    },
+    {
+      "epoch": 0.3664791407905312,
+      "grad_norm": 1.9777153348029328,
+      "learning_rate": 1e-05,
+      "loss": 1.2818,
+      "step": 627
+    },
+    {
+      "epoch": 0.3670636370278366,
+      "grad_norm": 1.6690279490476583,
+      "learning_rate": 1e-05,
+      "loss": 1.2875,
+      "step": 628
+    },
+    {
+      "epoch": 0.3676481332651421,
+      "grad_norm": 1.9342589718585115,
+      "learning_rate": 1e-05,
+      "loss": 1.2585,
+      "step": 629
+    },
+    {
+      "epoch": 0.36823262950244756,
+      "grad_norm": 1.935619320764398,
+      "learning_rate": 1e-05,
+      "loss": 1.2313,
+      "step": 630
+    },
+    {
+      "epoch": 0.36881712573975306,
+      "grad_norm": 2.0435952497926393,
+      "learning_rate": 1e-05,
+      "loss": 1.2145,
+      "step": 631
+    },
+    {
+      "epoch": 0.3694016219770585,
+      "grad_norm": 1.8490764722522108,
+      "learning_rate": 1e-05,
+      "loss": 1.2171,
+      "step": 632
+    },
+    {
+      "epoch": 0.369986118214364,
+      "grad_norm": 1.7368008255108502,
+      "learning_rate": 1e-05,
+      "loss": 1.1891,
+      "step": 633
+    },
+    {
+      "epoch": 0.37057061445166944,
+      "grad_norm": 2.075431113891477,
+      "learning_rate": 1e-05,
+      "loss": 1.2507,
+      "step": 634
+    },
+    {
+      "epoch": 0.37115511068897494,
+      "grad_norm": 1.9119771843791091,
+      "learning_rate": 1e-05,
+      "loss": 1.3025,
+      "step": 635
+    },
+    {
+      "epoch": 0.3717396069262804,
+      "grad_norm": 1.9263176207186385,
+      "learning_rate": 1e-05,
+      "loss": 1.2352,
+      "step": 636
+    },
+    {
+      "epoch": 0.3723241031635859,
+      "grad_norm": 1.8584207382145501,
+      "learning_rate": 1e-05,
+      "loss": 1.2333,
+      "step": 637
+    },
+    {
+      "epoch": 0.3729085994008914,
+      "grad_norm": 1.9005002776652664,
+      "learning_rate": 1e-05,
+      "loss": 1.3292,
+      "step": 638
+    },
+    {
+      "epoch": 0.3734930956381968,
+      "grad_norm": 2.0012911330332757,
+      "learning_rate": 1e-05,
+      "loss": 1.2474,
+      "step": 639
+    },
+    {
+      "epoch": 0.3740775918755023,
+      "grad_norm": 1.913295545682595,
+      "learning_rate": 1e-05,
+      "loss": 1.2353,
+      "step": 640
+    },
+    {
+      "epoch": 0.37466208811280777,
+      "grad_norm": 1.9449639004887131,
+      "learning_rate": 1e-05,
+      "loss": 1.264,
+      "step": 641
+    },
+    {
+      "epoch": 0.37524658435011327,
+      "grad_norm": 2.276761928306722,
+      "learning_rate": 1e-05,
+      "loss": 1.2098,
+      "step": 642
+    },
+    {
+      "epoch": 0.3758310805874187,
+      "grad_norm": 1.8117002598825125,
+      "learning_rate": 1e-05,
+      "loss": 1.2016,
+      "step": 643
+    },
+    {
+      "epoch": 0.3764155768247242,
+      "grad_norm": 1.8298569776519442,
+      "learning_rate": 1e-05,
+      "loss": 1.2866,
+      "step": 644
+    },
+    {
+      "epoch": 0.37700007306202965,
+      "grad_norm": 1.9364091577577964,
+      "learning_rate": 1e-05,
+      "loss": 1.2312,
+      "step": 645
+    },
+    {
+      "epoch": 0.37758456929933515,
+      "grad_norm": 1.8106934301264197,
+      "learning_rate": 1e-05,
+      "loss": 1.2307,
+      "step": 646
+    },
+    {
+      "epoch": 0.3781690655366406,
+      "grad_norm": 1.924913664685263,
+      "learning_rate": 1e-05,
+      "loss": 1.2393,
+      "step": 647
+    },
+    {
+      "epoch": 0.3787535617739461,
+      "grad_norm": 2.0858848290298146,
+      "learning_rate": 1e-05,
+      "loss": 1.2561,
+      "step": 648
+    },
+    {
+      "epoch": 0.37933805801125153,
+      "grad_norm": 1.8308681187458808,
+      "learning_rate": 1e-05,
+      "loss": 1.2348,
+      "step": 649
+    },
+    {
+      "epoch": 0.37992255424855703,
+      "grad_norm": 1.8366109468522391,
+      "learning_rate": 1e-05,
+      "loss": 1.2437,
+      "step": 650
+    },
+    {
+      "epoch": 0.3805070504858625,
+      "grad_norm": 1.9462140897126792,
+      "learning_rate": 1e-05,
+      "loss": 1.2097,
+      "step": 651
+    },
+    {
+      "epoch": 0.381091546723168,
+      "grad_norm": 1.8578736466938333,
+      "learning_rate": 1e-05,
+      "loss": 1.2301,
+      "step": 652
+    },
+    {
+      "epoch": 0.3816760429604734,
+      "grad_norm": 1.931867483381843,
+      "learning_rate": 1e-05,
+      "loss": 1.283,
+      "step": 653
+    },
+    {
+      "epoch": 0.3822605391977789,
+      "grad_norm": 1.777931282878353,
+      "learning_rate": 1e-05,
+      "loss": 1.2038,
+      "step": 654
+    },
+    {
+      "epoch": 0.3828450354350844,
+      "grad_norm": 1.8794604815422231,
+      "learning_rate": 1e-05,
+      "loss": 1.246,
+      "step": 655
+    },
+    {
+      "epoch": 0.38342953167238986,
+      "grad_norm": 2.0749887865253256,
+      "learning_rate": 1e-05,
+      "loss": 1.2365,
+      "step": 656
+    },
+    {
+      "epoch": 0.38401402790969535,
+      "grad_norm": 1.8895443228989306,
+      "learning_rate": 1e-05,
+      "loss": 1.2403,
+      "step": 657
+    },
+    {
+      "epoch": 0.3845985241470008,
+      "grad_norm": 1.8160183405775607,
+      "learning_rate": 1e-05,
+      "loss": 1.2527,
+      "step": 658
+    },
+    {
+      "epoch": 0.3851830203843063,
+      "grad_norm": 2.052491035576922,
+      "learning_rate": 1e-05,
+      "loss": 1.2794,
+      "step": 659
+    },
+    {
+      "epoch": 0.38576751662161174,
+      "grad_norm": 1.8569726609380135,
+      "learning_rate": 1e-05,
+      "loss": 1.2902,
+      "step": 660
+    },
+    {
+      "epoch": 0.38635201285891724,
+      "grad_norm": 1.840454458681519,
+      "learning_rate": 1e-05,
+      "loss": 1.2289,
+      "step": 661
+    },
+    {
+      "epoch": 0.3869365090962227,
+      "grad_norm": 1.901069771909571,
+      "learning_rate": 1e-05,
+      "loss": 1.2716,
+      "step": 662
+    },
+    {
+      "epoch": 0.3875210053335282,
+      "grad_norm": 1.7738121845246662,
+      "learning_rate": 1e-05,
+      "loss": 1.233,
+      "step": 663
+    },
+    {
+      "epoch": 0.3881055015708336,
+      "grad_norm": 1.847094006908835,
+      "learning_rate": 1e-05,
+      "loss": 1.2504,
+      "step": 664
+    },
+    {
+      "epoch": 0.3886899978081391,
+      "grad_norm": 2.222123223373021,
+      "learning_rate": 1e-05,
+      "loss": 1.2967,
+      "step": 665
+    },
+    {
+      "epoch": 0.38927449404544456,
+      "grad_norm": 1.7474327816377886,
+      "learning_rate": 1e-05,
+      "loss": 1.2153,
+      "step": 666
+    },
+    {
+      "epoch": 0.38985899028275006,
+      "grad_norm": 1.921856890606811,
+      "learning_rate": 1e-05,
+      "loss": 1.1871,
+      "step": 667
+    },
+    {
+      "epoch": 0.3904434865200555,
+      "grad_norm": 1.781174353704435,
+      "learning_rate": 1e-05,
+      "loss": 1.2525,
+      "step": 668
+    },
+    {
+      "epoch": 0.391027982757361,
+      "grad_norm": 2.0156676455832137,
+      "learning_rate": 1e-05,
+      "loss": 1.2447,
+      "step": 669
+    },
+    {
+      "epoch": 0.39161247899466645,
+      "grad_norm": 1.8489239462181082,
+      "learning_rate": 1e-05,
+      "loss": 1.2035,
+      "step": 670
+    },
+    {
+      "epoch": 0.39219697523197194,
+      "grad_norm": 2.14286997703903,
+      "learning_rate": 1e-05,
+      "loss": 1.2207,
+      "step": 671
+    },
+    {
+      "epoch": 0.39278147146927744,
+      "grad_norm": 2.034883722293709,
+      "learning_rate": 1e-05,
+      "loss": 1.2503,
+      "step": 672
+    },
+    {
+      "epoch": 0.3933659677065829,
+      "grad_norm": 1.906477785311283,
+      "learning_rate": 1e-05,
+      "loss": 1.2415,
+      "step": 673
+    },
+    {
+      "epoch": 0.3939504639438884,
+      "grad_norm": 1.8540813971923493,
+      "learning_rate": 1e-05,
+      "loss": 1.2923,
+      "step": 674
+    },
+    {
+      "epoch": 0.39453496018119383,
+      "grad_norm": 1.8490943580735069,
+      "learning_rate": 1e-05,
+      "loss": 1.2635,
+      "step": 675
+    },
+    {
+      "epoch": 0.3951194564184993,
+      "grad_norm": 1.7066437465862272,
+      "learning_rate": 1e-05,
+      "loss": 1.2242,
+      "step": 676
+    },
+    {
+      "epoch": 0.39570395265580477,
+      "grad_norm": 2.1317908139823225,
+      "learning_rate": 1e-05,
+      "loss": 1.2215,
+      "step": 677
+    },
+    {
+      "epoch": 0.39628844889311027,
+      "grad_norm": 1.9394067634399133,
+      "learning_rate": 1e-05,
+      "loss": 1.2882,
+      "step": 678
+    },
+    {
+      "epoch": 0.3968729451304157,
+      "grad_norm": 1.8154764577269566,
+      "learning_rate": 1e-05,
+      "loss": 1.2085,
+      "step": 679
+    },
+    {
+      "epoch": 0.3974574413677212,
+      "grad_norm": 1.965689369102174,
+      "learning_rate": 1e-05,
+      "loss": 1.2376,
+      "step": 680
+    },
+    {
+      "epoch": 0.39804193760502665,
+      "grad_norm": 1.8746805207228199,
+      "learning_rate": 1e-05,
+      "loss": 1.2118,
+      "step": 681
+    },
+    {
+      "epoch": 0.39862643384233215,
+      "grad_norm": 1.9209373635110283,
+      "learning_rate": 1e-05,
+      "loss": 1.2189,
+      "step": 682
+    },
+    {
+      "epoch": 0.3992109300796376,
+      "grad_norm": 1.8652225797607809,
+      "learning_rate": 1e-05,
+      "loss": 1.1738,
+      "step": 683
+    },
+    {
+      "epoch": 0.3997954263169431,
+      "grad_norm": 1.8231619406006876,
+      "learning_rate": 1e-05,
+      "loss": 1.2266,
+      "step": 684
+    },
+    {
+      "epoch": 0.40037992255424854,
+      "grad_norm": 2.1299043867461074,
+      "learning_rate": 1e-05,
+      "loss": 1.1873,
+      "step": 685
+    },
+    {
+      "epoch": 0.40096441879155403,
+      "grad_norm": 1.7968942125410932,
+      "learning_rate": 1e-05,
+      "loss": 1.1962,
+      "step": 686
+    },
+    {
+      "epoch": 0.4015489150288595,
+      "grad_norm": 1.9287311560205092,
+      "learning_rate": 1e-05,
+      "loss": 1.216,
+      "step": 687
+    },
+    {
+      "epoch": 0.402133411266165,
+      "grad_norm": 1.928305113543552,
+      "learning_rate": 1e-05,
+      "loss": 1.1889,
+      "step": 688
+    },
+    {
+      "epoch": 0.4027179075034705,
+      "grad_norm": 1.7560586077279643,
+      "learning_rate": 1e-05,
+      "loss": 1.2301,
+      "step": 689
+    },
+    {
+      "epoch": 0.4033024037407759,
+      "grad_norm": 2.0032247368125415,
+      "learning_rate": 1e-05,
+      "loss": 1.2478,
+      "step": 690
+    },
+    {
+      "epoch": 0.4038868999780814,
+      "grad_norm": 1.9013228158264661,
+      "learning_rate": 1e-05,
+      "loss": 1.2152,
+      "step": 691
+    },
+    {
+      "epoch": 0.40447139621538686,
+      "grad_norm": 1.941603427569868,
+      "learning_rate": 1e-05,
+      "loss": 1.2536,
+      "step": 692
+    },
+    {
+      "epoch": 0.40505589245269236,
+      "grad_norm": 1.9329019080134706,
+      "learning_rate": 1e-05,
+      "loss": 1.186,
+      "step": 693
+    },
+    {
+      "epoch": 0.4056403886899978,
+      "grad_norm": 2.0039273870696666,
+      "learning_rate": 1e-05,
+      "loss": 1.2344,
+      "step": 694
+    },
+    {
+      "epoch": 0.4062248849273033,
+      "grad_norm": 1.9766477880282185,
+      "learning_rate": 1e-05,
+      "loss": 1.2055,
+      "step": 695
+    },
+    {
+      "epoch": 0.40680938116460874,
+      "grad_norm": 1.743434077256146,
+      "learning_rate": 1e-05,
+      "loss": 1.1977,
+      "step": 696
+    },
+    {
+      "epoch": 0.40739387740191424,
+      "grad_norm": 1.9213664779143957,
+      "learning_rate": 1e-05,
+      "loss": 1.2387,
+      "step": 697
+    },
+    {
+      "epoch": 0.4079783736392197,
+      "grad_norm": 1.8090980903160492,
+      "learning_rate": 1e-05,
+      "loss": 1.2499,
+      "step": 698
+    },
+    {
+      "epoch": 0.4085628698765252,
+      "grad_norm": 1.8849908628270615,
+      "learning_rate": 1e-05,
+      "loss": 1.247,
+      "step": 699
+    },
+    {
+      "epoch": 0.4091473661138306,
+      "grad_norm": 1.8697568238282716,
+      "learning_rate": 1e-05,
+      "loss": 1.2458,
+      "step": 700
+    },
+    {
+      "epoch": 0.4097318623511361,
+      "grad_norm": 1.9443227186028607,
+      "learning_rate": 1e-05,
+      "loss": 1.3071,
+      "step": 701
+    },
+    {
+      "epoch": 0.41031635858844157,
+      "grad_norm": 2.008068180639773,
+      "learning_rate": 1e-05,
+      "loss": 1.2594,
+      "step": 702
+    },
+    {
+      "epoch": 0.41090085482574706,
+      "grad_norm": 1.9980287247329434,
+      "learning_rate": 1e-05,
+      "loss": 1.3015,
+      "step": 703
+    },
+    {
+      "epoch": 0.4114853510630525,
+      "grad_norm": 2.080489187818687,
+      "learning_rate": 1e-05,
+      "loss": 1.2064,
+      "step": 704
+    },
+    {
+      "epoch": 0.412069847300358,
+      "grad_norm": 1.7138496866537474,
+      "learning_rate": 1e-05,
+      "loss": 1.2913,
+      "step": 705
+    },
+    {
+      "epoch": 0.4126543435376635,
+      "grad_norm": 1.9693994971872804,
+      "learning_rate": 1e-05,
+      "loss": 1.2163,
+      "step": 706
+    },
+    {
+      "epoch": 0.41323883977496895,
+      "grad_norm": 1.9685248583455057,
+      "learning_rate": 1e-05,
+      "loss": 1.1954,
+      "step": 707
+    },
+    {
+      "epoch": 0.41382333601227445,
+      "grad_norm": 1.7064073827234496,
+      "learning_rate": 1e-05,
+      "loss": 1.3042,
+      "step": 708
+    },
+    {
+      "epoch": 0.4144078322495799,
+      "grad_norm": 1.8706243324213678,
+      "learning_rate": 1e-05,
+      "loss": 1.2555,
+      "step": 709
+    },
+    {
+      "epoch": 0.4149923284868854,
+      "grad_norm": 1.8943424028650786,
+      "learning_rate": 1e-05,
+      "loss": 1.2178,
+      "step": 710
+    },
+    {
+      "epoch": 0.41557682472419083,
+      "grad_norm": 1.8251827881996059,
+      "learning_rate": 1e-05,
+      "loss": 1.239,
+      "step": 711
+    },
+    {
+      "epoch": 0.41616132096149633,
+      "grad_norm": 1.7444670865548013,
+      "learning_rate": 1e-05,
+      "loss": 1.2236,
+      "step": 712
+    },
+    {
+      "epoch": 0.41674581719880177,
+      "grad_norm": 1.958686899203161,
+      "learning_rate": 1e-05,
+      "loss": 1.2045,
+      "step": 713
+    },
+    {
+      "epoch": 0.41733031343610727,
+      "grad_norm": 1.859148273145525,
+      "learning_rate": 1e-05,
+      "loss": 1.221,
+      "step": 714
+    },
+    {
+      "epoch": 0.4179148096734127,
+      "grad_norm": 1.944161000215102,
+      "learning_rate": 1e-05,
+      "loss": 1.2539,
+      "step": 715
+    },
+    {
+      "epoch": 0.4184993059107182,
+      "grad_norm": 1.9812461308754372,
+      "learning_rate": 1e-05,
+      "loss": 1.245,
+      "step": 716
+    },
+    {
+      "epoch": 0.41908380214802365,
+      "grad_norm": 1.9121073186935142,
+      "learning_rate": 1e-05,
+      "loss": 1.2076,
+      "step": 717
+    },
+    {
+      "epoch": 0.41966829838532915,
+      "grad_norm": 1.6956781781329928,
+      "learning_rate": 1e-05,
+      "loss": 1.2473,
+      "step": 718
+    },
+    {
+      "epoch": 0.4202527946226346,
+      "grad_norm": 1.6005519029429327,
+      "learning_rate": 1e-05,
+      "loss": 1.2178,
+      "step": 719
+    },
+    {
+      "epoch": 0.4208372908599401,
+      "grad_norm": 1.8248676851723231,
+      "learning_rate": 1e-05,
+      "loss": 1.1737,
+      "step": 720
+    },
+    {
+      "epoch": 0.42142178709724554,
+      "grad_norm": 1.9595090307520382,
+      "learning_rate": 1e-05,
+      "loss": 1.2207,
+      "step": 721
+    },
+    {
+      "epoch": 0.42200628333455104,
+      "grad_norm": 1.873630180921612,
+      "learning_rate": 1e-05,
+      "loss": 1.2513,
+      "step": 722
+    },
+    {
+      "epoch": 0.4225907795718565,
+      "grad_norm": 2.008890531902992,
+      "learning_rate": 1e-05,
+      "loss": 1.2673,
+      "step": 723
+    },
+    {
+      "epoch": 0.423175275809162,
+      "grad_norm": 1.8096057907409144,
+      "learning_rate": 1e-05,
+      "loss": 1.2268,
+      "step": 724
+    },
+    {
+      "epoch": 0.4237597720464675,
+      "grad_norm": 1.6941903099749094,
+      "learning_rate": 1e-05,
+      "loss": 1.2852,
+      "step": 725
+    },
+    {
+      "epoch": 0.4243442682837729,
+      "grad_norm": 1.9380190226039893,
+      "learning_rate": 1e-05,
+      "loss": 1.252,
+      "step": 726
+    },
+    {
+      "epoch": 0.4249287645210784,
+      "grad_norm": 2.0245940951133554,
+      "learning_rate": 1e-05,
+      "loss": 1.2277,
+      "step": 727
+    },
+    {
+      "epoch": 0.42551326075838386,
+      "grad_norm": 1.9277066114540902,
+      "learning_rate": 1e-05,
+      "loss": 1.2174,
+      "step": 728
+    },
+    {
+      "epoch": 0.42609775699568936,
+      "grad_norm": 1.9138551026174433,
+      "learning_rate": 1e-05,
+      "loss": 1.2359,
+      "step": 729
+    },
+    {
+      "epoch": 0.4266822532329948,
+      "grad_norm": 1.8545619633376635,
+      "learning_rate": 1e-05,
+      "loss": 1.2653,
+      "step": 730
+    },
+    {
+      "epoch": 0.4272667494703003,
+      "grad_norm": 1.9900427272526369,
+      "learning_rate": 1e-05,
+      "loss": 1.2287,
+      "step": 731
+    },
+    {
+      "epoch": 0.42785124570760574,
+      "grad_norm": 2.0590415791399526,
+      "learning_rate": 1e-05,
+      "loss": 1.2186,
+      "step": 732
+    },
+    {
+      "epoch": 0.42843574194491124,
+      "grad_norm": 1.7328955476962118,
+      "learning_rate": 1e-05,
+      "loss": 1.299,
+      "step": 733
+    },
+    {
+      "epoch": 0.4290202381822167,
+      "grad_norm": 1.9072737704002138,
+      "learning_rate": 1e-05,
+      "loss": 1.1626,
+      "step": 734
+    },
+    {
+      "epoch": 0.4296047344195222,
+      "grad_norm": 2.1072777892029566,
+      "learning_rate": 1e-05,
+      "loss": 1.1974,
+      "step": 735
+    },
+    {
+      "epoch": 0.4301892306568276,
+      "grad_norm": 1.8441091086782104,
+      "learning_rate": 1e-05,
+      "loss": 1.2612,
+      "step": 736
+    },
+    {
+      "epoch": 0.4307737268941331,
+      "grad_norm": 1.931347655645402,
+      "learning_rate": 1e-05,
+      "loss": 1.2295,
+      "step": 737
+    },
+    {
+      "epoch": 0.43135822313143857,
+      "grad_norm": 1.864772095860545,
+      "learning_rate": 1e-05,
+      "loss": 1.2262,
+      "step": 738
+    },
+    {
+      "epoch": 0.43194271936874407,
+      "grad_norm": 1.8607458254865599,
+      "learning_rate": 1e-05,
+      "loss": 1.2399,
+      "step": 739
+    },
+    {
+      "epoch": 0.4325272156060495,
+      "grad_norm": 1.9881949689188216,
+      "learning_rate": 1e-05,
+      "loss": 1.216,
+      "step": 740
+    },
+    {
+      "epoch": 0.433111711843355,
+      "grad_norm": 1.8796877547328115,
+      "learning_rate": 1e-05,
+      "loss": 1.2507,
+      "step": 741
+    },
+    {
+      "epoch": 0.4336962080806605,
+      "grad_norm": 1.7568598382673608,
+      "learning_rate": 1e-05,
+      "loss": 1.2667,
+      "step": 742
+    },
+    {
+      "epoch": 0.43428070431796595,
+      "grad_norm": 2.0151820317668743,
+      "learning_rate": 1e-05,
+      "loss": 1.2435,
+      "step": 743
+    },
+    {
+      "epoch": 0.43486520055527145,
+      "grad_norm": 1.9004953312860409,
+      "learning_rate": 1e-05,
+      "loss": 1.2727,
+      "step": 744
+    },
+    {
+      "epoch": 0.4354496967925769,
+      "grad_norm": 1.9105800362940384,
+      "learning_rate": 1e-05,
+      "loss": 1.2456,
+      "step": 745
+    },
+    {
+      "epoch": 0.4360341930298824,
+      "grad_norm": 1.9386144555377904,
+      "learning_rate": 1e-05,
+      "loss": 1.246,
+      "step": 746
+    },
+    {
+      "epoch": 0.43661868926718783,
+      "grad_norm": 1.8647250983499248,
+      "learning_rate": 1e-05,
+      "loss": 1.1995,
+      "step": 747
+    },
+    {
+      "epoch": 0.43720318550449333,
+      "grad_norm": 1.8621724620775402,
+      "learning_rate": 1e-05,
+      "loss": 1.2634,
+      "step": 748
+    },
+    {
+      "epoch": 0.4377876817417988,
+      "grad_norm": 1.916113665047828,
+      "learning_rate": 1e-05,
+      "loss": 1.2206,
+      "step": 749
+    },
+    {
+      "epoch": 0.43837217797910427,
+      "grad_norm": 1.819290548708105,
+      "learning_rate": 1e-05,
+      "loss": 1.1967,
+      "step": 750
+    },
+    {
+      "epoch": 0.4389566742164097,
+      "grad_norm": 1.910094004292115,
+      "learning_rate": 1e-05,
+      "loss": 1.2558,
+      "step": 751
+    },
+    {
+      "epoch": 0.4395411704537152,
+      "grad_norm": 1.8794307843371594,
+      "learning_rate": 1e-05,
+      "loss": 1.2468,
+      "step": 752
+    },
+    {
+      "epoch": 0.44012566669102066,
+      "grad_norm": 1.9724829588484902,
+      "learning_rate": 1e-05,
+      "loss": 1.1938,
+      "step": 753
+    },
+    {
+      "epoch": 0.44071016292832615,
+      "grad_norm": 1.7641928400978335,
+      "learning_rate": 1e-05,
+      "loss": 1.1626,
+      "step": 754
+    },
+    {
+      "epoch": 0.4412946591656316,
+      "grad_norm": 1.7586336341202051,
+      "learning_rate": 1e-05,
+      "loss": 1.2313,
+      "step": 755
+    },
+    {
+      "epoch": 0.4418791554029371,
+      "grad_norm": 2.0151715890654067,
+      "learning_rate": 1e-05,
+      "loss": 1.2319,
+      "step": 756
+    },
+    {
+      "epoch": 0.44246365164024254,
+      "grad_norm": 1.8669169618713686,
+      "learning_rate": 1e-05,
+      "loss": 1.2041,
+      "step": 757
+    },
+    {
+      "epoch": 0.44304814787754804,
+      "grad_norm": 2.050255288162073,
+      "learning_rate": 1e-05,
+      "loss": 1.1614,
+      "step": 758
+    },
+    {
+      "epoch": 0.44363264411485354,
+      "grad_norm": 1.7735814552833782,
+      "learning_rate": 1e-05,
+      "loss": 1.2114,
+      "step": 759
+    },
+    {
+      "epoch": 0.444217140352159,
+      "grad_norm": 2.0875381098461943,
+      "learning_rate": 1e-05,
+      "loss": 1.2872,
+      "step": 760
+    },
+    {
+      "epoch": 0.4448016365894645,
+      "grad_norm": 1.822066365618362,
+      "learning_rate": 1e-05,
+      "loss": 1.278,
+      "step": 761
+    },
+    {
+      "epoch": 0.4453861328267699,
+      "grad_norm": 1.9792474334240213,
+      "learning_rate": 1e-05,
+      "loss": 1.2357,
+      "step": 762
+    },
+    {
+      "epoch": 0.4459706290640754,
+      "grad_norm": 1.9006122900068625,
+      "learning_rate": 1e-05,
+      "loss": 1.255,
+      "step": 763
+    },
+    {
+      "epoch": 0.44655512530138086,
+      "grad_norm": 1.9844510835465001,
+      "learning_rate": 1e-05,
+      "loss": 1.2075,
+      "step": 764
+    },
+    {
+      "epoch": 0.44713962153868636,
+      "grad_norm": 1.9952466472355679,
+      "learning_rate": 1e-05,
+      "loss": 1.2639,
+      "step": 765
+    },
+    {
+      "epoch": 0.4477241177759918,
+      "grad_norm": 1.927759531766671,
+      "learning_rate": 1e-05,
+      "loss": 1.1951,
+      "step": 766
+    },
+    {
+      "epoch": 0.4483086140132973,
+      "grad_norm": 1.736545460882305,
+      "learning_rate": 1e-05,
+      "loss": 1.2269,
+      "step": 767
+    },
+    {
+      "epoch": 0.44889311025060274,
+      "grad_norm": 1.8490664348091488,
+      "learning_rate": 1e-05,
+      "loss": 1.2189,
+      "step": 768
+    },
+    {
+      "epoch": 0.44947760648790824,
+      "grad_norm": 1.9670110494137514,
+      "learning_rate": 1e-05,
+      "loss": 1.2168,
+      "step": 769
+    },
+    {
+      "epoch": 0.4500621027252137,
+      "grad_norm": 1.9588826645587833,
+      "learning_rate": 1e-05,
+      "loss": 1.2086,
+      "step": 770
+    },
+    {
+      "epoch": 0.4506465989625192,
+      "grad_norm": 1.9809847449194562,
+      "learning_rate": 1e-05,
+      "loss": 1.2204,
+      "step": 771
+    },
+    {
+      "epoch": 0.45123109519982463,
+      "grad_norm": 1.9425784970980797,
+      "learning_rate": 1e-05,
+      "loss": 1.1921,
+      "step": 772
+    },
+    {
+      "epoch": 0.4518155914371301,
+      "grad_norm": 1.8780077504278296,
+      "learning_rate": 1e-05,
+      "loss": 1.2492,
+      "step": 773
+    },
+    {
+      "epoch": 0.45240008767443557,
+      "grad_norm": 1.7180506482447975,
+      "learning_rate": 1e-05,
+      "loss": 1.223,
+      "step": 774
+    },
+    {
+      "epoch": 0.45298458391174107,
+      "grad_norm": 1.8218591878539512,
+      "learning_rate": 1e-05,
+      "loss": 1.2139,
+      "step": 775
+    },
+    {
+      "epoch": 0.45356908014904657,
+      "grad_norm": 1.8271810863109375,
+      "learning_rate": 1e-05,
+      "loss": 1.2017,
+      "step": 776
+    },
+    {
+      "epoch": 0.454153576386352,
+      "grad_norm": 1.8766596842782723,
+      "learning_rate": 1e-05,
+      "loss": 1.2167,
+      "step": 777
+    },
+    {
+      "epoch": 0.4547380726236575,
+      "grad_norm": 1.7593982123632694,
+      "learning_rate": 1e-05,
+      "loss": 1.2243,
+      "step": 778
+    },
+    {
+      "epoch": 0.45532256886096295,
+      "grad_norm": 1.8799361652527222,
+      "learning_rate": 1e-05,
+      "loss": 1.2783,
+      "step": 779
+    },
+    {
+      "epoch": 0.45590706509826845,
+      "grad_norm": 1.86508085342816,
+      "learning_rate": 1e-05,
+      "loss": 1.2472,
+      "step": 780
+    },
+    {
+      "epoch": 0.4564915613355739,
+      "grad_norm": 1.8387018083583029,
+      "learning_rate": 1e-05,
+      "loss": 1.2944,
+      "step": 781
+    },
+    {
+      "epoch": 0.4570760575728794,
+      "grad_norm": 1.804106967947994,
+      "learning_rate": 1e-05,
+      "loss": 1.225,
+      "step": 782
+    },
+    {
+      "epoch": 0.45766055381018483,
+      "grad_norm": 1.7784579436931307,
+      "learning_rate": 1e-05,
+      "loss": 1.2394,
+      "step": 783
+    },
+    {
+      "epoch": 0.45824505004749033,
+      "grad_norm": 2.0392400589176525,
+      "learning_rate": 1e-05,
+      "loss": 1.2231,
+      "step": 784
+    },
+    {
+      "epoch": 0.4588295462847958,
+      "grad_norm": 1.8903717738523407,
+      "learning_rate": 1e-05,
+      "loss": 1.2239,
+      "step": 785
+    },
+    {
+      "epoch": 0.4594140425221013,
+      "grad_norm": 1.7919941885532256,
+      "learning_rate": 1e-05,
+      "loss": 1.2783,
+      "step": 786
+    },
+    {
+      "epoch": 0.4599985387594067,
+      "grad_norm": 1.9745210572326655,
+      "learning_rate": 1e-05,
+      "loss": 1.2693,
+      "step": 787
+    },
+    {
+      "epoch": 0.4605830349967122,
+      "grad_norm": 1.792133242636464,
+      "learning_rate": 1e-05,
+      "loss": 1.2276,
+      "step": 788
+    },
+    {
+      "epoch": 0.46116753123401766,
+      "grad_norm": 1.825988293264415,
+      "learning_rate": 1e-05,
+      "loss": 1.2551,
+      "step": 789
+    },
+    {
+      "epoch": 0.46175202747132316,
+      "grad_norm": 1.760055203502491,
+      "learning_rate": 1e-05,
+      "loss": 1.2275,
+      "step": 790
+    },
+    {
+      "epoch": 0.4623365237086286,
+      "grad_norm": 2.096076688031519,
+      "learning_rate": 1e-05,
+      "loss": 1.1848,
+      "step": 791
+    },
+    {
+      "epoch": 0.4629210199459341,
+      "grad_norm": 1.8076365250342177,
+      "learning_rate": 1e-05,
+      "loss": 1.2134,
+      "step": 792
+    },
+    {
+      "epoch": 0.4635055161832396,
+      "grad_norm": 2.031984230126923,
+      "learning_rate": 1e-05,
+      "loss": 1.225,
+      "step": 793
+    },
+    {
+      "epoch": 0.46409001242054504,
+      "grad_norm": 1.870425721883526,
+      "learning_rate": 1e-05,
+      "loss": 1.1946,
+      "step": 794
+    },
+    {
+      "epoch": 0.46467450865785054,
+      "grad_norm": 1.9909996618656294,
+      "learning_rate": 1e-05,
+      "loss": 1.239,
+      "step": 795
+    },
+    {
+      "epoch": 0.465259004895156,
+      "grad_norm": 1.8445178413273005,
+      "learning_rate": 1e-05,
+      "loss": 1.2205,
+      "step": 796
+    },
+    {
+      "epoch": 0.4658435011324615,
+      "grad_norm": 1.8384733982294081,
+      "learning_rate": 1e-05,
+      "loss": 1.2403,
+      "step": 797
+    },
+    {
+      "epoch": 0.4664279973697669,
+      "grad_norm": 1.8435141744624612,
+      "learning_rate": 1e-05,
+      "loss": 1.2892,
+      "step": 798
+    },
+    {
+      "epoch": 0.4670124936070724,
+      "grad_norm": 1.8877113689564176,
+      "learning_rate": 1e-05,
+      "loss": 1.2293,
+      "step": 799
+    },
+    {
+      "epoch": 0.46759698984437786,
+      "grad_norm": 1.840810713851528,
+      "learning_rate": 1e-05,
+      "loss": 1.3008,
+      "step": 800
+    },
+    {
+      "epoch": 0.46818148608168336,
+      "grad_norm": 1.9501369739717633,
+      "learning_rate": 1e-05,
+      "loss": 1.2724,
+      "step": 801
+    },
+    {
+      "epoch": 0.4687659823189888,
+      "grad_norm": 1.8069106987428811,
+      "learning_rate": 1e-05,
+      "loss": 1.2424,
+      "step": 802
+    },
+    {
+      "epoch": 0.4693504785562943,
+      "grad_norm": 1.734556294898529,
+      "learning_rate": 1e-05,
+      "loss": 1.2352,
+      "step": 803
+    },
+    {
+      "epoch": 0.46993497479359975,
+      "grad_norm": 2.0169072883660943,
+      "learning_rate": 1e-05,
+      "loss": 1.2501,
+      "step": 804
+    },
+    {
+      "epoch": 0.47051947103090525,
+      "grad_norm": 1.8653784847688415,
+      "learning_rate": 1e-05,
+      "loss": 1.2323,
+      "step": 805
+    },
+    {
+      "epoch": 0.4711039672682107,
+      "grad_norm": 1.7052497626715415,
+      "learning_rate": 1e-05,
+      "loss": 1.2061,
+      "step": 806
+    },
+    {
+      "epoch": 0.4716884635055162,
+      "grad_norm": 1.9688195280422442,
+      "learning_rate": 1e-05,
+      "loss": 1.1884,
+      "step": 807
+    },
+    {
+      "epoch": 0.47227295974282163,
+      "grad_norm": 2.0637169220589944,
+      "learning_rate": 1e-05,
+      "loss": 1.2327,
+      "step": 808
+    },
+    {
+      "epoch": 0.47285745598012713,
+      "grad_norm": 1.7063021858838179,
+      "learning_rate": 1e-05,
+      "loss": 1.276,
+      "step": 809
+    },
+    {
+      "epoch": 0.4734419522174326,
+      "grad_norm": 1.6927660750583429,
+      "learning_rate": 1e-05,
+      "loss": 1.2303,
+      "step": 810
+    },
+    {
+      "epoch": 0.47402644845473807,
+      "grad_norm": 1.8971605234433193,
+      "learning_rate": 1e-05,
+      "loss": 1.2351,
+      "step": 811
+    },
+    {
+      "epoch": 0.47461094469204357,
+      "grad_norm": 1.9371955640216165,
+      "learning_rate": 1e-05,
+      "loss": 1.2246,
+      "step": 812
+    },
+    {
+      "epoch": 0.475195440929349,
+      "grad_norm": 1.965740065974074,
+      "learning_rate": 1e-05,
+      "loss": 1.2436,
+      "step": 813
+    },
+    {
+      "epoch": 0.4757799371666545,
+      "grad_norm": 1.9738250498616596,
+      "learning_rate": 1e-05,
+      "loss": 1.2192,
+      "step": 814
+    },
+    {
+      "epoch": 0.47636443340395995,
+      "grad_norm": 1.9045701316624748,
+      "learning_rate": 1e-05,
+      "loss": 1.2493,
+      "step": 815
+    },
+    {
+      "epoch": 0.47694892964126545,
+      "grad_norm": 1.8736213966087798,
+      "learning_rate": 1e-05,
+      "loss": 1.2295,
+      "step": 816
+    },
+    {
+      "epoch": 0.4775334258785709,
+      "grad_norm": 1.8377617122398722,
+      "learning_rate": 1e-05,
+      "loss": 1.2435,
+      "step": 817
+    },
+    {
+      "epoch": 0.4781179221158764,
+      "grad_norm": 1.8125474350573036,
+      "learning_rate": 1e-05,
+      "loss": 1.229,
+      "step": 818
+    },
+    {
+      "epoch": 0.47870241835318184,
+      "grad_norm": 1.767671568434817,
+      "learning_rate": 1e-05,
+      "loss": 1.2525,
+      "step": 819
+    },
+    {
+      "epoch": 0.47928691459048733,
+      "grad_norm": 1.9118966009848244,
+      "learning_rate": 1e-05,
+      "loss": 1.2739,
+      "step": 820
+    },
+    {
+      "epoch": 0.4798714108277928,
+      "grad_norm": 1.971933285459566,
+      "learning_rate": 1e-05,
+      "loss": 1.2618,
+      "step": 821
+    },
+    {
+      "epoch": 0.4804559070650983,
+      "grad_norm": 2.0664895224462723,
+      "learning_rate": 1e-05,
+      "loss": 1.2341,
+      "step": 822
+    },
+    {
+      "epoch": 0.4810404033024037,
+      "grad_norm": 1.995514305362607,
+      "learning_rate": 1e-05,
+      "loss": 1.2182,
+      "step": 823
+    },
+    {
+      "epoch": 0.4816248995397092,
+      "grad_norm": 1.929117402546271,
+      "learning_rate": 1e-05,
+      "loss": 1.2475,
+      "step": 824
+    },
+    {
+      "epoch": 0.48220939577701466,
+      "grad_norm": 2.0266424745437335,
+      "learning_rate": 1e-05,
+      "loss": 1.2117,
+      "step": 825
+    },
+    {
+      "epoch": 0.48279389201432016,
+      "grad_norm": 1.7801795866153023,
+      "learning_rate": 1e-05,
+      "loss": 1.2604,
+      "step": 826
+    },
+    {
+      "epoch": 0.48337838825162566,
+      "grad_norm": 1.8278162795333812,
+      "learning_rate": 1e-05,
+      "loss": 1.2351,
+      "step": 827
+    },
+    {
+      "epoch": 0.4839628844889311,
+      "grad_norm": 1.8653854207160967,
+      "learning_rate": 1e-05,
+      "loss": 1.2501,
+      "step": 828
+    },
+    {
+      "epoch": 0.4845473807262366,
+      "grad_norm": 1.8186000881738722,
+      "learning_rate": 1e-05,
+      "loss": 1.2137,
+      "step": 829
+    },
+    {
+      "epoch": 0.48513187696354204,
+      "grad_norm": 1.7026939707173758,
+      "learning_rate": 1e-05,
+      "loss": 1.2168,
+      "step": 830
+    },
+    {
+      "epoch": 0.48571637320084754,
+      "grad_norm": 1.8417391383769088,
+      "learning_rate": 1e-05,
+      "loss": 1.2301,
+      "step": 831
+    },
+    {
+      "epoch": 0.486300869438153,
+      "grad_norm": 1.782413155957072,
+      "learning_rate": 1e-05,
+      "loss": 1.2582,
+      "step": 832
+    },
+    {
+      "epoch": 0.4868853656754585,
+      "grad_norm": 1.8509099369982414,
+      "learning_rate": 1e-05,
+      "loss": 1.2591,
+      "step": 833
+    },
+    {
+      "epoch": 0.4874698619127639,
+      "grad_norm": 1.8338688007306392,
+      "learning_rate": 1e-05,
+      "loss": 1.2328,
+      "step": 834
+    },
+    {
+      "epoch": 0.4880543581500694,
+      "grad_norm": 1.944271094882053,
+      "learning_rate": 1e-05,
+      "loss": 1.2076,
+      "step": 835
+    },
+    {
+      "epoch": 0.48863885438737487,
+      "grad_norm": 1.8344515941299722,
+      "learning_rate": 1e-05,
+      "loss": 1.2123,
+      "step": 836
+    },
+    {
+      "epoch": 0.48922335062468036,
+      "grad_norm": 1.8554131022584306,
+      "learning_rate": 1e-05,
+      "loss": 1.1975,
+      "step": 837
+    },
+    {
+      "epoch": 0.4898078468619858,
+      "grad_norm": 2.1491969952496883,
+      "learning_rate": 1e-05,
+      "loss": 1.2213,
+      "step": 838
+    },
+    {
+      "epoch": 0.4903923430992913,
+      "grad_norm": 1.8110625795099486,
+      "learning_rate": 1e-05,
+      "loss": 1.2603,
+      "step": 839
+    },
+    {
+      "epoch": 0.49097683933659675,
+      "grad_norm": 1.9481177836382817,
+      "learning_rate": 1e-05,
+      "loss": 1.2316,
+      "step": 840
+    },
+    {
+      "epoch": 0.49156133557390225,
+      "grad_norm": 1.8519093110672213,
+      "learning_rate": 1e-05,
+      "loss": 1.2528,
+      "step": 841
+    },
+    {
+      "epoch": 0.4921458318112077,
+      "grad_norm": 1.9292108161987345,
+      "learning_rate": 1e-05,
+      "loss": 1.1871,
+      "step": 842
+    },
+    {
+      "epoch": 0.4927303280485132,
+      "grad_norm": 2.136233194848059,
+      "learning_rate": 1e-05,
+      "loss": 1.256,
+      "step": 843
+    },
+    {
+      "epoch": 0.4933148242858187,
+      "grad_norm": 2.258687790332659,
+      "learning_rate": 1e-05,
+      "loss": 1.1939,
+      "step": 844
+    },
+    {
+      "epoch": 0.49389932052312413,
+      "grad_norm": 1.579978317526749,
+      "learning_rate": 1e-05,
+      "loss": 1.2292,
+      "step": 845
+    },
+    {
+      "epoch": 0.49448381676042963,
+      "grad_norm": 1.9161605264417305,
+      "learning_rate": 1e-05,
+      "loss": 1.1936,
+      "step": 846
+    },
+    {
+      "epoch": 0.49506831299773507,
+      "grad_norm": 1.9984041286835748,
+      "learning_rate": 1e-05,
+      "loss": 1.2672,
+      "step": 847
+    },
+    {
+      "epoch": 0.49565280923504057,
+      "grad_norm": 1.8010127041278952,
+      "learning_rate": 1e-05,
+      "loss": 1.2202,
+      "step": 848
+    },
+    {
+      "epoch": 0.496237305472346,
+      "grad_norm": 1.810058534377188,
+      "learning_rate": 1e-05,
+      "loss": 1.2234,
+      "step": 849
+    },
+    {
+      "epoch": 0.4968218017096515,
+      "grad_norm": 1.8346998853464642,
+      "learning_rate": 1e-05,
+      "loss": 1.243,
+      "step": 850
+    },
+    {
+      "epoch": 0.49740629794695695,
+      "grad_norm": 1.9012321578836708,
+      "learning_rate": 1e-05,
+      "loss": 1.2147,
+      "step": 851
+    },
+    {
+      "epoch": 0.49799079418426245,
+      "grad_norm": 1.9100061555125138,
+      "learning_rate": 1e-05,
+      "loss": 1.2446,
+      "step": 852
+    },
+    {
+      "epoch": 0.4985752904215679,
+      "grad_norm": 1.9708025200888237,
+      "learning_rate": 1e-05,
+      "loss": 1.246,
+      "step": 853
+    },
+    {
+      "epoch": 0.4991597866588734,
+      "grad_norm": 1.7993343052241861,
+      "learning_rate": 1e-05,
+      "loss": 1.228,
+      "step": 854
+    },
+    {
+      "epoch": 0.49974428289617884,
+      "grad_norm": 1.8029605906264032,
+      "learning_rate": 1e-05,
+      "loss": 1.1943,
+      "step": 855
+    },
+    {
+      "epoch": 0.5003287791334843,
+      "grad_norm": 1.8684325597863052,
+      "learning_rate": 1e-05,
+      "loss": 1.2663,
+      "step": 856
+    },
+    {
+      "epoch": 0.5009132753707898,
+      "grad_norm": 1.8073593231096123,
+      "learning_rate": 1e-05,
+      "loss": 1.2302,
+      "step": 857
+    },
+    {
+      "epoch": 0.5014977716080953,
+      "grad_norm": 1.716012293531449,
+      "learning_rate": 1e-05,
+      "loss": 1.2422,
+      "step": 858
+    },
+    {
+      "epoch": 0.5020822678454008,
+      "grad_norm": 1.8301317828870955,
+      "learning_rate": 1e-05,
+      "loss": 1.3029,
+      "step": 859
+    },
+    {
+      "epoch": 0.5026667640827062,
+      "grad_norm": 1.8861494642199157,
+      "learning_rate": 1e-05,
+      "loss": 1.2452,
+      "step": 860
+    },
+    {
+      "epoch": 0.5032512603200117,
+      "grad_norm": 1.7322899277540442,
+      "learning_rate": 1e-05,
+      "loss": 1.1834,
+      "step": 861
+    },
+    {
+      "epoch": 0.5038357565573172,
+      "grad_norm": 1.8240962136270111,
+      "learning_rate": 1e-05,
+      "loss": 1.2111,
+      "step": 862
+    },
+    {
+      "epoch": 0.5044202527946227,
+      "grad_norm": 1.8422011330406862,
+      "learning_rate": 1e-05,
+      "loss": 1.2225,
+      "step": 863
+    },
+    {
+      "epoch": 0.5050047490319282,
+      "grad_norm": 1.8558676352373913,
+      "learning_rate": 1e-05,
+      "loss": 1.2068,
+      "step": 864
+    },
+    {
+      "epoch": 0.5055892452692335,
+      "grad_norm": 1.824594742315542,
+      "learning_rate": 1e-05,
+      "loss": 1.2216,
+      "step": 865
+    },
+    {
+      "epoch": 0.506173741506539,
+      "grad_norm": 2.0060714876629393,
+      "learning_rate": 1e-05,
+      "loss": 1.233,
+      "step": 866
+    },
+    {
+      "epoch": 0.5067582377438445,
+      "grad_norm": 1.8019886994353047,
+      "learning_rate": 1e-05,
+      "loss": 1.2552,
+      "step": 867
+    },
+    {
+      "epoch": 0.50734273398115,
+      "grad_norm": 1.813053075502338,
+      "learning_rate": 1e-05,
+      "loss": 1.2576,
+      "step": 868
+    },
+    {
+      "epoch": 0.5079272302184554,
+      "grad_norm": 1.7702745924299317,
+      "learning_rate": 1e-05,
+      "loss": 1.2207,
+      "step": 869
+    },
+    {
+      "epoch": 0.5085117264557609,
+      "grad_norm": 2.08241153454326,
+      "learning_rate": 1e-05,
+      "loss": 1.1829,
+      "step": 870
+    },
+    {
+      "epoch": 0.5090962226930664,
+      "grad_norm": 1.8789487073341697,
+      "learning_rate": 1e-05,
+      "loss": 1.2257,
+      "step": 871
+    },
+    {
+      "epoch": 0.5096807189303719,
+      "grad_norm": 1.7157607053781747,
+      "learning_rate": 1e-05,
+      "loss": 1.2163,
+      "step": 872
+    },
+    {
+      "epoch": 0.5102652151676773,
+      "grad_norm": 1.9574484909093388,
+      "learning_rate": 1e-05,
+      "loss": 1.2009,
+      "step": 873
+    },
+    {
+      "epoch": 0.5108497114049828,
+      "grad_norm": 1.9686893085121924,
+      "learning_rate": 1e-05,
+      "loss": 1.2479,
+      "step": 874
+    },
+    {
+      "epoch": 0.5114342076422883,
+      "grad_norm": 1.6937176259166784,
+      "learning_rate": 1e-05,
+      "loss": 1.21,
+      "step": 875
+    },
+    {
+      "epoch": 0.5120187038795938,
+      "grad_norm": 1.861982584190904,
+      "learning_rate": 1e-05,
+      "loss": 1.2128,
+      "step": 876
+    },
+    {
+      "epoch": 0.5126032001168992,
+      "grad_norm": 1.9900273976628204,
+      "learning_rate": 1e-05,
+      "loss": 1.2393,
+      "step": 877
+    },
+    {
+      "epoch": 0.5131876963542047,
+      "grad_norm": 2.1322395699865226,
+      "learning_rate": 1e-05,
+      "loss": 1.2917,
+      "step": 878
+    },
+    {
+      "epoch": 0.5137721925915102,
+      "grad_norm": 1.995024745430685,
+      "learning_rate": 1e-05,
+      "loss": 1.2451,
+      "step": 879
+    },
+    {
+      "epoch": 0.5143566888288157,
+      "grad_norm": 1.9358407494247944,
+      "learning_rate": 1e-05,
+      "loss": 1.2693,
+      "step": 880
+    },
+    {
+      "epoch": 0.5149411850661212,
+      "grad_norm": 1.7508281874022855,
+      "learning_rate": 1e-05,
+      "loss": 1.2399,
+      "step": 881
+    },
+    {
+      "epoch": 0.5155256813034266,
+      "grad_norm": 1.7698557979497644,
+      "learning_rate": 1e-05,
+      "loss": 1.2497,
+      "step": 882
+    },
+    {
+      "epoch": 0.5161101775407321,
+      "grad_norm": 1.8180156809893435,
+      "learning_rate": 1e-05,
+      "loss": 1.2153,
+      "step": 883
+    },
+    {
+      "epoch": 0.5166946737780376,
+      "grad_norm": 1.857180040521379,
+      "learning_rate": 1e-05,
+      "loss": 1.2147,
+      "step": 884
+    },
+    {
+      "epoch": 0.5172791700153431,
+      "grad_norm": 1.983350463397269,
+      "learning_rate": 1e-05,
+      "loss": 1.2337,
+      "step": 885
+    },
+    {
+      "epoch": 0.5178636662526485,
+      "grad_norm": 1.674711788894685,
+      "learning_rate": 1e-05,
+      "loss": 1.2604,
+      "step": 886
+    },
+    {
+      "epoch": 0.518448162489954,
+      "grad_norm": 1.8649435257541165,
+      "learning_rate": 1e-05,
+      "loss": 1.2134,
+      "step": 887
+    },
+    {
+      "epoch": 0.5190326587272595,
+      "grad_norm": 1.774125723816856,
+      "learning_rate": 1e-05,
+      "loss": 1.2944,
+      "step": 888
+    },
+    {
+      "epoch": 0.519617154964565,
+      "grad_norm": 2.0942247644094,
+      "learning_rate": 1e-05,
+      "loss": 1.234,
+      "step": 889
+    },
+    {
+      "epoch": 0.5202016512018703,
+      "grad_norm": 1.9683958101850505,
+      "learning_rate": 1e-05,
+      "loss": 1.2505,
+      "step": 890
+    },
+    {
+      "epoch": 0.5207861474391758,
+      "grad_norm": 1.6383063710962713,
+      "learning_rate": 1e-05,
+      "loss": 1.2134,
+      "step": 891
+    },
+    {
+      "epoch": 0.5213706436764813,
+      "grad_norm": 1.7664506444282555,
+      "learning_rate": 1e-05,
+      "loss": 1.2187,
+      "step": 892
+    },
+    {
+      "epoch": 0.5219551399137868,
+      "grad_norm": 1.7839465040406772,
+      "learning_rate": 1e-05,
+      "loss": 1.2304,
+      "step": 893
+    },
+    {
+      "epoch": 0.5225396361510922,
+      "grad_norm": 1.7207994296345346,
+      "learning_rate": 1e-05,
+      "loss": 1.2468,
+      "step": 894
+    },
+    {
+      "epoch": 0.5231241323883977,
+      "grad_norm": 1.6946351917089502,
+      "learning_rate": 1e-05,
+      "loss": 1.2627,
+      "step": 895
+    },
+    {
+      "epoch": 0.5237086286257032,
+      "grad_norm": 1.9419361740948922,
+      "learning_rate": 1e-05,
+      "loss": 1.2082,
+      "step": 896
+    },
+    {
+      "epoch": 0.5242931248630087,
+      "grad_norm": 1.7279655053036223,
+      "learning_rate": 1e-05,
+      "loss": 1.2177,
+      "step": 897
+    },
+    {
+      "epoch": 0.5248776211003142,
+      "grad_norm": 1.7155077313122193,
+      "learning_rate": 1e-05,
+      "loss": 1.2646,
+      "step": 898
+    },
+    {
+      "epoch": 0.5254621173376196,
+      "grad_norm": 1.8700477082077807,
+      "learning_rate": 1e-05,
+      "loss": 1.2461,
+      "step": 899
+    },
+    {
+      "epoch": 0.5260466135749251,
+      "grad_norm": 1.8083403608539446,
+      "learning_rate": 1e-05,
+      "loss": 1.2804,
+      "step": 900
+    },
+    {
+      "epoch": 0.5266311098122306,
+      "grad_norm": 1.7754060948525496,
+      "learning_rate": 1e-05,
+      "loss": 1.1991,
+      "step": 901
+    },
+    {
+      "epoch": 0.5272156060495361,
+      "grad_norm": 2.0056105182648905,
+      "learning_rate": 1e-05,
+      "loss": 1.199,
+      "step": 902
+    },
+    {
+      "epoch": 0.5278001022868415,
+      "grad_norm": 2.016296394243527,
+      "learning_rate": 1e-05,
+      "loss": 1.2235,
+      "step": 903
+    },
+    {
+      "epoch": 0.528384598524147,
+      "grad_norm": 2.038548294184711,
+      "learning_rate": 1e-05,
+      "loss": 1.1982,
+      "step": 904
+    },
+    {
+      "epoch": 0.5289690947614525,
+      "grad_norm": 2.1119700204488465,
+      "learning_rate": 1e-05,
+      "loss": 1.2171,
+      "step": 905
+    },
+    {
+      "epoch": 0.529553590998758,
+      "grad_norm": 1.733101626678849,
+      "learning_rate": 1e-05,
+      "loss": 1.1852,
+      "step": 906
+    },
+    {
+      "epoch": 0.5301380872360634,
+      "grad_norm": 1.6479006347921163,
+      "learning_rate": 1e-05,
+      "loss": 1.2285,
+      "step": 907
+    },
+    {
+      "epoch": 0.5307225834733689,
+      "grad_norm": 1.9322597600972746,
+      "learning_rate": 1e-05,
+      "loss": 1.2719,
+      "step": 908
+    },
+    {
+      "epoch": 0.5313070797106744,
+      "grad_norm": 1.8442128927056565,
+      "learning_rate": 1e-05,
+      "loss": 1.2737,
+      "step": 909
+    },
+    {
+      "epoch": 0.5318915759479799,
+      "grad_norm": 1.750115838602223,
+      "learning_rate": 1e-05,
+      "loss": 1.1906,
+      "step": 910
+    },
+    {
+      "epoch": 0.5324760721852853,
+      "grad_norm": 1.8977844110297504,
+      "learning_rate": 1e-05,
+      "loss": 1.1918,
+      "step": 911
+    },
+    {
+      "epoch": 0.5330605684225908,
+      "grad_norm": 1.7452510899591185,
+      "learning_rate": 1e-05,
+      "loss": 1.225,
+      "step": 912
+    },
+    {
+      "epoch": 0.5336450646598963,
+      "grad_norm": 1.8476156153713907,
+      "learning_rate": 1e-05,
+      "loss": 1.2429,
+      "step": 913
+    },
+    {
+      "epoch": 0.5342295608972017,
+      "grad_norm": 1.8358998200303784,
+      "learning_rate": 1e-05,
+      "loss": 1.2586,
+      "step": 914
+    },
+    {
+      "epoch": 0.5348140571345072,
+      "grad_norm": 2.071238496741492,
+      "learning_rate": 1e-05,
+      "loss": 1.2276,
+      "step": 915
+    },
+    {
+      "epoch": 0.5353985533718126,
+      "grad_norm": 1.8373107317137982,
+      "learning_rate": 1e-05,
+      "loss": 1.2371,
+      "step": 916
+    },
+    {
+      "epoch": 0.5359830496091181,
+      "grad_norm": 1.8496344378486158,
+      "learning_rate": 1e-05,
+      "loss": 1.2303,
+      "step": 917
+    },
+    {
+      "epoch": 0.5365675458464236,
+      "grad_norm": 1.9613734424223,
+      "learning_rate": 1e-05,
+      "loss": 1.2403,
+      "step": 918
+    },
+    {
+      "epoch": 0.5371520420837291,
+      "grad_norm": 1.713940570167926,
+      "learning_rate": 1e-05,
+      "loss": 1.2041,
+      "step": 919
+    },
+    {
+      "epoch": 0.5377365383210345,
+      "grad_norm": 1.70569985075408,
+      "learning_rate": 1e-05,
+      "loss": 1.1861,
+      "step": 920
+    },
+    {
+      "epoch": 0.53832103455834,
+      "grad_norm": 1.7929725044320473,
+      "learning_rate": 1e-05,
+      "loss": 1.208,
+      "step": 921
+    },
+    {
+      "epoch": 0.5389055307956455,
+      "grad_norm": 1.7652854883090803,
+      "learning_rate": 1e-05,
+      "loss": 1.1796,
+      "step": 922
+    },
+    {
+      "epoch": 0.539490027032951,
+      "grad_norm": 1.736580823382608,
+      "learning_rate": 1e-05,
+      "loss": 1.2831,
+      "step": 923
+    },
+    {
+      "epoch": 0.5400745232702564,
+      "grad_norm": 1.7413391897176682,
+      "learning_rate": 1e-05,
+      "loss": 1.207,
+      "step": 924
+    },
+    {
+      "epoch": 0.5406590195075619,
+      "grad_norm": 1.7932760109786618,
+      "learning_rate": 1e-05,
+      "loss": 1.2226,
+      "step": 925
+    },
+    {
+      "epoch": 0.5412435157448674,
+      "grad_norm": 1.8429759437299793,
+      "learning_rate": 1e-05,
+      "loss": 1.2042,
+      "step": 926
+    },
+    {
+      "epoch": 0.5418280119821729,
+      "grad_norm": 1.7949863774866406,
+      "learning_rate": 1e-05,
+      "loss": 1.1953,
+      "step": 927
+    },
+    {
+      "epoch": 0.5424125082194783,
+      "grad_norm": 1.928593326318846,
+      "learning_rate": 1e-05,
+      "loss": 1.2092,
+      "step": 928
+    },
+    {
+      "epoch": 0.5429970044567838,
+      "grad_norm": 1.9000119328022973,
+      "learning_rate": 1e-05,
+      "loss": 1.1688,
+      "step": 929
+    },
+    {
+      "epoch": 0.5435815006940893,
+      "grad_norm": 1.8491938599888895,
+      "learning_rate": 1e-05,
+      "loss": 1.2277,
+      "step": 930
+    },
+    {
+      "epoch": 0.5441659969313948,
+      "grad_norm": 1.9414930195624969,
+      "learning_rate": 1e-05,
+      "loss": 1.2154,
+      "step": 931
+    },
+    {
+      "epoch": 0.5447504931687003,
+      "grad_norm": 1.8275855281963016,
+      "learning_rate": 1e-05,
+      "loss": 1.1767,
+      "step": 932
+    },
+    {
+      "epoch": 0.5453349894060057,
+      "grad_norm": 1.9466099276684548,
+      "learning_rate": 1e-05,
+      "loss": 1.2275,
+      "step": 933
+    },
+    {
+      "epoch": 0.5459194856433112,
+      "grad_norm": 1.9088766535428652,
+      "learning_rate": 1e-05,
+      "loss": 1.1899,
+      "step": 934
+    },
+    {
+      "epoch": 0.5465039818806167,
+      "grad_norm": 1.787170847404954,
+      "learning_rate": 1e-05,
+      "loss": 1.2493,
+      "step": 935
+    },
+    {
+      "epoch": 0.5470884781179222,
+      "grad_norm": 1.8032848775212063,
+      "learning_rate": 1e-05,
+      "loss": 1.2055,
+      "step": 936
+    },
+    {
+      "epoch": 0.5476729743552275,
+      "grad_norm": 1.7358436744228711,
+      "learning_rate": 1e-05,
+      "loss": 1.2079,
+      "step": 937
+    },
+    {
+      "epoch": 0.548257470592533,
+      "grad_norm": 1.8009403188373418,
+      "learning_rate": 1e-05,
+      "loss": 1.2339,
+      "step": 938
+    },
+    {
+      "epoch": 0.5488419668298385,
+      "grad_norm": 1.8333668426293745,
+      "learning_rate": 1e-05,
+      "loss": 1.2483,
+      "step": 939
+    },
+    {
+      "epoch": 0.549426463067144,
+      "grad_norm": 1.7628083143483924,
+      "learning_rate": 1e-05,
+      "loss": 1.2467,
+      "step": 940
+    },
+    {
+      "epoch": 0.5500109593044494,
+      "grad_norm": 1.7452942664201216,
+      "learning_rate": 1e-05,
+      "loss": 1.2263,
+      "step": 941
+    },
+    {
+      "epoch": 0.5505954555417549,
+      "grad_norm": 1.7564706403983135,
+      "learning_rate": 1e-05,
+      "loss": 1.2364,
+      "step": 942
+    },
+    {
+      "epoch": 0.5511799517790604,
+      "grad_norm": 1.9284719847032055,
+      "learning_rate": 1e-05,
+      "loss": 1.266,
+      "step": 943
+    },
+    {
+      "epoch": 0.5517644480163659,
+      "grad_norm": 1.8407149833158474,
+      "learning_rate": 1e-05,
+      "loss": 1.1912,
+      "step": 944
+    },
+    {
+      "epoch": 0.5523489442536713,
+      "grad_norm": 1.657447001037217,
+      "learning_rate": 1e-05,
+      "loss": 1.2368,
+      "step": 945
+    },
+    {
+      "epoch": 0.5529334404909768,
+      "grad_norm": 1.7430001297901578,
+      "learning_rate": 1e-05,
+      "loss": 1.2121,
+      "step": 946
+    },
+    {
+      "epoch": 0.5535179367282823,
+      "grad_norm": 1.800523782313942,
+      "learning_rate": 1e-05,
+      "loss": 1.2183,
+      "step": 947
+    },
+    {
+      "epoch": 0.5541024329655878,
+      "grad_norm": 1.916370067772338,
+      "learning_rate": 1e-05,
+      "loss": 1.2042,
+      "step": 948
+    },
+    {
+      "epoch": 0.5546869292028933,
+      "grad_norm": 1.7383116965593701,
+      "learning_rate": 1e-05,
+      "loss": 1.2258,
+      "step": 949
+    },
+    {
+      "epoch": 0.5552714254401987,
+      "grad_norm": 1.7961721311832999,
+      "learning_rate": 1e-05,
+      "loss": 1.2648,
+      "step": 950
+    },
+    {
+      "epoch": 0.5558559216775042,
+      "grad_norm": 1.7620243986558444,
+      "learning_rate": 1e-05,
+      "loss": 1.1994,
+      "step": 951
+    },
+    {
+      "epoch": 0.5564404179148097,
+      "grad_norm": 1.7755328342804848,
+      "learning_rate": 1e-05,
+      "loss": 1.2079,
+      "step": 952
+    },
+    {
+      "epoch": 0.5570249141521152,
+      "grad_norm": 2.0858331838123525,
+      "learning_rate": 1e-05,
+      "loss": 1.2315,
+      "step": 953
+    },
+    {
+      "epoch": 0.5576094103894206,
+      "grad_norm": 2.1018151530745635,
+      "learning_rate": 1e-05,
+      "loss": 1.2701,
+      "step": 954
+    },
+    {
+      "epoch": 0.5581939066267261,
+      "grad_norm": 1.8834230348683787,
+      "learning_rate": 1e-05,
+      "loss": 1.2088,
+      "step": 955
+    },
+    {
+      "epoch": 0.5587784028640316,
+      "grad_norm": 1.6916296509832347,
+      "learning_rate": 1e-05,
+      "loss": 1.2254,
+      "step": 956
+    },
+    {
+      "epoch": 0.5593628991013371,
+      "grad_norm": 1.7600505216658389,
+      "learning_rate": 1e-05,
+      "loss": 1.2102,
+      "step": 957
+    },
+    {
+      "epoch": 0.5599473953386425,
+      "grad_norm": 1.7895746896103744,
+      "learning_rate": 1e-05,
+      "loss": 1.2528,
+      "step": 958
+    },
+    {
+      "epoch": 0.560531891575948,
+      "grad_norm": 1.6880845785651681,
+      "learning_rate": 1e-05,
+      "loss": 1.2008,
+      "step": 959
+    },
+    {
+      "epoch": 0.5611163878132535,
+      "grad_norm": 1.8220713745019033,
+      "learning_rate": 1e-05,
+      "loss": 1.243,
+      "step": 960
+    },
+    {
+      "epoch": 0.561700884050559,
+      "grad_norm": 1.8147363665166483,
+      "learning_rate": 1e-05,
+      "loss": 1.1985,
+      "step": 961
+    },
+    {
+      "epoch": 0.5622853802878643,
+      "grad_norm": 1.7796501643766898,
+      "learning_rate": 1e-05,
+      "loss": 1.2315,
+      "step": 962
+    },
+    {
+      "epoch": 0.5628698765251698,
+      "grad_norm": 1.7312163280252473,
+      "learning_rate": 1e-05,
+      "loss": 1.2009,
+      "step": 963
+    },
+    {
+      "epoch": 0.5634543727624753,
+      "grad_norm": 1.8337062987070347,
+      "learning_rate": 1e-05,
+      "loss": 1.2075,
+      "step": 964
+    },
+    {
+      "epoch": 0.5640388689997808,
+      "grad_norm": 1.7248119309244083,
+      "learning_rate": 1e-05,
+      "loss": 1.2533,
+      "step": 965
+    },
+    {
+      "epoch": 0.5646233652370863,
+      "grad_norm": 1.8176328067191592,
+      "learning_rate": 1e-05,
+      "loss": 1.2336,
+      "step": 966
+    },
+    {
+      "epoch": 0.5652078614743917,
+      "grad_norm": 1.756808384474707,
+      "learning_rate": 1e-05,
+      "loss": 1.1961,
+      "step": 967
+    },
+    {
+      "epoch": 0.5657923577116972,
+      "grad_norm": 1.8824786664573994,
+      "learning_rate": 1e-05,
+      "loss": 1.1988,
+      "step": 968
+    },
+    {
+      "epoch": 0.5663768539490027,
+      "grad_norm": 1.9166130540972195,
+      "learning_rate": 1e-05,
+      "loss": 1.2206,
+      "step": 969
+    },
+    {
+      "epoch": 0.5669613501863082,
+      "grad_norm": 1.8571849745607305,
+      "learning_rate": 1e-05,
+      "loss": 1.2776,
+      "step": 970
+    },
+    {
+      "epoch": 0.5675458464236136,
+      "grad_norm": 1.7890888120361557,
+      "learning_rate": 1e-05,
+      "loss": 1.2252,
+      "step": 971
+    },
+    {
+      "epoch": 0.5681303426609191,
+      "grad_norm": 1.839206337630607,
+      "learning_rate": 1e-05,
+      "loss": 1.1633,
+      "step": 972
+    },
+    {
+      "epoch": 0.5687148388982246,
+      "grad_norm": 2.024623225772148,
+      "learning_rate": 1e-05,
+      "loss": 1.2346,
+      "step": 973
+    },
+    {
+      "epoch": 0.5692993351355301,
+      "grad_norm": 1.870235968805569,
+      "learning_rate": 1e-05,
+      "loss": 1.1938,
+      "step": 974
+    },
+    {
+      "epoch": 0.5698838313728355,
+      "grad_norm": 1.882810319646063,
+      "learning_rate": 1e-05,
+      "loss": 1.1919,
+      "step": 975
+    },
+    {
+      "epoch": 0.570468327610141,
+      "grad_norm": 1.7428992921221593,
+      "learning_rate": 1e-05,
+      "loss": 1.1479,
+      "step": 976
+    },
+    {
+      "epoch": 0.5710528238474465,
+      "grad_norm": 1.9353054194246875,
+      "learning_rate": 1e-05,
+      "loss": 1.2291,
+      "step": 977
+    },
+    {
+      "epoch": 0.571637320084752,
+      "grad_norm": 1.9055986280257697,
+      "learning_rate": 1e-05,
+      "loss": 1.2401,
+      "step": 978
+    },
+    {
+      "epoch": 0.5722218163220574,
+      "grad_norm": 1.8410807313708435,
+      "learning_rate": 1e-05,
+      "loss": 1.2106,
+      "step": 979
+    },
+    {
+      "epoch": 0.5728063125593629,
+      "grad_norm": 1.8597416887703642,
+      "learning_rate": 1e-05,
+      "loss": 1.1826,
+      "step": 980
+    },
+    {
+      "epoch": 0.5733908087966684,
+      "grad_norm": 1.8671064291524868,
+      "learning_rate": 1e-05,
+      "loss": 1.2174,
+      "step": 981
+    },
+    {
+      "epoch": 0.5739753050339739,
+      "grad_norm": 1.6871999267593787,
+      "learning_rate": 1e-05,
+      "loss": 1.2674,
+      "step": 982
+    },
+    {
+      "epoch": 0.5745598012712794,
+      "grad_norm": 1.8292757333146792,
+      "learning_rate": 1e-05,
+      "loss": 1.2398,
+      "step": 983
+    },
+    {
+      "epoch": 0.5751442975085848,
+      "grad_norm": 1.7397372144407461,
+      "learning_rate": 1e-05,
+      "loss": 1.1708,
+      "step": 984
+    },
+    {
+      "epoch": 0.5757287937458903,
+      "grad_norm": 1.7099020679641075,
+      "learning_rate": 1e-05,
+      "loss": 1.2136,
+      "step": 985
+    },
+    {
+      "epoch": 0.5763132899831958,
+      "grad_norm": 1.776583504541904,
+      "learning_rate": 1e-05,
+      "loss": 1.2262,
+      "step": 986
+    },
+    {
+      "epoch": 0.5768977862205013,
+      "grad_norm": 1.6699207470561743,
+      "learning_rate": 1e-05,
+      "loss": 1.1978,
+      "step": 987
+    },
+    {
+      "epoch": 0.5774822824578066,
+      "grad_norm": 1.7799329051805064,
+      "learning_rate": 1e-05,
+      "loss": 1.2109,
+      "step": 988
+    },
+    {
+      "epoch": 0.5780667786951121,
+      "grad_norm": 1.7771601046227663,
+      "learning_rate": 1e-05,
+      "loss": 1.2144,
+      "step": 989
+    },
+    {
+      "epoch": 0.5786512749324176,
+      "grad_norm": 1.9566137074065644,
+      "learning_rate": 1e-05,
+      "loss": 1.247,
+      "step": 990
+    },
+    {
+      "epoch": 0.5792357711697231,
+      "grad_norm": 1.863080151692622,
+      "learning_rate": 1e-05,
+      "loss": 1.2578,
+      "step": 991
+    },
+    {
+      "epoch": 0.5798202674070285,
+      "grad_norm": 1.8875576722558056,
+      "learning_rate": 1e-05,
+      "loss": 1.2268,
+      "step": 992
+    },
+    {
+      "epoch": 0.580404763644334,
+      "grad_norm": 1.806488288677423,
+      "learning_rate": 1e-05,
+      "loss": 1.2844,
+      "step": 993
+    },
+    {
+      "epoch": 0.5809892598816395,
+      "grad_norm": 1.6814565283921035,
+      "learning_rate": 1e-05,
+      "loss": 1.2245,
+      "step": 994
+    },
+    {
+      "epoch": 0.581573756118945,
+      "grad_norm": 1.753203134347971,
+      "learning_rate": 1e-05,
+      "loss": 1.2555,
+      "step": 995
+    },
+    {
+      "epoch": 0.5821582523562504,
+      "grad_norm": 1.790883664368853,
+      "learning_rate": 1e-05,
+      "loss": 1.2143,
+      "step": 996
+    },
+    {
+      "epoch": 0.5827427485935559,
+      "grad_norm": 1.8319263041334297,
+      "learning_rate": 1e-05,
+      "loss": 1.2233,
+      "step": 997
+    },
+    {
+      "epoch": 0.5833272448308614,
+      "grad_norm": 1.783345981404674,
+      "learning_rate": 1e-05,
+      "loss": 1.2662,
+      "step": 998
+    },
+    {
+      "epoch": 0.5839117410681669,
+      "grad_norm": 1.974706745177003,
+      "learning_rate": 1e-05,
+      "loss": 1.2149,
+      "step": 999
+    },
+    {
+      "epoch": 0.5844962373054724,
+      "grad_norm": 1.7472151025809393,
+      "learning_rate": 1e-05,
+      "loss": 1.2414,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5850807335427778,
+      "grad_norm": 1.798386339939144,
+      "learning_rate": 1e-05,
+      "loss": 1.2432,
+      "step": 1001
+    },
+    {
+      "epoch": 0.5856652297800833,
+      "grad_norm": 1.6854222385255158,
+      "learning_rate": 1e-05,
+      "loss": 1.2108,
+      "step": 1002
+    },
+    {
+      "epoch": 0.5862497260173888,
+      "grad_norm": 1.8086282988766742,
+      "learning_rate": 1e-05,
+      "loss": 1.2358,
+      "step": 1003
+    },
+    {
+      "epoch": 0.5868342222546943,
+      "grad_norm": 1.79456953205842,
+      "learning_rate": 1e-05,
+      "loss": 1.1815,
+      "step": 1004
+    },
+    {
+      "epoch": 0.5874187184919997,
+      "grad_norm": 1.8444395853711308,
+      "learning_rate": 1e-05,
+      "loss": 1.2382,
+      "step": 1005
+    },
+    {
+      "epoch": 0.5880032147293052,
+      "grad_norm": 1.837698690001994,
+      "learning_rate": 1e-05,
+      "loss": 1.2156,
+      "step": 1006
+    },
+    {
+      "epoch": 0.5885877109666107,
+      "grad_norm": 1.736064328998977,
+      "learning_rate": 1e-05,
+      "loss": 1.1596,
+      "step": 1007
+    },
+    {
+      "epoch": 0.5891722072039162,
+      "grad_norm": 1.793237277517982,
+      "learning_rate": 1e-05,
+      "loss": 1.2369,
+      "step": 1008
+    },
+    {
+      "epoch": 0.5897567034412216,
+      "grad_norm": 2.1144472577744575,
+      "learning_rate": 1e-05,
+      "loss": 1.1598,
+      "step": 1009
+    },
+    {
+      "epoch": 0.590341199678527,
+      "grad_norm": 1.7715262539833763,
+      "learning_rate": 1e-05,
+      "loss": 1.1791,
+      "step": 1010
+    },
+    {
+      "epoch": 0.5909256959158325,
+      "grad_norm": 1.8837183735184628,
+      "learning_rate": 1e-05,
+      "loss": 1.2295,
+      "step": 1011
+    },
+    {
+      "epoch": 0.591510192153138,
+      "grad_norm": 1.8895273969528403,
+      "learning_rate": 1e-05,
+      "loss": 1.2283,
+      "step": 1012
+    },
+    {
+      "epoch": 0.5920946883904434,
+      "grad_norm": 1.6359283347179714,
+      "learning_rate": 1e-05,
+      "loss": 1.1968,
+      "step": 1013
+    },
+    {
+      "epoch": 0.5926791846277489,
+      "grad_norm": 1.7762623321419215,
+      "learning_rate": 1e-05,
+      "loss": 1.248,
+      "step": 1014
+    },
+    {
+      "epoch": 0.5932636808650544,
+      "grad_norm": 1.6667193962991542,
+      "learning_rate": 1e-05,
+      "loss": 1.2387,
+      "step": 1015
+    },
+    {
+      "epoch": 0.5938481771023599,
+      "grad_norm": 1.6407903736650542,
+      "learning_rate": 1e-05,
+      "loss": 1.1899,
+      "step": 1016
+    },
+    {
+      "epoch": 0.5944326733396654,
+      "grad_norm": 1.836372255544499,
+      "learning_rate": 1e-05,
+      "loss": 1.2147,
+      "step": 1017
+    },
+    {
+      "epoch": 0.5950171695769708,
+      "grad_norm": 1.9304808904350268,
+      "learning_rate": 1e-05,
+      "loss": 1.2073,
+      "step": 1018
+    },
+    {
+      "epoch": 0.5956016658142763,
+      "grad_norm": 1.7313233622795774,
+      "learning_rate": 1e-05,
+      "loss": 1.1915,
+      "step": 1019
+    },
+    {
+      "epoch": 0.5961861620515818,
+      "grad_norm": 1.7315292231236843,
+      "learning_rate": 1e-05,
+      "loss": 1.189,
+      "step": 1020
+    },
+    {
+      "epoch": 0.5967706582888873,
+      "grad_norm": 1.9178359049325022,
+      "learning_rate": 1e-05,
+      "loss": 1.2095,
+      "step": 1021
+    },
+    {
+      "epoch": 0.5973551545261927,
+      "grad_norm": 1.835761990740148,
+      "learning_rate": 1e-05,
+      "loss": 1.2136,
+      "step": 1022
+    },
+    {
+      "epoch": 0.5979396507634982,
+      "grad_norm": 1.7291947671134587,
+      "learning_rate": 1e-05,
+      "loss": 1.2676,
+      "step": 1023
+    },
+    {
+      "epoch": 0.5985241470008037,
+      "grad_norm": 1.7552245857948487,
+      "learning_rate": 1e-05,
+      "loss": 1.223,
+      "step": 1024
+    },
+    {
+      "epoch": 0.5991086432381092,
+      "grad_norm": 1.7331740250283552,
+      "learning_rate": 1e-05,
+      "loss": 1.2114,
+      "step": 1025
+    },
+    {
+      "epoch": 0.5996931394754146,
+      "grad_norm": 1.618442440432855,
+      "learning_rate": 1e-05,
+      "loss": 1.1919,
+      "step": 1026
+    },
+    {
+      "epoch": 0.6002776357127201,
+      "grad_norm": 1.6758071493011195,
+      "learning_rate": 1e-05,
+      "loss": 1.1733,
+      "step": 1027
+    },
+    {
+      "epoch": 0.6008621319500256,
+      "grad_norm": 1.75694952391172,
+      "learning_rate": 1e-05,
+      "loss": 1.2295,
+      "step": 1028
+    },
+    {
+      "epoch": 0.6014466281873311,
+      "grad_norm": 1.807074198267752,
+      "learning_rate": 1e-05,
+      "loss": 1.1966,
+      "step": 1029
+    },
+    {
+      "epoch": 0.6020311244246365,
+      "grad_norm": 1.6621139414574793,
+      "learning_rate": 1e-05,
+      "loss": 1.2136,
+      "step": 1030
+    },
+    {
+      "epoch": 0.602615620661942,
+      "grad_norm": 1.8035407260036178,
+      "learning_rate": 1e-05,
+      "loss": 1.2342,
+      "step": 1031
+    },
+    {
+      "epoch": 0.6032001168992475,
+      "grad_norm": 1.742933356973481,
+      "learning_rate": 1e-05,
+      "loss": 1.216,
+      "step": 1032
+    },
+    {
+      "epoch": 0.603784613136553,
+      "grad_norm": 1.6795772158560993,
+      "learning_rate": 1e-05,
+      "loss": 1.234,
+      "step": 1033
+    },
+    {
+      "epoch": 0.6043691093738585,
+      "grad_norm": 1.8279919955707504,
+      "learning_rate": 1e-05,
+      "loss": 1.2592,
+      "step": 1034
+    },
+    {
+      "epoch": 0.6049536056111638,
+      "grad_norm": 1.886031863194461,
+      "learning_rate": 1e-05,
+      "loss": 1.1769,
+      "step": 1035
+    },
+    {
+      "epoch": 0.6055381018484693,
+      "grad_norm": 1.6220860102064105,
+      "learning_rate": 1e-05,
+      "loss": 1.206,
+      "step": 1036
+    },
+    {
+      "epoch": 0.6061225980857748,
+      "grad_norm": 1.8952114233856132,
+      "learning_rate": 1e-05,
+      "loss": 1.1799,
+      "step": 1037
+    },
+    {
+      "epoch": 0.6067070943230803,
+      "grad_norm": 1.8384235405172846,
+      "learning_rate": 1e-05,
+      "loss": 1.1838,
+      "step": 1038
+    },
+    {
+      "epoch": 0.6072915905603857,
+      "grad_norm": 1.746760550223795,
+      "learning_rate": 1e-05,
+      "loss": 1.1562,
+      "step": 1039
+    },
+    {
+      "epoch": 0.6078760867976912,
+      "grad_norm": 1.9540488312353554,
+      "learning_rate": 1e-05,
+      "loss": 1.2024,
+      "step": 1040
+    },
+    {
+      "epoch": 0.6084605830349967,
+      "grad_norm": 1.755211707390591,
+      "learning_rate": 1e-05,
+      "loss": 1.2054,
+      "step": 1041
+    },
+    {
+      "epoch": 0.6090450792723022,
+      "grad_norm": 1.782848177126256,
+      "learning_rate": 1e-05,
+      "loss": 1.1973,
+      "step": 1042
+    },
+    {
+      "epoch": 0.6096295755096076,
+      "grad_norm": 1.8891728370029055,
+      "learning_rate": 1e-05,
+      "loss": 1.2199,
+      "step": 1043
+    },
+    {
+      "epoch": 0.6102140717469131,
+      "grad_norm": 1.842413847600614,
+      "learning_rate": 1e-05,
+      "loss": 1.1932,
+      "step": 1044
+    },
+    {
+      "epoch": 0.6107985679842186,
+      "grad_norm": 1.7356260837645192,
+      "learning_rate": 1e-05,
+      "loss": 1.2555,
+      "step": 1045
+    },
+    {
+      "epoch": 0.6113830642215241,
+      "grad_norm": 1.8194309063946728,
+      "learning_rate": 1e-05,
+      "loss": 1.2079,
+      "step": 1046
+    },
+    {
+      "epoch": 0.6119675604588295,
+      "grad_norm": 1.7912957473769857,
+      "learning_rate": 1e-05,
+      "loss": 1.2115,
+      "step": 1047
+    },
+    {
+      "epoch": 0.612552056696135,
+      "grad_norm": 1.7059722590052164,
+      "learning_rate": 1e-05,
+      "loss": 1.2405,
+      "step": 1048
+    },
+    {
+      "epoch": 0.6131365529334405,
+      "grad_norm": 1.637905586711982,
+      "learning_rate": 1e-05,
+      "loss": 1.1689,
+      "step": 1049
+    },
+    {
+      "epoch": 0.613721049170746,
+      "grad_norm": 1.7239511224253765,
+      "learning_rate": 1e-05,
+      "loss": 1.2164,
+      "step": 1050
+    },
+    {
+      "epoch": 0.6143055454080515,
+      "grad_norm": 1.994320201627745,
+      "learning_rate": 1e-05,
+      "loss": 1.2482,
+      "step": 1051
+    },
+    {
+      "epoch": 0.6148900416453569,
+      "grad_norm": 1.969360474198128,
+      "learning_rate": 1e-05,
+      "loss": 1.2187,
+      "step": 1052
+    },
+    {
+      "epoch": 0.6154745378826624,
+      "grad_norm": 1.8509725016077694,
+      "learning_rate": 1e-05,
+      "loss": 1.2466,
+      "step": 1053
+    },
+    {
+      "epoch": 0.6160590341199679,
+      "grad_norm": 1.787353098097567,
+      "learning_rate": 1e-05,
+      "loss": 1.2107,
+      "step": 1054
+    },
+    {
+      "epoch": 0.6166435303572734,
+      "grad_norm": 1.8558317585558846,
+      "learning_rate": 1e-05,
+      "loss": 1.2272,
+      "step": 1055
+    },
+    {
+      "epoch": 0.6172280265945788,
+      "grad_norm": 1.8172309906021091,
+      "learning_rate": 1e-05,
+      "loss": 1.2169,
+      "step": 1056
+    },
+    {
+      "epoch": 0.6178125228318843,
+      "grad_norm": 1.7396551573057206,
+      "learning_rate": 1e-05,
+      "loss": 1.2304,
+      "step": 1057
+    },
+    {
+      "epoch": 0.6183970190691898,
+      "grad_norm": 1.6718685274305252,
+      "learning_rate": 1e-05,
+      "loss": 1.2151,
+      "step": 1058
+    },
+    {
+      "epoch": 0.6189815153064953,
+      "grad_norm": 1.7942900348413684,
+      "learning_rate": 1e-05,
+      "loss": 1.2148,
+      "step": 1059
+    },
+    {
+      "epoch": 0.6195660115438006,
+      "grad_norm": 2.01531040018434,
+      "learning_rate": 1e-05,
+      "loss": 1.2337,
+      "step": 1060
+    },
+    {
+      "epoch": 0.6201505077811061,
+      "grad_norm": 1.8142099681093193,
+      "learning_rate": 1e-05,
+      "loss": 1.2069,
+      "step": 1061
+    },
+    {
+      "epoch": 0.6207350040184116,
+      "grad_norm": 1.6909262760365051,
+      "learning_rate": 1e-05,
+      "loss": 1.2031,
+      "step": 1062
+    },
+    {
+      "epoch": 0.6213195002557171,
+      "grad_norm": 1.7763296553454597,
+      "learning_rate": 1e-05,
+      "loss": 1.1487,
+      "step": 1063
+    },
+    {
+      "epoch": 0.6219039964930225,
+      "grad_norm": 1.6402658599110567,
+      "learning_rate": 1e-05,
+      "loss": 1.1851,
+      "step": 1064
+    },
+    {
+      "epoch": 0.622488492730328,
+      "grad_norm": 1.758562320413981,
+      "learning_rate": 1e-05,
+      "loss": 1.2233,
+      "step": 1065
+    },
+    {
+      "epoch": 0.6230729889676335,
+      "grad_norm": 1.7977562976046755,
+      "learning_rate": 1e-05,
+      "loss": 1.1844,
+      "step": 1066
+    },
+    {
+      "epoch": 0.623657485204939,
+      "grad_norm": 2.023526845939143,
+      "learning_rate": 1e-05,
+      "loss": 1.1589,
+      "step": 1067
+    },
+    {
+      "epoch": 0.6242419814422445,
+      "grad_norm": 1.5552841892142317,
+      "learning_rate": 1e-05,
+      "loss": 1.2279,
+      "step": 1068
+    },
+    {
+      "epoch": 0.6248264776795499,
+      "grad_norm": 1.9821831022312288,
+      "learning_rate": 1e-05,
+      "loss": 1.1733,
+      "step": 1069
+    },
+    {
+      "epoch": 0.6254109739168554,
+      "grad_norm": 1.735942832059775,
+      "learning_rate": 1e-05,
+      "loss": 1.2219,
+      "step": 1070
+    },
+    {
+      "epoch": 0.6259954701541609,
+      "grad_norm": 1.8715786335989917,
+      "learning_rate": 1e-05,
+      "loss": 1.2586,
+      "step": 1071
+    },
+    {
+      "epoch": 0.6265799663914664,
+      "grad_norm": 1.6789552063627022,
+      "learning_rate": 1e-05,
+      "loss": 1.2316,
+      "step": 1072
+    },
+    {
+      "epoch": 0.6271644626287718,
+      "grad_norm": 1.833203783685044,
+      "learning_rate": 1e-05,
+      "loss": 1.1559,
+      "step": 1073
+    },
+    {
+      "epoch": 0.6277489588660773,
+      "grad_norm": 1.9579930515803716,
+      "learning_rate": 1e-05,
+      "loss": 1.1499,
+      "step": 1074
+    },
+    {
+      "epoch": 0.6283334551033828,
+      "grad_norm": 1.7531227472515156,
+      "learning_rate": 1e-05,
+      "loss": 1.2282,
+      "step": 1075
+    },
+    {
+      "epoch": 0.6289179513406883,
+      "grad_norm": 1.9140440726815906,
+      "learning_rate": 1e-05,
+      "loss": 1.1932,
+      "step": 1076
+    },
+    {
+      "epoch": 0.6295024475779937,
+      "grad_norm": 1.821090860497006,
+      "learning_rate": 1e-05,
+      "loss": 1.2286,
+      "step": 1077
+    },
+    {
+      "epoch": 0.6300869438152992,
+      "grad_norm": 1.9811709504195893,
+      "learning_rate": 1e-05,
+      "loss": 1.2298,
+      "step": 1078
+    },
+    {
+      "epoch": 0.6306714400526047,
+      "grad_norm": 1.8198380124506728,
+      "learning_rate": 1e-05,
+      "loss": 1.2466,
+      "step": 1079
+    },
+    {
+      "epoch": 0.6312559362899102,
+      "grad_norm": 1.8852076459562477,
+      "learning_rate": 1e-05,
+      "loss": 1.2118,
+      "step": 1080
+    },
+    {
+      "epoch": 0.6318404325272156,
+      "grad_norm": 1.8307350058447143,
+      "learning_rate": 1e-05,
+      "loss": 1.2385,
+      "step": 1081
+    },
+    {
+      "epoch": 0.632424928764521,
+      "grad_norm": 1.8330280736939457,
+      "learning_rate": 1e-05,
+      "loss": 1.2624,
+      "step": 1082
+    },
+    {
+      "epoch": 0.6330094250018266,
+      "grad_norm": 1.772632012181666,
+      "learning_rate": 1e-05,
+      "loss": 1.1862,
+      "step": 1083
+    },
+    {
+      "epoch": 0.633593921239132,
+      "grad_norm": 1.6596790042462695,
+      "learning_rate": 1e-05,
+      "loss": 1.1678,
+      "step": 1084
+    },
+    {
+      "epoch": 0.6341784174764376,
+      "grad_norm": 1.8868942081649265,
+      "learning_rate": 1e-05,
+      "loss": 1.1766,
+      "step": 1085
+    },
+    {
+      "epoch": 0.6347629137137429,
+      "grad_norm": 1.8892384640907125,
+      "learning_rate": 1e-05,
+      "loss": 1.249,
+      "step": 1086
+    },
+    {
+      "epoch": 0.6353474099510484,
+      "grad_norm": 1.8544625697685082,
+      "learning_rate": 1e-05,
+      "loss": 1.1607,
+      "step": 1087
+    },
+    {
+      "epoch": 0.6359319061883539,
+      "grad_norm": 1.761321518862312,
+      "learning_rate": 1e-05,
+      "loss": 1.2196,
+      "step": 1088
+    },
+    {
+      "epoch": 0.6365164024256594,
+      "grad_norm": 1.840265611568339,
+      "learning_rate": 1e-05,
+      "loss": 1.1743,
+      "step": 1089
+    },
+    {
+      "epoch": 0.6371008986629648,
+      "grad_norm": 1.5713961560625507,
+      "learning_rate": 1e-05,
+      "loss": 1.2092,
+      "step": 1090
+    },
+    {
+      "epoch": 0.6376853949002703,
+      "grad_norm": 1.767621027402041,
+      "learning_rate": 1e-05,
+      "loss": 1.1978,
+      "step": 1091
+    },
+    {
+      "epoch": 0.6382698911375758,
+      "grad_norm": 1.8297235560490954,
+      "learning_rate": 1e-05,
+      "loss": 1.2122,
+      "step": 1092
+    },
+    {
+      "epoch": 0.6388543873748813,
+      "grad_norm": 1.917055869977204,
+      "learning_rate": 1e-05,
+      "loss": 1.1865,
+      "step": 1093
+    },
+    {
+      "epoch": 0.6394388836121867,
+      "grad_norm": 1.7763383229784637,
+      "learning_rate": 1e-05,
+      "loss": 1.177,
+      "step": 1094
+    },
+    {
+      "epoch": 0.6400233798494922,
+      "grad_norm": 1.7923215637739152,
+      "learning_rate": 1e-05,
+      "loss": 1.2208,
+      "step": 1095
+    },
+    {
+      "epoch": 0.6406078760867977,
+      "grad_norm": 1.8192861012029193,
+      "learning_rate": 1e-05,
+      "loss": 1.2408,
+      "step": 1096
+    },
+    {
+      "epoch": 0.6411923723241032,
+      "grad_norm": 1.8676454095330872,
+      "learning_rate": 1e-05,
+      "loss": 1.2206,
+      "step": 1097
+    },
+    {
+      "epoch": 0.6417768685614086,
+      "grad_norm": 1.7855361447374491,
+      "learning_rate": 1e-05,
+      "loss": 1.2492,
+      "step": 1098
+    },
+    {
+      "epoch": 0.6423613647987141,
+      "grad_norm": 1.7844996837822784,
+      "learning_rate": 1e-05,
+      "loss": 1.2089,
+      "step": 1099
+    },
+    {
+      "epoch": 0.6429458610360196,
+      "grad_norm": 1.8143013212107542,
+      "learning_rate": 1e-05,
+      "loss": 1.1897,
+      "step": 1100
+    },
+    {
+      "epoch": 0.6435303572733251,
+      "grad_norm": 1.7573894249913986,
+      "learning_rate": 1e-05,
+      "loss": 1.1878,
+      "step": 1101
+    },
+    {
+      "epoch": 0.6441148535106306,
+      "grad_norm": 1.8297951913521566,
+      "learning_rate": 1e-05,
+      "loss": 1.1812,
+      "step": 1102
+    },
+    {
+      "epoch": 0.644699349747936,
+      "grad_norm": 1.7582414485452247,
+      "learning_rate": 1e-05,
+      "loss": 1.2063,
+      "step": 1103
+    },
+    {
+      "epoch": 0.6452838459852415,
+      "grad_norm": 1.7742915703287798,
+      "learning_rate": 1e-05,
+      "loss": 1.2136,
+      "step": 1104
+    },
+    {
+      "epoch": 0.645868342222547,
+      "grad_norm": 1.7100192684464308,
+      "learning_rate": 1e-05,
+      "loss": 1.218,
+      "step": 1105
+    },
+    {
+      "epoch": 0.6464528384598525,
+      "grad_norm": 1.9224375009538148,
+      "learning_rate": 1e-05,
+      "loss": 1.1959,
+      "step": 1106
+    },
+    {
+      "epoch": 0.6470373346971579,
+      "grad_norm": 1.7860826966397658,
+      "learning_rate": 1e-05,
+      "loss": 1.2567,
+      "step": 1107
+    },
+    {
+      "epoch": 0.6476218309344633,
+      "grad_norm": 1.9802396241144813,
+      "learning_rate": 1e-05,
+      "loss": 1.2503,
+      "step": 1108
+    },
+    {
+      "epoch": 0.6482063271717688,
+      "grad_norm": 1.8405678613789989,
+      "learning_rate": 1e-05,
+      "loss": 1.1895,
+      "step": 1109
+    },
+    {
+      "epoch": 0.6487908234090743,
+      "grad_norm": 1.7677935152448914,
+      "learning_rate": 1e-05,
+      "loss": 1.1823,
+      "step": 1110
+    },
+    {
+      "epoch": 0.6493753196463797,
+      "grad_norm": 1.812675049745238,
+      "learning_rate": 1e-05,
+      "loss": 1.2287,
+      "step": 1111
+    },
+    {
+      "epoch": 0.6499598158836852,
+      "grad_norm": 1.9260950924255869,
+      "learning_rate": 1e-05,
+      "loss": 1.2497,
+      "step": 1112
+    },
+    {
+      "epoch": 0.6505443121209907,
+      "grad_norm": 1.813068463040134,
+      "learning_rate": 1e-05,
+      "loss": 1.2085,
+      "step": 1113
+    },
+    {
+      "epoch": 0.6511288083582962,
+      "grad_norm": 1.7159128143916689,
+      "learning_rate": 1e-05,
+      "loss": 1.1933,
+      "step": 1114
+    },
+    {
+      "epoch": 0.6517133045956016,
+      "grad_norm": 1.7590171486351085,
+      "learning_rate": 1e-05,
+      "loss": 1.1761,
+      "step": 1115
+    },
+    {
+      "epoch": 0.6522978008329071,
+      "grad_norm": 1.7709525727468818,
+      "learning_rate": 1e-05,
+      "loss": 1.1838,
+      "step": 1116
+    },
+    {
+      "epoch": 0.6528822970702126,
+      "grad_norm": 1.704962862303646,
+      "learning_rate": 1e-05,
+      "loss": 1.1433,
+      "step": 1117
+    },
+    {
+      "epoch": 0.6534667933075181,
+      "grad_norm": 1.7039138693927995,
+      "learning_rate": 1e-05,
+      "loss": 1.1823,
+      "step": 1118
+    },
+    {
+      "epoch": 0.6540512895448235,
+      "grad_norm": 2.1552233326562997,
+      "learning_rate": 1e-05,
+      "loss": 1.2273,
+      "step": 1119
+    },
+    {
+      "epoch": 0.654635785782129,
+      "grad_norm": 1.7151026380297445,
+      "learning_rate": 1e-05,
+      "loss": 1.2409,
+      "step": 1120
+    },
+    {
+      "epoch": 0.6552202820194345,
+      "grad_norm": 1.6834871990665032,
+      "learning_rate": 1e-05,
+      "loss": 1.2294,
+      "step": 1121
+    },
+    {
+      "epoch": 0.65580477825674,
+      "grad_norm": 1.7262455130016736,
+      "learning_rate": 1e-05,
+      "loss": 1.1876,
+      "step": 1122
+    },
+    {
+      "epoch": 0.6563892744940455,
+      "grad_norm": 1.7506704410058076,
+      "learning_rate": 1e-05,
+      "loss": 1.2018,
+      "step": 1123
+    },
+    {
+      "epoch": 0.6569737707313509,
+      "grad_norm": 1.6725951221392303,
+      "learning_rate": 1e-05,
+      "loss": 1.205,
+      "step": 1124
+    },
+    {
+      "epoch": 0.6575582669686564,
+      "grad_norm": 1.81951316402501,
+      "learning_rate": 1e-05,
+      "loss": 1.2441,
+      "step": 1125
+    },
+    {
+      "epoch": 0.6581427632059619,
+      "grad_norm": 1.8040927726554423,
+      "learning_rate": 1e-05,
+      "loss": 1.2579,
+      "step": 1126
+    },
+    {
+      "epoch": 0.6587272594432674,
+      "grad_norm": 1.9282113964958156,
+      "learning_rate": 1e-05,
+      "loss": 1.2921,
+      "step": 1127
+    },
+    {
+      "epoch": 0.6593117556805728,
+      "grad_norm": 1.6361409410775598,
+      "learning_rate": 1e-05,
+      "loss": 1.2424,
+      "step": 1128
+    },
+    {
+      "epoch": 0.6598962519178783,
+      "grad_norm": 1.8164366277517454,
+      "learning_rate": 1e-05,
+      "loss": 1.2023,
+      "step": 1129
+    },
+    {
+      "epoch": 0.6604807481551838,
+      "grad_norm": 2.0053450769498697,
+      "learning_rate": 1e-05,
+      "loss": 1.2203,
+      "step": 1130
+    },
+    {
+      "epoch": 0.6610652443924893,
+      "grad_norm": 1.7473423788286364,
+      "learning_rate": 1e-05,
+      "loss": 1.2142,
+      "step": 1131
+    },
+    {
+      "epoch": 0.6616497406297946,
+      "grad_norm": 1.7220772341276898,
+      "learning_rate": 1e-05,
+      "loss": 1.223,
+      "step": 1132
+    },
+    {
+      "epoch": 0.6622342368671001,
+      "grad_norm": 1.6681543117501672,
+      "learning_rate": 1e-05,
+      "loss": 1.227,
+      "step": 1133
+    },
+    {
+      "epoch": 0.6628187331044056,
+      "grad_norm": 1.9948363548902652,
+      "learning_rate": 1e-05,
+      "loss": 1.2276,
+      "step": 1134
+    },
+    {
+      "epoch": 0.6634032293417111,
+      "grad_norm": 1.7652959456116568,
+      "learning_rate": 1e-05,
+      "loss": 1.2115,
+      "step": 1135
+    },
+    {
+      "epoch": 0.6639877255790165,
+      "grad_norm": 1.7685916070049799,
+      "learning_rate": 1e-05,
+      "loss": 1.1544,
+      "step": 1136
+    },
+    {
+      "epoch": 0.664572221816322,
+      "grad_norm": 1.6984148541688373,
+      "learning_rate": 1e-05,
+      "loss": 1.2087,
+      "step": 1137
+    },
+    {
+      "epoch": 0.6651567180536275,
+      "grad_norm": 1.8205056013138963,
+      "learning_rate": 1e-05,
+      "loss": 1.2282,
+      "step": 1138
+    },
+    {
+      "epoch": 0.665741214290933,
+      "grad_norm": 1.6952404544338657,
+      "learning_rate": 1e-05,
+      "loss": 1.2119,
+      "step": 1139
+    },
+    {
+      "epoch": 0.6663257105282385,
+      "grad_norm": 1.8269656849982714,
+      "learning_rate": 1e-05,
+      "loss": 1.2327,
+      "step": 1140
+    },
+    {
+      "epoch": 0.6669102067655439,
+      "grad_norm": 1.7837755585206923,
+      "learning_rate": 1e-05,
+      "loss": 1.2295,
+      "step": 1141
+    },
+    {
+      "epoch": 0.6674947030028494,
+      "grad_norm": 1.7798495808217498,
+      "learning_rate": 1e-05,
+      "loss": 1.1694,
+      "step": 1142
+    },
+    {
+      "epoch": 0.6680791992401549,
+      "grad_norm": 1.868182438288904,
+      "learning_rate": 1e-05,
+      "loss": 1.226,
+      "step": 1143
+    },
+    {
+      "epoch": 0.6686636954774604,
+      "grad_norm": 1.6518873418426383,
+      "learning_rate": 1e-05,
+      "loss": 1.2067,
+      "step": 1144
+    },
+    {
+      "epoch": 0.6692481917147658,
+      "grad_norm": 1.7692891175964465,
+      "learning_rate": 1e-05,
+      "loss": 1.2213,
+      "step": 1145
+    },
+    {
+      "epoch": 0.6698326879520713,
+      "grad_norm": 1.7657247286376456,
+      "learning_rate": 1e-05,
+      "loss": 1.219,
+      "step": 1146
+    },
+    {
+      "epoch": 0.6704171841893768,
+      "grad_norm": 1.705486589316931,
+      "learning_rate": 1e-05,
+      "loss": 1.2366,
+      "step": 1147
+    },
+    {
+      "epoch": 0.6710016804266823,
+      "grad_norm": 1.8138405826156871,
+      "learning_rate": 1e-05,
+      "loss": 1.2095,
+      "step": 1148
+    },
+    {
+      "epoch": 0.6715861766639877,
+      "grad_norm": 1.815193930471276,
+      "learning_rate": 1e-05,
+      "loss": 1.156,
+      "step": 1149
+    },
+    {
+      "epoch": 0.6721706729012932,
+      "grad_norm": 1.8270551321884119,
+      "learning_rate": 1e-05,
+      "loss": 1.2361,
+      "step": 1150
+    },
+    {
+      "epoch": 0.6727551691385987,
+      "grad_norm": 1.8110668506052845,
+      "learning_rate": 1e-05,
+      "loss": 1.2207,
+      "step": 1151
+    },
+    {
+      "epoch": 0.6733396653759042,
+      "grad_norm": 1.7087693593326045,
+      "learning_rate": 1e-05,
+      "loss": 1.1873,
+      "step": 1152
+    },
+    {
+      "epoch": 0.6739241616132096,
+      "grad_norm": 1.7248427264073156,
+      "learning_rate": 1e-05,
+      "loss": 1.1919,
+      "step": 1153
+    },
+    {
+      "epoch": 0.6745086578505151,
+      "grad_norm": 1.8291044244770045,
+      "learning_rate": 1e-05,
+      "loss": 1.2109,
+      "step": 1154
+    },
+    {
+      "epoch": 0.6750931540878206,
+      "grad_norm": 1.73035071263532,
+      "learning_rate": 1e-05,
+      "loss": 1.1926,
+      "step": 1155
+    },
+    {
+      "epoch": 0.6756776503251261,
+      "grad_norm": 1.774539386494213,
+      "learning_rate": 1e-05,
+      "loss": 1.222,
+      "step": 1156
+    },
+    {
+      "epoch": 0.6762621465624316,
+      "grad_norm": 1.8248546275991542,
+      "learning_rate": 1e-05,
+      "loss": 1.19,
+      "step": 1157
+    },
+    {
+      "epoch": 0.6768466427997369,
+      "grad_norm": 1.9006035550437954,
+      "learning_rate": 1e-05,
+      "loss": 1.1768,
+      "step": 1158
+    },
+    {
+      "epoch": 0.6774311390370424,
+      "grad_norm": 1.751697841455201,
+      "learning_rate": 1e-05,
+      "loss": 1.1879,
+      "step": 1159
+    },
+    {
+      "epoch": 0.6780156352743479,
+      "grad_norm": 1.741820107273918,
+      "learning_rate": 1e-05,
+      "loss": 1.2717,
+      "step": 1160
+    },
+    {
+      "epoch": 0.6786001315116534,
+      "grad_norm": 1.7451926991432425,
+      "learning_rate": 1e-05,
+      "loss": 1.2016,
+      "step": 1161
+    },
+    {
+      "epoch": 0.6791846277489588,
+      "grad_norm": 1.7205587777076001,
+      "learning_rate": 1e-05,
+      "loss": 1.2182,
+      "step": 1162
+    },
+    {
+      "epoch": 0.6797691239862643,
+      "grad_norm": 1.7276879534031306,
+      "learning_rate": 1e-05,
+      "loss": 1.237,
+      "step": 1163
+    },
+    {
+      "epoch": 0.6803536202235698,
+      "grad_norm": 1.7073488830677723,
+      "learning_rate": 1e-05,
+      "loss": 1.2115,
+      "step": 1164
+    },
+    {
+      "epoch": 0.6809381164608753,
+      "grad_norm": 1.8924923650219723,
+      "learning_rate": 1e-05,
+      "loss": 1.2039,
+      "step": 1165
+    },
+    {
+      "epoch": 0.6815226126981807,
+      "grad_norm": 1.8134061909713377,
+      "learning_rate": 1e-05,
+      "loss": 1.1656,
+      "step": 1166
+    },
+    {
+      "epoch": 0.6821071089354862,
+      "grad_norm": 1.8066973872243488,
+      "learning_rate": 1e-05,
+      "loss": 1.2025,
+      "step": 1167
+    },
+    {
+      "epoch": 0.6826916051727917,
+      "grad_norm": 1.8939543701439574,
+      "learning_rate": 1e-05,
+      "loss": 1.1825,
+      "step": 1168
+    },
+    {
+      "epoch": 0.6832761014100972,
+      "grad_norm": 1.7377650149582842,
+      "learning_rate": 1e-05,
+      "loss": 1.2167,
+      "step": 1169
+    },
+    {
+      "epoch": 0.6838605976474026,
+      "grad_norm": 1.5731144234045429,
+      "learning_rate": 1e-05,
+      "loss": 1.1957,
+      "step": 1170
+    },
+    {
+      "epoch": 0.6844450938847081,
+      "grad_norm": 1.8260979628843774,
+      "learning_rate": 1e-05,
+      "loss": 1.1686,
+      "step": 1171
+    },
+    {
+      "epoch": 0.6850295901220136,
+      "grad_norm": 1.8728366721387752,
+      "learning_rate": 1e-05,
+      "loss": 1.2059,
+      "step": 1172
+    },
+    {
+      "epoch": 0.6856140863593191,
+      "grad_norm": 1.9565359662287622,
+      "learning_rate": 1e-05,
+      "loss": 1.2088,
+      "step": 1173
+    },
+    {
+      "epoch": 0.6861985825966246,
+      "grad_norm": 1.703347388331172,
+      "learning_rate": 1e-05,
+      "loss": 1.2411,
+      "step": 1174
+    },
+    {
+      "epoch": 0.68678307883393,
+      "grad_norm": 1.8333358209332369,
+      "learning_rate": 1e-05,
+      "loss": 1.2081,
+      "step": 1175
+    },
+    {
+      "epoch": 0.6873675750712355,
+      "grad_norm": 1.991859223482263,
+      "learning_rate": 1e-05,
+      "loss": 1.2183,
+      "step": 1176
+    },
+    {
+      "epoch": 0.687952071308541,
+      "grad_norm": 1.8572822720038273,
+      "learning_rate": 1e-05,
+      "loss": 1.1694,
+      "step": 1177
+    },
+    {
+      "epoch": 0.6885365675458465,
+      "grad_norm": 1.8867264562144144,
+      "learning_rate": 1e-05,
+      "loss": 1.2458,
+      "step": 1178
+    },
+    {
+      "epoch": 0.6891210637831519,
+      "grad_norm": 1.7465603930595364,
+      "learning_rate": 1e-05,
+      "loss": 1.2226,
+      "step": 1179
+    },
+    {
+      "epoch": 0.6897055600204574,
+      "grad_norm": 1.65386699496299,
+      "learning_rate": 1e-05,
+      "loss": 1.1906,
+      "step": 1180
+    },
+    {
+      "epoch": 0.6902900562577629,
+      "grad_norm": 1.7465721595462453,
+      "learning_rate": 1e-05,
+      "loss": 1.1973,
+      "step": 1181
+    },
+    {
+      "epoch": 0.6908745524950684,
+      "grad_norm": 1.6766354151250291,
+      "learning_rate": 1e-05,
+      "loss": 1.2441,
+      "step": 1182
+    },
+    {
+      "epoch": 0.6914590487323737,
+      "grad_norm": 1.7195514816556186,
+      "learning_rate": 1e-05,
+      "loss": 1.1778,
+      "step": 1183
+    },
+    {
+      "epoch": 0.6920435449696792,
+      "grad_norm": 1.6881620075768884,
+      "learning_rate": 1e-05,
+      "loss": 1.2271,
+      "step": 1184
+    },
+    {
+      "epoch": 0.6926280412069847,
+      "grad_norm": 1.7861818575103325,
+      "learning_rate": 1e-05,
+      "loss": 1.2192,
+      "step": 1185
+    },
+    {
+      "epoch": 0.6932125374442902,
+      "grad_norm": 1.9263250540969543,
+      "learning_rate": 1e-05,
+      "loss": 1.2114,
+      "step": 1186
+    },
+    {
+      "epoch": 0.6937970336815956,
+      "grad_norm": 1.7742300674045546,
+      "learning_rate": 1e-05,
+      "loss": 1.2178,
+      "step": 1187
+    },
+    {
+      "epoch": 0.6943815299189011,
+      "grad_norm": 1.7089168377501898,
+      "learning_rate": 1e-05,
+      "loss": 1.2316,
+      "step": 1188
+    },
+    {
+      "epoch": 0.6949660261562066,
+      "grad_norm": 1.923641131843193,
+      "learning_rate": 1e-05,
+      "loss": 1.1925,
+      "step": 1189
+    },
+    {
+      "epoch": 0.6955505223935121,
+      "grad_norm": 2.0742008760724415,
+      "learning_rate": 1e-05,
+      "loss": 1.229,
+      "step": 1190
+    },
+    {
+      "epoch": 0.6961350186308176,
+      "grad_norm": 1.694314733891756,
+      "learning_rate": 1e-05,
+      "loss": 1.1777,
+      "step": 1191
+    },
+    {
+      "epoch": 0.696719514868123,
+      "grad_norm": 1.7854399903861369,
+      "learning_rate": 1e-05,
+      "loss": 1.1509,
+      "step": 1192
+    },
+    {
+      "epoch": 0.6973040111054285,
+      "grad_norm": 1.6878099520707779,
+      "learning_rate": 1e-05,
+      "loss": 1.2174,
+      "step": 1193
+    },
+    {
+      "epoch": 0.697888507342734,
+      "grad_norm": 1.797106370003288,
+      "learning_rate": 1e-05,
+      "loss": 1.2013,
+      "step": 1194
+    },
+    {
+      "epoch": 0.6984730035800395,
+      "grad_norm": 1.6348604179257666,
+      "learning_rate": 1e-05,
+      "loss": 1.2623,
+      "step": 1195
+    },
+    {
+      "epoch": 0.6990574998173449,
+      "grad_norm": 1.7986745968056432,
+      "learning_rate": 1e-05,
+      "loss": 1.1855,
+      "step": 1196
+    },
+    {
+      "epoch": 0.6996419960546504,
+      "grad_norm": 1.9016875728475235,
+      "learning_rate": 1e-05,
+      "loss": 1.2266,
+      "step": 1197
+    },
+    {
+      "epoch": 0.7002264922919559,
+      "grad_norm": 1.7546757900422876,
+      "learning_rate": 1e-05,
+      "loss": 1.1789,
+      "step": 1198
+    },
+    {
+      "epoch": 0.7008109885292614,
+      "grad_norm": 1.651360675579258,
+      "learning_rate": 1e-05,
+      "loss": 1.1108,
+      "step": 1199
+    },
+    {
+      "epoch": 0.7013954847665668,
+      "grad_norm": 1.9037269126937781,
+      "learning_rate": 1e-05,
+      "loss": 1.1935,
+      "step": 1200
+    },
+    {
+      "epoch": 0.7019799810038723,
+      "grad_norm": 1.7679617015779632,
+      "learning_rate": 1e-05,
+      "loss": 1.2312,
+      "step": 1201
+    },
+    {
+      "epoch": 0.7025644772411778,
+      "grad_norm": 1.5959896092536294,
+      "learning_rate": 1e-05,
+      "loss": 1.2283,
+      "step": 1202
+    },
+    {
+      "epoch": 0.7031489734784833,
+      "grad_norm": 1.6502251458737838,
+      "learning_rate": 1e-05,
+      "loss": 1.1741,
+      "step": 1203
+    },
+    {
+      "epoch": 0.7037334697157887,
+      "grad_norm": 1.747351292950232,
+      "learning_rate": 1e-05,
+      "loss": 1.2361,
+      "step": 1204
+    },
+    {
+      "epoch": 0.7043179659530941,
+      "grad_norm": 1.8071310367691027,
+      "learning_rate": 1e-05,
+      "loss": 1.2317,
+      "step": 1205
+    },
+    {
+      "epoch": 0.7049024621903996,
+      "grad_norm": 1.8571177039391735,
+      "learning_rate": 1e-05,
+      "loss": 1.1919,
+      "step": 1206
+    },
+    {
+      "epoch": 0.7054869584277051,
+      "grad_norm": 1.8566204086615834,
+      "learning_rate": 1e-05,
+      "loss": 1.2134,
+      "step": 1207
+    },
+    {
+      "epoch": 0.7060714546650106,
+      "grad_norm": 1.6905178715659543,
+      "learning_rate": 1e-05,
+      "loss": 1.2375,
+      "step": 1208
+    },
+    {
+      "epoch": 0.706655950902316,
+      "grad_norm": 1.7718288992224764,
+      "learning_rate": 1e-05,
+      "loss": 1.2516,
+      "step": 1209
+    },
+    {
+      "epoch": 0.7072404471396215,
+      "grad_norm": 1.9259671727927958,
+      "learning_rate": 1e-05,
+      "loss": 1.1898,
+      "step": 1210
+    },
+    {
+      "epoch": 0.707824943376927,
+      "grad_norm": 1.7288657441640312,
+      "learning_rate": 1e-05,
+      "loss": 1.2299,
+      "step": 1211
+    },
+    {
+      "epoch": 0.7084094396142325,
+      "grad_norm": 1.6185171831052487,
+      "learning_rate": 1e-05,
+      "loss": 1.2297,
+      "step": 1212
+    },
+    {
+      "epoch": 0.7089939358515379,
+      "grad_norm": 1.6932515604867702,
+      "learning_rate": 1e-05,
+      "loss": 1.145,
+      "step": 1213
+    },
+    {
+      "epoch": 0.7095784320888434,
+      "grad_norm": 1.8238152777228331,
+      "learning_rate": 1e-05,
+      "loss": 1.2425,
+      "step": 1214
+    },
+    {
+      "epoch": 0.7101629283261489,
+      "grad_norm": 1.9068980592067355,
+      "learning_rate": 1e-05,
+      "loss": 1.271,
+      "step": 1215
+    },
+    {
+      "epoch": 0.7107474245634544,
+      "grad_norm": 1.7156369985929358,
+      "learning_rate": 1e-05,
+      "loss": 1.169,
+      "step": 1216
+    },
+    {
+      "epoch": 0.7113319208007598,
+      "grad_norm": 1.834796458050961,
+      "learning_rate": 1e-05,
+      "loss": 1.1911,
+      "step": 1217
+    },
+    {
+      "epoch": 0.7119164170380653,
+      "grad_norm": 1.922908041800173,
+      "learning_rate": 1e-05,
+      "loss": 1.2384,
+      "step": 1218
+    },
+    {
+      "epoch": 0.7125009132753708,
+      "grad_norm": 1.6852736623221183,
+      "learning_rate": 1e-05,
+      "loss": 1.2456,
+      "step": 1219
+    },
+    {
+      "epoch": 0.7130854095126763,
+      "grad_norm": 1.7572257384471976,
+      "learning_rate": 1e-05,
+      "loss": 1.1986,
+      "step": 1220
+    },
+    {
+      "epoch": 0.7136699057499817,
+      "grad_norm": 1.6785216853625138,
+      "learning_rate": 1e-05,
+      "loss": 1.1833,
+      "step": 1221
+    },
+    {
+      "epoch": 0.7142544019872872,
+      "grad_norm": 1.8167946880408583,
+      "learning_rate": 1e-05,
+      "loss": 1.1963,
+      "step": 1222
+    },
+    {
+      "epoch": 0.7148388982245927,
+      "grad_norm": 1.6785211837612601,
+      "learning_rate": 1e-05,
+      "loss": 1.2225,
+      "step": 1223
+    },
+    {
+      "epoch": 0.7154233944618982,
+      "grad_norm": 1.7826377845724366,
+      "learning_rate": 1e-05,
+      "loss": 1.2202,
+      "step": 1224
+    },
+    {
+      "epoch": 0.7160078906992037,
+      "grad_norm": 1.8265081500703284,
+      "learning_rate": 1e-05,
+      "loss": 1.2174,
+      "step": 1225
+    },
+    {
+      "epoch": 0.7165923869365091,
+      "grad_norm": 1.7058818132129345,
+      "learning_rate": 1e-05,
+      "loss": 1.178,
+      "step": 1226
+    },
+    {
+      "epoch": 0.7171768831738146,
+      "grad_norm": 1.8467386316877836,
+      "learning_rate": 1e-05,
+      "loss": 1.2609,
+      "step": 1227
+    },
+    {
+      "epoch": 0.7177613794111201,
+      "grad_norm": 1.97710557218607,
+      "learning_rate": 1e-05,
+      "loss": 1.1944,
+      "step": 1228
+    },
+    {
+      "epoch": 0.7183458756484256,
+      "grad_norm": 1.7356836136652662,
+      "learning_rate": 1e-05,
+      "loss": 1.2407,
+      "step": 1229
+    },
+    {
+      "epoch": 0.718930371885731,
+      "grad_norm": 1.6116395258109857,
+      "learning_rate": 1e-05,
+      "loss": 1.1985,
+      "step": 1230
+    },
+    {
+      "epoch": 0.7195148681230364,
+      "grad_norm": 1.8875775128614178,
+      "learning_rate": 1e-05,
+      "loss": 1.1922,
+      "step": 1231
+    },
+    {
+      "epoch": 0.7200993643603419,
+      "grad_norm": 1.6408195433392008,
+      "learning_rate": 1e-05,
+      "loss": 1.2271,
+      "step": 1232
+    },
+    {
+      "epoch": 0.7206838605976474,
+      "grad_norm": 1.9250613729817625,
+      "learning_rate": 1e-05,
+      "loss": 1.1976,
+      "step": 1233
+    },
+    {
+      "epoch": 0.7212683568349528,
+      "grad_norm": 1.8096879422326255,
+      "learning_rate": 1e-05,
+      "loss": 1.2035,
+      "step": 1234
+    },
+    {
+      "epoch": 0.7218528530722583,
+      "grad_norm": 1.7730482570051076,
+      "learning_rate": 1e-05,
+      "loss": 1.2116,
+      "step": 1235
+    },
+    {
+      "epoch": 0.7224373493095638,
+      "grad_norm": 1.848326206221338,
+      "learning_rate": 1e-05,
+      "loss": 1.2119,
+      "step": 1236
+    },
+    {
+      "epoch": 0.7230218455468693,
+      "grad_norm": 1.6542109287508988,
+      "learning_rate": 1e-05,
+      "loss": 1.2184,
+      "step": 1237
+    },
+    {
+      "epoch": 0.7236063417841747,
+      "grad_norm": 1.7684235035102427,
+      "learning_rate": 1e-05,
+      "loss": 1.2201,
+      "step": 1238
+    },
+    {
+      "epoch": 0.7241908380214802,
+      "grad_norm": 1.7303223592016879,
+      "learning_rate": 1e-05,
+      "loss": 1.2059,
+      "step": 1239
+    },
+    {
+      "epoch": 0.7247753342587857,
+      "grad_norm": 1.6595710325596602,
+      "learning_rate": 1e-05,
+      "loss": 1.1814,
+      "step": 1240
+    },
+    {
+      "epoch": 0.7253598304960912,
+      "grad_norm": 1.7361807658952841,
+      "learning_rate": 1e-05,
+      "loss": 1.169,
+      "step": 1241
+    },
+    {
+      "epoch": 0.7259443267333967,
+      "grad_norm": 1.7463690879950886,
+      "learning_rate": 1e-05,
+      "loss": 1.1794,
+      "step": 1242
+    },
+    {
+      "epoch": 0.7265288229707021,
+      "grad_norm": 1.5942977689918427,
+      "learning_rate": 1e-05,
+      "loss": 1.1661,
+      "step": 1243
+    },
+    {
+      "epoch": 0.7271133192080076,
+      "grad_norm": 1.6428473156206063,
+      "learning_rate": 1e-05,
+      "loss": 1.1873,
+      "step": 1244
+    },
+    {
+      "epoch": 0.7276978154453131,
+      "grad_norm": 1.6730594528295026,
+      "learning_rate": 1e-05,
+      "loss": 1.2416,
+      "step": 1245
+    },
+    {
+      "epoch": 0.7282823116826186,
+      "grad_norm": 1.8461753387807138,
+      "learning_rate": 1e-05,
+      "loss": 1.2569,
+      "step": 1246
+    },
+    {
+      "epoch": 0.728866807919924,
+      "grad_norm": 1.7686919852720164,
+      "learning_rate": 1e-05,
+      "loss": 1.2158,
+      "step": 1247
+    },
+    {
+      "epoch": 0.7294513041572295,
+      "grad_norm": 1.809347153986535,
+      "learning_rate": 1e-05,
+      "loss": 1.2221,
+      "step": 1248
+    },
+    {
+      "epoch": 0.730035800394535,
+      "grad_norm": 1.8611715696849878,
+      "learning_rate": 1e-05,
+      "loss": 1.2391,
+      "step": 1249
+    },
+    {
+      "epoch": 0.7306202966318405,
+      "grad_norm": 1.8963970752781283,
+      "learning_rate": 1e-05,
+      "loss": 1.2312,
+      "step": 1250
+    },
+    {
+      "epoch": 0.7312047928691459,
+      "grad_norm": 1.8907371613870927,
+      "learning_rate": 1e-05,
+      "loss": 1.2403,
+      "step": 1251
+    },
+    {
+      "epoch": 0.7317892891064514,
+      "grad_norm": 1.8052839992793395,
+      "learning_rate": 1e-05,
+      "loss": 1.245,
+      "step": 1252
+    },
+    {
+      "epoch": 0.7323737853437569,
+      "grad_norm": 1.8514499932471362,
+      "learning_rate": 1e-05,
+      "loss": 1.1928,
+      "step": 1253
+    },
+    {
+      "epoch": 0.7329582815810624,
+      "grad_norm": 1.718581925842705,
+      "learning_rate": 1e-05,
+      "loss": 1.246,
+      "step": 1254
+    },
+    {
+      "epoch": 0.7335427778183677,
+      "grad_norm": 1.884643775182727,
+      "learning_rate": 1e-05,
+      "loss": 1.2069,
+      "step": 1255
+    },
+    {
+      "epoch": 0.7341272740556732,
+      "grad_norm": 1.6222655148048355,
+      "learning_rate": 1e-05,
+      "loss": 1.217,
+      "step": 1256
+    },
+    {
+      "epoch": 0.7347117702929787,
+      "grad_norm": 1.7661212652828964,
+      "learning_rate": 1e-05,
+      "loss": 1.1814,
+      "step": 1257
+    },
+    {
+      "epoch": 0.7352962665302842,
+      "grad_norm": 1.588965640486027,
+      "learning_rate": 1e-05,
+      "loss": 1.2202,
+      "step": 1258
+    },
+    {
+      "epoch": 0.7358807627675897,
+      "grad_norm": 1.6910603004695342,
+      "learning_rate": 1e-05,
+      "loss": 1.2092,
+      "step": 1259
+    },
+    {
+      "epoch": 0.7364652590048951,
+      "grad_norm": 1.8688561579842795,
+      "learning_rate": 1e-05,
+      "loss": 1.1538,
+      "step": 1260
+    },
+    {
+      "epoch": 0.7370497552422006,
+      "grad_norm": 1.7690320111670608,
+      "learning_rate": 1e-05,
+      "loss": 1.2007,
+      "step": 1261
+    },
+    {
+      "epoch": 0.7376342514795061,
+      "grad_norm": 1.7237610247026385,
+      "learning_rate": 1e-05,
+      "loss": 1.1943,
+      "step": 1262
+    },
+    {
+      "epoch": 0.7382187477168116,
+      "grad_norm": 1.9245880106494826,
+      "learning_rate": 1e-05,
+      "loss": 1.2024,
+      "step": 1263
+    },
+    {
+      "epoch": 0.738803243954117,
+      "grad_norm": 1.941515535857357,
+      "learning_rate": 1e-05,
+      "loss": 1.145,
+      "step": 1264
+    },
+    {
+      "epoch": 0.7393877401914225,
+      "grad_norm": 1.7750625230363766,
+      "learning_rate": 1e-05,
+      "loss": 1.1954,
+      "step": 1265
+    },
+    {
+      "epoch": 0.739972236428728,
+      "grad_norm": 1.7930456519317548,
+      "learning_rate": 1e-05,
+      "loss": 1.2232,
+      "step": 1266
+    },
+    {
+      "epoch": 0.7405567326660335,
+      "grad_norm": 1.882006961202725,
+      "learning_rate": 1e-05,
+      "loss": 1.2272,
+      "step": 1267
+    },
+    {
+      "epoch": 0.7411412289033389,
+      "grad_norm": 1.7192592721532256,
+      "learning_rate": 1e-05,
+      "loss": 1.2354,
+      "step": 1268
+    },
+    {
+      "epoch": 0.7417257251406444,
+      "grad_norm": 1.7207683417559148,
+      "learning_rate": 1e-05,
+      "loss": 1.1954,
+      "step": 1269
+    },
+    {
+      "epoch": 0.7423102213779499,
+      "grad_norm": 2.0844351712742446,
+      "learning_rate": 1e-05,
+      "loss": 1.2047,
+      "step": 1270
+    },
+    {
+      "epoch": 0.7428947176152554,
+      "grad_norm": 1.7877114877381755,
+      "learning_rate": 1e-05,
+      "loss": 1.2032,
+      "step": 1271
+    },
+    {
+      "epoch": 0.7434792138525608,
+      "grad_norm": 1.7887545930197937,
+      "learning_rate": 1e-05,
+      "loss": 1.1877,
+      "step": 1272
+    },
+    {
+      "epoch": 0.7440637100898663,
+      "grad_norm": 1.7421202235212128,
+      "learning_rate": 1e-05,
+      "loss": 1.1363,
+      "step": 1273
+    },
+    {
+      "epoch": 0.7446482063271718,
+      "grad_norm": 1.751056755695512,
+      "learning_rate": 1e-05,
+      "loss": 1.2135,
+      "step": 1274
+    },
+    {
+      "epoch": 0.7452327025644773,
+      "grad_norm": 1.7483481434511396,
+      "learning_rate": 1e-05,
+      "loss": 1.245,
+      "step": 1275
+    },
+    {
+      "epoch": 0.7458171988017828,
+      "grad_norm": 1.8149277271597957,
+      "learning_rate": 1e-05,
+      "loss": 1.2164,
+      "step": 1276
+    },
+    {
+      "epoch": 0.7464016950390882,
+      "grad_norm": 1.6153706285910812,
+      "learning_rate": 1e-05,
+      "loss": 1.1907,
+      "step": 1277
+    },
+    {
+      "epoch": 0.7469861912763937,
+      "grad_norm": 1.6960195319940972,
+      "learning_rate": 1e-05,
+      "loss": 1.2103,
+      "step": 1278
+    },
+    {
+      "epoch": 0.7475706875136992,
+      "grad_norm": 1.7731348510465754,
+      "learning_rate": 1e-05,
+      "loss": 1.1993,
+      "step": 1279
+    },
+    {
+      "epoch": 0.7481551837510046,
+      "grad_norm": 1.709041785765991,
+      "learning_rate": 1e-05,
+      "loss": 1.212,
+      "step": 1280
+    },
+    {
+      "epoch": 0.74873967998831,
+      "grad_norm": 1.6518800447289317,
+      "learning_rate": 1e-05,
+      "loss": 1.1986,
+      "step": 1281
+    },
+    {
+      "epoch": 0.7493241762256155,
+      "grad_norm": 1.9323805032965706,
+      "learning_rate": 1e-05,
+      "loss": 1.1789,
+      "step": 1282
+    },
+    {
+      "epoch": 0.749908672462921,
+      "grad_norm": 1.7784135278331827,
+      "learning_rate": 1e-05,
+      "loss": 1.1866,
+      "step": 1283
+    },
+    {
+      "epoch": 0.7504931687002265,
+      "grad_norm": 1.7183890143774414,
+      "learning_rate": 1e-05,
+      "loss": 1.2156,
+      "step": 1284
+    },
+    {
+      "epoch": 0.7510776649375319,
+      "grad_norm": 1.7082359474812043,
+      "learning_rate": 1e-05,
+      "loss": 1.2184,
+      "step": 1285
+    },
+    {
+      "epoch": 0.7516621611748374,
+      "grad_norm": 1.8297557527886366,
+      "learning_rate": 1e-05,
+      "loss": 1.2218,
+      "step": 1286
+    },
+    {
+      "epoch": 0.7522466574121429,
+      "grad_norm": 1.8181367101397234,
+      "learning_rate": 1e-05,
+      "loss": 1.1545,
+      "step": 1287
+    },
+    {
+      "epoch": 0.7528311536494484,
+      "grad_norm": 1.7054194374674427,
+      "learning_rate": 1e-05,
+      "loss": 1.2298,
+      "step": 1288
+    },
+    {
+      "epoch": 0.7534156498867538,
+      "grad_norm": 1.7309758187700068,
+      "learning_rate": 1e-05,
+      "loss": 1.2029,
+      "step": 1289
+    },
+    {
+      "epoch": 0.7540001461240593,
+      "grad_norm": 1.8858872782411105,
+      "learning_rate": 1e-05,
+      "loss": 1.212,
+      "step": 1290
+    },
+    {
+      "epoch": 0.7545846423613648,
+      "grad_norm": 1.839397745586569,
+      "learning_rate": 1e-05,
+      "loss": 1.2546,
+      "step": 1291
+    },
+    {
+      "epoch": 0.7551691385986703,
+      "grad_norm": 1.6430085933113683,
+      "learning_rate": 1e-05,
+      "loss": 1.1865,
+      "step": 1292
+    },
+    {
+      "epoch": 0.7557536348359758,
+      "grad_norm": 1.6999189982554415,
+      "learning_rate": 1e-05,
+      "loss": 1.2375,
+      "step": 1293
+    },
+    {
+      "epoch": 0.7563381310732812,
+      "grad_norm": 1.8044698053984058,
+      "learning_rate": 1e-05,
+      "loss": 1.2508,
+      "step": 1294
+    },
+    {
+      "epoch": 0.7569226273105867,
+      "grad_norm": 1.7595509416186388,
+      "learning_rate": 1e-05,
+      "loss": 1.2238,
+      "step": 1295
+    },
+    {
+      "epoch": 0.7575071235478922,
+      "grad_norm": 1.6811468931265743,
+      "learning_rate": 1e-05,
+      "loss": 1.2063,
+      "step": 1296
+    },
+    {
+      "epoch": 0.7580916197851977,
+      "grad_norm": 1.7720477167041875,
+      "learning_rate": 1e-05,
+      "loss": 1.2058,
+      "step": 1297
+    },
+    {
+      "epoch": 0.7586761160225031,
+      "grad_norm": 1.9062735621771278,
+      "learning_rate": 1e-05,
+      "loss": 1.1823,
+      "step": 1298
+    },
+    {
+      "epoch": 0.7592606122598086,
+      "grad_norm": 1.8904593930287228,
+      "learning_rate": 1e-05,
+      "loss": 1.1963,
+      "step": 1299
+    },
+    {
+      "epoch": 0.7598451084971141,
+      "grad_norm": 1.681403713069301,
+      "learning_rate": 1e-05,
+      "loss": 1.2139,
+      "step": 1300
+    },
+    {
+      "epoch": 0.7604296047344196,
+      "grad_norm": 1.619736279309958,
+      "learning_rate": 1e-05,
+      "loss": 1.2335,
+      "step": 1301
+    },
+    {
+      "epoch": 0.761014100971725,
+      "grad_norm": 1.8262078156275081,
+      "learning_rate": 1e-05,
+      "loss": 1.1794,
+      "step": 1302
+    },
+    {
+      "epoch": 0.7615985972090304,
+      "grad_norm": 2.3846693235939997,
+      "learning_rate": 1e-05,
+      "loss": 1.2252,
+      "step": 1303
+    },
+    {
+      "epoch": 0.762183093446336,
+      "grad_norm": 1.9608211918364657,
+      "learning_rate": 1e-05,
+      "loss": 1.1648,
+      "step": 1304
+    },
+    {
+      "epoch": 0.7627675896836414,
+      "grad_norm": 1.8622237172471543,
+      "learning_rate": 1e-05,
+      "loss": 1.2044,
+      "step": 1305
+    },
+    {
+      "epoch": 0.7633520859209468,
+      "grad_norm": 1.8267283778074421,
+      "learning_rate": 1e-05,
+      "loss": 1.217,
+      "step": 1306
+    },
+    {
+      "epoch": 0.7639365821582523,
+      "grad_norm": 1.6990709268248192,
+      "learning_rate": 1e-05,
+      "loss": 1.1911,
+      "step": 1307
+    },
+    {
+      "epoch": 0.7645210783955578,
+      "grad_norm": 1.8659930993930636,
+      "learning_rate": 1e-05,
+      "loss": 1.2259,
+      "step": 1308
+    },
+    {
+      "epoch": 0.7651055746328633,
+      "grad_norm": 1.792322742316284,
+      "learning_rate": 1e-05,
+      "loss": 1.191,
+      "step": 1309
+    },
+    {
+      "epoch": 0.7656900708701688,
+      "grad_norm": 1.9352499628761268,
+      "learning_rate": 1e-05,
+      "loss": 1.2262,
+      "step": 1310
+    },
+    {
+      "epoch": 0.7662745671074742,
+      "grad_norm": 1.8644002858747164,
+      "learning_rate": 1e-05,
+      "loss": 1.2466,
+      "step": 1311
+    },
+    {
+      "epoch": 0.7668590633447797,
+      "grad_norm": 1.7734355271488278,
+      "learning_rate": 1e-05,
+      "loss": 1.2038,
+      "step": 1312
+    },
+    {
+      "epoch": 0.7674435595820852,
+      "grad_norm": 1.8908724409462652,
+      "learning_rate": 1e-05,
+      "loss": 1.2454,
+      "step": 1313
+    },
+    {
+      "epoch": 0.7680280558193907,
+      "grad_norm": 1.7797722648290537,
+      "learning_rate": 1e-05,
+      "loss": 1.2043,
+      "step": 1314
+    },
+    {
+      "epoch": 0.7686125520566961,
+      "grad_norm": 1.7178371729990909,
+      "learning_rate": 1e-05,
+      "loss": 1.2119,
+      "step": 1315
+    },
+    {
+      "epoch": 0.7691970482940016,
+      "grad_norm": 1.9134591476185028,
+      "learning_rate": 1e-05,
+      "loss": 1.1515,
+      "step": 1316
+    },
+    {
+      "epoch": 0.7697815445313071,
+      "grad_norm": 1.7944534912042929,
+      "learning_rate": 1e-05,
+      "loss": 1.1998,
+      "step": 1317
+    },
+    {
+      "epoch": 0.7703660407686126,
+      "grad_norm": 1.7721283404766524,
+      "learning_rate": 1e-05,
+      "loss": 1.1723,
+      "step": 1318
+    },
+    {
+      "epoch": 0.770950537005918,
+      "grad_norm": 1.7728627662934542,
+      "learning_rate": 1e-05,
+      "loss": 1.2024,
+      "step": 1319
+    },
+    {
+      "epoch": 0.7715350332432235,
+      "grad_norm": 1.7508502203495642,
+      "learning_rate": 1e-05,
+      "loss": 1.1867,
+      "step": 1320
+    },
+    {
+      "epoch": 0.772119529480529,
+      "grad_norm": 1.8178407053959227,
+      "learning_rate": 1e-05,
+      "loss": 1.2174,
+      "step": 1321
+    },
+    {
+      "epoch": 0.7727040257178345,
+      "grad_norm": 1.7240911614990535,
+      "learning_rate": 1e-05,
+      "loss": 1.2002,
+      "step": 1322
+    },
+    {
+      "epoch": 0.7732885219551399,
+      "grad_norm": 2.0566057694739395,
+      "learning_rate": 1e-05,
+      "loss": 1.184,
+      "step": 1323
+    },
+    {
+      "epoch": 0.7738730181924454,
+      "grad_norm": 1.8208263779937444,
+      "learning_rate": 1e-05,
+      "loss": 1.2558,
+      "step": 1324
+    },
+    {
+      "epoch": 0.7744575144297509,
+      "grad_norm": 1.6616003426222323,
+      "learning_rate": 1e-05,
+      "loss": 1.147,
+      "step": 1325
+    },
+    {
+      "epoch": 0.7750420106670564,
+      "grad_norm": 1.9213292695423028,
+      "learning_rate": 1e-05,
+      "loss": 1.1864,
+      "step": 1326
+    },
+    {
+      "epoch": 0.7756265069043619,
+      "grad_norm": 1.7782939375589157,
+      "learning_rate": 1e-05,
+      "loss": 1.2053,
+      "step": 1327
+    },
+    {
+      "epoch": 0.7762110031416672,
+      "grad_norm": 1.711881228190048,
+      "learning_rate": 1e-05,
+      "loss": 1.1593,
+      "step": 1328
+    },
+    {
+      "epoch": 0.7767954993789727,
+      "grad_norm": 1.8161102335211012,
+      "learning_rate": 1e-05,
+      "loss": 1.1829,
+      "step": 1329
+    },
+    {
+      "epoch": 0.7773799956162782,
+      "grad_norm": 1.798278572715019,
+      "learning_rate": 1e-05,
+      "loss": 1.1933,
+      "step": 1330
+    },
+    {
+      "epoch": 0.7779644918535837,
+      "grad_norm": 1.6489234388324376,
+      "learning_rate": 1e-05,
+      "loss": 1.1586,
+      "step": 1331
+    },
+    {
+      "epoch": 0.7785489880908891,
+      "grad_norm": 2.0465780539850713,
+      "learning_rate": 1e-05,
+      "loss": 1.2349,
+      "step": 1332
+    },
+    {
+      "epoch": 0.7791334843281946,
+      "grad_norm": 1.8203302893545938,
+      "learning_rate": 1e-05,
+      "loss": 1.186,
+      "step": 1333
+    },
+    {
+      "epoch": 0.7797179805655001,
+      "grad_norm": 1.9146839999688021,
+      "learning_rate": 1e-05,
+      "loss": 1.2164,
+      "step": 1334
+    },
+    {
+      "epoch": 0.7803024768028056,
+      "grad_norm": 1.6943262887147528,
+      "learning_rate": 1e-05,
+      "loss": 1.2018,
+      "step": 1335
+    },
+    {
+      "epoch": 0.780886973040111,
+      "grad_norm": 1.731224397462592,
+      "learning_rate": 1e-05,
+      "loss": 1.2778,
+      "step": 1336
+    },
+    {
+      "epoch": 0.7814714692774165,
+      "grad_norm": 1.9847798564085881,
+      "learning_rate": 1e-05,
+      "loss": 1.2213,
+      "step": 1337
+    },
+    {
+      "epoch": 0.782055965514722,
+      "grad_norm": 1.6803976609930487,
+      "learning_rate": 1e-05,
+      "loss": 1.1876,
+      "step": 1338
+    },
+    {
+      "epoch": 0.7826404617520275,
+      "grad_norm": 1.6332720643138157,
+      "learning_rate": 1e-05,
+      "loss": 1.192,
+      "step": 1339
+    },
+    {
+      "epoch": 0.7832249579893329,
+      "grad_norm": 1.6982638645580401,
+      "learning_rate": 1e-05,
+      "loss": 1.2305,
+      "step": 1340
+    },
+    {
+      "epoch": 0.7838094542266384,
+      "grad_norm": 1.7604625644215484,
+      "learning_rate": 1e-05,
+      "loss": 1.1881,
+      "step": 1341
+    },
+    {
+      "epoch": 0.7843939504639439,
+      "grad_norm": 1.654891347668368,
+      "learning_rate": 1e-05,
+      "loss": 1.2162,
+      "step": 1342
+    },
+    {
+      "epoch": 0.7849784467012494,
+      "grad_norm": 1.7700275200364886,
+      "learning_rate": 1e-05,
+      "loss": 1.1551,
+      "step": 1343
+    },
+    {
+      "epoch": 0.7855629429385549,
+      "grad_norm": 1.8544144922544,
+      "learning_rate": 1e-05,
+      "loss": 1.238,
+      "step": 1344
+    },
+    {
+      "epoch": 0.7861474391758603,
+      "grad_norm": 1.8220822940089572,
+      "learning_rate": 1e-05,
+      "loss": 1.1686,
+      "step": 1345
+    },
+    {
+      "epoch": 0.7867319354131658,
+      "grad_norm": 1.608485570183546,
+      "learning_rate": 1e-05,
+      "loss": 1.2507,
+      "step": 1346
+    },
+    {
+      "epoch": 0.7873164316504713,
+      "grad_norm": 1.627750564012596,
+      "learning_rate": 1e-05,
+      "loss": 1.2119,
+      "step": 1347
+    },
+    {
+      "epoch": 0.7879009278877768,
+      "grad_norm": 1.8612318480165047,
+      "learning_rate": 1e-05,
+      "loss": 1.2011,
+      "step": 1348
+    },
+    {
+      "epoch": 0.7884854241250822,
+      "grad_norm": 1.655141839368859,
+      "learning_rate": 1e-05,
+      "loss": 1.1849,
+      "step": 1349
+    },
+    {
+      "epoch": 0.7890699203623877,
+      "grad_norm": 1.8323707887253349,
+      "learning_rate": 1e-05,
+      "loss": 1.1669,
+      "step": 1350
+    },
+    {
+      "epoch": 0.7896544165996932,
+      "grad_norm": 1.6669208023128832,
+      "learning_rate": 1e-05,
+      "loss": 1.1722,
+      "step": 1351
+    },
+    {
+      "epoch": 0.7902389128369987,
+      "grad_norm": 1.6825031711840566,
+      "learning_rate": 1e-05,
+      "loss": 1.1973,
+      "step": 1352
+    },
+    {
+      "epoch": 0.790823409074304,
+      "grad_norm": 1.7994755330324355,
+      "learning_rate": 1e-05,
+      "loss": 1.1723,
+      "step": 1353
+    },
+    {
+      "epoch": 0.7914079053116095,
+      "grad_norm": 1.8163506310725106,
+      "learning_rate": 1e-05,
+      "loss": 1.1968,
+      "step": 1354
+    },
+    {
+      "epoch": 0.791992401548915,
+      "grad_norm": 1.7797980290964028,
+      "learning_rate": 1e-05,
+      "loss": 1.2057,
+      "step": 1355
+    },
+    {
+      "epoch": 0.7925768977862205,
+      "grad_norm": 1.7676173772446113,
+      "learning_rate": 1e-05,
+      "loss": 1.2646,
+      "step": 1356
+    },
+    {
+      "epoch": 0.7931613940235259,
+      "grad_norm": 1.8175584057311476,
+      "learning_rate": 1e-05,
+      "loss": 1.1815,
+      "step": 1357
+    },
+    {
+      "epoch": 0.7937458902608314,
+      "grad_norm": 1.8482468646615078,
+      "learning_rate": 1e-05,
+      "loss": 1.241,
+      "step": 1358
+    },
+    {
+      "epoch": 0.7943303864981369,
+      "grad_norm": 1.7773183776515273,
+      "learning_rate": 1e-05,
+      "loss": 1.158,
+      "step": 1359
+    },
+    {
+      "epoch": 0.7949148827354424,
+      "grad_norm": 1.6678662692918516,
+      "learning_rate": 1e-05,
+      "loss": 1.2372,
+      "step": 1360
+    },
+    {
+      "epoch": 0.7954993789727479,
+      "grad_norm": 1.5397730507162668,
+      "learning_rate": 1e-05,
+      "loss": 1.2164,
+      "step": 1361
+    },
+    {
+      "epoch": 0.7960838752100533,
+      "grad_norm": 1.7902568487914166,
+      "learning_rate": 1e-05,
+      "loss": 1.2434,
+      "step": 1362
+    },
+    {
+      "epoch": 0.7966683714473588,
+      "grad_norm": 1.7527239710483415,
+      "learning_rate": 1e-05,
+      "loss": 1.1695,
+      "step": 1363
+    },
+    {
+      "epoch": 0.7972528676846643,
+      "grad_norm": 1.8666151117518406,
+      "learning_rate": 1e-05,
+      "loss": 1.1593,
+      "step": 1364
+    },
+    {
+      "epoch": 0.7978373639219698,
+      "grad_norm": 1.706677112454,
+      "learning_rate": 1e-05,
+      "loss": 1.2235,
+      "step": 1365
+    },
+    {
+      "epoch": 0.7984218601592752,
+      "grad_norm": 1.6916785302821469,
+      "learning_rate": 1e-05,
+      "loss": 1.1937,
+      "step": 1366
+    },
+    {
+      "epoch": 0.7990063563965807,
+      "grad_norm": 1.8491501833474746,
+      "learning_rate": 1e-05,
+      "loss": 1.1626,
+      "step": 1367
+    },
+    {
+      "epoch": 0.7995908526338862,
+      "grad_norm": 1.8126117484896938,
+      "learning_rate": 1e-05,
+      "loss": 1.1853,
+      "step": 1368
+    },
+    {
+      "epoch": 0.8001753488711917,
+      "grad_norm": 1.72427974907637,
+      "learning_rate": 1e-05,
+      "loss": 1.1939,
+      "step": 1369
+    },
+    {
+      "epoch": 0.8007598451084971,
+      "grad_norm": 1.9730135116569565,
+      "learning_rate": 1e-05,
+      "loss": 1.214,
+      "step": 1370
+    },
+    {
+      "epoch": 0.8013443413458026,
+      "grad_norm": 1.7737484894893984,
+      "learning_rate": 1e-05,
+      "loss": 1.2274,
+      "step": 1371
+    },
+    {
+      "epoch": 0.8019288375831081,
+      "grad_norm": 1.774438080856986,
+      "learning_rate": 1e-05,
+      "loss": 1.2188,
+      "step": 1372
+    },
+    {
+      "epoch": 0.8025133338204136,
+      "grad_norm": 1.8060321353680795,
+      "learning_rate": 1e-05,
+      "loss": 1.1835,
+      "step": 1373
+    },
+    {
+      "epoch": 0.803097830057719,
+      "grad_norm": 1.7555686629067013,
+      "learning_rate": 1e-05,
+      "loss": 1.2003,
+      "step": 1374
+    },
+    {
+      "epoch": 0.8036823262950245,
+      "grad_norm": 1.65914755863923,
+      "learning_rate": 1e-05,
+      "loss": 1.2456,
+      "step": 1375
+    },
+    {
+      "epoch": 0.80426682253233,
+      "grad_norm": 1.8106252352791725,
+      "learning_rate": 1e-05,
+      "loss": 1.1821,
+      "step": 1376
+    },
+    {
+      "epoch": 0.8048513187696354,
+      "grad_norm": 1.6968487858889323,
+      "learning_rate": 1e-05,
+      "loss": 1.2492,
+      "step": 1377
+    },
+    {
+      "epoch": 0.805435815006941,
+      "grad_norm": 1.804932411641239,
+      "learning_rate": 1e-05,
+      "loss": 1.1721,
+      "step": 1378
+    },
+    {
+      "epoch": 0.8060203112442463,
+      "grad_norm": 1.6476475309284202,
+      "learning_rate": 1e-05,
+      "loss": 1.2038,
+      "step": 1379
+    },
+    {
+      "epoch": 0.8066048074815518,
+      "grad_norm": 1.6717744188855592,
+      "learning_rate": 1e-05,
+      "loss": 1.1727,
+      "step": 1380
+    },
+    {
+      "epoch": 0.8071893037188573,
+      "grad_norm": 1.8844750049268786,
+      "learning_rate": 1e-05,
+      "loss": 1.2417,
+      "step": 1381
+    },
+    {
+      "epoch": 0.8077737999561628,
+      "grad_norm": 1.708903302901378,
+      "learning_rate": 1e-05,
+      "loss": 1.1774,
+      "step": 1382
+    },
+    {
+      "epoch": 0.8083582961934682,
+      "grad_norm": 1.5900628600208675,
+      "learning_rate": 1e-05,
+      "loss": 1.2296,
+      "step": 1383
+    },
+    {
+      "epoch": 0.8089427924307737,
+      "grad_norm": 1.6390475034785486,
+      "learning_rate": 1e-05,
+      "loss": 1.2062,
+      "step": 1384
+    },
+    {
+      "epoch": 0.8095272886680792,
+      "grad_norm": 1.7826902175248391,
+      "learning_rate": 1e-05,
+      "loss": 1.2318,
+      "step": 1385
+    },
+    {
+      "epoch": 0.8101117849053847,
+      "grad_norm": 1.645241698140373,
+      "learning_rate": 1e-05,
+      "loss": 1.121,
+      "step": 1386
+    },
+    {
+      "epoch": 0.8106962811426901,
+      "grad_norm": 1.9155315020704324,
+      "learning_rate": 1e-05,
+      "loss": 1.2083,
+      "step": 1387
+    },
+    {
+      "epoch": 0.8112807773799956,
+      "grad_norm": 1.6618258394642649,
+      "learning_rate": 1e-05,
+      "loss": 1.1768,
+      "step": 1388
+    },
+    {
+      "epoch": 0.8118652736173011,
+      "grad_norm": 1.734020696166433,
+      "learning_rate": 1e-05,
+      "loss": 1.204,
+      "step": 1389
+    },
+    {
+      "epoch": 0.8124497698546066,
+      "grad_norm": 1.6961825449184114,
+      "learning_rate": 1e-05,
+      "loss": 1.2053,
+      "step": 1390
+    },
+    {
+      "epoch": 0.813034266091912,
+      "grad_norm": 2.0358781594587128,
+      "learning_rate": 1e-05,
+      "loss": 1.1995,
+      "step": 1391
+    },
+    {
+      "epoch": 0.8136187623292175,
+      "grad_norm": 1.9496722497829337,
+      "learning_rate": 1e-05,
+      "loss": 1.2021,
+      "step": 1392
+    },
+    {
+      "epoch": 0.814203258566523,
+      "grad_norm": 1.6771508126272237,
+      "learning_rate": 1e-05,
+      "loss": 1.2032,
+      "step": 1393
+    },
+    {
+      "epoch": 0.8147877548038285,
+      "grad_norm": 1.8532928894817202,
+      "learning_rate": 1e-05,
+      "loss": 1.1645,
+      "step": 1394
+    },
+    {
+      "epoch": 0.815372251041134,
+      "grad_norm": 1.8561955456339347,
+      "learning_rate": 1e-05,
+      "loss": 1.2019,
+      "step": 1395
+    },
+    {
+      "epoch": 0.8159567472784394,
+      "grad_norm": 1.864608968467544,
+      "learning_rate": 1e-05,
+      "loss": 1.1628,
+      "step": 1396
+    },
+    {
+      "epoch": 0.8165412435157449,
+      "grad_norm": 1.6945077638271562,
+      "learning_rate": 1e-05,
+      "loss": 1.1791,
+      "step": 1397
+    },
+    {
+      "epoch": 0.8171257397530504,
+      "grad_norm": 1.7931038980495606,
+      "learning_rate": 1e-05,
+      "loss": 1.1661,
+      "step": 1398
+    },
+    {
+      "epoch": 0.8177102359903559,
+      "grad_norm": 1.7161877879324599,
+      "learning_rate": 1e-05,
+      "loss": 1.2102,
+      "step": 1399
+    },
+    {
+      "epoch": 0.8182947322276612,
+      "grad_norm": 1.9385113754043772,
+      "learning_rate": 1e-05,
+      "loss": 1.2002,
+      "step": 1400
+    },
+    {
+      "epoch": 0.8188792284649667,
+      "grad_norm": 1.6114440422227254,
+      "learning_rate": 1e-05,
+      "loss": 1.2833,
+      "step": 1401
+    },
+    {
+      "epoch": 0.8194637247022722,
+      "grad_norm": 1.8019610611972545,
+      "learning_rate": 1e-05,
+      "loss": 1.1711,
+      "step": 1402
+    },
+    {
+      "epoch": 0.8200482209395777,
+      "grad_norm": 1.79747554477766,
+      "learning_rate": 1e-05,
+      "loss": 1.1635,
+      "step": 1403
+    },
+    {
+      "epoch": 0.8206327171768831,
+      "grad_norm": 1.78241292778234,
+      "learning_rate": 1e-05,
+      "loss": 1.2008,
+      "step": 1404
+    },
+    {
+      "epoch": 0.8212172134141886,
+      "grad_norm": 1.6534313401695546,
+      "learning_rate": 1e-05,
+      "loss": 1.1463,
+      "step": 1405
+    },
+    {
+      "epoch": 0.8218017096514941,
+      "grad_norm": 1.632942038806415,
+      "learning_rate": 1e-05,
+      "loss": 1.2303,
+      "step": 1406
+    },
+    {
+      "epoch": 0.8223862058887996,
+      "grad_norm": 1.7848785019495081,
+      "learning_rate": 1e-05,
+      "loss": 1.2154,
+      "step": 1407
+    },
+    {
+      "epoch": 0.822970702126105,
+      "grad_norm": 1.666295425065476,
+      "learning_rate": 1e-05,
+      "loss": 1.1808,
+      "step": 1408
+    },
+    {
+      "epoch": 0.8235551983634105,
+      "grad_norm": 1.7283295097949425,
+      "learning_rate": 1e-05,
+      "loss": 1.2157,
+      "step": 1409
+    },
+    {
+      "epoch": 0.824139694600716,
+      "grad_norm": 1.7936464003981827,
+      "learning_rate": 1e-05,
+      "loss": 1.2612,
+      "step": 1410
+    },
+    {
+      "epoch": 0.8247241908380215,
+      "grad_norm": 1.7389581906947083,
+      "learning_rate": 1e-05,
+      "loss": 1.2036,
+      "step": 1411
+    },
+    {
+      "epoch": 0.825308687075327,
+      "grad_norm": 1.8860634740238642,
+      "learning_rate": 1e-05,
+      "loss": 1.194,
+      "step": 1412
+    },
+    {
+      "epoch": 0.8258931833126324,
+      "grad_norm": 1.6875974262964184,
+      "learning_rate": 1e-05,
+      "loss": 1.1652,
+      "step": 1413
+    },
+    {
+      "epoch": 0.8264776795499379,
+      "grad_norm": 1.6606930403024354,
+      "learning_rate": 1e-05,
+      "loss": 1.187,
+      "step": 1414
+    },
+    {
+      "epoch": 0.8270621757872434,
+      "grad_norm": 1.7627927217387547,
+      "learning_rate": 1e-05,
+      "loss": 1.2048,
+      "step": 1415
+    },
+    {
+      "epoch": 0.8276466720245489,
+      "grad_norm": 1.6802154429848473,
+      "learning_rate": 1e-05,
+      "loss": 1.1634,
+      "step": 1416
+    },
+    {
+      "epoch": 0.8282311682618543,
+      "grad_norm": 1.7991425530477236,
+      "learning_rate": 1e-05,
+      "loss": 1.1747,
+      "step": 1417
+    },
+    {
+      "epoch": 0.8288156644991598,
+      "grad_norm": 1.6558307818855331,
+      "learning_rate": 1e-05,
+      "loss": 1.2391,
+      "step": 1418
+    },
+    {
+      "epoch": 0.8294001607364653,
+      "grad_norm": 1.742236017289666,
+      "learning_rate": 1e-05,
+      "loss": 1.2348,
+      "step": 1419
+    },
+    {
+      "epoch": 0.8299846569737708,
+      "grad_norm": 1.8025528022053625,
+      "learning_rate": 1e-05,
+      "loss": 1.2574,
+      "step": 1420
+    },
+    {
+      "epoch": 0.8305691532110762,
+      "grad_norm": 1.9068814914156604,
+      "learning_rate": 1e-05,
+      "loss": 1.1879,
+      "step": 1421
+    },
+    {
+      "epoch": 0.8311536494483817,
+      "grad_norm": 1.7160003176223484,
+      "learning_rate": 1e-05,
+      "loss": 1.1836,
+      "step": 1422
+    },
+    {
+      "epoch": 0.8317381456856872,
+      "grad_norm": 1.6088727099425113,
+      "learning_rate": 1e-05,
+      "loss": 1.2622,
+      "step": 1423
+    },
+    {
+      "epoch": 0.8323226419229927,
+      "grad_norm": 1.5911765703919147,
+      "learning_rate": 1e-05,
+      "loss": 1.1549,
+      "step": 1424
+    },
+    {
+      "epoch": 0.832907138160298,
+      "grad_norm": 1.7356418294027836,
+      "learning_rate": 1e-05,
+      "loss": 1.1818,
+      "step": 1425
+    },
+    {
+      "epoch": 0.8334916343976035,
+      "grad_norm": 1.763290160859928,
+      "learning_rate": 1e-05,
+      "loss": 1.1833,
+      "step": 1426
+    },
+    {
+      "epoch": 0.834076130634909,
+      "grad_norm": 1.8221057518648887,
+      "learning_rate": 1e-05,
+      "loss": 1.2145,
+      "step": 1427
+    },
+    {
+      "epoch": 0.8346606268722145,
+      "grad_norm": 1.6761459738721343,
+      "learning_rate": 1e-05,
+      "loss": 1.2052,
+      "step": 1428
+    },
+    {
+      "epoch": 0.83524512310952,
+      "grad_norm": 1.8464266546516155,
+      "learning_rate": 1e-05,
+      "loss": 1.176,
+      "step": 1429
+    },
+    {
+      "epoch": 0.8358296193468254,
+      "grad_norm": 1.7357473149023501,
+      "learning_rate": 1e-05,
+      "loss": 1.1677,
+      "step": 1430
+    },
+    {
+      "epoch": 0.8364141155841309,
+      "grad_norm": 1.6972354376088823,
+      "learning_rate": 1e-05,
+      "loss": 1.2074,
+      "step": 1431
+    },
+    {
+      "epoch": 0.8369986118214364,
+      "grad_norm": 2.01187317492466,
+      "learning_rate": 1e-05,
+      "loss": 1.2163,
+      "step": 1432
+    },
+    {
+      "epoch": 0.8375831080587419,
+      "grad_norm": 1.8980933744908484,
+      "learning_rate": 1e-05,
+      "loss": 1.2346,
+      "step": 1433
+    },
+    {
+      "epoch": 0.8381676042960473,
+      "grad_norm": 1.784164509497389,
+      "learning_rate": 1e-05,
+      "loss": 1.2028,
+      "step": 1434
+    },
+    {
+      "epoch": 0.8387521005333528,
+      "grad_norm": 1.7541700813785006,
+      "learning_rate": 1e-05,
+      "loss": 1.2202,
+      "step": 1435
+    },
+    {
+      "epoch": 0.8393365967706583,
+      "grad_norm": 1.6982359358861383,
+      "learning_rate": 1e-05,
+      "loss": 1.1943,
+      "step": 1436
+    },
+    {
+      "epoch": 0.8399210930079638,
+      "grad_norm": 1.7802940686376165,
+      "learning_rate": 1e-05,
+      "loss": 1.1802,
+      "step": 1437
+    },
+    {
+      "epoch": 0.8405055892452692,
+      "grad_norm": 1.622015993240893,
+      "learning_rate": 1e-05,
+      "loss": 1.2037,
+      "step": 1438
+    },
+    {
+      "epoch": 0.8410900854825747,
+      "grad_norm": 1.7703779223534404,
+      "learning_rate": 1e-05,
+      "loss": 1.1977,
+      "step": 1439
+    },
+    {
+      "epoch": 0.8416745817198802,
+      "grad_norm": 1.717220741619776,
+      "learning_rate": 1e-05,
+      "loss": 1.1641,
+      "step": 1440
+    },
+    {
+      "epoch": 0.8422590779571857,
+      "grad_norm": 1.7186291709395878,
+      "learning_rate": 1e-05,
+      "loss": 1.151,
+      "step": 1441
+    },
+    {
+      "epoch": 0.8428435741944911,
+      "grad_norm": 1.7857716647599395,
+      "learning_rate": 1e-05,
+      "loss": 1.268,
+      "step": 1442
+    },
+    {
+      "epoch": 0.8434280704317966,
+      "grad_norm": 1.8921881876013178,
+      "learning_rate": 1e-05,
+      "loss": 1.1941,
+      "step": 1443
+    },
+    {
+      "epoch": 0.8440125666691021,
+      "grad_norm": 1.7529726871764004,
+      "learning_rate": 1e-05,
+      "loss": 1.1963,
+      "step": 1444
+    },
+    {
+      "epoch": 0.8445970629064076,
+      "grad_norm": 1.7150267139262059,
+      "learning_rate": 1e-05,
+      "loss": 1.175,
+      "step": 1445
+    },
+    {
+      "epoch": 0.845181559143713,
+      "grad_norm": 1.654032907006136,
+      "learning_rate": 1e-05,
+      "loss": 1.199,
+      "step": 1446
+    },
+    {
+      "epoch": 0.8457660553810185,
+      "grad_norm": 1.8179723245739183,
+      "learning_rate": 1e-05,
+      "loss": 1.179,
+      "step": 1447
+    },
+    {
+      "epoch": 0.846350551618324,
+      "grad_norm": 1.769733808155264,
+      "learning_rate": 1e-05,
+      "loss": 1.1766,
+      "step": 1448
+    },
+    {
+      "epoch": 0.8469350478556295,
+      "grad_norm": 1.6969897714414874,
+      "learning_rate": 1e-05,
+      "loss": 1.2077,
+      "step": 1449
+    },
+    {
+      "epoch": 0.847519544092935,
+      "grad_norm": 1.603989185342847,
+      "learning_rate": 1e-05,
+      "loss": 1.2145,
+      "step": 1450
+    },
+    {
+      "epoch": 0.8481040403302403,
+      "grad_norm": 1.6892024867450706,
+      "learning_rate": 1e-05,
+      "loss": 1.187,
+      "step": 1451
+    },
+    {
+      "epoch": 0.8486885365675458,
+      "grad_norm": 1.7663972121434868,
+      "learning_rate": 1e-05,
+      "loss": 1.1909,
+      "step": 1452
+    },
+    {
+      "epoch": 0.8492730328048513,
+      "grad_norm": 1.7663890589077151,
+      "learning_rate": 1e-05,
+      "loss": 1.162,
+      "step": 1453
+    },
+    {
+      "epoch": 0.8498575290421568,
+      "grad_norm": 1.840749357401332,
+      "learning_rate": 1e-05,
+      "loss": 1.2277,
+      "step": 1454
+    },
+    {
+      "epoch": 0.8504420252794622,
+      "grad_norm": 1.8481764306219846,
+      "learning_rate": 1e-05,
+      "loss": 1.2437,
+      "step": 1455
+    },
+    {
+      "epoch": 0.8510265215167677,
+      "grad_norm": 2.0056806317008546,
+      "learning_rate": 1e-05,
+      "loss": 1.1801,
+      "step": 1456
+    },
+    {
+      "epoch": 0.8516110177540732,
+      "grad_norm": 1.7421275740354807,
+      "learning_rate": 1e-05,
+      "loss": 1.1784,
+      "step": 1457
+    },
+    {
+      "epoch": 0.8521955139913787,
+      "grad_norm": 1.6463946571742811,
+      "learning_rate": 1e-05,
+      "loss": 1.2047,
+      "step": 1458
+    },
+    {
+      "epoch": 0.8527800102286841,
+      "grad_norm": 1.8534292804865933,
+      "learning_rate": 1e-05,
+      "loss": 1.1072,
+      "step": 1459
+    },
+    {
+      "epoch": 0.8533645064659896,
+      "grad_norm": 1.660493440788436,
+      "learning_rate": 1e-05,
+      "loss": 1.2044,
+      "step": 1460
+    },
+    {
+      "epoch": 0.8539490027032951,
+      "grad_norm": 1.7178657814621334,
+      "learning_rate": 1e-05,
+      "loss": 1.1934,
+      "step": 1461
+    },
+    {
+      "epoch": 0.8545334989406006,
+      "grad_norm": 1.8034664255346757,
+      "learning_rate": 1e-05,
+      "loss": 1.1954,
+      "step": 1462
+    },
+    {
+      "epoch": 0.855117995177906,
+      "grad_norm": 1.8219751950773537,
+      "learning_rate": 1e-05,
+      "loss": 1.22,
+      "step": 1463
+    },
+    {
+      "epoch": 0.8557024914152115,
+      "grad_norm": 1.757679827504922,
+      "learning_rate": 1e-05,
+      "loss": 1.194,
+      "step": 1464
+    },
+    {
+      "epoch": 0.856286987652517,
+      "grad_norm": 1.8478299806235041,
+      "learning_rate": 1e-05,
+      "loss": 1.1999,
+      "step": 1465
+    },
+    {
+      "epoch": 0.8568714838898225,
+      "grad_norm": 1.8342999628569372,
+      "learning_rate": 1e-05,
+      "loss": 1.1523,
+      "step": 1466
+    },
+    {
+      "epoch": 0.857455980127128,
+      "grad_norm": 1.8058711202194881,
+      "learning_rate": 1e-05,
+      "loss": 1.1888,
+      "step": 1467
+    },
+    {
+      "epoch": 0.8580404763644334,
+      "grad_norm": 1.7549154135715819,
+      "learning_rate": 1e-05,
+      "loss": 1.1989,
+      "step": 1468
+    },
+    {
+      "epoch": 0.8586249726017389,
+      "grad_norm": 1.632177316470774,
+      "learning_rate": 1e-05,
+      "loss": 1.1618,
+      "step": 1469
+    },
+    {
+      "epoch": 0.8592094688390444,
+      "grad_norm": 1.6489017443634848,
+      "learning_rate": 1e-05,
+      "loss": 1.1718,
+      "step": 1470
+    },
+    {
+      "epoch": 0.8597939650763499,
+      "grad_norm": 1.7886179209076352,
+      "learning_rate": 1e-05,
+      "loss": 1.1854,
+      "step": 1471
+    },
+    {
+      "epoch": 0.8603784613136553,
+      "grad_norm": 1.7826112119646282,
+      "learning_rate": 1e-05,
+      "loss": 1.1955,
+      "step": 1472
+    },
+    {
+      "epoch": 0.8609629575509608,
+      "grad_norm": 1.7745476860628218,
+      "learning_rate": 1e-05,
+      "loss": 1.1501,
+      "step": 1473
+    },
+    {
+      "epoch": 0.8615474537882662,
+      "grad_norm": 1.7811391749638104,
+      "learning_rate": 1e-05,
+      "loss": 1.2271,
+      "step": 1474
+    },
+    {
+      "epoch": 0.8621319500255717,
+      "grad_norm": 1.6654214097424886,
+      "learning_rate": 1e-05,
+      "loss": 1.1706,
+      "step": 1475
+    },
+    {
+      "epoch": 0.8627164462628771,
+      "grad_norm": 1.6425116721877637,
+      "learning_rate": 1e-05,
+      "loss": 1.2291,
+      "step": 1476
+    },
+    {
+      "epoch": 0.8633009425001826,
+      "grad_norm": 1.8014718407130754,
+      "learning_rate": 1e-05,
+      "loss": 1.1503,
+      "step": 1477
+    },
+    {
+      "epoch": 0.8638854387374881,
+      "grad_norm": 1.64969830960317,
+      "learning_rate": 1e-05,
+      "loss": 1.1684,
+      "step": 1478
+    },
+    {
+      "epoch": 0.8644699349747936,
+      "grad_norm": 1.8626447481456212,
+      "learning_rate": 1e-05,
+      "loss": 1.1911,
+      "step": 1479
+    },
+    {
+      "epoch": 0.865054431212099,
+      "grad_norm": 1.7999013699572606,
+      "learning_rate": 1e-05,
+      "loss": 1.1589,
+      "step": 1480
+    },
+    {
+      "epoch": 0.8656389274494045,
+      "grad_norm": 1.7350189879619662,
+      "learning_rate": 1e-05,
+      "loss": 1.2509,
+      "step": 1481
+    },
+    {
+      "epoch": 0.86622342368671,
+      "grad_norm": 1.6442261704577075,
+      "learning_rate": 1e-05,
+      "loss": 1.2048,
+      "step": 1482
+    },
+    {
+      "epoch": 0.8668079199240155,
+      "grad_norm": 1.708758835990307,
+      "learning_rate": 1e-05,
+      "loss": 1.1914,
+      "step": 1483
+    },
+    {
+      "epoch": 0.867392416161321,
+      "grad_norm": 1.5337341987781408,
+      "learning_rate": 1e-05,
+      "loss": 1.199,
+      "step": 1484
+    },
+    {
+      "epoch": 0.8679769123986264,
+      "grad_norm": 1.7605320517392353,
+      "learning_rate": 1e-05,
+      "loss": 1.2022,
+      "step": 1485
+    },
+    {
+      "epoch": 0.8685614086359319,
+      "grad_norm": 1.7375158510982913,
+      "learning_rate": 1e-05,
+      "loss": 1.2335,
+      "step": 1486
+    },
+    {
+      "epoch": 0.8691459048732374,
+      "grad_norm": 1.7617254099001478,
+      "learning_rate": 1e-05,
+      "loss": 1.2285,
+      "step": 1487
+    },
+    {
+      "epoch": 0.8697304011105429,
+      "grad_norm": 1.8234824523974191,
+      "learning_rate": 1e-05,
+      "loss": 1.2743,
+      "step": 1488
+    },
+    {
+      "epoch": 0.8703148973478483,
+      "grad_norm": 1.8749402867546714,
+      "learning_rate": 1e-05,
+      "loss": 1.2073,
+      "step": 1489
+    },
+    {
+      "epoch": 0.8708993935851538,
+      "grad_norm": 1.7270080102348413,
+      "learning_rate": 1e-05,
+      "loss": 1.2167,
+      "step": 1490
+    },
+    {
+      "epoch": 0.8714838898224593,
+      "grad_norm": 1.8321914362303175,
+      "learning_rate": 1e-05,
+      "loss": 1.2103,
+      "step": 1491
+    },
+    {
+      "epoch": 0.8720683860597648,
+      "grad_norm": 1.8599349367242197,
+      "learning_rate": 1e-05,
+      "loss": 1.2081,
+      "step": 1492
+    },
+    {
+      "epoch": 0.8726528822970702,
+      "grad_norm": 1.763867288237875,
+      "learning_rate": 1e-05,
+      "loss": 1.2616,
+      "step": 1493
+    },
+    {
+      "epoch": 0.8732373785343757,
+      "grad_norm": 1.7555520730482963,
+      "learning_rate": 1e-05,
+      "loss": 1.2048,
+      "step": 1494
+    },
+    {
+      "epoch": 0.8738218747716812,
+      "grad_norm": 1.837358069670692,
+      "learning_rate": 1e-05,
+      "loss": 1.1601,
+      "step": 1495
+    },
+    {
+      "epoch": 0.8744063710089867,
+      "grad_norm": 1.7381696104513737,
+      "learning_rate": 1e-05,
+      "loss": 1.2335,
+      "step": 1496
+    },
+    {
+      "epoch": 0.874990867246292,
+      "grad_norm": 1.7934773859646869,
+      "learning_rate": 1e-05,
+      "loss": 1.2112,
+      "step": 1497
+    },
+    {
+      "epoch": 0.8755753634835975,
+      "grad_norm": 1.7551150394127193,
+      "learning_rate": 1e-05,
+      "loss": 1.1959,
+      "step": 1498
+    },
+    {
+      "epoch": 0.876159859720903,
+      "grad_norm": 1.8248764059582911,
+      "learning_rate": 1e-05,
+      "loss": 1.1852,
+      "step": 1499
+    },
+    {
+      "epoch": 0.8767443559582085,
+      "grad_norm": 1.7698285804645428,
+      "learning_rate": 1e-05,
+      "loss": 1.2238,
+      "step": 1500
+    },
+    {
+      "epoch": 0.877328852195514,
+      "grad_norm": 1.6813281063883738,
+      "learning_rate": 1e-05,
+      "loss": 1.1619,
+      "step": 1501
+    },
+    {
+      "epoch": 0.8779133484328194,
+      "grad_norm": 1.8518876660023262,
+      "learning_rate": 1e-05,
+      "loss": 1.2068,
+      "step": 1502
+    },
+    {
+      "epoch": 0.8784978446701249,
+      "grad_norm": 1.8045401624819568,
+      "learning_rate": 1e-05,
+      "loss": 1.2104,
+      "step": 1503
+    },
+    {
+      "epoch": 0.8790823409074304,
+      "grad_norm": 1.7106974229104552,
+      "learning_rate": 1e-05,
+      "loss": 1.2117,
+      "step": 1504
+    },
+    {
+      "epoch": 0.8796668371447359,
+      "grad_norm": 1.7087084922026714,
+      "learning_rate": 1e-05,
+      "loss": 1.1715,
+      "step": 1505
+    },
+    {
+      "epoch": 0.8802513333820413,
+      "grad_norm": 1.6411719742018618,
+      "learning_rate": 1e-05,
+      "loss": 1.2081,
+      "step": 1506
+    },
+    {
+      "epoch": 0.8808358296193468,
+      "grad_norm": 1.6320202052199069,
+      "learning_rate": 1e-05,
+      "loss": 1.2033,
+      "step": 1507
+    },
+    {
+      "epoch": 0.8814203258566523,
+      "grad_norm": 1.7373922957250891,
+      "learning_rate": 1e-05,
+      "loss": 1.1834,
+      "step": 1508
+    },
+    {
+      "epoch": 0.8820048220939578,
+      "grad_norm": 1.700921427859447,
+      "learning_rate": 1e-05,
+      "loss": 1.2035,
+      "step": 1509
+    },
+    {
+      "epoch": 0.8825893183312632,
+      "grad_norm": 1.7091635861720555,
+      "learning_rate": 1e-05,
+      "loss": 1.1898,
+      "step": 1510
+    },
+    {
+      "epoch": 0.8831738145685687,
+      "grad_norm": 1.6843301562488873,
+      "learning_rate": 1e-05,
+      "loss": 1.2434,
+      "step": 1511
+    },
+    {
+      "epoch": 0.8837583108058742,
+      "grad_norm": 1.7335478895648664,
+      "learning_rate": 1e-05,
+      "loss": 1.1461,
+      "step": 1512
+    },
+    {
+      "epoch": 0.8843428070431797,
+      "grad_norm": 1.8600514467247926,
+      "learning_rate": 1e-05,
+      "loss": 1.1909,
+      "step": 1513
+    },
+    {
+      "epoch": 0.8849273032804851,
+      "grad_norm": 1.8569383167981055,
+      "learning_rate": 1e-05,
+      "loss": 1.1505,
+      "step": 1514
+    },
+    {
+      "epoch": 0.8855117995177906,
+      "grad_norm": 1.6553137384864898,
+      "learning_rate": 1e-05,
+      "loss": 1.2131,
+      "step": 1515
+    },
+    {
+      "epoch": 0.8860962957550961,
+      "grad_norm": 1.7023760208050596,
+      "learning_rate": 1e-05,
+      "loss": 1.2554,
+      "step": 1516
+    },
+    {
+      "epoch": 0.8866807919924016,
+      "grad_norm": 1.66969722053932,
+      "learning_rate": 1e-05,
+      "loss": 1.1574,
+      "step": 1517
+    },
+    {
+      "epoch": 0.8872652882297071,
+      "grad_norm": 1.8662392809335355,
+      "learning_rate": 1e-05,
+      "loss": 1.1836,
+      "step": 1518
+    },
+    {
+      "epoch": 0.8878497844670125,
+      "grad_norm": 1.786452663154555,
+      "learning_rate": 1e-05,
+      "loss": 1.1535,
+      "step": 1519
+    },
+    {
+      "epoch": 0.888434280704318,
+      "grad_norm": 1.609114571341437,
+      "learning_rate": 1e-05,
+      "loss": 1.181,
+      "step": 1520
+    },
+    {
+      "epoch": 0.8890187769416235,
+      "grad_norm": 1.7629949248296455,
+      "learning_rate": 1e-05,
+      "loss": 1.2253,
+      "step": 1521
+    },
+    {
+      "epoch": 0.889603273178929,
+      "grad_norm": 1.7826200213380425,
+      "learning_rate": 1e-05,
+      "loss": 1.1835,
+      "step": 1522
+    },
+    {
+      "epoch": 0.8901877694162343,
+      "grad_norm": 1.7855922320808801,
+      "learning_rate": 1e-05,
+      "loss": 1.2082,
+      "step": 1523
+    },
+    {
+      "epoch": 0.8907722656535398,
+      "grad_norm": 1.7536629119974665,
+      "learning_rate": 1e-05,
+      "loss": 1.1671,
+      "step": 1524
+    },
+    {
+      "epoch": 0.8913567618908453,
+      "grad_norm": 1.6217788799725263,
+      "learning_rate": 1e-05,
+      "loss": 1.194,
+      "step": 1525
+    },
+    {
+      "epoch": 0.8919412581281508,
+      "grad_norm": 1.8564457752899315,
+      "learning_rate": 1e-05,
+      "loss": 1.1911,
+      "step": 1526
+    },
+    {
+      "epoch": 0.8925257543654562,
+      "grad_norm": 1.9067853077629826,
+      "learning_rate": 1e-05,
+      "loss": 1.2741,
+      "step": 1527
+    },
+    {
+      "epoch": 0.8931102506027617,
+      "grad_norm": 2.002773240616068,
+      "learning_rate": 1e-05,
+      "loss": 1.1514,
+      "step": 1528
+    },
+    {
+      "epoch": 0.8936947468400672,
+      "grad_norm": 1.9951706779061447,
+      "learning_rate": 1e-05,
+      "loss": 1.1521,
+      "step": 1529
+    },
+    {
+      "epoch": 0.8942792430773727,
+      "grad_norm": 1.8377596836354564,
+      "learning_rate": 1e-05,
+      "loss": 1.216,
+      "step": 1530
+    },
+    {
+      "epoch": 0.8948637393146781,
+      "grad_norm": 1.6861347772096185,
+      "learning_rate": 1e-05,
+      "loss": 1.2308,
+      "step": 1531
+    },
+    {
+      "epoch": 0.8954482355519836,
+      "grad_norm": 1.77258148275699,
+      "learning_rate": 1e-05,
+      "loss": 1.226,
+      "step": 1532
+    },
+    {
+      "epoch": 0.8960327317892891,
+      "grad_norm": 1.694276224178464,
+      "learning_rate": 1e-05,
+      "loss": 1.2059,
+      "step": 1533
+    },
+    {
+      "epoch": 0.8966172280265946,
+      "grad_norm": 1.6577557512135705,
+      "learning_rate": 1e-05,
+      "loss": 1.1629,
+      "step": 1534
+    },
+    {
+      "epoch": 0.8972017242639001,
+      "grad_norm": 1.7817477564009823,
+      "learning_rate": 1e-05,
+      "loss": 1.1871,
+      "step": 1535
+    },
+    {
+      "epoch": 0.8977862205012055,
+      "grad_norm": 1.7911632367306558,
+      "learning_rate": 1e-05,
+      "loss": 1.2126,
+      "step": 1536
+    },
+    {
+      "epoch": 0.898370716738511,
+      "grad_norm": 1.847149739355391,
+      "learning_rate": 1e-05,
+      "loss": 1.199,
+      "step": 1537
+    },
+    {
+      "epoch": 0.8989552129758165,
+      "grad_norm": 1.7724478102978882,
+      "learning_rate": 1e-05,
+      "loss": 1.2183,
+      "step": 1538
+    },
+    {
+      "epoch": 0.899539709213122,
+      "grad_norm": 1.6561962276286235,
+      "learning_rate": 1e-05,
+      "loss": 1.1871,
+      "step": 1539
+    },
+    {
+      "epoch": 0.9001242054504274,
+      "grad_norm": 1.9632724978374188,
+      "learning_rate": 1e-05,
+      "loss": 1.248,
+      "step": 1540
+    },
+    {
+      "epoch": 0.9007087016877329,
+      "grad_norm": 1.7862597856166642,
+      "learning_rate": 1e-05,
+      "loss": 1.2356,
+      "step": 1541
+    },
+    {
+      "epoch": 0.9012931979250384,
+      "grad_norm": 2.0590736057381585,
+      "learning_rate": 1e-05,
+      "loss": 1.1789,
+      "step": 1542
+    },
+    {
+      "epoch": 0.9018776941623439,
+      "grad_norm": 1.6584884845161934,
+      "learning_rate": 1e-05,
+      "loss": 1.1869,
+      "step": 1543
+    },
+    {
+      "epoch": 0.9024621903996493,
+      "grad_norm": 1.8254291278532877,
+      "learning_rate": 1e-05,
+      "loss": 1.1445,
+      "step": 1544
+    },
+    {
+      "epoch": 0.9030466866369548,
+      "grad_norm": 1.7186489264665739,
+      "learning_rate": 1e-05,
+      "loss": 1.242,
+      "step": 1545
+    },
+    {
+      "epoch": 0.9036311828742603,
+      "grad_norm": 1.8191200522845,
+      "learning_rate": 1e-05,
+      "loss": 1.2744,
+      "step": 1546
+    },
+    {
+      "epoch": 0.9042156791115658,
+      "grad_norm": 1.7629604193670823,
+      "learning_rate": 1e-05,
+      "loss": 1.1793,
+      "step": 1547
+    },
+    {
+      "epoch": 0.9048001753488711,
+      "grad_norm": 1.8258507188430813,
+      "learning_rate": 1e-05,
+      "loss": 1.1604,
+      "step": 1548
+    },
+    {
+      "epoch": 0.9053846715861766,
+      "grad_norm": 1.7167444487696069,
+      "learning_rate": 1e-05,
+      "loss": 1.1955,
+      "step": 1549
+    },
+    {
+      "epoch": 0.9059691678234821,
+      "grad_norm": 1.7202343068236075,
+      "learning_rate": 1e-05,
+      "loss": 1.1796,
+      "step": 1550
+    },
+    {
+      "epoch": 0.9065536640607876,
+      "grad_norm": 1.7612301097942817,
+      "learning_rate": 1e-05,
+      "loss": 1.1868,
+      "step": 1551
+    },
+    {
+      "epoch": 0.9071381602980931,
+      "grad_norm": 1.733376373797865,
+      "learning_rate": 1e-05,
+      "loss": 1.2079,
+      "step": 1552
+    },
+    {
+      "epoch": 0.9077226565353985,
+      "grad_norm": 1.7480514554724753,
+      "learning_rate": 1e-05,
+      "loss": 1.209,
+      "step": 1553
+    },
+    {
+      "epoch": 0.908307152772704,
+      "grad_norm": 1.7410740118704697,
+      "learning_rate": 1e-05,
+      "loss": 1.2132,
+      "step": 1554
+    },
+    {
+      "epoch": 0.9088916490100095,
+      "grad_norm": 1.6676897811800806,
+      "learning_rate": 1e-05,
+      "loss": 1.1811,
+      "step": 1555
+    },
+    {
+      "epoch": 0.909476145247315,
+      "grad_norm": 1.917350182182316,
+      "learning_rate": 1e-05,
+      "loss": 1.2021,
+      "step": 1556
+    },
+    {
+      "epoch": 0.9100606414846204,
+      "grad_norm": 1.9135898782055463,
+      "learning_rate": 1e-05,
+      "loss": 1.2023,
+      "step": 1557
+    },
+    {
+      "epoch": 0.9106451377219259,
+      "grad_norm": 1.8177452492525252,
+      "learning_rate": 1e-05,
+      "loss": 1.1587,
+      "step": 1558
+    },
+    {
+      "epoch": 0.9112296339592314,
+      "grad_norm": 1.6933062754760786,
+      "learning_rate": 1e-05,
+      "loss": 1.222,
+      "step": 1559
+    },
+    {
+      "epoch": 0.9118141301965369,
+      "grad_norm": 1.649612171791171,
+      "learning_rate": 1e-05,
+      "loss": 1.2014,
+      "step": 1560
+    },
+    {
+      "epoch": 0.9123986264338423,
+      "grad_norm": 1.687378415104643,
+      "learning_rate": 1e-05,
+      "loss": 1.239,
+      "step": 1561
+    },
+    {
+      "epoch": 0.9129831226711478,
+      "grad_norm": 1.6112986049469233,
+      "learning_rate": 1e-05,
+      "loss": 1.2241,
+      "step": 1562
+    },
+    {
+      "epoch": 0.9135676189084533,
+      "grad_norm": 1.6243173506610908,
+      "learning_rate": 1e-05,
+      "loss": 1.1971,
+      "step": 1563
+    },
+    {
+      "epoch": 0.9141521151457588,
+      "grad_norm": 1.7401528723316053,
+      "learning_rate": 1e-05,
+      "loss": 1.2163,
+      "step": 1564
+    },
+    {
+      "epoch": 0.9147366113830642,
+      "grad_norm": 1.8258188960018635,
+      "learning_rate": 1e-05,
+      "loss": 1.2277,
+      "step": 1565
+    },
+    {
+      "epoch": 0.9153211076203697,
+      "grad_norm": 1.8561276609663073,
+      "learning_rate": 1e-05,
+      "loss": 1.1385,
+      "step": 1566
+    },
+    {
+      "epoch": 0.9159056038576752,
+      "grad_norm": 1.634527996063688,
+      "learning_rate": 1e-05,
+      "loss": 1.2481,
+      "step": 1567
+    },
+    {
+      "epoch": 0.9164901000949807,
+      "grad_norm": 1.765039557221531,
+      "learning_rate": 1e-05,
+      "loss": 1.1624,
+      "step": 1568
+    },
+    {
+      "epoch": 0.9170745963322862,
+      "grad_norm": 1.7519448604280983,
+      "learning_rate": 1e-05,
+      "loss": 1.2113,
+      "step": 1569
+    },
+    {
+      "epoch": 0.9176590925695916,
+      "grad_norm": 2.0686593172853636,
+      "learning_rate": 1e-05,
+      "loss": 1.2111,
+      "step": 1570
+    },
+    {
+      "epoch": 0.918243588806897,
+      "grad_norm": 1.9185621410865406,
+      "learning_rate": 1e-05,
+      "loss": 1.2306,
+      "step": 1571
+    },
+    {
+      "epoch": 0.9188280850442025,
+      "grad_norm": 1.827584682952987,
+      "learning_rate": 1e-05,
+      "loss": 1.2544,
+      "step": 1572
+    },
+    {
+      "epoch": 0.919412581281508,
+      "grad_norm": 1.756405991481828,
+      "learning_rate": 1e-05,
+      "loss": 1.1802,
+      "step": 1573
+    },
+    {
+      "epoch": 0.9199970775188134,
+      "grad_norm": 1.6982412876652253,
+      "learning_rate": 1e-05,
+      "loss": 1.2118,
+      "step": 1574
+    },
+    {
+      "epoch": 0.9205815737561189,
+      "grad_norm": 1.6670679655642453,
+      "learning_rate": 1e-05,
+      "loss": 1.1891,
+      "step": 1575
+    },
+    {
+      "epoch": 0.9211660699934244,
+      "grad_norm": 1.662399606238649,
+      "learning_rate": 1e-05,
+      "loss": 1.1874,
+      "step": 1576
+    },
+    {
+      "epoch": 0.9217505662307299,
+      "grad_norm": 1.8098581583231792,
+      "learning_rate": 1e-05,
+      "loss": 1.231,
+      "step": 1577
+    },
+    {
+      "epoch": 0.9223350624680353,
+      "grad_norm": 1.6249756923215115,
+      "learning_rate": 1e-05,
+      "loss": 1.2109,
+      "step": 1578
+    },
+    {
+      "epoch": 0.9229195587053408,
+      "grad_norm": 1.6734052140044278,
+      "learning_rate": 1e-05,
+      "loss": 1.2111,
+      "step": 1579
+    },
+    {
+      "epoch": 0.9235040549426463,
+      "grad_norm": 1.7840154091191656,
+      "learning_rate": 1e-05,
+      "loss": 1.1704,
+      "step": 1580
+    },
+    {
+      "epoch": 0.9240885511799518,
+      "grad_norm": 1.6805556666556074,
+      "learning_rate": 1e-05,
+      "loss": 1.1994,
+      "step": 1581
+    },
+    {
+      "epoch": 0.9246730474172572,
+      "grad_norm": 1.6968210056812834,
+      "learning_rate": 1e-05,
+      "loss": 1.2131,
+      "step": 1582
+    },
+    {
+      "epoch": 0.9252575436545627,
+      "grad_norm": 1.6883202411719807,
+      "learning_rate": 1e-05,
+      "loss": 1.1849,
+      "step": 1583
+    },
+    {
+      "epoch": 0.9258420398918682,
+      "grad_norm": 1.8720560664906865,
+      "learning_rate": 1e-05,
+      "loss": 1.2338,
+      "step": 1584
+    },
+    {
+      "epoch": 0.9264265361291737,
+      "grad_norm": 1.7068094525260242,
+      "learning_rate": 1e-05,
+      "loss": 1.1446,
+      "step": 1585
+    },
+    {
+      "epoch": 0.9270110323664792,
+      "grad_norm": 1.6160136423251923,
+      "learning_rate": 1e-05,
+      "loss": 1.1609,
+      "step": 1586
+    },
+    {
+      "epoch": 0.9275955286037846,
+      "grad_norm": 1.7555269063981858,
+      "learning_rate": 1e-05,
+      "loss": 1.2237,
+      "step": 1587
+    },
+    {
+      "epoch": 0.9281800248410901,
+      "grad_norm": 1.8301406665398008,
+      "learning_rate": 1e-05,
+      "loss": 1.1738,
+      "step": 1588
+    },
+    {
+      "epoch": 0.9287645210783956,
+      "grad_norm": 1.972893377114612,
+      "learning_rate": 1e-05,
+      "loss": 1.186,
+      "step": 1589
+    },
+    {
+      "epoch": 0.9293490173157011,
+      "grad_norm": 1.8002422111053789,
+      "learning_rate": 1e-05,
+      "loss": 1.1522,
+      "step": 1590
+    },
+    {
+      "epoch": 0.9299335135530065,
+      "grad_norm": 1.7990476991408504,
+      "learning_rate": 1e-05,
+      "loss": 1.1656,
+      "step": 1591
+    },
+    {
+      "epoch": 0.930518009790312,
+      "grad_norm": 1.7577788500412894,
+      "learning_rate": 1e-05,
+      "loss": 1.2132,
+      "step": 1592
+    },
+    {
+      "epoch": 0.9311025060276175,
+      "grad_norm": 1.6964683749270157,
+      "learning_rate": 1e-05,
+      "loss": 1.17,
+      "step": 1593
+    },
+    {
+      "epoch": 0.931687002264923,
+      "grad_norm": 1.774977883444026,
+      "learning_rate": 1e-05,
+      "loss": 1.1749,
+      "step": 1594
+    },
+    {
+      "epoch": 0.9322714985022283,
+      "grad_norm": 1.6750779118066412,
+      "learning_rate": 1e-05,
+      "loss": 1.2247,
+      "step": 1595
+    },
+    {
+      "epoch": 0.9328559947395338,
+      "grad_norm": 1.7534349384350585,
+      "learning_rate": 1e-05,
+      "loss": 1.1894,
+      "step": 1596
+    },
+    {
+      "epoch": 0.9334404909768393,
+      "grad_norm": 1.8116106160777448,
+      "learning_rate": 1e-05,
+      "loss": 1.217,
+      "step": 1597
+    },
+    {
+      "epoch": 0.9340249872141448,
+      "grad_norm": 1.7111422505896219,
+      "learning_rate": 1e-05,
+      "loss": 1.1961,
+      "step": 1598
+    },
+    {
+      "epoch": 0.9346094834514502,
+      "grad_norm": 1.6499846244196101,
+      "learning_rate": 1e-05,
+      "loss": 1.2078,
+      "step": 1599
+    },
+    {
+      "epoch": 0.9351939796887557,
+      "grad_norm": 1.6776434241850748,
+      "learning_rate": 1e-05,
+      "loss": 1.2319,
+      "step": 1600
+    },
+    {
+      "epoch": 0.9357784759260612,
+      "grad_norm": 1.7555947307191049,
+      "learning_rate": 1e-05,
+      "loss": 1.1802,
+      "step": 1601
+    },
+    {
+      "epoch": 0.9363629721633667,
+      "grad_norm": 1.5978819743743424,
+      "learning_rate": 1e-05,
+      "loss": 1.1998,
+      "step": 1602
+    },
+    {
+      "epoch": 0.9369474684006722,
+      "grad_norm": 1.756398299835865,
+      "learning_rate": 1e-05,
+      "loss": 1.1939,
+      "step": 1603
+    },
+    {
+      "epoch": 0.9375319646379776,
+      "grad_norm": 1.8101686750346215,
+      "learning_rate": 1e-05,
+      "loss": 1.1782,
+      "step": 1604
+    },
+    {
+      "epoch": 0.9381164608752831,
+      "grad_norm": 1.8367090452789894,
+      "learning_rate": 1e-05,
+      "loss": 1.1922,
+      "step": 1605
+    },
+    {
+      "epoch": 0.9387009571125886,
+      "grad_norm": 1.9609883197890083,
+      "learning_rate": 1e-05,
+      "loss": 1.1611,
+      "step": 1606
+    },
+    {
+      "epoch": 0.9392854533498941,
+      "grad_norm": 1.7246079687777867,
+      "learning_rate": 1e-05,
+      "loss": 1.166,
+      "step": 1607
+    },
+    {
+      "epoch": 0.9398699495871995,
+      "grad_norm": 1.7705828333142812,
+      "learning_rate": 1e-05,
+      "loss": 1.2037,
+      "step": 1608
+    },
+    {
+      "epoch": 0.940454445824505,
+      "grad_norm": 1.7868850205936149,
+      "learning_rate": 1e-05,
+      "loss": 1.1673,
+      "step": 1609
+    },
+    {
+      "epoch": 0.9410389420618105,
+      "grad_norm": 1.6616957307274993,
+      "learning_rate": 1e-05,
+      "loss": 1.1956,
+      "step": 1610
+    },
+    {
+      "epoch": 0.941623438299116,
+      "grad_norm": 1.6918380615998256,
+      "learning_rate": 1e-05,
+      "loss": 1.1958,
+      "step": 1611
+    },
+    {
+      "epoch": 0.9422079345364214,
+      "grad_norm": 1.7842070915390702,
+      "learning_rate": 1e-05,
+      "loss": 1.183,
+      "step": 1612
+    },
+    {
+      "epoch": 0.9427924307737269,
+      "grad_norm": 1.5460219717795773,
+      "learning_rate": 1e-05,
+      "loss": 1.213,
+      "step": 1613
+    },
+    {
+      "epoch": 0.9433769270110324,
+      "grad_norm": 1.8760980018373106,
+      "learning_rate": 1e-05,
+      "loss": 1.2159,
+      "step": 1614
+    },
+    {
+      "epoch": 0.9439614232483379,
+      "grad_norm": 1.7295987899624417,
+      "learning_rate": 1e-05,
+      "loss": 1.2196,
+      "step": 1615
+    },
+    {
+      "epoch": 0.9445459194856433,
+      "grad_norm": 1.8362364660851174,
+      "learning_rate": 1e-05,
+      "loss": 1.2024,
+      "step": 1616
+    },
+    {
+      "epoch": 0.9451304157229488,
+      "grad_norm": 1.7679050909077925,
+      "learning_rate": 1e-05,
+      "loss": 1.1837,
+      "step": 1617
+    },
+    {
+      "epoch": 0.9457149119602543,
+      "grad_norm": 1.7000142207762792,
+      "learning_rate": 1e-05,
+      "loss": 1.2127,
+      "step": 1618
+    },
+    {
+      "epoch": 0.9462994081975598,
+      "grad_norm": 1.540461766161839,
+      "learning_rate": 1e-05,
+      "loss": 1.2304,
+      "step": 1619
+    },
+    {
+      "epoch": 0.9468839044348653,
+      "grad_norm": 1.7275614094166578,
+      "learning_rate": 1e-05,
+      "loss": 1.1718,
+      "step": 1620
+    },
+    {
+      "epoch": 0.9474684006721706,
+      "grad_norm": 1.721359482635086,
+      "learning_rate": 1e-05,
+      "loss": 1.1948,
+      "step": 1621
+    },
+    {
+      "epoch": 0.9480528969094761,
+      "grad_norm": 1.5641267924099658,
+      "learning_rate": 1e-05,
+      "loss": 1.1898,
+      "step": 1622
+    },
+    {
+      "epoch": 0.9486373931467816,
+      "grad_norm": 1.6471546827331278,
+      "learning_rate": 1e-05,
+      "loss": 1.1727,
+      "step": 1623
+    },
+    {
+      "epoch": 0.9492218893840871,
+      "grad_norm": 1.7342504326556558,
+      "learning_rate": 1e-05,
+      "loss": 1.2661,
+      "step": 1624
+    },
+    {
+      "epoch": 0.9498063856213925,
+      "grad_norm": 1.6210840157637734,
+      "learning_rate": 1e-05,
+      "loss": 1.1894,
+      "step": 1625
+    },
+    {
+      "epoch": 0.950390881858698,
+      "grad_norm": 1.7355973990781115,
+      "learning_rate": 1e-05,
+      "loss": 1.1947,
+      "step": 1626
+    },
+    {
+      "epoch": 0.9509753780960035,
+      "grad_norm": 1.6955967054828527,
+      "learning_rate": 1e-05,
+      "loss": 1.1489,
+      "step": 1627
+    },
+    {
+      "epoch": 0.951559874333309,
+      "grad_norm": 1.636645132192986,
+      "learning_rate": 1e-05,
+      "loss": 1.1762,
+      "step": 1628
+    },
+    {
+      "epoch": 0.9521443705706144,
+      "grad_norm": 1.7192252963533257,
+      "learning_rate": 1e-05,
+      "loss": 1.2309,
+      "step": 1629
+    },
+    {
+      "epoch": 0.9527288668079199,
+      "grad_norm": 1.6031192378494148,
+      "learning_rate": 1e-05,
+      "loss": 1.1373,
+      "step": 1630
+    },
+    {
+      "epoch": 0.9533133630452254,
+      "grad_norm": 1.6061839500907045,
+      "learning_rate": 1e-05,
+      "loss": 1.2281,
+      "step": 1631
+    },
+    {
+      "epoch": 0.9538978592825309,
+      "grad_norm": 1.7723821888130855,
+      "learning_rate": 1e-05,
+      "loss": 1.1746,
+      "step": 1632
+    },
+    {
+      "epoch": 0.9544823555198363,
+      "grad_norm": 1.6666351279717988,
+      "learning_rate": 1e-05,
+      "loss": 1.1916,
+      "step": 1633
+    },
+    {
+      "epoch": 0.9550668517571418,
+      "grad_norm": 1.7502882725962072,
+      "learning_rate": 1e-05,
+      "loss": 1.1968,
+      "step": 1634
+    },
+    {
+      "epoch": 0.9556513479944473,
+      "grad_norm": 1.5670414700988906,
+      "learning_rate": 1e-05,
+      "loss": 1.1953,
+      "step": 1635
+    },
+    {
+      "epoch": 0.9562358442317528,
+      "grad_norm": 2.029027043004617,
+      "learning_rate": 1e-05,
+      "loss": 1.1599,
+      "step": 1636
+    },
+    {
+      "epoch": 0.9568203404690583,
+      "grad_norm": 1.6286618844169378,
+      "learning_rate": 1e-05,
+      "loss": 1.2158,
+      "step": 1637
+    },
+    {
+      "epoch": 0.9574048367063637,
+      "grad_norm": 1.657724248567447,
+      "learning_rate": 1e-05,
+      "loss": 1.2083,
+      "step": 1638
+    },
+    {
+      "epoch": 0.9579893329436692,
+      "grad_norm": 1.8434349212766747,
+      "learning_rate": 1e-05,
+      "loss": 1.1669,
+      "step": 1639
+    },
+    {
+      "epoch": 0.9585738291809747,
+      "grad_norm": 1.5843915556766088,
+      "learning_rate": 1e-05,
+      "loss": 1.2229,
+      "step": 1640
+    },
+    {
+      "epoch": 0.9591583254182802,
+      "grad_norm": 1.7482988413995146,
+      "learning_rate": 1e-05,
+      "loss": 1.1533,
+      "step": 1641
+    },
+    {
+      "epoch": 0.9597428216555856,
+      "grad_norm": 1.7405759766154743,
+      "learning_rate": 1e-05,
+      "loss": 1.1545,
+      "step": 1642
+    },
+    {
+      "epoch": 0.960327317892891,
+      "grad_norm": 1.7077609820717727,
+      "learning_rate": 1e-05,
+      "loss": 1.1963,
+      "step": 1643
+    },
+    {
+      "epoch": 0.9609118141301966,
+      "grad_norm": 1.5681675337346077,
+      "learning_rate": 1e-05,
+      "loss": 1.1876,
+      "step": 1644
+    },
+    {
+      "epoch": 0.961496310367502,
+      "grad_norm": 1.6627577454497493,
+      "learning_rate": 1e-05,
+      "loss": 1.1721,
+      "step": 1645
+    },
+    {
+      "epoch": 0.9620808066048074,
+      "grad_norm": 1.6842626351275243,
+      "learning_rate": 1e-05,
+      "loss": 1.2109,
+      "step": 1646
+    },
+    {
+      "epoch": 0.9626653028421129,
+      "grad_norm": 1.649166402367081,
+      "learning_rate": 1e-05,
+      "loss": 1.1622,
+      "step": 1647
+    },
+    {
+      "epoch": 0.9632497990794184,
+      "grad_norm": 1.7060224540481428,
+      "learning_rate": 1e-05,
+      "loss": 1.2099,
+      "step": 1648
+    },
+    {
+      "epoch": 0.9638342953167239,
+      "grad_norm": 1.8266787524575958,
+      "learning_rate": 1e-05,
+      "loss": 1.1822,
+      "step": 1649
+    },
+    {
+      "epoch": 0.9644187915540293,
+      "grad_norm": 1.6524626287021558,
+      "learning_rate": 1e-05,
+      "loss": 1.1519,
+      "step": 1650
+    },
+    {
+      "epoch": 0.9650032877913348,
+      "grad_norm": 1.657767346606179,
+      "learning_rate": 1e-05,
+      "loss": 1.1871,
+      "step": 1651
+    },
+    {
+      "epoch": 0.9655877840286403,
+      "grad_norm": 1.919115726585691,
+      "learning_rate": 1e-05,
+      "loss": 1.1806,
+      "step": 1652
+    },
+    {
+      "epoch": 0.9661722802659458,
+      "grad_norm": 1.8643008351103663,
+      "learning_rate": 1e-05,
+      "loss": 1.2167,
+      "step": 1653
+    },
+    {
+      "epoch": 0.9667567765032513,
+      "grad_norm": 1.5904612573531929,
+      "learning_rate": 1e-05,
+      "loss": 1.2557,
+      "step": 1654
+    },
+    {
+      "epoch": 0.9673412727405567,
+      "grad_norm": 1.7023973505942323,
+      "learning_rate": 1e-05,
+      "loss": 1.1986,
+      "step": 1655
+    },
+    {
+      "epoch": 0.9679257689778622,
+      "grad_norm": 1.8558390884338123,
+      "learning_rate": 1e-05,
+      "loss": 1.2216,
+      "step": 1656
+    },
+    {
+      "epoch": 0.9685102652151677,
+      "grad_norm": 1.7569795360514773,
+      "learning_rate": 1e-05,
+      "loss": 1.2423,
+      "step": 1657
+    },
+    {
+      "epoch": 0.9690947614524732,
+      "grad_norm": 1.8343219902790653,
+      "learning_rate": 1e-05,
+      "loss": 1.232,
+      "step": 1658
+    },
+    {
+      "epoch": 0.9696792576897786,
+      "grad_norm": 1.8174531982167323,
+      "learning_rate": 1e-05,
+      "loss": 1.1755,
+      "step": 1659
+    },
+    {
+      "epoch": 0.9702637539270841,
+      "grad_norm": 1.544198845647354,
+      "learning_rate": 1e-05,
+      "loss": 1.1616,
+      "step": 1660
+    },
+    {
+      "epoch": 0.9708482501643896,
+      "grad_norm": 1.7237387405392666,
+      "learning_rate": 1e-05,
+      "loss": 1.2017,
+      "step": 1661
+    },
+    {
+      "epoch": 0.9714327464016951,
+      "grad_norm": 1.9108875255728657,
+      "learning_rate": 1e-05,
+      "loss": 1.1793,
+      "step": 1662
+    },
+    {
+      "epoch": 0.9720172426390005,
+      "grad_norm": 1.5802604118972206,
+      "learning_rate": 1e-05,
+      "loss": 1.2271,
+      "step": 1663
+    },
+    {
+      "epoch": 0.972601738876306,
+      "grad_norm": 1.6364552984507852,
+      "learning_rate": 1e-05,
+      "loss": 1.1791,
+      "step": 1664
+    },
+    {
+      "epoch": 0.9731862351136115,
+      "grad_norm": 1.6331100542799128,
+      "learning_rate": 1e-05,
+      "loss": 1.1931,
+      "step": 1665
+    },
+    {
+      "epoch": 0.973770731350917,
+      "grad_norm": 1.5337871118122632,
+      "learning_rate": 1e-05,
+      "loss": 1.1774,
+      "step": 1666
+    },
+    {
+      "epoch": 0.9743552275882224,
+      "grad_norm": 1.73409818774021,
+      "learning_rate": 1e-05,
+      "loss": 1.1673,
+      "step": 1667
+    },
+    {
+      "epoch": 0.9749397238255278,
+      "grad_norm": 1.7422936671001887,
+      "learning_rate": 1e-05,
+      "loss": 1.1847,
+      "step": 1668
+    },
+    {
+      "epoch": 0.9755242200628333,
+      "grad_norm": 1.7730411594098248,
+      "learning_rate": 1e-05,
+      "loss": 1.2572,
+      "step": 1669
+    },
+    {
+      "epoch": 0.9761087163001388,
+      "grad_norm": 1.7187452539160741,
+      "learning_rate": 1e-05,
+      "loss": 1.1777,
+      "step": 1670
+    },
+    {
+      "epoch": 0.9766932125374443,
+      "grad_norm": 1.7674086328167722,
+      "learning_rate": 1e-05,
+      "loss": 1.1905,
+      "step": 1671
+    },
+    {
+      "epoch": 0.9772777087747497,
+      "grad_norm": 1.6794295963883485,
+      "learning_rate": 1e-05,
+      "loss": 1.2033,
+      "step": 1672
+    },
+    {
+      "epoch": 0.9778622050120552,
+      "grad_norm": 1.9360254933592387,
+      "learning_rate": 1e-05,
+      "loss": 1.1694,
+      "step": 1673
+    },
+    {
+      "epoch": 0.9784467012493607,
+      "grad_norm": 1.6669976675433884,
+      "learning_rate": 1e-05,
+      "loss": 1.1915,
+      "step": 1674
+    },
+    {
+      "epoch": 0.9790311974866662,
+      "grad_norm": 1.5801037657595254,
+      "learning_rate": 1e-05,
+      "loss": 1.184,
+      "step": 1675
+    },
+    {
+      "epoch": 0.9796156937239716,
+      "grad_norm": 1.78580582508938,
+      "learning_rate": 1e-05,
+      "loss": 1.1913,
+      "step": 1676
+    },
+    {
+      "epoch": 0.9802001899612771,
+      "grad_norm": 1.9138785723653597,
+      "learning_rate": 1e-05,
+      "loss": 1.2033,
+      "step": 1677
+    },
+    {
+      "epoch": 0.9807846861985826,
+      "grad_norm": 1.7522172024637206,
+      "learning_rate": 1e-05,
+      "loss": 1.1734,
+      "step": 1678
+    },
+    {
+      "epoch": 0.9813691824358881,
+      "grad_norm": 1.714065155444966,
+      "learning_rate": 1e-05,
+      "loss": 1.2335,
+      "step": 1679
+    },
+    {
+      "epoch": 0.9819536786731935,
+      "grad_norm": 1.6451326836715672,
+      "learning_rate": 1e-05,
+      "loss": 1.1695,
+      "step": 1680
+    },
+    {
+      "epoch": 0.982538174910499,
+      "grad_norm": 1.5582861688740288,
+      "learning_rate": 1e-05,
+      "loss": 1.1642,
+      "step": 1681
+    },
+    {
+      "epoch": 0.9831226711478045,
+      "grad_norm": 1.8591994286387752,
+      "learning_rate": 1e-05,
+      "loss": 1.1933,
+      "step": 1682
+    },
+    {
+      "epoch": 0.98370716738511,
+      "grad_norm": 1.7652271529114363,
+      "learning_rate": 1e-05,
+      "loss": 1.2167,
+      "step": 1683
+    },
+    {
+      "epoch": 0.9842916636224154,
+      "grad_norm": 1.7271931571105024,
+      "learning_rate": 1e-05,
+      "loss": 1.2419,
+      "step": 1684
+    },
+    {
+      "epoch": 0.9848761598597209,
+      "grad_norm": 1.4377545314287328,
+      "learning_rate": 1e-05,
+      "loss": 1.1666,
+      "step": 1685
+    },
+    {
+      "epoch": 0.9854606560970264,
+      "grad_norm": 1.5709719567812213,
+      "learning_rate": 1e-05,
+      "loss": 1.2386,
+      "step": 1686
+    },
+    {
+      "epoch": 0.9860451523343319,
+      "grad_norm": 1.6961975777595968,
+      "learning_rate": 1e-05,
+      "loss": 1.2008,
+      "step": 1687
+    },
+    {
+      "epoch": 0.9866296485716374,
+      "grad_norm": 1.6979346414611487,
+      "learning_rate": 1e-05,
+      "loss": 1.2029,
+      "step": 1688
+    },
+    {
+      "epoch": 0.9872141448089428,
+      "grad_norm": 1.7227452704769806,
+      "learning_rate": 1e-05,
+      "loss": 1.1524,
+      "step": 1689
+    },
+    {
+      "epoch": 0.9877986410462483,
+      "grad_norm": 1.7751521903978762,
+      "learning_rate": 1e-05,
+      "loss": 1.2307,
+      "step": 1690
+    },
+    {
+      "epoch": 0.9883831372835538,
+      "grad_norm": 1.7977896803110063,
+      "learning_rate": 1e-05,
+      "loss": 1.1784,
+      "step": 1691
+    },
+    {
+      "epoch": 0.9889676335208593,
+      "grad_norm": 1.8415510067465632,
+      "learning_rate": 1e-05,
+      "loss": 1.2221,
+      "step": 1692
+    },
+    {
+      "epoch": 0.9895521297581646,
+      "grad_norm": 1.7247836945654178,
+      "learning_rate": 1e-05,
+      "loss": 1.1835,
+      "step": 1693
+    },
+    {
+      "epoch": 0.9901366259954701,
+      "grad_norm": 1.750283984427814,
+      "learning_rate": 1e-05,
+      "loss": 1.1987,
+      "step": 1694
+    },
+    {
+      "epoch": 0.9907211222327756,
+      "grad_norm": 1.749488002924086,
+      "learning_rate": 1e-05,
+      "loss": 1.2279,
+      "step": 1695
+    },
+    {
+      "epoch": 0.9913056184700811,
+      "grad_norm": 1.7152546214127649,
+      "learning_rate": 1e-05,
+      "loss": 1.1366,
+      "step": 1696
+    },
+    {
+      "epoch": 0.9918901147073865,
+      "grad_norm": 1.6479372023195673,
+      "learning_rate": 1e-05,
+      "loss": 1.1632,
+      "step": 1697
+    },
+    {
+      "epoch": 0.992474610944692,
+      "grad_norm": 1.8278281149593754,
+      "learning_rate": 1e-05,
+      "loss": 1.1928,
+      "step": 1698
+    },
+    {
+      "epoch": 0.9930591071819975,
+      "grad_norm": 1.5480083946013776,
+      "learning_rate": 1e-05,
+      "loss": 1.2006,
+      "step": 1699
+    },
+    {
+      "epoch": 0.993643603419303,
+      "grad_norm": 1.7963414485687306,
+      "learning_rate": 1e-05,
+      "loss": 1.1462,
+      "step": 1700
+    },
+    {
+      "epoch": 0.9942280996566084,
+      "grad_norm": 1.856016675177229,
+      "learning_rate": 1e-05,
+      "loss": 1.1666,
+      "step": 1701
+    },
+    {
+      "epoch": 0.9948125958939139,
+      "grad_norm": 1.7557833220839716,
+      "learning_rate": 1e-05,
+      "loss": 1.2103,
+      "step": 1702
+    },
+    {
+      "epoch": 0.9953970921312194,
+      "grad_norm": 1.7669302831688125,
+      "learning_rate": 1e-05,
+      "loss": 1.1961,
+      "step": 1703
+    },
+    {
+      "epoch": 0.9959815883685249,
+      "grad_norm": 1.8319550494315788,
+      "learning_rate": 1e-05,
+      "loss": 1.1953,
+      "step": 1704
+    },
+    {
+      "epoch": 0.9965660846058304,
+      "grad_norm": 1.7417184930651384,
+      "learning_rate": 1e-05,
+      "loss": 1.1771,
+      "step": 1705
+    },
+    {
+      "epoch": 0.9971505808431358,
+      "grad_norm": 1.5849722230874825,
+      "learning_rate": 1e-05,
+      "loss": 1.194,
+      "step": 1706
+    },
+    {
+      "epoch": 0.9977350770804413,
+      "grad_norm": 1.8091580114065404,
+      "learning_rate": 1e-05,
+      "loss": 1.2121,
+      "step": 1707
+    },
+    {
+      "epoch": 0.9983195733177468,
+      "grad_norm": 1.755106482112034,
+      "learning_rate": 1e-05,
+      "loss": 1.1896,
+      "step": 1708
+    },
+    {
+      "epoch": 0.9989040695550523,
+      "grad_norm": 1.7018192909609253,
+      "learning_rate": 1e-05,
+      "loss": 1.1746,
+      "step": 1709
+    },
+    {
+      "epoch": 0.9994885657923577,
+      "grad_norm": 1.6485292105949847,
+      "learning_rate": 1e-05,
+      "loss": 1.1914,
+      "step": 1710
+    },
+    {
+      "epoch": 0.9994885657923577,
+      "step": 1710,
+      "total_flos": 2767817350840320.0,
+      "train_loss": 1.2463982903469375,
+      "train_runtime": 242916.4328,
+      "train_samples_per_second": 0.901,
+      "train_steps_per_second": 0.007
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1710,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 10,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2767817350840320.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}