diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9719 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9996375498368975,
+  "eval_steps": 345,
+  "global_step": 1379,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0007249003262051468,
+      "grad_norm": 44.75750397895495,
+      "learning_rate": 4.7619047619047623e-07,
+      "loss": 2.3443,
+      "step": 1
+    },
+    {
+      "epoch": 0.0014498006524102935,
+      "grad_norm": 45.205818482220735,
+      "learning_rate": 9.523809523809525e-07,
+      "loss": 2.1445,
+      "step": 2
+    },
+    {
+      "epoch": 0.0021747009786154403,
+      "grad_norm": 56.724033240854,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 2.5876,
+      "step": 3
+    },
+    {
+      "epoch": 0.002899601304820587,
+      "grad_norm": 53.17611331537967,
+      "learning_rate": 1.904761904761905e-06,
+      "loss": 2.3983,
+      "step": 4
+    },
+    {
+      "epoch": 0.003624501631025734,
+      "grad_norm": 57.8664359648513,
+      "learning_rate": 2.380952380952381e-06,
+      "loss": 2.1637,
+      "step": 5
+    },
+    {
+      "epoch": 0.004349401957230881,
+      "grad_norm": 14.246129965796554,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 1.9145,
+      "step": 6
+    },
+    {
+      "epoch": 0.005074302283436027,
+      "grad_norm": 11.159860869255928,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.7268,
+      "step": 7
+    },
+    {
+      "epoch": 0.005799202609641174,
+      "grad_norm": 8.887448463820203,
+      "learning_rate": 3.80952380952381e-06,
+      "loss": 1.5522,
+      "step": 8
+    },
+    {
+      "epoch": 0.006524102935846321,
+      "grad_norm": 6.753347075425067,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 1.5248,
+      "step": 9
+    },
+    {
+      "epoch": 0.007249003262051468,
+      "grad_norm": 11.059402659834483,
+      "learning_rate": 4.761904761904762e-06,
+      "loss": 1.5304,
+      "step": 10
+    },
+    {
+      "epoch": 0.007973903588256614,
+      "grad_norm": 9.28275165512847,
+      "learning_rate": 5.2380952380952384e-06,
+      "loss": 1.4316,
+      "step": 11
+    },
+    {
+      "epoch": 0.008698803914461761,
+      "grad_norm": 4.714004999293759,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 1.331,
+      "step": 12
+    },
+    {
+      "epoch": 0.009423704240666908,
+      "grad_norm": 4.708567041689236,
+      "learning_rate": 6.1904761904761914e-06,
+      "loss": 1.4035,
+      "step": 13
+    },
+    {
+      "epoch": 0.010148604566872055,
+      "grad_norm": 7.8340994412560665,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.4499,
+      "step": 14
+    },
+    {
+      "epoch": 0.010873504893077202,
+      "grad_norm": 5.7926127365771105,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 1.2915,
+      "step": 15
+    },
+    {
+      "epoch": 0.011598405219282348,
+      "grad_norm": 4.338940169766625,
+      "learning_rate": 7.61904761904762e-06,
+      "loss": 1.2546,
+      "step": 16
+    },
+    {
+      "epoch": 0.012323305545487495,
+      "grad_norm": 5.3158492494198875,
+      "learning_rate": 8.095238095238097e-06,
+      "loss": 1.3437,
+      "step": 17
+    },
+    {
+      "epoch": 0.013048205871692642,
+      "grad_norm": 4.267050487387928,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 1.2076,
+      "step": 18
+    },
+    {
+      "epoch": 0.013773106197897789,
+      "grad_norm": 4.289501496589146,
+      "learning_rate": 9.047619047619049e-06,
+      "loss": 1.2095,
+      "step": 19
+    },
+    {
+      "epoch": 0.014498006524102935,
+      "grad_norm": 4.475865685170405,
+      "learning_rate": 9.523809523809525e-06,
+      "loss": 1.2499,
+      "step": 20
+    },
+    {
+      "epoch": 0.015222906850308082,
+      "grad_norm": 4.341512982701414,
+      "learning_rate": 1e-05,
+      "loss": 1.3725,
+      "step": 21
+    },
+    {
+      "epoch": 0.01594780717651323,
+      "grad_norm": 3.9777632208757443,
+      "learning_rate": 1.0476190476190477e-05,
+      "loss": 1.2679,
+      "step": 22
+    },
+    {
+      "epoch": 0.016672707502718376,
+      "grad_norm": 3.971780990603446,
+      "learning_rate": 1.0952380952380955e-05,
+      "loss": 1.3049,
+      "step": 23
+    },
+    {
+      "epoch": 0.017397607828923523,
+      "grad_norm": 4.355148301567439,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 1.1696,
+      "step": 24
+    },
+    {
+      "epoch": 0.01812250815512867,
+      "grad_norm": 5.700836460033174,
+      "learning_rate": 1.1904761904761905e-05,
+      "loss": 1.3116,
+      "step": 25
+    },
+    {
+      "epoch": 0.018847408481333816,
+      "grad_norm": 4.266676186480929,
+      "learning_rate": 1.2380952380952383e-05,
+      "loss": 1.2132,
+      "step": 26
+    },
+    {
+      "epoch": 0.019572308807538963,
+      "grad_norm": 4.685688044042141,
+      "learning_rate": 1.2857142857142859e-05,
+      "loss": 1.2743,
+      "step": 27
+    },
+    {
+      "epoch": 0.02029720913374411,
+      "grad_norm": 3.8278571816897977,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 1.3062,
+      "step": 28
+    },
+    {
+      "epoch": 0.021022109459949256,
+      "grad_norm": 4.207466080048964,
+      "learning_rate": 1.3809523809523811e-05,
+      "loss": 1.2238,
+      "step": 29
+    },
+    {
+      "epoch": 0.021747009786154403,
+      "grad_norm": 3.772755917601343,
+      "learning_rate": 1.4285714285714287e-05,
+      "loss": 1.2333,
+      "step": 30
+    },
+    {
+      "epoch": 0.02247191011235955,
+      "grad_norm": 3.2903338340660095,
+      "learning_rate": 1.4761904761904763e-05,
+      "loss": 1.1223,
+      "step": 31
+    },
+    {
+      "epoch": 0.023196810438564697,
+      "grad_norm": 3.2688442989444746,
+      "learning_rate": 1.523809523809524e-05,
+      "loss": 1.1476,
+      "step": 32
+    },
+    {
+      "epoch": 0.023921710764769843,
+      "grad_norm": 3.6401775542496284,
+      "learning_rate": 1.5714285714285715e-05,
+      "loss": 1.1805,
+      "step": 33
+    },
+    {
+      "epoch": 0.02464661109097499,
+      "grad_norm": 3.4329875974688004,
+      "learning_rate": 1.6190476190476193e-05,
+      "loss": 1.1406,
+      "step": 34
+    },
+    {
+      "epoch": 0.025371511417180137,
+      "grad_norm": 3.457570825026417,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 1.1447,
+      "step": 35
+    },
+    {
+      "epoch": 0.026096411743385284,
+      "grad_norm": 3.503397413974539,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 1.13,
+      "step": 36
+    },
+    {
+      "epoch": 0.02682131206959043,
+      "grad_norm": 2.880316116737165,
+      "learning_rate": 1.761904761904762e-05,
+      "loss": 1.2275,
+      "step": 37
+    },
+    {
+      "epoch": 0.027546212395795577,
+      "grad_norm": 3.4491978928897686,
+      "learning_rate": 1.8095238095238097e-05,
+      "loss": 1.2675,
+      "step": 38
+    },
+    {
+      "epoch": 0.028271112722000724,
+      "grad_norm": 3.4801420565131695,
+      "learning_rate": 1.8571428571428575e-05,
+      "loss": 1.1656,
+      "step": 39
+    },
+    {
+      "epoch": 0.02899601304820587,
+      "grad_norm": 3.154782654672038,
+      "learning_rate": 1.904761904761905e-05,
+      "loss": 1.157,
+      "step": 40
+    },
+    {
+      "epoch": 0.029720913374411018,
+      "grad_norm": 3.0915926527583633,
+      "learning_rate": 1.9523809523809524e-05,
+      "loss": 1.1432,
+      "step": 41
+    },
+    {
+      "epoch": 0.030445813700616164,
+      "grad_norm": 2.772154104430272,
+      "learning_rate": 2e-05,
+      "loss": 1.0994,
+      "step": 42
+    },
+    {
+      "epoch": 0.03117071402682131,
+      "grad_norm": 3.0517198878965304,
+      "learning_rate": 1.9999972393793303e-05,
+      "loss": 1.2549,
+      "step": 43
+    },
+    {
+      "epoch": 0.03189561435302646,
+      "grad_norm": 2.6750901532162987,
+      "learning_rate": 1.9999889575325633e-05,
+      "loss": 1.143,
+      "step": 44
+    },
+    {
+      "epoch": 0.032620514679231605,
+      "grad_norm": 3.6101390652119516,
+      "learning_rate": 1.9999751545054247e-05,
+      "loss": 1.2469,
+      "step": 45
+    },
+    {
+      "epoch": 0.03334541500543675,
+      "grad_norm": 2.843300653538732,
+      "learning_rate": 1.9999558303741244e-05,
+      "loss": 1.061,
+      "step": 46
+    },
+    {
+      "epoch": 0.0340703153316419,
+      "grad_norm": 3.9392334069856907,
+      "learning_rate": 1.9999309852453556e-05,
+      "loss": 1.2809,
+      "step": 47
+    },
+    {
+      "epoch": 0.034795215657847045,
+      "grad_norm": 3.773339147888779,
+      "learning_rate": 1.9999006192562948e-05,
+      "loss": 1.3008,
+      "step": 48
+    },
+    {
+      "epoch": 0.03552011598405219,
+      "grad_norm": 2.7832080749846453,
+      "learning_rate": 1.9998647325745995e-05,
+      "loss": 1.1342,
+      "step": 49
+    },
+    {
+      "epoch": 0.03624501631025734,
+      "grad_norm": 3.7382629441207667,
+      "learning_rate": 1.9998233253984088e-05,
+      "loss": 1.2641,
+      "step": 50
+    },
+    {
+      "epoch": 0.036969916636462485,
+      "grad_norm": 2.799440280567736,
+      "learning_rate": 1.9997763979563418e-05,
+      "loss": 1.2242,
+      "step": 51
+    },
+    {
+      "epoch": 0.03769481696266763,
+      "grad_norm": 3.3316333443131696,
+      "learning_rate": 1.9997239505074954e-05,
+      "loss": 1.2312,
+      "step": 52
+    },
+    {
+      "epoch": 0.03841971728887278,
+      "grad_norm": 5.725082920380033,
+      "learning_rate": 1.999665983341446e-05,
+      "loss": 1.4632,
+      "step": 53
+    },
+    {
+      "epoch": 0.039144617615077926,
+      "grad_norm": 2.553768110728692,
+      "learning_rate": 1.9996024967782436e-05,
+      "loss": 1.2164,
+      "step": 54
+    },
+    {
+      "epoch": 0.03986951794128307,
+      "grad_norm": 3.041402503331704,
+      "learning_rate": 1.9995334911684127e-05,
+      "loss": 1.2156,
+      "step": 55
+    },
+    {
+      "epoch": 0.04059441826748822,
+      "grad_norm": 2.595475331293293,
+      "learning_rate": 1.99945896689295e-05,
+      "loss": 1.1247,
+      "step": 56
+    },
+    {
+      "epoch": 0.041319318593693366,
+      "grad_norm": 2.939590634487055,
+      "learning_rate": 1.9993789243633227e-05,
+      "loss": 1.2171,
+      "step": 57
+    },
+    {
+      "epoch": 0.04204421891989851,
+      "grad_norm": 2.6474772321680606,
+      "learning_rate": 1.999293364021464e-05,
+      "loss": 1.187,
+      "step": 58
+    },
+    {
+      "epoch": 0.04276911924610366,
+      "grad_norm": 2.901881436981748,
+      "learning_rate": 1.9992022863397737e-05,
+      "loss": 1.1224,
+      "step": 59
+    },
+    {
+      "epoch": 0.043494019572308806,
+      "grad_norm": 2.9320116081953835,
+      "learning_rate": 1.999105691821113e-05,
+      "loss": 1.1692,
+      "step": 60
+    },
+    {
+      "epoch": 0.04421891989851395,
+      "grad_norm": 3.5247973111889004,
+      "learning_rate": 1.9990035809988045e-05,
+      "loss": 1.1575,
+      "step": 61
+    },
+    {
+      "epoch": 0.0449438202247191,
+      "grad_norm": 2.669047070211342,
+      "learning_rate": 1.9988959544366265e-05,
+      "loss": 1.1433,
+      "step": 62
+    },
+    {
+      "epoch": 0.04566872055092425,
+      "grad_norm": 3.213177659539329,
+      "learning_rate": 1.9987828127288105e-05,
+      "loss": 1.2635,
+      "step": 63
+    },
+    {
+      "epoch": 0.04639362087712939,
+      "grad_norm": 3.3559565945133554,
+      "learning_rate": 1.99866415650004e-05,
+      "loss": 1.3022,
+      "step": 64
+    },
+    {
+      "epoch": 0.04711852120333454,
+      "grad_norm": 2.3670279547756894,
+      "learning_rate": 1.998539986405444e-05,
+      "loss": 1.0142,
+      "step": 65
+    },
+    {
+      "epoch": 0.04784342152953969,
+      "grad_norm": 3.037710920162538,
+      "learning_rate": 1.998410303130596e-05,
+      "loss": 1.2591,
+      "step": 66
+    },
+    {
+      "epoch": 0.048568321855744834,
+      "grad_norm": 2.865611039939188,
+      "learning_rate": 1.9982751073915087e-05,
+      "loss": 1.1993,
+      "step": 67
+    },
+    {
+      "epoch": 0.04929322218194998,
+      "grad_norm": 3.48123837358242,
+      "learning_rate": 1.9981343999346302e-05,
+      "loss": 1.2121,
+      "step": 68
+    },
+    {
+      "epoch": 0.05001812250815513,
+      "grad_norm": 3.2299518093080977,
+      "learning_rate": 1.9979881815368406e-05,
+      "loss": 1.2152,
+      "step": 69
+    },
+    {
+      "epoch": 0.050743022834360274,
+      "grad_norm": 3.0034023909101233,
+      "learning_rate": 1.9978364530054465e-05,
+      "loss": 1.1339,
+      "step": 70
+    },
+    {
+      "epoch": 0.05146792316056542,
+      "grad_norm": 3.4428257188696794,
+      "learning_rate": 1.9976792151781778e-05,
+      "loss": 1.2363,
+      "step": 71
+    },
+    {
+      "epoch": 0.05219282348677057,
+      "grad_norm": 2.9622416635006408,
+      "learning_rate": 1.997516468923183e-05,
+      "loss": 1.274,
+      "step": 72
+    },
+    {
+      "epoch": 0.052917723812975714,
+      "grad_norm": 2.5702004303735455,
+      "learning_rate": 1.9973482151390228e-05,
+      "loss": 1.1873,
+      "step": 73
+    },
+    {
+      "epoch": 0.05364262413918086,
+      "grad_norm": 3.3076211834809492,
+      "learning_rate": 1.9971744547546676e-05,
+      "loss": 1.228,
+      "step": 74
+    },
+    {
+      "epoch": 0.05436752446538601,
+      "grad_norm": 2.6091996272999243,
+      "learning_rate": 1.99699518872949e-05,
+      "loss": 1.0701,
+      "step": 75
+    },
+    {
+      "epoch": 0.055092424791591155,
+      "grad_norm": 2.8268323348853044,
+      "learning_rate": 1.996810418053261e-05,
+      "loss": 1.2859,
+      "step": 76
+    },
+    {
+      "epoch": 0.0558173251177963,
+      "grad_norm": 2.5281726771474244,
+      "learning_rate": 1.996620143746144e-05,
+      "loss": 1.1387,
+      "step": 77
+    },
+    {
+      "epoch": 0.05654222544400145,
+      "grad_norm": 2.300493492648369,
+      "learning_rate": 1.99642436685869e-05,
+      "loss": 1.0669,
+      "step": 78
+    },
+    {
+      "epoch": 0.057267125770206595,
+      "grad_norm": 3.493223728132168,
+      "learning_rate": 1.9962230884718296e-05,
+      "loss": 1.1832,
+      "step": 79
+    },
+    {
+      "epoch": 0.05799202609641174,
+      "grad_norm": 2.7188006009604293,
+      "learning_rate": 1.9960163096968702e-05,
+      "loss": 1.1356,
+      "step": 80
+    },
+    {
+      "epoch": 0.05871692642261689,
+      "grad_norm": 2.7364057030270956,
+      "learning_rate": 1.9958040316754866e-05,
+      "loss": 1.2446,
+      "step": 81
+    },
+    {
+      "epoch": 0.059441826748822035,
+      "grad_norm": 2.7307874925286475,
+      "learning_rate": 1.995586255579717e-05,
+      "loss": 1.1808,
+      "step": 82
+    },
+    {
+      "epoch": 0.06016672707502718,
+      "grad_norm": 2.4215808753597226,
+      "learning_rate": 1.9953629826119562e-05,
+      "loss": 1.1579,
+      "step": 83
+    },
+    {
+      "epoch": 0.06089162740123233,
+      "grad_norm": 2.631267713200398,
+      "learning_rate": 1.9951342140049483e-05,
+      "loss": 1.1941,
+      "step": 84
+    },
+    {
+      "epoch": 0.061616527727437476,
+      "grad_norm": 2.673115849546717,
+      "learning_rate": 1.994899951021779e-05,
+      "loss": 1.2284,
+      "step": 85
+    },
+    {
+      "epoch": 0.06234142805364262,
+      "grad_norm": 2.6934406037224807,
+      "learning_rate": 1.9946601949558722e-05,
+      "loss": 1.206,
+      "step": 86
+    },
+    {
+      "epoch": 0.06306632837984777,
+      "grad_norm": 3.7302320786317673,
+      "learning_rate": 1.9944149471309777e-05,
+      "loss": 1.2247,
+      "step": 87
+    },
+    {
+      "epoch": 0.06379122870605292,
+      "grad_norm": 2.4550165493932363,
+      "learning_rate": 1.9941642089011684e-05,
+      "loss": 1.0596,
+      "step": 88
+    },
+    {
+      "epoch": 0.06451612903225806,
+      "grad_norm": 2.868687433143366,
+      "learning_rate": 1.9939079816508312e-05,
+      "loss": 1.1435,
+      "step": 89
+    },
+    {
+      "epoch": 0.06524102935846321,
+      "grad_norm": 3.0984151506536586,
+      "learning_rate": 1.9936462667946574e-05,
+      "loss": 1.1143,
+      "step": 90
+    },
+    {
+      "epoch": 0.06596592968466836,
+      "grad_norm": 2.483528934101709,
+      "learning_rate": 1.993379065777639e-05,
+      "loss": 1.1165,
+      "step": 91
+    },
+    {
+      "epoch": 0.0666908300108735,
+      "grad_norm": 2.4080631755195925,
+      "learning_rate": 1.9931063800750564e-05,
+      "loss": 1.1144,
+      "step": 92
+    },
+    {
+      "epoch": 0.06741573033707865,
+      "grad_norm": 2.8973853952983375,
+      "learning_rate": 1.9928282111924736e-05,
+      "loss": 1.2365,
+      "step": 93
+    },
+    {
+      "epoch": 0.0681406306632838,
+      "grad_norm": 3.3833221119081656,
+      "learning_rate": 1.9925445606657286e-05,
+      "loss": 1.2146,
+      "step": 94
+    },
+    {
+      "epoch": 0.06886553098948894,
+      "grad_norm": 2.908934165124817,
+      "learning_rate": 1.9922554300609237e-05,
+      "loss": 1.1916,
+      "step": 95
+    },
+    {
+      "epoch": 0.06959043131569409,
+      "grad_norm": 2.512496806932867,
+      "learning_rate": 1.991960820974419e-05,
+      "loss": 1.0935,
+      "step": 96
+    },
+    {
+      "epoch": 0.07031533164189924,
+      "grad_norm": 2.825963677244521,
+      "learning_rate": 1.991660735032822e-05,
+      "loss": 1.0931,
+      "step": 97
+    },
+    {
+      "epoch": 0.07104023196810438,
+      "grad_norm": 2.8818264863814296,
+      "learning_rate": 1.9913551738929803e-05,
+      "loss": 1.2153,
+      "step": 98
+    },
+    {
+      "epoch": 0.07176513229430953,
+      "grad_norm": 2.7275750720550147,
+      "learning_rate": 1.9910441392419702e-05,
+      "loss": 1.193,
+      "step": 99
+    },
+    {
+      "epoch": 0.07249003262051468,
+      "grad_norm": 3.462468824391245,
+      "learning_rate": 1.9907276327970892e-05,
+      "loss": 1.1469,
+      "step": 100
+    },
+    {
+      "epoch": 0.07321493294671982,
+      "grad_norm": 3.3377754324869224,
+      "learning_rate": 1.990405656305846e-05,
+      "loss": 1.2357,
+      "step": 101
+    },
+    {
+      "epoch": 0.07393983327292497,
+      "grad_norm": 2.839009968658976,
+      "learning_rate": 1.9900782115459503e-05,
+      "loss": 1.1848,
+      "step": 102
+    },
+    {
+      "epoch": 0.07466473359913012,
+      "grad_norm": 2.1391099705614063,
+      "learning_rate": 1.9897453003253035e-05,
+      "loss": 1.1225,
+      "step": 103
+    },
+    {
+      "epoch": 0.07538963392533526,
+      "grad_norm": 2.7719189493972243,
+      "learning_rate": 1.9894069244819893e-05,
+      "loss": 1.1538,
+      "step": 104
+    },
+    {
+      "epoch": 0.07611453425154041,
+      "grad_norm": 2.3940421507214933,
+      "learning_rate": 1.9890630858842614e-05,
+      "loss": 1.2852,
+      "step": 105
+    },
+    {
+      "epoch": 0.07683943457774556,
+      "grad_norm": 2.341837519488398,
+      "learning_rate": 1.9887137864305365e-05,
+      "loss": 1.1116,
+      "step": 106
+    },
+    {
+      "epoch": 0.0775643349039507,
+      "grad_norm": 2.403845819495617,
+      "learning_rate": 1.9883590280493815e-05,
+      "loss": 1.1509,
+      "step": 107
+    },
+    {
+      "epoch": 0.07828923523015585,
+      "grad_norm": 3.348129496258059,
+      "learning_rate": 1.9879988126995023e-05,
+      "loss": 1.1837,
+      "step": 108
+    },
+    {
+      "epoch": 0.079014135556361,
+      "grad_norm": 2.591536638052907,
+      "learning_rate": 1.9876331423697345e-05,
+      "loss": 1.1223,
+      "step": 109
+    },
+    {
+      "epoch": 0.07973903588256614,
+      "grad_norm": 2.3592029882899648,
+      "learning_rate": 1.9872620190790334e-05,
+      "loss": 1.143,
+      "step": 110
+    },
+    {
+      "epoch": 0.08046393620877129,
+      "grad_norm": 2.1392569469392346,
+      "learning_rate": 1.9868854448764594e-05,
+      "loss": 1.0607,
+      "step": 111
+    },
+    {
+      "epoch": 0.08118883653497644,
+      "grad_norm": 2.863527881373612,
+      "learning_rate": 1.9865034218411698e-05,
+      "loss": 1.1717,
+      "step": 112
+    },
+    {
+      "epoch": 0.08191373686118159,
+      "grad_norm": 2.924899234680409,
+      "learning_rate": 1.986115952082406e-05,
+      "loss": 1.2318,
+      "step": 113
+    },
+    {
+      "epoch": 0.08263863718738673,
+      "grad_norm": 2.433809686371776,
+      "learning_rate": 1.985723037739482e-05,
+      "loss": 1.2658,
+      "step": 114
+    },
+    {
+      "epoch": 0.08336353751359188,
+      "grad_norm": 2.8856283656848905,
+      "learning_rate": 1.9853246809817725e-05,
+      "loss": 1.3125,
+      "step": 115
+    },
+    {
+      "epoch": 0.08408843783979703,
+      "grad_norm": 2.5486082662488627,
+      "learning_rate": 1.9849208840087018e-05,
+      "loss": 1.1655,
+      "step": 116
+    },
+    {
+      "epoch": 0.08481333816600217,
+      "grad_norm": 2.769678541141219,
+      "learning_rate": 1.98451164904973e-05,
+      "loss": 1.1953,
+      "step": 117
+    },
+    {
+      "epoch": 0.08553823849220732,
+      "grad_norm": 2.473525065101367,
+      "learning_rate": 1.984096978364342e-05,
+      "loss": 1.1201,
+      "step": 118
+    },
+    {
+      "epoch": 0.08626313881841247,
+      "grad_norm": 2.7533405016925583,
+      "learning_rate": 1.9836768742420355e-05,
+      "loss": 1.2437,
+      "step": 119
+    },
+    {
+      "epoch": 0.08698803914461761,
+      "grad_norm": 2.4288996024121605,
+      "learning_rate": 1.983251339002306e-05,
+      "loss": 1.2333,
+      "step": 120
+    },
+    {
+      "epoch": 0.08771293947082276,
+      "grad_norm": 2.8989781107762616,
+      "learning_rate": 1.9828203749946362e-05,
+      "loss": 1.1184,
+      "step": 121
+    },
+    {
+      "epoch": 0.0884378397970279,
+      "grad_norm": 2.3092483374926527,
+      "learning_rate": 1.982383984598483e-05,
+      "loss": 1.0504,
+      "step": 122
+    },
+    {
+      "epoch": 0.08916274012323305,
+      "grad_norm": 2.714849936506903,
+      "learning_rate": 1.9819421702232624e-05,
+      "loss": 1.2218,
+      "step": 123
+    },
+    {
+      "epoch": 0.0898876404494382,
+      "grad_norm": 2.8105161825117286,
+      "learning_rate": 1.9814949343083383e-05,
+      "loss": 1.0007,
+      "step": 124
+    },
+    {
+      "epoch": 0.09061254077564335,
+      "grad_norm": 4.147738573133108,
+      "learning_rate": 1.981042279323009e-05,
+      "loss": 1.2543,
+      "step": 125
+    },
+    {
+      "epoch": 0.0913374411018485,
+      "grad_norm": 2.8310310979565783,
+      "learning_rate": 1.9805842077664913e-05,
+      "loss": 1.146,
+      "step": 126
+    },
+    {
+      "epoch": 0.09206234142805364,
+      "grad_norm": 3.0754511920919376,
+      "learning_rate": 1.9801207221679085e-05,
+      "loss": 1.203,
+      "step": 127
+    },
+    {
+      "epoch": 0.09278724175425879,
+      "grad_norm": 2.7063407512582787,
+      "learning_rate": 1.979651825086277e-05,
+      "loss": 1.0783,
+      "step": 128
+    },
+    {
+      "epoch": 0.09351214208046393,
+      "grad_norm": 2.511083096485902,
+      "learning_rate": 1.97917751911049e-05,
+      "loss": 1.1531,
+      "step": 129
+    },
+    {
+      "epoch": 0.09423704240666908,
+      "grad_norm": 2.4741075426355614,
+      "learning_rate": 1.9786978068593062e-05,
+      "loss": 1.1428,
+      "step": 130
+    },
+    {
+      "epoch": 0.09496194273287423,
+      "grad_norm": 2.5626898456386815,
+      "learning_rate": 1.978212690981332e-05,
+      "loss": 1.1239,
+      "step": 131
+    },
+    {
+      "epoch": 0.09568684305907937,
+      "grad_norm": 2.2391021429874263,
+      "learning_rate": 1.9777221741550096e-05,
+      "loss": 1.0702,
+      "step": 132
+    },
+    {
+      "epoch": 0.09641174338528452,
+      "grad_norm": 2.4369139499586074,
+      "learning_rate": 1.9772262590886006e-05,
+      "loss": 1.1692,
+      "step": 133
+    },
+    {
+      "epoch": 0.09713664371148967,
+      "grad_norm": 2.7781914212064747,
+      "learning_rate": 1.976724948520172e-05,
+      "loss": 1.1412,
+      "step": 134
+    },
+    {
+      "epoch": 0.09786154403769481,
+      "grad_norm": 2.5311023412287623,
+      "learning_rate": 1.9762182452175806e-05,
+      "loss": 1.2879,
+      "step": 135
+    },
+    {
+      "epoch": 0.09858644436389996,
+      "grad_norm": 3.2212943074374327,
+      "learning_rate": 1.9757061519784568e-05,
+      "loss": 1.1971,
+      "step": 136
+    },
+    {
+      "epoch": 0.09931134469010511,
+      "grad_norm": 2.549155447627049,
+      "learning_rate": 1.9751886716301914e-05,
+      "loss": 1.1439,
+      "step": 137
+    },
+    {
+      "epoch": 0.10003624501631025,
+      "grad_norm": 2.59925115990486,
+      "learning_rate": 1.9746658070299188e-05,
+      "loss": 1.1126,
+      "step": 138
+    },
+    {
+      "epoch": 0.1007611453425154,
+      "grad_norm": 3.1442581997348364,
+      "learning_rate": 1.9741375610645e-05,
+      "loss": 1.166,
+      "step": 139
+    },
+    {
+      "epoch": 0.10148604566872055,
+      "grad_norm": 2.7390978310814456,
+      "learning_rate": 1.9736039366505087e-05,
+      "loss": 1.1593,
+      "step": 140
+    },
+    {
+      "epoch": 0.1022109459949257,
+      "grad_norm": 2.68099301776624,
+      "learning_rate": 1.9730649367342137e-05,
+      "loss": 1.091,
+      "step": 141
+    },
+    {
+      "epoch": 0.10293584632113084,
+      "grad_norm": 2.5055718294224474,
+      "learning_rate": 1.9725205642915645e-05,
+      "loss": 1.1567,
+      "step": 142
+    },
+    {
+      "epoch": 0.10366074664733599,
+      "grad_norm": 2.4806609688516064,
+      "learning_rate": 1.971970822328172e-05,
+      "loss": 1.1877,
+      "step": 143
+    },
+    {
+      "epoch": 0.10438564697354114,
+      "grad_norm": 2.446019560687823,
+      "learning_rate": 1.9714157138792946e-05,
+      "loss": 0.8962,
+      "step": 144
+    },
+    {
+      "epoch": 0.10511054729974628,
+      "grad_norm": 2.6145362990605006,
+      "learning_rate": 1.9708552420098196e-05,
+      "loss": 1.1652,
+      "step": 145
+    },
+    {
+      "epoch": 0.10583544762595143,
+      "grad_norm": 2.8215724838180303,
+      "learning_rate": 1.9702894098142476e-05,
+      "loss": 1.1084,
+      "step": 146
+    },
+    {
+      "epoch": 0.10656034795215658,
+      "grad_norm": 2.3497120281958503,
+      "learning_rate": 1.969718220416675e-05,
+      "loss": 1.167,
+      "step": 147
+    },
+    {
+      "epoch": 0.10728524827836172,
+      "grad_norm": 2.583947968920025,
+      "learning_rate": 1.9691416769707763e-05,
+      "loss": 1.1908,
+      "step": 148
+    },
+    {
+      "epoch": 0.10801014860456687,
+      "grad_norm": 2.751597734547836,
+      "learning_rate": 1.968559782659787e-05,
+      "loss": 1.2345,
+      "step": 149
+    },
+    {
+      "epoch": 0.10873504893077202,
+      "grad_norm": 2.6976314838665685,
+      "learning_rate": 1.9679725406964853e-05,
+      "loss": 1.1939,
+      "step": 150
+    },
+    {
+      "epoch": 0.10945994925697716,
+      "grad_norm": 2.1954623751957905,
+      "learning_rate": 1.967379954323177e-05,
+      "loss": 1.1555,
+      "step": 151
+    },
+    {
+      "epoch": 0.11018484958318231,
+      "grad_norm": 2.7043983207505624,
+      "learning_rate": 1.9667820268116732e-05,
+      "loss": 1.1899,
+      "step": 152
+    },
+    {
+      "epoch": 0.11090974990938746,
+      "grad_norm": 2.076332249665599,
+      "learning_rate": 1.966178761463277e-05,
+      "loss": 1.1009,
+      "step": 153
+    },
+    {
+      "epoch": 0.1116346502355926,
+      "grad_norm": 2.3441526940711097,
+      "learning_rate": 1.965570161608762e-05,
+      "loss": 1.1319,
+      "step": 154
+    },
+    {
+      "epoch": 0.11235955056179775,
+      "grad_norm": 2.2285061151426113,
+      "learning_rate": 1.964956230608354e-05,
+      "loss": 1.1211,
+      "step": 155
+    },
+    {
+      "epoch": 0.1130844508880029,
+      "grad_norm": 2.3200577550605064,
+      "learning_rate": 1.9643369718517152e-05,
+      "loss": 1.1515,
+      "step": 156
+    },
+    {
+      "epoch": 0.11380935121420804,
+      "grad_norm": 2.5448094549925844,
+      "learning_rate": 1.9637123887579217e-05,
+      "loss": 1.302,
+      "step": 157
+    },
+    {
+      "epoch": 0.11453425154041319,
+      "grad_norm": 2.1389116260296714,
+      "learning_rate": 1.963082484775448e-05,
+      "loss": 1.1653,
+      "step": 158
+    },
+    {
+      "epoch": 0.11525915186661834,
+      "grad_norm": 2.752841944773758,
+      "learning_rate": 1.9624472633821464e-05,
+      "loss": 1.1178,
+      "step": 159
+    },
+    {
+      "epoch": 0.11598405219282348,
+      "grad_norm": 2.3858753840621345,
+      "learning_rate": 1.961806728085227e-05,
+      "loss": 1.0138,
+      "step": 160
+    },
+    {
+      "epoch": 0.11670895251902863,
+      "grad_norm": 2.6783021219358694,
+      "learning_rate": 1.9611608824212395e-05,
+      "loss": 1.2371,
+      "step": 161
+    },
+    {
+      "epoch": 0.11743385284523378,
+      "grad_norm": 2.626777858203357,
+      "learning_rate": 1.9605097299560545e-05,
+      "loss": 1.0034,
+      "step": 162
+    },
+    {
+      "epoch": 0.11815875317143892,
+      "grad_norm": 2.333729825834433,
+      "learning_rate": 1.959853274284841e-05,
+      "loss": 1.1103,
+      "step": 163
+    },
+    {
+      "epoch": 0.11888365349764407,
+      "grad_norm": 2.5562196927508425,
+      "learning_rate": 1.95919151903205e-05,
+      "loss": 1.2076,
+      "step": 164
+    },
+    {
+      "epoch": 0.11960855382384922,
+      "grad_norm": 3.2522325367194425,
+      "learning_rate": 1.9585244678513913e-05,
+      "loss": 1.258,
+      "step": 165
+    },
+    {
+      "epoch": 0.12033345415005436,
+      "grad_norm": 3.792145378660367,
+      "learning_rate": 1.9578521244258162e-05,
+      "loss": 1.1639,
+      "step": 166
+    },
+    {
+      "epoch": 0.12105835447625951,
+      "grad_norm": 3.4172955897048207,
+      "learning_rate": 1.9571744924674943e-05,
+      "loss": 1.1899,
+      "step": 167
+    },
+    {
+      "epoch": 0.12178325480246466,
+      "grad_norm": 2.884712209279333,
+      "learning_rate": 1.9564915757177955e-05,
+      "loss": 1.1658,
+      "step": 168
+    },
+    {
+      "epoch": 0.1225081551286698,
+      "grad_norm": 2.584866024173676,
+      "learning_rate": 1.955803377947268e-05,
+      "loss": 1.1995,
+      "step": 169
+    },
+    {
+      "epoch": 0.12323305545487495,
+      "grad_norm": 2.608204074681373,
+      "learning_rate": 1.955109902955617e-05,
+      "loss": 1.1891,
+      "step": 170
+    },
+    {
+      "epoch": 0.1239579557810801,
+      "grad_norm": 3.021997715506911,
+      "learning_rate": 1.954411154571687e-05,
+      "loss": 1.1485,
+      "step": 171
+    },
+    {
+      "epoch": 0.12468285610728524,
+      "grad_norm": 2.136848876157592,
+      "learning_rate": 1.9537071366534352e-05,
+      "loss": 1.0598,
+      "step": 172
+    },
+    {
+      "epoch": 0.1254077564334904,
+      "grad_norm": 2.4835581901419332,
+      "learning_rate": 1.9529978530879144e-05,
+      "loss": 1.1085,
+      "step": 173
+    },
+    {
+      "epoch": 0.12613265675969554,
+      "grad_norm": 2.3827179769931823,
+      "learning_rate": 1.9522833077912512e-05,
+      "loss": 1.1484,
+      "step": 174
+    },
+    {
+      "epoch": 0.12685755708590069,
+      "grad_norm": 2.256838168033284,
+      "learning_rate": 1.951563504708622e-05,
+      "loss": 1.1174,
+      "step": 175
+    },
+    {
+      "epoch": 0.12758245741210583,
+      "grad_norm": 2.677664336590985,
+      "learning_rate": 1.950838447814233e-05,
+      "loss": 1.1908,
+      "step": 176
+    },
+    {
+      "epoch": 0.12830735773831098,
+      "grad_norm": 2.851750128481314,
+      "learning_rate": 1.9501081411112987e-05,
+      "loss": 1.2887,
+      "step": 177
+    },
+    {
+      "epoch": 0.12903225806451613,
+      "grad_norm": 2.299218668392714,
+      "learning_rate": 1.9493725886320193e-05,
+      "loss": 1.0708,
+      "step": 178
+    },
+    {
+      "epoch": 0.12975715839072127,
+      "grad_norm": 2.654535712782758,
+      "learning_rate": 1.9486317944375563e-05,
+      "loss": 1.1428,
+      "step": 179
+    },
+    {
+      "epoch": 0.13048205871692642,
+      "grad_norm": 3.1562180739485184,
+      "learning_rate": 1.947885762618014e-05,
+      "loss": 1.1446,
+      "step": 180
+    },
+    {
+      "epoch": 0.13120695904313157,
+      "grad_norm": 2.4022359841083762,
+      "learning_rate": 1.9471344972924142e-05,
+      "loss": 1.1913,
+      "step": 181
+    },
+    {
+      "epoch": 0.1319318593693367,
+      "grad_norm": 1.9874430354522221,
+      "learning_rate": 1.9463780026086735e-05,
+      "loss": 1.1536,
+      "step": 182
+    },
+    {
+      "epoch": 0.13265675969554186,
+      "grad_norm": 2.6028657856098762,
+      "learning_rate": 1.945616282743582e-05,
+      "loss": 1.0459,
+      "step": 183
+    },
+    {
+      "epoch": 0.133381660021747,
+      "grad_norm": 2.4961285769674584,
+      "learning_rate": 1.9448493419027795e-05,
+      "loss": 1.2184,
+      "step": 184
+    },
+    {
+      "epoch": 0.13410656034795215,
+      "grad_norm": 2.4545736340279607,
+      "learning_rate": 1.9440771843207305e-05,
+      "loss": 1.0966,
+      "step": 185
+    },
+    {
+      "epoch": 0.1348314606741573,
+      "grad_norm": 2.4064312104977796,
+      "learning_rate": 1.9432998142607036e-05,
+      "loss": 1.1427,
+      "step": 186
+    },
+    {
+      "epoch": 0.13555636100036245,
+      "grad_norm": 2.10260997961401,
+      "learning_rate": 1.9425172360147467e-05,
+      "loss": 1.153,
+      "step": 187
+    },
+    {
+      "epoch": 0.1362812613265676,
+      "grad_norm": 2.2184620974349,
+      "learning_rate": 1.9417294539036634e-05,
+      "loss": 1.1356,
+      "step": 188
+    },
+    {
+      "epoch": 0.13700616165277274,
+      "grad_norm": 2.715658298293896,
+      "learning_rate": 1.9409364722769882e-05,
+      "loss": 1.2261,
+      "step": 189
+    },
+    {
+      "epoch": 0.1377310619789779,
+      "grad_norm": 2.240649838269821,
+      "learning_rate": 1.9401382955129646e-05,
+      "loss": 1.0599,
+      "step": 190
+    },
+    {
+      "epoch": 0.13845596230518303,
+      "grad_norm": 3.0178674865097728,
+      "learning_rate": 1.9393349280185187e-05,
+      "loss": 1.2233,
+      "step": 191
+    },
+    {
+      "epoch": 0.13918086263138818,
+      "grad_norm": 2.132340826672663,
+      "learning_rate": 1.9385263742292368e-05,
+      "loss": 1.0894,
+      "step": 192
+    },
+    {
+      "epoch": 0.13990576295759333,
+      "grad_norm": 2.4647699353823733,
+      "learning_rate": 1.937712638609339e-05,
+      "loss": 1.2108,
+      "step": 193
+    },
+    {
+      "epoch": 0.14063066328379847,
+      "grad_norm": 2.4490272195884852,
+      "learning_rate": 1.9368937256516567e-05,
+      "loss": 1.1608,
+      "step": 194
+    },
+    {
+      "epoch": 0.14135556361000362,
+      "grad_norm": 2.1989447827390665,
+      "learning_rate": 1.9360696398776056e-05,
+      "loss": 1.1457,
+      "step": 195
+    },
+    {
+      "epoch": 0.14208046393620877,
+      "grad_norm": 2.068383189183824,
+      "learning_rate": 1.9352403858371618e-05,
+      "loss": 1.1081,
+      "step": 196
+    },
+    {
+      "epoch": 0.14280536426241391,
+      "grad_norm": 2.0311982289386474,
+      "learning_rate": 1.9344059681088372e-05,
+      "loss": 1.1035,
+      "step": 197
+    },
+    {
+      "epoch": 0.14353026458861906,
+      "grad_norm": 2.486963653282551,
+      "learning_rate": 1.933566391299654e-05,
+      "loss": 1.0822,
+      "step": 198
+    },
+    {
+      "epoch": 0.1442551649148242,
+      "grad_norm": 2.3367902123479314,
+      "learning_rate": 1.9327216600451177e-05,
+      "loss": 1.017,
+      "step": 199
+    },
+    {
+      "epoch": 0.14498006524102935,
+      "grad_norm": 2.390926126929064,
+      "learning_rate": 1.931871779009194e-05,
+      "loss": 1.0842,
+      "step": 200
+    },
+    {
+      "epoch": 0.1457049655672345,
+      "grad_norm": 3.3906627893855936,
+      "learning_rate": 1.9310167528842808e-05,
+      "loss": 1.2239,
+      "step": 201
+    },
+    {
+      "epoch": 0.14642986589343965,
+      "grad_norm": 2.36160531790082,
+      "learning_rate": 1.9301565863911836e-05,
+      "loss": 1.1343,
+      "step": 202
+    },
+    {
+      "epoch": 0.1471547662196448,
+      "grad_norm": 2.4172767818260885,
+      "learning_rate": 1.9292912842790893e-05,
+      "loss": 1.1009,
+      "step": 203
+    },
+    {
+      "epoch": 0.14787966654584994,
+      "grad_norm": 2.391111956498114,
+      "learning_rate": 1.92842085132554e-05,
+      "loss": 1.1245,
+      "step": 204
+    },
+    {
+      "epoch": 0.1486045668720551,
+      "grad_norm": 2.9512769517526527,
+      "learning_rate": 1.927545292336406e-05,
+      "loss": 1.1341,
+      "step": 205
+    },
+    {
+      "epoch": 0.14932946719826024,
+      "grad_norm": 2.247325705969668,
+      "learning_rate": 1.9266646121458597e-05,
+      "loss": 1.1099,
+      "step": 206
+    },
+    {
+      "epoch": 0.15005436752446538,
+      "grad_norm": 4.45573844838241,
+      "learning_rate": 1.9257788156163484e-05,
+      "loss": 1.2177,
+      "step": 207
+    },
+    {
+      "epoch": 0.15077926785067053,
+      "grad_norm": 2.308206270902384,
+      "learning_rate": 1.9248879076385696e-05,
+      "loss": 1.1115,
+      "step": 208
+    },
+    {
+      "epoch": 0.15150416817687568,
+      "grad_norm": 2.655262555875558,
+      "learning_rate": 1.9239918931314404e-05,
+      "loss": 1.1081,
+      "step": 209
+    },
+    {
+      "epoch": 0.15222906850308082,
+      "grad_norm": 2.939655738673358,
+      "learning_rate": 1.9230907770420737e-05,
+      "loss": 1.1192,
+      "step": 210
+    },
+    {
+      "epoch": 0.15295396882928597,
+      "grad_norm": 2.7252753733179564,
+      "learning_rate": 1.9221845643457485e-05,
+      "loss": 1.1471,
+      "step": 211
+    },
+    {
+      "epoch": 0.15367886915549112,
+      "grad_norm": 2.7708259653245677,
+      "learning_rate": 1.9212732600458845e-05,
+      "loss": 1.0776,
+      "step": 212
+    },
+    {
+      "epoch": 0.15440376948169626,
+      "grad_norm": 1.9749460870383841,
+      "learning_rate": 1.9203568691740115e-05,
+      "loss": 1.0709,
+      "step": 213
+    },
+    {
+      "epoch": 0.1551286698079014,
+      "grad_norm": 2.0889553423500353,
+      "learning_rate": 1.9194353967897453e-05,
+      "loss": 1.1986,
+      "step": 214
+    },
+    {
+      "epoch": 0.15585357013410656,
+      "grad_norm": 2.5179546631862624,
+      "learning_rate": 1.9185088479807577e-05,
+      "loss": 1.2635,
+      "step": 215
+    },
+    {
+      "epoch": 0.1565784704603117,
+      "grad_norm": 2.7700789430710424,
+      "learning_rate": 1.9175772278627477e-05,
+      "loss": 1.2196,
+      "step": 216
+    },
+    {
+      "epoch": 0.15730337078651685,
+      "grad_norm": 2.3103773826109313,
+      "learning_rate": 1.916640541579415e-05,
+      "loss": 1.1924,
+      "step": 217
+    },
+    {
+      "epoch": 0.158028271112722,
+      "grad_norm": 3.7179435673997587,
+      "learning_rate": 1.9156987943024305e-05,
+      "loss": 1.0714,
+      "step": 218
+    },
+    {
+      "epoch": 0.15875317143892714,
+      "grad_norm": 3.047980163715747,
+      "learning_rate": 1.9147519912314086e-05,
+      "loss": 1.0348,
+      "step": 219
+    },
+    {
+      "epoch": 0.1594780717651323,
+      "grad_norm": 2.3355941636809323,
+      "learning_rate": 1.9138001375938773e-05,
+      "loss": 1.0714,
+      "step": 220
+    },
+    {
+      "epoch": 0.16020297209133744,
+      "grad_norm": 2.4367237951314764,
+      "learning_rate": 1.9128432386452503e-05,
+      "loss": 1.0923,
+      "step": 221
+    },
+    {
+      "epoch": 0.16092787241754258,
+      "grad_norm": 2.534850993992721,
+      "learning_rate": 1.9118812996687975e-05,
+      "loss": 1.2025,
+      "step": 222
+    },
+    {
+      "epoch": 0.16165277274374773,
+      "grad_norm": 2.6850585710213637,
+      "learning_rate": 1.910914325975616e-05,
+      "loss": 1.2315,
+      "step": 223
+    },
+    {
+      "epoch": 0.16237767306995288,
+      "grad_norm": 2.6390954960986,
+      "learning_rate": 1.9099423229046015e-05,
+      "loss": 1.171,
+      "step": 224
+    },
+    {
+      "epoch": 0.16310257339615802,
+      "grad_norm": 3.674303992733298,
+      "learning_rate": 1.908965295822417e-05,
+      "loss": 1.22,
+      "step": 225
+    },
+    {
+      "epoch": 0.16382747372236317,
+      "grad_norm": 2.3198272055347413,
+      "learning_rate": 1.907983250123465e-05,
+      "loss": 0.9999,
+      "step": 226
+    },
+    {
+      "epoch": 0.16455237404856832,
+      "grad_norm": 2.367941394706144,
+      "learning_rate": 1.906996191229857e-05,
+      "loss": 1.1324,
+      "step": 227
+    },
+    {
+      "epoch": 0.16527727437477346,
+      "grad_norm": 2.2348457223954536,
+      "learning_rate": 1.906004124591383e-05,
+      "loss": 1.1695,
+      "step": 228
+    },
+    {
+      "epoch": 0.1660021747009786,
+      "grad_norm": 2.5357413278993373,
+      "learning_rate": 1.905007055685483e-05,
+      "loss": 1.2143,
+      "step": 229
+    },
+    {
+      "epoch": 0.16672707502718376,
+      "grad_norm": 2.197512694165324,
+      "learning_rate": 1.904004990017214e-05,
+      "loss": 1.059,
+      "step": 230
+    },
+    {
+      "epoch": 0.1674519753533889,
+      "grad_norm": 2.246714216934824,
+      "learning_rate": 1.902997933119223e-05,
+      "loss": 1.0306,
+      "step": 231
+    },
+    {
+      "epoch": 0.16817687567959405,
+      "grad_norm": 2.3508232503422257,
+      "learning_rate": 1.9019858905517146e-05,
+      "loss": 1.232,
+      "step": 232
+    },
+    {
+      "epoch": 0.1689017760057992,
+      "grad_norm": 3.2397999719968706,
+      "learning_rate": 1.900968867902419e-05,
+      "loss": 1.1642,
+      "step": 233
+    },
+    {
+      "epoch": 0.16962667633200434,
+      "grad_norm": 2.2082937384067134,
+      "learning_rate": 1.899946870786565e-05,
+      "loss": 1.0615,
+      "step": 234
+    },
+    {
+      "epoch": 0.1703515766582095,
+      "grad_norm": 2.6489946010926437,
+      "learning_rate": 1.8989199048468443e-05,
+      "loss": 1.2625,
+      "step": 235
+    },
+    {
+      "epoch": 0.17107647698441464,
+      "grad_norm": 2.243711710470245,
+      "learning_rate": 1.8978879757533838e-05,
+      "loss": 1.1316,
+      "step": 236
+    },
+    {
+      "epoch": 0.17180137731061979,
+      "grad_norm": 2.846092716410688,
+      "learning_rate": 1.8968510892037136e-05,
+      "loss": 1.2307,
+      "step": 237
+    },
+    {
+      "epoch": 0.17252627763682493,
+      "grad_norm": 2.7609914990765128,
+      "learning_rate": 1.8958092509227347e-05,
+      "loss": 1.1384,
+      "step": 238
+    },
+    {
+      "epoch": 0.17325117796303008,
+      "grad_norm": 1.8868238757930658,
+      "learning_rate": 1.8947624666626866e-05,
+      "loss": 1.0692,
+      "step": 239
+    },
+    {
+      "epoch": 0.17397607828923523,
+      "grad_norm": 2.5980341998176146,
+      "learning_rate": 1.893710742203119e-05,
+      "loss": 1.1284,
+      "step": 240
+    },
+    {
+      "epoch": 0.17470097861544037,
+      "grad_norm": 2.7513723878066982,
+      "learning_rate": 1.8926540833508557e-05,
+      "loss": 1.0836,
+      "step": 241
+    },
+    {
+      "epoch": 0.17542587894164552,
+      "grad_norm": 2.6788196765201535,
+      "learning_rate": 1.8915924959399656e-05,
+      "loss": 1.1034,
+      "step": 242
+    },
+    {
+      "epoch": 0.17615077926785067,
+      "grad_norm": 2.5118625400183023,
+      "learning_rate": 1.8905259858317287e-05,
+      "loss": 1.2339,
+      "step": 243
+    },
+    {
+      "epoch": 0.1768756795940558,
+      "grad_norm": 2.126239831693274,
+      "learning_rate": 1.889454558914605e-05,
+      "loss": 1.0637,
+      "step": 244
+    },
+    {
+      "epoch": 0.17760057992026096,
+      "grad_norm": 2.604602180586931,
+      "learning_rate": 1.888378221104201e-05,
+      "loss": 1.2855,
+      "step": 245
+    },
+    {
+      "epoch": 0.1783254802464661,
+      "grad_norm": 2.4753012194229562,
+      "learning_rate": 1.8872969783432376e-05,
+      "loss": 1.0602,
+      "step": 246
+    },
+    {
+      "epoch": 0.17905038057267125,
+      "grad_norm": 2.6860064054799584,
+      "learning_rate": 1.886210836601517e-05,
+      "loss": 1.1418,
+      "step": 247
+    },
+    {
+      "epoch": 0.1797752808988764,
+      "grad_norm": 2.5126660296525833,
+      "learning_rate": 1.8851198018758898e-05,
+      "loss": 1.1937,
+      "step": 248
+    },
+    {
+      "epoch": 0.18050018122508155,
+      "grad_norm": 1.7715675237469672,
+      "learning_rate": 1.884023880190222e-05,
+      "loss": 1.0683,
+      "step": 249
+    },
+    {
+      "epoch": 0.1812250815512867,
+      "grad_norm": 2.3549127021241048,
+      "learning_rate": 1.8829230775953616e-05,
+      "loss": 1.1331,
+      "step": 250
+    },
+    {
+      "epoch": 0.18194998187749184,
+      "grad_norm": 2.043319865990913,
+      "learning_rate": 1.8818174001691055e-05,
+      "loss": 1.009,
+      "step": 251
+    },
+    {
+      "epoch": 0.182674882203697,
+      "grad_norm": 2.5088376984194682,
+      "learning_rate": 1.880706854016166e-05,
+      "loss": 1.0838,
+      "step": 252
+    },
+    {
+      "epoch": 0.18339978252990213,
+      "grad_norm": 2.095009396399693,
+      "learning_rate": 1.8795914452681362e-05,
+      "loss": 1.0487,
+      "step": 253
+    },
+    {
+      "epoch": 0.18412468285610728,
+      "grad_norm": 2.3982978050324073,
+      "learning_rate": 1.8784711800834564e-05,
+      "loss": 1.0634,
+      "step": 254
+    },
+    {
+      "epoch": 0.18484958318231243,
+      "grad_norm": 2.1820663845429524,
+      "learning_rate": 1.877346064647382e-05,
+      "loss": 1.0621,
+      "step": 255
+    },
+    {
+      "epoch": 0.18557448350851757,
+      "grad_norm": 2.9905602073691147,
+      "learning_rate": 1.8762161051719464e-05,
+      "loss": 1.2322,
+      "step": 256
+    },
+    {
+      "epoch": 0.18629938383472272,
+      "grad_norm": 2.2435538559731505,
+      "learning_rate": 1.8750813078959282e-05,
+      "loss": 1.0788,
+      "step": 257
+    },
+    {
+      "epoch": 0.18702428416092787,
+      "grad_norm": 2.2367444808107906,
+      "learning_rate": 1.8739416790848177e-05,
+      "loss": 1.0598,
+      "step": 258
+    },
+    {
+      "epoch": 0.18774918448713301,
+      "grad_norm": 2.472646715254004,
+      "learning_rate": 1.87279722503078e-05,
+      "loss": 1.0482,
+      "step": 259
+    },
+    {
+      "epoch": 0.18847408481333816,
+      "grad_norm": 2.111989985490911,
+      "learning_rate": 1.8716479520526227e-05,
+      "loss": 1.1027,
+      "step": 260
+    },
+    {
+      "epoch": 0.1891989851395433,
+      "grad_norm": 2.510087294479276,
+      "learning_rate": 1.870493866495759e-05,
+      "loss": 1.1856,
+      "step": 261
+    },
+    {
+      "epoch": 0.18992388546574845,
+      "grad_norm": 2.5133146185218904,
+      "learning_rate": 1.8693349747321737e-05,
+      "loss": 1.1478,
+      "step": 262
+    },
+    {
+      "epoch": 0.1906487857919536,
+      "grad_norm": 2.3395156482299972,
+      "learning_rate": 1.868171283160388e-05,
+      "loss": 1.1815,
+      "step": 263
+    },
+    {
+      "epoch": 0.19137368611815875,
+      "grad_norm": 2.3764168972959707,
+      "learning_rate": 1.867002798205424e-05,
+      "loss": 1.2432,
+      "step": 264
+    },
+    {
+      "epoch": 0.1920985864443639,
+      "grad_norm": 2.559923113432804,
+      "learning_rate": 1.8658295263187692e-05,
+      "loss": 1.1154,
+      "step": 265
+    },
+    {
+      "epoch": 0.19282348677056904,
+      "grad_norm": 2.8783556981475544,
+      "learning_rate": 1.8646514739783404e-05,
+      "loss": 1.1236,
+      "step": 266
+    },
+    {
+      "epoch": 0.1935483870967742,
+      "grad_norm": 2.6502413673324985,
+      "learning_rate": 1.8634686476884497e-05,
+      "loss": 1.2676,
+      "step": 267
+    },
+    {
+      "epoch": 0.19427328742297933,
+      "grad_norm": 3.3520733595809857,
+      "learning_rate": 1.8622810539797656e-05,
+      "loss": 1.0714,
+      "step": 268
+    },
+    {
+      "epoch": 0.19499818774918448,
+      "grad_norm": 2.9157405583997353,
+      "learning_rate": 1.8610886994092802e-05,
+      "loss": 1.1882,
+      "step": 269
+    },
+    {
+      "epoch": 0.19572308807538963,
+      "grad_norm": 2.118069033541846,
+      "learning_rate": 1.8598915905602704e-05,
+      "loss": 1.1082,
+      "step": 270
+    },
+    {
+      "epoch": 0.19644798840159478,
+      "grad_norm": 2.3662623126360693,
+      "learning_rate": 1.858689734042263e-05,
+      "loss": 1.0517,
+      "step": 271
+    },
+    {
+      "epoch": 0.19717288872779992,
+      "grad_norm": 2.3124998904314658,
+      "learning_rate": 1.8574831364909988e-05,
+      "loss": 1.1954,
+      "step": 272
+    },
+    {
+      "epoch": 0.19789778905400507,
+      "grad_norm": 2.2466176708447576,
+      "learning_rate": 1.8562718045683933e-05,
+      "loss": 1.153,
+      "step": 273
+    },
+    {
+      "epoch": 0.19862268938021022,
+      "grad_norm": 3.186725132878171,
+      "learning_rate": 1.855055744962502e-05,
+      "loss": 1.0749,
+      "step": 274
+    },
+    {
+      "epoch": 0.19934758970641536,
+      "grad_norm": 1.9797977750729439,
+      "learning_rate": 1.8538349643874845e-05,
+      "loss": 1.1717,
+      "step": 275
+    },
+    {
+      "epoch": 0.2000724900326205,
+      "grad_norm": 2.650096534254559,
+      "learning_rate": 1.8526094695835647e-05,
+      "loss": 1.0998,
+      "step": 276
+    },
+    {
+      "epoch": 0.20079739035882566,
+      "grad_norm": 2.242222036665605,
+      "learning_rate": 1.851379267316995e-05,
+      "loss": 1.12,
+      "step": 277
+    },
+    {
+      "epoch": 0.2015222906850308,
+      "grad_norm": 3.0768668231596763,
+      "learning_rate": 1.8501443643800185e-05,
+      "loss": 1.2119,
+      "step": 278
+    },
+    {
+      "epoch": 0.20224719101123595,
+      "grad_norm": 2.4142131380970655,
+      "learning_rate": 1.8489047675908328e-05,
+      "loss": 1.1211,
+      "step": 279
+    },
+    {
+      "epoch": 0.2029720913374411,
+      "grad_norm": 2.494879063057067,
+      "learning_rate": 1.8476604837935515e-05,
+      "loss": 1.1135,
+      "step": 280
+    },
+    {
+      "epoch": 0.20369699166364624,
+      "grad_norm": 2.105871660944368,
+      "learning_rate": 1.846411519858165e-05,
+      "loss": 1.1952,
+      "step": 281
+    },
+    {
+      "epoch": 0.2044218919898514,
+      "grad_norm": 2.2793551122599847,
+      "learning_rate": 1.8451578826805046e-05,
+      "loss": 1.148,
+      "step": 282
+    },
+    {
+      "epoch": 0.20514679231605654,
+      "grad_norm": 2.3515542983789794,
+      "learning_rate": 1.843899579182204e-05,
+      "loss": 1.078,
+      "step": 283
+    },
+    {
+      "epoch": 0.20587169264226168,
+      "grad_norm": 2.3215932574696043,
+      "learning_rate": 1.8426366163106603e-05,
+      "loss": 0.9794,
+      "step": 284
+    },
+    {
+      "epoch": 0.20659659296846683,
+      "grad_norm": 2.9004041098455122,
+      "learning_rate": 1.841369001038997e-05,
+      "loss": 1.1092,
+      "step": 285
+    },
+    {
+      "epoch": 0.20732149329467198,
+      "grad_norm": 2.503251753777213,
+      "learning_rate": 1.8400967403660228e-05,
+      "loss": 1.1586,
+      "step": 286
+    },
+    {
+      "epoch": 0.20804639362087712,
+      "grad_norm": 2.6266283769571546,
+      "learning_rate": 1.8388198413161962e-05,
+      "loss": 1.1247,
+      "step": 287
+    },
+    {
+      "epoch": 0.20877129394708227,
+      "grad_norm": 2.6053044014908644,
+      "learning_rate": 1.837538310939586e-05,
+      "loss": 1.0471,
+      "step": 288
+    },
+    {
+      "epoch": 0.20949619427328742,
+      "grad_norm": 2.2521209365507766,
+      "learning_rate": 1.8362521563118298e-05,
+      "loss": 1.0282,
+      "step": 289
+    },
+    {
+      "epoch": 0.21022109459949256,
+      "grad_norm": 2.086975827241926,
+      "learning_rate": 1.834961384534098e-05,
+      "loss": 1.1105,
+      "step": 290
+    },
+    {
+      "epoch": 0.2109459949256977,
+      "grad_norm": 2.282332985227685,
+      "learning_rate": 1.8336660027330525e-05,
+      "loss": 1.1147,
+      "step": 291
+    },
+    {
+      "epoch": 0.21167089525190286,
+      "grad_norm": 2.4541255599254326,
+      "learning_rate": 1.83236601806081e-05,
+      "loss": 1.0968,
+      "step": 292
+    },
+    {
+      "epoch": 0.212395795578108,
+      "grad_norm": 2.5576128314214763,
+      "learning_rate": 1.8310614376948986e-05,
+      "loss": 1.2477,
+      "step": 293
+    },
+    {
+      "epoch": 0.21312069590431315,
+      "grad_norm": 2.073918446338343,
+      "learning_rate": 1.829752268838222e-05,
+      "loss": 1.0705,
+      "step": 294
+    },
+    {
+      "epoch": 0.2138455962305183,
+      "grad_norm": 3.0383016771932168,
+      "learning_rate": 1.8284385187190168e-05,
+      "loss": 1.2738,
+      "step": 295
+    },
+    {
+      "epoch": 0.21457049655672344,
+      "grad_norm": 2.5327376795510714,
+      "learning_rate": 1.8271201945908152e-05,
+      "loss": 1.0948,
+      "step": 296
+    },
+    {
+      "epoch": 0.2152953968829286,
+      "grad_norm": 2.913601708623835,
+      "learning_rate": 1.825797303732402e-05,
+      "loss": 1.1031,
+      "step": 297
+    },
+    {
+      "epoch": 0.21602029720913374,
+      "grad_norm": 2.7092322107658933,
+      "learning_rate": 1.8244698534477776e-05,
+      "loss": 1.0646,
+      "step": 298
+    },
+    {
+      "epoch": 0.21674519753533888,
+      "grad_norm": 2.2272013984412418,
+      "learning_rate": 1.823137851066115e-05,
+      "loss": 0.9988,
+      "step": 299
+    },
+    {
+      "epoch": 0.21747009786154403,
+      "grad_norm": 1.897477153962133,
+      "learning_rate": 1.821801303941721e-05,
+      "loss": 1.086,
+      "step": 300
+    },
+    {
+      "epoch": 0.21819499818774918,
+      "grad_norm": 2.055450943656558,
+      "learning_rate": 1.8204602194539948e-05,
+      "loss": 1.1499,
+      "step": 301
+    },
+    {
+      "epoch": 0.21891989851395433,
+      "grad_norm": 2.4909766282023016,
+      "learning_rate": 1.819114605007388e-05,
+      "loss": 1.1902,
+      "step": 302
+    },
+    {
+      "epoch": 0.21964479884015947,
+      "grad_norm": 2.155603573794576,
+      "learning_rate": 1.8177644680313618e-05,
+      "loss": 1.1291,
+      "step": 303
+    },
+    {
+      "epoch": 0.22036969916636462,
+      "grad_norm": 2.3903217112059845,
+      "learning_rate": 1.8164098159803485e-05,
+      "loss": 1.074,
+      "step": 304
+    },
+    {
+      "epoch": 0.22109459949256977,
+      "grad_norm": 2.4252393061914566,
+      "learning_rate": 1.8150506563337095e-05,
+      "loss": 1.1277,
+      "step": 305
+    },
+    {
+      "epoch": 0.2218194998187749,
+      "grad_norm": 2.390835505304097,
+      "learning_rate": 1.813686996595693e-05,
+      "loss": 1.0942,
+      "step": 306
+    },
+    {
+      "epoch": 0.22254440014498006,
+      "grad_norm": 2.3599664718135087,
+      "learning_rate": 1.8123188442953935e-05,
+      "loss": 1.1321,
+      "step": 307
+    },
+    {
+      "epoch": 0.2232693004711852,
+      "grad_norm": 3.4063853490696356,
+      "learning_rate": 1.81094620698671e-05,
+      "loss": 1.2013,
+      "step": 308
+    },
+    {
+      "epoch": 0.22399420079739035,
+      "grad_norm": 2.8215352516328682,
+      "learning_rate": 1.809569092248304e-05,
+      "loss": 1.1624,
+      "step": 309
+    },
+    {
+      "epoch": 0.2247191011235955,
+      "grad_norm": 2.692738685235975,
+      "learning_rate": 1.8081875076835587e-05,
+      "loss": 1.2877,
+      "step": 310
+    },
+    {
+      "epoch": 0.22544400144980065,
+      "grad_norm": 1.9994403713749573,
+      "learning_rate": 1.8068014609205363e-05,
+      "loss": 1.1385,
+      "step": 311
+    },
+    {
+      "epoch": 0.2261689017760058,
+      "grad_norm": 2.6062186167459864,
+      "learning_rate": 1.805410959611935e-05,
+      "loss": 1.17,
+      "step": 312
+    },
+    {
+      "epoch": 0.22689380210221094,
+      "grad_norm": 2.471082211880037,
+      "learning_rate": 1.804016011435048e-05,
+      "loss": 1.1468,
+      "step": 313
+    },
+    {
+      "epoch": 0.2276187024284161,
+      "grad_norm": 2.685017007295627,
+      "learning_rate": 1.802616624091721e-05,
+      "loss": 1.1734,
+      "step": 314
+    },
+    {
+      "epoch": 0.22834360275462123,
+      "grad_norm": 2.462530495353453,
+      "learning_rate": 1.8012128053083097e-05,
+      "loss": 1.1009,
+      "step": 315
+    },
+    {
+      "epoch": 0.22906850308082638,
+      "grad_norm": 2.2431993395283767,
+      "learning_rate": 1.7998045628356355e-05,
+      "loss": 1.1612,
+      "step": 316
+    },
+    {
+      "epoch": 0.22979340340703153,
+      "grad_norm": 2.3899511245334133,
+      "learning_rate": 1.7983919044489455e-05,
+      "loss": 1.1758,
+      "step": 317
+    },
+    {
+      "epoch": 0.23051830373323667,
+      "grad_norm": 2.259336195556954,
+      "learning_rate": 1.7969748379478675e-05,
+      "loss": 1.1089,
+      "step": 318
+    },
+    {
+      "epoch": 0.23124320405944182,
+      "grad_norm": 3.0578949924857013,
+      "learning_rate": 1.795553371156368e-05,
+      "loss": 1.2194,
+      "step": 319
+    },
+    {
+      "epoch": 0.23196810438564697,
+      "grad_norm": 2.4251956933972867,
+      "learning_rate": 1.7941275119227076e-05,
+      "loss": 1.1154,
+      "step": 320
+    },
+    {
+      "epoch": 0.2326930047118521,
+      "grad_norm": 2.3987899466375344,
+      "learning_rate": 1.7926972681193996e-05,
+      "loss": 1.1155,
+      "step": 321
+    },
+    {
+      "epoch": 0.23341790503805726,
+      "grad_norm": 2.0179495834873395,
+      "learning_rate": 1.7912626476431648e-05,
+      "loss": 1.1214,
+      "step": 322
+    },
+    {
+      "epoch": 0.2341428053642624,
+      "grad_norm": 2.2577973639628492,
+      "learning_rate": 1.78982365841489e-05,
+      "loss": 1.145,
+      "step": 323
+    },
+    {
+      "epoch": 0.23486770569046755,
+      "grad_norm": 2.5258056683443857,
+      "learning_rate": 1.788380308379581e-05,
+      "loss": 1.0736,
+      "step": 324
+    },
+    {
+      "epoch": 0.2355926060166727,
+      "grad_norm": 2.597512148415182,
+      "learning_rate": 1.7869326055063223e-05,
+      "loss": 1.0742,
+      "step": 325
+    },
+    {
+      "epoch": 0.23631750634287785,
+      "grad_norm": 2.4111146476669063,
+      "learning_rate": 1.7854805577882307e-05,
+      "loss": 1.1139,
+      "step": 326
+    },
+    {
+      "epoch": 0.237042406669083,
+      "grad_norm": 2.513312885502925,
+      "learning_rate": 1.784024173242412e-05,
+      "loss": 1.1792,
+      "step": 327
+    },
+    {
+      "epoch": 0.23776730699528814,
+      "grad_norm": 2.6103960099654446,
+      "learning_rate": 1.7825634599099167e-05,
+      "loss": 1.0752,
+      "step": 328
+    },
+    {
+      "epoch": 0.2384922073214933,
+      "grad_norm": 2.5153496087422136,
+      "learning_rate": 1.7810984258556955e-05,
+      "loss": 1.1024,
+      "step": 329
+    },
+    {
+      "epoch": 0.23921710764769843,
+      "grad_norm": 1.891594673564252,
+      "learning_rate": 1.779629079168556e-05,
+      "loss": 0.9794,
+      "step": 330
+    },
+    {
+      "epoch": 0.23994200797390358,
+      "grad_norm": 2.103561148427696,
+      "learning_rate": 1.7781554279611143e-05,
+      "loss": 0.9872,
+      "step": 331
+    },
+    {
+      "epoch": 0.24066690830010873,
+      "grad_norm": 2.258815019112651,
+      "learning_rate": 1.7766774803697555e-05,
+      "loss": 1.1095,
+      "step": 332
+    },
+    {
+      "epoch": 0.24139180862631388,
+      "grad_norm": 2.0559334388087938,
+      "learning_rate": 1.7751952445545846e-05,
+      "loss": 1.1,
+      "step": 333
+    },
+    {
+      "epoch": 0.24211670895251902,
+      "grad_norm": 2.382861239464527,
+      "learning_rate": 1.7737087286993832e-05,
+      "loss": 1.118,
+      "step": 334
+    },
+    {
+      "epoch": 0.24284160927872417,
+      "grad_norm": 2.0365687571462536,
+      "learning_rate": 1.7722179410115644e-05,
+      "loss": 1.1233,
+      "step": 335
+    },
+    {
+      "epoch": 0.24356650960492932,
+      "grad_norm": 2.5340895730401036,
+      "learning_rate": 1.770722889722126e-05,
+      "loss": 1.2206,
+      "step": 336
+    },
+    {
+      "epoch": 0.24429140993113446,
+      "grad_norm": 2.2828107095058865,
+      "learning_rate": 1.769223583085608e-05,
+      "loss": 1.144,
+      "step": 337
+    },
+    {
+      "epoch": 0.2450163102573396,
+      "grad_norm": 2.7476805734384495,
+      "learning_rate": 1.767720029380044e-05,
+      "loss": 1.0737,
+      "step": 338
+    },
+    {
+      "epoch": 0.24574121058354476,
+      "grad_norm": 2.1237162836325068,
+      "learning_rate": 1.7662122369069164e-05,
+      "loss": 1.0684,
+      "step": 339
+    },
+    {
+      "epoch": 0.2464661109097499,
+      "grad_norm": 2.5061834120195488,
+      "learning_rate": 1.764700213991111e-05,
+      "loss": 1.1518,
+      "step": 340
+    },
+    {
+      "epoch": 0.24719101123595505,
+      "grad_norm": 2.3826699101861353,
+      "learning_rate": 1.7631839689808724e-05,
+      "loss": 1.0676,
+      "step": 341
+    },
+    {
+      "epoch": 0.2479159115621602,
+      "grad_norm": 2.8748822951105235,
+      "learning_rate": 1.7616635102477545e-05,
+      "loss": 1.1453,
+      "step": 342
+    },
+    {
+      "epoch": 0.24864081188836534,
+      "grad_norm": 2.2078908610481434,
+      "learning_rate": 1.760138846186577e-05,
+      "loss": 0.9728,
+      "step": 343
+    },
+    {
+      "epoch": 0.2493657122145705,
+      "grad_norm": 2.750380120949678,
+      "learning_rate": 1.7586099852153778e-05,
+      "loss": 1.194,
+      "step": 344
+    },
+    {
+      "epoch": 0.25009061254077564,
+      "grad_norm": 1.9828736953069481,
+      "learning_rate": 1.7570769357753682e-05,
+      "loss": 1.122,
+      "step": 345
+    },
+    {
+      "epoch": 0.25009061254077564,
+      "eval_loss": 1.9538915157318115,
+      "eval_runtime": 1129.0599,
+      "eval_samples_per_second": 15.636,
+      "eval_steps_per_second": 0.489,
+      "step": 345
+    },
+    {
+      "epoch": 0.2508155128669808,
+      "grad_norm": 2.4573515309403184,
+      "learning_rate": 1.755539706330883e-05,
+      "loss": 1.1254,
+      "step": 346
+    },
+    {
+      "epoch": 0.25154041319318593,
+      "grad_norm": 2.382471730290679,
+      "learning_rate": 1.753998305369338e-05,
+      "loss": 1.0609,
+      "step": 347
+    },
+    {
+      "epoch": 0.2522653135193911,
+      "grad_norm": 2.0670497345618624,
+      "learning_rate": 1.752452741401179e-05,
+      "loss": 1.11,
+      "step": 348
+    },
+    {
+      "epoch": 0.2529902138455962,
+      "grad_norm": 2.826479622570429,
+      "learning_rate": 1.7509030229598387e-05,
+      "loss": 1.0688,
+      "step": 349
+    },
+    {
+      "epoch": 0.25371511417180137,
+      "grad_norm": 2.4410246210380797,
+      "learning_rate": 1.749349158601686e-05,
+      "loss": 1.0848,
+      "step": 350
+    },
+    {
+      "epoch": 0.2544400144980065,
+      "grad_norm": 2.1500648572615226,
+      "learning_rate": 1.7477911569059808e-05,
+      "loss": 1.1225,
+      "step": 351
+    },
+    {
+      "epoch": 0.25516491482421166,
+      "grad_norm": 2.303703660619773,
+      "learning_rate": 1.746229026474827e-05,
+      "loss": 1.0781,
+      "step": 352
+    },
+    {
+      "epoch": 0.2558898151504168,
+      "grad_norm": 2.6829872427143853,
+      "learning_rate": 1.744662775933123e-05,
+      "loss": 1.0742,
+      "step": 353
+    },
+    {
+      "epoch": 0.25661471547662196,
+      "grad_norm": 2.2217416215597448,
+      "learning_rate": 1.743092413928517e-05,
+      "loss": 1.142,
+      "step": 354
+    },
+    {
+      "epoch": 0.2573396158028271,
+      "grad_norm": 2.6300226414802594,
+      "learning_rate": 1.741517949131356e-05,
+      "loss": 1.0869,
+      "step": 355
+    },
+    {
+      "epoch": 0.25806451612903225,
+      "grad_norm": 2.094472680631636,
+      "learning_rate": 1.7399393902346403e-05,
+      "loss": 1.0797,
+      "step": 356
+    },
+    {
+      "epoch": 0.2587894164552374,
+      "grad_norm": 2.6696223277650097,
+      "learning_rate": 1.738356745953975e-05,
+      "loss": 1.1725,
+      "step": 357
+    },
+    {
+      "epoch": 0.25951431678144254,
+      "grad_norm": 2.560913641389733,
+      "learning_rate": 1.7367700250275196e-05,
+      "loss": 1.0524,
+      "step": 358
+    },
+    {
+      "epoch": 0.2602392171076477,
+      "grad_norm": 2.499190502628081,
+      "learning_rate": 1.735179236215945e-05,
+      "loss": 1.0864,
+      "step": 359
+    },
+    {
+      "epoch": 0.26096411743385284,
+      "grad_norm": 2.1752493153518127,
+      "learning_rate": 1.7335843883023792e-05,
+      "loss": 1.1682,
+      "step": 360
+    },
+    {
+      "epoch": 0.261689017760058,
+      "grad_norm": 2.193115771884426,
+      "learning_rate": 1.7319854900923627e-05,
+      "loss": 1.1538,
+      "step": 361
+    },
+    {
+      "epoch": 0.26241391808626313,
+      "grad_norm": 3.6037027339305228,
+      "learning_rate": 1.7303825504137985e-05,
+      "loss": 1.2803,
+      "step": 362
+    },
+    {
+      "epoch": 0.2631388184124683,
+      "grad_norm": 2.151895072296486,
+      "learning_rate": 1.728775578116903e-05,
+      "loss": 1.027,
+      "step": 363
+    },
+    {
+      "epoch": 0.2638637187386734,
+      "grad_norm": 2.0474799790136315,
+      "learning_rate": 1.7271645820741586e-05,
+      "loss": 1.0166,
+      "step": 364
+    },
+    {
+      "epoch": 0.26458861906487857,
+      "grad_norm": 1.9599613314591962,
+      "learning_rate": 1.7255495711802627e-05,
+      "loss": 1.022,
+      "step": 365
+    },
+    {
+      "epoch": 0.2653135193910837,
+      "grad_norm": 2.401594401551547,
+      "learning_rate": 1.723930554352081e-05,
+      "loss": 1.1917,
+      "step": 366
+    },
+    {
+      "epoch": 0.26603841971728887,
+      "grad_norm": 2.642087840213965,
+      "learning_rate": 1.7223075405285956e-05,
+      "loss": 1.2054,
+      "step": 367
+    },
+    {
+      "epoch": 0.266763320043494,
+      "grad_norm": 2.18604834719014,
+      "learning_rate": 1.7206805386708572e-05,
+      "loss": 1.1475,
+      "step": 368
+    },
+    {
+      "epoch": 0.26748822036969916,
+      "grad_norm": 2.0338413719031507,
+      "learning_rate": 1.7190495577619364e-05,
+      "loss": 1.1028,
+      "step": 369
+    },
+    {
+      "epoch": 0.2682131206959043,
+      "grad_norm": 2.1711132034769394,
+      "learning_rate": 1.717414606806872e-05,
+      "loss": 1.1044,
+      "step": 370
+    },
+    {
+      "epoch": 0.26893802102210945,
+      "grad_norm": 1.930408817282508,
+      "learning_rate": 1.715775694832623e-05,
+      "loss": 1.1185,
+      "step": 371
+    },
+    {
+      "epoch": 0.2696629213483146,
+      "grad_norm": 2.1748455201638643,
+      "learning_rate": 1.7141328308880178e-05,
+      "loss": 1.0437,
+      "step": 372
+    },
+    {
+      "epoch": 0.27038782167451975,
+      "grad_norm": 2.487178508999438,
+      "learning_rate": 1.712486024043705e-05,
+      "loss": 1.1261,
+      "step": 373
+    },
+    {
+      "epoch": 0.2711127220007249,
+      "grad_norm": 2.6620415854141077,
+      "learning_rate": 1.7108352833921022e-05,
+      "loss": 1.1125,
+      "step": 374
+    },
+    {
+      "epoch": 0.27183762232693004,
+      "grad_norm": 2.4779294920405333,
+      "learning_rate": 1.7091806180473472e-05,
+      "loss": 1.1376,
+      "step": 375
+    },
+    {
+      "epoch": 0.2725625226531352,
+      "grad_norm": 2.1914137672000433,
+      "learning_rate": 1.7075220371452465e-05,
+      "loss": 1.2255,
+      "step": 376
+    },
+    {
+      "epoch": 0.27328742297934033,
+      "grad_norm": 2.2406112575154404,
+      "learning_rate": 1.705859549843226e-05,
+      "loss": 1.1511,
+      "step": 377
+    },
+    {
+      "epoch": 0.2740123233055455,
+      "grad_norm": 2.0333226545718914,
+      "learning_rate": 1.7041931653202788e-05,
+      "loss": 1.1184,
+      "step": 378
+    },
+    {
+      "epoch": 0.2747372236317506,
+      "grad_norm": 2.8707122599293875,
+      "learning_rate": 1.702522892776916e-05,
+      "loss": 1.0367,
+      "step": 379
+    },
+    {
+      "epoch": 0.2754621239579558,
+      "grad_norm": 2.242723836798557,
+      "learning_rate": 1.7008487414351163e-05,
+      "loss": 1.1647,
+      "step": 380
+    },
+    {
+      "epoch": 0.2761870242841609,
+      "grad_norm": 3.052876110953187,
+      "learning_rate": 1.6991707205382723e-05,
+      "loss": 1.1574,
+      "step": 381
+    },
+    {
+      "epoch": 0.27691192461036607,
+      "grad_norm": 2.3302064287160955,
+      "learning_rate": 1.6974888393511425e-05,
+      "loss": 1.0833,
+      "step": 382
+    },
+    {
+      "epoch": 0.2776368249365712,
+      "grad_norm": 2.362447133340326,
+      "learning_rate": 1.695803107159799e-05,
+      "loss": 1.0142,
+      "step": 383
+    },
+    {
+      "epoch": 0.27836172526277636,
+      "grad_norm": 1.9262635584689904,
+      "learning_rate": 1.694113533271576e-05,
+      "loss": 0.9974,
+      "step": 384
+    },
+    {
+      "epoch": 0.2790866255889815,
+      "grad_norm": 2.6725693558247943,
+      "learning_rate": 1.6924201270150194e-05,
+      "loss": 1.1584,
+      "step": 385
+    },
+    {
+      "epoch": 0.27981152591518665,
+      "grad_norm": 2.3758190016297855,
+      "learning_rate": 1.6907228977398324e-05,
+      "loss": 1.1007,
+      "step": 386
+    },
+    {
+      "epoch": 0.2805364262413918,
+      "grad_norm": 2.02794227684942,
+      "learning_rate": 1.689021854816829e-05,
+      "loss": 0.9606,
+      "step": 387
+    },
+    {
+      "epoch": 0.28126132656759695,
+      "grad_norm": 2.182217806815362,
+      "learning_rate": 1.6873170076378764e-05,
+      "loss": 1.1012,
+      "step": 388
+    },
+    {
+      "epoch": 0.2819862268938021,
+      "grad_norm": 2.8274640981113714,
+      "learning_rate": 1.685608365615848e-05,
+      "loss": 1.1609,
+      "step": 389
+    },
+    {
+      "epoch": 0.28271112722000724,
+      "grad_norm": 2.0005395310858165,
+      "learning_rate": 1.683895938184569e-05,
+      "loss": 1.1083,
+      "step": 390
+    },
+    {
+      "epoch": 0.2834360275462124,
+      "grad_norm": 2.5469245481992195,
+      "learning_rate": 1.6821797347987637e-05,
+      "loss": 1.1813,
+      "step": 391
+    },
+    {
+      "epoch": 0.28416092787241753,
+      "grad_norm": 2.250275964552977,
+      "learning_rate": 1.680459764934006e-05,
+      "loss": 1.0717,
+      "step": 392
+    },
+    {
+      "epoch": 0.2848858281986227,
+      "grad_norm": 2.1541186966473314,
+      "learning_rate": 1.678736038086664e-05,
+      "loss": 1.0937,
+      "step": 393
+    },
+    {
+      "epoch": 0.28561072852482783,
+      "grad_norm": 2.2526083463595694,
+      "learning_rate": 1.67700856377385e-05,
+      "loss": 1.1692,
+      "step": 394
+    },
+    {
+      "epoch": 0.286335628851033,
+      "grad_norm": 2.360291496431212,
+      "learning_rate": 1.6752773515333665e-05,
+      "loss": 1.0956,
+      "step": 395
+    },
+    {
+      "epoch": 0.2870605291772381,
+      "grad_norm": 2.7484396524462693,
+      "learning_rate": 1.6735424109236538e-05,
+      "loss": 1.0348,
+      "step": 396
+    },
+    {
+      "epoch": 0.28778542950344327,
+      "grad_norm": 1.991944001354783,
+      "learning_rate": 1.671803751523738e-05,
+      "loss": 1.0371,
+      "step": 397
+    },
+    {
+      "epoch": 0.2885103298296484,
+      "grad_norm": 2.1121316874137794,
+      "learning_rate": 1.6700613829331778e-05,
+      "loss": 1.0483,
+      "step": 398
+    },
+    {
+      "epoch": 0.28923523015585356,
+      "grad_norm": 2.9850153218884405,
+      "learning_rate": 1.6683153147720098e-05,
+      "loss": 1.2002,
+      "step": 399
+    },
+    {
+      "epoch": 0.2899601304820587,
+      "grad_norm": 2.2268590912100987,
+      "learning_rate": 1.666565556680698e-05,
+      "loss": 1.1616,
+      "step": 400
+    },
+    {
+      "epoch": 0.29068503080826386,
+      "grad_norm": 2.135948670713283,
+      "learning_rate": 1.664812118320079e-05,
+      "loss": 0.9466,
+      "step": 401
+    },
+    {
+      "epoch": 0.291409931134469,
+      "grad_norm": 2.8208944014300603,
+      "learning_rate": 1.6630550093713096e-05,
+      "loss": 1.1814,
+      "step": 402
+    },
+    {
+      "epoch": 0.29213483146067415,
+      "grad_norm": 2.225373597938535,
+      "learning_rate": 1.661294239535812e-05,
+      "loss": 1.0444,
+      "step": 403
+    },
+    {
+      "epoch": 0.2928597317868793,
+      "grad_norm": 2.253251092388969,
+      "learning_rate": 1.6595298185352216e-05,
+      "loss": 1.092,
+      "step": 404
+    },
+    {
+      "epoch": 0.29358463211308444,
+      "grad_norm": 2.46062887701134,
+      "learning_rate": 1.6577617561113322e-05,
+      "loss": 1.0818,
+      "step": 405
+    },
+    {
+      "epoch": 0.2943095324392896,
+      "grad_norm": 3.129489147993457,
+      "learning_rate": 1.6559900620260435e-05,
+      "loss": 1.1676,
+      "step": 406
+    },
+    {
+      "epoch": 0.29503443276549474,
+      "grad_norm": 2.063039937217649,
+      "learning_rate": 1.654214746061306e-05,
+      "loss": 1.1648,
+      "step": 407
+    },
+    {
+      "epoch": 0.2957593330916999,
+      "grad_norm": 2.4465593098747256,
+      "learning_rate": 1.652435818019068e-05,
+      "loss": 1.0201,
+      "step": 408
+    },
+    {
+      "epoch": 0.29648423341790503,
+      "grad_norm": 2.399324464811924,
+      "learning_rate": 1.6506532877212195e-05,
+      "loss": 1.0413,
+      "step": 409
+    },
+    {
+      "epoch": 0.2972091337441102,
+      "grad_norm": 2.6482971890623546,
+      "learning_rate": 1.6488671650095416e-05,
+      "loss": 1.105,
+      "step": 410
+    },
+    {
+      "epoch": 0.2979340340703153,
+      "grad_norm": 3.034059815258785,
+      "learning_rate": 1.6470774597456482e-05,
+      "loss": 1.066,
+      "step": 411
+    },
+    {
+      "epoch": 0.29865893439652047,
+      "grad_norm": 2.0229310257745725,
+      "learning_rate": 1.6452841818109342e-05,
+      "loss": 1.0126,
+      "step": 412
+    },
+    {
+      "epoch": 0.2993838347227256,
+      "grad_norm": 2.09062623797439,
+      "learning_rate": 1.64348734110652e-05,
+      "loss": 1.0359,
+      "step": 413
+    },
+    {
+      "epoch": 0.30010873504893076,
+      "grad_norm": 2.261696086833709,
+      "learning_rate": 1.6416869475531964e-05,
+      "loss": 1.0387,
+      "step": 414
+    },
+    {
+      "epoch": 0.3008336353751359,
+      "grad_norm": 2.0194176082613686,
+      "learning_rate": 1.639883011091371e-05,
+      "loss": 1.0375,
+      "step": 415
+    },
+    {
+      "epoch": 0.30155853570134106,
+      "grad_norm": 3.0655992063503392,
+      "learning_rate": 1.6380755416810125e-05,
+      "loss": 1.1433,
+      "step": 416
+    },
+    {
+      "epoch": 0.3022834360275462,
+      "grad_norm": 2.200388293782514,
+      "learning_rate": 1.6362645493015955e-05,
+      "loss": 1.0384,
+      "step": 417
+    },
+    {
+      "epoch": 0.30300833635375135,
+      "grad_norm": 2.2340973013167646,
+      "learning_rate": 1.634450043952046e-05,
+      "loss": 1.1233,
+      "step": 418
+    },
+    {
+      "epoch": 0.3037332366799565,
+      "grad_norm": 2.4431864602307334,
+      "learning_rate": 1.6326320356506858e-05,
+      "loss": 1.1127,
+      "step": 419
+    },
+    {
+      "epoch": 0.30445813700616164,
+      "grad_norm": 2.5340829203747997,
+      "learning_rate": 1.6308105344351776e-05,
+      "loss": 1.1336,
+      "step": 420
+    },
+    {
+      "epoch": 0.3051830373323668,
+      "grad_norm": 2.4621749417333865,
+      "learning_rate": 1.6289855503624694e-05,
+      "loss": 1.18,
+      "step": 421
+    },
+    {
+      "epoch": 0.30590793765857194,
+      "grad_norm": 2.435171083757431,
+      "learning_rate": 1.6271570935087385e-05,
+      "loss": 1.1034,
+      "step": 422
+    },
+    {
+      "epoch": 0.3066328379847771,
+      "grad_norm": 2.56776282560331,
+      "learning_rate": 1.6253251739693366e-05,
+      "loss": 1.1227,
+      "step": 423
+    },
+    {
+      "epoch": 0.30735773831098223,
+      "grad_norm": 2.6606550385377172,
+      "learning_rate": 1.6234898018587336e-05,
+      "loss": 1.0606,
+      "step": 424
+    },
+    {
+      "epoch": 0.3080826386371874,
+      "grad_norm": 2.3905958945143584,
+      "learning_rate": 1.621650987310462e-05,
+      "loss": 1.0924,
+      "step": 425
+    },
+    {
+      "epoch": 0.3088075389633925,
+      "grad_norm": 2.16146889842216,
+      "learning_rate": 1.6198087404770602e-05,
+      "loss": 1.0924,
+      "step": 426
+    },
+    {
+      "epoch": 0.30953243928959767,
+      "grad_norm": 2.511397106934819,
+      "learning_rate": 1.617963071530018e-05,
+      "loss": 1.1366,
+      "step": 427
+    },
+    {
+      "epoch": 0.3102573396158028,
+      "grad_norm": 2.394538554344066,
+      "learning_rate": 1.616113990659719e-05,
+      "loss": 1.0009,
+      "step": 428
+    },
+    {
+      "epoch": 0.31098223994200797,
+      "grad_norm": 2.4259004458514006,
+      "learning_rate": 1.6142615080753846e-05,
+      "loss": 1.1315,
+      "step": 429
+    },
+    {
+      "epoch": 0.3117071402682131,
+      "grad_norm": 2.6037915854967726,
+      "learning_rate": 1.6124056340050186e-05,
+      "loss": 1.0235,
+      "step": 430
+    },
+    {
+      "epoch": 0.31243204059441826,
+      "grad_norm": 2.260088913738152,
+      "learning_rate": 1.6105463786953502e-05,
+      "loss": 1.016,
+      "step": 431
+    },
+    {
+      "epoch": 0.3131569409206234,
+      "grad_norm": 1.9542058053758606,
+      "learning_rate": 1.6086837524117755e-05,
+      "loss": 1.0546,
+      "step": 432
+    },
+    {
+      "epoch": 0.31388184124682855,
+      "grad_norm": 2.279040211460153,
+      "learning_rate": 1.6068177654383042e-05,
+      "loss": 1.1363,
+      "step": 433
+    },
+    {
+      "epoch": 0.3146067415730337,
+      "grad_norm": 2.4083685524014933,
+      "learning_rate": 1.6049484280775012e-05,
+      "loss": 1.0617,
+      "step": 434
+    },
+    {
+      "epoch": 0.31533164189923885,
+      "grad_norm": 2.3755540476745844,
+      "learning_rate": 1.6030757506504284e-05,
+      "loss": 1.0368,
+      "step": 435
+    },
+    {
+      "epoch": 0.316056542225444,
+      "grad_norm": 2.1752306039124565,
+      "learning_rate": 1.6011997434965906e-05,
+      "loss": 1.0696,
+      "step": 436
+    },
+    {
+      "epoch": 0.31678144255164914,
+      "grad_norm": 2.0792526403603557,
+      "learning_rate": 1.5993204169738757e-05,
+      "loss": 1.0583,
+      "step": 437
+    },
+    {
+      "epoch": 0.3175063428778543,
+      "grad_norm": 2.1158935057351838,
+      "learning_rate": 1.5974377814584986e-05,
+      "loss": 1.1241,
+      "step": 438
+    },
+    {
+      "epoch": 0.31823124320405943,
+      "grad_norm": 1.7878941702995481,
+      "learning_rate": 1.5955518473449448e-05,
+      "loss": 1.1129,
+      "step": 439
+    },
+    {
+      "epoch": 0.3189561435302646,
+      "grad_norm": 2.1238392656978404,
+      "learning_rate": 1.593662625045912e-05,
+      "loss": 1.0596,
+      "step": 440
+    },
+    {
+      "epoch": 0.3196810438564697,
+      "grad_norm": 2.1813159053603104,
+      "learning_rate": 1.591770124992252e-05,
+      "loss": 1.1104,
+      "step": 441
+    },
+    {
+      "epoch": 0.3204059441826749,
+      "grad_norm": 1.8368381622027763,
+      "learning_rate": 1.589874357632914e-05,
+      "loss": 1.0579,
+      "step": 442
+    },
+    {
+      "epoch": 0.32113084450888,
+      "grad_norm": 2.233702212234041,
+      "learning_rate": 1.587975333434888e-05,
+      "loss": 1.1181,
+      "step": 443
+    },
+    {
+      "epoch": 0.32185574483508517,
+      "grad_norm": 2.112351085565823,
+      "learning_rate": 1.586073062883144e-05,
+      "loss": 1.028,
+      "step": 444
+    },
+    {
+      "epoch": 0.3225806451612903,
+      "grad_norm": 3.4664697517256595,
+      "learning_rate": 1.5841675564805773e-05,
+      "loss": 1.0466,
+      "step": 445
+    },
+    {
+      "epoch": 0.32330554548749546,
+      "grad_norm": 1.9010894055087597,
+      "learning_rate": 1.582258824747949e-05,
+      "loss": 1.0831,
+      "step": 446
+    },
+    {
+      "epoch": 0.3240304458137006,
+      "grad_norm": 2.9306779332036714,
+      "learning_rate": 1.580346878223827e-05,
+      "loss": 1.2466,
+      "step": 447
+    },
+    {
+      "epoch": 0.32475534613990575,
+      "grad_norm": 1.9192009119584101,
+      "learning_rate": 1.5784317274645294e-05,
+      "loss": 1.1088,
+      "step": 448
+    },
+    {
+      "epoch": 0.3254802464661109,
+      "grad_norm": 2.242207905403273,
+      "learning_rate": 1.576513383044066e-05,
+      "loss": 1.1625,
+      "step": 449
+    },
+    {
+      "epoch": 0.32620514679231605,
+      "grad_norm": 2.224398160967933,
+      "learning_rate": 1.5745918555540796e-05,
+      "loss": 1.1488,
+      "step": 450
+    },
+    {
+      "epoch": 0.3269300471185212,
+      "grad_norm": 2.0861425073253503,
+      "learning_rate": 1.572667155603787e-05,
+      "loss": 0.9955,
+      "step": 451
+    },
+    {
+      "epoch": 0.32765494744472634,
+      "grad_norm": 1.92493109968563,
+      "learning_rate": 1.5707392938199208e-05,
+      "loss": 0.9444,
+      "step": 452
+    },
+    {
+      "epoch": 0.3283798477709315,
+      "grad_norm": 1.9283323786352506,
+      "learning_rate": 1.5688082808466714e-05,
+      "loss": 0.9664,
+      "step": 453
+    },
+    {
+      "epoch": 0.32910474809713663,
+      "grad_norm": 2.5055746794884093,
+      "learning_rate": 1.566874127345628e-05,
+      "loss": 1.0542,
+      "step": 454
+    },
+    {
+      "epoch": 0.3298296484233418,
+      "grad_norm": 2.4180849774657784,
+      "learning_rate": 1.5649368439957182e-05,
+      "loss": 1.1004,
+      "step": 455
+    },
+    {
+      "epoch": 0.33055454874954693,
+      "grad_norm": 3.1969300216088574,
+      "learning_rate": 1.562996441493151e-05,
+      "loss": 1.1926,
+      "step": 456
+    },
+    {
+      "epoch": 0.3312794490757521,
+      "grad_norm": 1.8450416397567433,
+      "learning_rate": 1.561052930551357e-05,
+      "loss": 0.9204,
+      "step": 457
+    },
+    {
+      "epoch": 0.3320043494019572,
+      "grad_norm": 2.1352988507915076,
+      "learning_rate": 1.5591063219009296e-05,
+      "loss": 1.1082,
+      "step": 458
+    },
+    {
+      "epoch": 0.33272924972816237,
+      "grad_norm": 2.0068487127258563,
+      "learning_rate": 1.5571566262895637e-05,
+      "loss": 1.0769,
+      "step": 459
+    },
+    {
+      "epoch": 0.3334541500543675,
+      "grad_norm": 2.158036051650137,
+      "learning_rate": 1.5552038544820008e-05,
+      "loss": 1.0965,
+      "step": 460
+    },
+    {
+      "epoch": 0.33417905038057266,
+      "grad_norm": 2.260378271679646,
+      "learning_rate": 1.5532480172599646e-05,
+      "loss": 1.08,
+      "step": 461
+    },
+    {
+      "epoch": 0.3349039507067778,
+      "grad_norm": 2.0768476079893317,
+      "learning_rate": 1.5512891254221046e-05,
+      "loss": 1.0548,
+      "step": 462
+    },
+    {
+      "epoch": 0.33562885103298296,
+      "grad_norm": 2.227359781654928,
+      "learning_rate": 1.549327189783935e-05,
+      "loss": 1.0364,
+      "step": 463
+    },
+    {
+      "epoch": 0.3363537513591881,
+      "grad_norm": 2.008303234643065,
+      "learning_rate": 1.5473622211777763e-05,
+      "loss": 1.0596,
+      "step": 464
+    },
+    {
+      "epoch": 0.33707865168539325,
+      "grad_norm": 2.1053343622780525,
+      "learning_rate": 1.5453942304526945e-05,
+      "loss": 1.0106,
+      "step": 465
+    },
+    {
+      "epoch": 0.3378035520115984,
+      "grad_norm": 2.2528412924953645,
+      "learning_rate": 1.543423228474441e-05,
+      "loss": 1.1579,
+      "step": 466
+    },
+    {
+      "epoch": 0.33852845233780354,
+      "grad_norm": 2.098527027839517,
+      "learning_rate": 1.5414492261253937e-05,
+      "loss": 1.0913,
+      "step": 467
+    },
+    {
+      "epoch": 0.3392533526640087,
+      "grad_norm": 2.3910084145103383,
+      "learning_rate": 1.539472234304496e-05,
+      "loss": 1.0481,
+      "step": 468
+    },
+    {
+      "epoch": 0.33997825299021384,
+      "grad_norm": 2.0843546569961937,
+      "learning_rate": 1.537492263927196e-05,
+      "loss": 1.1351,
+      "step": 469
+    },
+    {
+      "epoch": 0.340703153316419,
+      "grad_norm": 2.0698901517851294,
+      "learning_rate": 1.5355093259253897e-05,
+      "loss": 1.1095,
+      "step": 470
+    },
+    {
+      "epoch": 0.34142805364262413,
+      "grad_norm": 2.2916177664815867,
+      "learning_rate": 1.533523431247355e-05,
+      "loss": 1.094,
+      "step": 471
+    },
+    {
+      "epoch": 0.3421529539688293,
+      "grad_norm": 2.000471695373447,
+      "learning_rate": 1.531534590857696e-05,
+      "loss": 1.0647,
+      "step": 472
+    },
+    {
+      "epoch": 0.3428778542950344,
+      "grad_norm": 2.048745212856722,
+      "learning_rate": 1.529542815737281e-05,
+      "loss": 1.0448,
+      "step": 473
+    },
+    {
+      "epoch": 0.34360275462123957,
+      "grad_norm": 2.104486287385562,
+      "learning_rate": 1.52754811688318e-05,
+      "loss": 1.0525,
+      "step": 474
+    },
+    {
+      "epoch": 0.3443276549474447,
+      "grad_norm": 2.172054502687611,
+      "learning_rate": 1.525550505308608e-05,
+      "loss": 1.0437,
+      "step": 475
+    },
+    {
+      "epoch": 0.34505255527364986,
+      "grad_norm": 2.1612251883461355,
+      "learning_rate": 1.52354999204286e-05,
+      "loss": 1.1142,
+      "step": 476
+    },
+    {
+      "epoch": 0.345777455599855,
+      "grad_norm": 2.5598958893453987,
+      "learning_rate": 1.5215465881312526e-05,
+      "loss": 1.0597,
+      "step": 477
+    },
+    {
+      "epoch": 0.34650235592606016,
+      "grad_norm": 2.1402492854032875,
+      "learning_rate": 1.519540304635062e-05,
+      "loss": 1.0678,
+      "step": 478
+    },
+    {
+      "epoch": 0.3472272562522653,
+      "grad_norm": 2.386151874903598,
+      "learning_rate": 1.5175311526314642e-05,
+      "loss": 1.1297,
+      "step": 479
+    },
+    {
+      "epoch": 0.34795215657847045,
+      "grad_norm": 2.8147411958053734,
+      "learning_rate": 1.5155191432134721e-05,
+      "loss": 1.047,
+      "step": 480
+    },
+    {
+      "epoch": 0.3486770569046756,
+      "grad_norm": 2.185563612843593,
+      "learning_rate": 1.513504287489875e-05,
+      "loss": 1.0674,
+      "step": 481
+    },
+    {
+      "epoch": 0.34940195723088074,
+      "grad_norm": 2.1264088878183705,
+      "learning_rate": 1.5114865965851779e-05,
+      "loss": 1.1097,
+      "step": 482
+    },
+    {
+      "epoch": 0.3501268575570859,
+      "grad_norm": 2.294219589338203,
+      "learning_rate": 1.509466081639539e-05,
+      "loss": 0.9861,
+      "step": 483
+    },
+    {
+      "epoch": 0.35085175788329104,
+      "grad_norm": 2.2845015506312474,
+      "learning_rate": 1.507442753808709e-05,
+      "loss": 1.0646,
+      "step": 484
+    },
+    {
+      "epoch": 0.3515766582094962,
+      "grad_norm": 2.1560916077937606,
+      "learning_rate": 1.5054166242639694e-05,
+      "loss": 1.0056,
+      "step": 485
+    },
+    {
+      "epoch": 0.35230155853570133,
+      "grad_norm": 2.0570915443776654,
+      "learning_rate": 1.5033877041920703e-05,
+      "loss": 1.0926,
+      "step": 486
+    },
+    {
+      "epoch": 0.3530264588619065,
+      "grad_norm": 1.9130291874506817,
+      "learning_rate": 1.5013560047951689e-05,
+      "loss": 0.9783,
+      "step": 487
+    },
+    {
+      "epoch": 0.3537513591881116,
+      "grad_norm": 2.3476451826235625,
+      "learning_rate": 1.499321537290768e-05,
+      "loss": 1.0546,
+      "step": 488
+    },
+    {
+      "epoch": 0.35447625951431677,
+      "grad_norm": 2.3654378069877966,
+      "learning_rate": 1.4972843129116537e-05,
+      "loss": 1.0442,
+      "step": 489
+    },
+    {
+      "epoch": 0.3552011598405219,
+      "grad_norm": 1.9825769936244912,
+      "learning_rate": 1.4952443429058334e-05,
+      "loss": 1.0081,
+      "step": 490
+    },
+    {
+      "epoch": 0.35592606016672707,
+      "grad_norm": 2.2643797127980747,
+      "learning_rate": 1.493201638536474e-05,
+      "loss": 1.0317,
+      "step": 491
+    },
+    {
+      "epoch": 0.3566509604929322,
+      "grad_norm": 2.4516169299014465,
+      "learning_rate": 1.4911562110818392e-05,
+      "loss": 0.9924,
+      "step": 492
+    },
+    {
+      "epoch": 0.35737586081913736,
+      "grad_norm": 2.446561602895059,
+      "learning_rate": 1.4891080718352275e-05,
+      "loss": 1.0407,
+      "step": 493
+    },
+    {
+      "epoch": 0.3581007611453425,
+      "grad_norm": 2.0486524427646144,
+      "learning_rate": 1.4870572321049102e-05,
+      "loss": 1.0797,
+      "step": 494
+    },
+    {
+      "epoch": 0.35882566147154765,
+      "grad_norm": 2.485346536887422,
+      "learning_rate": 1.4850037032140683e-05,
+      "loss": 0.9983,
+      "step": 495
+    },
+    {
+      "epoch": 0.3595505617977528,
+      "grad_norm": 3.0439426290182094,
+      "learning_rate": 1.4829474965007303e-05,
+      "loss": 1.152,
+      "step": 496
+    },
+    {
+      "epoch": 0.36027546212395795,
+      "grad_norm": 2.540523235848105,
+      "learning_rate": 1.4808886233177096e-05,
+      "loss": 1.2146,
+      "step": 497
+    },
+    {
+      "epoch": 0.3610003624501631,
+      "grad_norm": 2.5194385485382207,
+      "learning_rate": 1.4788270950325424e-05,
+      "loss": 1.0989,
+      "step": 498
+    },
+    {
+      "epoch": 0.36172526277636824,
+      "grad_norm": 2.3200096809507693,
+      "learning_rate": 1.4767629230274236e-05,
+      "loss": 1.2226,
+      "step": 499
+    },
+    {
+      "epoch": 0.3624501631025734,
+      "grad_norm": 2.0411294097073363,
+      "learning_rate": 1.474696118699145e-05,
+      "loss": 1.0421,
+      "step": 500
+    },
+    {
+      "epoch": 0.36317506342877853,
+      "grad_norm": 2.3819141883855597,
+      "learning_rate": 1.4726266934590321e-05,
+      "loss": 1.0162,
+      "step": 501
+    },
+    {
+      "epoch": 0.3638999637549837,
+      "grad_norm": 1.9177330075859071,
+      "learning_rate": 1.4705546587328811e-05,
+      "loss": 1.0171,
+      "step": 502
+    },
+    {
+      "epoch": 0.3646248640811888,
+      "grad_norm": 2.06883784081889,
+      "learning_rate": 1.4684800259608958e-05,
+      "loss": 1.0567,
+      "step": 503
+    },
+    {
+      "epoch": 0.365349764407394,
+      "grad_norm": 2.796404840199928,
+      "learning_rate": 1.4664028065976245e-05,
+      "loss": 1.0567,
+      "step": 504
+    },
+    {
+      "epoch": 0.3660746647335991,
+      "grad_norm": 2.0344928788193433,
+      "learning_rate": 1.4643230121118965e-05,
+      "loss": 1.1192,
+      "step": 505
+    },
+    {
+      "epoch": 0.36679956505980427,
+      "grad_norm": 1.931575463278115,
+      "learning_rate": 1.4622406539867593e-05,
+      "loss": 1.0491,
+      "step": 506
+    },
+    {
+      "epoch": 0.3675244653860094,
+      "grad_norm": 2.5445474863300106,
+      "learning_rate": 1.460155743719414e-05,
+      "loss": 1.0213,
+      "step": 507
+    },
+    {
+      "epoch": 0.36824936571221456,
+      "grad_norm": 2.0670836939977564,
+      "learning_rate": 1.4580682928211545e-05,
+      "loss": 0.92,
+      "step": 508
+    },
+    {
+      "epoch": 0.3689742660384197,
+      "grad_norm": 2.575085198850707,
+      "learning_rate": 1.4559783128173e-05,
+      "loss": 1.0829,
+      "step": 509
+    },
+    {
+      "epoch": 0.36969916636462485,
+      "grad_norm": 2.7113350290504723,
+      "learning_rate": 1.4538858152471349e-05,
+      "loss": 1.1535,
+      "step": 510
+    },
+    {
+      "epoch": 0.37042406669083,
+      "grad_norm": 2.312117149467026,
+      "learning_rate": 1.4517908116638433e-05,
+      "loss": 1.0897,
+      "step": 511
+    },
+    {
+      "epoch": 0.37114896701703515,
+      "grad_norm": 2.3702995154115656,
+      "learning_rate": 1.4496933136344453e-05,
+      "loss": 1.0635,
+      "step": 512
+    },
+    {
+      "epoch": 0.3718738673432403,
+      "grad_norm": 2.4501757498607537,
+      "learning_rate": 1.4475933327397342e-05,
+      "loss": 1.0603,
+      "step": 513
+    },
+    {
+      "epoch": 0.37259876766944544,
+      "grad_norm": 2.023783253482957,
+      "learning_rate": 1.445490880574211e-05,
+      "loss": 1.0789,
+      "step": 514
+    },
+    {
+      "epoch": 0.3733236679956506,
+      "grad_norm": 3.2864848049242648,
+      "learning_rate": 1.4433859687460216e-05,
+      "loss": 1.0952,
+      "step": 515
+    },
+    {
+      "epoch": 0.37404856832185573,
+      "grad_norm": 2.4191595705564737,
+      "learning_rate": 1.4412786088768923e-05,
+      "loss": 1.1147,
+      "step": 516
+    },
+    {
+      "epoch": 0.3747734686480609,
+      "grad_norm": 2.4447277431874825,
+      "learning_rate": 1.439168812602065e-05,
+      "loss": 1.0393,
+      "step": 517
+    },
+    {
+      "epoch": 0.37549836897426603,
+      "grad_norm": 1.9206969249262023,
+      "learning_rate": 1.437056591570235e-05,
+      "loss": 1.0022,
+      "step": 518
+    },
+    {
+      "epoch": 0.3762232693004712,
+      "grad_norm": 2.359427663223057,
+      "learning_rate": 1.4349419574434838e-05,
+      "loss": 1.093,
+      "step": 519
+    },
+    {
+      "epoch": 0.3769481696266763,
+      "grad_norm": 2.3010718359012103,
+      "learning_rate": 1.4328249218972168e-05,
+      "loss": 1.0408,
+      "step": 520
+    },
+    {
+      "epoch": 0.37767306995288147,
+      "grad_norm": 2.2594128382319596,
+      "learning_rate": 1.4307054966200984e-05,
+      "loss": 0.9412,
+      "step": 521
+    },
+    {
+      "epoch": 0.3783979702790866,
+      "grad_norm": 2.1054269973126747,
+      "learning_rate": 1.4285836933139865e-05,
+      "loss": 1.0254,
+      "step": 522
+    },
+    {
+      "epoch": 0.37912287060529176,
+      "grad_norm": 2.168937643340505,
+      "learning_rate": 1.42645952369387e-05,
+      "loss": 1.0537,
+      "step": 523
+    },
+    {
+      "epoch": 0.3798477709314969,
+      "grad_norm": 2.061073976619602,
+      "learning_rate": 1.4243329994878017e-05,
+      "loss": 1.1168,
+      "step": 524
+    },
+    {
+      "epoch": 0.38057267125770206,
+      "grad_norm": 2.299186158706926,
+      "learning_rate": 1.4222041324368347e-05,
+      "loss": 1.1138,
+      "step": 525
+    },
+    {
+      "epoch": 0.3812975715839072,
+      "grad_norm": 2.251869960906443,
+      "learning_rate": 1.4200729342949577e-05,
+      "loss": 0.9248,
+      "step": 526
+    },
+    {
+      "epoch": 0.38202247191011235,
+      "grad_norm": 2.2506167583969856,
+      "learning_rate": 1.4179394168290309e-05,
+      "loss": 1.0757,
+      "step": 527
+    },
+    {
+      "epoch": 0.3827473722363175,
+      "grad_norm": 2.916531852779516,
+      "learning_rate": 1.4158035918187182e-05,
+      "loss": 1.1818,
+      "step": 528
+    },
+    {
+      "epoch": 0.38347227256252264,
+      "grad_norm": 2.3787214433218806,
+      "learning_rate": 1.4136654710564251e-05,
+      "loss": 1.0336,
+      "step": 529
+    },
+    {
+      "epoch": 0.3841971728887278,
+      "grad_norm": 2.4765072845885068,
+      "learning_rate": 1.4115250663472326e-05,
+      "loss": 1.0506,
+      "step": 530
+    },
+    {
+      "epoch": 0.38492207321493294,
+      "grad_norm": 2.407842302823844,
+      "learning_rate": 1.4093823895088315e-05,
+      "loss": 1.1,
+      "step": 531
+    },
+    {
+      "epoch": 0.3856469735411381,
+      "grad_norm": 2.3150861515619487,
+      "learning_rate": 1.4072374523714577e-05,
+      "loss": 1.0437,
+      "step": 532
+    },
+    {
+      "epoch": 0.38637187386734323,
+      "grad_norm": 2.0825864756582506,
+      "learning_rate": 1.4050902667778272e-05,
+      "loss": 1.1277,
+      "step": 533
+    },
+    {
+      "epoch": 0.3870967741935484,
+      "grad_norm": 2.243168272872563,
+      "learning_rate": 1.402940844583069e-05,
+      "loss": 1.0506,
+      "step": 534
+    },
+    {
+      "epoch": 0.3878216745197535,
+      "grad_norm": 2.185661024343008,
+      "learning_rate": 1.4007891976546627e-05,
+      "loss": 1.088,
+      "step": 535
+    },
+    {
+      "epoch": 0.38854657484595867,
+      "grad_norm": 2.2131084797482186,
+      "learning_rate": 1.3986353378723696e-05,
+      "loss": 1.0654,
+      "step": 536
+    },
+    {
+      "epoch": 0.3892714751721638,
+      "grad_norm": 2.4512297593144177,
+      "learning_rate": 1.3964792771281702e-05,
+      "loss": 1.2112,
+      "step": 537
+    },
+    {
+      "epoch": 0.38999637549836896,
+      "grad_norm": 1.9649805513289704,
+      "learning_rate": 1.394321027326195e-05,
+      "loss": 1.0487,
+      "step": 538
+    },
+    {
+      "epoch": 0.3907212758245741,
+      "grad_norm": 2.0806905976421985,
+      "learning_rate": 1.392160600382663e-05,
+      "loss": 1.1033,
+      "step": 539
+    },
+    {
+      "epoch": 0.39144617615077926,
+      "grad_norm": 2.117182296397933,
+      "learning_rate": 1.3899980082258122e-05,
+      "loss": 1.0957,
+      "step": 540
+    },
+    {
+      "epoch": 0.3921710764769844,
+      "grad_norm": 2.4101205011369777,
+      "learning_rate": 1.3878332627958365e-05,
+      "loss": 1.1324,
+      "step": 541
+    },
+    {
+      "epoch": 0.39289597680318955,
+      "grad_norm": 2.0554352528605255,
+      "learning_rate": 1.385666376044817e-05,
+      "loss": 1.0756,
+      "step": 542
+    },
+    {
+      "epoch": 0.3936208771293947,
+      "grad_norm": 2.8127997073836806,
+      "learning_rate": 1.3834973599366588e-05,
+      "loss": 1.203,
+      "step": 543
+    },
+    {
+      "epoch": 0.39434577745559984,
+      "grad_norm": 2.249229842195169,
+      "learning_rate": 1.3813262264470235e-05,
+      "loss": 1.0307,
+      "step": 544
+    },
+    {
+      "epoch": 0.395070677781805,
+      "grad_norm": 2.371447175626824,
+      "learning_rate": 1.3791529875632628e-05,
+      "loss": 1.1348,
+      "step": 545
+    },
+    {
+      "epoch": 0.39579557810801014,
+      "grad_norm": 2.2035574911903697,
+      "learning_rate": 1.3769776552843532e-05,
+      "loss": 1.1021,
+      "step": 546
+    },
+    {
+      "epoch": 0.3965204784342153,
+      "grad_norm": 2.749098638445283,
+      "learning_rate": 1.3748002416208292e-05,
+      "loss": 1.1599,
+      "step": 547
+    },
+    {
+      "epoch": 0.39724537876042043,
+      "grad_norm": 1.9103185758921004,
+      "learning_rate": 1.372620758594717e-05,
+      "loss": 0.9921,
+      "step": 548
+    },
+    {
+      "epoch": 0.3979702790866256,
+      "grad_norm": 2.1532622643710946,
+      "learning_rate": 1.3704392182394686e-05,
+      "loss": 1.0174,
+      "step": 549
+    },
+    {
+      "epoch": 0.3986951794128307,
+      "grad_norm": 2.0770483467925875,
+      "learning_rate": 1.3682556325998947e-05,
+      "loss": 1.1642,
+      "step": 550
+    },
+    {
+      "epoch": 0.39942007973903587,
+      "grad_norm": 2.0510030802176002,
+      "learning_rate": 1.3660700137320986e-05,
+      "loss": 1.0736,
+      "step": 551
+    },
+    {
+      "epoch": 0.400144980065241,
+      "grad_norm": 2.4143942712642583,
+      "learning_rate": 1.3638823737034095e-05,
+      "loss": 1.0679,
+      "step": 552
+    },
+    {
+      "epoch": 0.40086988039144617,
+      "grad_norm": 2.087827823927243,
+      "learning_rate": 1.3616927245923157e-05,
+      "loss": 1.077,
+      "step": 553
+    },
+    {
+      "epoch": 0.4015947807176513,
+      "grad_norm": 2.155634016832207,
+      "learning_rate": 1.359501078488399e-05,
+      "loss": 0.9344,
+      "step": 554
+    },
+    {
+      "epoch": 0.40231968104385646,
+      "grad_norm": 2.9410176032472104,
+      "learning_rate": 1.357307447492266e-05,
+      "loss": 0.9876,
+      "step": 555
+    },
+    {
+      "epoch": 0.4030445813700616,
+      "grad_norm": 2.3785327314059965,
+      "learning_rate": 1.3551118437154833e-05,
+      "loss": 0.8992,
+      "step": 556
+    },
+    {
+      "epoch": 0.40376948169626675,
+      "grad_norm": 2.206200464844771,
+      "learning_rate": 1.3529142792805087e-05,
+      "loss": 1.0472,
+      "step": 557
+    },
+    {
+      "epoch": 0.4044943820224719,
+      "grad_norm": 2.4336430655807058,
+      "learning_rate": 1.3507147663206257e-05,
+      "loss": 1.147,
+      "step": 558
+    },
+    {
+      "epoch": 0.40521928234867705,
+      "grad_norm": 2.0423284010474627,
+      "learning_rate": 1.348513316979877e-05,
+      "loss": 1.0467,
+      "step": 559
+    },
+    {
+      "epoch": 0.4059441826748822,
+      "grad_norm": 2.6261549064320215,
+      "learning_rate": 1.346309943412995e-05,
+      "loss": 0.954,
+      "step": 560
+    },
+    {
+      "epoch": 0.40666908300108734,
+      "grad_norm": 2.8231886882036386,
+      "learning_rate": 1.3441046577853371e-05,
+      "loss": 1.088,
+      "step": 561
+    },
+    {
+      "epoch": 0.4073939833272925,
+      "grad_norm": 2.5545065191928766,
+      "learning_rate": 1.3418974722728177e-05,
+      "loss": 1.1541,
+      "step": 562
+    },
+    {
+      "epoch": 0.40811888365349763,
+      "grad_norm": 2.714166874930808,
+      "learning_rate": 1.3396883990618404e-05,
+      "loss": 1.1053,
+      "step": 563
+    },
+    {
+      "epoch": 0.4088437839797028,
+      "grad_norm": 1.9784578326236464,
+      "learning_rate": 1.3374774503492316e-05,
+      "loss": 1.0065,
+      "step": 564
+    },
+    {
+      "epoch": 0.4095686843059079,
+      "grad_norm": 2.1486595608204357,
+      "learning_rate": 1.3352646383421728e-05,
+      "loss": 1.0184,
+      "step": 565
+    },
+    {
+      "epoch": 0.4102935846321131,
+      "grad_norm": 2.5242559447298483,
+      "learning_rate": 1.3330499752581331e-05,
+      "loss": 1.0812,
+      "step": 566
+    },
+    {
+      "epoch": 0.4110184849583182,
+      "grad_norm": 1.9089213008734895,
+      "learning_rate": 1.3308334733248019e-05,
+      "loss": 0.9932,
+      "step": 567
+    },
+    {
+      "epoch": 0.41174338528452337,
+      "grad_norm": 1.890310480638506,
+      "learning_rate": 1.3286151447800211e-05,
+      "loss": 0.9832,
+      "step": 568
+    },
+    {
+      "epoch": 0.4124682856107285,
+      "grad_norm": 2.038892051314124,
+      "learning_rate": 1.3263950018717184e-05,
+      "loss": 0.9662,
+      "step": 569
+    },
+    {
+      "epoch": 0.41319318593693366,
+      "grad_norm": 2.241417479572917,
+      "learning_rate": 1.3241730568578383e-05,
+      "loss": 1.0493,
+      "step": 570
+    },
+    {
+      "epoch": 0.4139180862631388,
+      "grad_norm": 2.260815460922618,
+      "learning_rate": 1.3219493220062756e-05,
+      "loss": 1.0101,
+      "step": 571
+    },
+    {
+      "epoch": 0.41464298658934395,
+      "grad_norm": 2.456266733687105,
+      "learning_rate": 1.3197238095948066e-05,
+      "loss": 1.0541,
+      "step": 572
+    },
+    {
+      "epoch": 0.4153678869155491,
+      "grad_norm": 2.5507405445700377,
+      "learning_rate": 1.3174965319110231e-05,
+      "loss": 1.0279,
+      "step": 573
+    },
+    {
+      "epoch": 0.41609278724175425,
+      "grad_norm": 2.847963262037451,
+      "learning_rate": 1.3152675012522629e-05,
+      "loss": 1.108,
+      "step": 574
+    },
+    {
+      "epoch": 0.4168176875679594,
+      "grad_norm": 2.1767458920508784,
+      "learning_rate": 1.313036729925541e-05,
+      "loss": 1.0479,
+      "step": 575
+    },
+    {
+      "epoch": 0.41754258789416454,
+      "grad_norm": 3.6818743459800234,
+      "learning_rate": 1.3108042302474858e-05,
+      "loss": 1.1633,
+      "step": 576
+    },
+    {
+      "epoch": 0.4182674882203697,
+      "grad_norm": 2.4735330534741595,
+      "learning_rate": 1.3085700145442653e-05,
+      "loss": 0.9817,
+      "step": 577
+    },
+    {
+      "epoch": 0.41899238854657483,
+      "grad_norm": 2.099259381744443,
+      "learning_rate": 1.3063340951515246e-05,
+      "loss": 0.9347,
+      "step": 578
+    },
+    {
+      "epoch": 0.41971728887278,
+      "grad_norm": 2.3256524631392916,
+      "learning_rate": 1.3040964844143144e-05,
+      "loss": 1.0972,
+      "step": 579
+    },
+    {
+      "epoch": 0.42044218919898513,
+      "grad_norm": 2.5886329967045385,
+      "learning_rate": 1.301857194687023e-05,
+      "loss": 1.1258,
+      "step": 580
+    },
+    {
+      "epoch": 0.4211670895251903,
+      "grad_norm": 2.0842353427178097,
+      "learning_rate": 1.2996162383333097e-05,
+      "loss": 1.0067,
+      "step": 581
+    },
+    {
+      "epoch": 0.4218919898513954,
+      "grad_norm": 2.0384216517802662,
+      "learning_rate": 1.2973736277260351e-05,
+      "loss": 1.0734,
+      "step": 582
+    },
+    {
+      "epoch": 0.42261689017760057,
+      "grad_norm": 2.4453213613902784,
+      "learning_rate": 1.295129375247194e-05,
+      "loss": 1.1793,
+      "step": 583
+    },
+    {
+      "epoch": 0.4233417905038057,
+      "grad_norm": 1.9025172070678824,
+      "learning_rate": 1.292883493287846e-05,
+      "loss": 1.083,
+      "step": 584
+    },
+    {
+      "epoch": 0.42406669083001086,
+      "grad_norm": 2.6363997339281595,
+      "learning_rate": 1.290635994248047e-05,
+      "loss": 1.0293,
+      "step": 585
+    },
+    {
+      "epoch": 0.424791591156216,
+      "grad_norm": 2.311748807183176,
+      "learning_rate": 1.2883868905367818e-05,
+      "loss": 1.1425,
+      "step": 586
+    },
+    {
+      "epoch": 0.42551649148242116,
+      "grad_norm": 2.357170974552937,
+      "learning_rate": 1.2861361945718947e-05,
+      "loss": 1.2254,
+      "step": 587
+    },
+    {
+      "epoch": 0.4262413918086263,
+      "grad_norm": 2.176179466256725,
+      "learning_rate": 1.2838839187800218e-05,
+      "loss": 1.0198,
+      "step": 588
+    },
+    {
+      "epoch": 0.42696629213483145,
+      "grad_norm": 2.9614059903538967,
+      "learning_rate": 1.2816300755965202e-05,
+      "loss": 1.1044,
+      "step": 589
+    },
+    {
+      "epoch": 0.4276911924610366,
+      "grad_norm": 2.089967789540345,
+      "learning_rate": 1.2793746774654034e-05,
+      "loss": 0.9743,
+      "step": 590
+    },
+    {
+      "epoch": 0.42841609278724174,
+      "grad_norm": 2.0512215889422114,
+      "learning_rate": 1.277117736839268e-05,
+      "loss": 1.0381,
+      "step": 591
+    },
+    {
+      "epoch": 0.4291409931134469,
+      "grad_norm": 3.3593984924680065,
+      "learning_rate": 1.2748592661792278e-05,
+      "loss": 1.0852,
+      "step": 592
+    },
+    {
+      "epoch": 0.42986589343965204,
+      "grad_norm": 1.8379649951016293,
+      "learning_rate": 1.2725992779548451e-05,
+      "loss": 1.0407,
+      "step": 593
+    },
+    {
+      "epoch": 0.4305907937658572,
+      "grad_norm": 2.0580648499060614,
+      "learning_rate": 1.2703377846440591e-05,
+      "loss": 0.9797,
+      "step": 594
+    },
+    {
+      "epoch": 0.43131569409206233,
+      "grad_norm": 2.547624173063941,
+      "learning_rate": 1.2680747987331215e-05,
+      "loss": 1.0976,
+      "step": 595
+    },
+    {
+      "epoch": 0.4320405944182675,
+      "grad_norm": 2.358740461402887,
+      "learning_rate": 1.2658103327165226e-05,
+      "loss": 0.9832,
+      "step": 596
+    },
+    {
+      "epoch": 0.4327654947444726,
+      "grad_norm": 2.2958356138544223,
+      "learning_rate": 1.2635443990969266e-05,
+      "loss": 1.0707,
+      "step": 597
+    },
+    {
+      "epoch": 0.43349039507067777,
+      "grad_norm": 2.2673074349057036,
+      "learning_rate": 1.2612770103850998e-05,
+      "loss": 1.155,
+      "step": 598
+    },
+    {
+      "epoch": 0.4342152953968829,
+      "grad_norm": 2.191734084847357,
+      "learning_rate": 1.2590081790998414e-05,
+      "loss": 1.1389,
+      "step": 599
+    },
+    {
+      "epoch": 0.43494019572308806,
+      "grad_norm": 2.2857227047676365,
+      "learning_rate": 1.2567379177679178e-05,
+      "loss": 0.9963,
+      "step": 600
+    },
+    {
+      "epoch": 0.4356650960492932,
+      "grad_norm": 2.1784697059963047,
+      "learning_rate": 1.254466238923989e-05,
+      "loss": 1.0586,
+      "step": 601
+    },
+    {
+      "epoch": 0.43638999637549836,
+      "grad_norm": 2.521226757264337,
+      "learning_rate": 1.2521931551105427e-05,
+      "loss": 1.1325,
+      "step": 602
+    },
+    {
+      "epoch": 0.4371148967017035,
+      "grad_norm": 2.061977865722029,
+      "learning_rate": 1.2499186788778223e-05,
+      "loss": 0.9365,
+      "step": 603
+    },
+    {
+      "epoch": 0.43783979702790865,
+      "grad_norm": 2.447645929802628,
+      "learning_rate": 1.2476428227837607e-05,
+      "loss": 1.0145,
+      "step": 604
+    },
+    {
+      "epoch": 0.4385646973541138,
+      "grad_norm": 2.267005489506538,
+      "learning_rate": 1.2453655993939088e-05,
+      "loss": 1.0399,
+      "step": 605
+    },
+    {
+      "epoch": 0.43928959768031894,
+      "grad_norm": 2.084407985939619,
+      "learning_rate": 1.2430870212813661e-05,
+      "loss": 0.9565,
+      "step": 606
+    },
+    {
+      "epoch": 0.4400144980065241,
+      "grad_norm": 2.6037410215898458,
+      "learning_rate": 1.2408071010267125e-05,
+      "loss": 1.0644,
+      "step": 607
+    },
+    {
+      "epoch": 0.44073939833272924,
+      "grad_norm": 2.2081346806030235,
+      "learning_rate": 1.2385258512179375e-05,
+      "loss": 0.9149,
+      "step": 608
+    },
+    {
+      "epoch": 0.4414642986589344,
+      "grad_norm": 2.3544684711758173,
+      "learning_rate": 1.2362432844503725e-05,
+      "loss": 1.0141,
+      "step": 609
+    },
+    {
+      "epoch": 0.44218919898513953,
+      "grad_norm": 2.0791622079935914,
+      "learning_rate": 1.233959413326619e-05,
+      "loss": 1.0256,
+      "step": 610
+    },
+    {
+      "epoch": 0.4429140993113447,
+      "grad_norm": 2.968477393667513,
+      "learning_rate": 1.2316742504564811e-05,
+      "loss": 1.1866,
+      "step": 611
+    },
+    {
+      "epoch": 0.4436389996375498,
+      "grad_norm": 2.076859528680039,
+      "learning_rate": 1.2293878084568944e-05,
+      "loss": 0.9385,
+      "step": 612
+    },
+    {
+      "epoch": 0.44436389996375497,
+      "grad_norm": 2.499714197915474,
+      "learning_rate": 1.2271000999518563e-05,
+      "loss": 1.1666,
+      "step": 613
+    },
+    {
+      "epoch": 0.4450888002899601,
+      "grad_norm": 2.003190354705559,
+      "learning_rate": 1.2248111375723585e-05,
+      "loss": 1.0057,
+      "step": 614
+    },
+    {
+      "epoch": 0.44581370061616526,
+      "grad_norm": 2.037354395792757,
+      "learning_rate": 1.2225209339563144e-05,
+      "loss": 0.9492,
+      "step": 615
+    },
+    {
+      "epoch": 0.4465386009423704,
+      "grad_norm": 2.586450689063457,
+      "learning_rate": 1.2202295017484911e-05,
+      "loss": 1.0752,
+      "step": 616
+    },
+    {
+      "epoch": 0.44726350126857556,
+      "grad_norm": 1.903194057594901,
+      "learning_rate": 1.2179368536004382e-05,
+      "loss": 0.9995,
+      "step": 617
+    },
+    {
+      "epoch": 0.4479884015947807,
+      "grad_norm": 2.3340177701835523,
+      "learning_rate": 1.2156430021704196e-05,
+      "loss": 1.0832,
+      "step": 618
+    },
+    {
+      "epoch": 0.44871330192098585,
+      "grad_norm": 2.181064156987608,
+      "learning_rate": 1.2133479601233431e-05,
+      "loss": 0.9726,
+      "step": 619
+    },
+    {
+      "epoch": 0.449438202247191,
+      "grad_norm": 2.1594915956868097,
+      "learning_rate": 1.2110517401306896e-05,
+      "loss": 0.9393,
+      "step": 620
+    },
+    {
+      "epoch": 0.45016310257339615,
+      "grad_norm": 2.5184901765324934,
+      "learning_rate": 1.2087543548704435e-05,
+      "loss": 1.1569,
+      "step": 621
+    },
+    {
+      "epoch": 0.4508880028996013,
+      "grad_norm": 2.2361203028149585,
+      "learning_rate": 1.2064558170270236e-05,
+      "loss": 1.0704,
+      "step": 622
+    },
+    {
+      "epoch": 0.45161290322580644,
+      "grad_norm": 1.9402238173723454,
+      "learning_rate": 1.2041561392912118e-05,
+      "loss": 1.0127,
+      "step": 623
+    },
+    {
+      "epoch": 0.4523378035520116,
+      "grad_norm": 2.6102129250515724,
+      "learning_rate": 1.2018553343600841e-05,
+      "loss": 1.1364,
+      "step": 624
+    },
+    {
+      "epoch": 0.45306270387821673,
+      "grad_norm": 2.0267045043502883,
+      "learning_rate": 1.1995534149369397e-05,
+      "loss": 1.0408,
+      "step": 625
+    },
+    {
+      "epoch": 0.4537876042044219,
+      "grad_norm": 2.0973318190957353,
+      "learning_rate": 1.1972503937312315e-05,
+      "loss": 0.994,
+      "step": 626
+    },
+    {
+      "epoch": 0.454512504530627,
+      "grad_norm": 1.9897533182088447,
+      "learning_rate": 1.194946283458495e-05,
+      "loss": 0.9902,
+      "step": 627
+    },
+    {
+      "epoch": 0.4552374048568322,
+      "grad_norm": 2.624450240222456,
+      "learning_rate": 1.1926410968402791e-05,
+      "loss": 1.1201,
+      "step": 628
+    },
+    {
+      "epoch": 0.4559623051830373,
+      "grad_norm": 1.9043526191630893,
+      "learning_rate": 1.1903348466040758e-05,
+      "loss": 1.0846,
+      "step": 629
+    },
+    {
+      "epoch": 0.45668720550924247,
+      "grad_norm": 1.9419173600857174,
+      "learning_rate": 1.1880275454832493e-05,
+      "loss": 1.0469,
+      "step": 630
+    },
+    {
+      "epoch": 0.4574121058354476,
+      "grad_norm": 2.2783826877349673,
+      "learning_rate": 1.1857192062169656e-05,
+      "loss": 0.9387,
+      "step": 631
+    },
+    {
+      "epoch": 0.45813700616165276,
+      "grad_norm": 2.093160188755417,
+      "learning_rate": 1.183409841550123e-05,
+      "loss": 1.1626,
+      "step": 632
+    },
+    {
+      "epoch": 0.4588619064878579,
+      "grad_norm": 2.203652115607489,
+      "learning_rate": 1.1810994642332808e-05,
+      "loss": 1.0042,
+      "step": 633
+    },
+    {
+      "epoch": 0.45958680681406305,
+      "grad_norm": 2.0594672175740008,
+      "learning_rate": 1.1787880870225905e-05,
+      "loss": 1.0613,
+      "step": 634
+    },
+    {
+      "epoch": 0.4603117071402682,
+      "grad_norm": 2.1686023194883055,
+      "learning_rate": 1.1764757226797233e-05,
+      "loss": 1.0283,
+      "step": 635
+    },
+    {
+      "epoch": 0.46103660746647335,
+      "grad_norm": 2.3713976995681327,
+      "learning_rate": 1.1741623839718002e-05,
+      "loss": 1.0412,
+      "step": 636
+    },
+    {
+      "epoch": 0.4617615077926785,
+      "grad_norm": 2.3932119999968005,
+      "learning_rate": 1.1718480836713228e-05,
+      "loss": 1.0953,
+      "step": 637
+    },
+    {
+      "epoch": 0.46248640811888364,
+      "grad_norm": 2.411654231766739,
+      "learning_rate": 1.169532834556102e-05,
+      "loss": 1.0146,
+      "step": 638
+    },
+    {
+      "epoch": 0.4632113084450888,
+      "grad_norm": 2.481106610277318,
+      "learning_rate": 1.167216649409187e-05,
+      "loss": 0.9666,
+      "step": 639
+    },
+    {
+      "epoch": 0.46393620877129393,
+      "grad_norm": 2.466302214254987,
+      "learning_rate": 1.164899541018794e-05,
+      "loss": 1.1808,
+      "step": 640
+    },
+    {
+      "epoch": 0.4646611090974991,
+      "grad_norm": 2.086736565800973,
+      "learning_rate": 1.1625815221782386e-05,
+      "loss": 1.066,
+      "step": 641
+    },
+    {
+      "epoch": 0.4653860094237042,
+      "grad_norm": 2.1129903671304344,
+      "learning_rate": 1.1602626056858617e-05,
+      "loss": 1.0851,
+      "step": 642
+    },
+    {
+      "epoch": 0.4661109097499094,
+      "grad_norm": 2.012359347751327,
+      "learning_rate": 1.1579428043449612e-05,
+      "loss": 0.9234,
+      "step": 643
+    },
+    {
+      "epoch": 0.4668358100761145,
+      "grad_norm": 1.731209443837663,
+      "learning_rate": 1.1556221309637204e-05,
+      "loss": 0.9405,
+      "step": 644
+    },
+    {
+      "epoch": 0.46756071040231967,
+      "grad_norm": 2.3738797051074587,
+      "learning_rate": 1.1533005983551361e-05,
+      "loss": 1.0442,
+      "step": 645
+    },
+    {
+      "epoch": 0.4682856107285248,
+      "grad_norm": 2.5124526996392333,
+      "learning_rate": 1.1509782193369514e-05,
+      "loss": 1.0362,
+      "step": 646
+    },
+    {
+      "epoch": 0.46901051105472996,
+      "grad_norm": 2.2245947511275403,
+      "learning_rate": 1.1486550067315803e-05,
+      "loss": 1.0992,
+      "step": 647
+    },
+    {
+      "epoch": 0.4697354113809351,
+      "grad_norm": 2.3135678458341484,
+      "learning_rate": 1.146330973366041e-05,
+      "loss": 1.0667,
+      "step": 648
+    },
+    {
+      "epoch": 0.47046031170714026,
+      "grad_norm": 1.8464379606767438,
+      "learning_rate": 1.1440061320718821e-05,
+      "loss": 1.094,
+      "step": 649
+    },
+    {
+      "epoch": 0.4711852120333454,
+      "grad_norm": 2.0764785227825864,
+      "learning_rate": 1.1416804956851138e-05,
+      "loss": 0.8967,
+      "step": 650
+    },
+    {
+      "epoch": 0.47191011235955055,
+      "grad_norm": 2.434020067387766,
+      "learning_rate": 1.1393540770461358e-05,
+      "loss": 0.9351,
+      "step": 651
+    },
+    {
+      "epoch": 0.4726350126857557,
+      "grad_norm": 2.2367136913648578,
+      "learning_rate": 1.1370268889996665e-05,
+      "loss": 0.9879,
+      "step": 652
+    },
+    {
+      "epoch": 0.47335991301196084,
+      "grad_norm": 2.16313174121554,
+      "learning_rate": 1.1346989443946734e-05,
+      "loss": 1.1194,
+      "step": 653
+    },
+    {
+      "epoch": 0.474084813338166,
+      "grad_norm": 2.165617974637285,
+      "learning_rate": 1.1323702560842998e-05,
+      "loss": 1.0268,
+      "step": 654
+    },
+    {
+      "epoch": 0.47480971366437114,
+      "grad_norm": 2.150563367555029,
+      "learning_rate": 1.1300408369257962e-05,
+      "loss": 1.1303,
+      "step": 655
+    },
+    {
+      "epoch": 0.4755346139905763,
+      "grad_norm": 2.1496140379016513,
+      "learning_rate": 1.127710699780448e-05,
+      "loss": 1.0181,
+      "step": 656
+    },
+    {
+      "epoch": 0.47625951431678143,
+      "grad_norm": 1.9308649623116936,
+      "learning_rate": 1.125379857513505e-05,
+      "loss": 1.0044,
+      "step": 657
+    },
+    {
+      "epoch": 0.4769844146429866,
+      "grad_norm": 2.520717730566298,
+      "learning_rate": 1.1230483229941092e-05,
+      "loss": 1.0142,
+      "step": 658
+    },
+    {
+      "epoch": 0.4777093149691917,
+      "grad_norm": 1.9969419284442884,
+      "learning_rate": 1.1207161090952255e-05,
+      "loss": 0.9586,
+      "step": 659
+    },
+    {
+      "epoch": 0.47843421529539687,
+      "grad_norm": 2.199551212478649,
+      "learning_rate": 1.1183832286935703e-05,
+      "loss": 1.0442,
+      "step": 660
+    },
+    {
+      "epoch": 0.479159115621602,
+      "grad_norm": 2.4903266005653433,
+      "learning_rate": 1.1160496946695388e-05,
+      "loss": 1.057,
+      "step": 661
+    },
+    {
+      "epoch": 0.47988401594780716,
+      "grad_norm": 2.0757922408522447,
+      "learning_rate": 1.1137155199071356e-05,
+      "loss": 1.0865,
+      "step": 662
+    },
+    {
+      "epoch": 0.4806089162740123,
+      "grad_norm": 1.9478487958266806,
+      "learning_rate": 1.1113807172939033e-05,
+      "loss": 0.9878,
+      "step": 663
+    },
+    {
+      "epoch": 0.48133381660021746,
+      "grad_norm": 2.5077394876515027,
+      "learning_rate": 1.1090452997208496e-05,
+      "loss": 1.001,
+      "step": 664
+    },
+    {
+      "epoch": 0.4820587169264226,
+      "grad_norm": 2.3619801940222542,
+      "learning_rate": 1.1067092800823798e-05,
+      "loss": 1.0786,
+      "step": 665
+    },
+    {
+      "epoch": 0.48278361725262775,
+      "grad_norm": 2.4608146249469485,
+      "learning_rate": 1.1043726712762213e-05,
+      "loss": 0.8398,
+      "step": 666
+    },
+    {
+      "epoch": 0.4835085175788329,
+      "grad_norm": 2.5682411824097273,
+      "learning_rate": 1.1020354862033553e-05,
+      "loss": 1.0639,
+      "step": 667
+    },
+    {
+      "epoch": 0.48423341790503804,
+      "grad_norm": 2.2377079055852636,
+      "learning_rate": 1.099697737767945e-05,
+      "loss": 1.0577,
+      "step": 668
+    },
+    {
+      "epoch": 0.4849583182312432,
+      "grad_norm": 2.4684218039338366,
+      "learning_rate": 1.097359438877263e-05,
+      "loss": 1.2059,
+      "step": 669
+    },
+    {
+      "epoch": 0.48568321855744834,
+      "grad_norm": 1.9986702143467359,
+      "learning_rate": 1.0950206024416228e-05,
+      "loss": 1.1099,
+      "step": 670
+    },
+    {
+      "epoch": 0.4864081188836535,
+      "grad_norm": 1.651255607571327,
+      "learning_rate": 1.0926812413743041e-05,
+      "loss": 1.0481,
+      "step": 671
+    },
+    {
+      "epoch": 0.48713301920985863,
+      "grad_norm": 2.1890746076308734,
+      "learning_rate": 1.0903413685914843e-05,
+      "loss": 0.9785,
+      "step": 672
+    },
+    {
+      "epoch": 0.4878579195360638,
+      "grad_norm": 1.983204547752865,
+      "learning_rate": 1.0880009970121655e-05,
+      "loss": 0.9511,
+      "step": 673
+    },
+    {
+      "epoch": 0.4885828198622689,
+      "grad_norm": 1.992684611945782,
+      "learning_rate": 1.0856601395581037e-05,
+      "loss": 1.0077,
+      "step": 674
+    },
+    {
+      "epoch": 0.48930772018847407,
+      "grad_norm": 2.155401888693332,
+      "learning_rate": 1.0833188091537386e-05,
+      "loss": 0.9072,
+      "step": 675
+    },
+    {
+      "epoch": 0.4900326205146792,
+      "grad_norm": 2.7840807715345406,
+      "learning_rate": 1.08097701872612e-05,
+      "loss": 1.0153,
+      "step": 676
+    },
+    {
+      "epoch": 0.49075752084088436,
+      "grad_norm": 2.3948835243364996,
+      "learning_rate": 1.0786347812048381e-05,
+      "loss": 1.0366,
+      "step": 677
+    },
+    {
+      "epoch": 0.4914824211670895,
+      "grad_norm": 2.3996815926961865,
+      "learning_rate": 1.0762921095219511e-05,
+      "loss": 1.1168,
+      "step": 678
+    },
+    {
+      "epoch": 0.49220732149329466,
+      "grad_norm": 2.186176414807414,
+      "learning_rate": 1.0739490166119155e-05,
+      "loss": 0.9772,
+      "step": 679
+    },
+    {
+      "epoch": 0.4929322218194998,
+      "grad_norm": 2.397287885437739,
+      "learning_rate": 1.0716055154115123e-05,
+      "loss": 1.0678,
+      "step": 680
+    },
+    {
+      "epoch": 0.49365712214570495,
+      "grad_norm": 2.257426550702264,
+      "learning_rate": 1.0692616188597775e-05,
+      "loss": 0.9975,
+      "step": 681
+    },
+    {
+      "epoch": 0.4943820224719101,
+      "grad_norm": 2.2392714970259324,
+      "learning_rate": 1.0669173398979292e-05,
+      "loss": 0.9476,
+      "step": 682
+    },
+    {
+      "epoch": 0.49510692279811525,
+      "grad_norm": 1.9243901111495054,
+      "learning_rate": 1.0645726914692973e-05,
+      "loss": 1.0169,
+      "step": 683
+    },
+    {
+      "epoch": 0.4958318231243204,
+      "grad_norm": 1.874543436460937,
+      "learning_rate": 1.062227686519252e-05,
+      "loss": 0.9925,
+      "step": 684
+    },
+    {
+      "epoch": 0.49655672345052554,
+      "grad_norm": 2.4206739807777744,
+      "learning_rate": 1.0598823379951315e-05,
+      "loss": 0.9653,
+      "step": 685
+    },
+    {
+      "epoch": 0.4972816237767307,
+      "grad_norm": 1.8452641728018429,
+      "learning_rate": 1.057536658846171e-05,
+      "loss": 1.1065,
+      "step": 686
+    },
+    {
+      "epoch": 0.49800652410293583,
+      "grad_norm": 2.615942811689107,
+      "learning_rate": 1.0551906620234312e-05,
+      "loss": 1.1092,
+      "step": 687
+    },
+    {
+      "epoch": 0.498731424429141,
+      "grad_norm": 2.7357084625917576,
+      "learning_rate": 1.0528443604797266e-05,
+      "loss": 1.1432,
+      "step": 688
+    },
+    {
+      "epoch": 0.4994563247553461,
+      "grad_norm": 2.2460612751605185,
+      "learning_rate": 1.0504977671695542e-05,
+      "loss": 0.9907,
+      "step": 689
+    },
+    {
+      "epoch": 0.5001812250815513,
+      "grad_norm": 2.753596180660813,
+      "learning_rate": 1.0481508950490222e-05,
+      "loss": 1.0457,
+      "step": 690
+    },
+    {
+      "epoch": 0.5001812250815513,
+      "eval_loss": 1.8697493076324463,
+      "eval_runtime": 1108.3592,
+      "eval_samples_per_second": 15.928,
+      "eval_steps_per_second": 0.498,
+      "step": 690
+    },
+    {
+      "epoch": 0.5009061254077565,
+      "grad_norm": 2.6999729213529817,
+      "learning_rate": 1.0458037570757781e-05,
+      "loss": 1.0595,
+      "step": 691
+    },
+    {
+      "epoch": 0.5016310257339616,
+      "grad_norm": 2.211696739230099,
+      "learning_rate": 1.0434563662089367e-05,
+      "loss": 0.9517,
+      "step": 692
+    },
+    {
+      "epoch": 0.5023559260601668,
+      "grad_norm": 2.4087448709687935,
+      "learning_rate": 1.04110873540901e-05,
+      "loss": 1.1667,
+      "step": 693
+    },
+    {
+      "epoch": 0.5030808263863719,
+      "grad_norm": 2.4196372563289645,
+      "learning_rate": 1.0387608776378337e-05,
+      "loss": 1.0019,
+      "step": 694
+    },
+    {
+      "epoch": 0.5038057267125771,
+      "grad_norm": 2.2271775487795424,
+      "learning_rate": 1.0364128058584974e-05,
+      "loss": 1.0789,
+      "step": 695
+    },
+    {
+      "epoch": 0.5045306270387822,
+      "grad_norm": 2.3449990407458,
+      "learning_rate": 1.0340645330352723e-05,
+      "loss": 1.0373,
+      "step": 696
+    },
+    {
+      "epoch": 0.5052555273649874,
+      "grad_norm": 2.0780920184441847,
+      "learning_rate": 1.0317160721335393e-05,
+      "loss": 1.0106,
+      "step": 697
+    },
+    {
+      "epoch": 0.5059804276911924,
+      "grad_norm": 2.5924431334903186,
+      "learning_rate": 1.0293674361197173e-05,
+      "loss": 1.0588,
+      "step": 698
+    },
+    {
+      "epoch": 0.5067053280173976,
+      "grad_norm": 2.1365135156985744,
+      "learning_rate": 1.0270186379611932e-05,
+      "loss": 1.023,
+      "step": 699
+    },
+    {
+      "epoch": 0.5074302283436027,
+      "grad_norm": 2.173447677361539,
+      "learning_rate": 1.0246696906262484e-05,
+      "loss": 1.0468,
+      "step": 700
+    },
+    {
+      "epoch": 0.5081551286698079,
+      "grad_norm": 2.245470686825682,
+      "learning_rate": 1.0223206070839878e-05,
+      "loss": 1.0525,
+      "step": 701
+    },
+    {
+      "epoch": 0.508880028996013,
+      "grad_norm": 2.2398665851941706,
+      "learning_rate": 1.0199714003042685e-05,
+      "loss": 1.0942,
+      "step": 702
+    },
+    {
+      "epoch": 0.5096049293222182,
+      "grad_norm": 2.5997096574935314,
+      "learning_rate": 1.017622083257628e-05,
+      "loss": 1.0766,
+      "step": 703
+    },
+    {
+      "epoch": 0.5103298296484233,
+      "grad_norm": 2.4146958737582644,
+      "learning_rate": 1.0152726689152135e-05,
+      "loss": 1.0928,
+      "step": 704
+    },
+    {
+      "epoch": 0.5110547299746285,
+      "grad_norm": 2.130801798872479,
+      "learning_rate": 1.0129231702487077e-05,
+      "loss": 0.972,
+      "step": 705
+    },
+    {
+      "epoch": 0.5117796303008336,
+      "grad_norm": 2.236352513835616,
+      "learning_rate": 1.01057360023026e-05,
+      "loss": 1.0103,
+      "step": 706
+    },
+    {
+      "epoch": 0.5125045306270388,
+      "grad_norm": 2.188186582922333,
+      "learning_rate": 1.0082239718324136e-05,
+      "loss": 1.0367,
+      "step": 707
+    },
+    {
+      "epoch": 0.5132294309532439,
+      "grad_norm": 2.198716617442744,
+      "learning_rate": 1.0058742980280341e-05,
+      "loss": 0.972,
+      "step": 708
+    },
+    {
+      "epoch": 0.5139543312794491,
+      "grad_norm": 2.294063462573355,
+      "learning_rate": 1.0035245917902376e-05,
+      "loss": 1.0778,
+      "step": 709
+    },
+    {
+      "epoch": 0.5146792316056542,
+      "grad_norm": 2.228364149473082,
+      "learning_rate": 1.001174866092319e-05,
+      "loss": 1.0271,
+      "step": 710
+    },
+    {
+      "epoch": 0.5154041319318594,
+      "grad_norm": 1.9300202276554854,
+      "learning_rate": 9.988251339076811e-06,
+      "loss": 0.9044,
+      "step": 711
+    },
+    {
+      "epoch": 0.5161290322580645,
+      "grad_norm": 2.3634423427900533,
+      "learning_rate": 9.964754082097626e-06,
+      "loss": 1.0843,
+      "step": 712
+    },
+    {
+      "epoch": 0.5168539325842697,
+      "grad_norm": 2.38754231763043,
+      "learning_rate": 9.941257019719662e-06,
+      "loss": 1.0854,
+      "step": 713
+    },
+    {
+      "epoch": 0.5175788329104748,
+      "grad_norm": 1.8044209443143941,
+      "learning_rate": 9.917760281675867e-06,
+      "loss": 0.9059,
+      "step": 714
+    },
+    {
+      "epoch": 0.51830373323668,
+      "grad_norm": 2.506057831976918,
+      "learning_rate": 9.894263997697405e-06,
+      "loss": 1.0033,
+      "step": 715
+    },
+    {
+      "epoch": 0.5190286335628851,
+      "grad_norm": 2.383694009514269,
+      "learning_rate": 9.870768297512924e-06,
+      "loss": 0.9435,
+      "step": 716
+    },
+    {
+      "epoch": 0.5197535338890903,
+      "grad_norm": 2.7218871091799195,
+      "learning_rate": 9.847273310847869e-06,
+      "loss": 1.013,
+      "step": 717
+    },
+    {
+      "epoch": 0.5204784342152954,
+      "grad_norm": 1.9488674447864536,
+      "learning_rate": 9.823779167423723e-06,
+      "loss": 1.064,
+      "step": 718
+    },
+    {
+      "epoch": 0.5212033345415006,
+      "grad_norm": 2.543494468466047,
+      "learning_rate": 9.80028599695732e-06,
+      "loss": 1.0396,
+      "step": 719
+    },
+    {
+      "epoch": 0.5219282348677057,
+      "grad_norm": 2.1642282148880976,
+      "learning_rate": 9.776793929160129e-06,
+      "loss": 0.9959,
+      "step": 720
+    },
+    {
+      "epoch": 0.5226531351939109,
+      "grad_norm": 1.9378193767391163,
+      "learning_rate": 9.753303093737518e-06,
+      "loss": 1.0852,
+      "step": 721
+    },
+    {
+      "epoch": 0.523378035520116,
+      "grad_norm": 2.5527944661940536,
+      "learning_rate": 9.729813620388071e-06,
+      "loss": 1.1522,
+      "step": 722
+    },
+    {
+      "epoch": 0.5241029358463212,
+      "grad_norm": 2.3217437760784074,
+      "learning_rate": 9.70632563880283e-06,
+      "loss": 1.0067,
+      "step": 723
+    },
+    {
+      "epoch": 0.5248278361725263,
+      "grad_norm": 3.1417602276811527,
+      "learning_rate": 9.682839278664614e-06,
+      "loss": 0.9952,
+      "step": 724
+    },
+    {
+      "epoch": 0.5255527364987315,
+      "grad_norm": 1.9148171819921405,
+      "learning_rate": 9.659354669647277e-06,
+      "loss": 0.9656,
+      "step": 725
+    },
+    {
+      "epoch": 0.5262776368249366,
+      "grad_norm": 2.449509825571267,
+      "learning_rate": 9.635871941415026e-06,
+      "loss": 0.9131,
+      "step": 726
+    },
+    {
+      "epoch": 0.5270025371511418,
+      "grad_norm": 2.1963443291334697,
+      "learning_rate": 9.612391223621667e-06,
+      "loss": 1.0032,
+      "step": 727
+    },
+    {
+      "epoch": 0.5277274374773469,
+      "grad_norm": 2.2338194520764074,
+      "learning_rate": 9.588912645909905e-06,
+      "loss": 1.0494,
+      "step": 728
+    },
+    {
+      "epoch": 0.528452337803552,
+      "grad_norm": 1.9959846236290153,
+      "learning_rate": 9.565436337910637e-06,
+      "loss": 0.947,
+      "step": 729
+    },
+    {
+      "epoch": 0.5291772381297571,
+      "grad_norm": 2.078163095936865,
+      "learning_rate": 9.54196242924222e-06,
+      "loss": 1.0311,
+      "step": 730
+    },
+    {
+      "epoch": 0.5299021384559623,
+      "grad_norm": 2.1980066549775024,
+      "learning_rate": 9.51849104950978e-06,
+      "loss": 1.0051,
+      "step": 731
+    },
+    {
+      "epoch": 0.5306270387821674,
+      "grad_norm": 2.4191287974234705,
+      "learning_rate": 9.495022328304461e-06,
+      "loss": 1.0469,
+      "step": 732
+    },
+    {
+      "epoch": 0.5313519391083726,
+      "grad_norm": 2.298052468185075,
+      "learning_rate": 9.47155639520274e-06,
+      "loss": 1.0115,
+      "step": 733
+    },
+    {
+      "epoch": 0.5320768394345777,
+      "grad_norm": 2.1232889455421406,
+      "learning_rate": 9.448093379765693e-06,
+      "loss": 1.0524,
+      "step": 734
+    },
+    {
+      "epoch": 0.5328017397607829,
+      "grad_norm": 2.44786208230678,
+      "learning_rate": 9.424633411538289e-06,
+      "loss": 1.0772,
+      "step": 735
+    },
+    {
+      "epoch": 0.533526640086988,
+      "grad_norm": 2.102354340691057,
+      "learning_rate": 9.401176620048687e-06,
+      "loss": 1.0344,
+      "step": 736
+    },
+    {
+      "epoch": 0.5342515404131932,
+      "grad_norm": 2.108353702137098,
+      "learning_rate": 9.377723134807482e-06,
+      "loss": 1.1669,
+      "step": 737
+    },
+    {
+      "epoch": 0.5349764407393983,
+      "grad_norm": 2.4286899167576412,
+      "learning_rate": 9.35427308530703e-06,
+      "loss": 1.0635,
+      "step": 738
+    },
+    {
+      "epoch": 0.5357013410656035,
+      "grad_norm": 2.2012613191456025,
+      "learning_rate": 9.330826601020713e-06,
+      "loss": 1.1047,
+      "step": 739
+    },
+    {
+      "epoch": 0.5364262413918086,
+      "grad_norm": 2.7290128431948646,
+      "learning_rate": 9.307383811402229e-06,
+      "loss": 1.0265,
+      "step": 740
+    },
+    {
+      "epoch": 0.5371511417180138,
+      "grad_norm": 2.3858907048476135,
+      "learning_rate": 9.283944845884878e-06,
+      "loss": 0.9387,
+      "step": 741
+    },
+    {
+      "epoch": 0.5378760420442189,
+      "grad_norm": 2.080475476085547,
+      "learning_rate": 9.260509833880848e-06,
+      "loss": 0.9617,
+      "step": 742
+    },
+    {
+      "epoch": 0.5386009423704241,
+      "grad_norm": 1.8615886294130846,
+      "learning_rate": 9.23707890478049e-06,
+      "loss": 0.8874,
+      "step": 743
+    },
+    {
+      "epoch": 0.5393258426966292,
+      "grad_norm": 1.9904190396103651,
+      "learning_rate": 9.213652187951624e-06,
+      "loss": 1.0746,
+      "step": 744
+    },
+    {
+      "epoch": 0.5400507430228344,
+      "grad_norm": 2.96232733164202,
+      "learning_rate": 9.190229812738802e-06,
+      "loss": 1.1089,
+      "step": 745
+    },
+    {
+      "epoch": 0.5407756433490395,
+      "grad_norm": 2.525943985627746,
+      "learning_rate": 9.166811908462616e-06,
+      "loss": 1.0642,
+      "step": 746
+    },
+    {
+      "epoch": 0.5415005436752447,
+      "grad_norm": 2.2418322634514167,
+      "learning_rate": 9.143398604418965e-06,
+      "loss": 1.1124,
+      "step": 747
+    },
+    {
+      "epoch": 0.5422254440014498,
+      "grad_norm": 1.9220381898458985,
+      "learning_rate": 9.11999002987835e-06,
+      "loss": 0.9473,
+      "step": 748
+    },
+    {
+      "epoch": 0.542950344327655,
+      "grad_norm": 2.6621473432063176,
+      "learning_rate": 9.096586314085162e-06,
+      "loss": 1.0125,
+      "step": 749
+    },
+    {
+      "epoch": 0.5436752446538601,
+      "grad_norm": 2.3340603683637156,
+      "learning_rate": 9.07318758625696e-06,
+      "loss": 1.0542,
+      "step": 750
+    },
+    {
+      "epoch": 0.5444001449800653,
+      "grad_norm": 2.1363055122068215,
+      "learning_rate": 9.049793975583775e-06,
+      "loss": 1.0586,
+      "step": 751
+    },
+    {
+      "epoch": 0.5451250453062704,
+      "grad_norm": 2.0326003541767803,
+      "learning_rate": 9.026405611227371e-06,
+      "loss": 0.9054,
+      "step": 752
+    },
+    {
+      "epoch": 0.5458499456324756,
+      "grad_norm": 1.6432957234320769,
+      "learning_rate": 9.003022622320555e-06,
+      "loss": 0.9625,
+      "step": 753
+    },
+    {
+      "epoch": 0.5465748459586807,
+      "grad_norm": 1.78897998635208,
+      "learning_rate": 8.97964513796645e-06,
+      "loss": 1.0138,
+      "step": 754
+    },
+    {
+      "epoch": 0.5472997462848859,
+      "grad_norm": 2.797841405037185,
+      "learning_rate": 8.95627328723779e-06,
+      "loss": 1.2344,
+      "step": 755
+    },
+    {
+      "epoch": 0.548024646611091,
+      "grad_norm": 2.0182212487321864,
+      "learning_rate": 8.932907199176206e-06,
+      "loss": 1.0249,
+      "step": 756
+    },
+    {
+      "epoch": 0.5487495469372962,
+      "grad_norm": 2.1185607535847586,
+      "learning_rate": 8.909547002791506e-06,
+      "loss": 1.0009,
+      "step": 757
+    },
+    {
+      "epoch": 0.5494744472635013,
+      "grad_norm": 2.5153165221241687,
+      "learning_rate": 8.886192827060974e-06,
+      "loss": 1.0271,
+      "step": 758
+    },
+    {
+      "epoch": 0.5501993475897065,
+      "grad_norm": 2.205784694298984,
+      "learning_rate": 8.862844800928645e-06,
+      "loss": 1.0508,
+      "step": 759
+    },
+    {
+      "epoch": 0.5509242479159115,
+      "grad_norm": 2.908166150479793,
+      "learning_rate": 8.839503053304614e-06,
+      "loss": 1.039,
+      "step": 760
+    },
+    {
+      "epoch": 0.5516491482421167,
+      "grad_norm": 1.9201627389676703,
+      "learning_rate": 8.8161677130643e-06,
+      "loss": 0.9532,
+      "step": 761
+    },
+    {
+      "epoch": 0.5523740485683218,
+      "grad_norm": 2.923787620312773,
+      "learning_rate": 8.792838909047747e-06,
+      "loss": 1.0188,
+      "step": 762
+    },
+    {
+      "epoch": 0.553098948894527,
+      "grad_norm": 3.157465367532413,
+      "learning_rate": 8.769516770058915e-06,
+      "loss": 1.0947,
+      "step": 763
+    },
+    {
+      "epoch": 0.5538238492207321,
+      "grad_norm": 2.060375498630697,
+      "learning_rate": 8.746201424864956e-06,
+      "loss": 0.9142,
+      "step": 764
+    },
+    {
+      "epoch": 0.5545487495469373,
+      "grad_norm": 1.9982518012339532,
+      "learning_rate": 8.722893002195523e-06,
+      "loss": 0.9624,
+      "step": 765
+    },
+    {
+      "epoch": 0.5552736498731424,
+      "grad_norm": 1.841122163960603,
+      "learning_rate": 8.699591630742042e-06,
+      "loss": 0.979,
+      "step": 766
+    },
+    {
+      "epoch": 0.5559985501993476,
+      "grad_norm": 2.0151604980838633,
+      "learning_rate": 8.676297439157007e-06,
+      "loss": 0.994,
+      "step": 767
+    },
+    {
+      "epoch": 0.5567234505255527,
+      "grad_norm": 2.5456721286611037,
+      "learning_rate": 8.653010556053271e-06,
+      "loss": 1.0495,
+      "step": 768
+    },
+    {
+      "epoch": 0.5574483508517579,
+      "grad_norm": 2.7097490860787183,
+      "learning_rate": 8.629731110003337e-06,
+      "loss": 1.124,
+      "step": 769
+    },
+    {
+      "epoch": 0.558173251177963,
+      "grad_norm": 1.949724143019753,
+      "learning_rate": 8.606459229538645e-06,
+      "loss": 0.997,
+      "step": 770
+    },
+    {
+      "epoch": 0.5588981515041682,
+      "grad_norm": 2.020801130677908,
+      "learning_rate": 8.583195043148864e-06,
+      "loss": 1.1449,
+      "step": 771
+    },
+    {
+      "epoch": 0.5596230518303733,
+      "grad_norm": 2.130589840488668,
+      "learning_rate": 8.55993867928118e-06,
+      "loss": 0.9983,
+      "step": 772
+    },
+    {
+      "epoch": 0.5603479521565785,
+      "grad_norm": 2.0836041709296316,
+      "learning_rate": 8.536690266339593e-06,
+      "loss": 0.9946,
+      "step": 773
+    },
+    {
+      "epoch": 0.5610728524827836,
+      "grad_norm": 2.3227053384824274,
+      "learning_rate": 8.513449932684198e-06,
+      "loss": 0.8824,
+      "step": 774
+    },
+    {
+      "epoch": 0.5617977528089888,
+      "grad_norm": 1.983623134191808,
+      "learning_rate": 8.490217806630489e-06,
+      "loss": 0.9447,
+      "step": 775
+    },
+    {
+      "epoch": 0.5625226531351939,
+      "grad_norm": 2.1427968901005725,
+      "learning_rate": 8.46699401644864e-06,
+      "loss": 1.0275,
+      "step": 776
+    },
+    {
+      "epoch": 0.5632475534613991,
+      "grad_norm": 2.263521727209926,
+      "learning_rate": 8.443778690362801e-06,
+      "loss": 0.9779,
+      "step": 777
+    },
+    {
+      "epoch": 0.5639724537876042,
+      "grad_norm": 1.784185254505844,
+      "learning_rate": 8.42057195655039e-06,
+      "loss": 0.9217,
+      "step": 778
+    },
+    {
+      "epoch": 0.5646973541138094,
+      "grad_norm": 2.5535009035102934,
+      "learning_rate": 8.397373943141384e-06,
+      "loss": 1.0259,
+      "step": 779
+    },
+    {
+      "epoch": 0.5654222544400145,
+      "grad_norm": 2.3274203754164873,
+      "learning_rate": 8.374184778217617e-06,
+      "loss": 1.0449,
+      "step": 780
+    },
+    {
+      "epoch": 0.5661471547662197,
+      "grad_norm": 2.751811026011905,
+      "learning_rate": 8.351004589812061e-06,
+      "loss": 1.0106,
+      "step": 781
+    },
+    {
+      "epoch": 0.5668720550924248,
+      "grad_norm": 2.0365364413353726,
+      "learning_rate": 8.327833505908135e-06,
+      "loss": 1.0282,
+      "step": 782
+    },
+    {
+      "epoch": 0.56759695541863,
+      "grad_norm": 2.3504702430826154,
+      "learning_rate": 8.30467165443898e-06,
+      "loss": 1.0629,
+      "step": 783
+    },
+    {
+      "epoch": 0.5683218557448351,
+      "grad_norm": 2.0575133385244646,
+      "learning_rate": 8.281519163286772e-06,
+      "loss": 1.0005,
+      "step": 784
+    },
+    {
+      "epoch": 0.5690467560710403,
+      "grad_norm": 2.5919383011426467,
+      "learning_rate": 8.258376160282001e-06,
+      "loss": 1.0555,
+      "step": 785
+    },
+    {
+      "epoch": 0.5697716563972454,
+      "grad_norm": 2.4060395092358102,
+      "learning_rate": 8.235242773202772e-06,
+      "loss": 1.0631,
+      "step": 786
+    },
+    {
+      "epoch": 0.5704965567234506,
+      "grad_norm": 1.9397501796747694,
+      "learning_rate": 8.212119129774098e-06,
+      "loss": 1.1304,
+      "step": 787
+    },
+    {
+      "epoch": 0.5712214570496557,
+      "grad_norm": 2.110530023102397,
+      "learning_rate": 8.189005357667191e-06,
+      "loss": 0.9567,
+      "step": 788
+    },
+    {
+      "epoch": 0.5719463573758609,
+      "grad_norm": 2.1870319191564658,
+      "learning_rate": 8.165901584498774e-06,
+      "loss": 1.0683,
+      "step": 789
+    },
+    {
+      "epoch": 0.572671257702066,
+      "grad_norm": 1.8495796212471047,
+      "learning_rate": 8.142807937830348e-06,
+      "loss": 0.9837,
+      "step": 790
+    },
+    {
+      "epoch": 0.5733961580282712,
+      "grad_norm": 2.64388618965451,
+      "learning_rate": 8.11972454516751e-06,
+      "loss": 0.9659,
+      "step": 791
+    },
+    {
+      "epoch": 0.5741210583544762,
+      "grad_norm": 2.0330246862165415,
+      "learning_rate": 8.096651533959244e-06,
+      "loss": 0.9591,
+      "step": 792
+    },
+    {
+      "epoch": 0.5748459586806814,
+      "grad_norm": 2.2631207779996827,
+      "learning_rate": 8.073589031597209e-06,
+      "loss": 0.9873,
+      "step": 793
+    },
+    {
+      "epoch": 0.5755708590068865,
+      "grad_norm": 2.3948632401763064,
+      "learning_rate": 8.050537165415053e-06,
+      "loss": 1.1371,
+      "step": 794
+    },
+    {
+      "epoch": 0.5762957593330917,
+      "grad_norm": 2.075607270336768,
+      "learning_rate": 8.027496062687688e-06,
+      "loss": 0.8908,
+      "step": 795
+    },
+    {
+      "epoch": 0.5770206596592968,
+      "grad_norm": 2.065746544851814,
+      "learning_rate": 8.004465850630605e-06,
+      "loss": 1.0025,
+      "step": 796
+    },
+    {
+      "epoch": 0.577745559985502,
+      "grad_norm": 2.2202432240869845,
+      "learning_rate": 7.981446656399162e-06,
+      "loss": 1.0429,
+      "step": 797
+    },
+    {
+      "epoch": 0.5784704603117071,
+      "grad_norm": 2.309075300778911,
+      "learning_rate": 7.958438607087884e-06,
+      "loss": 0.9851,
+      "step": 798
+    },
+    {
+      "epoch": 0.5791953606379123,
+      "grad_norm": 2.4033759237105166,
+      "learning_rate": 7.935441829729766e-06,
+      "loss": 1.032,
+      "step": 799
+    },
+    {
+      "epoch": 0.5799202609641174,
+      "grad_norm": 2.625985750403029,
+      "learning_rate": 7.912456451295567e-06,
+      "loss": 1.0102,
+      "step": 800
+    },
+    {
+      "epoch": 0.5806451612903226,
+      "grad_norm": 2.410099594821792,
+      "learning_rate": 7.889482598693108e-06,
+      "loss": 0.9639,
+      "step": 801
+    },
+    {
+      "epoch": 0.5813700616165277,
+      "grad_norm": 1.9619910669766136,
+      "learning_rate": 7.866520398766574e-06,
+      "loss": 0.9297,
+      "step": 802
+    },
+    {
+      "epoch": 0.5820949619427329,
+      "grad_norm": 2.930559344753908,
+      "learning_rate": 7.843569978295804e-06,
+      "loss": 1.0681,
+      "step": 803
+    },
+    {
+      "epoch": 0.582819862268938,
+      "grad_norm": 1.8579566089499246,
+      "learning_rate": 7.820631463995622e-06,
+      "loss": 0.9247,
+      "step": 804
+    },
+    {
+      "epoch": 0.5835447625951432,
+      "grad_norm": 1.783098743265068,
+      "learning_rate": 7.797704982515094e-06,
+      "loss": 0.9649,
+      "step": 805
+    },
+    {
+      "epoch": 0.5842696629213483,
+      "grad_norm": 1.9534992322071332,
+      "learning_rate": 7.774790660436857e-06,
+      "loss": 0.9234,
+      "step": 806
+    },
+    {
+      "epoch": 0.5849945632475535,
+      "grad_norm": 2.3775300305715428,
+      "learning_rate": 7.751888624276418e-06,
+      "loss": 1.0246,
+      "step": 807
+    },
+    {
+      "epoch": 0.5857194635737586,
+      "grad_norm": 2.0533804995629525,
+      "learning_rate": 7.728999000481436e-06,
+      "loss": 0.9705,
+      "step": 808
+    },
+    {
+      "epoch": 0.5864443638999638,
+      "grad_norm": 1.959728561512236,
+      "learning_rate": 7.706121915431059e-06,
+      "loss": 0.9359,
+      "step": 809
+    },
+    {
+      "epoch": 0.5871692642261689,
+      "grad_norm": 2.502678481308203,
+      "learning_rate": 7.68325749543519e-06,
+      "loss": 0.9802,
+      "step": 810
+    },
+    {
+      "epoch": 0.5878941645523741,
+      "grad_norm": 3.0385891335086352,
+      "learning_rate": 7.660405866733813e-06,
+      "loss": 1.1222,
+      "step": 811
+    },
+    {
+      "epoch": 0.5886190648785792,
+      "grad_norm": 2.040606391856475,
+      "learning_rate": 7.637567155496277e-06,
+      "loss": 1.02,
+      "step": 812
+    },
+    {
+      "epoch": 0.5893439652047844,
+      "grad_norm": 2.305055241806458,
+      "learning_rate": 7.614741487820626e-06,
+      "loss": 0.9964,
+      "step": 813
+    },
+    {
+      "epoch": 0.5900688655309895,
+      "grad_norm": 1.882568911490993,
+      "learning_rate": 7.5919289897328784e-06,
+      "loss": 0.9941,
+      "step": 814
+    },
+    {
+      "epoch": 0.5907937658571947,
+      "grad_norm": 2.5290386344366156,
+      "learning_rate": 7.569129787186342e-06,
+      "loss": 1.0734,
+      "step": 815
+    },
+    {
+      "epoch": 0.5915186661833998,
+      "grad_norm": 2.8819681752053454,
+      "learning_rate": 7.546344006060918e-06,
+      "loss": 1.0041,
+      "step": 816
+    },
+    {
+      "epoch": 0.592243566509605,
+      "grad_norm": 1.8804822924012536,
+      "learning_rate": 7.523571772162392e-06,
+      "loss": 0.9151,
+      "step": 817
+    },
+    {
+      "epoch": 0.5929684668358101,
+      "grad_norm": 2.255992660838785,
+      "learning_rate": 7.500813211221778e-06,
+      "loss": 1.0903,
+      "step": 818
+    },
+    {
+      "epoch": 0.5936933671620153,
+      "grad_norm": 2.9086758708738345,
+      "learning_rate": 7.478068448894577e-06,
+      "loss": 1.0814,
+      "step": 819
+    },
+    {
+      "epoch": 0.5944182674882204,
+      "grad_norm": 2.437481894199292,
+      "learning_rate": 7.455337610760114e-06,
+      "loss": 1.1113,
+      "step": 820
+    },
+    {
+      "epoch": 0.5951431678144256,
+      "grad_norm": 1.8660888898062387,
+      "learning_rate": 7.4326208223208274e-06,
+      "loss": 0.9695,
+      "step": 821
+    },
+    {
+      "epoch": 0.5958680681406306,
+      "grad_norm": 2.1229612134239995,
+      "learning_rate": 7.409918209001585e-06,
+      "loss": 1.0515,
+      "step": 822
+    },
+    {
+      "epoch": 0.5965929684668358,
+      "grad_norm": 2.3583730518901995,
+      "learning_rate": 7.387229896149006e-06,
+      "loss": 0.9714,
+      "step": 823
+    },
+    {
+      "epoch": 0.5973178687930409,
+      "grad_norm": 2.3153711669204413,
+      "learning_rate": 7.364556009030734e-06,
+      "loss": 1.059,
+      "step": 824
+    },
+    {
+      "epoch": 0.5980427691192461,
+      "grad_norm": 1.9967290233140527,
+      "learning_rate": 7.341896672834776e-06,
+      "loss": 1.047,
+      "step": 825
+    },
+    {
+      "epoch": 0.5987676694454512,
+      "grad_norm": 1.9339818611902988,
+      "learning_rate": 7.31925201266879e-06,
+      "loss": 0.972,
+      "step": 826
+    },
+    {
+      "epoch": 0.5994925697716564,
+      "grad_norm": 1.9457458272271317,
+      "learning_rate": 7.2966221535594085e-06,
+      "loss": 1.0206,
+      "step": 827
+    },
+    {
+      "epoch": 0.6002174700978615,
+      "grad_norm": 2.1792046814563495,
+      "learning_rate": 7.274007220451553e-06,
+      "loss": 1.0732,
+      "step": 828
+    },
+    {
+      "epoch": 0.6009423704240667,
+      "grad_norm": 2.029220407091385,
+      "learning_rate": 7.251407338207725e-06,
+      "loss": 0.8824,
+      "step": 829
+    },
+    {
+      "epoch": 0.6016672707502718,
+      "grad_norm": 1.915241829499402,
+      "learning_rate": 7.228822631607327e-06,
+      "loss": 0.9666,
+      "step": 830
+    },
+    {
+      "epoch": 0.602392171076477,
+      "grad_norm": 1.8518746441416718,
+      "learning_rate": 7.2062532253459714e-06,
+      "loss": 0.8742,
+      "step": 831
+    },
+    {
+      "epoch": 0.6031170714026821,
+      "grad_norm": 2.5929079919190547,
+      "learning_rate": 7.183699244034797e-06,
+      "loss": 1.1043,
+      "step": 832
+    },
+    {
+      "epoch": 0.6038419717288873,
+      "grad_norm": 2.2618493759552516,
+      "learning_rate": 7.161160812199785e-06,
+      "loss": 1.1689,
+      "step": 833
+    },
+    {
+      "epoch": 0.6045668720550924,
+      "grad_norm": 1.9916609519278592,
+      "learning_rate": 7.138638054281055e-06,
+      "loss": 1.0807,
+      "step": 834
+    },
+    {
+      "epoch": 0.6052917723812976,
+      "grad_norm": 1.8106802403260034,
+      "learning_rate": 7.116131094632186e-06,
+      "loss": 0.8694,
+      "step": 835
+    },
+    {
+      "epoch": 0.6060166727075027,
+      "grad_norm": 2.2006653848900015,
+      "learning_rate": 7.093640057519531e-06,
+      "loss": 1.0905,
+      "step": 836
+    },
+    {
+      "epoch": 0.6067415730337079,
+      "grad_norm": 2.3615298366429496,
+      "learning_rate": 7.0711650671215405e-06,
+      "loss": 1.0178,
+      "step": 837
+    },
+    {
+      "epoch": 0.607466473359913,
+      "grad_norm": 1.9549782714474917,
+      "learning_rate": 7.048706247528061e-06,
+      "loss": 1.0098,
+      "step": 838
+    },
+    {
+      "epoch": 0.6081913736861182,
+      "grad_norm": 2.2334786388231396,
+      "learning_rate": 7.02626372273965e-06,
+      "loss": 1.0028,
+      "step": 839
+    },
+    {
+      "epoch": 0.6089162740123233,
+      "grad_norm": 2.515375317083732,
+      "learning_rate": 7.003837616666906e-06,
+      "loss": 0.8934,
+      "step": 840
+    },
+    {
+      "epoch": 0.6096411743385285,
+      "grad_norm": 2.0714175972616347,
+      "learning_rate": 6.98142805312977e-06,
+      "loss": 1.008,
+      "step": 841
+    },
+    {
+      "epoch": 0.6103660746647336,
+      "grad_norm": 2.183667905495028,
+      "learning_rate": 6.959035155856857e-06,
+      "loss": 1.0119,
+      "step": 842
+    },
+    {
+      "epoch": 0.6110909749909388,
+      "grad_norm": 2.178274745085645,
+      "learning_rate": 6.936659048484755e-06,
+      "loss": 0.9791,
+      "step": 843
+    },
+    {
+      "epoch": 0.6118158753171439,
+      "grad_norm": 2.2299367836336903,
+      "learning_rate": 6.914299854557349e-06,
+      "loss": 1.0342,
+      "step": 844
+    },
+    {
+      "epoch": 0.6125407756433491,
+      "grad_norm": 1.870612790239624,
+      "learning_rate": 6.891957697525149e-06,
+      "loss": 1.0123,
+      "step": 845
+    },
+    {
+      "epoch": 0.6132656759695542,
+      "grad_norm": 2.2457853373416996,
+      "learning_rate": 6.869632700744588e-06,
+      "loss": 0.9599,
+      "step": 846
+    },
+    {
+      "epoch": 0.6139905762957594,
+      "grad_norm": 1.8964211823089199,
+      "learning_rate": 6.847324987477375e-06,
+      "loss": 1.0702,
+      "step": 847
+    },
+    {
+      "epoch": 0.6147154766219645,
+      "grad_norm": 2.2231409481558595,
+      "learning_rate": 6.82503468088977e-06,
+      "loss": 0.9252,
+      "step": 848
+    },
+    {
+      "epoch": 0.6154403769481697,
+      "grad_norm": 2.2158057236580575,
+      "learning_rate": 6.802761904051937e-06,
+      "loss": 0.8728,
+      "step": 849
+    },
+    {
+      "epoch": 0.6161652772743748,
+      "grad_norm": 1.984090771986885,
+      "learning_rate": 6.78050677993725e-06,
+      "loss": 0.9488,
+      "step": 850
+    },
+    {
+      "epoch": 0.61689017760058,
+      "grad_norm": 2.194669700982665,
+      "learning_rate": 6.758269431421618e-06,
+      "loss": 1.0144,
+      "step": 851
+    },
+    {
+      "epoch": 0.617615077926785,
+      "grad_norm": 2.391278122000099,
+      "learning_rate": 6.7360499812828195e-06,
+      "loss": 0.9911,
+      "step": 852
+    },
+    {
+      "epoch": 0.6183399782529903,
+      "grad_norm": 2.022013346401567,
+      "learning_rate": 6.713848552199791e-06,
+      "loss": 0.9898,
+      "step": 853
+    },
+    {
+      "epoch": 0.6190648785791953,
+      "grad_norm": 2.109272857624347,
+      "learning_rate": 6.6916652667519855e-06,
+      "loss": 0.9562,
+      "step": 854
+    },
+    {
+      "epoch": 0.6197897789054005,
+      "grad_norm": 2.829806104335344,
+      "learning_rate": 6.669500247418674e-06,
+      "loss": 1.0893,
+      "step": 855
+    },
+    {
+      "epoch": 0.6205146792316056,
+      "grad_norm": 2.2159865585012715,
+      "learning_rate": 6.647353616578274e-06,
+      "loss": 1.0566,
+      "step": 856
+    },
+    {
+      "epoch": 0.6212395795578108,
+      "grad_norm": 2.5699500401567272,
+      "learning_rate": 6.625225496507688e-06,
+      "loss": 0.9406,
+      "step": 857
+    },
+    {
+      "epoch": 0.6219644798840159,
+      "grad_norm": 1.9318589400149984,
+      "learning_rate": 6.6031160093816005e-06,
+      "loss": 0.9556,
+      "step": 858
+    },
+    {
+      "epoch": 0.6226893802102211,
+      "grad_norm": 2.046984039555971,
+      "learning_rate": 6.5810252772718276e-06,
+      "loss": 1.0834,
+      "step": 859
+    },
+    {
+      "epoch": 0.6234142805364262,
+      "grad_norm": 2.6992216221025087,
+      "learning_rate": 6.5589534221466324e-06,
+      "loss": 0.9793,
+      "step": 860
+    },
+    {
+      "epoch": 0.6241391808626314,
+      "grad_norm": 2.212396980492137,
+      "learning_rate": 6.536900565870052e-06,
+      "loss": 0.8663,
+      "step": 861
+    },
+    {
+      "epoch": 0.6248640811888365,
+      "grad_norm": 1.7999266895906552,
+      "learning_rate": 6.514866830201234e-06,
+      "loss": 0.8443,
+      "step": 862
+    },
+    {
+      "epoch": 0.6255889815150417,
+      "grad_norm": 2.1176385115675083,
+      "learning_rate": 6.4928523367937444e-06,
+      "loss": 0.9933,
+      "step": 863
+    },
+    {
+      "epoch": 0.6263138818412468,
+      "grad_norm": 2.2762631364540784,
+      "learning_rate": 6.470857207194918e-06,
+      "loss": 1.0051,
+      "step": 864
+    },
+    {
+      "epoch": 0.627038782167452,
+      "grad_norm": 2.0490123793497976,
+      "learning_rate": 6.448881562845169e-06,
+      "loss": 0.9627,
+      "step": 865
+    },
+    {
+      "epoch": 0.6277636824936571,
+      "grad_norm": 2.5600372472836117,
+      "learning_rate": 6.426925525077341e-06,
+      "loss": 1.0176,
+      "step": 866
+    },
+    {
+      "epoch": 0.6284885828198623,
+      "grad_norm": 2.8510918650806882,
+      "learning_rate": 6.404989215116012e-06,
+      "loss": 1.0215,
+      "step": 867
+    },
+    {
+      "epoch": 0.6292134831460674,
+      "grad_norm": 1.9769591446882206,
+      "learning_rate": 6.3830727540768445e-06,
+      "loss": 1.0687,
+      "step": 868
+    },
+    {
+      "epoch": 0.6299383834722726,
+      "grad_norm": 1.677393229029825,
+      "learning_rate": 6.361176262965909e-06,
+      "loss": 0.9384,
+      "step": 869
+    },
+    {
+      "epoch": 0.6306632837984777,
+      "grad_norm": 2.4132581003126705,
+      "learning_rate": 6.339299862679016e-06,
+      "loss": 0.973,
+      "step": 870
+    },
+    {
+      "epoch": 0.6313881841246829,
+      "grad_norm": 2.562250697801829,
+      "learning_rate": 6.317443674001055e-06,
+      "loss": 0.9105,
+      "step": 871
+    },
+    {
+      "epoch": 0.632113084450888,
+      "grad_norm": 2.0225187463112033,
+      "learning_rate": 6.295607817605316e-06,
+      "loss": 0.9315,
+      "step": 872
+    },
+    {
+      "epoch": 0.6328379847770932,
+      "grad_norm": 2.6962240997771167,
+      "learning_rate": 6.273792414052833e-06,
+      "loss": 1.058,
+      "step": 873
+    },
+    {
+      "epoch": 0.6335628851032983,
+      "grad_norm": 2.3293379147129705,
+      "learning_rate": 6.251997583791711e-06,
+      "loss": 1.0571,
+      "step": 874
+    },
+    {
+      "epoch": 0.6342877854295035,
+      "grad_norm": 2.17475653423949,
+      "learning_rate": 6.230223447156469e-06,
+      "loss": 1.0355,
+      "step": 875
+    },
+    {
+      "epoch": 0.6350126857557086,
+      "grad_norm": 2.293124315791374,
+      "learning_rate": 6.208470124367374e-06,
+      "loss": 1.0007,
+      "step": 876
+    },
+    {
+      "epoch": 0.6357375860819138,
+      "grad_norm": 2.1218951006607236,
+      "learning_rate": 6.186737735529769e-06,
+      "loss": 0.9075,
+      "step": 877
+    },
+    {
+      "epoch": 0.6364624864081189,
+      "grad_norm": 2.1947275964384114,
+      "learning_rate": 6.165026400633416e-06,
+      "loss": 1.0155,
+      "step": 878
+    },
+    {
+      "epoch": 0.6371873867343241,
+      "grad_norm": 1.8448468197236452,
+      "learning_rate": 6.143336239551836e-06,
+      "loss": 0.9704,
+      "step": 879
+    },
+    {
+      "epoch": 0.6379122870605292,
+      "grad_norm": 2.2203852100211496,
+      "learning_rate": 6.12166737204164e-06,
+      "loss": 1.0183,
+      "step": 880
+    },
+    {
+      "epoch": 0.6386371873867344,
+      "grad_norm": 1.9940936134783194,
+      "learning_rate": 6.10001991774188e-06,
+      "loss": 1.0362,
+      "step": 881
+    },
+    {
+      "epoch": 0.6393620877129395,
+      "grad_norm": 1.876178541905188,
+      "learning_rate": 6.078393996173375e-06,
+      "loss": 1.0958,
+      "step": 882
+    },
+    {
+      "epoch": 0.6400869880391447,
+      "grad_norm": 2.4300065422050174,
+      "learning_rate": 6.056789726738055e-06,
+      "loss": 1.0112,
+      "step": 883
+    },
+    {
+      "epoch": 0.6408118883653497,
+      "grad_norm": 2.182070128778044,
+      "learning_rate": 6.035207228718305e-06,
+      "loss": 1.0532,
+      "step": 884
+    },
+    {
+      "epoch": 0.641536788691555,
+      "grad_norm": 2.2790151746934826,
+      "learning_rate": 6.0136466212763055e-06,
+      "loss": 0.9994,
+      "step": 885
+    },
+    {
+      "epoch": 0.64226168901776,
+      "grad_norm": 2.0380124474521333,
+      "learning_rate": 5.992108023453376e-06,
+      "loss": 0.9901,
+      "step": 886
+    },
+    {
+      "epoch": 0.6429865893439652,
+      "grad_norm": 2.3256602048348234,
+      "learning_rate": 5.9705915541693114e-06,
+      "loss": 1.1098,
+      "step": 887
+    },
+    {
+      "epoch": 0.6437114896701703,
+      "grad_norm": 2.753858071131981,
+      "learning_rate": 5.949097332221734e-06,
+      "loss": 1.0406,
+      "step": 888
+    },
+    {
+      "epoch": 0.6444363899963755,
+      "grad_norm": 2.068578563869525,
+      "learning_rate": 5.927625476285426e-06,
+      "loss": 0.9272,
+      "step": 889
+    },
+    {
+      "epoch": 0.6451612903225806,
+      "grad_norm": 2.0943155971716947,
+      "learning_rate": 5.906176104911687e-06,
+      "loss": 1.0903,
+      "step": 890
+    },
+    {
+      "epoch": 0.6458861906487858,
+      "grad_norm": 2.244523642610638,
+      "learning_rate": 5.884749336527676e-06,
+      "loss": 1.0465,
+      "step": 891
+    },
+    {
+      "epoch": 0.6466110909749909,
+      "grad_norm": 2.2948461929009554,
+      "learning_rate": 5.8633452894357516e-06,
+      "loss": 0.9543,
+      "step": 892
+    },
+    {
+      "epoch": 0.6473359913011961,
+      "grad_norm": 2.316443702020551,
+      "learning_rate": 5.841964081812822e-06,
+      "loss": 1.0483,
+      "step": 893
+    },
+    {
+      "epoch": 0.6480608916274012,
+      "grad_norm": 2.210302183543952,
+      "learning_rate": 5.820605831709694e-06,
+      "loss": 0.8937,
+      "step": 894
+    },
+    {
+      "epoch": 0.6487857919536064,
+      "grad_norm": 2.1138527969671097,
+      "learning_rate": 5.799270657050422e-06,
+      "loss": 0.8968,
+      "step": 895
+    },
+    {
+      "epoch": 0.6495106922798115,
+      "grad_norm": 2.136178605345999,
+      "learning_rate": 5.777958675631657e-06,
+      "loss": 1.008,
+      "step": 896
+    },
+    {
+      "epoch": 0.6502355926060167,
+      "grad_norm": 1.9381106844139175,
+      "learning_rate": 5.756670005121987e-06,
+      "loss": 0.9595,
+      "step": 897
+    },
+    {
+      "epoch": 0.6509604929322218,
+      "grad_norm": 2.0173878555394604,
+      "learning_rate": 5.735404763061303e-06,
+      "loss": 0.9614,
+      "step": 898
+    },
+    {
+      "epoch": 0.651685393258427,
+      "grad_norm": 2.159725328780946,
+      "learning_rate": 5.714163066860132e-06,
+      "loss": 0.9797,
+      "step": 899
+    },
+    {
+      "epoch": 0.6524102935846321,
+      "grad_norm": 2.0644994410010713,
+      "learning_rate": 5.6929450337990175e-06,
+      "loss": 0.9167,
+      "step": 900
+    },
+    {
+      "epoch": 0.6531351939108373,
+      "grad_norm": 2.019906324854319,
+      "learning_rate": 5.671750781027836e-06,
+      "loss": 0.9609,
+      "step": 901
+    },
+    {
+      "epoch": 0.6538600942370424,
+      "grad_norm": 2.774207911473111,
+      "learning_rate": 5.650580425565166e-06,
+      "loss": 0.9905,
+      "step": 902
+    },
+    {
+      "epoch": 0.6545849945632476,
+      "grad_norm": 1.9230557138061348,
+      "learning_rate": 5.629434084297654e-06,
+      "loss": 1.053,
+      "step": 903
+    },
+    {
+      "epoch": 0.6553098948894527,
+      "grad_norm": 1.6916309404007928,
+      "learning_rate": 5.60831187397935e-06,
+      "loss": 0.9022,
+      "step": 904
+    },
+    {
+      "epoch": 0.6560347952156579,
+      "grad_norm": 2.292413850175932,
+      "learning_rate": 5.58721391123108e-06,
+      "loss": 0.9563,
+      "step": 905
+    },
+    {
+      "epoch": 0.656759695541863,
+      "grad_norm": 2.187934211851262,
+      "learning_rate": 5.566140312539786e-06,
+      "loss": 1.0045,
+      "step": 906
+    },
+    {
+      "epoch": 0.6574845958680682,
+      "grad_norm": 2.2765131164518726,
+      "learning_rate": 5.545091194257892e-06,
+      "loss": 1.121,
+      "step": 907
+    },
+    {
+      "epoch": 0.6582094961942733,
+      "grad_norm": 4.530518710071565,
+      "learning_rate": 5.524066672602662e-06,
+      "loss": 1.1574,
+      "step": 908
+    },
+    {
+      "epoch": 0.6589343965204785,
+      "grad_norm": 2.964421873486366,
+      "learning_rate": 5.503066863655546e-06,
+      "loss": 0.9753,
+      "step": 909
+    },
+    {
+      "epoch": 0.6596592968466836,
+      "grad_norm": 1.7482124392759353,
+      "learning_rate": 5.482091883361571e-06,
+      "loss": 0.9915,
+      "step": 910
+    },
+    {
+      "epoch": 0.6603841971728888,
+      "grad_norm": 2.224885148056625,
+      "learning_rate": 5.461141847528655e-06,
+      "loss": 0.9071,
+      "step": 911
+    },
+    {
+      "epoch": 0.6611090974990939,
+      "grad_norm": 3.186202824223461,
+      "learning_rate": 5.440216871827004e-06,
+      "loss": 1.0746,
+      "step": 912
+    },
+    {
+      "epoch": 0.6618339978252991,
+      "grad_norm": 1.7683023401240061,
+      "learning_rate": 5.419317071788461e-06,
+      "loss": 0.9636,
+      "step": 913
+    },
+    {
+      "epoch": 0.6625588981515041,
+      "grad_norm": 3.098972560392732,
+      "learning_rate": 5.398442562805859e-06,
+      "loss": 1.1527,
+      "step": 914
+    },
+    {
+      "epoch": 0.6632837984777094,
+      "grad_norm": 2.2445548524254946,
+      "learning_rate": 5.3775934601324094e-06,
+      "loss": 1.0563,
+      "step": 915
+    },
+    {
+      "epoch": 0.6640086988039144,
+      "grad_norm": 1.739464246848828,
+      "learning_rate": 5.3567698788810366e-06,
+      "loss": 1.0422,
+      "step": 916
+    },
+    {
+      "epoch": 0.6647335991301196,
+      "grad_norm": 2.0974710880958485,
+      "learning_rate": 5.335971934023757e-06,
+      "loss": 0.899,
+      "step": 917
+    },
+    {
+      "epoch": 0.6654584994563247,
+      "grad_norm": 2.029612124614373,
+      "learning_rate": 5.315199740391044e-06,
+      "loss": 0.9541,
+      "step": 918
+    },
+    {
+      "epoch": 0.6661833997825299,
+      "grad_norm": 2.2132638391788158,
+      "learning_rate": 5.294453412671192e-06,
+      "loss": 1.0131,
+      "step": 919
+    },
+    {
+      "epoch": 0.666908300108735,
+      "grad_norm": 2.5732318578474107,
+      "learning_rate": 5.273733065409683e-06,
+      "loss": 0.9976,
+      "step": 920
+    },
+    {
+      "epoch": 0.6676332004349402,
+      "grad_norm": 2.265338233038655,
+      "learning_rate": 5.2530388130085555e-06,
+      "loss": 0.9039,
+      "step": 921
+    },
+    {
+      "epoch": 0.6683581007611453,
+      "grad_norm": 2.006264114627049,
+      "learning_rate": 5.232370769725769e-06,
+      "loss": 0.9893,
+      "step": 922
+    },
+    {
+      "epoch": 0.6690830010873505,
+      "grad_norm": 2.0904960315209076,
+      "learning_rate": 5.2117290496745764e-06,
+      "loss": 0.9729,
+      "step": 923
+    },
+    {
+      "epoch": 0.6698079014135556,
+      "grad_norm": 1.943083190146926,
+      "learning_rate": 5.191113766822905e-06,
+      "loss": 0.9807,
+      "step": 924
+    },
+    {
+      "epoch": 0.6705328017397608,
+      "grad_norm": 2.661838718236442,
+      "learning_rate": 5.1705250349927004e-06,
+      "loss": 0.9773,
+      "step": 925
+    },
+    {
+      "epoch": 0.6712577020659659,
+      "grad_norm": 2.462190516757415,
+      "learning_rate": 5.149962967859321e-06,
+      "loss": 1.0769,
+      "step": 926
+    },
+    {
+      "epoch": 0.6719826023921711,
+      "grad_norm": 2.142823487160251,
+      "learning_rate": 5.129427678950902e-06,
+      "loss": 0.938,
+      "step": 927
+    },
+    {
+      "epoch": 0.6727075027183762,
+      "grad_norm": 1.972351101050343,
+      "learning_rate": 5.108919281647728e-06,
+      "loss": 0.9529,
+      "step": 928
+    },
+    {
+      "epoch": 0.6734324030445814,
+      "grad_norm": 1.9370633685514163,
+      "learning_rate": 5.088437889181612e-06,
+      "loss": 1.075,
+      "step": 929
+    },
+    {
+      "epoch": 0.6741573033707865,
+      "grad_norm": 1.975590280335086,
+      "learning_rate": 5.067983614635263e-06,
+      "loss": 0.9653,
+      "step": 930
+    },
+    {
+      "epoch": 0.6748822036969917,
+      "grad_norm": 2.187164922126263,
+      "learning_rate": 5.04755657094167e-06,
+      "loss": 1.1111,
+      "step": 931
+    },
+    {
+      "epoch": 0.6756071040231968,
+      "grad_norm": 1.9292793498955334,
+      "learning_rate": 5.027156870883468e-06,
+      "loss": 0.9759,
+      "step": 932
+    },
+    {
+      "epoch": 0.676332004349402,
+      "grad_norm": 1.9011667121841174,
+      "learning_rate": 5.006784627092322e-06,
+      "loss": 0.9786,
+      "step": 933
+    },
+    {
+      "epoch": 0.6770569046756071,
+      "grad_norm": 1.8408728752655816,
+      "learning_rate": 4.986439952048313e-06,
+      "loss": 0.8755,
+      "step": 934
+    },
+    {
+      "epoch": 0.6777818050018123,
+      "grad_norm": 2.2274071731605747,
+      "learning_rate": 4.9661229580792995e-06,
+      "loss": 0.9514,
+      "step": 935
+    },
+    {
+      "epoch": 0.6785067053280174,
+      "grad_norm": 2.11818671076986,
+      "learning_rate": 4.945833757360308e-06,
+      "loss": 0.9173,
+      "step": 936
+    },
+    {
+      "epoch": 0.6792316056542226,
+      "grad_norm": 2.244914641056037,
+      "learning_rate": 4.925572461912912e-06,
+      "loss": 0.9013,
+      "step": 937
+    },
+    {
+      "epoch": 0.6799565059804277,
+      "grad_norm": 2.175683562082561,
+      "learning_rate": 4.905339183604614e-06,
+      "loss": 1.0385,
+      "step": 938
+    },
+    {
+      "epoch": 0.6806814063066329,
+      "grad_norm": 1.8770079340690111,
+      "learning_rate": 4.885134034148225e-06,
+      "loss": 0.9151,
+      "step": 939
+    },
+    {
+      "epoch": 0.681406306632838,
+      "grad_norm": 2.071578077945829,
+      "learning_rate": 4.864957125101254e-06,
+      "loss": 1.0396,
+      "step": 940
+    },
+    {
+      "epoch": 0.6821312069590432,
+      "grad_norm": 2.6472781206388203,
+      "learning_rate": 4.844808567865283e-06,
+      "loss": 1.0126,
+      "step": 941
+    },
+    {
+      "epoch": 0.6828561072852483,
+      "grad_norm": 1.9436626348748187,
+      "learning_rate": 4.824688473685362e-06,
+      "loss": 1.0218,
+      "step": 942
+    },
+    {
+      "epoch": 0.6835810076114535,
+      "grad_norm": 2.986623329743054,
+      "learning_rate": 4.804596953649381e-06,
+      "loss": 1.0179,
+      "step": 943
+    },
+    {
+      "epoch": 0.6843059079376586,
+      "grad_norm": 2.0261869595811843,
+      "learning_rate": 4.784534118687477e-06,
+      "loss": 0.9552,
+      "step": 944
+    },
+    {
+      "epoch": 0.6850308082638638,
+      "grad_norm": 2.084576288901219,
+      "learning_rate": 4.764500079571403e-06,
+      "loss": 0.9645,
+      "step": 945
+    },
+    {
+      "epoch": 0.6857557085900688,
+      "grad_norm": 1.941612444910497,
+      "learning_rate": 4.744494946913923e-06,
+      "loss": 0.9273,
+      "step": 946
+    },
+    {
+      "epoch": 0.686480608916274,
+      "grad_norm": 2.3028054945708005,
+      "learning_rate": 4.724518831168203e-06,
+      "loss": 0.8653,
+      "step": 947
+    },
+    {
+      "epoch": 0.6872055092424791,
+      "grad_norm": 2.587974290597117,
+      "learning_rate": 4.704571842627196e-06,
+      "loss": 0.9328,
+      "step": 948
+    },
+    {
+      "epoch": 0.6879304095686843,
+      "grad_norm": 2.745061826423869,
+      "learning_rate": 4.684654091423043e-06,
+      "loss": 1.011,
+      "step": 949
+    },
+    {
+      "epoch": 0.6886553098948894,
+      "grad_norm": 2.084724138668754,
+      "learning_rate": 4.664765687526455e-06,
+      "loss": 1.031,
+      "step": 950
+    },
+    {
+      "epoch": 0.6893802102210946,
+      "grad_norm": 2.409381235205495,
+      "learning_rate": 4.644906740746108e-06,
+      "loss": 0.9139,
+      "step": 951
+    },
+    {
+      "epoch": 0.6901051105472997,
+      "grad_norm": 2.3347733533309905,
+      "learning_rate": 4.6250773607280375e-06,
+      "loss": 0.9462,
+      "step": 952
+    },
+    {
+      "epoch": 0.6908300108735049,
+      "grad_norm": 2.6210504366809095,
+      "learning_rate": 4.605277656955043e-06,
+      "loss": 1.0772,
+      "step": 953
+    },
+    {
+      "epoch": 0.69155491119971,
+      "grad_norm": 1.971545084083224,
+      "learning_rate": 4.585507738746064e-06,
+      "loss": 0.9594,
+      "step": 954
+    },
+    {
+      "epoch": 0.6922798115259152,
+      "grad_norm": 2.369021752491031,
+      "learning_rate": 4.5657677152555925e-06,
+      "loss": 1.0682,
+      "step": 955
+    },
+    {
+      "epoch": 0.6930047118521203,
+      "grad_norm": 2.714912007155751,
+      "learning_rate": 4.5460576954730585e-06,
+      "loss": 1.0267,
+      "step": 956
+    },
+    {
+      "epoch": 0.6937296121783255,
+      "grad_norm": 2.0355578348064727,
+      "learning_rate": 4.526377788222239e-06,
+      "loss": 1.0435,
+      "step": 957
+    },
+    {
+      "epoch": 0.6944545125045306,
+      "grad_norm": 2.1978211554878953,
+      "learning_rate": 4.506728102160653e-06,
+      "loss": 1.008,
+      "step": 958
+    },
+    {
+      "epoch": 0.6951794128307358,
+      "grad_norm": 2.007198364190231,
+      "learning_rate": 4.487108745778958e-06,
+      "loss": 1.0288,
+      "step": 959
+    },
+    {
+      "epoch": 0.6959043131569409,
+      "grad_norm": 2.019964605246266,
+      "learning_rate": 4.467519827400357e-06,
+      "loss": 1.1013,
+      "step": 960
+    },
+    {
+      "epoch": 0.6966292134831461,
+      "grad_norm": 2.0193639156586136,
+      "learning_rate": 4.447961455179995e-06,
+      "loss": 0.9918,
+      "step": 961
+    },
+    {
+      "epoch": 0.6973541138093512,
+      "grad_norm": 1.9471456399675218,
+      "learning_rate": 4.428433737104362e-06,
+      "loss": 0.9593,
+      "step": 962
+    },
+    {
+      "epoch": 0.6980790141355564,
+      "grad_norm": 2.1469665342189206,
+      "learning_rate": 4.408936780990708e-06,
+      "loss": 0.8953,
+      "step": 963
+    },
+    {
+      "epoch": 0.6988039144617615,
+      "grad_norm": 2.5540392388335453,
+      "learning_rate": 4.3894706944864305e-06,
+      "loss": 1.0014,
+      "step": 964
+    },
+    {
+      "epoch": 0.6995288147879667,
+      "grad_norm": 2.8104373863238177,
+      "learning_rate": 4.370035585068493e-06,
+      "loss": 1.1114,
+      "step": 965
+    },
+    {
+      "epoch": 0.7002537151141718,
+      "grad_norm": 2.30574506834477,
+      "learning_rate": 4.350631560042821e-06,
+      "loss": 0.9957,
+      "step": 966
+    },
+    {
+      "epoch": 0.700978615440377,
+      "grad_norm": 2.4431897548258523,
+      "learning_rate": 4.331258726543723e-06,
+      "loss": 1.007,
+      "step": 967
+    },
+    {
+      "epoch": 0.7017035157665821,
+      "grad_norm": 2.2915344664162918,
+      "learning_rate": 4.311917191533288e-06,
+      "loss": 1.022,
+      "step": 968
+    },
+    {
+      "epoch": 0.7024284160927873,
+      "grad_norm": 1.6307353345438238,
+      "learning_rate": 4.292607061800796e-06,
+      "loss": 0.9134,
+      "step": 969
+    },
+    {
+      "epoch": 0.7031533164189924,
+      "grad_norm": 2.094209654163443,
+      "learning_rate": 4.273328443962137e-06,
+      "loss": 1.0199,
+      "step": 970
+    },
+    {
+      "epoch": 0.7038782167451976,
+      "grad_norm": 2.6089783909619397,
+      "learning_rate": 4.254081444459209e-06,
+      "loss": 1.034,
+      "step": 971
+    },
+    {
+      "epoch": 0.7046031170714027,
+      "grad_norm": 2.756310969599043,
+      "learning_rate": 4.234866169559342e-06,
+      "loss": 0.9988,
+      "step": 972
+    },
+    {
+      "epoch": 0.7053280173976079,
+      "grad_norm": 2.1627058695633554,
+      "learning_rate": 4.2156827253547095e-06,
+      "loss": 0.8527,
+      "step": 973
+    },
+    {
+      "epoch": 0.706052917723813,
+      "grad_norm": 2.128748230219529,
+      "learning_rate": 4.1965312177617355e-06,
+      "loss": 0.9118,
+      "step": 974
+    },
+    {
+      "epoch": 0.7067778180500182,
+      "grad_norm": 2.3974561539624117,
+      "learning_rate": 4.177411752520514e-06,
+      "loss": 1.0645,
+      "step": 975
+    },
+    {
+      "epoch": 0.7075027183762232,
+      "grad_norm": 2.3317825972998016,
+      "learning_rate": 4.1583244351942275e-06,
+      "loss": 1.0144,
+      "step": 976
+    },
+    {
+      "epoch": 0.7082276187024285,
+      "grad_norm": 2.2782371205197585,
+      "learning_rate": 4.139269371168562e-06,
+      "loss": 0.9467,
+      "step": 977
+    },
+    {
+      "epoch": 0.7089525190286335,
+      "grad_norm": 1.9702225529384383,
+      "learning_rate": 4.1202466656511245e-06,
+      "loss": 0.9395,
+      "step": 978
+    },
+    {
+      "epoch": 0.7096774193548387,
+      "grad_norm": 2.056174409141514,
+      "learning_rate": 4.101256423670862e-06,
+      "loss": 0.942,
+      "step": 979
+    },
+    {
+      "epoch": 0.7104023196810438,
+      "grad_norm": 2.7573170909057474,
+      "learning_rate": 4.082298750077485e-06,
+      "loss": 0.8837,
+      "step": 980
+    },
+    {
+      "epoch": 0.711127220007249,
+      "grad_norm": 2.79548321777382,
+      "learning_rate": 4.06337374954088e-06,
+      "loss": 0.9387,
+      "step": 981
+    },
+    {
+      "epoch": 0.7118521203334541,
+      "grad_norm": 2.7061864575287875,
+      "learning_rate": 4.04448152655055e-06,
+      "loss": 1.114,
+      "step": 982
+    },
+    {
+      "epoch": 0.7125770206596593,
+      "grad_norm": 2.2826349591175488,
+      "learning_rate": 4.025622185415014e-06,
+      "loss": 0.9945,
+      "step": 983
+    },
+    {
+      "epoch": 0.7133019209858644,
+      "grad_norm": 2.2626152003002176,
+      "learning_rate": 4.006795830261247e-06,
+      "loss": 0.9539,
+      "step": 984
+    },
+    {
+      "epoch": 0.7140268213120696,
+      "grad_norm": 1.9616752319508053,
+      "learning_rate": 3.988002565034096e-06,
+      "loss": 0.9027,
+      "step": 985
+    },
+    {
+      "epoch": 0.7147517216382747,
+      "grad_norm": 2.2013038556275704,
+      "learning_rate": 3.9692424934957175e-06,
+      "loss": 0.9855,
+      "step": 986
+    },
+    {
+      "epoch": 0.7154766219644799,
+      "grad_norm": 2.2452819506033697,
+      "learning_rate": 3.950515719224991e-06,
+      "loss": 0.9616,
+      "step": 987
+    },
+    {
+      "epoch": 0.716201522290685,
+      "grad_norm": 2.298977632173436,
+      "learning_rate": 3.9318223456169604e-06,
+      "loss": 0.9073,
+      "step": 988
+    },
+    {
+      "epoch": 0.7169264226168902,
+      "grad_norm": 2.838544913848763,
+      "learning_rate": 3.91316247588225e-06,
+      "loss": 1.1417,
+      "step": 989
+    },
+    {
+      "epoch": 0.7176513229430953,
+      "grad_norm": 2.131182681333421,
+      "learning_rate": 3.8945362130465035e-06,
+      "loss": 0.9419,
+      "step": 990
+    },
+    {
+      "epoch": 0.7183762232693005,
+      "grad_norm": 2.065261545069331,
+      "learning_rate": 3.8759436599498125e-06,
+      "loss": 0.9205,
+      "step": 991
+    },
+    {
+      "epoch": 0.7191011235955056,
+      "grad_norm": 2.1618911363647086,
+      "learning_rate": 3.857384919246154e-06,
+      "loss": 0.9108,
+      "step": 992
+    },
+    {
+      "epoch": 0.7198260239217108,
+      "grad_norm": 2.7172592088098044,
+      "learning_rate": 3.838860093402813e-06,
+      "loss": 1.0987,
+      "step": 993
+    },
+    {
+      "epoch": 0.7205509242479159,
+      "grad_norm": 2.6032460817441203,
+      "learning_rate": 3.820369284699823e-06,
+      "loss": 0.9295,
+      "step": 994
+    },
+    {
+      "epoch": 0.7212758245741211,
+      "grad_norm": 2.0455131957130717,
+      "learning_rate": 3.8019125952294045e-06,
+      "loss": 0.8968,
+      "step": 995
+    },
+    {
+      "epoch": 0.7220007249003262,
+      "grad_norm": 2.4549976135759866,
+      "learning_rate": 3.783490126895384e-06,
+      "loss": 0.9866,
+      "step": 996
+    },
+    {
+      "epoch": 0.7227256252265314,
+      "grad_norm": 2.38725429694958,
+      "learning_rate": 3.7651019814126656e-06,
+      "loss": 1.0763,
+      "step": 997
+    },
+    {
+      "epoch": 0.7234505255527365,
+      "grad_norm": 2.447633489268824,
+      "learning_rate": 3.746748260306635e-06,
+      "loss": 1.0442,
+      "step": 998
+    },
+    {
+      "epoch": 0.7241754258789417,
+      "grad_norm": 2.015059323195586,
+      "learning_rate": 3.7284290649126176e-06,
+      "loss": 1.0369,
+      "step": 999
+    },
+    {
+      "epoch": 0.7249003262051468,
+      "grad_norm": 2.1439701734436944,
+      "learning_rate": 3.7101444963753096e-06,
+      "loss": 0.9838,
+      "step": 1000
+    },
+    {
+      "epoch": 0.725625226531352,
+      "grad_norm": 2.196165996622997,
+      "learning_rate": 3.691894655648225e-06,
+      "loss": 1.0422,
+      "step": 1001
+    },
+    {
+      "epoch": 0.7263501268575571,
+      "grad_norm": 2.1072970110294498,
+      "learning_rate": 3.6736796434931443e-06,
+      "loss": 1.0379,
+      "step": 1002
+    },
+    {
+      "epoch": 0.7270750271837623,
+      "grad_norm": 2.052889977594413,
+      "learning_rate": 3.6554995604795427e-06,
+      "loss": 0.9731,
+      "step": 1003
+    },
+    {
+      "epoch": 0.7277999275099674,
+      "grad_norm": 2.1849621710221605,
+      "learning_rate": 3.637354506984051e-06,
+      "loss": 0.9397,
+      "step": 1004
+    },
+    {
+      "epoch": 0.7285248278361726,
+      "grad_norm": 2.302003380708732,
+      "learning_rate": 3.619244583189877e-06,
+      "loss": 1.0793,
+      "step": 1005
+    },
+    {
+      "epoch": 0.7292497281623777,
+      "grad_norm": 2.4873231398689692,
+      "learning_rate": 3.6011698890862913e-06,
+      "loss": 0.9777,
+      "step": 1006
+    },
+    {
+      "epoch": 0.7299746284885829,
+      "grad_norm": 2.06017307338732,
+      "learning_rate": 3.583130524468038e-06,
+      "loss": 0.8838,
+      "step": 1007
+    },
+    {
+      "epoch": 0.730699528814788,
+      "grad_norm": 1.9994143545054046,
+      "learning_rate": 3.565126588934803e-06,
+      "loss": 0.9971,
+      "step": 1008
+    },
+    {
+      "epoch": 0.7314244291409931,
+      "grad_norm": 1.8151755246614727,
+      "learning_rate": 3.547158181890661e-06,
+      "loss": 0.9291,
+      "step": 1009
+    },
+    {
+      "epoch": 0.7321493294671982,
+      "grad_norm": 1.9292646278494385,
+      "learning_rate": 3.529225402543518e-06,
+      "loss": 1.0508,
+      "step": 1010
+    },
+    {
+      "epoch": 0.7328742297934034,
+      "grad_norm": 1.9667040655974664,
+      "learning_rate": 3.5113283499045846e-06,
+      "loss": 1.0381,
+      "step": 1011
+    },
+    {
+      "epoch": 0.7335991301196085,
+      "grad_norm": 2.490956546908503,
+      "learning_rate": 3.4934671227878046e-06,
+      "loss": 0.9943,
+      "step": 1012
+    },
+    {
+      "epoch": 0.7343240304458137,
+      "grad_norm": 1.7574463836197582,
+      "learning_rate": 3.475641819809327e-06,
+      "loss": 0.9338,
+      "step": 1013
+    },
+    {
+      "epoch": 0.7350489307720188,
+      "grad_norm": 2.8098119938777795,
+      "learning_rate": 3.4578525393869444e-06,
+      "loss": 1.0599,
+      "step": 1014
+    },
+    {
+      "epoch": 0.735773831098224,
+      "grad_norm": 2.1547229502003877,
+      "learning_rate": 3.4400993797395664e-06,
+      "loss": 0.9071,
+      "step": 1015
+    },
+    {
+      "epoch": 0.7364987314244291,
+      "grad_norm": 2.3529684580357673,
+      "learning_rate": 3.42238243888668e-06,
+      "loss": 0.8903,
+      "step": 1016
+    },
+    {
+      "epoch": 0.7372236317506343,
+      "grad_norm": 3.3656369614826076,
+      "learning_rate": 3.4047018146477873e-06,
+      "loss": 0.9104,
+      "step": 1017
+    },
+    {
+      "epoch": 0.7379485320768394,
+      "grad_norm": 2.3603322866432315,
+      "learning_rate": 3.3870576046418825e-06,
+      "loss": 0.9784,
+      "step": 1018
+    },
+    {
+      "epoch": 0.7386734324030446,
+      "grad_norm": 2.321872140130776,
+      "learning_rate": 3.369449906286907e-06,
+      "loss": 1.0385,
+      "step": 1019
+    },
+    {
+      "epoch": 0.7393983327292497,
+      "grad_norm": 2.1167495374440315,
+      "learning_rate": 3.351878816799209e-06,
+      "loss": 1.0955,
+      "step": 1020
+    },
+    {
+      "epoch": 0.7401232330554549,
+      "grad_norm": 1.786896281579396,
+      "learning_rate": 3.3343444331930207e-06,
+      "loss": 0.991,
+      "step": 1021
+    },
+    {
+      "epoch": 0.74084813338166,
+      "grad_norm": 2.2746758416106427,
+      "learning_rate": 3.316846852279907e-06,
+      "loss": 0.9766,
+      "step": 1022
+    },
+    {
+      "epoch": 0.7415730337078652,
+      "grad_norm": 2.6138994872082097,
+      "learning_rate": 3.2993861706682274e-06,
+      "loss": 1.1535,
+      "step": 1023
+    },
+    {
+      "epoch": 0.7422979340340703,
+      "grad_norm": 2.6165034172242176,
+      "learning_rate": 3.281962484762622e-06,
+      "loss": 0.9753,
+      "step": 1024
+    },
+    {
+      "epoch": 0.7430228343602755,
+      "grad_norm": 1.9155784149193085,
+      "learning_rate": 3.264575890763463e-06,
+      "loss": 0.9675,
+      "step": 1025
+    },
+    {
+      "epoch": 0.7437477346864806,
+      "grad_norm": 2.0624358920096215,
+      "learning_rate": 3.247226484666339e-06,
+      "loss": 1.0538,
+      "step": 1026
+    },
+    {
+      "epoch": 0.7444726350126858,
+      "grad_norm": 3.5846085824449987,
+      "learning_rate": 3.2299143622615036e-06,
+      "loss": 1.107,
+      "step": 1027
+    },
+    {
+      "epoch": 0.7451975353388909,
+      "grad_norm": 2.46962567912047,
+      "learning_rate": 3.2126396191333643e-06,
+      "loss": 0.9327,
+      "step": 1028
+    },
+    {
+      "epoch": 0.7459224356650961,
+      "grad_norm": 2.4139413136075225,
+      "learning_rate": 3.195402350659945e-06,
+      "loss": 1.098,
+      "step": 1029
+    },
+    {
+      "epoch": 0.7466473359913012,
+      "grad_norm": 2.017065330116569,
+      "learning_rate": 3.178202652012363e-06,
+      "loss": 0.9482,
+      "step": 1030
+    },
+    {
+      "epoch": 0.7473722363175064,
+      "grad_norm": 2.404901697722625,
+      "learning_rate": 3.161040618154315e-06,
+      "loss": 1.0855,
+      "step": 1031
+    },
+    {
+      "epoch": 0.7480971366437115,
+      "grad_norm": 1.8467098745774249,
+      "learning_rate": 3.143916343841523e-06,
+      "loss": 1.0062,
+      "step": 1032
+    },
+    {
+      "epoch": 0.7488220369699167,
+      "grad_norm": 2.5314983525811883,
+      "learning_rate": 3.1268299236212395e-06,
+      "loss": 0.8812,
+      "step": 1033
+    },
+    {
+      "epoch": 0.7495469372961218,
+      "grad_norm": 2.3038281210461697,
+      "learning_rate": 3.1097814518317127e-06,
+      "loss": 1.0849,
+      "step": 1034
+    },
+    {
+      "epoch": 0.750271837622327,
+      "grad_norm": 2.4284883217355846,
+      "learning_rate": 3.092771022601676e-06,
+      "loss": 1.0175,
+      "step": 1035
+    },
+    {
+      "epoch": 0.750271837622327,
+      "eval_loss": 1.7681310176849365,
+      "eval_runtime": 1124.5222,
+      "eval_samples_per_second": 15.699,
+      "eval_steps_per_second": 0.245,
+      "step": 1035
+    },
+    {
+      "epoch": 0.7509967379485321,
+      "grad_norm": 2.1284406550026986,
+      "learning_rate": 3.0757987298498106e-06,
+      "loss": 1.0076,
+      "step": 1036
+    },
+    {
+      "epoch": 0.7517216382747373,
+      "grad_norm": 2.22165220425758,
+      "learning_rate": 3.0588646672842404e-06,
+      "loss": 0.951,
+      "step": 1037
+    },
+    {
+      "epoch": 0.7524465386009423,
+      "grad_norm": 2.5711481889523804,
+      "learning_rate": 3.0419689284020126e-06,
+      "loss": 1.0273,
+      "step": 1038
+    },
+    {
+      "epoch": 0.7531714389271476,
+      "grad_norm": 2.0814034949241202,
+      "learning_rate": 3.0251116064885767e-06,
+      "loss": 0.8638,
+      "step": 1039
+    },
+    {
+      "epoch": 0.7538963392533526,
+      "grad_norm": 2.3755493898944664,
+      "learning_rate": 3.0082927946172837e-06,
+      "loss": 0.9276,
+      "step": 1040
+    },
+    {
+      "epoch": 0.7546212395795578,
+      "grad_norm": 1.860045261304839,
+      "learning_rate": 2.9915125856488424e-06,
+      "loss": 0.8793,
+      "step": 1041
+    },
+    {
+      "epoch": 0.7553461399057629,
+      "grad_norm": 2.166915926531366,
+      "learning_rate": 2.974771072230842e-06,
+      "loss": 1.035,
+      "step": 1042
+    },
+    {
+      "epoch": 0.7560710402319681,
+      "grad_norm": 2.1712284867002745,
+      "learning_rate": 2.958068346797217e-06,
+      "loss": 0.9518,
+      "step": 1043
+    },
+    {
+      "epoch": 0.7567959405581732,
+      "grad_norm": 2.888919414338719,
+      "learning_rate": 2.941404501567743e-06,
+      "loss": 1.0718,
+      "step": 1044
+    },
+    {
+      "epoch": 0.7575208408843784,
+      "grad_norm": 2.230289529665858,
+      "learning_rate": 2.9247796285475362e-06,
+      "loss": 1.0045,
+      "step": 1045
+    },
+    {
+      "epoch": 0.7582457412105835,
+      "grad_norm": 2.4680966008183534,
+      "learning_rate": 2.90819381952653e-06,
+      "loss": 1.1622,
+      "step": 1046
+    },
+    {
+      "epoch": 0.7589706415367887,
+      "grad_norm": 2.965957230775603,
+      "learning_rate": 2.8916471660789813e-06,
+      "loss": 0.9385,
+      "step": 1047
+    },
+    {
+      "epoch": 0.7596955418629938,
+      "grad_norm": 2.3702554177333104,
+      "learning_rate": 2.875139759562954e-06,
+      "loss": 0.9725,
+      "step": 1048
+    },
+    {
+      "epoch": 0.760420442189199,
+      "grad_norm": 2.1238051733926113,
+      "learning_rate": 2.8586716911198244e-06,
+      "loss": 0.9945,
+      "step": 1049
+    },
+    {
+      "epoch": 0.7611453425154041,
+      "grad_norm": 1.8445910518867181,
+      "learning_rate": 2.8422430516737733e-06,
+      "loss": 0.959,
+      "step": 1050
+    },
+    {
+      "epoch": 0.7618702428416093,
+      "grad_norm": 2.4547009955302177,
+      "learning_rate": 2.825853931931283e-06,
+      "loss": 0.9879,
+      "step": 1051
+    },
+    {
+      "epoch": 0.7625951431678144,
+      "grad_norm": 2.002478666013241,
+      "learning_rate": 2.80950442238064e-06,
+      "loss": 0.9518,
+      "step": 1052
+    },
+    {
+      "epoch": 0.7633200434940196,
+      "grad_norm": 2.3705241165870197,
+      "learning_rate": 2.793194613291431e-06,
+      "loss": 1.0771,
+      "step": 1053
+    },
+    {
+      "epoch": 0.7640449438202247,
+      "grad_norm": 2.5129264862370957,
+      "learning_rate": 2.7769245947140468e-06,
+      "loss": 1.0603,
+      "step": 1054
+    },
+    {
+      "epoch": 0.7647698441464299,
+      "grad_norm": 1.8777445092267904,
+      "learning_rate": 2.7606944564791914e-06,
+      "loss": 0.8863,
+      "step": 1055
+    },
+    {
+      "epoch": 0.765494744472635,
+      "grad_norm": 2.0066823550902164,
+      "learning_rate": 2.7445042881973727e-06,
+      "loss": 0.9708,
+      "step": 1056
+    },
+    {
+      "epoch": 0.7662196447988402,
+      "grad_norm": 1.9969442657507075,
+      "learning_rate": 2.7283541792584165e-06,
+      "loss": 1.0216,
+      "step": 1057
+    },
+    {
+      "epoch": 0.7669445451250453,
+      "grad_norm": 2.043307797498135,
+      "learning_rate": 2.712244218830973e-06,
+      "loss": 1.0059,
+      "step": 1058
+    },
+    {
+      "epoch": 0.7676694454512505,
+      "grad_norm": 2.05640044579258,
+      "learning_rate": 2.6961744958620195e-06,
+      "loss": 0.977,
+      "step": 1059
+    },
+    {
+      "epoch": 0.7683943457774556,
+      "grad_norm": 2.0705940087415353,
+      "learning_rate": 2.6801450990763767e-06,
+      "loss": 1.0993,
+      "step": 1060
+    },
+    {
+      "epoch": 0.7691192461036608,
+      "grad_norm": 2.2801073634969757,
+      "learning_rate": 2.664156116976212e-06,
+      "loss": 1.1062,
+      "step": 1061
+    },
+    {
+      "epoch": 0.7698441464298659,
+      "grad_norm": 1.8809467398522293,
+      "learning_rate": 2.648207637840554e-06,
+      "loss": 0.9475,
+      "step": 1062
+    },
+    {
+      "epoch": 0.7705690467560711,
+      "grad_norm": 1.9284658912112769,
+      "learning_rate": 2.6322997497248036e-06,
+      "loss": 0.8972,
+      "step": 1063
+    },
+    {
+      "epoch": 0.7712939470822762,
+      "grad_norm": 2.1543128958058384,
+      "learning_rate": 2.616432540460255e-06,
+      "loss": 0.9557,
+      "step": 1064
+    },
+    {
+      "epoch": 0.7720188474084814,
+      "grad_norm": 1.9290768627484227,
+      "learning_rate": 2.6006060976535974e-06,
+      "loss": 0.9678,
+      "step": 1065
+    },
+    {
+      "epoch": 0.7727437477346865,
+      "grad_norm": 3.3836187115327174,
+      "learning_rate": 2.58482050868644e-06,
+      "loss": 1.0705,
+      "step": 1066
+    },
+    {
+      "epoch": 0.7734686480608917,
+      "grad_norm": 2.379431503328967,
+      "learning_rate": 2.5690758607148304e-06,
+      "loss": 1.034,
+      "step": 1067
+    },
+    {
+      "epoch": 0.7741935483870968,
+      "grad_norm": 2.1978701615866667,
+      "learning_rate": 2.553372240668769e-06,
+      "loss": 0.9683,
+      "step": 1068
+    },
+    {
+      "epoch": 0.774918448713302,
+      "grad_norm": 2.334219947775647,
+      "learning_rate": 2.5377097352517332e-06,
+      "loss": 0.9329,
+      "step": 1069
+    },
+    {
+      "epoch": 0.775643349039507,
+      "grad_norm": 1.8982866415446444,
+      "learning_rate": 2.522088430940195e-06,
+      "loss": 0.843,
+      "step": 1070
+    },
+    {
+      "epoch": 0.7763682493657122,
+      "grad_norm": 1.8718401538017733,
+      "learning_rate": 2.5065084139831443e-06,
+      "loss": 0.872,
+      "step": 1071
+    },
+    {
+      "epoch": 0.7770931496919173,
+      "grad_norm": 2.1329592476566663,
+      "learning_rate": 2.4909697704016158e-06,
+      "loss": 0.9499,
+      "step": 1072
+    },
+    {
+      "epoch": 0.7778180500181225,
+      "grad_norm": 2.3116722160450855,
+      "learning_rate": 2.4754725859882077e-06,
+      "loss": 1.071,
+      "step": 1073
+    },
+    {
+      "epoch": 0.7785429503443276,
+      "grad_norm": 2.411486503556151,
+      "learning_rate": 2.4600169463066213e-06,
+      "loss": 0.9443,
+      "step": 1074
+    },
+    {
+      "epoch": 0.7792678506705328,
+      "grad_norm": 1.8955688908325088,
+      "learning_rate": 2.4446029366911708e-06,
+      "loss": 1.0264,
+      "step": 1075
+    },
+    {
+      "epoch": 0.7799927509967379,
+      "grad_norm": 2.2466514583663972,
+      "learning_rate": 2.4292306422463218e-06,
+      "loss": 1.0286,
+      "step": 1076
+    },
+    {
+      "epoch": 0.7807176513229431,
+      "grad_norm": 2.605174653703314,
+      "learning_rate": 2.4139001478462233e-06,
+      "loss": 1.0762,
+      "step": 1077
+    },
+    {
+      "epoch": 0.7814425516491482,
+      "grad_norm": 2.125905431943264,
+      "learning_rate": 2.3986115381342347e-06,
+      "loss": 1.0855,
+      "step": 1078
+    },
+    {
+      "epoch": 0.7821674519753534,
+      "grad_norm": 2.1131544647871685,
+      "learning_rate": 2.3833648975224588e-06,
+      "loss": 0.9731,
+      "step": 1079
+    },
+    {
+      "epoch": 0.7828923523015585,
+      "grad_norm": 2.1666492164090068,
+      "learning_rate": 2.3681603101912785e-06,
+      "loss": 1.1026,
+      "step": 1080
+    },
+    {
+      "epoch": 0.7836172526277637,
+      "grad_norm": 1.902794404555453,
+      "learning_rate": 2.3529978600888915e-06,
+      "loss": 0.9013,
+      "step": 1081
+    },
+    {
+      "epoch": 0.7843421529539688,
+      "grad_norm": 1.904459392353737,
+      "learning_rate": 2.3378776309308414e-06,
+      "loss": 0.9706,
+      "step": 1082
+    },
+    {
+      "epoch": 0.785067053280174,
+      "grad_norm": 1.6870600380072511,
+      "learning_rate": 2.3227997061995622e-06,
+      "loss": 0.8814,
+      "step": 1083
+    },
+    {
+      "epoch": 0.7857919536063791,
+      "grad_norm": 2.034799823983302,
+      "learning_rate": 2.307764169143919e-06,
+      "loss": 0.9905,
+      "step": 1084
+    },
+    {
+      "epoch": 0.7865168539325843,
+      "grad_norm": 2.148503285324599,
+      "learning_rate": 2.292771102778739e-06,
+      "loss": 1.0019,
+      "step": 1085
+    },
+    {
+      "epoch": 0.7872417542587894,
+      "grad_norm": 1.8765819842880695,
+      "learning_rate": 2.277820589884359e-06,
+      "loss": 0.9529,
+      "step": 1086
+    },
+    {
+      "epoch": 0.7879666545849946,
+      "grad_norm": 2.6372186463117666,
+      "learning_rate": 2.26291271300617e-06,
+      "loss": 0.9756,
+      "step": 1087
+    },
+    {
+      "epoch": 0.7886915549111997,
+      "grad_norm": 2.129327741523149,
+      "learning_rate": 2.248047554454157e-06,
+      "loss": 0.9456,
+      "step": 1088
+    },
+    {
+      "epoch": 0.7894164552374049,
+      "grad_norm": 1.9575868058617358,
+      "learning_rate": 2.2332251963024475e-06,
+      "loss": 0.9592,
+      "step": 1089
+    },
+    {
+      "epoch": 0.79014135556361,
+      "grad_norm": 2.239457152716971,
+      "learning_rate": 2.2184457203888598e-06,
+      "loss": 1.0179,
+      "step": 1090
+    },
+    {
+      "epoch": 0.7908662558898152,
+      "grad_norm": 2.0380730008597463,
+      "learning_rate": 2.2037092083144463e-06,
+      "loss": 1.0047,
+      "step": 1091
+    },
+    {
+      "epoch": 0.7915911562160203,
+      "grad_norm": 2.651283946960534,
+      "learning_rate": 2.1890157414430448e-06,
+      "loss": 0.9971,
+      "step": 1092
+    },
+    {
+      "epoch": 0.7923160565422255,
+      "grad_norm": 1.9224733064527182,
+      "learning_rate": 2.1743654009008362e-06,
+      "loss": 0.9758,
+      "step": 1093
+    },
+    {
+      "epoch": 0.7930409568684306,
+      "grad_norm": 2.042210473646339,
+      "learning_rate": 2.1597582675758833e-06,
+      "loss": 0.9157,
+      "step": 1094
+    },
+    {
+      "epoch": 0.7937658571946358,
+      "grad_norm": 1.8843195169105176,
+      "learning_rate": 2.1451944221176966e-06,
+      "loss": 0.8876,
+      "step": 1095
+    },
+    {
+      "epoch": 0.7944907575208409,
+      "grad_norm": 1.7432937559280743,
+      "learning_rate": 2.1306739449367796e-06,
+      "loss": 0.966,
+      "step": 1096
+    },
+    {
+      "epoch": 0.7952156578470461,
+      "grad_norm": 2.4904321442188073,
+      "learning_rate": 2.116196916204193e-06,
+      "loss": 1.0147,
+      "step": 1097
+    },
+    {
+      "epoch": 0.7959405581732512,
+      "grad_norm": 1.867849742985621,
+      "learning_rate": 2.1017634158511037e-06,
+      "loss": 0.8396,
+      "step": 1098
+    },
+    {
+      "epoch": 0.7966654584994564,
+      "grad_norm": 2.176037217510385,
+      "learning_rate": 2.0873735235683535e-06,
+      "loss": 1.0433,
+      "step": 1099
+    },
+    {
+      "epoch": 0.7973903588256614,
+      "grad_norm": 1.8093979197769543,
+      "learning_rate": 2.073027318806008e-06,
+      "loss": 1.0163,
+      "step": 1100
+    },
+    {
+      "epoch": 0.7981152591518667,
+      "grad_norm": 2.317013372213296,
+      "learning_rate": 2.058724880772929e-06,
+      "loss": 0.934,
+      "step": 1101
+    },
+    {
+      "epoch": 0.7988401594780717,
+      "grad_norm": 1.8950042969585994,
+      "learning_rate": 2.044466288436322e-06,
+      "loss": 0.9235,
+      "step": 1102
+    },
+    {
+      "epoch": 0.799565059804277,
+      "grad_norm": 2.26981624646591,
+      "learning_rate": 2.030251620521325e-06,
+      "loss": 0.9664,
+      "step": 1103
+    },
+    {
+      "epoch": 0.800289960130482,
+      "grad_norm": 1.9558043676284143,
+      "learning_rate": 2.0160809555105467e-06,
+      "loss": 1.0178,
+      "step": 1104
+    },
+    {
+      "epoch": 0.8010148604566872,
+      "grad_norm": 2.919669591456051,
+      "learning_rate": 2.0019543716436474e-06,
+      "loss": 0.9538,
+      "step": 1105
+    },
+    {
+      "epoch": 0.8017397607828923,
+      "grad_norm": 2.0335272734209924,
+      "learning_rate": 1.9878719469169104e-06,
+      "loss": 0.8791,
+      "step": 1106
+    },
+    {
+      "epoch": 0.8024646611090975,
+      "grad_norm": 2.171132136552404,
+      "learning_rate": 1.9738337590827918e-06,
+      "loss": 0.9331,
+      "step": 1107
+    },
+    {
+      "epoch": 0.8031895614353026,
+      "grad_norm": 1.942377740059999,
+      "learning_rate": 1.9598398856495227e-06,
+      "loss": 0.8755,
+      "step": 1108
+    },
+    {
+      "epoch": 0.8039144617615078,
+      "grad_norm": 2.135810469663323,
+      "learning_rate": 1.9458904038806537e-06,
+      "loss": 1.0203,
+      "step": 1109
+    },
+    {
+      "epoch": 0.8046393620877129,
+      "grad_norm": 2.3944483070855362,
+      "learning_rate": 1.9319853907946393e-06,
+      "loss": 0.9595,
+      "step": 1110
+    },
+    {
+      "epoch": 0.8053642624139181,
+      "grad_norm": 1.9622747573911234,
+      "learning_rate": 1.918124923164414e-06,
+      "loss": 0.9016,
+      "step": 1111
+    },
+    {
+      "epoch": 0.8060891627401232,
+      "grad_norm": 2.703351031078299,
+      "learning_rate": 1.9043090775169604e-06,
+      "loss": 1.0531,
+      "step": 1112
+    },
+    {
+      "epoch": 0.8068140630663284,
+      "grad_norm": 2.1339961008284063,
+      "learning_rate": 1.890537930132903e-06,
+      "loss": 1.0974,
+      "step": 1113
+    },
+    {
+      "epoch": 0.8075389633925335,
+      "grad_norm": 2.6633749557025896,
+      "learning_rate": 1.8768115570460677e-06,
+      "loss": 1.0197,
+      "step": 1114
+    },
+    {
+      "epoch": 0.8082638637187387,
+      "grad_norm": 1.7899870312478274,
+      "learning_rate": 1.8631300340430747e-06,
+      "loss": 1.0164,
+      "step": 1115
+    },
+    {
+      "epoch": 0.8089887640449438,
+      "grad_norm": 1.786288489816615,
+      "learning_rate": 1.8494934366629057e-06,
+      "loss": 1.0085,
+      "step": 1116
+    },
+    {
+      "epoch": 0.809713664371149,
+      "grad_norm": 2.213791559175714,
+      "learning_rate": 1.8359018401965156e-06,
+      "loss": 1.0711,
+      "step": 1117
+    },
+    {
+      "epoch": 0.8104385646973541,
+      "grad_norm": 2.227118578843932,
+      "learning_rate": 1.8223553196863852e-06,
+      "loss": 1.0198,
+      "step": 1118
+    },
+    {
+      "epoch": 0.8111634650235593,
+      "grad_norm": 1.9847568274305198,
+      "learning_rate": 1.808853949926125e-06,
+      "loss": 0.9065,
+      "step": 1119
+    },
+    {
+      "epoch": 0.8118883653497644,
+      "grad_norm": 1.9370018576926253,
+      "learning_rate": 1.795397805460053e-06,
+      "loss": 0.9994,
+      "step": 1120
+    },
+    {
+      "epoch": 0.8126132656759696,
+      "grad_norm": 2.2601268831652424,
+      "learning_rate": 1.781986960582789e-06,
+      "loss": 0.9872,
+      "step": 1121
+    },
+    {
+      "epoch": 0.8133381660021747,
+      "grad_norm": 2.3214803090685265,
+      "learning_rate": 1.7686214893388509e-06,
+      "loss": 0.9926,
+      "step": 1122
+    },
+    {
+      "epoch": 0.8140630663283799,
+      "grad_norm": 2.5777964336742647,
+      "learning_rate": 1.7553014655222255e-06,
+      "loss": 0.9657,
+      "step": 1123
+    },
+    {
+      "epoch": 0.814787966654585,
+      "grad_norm": 1.9895856551523463,
+      "learning_rate": 1.7420269626759822e-06,
+      "loss": 0.8018,
+      "step": 1124
+    },
+    {
+      "epoch": 0.8155128669807902,
+      "grad_norm": 2.071389074929151,
+      "learning_rate": 1.7287980540918548e-06,
+      "loss": 1.1168,
+      "step": 1125
+    },
+    {
+      "epoch": 0.8162377673069953,
+      "grad_norm": 2.3768284586683786,
+      "learning_rate": 1.7156148128098337e-06,
+      "loss": 1.0561,
+      "step": 1126
+    },
+    {
+      "epoch": 0.8169626676332005,
+      "grad_norm": 2.108829554647349,
+      "learning_rate": 1.7024773116177839e-06,
+      "loss": 1.0184,
+      "step": 1127
+    },
+    {
+      "epoch": 0.8176875679594056,
+      "grad_norm": 1.9145136763918424,
+      "learning_rate": 1.6893856230510152e-06,
+      "loss": 1.0269,
+      "step": 1128
+    },
+    {
+      "epoch": 0.8184124682856108,
+      "grad_norm": 1.984940928437078,
+      "learning_rate": 1.6763398193919034e-06,
+      "loss": 0.9718,
+      "step": 1129
+    },
+    {
+      "epoch": 0.8191373686118159,
+      "grad_norm": 2.491917383583085,
+      "learning_rate": 1.6633399726694755e-06,
+      "loss": 0.9837,
+      "step": 1130
+    },
+    {
+      "epoch": 0.819862268938021,
+      "grad_norm": 2.1889459447102295,
+      "learning_rate": 1.6503861546590216e-06,
+      "loss": 0.8859,
+      "step": 1131
+    },
+    {
+      "epoch": 0.8205871692642261,
+      "grad_norm": 2.2855735715808807,
+      "learning_rate": 1.6374784368817032e-06,
+      "loss": 0.8568,
+      "step": 1132
+    },
+    {
+      "epoch": 0.8213120695904313,
+      "grad_norm": 2.4119276770093245,
+      "learning_rate": 1.6246168906041405e-06,
+      "loss": 1.0357,
+      "step": 1133
+    },
+    {
+      "epoch": 0.8220369699166364,
+      "grad_norm": 2.2497941042194207,
+      "learning_rate": 1.6118015868380387e-06,
+      "loss": 1.0791,
+      "step": 1134
+    },
+    {
+      "epoch": 0.8227618702428416,
+      "grad_norm": 2.6543479685385036,
+      "learning_rate": 1.5990325963397779e-06,
+      "loss": 1.0276,
+      "step": 1135
+    },
+    {
+      "epoch": 0.8234867705690467,
+      "grad_norm": 1.9249733038058623,
+      "learning_rate": 1.5863099896100343e-06,
+      "loss": 0.966,
+      "step": 1136
+    },
+    {
+      "epoch": 0.8242116708952519,
+      "grad_norm": 2.2761621468500497,
+      "learning_rate": 1.5736338368933978e-06,
+      "loss": 0.8657,
+      "step": 1137
+    },
+    {
+      "epoch": 0.824936571221457,
+      "grad_norm": 1.9374428897603755,
+      "learning_rate": 1.561004208177962e-06,
+      "loss": 1.0411,
+      "step": 1138
+    },
+    {
+      "epoch": 0.8256614715476622,
+      "grad_norm": 2.13221264820333,
+      "learning_rate": 1.5484211731949572e-06,
+      "loss": 1.0386,
+      "step": 1139
+    },
+    {
+      "epoch": 0.8263863718738673,
+      "grad_norm": 2.1816254648738194,
+      "learning_rate": 1.5358848014183547e-06,
+      "loss": 1.0673,
+      "step": 1140
+    },
+    {
+      "epoch": 0.8271112722000725,
+      "grad_norm": 2.333905932739388,
+      "learning_rate": 1.523395162064486e-06,
+      "loss": 0.8971,
+      "step": 1141
+    },
+    {
+      "epoch": 0.8278361725262776,
+      "grad_norm": 1.9732510886779988,
+      "learning_rate": 1.5109523240916702e-06,
+      "loss": 0.9907,
+      "step": 1142
+    },
+    {
+      "epoch": 0.8285610728524828,
+      "grad_norm": 2.0768050361793606,
+      "learning_rate": 1.4985563561998185e-06,
+      "loss": 1.0126,
+      "step": 1143
+    },
+    {
+      "epoch": 0.8292859731786879,
+      "grad_norm": 2.0071930948425933,
+      "learning_rate": 1.4862073268300558e-06,
+      "loss": 0.9991,
+      "step": 1144
+    },
+    {
+      "epoch": 0.8300108735048931,
+      "grad_norm": 2.0782467893964593,
+      "learning_rate": 1.4739053041643536e-06,
+      "loss": 0.9887,
+      "step": 1145
+    },
+    {
+      "epoch": 0.8307357738310982,
+      "grad_norm": 2.171294999994677,
+      "learning_rate": 1.4616503561251538e-06,
+      "loss": 0.9402,
+      "step": 1146
+    },
+    {
+      "epoch": 0.8314606741573034,
+      "grad_norm": 2.407000985423698,
+      "learning_rate": 1.4494425503749788e-06,
+      "loss": 0.981,
+      "step": 1147
+    },
+    {
+      "epoch": 0.8321855744835085,
+      "grad_norm": 1.9223688184324994,
+      "learning_rate": 1.437281954316071e-06,
+      "loss": 0.9358,
+      "step": 1148
+    },
+    {
+      "epoch": 0.8329104748097137,
+      "grad_norm": 1.9968819662132635,
+      "learning_rate": 1.4251686350900152e-06,
+      "loss": 0.9298,
+      "step": 1149
+    },
+    {
+      "epoch": 0.8336353751359188,
+      "grad_norm": 2.136971059899701,
+      "learning_rate": 1.4131026595773689e-06,
+      "loss": 0.9215,
+      "step": 1150
+    },
+    {
+      "epoch": 0.834360275462124,
+      "grad_norm": 1.8337681676570194,
+      "learning_rate": 1.4010840943972982e-06,
+      "loss": 1.0268,
+      "step": 1151
+    },
+    {
+      "epoch": 0.8350851757883291,
+      "grad_norm": 2.1977568849453744,
+      "learning_rate": 1.3891130059072032e-06,
+      "loss": 0.9156,
+      "step": 1152
+    },
+    {
+      "epoch": 0.8358100761145343,
+      "grad_norm": 2.255344231343079,
+      "learning_rate": 1.3771894602023483e-06,
+      "loss": 1.087,
+      "step": 1153
+    },
+    {
+      "epoch": 0.8365349764407394,
+      "grad_norm": 2.2385193595399606,
+      "learning_rate": 1.3653135231155079e-06,
+      "loss": 1.0131,
+      "step": 1154
+    },
+    {
+      "epoch": 0.8372598767669446,
+      "grad_norm": 1.9074930266008068,
+      "learning_rate": 1.353485260216596e-06,
+      "loss": 1.046,
+      "step": 1155
+    },
+    {
+      "epoch": 0.8379847770931497,
+      "grad_norm": 2.3457699506131124,
+      "learning_rate": 1.3417047368123103e-06,
+      "loss": 1.0063,
+      "step": 1156
+    },
+    {
+      "epoch": 0.8387096774193549,
+      "grad_norm": 2.195479709372169,
+      "learning_rate": 1.329972017945761e-06,
+      "loss": 0.9382,
+      "step": 1157
+    },
+    {
+      "epoch": 0.83943457774556,
+      "grad_norm": 2.815541285577747,
+      "learning_rate": 1.3182871683961217e-06,
+      "loss": 0.9169,
+      "step": 1158
+    },
+    {
+      "epoch": 0.8401594780717652,
+      "grad_norm": 2.473680572699823,
+      "learning_rate": 1.3066502526782653e-06,
+      "loss": 0.984,
+      "step": 1159
+    },
+    {
+      "epoch": 0.8408843783979703,
+      "grad_norm": 2.3309597371635995,
+      "learning_rate": 1.2950613350424113e-06,
+      "loss": 0.9095,
+      "step": 1160
+    },
+    {
+      "epoch": 0.8416092787241755,
+      "grad_norm": 2.6521093626394148,
+      "learning_rate": 1.2835204794737755e-06,
+      "loss": 0.9806,
+      "step": 1161
+    },
+    {
+      "epoch": 0.8423341790503805,
+      "grad_norm": 2.2183368928655134,
+      "learning_rate": 1.272027749692203e-06,
+      "loss": 1.0063,
+      "step": 1162
+    },
+    {
+      "epoch": 0.8430590793765858,
+      "grad_norm": 2.335358593922603,
+      "learning_rate": 1.2605832091518277e-06,
+      "loss": 1.0009,
+      "step": 1163
+    },
+    {
+      "epoch": 0.8437839797027908,
+      "grad_norm": 3.1347973939628586,
+      "learning_rate": 1.2491869210407215e-06,
+      "loss": 1.0215,
+      "step": 1164
+    },
+    {
+      "epoch": 0.844508880028996,
+      "grad_norm": 2.2148770630869667,
+      "learning_rate": 1.2378389482805397e-06,
+      "loss": 0.9492,
+      "step": 1165
+    },
+    {
+      "epoch": 0.8452337803552011,
+      "grad_norm": 1.8250263747397901,
+      "learning_rate": 1.2265393535261827e-06,
+      "loss": 0.9507,
+      "step": 1166
+    },
+    {
+      "epoch": 0.8459586806814063,
+      "grad_norm": 2.5651175153179806,
+      "learning_rate": 1.2152881991654364e-06,
+      "loss": 1.0011,
+      "step": 1167
+    },
+    {
+      "epoch": 0.8466835810076114,
+      "grad_norm": 1.9418141541403875,
+      "learning_rate": 1.204085547318642e-06,
+      "loss": 1.0761,
+      "step": 1168
+    },
+    {
+      "epoch": 0.8474084813338166,
+      "grad_norm": 2.0646032693286833,
+      "learning_rate": 1.1929314598383423e-06,
+      "loss": 0.9781,
+      "step": 1169
+    },
+    {
+      "epoch": 0.8481333816600217,
+      "grad_norm": 1.8004575813318966,
+      "learning_rate": 1.1818259983089452e-06,
+      "loss": 0.9592,
+      "step": 1170
+    },
+    {
+      "epoch": 0.8488582819862269,
+      "grad_norm": 2.589648686714714,
+      "learning_rate": 1.1707692240463876e-06,
+      "loss": 1.1313,
+      "step": 1171
+    },
+    {
+      "epoch": 0.849583182312432,
+      "grad_norm": 2.0114408256271075,
+      "learning_rate": 1.1597611980977853e-06,
+      "loss": 0.9651,
+      "step": 1172
+    },
+    {
+      "epoch": 0.8503080826386372,
+      "grad_norm": 1.7560110108173013,
+      "learning_rate": 1.1488019812411067e-06,
+      "loss": 0.969,
+      "step": 1173
+    },
+    {
+      "epoch": 0.8510329829648423,
+      "grad_norm": 2.0223370350736958,
+      "learning_rate": 1.1378916339848323e-06,
+      "loss": 0.896,
+      "step": 1174
+    },
+    {
+      "epoch": 0.8517578832910475,
+      "grad_norm": 2.2941607223768075,
+      "learning_rate": 1.127030216567625e-06,
+      "loss": 1.1139,
+      "step": 1175
+    },
+    {
+      "epoch": 0.8524827836172526,
+      "grad_norm": 1.7519205543151821,
+      "learning_rate": 1.1162177889579906e-06,
+      "loss": 0.9145,
+      "step": 1176
+    },
+    {
+      "epoch": 0.8532076839434578,
+      "grad_norm": 2.0778773139552094,
+      "learning_rate": 1.1054544108539511e-06,
+      "loss": 0.8809,
+      "step": 1177
+    },
+    {
+      "epoch": 0.8539325842696629,
+      "grad_norm": 1.8224740916196427,
+      "learning_rate": 1.0947401416827152e-06,
+      "loss": 0.9089,
+      "step": 1178
+    },
+    {
+      "epoch": 0.8546574845958681,
+      "grad_norm": 3.369653427250154,
+      "learning_rate": 1.0840750406003486e-06,
+      "loss": 0.9113,
+      "step": 1179
+    },
+    {
+      "epoch": 0.8553823849220732,
+      "grad_norm": 3.0341804857536254,
+      "learning_rate": 1.0734591664914463e-06,
+      "loss": 1.0581,
+      "step": 1180
+    },
+    {
+      "epoch": 0.8561072852482784,
+      "grad_norm": 2.176890160993487,
+      "learning_rate": 1.0628925779688148e-06,
+      "loss": 0.8511,
+      "step": 1181
+    },
+    {
+      "epoch": 0.8568321855744835,
+      "grad_norm": 2.505948283667956,
+      "learning_rate": 1.0523753333731357e-06,
+      "loss": 1.023,
+      "step": 1182
+    },
+    {
+      "epoch": 0.8575570859006887,
+      "grad_norm": 2.8808186692504423,
+      "learning_rate": 1.041907490772658e-06,
+      "loss": 0.9036,
+      "step": 1183
+    },
+    {
+      "epoch": 0.8582819862268938,
+      "grad_norm": 2.378412246151113,
+      "learning_rate": 1.031489107962863e-06,
+      "loss": 0.951,
+      "step": 1184
+    },
+    {
+      "epoch": 0.859006886553099,
+      "grad_norm": 2.4071476437909594,
+      "learning_rate": 1.0211202424661604e-06,
+      "loss": 0.9586,
+      "step": 1185
+    },
+    {
+      "epoch": 0.8597317868793041,
+      "grad_norm": 2.6658079886381785,
+      "learning_rate": 1.0108009515315586e-06,
+      "loss": 1.1094,
+      "step": 1186
+    },
+    {
+      "epoch": 0.8604566872055093,
+      "grad_norm": 2.2215584462527596,
+      "learning_rate": 1.0005312921343523e-06,
+      "loss": 0.8504,
+      "step": 1187
+    },
+    {
+      "epoch": 0.8611815875317144,
+      "grad_norm": 2.0187808429022107,
+      "learning_rate": 9.903113209758098e-07,
+      "loss": 0.9252,
+      "step": 1188
+    },
+    {
+      "epoch": 0.8619064878579196,
+      "grad_norm": 2.4590642711919006,
+      "learning_rate": 9.801410944828572e-07,
+      "loss": 0.9505,
+      "step": 1189
+    },
+    {
+      "epoch": 0.8626313881841247,
+      "grad_norm": 1.9067553771838575,
+      "learning_rate": 9.700206688077707e-07,
+      "loss": 0.9373,
+      "step": 1190
+    },
+    {
+      "epoch": 0.8633562885103299,
+      "grad_norm": 2.206921164430631,
+      "learning_rate": 9.599500998278632e-07,
+      "loss": 1.0474,
+      "step": 1191
+    },
+    {
+      "epoch": 0.864081188836535,
+      "grad_norm": 2.2333129821917095,
+      "learning_rate": 9.499294431451755e-07,
+      "loss": 1.0328,
+      "step": 1192
+    },
+    {
+      "epoch": 0.8648060891627402,
+      "grad_norm": 2.482019605419004,
+      "learning_rate": 9.399587540861721e-07,
+      "loss": 0.9418,
+      "step": 1193
+    },
+    {
+      "epoch": 0.8655309894889452,
+      "grad_norm": 1.998307923726419,
+      "learning_rate": 9.300380877014315e-07,
+      "loss": 0.979,
+      "step": 1194
+    },
+    {
+      "epoch": 0.8662558898151504,
+      "grad_norm": 1.9225932853992822,
+      "learning_rate": 9.201674987653509e-07,
+      "loss": 0.9395,
+      "step": 1195
+    },
+    {
+      "epoch": 0.8669807901413555,
+      "grad_norm": 2.9091365349477796,
+      "learning_rate": 9.10347041775832e-07,
+      "loss": 0.9469,
+      "step": 1196
+    },
+    {
+      "epoch": 0.8677056904675607,
+      "grad_norm": 1.9149962880721147,
+      "learning_rate": 9.00576770953987e-07,
+      "loss": 1.0476,
+      "step": 1197
+    },
+    {
+      "epoch": 0.8684305907937658,
+      "grad_norm": 1.9887916713607694,
+      "learning_rate": 8.908567402438417e-07,
+      "loss": 0.8714,
+      "step": 1198
+    },
+    {
+      "epoch": 0.869155491119971,
+      "grad_norm": 2.245816854601276,
+      "learning_rate": 8.811870033120284e-07,
+      "loss": 0.9797,
+      "step": 1199
+    },
+    {
+      "epoch": 0.8698803914461761,
+      "grad_norm": 2.2713836769286595,
+      "learning_rate": 8.715676135475004e-07,
+      "loss": 0.9377,
+      "step": 1200
+    },
+    {
+      "epoch": 0.8706052917723813,
+      "grad_norm": 2.5205170405942905,
+      "learning_rate": 8.6199862406123e-07,
+      "loss": 0.9551,
+      "step": 1201
+    },
+    {
+      "epoch": 0.8713301920985864,
+      "grad_norm": 2.35062405333527,
+      "learning_rate": 8.524800876859162e-07,
+      "loss": 0.997,
+      "step": 1202
+    },
+    {
+      "epoch": 0.8720550924247916,
+      "grad_norm": 2.214767113139861,
+      "learning_rate": 8.430120569756949e-07,
+      "loss": 0.9209,
+      "step": 1203
+    },
+    {
+      "epoch": 0.8727799927509967,
+      "grad_norm": 1.8210526020082338,
+      "learning_rate": 8.335945842058524e-07,
+      "loss": 0.8505,
+      "step": 1204
+    },
+    {
+      "epoch": 0.8735048930772019,
+      "grad_norm": 3.169940456082299,
+      "learning_rate": 8.242277213725258e-07,
+      "loss": 1.0918,
+      "step": 1205
+    },
+    {
+      "epoch": 0.874229793403407,
+      "grad_norm": 2.466935646079598,
+      "learning_rate": 8.149115201924263e-07,
+      "loss": 0.9271,
+      "step": 1206
+    },
+    {
+      "epoch": 0.8749546937296122,
+      "grad_norm": 1.8036637263972901,
+      "learning_rate": 8.056460321025483e-07,
+      "loss": 0.9608,
+      "step": 1207
+    },
+    {
+      "epoch": 0.8756795940558173,
+      "grad_norm": 1.9639970375097486,
+      "learning_rate": 7.964313082598884e-07,
+      "loss": 0.863,
+      "step": 1208
+    },
+    {
+      "epoch": 0.8764044943820225,
+      "grad_norm": 1.8756276620751722,
+      "learning_rate": 7.872673995411606e-07,
+      "loss": 0.8938,
+      "step": 1209
+    },
+    {
+      "epoch": 0.8771293947082276,
+      "grad_norm": 2.2245625435715577,
+      "learning_rate": 7.781543565425153e-07,
+      "loss": 0.8795,
+      "step": 1210
+    },
+    {
+      "epoch": 0.8778542950344328,
+      "grad_norm": 2.5233957402711105,
+      "learning_rate": 7.690922295792647e-07,
+      "loss": 0.9564,
+      "step": 1211
+    },
+    {
+      "epoch": 0.8785791953606379,
+      "grad_norm": 2.325154347559063,
+      "learning_rate": 7.600810686855975e-07,
+      "loss": 0.9919,
+      "step": 1212
+    },
+    {
+      "epoch": 0.8793040956868431,
+      "grad_norm": 2.5256436615222437,
+      "learning_rate": 7.511209236143047e-07,
+      "loss": 1.0318,
+      "step": 1213
+    },
+    {
+      "epoch": 0.8800289960130482,
+      "grad_norm": 2.709568964312284,
+      "learning_rate": 7.422118438365156e-07,
+      "loss": 1.0083,
+      "step": 1214
+    },
+    {
+      "epoch": 0.8807538963392534,
+      "grad_norm": 2.2518723793524384,
+      "learning_rate": 7.333538785414062e-07,
+      "loss": 0.894,
+      "step": 1215
+    },
+    {
+      "epoch": 0.8814787966654585,
+      "grad_norm": 2.443491219954294,
+      "learning_rate": 7.24547076635942e-07,
+      "loss": 0.9463,
+      "step": 1216
+    },
+    {
+      "epoch": 0.8822036969916637,
+      "grad_norm": 2.134388443555549,
+      "learning_rate": 7.157914867446003e-07,
+      "loss": 0.933,
+      "step": 1217
+    },
+    {
+      "epoch": 0.8829285973178688,
+      "grad_norm": 2.512674514907188,
+      "learning_rate": 7.070871572091076e-07,
+      "loss": 0.893,
+      "step": 1218
+    },
+    {
+      "epoch": 0.883653497644074,
+      "grad_norm": 2.453936234251306,
+      "learning_rate": 6.984341360881664e-07,
+      "loss": 0.9721,
+      "step": 1219
+    },
+    {
+      "epoch": 0.8843783979702791,
+      "grad_norm": 2.1489162189409337,
+      "learning_rate": 6.898324711571958e-07,
+      "loss": 0.9212,
+      "step": 1220
+    },
+    {
+      "epoch": 0.8851032982964843,
+      "grad_norm": 2.510668864545298,
+      "learning_rate": 6.812822099080629e-07,
+      "loss": 0.9371,
+      "step": 1221
+    },
+    {
+      "epoch": 0.8858281986226894,
+      "grad_norm": 1.9545987636324476,
+      "learning_rate": 6.727833995488242e-07,
+      "loss": 0.9738,
+      "step": 1222
+    },
+    {
+      "epoch": 0.8865530989488946,
+      "grad_norm": 2.1706699023503604,
+      "learning_rate": 6.64336087003461e-07,
+      "loss": 1.1229,
+      "step": 1223
+    },
+    {
+      "epoch": 0.8872779992750996,
+      "grad_norm": 1.8682535430546066,
+      "learning_rate": 6.559403189116275e-07,
+      "loss": 0.8849,
+      "step": 1224
+    },
+    {
+      "epoch": 0.8880028996013049,
+      "grad_norm": 2.159689783926262,
+      "learning_rate": 6.475961416283838e-07,
+      "loss": 0.9401,
+      "step": 1225
+    },
+    {
+      "epoch": 0.8887277999275099,
+      "grad_norm": 2.7767173791052913,
+      "learning_rate": 6.393036012239473e-07,
+      "loss": 0.8924,
+      "step": 1226
+    },
+    {
+      "epoch": 0.8894527002537151,
+      "grad_norm": 2.4204337218238905,
+      "learning_rate": 6.310627434834349e-07,
+      "loss": 0.9524,
+      "step": 1227
+    },
+    {
+      "epoch": 0.8901776005799202,
+      "grad_norm": 1.8765448611106061,
+      "learning_rate": 6.228736139066105e-07,
+      "loss": 0.859,
+      "step": 1228
+    },
+    {
+      "epoch": 0.8909025009061254,
+      "grad_norm": 1.919204528359367,
+      "learning_rate": 6.147362577076343e-07,
+      "loss": 0.8362,
+      "step": 1229
+    },
+    {
+      "epoch": 0.8916274012323305,
+      "grad_norm": 1.9433910718170642,
+      "learning_rate": 6.066507198148142e-07,
+      "loss": 0.8749,
+      "step": 1230
+    },
+    {
+      "epoch": 0.8923523015585357,
+      "grad_norm": 2.9697875554755884,
+      "learning_rate": 5.986170448703577e-07,
+      "loss": 0.943,
+      "step": 1231
+    },
+    {
+      "epoch": 0.8930772018847408,
+      "grad_norm": 2.353044622771824,
+      "learning_rate": 5.906352772301193e-07,
+      "loss": 1.1478,
+      "step": 1232
+    },
+    {
+      "epoch": 0.893802102210946,
+      "grad_norm": 1.850720536456356,
+      "learning_rate": 5.827054609633686e-07,
+      "loss": 0.9644,
+      "step": 1233
+    },
+    {
+      "epoch": 0.8945270025371511,
+      "grad_norm": 1.8013108857133973,
+      "learning_rate": 5.748276398525332e-07,
+      "loss": 1.0246,
+      "step": 1234
+    },
+    {
+      "epoch": 0.8952519028633563,
+      "grad_norm": 2.469890038185596,
+      "learning_rate": 5.670018573929647e-07,
+      "loss": 0.8624,
+      "step": 1235
+    },
+    {
+      "epoch": 0.8959768031895614,
+      "grad_norm": 2.2167116711922175,
+      "learning_rate": 5.592281567926983e-07,
+      "loss": 0.9879,
+      "step": 1236
+    },
+    {
+      "epoch": 0.8967017035157666,
+      "grad_norm": 2.4449293680021236,
+      "learning_rate": 5.515065809722064e-07,
+      "loss": 0.9898,
+      "step": 1237
+    },
+    {
+      "epoch": 0.8974266038419717,
+      "grad_norm": 2.746110084978734,
+      "learning_rate": 5.438371725641778e-07,
+      "loss": 1.0036,
+      "step": 1238
+    },
+    {
+      "epoch": 0.8981515041681769,
+      "grad_norm": 2.4639823270799983,
+      "learning_rate": 5.362199739132656e-07,
+      "loss": 1.0811,
+      "step": 1239
+    },
+    {
+      "epoch": 0.898876404494382,
+      "grad_norm": 1.9704675962545042,
+      "learning_rate": 5.286550270758617e-07,
+      "loss": 0.8005,
+      "step": 1240
+    },
+    {
+      "epoch": 0.8996013048205872,
+      "grad_norm": 2.2146310697220546,
+      "learning_rate": 5.211423738198618e-07,
+      "loss": 0.9509,
+      "step": 1241
+    },
+    {
+      "epoch": 0.9003262051467923,
+      "grad_norm": 1.8720450871060204,
+      "learning_rate": 5.136820556244393e-07,
+      "loss": 0.9735,
+      "step": 1242
+    },
+    {
+      "epoch": 0.9010511054729975,
+      "grad_norm": 1.8982694817893693,
+      "learning_rate": 5.062741136798111e-07,
+      "loss": 0.9381,
+      "step": 1243
+    },
+    {
+      "epoch": 0.9017760057992026,
+      "grad_norm": 1.858714766448879,
+      "learning_rate": 4.98918588887013e-07,
+      "loss": 0.9523,
+      "step": 1244
+    },
+    {
+      "epoch": 0.9025009061254078,
+      "grad_norm": 2.4016041132668344,
+      "learning_rate": 4.916155218576734e-07,
+      "loss": 1.0073,
+      "step": 1245
+    },
+    {
+      "epoch": 0.9032258064516129,
+      "grad_norm": 2.347954663659916,
+      "learning_rate": 4.843649529137861e-07,
+      "loss": 1.0266,
+      "step": 1246
+    },
+    {
+      "epoch": 0.9039507067778181,
+      "grad_norm": 2.1409896869995193,
+      "learning_rate": 4.771669220874908e-07,
+      "loss": 0.9691,
+      "step": 1247
+    },
+    {
+      "epoch": 0.9046756071040232,
+      "grad_norm": 2.49959014733721,
+      "learning_rate": 4.70021469120856e-07,
+      "loss": 1.0186,
+      "step": 1248
+    },
+    {
+      "epoch": 0.9054005074302284,
+      "grad_norm": 2.1558155360425513,
+      "learning_rate": 4.629286334656502e-07,
+      "loss": 0.9444,
+      "step": 1249
+    },
+    {
+      "epoch": 0.9061254077564335,
+      "grad_norm": 1.6872092574160515,
+      "learning_rate": 4.55888454283131e-07,
+      "loss": 0.8999,
+      "step": 1250
+    },
+    {
+      "epoch": 0.9068503080826387,
+      "grad_norm": 2.4896678483642445,
+      "learning_rate": 4.489009704438274e-07,
+      "loss": 0.9269,
+      "step": 1251
+    },
+    {
+      "epoch": 0.9075752084088438,
+      "grad_norm": 2.2948377479813344,
+      "learning_rate": 4.419662205273234e-07,
+      "loss": 1.0014,
+      "step": 1252
+    },
+    {
+      "epoch": 0.908300108735049,
+      "grad_norm": 2.287777069195488,
+      "learning_rate": 4.35084242822047e-07,
+      "loss": 0.9587,
+      "step": 1253
+    },
+    {
+      "epoch": 0.909025009061254,
+      "grad_norm": 2.32897276572487,
+      "learning_rate": 4.282550753250603e-07,
+      "loss": 0.9474,
+      "step": 1254
+    },
+    {
+      "epoch": 0.9097499093874593,
+      "grad_norm": 2.314915920284867,
+      "learning_rate": 4.2147875574184095e-07,
+      "loss": 0.981,
+      "step": 1255
+    },
+    {
+      "epoch": 0.9104748097136643,
+      "grad_norm": 1.9516096065203952,
+      "learning_rate": 4.1475532148608665e-07,
+      "loss": 0.9547,
+      "step": 1256
+    },
+    {
+      "epoch": 0.9111997100398695,
+      "grad_norm": 2.1927356687145227,
+      "learning_rate": 4.08084809679502e-07,
+      "loss": 0.9606,
+      "step": 1257
+    },
+    {
+      "epoch": 0.9119246103660746,
+      "grad_norm": 2.0841597477886924,
+      "learning_rate": 4.0146725715159096e-07,
+      "loss": 0.9804,
+      "step": 1258
+    },
+    {
+      "epoch": 0.9126495106922798,
+      "grad_norm": 2.5613066364600914,
+      "learning_rate": 3.9490270043945787e-07,
+      "loss": 1.0065,
+      "step": 1259
+    },
+    {
+      "epoch": 0.9133744110184849,
+      "grad_norm": 2.4125648530924497,
+      "learning_rate": 3.883911757876058e-07,
+      "loss": 0.9775,
+      "step": 1260
+    },
+    {
+      "epoch": 0.9140993113446901,
+      "grad_norm": 1.939270842406947,
+      "learning_rate": 3.8193271914773287e-07,
+      "loss": 0.9695,
+      "step": 1261
+    },
+    {
+      "epoch": 0.9148242116708952,
+      "grad_norm": 2.2535052888130336,
+      "learning_rate": 3.7552736617853745e-07,
+      "loss": 0.8691,
+      "step": 1262
+    },
+    {
+      "epoch": 0.9155491119971004,
+      "grad_norm": 2.06134443104641,
+      "learning_rate": 3.6917515224552115e-07,
+      "loss": 0.9972,
+      "step": 1263
+    },
+    {
+      "epoch": 0.9162740123233055,
+      "grad_norm": 2.187718479642105,
+      "learning_rate": 3.628761124207858e-07,
+      "loss": 0.9614,
+      "step": 1264
+    },
+    {
+      "epoch": 0.9169989126495107,
+      "grad_norm": 2.728957931329134,
+      "learning_rate": 3.566302814828526e-07,
+      "loss": 1.0619,
+      "step": 1265
+    },
+    {
+      "epoch": 0.9177238129757158,
+      "grad_norm": 2.295000346358395,
+      "learning_rate": 3.504376939164611e-07,
+      "loss": 1.0473,
+      "step": 1266
+    },
+    {
+      "epoch": 0.918448713301921,
+      "grad_norm": 2.077400704239704,
+      "learning_rate": 3.442983839123826e-07,
+      "loss": 1.0098,
+      "step": 1267
+    },
+    {
+      "epoch": 0.9191736136281261,
+      "grad_norm": 2.3680349659705016,
+      "learning_rate": 3.382123853672303e-07,
+      "loss": 1.0091,
+      "step": 1268
+    },
+    {
+      "epoch": 0.9198985139543313,
+      "grad_norm": 2.392219243933128,
+      "learning_rate": 3.321797318832687e-07,
+      "loss": 1.0328,
+      "step": 1269
+    },
+    {
+      "epoch": 0.9206234142805364,
+      "grad_norm": 1.933874299943873,
+      "learning_rate": 3.262004567682342e-07,
+      "loss": 0.9634,
+      "step": 1270
+    },
+    {
+      "epoch": 0.9213483146067416,
+      "grad_norm": 2.1317697728223997,
+      "learning_rate": 3.20274593035147e-07,
+      "loss": 0.9766,
+      "step": 1271
+    },
+    {
+      "epoch": 0.9220732149329467,
+      "grad_norm": 1.7882374391836628,
+      "learning_rate": 3.144021734021352e-07,
+      "loss": 0.9166,
+      "step": 1272
+    },
+    {
+      "epoch": 0.9227981152591519,
+      "grad_norm": 2.053906284134757,
+      "learning_rate": 3.085832302922398e-07,
+      "loss": 0.9921,
+      "step": 1273
+    },
+    {
+      "epoch": 0.923523015585357,
+      "grad_norm": 2.130994971696479,
+      "learning_rate": 3.028177958332512e-07,
+      "loss": 0.9309,
+      "step": 1274
+    },
+    {
+      "epoch": 0.9242479159115622,
+      "grad_norm": 2.070911861584657,
+      "learning_rate": 2.971059018575262e-07,
+      "loss": 0.9546,
+      "step": 1275
+    },
+    {
+      "epoch": 0.9249728162377673,
+      "grad_norm": 2.180354582475891,
+      "learning_rate": 2.9144757990180685e-07,
+      "loss": 0.9385,
+      "step": 1276
+    },
+    {
+      "epoch": 0.9256977165639725,
+      "grad_norm": 2.101379134926334,
+      "learning_rate": 2.858428612070585e-07,
+      "loss": 0.9473,
+      "step": 1277
+    },
+    {
+      "epoch": 0.9264226168901776,
+      "grad_norm": 2.941363671739184,
+      "learning_rate": 2.802917767182822e-07,
+      "loss": 1.0476,
+      "step": 1278
+    },
+    {
+      "epoch": 0.9271475172163828,
+      "grad_norm": 2.099272282487294,
+      "learning_rate": 2.74794357084357e-07,
+      "loss": 1.135,
+      "step": 1279
+    },
+    {
+      "epoch": 0.9278724175425879,
+      "grad_norm": 2.0386416225655744,
+      "learning_rate": 2.693506326578632e-07,
+      "loss": 0.9461,
+      "step": 1280
+    },
+    {
+      "epoch": 0.9285973178687931,
+      "grad_norm": 2.158374614724584,
+      "learning_rate": 2.639606334949163e-07,
+      "loss": 0.8921,
+      "step": 1281
+    },
+    {
+      "epoch": 0.9293222181949982,
+      "grad_norm": 1.8765718966068676,
+      "learning_rate": 2.586243893550033e-07,
+      "loss": 0.9305,
+      "step": 1282
+    },
+    {
+      "epoch": 0.9300471185212034,
+      "grad_norm": 2.3160935821533144,
+      "learning_rate": 2.533419297008155e-07,
+      "loss": 0.9176,
+      "step": 1283
+    },
+    {
+      "epoch": 0.9307720188474085,
+      "grad_norm": 2.1878330344349313,
+      "learning_rate": 2.481132836980871e-07,
+      "loss": 0.955,
+      "step": 1284
+    },
+    {
+      "epoch": 0.9314969191736137,
+      "grad_norm": 2.195724525034424,
+      "learning_rate": 2.4293848021543443e-07,
+      "loss": 0.8916,
+      "step": 1285
+    },
+    {
+      "epoch": 0.9322218194998187,
+      "grad_norm": 1.7464619703681252,
+      "learning_rate": 2.378175478241984e-07,
+      "loss": 0.9268,
+      "step": 1286
+    },
+    {
+      "epoch": 0.932946719826024,
+      "grad_norm": 2.5869309705425465,
+      "learning_rate": 2.3275051479828115e-07,
+      "loss": 0.9683,
+      "step": 1287
+    },
+    {
+      "epoch": 0.933671620152229,
+      "grad_norm": 1.9334780055162688,
+      "learning_rate": 2.27737409113995e-07,
+      "loss": 0.9208,
+      "step": 1288
+    },
+    {
+      "epoch": 0.9343965204784342,
+      "grad_norm": 1.8929021569949063,
+      "learning_rate": 2.2277825844990607e-07,
+      "loss": 0.912,
+      "step": 1289
+    },
+    {
+      "epoch": 0.9351214208046393,
+      "grad_norm": 2.2127598690640564,
+      "learning_rate": 2.1787309018668213e-07,
+      "loss": 1.0383,
+      "step": 1290
+    },
+    {
+      "epoch": 0.9358463211308445,
+      "grad_norm": 2.210748449295597,
+      "learning_rate": 2.130219314069415e-07,
+      "loss": 0.9721,
+      "step": 1291
+    },
+    {
+      "epoch": 0.9365712214570496,
+      "grad_norm": 2.115265811779472,
+      "learning_rate": 2.0822480889510222e-07,
+      "loss": 1.0026,
+      "step": 1292
+    },
+    {
+      "epoch": 0.9372961217832548,
+      "grad_norm": 2.040211532276065,
+      "learning_rate": 2.034817491372354e-07,
+      "loss": 0.9751,
+      "step": 1293
+    },
+    {
+      "epoch": 0.9380210221094599,
+      "grad_norm": 2.3955059923006483,
+      "learning_rate": 1.9879277832091758e-07,
+      "loss": 1.0528,
+      "step": 1294
+    },
+    {
+      "epoch": 0.9387459224356651,
+      "grad_norm": 2.0551722313954297,
+      "learning_rate": 1.941579223350898e-07,
+      "loss": 0.9919,
+      "step": 1295
+    },
+    {
+      "epoch": 0.9394708227618702,
+      "grad_norm": 1.8511261527636662,
+      "learning_rate": 1.8957720676991087e-07,
+      "loss": 0.957,
+      "step": 1296
+    },
+    {
+      "epoch": 0.9401957230880754,
+      "grad_norm": 1.986865247006612,
+      "learning_rate": 1.850506569166155e-07,
+      "loss": 1.0243,
+      "step": 1297
+    },
+    {
+      "epoch": 0.9409206234142805,
+      "grad_norm": 1.9332096525646203,
+      "learning_rate": 1.8057829776737867e-07,
+      "loss": 1.0149,
+      "step": 1298
+    },
+    {
+      "epoch": 0.9416455237404857,
+      "grad_norm": 2.0914992940185724,
+      "learning_rate": 1.761601540151736e-07,
+      "loss": 1.0177,
+      "step": 1299
+    },
+    {
+      "epoch": 0.9423704240666908,
+      "grad_norm": 2.4468777213087205,
+      "learning_rate": 1.7179625005364076e-07,
+      "loss": 1.072,
+      "step": 1300
+    },
+    {
+      "epoch": 0.943095324392896,
+      "grad_norm": 1.982397014057623,
+      "learning_rate": 1.6748660997694454e-07,
+      "loss": 0.862,
+      "step": 1301
+    },
+    {
+      "epoch": 0.9438202247191011,
+      "grad_norm": 2.4635792769761005,
+      "learning_rate": 1.6323125757964799e-07,
+      "loss": 0.9996,
+      "step": 1302
+    },
+    {
+      "epoch": 0.9445451250453063,
+      "grad_norm": 2.1226326530971704,
+      "learning_rate": 1.5903021635658044e-07,
+      "loss": 1.0194,
+      "step": 1303
+    },
+    {
+      "epoch": 0.9452700253715114,
+      "grad_norm": 2.2271194959547116,
+      "learning_rate": 1.5488350950270237e-07,
+      "loss": 0.9331,
+      "step": 1304
+    },
+    {
+      "epoch": 0.9459949256977166,
+      "grad_norm": 2.472581051188092,
+      "learning_rate": 1.507911599129841e-07,
+      "loss": 0.9543,
+      "step": 1305
+    },
+    {
+      "epoch": 0.9467198260239217,
+      "grad_norm": 1.872694939056278,
+      "learning_rate": 1.4675319018227608e-07,
+      "loss": 1.0782,
+      "step": 1306
+    },
+    {
+      "epoch": 0.9474447263501269,
+      "grad_norm": 2.0981456286626297,
+      "learning_rate": 1.4276962260518223e-07,
+      "loss": 0.9838,
+      "step": 1307
+    },
+    {
+      "epoch": 0.948169626676332,
+      "grad_norm": 2.1938985740167483,
+      "learning_rate": 1.3884047917594236e-07,
+      "loss": 1.0023,
+      "step": 1308
+    },
+    {
+      "epoch": 0.9488945270025372,
+      "grad_norm": 2.342040537819885,
+      "learning_rate": 1.349657815883032e-07,
+      "loss": 0.9773,
+      "step": 1309
+    },
+    {
+      "epoch": 0.9496194273287423,
+      "grad_norm": 2.0373395983000795,
+      "learning_rate": 1.3114555123540762e-07,
+      "loss": 0.9412,
+      "step": 1310
+    },
+    {
+      "epoch": 0.9503443276549475,
+      "grad_norm": 2.3706302014168714,
+      "learning_rate": 1.2737980920966785e-07,
+      "loss": 1.0296,
+      "step": 1311
+    },
+    {
+      "epoch": 0.9510692279811526,
+      "grad_norm": 2.130493636831118,
+      "learning_rate": 1.236685763026546e-07,
+      "loss": 0.9575,
+      "step": 1312
+    },
+    {
+      "epoch": 0.9517941283073578,
+      "grad_norm": 2.257925035976561,
+      "learning_rate": 1.2001187300498153e-07,
+      "loss": 0.9083,
+      "step": 1313
+    },
+    {
+      "epoch": 0.9525190286335629,
+      "grad_norm": 2.1864022493275783,
+      "learning_rate": 1.1640971950618752e-07,
+      "loss": 0.923,
+      "step": 1314
+    },
+    {
+      "epoch": 0.9532439289597681,
+      "grad_norm": 2.5860606750284734,
+      "learning_rate": 1.1286213569463467e-07,
+      "loss": 1.0475,
+      "step": 1315
+    },
+    {
+      "epoch": 0.9539688292859732,
+      "grad_norm": 2.258441636325009,
+      "learning_rate": 1.0936914115738717e-07,
+      "loss": 0.8783,
+      "step": 1316
+    },
+    {
+      "epoch": 0.9546937296121784,
+      "grad_norm": 1.8730597641684372,
+      "learning_rate": 1.0593075518011143e-07,
+      "loss": 0.9545,
+      "step": 1317
+    },
+    {
+      "epoch": 0.9554186299383834,
+      "grad_norm": 2.3138211525804584,
+      "learning_rate": 1.0254699674696611e-07,
+      "loss": 1.016,
+      "step": 1318
+    },
+    {
+      "epoch": 0.9561435302645886,
+      "grad_norm": 1.9117655582358168,
+      "learning_rate": 9.921788454049896e-08,
+      "loss": 1.0075,
+      "step": 1319
+    },
+    {
+      "epoch": 0.9568684305907937,
+      "grad_norm": 2.0487380925988368,
+      "learning_rate": 9.594343694154129e-08,
+      "loss": 0.9545,
+      "step": 1320
+    },
+    {
+      "epoch": 0.9575933309169989,
+      "grad_norm": 1.808692209003071,
+      "learning_rate": 9.272367202910804e-08,
+      "loss": 0.8925,
+      "step": 1321
+    },
+    {
+      "epoch": 0.958318231243204,
+      "grad_norm": 2.7015235278624194,
+      "learning_rate": 8.9558607580299e-08,
+      "loss": 0.9318,
+      "step": 1322
+    },
+    {
+      "epoch": 0.9590431315694092,
+      "grad_norm": 2.2921044642024757,
+      "learning_rate": 8.644826107019888e-08,
+      "loss": 1.0238,
+      "step": 1323
+    },
+    {
+      "epoch": 0.9597680318956143,
+      "grad_norm": 1.9280021670942513,
+      "learning_rate": 8.339264967178074e-08,
+      "loss": 0.971,
+      "step": 1324
+    },
+    {
+      "epoch": 0.9604929322218195,
+      "grad_norm": 1.8067867464540794,
+      "learning_rate": 8.039179025581379e-08,
+      "loss": 0.8251,
+      "step": 1325
+    },
+    {
+      "epoch": 0.9612178325480246,
+      "grad_norm": 2.7036098510140016,
+      "learning_rate": 7.744569939076574e-08,
+      "loss": 1.0214,
+      "step": 1326
+    },
+    {
+      "epoch": 0.9619427328742298,
+      "grad_norm": 1.7075937015114786,
+      "learning_rate": 7.455439334271619e-08,
+      "loss": 0.9234,
+      "step": 1327
+    },
+    {
+      "epoch": 0.9626676332004349,
+      "grad_norm": 1.8777145955794914,
+      "learning_rate": 7.171788807526336e-08,
+      "loss": 0.8121,
+      "step": 1328
+    },
+    {
+      "epoch": 0.9633925335266401,
+      "grad_norm": 2.0244110553223695,
+      "learning_rate": 6.89361992494375e-08,
+      "loss": 0.9208,
+      "step": 1329
+    },
+    {
+      "epoch": 0.9641174338528452,
+      "grad_norm": 2.542844240335413,
+      "learning_rate": 6.62093422236132e-08,
+      "loss": 1.0046,
+      "step": 1330
+    },
+    {
+      "epoch": 0.9648423341790504,
+      "grad_norm": 2.543683044105189,
+      "learning_rate": 6.353733205342715e-08,
+      "loss": 1.0122,
+      "step": 1331
+    },
+    {
+      "epoch": 0.9655672345052555,
+      "grad_norm": 2.1697924274060245,
+      "learning_rate": 6.092018349169171e-08,
+      "loss": 1.0031,
+      "step": 1332
+    },
+    {
+      "epoch": 0.9662921348314607,
+      "grad_norm": 2.0531262536255985,
+      "learning_rate": 5.8357910988317e-08,
+      "loss": 0.8397,
+      "step": 1333
+    },
+    {
+      "epoch": 0.9670170351576658,
+      "grad_norm": 2.913557846619087,
+      "learning_rate": 5.585052869022556e-08,
+      "loss": 0.9297,
+      "step": 1334
+    },
+    {
+      "epoch": 0.967741935483871,
+      "grad_norm": 2.204845207370084,
+      "learning_rate": 5.339805044128121e-08,
+      "loss": 1.003,
+      "step": 1335
+    },
+    {
+      "epoch": 0.9684668358100761,
+      "grad_norm": 2.644048528912831,
+      "learning_rate": 5.1000489782210284e-08,
+      "loss": 1.0444,
+      "step": 1336
+    },
+    {
+      "epoch": 0.9691917361362813,
+      "grad_norm": 2.402838017572579,
+      "learning_rate": 4.8657859950520524e-08,
+      "loss": 0.9235,
+      "step": 1337
+    },
+    {
+      "epoch": 0.9699166364624864,
+      "grad_norm": 1.9654858262351746,
+      "learning_rate": 4.6370173880438964e-08,
+      "loss": 0.958,
+      "step": 1338
+    },
+    {
+      "epoch": 0.9706415367886916,
+      "grad_norm": 2.7707673392332746,
+      "learning_rate": 4.413744420283195e-08,
+      "loss": 0.8956,
+      "step": 1339
+    },
+    {
+      "epoch": 0.9713664371148967,
+      "grad_norm": 1.9709988493814223,
+      "learning_rate": 4.195968324513744e-08,
+      "loss": 0.9951,
+      "step": 1340
+    },
+    {
+      "epoch": 0.9720913374411019,
+      "grad_norm": 2.1055256777177958,
+      "learning_rate": 3.983690303130172e-08,
+      "loss": 0.9204,
+      "step": 1341
+    },
+    {
+      "epoch": 0.972816237767307,
+      "grad_norm": 2.218819032192361,
+      "learning_rate": 3.776911528170391e-08,
+      "loss": 0.8971,
+      "step": 1342
+    },
+    {
+      "epoch": 0.9735411380935122,
+      "grad_norm": 2.0858300430985146,
+      "learning_rate": 3.575633141310153e-08,
+      "loss": 0.897,
+      "step": 1343
+    },
+    {
+      "epoch": 0.9742660384197173,
+      "grad_norm": 2.3972281459872713,
+      "learning_rate": 3.379856253855951e-08,
+      "loss": 0.9454,
+      "step": 1344
+    },
+    {
+      "epoch": 0.9749909387459225,
+      "grad_norm": 2.488073794270114,
+      "learning_rate": 3.1895819467391286e-08,
+      "loss": 0.8999,
+      "step": 1345
+    },
+    {
+      "epoch": 0.9757158390721276,
+      "grad_norm": 1.988995529917583,
+      "learning_rate": 3.0048112705102215e-08,
+      "loss": 0.971,
+      "step": 1346
+    },
+    {
+      "epoch": 0.9764407393983328,
+      "grad_norm": 2.077919130586888,
+      "learning_rate": 2.8255452453326282e-08,
+      "loss": 1.003,
+      "step": 1347
+    },
+    {
+      "epoch": 0.9771656397245378,
+      "grad_norm": 2.028713109556211,
+      "learning_rate": 2.65178486097728e-08,
+      "loss": 1.049,
+      "step": 1348
+    },
+    {
+      "epoch": 0.977890540050743,
+      "grad_norm": 1.6908367188168048,
+      "learning_rate": 2.4835310768172026e-08,
+      "loss": 0.8732,
+      "step": 1349
+    },
+    {
+      "epoch": 0.9786154403769481,
+      "grad_norm": 2.4804402514448127,
+      "learning_rate": 2.3207848218222974e-08,
+      "loss": 0.9352,
+      "step": 1350
+    },
+    {
+      "epoch": 0.9793403407031533,
+      "grad_norm": 2.1542622951441857,
+      "learning_rate": 2.163546994553789e-08,
+      "loss": 0.9104,
+      "step": 1351
+    },
+    {
+      "epoch": 0.9800652410293584,
+      "grad_norm": 2.527194109345364,
+      "learning_rate": 2.011818463159676e-08,
+      "loss": 1.0832,
+      "step": 1352
+    },
+    {
+      "epoch": 0.9807901413555636,
+      "grad_norm": 2.317833340426504,
+      "learning_rate": 1.8656000653698436e-08,
+      "loss": 1.0094,
+      "step": 1353
+    },
+    {
+      "epoch": 0.9815150416817687,
+      "grad_norm": 2.343407752981275,
+      "learning_rate": 1.724892608491291e-08,
+      "loss": 1.0245,
+      "step": 1354
+    },
+    {
+      "epoch": 0.9822399420079739,
+      "grad_norm": 1.8989849927794842,
+      "learning_rate": 1.5896968694040228e-08,
+      "loss": 0.9417,
+      "step": 1355
+    },
+    {
+      "epoch": 0.982964842334179,
+      "grad_norm": 2.770976661865063,
+      "learning_rate": 1.4600135945560534e-08,
+      "loss": 0.99,
+      "step": 1356
+    },
+    {
+      "epoch": 0.9836897426603842,
+      "grad_norm": 1.7947478635260772,
+      "learning_rate": 1.3358434999602987e-08,
+      "loss": 0.892,
+      "step": 1357
+    },
+    {
+      "epoch": 0.9844146429865893,
+      "grad_norm": 2.310601190306124,
+      "learning_rate": 1.2171872711895794e-08,
+      "loss": 1.0248,
+      "step": 1358
+    },
+    {
+      "epoch": 0.9851395433127945,
+      "grad_norm": 2.287100614820235,
+      "learning_rate": 1.1040455633738457e-08,
+      "loss": 0.9974,
+      "step": 1359
+    },
+    {
+      "epoch": 0.9858644436389996,
+      "grad_norm": 2.107445997393385,
+      "learning_rate": 9.964190011955144e-09,
+      "loss": 0.9899,
+      "step": 1360
+    },
+    {
+      "epoch": 0.9865893439652048,
+      "grad_norm": 2.447737679629097,
+      "learning_rate": 8.943081788869157e-09,
+      "loss": 0.9687,
+      "step": 1361
+    },
+    {
+      "epoch": 0.9873142442914099,
+      "grad_norm": 2.151832291830881,
+      "learning_rate": 7.97713660226629e-09,
+      "loss": 1.0199,
+      "step": 1362
+    },
+    {
+      "epoch": 0.9880391446176151,
+      "grad_norm": 2.0385527802522976,
+      "learning_rate": 7.066359785362631e-09,
+      "loss": 1.021,
+      "step": 1363
+    },
+    {
+      "epoch": 0.9887640449438202,
+      "grad_norm": 2.3197757332833415,
+      "learning_rate": 6.210756366775705e-09,
+      "loss": 0.9591,
+      "step": 1364
+    },
+    {
+      "epoch": 0.9894889452700254,
+      "grad_norm": 2.3203612246956924,
+      "learning_rate": 5.410331070498931e-09,
+      "loss": 0.9674,
+      "step": 1365
+    },
+    {
+      "epoch": 0.9902138455962305,
+      "grad_norm": 1.8262826260586555,
+      "learning_rate": 4.665088315874977e-09,
+      "loss": 1.0296,
+      "step": 1366
+    },
+    {
+      "epoch": 0.9909387459224357,
+      "grad_norm": 1.961319160005979,
+      "learning_rate": 3.975032217566899e-09,
+      "loss": 0.9492,
+      "step": 1367
+    },
+    {
+      "epoch": 0.9916636462486408,
+      "grad_norm": 1.8126619362913512,
+      "learning_rate": 3.3401665855414824e-09,
+      "loss": 0.8814,
+      "step": 1368
+    },
+    {
+      "epoch": 0.992388546574846,
+      "grad_norm": 2.0882014952836916,
+      "learning_rate": 2.760494925045931e-09,
+      "loss": 0.8576,
+      "step": 1369
+    },
+    {
+      "epoch": 0.9931134469010511,
+      "grad_norm": 2.2450600007272037,
+      "learning_rate": 2.236020436586772e-09,
+      "loss": 0.9932,
+      "step": 1370
+    },
+    {
+      "epoch": 0.9938383472272563,
+      "grad_norm": 2.44028976343664,
+      "learning_rate": 1.7667460159143113e-09,
+      "loss": 0.8992,
+      "step": 1371
+    },
+    {
+      "epoch": 0.9945632475534614,
+      "grad_norm": 2.0146202920467293,
+      "learning_rate": 1.3526742540070913e-09,
+      "loss": 1.006,
+      "step": 1372
+    },
+    {
+      "epoch": 0.9952881478796666,
+      "grad_norm": 1.9601240017014594,
+      "learning_rate": 9.938074370530182e-10,
+      "loss": 0.9297,
+      "step": 1373
+    },
+    {
+      "epoch": 0.9960130482058717,
+      "grad_norm": 1.8087006446044052,
+      "learning_rate": 6.90147546443809e-10,
+      "loss": 0.8517,
+      "step": 1374
+    },
+    {
+      "epoch": 0.9967379485320769,
+      "grad_norm": 2.3390439919559936,
+      "learning_rate": 4.416962587583395e-10,
+      "loss": 0.9555,
+      "step": 1375
+    },
+    {
+      "epoch": 0.997462848858282,
+      "grad_norm": 1.9176591785146089,
+      "learning_rate": 2.484549457559826e-10,
+      "loss": 0.9623,
+      "step": 1376
+    },
+    {
+      "epoch": 0.9981877491844872,
+      "grad_norm": 2.215362657603315,
+      "learning_rate": 1.1042467436994664e-10,
+      "loss": 0.8946,
+      "step": 1377
+    },
+    {
+      "epoch": 0.9989126495106923,
+      "grad_norm": 2.000196078224186,
+      "learning_rate": 2.7606206697283755e-11,
+      "loss": 0.9503,
+      "step": 1378
+    },
+    {
+      "epoch": 0.9996375498368975,
+      "grad_norm": 2.2467269541867663,
+      "learning_rate": 0.0,
+      "loss": 0.9984,
+      "step": 1379
+    },
+    {
+      "epoch": 0.9996375498368975,
+      "step": 1379,
+      "total_flos": 1.3996327755055104e+16,
+      "train_loss": 0.49300233914940317,
+      "train_runtime": 19998.664,
+      "train_samples_per_second": 8.828,
+      "train_steps_per_second": 0.069
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1379,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 345,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.3996327755055104e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}