diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6748 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 958,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0020876826722338203,
+      "grad_norm": 0.8805100452450126,
+      "learning_rate": 1.3793103448275863e-05,
+      "loss": 3.4375,
+      "step": 1
+    },
+    {
+      "epoch": 0.0041753653444676405,
+      "grad_norm": 0.8962056177448029,
+      "learning_rate": 2.7586206896551727e-05,
+      "loss": 3.4531,
+      "step": 2
+    },
+    {
+      "epoch": 0.006263048016701462,
+      "grad_norm": 0.8975533488683494,
+      "learning_rate": 4.1379310344827587e-05,
+      "loss": 3.375,
+      "step": 3
+    },
+    {
+      "epoch": 0.008350730688935281,
+      "grad_norm": 0.8562154080719621,
+      "learning_rate": 5.517241379310345e-05,
+      "loss": 3.0312,
+      "step": 4
+    },
+    {
+      "epoch": 0.010438413361169102,
+      "grad_norm": 0.7722844092130382,
+      "learning_rate": 6.896551724137931e-05,
+      "loss": 2.0938,
+      "step": 5
+    },
+    {
+      "epoch": 0.012526096033402923,
+      "grad_norm": 0.5469423252625637,
+      "learning_rate": 8.275862068965517e-05,
+      "loss": 1.5391,
+      "step": 6
+    },
+    {
+      "epoch": 0.014613778705636743,
+      "grad_norm": 0.269527877565377,
+      "learning_rate": 9.655172413793105e-05,
+      "loss": 1.3281,
+      "step": 7
+    },
+    {
+      "epoch": 0.016701461377870562,
+      "grad_norm": 0.2129561647095634,
+      "learning_rate": 0.0001103448275862069,
+      "loss": 1.0547,
+      "step": 8
+    },
+    {
+      "epoch": 0.018789144050104383,
+      "grad_norm": 0.209174835903919,
+      "learning_rate": 0.00012413793103448277,
+      "loss": 0.8672,
+      "step": 9
+    },
+    {
+      "epoch": 0.020876826722338204,
+      "grad_norm": 0.16723166688980518,
+      "learning_rate": 0.00013793103448275863,
+      "loss": 0.7461,
+      "step": 10
+    },
+    {
+      "epoch": 0.022964509394572025,
+      "grad_norm": 0.13523071310549678,
+      "learning_rate": 0.00015172413793103449,
+      "loss": 0.75,
+      "step": 11
+    },
+    {
+      "epoch": 0.025052192066805846,
+      "grad_norm": 0.13783786374844356,
+      "learning_rate": 0.00016551724137931035,
+      "loss": 0.7578,
+      "step": 12
+    },
+    {
+      "epoch": 0.027139874739039668,
+      "grad_norm": 0.11159117473743674,
+      "learning_rate": 0.0001793103448275862,
+      "loss": 0.6914,
+      "step": 13
+    },
+    {
+      "epoch": 0.029227557411273485,
+      "grad_norm": 0.10486826556036052,
+      "learning_rate": 0.0001931034482758621,
+      "loss": 0.6484,
+      "step": 14
+    },
+    {
+      "epoch": 0.031315240083507306,
+      "grad_norm": 0.1012538060823961,
+      "learning_rate": 0.00020689655172413795,
+      "loss": 0.6562,
+      "step": 15
+    },
+    {
+      "epoch": 0.033402922755741124,
+      "grad_norm": 0.10310591282634196,
+      "learning_rate": 0.0002206896551724138,
+      "loss": 0.6602,
+      "step": 16
+    },
+    {
+      "epoch": 0.03549060542797495,
+      "grad_norm": 0.07838991965173675,
+      "learning_rate": 0.00023448275862068965,
+      "loss": 0.5469,
+      "step": 17
+    },
+    {
+      "epoch": 0.037578288100208766,
+      "grad_norm": 0.07804122247550843,
+      "learning_rate": 0.00024827586206896553,
+      "loss": 0.543,
+      "step": 18
+    },
+    {
+      "epoch": 0.03966597077244259,
+      "grad_norm": 0.08565464666867292,
+      "learning_rate": 0.00026206896551724137,
+      "loss": 0.5625,
+      "step": 19
+    },
+    {
+      "epoch": 0.04175365344467641,
+      "grad_norm": 0.07879295016594862,
+      "learning_rate": 0.00027586206896551725,
+      "loss": 0.4551,
+      "step": 20
+    },
+    {
+      "epoch": 0.04384133611691023,
+      "grad_norm": 0.07549627256843491,
+      "learning_rate": 0.00028965517241379314,
+      "loss": 0.4863,
+      "step": 21
+    },
+    {
+      "epoch": 0.04592901878914405,
+      "grad_norm": 0.07816061161931821,
+      "learning_rate": 0.00030344827586206897,
+      "loss": 0.459,
+      "step": 22
+    },
+    {
+      "epoch": 0.04801670146137787,
+      "grad_norm": 0.07796400588496231,
+      "learning_rate": 0.00031724137931034486,
+      "loss": 0.5156,
+      "step": 23
+    },
+    {
+      "epoch": 0.05010438413361169,
+      "grad_norm": 0.06699113066481945,
+      "learning_rate": 0.0003310344827586207,
+      "loss": 0.5508,
+      "step": 24
+    },
+    {
+      "epoch": 0.05219206680584551,
+      "grad_norm": 0.07261576114010894,
+      "learning_rate": 0.0003448275862068965,
+      "loss": 0.5469,
+      "step": 25
+    },
+    {
+      "epoch": 0.054279749478079335,
+      "grad_norm": 0.07464032217874282,
+      "learning_rate": 0.0003586206896551724,
+      "loss": 0.4805,
+      "step": 26
+    },
+    {
+      "epoch": 0.05636743215031315,
+      "grad_norm": 0.06913377612077472,
+      "learning_rate": 0.0003724137931034483,
+      "loss": 0.5117,
+      "step": 27
+    },
+    {
+      "epoch": 0.05845511482254697,
+      "grad_norm": 0.061903733751750956,
+      "learning_rate": 0.0003862068965517242,
+      "loss": 0.4844,
+      "step": 28
+    },
+    {
+      "epoch": 0.060542797494780795,
+      "grad_norm": 0.06908123825325603,
+      "learning_rate": 0.0004,
+      "loss": 0.4434,
+      "step": 29
+    },
+    {
+      "epoch": 0.06263048016701461,
+      "grad_norm": 0.05444037852658766,
+      "learning_rate": 0.00039999885641643965,
+      "loss": 0.4336,
+      "step": 30
+    },
+    {
+      "epoch": 0.06471816283924843,
+      "grad_norm": 0.06329218054783699,
+      "learning_rate": 0.00039999542567883625,
+      "loss": 0.5078,
+      "step": 31
+    },
+    {
+      "epoch": 0.06680584551148225,
+      "grad_norm": 0.05617311135086726,
+      "learning_rate": 0.0003999897078264233,
+      "loss": 0.3418,
+      "step": 32
+    },
+    {
+      "epoch": 0.06889352818371608,
+      "grad_norm": 0.0631466065976358,
+      "learning_rate": 0.00039998170292458913,
+      "loss": 0.4512,
+      "step": 33
+    },
+    {
+      "epoch": 0.0709812108559499,
+      "grad_norm": 0.05939776609982831,
+      "learning_rate": 0.0003999714110648765,
+      "loss": 0.4727,
+      "step": 34
+    },
+    {
+      "epoch": 0.07306889352818371,
+      "grad_norm": 0.07015310870913878,
+      "learning_rate": 0.0003999588323649815,
+      "loss": 0.4316,
+      "step": 35
+    },
+    {
+      "epoch": 0.07515657620041753,
+      "grad_norm": 0.05541574151493628,
+      "learning_rate": 0.00039994396696875197,
+      "loss": 0.3691,
+      "step": 36
+    },
+    {
+      "epoch": 0.07724425887265135,
+      "grad_norm": 0.04781737127842468,
+      "learning_rate": 0.0003999268150461862,
+      "loss": 0.3711,
+      "step": 37
+    },
+    {
+      "epoch": 0.07933194154488518,
+      "grad_norm": 0.052081481582784105,
+      "learning_rate": 0.00039990737679343073,
+      "loss": 0.3633,
+      "step": 38
+    },
+    {
+      "epoch": 0.081419624217119,
+      "grad_norm": 0.04824424451620144,
+      "learning_rate": 0.0003998856524327782,
+      "loss": 0.3867,
+      "step": 39
+    },
+    {
+      "epoch": 0.08350730688935282,
+      "grad_norm": 0.048331445388276904,
+      "learning_rate": 0.0003998616422126649,
+      "loss": 0.4199,
+      "step": 40
+    },
+    {
+      "epoch": 0.08559498956158663,
+      "grad_norm": 0.06796823383551855,
+      "learning_rate": 0.00039983534640766766,
+      "loss": 0.4629,
+      "step": 41
+    },
+    {
+      "epoch": 0.08768267223382047,
+      "grad_norm": 0.04720979824092969,
+      "learning_rate": 0.0003998067653185011,
+      "loss": 0.3594,
+      "step": 42
+    },
+    {
+      "epoch": 0.08977035490605428,
+      "grad_norm": 0.05374730136315078,
+      "learning_rate": 0.00039977589927201373,
+      "loss": 0.4414,
+      "step": 43
+    },
+    {
+      "epoch": 0.0918580375782881,
+      "grad_norm": 0.050894908072710146,
+      "learning_rate": 0.0003997427486211847,
+      "loss": 0.4043,
+      "step": 44
+    },
+    {
+      "epoch": 0.09394572025052192,
+      "grad_norm": 0.04383506588833311,
+      "learning_rate": 0.0003997073137451194,
+      "loss": 0.3281,
+      "step": 45
+    },
+    {
+      "epoch": 0.09603340292275574,
+      "grad_norm": 0.055686594769476794,
+      "learning_rate": 0.00039966959504904515,
+      "loss": 0.4062,
+      "step": 46
+    },
+    {
+      "epoch": 0.09812108559498957,
+      "grad_norm": 0.05244693300757331,
+      "learning_rate": 0.00039962959296430676,
+      "loss": 0.3652,
+      "step": 47
+    },
+    {
+      "epoch": 0.10020876826722339,
+      "grad_norm": 0.05926741920735249,
+      "learning_rate": 0.0003995873079483616,
+      "loss": 0.4844,
+      "step": 48
+    },
+    {
+      "epoch": 0.1022964509394572,
+      "grad_norm": 0.04551719284385615,
+      "learning_rate": 0.00039954274048477405,
+      "loss": 0.3223,
+      "step": 49
+    },
+    {
+      "epoch": 0.10438413361169102,
+      "grad_norm": 0.04385484814869444,
+      "learning_rate": 0.0003994958910832104,
+      "loss": 0.4004,
+      "step": 50
+    },
+    {
+      "epoch": 0.10647181628392484,
+      "grad_norm": 0.04780597607586747,
+      "learning_rate": 0.0003994467602794327,
+      "loss": 0.3125,
+      "step": 51
+    },
+    {
+      "epoch": 0.10855949895615867,
+      "grad_norm": 0.08970784709982177,
+      "learning_rate": 0.00039939534863529265,
+      "loss": 0.3359,
+      "step": 52
+    },
+    {
+      "epoch": 0.11064718162839249,
+      "grad_norm": 0.055961833035488864,
+      "learning_rate": 0.00039934165673872536,
+      "loss": 0.4023,
+      "step": 53
+    },
+    {
+      "epoch": 0.1127348643006263,
+      "grad_norm": 0.051294633986921585,
+      "learning_rate": 0.00039928568520374263,
+      "loss": 0.3848,
+      "step": 54
+    },
+    {
+      "epoch": 0.11482254697286012,
+      "grad_norm": 0.05439699876212045,
+      "learning_rate": 0.0003992274346704257,
+      "loss": 0.4785,
+      "step": 55
+    },
+    {
+      "epoch": 0.11691022964509394,
+      "grad_norm": 0.04990096696869215,
+      "learning_rate": 0.00039916690580491806,
+      "loss": 0.377,
+      "step": 56
+    },
+    {
+      "epoch": 0.11899791231732777,
+      "grad_norm": 0.04880913842131259,
+      "learning_rate": 0.00039910409929941787,
+      "loss": 0.3809,
+      "step": 57
+    },
+    {
+      "epoch": 0.12108559498956159,
+      "grad_norm": 0.048110591694635114,
+      "learning_rate": 0.0003990390158721701,
+      "loss": 0.375,
+      "step": 58
+    },
+    {
+      "epoch": 0.12317327766179541,
+      "grad_norm": 0.05425382131888079,
+      "learning_rate": 0.0003989716562674579,
+      "loss": 0.4297,
+      "step": 59
+    },
+    {
+      "epoch": 0.12526096033402923,
+      "grad_norm": 0.04140427495861459,
+      "learning_rate": 0.0003989020212555948,
+      "loss": 0.3164,
+      "step": 60
+    },
+    {
+      "epoch": 0.12734864300626306,
+      "grad_norm": 0.055270560900777276,
+      "learning_rate": 0.00039883011163291545,
+      "loss": 0.418,
+      "step": 61
+    },
+    {
+      "epoch": 0.12943632567849686,
+      "grad_norm": 0.04546442173482473,
+      "learning_rate": 0.0003987559282217662,
+      "loss": 0.3125,
+      "step": 62
+    },
+    {
+      "epoch": 0.1315240083507307,
+      "grad_norm": 0.05835477454974449,
+      "learning_rate": 0.00039867947187049654,
+      "loss": 0.4199,
+      "step": 63
+    },
+    {
+      "epoch": 0.1336116910229645,
+      "grad_norm": 0.06512707863408194,
+      "learning_rate": 0.00039860074345344875,
+      "loss": 0.4238,
+      "step": 64
+    },
+    {
+      "epoch": 0.13569937369519833,
+      "grad_norm": 0.04896390157005381,
+      "learning_rate": 0.0003985197438709479,
+      "loss": 0.2598,
+      "step": 65
+    },
+    {
+      "epoch": 0.13778705636743216,
+      "grad_norm": 0.048613271081049014,
+      "learning_rate": 0.00039843647404929207,
+      "loss": 0.3633,
+      "step": 66
+    },
+    {
+      "epoch": 0.13987473903966596,
+      "grad_norm": 0.050233778334495785,
+      "learning_rate": 0.0003983509349407412,
+      "loss": 0.3926,
+      "step": 67
+    },
+    {
+      "epoch": 0.1419624217118998,
+      "grad_norm": 0.051442044526522436,
+      "learning_rate": 0.0003982631275235064,
+      "loss": 0.4531,
+      "step": 68
+    },
+    {
+      "epoch": 0.1440501043841336,
+      "grad_norm": 0.04624056126157752,
+      "learning_rate": 0.00039817305280173893,
+      "loss": 0.373,
+      "step": 69
+    },
+    {
+      "epoch": 0.14613778705636743,
+      "grad_norm": 0.04350984771291989,
+      "learning_rate": 0.0003980807118055185,
+      "loss": 0.377,
+      "step": 70
+    },
+    {
+      "epoch": 0.14822546972860126,
+      "grad_norm": 0.04395725697208839,
+      "learning_rate": 0.0003979861055908415,
+      "loss": 0.3633,
+      "step": 71
+    },
+    {
+      "epoch": 0.15031315240083507,
+      "grad_norm": 0.04706297130636851,
+      "learning_rate": 0.0003978892352396091,
+      "loss": 0.4199,
+      "step": 72
+    },
+    {
+      "epoch": 0.1524008350730689,
+      "grad_norm": 0.044497773696428565,
+      "learning_rate": 0.00039779010185961473,
+      "loss": 0.4707,
+      "step": 73
+    },
+    {
+      "epoch": 0.1544885177453027,
+      "grad_norm": 0.0411155822363719,
+      "learning_rate": 0.0003976887065845314,
+      "loss": 0.3262,
+      "step": 74
+    },
+    {
+      "epoch": 0.15657620041753653,
+      "grad_norm": 0.043812847498547546,
+      "learning_rate": 0.0003975850505738988,
+      "loss": 0.3359,
+      "step": 75
+    },
+    {
+      "epoch": 0.15866388308977036,
+      "grad_norm": 0.0554940446413714,
+      "learning_rate": 0.0003974791350131101,
+      "loss": 0.3867,
+      "step": 76
+    },
+    {
+      "epoch": 0.16075156576200417,
+      "grad_norm": 0.04898321080003146,
+      "learning_rate": 0.00039737096111339804,
+      "loss": 0.4648,
+      "step": 77
+    },
+    {
+      "epoch": 0.162839248434238,
+      "grad_norm": 0.04703455829390756,
+      "learning_rate": 0.00039726053011182176,
+      "loss": 0.3945,
+      "step": 78
+    },
+    {
+      "epoch": 0.1649269311064718,
+      "grad_norm": 0.048291736323549404,
+      "learning_rate": 0.00039714784327125196,
+      "loss": 0.3555,
+      "step": 79
+    },
+    {
+      "epoch": 0.16701461377870563,
+      "grad_norm": 0.048158996103635526,
+      "learning_rate": 0.00039703290188035687,
+      "loss": 0.3848,
+      "step": 80
+    },
+    {
+      "epoch": 0.16910229645093947,
+      "grad_norm": 0.041070665599436396,
+      "learning_rate": 0.00039691570725358726,
+      "loss": 0.3672,
+      "step": 81
+    },
+    {
+      "epoch": 0.17118997912317327,
+      "grad_norm": 0.04514739709583794,
+      "learning_rate": 0.0003967962607311617,
+      "loss": 0.416,
+      "step": 82
+    },
+    {
+      "epoch": 0.1732776617954071,
+      "grad_norm": 0.05018154319088866,
+      "learning_rate": 0.0003966745636790509,
+      "loss": 0.4219,
+      "step": 83
+    },
+    {
+      "epoch": 0.17536534446764093,
+      "grad_norm": 0.04615841699661321,
+      "learning_rate": 0.00039655061748896243,
+      "loss": 0.3594,
+      "step": 84
+    },
+    {
+      "epoch": 0.17745302713987474,
+      "grad_norm": 0.04770173243618234,
+      "learning_rate": 0.00039642442357832447,
+      "loss": 0.3789,
+      "step": 85
+    },
+    {
+      "epoch": 0.17954070981210857,
+      "grad_norm": 0.0529492248920703,
+      "learning_rate": 0.00039629598339026984,
+      "loss": 0.4102,
+      "step": 86
+    },
+    {
+      "epoch": 0.18162839248434237,
+      "grad_norm": 0.04108643805111346,
+      "learning_rate": 0.00039616529839361946,
+      "loss": 0.3691,
+      "step": 87
+    },
+    {
+      "epoch": 0.1837160751565762,
+      "grad_norm": 0.04230844261748932,
+      "learning_rate": 0.00039603237008286544,
+      "loss": 0.3672,
+      "step": 88
+    },
+    {
+      "epoch": 0.18580375782881003,
+      "grad_norm": 0.057536412046600406,
+      "learning_rate": 0.00039589719997815413,
+      "loss": 0.3574,
+      "step": 89
+    },
+    {
+      "epoch": 0.18789144050104384,
+      "grad_norm": 0.03943187889859641,
+      "learning_rate": 0.0003957597896252685,
+      "loss": 0.332,
+      "step": 90
+    },
+    {
+      "epoch": 0.18997912317327767,
+      "grad_norm": 0.04024451449784526,
+      "learning_rate": 0.000395620140595611,
+      "loss": 0.3438,
+      "step": 91
+    },
+    {
+      "epoch": 0.19206680584551147,
+      "grad_norm": 0.04428176826646225,
+      "learning_rate": 0.0003954782544861847,
+      "loss": 0.3574,
+      "step": 92
+    },
+    {
+      "epoch": 0.1941544885177453,
+      "grad_norm": 0.045692937481720604,
+      "learning_rate": 0.00039533413291957596,
+      "loss": 0.3574,
+      "step": 93
+    },
+    {
+      "epoch": 0.19624217118997914,
+      "grad_norm": 0.04688683747789243,
+      "learning_rate": 0.00039518777754393535,
+      "loss": 0.3867,
+      "step": 94
+    },
+    {
+      "epoch": 0.19832985386221294,
+      "grad_norm": 0.05536707488449957,
+      "learning_rate": 0.0003950391900329588,
+      "loss": 0.3223,
+      "step": 95
+    },
+    {
+      "epoch": 0.20041753653444677,
+      "grad_norm": 0.04606231136571311,
+      "learning_rate": 0.0003948883720858687,
+      "loss": 0.3535,
+      "step": 96
+    },
+    {
+      "epoch": 0.20250521920668058,
+      "grad_norm": 0.04099133317121465,
+      "learning_rate": 0.00039473532542739435,
+      "loss": 0.3066,
+      "step": 97
+    },
+    {
+      "epoch": 0.2045929018789144,
+      "grad_norm": 0.055252229005040635,
+      "learning_rate": 0.0003945800518077521,
+      "loss": 0.3809,
+      "step": 98
+    },
+    {
+      "epoch": 0.20668058455114824,
+      "grad_norm": 0.04490759353014913,
+      "learning_rate": 0.00039442255300262553,
+      "loss": 0.373,
+      "step": 99
+    },
+    {
+      "epoch": 0.20876826722338204,
+      "grad_norm": 0.04889359614788617,
+      "learning_rate": 0.0003942628308131452,
+      "loss": 0.377,
+      "step": 100
+    },
+    {
+      "epoch": 0.21085594989561587,
+      "grad_norm": 0.04331101172153666,
+      "learning_rate": 0.0003941008870658677,
+      "loss": 0.3926,
+      "step": 101
+    },
+    {
+      "epoch": 0.21294363256784968,
+      "grad_norm": 0.039340556954521134,
+      "learning_rate": 0.00039393672361275506,
+      "loss": 0.3379,
+      "step": 102
+    },
+    {
+      "epoch": 0.2150313152400835,
+      "grad_norm": 0.047692728640419506,
+      "learning_rate": 0.0003937703423311537,
+      "loss": 0.375,
+      "step": 103
+    },
+    {
+      "epoch": 0.21711899791231734,
+      "grad_norm": 0.04193899617132502,
+      "learning_rate": 0.00039360174512377246,
+      "loss": 0.3203,
+      "step": 104
+    },
+    {
+      "epoch": 0.21920668058455114,
+      "grad_norm": 0.04450971768791051,
+      "learning_rate": 0.0003934309339186614,
+      "loss": 0.3789,
+      "step": 105
+    },
+    {
+      "epoch": 0.22129436325678498,
+      "grad_norm": 0.04610401641341007,
+      "learning_rate": 0.00039325791066918925,
+      "loss": 0.4004,
+      "step": 106
+    },
+    {
+      "epoch": 0.22338204592901878,
+      "grad_norm": 0.0394316515696973,
+      "learning_rate": 0.00039308267735402156,
+      "loss": 0.2734,
+      "step": 107
+    },
+    {
+      "epoch": 0.2254697286012526,
+      "grad_norm": 0.04552211598084851,
+      "learning_rate": 0.00039290523597709765,
+      "loss": 0.3555,
+      "step": 108
+    },
+    {
+      "epoch": 0.22755741127348644,
+      "grad_norm": 0.044355074180719295,
+      "learning_rate": 0.00039272558856760797,
+      "loss": 0.3418,
+      "step": 109
+    },
+    {
+      "epoch": 0.22964509394572025,
+      "grad_norm": 0.04664869616431499,
+      "learning_rate": 0.0003925437371799708,
+      "loss": 0.3301,
+      "step": 110
+    },
+    {
+      "epoch": 0.23173277661795408,
+      "grad_norm": 0.04421582409863553,
+      "learning_rate": 0.0003923596838938086,
+      "loss": 0.3242,
+      "step": 111
+    },
+    {
+      "epoch": 0.23382045929018788,
+      "grad_norm": 0.04914905461390972,
+      "learning_rate": 0.00039217343081392463,
+      "loss": 0.3027,
+      "step": 112
+    },
+    {
+      "epoch": 0.2359081419624217,
+      "grad_norm": 0.04801403896872833,
+      "learning_rate": 0.00039198498007027836,
+      "loss": 0.377,
+      "step": 113
+    },
+    {
+      "epoch": 0.23799582463465555,
+      "grad_norm": 0.039909379565673524,
+      "learning_rate": 0.0003917943338179616,
+      "loss": 0.2949,
+      "step": 114
+    },
+    {
+      "epoch": 0.24008350730688935,
+      "grad_norm": 0.04562132710784955,
+      "learning_rate": 0.00039160149423717353,
+      "loss": 0.3828,
+      "step": 115
+    },
+    {
+      "epoch": 0.24217118997912318,
+      "grad_norm": 0.04218187984586478,
+      "learning_rate": 0.00039140646353319584,
+      "loss": 0.3379,
+      "step": 116
+    },
+    {
+      "epoch": 0.24425887265135698,
+      "grad_norm": 0.13995497079449562,
+      "learning_rate": 0.0003912092439363677,
+      "loss": 0.3887,
+      "step": 117
+    },
+    {
+      "epoch": 0.24634655532359082,
+      "grad_norm": 0.03956892044992931,
+      "learning_rate": 0.00039100983770205995,
+      "loss": 0.3027,
+      "step": 118
+    },
+    {
+      "epoch": 0.24843423799582465,
+      "grad_norm": 0.043965744196170775,
+      "learning_rate": 0.00039080824711064946,
+      "loss": 0.3223,
+      "step": 119
+    },
+    {
+      "epoch": 0.25052192066805845,
+      "grad_norm": 0.044644460638659476,
+      "learning_rate": 0.00039060447446749314,
+      "loss": 0.3496,
+      "step": 120
+    },
+    {
+      "epoch": 0.25260960334029225,
+      "grad_norm": 0.05275646603548191,
+      "learning_rate": 0.00039039852210290147,
+      "loss": 0.3438,
+      "step": 121
+    },
+    {
+      "epoch": 0.2546972860125261,
+      "grad_norm": 0.04064451365674439,
+      "learning_rate": 0.00039019039237211175,
+      "loss": 0.3359,
+      "step": 122
+    },
+    {
+      "epoch": 0.2567849686847599,
+      "grad_norm": 0.048323141931080164,
+      "learning_rate": 0.00038998008765526136,
+      "loss": 0.377,
+      "step": 123
+    },
+    {
+      "epoch": 0.2588726513569937,
+      "grad_norm": 0.03725290930480903,
+      "learning_rate": 0.00038976761035736064,
+      "loss": 0.2988,
+      "step": 124
+    },
+    {
+      "epoch": 0.2609603340292276,
+      "grad_norm": 0.04052181522164268,
+      "learning_rate": 0.00038955296290826483,
+      "loss": 0.3203,
+      "step": 125
+    },
+    {
+      "epoch": 0.2630480167014614,
+      "grad_norm": 0.04516920645500193,
+      "learning_rate": 0.00038933614776264695,
+      "loss": 0.3496,
+      "step": 126
+    },
+    {
+      "epoch": 0.2651356993736952,
+      "grad_norm": 0.04887108820007405,
+      "learning_rate": 0.00038911716739996945,
+      "loss": 0.2734,
+      "step": 127
+    },
+    {
+      "epoch": 0.267223382045929,
+      "grad_norm": 0.04662438292740599,
+      "learning_rate": 0.0003888960243244556,
+      "loss": 0.2949,
+      "step": 128
+    },
+    {
+      "epoch": 0.26931106471816285,
+      "grad_norm": 0.0462101738931956,
+      "learning_rate": 0.0003886727210650613,
+      "loss": 0.3711,
+      "step": 129
+    },
+    {
+      "epoch": 0.27139874739039666,
+      "grad_norm": 0.045514397288015096,
+      "learning_rate": 0.00038844726017544606,
+      "loss": 0.3945,
+      "step": 130
+    },
+    {
+      "epoch": 0.27348643006263046,
+      "grad_norm": 0.04564176989369504,
+      "learning_rate": 0.00038821964423394335,
+      "loss": 0.3262,
+      "step": 131
+    },
+    {
+      "epoch": 0.2755741127348643,
+      "grad_norm": 0.0480634271209225,
+      "learning_rate": 0.00038798987584353175,
+      "loss": 0.4297,
+      "step": 132
+    },
+    {
+      "epoch": 0.2776617954070981,
+      "grad_norm": 0.04064367156377542,
+      "learning_rate": 0.0003877579576318048,
+      "loss": 0.3066,
+      "step": 133
+    },
+    {
+      "epoch": 0.2797494780793319,
+      "grad_norm": 0.06187073759126634,
+      "learning_rate": 0.000387523892250941,
+      "loss": 0.3828,
+      "step": 134
+    },
+    {
+      "epoch": 0.2818371607515658,
+      "grad_norm": 0.047235774434653544,
+      "learning_rate": 0.0003872876823776737,
+      "loss": 0.4551,
+      "step": 135
+    },
+    {
+      "epoch": 0.2839248434237996,
+      "grad_norm": 0.04613794323687757,
+      "learning_rate": 0.00038704933071326004,
+      "loss": 0.3555,
+      "step": 136
+    },
+    {
+      "epoch": 0.2860125260960334,
+      "grad_norm": 0.04123661418763717,
+      "learning_rate": 0.00038680883998345046,
+      "loss": 0.3848,
+      "step": 137
+    },
+    {
+      "epoch": 0.2881002087682672,
+      "grad_norm": 0.03990534224519547,
+      "learning_rate": 0.0003865662129384575,
+      "loss": 0.3223,
+      "step": 138
+    },
+    {
+      "epoch": 0.29018789144050106,
+      "grad_norm": 0.03964108882883799,
+      "learning_rate": 0.00038632145235292407,
+      "loss": 0.3145,
+      "step": 139
+    },
+    {
+      "epoch": 0.29227557411273486,
+      "grad_norm": 0.04771985690919652,
+      "learning_rate": 0.00038607456102589217,
+      "loss": 0.3555,
+      "step": 140
+    },
+    {
+      "epoch": 0.29436325678496866,
+      "grad_norm": 0.041459959268485676,
+      "learning_rate": 0.0003858255417807701,
+      "loss": 0.3203,
+      "step": 141
+    },
+    {
+      "epoch": 0.2964509394572025,
+      "grad_norm": 0.04488781487797441,
+      "learning_rate": 0.0003855743974653014,
+      "loss": 0.3516,
+      "step": 142
+    },
+    {
+      "epoch": 0.2985386221294363,
+      "grad_norm": 0.045595447329589846,
+      "learning_rate": 0.00038532113095153086,
+      "loss": 0.4121,
+      "step": 143
+    },
+    {
+      "epoch": 0.30062630480167013,
+      "grad_norm": 0.051789411309846184,
+      "learning_rate": 0.0003850657451357729,
+      "loss": 0.3945,
+      "step": 144
+    },
+    {
+      "epoch": 0.302713987473904,
+      "grad_norm": 0.04655813617947128,
+      "learning_rate": 0.00038480824293857756,
+      "loss": 0.377,
+      "step": 145
+    },
+    {
+      "epoch": 0.3048016701461378,
+      "grad_norm": 0.07648712725066147,
+      "learning_rate": 0.00038454862730469773,
+      "loss": 0.3438,
+      "step": 146
+    },
+    {
+      "epoch": 0.3068893528183716,
+      "grad_norm": 0.05310195833383878,
+      "learning_rate": 0.0003842869012030551,
+      "loss": 0.3594,
+      "step": 147
+    },
+    {
+      "epoch": 0.3089770354906054,
+      "grad_norm": 0.04391887584861711,
+      "learning_rate": 0.00038402306762670636,
+      "loss": 0.3555,
+      "step": 148
+    },
+    {
+      "epoch": 0.31106471816283926,
+      "grad_norm": 0.038985625260836355,
+      "learning_rate": 0.000383757129592809,
+      "loss": 0.2969,
+      "step": 149
+    },
+    {
+      "epoch": 0.31315240083507306,
+      "grad_norm": 0.050635143690619895,
+      "learning_rate": 0.00038348909014258647,
+      "loss": 0.3398,
+      "step": 150
+    },
+    {
+      "epoch": 0.31524008350730687,
+      "grad_norm": 0.04475955558328409,
+      "learning_rate": 0.00038321895234129394,
+      "loss": 0.3594,
+      "step": 151
+    },
+    {
+      "epoch": 0.3173277661795407,
+      "grad_norm": 0.046705792421880006,
+      "learning_rate": 0.0003829467192781829,
+      "loss": 0.3906,
+      "step": 152
+    },
+    {
+      "epoch": 0.31941544885177453,
+      "grad_norm": 0.04135295822439203,
+      "learning_rate": 0.00038267239406646595,
+      "loss": 0.3203,
+      "step": 153
+    },
+    {
+      "epoch": 0.32150313152400833,
+      "grad_norm": 0.03936648551856702,
+      "learning_rate": 0.000382395979843281,
+      "loss": 0.3086,
+      "step": 154
+    },
+    {
+      "epoch": 0.3235908141962422,
+      "grad_norm": 0.04411342571947294,
+      "learning_rate": 0.0003821174797696558,
+      "loss": 0.3145,
+      "step": 155
+    },
+    {
+      "epoch": 0.325678496868476,
+      "grad_norm": 0.0440989009123292,
+      "learning_rate": 0.0003818368970304713,
+      "loss": 0.3398,
+      "step": 156
+    },
+    {
+      "epoch": 0.3277661795407098,
+      "grad_norm": 0.04623084733338215,
+      "learning_rate": 0.00038155423483442565,
+      "loss": 0.373,
+      "step": 157
+    },
+    {
+      "epoch": 0.3298538622129436,
+      "grad_norm": 0.03851250195995776,
+      "learning_rate": 0.00038126949641399725,
+      "loss": 0.2773,
+      "step": 158
+    },
+    {
+      "epoch": 0.33194154488517746,
+      "grad_norm": 0.04376944212115029,
+      "learning_rate": 0.0003809826850254078,
+      "loss": 0.3613,
+      "step": 159
+    },
+    {
+      "epoch": 0.33402922755741127,
+      "grad_norm": 0.044952970087986426,
+      "learning_rate": 0.00038069380394858526,
+      "loss": 0.2734,
+      "step": 160
+    },
+    {
+      "epoch": 0.33611691022964507,
+      "grad_norm": 0.04338817418637471,
+      "learning_rate": 0.0003804028564871262,
+      "loss": 0.334,
+      "step": 161
+    },
+    {
+      "epoch": 0.33820459290187893,
+      "grad_norm": 0.04540609013390802,
+      "learning_rate": 0.00038010984596825774,
+      "loss": 0.3086,
+      "step": 162
+    },
+    {
+      "epoch": 0.34029227557411273,
+      "grad_norm": 0.040675218084428574,
+      "learning_rate": 0.00037981477574280015,
+      "loss": 0.3105,
+      "step": 163
+    },
+    {
+      "epoch": 0.34237995824634654,
+      "grad_norm": 0.04628740772476471,
+      "learning_rate": 0.000379517649185128,
+      "loss": 0.375,
+      "step": 164
+    },
+    {
+      "epoch": 0.3444676409185804,
+      "grad_norm": 0.045320126800110626,
+      "learning_rate": 0.0003792184696931318,
+      "loss": 0.3926,
+      "step": 165
+    },
+    {
+      "epoch": 0.3465553235908142,
+      "grad_norm": 0.047079880323708634,
+      "learning_rate": 0.00037891724068817896,
+      "loss": 0.3613,
+      "step": 166
+    },
+    {
+      "epoch": 0.348643006263048,
+      "grad_norm": 0.042006341910680284,
+      "learning_rate": 0.0003786139656150749,
+      "loss": 0.3848,
+      "step": 167
+    },
+    {
+      "epoch": 0.35073068893528186,
+      "grad_norm": 0.042457959635259945,
+      "learning_rate": 0.0003783086479420235,
+      "loss": 0.3594,
+      "step": 168
+    },
+    {
+      "epoch": 0.35281837160751567,
+      "grad_norm": 0.04729445116359038,
+      "learning_rate": 0.0003780012911605875,
+      "loss": 0.4023,
+      "step": 169
+    },
+    {
+      "epoch": 0.35490605427974947,
+      "grad_norm": 0.04490557924422057,
+      "learning_rate": 0.00037769189878564843,
+      "loss": 0.3184,
+      "step": 170
+    },
+    {
+      "epoch": 0.3569937369519833,
+      "grad_norm": 0.04659805592024877,
+      "learning_rate": 0.0003773804743553667,
+      "loss": 0.3613,
+      "step": 171
+    },
+    {
+      "epoch": 0.35908141962421714,
+      "grad_norm": 0.0457558395135587,
+      "learning_rate": 0.00037706702143114085,
+      "loss": 0.3848,
+      "step": 172
+    },
+    {
+      "epoch": 0.36116910229645094,
+      "grad_norm": 0.042225345661087835,
+      "learning_rate": 0.0003767515435975671,
+      "loss": 0.3652,
+      "step": 173
+    },
+    {
+      "epoch": 0.36325678496868474,
+      "grad_norm": 0.03862535581909019,
+      "learning_rate": 0.00037643404446239805,
+      "loss": 0.2578,
+      "step": 174
+    },
+    {
+      "epoch": 0.3653444676409186,
+      "grad_norm": 0.04511245438490728,
+      "learning_rate": 0.0003761145276565015,
+      "loss": 0.3789,
+      "step": 175
+    },
+    {
+      "epoch": 0.3674321503131524,
+      "grad_norm": 0.04418007608895449,
+      "learning_rate": 0.0003757929968338193,
+      "loss": 0.375,
+      "step": 176
+    },
+    {
+      "epoch": 0.3695198329853862,
+      "grad_norm": 0.04131570246472954,
+      "learning_rate": 0.00037546945567132486,
+      "loss": 0.3262,
+      "step": 177
+    },
+    {
+      "epoch": 0.37160751565762007,
+      "grad_norm": 0.04432885957928793,
+      "learning_rate": 0.0003751439078689819,
+      "loss": 0.3105,
+      "step": 178
+    },
+    {
+      "epoch": 0.3736951983298539,
+      "grad_norm": 0.04130281228669527,
+      "learning_rate": 0.0003748163571497015,
+      "loss": 0.3145,
+      "step": 179
+    },
+    {
+      "epoch": 0.3757828810020877,
+      "grad_norm": 0.041235332477716484,
+      "learning_rate": 0.00037448680725929987,
+      "loss": 0.3086,
+      "step": 180
+    },
+    {
+      "epoch": 0.3778705636743215,
+      "grad_norm": 0.046591187622845885,
+      "learning_rate": 0.00037415526196645535,
+      "loss": 0.377,
+      "step": 181
+    },
+    {
+      "epoch": 0.37995824634655534,
+      "grad_norm": 0.043131270347982746,
+      "learning_rate": 0.00037382172506266544,
+      "loss": 0.3223,
+      "step": 182
+    },
+    {
+      "epoch": 0.38204592901878914,
+      "grad_norm": 0.04113286864493325,
+      "learning_rate": 0.00037348620036220325,
+      "loss": 0.3379,
+      "step": 183
+    },
+    {
+      "epoch": 0.38413361169102295,
+      "grad_norm": 0.045126082835651934,
+      "learning_rate": 0.00037314869170207427,
+      "loss": 0.3906,
+      "step": 184
+    },
+    {
+      "epoch": 0.3862212943632568,
+      "grad_norm": 0.0468308767634132,
+      "learning_rate": 0.00037280920294197186,
+      "loss": 0.418,
+      "step": 185
+    },
+    {
+      "epoch": 0.3883089770354906,
+      "grad_norm": 0.041909756539268275,
+      "learning_rate": 0.0003724677379642338,
+      "loss": 0.3008,
+      "step": 186
+    },
+    {
+      "epoch": 0.3903966597077244,
+      "grad_norm": 0.039249830774930286,
+      "learning_rate": 0.0003721243006737973,
+      "loss": 0.293,
+      "step": 187
+    },
+    {
+      "epoch": 0.3924843423799583,
+      "grad_norm": 0.043933299859950536,
+      "learning_rate": 0.0003717788949981549,
+      "loss": 0.3281,
+      "step": 188
+    },
+    {
+      "epoch": 0.3945720250521921,
+      "grad_norm": 0.045457197762689416,
+      "learning_rate": 0.0003714315248873091,
+      "loss": 0.416,
+      "step": 189
+    },
+    {
+      "epoch": 0.3966597077244259,
+      "grad_norm": 0.038856886092809004,
+      "learning_rate": 0.00037108219431372734,
+      "loss": 0.3242,
+      "step": 190
+    },
+    {
+      "epoch": 0.3987473903966597,
+      "grad_norm": 0.03887481602609131,
+      "learning_rate": 0.0003707309072722966,
+      "loss": 0.3047,
+      "step": 191
+    },
+    {
+      "epoch": 0.40083507306889354,
+      "grad_norm": 0.04338847469385484,
+      "learning_rate": 0.0003703776677802778,
+      "loss": 0.3945,
+      "step": 192
+    },
+    {
+      "epoch": 0.40292275574112735,
+      "grad_norm": 0.04874209932196092,
+      "learning_rate": 0.0003700224798772597,
+      "loss": 0.4316,
+      "step": 193
+    },
+    {
+      "epoch": 0.40501043841336115,
+      "grad_norm": 0.0409033493154823,
+      "learning_rate": 0.00036966534762511276,
+      "loss": 0.3301,
+      "step": 194
+    },
+    {
+      "epoch": 0.407098121085595,
+      "grad_norm": 0.040056909386716535,
+      "learning_rate": 0.0003693062751079427,
+      "loss": 0.3887,
+      "step": 195
+    },
+    {
+      "epoch": 0.4091858037578288,
+      "grad_norm": 0.03764639519212107,
+      "learning_rate": 0.0003689452664320437,
+      "loss": 0.3125,
+      "step": 196
+    },
+    {
+      "epoch": 0.4112734864300626,
+      "grad_norm": 0.04012224378405559,
+      "learning_rate": 0.00036858232572585173,
+      "loss": 0.3574,
+      "step": 197
+    },
+    {
+      "epoch": 0.4133611691022965,
+      "grad_norm": 0.04149192047957985,
+      "learning_rate": 0.00036821745713989706,
+      "loss": 0.3242,
+      "step": 198
+    },
+    {
+      "epoch": 0.4154488517745303,
+      "grad_norm": 0.05017075921477889,
+      "learning_rate": 0.00036785066484675683,
+      "loss": 0.3652,
+      "step": 199
+    },
+    {
+      "epoch": 0.4175365344467641,
+      "grad_norm": 0.04227941718590804,
+      "learning_rate": 0.00036748195304100735,
+      "loss": 0.3242,
+      "step": 200
+    },
+    {
+      "epoch": 0.4196242171189979,
+      "grad_norm": 0.03768968002704521,
+      "learning_rate": 0.00036711132593917635,
+      "loss": 0.3105,
+      "step": 201
+    },
+    {
+      "epoch": 0.42171189979123175,
+      "grad_norm": 0.04194019808640771,
+      "learning_rate": 0.00036673878777969426,
+      "loss": 0.2891,
+      "step": 202
+    },
+    {
+      "epoch": 0.42379958246346555,
+      "grad_norm": 0.04542209099244845,
+      "learning_rate": 0.00036636434282284633,
+      "loss": 0.3359,
+      "step": 203
+    },
+    {
+      "epoch": 0.42588726513569936,
+      "grad_norm": 0.03795072966054866,
+      "learning_rate": 0.0003659879953507235,
+      "loss": 0.2178,
+      "step": 204
+    },
+    {
+      "epoch": 0.4279749478079332,
+      "grad_norm": 0.0676312462445889,
+      "learning_rate": 0.0003656097496671737,
+      "loss": 0.4004,
+      "step": 205
+    },
+    {
+      "epoch": 0.430062630480167,
+      "grad_norm": 0.06595187551394353,
+      "learning_rate": 0.00036522961009775217,
+      "loss": 0.3359,
+      "step": 206
+    },
+    {
+      "epoch": 0.4321503131524008,
+      "grad_norm": 0.048238875202028735,
+      "learning_rate": 0.0003648475809896727,
+      "loss": 0.375,
+      "step": 207
+    },
+    {
+      "epoch": 0.4342379958246347,
+      "grad_norm": 0.04774287102162698,
+      "learning_rate": 0.00036446366671175735,
+      "loss": 0.3613,
+      "step": 208
+    },
+    {
+      "epoch": 0.4363256784968685,
+      "grad_norm": 0.04447429752378728,
+      "learning_rate": 0.0003640778716543866,
+      "loss": 0.3594,
+      "step": 209
+    },
+    {
+      "epoch": 0.4384133611691023,
+      "grad_norm": 0.04069940254145397,
+      "learning_rate": 0.00036369020022944946,
+      "loss": 0.3105,
+      "step": 210
+    },
+    {
+      "epoch": 0.4405010438413361,
+      "grad_norm": 0.03646181755503713,
+      "learning_rate": 0.0003633006568702924,
+      "loss": 0.3047,
+      "step": 211
+    },
+    {
+      "epoch": 0.44258872651356995,
+      "grad_norm": 0.041622368664595316,
+      "learning_rate": 0.00036290924603166945,
+      "loss": 0.3418,
+      "step": 212
+    },
+    {
+      "epoch": 0.44467640918580376,
+      "grad_norm": 0.04687265549200718,
+      "learning_rate": 0.00036251597218969045,
+      "loss": 0.3926,
+      "step": 213
+    },
+    {
+      "epoch": 0.44676409185803756,
+      "grad_norm": 0.05629699568400707,
+      "learning_rate": 0.0003621208398417705,
+      "loss": 0.3848,
+      "step": 214
+    },
+    {
+      "epoch": 0.4488517745302714,
+      "grad_norm": 0.04583925885988264,
+      "learning_rate": 0.0003617238535065782,
+      "loss": 0.334,
+      "step": 215
+    },
+    {
+      "epoch": 0.4509394572025052,
+      "grad_norm": 0.04693126530215214,
+      "learning_rate": 0.0003613250177239838,
+      "loss": 0.3691,
+      "step": 216
+    },
+    {
+      "epoch": 0.453027139874739,
+      "grad_norm": 0.05342936402981673,
+      "learning_rate": 0.00036092433705500797,
+      "loss": 0.3301,
+      "step": 217
+    },
+    {
+      "epoch": 0.4551148225469729,
+      "grad_norm": 0.04089550326814837,
+      "learning_rate": 0.000360521816081769,
+      "loss": 0.2793,
+      "step": 218
+    },
+    {
+      "epoch": 0.4572025052192067,
+      "grad_norm": 0.04725556069406023,
+      "learning_rate": 0.00036011745940743036,
+      "loss": 0.3887,
+      "step": 219
+    },
+    {
+      "epoch": 0.4592901878914405,
+      "grad_norm": 0.04629637553798221,
+      "learning_rate": 0.0003597112716561486,
+      "loss": 0.3477,
+      "step": 220
+    },
+    {
+      "epoch": 0.4613778705636743,
+      "grad_norm": 0.042829084723361624,
+      "learning_rate": 0.0003593032574730201,
+      "loss": 0.3379,
+      "step": 221
+    },
+    {
+      "epoch": 0.46346555323590816,
+      "grad_norm": 0.04290003185482568,
+      "learning_rate": 0.000358893421524028,
+      "loss": 0.3828,
+      "step": 222
+    },
+    {
+      "epoch": 0.46555323590814196,
+      "grad_norm": 0.03695634767631775,
+      "learning_rate": 0.0003584817684959888,
+      "loss": 0.2539,
+      "step": 223
+    },
+    {
+      "epoch": 0.46764091858037576,
+      "grad_norm": 0.043249882204696076,
+      "learning_rate": 0.00035806830309649887,
+      "loss": 0.291,
+      "step": 224
+    },
+    {
+      "epoch": 0.4697286012526096,
+      "grad_norm": 0.0426499927716365,
+      "learning_rate": 0.0003576530300538805,
+      "loss": 0.3262,
+      "step": 225
+    },
+    {
+      "epoch": 0.4718162839248434,
+      "grad_norm": 0.038269992068262225,
+      "learning_rate": 0.00035723595411712793,
+      "loss": 0.2598,
+      "step": 226
+    },
+    {
+      "epoch": 0.47390396659707723,
+      "grad_norm": 0.039227521535724244,
+      "learning_rate": 0.0003568170800558531,
+      "loss": 0.2812,
+      "step": 227
+    },
+    {
+      "epoch": 0.4759916492693111,
+      "grad_norm": 0.0414687007362602,
+      "learning_rate": 0.0003563964126602308,
+      "loss": 0.3184,
+      "step": 228
+    },
+    {
+      "epoch": 0.4780793319415449,
+      "grad_norm": 0.06563073201350628,
+      "learning_rate": 0.0003559739567409443,
+      "loss": 0.2969,
+      "step": 229
+    },
+    {
+      "epoch": 0.4801670146137787,
+      "grad_norm": 0.04223079865704053,
+      "learning_rate": 0.00035554971712913004,
+      "loss": 0.334,
+      "step": 230
+    },
+    {
+      "epoch": 0.4822546972860125,
+      "grad_norm": 0.042708614625140126,
+      "learning_rate": 0.00035512369867632236,
+      "loss": 0.3418,
+      "step": 231
+    },
+    {
+      "epoch": 0.48434237995824636,
+      "grad_norm": 0.04063103807491919,
+      "learning_rate": 0.00035469590625439843,
+      "loss": 0.3477,
+      "step": 232
+    },
+    {
+      "epoch": 0.48643006263048016,
+      "grad_norm": 0.045147549230686396,
+      "learning_rate": 0.00035426634475552187,
+      "loss": 0.3516,
+      "step": 233
+    },
+    {
+      "epoch": 0.48851774530271397,
+      "grad_norm": 0.04106108808167586,
+      "learning_rate": 0.00035383501909208755,
+      "loss": 0.2812,
+      "step": 234
+    },
+    {
+      "epoch": 0.4906054279749478,
+      "grad_norm": 0.0428728366141268,
+      "learning_rate": 0.0003534019341966647,
+      "loss": 0.3398,
+      "step": 235
+    },
+    {
+      "epoch": 0.49269311064718163,
+      "grad_norm": 0.047772722106343216,
+      "learning_rate": 0.00035296709502194115,
+      "loss": 0.3848,
+      "step": 236
+    },
+    {
+      "epoch": 0.49478079331941544,
+      "grad_norm": 0.04629142912114304,
+      "learning_rate": 0.000352530506540666,
+      "loss": 0.2715,
+      "step": 237
+    },
+    {
+      "epoch": 0.4968684759916493,
+      "grad_norm": 0.05018937450568595,
+      "learning_rate": 0.0003520921737455935,
+      "loss": 0.3027,
+      "step": 238
+    },
+    {
+      "epoch": 0.4989561586638831,
+      "grad_norm": 0.04431763914166162,
+      "learning_rate": 0.00035165210164942535,
+      "loss": 0.3477,
+      "step": 239
+    },
+    {
+      "epoch": 0.5010438413361169,
+      "grad_norm": 0.04080914710858152,
+      "learning_rate": 0.00035121029528475386,
+      "loss": 0.3086,
+      "step": 240
+    },
+    {
+      "epoch": 0.5031315240083507,
+      "grad_norm": 0.04222028074641329,
+      "learning_rate": 0.0003507667597040038,
+      "loss": 0.3672,
+      "step": 241
+    },
+    {
+      "epoch": 0.5052192066805845,
+      "grad_norm": 0.04367534710869874,
+      "learning_rate": 0.0003503214999793752,
+      "loss": 0.3223,
+      "step": 242
+    },
+    {
+      "epoch": 0.5073068893528184,
+      "grad_norm": 0.04813953280858226,
+      "learning_rate": 0.00034987452120278505,
+      "loss": 0.3984,
+      "step": 243
+    },
+    {
+      "epoch": 0.5093945720250522,
+      "grad_norm": 0.046413128010797784,
+      "learning_rate": 0.00034942582848580925,
+      "loss": 0.3848,
+      "step": 244
+    },
+    {
+      "epoch": 0.511482254697286,
+      "grad_norm": 0.036986531982911595,
+      "learning_rate": 0.0003489754269596239,
+      "loss": 0.3066,
+      "step": 245
+    },
+    {
+      "epoch": 0.5135699373695198,
+      "grad_norm": 0.040001114155947204,
+      "learning_rate": 0.00034852332177494685,
+      "loss": 0.2852,
+      "step": 246
+    },
+    {
+      "epoch": 0.5156576200417536,
+      "grad_norm": 0.04605884877382518,
+      "learning_rate": 0.0003480695181019786,
+      "loss": 0.3535,
+      "step": 247
+    },
+    {
+      "epoch": 0.5177453027139874,
+      "grad_norm": 0.0406658437248984,
+      "learning_rate": 0.0003476140211303433,
+      "loss": 0.334,
+      "step": 248
+    },
+    {
+      "epoch": 0.5198329853862212,
+      "grad_norm": 0.03909067687413047,
+      "learning_rate": 0.0003471568360690296,
+      "loss": 0.2871,
+      "step": 249
+    },
+    {
+      "epoch": 0.5219206680584552,
+      "grad_norm": 0.04831389973643467,
+      "learning_rate": 0.00034669796814633055,
+      "loss": 0.3398,
+      "step": 250
+    },
+    {
+      "epoch": 0.524008350730689,
+      "grad_norm": 0.04408464727595837,
+      "learning_rate": 0.0003462374226097844,
+      "loss": 0.3242,
+      "step": 251
+    },
+    {
+      "epoch": 0.5260960334029228,
+      "grad_norm": 0.03798160874377182,
+      "learning_rate": 0.00034577520472611414,
+      "loss": 0.252,
+      "step": 252
+    },
+    {
+      "epoch": 0.5281837160751566,
+      "grad_norm": 0.04625762493599334,
+      "learning_rate": 0.0003453113197811675,
+      "loss": 0.334,
+      "step": 253
+    },
+    {
+      "epoch": 0.5302713987473904,
+      "grad_norm": 0.03757333278335248,
+      "learning_rate": 0.0003448457730798564,
+      "loss": 0.2734,
+      "step": 254
+    },
+    {
+      "epoch": 0.5323590814196242,
+      "grad_norm": 0.038808091023111425,
+      "learning_rate": 0.0003443785699460965,
+      "loss": 0.3652,
+      "step": 255
+    },
+    {
+      "epoch": 0.534446764091858,
+      "grad_norm": 0.03797009162534055,
+      "learning_rate": 0.0003439097157227459,
+      "loss": 0.3223,
+      "step": 256
+    },
+    {
+      "epoch": 0.5365344467640919,
+      "grad_norm": 0.038601658815291935,
+      "learning_rate": 0.0003434392157715445,
+      "loss": 0.373,
+      "step": 257
+    },
+    {
+      "epoch": 0.5386221294363257,
+      "grad_norm": 0.04680506137895324,
+      "learning_rate": 0.00034296707547305235,
+      "loss": 0.334,
+      "step": 258
+    },
+    {
+      "epoch": 0.5407098121085595,
+      "grad_norm": 0.041499005757660336,
+      "learning_rate": 0.0003424933002265883,
+      "loss": 0.3457,
+      "step": 259
+    },
+    {
+      "epoch": 0.5427974947807933,
+      "grad_norm": 0.04115196540841461,
+      "learning_rate": 0.00034201789545016824,
+      "loss": 0.3242,
+      "step": 260
+    },
+    {
+      "epoch": 0.5448851774530271,
+      "grad_norm": 0.04340048389863411,
+      "learning_rate": 0.00034154086658044297,
+      "loss": 0.2949,
+      "step": 261
+    },
+    {
+      "epoch": 0.5469728601252609,
+      "grad_norm": 0.04378711119505926,
+      "learning_rate": 0.0003410622190726362,
+      "loss": 0.3066,
+      "step": 262
+    },
+    {
+      "epoch": 0.5490605427974948,
+      "grad_norm": 0.04643412727247773,
+      "learning_rate": 0.00034058195840048225,
+      "loss": 0.3652,
+      "step": 263
+    },
+    {
+      "epoch": 0.5511482254697286,
+      "grad_norm": 0.042043153196747375,
+      "learning_rate": 0.0003401000900561631,
+      "loss": 0.3789,
+      "step": 264
+    },
+    {
+      "epoch": 0.5532359081419624,
+      "grad_norm": 0.040040695052475275,
+      "learning_rate": 0.00033961661955024595,
+      "loss": 0.3145,
+      "step": 265
+    },
+    {
+      "epoch": 0.5553235908141962,
+      "grad_norm": 0.038423365462145156,
+      "learning_rate": 0.0003391315524116202,
+      "loss": 0.3223,
+      "step": 266
+    },
+    {
+      "epoch": 0.55741127348643,
+      "grad_norm": 0.038539884775496434,
+      "learning_rate": 0.00033864489418743355,
+      "loss": 0.3086,
+      "step": 267
+    },
+    {
+      "epoch": 0.5594989561586639,
+      "grad_norm": 0.04422443806977669,
+      "learning_rate": 0.0003381566504430298,
+      "loss": 0.3945,
+      "step": 268
+    },
+    {
+      "epoch": 0.5615866388308977,
+      "grad_norm": 0.0377661934221125,
+      "learning_rate": 0.00033766682676188386,
+      "loss": 0.3398,
+      "step": 269
+    },
+    {
+      "epoch": 0.5636743215031316,
+      "grad_norm": 0.036534819703584714,
+      "learning_rate": 0.00033717542874553896,
+      "loss": 0.3418,
+      "step": 270
+    },
+    {
+      "epoch": 0.5657620041753654,
+      "grad_norm": 0.04440346943789282,
+      "learning_rate": 0.0003366824620135421,
+      "loss": 0.3984,
+      "step": 271
+    },
+    {
+      "epoch": 0.5678496868475992,
+      "grad_norm": 0.037034787962705526,
+      "learning_rate": 0.00033618793220337965,
+      "loss": 0.2812,
+      "step": 272
+    },
+    {
+      "epoch": 0.569937369519833,
+      "grad_norm": 0.03927620752143223,
+      "learning_rate": 0.0003356918449704133,
+      "loss": 0.2695,
+      "step": 273
+    },
+    {
+      "epoch": 0.5720250521920668,
+      "grad_norm": 0.04369797108825467,
+      "learning_rate": 0.00033519420598781505,
+      "loss": 0.3477,
+      "step": 274
+    },
+    {
+      "epoch": 0.5741127348643006,
+      "grad_norm": 0.0379874901088521,
+      "learning_rate": 0.0003346950209465025,
+      "loss": 0.248,
+      "step": 275
+    },
+    {
+      "epoch": 0.5762004175365344,
+      "grad_norm": 0.042755557159503056,
+      "learning_rate": 0.0003341942955550738,
+      "loss": 0.3242,
+      "step": 276
+    },
+    {
+      "epoch": 0.5782881002087683,
+      "grad_norm": 0.039892604797522456,
+      "learning_rate": 0.0003336920355397421,
+      "loss": 0.3125,
+      "step": 277
+    },
+    {
+      "epoch": 0.5803757828810021,
+      "grad_norm": 0.037020438684393274,
+      "learning_rate": 0.0003331882466442704,
+      "loss": 0.2832,
+      "step": 278
+    },
+    {
+      "epoch": 0.5824634655532359,
+      "grad_norm": 0.0448544789689836,
+      "learning_rate": 0.0003326829346299057,
+      "loss": 0.3613,
+      "step": 279
+    },
+    {
+      "epoch": 0.5845511482254697,
+      "grad_norm": 0.037923744133296695,
+      "learning_rate": 0.00033217610527531315,
+      "loss": 0.3262,
+      "step": 280
+    },
+    {
+      "epoch": 0.5866388308977035,
+      "grad_norm": 0.039536585438320955,
+      "learning_rate": 0.00033166776437650985,
+      "loss": 0.3301,
+      "step": 281
+    },
+    {
+      "epoch": 0.5887265135699373,
+      "grad_norm": 0.042309874793804436,
+      "learning_rate": 0.0003311579177467988,
+      "loss": 0.3281,
+      "step": 282
+    },
+    {
+      "epoch": 0.5908141962421712,
+      "grad_norm": 0.04247524788442754,
+      "learning_rate": 0.0003306465712167023,
+      "loss": 0.3281,
+      "step": 283
+    },
+    {
+      "epoch": 0.592901878914405,
+      "grad_norm": 0.038298773282535274,
+      "learning_rate": 0.0003301337306338951,
+      "loss": 0.3066,
+      "step": 284
+    },
+    {
+      "epoch": 0.5949895615866388,
+      "grad_norm": 0.04845680190635201,
+      "learning_rate": 0.00032961940186313784,
+      "loss": 0.3867,
+      "step": 285
+    },
+    {
+      "epoch": 0.5970772442588727,
+      "grad_norm": 0.04425999769437242,
+      "learning_rate": 0.00032910359078620974,
+      "loss": 0.3848,
+      "step": 286
+    },
+    {
+      "epoch": 0.5991649269311065,
+      "grad_norm": 0.04350677333004508,
+      "learning_rate": 0.0003285863033018416,
+      "loss": 0.3359,
+      "step": 287
+    },
+    {
+      "epoch": 0.6012526096033403,
+      "grad_norm": 0.04313248762187342,
+      "learning_rate": 0.00032806754532564795,
+      "loss": 0.3281,
+      "step": 288
+    },
+    {
+      "epoch": 0.6033402922755741,
+      "grad_norm": 0.04172934704529679,
+      "learning_rate": 0.00032754732279005974,
+      "loss": 0.3887,
+      "step": 289
+    },
+    {
+      "epoch": 0.605427974947808,
+      "grad_norm": 0.053530546182614214,
+      "learning_rate": 0.0003270256416442564,
+      "loss": 0.4512,
+      "step": 290
+    },
+    {
+      "epoch": 0.6075156576200418,
+      "grad_norm": 0.0585588589735363,
+      "learning_rate": 0.0003265025078540977,
+      "loss": 0.4023,
+      "step": 291
+    },
+    {
+      "epoch": 0.6096033402922756,
+      "grad_norm": 0.046794215765212366,
+      "learning_rate": 0.00032597792740205573,
+      "loss": 0.2871,
+      "step": 292
+    },
+    {
+      "epoch": 0.6116910229645094,
+      "grad_norm": 0.04076085980117636,
+      "learning_rate": 0.0003254519062871462,
+      "loss": 0.3027,
+      "step": 293
+    },
+    {
+      "epoch": 0.6137787056367432,
+      "grad_norm": 0.039930806408977594,
+      "learning_rate": 0.0003249244505248603,
+      "loss": 0.3086,
+      "step": 294
+    },
+    {
+      "epoch": 0.615866388308977,
+      "grad_norm": 0.047339086645206876,
+      "learning_rate": 0.0003243955661470952,
+      "loss": 0.373,
+      "step": 295
+    },
+    {
+      "epoch": 0.6179540709812108,
+      "grad_norm": 0.041035006122051215,
+      "learning_rate": 0.00032386525920208574,
+      "loss": 0.3281,
+      "step": 296
+    },
+    {
+      "epoch": 0.6200417536534447,
+      "grad_norm": 0.05042533676976384,
+      "learning_rate": 0.000323333535754335,
+      "loss": 0.293,
+      "step": 297
+    },
+    {
+      "epoch": 0.6221294363256785,
+      "grad_norm": 0.043959165085212505,
+      "learning_rate": 0.00032280040188454494,
+      "loss": 0.3125,
+      "step": 298
+    },
+    {
+      "epoch": 0.6242171189979123,
+      "grad_norm": 0.04458635918949398,
+      "learning_rate": 0.0003222658636895469,
+      "loss": 0.3105,
+      "step": 299
+    },
+    {
+      "epoch": 0.6263048016701461,
+      "grad_norm": 0.04146656432155535,
+      "learning_rate": 0.0003217299272822316,
+      "loss": 0.3672,
+      "step": 300
+    },
+    {
+      "epoch": 0.6283924843423799,
+      "grad_norm": 0.04285146681080291,
+      "learning_rate": 0.00032119259879147987,
+      "loss": 0.2773,
+      "step": 301
+    },
+    {
+      "epoch": 0.6304801670146137,
+      "grad_norm": 0.04012756626164825,
+      "learning_rate": 0.0003206538843620919,
+      "loss": 0.3145,
+      "step": 302
+    },
+    {
+      "epoch": 0.6325678496868476,
+      "grad_norm": 0.03395547556102837,
+      "learning_rate": 0.0003201137901547175,
+      "loss": 0.2344,
+      "step": 303
+    },
+    {
+      "epoch": 0.6346555323590815,
+      "grad_norm": 0.038474167468473125,
+      "learning_rate": 0.00031957232234578506,
+      "loss": 0.3105,
+      "step": 304
+    },
+    {
+      "epoch": 0.6367432150313153,
+      "grad_norm": 0.04345850995580467,
+      "learning_rate": 0.00031902948712743146,
+      "loss": 0.3711,
+      "step": 305
+    },
+    {
+      "epoch": 0.6388308977035491,
+      "grad_norm": 0.04561991595493648,
+      "learning_rate": 0.00031848529070743113,
+      "loss": 0.3281,
+      "step": 306
+    },
+    {
+      "epoch": 0.6409185803757829,
+      "grad_norm": 0.043639261730465534,
+      "learning_rate": 0.00031793973930912465,
+      "loss": 0.3535,
+      "step": 307
+    },
+    {
+      "epoch": 0.6430062630480167,
+      "grad_norm": 0.041614849807266444,
+      "learning_rate": 0.00031739283917134835,
+      "loss": 0.293,
+      "step": 308
+    },
+    {
+      "epoch": 0.6450939457202505,
+      "grad_norm": 0.037829924843835225,
+      "learning_rate": 0.00031684459654836224,
+      "loss": 0.2793,
+      "step": 309
+    },
+    {
+      "epoch": 0.6471816283924844,
+      "grad_norm": 0.043593747485495736,
+      "learning_rate": 0.00031629501770977865,
+      "loss": 0.2871,
+      "step": 310
+    },
+    {
+      "epoch": 0.6492693110647182,
+      "grad_norm": 0.043504619259611736,
+      "learning_rate": 0.0003157441089404911,
+      "loss": 0.377,
+      "step": 311
+    },
+    {
+      "epoch": 0.651356993736952,
+      "grad_norm": 0.042969570910675,
+      "learning_rate": 0.00031519187654060153,
+      "loss": 0.3535,
+      "step": 312
+    },
+    {
+      "epoch": 0.6534446764091858,
+      "grad_norm": 0.04444146348618118,
+      "learning_rate": 0.0003146383268253489,
+      "loss": 0.3633,
+      "step": 313
+    },
+    {
+      "epoch": 0.6555323590814196,
+      "grad_norm": 0.0579233541293356,
+      "learning_rate": 0.00031408346612503684,
+      "loss": 0.2852,
+      "step": 314
+    },
+    {
+      "epoch": 0.6576200417536534,
+      "grad_norm": 0.04116612592402649,
+      "learning_rate": 0.00031352730078496103,
+      "loss": 0.3184,
+      "step": 315
+    },
+    {
+      "epoch": 0.6597077244258872,
+      "grad_norm": 0.042619485954835536,
+      "learning_rate": 0.0003129698371653368,
+      "loss": 0.3555,
+      "step": 316
+    },
+    {
+      "epoch": 0.6617954070981211,
+      "grad_norm": 0.03978568537420636,
+      "learning_rate": 0.00031241108164122667,
+      "loss": 0.3184,
+      "step": 317
+    },
+    {
+      "epoch": 0.6638830897703549,
+      "grad_norm": 0.04398595731711137,
+      "learning_rate": 0.0003118510406024667,
+      "loss": 0.3262,
+      "step": 318
+    },
+    {
+      "epoch": 0.6659707724425887,
+      "grad_norm": 0.03557319884714253,
+      "learning_rate": 0.0003112897204535943,
+      "loss": 0.2539,
+      "step": 319
+    },
+    {
+      "epoch": 0.6680584551148225,
+      "grad_norm": 0.04098653755334443,
+      "learning_rate": 0.0003107271276137745,
+      "loss": 0.3711,
+      "step": 320
+    },
+    {
+      "epoch": 0.6701461377870563,
+      "grad_norm": 0.0377819264468114,
+      "learning_rate": 0.0003101632685167263,
+      "loss": 0.2891,
+      "step": 321
+    },
+    {
+      "epoch": 0.6722338204592901,
+      "grad_norm": 0.038417025071128603,
+      "learning_rate": 0.00030959814961064973,
+      "loss": 0.2891,
+      "step": 322
+    },
+    {
+      "epoch": 0.6743215031315241,
+      "grad_norm": 0.038602243878265344,
+      "learning_rate": 0.0003090317773581518,
+      "loss": 0.2734,
+      "step": 323
+    },
+    {
+      "epoch": 0.6764091858037579,
+      "grad_norm": 0.0487821791964808,
+      "learning_rate": 0.0003084641582361723,
+      "loss": 0.3574,
+      "step": 324
+    },
+    {
+      "epoch": 0.6784968684759917,
+      "grad_norm": 0.04544872266451946,
+      "learning_rate": 0.00030789529873591034,
+      "loss": 0.3008,
+      "step": 325
+    },
+    {
+      "epoch": 0.6805845511482255,
+      "grad_norm": 0.0373160417363858,
+      "learning_rate": 0.0003073252053627496,
+      "loss": 0.2871,
+      "step": 326
+    },
+    {
+      "epoch": 0.6826722338204593,
+      "grad_norm": 0.04358868533626725,
+      "learning_rate": 0.0003067538846361841,
+      "loss": 0.3066,
+      "step": 327
+    },
+    {
+      "epoch": 0.6847599164926931,
+      "grad_norm": 0.040357136071415334,
+      "learning_rate": 0.0003061813430897439,
+      "loss": 0.3516,
+      "step": 328
+    },
+    {
+      "epoch": 0.6868475991649269,
+      "grad_norm": 0.04862770193509928,
+      "learning_rate": 0.0003056075872709199,
+      "loss": 0.4062,
+      "step": 329
+    },
+    {
+      "epoch": 0.6889352818371608,
+      "grad_norm": 0.04688881879011172,
+      "learning_rate": 0.00030503262374108927,
+      "loss": 0.4434,
+      "step": 330
+    },
+    {
+      "epoch": 0.6910229645093946,
+      "grad_norm": 0.04276287817278344,
+      "learning_rate": 0.00030445645907544057,
+      "loss": 0.3105,
+      "step": 331
+    },
+    {
+      "epoch": 0.6931106471816284,
+      "grad_norm": 0.03644902974160357,
+      "learning_rate": 0.0003038790998628981,
+      "loss": 0.249,
+      "step": 332
+    },
+    {
+      "epoch": 0.6951983298538622,
+      "grad_norm": 0.03602308126572102,
+      "learning_rate": 0.0003033005527060469,
+      "loss": 0.2891,
+      "step": 333
+    },
+    {
+      "epoch": 0.697286012526096,
+      "grad_norm": 0.0397502232733115,
+      "learning_rate": 0.0003027208242210571,
+      "loss": 0.334,
+      "step": 334
+    },
+    {
+      "epoch": 0.6993736951983298,
+      "grad_norm": 0.042614553937913115,
+      "learning_rate": 0.0003021399210376084,
+      "loss": 0.3281,
+      "step": 335
+    },
+    {
+      "epoch": 0.7014613778705637,
+      "grad_norm": 0.040360864280806166,
+      "learning_rate": 0.0003015578497988142,
+      "loss": 0.3145,
+      "step": 336
+    },
+    {
+      "epoch": 0.7035490605427975,
+      "grad_norm": 0.03697519450333998,
+      "learning_rate": 0.0003009746171611454,
+      "loss": 0.2852,
+      "step": 337
+    },
+    {
+      "epoch": 0.7056367432150313,
+      "grad_norm": 0.03920168369252357,
+      "learning_rate": 0.0003003902297943545,
+      "loss": 0.2578,
+      "step": 338
+    },
+    {
+      "epoch": 0.7077244258872651,
+      "grad_norm": 0.04596701530986532,
+      "learning_rate": 0.00029980469438139947,
+      "loss": 0.3066,
+      "step": 339
+    },
+    {
+      "epoch": 0.7098121085594989,
+      "grad_norm": 0.03590053569859233,
+      "learning_rate": 0.00029921801761836696,
+      "loss": 0.2598,
+      "step": 340
+    },
+    {
+      "epoch": 0.7118997912317327,
+      "grad_norm": 0.039137959610818826,
+      "learning_rate": 0.0002986302062143961,
+      "loss": 0.3398,
+      "step": 341
+    },
+    {
+      "epoch": 0.7139874739039666,
+      "grad_norm": 0.04445491792602957,
+      "learning_rate": 0.0002980412668916013,
+      "loss": 0.3398,
+      "step": 342
+    },
+    {
+      "epoch": 0.7160751565762005,
+      "grad_norm": 0.040340445047287156,
+      "learning_rate": 0.00029745120638499596,
+      "loss": 0.3516,
+      "step": 343
+    },
+    {
+      "epoch": 0.7181628392484343,
+      "grad_norm": 0.04520397395484933,
+      "learning_rate": 0.00029686003144241494,
+      "loss": 0.3555,
+      "step": 344
+    },
+    {
+      "epoch": 0.7202505219206681,
+      "grad_norm": 0.045695108134425295,
+      "learning_rate": 0.0002962677488244378,
+      "loss": 0.2871,
+      "step": 345
+    },
+    {
+      "epoch": 0.7223382045929019,
+      "grad_norm": 0.0414553610811136,
+      "learning_rate": 0.0002956743653043111,
+      "loss": 0.3828,
+      "step": 346
+    },
+    {
+      "epoch": 0.7244258872651357,
+      "grad_norm": 0.043267589906942225,
+      "learning_rate": 0.00029507988766787135,
+      "loss": 0.3262,
+      "step": 347
+    },
+    {
+      "epoch": 0.7265135699373695,
+      "grad_norm": 0.03665181460868117,
+      "learning_rate": 0.0002944843227134669,
+      "loss": 0.3301,
+      "step": 348
+    },
+    {
+      "epoch": 0.7286012526096033,
+      "grad_norm": 0.04220684212991339,
+      "learning_rate": 0.00029388767725188077,
+      "loss": 0.3223,
+      "step": 349
+    },
+    {
+      "epoch": 0.7306889352818372,
+      "grad_norm": 0.04468007941423499,
+      "learning_rate": 0.0002932899581062524,
+      "loss": 0.2949,
+      "step": 350
+    },
+    {
+      "epoch": 0.732776617954071,
+      "grad_norm": 0.044433764048658723,
+      "learning_rate": 0.00029269117211199955,
+      "loss": 0.377,
+      "step": 351
+    },
+    {
+      "epoch": 0.7348643006263048,
+      "grad_norm": 0.046209035174827495,
+      "learning_rate": 0.0002920913261167405,
+      "loss": 0.3691,
+      "step": 352
+    },
+    {
+      "epoch": 0.7369519832985386,
+      "grad_norm": 0.040034133910176115,
+      "learning_rate": 0.00029149042698021544,
+      "loss": 0.3105,
+      "step": 353
+    },
+    {
+      "epoch": 0.7390396659707724,
+      "grad_norm": 0.04609016944022695,
+      "learning_rate": 0.00029088848157420813,
+      "loss": 0.3984,
+      "step": 354
+    },
+    {
+      "epoch": 0.7411273486430062,
+      "grad_norm": 0.04480775819697267,
+      "learning_rate": 0.00029028549678246715,
+      "loss": 0.332,
+      "step": 355
+    },
+    {
+      "epoch": 0.7432150313152401,
+      "grad_norm": 0.04387465996335384,
+      "learning_rate": 0.0002896814795006277,
+      "loss": 0.3984,
+      "step": 356
+    },
+    {
+      "epoch": 0.7453027139874739,
+      "grad_norm": 0.041864619992452806,
+      "learning_rate": 0.0002890764366361318,
+      "loss": 0.3633,
+      "step": 357
+    },
+    {
+      "epoch": 0.7473903966597077,
+      "grad_norm": 0.03717012201708029,
+      "learning_rate": 0.0002884703751081504,
+      "loss": 0.2324,
+      "step": 358
+    },
+    {
+      "epoch": 0.7494780793319415,
+      "grad_norm": 0.043251857860819434,
+      "learning_rate": 0.0002878633018475035,
+      "loss": 0.3066,
+      "step": 359
+    },
+    {
+      "epoch": 0.7515657620041754,
+      "grad_norm": 0.03591327407860244,
+      "learning_rate": 0.00028725522379658097,
+      "loss": 0.2852,
+      "step": 360
+    },
+    {
+      "epoch": 0.7536534446764092,
+      "grad_norm": 0.0381908896371197,
+      "learning_rate": 0.00028664614790926355,
+      "loss": 0.3477,
+      "step": 361
+    },
+    {
+      "epoch": 0.755741127348643,
+      "grad_norm": 0.04392871145759733,
+      "learning_rate": 0.000286036081150843,
+      "loss": 0.4414,
+      "step": 362
+    },
+    {
+      "epoch": 0.7578288100208769,
+      "grad_norm": 0.04392250979416697,
+      "learning_rate": 0.0002854250304979423,
+      "loss": 0.2832,
+      "step": 363
+    },
+    {
+      "epoch": 0.7599164926931107,
+      "grad_norm": 0.03445036859255453,
+      "learning_rate": 0.0002848130029384363,
+      "loss": 0.3027,
+      "step": 364
+    },
+    {
+      "epoch": 0.7620041753653445,
+      "grad_norm": 0.03588315051937214,
+      "learning_rate": 0.00028420000547137175,
+      "loss": 0.2793,
+      "step": 365
+    },
+    {
+      "epoch": 0.7640918580375783,
+      "grad_norm": 0.043623631697390225,
+      "learning_rate": 0.00028358604510688666,
+      "loss": 0.3301,
+      "step": 366
+    },
+    {
+      "epoch": 0.7661795407098121,
+      "grad_norm": 0.044899266192686,
+      "learning_rate": 0.000282971128866131,
+      "loss": 0.2988,
+      "step": 367
+    },
+    {
+      "epoch": 0.7682672233820459,
+      "grad_norm": 0.044441044360934975,
+      "learning_rate": 0.0002823552637811857,
+      "loss": 0.3262,
+      "step": 368
+    },
+    {
+      "epoch": 0.7703549060542797,
+      "grad_norm": 0.04617336800060674,
+      "learning_rate": 0.00028173845689498286,
+      "loss": 0.2812,
+      "step": 369
+    },
+    {
+      "epoch": 0.7724425887265136,
+      "grad_norm": 0.03481084134760086,
+      "learning_rate": 0.0002811207152612243,
+      "loss": 0.2637,
+      "step": 370
+    },
+    {
+      "epoch": 0.7745302713987474,
+      "grad_norm": 0.04359599728927467,
+      "learning_rate": 0.000280502045944302,
+      "loss": 0.3516,
+      "step": 371
+    },
+    {
+      "epoch": 0.7766179540709812,
+      "grad_norm": 0.03910171209605865,
+      "learning_rate": 0.0002798824560192164,
+      "loss": 0.2793,
+      "step": 372
+    },
+    {
+      "epoch": 0.778705636743215,
+      "grad_norm": 0.04498875604580853,
+      "learning_rate": 0.0002792619525714962,
+      "loss": 0.334,
+      "step": 373
+    },
+    {
+      "epoch": 0.7807933194154488,
+      "grad_norm": 0.046945206399684386,
+      "learning_rate": 0.00027864054269711666,
+      "loss": 0.3262,
+      "step": 374
+    },
+    {
+      "epoch": 0.7828810020876826,
+      "grad_norm": 0.03720210905131818,
+      "learning_rate": 0.0002780182335024191,
+      "loss": 0.252,
+      "step": 375
+    },
+    {
+      "epoch": 0.7849686847599165,
+      "grad_norm": 0.03971464889707817,
+      "learning_rate": 0.00027739503210402907,
+      "loss": 0.334,
+      "step": 376
+    },
+    {
+      "epoch": 0.7870563674321504,
+      "grad_norm": 0.047941221203412314,
+      "learning_rate": 0.00027677094562877534,
+      "loss": 0.3066,
+      "step": 377
+    },
+    {
+      "epoch": 0.7891440501043842,
+      "grad_norm": 0.04116312254250454,
+      "learning_rate": 0.0002761459812136083,
+      "loss": 0.3652,
+      "step": 378
+    },
+    {
+      "epoch": 0.791231732776618,
+      "grad_norm": 0.0408608818482939,
+      "learning_rate": 0.00027552014600551813,
+      "loss": 0.3359,
+      "step": 379
+    },
+    {
+      "epoch": 0.7933194154488518,
+      "grad_norm": 0.039519399095894796,
+      "learning_rate": 0.00027489344716145344,
+      "loss": 0.3105,
+      "step": 380
+    },
+    {
+      "epoch": 0.7954070981210856,
+      "grad_norm": 0.04468756187207934,
+      "learning_rate": 0.0002742658918482392,
+      "loss": 0.334,
+      "step": 381
+    },
+    {
+      "epoch": 0.7974947807933194,
+      "grad_norm": 0.03799026650779495,
+      "learning_rate": 0.00027363748724249485,
+      "loss": 0.3125,
+      "step": 382
+    },
+    {
+      "epoch": 0.7995824634655533,
+      "grad_norm": 0.05260052294591301,
+      "learning_rate": 0.000273008240530552,
+      "loss": 0.4316,
+      "step": 383
+    },
+    {
+      "epoch": 0.8016701461377871,
+      "grad_norm": 0.04069771345832873,
+      "learning_rate": 0.00027237815890837283,
+      "loss": 0.2969,
+      "step": 384
+    },
+    {
+      "epoch": 0.8037578288100209,
+      "grad_norm": 0.039739122060491815,
+      "learning_rate": 0.000271747249581467,
+      "loss": 0.3555,
+      "step": 385
+    },
+    {
+      "epoch": 0.8058455114822547,
+      "grad_norm": 0.04610276901781518,
+      "learning_rate": 0.00027111551976480996,
+      "loss": 0.3574,
+      "step": 386
+    },
+    {
+      "epoch": 0.8079331941544885,
+      "grad_norm": 0.04374672870028719,
+      "learning_rate": 0.00027048297668276,
+      "loss": 0.3047,
+      "step": 387
+    },
+    {
+      "epoch": 0.8100208768267223,
+      "grad_norm": 0.04631564065384605,
+      "learning_rate": 0.0002698496275689758,
+      "loss": 0.2871,
+      "step": 388
+    },
+    {
+      "epoch": 0.8121085594989561,
+      "grad_norm": 0.049176525271089545,
+      "learning_rate": 0.00026921547966633373,
+      "loss": 0.2871,
+      "step": 389
+    },
+    {
+      "epoch": 0.81419624217119,
+      "grad_norm": 0.040234387326007176,
+      "learning_rate": 0.0002685805402268449,
+      "loss": 0.3301,
+      "step": 390
+    },
+    {
+      "epoch": 0.8162839248434238,
+      "grad_norm": 0.044678234660364975,
+      "learning_rate": 0.0002679448165115725,
+      "loss": 0.2852,
+      "step": 391
+    },
+    {
+      "epoch": 0.8183716075156576,
+      "grad_norm": 0.043749807508504976,
+      "learning_rate": 0.00026730831579054823,
+      "loss": 0.3418,
+      "step": 392
+    },
+    {
+      "epoch": 0.8204592901878914,
+      "grad_norm": 0.03702597280717622,
+      "learning_rate": 0.00026667104534268984,
+      "loss": 0.334,
+      "step": 393
+    },
+    {
+      "epoch": 0.8225469728601252,
+      "grad_norm": 0.04051649281576064,
+      "learning_rate": 0.00026603301245571733,
+      "loss": 0.3652,
+      "step": 394
+    },
+    {
+      "epoch": 0.824634655532359,
+      "grad_norm": 0.04070349409408308,
+      "learning_rate": 0.00026539422442607,
+      "loss": 0.3164,
+      "step": 395
+    },
+    {
+      "epoch": 0.826722338204593,
+      "grad_norm": 0.03685594277413145,
+      "learning_rate": 0.0002647546885588227,
+      "loss": 0.2695,
+      "step": 396
+    },
+    {
+      "epoch": 0.8288100208768268,
+      "grad_norm": 0.033959522628423805,
+      "learning_rate": 0.00026411441216760243,
+      "loss": 0.2422,
+      "step": 397
+    },
+    {
+      "epoch": 0.8308977035490606,
+      "grad_norm": 0.03649515909769935,
+      "learning_rate": 0.00026347340257450473,
+      "loss": 0.252,
+      "step": 398
+    },
+    {
+      "epoch": 0.8329853862212944,
+      "grad_norm": 0.05036202800279952,
+      "learning_rate": 0.00026283166711001,
+      "loss": 0.3203,
+      "step": 399
+    },
+    {
+      "epoch": 0.8350730688935282,
+      "grad_norm": 0.04724871652205951,
+      "learning_rate": 0.0002621892131128994,
+      "loss": 0.3555,
+      "step": 400
+    },
+    {
+      "epoch": 0.837160751565762,
+      "grad_norm": 0.04523467438521327,
+      "learning_rate": 0.0002615460479301714,
+      "loss": 0.3809,
+      "step": 401
+    },
+    {
+      "epoch": 0.8392484342379958,
+      "grad_norm": 0.04240307801356352,
+      "learning_rate": 0.0002609021789169571,
+      "loss": 0.3574,
+      "step": 402
+    },
+    {
+      "epoch": 0.8413361169102297,
+      "grad_norm": 0.03998645664485665,
+      "learning_rate": 0.0002602576134364369,
+      "loss": 0.3008,
+      "step": 403
+    },
+    {
+      "epoch": 0.8434237995824635,
+      "grad_norm": 0.045774083398434376,
+      "learning_rate": 0.0002596123588597555,
+      "loss": 0.4043,
+      "step": 404
+    },
+    {
+      "epoch": 0.8455114822546973,
+      "grad_norm": 0.043894774464113946,
+      "learning_rate": 0.0002589664225659382,
+      "loss": 0.3301,
+      "step": 405
+    },
+    {
+      "epoch": 0.8475991649269311,
+      "grad_norm": 0.04133334314623751,
+      "learning_rate": 0.0002583198119418063,
+      "loss": 0.2637,
+      "step": 406
+    },
+    {
+      "epoch": 0.8496868475991649,
+      "grad_norm": 0.037286338050560974,
+      "learning_rate": 0.00025767253438189255,
+      "loss": 0.3164,
+      "step": 407
+    },
+    {
+      "epoch": 0.8517745302713987,
+      "grad_norm": 0.04288387043023777,
+      "learning_rate": 0.0002570245972883568,
+      "loss": 0.3164,
+      "step": 408
+    },
+    {
+      "epoch": 0.8538622129436325,
+      "grad_norm": 0.04117605927774004,
+      "learning_rate": 0.000256376008070901,
+      "loss": 0.3457,
+      "step": 409
+    },
+    {
+      "epoch": 0.8559498956158664,
+      "grad_norm": 0.040660730568059064,
+      "learning_rate": 0.00025572677414668497,
+      "loss": 0.3691,
+      "step": 410
+    },
+    {
+      "epoch": 0.8580375782881002,
+      "grad_norm": 0.03473424285365033,
+      "learning_rate": 0.00025507690294024105,
+      "loss": 0.252,
+      "step": 411
+    },
+    {
+      "epoch": 0.860125260960334,
+      "grad_norm": 0.04106073793621654,
+      "learning_rate": 0.00025442640188338945,
+      "loss": 0.2871,
+      "step": 412
+    },
+    {
+      "epoch": 0.8622129436325678,
+      "grad_norm": 0.03780921487044816,
+      "learning_rate": 0.00025377527841515343,
+      "loss": 0.2969,
+      "step": 413
+    },
+    {
+      "epoch": 0.8643006263048016,
+      "grad_norm": 0.03779696762506096,
+      "learning_rate": 0.00025312353998167397,
+      "loss": 0.291,
+      "step": 414
+    },
+    {
+      "epoch": 0.8663883089770354,
+      "grad_norm": 0.04781732258675552,
+      "learning_rate": 0.00025247119403612456,
+      "loss": 0.3301,
+      "step": 415
+    },
+    {
+      "epoch": 0.8684759916492694,
+      "grad_norm": 0.03912752520944136,
+      "learning_rate": 0.0002518182480386261,
+      "loss": 0.3477,
+      "step": 416
+    },
+    {
+      "epoch": 0.8705636743215032,
+      "grad_norm": 0.043540503401128916,
+      "learning_rate": 0.0002511647094561619,
+      "loss": 0.3398,
+      "step": 417
+    },
+    {
+      "epoch": 0.872651356993737,
+      "grad_norm": 0.04207586943763798,
+      "learning_rate": 0.00025051058576249153,
+      "loss": 0.3398,
+      "step": 418
+    },
+    {
+      "epoch": 0.8747390396659708,
+      "grad_norm": 0.03596881591749821,
+      "learning_rate": 0.00024985588443806624,
+      "loss": 0.2715,
+      "step": 419
+    },
+    {
+      "epoch": 0.8768267223382046,
+      "grad_norm": 0.042287007835322306,
+      "learning_rate": 0.00024920061296994253,
+      "loss": 0.3086,
+      "step": 420
+    },
+    {
+      "epoch": 0.8789144050104384,
+      "grad_norm": 0.03856865323635185,
+      "learning_rate": 0.00024854477885169734,
+      "loss": 0.2871,
+      "step": 421
+    },
+    {
+      "epoch": 0.8810020876826722,
+      "grad_norm": 0.03879769693027103,
+      "learning_rate": 0.0002478883895833417,
+      "loss": 0.2617,
+      "step": 422
+    },
+    {
+      "epoch": 0.8830897703549061,
+      "grad_norm": 0.04500551833152804,
+      "learning_rate": 0.00024723145267123556,
+      "loss": 0.2793,
+      "step": 423
+    },
+    {
+      "epoch": 0.8851774530271399,
+      "grad_norm": 0.03925914590146329,
+      "learning_rate": 0.0002465739756280013,
+      "loss": 0.3086,
+      "step": 424
+    },
+    {
+      "epoch": 0.8872651356993737,
+      "grad_norm": 0.04390031842385643,
+      "learning_rate": 0.0002459159659724383,
+      "loss": 0.3594,
+      "step": 425
+    },
+    {
+      "epoch": 0.8893528183716075,
+      "grad_norm": 0.04015458555102483,
+      "learning_rate": 0.00024525743122943684,
+      "loss": 0.2773,
+      "step": 426
+    },
+    {
+      "epoch": 0.8914405010438413,
+      "grad_norm": 0.043251502572368115,
+      "learning_rate": 0.000244598378929892,
+      "loss": 0.3633,
+      "step": 427
+    },
+    {
+      "epoch": 0.8935281837160751,
+      "grad_norm": 0.041718720849523466,
+      "learning_rate": 0.00024393881661061747,
+      "loss": 0.3418,
+      "step": 428
+    },
+    {
+      "epoch": 0.8956158663883089,
+      "grad_norm": 0.03667112112306563,
+      "learning_rate": 0.0002432787518142596,
+      "loss": 0.2754,
+      "step": 429
+    },
+    {
+      "epoch": 0.8977035490605428,
+      "grad_norm": 0.04034875824232465,
+      "learning_rate": 0.0002426181920892108,
+      "loss": 0.3242,
+      "step": 430
+    },
+    {
+      "epoch": 0.8997912317327766,
+      "grad_norm": 0.0416779068349274,
+      "learning_rate": 0.0002419571449895236,
+      "loss": 0.2793,
+      "step": 431
+    },
+    {
+      "epoch": 0.9018789144050104,
+      "grad_norm": 0.04492298155955886,
+      "learning_rate": 0.00024129561807482378,
+      "loss": 0.373,
+      "step": 432
+    },
+    {
+      "epoch": 0.9039665970772442,
+      "grad_norm": 0.0358568800030184,
+      "learning_rate": 0.00024063361891022455,
+      "loss": 0.293,
+      "step": 433
+    },
+    {
+      "epoch": 0.906054279749478,
+      "grad_norm": 0.041240679259523606,
+      "learning_rate": 0.00023997115506623948,
+      "loss": 0.3477,
+      "step": 434
+    },
+    {
+      "epoch": 0.9081419624217119,
+      "grad_norm": 0.04776747282663658,
+      "learning_rate": 0.00023930823411869617,
+      "loss": 0.3633,
+      "step": 435
+    },
+    {
+      "epoch": 0.9102296450939458,
+      "grad_norm": 0.03739632540656231,
+      "learning_rate": 0.00023864486364864963,
+      "loss": 0.3027,
+      "step": 436
+    },
+    {
+      "epoch": 0.9123173277661796,
+      "grad_norm": 0.04013506353404083,
+      "learning_rate": 0.00023798105124229536,
+      "loss": 0.2793,
+      "step": 437
+    },
+    {
+      "epoch": 0.9144050104384134,
+      "grad_norm": 0.044295244931196354,
+      "learning_rate": 0.00023731680449088312,
+      "loss": 0.377,
+      "step": 438
+    },
+    {
+      "epoch": 0.9164926931106472,
+      "grad_norm": 0.03805031241226096,
+      "learning_rate": 0.00023665213099062938,
+      "loss": 0.2559,
+      "step": 439
+    },
+    {
+      "epoch": 0.918580375782881,
+      "grad_norm": 0.03767845062036613,
+      "learning_rate": 0.00023598703834263113,
+      "loss": 0.3125,
+      "step": 440
+    },
+    {
+      "epoch": 0.9206680584551148,
+      "grad_norm": 0.04968215089246422,
+      "learning_rate": 0.0002353215341527785,
+      "loss": 0.3223,
+      "step": 441
+    },
+    {
+      "epoch": 0.9227557411273486,
+      "grad_norm": 0.036246340480605606,
+      "learning_rate": 0.00023465562603166808,
+      "loss": 0.2734,
+      "step": 442
+    },
+    {
+      "epoch": 0.9248434237995825,
+      "grad_norm": 0.034358684978586476,
+      "learning_rate": 0.00023398932159451557,
+      "loss": 0.3457,
+      "step": 443
+    },
+    {
+      "epoch": 0.9269311064718163,
+      "grad_norm": 0.03631746144581004,
+      "learning_rate": 0.000233322628461069,
+      "loss": 0.3672,
+      "step": 444
+    },
+    {
+      "epoch": 0.9290187891440501,
+      "grad_norm": 0.037623854588027666,
+      "learning_rate": 0.0002326555542555215,
+      "loss": 0.3047,
+      "step": 445
+    },
+    {
+      "epoch": 0.9311064718162839,
+      "grad_norm": 0.037886888034629966,
+      "learning_rate": 0.0002319881066064239,
+      "loss": 0.2812,
+      "step": 446
+    },
+    {
+      "epoch": 0.9331941544885177,
+      "grad_norm": 0.039683782098775454,
+      "learning_rate": 0.00023132029314659795,
+      "loss": 0.2969,
+      "step": 447
+    },
+    {
+      "epoch": 0.9352818371607515,
+      "grad_norm": 0.032859528564560474,
+      "learning_rate": 0.00023065212151304843,
+      "loss": 0.2617,
+      "step": 448
+    },
+    {
+      "epoch": 0.9373695198329853,
+      "grad_norm": 0.032092006779359715,
+      "learning_rate": 0.00022998359934687645,
+      "loss": 0.2656,
+      "step": 449
+    },
+    {
+      "epoch": 0.9394572025052192,
+      "grad_norm": 0.03973715339510534,
+      "learning_rate": 0.00022931473429319143,
+      "loss": 0.3242,
+      "step": 450
+    },
+    {
+      "epoch": 0.941544885177453,
+      "grad_norm": 0.040296933695116964,
+      "learning_rate": 0.00022864553400102437,
+      "loss": 0.3281,
+      "step": 451
+    },
+    {
+      "epoch": 0.9436325678496869,
+      "grad_norm": 0.03972360152628702,
+      "learning_rate": 0.00022797600612323965,
+      "loss": 0.3301,
+      "step": 452
+    },
+    {
+      "epoch": 0.9457202505219207,
+      "grad_norm": 0.03196535128099967,
+      "learning_rate": 0.00022730615831644807,
+      "loss": 0.2617,
+      "step": 453
+    },
+    {
+      "epoch": 0.9478079331941545,
+      "grad_norm": 0.041518745854850714,
+      "learning_rate": 0.00022663599824091892,
+      "loss": 0.3535,
+      "step": 454
+    },
+    {
+      "epoch": 0.9498956158663883,
+      "grad_norm": 0.03987544052335102,
+      "learning_rate": 0.00022596553356049283,
+      "loss": 0.3184,
+      "step": 455
+    },
+    {
+      "epoch": 0.9519832985386222,
+      "grad_norm": 0.03790469940548919,
+      "learning_rate": 0.0002252947719424935,
+      "loss": 0.2988,
+      "step": 456
+    },
+    {
+      "epoch": 0.954070981210856,
+      "grad_norm": 0.03494334490842234,
+      "learning_rate": 0.00022462372105764065,
+      "loss": 0.2793,
+      "step": 457
+    },
+    {
+      "epoch": 0.9561586638830898,
+      "grad_norm": 0.03463196563976093,
+      "learning_rate": 0.0002239523885799618,
+      "loss": 0.2637,
+      "step": 458
+    },
+    {
+      "epoch": 0.9582463465553236,
+      "grad_norm": 0.03593899157528181,
+      "learning_rate": 0.0002232807821867048,
+      "loss": 0.3184,
+      "step": 459
+    },
+    {
+      "epoch": 0.9603340292275574,
+      "grad_norm": 0.03890353431042402,
+      "learning_rate": 0.00022260890955825001,
+      "loss": 0.3008,
+      "step": 460
+    },
+    {
+      "epoch": 0.9624217118997912,
+      "grad_norm": 0.033343558090590815,
+      "learning_rate": 0.00022193677837802235,
+      "loss": 0.2891,
+      "step": 461
+    },
+    {
+      "epoch": 0.964509394572025,
+      "grad_norm": 0.037945187747727475,
+      "learning_rate": 0.0002212643963324035,
+      "loss": 0.3574,
+      "step": 462
+    },
+    {
+      "epoch": 0.9665970772442589,
+      "grad_norm": 0.03770022520852902,
+      "learning_rate": 0.00022059177111064391,
+      "loss": 0.3262,
+      "step": 463
+    },
+    {
+      "epoch": 0.9686847599164927,
+      "grad_norm": 0.03875032196468757,
+      "learning_rate": 0.00021991891040477516,
+      "loss": 0.3145,
+      "step": 464
+    },
+    {
+      "epoch": 0.9707724425887265,
+      "grad_norm": 0.03788954429707523,
+      "learning_rate": 0.00021924582190952153,
+      "loss": 0.3008,
+      "step": 465
+    },
+    {
+      "epoch": 0.9728601252609603,
+      "grad_norm": 0.035023715952115834,
+      "learning_rate": 0.00021857251332221256,
+      "loss": 0.2412,
+      "step": 466
+    },
+    {
+      "epoch": 0.9749478079331941,
+      "grad_norm": 0.037210987286251065,
+      "learning_rate": 0.0002178989923426944,
+      "loss": 0.3398,
+      "step": 467
+    },
+    {
+      "epoch": 0.9770354906054279,
+      "grad_norm": 0.03207771649660305,
+      "learning_rate": 0.0002172252666732424,
+      "loss": 0.2578,
+      "step": 468
+    },
+    {
+      "epoch": 0.9791231732776617,
+      "grad_norm": 0.031951869684034837,
+      "learning_rate": 0.00021655134401847245,
+      "loss": 0.2695,
+      "step": 469
+    },
+    {
+      "epoch": 0.9812108559498957,
+      "grad_norm": 0.03674391020210522,
+      "learning_rate": 0.00021587723208525334,
+      "loss": 0.3066,
+      "step": 470
+    },
+    {
+      "epoch": 0.9832985386221295,
+      "grad_norm": 0.035932078354231324,
+      "learning_rate": 0.00021520293858261823,
+      "loss": 0.3086,
+      "step": 471
+    },
+    {
+      "epoch": 0.9853862212943633,
+      "grad_norm": 0.03876657300516048,
+      "learning_rate": 0.00021452847122167677,
+      "loss": 0.3008,
+      "step": 472
+    },
+    {
+      "epoch": 0.9874739039665971,
+      "grad_norm": 0.032766584147543416,
+      "learning_rate": 0.00021385383771552684,
+      "loss": 0.3047,
+      "step": 473
+    },
+    {
+      "epoch": 0.9895615866388309,
+      "grad_norm": 0.04086665364700512,
+      "learning_rate": 0.00021317904577916635,
+      "loss": 0.3223,
+      "step": 474
+    },
+    {
+      "epoch": 0.9916492693110647,
+      "grad_norm": 0.043130532139654114,
+      "learning_rate": 0.0002125041031294049,
+      "loss": 0.3242,
+      "step": 475
+    },
+    {
+      "epoch": 0.9937369519832986,
+      "grad_norm": 0.04273564060035245,
+      "learning_rate": 0.00021182901748477564,
+      "loss": 0.4023,
+      "step": 476
+    },
+    {
+      "epoch": 0.9958246346555324,
+      "grad_norm": 0.03628393896960175,
+      "learning_rate": 0.00021115379656544715,
+      "loss": 0.2969,
+      "step": 477
+    },
+    {
+      "epoch": 0.9979123173277662,
+      "grad_norm": 0.03800046175114251,
+      "learning_rate": 0.00021047844809313467,
+      "loss": 0.332,
+      "step": 478
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.038197115177252855,
+      "learning_rate": 0.00020980297979101252,
+      "loss": 0.2695,
+      "step": 479
+    },
+    {
+      "epoch": 1.0020876826722338,
+      "grad_norm": 0.032423776045183476,
+      "learning_rate": 0.000209127399383625,
+      "loss": 0.2021,
+      "step": 480
+    },
+    {
+      "epoch": 1.0041753653444676,
+      "grad_norm": 0.03224992988614844,
+      "learning_rate": 0.00020845171459679866,
+      "loss": 0.2139,
+      "step": 481
+    },
+    {
+      "epoch": 1.0062630480167014,
+      "grad_norm": 0.03676757603527534,
+      "learning_rate": 0.00020777593315755358,
+      "loss": 0.249,
+      "step": 482
+    },
+    {
+      "epoch": 1.0083507306889352,
+      "grad_norm": 0.03889498255916948,
+      "learning_rate": 0.00020710006279401531,
+      "loss": 0.3047,
+      "step": 483
+    },
+    {
+      "epoch": 1.010438413361169,
+      "grad_norm": 0.03394341377626214,
+      "learning_rate": 0.00020642411123532605,
+      "loss": 0.2598,
+      "step": 484
+    },
+    {
+      "epoch": 1.0125260960334028,
+      "grad_norm": 0.03195140878316827,
+      "learning_rate": 0.0002057480862115569,
+      "loss": 0.2266,
+      "step": 485
+    },
+    {
+      "epoch": 1.0146137787056368,
+      "grad_norm": 0.03236779419185144,
+      "learning_rate": 0.0002050719954536187,
+      "loss": 0.2373,
+      "step": 486
+    },
+    {
+      "epoch": 1.0167014613778707,
+      "grad_norm": 0.033762294027972366,
+      "learning_rate": 0.0002043958466931744,
+      "loss": 0.2324,
+      "step": 487
+    },
+    {
+      "epoch": 1.0187891440501045,
+      "grad_norm": 0.03225364955535692,
+      "learning_rate": 0.00020371964766254988,
+      "loss": 0.2246,
+      "step": 488
+    },
+    {
+      "epoch": 1.0208768267223383,
+      "grad_norm": 0.03549374570066917,
+      "learning_rate": 0.00020304340609464627,
+      "loss": 0.2559,
+      "step": 489
+    },
+    {
+      "epoch": 1.022964509394572,
+      "grad_norm": 0.039376031742883684,
+      "learning_rate": 0.00020236712972285088,
+      "loss": 0.291,
+      "step": 490
+    },
+    {
+      "epoch": 1.0250521920668059,
+      "grad_norm": 0.033112169954733074,
+      "learning_rate": 0.0002016908262809491,
+      "loss": 0.2188,
+      "step": 491
+    },
+    {
+      "epoch": 1.0271398747390397,
+      "grad_norm": 0.03556499196371877,
+      "learning_rate": 0.00020101450350303596,
+      "loss": 0.2363,
+      "step": 492
+    },
+    {
+      "epoch": 1.0292275574112735,
+      "grad_norm": 0.03747952002586393,
+      "learning_rate": 0.0002003381691234275,
+      "loss": 0.2734,
+      "step": 493
+    },
+    {
+      "epoch": 1.0313152400835073,
+      "grad_norm": 0.039212245318626486,
+      "learning_rate": 0.00019966183087657255,
+      "loss": 0.2812,
+      "step": 494
+    },
+    {
+      "epoch": 1.033402922755741,
+      "grad_norm": 0.038556823852449205,
+      "learning_rate": 0.00019898549649696409,
+      "loss": 0.252,
+      "step": 495
+    },
+    {
+      "epoch": 1.0354906054279749,
+      "grad_norm": 0.03678903694409147,
+      "learning_rate": 0.00019830917371905095,
+      "loss": 0.2227,
+      "step": 496
+    },
+    {
+      "epoch": 1.0375782881002087,
+      "grad_norm": 0.035536317646152164,
+      "learning_rate": 0.00019763287027714917,
+      "loss": 0.2451,
+      "step": 497
+    },
+    {
+      "epoch": 1.0396659707724425,
+      "grad_norm": 0.03669268286478292,
+      "learning_rate": 0.00019695659390535377,
+      "loss": 0.2158,
+      "step": 498
+    },
+    {
+      "epoch": 1.0417536534446765,
+      "grad_norm": 0.0365044069944666,
+      "learning_rate": 0.0001962803523374501,
+      "loss": 0.2334,
+      "step": 499
+    },
+    {
+      "epoch": 1.0438413361169103,
+      "grad_norm": 0.04311308542212124,
+      "learning_rate": 0.00019560415330682567,
+      "loss": 0.2754,
+      "step": 500
+    },
+    {
+      "epoch": 1.0459290187891441,
+      "grad_norm": 0.03807877662781654,
+      "learning_rate": 0.00019492800454638128,
+      "loss": 0.2207,
+      "step": 501
+    },
+    {
+      "epoch": 1.048016701461378,
+      "grad_norm": 0.037347719889777004,
+      "learning_rate": 0.00019425191378844315,
+      "loss": 0.2598,
+      "step": 502
+    },
+    {
+      "epoch": 1.0501043841336117,
+      "grad_norm": 0.041720879231187474,
+      "learning_rate": 0.00019357588876467403,
+      "loss": 0.2832,
+      "step": 503
+    },
+    {
+      "epoch": 1.0521920668058455,
+      "grad_norm": 0.03611878979996833,
+      "learning_rate": 0.0001928999372059848,
+      "loss": 0.2305,
+      "step": 504
+    },
+    {
+      "epoch": 1.0542797494780793,
+      "grad_norm": 0.03914121207636423,
+      "learning_rate": 0.00019222406684244652,
+      "loss": 0.2715,
+      "step": 505
+    },
+    {
+      "epoch": 1.0563674321503131,
+      "grad_norm": 0.045008260139354196,
+      "learning_rate": 0.00019154828540320141,
+      "loss": 0.3496,
+      "step": 506
+    },
+    {
+      "epoch": 1.058455114822547,
+      "grad_norm": 0.06925075294932237,
+      "learning_rate": 0.00019087260061637507,
+      "loss": 0.21,
+      "step": 507
+    },
+    {
+      "epoch": 1.0605427974947808,
+      "grad_norm": 0.042949891828889544,
+      "learning_rate": 0.00019019702020898753,
+      "loss": 0.3164,
+      "step": 508
+    },
+    {
+      "epoch": 1.0626304801670146,
+      "grad_norm": 0.03336924307296428,
+      "learning_rate": 0.00018952155190686535,
+      "loss": 0.2334,
+      "step": 509
+    },
+    {
+      "epoch": 1.0647181628392484,
+      "grad_norm": 0.037171672304770545,
+      "learning_rate": 0.00018884620343455292,
+      "loss": 0.2432,
+      "step": 510
+    },
+    {
+      "epoch": 1.0668058455114822,
+      "grad_norm": 0.03845099190077751,
+      "learning_rate": 0.0001881709825152244,
+      "loss": 0.2793,
+      "step": 511
+    },
+    {
+      "epoch": 1.068893528183716,
+      "grad_norm": 0.03411076783878538,
+      "learning_rate": 0.00018749589687059513,
+      "loss": 0.2041,
+      "step": 512
+    },
+    {
+      "epoch": 1.07098121085595,
+      "grad_norm": 0.03701566955647055,
+      "learning_rate": 0.00018682095422083373,
+      "loss": 0.2305,
+      "step": 513
+    },
+    {
+      "epoch": 1.0730688935281838,
+      "grad_norm": 0.03954320852756579,
+      "learning_rate": 0.00018614616228447318,
+      "loss": 0.2354,
+      "step": 514
+    },
+    {
+      "epoch": 1.0751565762004176,
+      "grad_norm": 0.04313297344221612,
+      "learning_rate": 0.00018547152877832328,
+      "loss": 0.2656,
+      "step": 515
+    },
+    {
+      "epoch": 1.0772442588726514,
+      "grad_norm": 0.03799814412465588,
+      "learning_rate": 0.0001847970614173818,
+      "loss": 0.2285,
+      "step": 516
+    },
+    {
+      "epoch": 1.0793319415448852,
+      "grad_norm": 0.04096742193043013,
+      "learning_rate": 0.0001841227679147467,
+      "loss": 0.25,
+      "step": 517
+    },
+    {
+      "epoch": 1.081419624217119,
+      "grad_norm": 0.03795069745138144,
+      "learning_rate": 0.00018344865598152754,
+      "loss": 0.2471,
+      "step": 518
+    },
+    {
+      "epoch": 1.0835073068893528,
+      "grad_norm": 0.04325821353197005,
+      "learning_rate": 0.00018277473332675758,
+      "loss": 0.2754,
+      "step": 519
+    },
+    {
+      "epoch": 1.0855949895615866,
+      "grad_norm": 0.04070425486872679,
+      "learning_rate": 0.0001821010076573056,
+      "loss": 0.2617,
+      "step": 520
+    },
+    {
+      "epoch": 1.0876826722338204,
+      "grad_norm": 0.031587246842812465,
+      "learning_rate": 0.00018142748667778746,
+      "loss": 0.1777,
+      "step": 521
+    },
+    {
+      "epoch": 1.0897703549060542,
+      "grad_norm": 0.036539737613616244,
+      "learning_rate": 0.00018075417809047857,
+      "loss": 0.2773,
+      "step": 522
+    },
+    {
+      "epoch": 1.091858037578288,
+      "grad_norm": 0.041178092025696736,
+      "learning_rate": 0.00018008108959522497,
+      "loss": 0.2676,
+      "step": 523
+    },
+    {
+      "epoch": 1.0939457202505218,
+      "grad_norm": 0.04046942553591543,
+      "learning_rate": 0.00017940822888935616,
+      "loss": 0.2217,
+      "step": 524
+    },
+    {
+      "epoch": 1.0960334029227556,
+      "grad_norm": 0.03449629028853095,
+      "learning_rate": 0.0001787356036675966,
+      "loss": 0.2061,
+      "step": 525
+    },
+    {
+      "epoch": 1.0981210855949897,
+      "grad_norm": 0.03986588393895996,
+      "learning_rate": 0.0001780632216219777,
+      "loss": 0.3125,
+      "step": 526
+    },
+    {
+      "epoch": 1.1002087682672235,
+      "grad_norm": 0.0333875432706545,
+      "learning_rate": 0.00017739109044175003,
+      "loss": 0.1787,
+      "step": 527
+    },
+    {
+      "epoch": 1.1022964509394573,
+      "grad_norm": 0.04046361146111627,
+      "learning_rate": 0.00017671921781329522,
+      "loss": 0.2715,
+      "step": 528
+    },
+    {
+      "epoch": 1.104384133611691,
+      "grad_norm": 0.04262435868048576,
+      "learning_rate": 0.00017604761142003827,
+      "loss": 0.2559,
+      "step": 529
+    },
+    {
+      "epoch": 1.1064718162839249,
+      "grad_norm": 0.03428115599372914,
+      "learning_rate": 0.0001753762789423594,
+      "loss": 0.2061,
+      "step": 530
+    },
+    {
+      "epoch": 1.1085594989561587,
+      "grad_norm": 0.03868138089293832,
+      "learning_rate": 0.00017470522805750654,
+      "loss": 0.2236,
+      "step": 531
+    },
+    {
+      "epoch": 1.1106471816283925,
+      "grad_norm": 0.03475622072699226,
+      "learning_rate": 0.00017403446643950724,
+      "loss": 0.248,
+      "step": 532
+    },
+    {
+      "epoch": 1.1127348643006263,
+      "grad_norm": 0.03979226115172334,
+      "learning_rate": 0.00017336400175908112,
+      "loss": 0.2988,
+      "step": 533
+    },
+    {
+      "epoch": 1.11482254697286,
+      "grad_norm": 0.036336277616070035,
+      "learning_rate": 0.000172693841683552,
+      "loss": 0.2236,
+      "step": 534
+    },
+    {
+      "epoch": 1.116910229645094,
+      "grad_norm": 0.039263581746387936,
+      "learning_rate": 0.0001720239938767604,
+      "loss": 0.332,
+      "step": 535
+    },
+    {
+      "epoch": 1.1189979123173277,
+      "grad_norm": 0.04395872414762282,
+      "learning_rate": 0.00017135446599897565,
+      "loss": 0.2734,
+      "step": 536
+    },
+    {
+      "epoch": 1.1210855949895615,
+      "grad_norm": 0.03867598581621191,
+      "learning_rate": 0.00017068526570680853,
+      "loss": 0.249,
+      "step": 537
+    },
+    {
+      "epoch": 1.1231732776617953,
+      "grad_norm": 0.035662617079617494,
+      "learning_rate": 0.00017001640065312357,
+      "loss": 0.2041,
+      "step": 538
+    },
+    {
+      "epoch": 1.1252609603340291,
+      "grad_norm": 0.04078186211455469,
+      "learning_rate": 0.00016934787848695156,
+      "loss": 0.2119,
+      "step": 539
+    },
+    {
+      "epoch": 1.1273486430062631,
+      "grad_norm": 0.03947326929596921,
+      "learning_rate": 0.00016867970685340215,
+      "loss": 0.2393,
+      "step": 540
+    },
+    {
+      "epoch": 1.129436325678497,
+      "grad_norm": 0.04449645249301931,
+      "learning_rate": 0.00016801189339357616,
+      "loss": 0.3203,
+      "step": 541
+    },
+    {
+      "epoch": 1.1315240083507307,
+      "grad_norm": 0.046340346999829446,
+      "learning_rate": 0.00016734444574447862,
+      "loss": 0.2598,
+      "step": 542
+    },
+    {
+      "epoch": 1.1336116910229646,
+      "grad_norm": 0.03957931201804222,
+      "learning_rate": 0.00016667737153893105,
+      "loss": 0.2402,
+      "step": 543
+    },
+    {
+      "epoch": 1.1356993736951984,
+      "grad_norm": 0.03703757920078097,
+      "learning_rate": 0.00016601067840548453,
+      "loss": 0.1924,
+      "step": 544
+    },
+    {
+      "epoch": 1.1377870563674322,
+      "grad_norm": 0.03682994100461504,
+      "learning_rate": 0.000165344373968332,
+      "loss": 0.2383,
+      "step": 545
+    },
+    {
+      "epoch": 1.139874739039666,
+      "grad_norm": 0.03920731523127231,
+      "learning_rate": 0.00016467846584722153,
+      "loss": 0.2734,
+      "step": 546
+    },
+    {
+      "epoch": 1.1419624217118998,
+      "grad_norm": 0.04276682539134295,
+      "learning_rate": 0.0001640129616573689,
+      "loss": 0.2852,
+      "step": 547
+    },
+    {
+      "epoch": 1.1440501043841336,
+      "grad_norm": 0.03614032190490994,
+      "learning_rate": 0.00016334786900937067,
+      "loss": 0.2354,
+      "step": 548
+    },
+    {
+      "epoch": 1.1461377870563674,
+      "grad_norm": 0.041006439179447914,
+      "learning_rate": 0.00016268319550911695,
+      "loss": 0.2432,
+      "step": 549
+    },
+    {
+      "epoch": 1.1482254697286012,
+      "grad_norm": 0.044517563506872825,
+      "learning_rate": 0.00016201894875770466,
+      "loss": 0.2217,
+      "step": 550
+    },
+    {
+      "epoch": 1.150313152400835,
+      "grad_norm": 0.03655321665381233,
+      "learning_rate": 0.00016135513635135045,
+      "loss": 0.2031,
+      "step": 551
+    },
+    {
+      "epoch": 1.152400835073069,
+      "grad_norm": 0.04079189189547784,
+      "learning_rate": 0.00016069176588130388,
+      "loss": 0.3027,
+      "step": 552
+    },
+    {
+      "epoch": 1.1544885177453028,
+      "grad_norm": 0.03485321031694839,
+      "learning_rate": 0.00016002884493376054,
+      "loss": 0.2275,
+      "step": 553
+    },
+    {
+      "epoch": 1.1565762004175366,
+      "grad_norm": 0.03052141642274577,
+      "learning_rate": 0.00015936638108977546,
+      "loss": 0.1992,
+      "step": 554
+    },
+    {
+      "epoch": 1.1586638830897704,
+      "grad_norm": 0.03724204209659217,
+      "learning_rate": 0.0001587043819251762,
+      "loss": 0.2207,
+      "step": 555
+    },
+    {
+      "epoch": 1.1607515657620042,
+      "grad_norm": 0.03701626710699115,
+      "learning_rate": 0.00015804285501047646,
+      "loss": 0.2041,
+      "step": 556
+    },
+    {
+      "epoch": 1.162839248434238,
+      "grad_norm": 0.04171363907307022,
+      "learning_rate": 0.00015738180791078922,
+      "loss": 0.2715,
+      "step": 557
+    },
+    {
+      "epoch": 1.1649269311064718,
+      "grad_norm": 0.0392243815886844,
+      "learning_rate": 0.00015672124818574042,
+      "loss": 0.2734,
+      "step": 558
+    },
+    {
+      "epoch": 1.1670146137787056,
+      "grad_norm": 0.034576244303318175,
+      "learning_rate": 0.0001560611833893826,
+      "loss": 0.207,
+      "step": 559
+    },
+    {
+      "epoch": 1.1691022964509394,
+      "grad_norm": 0.03659195217316462,
+      "learning_rate": 0.0001554016210701081,
+      "loss": 0.2178,
+      "step": 560
+    },
+    {
+      "epoch": 1.1711899791231732,
+      "grad_norm": 0.04227483515507797,
+      "learning_rate": 0.00015474256877056326,
+      "loss": 0.3086,
+      "step": 561
+    },
+    {
+      "epoch": 1.173277661795407,
+      "grad_norm": 0.0359857035728927,
+      "learning_rate": 0.00015408403402756177,
+      "loss": 0.2617,
+      "step": 562
+    },
+    {
+      "epoch": 1.1753653444676408,
+      "grad_norm": 0.03680442854270533,
+      "learning_rate": 0.0001534260243719988,
+      "loss": 0.209,
+      "step": 563
+    },
+    {
+      "epoch": 1.1774530271398747,
+      "grad_norm": 0.03670404417452906,
+      "learning_rate": 0.0001527685473287645,
+      "loss": 0.2197,
+      "step": 564
+    },
+    {
+      "epoch": 1.1795407098121085,
+      "grad_norm": 0.0393433887764882,
+      "learning_rate": 0.00015211161041665832,
+      "loss": 0.2324,
+      "step": 565
+    },
+    {
+      "epoch": 1.1816283924843423,
+      "grad_norm": 0.03642939621217633,
+      "learning_rate": 0.00015145522114830268,
+      "loss": 0.2393,
+      "step": 566
+    },
+    {
+      "epoch": 1.1837160751565763,
+      "grad_norm": 0.03878579783362816,
+      "learning_rate": 0.00015079938703005752,
+      "loss": 0.2559,
+      "step": 567
+    },
+    {
+      "epoch": 1.18580375782881,
+      "grad_norm": 0.040970554736426454,
+      "learning_rate": 0.0001501441155619338,
+      "loss": 0.2139,
+      "step": 568
+    },
+    {
+      "epoch": 1.187891440501044,
+      "grad_norm": 0.036820951922352504,
+      "learning_rate": 0.0001494894142375085,
+      "loss": 0.1719,
+      "step": 569
+    },
+    {
+      "epoch": 1.1899791231732777,
+      "grad_norm": 0.04473872009108077,
+      "learning_rate": 0.00014883529054383817,
+      "loss": 0.2734,
+      "step": 570
+    },
+    {
+      "epoch": 1.1920668058455115,
+      "grad_norm": 0.037460855453091094,
+      "learning_rate": 0.00014818175196137392,
+      "loss": 0.2334,
+      "step": 571
+    },
+    {
+      "epoch": 1.1941544885177453,
+      "grad_norm": 0.03980728407908227,
+      "learning_rate": 0.00014752880596387551,
+      "loss": 0.2295,
+      "step": 572
+    },
+    {
+      "epoch": 1.196242171189979,
+      "grad_norm": 0.03562870229391469,
+      "learning_rate": 0.00014687646001832608,
+      "loss": 0.1953,
+      "step": 573
+    },
+    {
+      "epoch": 1.198329853862213,
+      "grad_norm": 0.0412827193725494,
+      "learning_rate": 0.00014622472158484654,
+      "loss": 0.25,
+      "step": 574
+    },
+    {
+      "epoch": 1.2004175365344467,
+      "grad_norm": 0.03634424947703343,
+      "learning_rate": 0.00014557359811661054,
+      "loss": 0.207,
+      "step": 575
+    },
+    {
+      "epoch": 1.2025052192066805,
+      "grad_norm": 0.036441371511458535,
+      "learning_rate": 0.000144923097059759,
+      "loss": 0.249,
+      "step": 576
+    },
+    {
+      "epoch": 1.2045929018789143,
+      "grad_norm": 0.04200794927737242,
+      "learning_rate": 0.00014427322585331505,
+      "loss": 0.3027,
+      "step": 577
+    },
+    {
+      "epoch": 1.2066805845511483,
+      "grad_norm": 0.03620959513479551,
+      "learning_rate": 0.00014362399192909908,
+      "loss": 0.2324,
+      "step": 578
+    },
+    {
+      "epoch": 1.2087682672233822,
+      "grad_norm": 0.03930165573165628,
+      "learning_rate": 0.0001429754027116433,
+      "loss": 0.2812,
+      "step": 579
+    },
+    {
+      "epoch": 1.210855949895616,
+      "grad_norm": 0.040322537436277076,
+      "learning_rate": 0.00014232746561810747,
+      "loss": 0.2773,
+      "step": 580
+    },
+    {
+      "epoch": 1.2129436325678498,
+      "grad_norm": 0.03781024017014511,
+      "learning_rate": 0.00014168018805819376,
+      "loss": 0.2354,
+      "step": 581
+    },
+    {
+      "epoch": 1.2150313152400836,
+      "grad_norm": 0.038595664424750396,
+      "learning_rate": 0.00014103357743406182,
+      "loss": 0.2715,
+      "step": 582
+    },
+    {
+      "epoch": 1.2171189979123174,
+      "grad_norm": 0.034975311451833,
+      "learning_rate": 0.00014038764114024456,
+      "loss": 0.2041,
+      "step": 583
+    },
+    {
+      "epoch": 1.2192066805845512,
+      "grad_norm": 0.03825635258528123,
+      "learning_rate": 0.00013974238656356313,
+      "loss": 0.2344,
+      "step": 584
+    },
+    {
+      "epoch": 1.221294363256785,
+      "grad_norm": 0.03735073552974533,
+      "learning_rate": 0.0001390978210830429,
+      "loss": 0.2559,
+      "step": 585
+    },
+    {
+      "epoch": 1.2233820459290188,
+      "grad_norm": 0.040138739862162154,
+      "learning_rate": 0.00013845395206982863,
+      "loss": 0.2656,
+      "step": 586
+    },
+    {
+      "epoch": 1.2254697286012526,
+      "grad_norm": 0.04418658015359039,
+      "learning_rate": 0.00013781078688710063,
+      "loss": 0.291,
+      "step": 587
+    },
+    {
+      "epoch": 1.2275574112734864,
+      "grad_norm": 0.04239030702531795,
+      "learning_rate": 0.00013716833288999005,
+      "loss": 0.2637,
+      "step": 588
+    },
+    {
+      "epoch": 1.2296450939457202,
+      "grad_norm": 0.03818928317637622,
+      "learning_rate": 0.00013652659742549532,
+      "loss": 0.252,
+      "step": 589
+    },
+    {
+      "epoch": 1.231732776617954,
+      "grad_norm": 0.033789428188020935,
+      "learning_rate": 0.00013588558783239762,
+      "loss": 0.1826,
+      "step": 590
+    },
+    {
+      "epoch": 1.2338204592901878,
+      "grad_norm": 0.03479799715006895,
+      "learning_rate": 0.00013524531144117736,
+      "loss": 0.208,
+      "step": 591
+    },
+    {
+      "epoch": 1.2359081419624216,
+      "grad_norm": 0.0407139048904091,
+      "learning_rate": 0.00013460577557393,
+      "loss": 0.2832,
+      "step": 592
+    },
+    {
+      "epoch": 1.2379958246346556,
+      "grad_norm": 0.04211809721321409,
+      "learning_rate": 0.00013396698754428266,
+      "loss": 0.2402,
+      "step": 593
+    },
+    {
+      "epoch": 1.2400835073068894,
+      "grad_norm": 0.038313063106902565,
+      "learning_rate": 0.00013332895465731018,
+      "loss": 0.2539,
+      "step": 594
+    },
+    {
+      "epoch": 1.2421711899791232,
+      "grad_norm": 0.034451480337492464,
+      "learning_rate": 0.00013269168420945178,
+      "loss": 0.209,
+      "step": 595
+    },
+    {
+      "epoch": 1.244258872651357,
+      "grad_norm": 0.042570134603052,
+      "learning_rate": 0.0001320551834884276,
+      "loss": 0.2734,
+      "step": 596
+    },
+    {
+      "epoch": 1.2463465553235908,
+      "grad_norm": 0.041331289867428476,
+      "learning_rate": 0.00013141945977315515,
+      "loss": 0.3066,
+      "step": 597
+    },
+    {
+      "epoch": 1.2484342379958246,
+      "grad_norm": 0.03567452834168155,
+      "learning_rate": 0.00013078452033366635,
+      "loss": 0.2129,
+      "step": 598
+    },
+    {
+      "epoch": 1.2505219206680585,
+      "grad_norm": 0.03702357216857016,
+      "learning_rate": 0.00013015037243102425,
+      "loss": 0.2471,
+      "step": 599
+    },
+    {
+      "epoch": 1.2526096033402923,
+      "grad_norm": 0.03423169810999704,
+      "learning_rate": 0.00012951702331724007,
+      "loss": 0.2236,
+      "step": 600
+    },
+    {
+      "epoch": 1.254697286012526,
+      "grad_norm": 0.04865070295457581,
+      "learning_rate": 0.00012888448023519006,
+      "loss": 0.3223,
+      "step": 601
+    },
+    {
+      "epoch": 1.2567849686847599,
+      "grad_norm": 0.031921521655174964,
+      "learning_rate": 0.000128252750418533,
+      "loss": 0.1934,
+      "step": 602
+    },
+    {
+      "epoch": 1.2588726513569937,
+      "grad_norm": 0.03207714489804328,
+      "learning_rate": 0.0001276218410916272,
+      "loss": 0.1768,
+      "step": 603
+    },
+    {
+      "epoch": 1.2609603340292277,
+      "grad_norm": 0.03764702756962339,
+      "learning_rate": 0.000126991759469448,
+      "loss": 0.207,
+      "step": 604
+    },
+    {
+      "epoch": 1.2630480167014615,
+      "grad_norm": 0.03494076286574369,
+      "learning_rate": 0.0001263625127575052,
+      "loss": 0.2129,
+      "step": 605
+    },
+    {
+      "epoch": 1.2651356993736953,
+      "grad_norm": 0.03530986655804941,
+      "learning_rate": 0.00012573410815176083,
+      "loss": 0.2617,
+      "step": 606
+    },
+    {
+      "epoch": 1.267223382045929,
+      "grad_norm": 0.03741233532866471,
+      "learning_rate": 0.00012510655283854658,
+      "loss": 0.1914,
+      "step": 607
+    },
+    {
+      "epoch": 1.269311064718163,
+      "grad_norm": 0.03133581344814849,
+      "learning_rate": 0.00012447985399448194,
+      "loss": 0.1631,
+      "step": 608
+    },
+    {
+      "epoch": 1.2713987473903967,
+      "grad_norm": 0.03819672658693745,
+      "learning_rate": 0.00012385401878639175,
+      "loss": 0.2891,
+      "step": 609
+    },
+    {
+      "epoch": 1.2734864300626305,
+      "grad_norm": 0.04476513417542424,
+      "learning_rate": 0.00012322905437122468,
+      "loss": 0.3379,
+      "step": 610
+    },
+    {
+      "epoch": 1.2755741127348643,
+      "grad_norm": 0.037096901415977844,
+      "learning_rate": 0.00012260496789597093,
+      "loss": 0.2637,
+      "step": 611
+    },
+    {
+      "epoch": 1.2776617954070981,
+      "grad_norm": 0.033560968570005754,
+      "learning_rate": 0.00012198176649758091,
+      "loss": 0.1768,
+      "step": 612
+    },
+    {
+      "epoch": 1.279749478079332,
+      "grad_norm": 0.03915286043749694,
+      "learning_rate": 0.00012135945730288333,
+      "loss": 0.2402,
+      "step": 613
+    },
+    {
+      "epoch": 1.2818371607515657,
+      "grad_norm": 0.04243780465562693,
+      "learning_rate": 0.00012073804742850385,
+      "loss": 0.2266,
+      "step": 614
+    },
+    {
+      "epoch": 1.2839248434237995,
+      "grad_norm": 0.035340613835254,
+      "learning_rate": 0.00012011754398078369,
+      "loss": 0.2021,
+      "step": 615
+    },
+    {
+      "epoch": 1.2860125260960333,
+      "grad_norm": 0.038064762794890517,
+      "learning_rate": 0.00011949795405569813,
+      "loss": 0.2373,
+      "step": 616
+    },
+    {
+      "epoch": 1.2881002087682671,
+      "grad_norm": 0.03189685893047627,
+      "learning_rate": 0.00011887928473877581,
+      "loss": 0.2178,
+      "step": 617
+    },
+    {
+      "epoch": 1.290187891440501,
+      "grad_norm": 0.04119691632833074,
+      "learning_rate": 0.00011826154310501723,
+      "loss": 0.2832,
+      "step": 618
+    },
+    {
+      "epoch": 1.2922755741127347,
+      "grad_norm": 0.035970451962649,
+      "learning_rate": 0.00011764473621881427,
+      "loss": 0.2295,
+      "step": 619
+    },
+    {
+      "epoch": 1.2943632567849686,
+      "grad_norm": 0.03644862411137844,
+      "learning_rate": 0.000117028871133869,
+      "loss": 0.293,
+      "step": 620
+    },
+    {
+      "epoch": 1.2964509394572026,
+      "grad_norm": 0.04232916111512329,
+      "learning_rate": 0.00011641395489311334,
+      "loss": 0.2432,
+      "step": 621
+    },
+    {
+      "epoch": 1.2985386221294364,
+      "grad_norm": 0.03815713518197317,
+      "learning_rate": 0.00011579999452862834,
+      "loss": 0.2363,
+      "step": 622
+    },
+    {
+      "epoch": 1.3006263048016702,
+      "grad_norm": 0.03522231596610688,
+      "learning_rate": 0.00011518699706156373,
+      "loss": 0.2275,
+      "step": 623
+    },
+    {
+      "epoch": 1.302713987473904,
+      "grad_norm": 0.03763846715271956,
+      "learning_rate": 0.00011457496950205784,
+      "loss": 0.2139,
+      "step": 624
+    },
+    {
+      "epoch": 1.3048016701461378,
+      "grad_norm": 0.03937303225722648,
+      "learning_rate": 0.00011396391884915707,
+      "loss": 0.2246,
+      "step": 625
+    },
+    {
+      "epoch": 1.3068893528183716,
+      "grad_norm": 0.03259155185424707,
+      "learning_rate": 0.00011335385209073645,
+      "loss": 0.2148,
+      "step": 626
+    },
+    {
+      "epoch": 1.3089770354906054,
+      "grad_norm": 0.036437191147767146,
+      "learning_rate": 0.00011274477620341906,
+      "loss": 0.2168,
+      "step": 627
+    },
+    {
+      "epoch": 1.3110647181628392,
+      "grad_norm": 0.03676944565811173,
+      "learning_rate": 0.00011213669815249659,
+      "loss": 0.21,
+      "step": 628
+    },
+    {
+      "epoch": 1.313152400835073,
+      "grad_norm": 0.03659844810941453,
+      "learning_rate": 0.00011152962489184955,
+      "loss": 0.2139,
+      "step": 629
+    },
+    {
+      "epoch": 1.3152400835073068,
+      "grad_norm": 0.03855990494361526,
+      "learning_rate": 0.0001109235633638682,
+      "loss": 0.2373,
+      "step": 630
+    },
+    {
+      "epoch": 1.3173277661795408,
+      "grad_norm": 0.034719164079528235,
+      "learning_rate": 0.00011031852049937237,
+      "loss": 0.2236,
+      "step": 631
+    },
+    {
+      "epoch": 1.3194154488517746,
+      "grad_norm": 0.03650236752350518,
+      "learning_rate": 0.00010971450321753276,
+      "loss": 0.2197,
+      "step": 632
+    },
+    {
+      "epoch": 1.3215031315240084,
+      "grad_norm": 0.035674494077417296,
+      "learning_rate": 0.00010911151842579195,
+      "loss": 0.25,
+      "step": 633
+    },
+    {
+      "epoch": 1.3235908141962422,
+      "grad_norm": 0.040631215715819284,
+      "learning_rate": 0.00010850957301978462,
+      "loss": 0.2891,
+      "step": 634
+    },
+    {
+      "epoch": 1.325678496868476,
+      "grad_norm": 0.03124048788618308,
+      "learning_rate": 0.00010790867388325951,
+      "loss": 0.208,
+      "step": 635
+    },
+    {
+      "epoch": 1.3277661795407099,
+      "grad_norm": 0.04271594640915922,
+      "learning_rate": 0.0001073088278880005,
+      "loss": 0.3184,
+      "step": 636
+    },
+    {
+      "epoch": 1.3298538622129437,
+      "grad_norm": 0.03507030432270385,
+      "learning_rate": 0.0001067100418937477,
+      "loss": 0.208,
+      "step": 637
+    },
+    {
+      "epoch": 1.3319415448851775,
+      "grad_norm": 0.04020676296142397,
+      "learning_rate": 0.00010611232274811929,
+      "loss": 0.2988,
+      "step": 638
+    },
+    {
+      "epoch": 1.3340292275574113,
+      "grad_norm": 0.034794700042634165,
+      "learning_rate": 0.00010551567728653311,
+      "loss": 0.2061,
+      "step": 639
+    },
+    {
+      "epoch": 1.336116910229645,
+      "grad_norm": 0.04251059198573156,
+      "learning_rate": 0.00010492011233212871,
+      "loss": 0.2578,
+      "step": 640
+    },
+    {
+      "epoch": 1.3382045929018789,
+      "grad_norm": 0.031789932037755535,
+      "learning_rate": 0.0001043256346956889,
+      "loss": 0.2109,
+      "step": 641
+    },
+    {
+      "epoch": 1.3402922755741127,
+      "grad_norm": 0.039155457576247846,
+      "learning_rate": 0.00010373225117556224,
+      "loss": 0.2422,
+      "step": 642
+    },
+    {
+      "epoch": 1.3423799582463465,
+      "grad_norm": 0.035388665066062634,
+      "learning_rate": 0.00010313996855758504,
+      "loss": 0.2002,
+      "step": 643
+    },
+    {
+      "epoch": 1.3444676409185803,
+      "grad_norm": 0.0385841616358797,
+      "learning_rate": 0.00010254879361500407,
+      "loss": 0.2041,
+      "step": 644
+    },
+    {
+      "epoch": 1.346555323590814,
+      "grad_norm": 0.03971686342813934,
+      "learning_rate": 0.00010195873310839874,
+      "loss": 0.2256,
+      "step": 645
+    },
+    {
+      "epoch": 1.348643006263048,
+      "grad_norm": 0.04152102958645101,
+      "learning_rate": 0.00010136979378560398,
+      "loss": 0.2734,
+      "step": 646
+    },
+    {
+      "epoch": 1.350730688935282,
+      "grad_norm": 0.036651673265288884,
+      "learning_rate": 0.00010078198238163299,
+      "loss": 0.2148,
+      "step": 647
+    },
+    {
+      "epoch": 1.3528183716075157,
+      "grad_norm": 0.032824107867213306,
+      "learning_rate": 0.00010019530561860051,
+      "loss": 0.208,
+      "step": 648
+    },
+    {
+      "epoch": 1.3549060542797495,
+      "grad_norm": 0.03503308809553616,
+      "learning_rate": 9.96097702056455e-05,
+      "loss": 0.2041,
+      "step": 649
+    },
+    {
+      "epoch": 1.3569937369519833,
+      "grad_norm": 0.0404355664516907,
+      "learning_rate": 9.902538283885465e-05,
+      "loss": 0.2021,
+      "step": 650
+    },
+    {
+      "epoch": 1.3590814196242171,
+      "grad_norm": 0.047056225295053385,
+      "learning_rate": 9.844215020118576e-05,
+      "loss": 0.3125,
+      "step": 651
+    },
+    {
+      "epoch": 1.361169102296451,
+      "grad_norm": 0.0328129234031199,
+      "learning_rate": 9.78600789623916e-05,
+      "loss": 0.1973,
+      "step": 652
+    },
+    {
+      "epoch": 1.3632567849686847,
+      "grad_norm": 0.03175285628250321,
+      "learning_rate": 9.727917577894297e-05,
+      "loss": 0.1914,
+      "step": 653
+    },
+    {
+      "epoch": 1.3653444676409185,
+      "grad_norm": 0.04520466671433204,
+      "learning_rate": 9.669944729395316e-05,
+      "loss": 0.3125,
+      "step": 654
+    },
+    {
+      "epoch": 1.3674321503131524,
+      "grad_norm": 0.03901825127204986,
+      "learning_rate": 9.612090013710195e-05,
+      "loss": 0.2412,
+      "step": 655
+    },
+    {
+      "epoch": 1.3695198329853862,
+      "grad_norm": 0.037647523878022705,
+      "learning_rate": 9.554354092455949e-05,
+      "loss": 0.2227,
+      "step": 656
+    },
+    {
+      "epoch": 1.3716075156576202,
+      "grad_norm": 0.03796904304768066,
+      "learning_rate": 9.496737625891076e-05,
+      "loss": 0.2402,
+      "step": 657
+    },
+    {
+      "epoch": 1.373695198329854,
+      "grad_norm": 0.036620990435136655,
+      "learning_rate": 9.439241272908012e-05,
+      "loss": 0.21,
+      "step": 658
+    },
+    {
+      "epoch": 1.3757828810020878,
+      "grad_norm": 0.03753406391892656,
+      "learning_rate": 9.381865691025613e-05,
+      "loss": 0.2471,
+      "step": 659
+    },
+    {
+      "epoch": 1.3778705636743216,
+      "grad_norm": 0.03675743788954959,
+      "learning_rate": 9.324611536381591e-05,
+      "loss": 0.2256,
+      "step": 660
+    },
+    {
+      "epoch": 1.3799582463465554,
+      "grad_norm": 0.04379732725831618,
+      "learning_rate": 9.267479463725048e-05,
+      "loss": 0.3184,
+      "step": 661
+    },
+    {
+      "epoch": 1.3820459290187892,
+      "grad_norm": 0.04108282244231379,
+      "learning_rate": 9.210470126408966e-05,
+      "loss": 0.3145,
+      "step": 662
+    },
+    {
+      "epoch": 1.384133611691023,
+      "grad_norm": 0.033523837380523475,
+      "learning_rate": 9.15358417638277e-05,
+      "loss": 0.2021,
+      "step": 663
+    },
+    {
+      "epoch": 1.3862212943632568,
+      "grad_norm": 0.03998207042720944,
+      "learning_rate": 9.096822264184825e-05,
+      "loss": 0.2441,
+      "step": 664
+    },
+    {
+      "epoch": 1.3883089770354906,
+      "grad_norm": 0.037635714292856164,
+      "learning_rate": 9.040185038935029e-05,
+      "loss": 0.1934,
+      "step": 665
+    },
+    {
+      "epoch": 1.3903966597077244,
+      "grad_norm": 0.04063187756459359,
+      "learning_rate": 8.983673148327369e-05,
+      "loss": 0.2812,
+      "step": 666
+    },
+    {
+      "epoch": 1.3924843423799582,
+      "grad_norm": 0.03568609472936473,
+      "learning_rate": 8.927287238622555e-05,
+      "loss": 0.2578,
+      "step": 667
+    },
+    {
+      "epoch": 1.394572025052192,
+      "grad_norm": 0.044075873922673904,
+      "learning_rate": 8.871027954640567e-05,
+      "loss": 0.2949,
+      "step": 668
+    },
+    {
+      "epoch": 1.3966597077244258,
+      "grad_norm": 0.04018365813830774,
+      "learning_rate": 8.814895939753331e-05,
+      "loss": 0.3203,
+      "step": 669
+    },
+    {
+      "epoch": 1.3987473903966596,
+      "grad_norm": 0.038371138538880856,
+      "learning_rate": 8.758891835877335e-05,
+      "loss": 0.2451,
+      "step": 670
+    },
+    {
+      "epoch": 1.4008350730688934,
+      "grad_norm": 0.03644115907045684,
+      "learning_rate": 8.703016283466323e-05,
+      "loss": 0.2109,
+      "step": 671
+    },
+    {
+      "epoch": 1.4029227557411272,
+      "grad_norm": 0.0355184313530291,
+      "learning_rate": 8.64726992150391e-05,
+      "loss": 0.2256,
+      "step": 672
+    },
+    {
+      "epoch": 1.405010438413361,
+      "grad_norm": 0.032473676946887235,
+      "learning_rate": 8.59165338749632e-05,
+      "loss": 0.2266,
+      "step": 673
+    },
+    {
+      "epoch": 1.407098121085595,
+      "grad_norm": 0.03821259918901156,
+      "learning_rate": 8.536167317465114e-05,
+      "loss": 0.2305,
+      "step": 674
+    },
+    {
+      "epoch": 1.4091858037578289,
+      "grad_norm": 0.035842354722894494,
+      "learning_rate": 8.480812345939855e-05,
+      "loss": 0.2363,
+      "step": 675
+    },
+    {
+      "epoch": 1.4112734864300627,
+      "grad_norm": 0.034936854456220405,
+      "learning_rate": 8.4255891059509e-05,
+      "loss": 0.2344,
+      "step": 676
+    },
+    {
+      "epoch": 1.4133611691022965,
+      "grad_norm": 0.0327536953633252,
+      "learning_rate": 8.370498229022133e-05,
+      "loss": 0.2354,
+      "step": 677
+    },
+    {
+      "epoch": 1.4154488517745303,
+      "grad_norm": 0.03332967518468726,
+      "learning_rate": 8.315540345163783e-05,
+      "loss": 0.21,
+      "step": 678
+    },
+    {
+      "epoch": 1.417536534446764,
+      "grad_norm": 0.039757185578627195,
+      "learning_rate": 8.26071608286517e-05,
+      "loss": 0.2451,
+      "step": 679
+    },
+    {
+      "epoch": 1.4196242171189979,
+      "grad_norm": 0.033351639676043855,
+      "learning_rate": 8.206026069087538e-05,
+      "loss": 0.1963,
+      "step": 680
+    },
+    {
+      "epoch": 1.4217118997912317,
+      "grad_norm": 0.036626791510792185,
+      "learning_rate": 8.151470929256893e-05,
+      "loss": 0.2471,
+      "step": 681
+    },
+    {
+      "epoch": 1.4237995824634655,
+      "grad_norm": 0.03663750757496257,
+      "learning_rate": 8.097051287256854e-05,
+      "loss": 0.2559,
+      "step": 682
+    },
+    {
+      "epoch": 1.4258872651356993,
+      "grad_norm": 0.03533804105379715,
+      "learning_rate": 8.042767765421499e-05,
+      "loss": 0.2334,
+      "step": 683
+    },
+    {
+      "epoch": 1.4279749478079333,
+      "grad_norm": 0.03548267765082115,
+      "learning_rate": 7.988620984528248e-05,
+      "loss": 0.1992,
+      "step": 684
+    },
+    {
+      "epoch": 1.4300626304801671,
+      "grad_norm": 0.03365970412354974,
+      "learning_rate": 7.934611563790803e-05,
+      "loss": 0.2129,
+      "step": 685
+    },
+    {
+      "epoch": 1.432150313152401,
+      "grad_norm": 0.036222173404344804,
+      "learning_rate": 7.880740120852012e-05,
+      "loss": 0.2441,
+      "step": 686
+    },
+    {
+      "epoch": 1.4342379958246347,
+      "grad_norm": 0.03545069244831471,
+      "learning_rate": 7.827007271776843e-05,
+      "loss": 0.2246,
+      "step": 687
+    },
+    {
+      "epoch": 1.4363256784968685,
+      "grad_norm": 0.036141104648819485,
+      "learning_rate": 7.773413631045314e-05,
+      "loss": 0.2334,
+      "step": 688
+    },
+    {
+      "epoch": 1.4384133611691023,
+      "grad_norm": 0.03552223808507066,
+      "learning_rate": 7.719959811545512e-05,
+      "loss": 0.207,
+      "step": 689
+    },
+    {
+      "epoch": 1.4405010438413361,
+      "grad_norm": 0.040531388046511914,
+      "learning_rate": 7.666646424566508e-05,
+      "loss": 0.25,
+      "step": 690
+    },
+    {
+      "epoch": 1.44258872651357,
+      "grad_norm": 0.03642114412895087,
+      "learning_rate": 7.613474079791432e-05,
+      "loss": 0.2275,
+      "step": 691
+    },
+    {
+      "epoch": 1.4446764091858038,
+      "grad_norm": 0.03334766371093395,
+      "learning_rate": 7.56044338529049e-05,
+      "loss": 0.2246,
+      "step": 692
+    },
+    {
+      "epoch": 1.4467640918580376,
+      "grad_norm": 0.038435220776801955,
+      "learning_rate": 7.50755494751398e-05,
+      "loss": 0.2871,
+      "step": 693
+    },
+    {
+      "epoch": 1.4488517745302714,
+      "grad_norm": 0.03639976375126217,
+      "learning_rate": 7.454809371285381e-05,
+      "loss": 0.2236,
+      "step": 694
+    },
+    {
+      "epoch": 1.4509394572025052,
+      "grad_norm": 0.04139491524391803,
+      "learning_rate": 7.402207259794428e-05,
+      "loss": 0.2617,
+      "step": 695
+    },
+    {
+      "epoch": 1.453027139874739,
+      "grad_norm": 0.031147759796875986,
+      "learning_rate": 7.34974921459023e-05,
+      "loss": 0.1904,
+      "step": 696
+    },
+    {
+      "epoch": 1.4551148225469728,
+      "grad_norm": 0.04026433329347746,
+      "learning_rate": 7.297435835574362e-05,
+      "loss": 0.2393,
+      "step": 697
+    },
+    {
+      "epoch": 1.4572025052192066,
+      "grad_norm": 0.03965005806725716,
+      "learning_rate": 7.24526772099403e-05,
+      "loss": 0.2715,
+      "step": 698
+    },
+    {
+      "epoch": 1.4592901878914404,
+      "grad_norm": 0.040893406654203844,
+      "learning_rate": 7.193245467435206e-05,
+      "loss": 0.2471,
+      "step": 699
+    },
+    {
+      "epoch": 1.4613778705636742,
+      "grad_norm": 0.03636456124885245,
+      "learning_rate": 7.141369669815841e-05,
+      "loss": 0.2168,
+      "step": 700
+    },
+    {
+      "epoch": 1.4634655532359082,
+      "grad_norm": 0.03647208734905981,
+      "learning_rate": 7.089640921379026e-05,
+      "loss": 0.2178,
+      "step": 701
+    },
+    {
+      "epoch": 1.465553235908142,
+      "grad_norm": 0.0398294556445215,
+      "learning_rate": 7.038059813686224e-05,
+      "loss": 0.2695,
+      "step": 702
+    },
+    {
+      "epoch": 1.4676409185803758,
+      "grad_norm": 0.03444283718095077,
+      "learning_rate": 6.986626936610491e-05,
+      "loss": 0.2109,
+      "step": 703
+    },
+    {
+      "epoch": 1.4697286012526096,
+      "grad_norm": 0.036117772890174496,
+      "learning_rate": 6.935342878329774e-05,
+      "loss": 0.1855,
+      "step": 704
+    },
+    {
+      "epoch": 1.4718162839248434,
+      "grad_norm": 0.03485285291792577,
+      "learning_rate": 6.884208225320121e-05,
+      "loss": 0.1953,
+      "step": 705
+    },
+    {
+      "epoch": 1.4739039665970772,
+      "grad_norm": 0.03913119058204658,
+      "learning_rate": 6.833223562349018e-05,
+      "loss": 0.291,
+      "step": 706
+    },
+    {
+      "epoch": 1.475991649269311,
+      "grad_norm": 0.03619007457603297,
+      "learning_rate": 6.782389472468687e-05,
+      "loss": 0.2256,
+      "step": 707
+    },
+    {
+      "epoch": 1.4780793319415448,
+      "grad_norm": 0.03954136112257214,
+      "learning_rate": 6.731706537009437e-05,
+      "loss": 0.2148,
+      "step": 708
+    },
+    {
+      "epoch": 1.4801670146137786,
+      "grad_norm": 0.039048253382048426,
+      "learning_rate": 6.68117533557297e-05,
+      "loss": 0.2119,
+      "step": 709
+    },
+    {
+      "epoch": 1.4822546972860124,
+      "grad_norm": 0.03633979845664376,
+      "learning_rate": 6.630796446025793e-05,
+      "loss": 0.207,
+      "step": 710
+    },
+    {
+      "epoch": 1.4843423799582465,
+      "grad_norm": 0.04042749139792089,
+      "learning_rate": 6.580570444492626e-05,
+      "loss": 0.2617,
+      "step": 711
+    },
+    {
+      "epoch": 1.4864300626304803,
+      "grad_norm": 0.04112986991098296,
+      "learning_rate": 6.530497905349753e-05,
+      "loss": 0.2578,
+      "step": 712
+    },
+    {
+      "epoch": 1.488517745302714,
+      "grad_norm": 0.04014939278609065,
+      "learning_rate": 6.480579401218502e-05,
+      "loss": 0.2324,
+      "step": 713
+    },
+    {
+      "epoch": 1.4906054279749479,
+      "grad_norm": 0.03662270065596823,
+      "learning_rate": 6.430815502958674e-05,
+      "loss": 0.1689,
+      "step": 714
+    },
+    {
+      "epoch": 1.4926931106471817,
+      "grad_norm": 0.033001164537146806,
+      "learning_rate": 6.381206779662039e-05,
+      "loss": 0.1826,
+      "step": 715
+    },
+    {
+      "epoch": 1.4947807933194155,
+      "grad_norm": 0.04093484928652971,
+      "learning_rate": 6.331753798645796e-05,
+      "loss": 0.249,
+      "step": 716
+    },
+    {
+      "epoch": 1.4968684759916493,
+      "grad_norm": 0.03400770412940897,
+      "learning_rate": 6.282457125446109e-05,
+      "loss": 0.1709,
+      "step": 717
+    },
+    {
+      "epoch": 1.498956158663883,
+      "grad_norm": 0.03449610130369155,
+      "learning_rate": 6.233317323811615e-05,
+      "loss": 0.2168,
+      "step": 718
+    },
+    {
+      "epoch": 1.501043841336117,
+      "grad_norm": 0.03943007646918304,
+      "learning_rate": 6.184334955697028e-05,
+      "loss": 0.2617,
+      "step": 719
+    },
+    {
+      "epoch": 1.5031315240083507,
+      "grad_norm": 0.04046270086854637,
+      "learning_rate": 6.135510581256647e-05,
+      "loss": 0.2715,
+      "step": 720
+    },
+    {
+      "epoch": 1.5052192066805845,
+      "grad_norm": 0.036042076960770504,
+      "learning_rate": 6.086844758837991e-05,
+      "loss": 0.2275,
+      "step": 721
+    },
+    {
+      "epoch": 1.5073068893528183,
+      "grad_norm": 0.038478916176100005,
+      "learning_rate": 6.0383380449754004e-05,
+      "loss": 0.1924,
+      "step": 722
+    },
+    {
+      "epoch": 1.5093945720250521,
+      "grad_norm": 0.0578411545976244,
+      "learning_rate": 5.98999099438369e-05,
+      "loss": 0.2852,
+      "step": 723
+    },
+    {
+      "epoch": 1.511482254697286,
+      "grad_norm": 0.03739145944754586,
+      "learning_rate": 5.941804159951778e-05,
+      "loss": 0.2197,
+      "step": 724
+    },
+    {
+      "epoch": 1.5135699373695197,
+      "grad_norm": 0.036832588153257496,
+      "learning_rate": 5.893778092736382e-05,
+      "loss": 0.2422,
+      "step": 725
+    },
+    {
+      "epoch": 1.5156576200417535,
+      "grad_norm": 0.03671021108564163,
+      "learning_rate": 5.845913341955711e-05,
+      "loss": 0.1572,
+      "step": 726
+    },
+    {
+      "epoch": 1.5177453027139873,
+      "grad_norm": 0.03573073453794959,
+      "learning_rate": 5.798210454983184e-05,
+      "loss": 0.2168,
+      "step": 727
+    },
+    {
+      "epoch": 1.5198329853862211,
+      "grad_norm": 0.035959632617295005,
+      "learning_rate": 5.7506699773411764e-05,
+      "loss": 0.2451,
+      "step": 728
+    },
+    {
+      "epoch": 1.5219206680584552,
+      "grad_norm": 0.038728578675763334,
+      "learning_rate": 5.703292452694771e-05,
+      "loss": 0.2139,
+      "step": 729
+    },
+    {
+      "epoch": 1.524008350730689,
+      "grad_norm": 0.035308915371426546,
+      "learning_rate": 5.6560784228455586e-05,
+      "loss": 0.2021,
+      "step": 730
+    },
+    {
+      "epoch": 1.5260960334029228,
+      "grad_norm": 0.039840996287124834,
+      "learning_rate": 5.609028427725418e-05,
+      "loss": 0.1953,
+      "step": 731
+    },
+    {
+      "epoch": 1.5281837160751566,
+      "grad_norm": 0.04087634507351788,
+      "learning_rate": 5.562143005390361e-05,
+      "loss": 0.2617,
+      "step": 732
+    },
+    {
+      "epoch": 1.5302713987473904,
+      "grad_norm": 0.03746689779610756,
+      "learning_rate": 5.5154226920143626e-05,
+      "loss": 0.2197,
+      "step": 733
+    },
+    {
+      "epoch": 1.5323590814196242,
+      "grad_norm": 0.04226091666932121,
+      "learning_rate": 5.468868021883256e-05,
+      "loss": 0.2773,
+      "step": 734
+    },
+    {
+      "epoch": 1.534446764091858,
+      "grad_norm": 0.03255698564490662,
+      "learning_rate": 5.422479527388591e-05,
+      "loss": 0.1816,
+      "step": 735
+    },
+    {
+      "epoch": 1.536534446764092,
+      "grad_norm": 0.03534998758506692,
+      "learning_rate": 5.376257739021564e-05,
+      "loss": 0.2158,
+      "step": 736
+    },
+    {
+      "epoch": 1.5386221294363258,
+      "grad_norm": 0.04286917750609741,
+      "learning_rate": 5.330203185366942e-05,
+      "loss": 0.2559,
+      "step": 737
+    },
+    {
+      "epoch": 1.5407098121085596,
+      "grad_norm": 0.03987124501813775,
+      "learning_rate": 5.284316393097042e-05,
+      "loss": 0.2197,
+      "step": 738
+    },
+    {
+      "epoch": 1.5427974947807934,
+      "grad_norm": 0.0357279862009252,
+      "learning_rate": 5.2385978869656705e-05,
+      "loss": 0.209,
+      "step": 739
+    },
+    {
+      "epoch": 1.5448851774530272,
+      "grad_norm": 0.03917856556644758,
+      "learning_rate": 5.193048189802143e-05,
+      "loss": 0.2402,
+      "step": 740
+    },
+    {
+      "epoch": 1.546972860125261,
+      "grad_norm": 0.05340411374410496,
+      "learning_rate": 5.147667822505318e-05,
+      "loss": 0.2695,
+      "step": 741
+    },
+    {
+      "epoch": 1.5490605427974948,
+      "grad_norm": 0.041333285407966595,
+      "learning_rate": 5.1024573040376115e-05,
+      "loss": 0.2969,
+      "step": 742
+    },
+    {
+      "epoch": 1.5511482254697286,
+      "grad_norm": 0.039402130485741114,
+      "learning_rate": 5.057417151419077e-05,
+      "loss": 0.2402,
+      "step": 743
+    },
+    {
+      "epoch": 1.5532359081419624,
+      "grad_norm": 0.035671385682246794,
+      "learning_rate": 5.012547879721494e-05,
+      "loss": 0.2158,
+      "step": 744
+    },
+    {
+      "epoch": 1.5553235908141962,
+      "grad_norm": 0.038417929606667435,
+      "learning_rate": 4.967850002062491e-05,
+      "loss": 0.2236,
+      "step": 745
+    },
+    {
+      "epoch": 1.55741127348643,
+      "grad_norm": 0.03949905474465816,
+      "learning_rate": 4.923324029599632e-05,
+      "loss": 0.2715,
+      "step": 746
+    },
+    {
+      "epoch": 1.5594989561586639,
+      "grad_norm": 0.0352045032702957,
+      "learning_rate": 4.878970471524622e-05,
+      "loss": 0.2275,
+      "step": 747
+    },
+    {
+      "epoch": 1.5615866388308977,
+      "grad_norm": 0.036303248719977955,
+      "learning_rate": 4.834789835057465e-05,
+      "loss": 0.2021,
+      "step": 748
+    },
+    {
+      "epoch": 1.5636743215031315,
+      "grad_norm": 0.03443800350271572,
+      "learning_rate": 4.790782625440655e-05,
+      "loss": 0.1992,
+      "step": 749
+    },
+    {
+      "epoch": 1.5657620041753653,
+      "grad_norm": 0.04231633067945213,
+      "learning_rate": 4.7469493459334066e-05,
+      "loss": 0.2617,
+      "step": 750
+    },
+    {
+      "epoch": 1.567849686847599,
+      "grad_norm": 0.037830454141616246,
+      "learning_rate": 4.70329049780589e-05,
+      "loss": 0.2217,
+      "step": 751
+    },
+    {
+      "epoch": 1.5699373695198329,
+      "grad_norm": 0.037662635732679015,
+      "learning_rate": 4.65980658033353e-05,
+      "loss": 0.2188,
+      "step": 752
+    },
+    {
+      "epoch": 1.5720250521920667,
+      "grad_norm": 0.04067293838178101,
+      "learning_rate": 4.616498090791248e-05,
+      "loss": 0.293,
+      "step": 753
+    },
+    {
+      "epoch": 1.5741127348643005,
+      "grad_norm": 0.03901581955251283,
+      "learning_rate": 4.5733655244478146e-05,
+      "loss": 0.2266,
+      "step": 754
+    },
+    {
+      "epoch": 1.5762004175365343,
+      "grad_norm": 0.03554948290229258,
+      "learning_rate": 4.5304093745601605e-05,
+      "loss": 0.1943,
+      "step": 755
+    },
+    {
+      "epoch": 1.5782881002087683,
+      "grad_norm": 0.0405280460733433,
+      "learning_rate": 4.487630132367764e-05,
+      "loss": 0.2617,
+      "step": 756
+    },
+    {
+      "epoch": 1.5803757828810021,
+      "grad_norm": 0.03823926366289705,
+      "learning_rate": 4.445028287087003e-05,
+      "loss": 0.252,
+      "step": 757
+    },
+    {
+      "epoch": 1.582463465553236,
+      "grad_norm": 0.03721532941881517,
+      "learning_rate": 4.402604325905575e-05,
+      "loss": 0.2129,
+      "step": 758
+    },
+    {
+      "epoch": 1.5845511482254697,
+      "grad_norm": 0.029782665453401677,
+      "learning_rate": 4.360358733976919e-05,
+      "loss": 0.1826,
+      "step": 759
+    },
+    {
+      "epoch": 1.5866388308977035,
+      "grad_norm": 0.03989589641175334,
+      "learning_rate": 4.3182919944146936e-05,
+      "loss": 0.2734,
+      "step": 760
+    },
+    {
+      "epoch": 1.5887265135699373,
+      "grad_norm": 0.04383942244265336,
+      "learning_rate": 4.2764045882872086e-05,
+      "loss": 0.3008,
+      "step": 761
+    },
+    {
+      "epoch": 1.5908141962421714,
+      "grad_norm": 0.034818289814213685,
+      "learning_rate": 4.2346969946119574e-05,
+      "loss": 0.2188,
+      "step": 762
+    },
+    {
+      "epoch": 1.5929018789144052,
+      "grad_norm": 0.03770186022731002,
+      "learning_rate": 4.1931696903501204e-05,
+      "loss": 0.2314,
+      "step": 763
+    },
+    {
+      "epoch": 1.594989561586639,
+      "grad_norm": 0.03973578816213652,
+      "learning_rate": 4.151823150401124e-05,
+      "loss": 0.2402,
+      "step": 764
+    },
+    {
+      "epoch": 1.5970772442588728,
+      "grad_norm": 0.03236007667945323,
+      "learning_rate": 4.1106578475972035e-05,
+      "loss": 0.1924,
+      "step": 765
+    },
+    {
+      "epoch": 1.5991649269311066,
+      "grad_norm": 0.03764508247239107,
+      "learning_rate": 4.069674252697988e-05,
+      "loss": 0.1748,
+      "step": 766
+    },
+    {
+      "epoch": 1.6012526096033404,
+      "grad_norm": 0.03731281581422325,
+      "learning_rate": 4.02887283438514e-05,
+      "loss": 0.2559,
+      "step": 767
+    },
+    {
+      "epoch": 1.6033402922755742,
+      "grad_norm": 0.0372624393342072,
+      "learning_rate": 3.988254059256971e-05,
+      "loss": 0.2188,
+      "step": 768
+    },
+    {
+      "epoch": 1.605427974947808,
+      "grad_norm": 0.039361210964991795,
+      "learning_rate": 3.9478183918231106e-05,
+      "loss": 0.2578,
+      "step": 769
+    },
+    {
+      "epoch": 1.6075156576200418,
+      "grad_norm": 0.03750381742110697,
+      "learning_rate": 3.907566294499201e-05,
+      "loss": 0.1982,
+      "step": 770
+    },
+    {
+      "epoch": 1.6096033402922756,
+      "grad_norm": 0.03771991482783373,
+      "learning_rate": 3.8674982276016205e-05,
+      "loss": 0.2412,
+      "step": 771
+    },
+    {
+      "epoch": 1.6116910229645094,
+      "grad_norm": 0.04091075279365844,
+      "learning_rate": 3.82761464934219e-05,
+      "loss": 0.2793,
+      "step": 772
+    },
+    {
+      "epoch": 1.6137787056367432,
+      "grad_norm": 0.03859340968976196,
+      "learning_rate": 3.787916015822954e-05,
+      "loss": 0.2471,
+      "step": 773
+    },
+    {
+      "epoch": 1.615866388308977,
+      "grad_norm": 0.03200544736921389,
+      "learning_rate": 3.748402781030955e-05,
+      "loss": 0.1826,
+      "step": 774
+    },
+    {
+      "epoch": 1.6179540709812108,
+      "grad_norm": 0.0340269199931104,
+      "learning_rate": 3.709075396833057e-05,
+      "loss": 0.1855,
+      "step": 775
+    },
+    {
+      "epoch": 1.6200417536534446,
+      "grad_norm": 0.038072978162411425,
+      "learning_rate": 3.66993431297076e-05,
+      "loss": 0.2432,
+      "step": 776
+    },
+    {
+      "epoch": 1.6221294363256784,
+      "grad_norm": 0.04102889623967079,
+      "learning_rate": 3.63097997705506e-05,
+      "loss": 0.3145,
+      "step": 777
+    },
+    {
+      "epoch": 1.6242171189979122,
+      "grad_norm": 0.04007477882751308,
+      "learning_rate": 3.592212834561337e-05,
+      "loss": 0.2432,
+      "step": 778
+    },
+    {
+      "epoch": 1.626304801670146,
+      "grad_norm": 0.03819036990721501,
+      "learning_rate": 3.553633328824266e-05,
+      "loss": 0.1973,
+      "step": 779
+    },
+    {
+      "epoch": 1.6283924843423798,
+      "grad_norm": 0.038350798236554316,
+      "learning_rate": 3.515241901032731e-05,
+      "loss": 0.2676,
+      "step": 780
+    },
+    {
+      "epoch": 1.6304801670146136,
+      "grad_norm": 0.03548951603993466,
+      "learning_rate": 3.477038990224786e-05,
+      "loss": 0.2002,
+      "step": 781
+    },
+    {
+      "epoch": 1.6325678496868476,
+      "grad_norm": 0.034615335672358714,
+      "learning_rate": 3.439025033282639e-05,
+      "loss": 0.1963,
+      "step": 782
+    },
+    {
+      "epoch": 1.6346555323590815,
+      "grad_norm": 0.04345337713073096,
+      "learning_rate": 3.401200464927654e-05,
+      "loss": 0.2578,
+      "step": 783
+    },
+    {
+      "epoch": 1.6367432150313153,
+      "grad_norm": 0.03791537054424221,
+      "learning_rate": 3.363565717715373e-05,
+      "loss": 0.2227,
+      "step": 784
+    },
+    {
+      "epoch": 1.638830897703549,
+      "grad_norm": 0.04113195871715081,
+      "learning_rate": 3.326121222030578e-05,
+      "loss": 0.2012,
+      "step": 785
+    },
+    {
+      "epoch": 1.6409185803757829,
+      "grad_norm": 0.041884239195247946,
+      "learning_rate": 3.288867406082372e-05,
+      "loss": 0.2236,
+      "step": 786
+    },
+    {
+      "epoch": 1.6430062630480167,
+      "grad_norm": 0.03652594444269533,
+      "learning_rate": 3.251804695899267e-05,
+      "loss": 0.2139,
+      "step": 787
+    },
+    {
+      "epoch": 1.6450939457202505,
+      "grad_norm": 0.05643896674418451,
+      "learning_rate": 3.214933515324323e-05,
+      "loss": 0.2422,
+      "step": 788
+    },
+    {
+      "epoch": 1.6471816283924845,
+      "grad_norm": 0.04098991740754866,
+      "learning_rate": 3.178254286010296e-05,
+      "loss": 0.2422,
+      "step": 789
+    },
+    {
+      "epoch": 1.6492693110647183,
+      "grad_norm": 0.04098022159134268,
+      "learning_rate": 3.1417674274148276e-05,
+      "loss": 0.2402,
+      "step": 790
+    },
+    {
+      "epoch": 1.651356993736952,
+      "grad_norm": 0.037087942509587524,
+      "learning_rate": 3.105473356795634e-05,
+      "loss": 0.2295,
+      "step": 791
+    },
+    {
+      "epoch": 1.653444676409186,
+      "grad_norm": 0.03775161487801411,
+      "learning_rate": 3.0693724892057396e-05,
+      "loss": 0.207,
+      "step": 792
+    },
+    {
+      "epoch": 1.6555323590814197,
+      "grad_norm": 0.04001811834655189,
+      "learning_rate": 3.0334652374887263e-05,
+      "loss": 0.2773,
+      "step": 793
+    },
+    {
+      "epoch": 1.6576200417536535,
+      "grad_norm": 0.03971800362165568,
+      "learning_rate": 2.997752012274031e-05,
+      "loss": 0.2139,
+      "step": 794
+    },
+    {
+      "epoch": 1.6597077244258873,
+      "grad_norm": 0.03971183328914648,
+      "learning_rate": 2.9622332219722193e-05,
+      "loss": 0.2422,
+      "step": 795
+    },
+    {
+      "epoch": 1.6617954070981211,
+      "grad_norm": 0.03937647373747184,
+      "learning_rate": 2.9269092727703375e-05,
+      "loss": 0.252,
+      "step": 796
+    },
+    {
+      "epoch": 1.663883089770355,
+      "grad_norm": 0.03693074285981622,
+      "learning_rate": 2.891780568627267e-05,
+      "loss": 0.1992,
+      "step": 797
+    },
+    {
+      "epoch": 1.6659707724425887,
+      "grad_norm": 0.03819446049329271,
+      "learning_rate": 2.8568475112690918e-05,
+      "loss": 0.2383,
+      "step": 798
+    },
+    {
+      "epoch": 1.6680584551148225,
+      "grad_norm": 0.03385681858741079,
+      "learning_rate": 2.8221105001845095e-05,
+      "loss": 0.1914,
+      "step": 799
+    },
+    {
+      "epoch": 1.6701461377870563,
+      "grad_norm": 0.03481284448136041,
+      "learning_rate": 2.7875699326202663e-05,
+      "loss": 0.1943,
+      "step": 800
+    },
+    {
+      "epoch": 1.6722338204592901,
+      "grad_norm": 0.038887635380198286,
+      "learning_rate": 2.7532262035766288e-05,
+      "loss": 0.2236,
+      "step": 801
+    },
+    {
+      "epoch": 1.674321503131524,
+      "grad_norm": 0.039268882753983984,
+      "learning_rate": 2.7190797058028207e-05,
+      "loss": 0.2402,
+      "step": 802
+    },
+    {
+      "epoch": 1.6764091858037578,
+      "grad_norm": 0.03561285649555011,
+      "learning_rate": 2.685130829792577e-05,
+      "loss": 0.2041,
+      "step": 803
+    },
+    {
+      "epoch": 1.6784968684759916,
+      "grad_norm": 0.03548955540319579,
+      "learning_rate": 2.6513799637796745e-05,
+      "loss": 0.2188,
+      "step": 804
+    },
+    {
+      "epoch": 1.6805845511482254,
+      "grad_norm": 0.03987654406934741,
+      "learning_rate": 2.61782749373346e-05,
+      "loss": 0.2334,
+      "step": 805
+    },
+    {
+      "epoch": 1.6826722338204592,
+      "grad_norm": 0.03591235670312746,
+      "learning_rate": 2.584473803354468e-05,
+      "loss": 0.2217,
+      "step": 806
+    },
+    {
+      "epoch": 1.684759916492693,
+      "grad_norm": 0.03563754130765487,
+      "learning_rate": 2.5513192740700144e-05,
+      "loss": 0.2061,
+      "step": 807
+    },
+    {
+      "epoch": 1.6868475991649268,
+      "grad_norm": 0.03453991302292082,
+      "learning_rate": 2.5183642850298505e-05,
+      "loss": 0.207,
+      "step": 808
+    },
+    {
+      "epoch": 1.6889352818371608,
+      "grad_norm": 0.040646656045912735,
+      "learning_rate": 2.4856092131018137e-05,
+      "loss": 0.293,
+      "step": 809
+    },
+    {
+      "epoch": 1.6910229645093946,
+      "grad_norm": 0.03639422717002704,
+      "learning_rate": 2.4530544328675186e-05,
+      "loss": 0.2236,
+      "step": 810
+    },
+    {
+      "epoch": 1.6931106471816284,
+      "grad_norm": 0.04124184507078617,
+      "learning_rate": 2.4207003166180765e-05,
+      "loss": 0.2344,
+      "step": 811
+    },
+    {
+      "epoch": 1.6951983298538622,
+      "grad_norm": 0.04821861121500109,
+      "learning_rate": 2.3885472343498515e-05,
+      "loss": 0.2227,
+      "step": 812
+    },
+    {
+      "epoch": 1.697286012526096,
+      "grad_norm": 0.036261622126831795,
+      "learning_rate": 2.3565955537602014e-05,
+      "loss": 0.2266,
+      "step": 813
+    },
+    {
+      "epoch": 1.6993736951983298,
+      "grad_norm": 0.03833360109114772,
+      "learning_rate": 2.3248456402432918e-05,
+      "loss": 0.2598,
+      "step": 814
+    },
+    {
+      "epoch": 1.7014613778705638,
+      "grad_norm": 0.03811210564749943,
+      "learning_rate": 2.293297856885912e-05,
+      "loss": 0.2197,
+      "step": 815
+    },
+    {
+      "epoch": 1.7035490605427976,
+      "grad_norm": 0.03440881800931742,
+      "learning_rate": 2.261952564463332e-05,
+      "loss": 0.1865,
+      "step": 816
+    },
+    {
+      "epoch": 1.7056367432150314,
+      "grad_norm": 0.03784666030148996,
+      "learning_rate": 2.2308101214351628e-05,
+      "loss": 0.2324,
+      "step": 817
+    },
+    {
+      "epoch": 1.7077244258872653,
+      "grad_norm": 0.03706179096561306,
+      "learning_rate": 2.1998708839412597e-05,
+      "loss": 0.21,
+      "step": 818
+    },
+    {
+      "epoch": 1.709812108559499,
+      "grad_norm": 0.036230331032483534,
+      "learning_rate": 2.1691352057976565e-05,
+      "loss": 0.2471,
+      "step": 819
+    },
+    {
+      "epoch": 1.7118997912317329,
+      "grad_norm": 0.04090973690460557,
+      "learning_rate": 2.138603438492517e-05,
+      "loss": 0.2637,
+      "step": 820
+    },
+    {
+      "epoch": 1.7139874739039667,
+      "grad_norm": 0.051662701812649245,
+      "learning_rate": 2.108275931182111e-05,
+      "loss": 0.1953,
+      "step": 821
+    },
+    {
+      "epoch": 1.7160751565762005,
+      "grad_norm": 0.038240986750146325,
+      "learning_rate": 2.0781530306868246e-05,
+      "loss": 0.2393,
+      "step": 822
+    },
+    {
+      "epoch": 1.7181628392484343,
+      "grad_norm": 0.04106815130974131,
+      "learning_rate": 2.048235081487202e-05,
+      "loss": 0.2637,
+      "step": 823
+    },
+    {
+      "epoch": 1.720250521920668,
+      "grad_norm": 0.03673610429332063,
+      "learning_rate": 2.0185224257199908e-05,
+      "loss": 0.2207,
+      "step": 824
+    },
+    {
+      "epoch": 1.7223382045929019,
+      "grad_norm": 0.04334868227291063,
+      "learning_rate": 1.9890154031742325e-05,
+      "loss": 0.3223,
+      "step": 825
+    },
+    {
+      "epoch": 1.7244258872651357,
+      "grad_norm": 0.04017653068637208,
+      "learning_rate": 1.9597143512873872e-05,
+      "loss": 0.2656,
+      "step": 826
+    },
+    {
+      "epoch": 1.7265135699373695,
+      "grad_norm": 0.04143141277717955,
+      "learning_rate": 1.9306196051414728e-05,
+      "loss": 0.252,
+      "step": 827
+    },
+    {
+      "epoch": 1.7286012526096033,
+      "grad_norm": 0.030689616701832047,
+      "learning_rate": 1.9017314974592203e-05,
+      "loss": 0.1982,
+      "step": 828
+    },
+    {
+      "epoch": 1.730688935281837,
+      "grad_norm": 0.04149364456774439,
+      "learning_rate": 1.873050358600279e-05,
+      "loss": 0.2451,
+      "step": 829
+    },
+    {
+      "epoch": 1.732776617954071,
+      "grad_norm": 0.037822515436121175,
+      "learning_rate": 1.8445765165574324e-05,
+      "loss": 0.2246,
+      "step": 830
+    },
+    {
+      "epoch": 1.7348643006263047,
+      "grad_norm": 0.036300250288018185,
+      "learning_rate": 1.8163102969528702e-05,
+      "loss": 0.1709,
+      "step": 831
+    },
+    {
+      "epoch": 1.7369519832985385,
+      "grad_norm": 0.03813579853432033,
+      "learning_rate": 1.7882520230344225e-05,
+      "loss": 0.1982,
+      "step": 832
+    },
+    {
+      "epoch": 1.7390396659707723,
+      "grad_norm": 0.03490543301641405,
+      "learning_rate": 1.7604020156719026e-05,
+      "loss": 0.2256,
+      "step": 833
+    },
+    {
+      "epoch": 1.7411273486430061,
+      "grad_norm": 0.03767801651422133,
+      "learning_rate": 1.7327605933534087e-05,
+      "loss": 0.248,
+      "step": 834
+    },
+    {
+      "epoch": 1.7432150313152401,
+      "grad_norm": 0.03559965744869205,
+      "learning_rate": 1.705328072181711e-05,
+      "loss": 0.1963,
+      "step": 835
+    },
+    {
+      "epoch": 1.745302713987474,
+      "grad_norm": 0.03620103246900227,
+      "learning_rate": 1.6781047658706094e-05,
+      "loss": 0.2363,
+      "step": 836
+    },
+    {
+      "epoch": 1.7473903966597077,
+      "grad_norm": 0.03363208108698153,
+      "learning_rate": 1.6510909857413593e-05,
+      "loss": 0.208,
+      "step": 837
+    },
+    {
+      "epoch": 1.7494780793319415,
+      "grad_norm": 0.038864402335742426,
+      "learning_rate": 1.624287040719108e-05,
+      "loss": 0.2217,
+      "step": 838
+    },
+    {
+      "epoch": 1.7515657620041754,
+      "grad_norm": 0.03604996210348334,
+      "learning_rate": 1.597693237329363e-05,
+      "loss": 0.2363,
+      "step": 839
+    },
+    {
+      "epoch": 1.7536534446764092,
+      "grad_norm": 0.03870805026122142,
+      "learning_rate": 1.5713098796944913e-05,
+      "loss": 0.2363,
+      "step": 840
+    },
+    {
+      "epoch": 1.755741127348643,
+      "grad_norm": 0.037967280022889786,
+      "learning_rate": 1.545137269530228e-05,
+      "loss": 0.2031,
+      "step": 841
+    },
+    {
+      "epoch": 1.757828810020877,
+      "grad_norm": 0.032100601105533956,
+      "learning_rate": 1.5191757061422484e-05,
+      "loss": 0.1963,
+      "step": 842
+    },
+    {
+      "epoch": 1.7599164926931108,
+      "grad_norm": 0.03677180115256404,
+      "learning_rate": 1.493425486422717e-05,
+      "loss": 0.2324,
+      "step": 843
+    },
+    {
+      "epoch": 1.7620041753653446,
+      "grad_norm": 0.04179333438984321,
+      "learning_rate": 1.467886904846918e-05,
+      "loss": 0.2197,
+      "step": 844
+    },
+    {
+      "epoch": 1.7640918580375784,
+      "grad_norm": 0.03330069993588093,
+      "learning_rate": 1.4425602534698645e-05,
+      "loss": 0.1992,
+      "step": 845
+    },
+    {
+      "epoch": 1.7661795407098122,
+      "grad_norm": 0.03921593097418753,
+      "learning_rate": 1.4174458219229868e-05,
+      "loss": 0.2734,
+      "step": 846
+    },
+    {
+      "epoch": 1.768267223382046,
+      "grad_norm": 0.0384869660070795,
+      "learning_rate": 1.3925438974107918e-05,
+      "loss": 0.2285,
+      "step": 847
+    },
+    {
+      "epoch": 1.7703549060542798,
+      "grad_norm": 0.03235612278723638,
+      "learning_rate": 1.3678547647075946e-05,
+      "loss": 0.2021,
+      "step": 848
+    },
+    {
+      "epoch": 1.7724425887265136,
+      "grad_norm": 0.034557637083813396,
+      "learning_rate": 1.3433787061542525e-05,
+      "loss": 0.1982,
+      "step": 849
+    },
+    {
+      "epoch": 1.7745302713987474,
+      "grad_norm": 0.036691512115311166,
+      "learning_rate": 1.3191160016549564e-05,
+      "loss": 0.2695,
+      "step": 850
+    },
+    {
+      "epoch": 1.7766179540709812,
+      "grad_norm": 0.036890816812044876,
+      "learning_rate": 1.2950669286740024e-05,
+      "loss": 0.2119,
+      "step": 851
+    },
+    {
+      "epoch": 1.778705636743215,
+      "grad_norm": 0.03801656481599902,
+      "learning_rate": 1.271231762232632e-05,
+      "loss": 0.2354,
+      "step": 852
+    },
+    {
+      "epoch": 1.7807933194154488,
+      "grad_norm": 0.0431132757880389,
+      "learning_rate": 1.2476107749058986e-05,
+      "loss": 0.2305,
+      "step": 853
+    },
+    {
+      "epoch": 1.7828810020876826,
+      "grad_norm": 0.034619924337342727,
+      "learning_rate": 1.2242042368195218e-05,
+      "loss": 0.1914,
+      "step": 854
+    },
+    {
+      "epoch": 1.7849686847599164,
+      "grad_norm": 0.034941322602160714,
+      "learning_rate": 1.2010124156468294e-05,
+      "loss": 0.2051,
+      "step": 855
+    },
+    {
+      "epoch": 1.7870563674321502,
+      "grad_norm": 0.03577282594283845,
+      "learning_rate": 1.1780355766056694e-05,
+      "loss": 0.2334,
+      "step": 856
+    },
+    {
+      "epoch": 1.789144050104384,
+      "grad_norm": 0.03923352027261708,
+      "learning_rate": 1.1552739824554026e-05,
+      "loss": 0.2217,
+      "step": 857
+    },
+    {
+      "epoch": 1.7912317327766178,
+      "grad_norm": 0.03464671275803716,
+      "learning_rate": 1.1327278934938723e-05,
+      "loss": 0.1992,
+      "step": 858
+    },
+    {
+      "epoch": 1.7933194154488517,
+      "grad_norm": 0.04157357259303964,
+      "learning_rate": 1.1103975675544443e-05,
+      "loss": 0.2461,
+      "step": 859
+    },
+    {
+      "epoch": 1.7954070981210855,
+      "grad_norm": 0.040227594429237123,
+      "learning_rate": 1.0882832600030624e-05,
+      "loss": 0.2441,
+      "step": 860
+    },
+    {
+      "epoch": 1.7974947807933193,
+      "grad_norm": 0.03488250123821703,
+      "learning_rate": 1.0663852237353067e-05,
+      "loss": 0.1963,
+      "step": 861
+    },
+    {
+      "epoch": 1.7995824634655533,
+      "grad_norm": 0.03979475512791435,
+      "learning_rate": 1.0447037091735223e-05,
+      "loss": 0.2471,
+      "step": 862
+    },
+    {
+      "epoch": 1.801670146137787,
+      "grad_norm": 0.03997223499064563,
+      "learning_rate": 1.0232389642639395e-05,
+      "loss": 0.2471,
+      "step": 863
+    },
+    {
+      "epoch": 1.803757828810021,
+      "grad_norm": 0.04038899286733726,
+      "learning_rate": 1.0019912344738625e-05,
+      "loss": 0.2598,
+      "step": 864
+    },
+    {
+      "epoch": 1.8058455114822547,
+      "grad_norm": 0.03748451896837008,
+      "learning_rate": 9.809607627888296e-06,
+      "loss": 0.2002,
+      "step": 865
+    },
+    {
+      "epoch": 1.8079331941544885,
+      "grad_norm": 0.03427034514456546,
+      "learning_rate": 9.601477897098576e-06,
+      "loss": 0.1729,
+      "step": 866
+    },
+    {
+      "epoch": 1.8100208768267223,
+      "grad_norm": 0.04471055633543089,
+      "learning_rate": 9.395525532506866e-06,
+      "loss": 0.2256,
+      "step": 867
+    },
+    {
+      "epoch": 1.812108559498956,
+      "grad_norm": 0.03475782341698621,
+      "learning_rate": 9.191752889350547e-06,
+      "loss": 0.2207,
+      "step": 868
+    },
+    {
+      "epoch": 1.8141962421711901,
+      "grad_norm": 0.03828798248246293,
+      "learning_rate": 8.990162297940097e-06,
+      "loss": 0.2109,
+      "step": 869
+    },
+    {
+      "epoch": 1.816283924843424,
+      "grad_norm": 0.04180601763110725,
+      "learning_rate": 8.79075606363231e-06,
+      "loss": 0.2676,
+      "step": 870
+    },
+    {
+      "epoch": 1.8183716075156577,
+      "grad_norm": 0.03646735664376503,
+      "learning_rate": 8.593536466804142e-06,
+      "loss": 0.207,
+      "step": 871
+    },
+    {
+      "epoch": 1.8204592901878915,
+      "grad_norm": 0.03218572550634406,
+      "learning_rate": 8.398505762826503e-06,
+      "loss": 0.1797,
+      "step": 872
+    },
+    {
+      "epoch": 1.8225469728601253,
+      "grad_norm": 0.03982709976184709,
+      "learning_rate": 8.205666182038418e-06,
+      "loss": 0.2432,
+      "step": 873
+    },
+    {
+      "epoch": 1.8246346555323592,
+      "grad_norm": 0.037982327489911684,
+      "learning_rate": 8.015019929721668e-06,
+      "loss": 0.2344,
+      "step": 874
+    },
+    {
+      "epoch": 1.826722338204593,
+      "grad_norm": 0.03580893689768604,
+      "learning_rate": 7.826569186075428e-06,
+      "loss": 0.252,
+      "step": 875
+    },
+    {
+      "epoch": 1.8288100208768268,
+      "grad_norm": 0.035854237119299116,
+      "learning_rate": 7.640316106191403e-06,
+      "loss": 0.1963,
+      "step": 876
+    },
+    {
+      "epoch": 1.8308977035490606,
+      "grad_norm": 0.03548915356785303,
+      "learning_rate": 7.4562628200292475e-06,
+      "loss": 0.2148,
+      "step": 877
+    },
+    {
+      "epoch": 1.8329853862212944,
+      "grad_norm": 0.03888900644009877,
+      "learning_rate": 7.274411432392026e-06,
+      "loss": 0.2812,
+      "step": 878
+    },
+    {
+      "epoch": 1.8350730688935282,
+      "grad_norm": 0.036848786139394796,
+      "learning_rate": 7.0947640229023675e-06,
+      "loss": 0.1904,
+      "step": 879
+    },
+    {
+      "epoch": 1.837160751565762,
+      "grad_norm": 0.032594039620939604,
+      "learning_rate": 6.917322645978463e-06,
+      "loss": 0.1904,
+      "step": 880
+    },
+    {
+      "epoch": 1.8392484342379958,
+      "grad_norm": 0.038807798544498055,
+      "learning_rate": 6.742089330810774e-06,
+      "loss": 0.2383,
+      "step": 881
+    },
+    {
+      "epoch": 1.8413361169102296,
+      "grad_norm": 0.042279119809553053,
+      "learning_rate": 6.5690660813386484e-06,
+      "loss": 0.3184,
+      "step": 882
+    },
+    {
+      "epoch": 1.8434237995824634,
+      "grad_norm": 0.03575523378084652,
+      "learning_rate": 6.398254876227561e-06,
+      "loss": 0.1875,
+      "step": 883
+    },
+    {
+      "epoch": 1.8455114822546972,
+      "grad_norm": 0.03438492023581685,
+      "learning_rate": 6.229657668846333e-06,
+      "loss": 0.21,
+      "step": 884
+    },
+    {
+      "epoch": 1.847599164926931,
+      "grad_norm": 0.04355612584591555,
+      "learning_rate": 6.0632763872449535e-06,
+      "loss": 0.2754,
+      "step": 885
+    },
+    {
+      "epoch": 1.8496868475991648,
+      "grad_norm": 0.03533181091838774,
+      "learning_rate": 5.899112934132345e-06,
+      "loss": 0.1689,
+      "step": 886
+    },
+    {
+      "epoch": 1.8517745302713986,
+      "grad_norm": 0.040660080641486476,
+      "learning_rate": 5.73716918685483e-06,
+      "loss": 0.2471,
+      "step": 887
+    },
+    {
+      "epoch": 1.8538622129436324,
+      "grad_norm": 0.03235795768355878,
+      "learning_rate": 5.577446997374458e-06,
+      "loss": 0.1826,
+      "step": 888
+    },
+    {
+      "epoch": 1.8559498956158664,
+      "grad_norm": 0.03609407610302125,
+      "learning_rate": 5.4199481922479146e-06,
+      "loss": 0.1816,
+      "step": 889
+    },
+    {
+      "epoch": 1.8580375782881002,
+      "grad_norm": 0.03647085365814291,
+      "learning_rate": 5.264674572605665e-06,
+      "loss": 0.2393,
+      "step": 890
+    },
+    {
+      "epoch": 1.860125260960334,
+      "grad_norm": 0.03892481200719314,
+      "learning_rate": 5.11162791413129e-06,
+      "loss": 0.2021,
+      "step": 891
+    },
+    {
+      "epoch": 1.8622129436325678,
+      "grad_norm": 0.03859819801659933,
+      "learning_rate": 4.9608099670412025e-06,
+      "loss": 0.2402,
+      "step": 892
+    },
+    {
+      "epoch": 1.8643006263048016,
+      "grad_norm": 0.03638768037937149,
+      "learning_rate": 4.812222456064697e-06,
+      "loss": 0.1729,
+      "step": 893
+    },
+    {
+      "epoch": 1.8663883089770354,
+      "grad_norm": 0.03578385304837865,
+      "learning_rate": 4.665867080424046e-06,
+      "loss": 0.2217,
+      "step": 894
+    },
+    {
+      "epoch": 1.8684759916492695,
+      "grad_norm": 0.03899470496061886,
+      "learning_rate": 4.521745513815345e-06,
+      "loss": 0.2236,
+      "step": 895
+    },
+    {
+      "epoch": 1.8705636743215033,
+      "grad_norm": 0.03565199656237529,
+      "learning_rate": 4.379859404389075e-06,
+      "loss": 0.2158,
+      "step": 896
+    },
+    {
+      "epoch": 1.872651356993737,
+      "grad_norm": 0.03700782502510176,
+      "learning_rate": 4.240210374731479e-06,
+      "loss": 0.2441,
+      "step": 897
+    },
+    {
+      "epoch": 1.8747390396659709,
+      "grad_norm": 0.034012775228610695,
+      "learning_rate": 4.10280002184591e-06,
+      "loss": 0.1934,
+      "step": 898
+    },
+    {
+      "epoch": 1.8768267223382047,
+      "grad_norm": 0.032231872294556166,
+      "learning_rate": 3.967629917134574e-06,
+      "loss": 0.1689,
+      "step": 899
+    },
+    {
+      "epoch": 1.8789144050104385,
+      "grad_norm": 0.039125390388194826,
+      "learning_rate": 3.834701606380575e-06,
+      "loss": 0.2285,
+      "step": 900
+    },
+    {
+      "epoch": 1.8810020876826723,
+      "grad_norm": 0.04416962700457997,
+      "learning_rate": 3.7040166097301877e-06,
+      "loss": 0.2656,
+      "step": 901
+    },
+    {
+      "epoch": 1.883089770354906,
+      "grad_norm": 0.04111412261545973,
+      "learning_rate": 3.575576421675586e-06,
+      "loss": 0.248,
+      "step": 902
+    },
+    {
+      "epoch": 1.88517745302714,
+      "grad_norm": 0.03397805823890809,
+      "learning_rate": 3.449382511037613e-06,
+      "loss": 0.1904,
+      "step": 903
+    },
+    {
+      "epoch": 1.8872651356993737,
+      "grad_norm": 0.041722415283545866,
+      "learning_rate": 3.3254363209491046e-06,
+      "loss": 0.3301,
+      "step": 904
+    },
+    {
+      "epoch": 1.8893528183716075,
+      "grad_norm": 0.047301660349510964,
+      "learning_rate": 3.203739268838324e-06,
+      "loss": 0.2988,
+      "step": 905
+    },
+    {
+      "epoch": 1.8914405010438413,
+      "grad_norm": 0.03091322302919706,
+      "learning_rate": 3.0842927464127537e-06,
+      "loss": 0.1592,
+      "step": 906
+    },
+    {
+      "epoch": 1.8935281837160751,
+      "grad_norm": 0.038250787995771175,
+      "learning_rate": 2.9670981196431525e-06,
+      "loss": 0.248,
+      "step": 907
+    },
+    {
+      "epoch": 1.895615866388309,
+      "grad_norm": 0.04184624767282992,
+      "learning_rate": 2.8521567287480344e-06,
+      "loss": 0.2461,
+      "step": 908
+    },
+    {
+      "epoch": 1.8977035490605427,
+      "grad_norm": 0.03669678971086905,
+      "learning_rate": 2.7394698881782367e-06,
+      "loss": 0.1895,
+      "step": 909
+    },
+    {
+      "epoch": 1.8997912317327765,
+      "grad_norm": 0.03332356389562326,
+      "learning_rate": 2.6290388866019533e-06,
+      "loss": 0.1934,
+      "step": 910
+    },
+    {
+      "epoch": 1.9018789144050103,
+      "grad_norm": 0.03567454628961636,
+      "learning_rate": 2.5208649868899704e-06,
+      "loss": 0.2207,
+      "step": 911
+    },
+    {
+      "epoch": 1.9039665970772441,
+      "grad_norm": 0.03723478614951914,
+      "learning_rate": 2.4149494261012097e-06,
+      "loss": 0.2002,
+      "step": 912
+    },
+    {
+      "epoch": 1.906054279749478,
+      "grad_norm": 0.03840886510699952,
+      "learning_rate": 2.3112934154686296e-06,
+      "loss": 0.1982,
+      "step": 913
+    },
+    {
+      "epoch": 1.9081419624217117,
+      "grad_norm": 0.03658052263392586,
+      "learning_rate": 2.2098981403852804e-06,
+      "loss": 0.2295,
+      "step": 914
+    },
+    {
+      "epoch": 1.9102296450939458,
+      "grad_norm": 0.042943855139820596,
+      "learning_rate": 2.1107647603908933e-06,
+      "loss": 0.2422,
+      "step": 915
+    },
+    {
+      "epoch": 1.9123173277661796,
+      "grad_norm": 0.038181568479852754,
+      "learning_rate": 2.0138944091585123e-06,
+      "loss": 0.1768,
+      "step": 916
+    },
+    {
+      "epoch": 1.9144050104384134,
+      "grad_norm": 0.04044402128707661,
+      "learning_rate": 1.91928819448155e-06,
+      "loss": 0.249,
+      "step": 917
+    },
+    {
+      "epoch": 1.9164926931106472,
+      "grad_norm": 0.04021345081714543,
+      "learning_rate": 1.8269471982611086e-06,
+      "loss": 0.2578,
+      "step": 918
+    },
+    {
+      "epoch": 1.918580375782881,
+      "grad_norm": 0.03765027742041835,
+      "learning_rate": 1.7368724764936118e-06,
+      "loss": 0.2275,
+      "step": 919
+    },
+    {
+      "epoch": 1.9206680584551148,
+      "grad_norm": 0.03276750255084186,
+      "learning_rate": 1.6490650592588363e-06,
+      "loss": 0.1797,
+      "step": 920
+    },
+    {
+      "epoch": 1.9227557411273486,
+      "grad_norm": 0.03836166999423338,
+      "learning_rate": 1.5635259507079224e-06,
+      "loss": 0.2656,
+      "step": 921
+    },
+    {
+      "epoch": 1.9248434237995826,
+      "grad_norm": 0.040716455351115934,
+      "learning_rate": 1.4802561290520934e-06,
+      "loss": 0.2373,
+      "step": 922
+    },
+    {
+      "epoch": 1.9269311064718164,
+      "grad_norm": 0.044722992742083724,
+      "learning_rate": 1.3992565465512863e-06,
+      "loss": 0.2832,
+      "step": 923
+    },
+    {
+      "epoch": 1.9290187891440502,
+      "grad_norm": 0.035718501429723326,
+      "learning_rate": 1.3205281295034512e-06,
+      "loss": 0.2109,
+      "step": 924
+    },
+    {
+      "epoch": 1.931106471816284,
+      "grad_norm": 0.040897036162254374,
+      "learning_rate": 1.2440717782338018e-06,
+      "loss": 0.248,
+      "step": 925
+    },
+    {
+      "epoch": 1.9331941544885178,
+      "grad_norm": 0.037701179364082175,
+      "learning_rate": 1.169888367084626e-06,
+      "loss": 0.2021,
+      "step": 926
+    },
+    {
+      "epoch": 1.9352818371607516,
+      "grad_norm": 0.03800126744938023,
+      "learning_rate": 1.0979787444051815e-06,
+      "loss": 0.2051,
+      "step": 927
+    },
+    {
+      "epoch": 1.9373695198329854,
+      "grad_norm": 0.0376053096363303,
+      "learning_rate": 1.0283437325421252e-06,
+      "loss": 0.25,
+      "step": 928
+    },
+    {
+      "epoch": 1.9394572025052192,
+      "grad_norm": 0.03852373604309701,
+      "learning_rate": 9.60984127829989e-07,
+      "loss": 0.1895,
+      "step": 929
+    },
+    {
+      "epoch": 1.941544885177453,
+      "grad_norm": 0.05141719488909021,
+      "learning_rate": 8.95900700582164e-07,
+      "loss": 0.2969,
+      "step": 930
+    },
+    {
+      "epoch": 1.9436325678496869,
+      "grad_norm": 0.0422557444402279,
+      "learning_rate": 8.330941950819738e-07,
+      "loss": 0.2793,
+      "step": 931
+    },
+    {
+      "epoch": 1.9457202505219207,
+      "grad_norm": 0.03878673278759856,
+      "learning_rate": 7.725653295743484e-07,
+      "loss": 0.2412,
+      "step": 932
+    },
+    {
+      "epoch": 1.9478079331941545,
+      "grad_norm": 0.04515324231545769,
+      "learning_rate": 7.14314796257387e-07,
+      "loss": 0.2715,
+      "step": 933
+    },
+    {
+      "epoch": 1.9498956158663883,
+      "grad_norm": 0.04060496658163304,
+      "learning_rate": 6.58343261274652e-07,
+      "loss": 0.2559,
+      "step": 934
+    },
+    {
+      "epoch": 1.951983298538622,
+      "grad_norm": 0.0410883973174675,
+      "learning_rate": 6.046513647074203e-07,
+      "loss": 0.2676,
+      "step": 935
+    },
+    {
+      "epoch": 1.9540709812108559,
+      "grad_norm": 0.037425826708713576,
+      "learning_rate": 5.532397205673556e-07,
+      "loss": 0.208,
+      "step": 936
+    },
+    {
+      "epoch": 1.9561586638830897,
+      "grad_norm": 0.03711360512013879,
+      "learning_rate": 5.04108916789603e-07,
+      "loss": 0.2246,
+      "step": 937
+    },
+    {
+      "epoch": 1.9582463465553235,
+      "grad_norm": 0.04129502522262085,
+      "learning_rate": 4.5725951522592734e-07,
+      "loss": 0.2812,
+      "step": 938
+    },
+    {
+      "epoch": 1.9603340292275573,
+      "grad_norm": 0.042768062565203456,
+      "learning_rate": 4.126920516384303e-07,
+      "loss": 0.2734,
+      "step": 939
+    },
+    {
+      "epoch": 1.962421711899791,
+      "grad_norm": 0.03636718305173902,
+      "learning_rate": 3.704070356932432e-07,
+      "loss": 0.2275,
+      "step": 940
+    },
+    {
+      "epoch": 1.964509394572025,
+      "grad_norm": 0.03914519289291409,
+      "learning_rate": 3.3040495095491006e-07,
+      "loss": 0.2051,
+      "step": 941
+    },
+    {
+      "epoch": 1.966597077244259,
+      "grad_norm": 0.036435492987632524,
+      "learning_rate": 2.926862548806364e-07,
+      "loss": 0.21,
+      "step": 942
+    },
+    {
+      "epoch": 1.9686847599164927,
+      "grad_norm": 0.04520438858168889,
+      "learning_rate": 2.5725137881529306e-07,
+      "loss": 0.2852,
+      "step": 943
+    },
+    {
+      "epoch": 1.9707724425887265,
+      "grad_norm": 0.036719387383711895,
+      "learning_rate": 2.2410072798624283e-07,
+      "loss": 0.2422,
+      "step": 944
+    },
+    {
+      "epoch": 1.9728601252609603,
+      "grad_norm": 0.03890839218462144,
+      "learning_rate": 1.9323468149892165e-07,
+      "loss": 0.2617,
+      "step": 945
+    },
+    {
+      "epoch": 1.9749478079331941,
+      "grad_norm": 0.04401800316091134,
+      "learning_rate": 1.6465359233233114e-07,
+      "loss": 0.2314,
+      "step": 946
+    },
+    {
+      "epoch": 1.977035490605428,
+      "grad_norm": 0.042526472328545395,
+      "learning_rate": 1.383577873351305e-07,
+      "loss": 0.2734,
+      "step": 947
+    },
+    {
+      "epoch": 1.9791231732776617,
+      "grad_norm": 0.04123640104260365,
+      "learning_rate": 1.143475672218175e-07,
+      "loss": 0.2715,
+      "step": 948
+    },
+    {
+      "epoch": 1.9812108559498958,
+      "grad_norm": 0.041143122691388076,
+      "learning_rate": 9.26232065693089e-08,
+      "loss": 0.2578,
+      "step": 949
+    },
+    {
+      "epoch": 1.9832985386221296,
+      "grad_norm": 0.03646932624947557,
+      "learning_rate": 7.318495381383184e-08,
+      "loss": 0.1953,
+      "step": 950
+    },
+    {
+      "epoch": 1.9853862212943634,
+      "grad_norm": 0.042625916797236235,
+      "learning_rate": 5.603303124803727e-08,
+      "loss": 0.2344,
+      "step": 951
+    },
+    {
+      "epoch": 1.9874739039665972,
+      "grad_norm": 0.039620902455050594,
+      "learning_rate": 4.1167635018513064e-08,
+      "loss": 0.2598,
+      "step": 952
+    },
+    {
+      "epoch": 1.989561586638831,
+      "grad_norm": 0.0329100227494744,
+      "learning_rate": 2.8588935123474714e-08,
+      "loss": 0.1758,
+      "step": 953
+    },
+    {
+      "epoch": 1.9916492693110648,
+      "grad_norm": 0.03829557846462901,
+      "learning_rate": 1.8297075410877997e-08,
+      "loss": 0.2676,
+      "step": 954
+    },
+    {
+      "epoch": 1.9937369519832986,
+      "grad_norm": 0.037593692719248624,
+      "learning_rate": 1.02921735767314e-08,
+      "loss": 0.1943,
+      "step": 955
+    },
+    {
+      "epoch": 1.9958246346555324,
+      "grad_norm": 0.03806855037144764,
+      "learning_rate": 4.574321163763884e-09,
+      "loss": 0.2383,
+      "step": 956
+    },
+    {
+      "epoch": 1.9979123173277662,
+      "grad_norm": 0.03559582629136701,
+      "learning_rate": 1.1435835604034495e-09,
+      "loss": 0.1787,
+      "step": 957
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.03239360059033573,
+      "learning_rate": 0.0,
+      "loss": 0.167,
+      "step": 958
+    },
+    {
+      "epoch": 2.0,
+      "step": 958,
+      "total_flos": 5.650597015940235e+18,
+      "train_loss": 0.3088689816022965,
+      "train_runtime": 6060.4121,
+      "train_samples_per_second": 90.951,
+      "train_steps_per_second": 0.158
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 958,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.650597015940235e+18,
+  "train_batch_size": 72,
+  "trial_name": null,
+  "trial_params": null
+}