diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4889 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9989126495106923,
+  "eval_steps": 173,
+  "global_step": 689,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0014498006524102935,
+      "grad_norm": 24.13447961507207,
+      "learning_rate": 9.523809523809525e-07,
+      "loss": 1.8998,
+      "step": 1
+    },
+    {
+      "epoch": 0.002899601304820587,
+      "grad_norm": 30.45295521252791,
+      "learning_rate": 1.904761904761905e-06,
+      "loss": 1.9966,
+      "step": 2
+    },
+    {
+      "epoch": 0.004349401957230881,
+      "grad_norm": 30.801408549502632,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 2.0608,
+      "step": 3
+    },
+    {
+      "epoch": 0.005799202609641174,
+      "grad_norm": 20.960305151544546,
+      "learning_rate": 3.80952380952381e-06,
+      "loss": 1.8596,
+      "step": 4
+    },
+    {
+      "epoch": 0.007249003262051468,
+      "grad_norm": 6.7837904483700076,
+      "learning_rate": 4.761904761904762e-06,
+      "loss": 1.6535,
+      "step": 5
+    },
+    {
+      "epoch": 0.008698803914461761,
+      "grad_norm": 5.304344788310996,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 1.54,
+      "step": 6
+    },
+    {
+      "epoch": 0.010148604566872055,
+      "grad_norm": 4.991108317534938,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.5388,
+      "step": 7
+    },
+    {
+      "epoch": 0.011598405219282348,
+      "grad_norm": 13.08572986273314,
+      "learning_rate": 7.61904761904762e-06,
+      "loss": 1.4991,
+      "step": 8
+    },
+    {
+      "epoch": 0.013048205871692642,
+      "grad_norm": 6.611519344484442,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 1.4207,
+      "step": 9
+    },
+    {
+      "epoch": 0.014498006524102935,
+      "grad_norm": 4.554494831650385,
+      "learning_rate": 9.523809523809525e-06,
+      "loss": 1.3406,
+      "step": 10
+    },
+    {
+      "epoch": 0.01594780717651323,
+      "grad_norm": 3.603321232525562,
+      "learning_rate": 1.0476190476190477e-05,
+      "loss": 1.4017,
+      "step": 11
+    },
+    {
+      "epoch": 0.017397607828923523,
+      "grad_norm": 3.1607551883424767,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 1.3311,
+      "step": 12
+    },
+    {
+      "epoch": 0.018847408481333816,
+      "grad_norm": 3.2895713843454857,
+      "learning_rate": 1.2380952380952383e-05,
+      "loss": 1.2544,
+      "step": 13
+    },
+    {
+      "epoch": 0.02029720913374411,
+      "grad_norm": 2.444984709641969,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 1.3479,
+      "step": 14
+    },
+    {
+      "epoch": 0.021747009786154403,
+      "grad_norm": 2.794438483738135,
+      "learning_rate": 1.4285714285714287e-05,
+      "loss": 1.2702,
+      "step": 15
+    },
+    {
+      "epoch": 0.023196810438564697,
+      "grad_norm": 2.4896429131191162,
+      "learning_rate": 1.523809523809524e-05,
+      "loss": 1.216,
+      "step": 16
+    },
+    {
+      "epoch": 0.02464661109097499,
+      "grad_norm": 2.592280520795497,
+      "learning_rate": 1.6190476190476193e-05,
+      "loss": 1.2603,
+      "step": 17
+    },
+    {
+      "epoch": 0.026096411743385284,
+      "grad_norm": 2.3089246064331728,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 1.1429,
+      "step": 18
+    },
+    {
+      "epoch": 0.027546212395795577,
+      "grad_norm": 2.0892561574154245,
+      "learning_rate": 1.8095238095238097e-05,
+      "loss": 1.2466,
+      "step": 19
+    },
+    {
+      "epoch": 0.02899601304820587,
+      "grad_norm": 2.1894860428896474,
+      "learning_rate": 1.904761904761905e-05,
+      "loss": 1.1649,
+      "step": 20
+    },
+    {
+      "epoch": 0.030445813700616164,
+      "grad_norm": 2.1260454838228187,
+      "learning_rate": 2e-05,
+      "loss": 1.2496,
+      "step": 21
+    },
+    {
+      "epoch": 0.03189561435302646,
+      "grad_norm": 2.036819786811594,
+      "learning_rate": 1.999988940995767e-05,
+      "loss": 1.2229,
+      "step": 22
+    },
+    {
+      "epoch": 0.03334541500543675,
+      "grad_norm": 2.0509146511361744,
+      "learning_rate": 1.9999557642276715e-05,
+      "loss": 1.2147,
+      "step": 23
+    },
+    {
+      "epoch": 0.034795215657847045,
+      "grad_norm": 1.686495679737353,
+      "learning_rate": 1.9999004704295167e-05,
+      "loss": 1.2724,
+      "step": 24
+    },
+    {
+      "epoch": 0.03624501631025734,
+      "grad_norm": 1.7535694438920035,
+      "learning_rate": 1.9998230608242917e-05,
+      "loss": 1.1696,
+      "step": 25
+    },
+    {
+      "epoch": 0.03769481696266763,
+      "grad_norm": 1.9704353717115366,
+      "learning_rate": 1.999723537124143e-05,
+      "loss": 1.2729,
+      "step": 26
+    },
+    {
+      "epoch": 0.039144617615077926,
+      "grad_norm": 1.9861051369947142,
+      "learning_rate": 1.9996019015303366e-05,
+      "loss": 1.3143,
+      "step": 27
+    },
+    {
+      "epoch": 0.04059441826748822,
+      "grad_norm": 1.881689644020345,
+      "learning_rate": 1.9994581567332095e-05,
+      "loss": 1.2242,
+      "step": 28
+    },
+    {
+      "epoch": 0.04204421891989851,
+      "grad_norm": 1.7304543958546825,
+      "learning_rate": 1.9992923059121107e-05,
+      "loss": 1.2306,
+      "step": 29
+    },
+    {
+      "epoch": 0.043494019572308806,
+      "grad_norm": 1.863183618720525,
+      "learning_rate": 1.999104352735329e-05,
+      "loss": 1.1695,
+      "step": 30
+    },
+    {
+      "epoch": 0.0449438202247191,
+      "grad_norm": 1.774724097711846,
+      "learning_rate": 1.9988943013600156e-05,
+      "loss": 1.1461,
+      "step": 31
+    },
+    {
+      "epoch": 0.04639362087712939,
+      "grad_norm": 1.8971820140948996,
+      "learning_rate": 1.9986621564320878e-05,
+      "loss": 1.2419,
+      "step": 32
+    },
+    {
+      "epoch": 0.04784342152953969,
+      "grad_norm": 1.7527857842521397,
+      "learning_rate": 1.9984079230861296e-05,
+      "loss": 1.1451,
+      "step": 33
+    },
+    {
+      "epoch": 0.04929322218194998,
+      "grad_norm": 1.7046754732151643,
+      "learning_rate": 1.998131606945276e-05,
+      "loss": 1.1962,
+      "step": 34
+    },
+    {
+      "epoch": 0.050743022834360274,
+      "grad_norm": 1.7600404155638136,
+      "learning_rate": 1.9978332141210894e-05,
+      "loss": 1.1604,
+      "step": 35
+    },
+    {
+      "epoch": 0.05219282348677057,
+      "grad_norm": 1.678514906913871,
+      "learning_rate": 1.997512751213426e-05,
+      "loss": 1.2596,
+      "step": 36
+    },
+    {
+      "epoch": 0.05364262413918086,
+      "grad_norm": 1.679724530623478,
+      "learning_rate": 1.9971702253102856e-05,
+      "loss": 1.2399,
+      "step": 37
+    },
+    {
+      "epoch": 0.055092424791591155,
+      "grad_norm": 1.4877311404077136,
+      "learning_rate": 1.99680564398766e-05,
+      "loss": 1.1509,
+      "step": 38
+    },
+    {
+      "epoch": 0.05654222544400145,
+      "grad_norm": 1.6332470683549214,
+      "learning_rate": 1.9964190153093613e-05,
+      "loss": 1.1343,
+      "step": 39
+    },
+    {
+      "epoch": 0.05799202609641174,
+      "grad_norm": 1.84538760509367,
+      "learning_rate": 1.996010347826847e-05,
+      "loss": 1.1419,
+      "step": 40
+    },
+    {
+      "epoch": 0.059441826748822035,
+      "grad_norm": 1.6655966095988817,
+      "learning_rate": 1.9955796505790266e-05,
+      "loss": 1.171,
+      "step": 41
+    },
+    {
+      "epoch": 0.06089162740123233,
+      "grad_norm": 1.6686217671799741,
+      "learning_rate": 1.9951269330920667e-05,
+      "loss": 1.1942,
+      "step": 42
+    },
+    {
+      "epoch": 0.06234142805364262,
+      "grad_norm": 1.6560839908920755,
+      "learning_rate": 1.9946522053791755e-05,
+      "loss": 1.1952,
+      "step": 43
+    },
+    {
+      "epoch": 0.06379122870605292,
+      "grad_norm": 1.7032868748569239,
+      "learning_rate": 1.9941554779403854e-05,
+      "loss": 1.1222,
+      "step": 44
+    },
+    {
+      "epoch": 0.06524102935846321,
+      "grad_norm": 1.8393238935309384,
+      "learning_rate": 1.993636761762317e-05,
+      "loss": 1.0861,
+      "step": 45
+    },
+    {
+      "epoch": 0.0666908300108735,
+      "grad_norm": 1.7714338397855247,
+      "learning_rate": 1.9930960683179404e-05,
+      "loss": 1.1484,
+      "step": 46
+    },
+    {
+      "epoch": 0.0681406306632838,
+      "grad_norm": 2.2410599411909975,
+      "learning_rate": 1.992533409566317e-05,
+      "loss": 1.244,
+      "step": 47
+    },
+    {
+      "epoch": 0.06959043131569409,
+      "grad_norm": 1.5313267603910532,
+      "learning_rate": 1.991948797952338e-05,
+      "loss": 1.1747,
+      "step": 48
+    },
+    {
+      "epoch": 0.07104023196810438,
+      "grad_norm": 1.5299254412788565,
+      "learning_rate": 1.991342246406448e-05,
+      "loss": 1.1817,
+      "step": 49
+    },
+    {
+      "epoch": 0.07249003262051468,
+      "grad_norm": 1.8915773926444455,
+      "learning_rate": 1.9907137683443593e-05,
+      "loss": 1.1324,
+      "step": 50
+    },
+    {
+      "epoch": 0.07393983327292497,
+      "grad_norm": 1.3490879234916116,
+      "learning_rate": 1.990063377666755e-05,
+      "loss": 1.1723,
+      "step": 51
+    },
+    {
+      "epoch": 0.07538963392533526,
+      "grad_norm": 1.4568194926297389,
+      "learning_rate": 1.9893910887589817e-05,
+      "loss": 1.1436,
+      "step": 52
+    },
+    {
+      "epoch": 0.07683943457774556,
+      "grad_norm": 1.4469636067398668,
+      "learning_rate": 1.9886969164907306e-05,
+      "loss": 1.2352,
+      "step": 53
+    },
+    {
+      "epoch": 0.07828923523015585,
+      "grad_norm": 1.5970835964860943,
+      "learning_rate": 1.9879808762157106e-05,
+      "loss": 1.1225,
+      "step": 54
+    },
+    {
+      "epoch": 0.07973903588256614,
+      "grad_norm": 1.781212415486643,
+      "learning_rate": 1.987242983771306e-05,
+      "loss": 1.154,
+      "step": 55
+    },
+    {
+      "epoch": 0.08118883653497644,
+      "grad_norm": 1.7378145254926347,
+      "learning_rate": 1.986483255478229e-05,
+      "loss": 1.1878,
+      "step": 56
+    },
+    {
+      "epoch": 0.08263863718738673,
+      "grad_norm": 1.4814443147300105,
+      "learning_rate": 1.985701708140155e-05,
+      "loss": 1.2252,
+      "step": 57
+    },
+    {
+      "epoch": 0.08408843783979703,
+      "grad_norm": 1.6062356336169084,
+      "learning_rate": 1.9848983590433555e-05,
+      "loss": 1.1983,
+      "step": 58
+    },
+    {
+      "epoch": 0.08553823849220732,
+      "grad_norm": 1.6978976055091966,
+      "learning_rate": 1.984073225956313e-05,
+      "loss": 1.0904,
+      "step": 59
+    },
+    {
+      "epoch": 0.08698803914461761,
+      "grad_norm": 1.5426256140619299,
+      "learning_rate": 1.983226327129327e-05,
+      "loss": 1.1949,
+      "step": 60
+    },
+    {
+      "epoch": 0.0884378397970279,
+      "grad_norm": 1.5601237512353299,
+      "learning_rate": 1.982357681294114e-05,
+      "loss": 1.0572,
+      "step": 61
+    },
+    {
+      "epoch": 0.0898876404494382,
+      "grad_norm": 1.5967484729572234,
+      "learning_rate": 1.981467307663389e-05,
+      "loss": 1.0921,
+      "step": 62
+    },
+    {
+      "epoch": 0.0913374411018485,
+      "grad_norm": 1.641500172879362,
+      "learning_rate": 1.9805552259304447e-05,
+      "loss": 1.1328,
+      "step": 63
+    },
+    {
+      "epoch": 0.09278724175425879,
+      "grad_norm": 1.6488330164710152,
+      "learning_rate": 1.9796214562687114e-05,
+      "loss": 1.1363,
+      "step": 64
+    },
+    {
+      "epoch": 0.09423704240666908,
+      "grad_norm": 1.4688665781902088,
+      "learning_rate": 1.978666019331315e-05,
+      "loss": 1.1283,
+      "step": 65
+    },
+    {
+      "epoch": 0.09568684305907937,
+      "grad_norm": 1.5154306356661489,
+      "learning_rate": 1.9776889362506175e-05,
+      "loss": 1.0697,
+      "step": 66
+    },
+    {
+      "epoch": 0.09713664371148967,
+      "grad_norm": 1.4911514238908088,
+      "learning_rate": 1.9766902286377507e-05,
+      "loss": 1.115,
+      "step": 67
+    },
+    {
+      "epoch": 0.09858644436389996,
+      "grad_norm": 1.5307558397629772,
+      "learning_rate": 1.9756699185821387e-05,
+      "loss": 1.2047,
+      "step": 68
+    },
+    {
+      "epoch": 0.10003624501631025,
+      "grad_norm": 1.5106988471187286,
+      "learning_rate": 1.974628028651007e-05,
+      "loss": 1.1261,
+      "step": 69
+    },
+    {
+      "epoch": 0.10148604566872055,
+      "grad_norm": 1.4552475316752103,
+      "learning_rate": 1.973564581888887e-05,
+      "loss": 1.1147,
+      "step": 70
+    },
+    {
+      "epoch": 0.10293584632113084,
+      "grad_norm": 1.53431209741887,
+      "learning_rate": 1.9724796018171022e-05,
+      "loss": 1.1902,
+      "step": 71
+    },
+    {
+      "epoch": 0.10438564697354114,
+      "grad_norm": 1.4015250181683467,
+      "learning_rate": 1.9713731124332513e-05,
+      "loss": 1.0722,
+      "step": 72
+    },
+    {
+      "epoch": 0.10583544762595143,
+      "grad_norm": 1.5534118963467138,
+      "learning_rate": 1.9702451382106763e-05,
+      "loss": 1.1466,
+      "step": 73
+    },
+    {
+      "epoch": 0.10728524827836172,
+      "grad_norm": 1.450611476626139,
+      "learning_rate": 1.96909570409792e-05,
+      "loss": 1.1608,
+      "step": 74
+    },
+    {
+      "epoch": 0.10873504893077202,
+      "grad_norm": 1.4459612852552788,
+      "learning_rate": 1.967924835518176e-05,
+      "loss": 1.2124,
+      "step": 75
+    },
+    {
+      "epoch": 0.11018484958318231,
+      "grad_norm": 1.4797644140322972,
+      "learning_rate": 1.9667325583687255e-05,
+      "loss": 1.198,
+      "step": 76
+    },
+    {
+      "epoch": 0.1116346502355926,
+      "grad_norm": 1.4882682302599846,
+      "learning_rate": 1.9655188990203648e-05,
+      "loss": 1.1202,
+      "step": 77
+    },
+    {
+      "epoch": 0.1130844508880029,
+      "grad_norm": 1.355157509169686,
+      "learning_rate": 1.9642838843168215e-05,
+      "loss": 1.0922,
+      "step": 78
+    },
+    {
+      "epoch": 0.11453425154041319,
+      "grad_norm": 1.3915247696009005,
+      "learning_rate": 1.963027541574161e-05,
+      "loss": 1.2374,
+      "step": 79
+    },
+    {
+      "epoch": 0.11598405219282348,
+      "grad_norm": 1.458141199446404,
+      "learning_rate": 1.9617498985801834e-05,
+      "loss": 1.0358,
+      "step": 80
+    },
+    {
+      "epoch": 0.11743385284523378,
+      "grad_norm": 1.5755642593090884,
+      "learning_rate": 1.9604509835938066e-05,
+      "loss": 1.1363,
+      "step": 81
+    },
+    {
+      "epoch": 0.11888365349764407,
+      "grad_norm": 1.3852311342190622,
+      "learning_rate": 1.9591308253444437e-05,
+      "loss": 1.1751,
+      "step": 82
+    },
+    {
+      "epoch": 0.12033345415005436,
+      "grad_norm": 1.3988027363976137,
+      "learning_rate": 1.9577894530313655e-05,
+      "loss": 1.1621,
+      "step": 83
+    },
+    {
+      "epoch": 0.12178325480246466,
+      "grad_norm": 1.6337073588999842,
+      "learning_rate": 1.9564268963230564e-05,
+      "loss": 1.1559,
+      "step": 84
+    },
+    {
+      "epoch": 0.12323305545487495,
+      "grad_norm": 1.531320376259164,
+      "learning_rate": 1.9550431853565577e-05,
+      "loss": 1.1818,
+      "step": 85
+    },
+    {
+      "epoch": 0.12468285610728524,
+      "grad_norm": 1.6433549398467089,
+      "learning_rate": 1.9536383507367994e-05,
+      "loss": 1.1229,
+      "step": 86
+    },
+    {
+      "epoch": 0.12613265675969554,
+      "grad_norm": 1.3895840180640682,
+      "learning_rate": 1.9522124235359265e-05,
+      "loss": 1.0912,
+      "step": 87
+    },
+    {
+      "epoch": 0.12758245741210583,
+      "grad_norm": 1.5848650229194654,
+      "learning_rate": 1.950765435292608e-05,
+      "loss": 1.1968,
+      "step": 88
+    },
+    {
+      "epoch": 0.12903225806451613,
+      "grad_norm": 1.7121315741169545,
+      "learning_rate": 1.9492974180113425e-05,
+      "loss": 1.1739,
+      "step": 89
+    },
+    {
+      "epoch": 0.13048205871692642,
+      "grad_norm": 1.4744411608527952,
+      "learning_rate": 1.9478084041617492e-05,
+      "loss": 1.1226,
+      "step": 90
+    },
+    {
+      "epoch": 0.1319318593693367,
+      "grad_norm": 1.4038484929114188,
+      "learning_rate": 1.946298426677848e-05,
+      "loss": 1.1597,
+      "step": 91
+    },
+    {
+      "epoch": 0.133381660021747,
+      "grad_norm": 1.5583842316972911,
+      "learning_rate": 1.944767518957334e-05,
+      "loss": 1.1542,
+      "step": 92
+    },
+    {
+      "epoch": 0.1348314606741573,
+      "grad_norm": 1.432692159389847,
+      "learning_rate": 1.943215714860838e-05,
+      "loss": 1.103,
+      "step": 93
+    },
+    {
+      "epoch": 0.1362812613265676,
+      "grad_norm": 1.2610863891108142,
+      "learning_rate": 1.941643048711175e-05,
+      "loss": 1.1572,
+      "step": 94
+    },
+    {
+      "epoch": 0.1377310619789779,
+      "grad_norm": 1.409082053656987,
+      "learning_rate": 1.9400495552925887e-05,
+      "loss": 1.0925,
+      "step": 95
+    },
+    {
+      "epoch": 0.13918086263138818,
+      "grad_norm": 1.5180883575660837,
+      "learning_rate": 1.93843526984998e-05,
+      "loss": 1.1747,
+      "step": 96
+    },
+    {
+      "epoch": 0.14063066328379847,
+      "grad_norm": 1.446188613038219,
+      "learning_rate": 1.936800228088128e-05,
+      "loss": 1.1694,
+      "step": 97
+    },
+    {
+      "epoch": 0.14208046393620877,
+      "grad_norm": 1.4730544927038625,
+      "learning_rate": 1.9351444661709e-05,
+      "loss": 1.1774,
+      "step": 98
+    },
+    {
+      "epoch": 0.14353026458861906,
+      "grad_norm": 1.4643917490874658,
+      "learning_rate": 1.9334680207204525e-05,
+      "loss": 1.1042,
+      "step": 99
+    },
+    {
+      "epoch": 0.14498006524102935,
+      "grad_norm": 1.4165444183523295,
+      "learning_rate": 1.9317709288164198e-05,
+      "loss": 1.0092,
+      "step": 100
+    },
+    {
+      "epoch": 0.14642986589343965,
+      "grad_norm": 1.5403469992965209,
+      "learning_rate": 1.930053227995095e-05,
+      "loss": 1.1421,
+      "step": 101
+    },
+    {
+      "epoch": 0.14787966654584994,
+      "grad_norm": 1.358715840826538,
+      "learning_rate": 1.9283149562485997e-05,
+      "loss": 1.1104,
+      "step": 102
+    },
+    {
+      "epoch": 0.14932946719826024,
+      "grad_norm": 1.6044345854353719,
+      "learning_rate": 1.9265561520240434e-05,
+      "loss": 1.1609,
+      "step": 103
+    },
+    {
+      "epoch": 0.15077926785067053,
+      "grad_norm": 1.443782398769455,
+      "learning_rate": 1.924776854222672e-05,
+      "loss": 1.1445,
+      "step": 104
+    },
+    {
+      "epoch": 0.15222906850308082,
+      "grad_norm": 1.5141352847459923,
+      "learning_rate": 1.9229771021990094e-05,
+      "loss": 1.1046,
+      "step": 105
+    },
+    {
+      "epoch": 0.15367886915549112,
+      "grad_norm": 1.4472178264696836,
+      "learning_rate": 1.9211569357599868e-05,
+      "loss": 1.1074,
+      "step": 106
+    },
+    {
+      "epoch": 0.1551286698079014,
+      "grad_norm": 1.4362039642645186,
+      "learning_rate": 1.9193163951640608e-05,
+      "loss": 1.1988,
+      "step": 107
+    },
+    {
+      "epoch": 0.1565784704603117,
+      "grad_norm": 1.5152580522698165,
+      "learning_rate": 1.9174555211203235e-05,
+      "loss": 1.2267,
+      "step": 108
+    },
+    {
+      "epoch": 0.158028271112722,
+      "grad_norm": 1.639305174550132,
+      "learning_rate": 1.9155743547876026e-05,
+      "loss": 1.1182,
+      "step": 109
+    },
+    {
+      "epoch": 0.1594780717651323,
+      "grad_norm": 1.4118479734974339,
+      "learning_rate": 1.9136729377735514e-05,
+      "loss": 1.0732,
+      "step": 110
+    },
+    {
+      "epoch": 0.16092787241754258,
+      "grad_norm": 1.5233666916040183,
+      "learning_rate": 1.9117513121337273e-05,
+      "loss": 1.1565,
+      "step": 111
+    },
+    {
+      "epoch": 0.16237767306995288,
+      "grad_norm": 1.5669668780786876,
+      "learning_rate": 1.9098095203706625e-05,
+      "loss": 1.1698,
+      "step": 112
+    },
+    {
+      "epoch": 0.16382747372236317,
+      "grad_norm": 1.8263559139612136,
+      "learning_rate": 1.9078476054329237e-05,
+      "loss": 1.1195,
+      "step": 113
+    },
+    {
+      "epoch": 0.16527727437477346,
+      "grad_norm": 1.640788164508231,
+      "learning_rate": 1.905865610714162e-05,
+      "loss": 1.1575,
+      "step": 114
+    },
+    {
+      "epoch": 0.16672707502718376,
+      "grad_norm": 1.4501965846013474,
+      "learning_rate": 1.9038635800521538e-05,
+      "loss": 1.1322,
+      "step": 115
+    },
+    {
+      "epoch": 0.16817687567959405,
+      "grad_norm": 1.51298564867844,
+      "learning_rate": 1.9018415577278295e-05,
+      "loss": 1.162,
+      "step": 116
+    },
+    {
+      "epoch": 0.16962667633200434,
+      "grad_norm": 1.513249352306865,
+      "learning_rate": 1.8997995884642968e-05,
+      "loss": 1.1342,
+      "step": 117
+    },
+    {
+      "epoch": 0.17107647698441464,
+      "grad_norm": 1.6015751718989664,
+      "learning_rate": 1.8977377174258485e-05,
+      "loss": 1.2142,
+      "step": 118
+    },
+    {
+      "epoch": 0.17252627763682493,
+      "grad_norm": 1.501950267018482,
+      "learning_rate": 1.895655990216966e-05,
+      "loss": 1.1497,
+      "step": 119
+    },
+    {
+      "epoch": 0.17397607828923523,
+      "grad_norm": 1.3779583455375068,
+      "learning_rate": 1.893554452881309e-05,
+      "loss": 1.0982,
+      "step": 120
+    },
+    {
+      "epoch": 0.17542587894164552,
+      "grad_norm": 1.3519192766579768,
+      "learning_rate": 1.8914331519006985e-05,
+      "loss": 1.0864,
+      "step": 121
+    },
+    {
+      "epoch": 0.1768756795940558,
+      "grad_norm": 1.3860532383536595,
+      "learning_rate": 1.8892921341940873e-05,
+      "loss": 1.1255,
+      "step": 122
+    },
+    {
+      "epoch": 0.1783254802464661,
+      "grad_norm": 1.3825959683141158,
+      "learning_rate": 1.8871314471165233e-05,
+      "loss": 1.1342,
+      "step": 123
+    },
+    {
+      "epoch": 0.1797752808988764,
+      "grad_norm": 1.483862180977746,
+      "learning_rate": 1.8849511384581015e-05,
+      "loss": 1.1137,
+      "step": 124
+    },
+    {
+      "epoch": 0.1812250815512867,
+      "grad_norm": 1.2796802416681439,
+      "learning_rate": 1.8827512564429076e-05,
+      "loss": 1.1129,
+      "step": 125
+    },
+    {
+      "epoch": 0.182674882203697,
+      "grad_norm": 1.2656064431036425,
+      "learning_rate": 1.8805318497279498e-05,
+      "loss": 1.0582,
+      "step": 126
+    },
+    {
+      "epoch": 0.18412468285610728,
+      "grad_norm": 1.4861590500158623,
+      "learning_rate": 1.8782929674020856e-05,
+      "loss": 1.0994,
+      "step": 127
+    },
+    {
+      "epoch": 0.18557448350851757,
+      "grad_norm": 1.5878571152116014,
+      "learning_rate": 1.8760346589849323e-05,
+      "loss": 1.1222,
+      "step": 128
+    },
+    {
+      "epoch": 0.18702428416092787,
+      "grad_norm": 1.5139654532058364,
+      "learning_rate": 1.8737569744257754e-05,
+      "loss": 1.0915,
+      "step": 129
+    },
+    {
+      "epoch": 0.18847408481333816,
+      "grad_norm": 1.3248045858003061,
+      "learning_rate": 1.871459964102461e-05,
+      "loss": 1.0953,
+      "step": 130
+    },
+    {
+      "epoch": 0.18992388546574845,
+      "grad_norm": 1.6134962975888272,
+      "learning_rate": 1.8691436788202823e-05,
+      "loss": 1.179,
+      "step": 131
+    },
+    {
+      "epoch": 0.19137368611815875,
+      "grad_norm": 1.755144132755189,
+      "learning_rate": 1.8668081698108576e-05,
+      "loss": 1.2302,
+      "step": 132
+    },
+    {
+      "epoch": 0.19282348677056904,
+      "grad_norm": 1.4801616489839033,
+      "learning_rate": 1.8644534887309947e-05,
+      "loss": 1.108,
+      "step": 133
+    },
+    {
+      "epoch": 0.19427328742297933,
+      "grad_norm": 1.614620294409322,
+      "learning_rate": 1.8620796876615494e-05,
+      "loss": 1.1353,
+      "step": 134
+    },
+    {
+      "epoch": 0.19572308807538963,
+      "grad_norm": 1.7370142366361634,
+      "learning_rate": 1.859686819106274e-05,
+      "loss": 1.1719,
+      "step": 135
+    },
+    {
+      "epoch": 0.19717288872779992,
+      "grad_norm": 1.3062716789125641,
+      "learning_rate": 1.8572749359906556e-05,
+      "loss": 1.1363,
+      "step": 136
+    },
+    {
+      "epoch": 0.19862268938021022,
+      "grad_norm": 1.3365557550793516,
+      "learning_rate": 1.8548440916607454e-05,
+      "loss": 1.0996,
+      "step": 137
+    },
+    {
+      "epoch": 0.2000724900326205,
+      "grad_norm": 1.4109754447602447,
+      "learning_rate": 1.8523943398819786e-05,
+      "loss": 1.131,
+      "step": 138
+    },
+    {
+      "epoch": 0.2015222906850308,
+      "grad_norm": 1.4505137821937386,
+      "learning_rate": 1.8499257348379857e-05,
+      "loss": 1.128,
+      "step": 139
+    },
+    {
+      "epoch": 0.2029720913374411,
+      "grad_norm": 1.372524236653458,
+      "learning_rate": 1.8474383311293943e-05,
+      "loss": 1.1371,
+      "step": 140
+    },
+    {
+      "epoch": 0.2044218919898514,
+      "grad_norm": 1.4076947769304085,
+      "learning_rate": 1.8449321837726207e-05,
+      "loss": 1.1929,
+      "step": 141
+    },
+    {
+      "epoch": 0.20587169264226168,
+      "grad_norm": 1.3802406772528608,
+      "learning_rate": 1.842407348198653e-05,
+      "loss": 1.0398,
+      "step": 142
+    },
+    {
+      "epoch": 0.20732149329467198,
+      "grad_norm": 1.5087433128983168,
+      "learning_rate": 1.8398638802518263e-05,
+      "loss": 1.0679,
+      "step": 143
+    },
+    {
+      "epoch": 0.20877129394708227,
+      "grad_norm": 1.5319722218764114,
+      "learning_rate": 1.8373018361885856e-05,
+      "loss": 1.066,
+      "step": 144
+    },
+    {
+      "epoch": 0.21022109459949256,
+      "grad_norm": 1.4608227726445189,
+      "learning_rate": 1.8347212726762436e-05,
+      "loss": 1.0817,
+      "step": 145
+    },
+    {
+      "epoch": 0.21167089525190286,
+      "grad_norm": 1.4830268366025634,
+      "learning_rate": 1.832122246791726e-05,
+      "loss": 1.104,
+      "step": 146
+    },
+    {
+      "epoch": 0.21312069590431315,
+      "grad_norm": 1.4099537211404467,
+      "learning_rate": 1.8295048160203087e-05,
+      "loss": 1.1645,
+      "step": 147
+    },
+    {
+      "epoch": 0.21457049655672344,
+      "grad_norm": 1.4557844872919459,
+      "learning_rate": 1.8268690382543484e-05,
+      "loss": 1.141,
+      "step": 148
+    },
+    {
+      "epoch": 0.21602029720913374,
+      "grad_norm": 1.270168457505729,
+      "learning_rate": 1.8242149717919993e-05,
+      "loss": 1.0577,
+      "step": 149
+    },
+    {
+      "epoch": 0.21747009786154403,
+      "grad_norm": 1.2697491363148088,
+      "learning_rate": 1.8215426753359268e-05,
+      "loss": 1.04,
+      "step": 150
+    },
+    {
+      "epoch": 0.21891989851395433,
+      "grad_norm": 1.3442828497425443,
+      "learning_rate": 1.8188522079920056e-05,
+      "loss": 1.1662,
+      "step": 151
+    },
+    {
+      "epoch": 0.22036969916636462,
+      "grad_norm": 1.2276541689878522,
+      "learning_rate": 1.816143629268016e-05,
+      "loss": 1.0708,
+      "step": 152
+    },
+    {
+      "epoch": 0.2218194998187749,
+      "grad_norm": 1.3323604772774698,
+      "learning_rate": 1.8134169990723246e-05,
+      "loss": 1.0824,
+      "step": 153
+    },
+    {
+      "epoch": 0.2232693004711852,
+      "grad_norm": 1.4557355167897452,
+      "learning_rate": 1.8106723777125614e-05,
+      "loss": 1.1396,
+      "step": 154
+    },
+    {
+      "epoch": 0.2247191011235955,
+      "grad_norm": 1.5218641501565704,
+      "learning_rate": 1.8079098258942846e-05,
+      "loss": 1.1925,
+      "step": 155
+    },
+    {
+      "epoch": 0.2261689017760058,
+      "grad_norm": 1.5207834973985892,
+      "learning_rate": 1.8051294047196393e-05,
+      "loss": 1.1745,
+      "step": 156
+    },
+    {
+      "epoch": 0.2276187024284161,
+      "grad_norm": 1.5074176278501161,
+      "learning_rate": 1.8023311756860036e-05,
+      "loss": 1.1292,
+      "step": 157
+    },
+    {
+      "epoch": 0.22906850308082638,
+      "grad_norm": 1.4716520653852196,
+      "learning_rate": 1.799515200684632e-05,
+      "loss": 1.1131,
+      "step": 158
+    },
+    {
+      "epoch": 0.23051830373323667,
+      "grad_norm": 1.4845865930384103,
+      "learning_rate": 1.796681541999283e-05,
+      "loss": 1.1263,
+      "step": 159
+    },
+    {
+      "epoch": 0.23196810438564697,
+      "grad_norm": 1.3751923391802785,
+      "learning_rate": 1.793830262304843e-05,
+      "loss": 1.1238,
+      "step": 160
+    },
+    {
+      "epoch": 0.23341790503805726,
+      "grad_norm": 1.3201509206287807,
+      "learning_rate": 1.7909614246659406e-05,
+      "loss": 1.1805,
+      "step": 161
+    },
+    {
+      "epoch": 0.23486770569046755,
+      "grad_norm": 1.3214787783622646,
+      "learning_rate": 1.7880750925355516e-05,
+      "loss": 1.1013,
+      "step": 162
+    },
+    {
+      "epoch": 0.23631750634287785,
+      "grad_norm": 1.4751076809172468,
+      "learning_rate": 1.785171329753594e-05,
+      "loss": 1.0894,
+      "step": 163
+    },
+    {
+      "epoch": 0.23776730699528814,
+      "grad_norm": 1.464669067828848,
+      "learning_rate": 1.7822502005455176e-05,
+      "loss": 1.0642,
+      "step": 164
+    },
+    {
+      "epoch": 0.23921710764769843,
+      "grad_norm": 1.3955975778875378,
+      "learning_rate": 1.7793117695208828e-05,
+      "loss": 1.1109,
+      "step": 165
+    },
+    {
+      "epoch": 0.24066690830010873,
+      "grad_norm": 1.51151089334327,
+      "learning_rate": 1.7763561016719323e-05,
+      "loss": 1.0363,
+      "step": 166
+    },
+    {
+      "epoch": 0.24211670895251902,
+      "grad_norm": 1.4084955117317381,
+      "learning_rate": 1.773383262372152e-05,
+      "loss": 1.1257,
+      "step": 167
+    },
+    {
+      "epoch": 0.24356650960492932,
+      "grad_norm": 1.3408461562496001,
+      "learning_rate": 1.7703933173748276e-05,
+      "loss": 1.1642,
+      "step": 168
+    },
+    {
+      "epoch": 0.2450163102573396,
+      "grad_norm": 1.4577535476860215,
+      "learning_rate": 1.767386332811587e-05,
+      "loss": 1.0761,
+      "step": 169
+    },
+    {
+      "epoch": 0.2464661109097499,
+      "grad_norm": 1.436094655941693,
+      "learning_rate": 1.764362375190941e-05,
+      "loss": 1.1197,
+      "step": 170
+    },
+    {
+      "epoch": 0.2479159115621602,
+      "grad_norm": 1.3810690133360064,
+      "learning_rate": 1.761321511396809e-05,
+      "loss": 1.1025,
+      "step": 171
+    },
+    {
+      "epoch": 0.2493657122145705,
+      "grad_norm": 1.5272570499192768,
+      "learning_rate": 1.7582638086870426e-05,
+      "loss": 1.1389,
+      "step": 172
+    },
+    {
+      "epoch": 0.2508155128669808,
+      "grad_norm": 1.4332805828816457,
+      "learning_rate": 1.7551893346919363e-05,
+      "loss": 1.131,
+      "step": 173
+    },
+    {
+      "epoch": 0.2508155128669808,
+      "eval_loss": 1.841679334640503,
+      "eval_runtime": 352.3415,
+      "eval_samples_per_second": 50.105,
+      "eval_steps_per_second": 0.783,
+      "step": 173
+    },
+    {
+      "epoch": 0.2522653135193911,
+      "grad_norm": 1.4330820634024624,
+      "learning_rate": 1.752098157412732e-05,
+      "loss": 1.0607,
+      "step": 174
+    },
+    {
+      "epoch": 0.25371511417180137,
+      "grad_norm": 1.4201461220688676,
+      "learning_rate": 1.7489903452201144e-05,
+      "loss": 1.0765,
+      "step": 175
+    },
+    {
+      "epoch": 0.25516491482421166,
+      "grad_norm": 1.2784340155008505,
+      "learning_rate": 1.7458659668527005e-05,
+      "loss": 1.0972,
+      "step": 176
+    },
+    {
+      "epoch": 0.25661471547662196,
+      "grad_norm": 1.4100506654534135,
+      "learning_rate": 1.742725091415517e-05,
+      "loss": 1.0948,
+      "step": 177
+    },
+    {
+      "epoch": 0.25806451612903225,
+      "grad_norm": 1.5152898205746248,
+      "learning_rate": 1.7395677883784742e-05,
+      "loss": 1.1025,
+      "step": 178
+    },
+    {
+      "epoch": 0.25951431678144254,
+      "grad_norm": 1.374892883843552,
+      "learning_rate": 1.7363941275748263e-05,
+      "loss": 1.1152,
+      "step": 179
+    },
+    {
+      "epoch": 0.26096411743385284,
+      "grad_norm": 1.6580439963300477,
+      "learning_rate": 1.733204179199631e-05,
+      "loss": 1.1579,
+      "step": 180
+    },
+    {
+      "epoch": 0.26241391808626313,
+      "grad_norm": 1.451092995531561,
+      "learning_rate": 1.729998013808192e-05,
+      "loss": 1.1943,
+      "step": 181
+    },
+    {
+      "epoch": 0.2638637187386734,
+      "grad_norm": 1.277688366523427,
+      "learning_rate": 1.726775702314504e-05,
+      "loss": 1.0366,
+      "step": 182
+    },
+    {
+      "epoch": 0.2653135193910837,
+      "grad_norm": 1.291568050166545,
+      "learning_rate": 1.7235373159896792e-05,
+      "loss": 1.0691,
+      "step": 183
+    },
+    {
+      "epoch": 0.266763320043494,
+      "grad_norm": 1.371629627865818,
+      "learning_rate": 1.720282926460374e-05,
+      "loss": 1.1668,
+      "step": 184
+    },
+    {
+      "epoch": 0.2682131206959043,
+      "grad_norm": 1.4259431105964036,
+      "learning_rate": 1.717012605707203e-05,
+      "loss": 1.1126,
+      "step": 185
+    },
+    {
+      "epoch": 0.2696629213483146,
+      "grad_norm": 1.192260375740259,
+      "learning_rate": 1.7137264260631493e-05,
+      "loss": 1.068,
+      "step": 186
+    },
+    {
+      "epoch": 0.2711127220007249,
+      "grad_norm": 1.4092264976869975,
+      "learning_rate": 1.7104244602119613e-05,
+      "loss": 1.1454,
+      "step": 187
+    },
+    {
+      "epoch": 0.2725625226531352,
+      "grad_norm": 1.3111363306139812,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 1.1602,
+      "step": 188
+    },
+    {
+      "epoch": 0.2740123233055455,
+      "grad_norm": 1.382222270750256,
+      "learning_rate": 1.7037734623673616e-05,
+      "loss": 1.1134,
+      "step": 189
+    },
+    {
+      "epoch": 0.2754621239579558,
+      "grad_norm": 1.2995650712044595,
+      "learning_rate": 1.7004245774807764e-05,
+      "loss": 1.0502,
+      "step": 190
+    },
+    {
+      "epoch": 0.27691192461036607,
+      "grad_norm": 1.5471631444598317,
+      "learning_rate": 1.697060200597457e-05,
+      "loss": 1.1771,
+      "step": 191
+    },
+    {
+      "epoch": 0.27836172526277636,
+      "grad_norm": 1.4439447079239869,
+      "learning_rate": 1.6936804061307196e-05,
+      "loss": 1.0014,
+      "step": 192
+    },
+    {
+      "epoch": 0.27981152591518665,
+      "grad_norm": 1.5273892731490175,
+      "learning_rate": 1.6902852688348866e-05,
+      "loss": 1.1393,
+      "step": 193
+    },
+    {
+      "epoch": 0.28126132656759695,
+      "grad_norm": 1.3454897882417334,
+      "learning_rate": 1.686874863803633e-05,
+      "loss": 1.0892,
+      "step": 194
+    },
+    {
+      "epoch": 0.28271112722000724,
+      "grad_norm": 1.3405406430571376,
+      "learning_rate": 1.6834492664683275e-05,
+      "loss": 1.1305,
+      "step": 195
+    },
+    {
+      "epoch": 0.28416092787241753,
+      "grad_norm": 1.4318362570035057,
+      "learning_rate": 1.68000855259636e-05,
+      "loss": 1.1346,
+      "step": 196
+    },
+    {
+      "epoch": 0.28561072852482783,
+      "grad_norm": 1.3779026744388763,
+      "learning_rate": 1.676552798289469e-05,
+      "loss": 1.1185,
+      "step": 197
+    },
+    {
+      "epoch": 0.2870605291772381,
+      "grad_norm": 1.4515408761124213,
+      "learning_rate": 1.6730820799820575e-05,
+      "loss": 1.0151,
+      "step": 198
+    },
+    {
+      "epoch": 0.2885103298296484,
+      "grad_norm": 1.382940350476588,
+      "learning_rate": 1.669596474439503e-05,
+      "loss": 1.064,
+      "step": 199
+    },
+    {
+      "epoch": 0.2899601304820587,
+      "grad_norm": 1.5067833204567096,
+      "learning_rate": 1.666096058756458e-05,
+      "loss": 1.1951,
+      "step": 200
+    },
+    {
+      "epoch": 0.291409931134469,
+      "grad_norm": 1.382313596426829,
+      "learning_rate": 1.6625809103551462e-05,
+      "loss": 1.0388,
+      "step": 201
+    },
+    {
+      "epoch": 0.2928597317868793,
+      "grad_norm": 1.4284878703158121,
+      "learning_rate": 1.6590511069836498e-05,
+      "loss": 1.127,
+      "step": 202
+    },
+    {
+      "epoch": 0.2943095324392896,
+      "grad_norm": 1.3765430448687064,
+      "learning_rate": 1.6555067267141896e-05,
+      "loss": 1.0894,
+      "step": 203
+    },
+    {
+      "epoch": 0.2957593330916999,
+      "grad_norm": 1.5344983191587576,
+      "learning_rate": 1.651947847941398e-05,
+      "loss": 1.1576,
+      "step": 204
+    },
+    {
+      "epoch": 0.2972091337441102,
+      "grad_norm": 1.5095048776193207,
+      "learning_rate": 1.648374549380587e-05,
+      "loss": 1.0476,
+      "step": 205
+    },
+    {
+      "epoch": 0.29865893439652047,
+      "grad_norm": 1.408058843216124,
+      "learning_rate": 1.644786910066003e-05,
+      "loss": 1.0792,
+      "step": 206
+    },
+    {
+      "epoch": 0.30010873504893076,
+      "grad_norm": 1.4031374392112927,
+      "learning_rate": 1.641185009349084e-05,
+      "loss": 1.0487,
+      "step": 207
+    },
+    {
+      "epoch": 0.30155853570134106,
+      "grad_norm": 1.3893476293660785,
+      "learning_rate": 1.637568926896699e-05,
+      "loss": 1.0713,
+      "step": 208
+    },
+    {
+      "epoch": 0.30300833635375135,
+      "grad_norm": 1.4103093205858785,
+      "learning_rate": 1.633938742689392e-05,
+      "loss": 1.075,
+      "step": 209
+    },
+    {
+      "epoch": 0.30445813700616164,
+      "grad_norm": 1.3745055025730226,
+      "learning_rate": 1.6302945370196074e-05,
+      "loss": 1.1094,
+      "step": 210
+    },
+    {
+      "epoch": 0.30590793765857194,
+      "grad_norm": 1.3694261204194211,
+      "learning_rate": 1.626636390489916e-05,
+      "loss": 1.1072,
+      "step": 211
+    },
+    {
+      "epoch": 0.30735773831098223,
+      "grad_norm": 1.4800523539333257,
+      "learning_rate": 1.6229643840112355e-05,
+      "loss": 1.1181,
+      "step": 212
+    },
+    {
+      "epoch": 0.3088075389633925,
+      "grad_norm": 1.502911751017136,
+      "learning_rate": 1.6192785988010354e-05,
+      "loss": 1.1193,
+      "step": 213
+    },
+    {
+      "epoch": 0.3102573396158028,
+      "grad_norm": 1.3613301317236974,
+      "learning_rate": 1.6155791163815438e-05,
+      "loss": 1.0224,
+      "step": 214
+    },
+    {
+      "epoch": 0.3117071402682131,
+      "grad_norm": 1.38369711210631,
+      "learning_rate": 1.6118660185779448e-05,
+      "loss": 1.0496,
+      "step": 215
+    },
+    {
+      "epoch": 0.3131569409206234,
+      "grad_norm": 1.3539618212963933,
+      "learning_rate": 1.6081393875165665e-05,
+      "loss": 1.0669,
+      "step": 216
+    },
+    {
+      "epoch": 0.3146067415730337,
+      "grad_norm": 1.4020038827525751,
+      "learning_rate": 1.604399305623067e-05,
+      "loss": 1.1059,
+      "step": 217
+    },
+    {
+      "epoch": 0.316056542225444,
+      "grad_norm": 1.4108834565634198,
+      "learning_rate": 1.6006458556206087e-05,
+      "loss": 1.0657,
+      "step": 218
+    },
+    {
+      "epoch": 0.3175063428778543,
+      "grad_norm": 1.326125815596232,
+      "learning_rate": 1.5968791205280306e-05,
+      "loss": 1.0886,
+      "step": 219
+    },
+    {
+      "epoch": 0.3189561435302646,
+      "grad_norm": 1.2936076784563293,
+      "learning_rate": 1.593099183658012e-05,
+      "loss": 1.0675,
+      "step": 220
+    },
+    {
+      "epoch": 0.3204059441826749,
+      "grad_norm": 1.2687920081844093,
+      "learning_rate": 1.5893061286152275e-05,
+      "loss": 1.1078,
+      "step": 221
+    },
+    {
+      "epoch": 0.32185574483508517,
+      "grad_norm": 1.3201305501410263,
+      "learning_rate": 1.585500039294501e-05,
+      "loss": 1.0779,
+      "step": 222
+    },
+    {
+      "epoch": 0.32330554548749546,
+      "grad_norm": 1.3159779799020028,
+      "learning_rate": 1.5816809998789493e-05,
+      "loss": 1.0612,
+      "step": 223
+    },
+    {
+      "epoch": 0.32475534613990575,
+      "grad_norm": 1.304975228889699,
+      "learning_rate": 1.5778490948381175e-05,
+      "loss": 1.1662,
+      "step": 224
+    },
+    {
+      "epoch": 0.32620514679231605,
+      "grad_norm": 1.354747863212415,
+      "learning_rate": 1.5740044089261137e-05,
+      "loss": 1.1426,
+      "step": 225
+    },
+    {
+      "epoch": 0.32765494744472634,
+      "grad_norm": 1.4184761333698903,
+      "learning_rate": 1.570147027179734e-05,
+      "loss": 0.9907,
+      "step": 226
+    },
+    {
+      "epoch": 0.32910474809713663,
+      "grad_norm": 1.3180474937954418,
+      "learning_rate": 1.5662770349165802e-05,
+      "loss": 1.0448,
+      "step": 227
+    },
+    {
+      "epoch": 0.33055454874954693,
+      "grad_norm": 1.5730458083431538,
+      "learning_rate": 1.562394517733174e-05,
+      "loss": 1.1055,
+      "step": 228
+    },
+    {
+      "epoch": 0.3320043494019572,
+      "grad_norm": 1.2578608779840017,
+      "learning_rate": 1.5584995615030634e-05,
+      "loss": 1.0582,
+      "step": 229
+    },
+    {
+      "epoch": 0.3334541500543675,
+      "grad_norm": 1.3028312790422167,
+      "learning_rate": 1.554592252374923e-05,
+      "loss": 1.1221,
+      "step": 230
+    },
+    {
+      "epoch": 0.3349039507067778,
+      "grad_norm": 1.4226986687796697,
+      "learning_rate": 1.55067267677065e-05,
+      "loss": 1.0991,
+      "step": 231
+    },
+    {
+      "epoch": 0.3363537513591881,
+      "grad_norm": 1.3528185911897137,
+      "learning_rate": 1.5467409213834496e-05,
+      "loss": 1.0415,
+      "step": 232
+    },
+    {
+      "epoch": 0.3378035520115984,
+      "grad_norm": 1.378923222061165,
+      "learning_rate": 1.5427970731759214e-05,
+      "loss": 1.1118,
+      "step": 233
+    },
+    {
+      "epoch": 0.3392533526640087,
+      "grad_norm": 1.3034689008803908,
+      "learning_rate": 1.5388412193781334e-05,
+      "loss": 1.079,
+      "step": 234
+    },
+    {
+      "epoch": 0.340703153316419,
+      "grad_norm": 1.3482623877302018,
+      "learning_rate": 1.5348734474856934e-05,
+      "loss": 1.0855,
+      "step": 235
+    },
+    {
+      "epoch": 0.3421529539688293,
+      "grad_norm": 1.2894162016498965,
+      "learning_rate": 1.5308938452578136e-05,
+      "loss": 1.1034,
+      "step": 236
+    },
+    {
+      "epoch": 0.34360275462123957,
+      "grad_norm": 1.3262428062937586,
+      "learning_rate": 1.52690250071537e-05,
+      "loss": 1.1017,
+      "step": 237
+    },
+    {
+      "epoch": 0.34505255527364986,
+      "grad_norm": 1.2838917861888357,
+      "learning_rate": 1.5228995021389548e-05,
+      "loss": 1.1039,
+      "step": 238
+    },
+    {
+      "epoch": 0.34650235592606016,
+      "grad_norm": 1.3543990371654708,
+      "learning_rate": 1.5188849380669241e-05,
+      "loss": 1.0308,
+      "step": 239
+    },
+    {
+      "epoch": 0.34795215657847045,
+      "grad_norm": 1.4052033380855233,
+      "learning_rate": 1.5148588972934408e-05,
+      "loss": 1.1384,
+      "step": 240
+    },
+    {
+      "epoch": 0.34940195723088074,
+      "grad_norm": 1.349865692433653,
+      "learning_rate": 1.510821468866508e-05,
+      "loss": 1.1298,
+      "step": 241
+    },
+    {
+      "epoch": 0.35085175788329104,
+      "grad_norm": 1.2557353783353449,
+      "learning_rate": 1.5067727420860023e-05,
+      "loss": 1.0146,
+      "step": 242
+    },
+    {
+      "epoch": 0.35230155853570133,
+      "grad_norm": 1.3200472632000788,
+      "learning_rate": 1.5027128065016969e-05,
+      "loss": 1.0429,
+      "step": 243
+    },
+    {
+      "epoch": 0.3537513591881116,
+      "grad_norm": 1.3679553593283058,
+      "learning_rate": 1.4986417519112813e-05,
+      "loss": 1.0228,
+      "step": 244
+    },
+    {
+      "epoch": 0.3552011598405219,
+      "grad_norm": 1.292270014911043,
+      "learning_rate": 1.4945596683583754e-05,
+      "loss": 1.0589,
+      "step": 245
+    },
+    {
+      "epoch": 0.3566509604929322,
+      "grad_norm": 1.2528597538661148,
+      "learning_rate": 1.4904666461305377e-05,
+      "loss": 0.984,
+      "step": 246
+    },
+    {
+      "epoch": 0.3581007611453425,
+      "grad_norm": 1.3423177010754677,
+      "learning_rate": 1.4863627757572687e-05,
+      "loss": 1.0421,
+      "step": 247
+    },
+    {
+      "epoch": 0.3595505617977528,
+      "grad_norm": 1.6296373281468828,
+      "learning_rate": 1.4822481480080079e-05,
+      "loss": 1.0831,
+      "step": 248
+    },
+    {
+      "epoch": 0.3610003624501631,
+      "grad_norm": 1.4633418165145986,
+      "learning_rate": 1.4781228538901267e-05,
+      "loss": 1.1172,
+      "step": 249
+    },
+    {
+      "epoch": 0.3624501631025734,
+      "grad_norm": 1.3688765477502276,
+      "learning_rate": 1.4739869846469155e-05,
+      "loss": 1.1529,
+      "step": 250
+    },
+    {
+      "epoch": 0.3638999637549837,
+      "grad_norm": 1.2665716310170894,
+      "learning_rate": 1.4698406317555652e-05,
+      "loss": 1.0345,
+      "step": 251
+    },
+    {
+      "epoch": 0.365349764407394,
+      "grad_norm": 1.3561335644837103,
+      "learning_rate": 1.465683886925144e-05,
+      "loss": 1.0893,
+      "step": 252
+    },
+    {
+      "epoch": 0.36679956505980427,
+      "grad_norm": 1.3418759970034684,
+      "learning_rate": 1.4615168420945689e-05,
+      "loss": 1.1062,
+      "step": 253
+    },
+    {
+      "epoch": 0.36824936571221456,
+      "grad_norm": 1.3616195889628302,
+      "learning_rate": 1.4573395894305736e-05,
+      "loss": 0.9982,
+      "step": 254
+    },
+    {
+      "epoch": 0.36969916636462485,
+      "grad_norm": 1.432382317426238,
+      "learning_rate": 1.4531522213256675e-05,
+      "loss": 1.0987,
+      "step": 255
+    },
+    {
+      "epoch": 0.37114896701703515,
+      "grad_norm": 1.3290769764085317,
+      "learning_rate": 1.4489548303960931e-05,
+      "loss": 1.03,
+      "step": 256
+    },
+    {
+      "epoch": 0.37259876766944544,
+      "grad_norm": 1.3458581991423038,
+      "learning_rate": 1.4447475094797796e-05,
+      "loss": 1.1052,
+      "step": 257
+    },
+    {
+      "epoch": 0.37404856832185573,
+      "grad_norm": 1.2532153084860576,
+      "learning_rate": 1.440530351634286e-05,
+      "loss": 1.1059,
+      "step": 258
+    },
+    {
+      "epoch": 0.37549836897426603,
+      "grad_norm": 1.419815062163657,
+      "learning_rate": 1.4363034501347457e-05,
+      "loss": 1.0728,
+      "step": 259
+    },
+    {
+      "epoch": 0.3769481696266763,
+      "grad_norm": 1.4581582525201235,
+      "learning_rate": 1.4320668984718013e-05,
+      "loss": 1.0366,
+      "step": 260
+    },
+    {
+      "epoch": 0.3783979702790866,
+      "grad_norm": 1.3628539921700007,
+      "learning_rate": 1.4278207903495388e-05,
+      "loss": 1.0115,
+      "step": 261
+    },
+    {
+      "epoch": 0.3798477709314969,
+      "grad_norm": 1.1732811856250036,
+      "learning_rate": 1.4235652196834135e-05,
+      "loss": 1.0383,
+      "step": 262
+    },
+    {
+      "epoch": 0.3812975715839072,
+      "grad_norm": 1.4172177383419562,
+      "learning_rate": 1.4193002805981731e-05,
+      "loss": 1.0211,
+      "step": 263
+    },
+    {
+      "epoch": 0.3827473722363175,
+      "grad_norm": 1.3663936582239447,
+      "learning_rate": 1.4150260674257767e-05,
+      "loss": 1.1243,
+      "step": 264
+    },
+    {
+      "epoch": 0.3841971728887278,
+      "grad_norm": 1.4498394639726826,
+      "learning_rate": 1.4107426747033075e-05,
+      "loss": 1.0731,
+      "step": 265
+    },
+    {
+      "epoch": 0.3856469735411381,
+      "grad_norm": 1.3201058485913977,
+      "learning_rate": 1.406450197170882e-05,
+      "loss": 1.0852,
+      "step": 266
+    },
+    {
+      "epoch": 0.3870967741935484,
+      "grad_norm": 1.47121604115621,
+      "learning_rate": 1.4021487297695539e-05,
+      "loss": 1.1067,
+      "step": 267
+    },
+    {
+      "epoch": 0.38854657484595867,
+      "grad_norm": 1.3140462827328072,
+      "learning_rate": 1.3978383676392166e-05,
+      "loss": 1.0835,
+      "step": 268
+    },
+    {
+      "epoch": 0.38999637549836896,
+      "grad_norm": 1.3262584010206409,
+      "learning_rate": 1.3935192061164957e-05,
+      "loss": 1.141,
+      "step": 269
+    },
+    {
+      "epoch": 0.39144617615077926,
+      "grad_norm": 1.2562209901458021,
+      "learning_rate": 1.3891913407326424e-05,
+      "loss": 1.0861,
+      "step": 270
+    },
+    {
+      "epoch": 0.39289597680318955,
+      "grad_norm": 1.299920602202958,
+      "learning_rate": 1.3848548672114197e-05,
+      "loss": 1.1055,
+      "step": 271
+    },
+    {
+      "epoch": 0.39434577745559984,
+      "grad_norm": 1.5533447336640442,
+      "learning_rate": 1.3805098814669857e-05,
+      "loss": 1.1142,
+      "step": 272
+    },
+    {
+      "epoch": 0.39579557810801014,
+      "grad_norm": 1.4509679929686607,
+      "learning_rate": 1.3761564796017724e-05,
+      "loss": 1.1048,
+      "step": 273
+    },
+    {
+      "epoch": 0.39724537876042043,
+      "grad_norm": 1.3839245284766264,
+      "learning_rate": 1.3717947579043584e-05,
+      "loss": 1.0982,
+      "step": 274
+    },
+    {
+      "epoch": 0.3986951794128307,
+      "grad_norm": 1.2758002292658532,
+      "learning_rate": 1.3674248128473417e-05,
+      "loss": 1.0631,
+      "step": 275
+    },
+    {
+      "epoch": 0.400144980065241,
+      "grad_norm": 1.3235081113421823,
+      "learning_rate": 1.3630467410852038e-05,
+      "loss": 1.0509,
+      "step": 276
+    },
+    {
+      "epoch": 0.4015947807176513,
+      "grad_norm": 1.3534599840626527,
+      "learning_rate": 1.3586606394521729e-05,
+      "loss": 1.0007,
+      "step": 277
+    },
+    {
+      "epoch": 0.4030445813700616,
+      "grad_norm": 1.5338787417623796,
+      "learning_rate": 1.3542666049600822e-05,
+      "loss": 0.9811,
+      "step": 278
+    },
+    {
+      "epoch": 0.4044943820224719,
+      "grad_norm": 1.3526061961306781,
+      "learning_rate": 1.349864734796224e-05,
+      "loss": 1.1228,
+      "step": 279
+    },
+    {
+      "epoch": 0.4059441826748822,
+      "grad_norm": 1.347353206438488,
+      "learning_rate": 1.3454551263211993e-05,
+      "loss": 1.019,
+      "step": 280
+    },
+    {
+      "epoch": 0.4073939833272925,
+      "grad_norm": 1.4186421852650903,
+      "learning_rate": 1.341037877066766e-05,
+      "loss": 1.0659,
+      "step": 281
+    },
+    {
+      "epoch": 0.4088437839797028,
+      "grad_norm": 1.4218416867360633,
+      "learning_rate": 1.3366130847336806e-05,
+      "loss": 1.0415,
+      "step": 282
+    },
+    {
+      "epoch": 0.4102935846321131,
+      "grad_norm": 1.8894860663570074,
+      "learning_rate": 1.3321808471895374e-05,
+      "loss": 1.0502,
+      "step": 283
+    },
+    {
+      "epoch": 0.41174338528452337,
+      "grad_norm": 1.2636218428805208,
+      "learning_rate": 1.3277412624666037e-05,
+      "loss": 1.0346,
+      "step": 284
+    },
+    {
+      "epoch": 0.41319318593693366,
+      "grad_norm": 1.3069669351148674,
+      "learning_rate": 1.323294428759652e-05,
+      "loss": 1.0241,
+      "step": 285
+    },
+    {
+      "epoch": 0.41464298658934395,
+      "grad_norm": 1.3510826056533585,
+      "learning_rate": 1.3188404444237883e-05,
+      "loss": 1.0578,
+      "step": 286
+    },
+    {
+      "epoch": 0.41609278724175425,
+      "grad_norm": 1.2829266217180768,
+      "learning_rate": 1.3143794079722754e-05,
+      "loss": 1.0593,
+      "step": 287
+    },
+    {
+      "epoch": 0.41754258789416454,
+      "grad_norm": 1.3058530594657887,
+      "learning_rate": 1.309911418074355e-05,
+      "loss": 1.1643,
+      "step": 288
+    },
+    {
+      "epoch": 0.41899238854657483,
+      "grad_norm": 1.3418953016132162,
+      "learning_rate": 1.3054365735530666e-05,
+      "loss": 0.9297,
+      "step": 289
+    },
+    {
+      "epoch": 0.42044218919898513,
+      "grad_norm": 1.5001549087946748,
+      "learning_rate": 1.3009549733830582e-05,
+      "loss": 1.1337,
+      "step": 290
+    },
+    {
+      "epoch": 0.4218919898513954,
+      "grad_norm": 1.3258715314662795,
+      "learning_rate": 1.2964667166884005e-05,
+      "loss": 1.0307,
+      "step": 291
+    },
+    {
+      "epoch": 0.4233417905038057,
+      "grad_norm": 1.3576369399235355,
+      "learning_rate": 1.2919719027403935e-05,
+      "loss": 1.0963,
+      "step": 292
+    },
+    {
+      "epoch": 0.424791591156216,
+      "grad_norm": 1.3252741449377512,
+      "learning_rate": 1.2874706309553697e-05,
+      "loss": 1.1408,
+      "step": 293
+    },
+    {
+      "epoch": 0.4262413918086263,
+      "grad_norm": 1.4908231168535822,
+      "learning_rate": 1.2829630008924966e-05,
+      "loss": 1.1294,
+      "step": 294
+    },
+    {
+      "epoch": 0.4276911924610366,
+      "grad_norm": 1.424295780363766,
+      "learning_rate": 1.2784491122515743e-05,
+      "loss": 1.0084,
+      "step": 295
+    },
+    {
+      "epoch": 0.4291409931134469,
+      "grad_norm": 1.3848017149501615,
+      "learning_rate": 1.2739290648708298e-05,
+      "loss": 1.042,
+      "step": 296
+    },
+    {
+      "epoch": 0.4305907937658572,
+      "grad_norm": 1.299106871263591,
+      "learning_rate": 1.2694029587247097e-05,
+      "loss": 1.0464,
+      "step": 297
+    },
+    {
+      "epoch": 0.4320405944182675,
+      "grad_norm": 1.3119601070089786,
+      "learning_rate": 1.2648708939216671e-05,
+      "loss": 1.0395,
+      "step": 298
+    },
+    {
+      "epoch": 0.43349039507067777,
+      "grad_norm": 1.402639991172934,
+      "learning_rate": 1.260332970701951e-05,
+      "loss": 1.1025,
+      "step": 299
+    },
+    {
+      "epoch": 0.43494019572308806,
+      "grad_norm": 1.4212208708281426,
+      "learning_rate": 1.2557892894353848e-05,
+      "loss": 1.0808,
+      "step": 300
+    },
+    {
+      "epoch": 0.43638999637549836,
+      "grad_norm": 1.3107344640436431,
+      "learning_rate": 1.251239950619149e-05,
+      "loss": 1.1174,
+      "step": 301
+    },
+    {
+      "epoch": 0.43783979702790865,
+      "grad_norm": 1.5588219140088742,
+      "learning_rate": 1.2466850548755586e-05,
+      "loss": 1.0293,
+      "step": 302
+    },
+    {
+      "epoch": 0.43928959768031894,
+      "grad_norm": 1.50956276436452,
+      "learning_rate": 1.242124702949836e-05,
+      "loss": 1.0333,
+      "step": 303
+    },
+    {
+      "epoch": 0.44073939833272924,
+      "grad_norm": 1.3379490165652188,
+      "learning_rate": 1.237558995707884e-05,
+      "loss": 1.0095,
+      "step": 304
+    },
+    {
+      "epoch": 0.44218919898513953,
+      "grad_norm": 1.4039594834230462,
+      "learning_rate": 1.2329880341340533e-05,
+      "loss": 1.0551,
+      "step": 305
+    },
+    {
+      "epoch": 0.4436389996375498,
+      "grad_norm": 1.3479709521521714,
+      "learning_rate": 1.2284119193289114e-05,
+      "loss": 1.0486,
+      "step": 306
+    },
+    {
+      "epoch": 0.4450888002899601,
+      "grad_norm": 1.3969199369845706,
+      "learning_rate": 1.223830752507004e-05,
+      "loss": 1.0763,
+      "step": 307
+    },
+    {
+      "epoch": 0.4465386009423704,
+      "grad_norm": 1.4103637429370277,
+      "learning_rate": 1.2192446349946181e-05,
+      "loss": 1.0086,
+      "step": 308
+    },
+    {
+      "epoch": 0.4479884015947807,
+      "grad_norm": 1.3755262780276987,
+      "learning_rate": 1.2146536682275388e-05,
+      "loss": 1.0495,
+      "step": 309
+    },
+    {
+      "epoch": 0.449438202247191,
+      "grad_norm": 1.4681165441323079,
+      "learning_rate": 1.2100579537488084e-05,
+      "loss": 0.939,
+      "step": 310
+    },
+    {
+      "epoch": 0.4508880028996013,
+      "grad_norm": 1.505040124294731,
+      "learning_rate": 1.2054575932064787e-05,
+      "loss": 1.1042,
+      "step": 311
+    },
+    {
+      "epoch": 0.4523378035520116,
+      "grad_norm": 1.2696631491153292,
+      "learning_rate": 1.200852688351363e-05,
+      "loss": 1.0819,
+      "step": 312
+    },
+    {
+      "epoch": 0.4537876042044219,
+      "grad_norm": 1.404807604951442,
+      "learning_rate": 1.196243341034786e-05,
+      "loss": 1.0588,
+      "step": 313
+    },
+    {
+      "epoch": 0.4552374048568322,
+      "grad_norm": 1.435086378865791,
+      "learning_rate": 1.1916296532063308e-05,
+      "loss": 1.096,
+      "step": 314
+    },
+    {
+      "epoch": 0.45668720550924247,
+      "grad_norm": 1.3561249463865293,
+      "learning_rate": 1.1870117269115833e-05,
+      "loss": 1.1217,
+      "step": 315
+    },
+    {
+      "epoch": 0.45813700616165276,
+      "grad_norm": 1.33752541435148,
+      "learning_rate": 1.1823896642898769e-05,
+      "loss": 1.0993,
+      "step": 316
+    },
+    {
+      "epoch": 0.45958680681406305,
+      "grad_norm": 1.3115033949341803,
+      "learning_rate": 1.1777635675720313e-05,
+      "loss": 1.0268,
+      "step": 317
+    },
+    {
+      "epoch": 0.46103660746647335,
+      "grad_norm": 1.4765418406375446,
+      "learning_rate": 1.1731335390780935e-05,
+      "loss": 1.0539,
+      "step": 318
+    },
+    {
+      "epoch": 0.46248640811888364,
+      "grad_norm": 1.5509309922226189,
+      "learning_rate": 1.1684996812150726e-05,
+      "loss": 1.1008,
+      "step": 319
+    },
+    {
+      "epoch": 0.46393620877129393,
+      "grad_norm": 1.4328064073112206,
+      "learning_rate": 1.1638620964746761e-05,
+      "loss": 1.099,
+      "step": 320
+    },
+    {
+      "epoch": 0.4653860094237042,
+      "grad_norm": 1.3290715693303645,
+      "learning_rate": 1.1592208874310426e-05,
+      "loss": 1.0526,
+      "step": 321
+    },
+    {
+      "epoch": 0.4668358100761145,
+      "grad_norm": 1.2864531964749908,
+      "learning_rate": 1.1545761567384725e-05,
+      "loss": 0.9944,
+      "step": 322
+    },
+    {
+      "epoch": 0.4682856107285248,
+      "grad_norm": 1.3507427442705382,
+      "learning_rate": 1.1499280071291596e-05,
+      "loss": 1.0177,
+      "step": 323
+    },
+    {
+      "epoch": 0.4697354113809351,
+      "grad_norm": 1.228671083860629,
+      "learning_rate": 1.1452765414109155e-05,
+      "loss": 1.1236,
+      "step": 324
+    },
+    {
+      "epoch": 0.4711852120333454,
+      "grad_norm": 1.2785139150063036,
+      "learning_rate": 1.1406218624648986e-05,
+      "loss": 1.0512,
+      "step": 325
+    },
+    {
+      "epoch": 0.4726350126857557,
+      "grad_norm": 1.412993691271352,
+      "learning_rate": 1.1359640732433369e-05,
+      "loss": 0.9775,
+      "step": 326
+    },
+    {
+      "epoch": 0.474084813338166,
+      "grad_norm": 1.19591741403552,
+      "learning_rate": 1.1313032767672525e-05,
+      "loss": 1.1002,
+      "step": 327
+    },
+    {
+      "epoch": 0.4755346139905763,
+      "grad_norm": 1.242808153249587,
+      "learning_rate": 1.1266395761241809e-05,
+      "loss": 1.0603,
+      "step": 328
+    },
+    {
+      "epoch": 0.4769844146429866,
+      "grad_norm": 1.2303310879959044,
+      "learning_rate": 1.1219730744658921e-05,
+      "loss": 0.9991,
+      "step": 329
+    },
+    {
+      "epoch": 0.47843421529539687,
+      "grad_norm": 1.221527930479733,
+      "learning_rate": 1.1173038750061099e-05,
+      "loss": 0.9804,
+      "step": 330
+    },
+    {
+      "epoch": 0.47988401594780716,
+      "grad_norm": 1.4075532089201872,
+      "learning_rate": 1.1126320810182271e-05,
+      "loss": 1.0552,
+      "step": 331
+    },
+    {
+      "epoch": 0.48133381660021746,
+      "grad_norm": 1.3769455406866258,
+      "learning_rate": 1.107957795833023e-05,
+      "loss": 1.017,
+      "step": 332
+    },
+    {
+      "epoch": 0.48278361725262775,
+      "grad_norm": 1.247905110134898,
+      "learning_rate": 1.1032811228363766e-05,
+      "loss": 0.9647,
+      "step": 333
+    },
+    {
+      "epoch": 0.48423341790503804,
+      "grad_norm": 1.333854159973312,
+      "learning_rate": 1.0986021654669807e-05,
+      "loss": 1.0218,
+      "step": 334
+    },
+    {
+      "epoch": 0.48568321855744834,
+      "grad_norm": 1.31215798422068,
+      "learning_rate": 1.0939210272140545e-05,
+      "loss": 1.1878,
+      "step": 335
+    },
+    {
+      "epoch": 0.48713301920985863,
+      "grad_norm": 1.276405471136989,
+      "learning_rate": 1.089237811615053e-05,
+      "loss": 1.0415,
+      "step": 336
+    },
+    {
+      "epoch": 0.4885828198622689,
+      "grad_norm": 1.3852346792500927,
+      "learning_rate": 1.0845526222533791e-05,
+      "loss": 1.0113,
+      "step": 337
+    },
+    {
+      "epoch": 0.4900326205146792,
+      "grad_norm": 1.3612523288427314,
+      "learning_rate": 1.07986556275609e-05,
+      "loss": 0.9802,
+      "step": 338
+    },
+    {
+      "epoch": 0.4914824211670895,
+      "grad_norm": 1.4034844318979187,
+      "learning_rate": 1.0751767367916081e-05,
+      "loss": 1.0656,
+      "step": 339
+    },
+    {
+      "epoch": 0.4929322218194998,
+      "grad_norm": 1.3624420745089656,
+      "learning_rate": 1.0704862480674251e-05,
+      "loss": 1.0493,
+      "step": 340
+    },
+    {
+      "epoch": 0.4943820224719101,
+      "grad_norm": 1.3060794829555848,
+      "learning_rate": 1.0657942003278107e-05,
+      "loss": 1.0063,
+      "step": 341
+    },
+    {
+      "epoch": 0.4958318231243204,
+      "grad_norm": 1.200376057250413,
+      "learning_rate": 1.0611006973515165e-05,
+      "loss": 1.0114,
+      "step": 342
+    },
+    {
+      "epoch": 0.4972816237767307,
+      "grad_norm": 1.2724481286652172,
+      "learning_rate": 1.0564058429494808e-05,
+      "loss": 1.0728,
+      "step": 343
+    },
+    {
+      "epoch": 0.498731424429141,
+      "grad_norm": 1.3522641309772223,
+      "learning_rate": 1.0517097409625335e-05,
+      "loss": 1.1265,
+      "step": 344
+    },
+    {
+      "epoch": 0.5001812250815513,
+      "grad_norm": 1.571601256230779,
+      "learning_rate": 1.0470124952590976e-05,
+      "loss": 1.0285,
+      "step": 345
+    },
+    {
+      "epoch": 0.5016310257339616,
+      "grad_norm": 1.3172306892745251,
+      "learning_rate": 1.0423142097328938e-05,
+      "loss": 1.0355,
+      "step": 346
+    },
+    {
+      "epoch": 0.5016310257339616,
+      "eval_loss": 1.8076887130737305,
+      "eval_runtime": 336.9555,
+      "eval_samples_per_second": 52.393,
+      "eval_steps_per_second": 0.819,
+      "step": 346
+    },
+    {
+      "epoch": 0.5030808263863719,
+      "grad_norm": 1.448582832251904,
+      "learning_rate": 1.0376149883006406e-05,
+      "loss": 1.0642,
+      "step": 347
+    },
+    {
+      "epoch": 0.5045306270387822,
+      "grad_norm": 1.3531324873317523,
+      "learning_rate": 1.0329149348997576e-05,
+      "loss": 1.0621,
+      "step": 348
+    },
+    {
+      "epoch": 0.5059804276911924,
+      "grad_norm": 1.252411128408006,
+      "learning_rate": 1.0282141534860662e-05,
+      "loss": 1.0045,
+      "step": 349
+    },
+    {
+      "epoch": 0.5074302283436027,
+      "grad_norm": 1.3288435941350787,
+      "learning_rate": 1.0235127480314888e-05,
+      "loss": 1.0891,
+      "step": 350
+    },
+    {
+      "epoch": 0.508880028996013,
+      "grad_norm": 1.2319376588312267,
+      "learning_rate": 1.0188108225217516e-05,
+      "loss": 1.0598,
+      "step": 351
+    },
+    {
+      "epoch": 0.5103298296484233,
+      "grad_norm": 1.432271091703038,
+      "learning_rate": 1.0141084809540826e-05,
+      "loss": 1.1082,
+      "step": 352
+    },
+    {
+      "epoch": 0.5117796303008336,
+      "grad_norm": 1.2732706792889128,
+      "learning_rate": 1.0094058273349125e-05,
+      "loss": 1.0001,
+      "step": 353
+    },
+    {
+      "epoch": 0.5132294309532439,
+      "grad_norm": 1.4058301553796166,
+      "learning_rate": 1.0047029656775735e-05,
+      "loss": 1.0382,
+      "step": 354
+    },
+    {
+      "epoch": 0.5146792316056542,
+      "grad_norm": 1.265282899856481,
+      "learning_rate": 1e-05,
+      "loss": 1.0471,
+      "step": 355
+    },
+    {
+      "epoch": 0.5161290322580645,
+      "grad_norm": 1.3764437409384491,
+      "learning_rate": 9.952970343224268e-06,
+      "loss": 1.0262,
+      "step": 356
+    },
+    {
+      "epoch": 0.5175788329104748,
+      "grad_norm": 1.3534044766089743,
+      "learning_rate": 9.90594172665088e-06,
+      "loss": 1.0072,
+      "step": 357
+    },
+    {
+      "epoch": 0.5190286335628851,
+      "grad_norm": 1.3551460799960182,
+      "learning_rate": 9.858915190459179e-06,
+      "loss": 0.9829,
+      "step": 358
+    },
+    {
+      "epoch": 0.5204784342152954,
+      "grad_norm": 1.30627731426207,
+      "learning_rate": 9.811891774782484e-06,
+      "loss": 1.0645,
+      "step": 359
+    },
+    {
+      "epoch": 0.5219282348677057,
+      "grad_norm": 1.2938381830747794,
+      "learning_rate": 9.764872519685112e-06,
+      "loss": 0.9779,
+      "step": 360
+    },
+    {
+      "epoch": 0.523378035520116,
+      "grad_norm": 1.3917831737510398,
+      "learning_rate": 9.717858465139341e-06,
+      "loss": 1.1391,
+      "step": 361
+    },
+    {
+      "epoch": 0.5248278361725263,
+      "grad_norm": 1.2900463720917648,
+      "learning_rate": 9.670850651002427e-06,
+      "loss": 1.0437,
+      "step": 362
+    },
+    {
+      "epoch": 0.5262776368249366,
+      "grad_norm": 1.148673106741137,
+      "learning_rate": 9.623850116993598e-06,
+      "loss": 0.9422,
+      "step": 363
+    },
+    {
+      "epoch": 0.5277274374773469,
+      "grad_norm": 1.3375207113484797,
+      "learning_rate": 9.576857902671067e-06,
+      "loss": 1.0311,
+      "step": 364
+    },
+    {
+      "epoch": 0.5291772381297571,
+      "grad_norm": 1.4514970987707188,
+      "learning_rate": 9.529875047409027e-06,
+      "loss": 1.0235,
+      "step": 365
+    },
+    {
+      "epoch": 0.5306270387821674,
+      "grad_norm": 1.3146195007658552,
+      "learning_rate": 9.482902590374665e-06,
+      "loss": 1.054,
+      "step": 366
+    },
+    {
+      "epoch": 0.5320768394345777,
+      "grad_norm": 1.4465578134725754,
+      "learning_rate": 9.435941570505192e-06,
+      "loss": 1.0902,
+      "step": 367
+    },
+    {
+      "epoch": 0.533526640086988,
+      "grad_norm": 1.3149439811370123,
+      "learning_rate": 9.388993026484837e-06,
+      "loss": 1.0832,
+      "step": 368
+    },
+    {
+      "epoch": 0.5349764407393983,
+      "grad_norm": 1.3717860776183055,
+      "learning_rate": 9.342057996721894e-06,
+      "loss": 1.1313,
+      "step": 369
+    },
+    {
+      "epoch": 0.5364262413918086,
+      "grad_norm": 1.293824254802867,
+      "learning_rate": 9.295137519325752e-06,
+      "loss": 1.1008,
+      "step": 370
+    },
+    {
+      "epoch": 0.5378760420442189,
+      "grad_norm": 1.306197728205856,
+      "learning_rate": 9.248232632083924e-06,
+      "loss": 0.9883,
+      "step": 371
+    },
+    {
+      "epoch": 0.5393258426966292,
+      "grad_norm": 1.2825187891839944,
+      "learning_rate": 9.201344372439103e-06,
+      "loss": 1.0283,
+      "step": 372
+    },
+    {
+      "epoch": 0.5407756433490395,
+      "grad_norm": 1.3162089821796013,
+      "learning_rate": 9.154473777466209e-06,
+      "loss": 1.0649,
+      "step": 373
+    },
+    {
+      "epoch": 0.5422254440014498,
+      "grad_norm": 1.2640029408596194,
+      "learning_rate": 9.10762188384947e-06,
+      "loss": 1.0531,
+      "step": 374
+    },
+    {
+      "epoch": 0.5436752446538601,
+      "grad_norm": 1.3657812695251283,
+      "learning_rate": 9.060789727859457e-06,
+      "loss": 1.0382,
+      "step": 375
+    },
+    {
+      "epoch": 0.5451250453062704,
+      "grad_norm": 1.2322460555653036,
+      "learning_rate": 9.013978345330195e-06,
+      "loss": 1.0029,
+      "step": 376
+    },
+    {
+      "epoch": 0.5465748459586807,
+      "grad_norm": 1.2682895791722462,
+      "learning_rate": 8.967188771636237e-06,
+      "loss": 1.0692,
+      "step": 377
+    },
+    {
+      "epoch": 0.548024646611091,
+      "grad_norm": 1.521260849782353,
+      "learning_rate": 8.920422041669774e-06,
+      "loss": 1.1387,
+      "step": 378
+    },
+    {
+      "epoch": 0.5494744472635013,
+      "grad_norm": 1.3560842862084204,
+      "learning_rate": 8.873679189817732e-06,
+      "loss": 1.0457,
+      "step": 379
+    },
+    {
+      "epoch": 0.5509242479159115,
+      "grad_norm": 1.393292623957829,
+      "learning_rate": 8.826961249938903e-06,
+      "loss": 1.0099,
+      "step": 380
+    },
+    {
+      "epoch": 0.5523740485683218,
+      "grad_norm": 1.5176504362465584,
+      "learning_rate": 8.78026925534108e-06,
+      "loss": 1.0694,
+      "step": 381
+    },
+    {
+      "epoch": 0.5538238492207321,
+      "grad_norm": 1.3965155596241474,
+      "learning_rate": 8.733604238758196e-06,
+      "loss": 1.0683,
+      "step": 382
+    },
+    {
+      "epoch": 0.5552736498731424,
+      "grad_norm": 1.3693796520511425,
+      "learning_rate": 8.68696723232748e-06,
+      "loss": 1.0,
+      "step": 383
+    },
+    {
+      "epoch": 0.5567234505255527,
+      "grad_norm": 1.3249628611352424,
+      "learning_rate": 8.640359267566633e-06,
+      "loss": 1.0347,
+      "step": 384
+    },
+    {
+      "epoch": 0.558173251177963,
+      "grad_norm": 1.317777839549655,
+      "learning_rate": 8.59378137535102e-06,
+      "loss": 1.0645,
+      "step": 385
+    },
+    {
+      "epoch": 0.5596230518303733,
+      "grad_norm": 1.4006771241963718,
+      "learning_rate": 8.54723458589085e-06,
+      "loss": 1.125,
+      "step": 386
+    },
+    {
+      "epoch": 0.5610728524827836,
+      "grad_norm": 1.24960349625865,
+      "learning_rate": 8.500719928708406e-06,
+      "loss": 0.9478,
+      "step": 387
+    },
+    {
+      "epoch": 0.5625226531351939,
+      "grad_norm": 1.3561186145011772,
+      "learning_rate": 8.454238432615275e-06,
+      "loss": 1.0236,
+      "step": 388
+    },
+    {
+      "epoch": 0.5639724537876042,
+      "grad_norm": 1.4620185960733665,
+      "learning_rate": 8.407791125689577e-06,
+      "loss": 0.9982,
+      "step": 389
+    },
+    {
+      "epoch": 0.5654222544400145,
+      "grad_norm": 1.457736925823573,
+      "learning_rate": 8.36137903525324e-06,
+      "loss": 1.0525,
+      "step": 390
+    },
+    {
+      "epoch": 0.5668720550924248,
+      "grad_norm": 1.3870535512584614,
+      "learning_rate": 8.315003187849276e-06,
+      "loss": 1.088,
+      "step": 391
+    },
+    {
+      "epoch": 0.5683218557448351,
+      "grad_norm": 1.4705121841559596,
+      "learning_rate": 8.268664609219066e-06,
+      "loss": 1.0787,
+      "step": 392
+    },
+    {
+      "epoch": 0.5697716563972454,
+      "grad_norm": 1.3404706887210358,
+      "learning_rate": 8.222364324279689e-06,
+      "loss": 1.0631,
+      "step": 393
+    },
+    {
+      "epoch": 0.5712214570496557,
+      "grad_norm": 1.2900970416638202,
+      "learning_rate": 8.176103357101235e-06,
+      "loss": 1.0853,
+      "step": 394
+    },
+    {
+      "epoch": 0.572671257702066,
+      "grad_norm": 1.3595869736850361,
+      "learning_rate": 8.129882730884168e-06,
+      "loss": 1.0506,
+      "step": 395
+    },
+    {
+      "epoch": 0.5741210583544762,
+      "grad_norm": 1.505212604551167,
+      "learning_rate": 8.083703467936696e-06,
+      "loss": 0.9911,
+      "step": 396
+    },
+    {
+      "epoch": 0.5755708590068865,
+      "grad_norm": 1.3484771028026263,
+      "learning_rate": 8.037566589652141e-06,
+      "loss": 1.0766,
+      "step": 397
+    },
+    {
+      "epoch": 0.5770206596592968,
+      "grad_norm": 1.343964550355876,
+      "learning_rate": 7.991473116486372e-06,
+      "loss": 1.0144,
+      "step": 398
+    },
+    {
+      "epoch": 0.5784704603117071,
+      "grad_norm": 1.364356982221745,
+      "learning_rate": 7.945424067935216e-06,
+      "loss": 1.0378,
+      "step": 399
+    },
+    {
+      "epoch": 0.5799202609641174,
+      "grad_norm": 1.3258239693436584,
+      "learning_rate": 7.899420462511921e-06,
+      "loss": 1.0438,
+      "step": 400
+    },
+    {
+      "epoch": 0.5813700616165277,
+      "grad_norm": 1.3710707590736264,
+      "learning_rate": 7.853463317724614e-06,
+      "loss": 0.9477,
+      "step": 401
+    },
+    {
+      "epoch": 0.582819862268938,
+      "grad_norm": 1.6583685381554343,
+      "learning_rate": 7.807553650053824e-06,
+      "loss": 1.0053,
+      "step": 402
+    },
+    {
+      "epoch": 0.5842696629213483,
+      "grad_norm": 1.2431350846607607,
+      "learning_rate": 7.761692474929961e-06,
+      "loss": 0.9932,
+      "step": 403
+    },
+    {
+      "epoch": 0.5857194635737586,
+      "grad_norm": 1.4558285908873991,
+      "learning_rate": 7.715880806710888e-06,
+      "loss": 1.0134,
+      "step": 404
+    },
+    {
+      "epoch": 0.5871692642261689,
+      "grad_norm": 1.397858275795389,
+      "learning_rate": 7.670119658659469e-06,
+      "loss": 1.0292,
+      "step": 405
+    },
+    {
+      "epoch": 0.5886190648785792,
+      "grad_norm": 1.324682225646718,
+      "learning_rate": 7.6244100429211644e-06,
+      "loss": 1.1108,
+      "step": 406
+    },
+    {
+      "epoch": 0.5900688655309895,
+      "grad_norm": 1.297190226763353,
+      "learning_rate": 7.578752970501641e-06,
+      "loss": 1.0502,
+      "step": 407
+    },
+    {
+      "epoch": 0.5915186661833998,
+      "grad_norm": 1.4253621325556611,
+      "learning_rate": 7.533149451244416e-06,
+      "loss": 1.0062,
+      "step": 408
+    },
+    {
+      "epoch": 0.5929684668358101,
+      "grad_norm": 1.2843957601949123,
+      "learning_rate": 7.487600493808513e-06,
+      "loss": 1.0104,
+      "step": 409
+    },
+    {
+      "epoch": 0.5944182674882204,
+      "grad_norm": 1.2646268055936256,
+      "learning_rate": 7.442107105646156e-06,
+      "loss": 1.0627,
+      "step": 410
+    },
+    {
+      "epoch": 0.5958680681406306,
+      "grad_norm": 1.3074946851615719,
+      "learning_rate": 7.396670292980492e-06,
+      "loss": 1.0605,
+      "step": 411
+    },
+    {
+      "epoch": 0.5973178687930409,
+      "grad_norm": 1.374284698225198,
+      "learning_rate": 7.35129106078333e-06,
+      "loss": 1.0495,
+      "step": 412
+    },
+    {
+      "epoch": 0.5987676694454512,
+      "grad_norm": 1.2235812116447573,
+      "learning_rate": 7.30597041275291e-06,
+      "loss": 1.0223,
+      "step": 413
+    },
+    {
+      "epoch": 0.6002174700978615,
+      "grad_norm": 1.2965321779663,
+      "learning_rate": 7.260709351291706e-06,
+      "loss": 1.0764,
+      "step": 414
+    },
+    {
+      "epoch": 0.6016672707502718,
+      "grad_norm": 1.3480385370684136,
+      "learning_rate": 7.215508877484259e-06,
+      "loss": 0.9672,
+      "step": 415
+    },
+    {
+      "epoch": 0.6031170714026821,
+      "grad_norm": 1.2361558252999,
+      "learning_rate": 7.170369991075035e-06,
+      "loss": 1.0036,
+      "step": 416
+    },
+    {
+      "epoch": 0.6045668720550924,
+      "grad_norm": 1.2767108805725191,
+      "learning_rate": 7.125293690446307e-06,
+      "loss": 1.1183,
+      "step": 417
+    },
+    {
+      "epoch": 0.6060166727075027,
+      "grad_norm": 1.2530656367549364,
+      "learning_rate": 7.08028097259607e-06,
+      "loss": 0.9611,
+      "step": 418
+    },
+    {
+      "epoch": 0.607466473359913,
+      "grad_norm": 1.2483441140350913,
+      "learning_rate": 7.035332833115998e-06,
+      "loss": 1.0412,
+      "step": 419
+    },
+    {
+      "epoch": 0.6089162740123233,
+      "grad_norm": 1.351268565546029,
+      "learning_rate": 6.990450266169424e-06,
+      "loss": 0.9848,
+      "step": 420
+    },
+    {
+      "epoch": 0.6103660746647336,
+      "grad_norm": 1.409369585726132,
+      "learning_rate": 6.945634264469338e-06,
+      "loss": 1.052,
+      "step": 421
+    },
+    {
+      "epoch": 0.6118158753171439,
+      "grad_norm": 1.4747221037442269,
+      "learning_rate": 6.900885819256449e-06,
+      "loss": 1.0782,
+      "step": 422
+    },
+    {
+      "epoch": 0.6132656759695542,
+      "grad_norm": 1.499694952026654,
+      "learning_rate": 6.85620592027725e-06,
+      "loss": 1.0388,
+      "step": 423
+    },
+    {
+      "epoch": 0.6147154766219645,
+      "grad_norm": 1.2703410650691298,
+      "learning_rate": 6.811595555762118e-06,
+      "loss": 0.9975,
+      "step": 424
+    },
+    {
+      "epoch": 0.6161652772743748,
+      "grad_norm": 1.261516150924624,
+      "learning_rate": 6.767055712403481e-06,
+      "loss": 0.9403,
+      "step": 425
+    },
+    {
+      "epoch": 0.617615077926785,
+      "grad_norm": 1.2625249063597415,
+      "learning_rate": 6.722587375333967e-06,
+      "loss": 1.032,
+      "step": 426
+    },
+    {
+      "epoch": 0.6190648785791953,
+      "grad_norm": 1.3348915151430407,
+      "learning_rate": 6.678191528104629e-06,
+      "loss": 1.0424,
+      "step": 427
+    },
+    {
+      "epoch": 0.6205146792316056,
+      "grad_norm": 1.4958113972927707,
+      "learning_rate": 6.633869152663197e-06,
+      "loss": 1.0741,
+      "step": 428
+    },
+    {
+      "epoch": 0.6219644798840159,
+      "grad_norm": 1.2326273356689277,
+      "learning_rate": 6.58962122933234e-06,
+      "loss": 0.9413,
+      "step": 429
+    },
+    {
+      "epoch": 0.6234142805364262,
+      "grad_norm": 1.3040347595413484,
+      "learning_rate": 6.545448736788009e-06,
+      "loss": 1.0819,
+      "step": 430
+    },
+    {
+      "epoch": 0.6248640811888365,
+      "grad_norm": 1.1817607916335953,
+      "learning_rate": 6.501352652037765e-06,
+      "loss": 0.9096,
+      "step": 431
+    },
+    {
+      "epoch": 0.6263138818412468,
+      "grad_norm": 1.3566001110593033,
+      "learning_rate": 6.457333950399181e-06,
+      "loss": 0.9946,
+      "step": 432
+    },
+    {
+      "epoch": 0.6277636824936571,
+      "grad_norm": 1.3421079116970707,
+      "learning_rate": 6.413393605478275e-06,
+      "loss": 1.0104,
+      "step": 433
+    },
+    {
+      "epoch": 0.6292134831460674,
+      "grad_norm": 1.5462764870995802,
+      "learning_rate": 6.369532589147967e-06,
+      "loss": 1.0834,
+      "step": 434
+    },
+    {
+      "epoch": 0.6306632837984777,
+      "grad_norm": 1.267821394928889,
+      "learning_rate": 6.325751871526589e-06,
+      "loss": 0.9372,
+      "step": 435
+    },
+    {
+      "epoch": 0.632113084450888,
+      "grad_norm": 1.4170425335702417,
+      "learning_rate": 6.282052420956417e-06,
+      "loss": 0.9877,
+      "step": 436
+    },
+    {
+      "epoch": 0.6335628851032983,
+      "grad_norm": 1.4148091319251561,
+      "learning_rate": 6.238435203982278e-06,
+      "loss": 1.08,
+      "step": 437
+    },
+    {
+      "epoch": 0.6350126857557086,
+      "grad_norm": 1.2459107095516722,
+      "learning_rate": 6.194901185330145e-06,
+      "loss": 1.0294,
+      "step": 438
+    },
+    {
+      "epoch": 0.6364624864081189,
+      "grad_norm": 1.4494207729208453,
+      "learning_rate": 6.1514513278858055e-06,
+      "loss": 0.9989,
+      "step": 439
+    },
+    {
+      "epoch": 0.6379122870605292,
+      "grad_norm": 1.3852840193041625,
+      "learning_rate": 6.1080865926735814e-06,
+      "loss": 1.034,
+      "step": 440
+    },
+    {
+      "epoch": 0.6393620877129395,
+      "grad_norm": 1.2595777294589594,
+      "learning_rate": 6.064807938835046e-06,
+      "loss": 1.0758,
+      "step": 441
+    },
+    {
+      "epoch": 0.6408118883653497,
+      "grad_norm": 1.348458652590585,
+      "learning_rate": 6.021616323607834e-06,
+      "loss": 1.0769,
+      "step": 442
+    },
+    {
+      "epoch": 0.64226168901776,
+      "grad_norm": 1.3148084643010172,
+      "learning_rate": 5.978512702304461e-06,
+      "loss": 1.0039,
+      "step": 443
+    },
+    {
+      "epoch": 0.6437114896701703,
+      "grad_norm": 1.4509187805219024,
+      "learning_rate": 5.935498028291183e-06,
+      "loss": 1.0422,
+      "step": 444
+    },
+    {
+      "epoch": 0.6451612903225806,
+      "grad_norm": 1.3618080792391065,
+      "learning_rate": 5.892573252966926e-06,
+      "loss": 1.0356,
+      "step": 445
+    },
+    {
+      "epoch": 0.6466110909749909,
+      "grad_norm": 1.3989269424106427,
+      "learning_rate": 5.849739325742234e-06,
+      "loss": 1.0615,
+      "step": 446
+    },
+    {
+      "epoch": 0.6480608916274012,
+      "grad_norm": 1.240479674744889,
+      "learning_rate": 5.806997194018272e-06,
+      "loss": 1.0121,
+      "step": 447
+    },
+    {
+      "epoch": 0.6495106922798115,
+      "grad_norm": 1.3734387206797385,
+      "learning_rate": 5.7643478031658676e-06,
+      "loss": 0.981,
+      "step": 448
+    },
+    {
+      "epoch": 0.6509604929322218,
+      "grad_norm": 1.2657594345293466,
+      "learning_rate": 5.7217920965046115e-06,
+      "loss": 1.0068,
+      "step": 449
+    },
+    {
+      "epoch": 0.6524102935846321,
+      "grad_norm": 1.2288753201980551,
+      "learning_rate": 5.679331015281984e-06,
+      "loss": 0.9576,
+      "step": 450
+    },
+    {
+      "epoch": 0.6538600942370424,
+      "grad_norm": 1.2807404293524816,
+      "learning_rate": 5.636965498652544e-06,
+      "loss": 0.9926,
+      "step": 451
+    },
+    {
+      "epoch": 0.6553098948894527,
+      "grad_norm": 1.2835107746611496,
+      "learning_rate": 5.594696483657138e-06,
+      "loss": 1.0113,
+      "step": 452
+    },
+    {
+      "epoch": 0.656759695541863,
+      "grad_norm": 1.3601474178642867,
+      "learning_rate": 5.552524905202208e-06,
+      "loss": 1.0401,
+      "step": 453
+    },
+    {
+      "epoch": 0.6582094961942733,
+      "grad_norm": 1.3948276641269504,
+      "learning_rate": 5.5104516960390695e-06,
+      "loss": 1.1173,
+      "step": 454
+    },
+    {
+      "epoch": 0.6596592968466836,
+      "grad_norm": 1.1903638742292686,
+      "learning_rate": 5.468477786743334e-06,
+      "loss": 0.9996,
+      "step": 455
+    },
+    {
+      "epoch": 0.6611090974990939,
+      "grad_norm": 1.405709513195041,
+      "learning_rate": 5.4266041056942644e-06,
+      "loss": 0.9969,
+      "step": 456
+    },
+    {
+      "epoch": 0.6625588981515041,
+      "grad_norm": 1.3141799209288725,
+      "learning_rate": 5.384831579054313e-06,
+      "loss": 1.0837,
+      "step": 457
+    },
+    {
+      "epoch": 0.6640086988039144,
+      "grad_norm": 1.2093698715287247,
+      "learning_rate": 5.343161130748564e-06,
+      "loss": 1.0692,
+      "step": 458
+    },
+    {
+      "epoch": 0.6654584994563247,
+      "grad_norm": 1.2530804858141193,
+      "learning_rate": 5.3015936824443526e-06,
+      "loss": 0.9867,
+      "step": 459
+    },
+    {
+      "epoch": 0.666908300108735,
+      "grad_norm": 1.4599692630742325,
+      "learning_rate": 5.260130153530847e-06,
+      "loss": 1.0011,
+      "step": 460
+    },
+    {
+      "epoch": 0.6683581007611453,
+      "grad_norm": 1.2898082702683744,
+      "learning_rate": 5.218771461098733e-06,
+      "loss": 0.9377,
+      "step": 461
+    },
+    {
+      "epoch": 0.6698079014135556,
+      "grad_norm": 1.3834943292006796,
+      "learning_rate": 5.177518519919924e-06,
+      "loss": 1.0588,
+      "step": 462
+    },
+    {
+      "epoch": 0.6712577020659659,
+      "grad_norm": 1.2898439288467713,
+      "learning_rate": 5.136372242427316e-06,
+      "loss": 1.0437,
+      "step": 463
+    },
+    {
+      "epoch": 0.6727075027183762,
+      "grad_norm": 1.3735831016569546,
+      "learning_rate": 5.095333538694624e-06,
+      "loss": 0.9869,
+      "step": 464
+    },
+    {
+      "epoch": 0.6741573033707865,
+      "grad_norm": 1.2748893883683556,
+      "learning_rate": 5.054403316416247e-06,
+      "loss": 1.0656,
+      "step": 465
+    },
+    {
+      "epoch": 0.6756071040231968,
+      "grad_norm": 1.363590569907258,
+      "learning_rate": 5.0135824808871905e-06,
+      "loss": 1.1012,
+      "step": 466
+    },
+    {
+      "epoch": 0.6770569046756071,
+      "grad_norm": 1.331119604913152,
+      "learning_rate": 4.972871934983032e-06,
+      "loss": 0.9958,
+      "step": 467
+    },
+    {
+      "epoch": 0.6785067053280174,
+      "grad_norm": 1.23582130382964,
+      "learning_rate": 4.932272579139981e-06,
+      "loss": 0.9618,
+      "step": 468
+    },
+    {
+      "epoch": 0.6799565059804277,
+      "grad_norm": 1.3681264796520285,
+      "learning_rate": 4.891785311334924e-06,
+      "loss": 0.999,
+      "step": 469
+    },
+    {
+      "epoch": 0.681406306632838,
+      "grad_norm": 1.2674026204388216,
+      "learning_rate": 4.851411027065596e-06,
+      "loss": 1.0515,
+      "step": 470
+    },
+    {
+      "epoch": 0.6828561072852483,
+      "grad_norm": 1.3794573463974327,
+      "learning_rate": 4.811150619330759e-06,
+      "loss": 1.0494,
+      "step": 471
+    },
+    {
+      "epoch": 0.6843059079376586,
+      "grad_norm": 1.3291290323246547,
+      "learning_rate": 4.771004978610457e-06,
+      "loss": 0.9893,
+      "step": 472
+    },
+    {
+      "epoch": 0.6857557085900688,
+      "grad_norm": 1.316759948103253,
+      "learning_rate": 4.730974992846304e-06,
+      "loss": 1.0003,
+      "step": 473
+    },
+    {
+      "epoch": 0.6872055092424791,
+      "grad_norm": 1.2930117409598687,
+      "learning_rate": 4.6910615474218675e-06,
+      "loss": 0.883,
+      "step": 474
+    },
+    {
+      "epoch": 0.6886553098948894,
+      "grad_norm": 1.2498857911489059,
+      "learning_rate": 4.651265525143069e-06,
+      "loss": 1.0313,
+      "step": 475
+    },
+    {
+      "epoch": 0.6901051105472997,
+      "grad_norm": 1.4039764837348403,
+      "learning_rate": 4.611587806218669e-06,
+      "loss": 0.9196,
+      "step": 476
+    },
+    {
+      "epoch": 0.69155491119971,
+      "grad_norm": 1.4160641559583,
+      "learning_rate": 4.572029268240787e-06,
+      "loss": 1.0488,
+      "step": 477
+    },
+    {
+      "epoch": 0.6930047118521203,
+      "grad_norm": 1.3282490304481793,
+      "learning_rate": 4.532590786165509e-06,
+      "loss": 1.0459,
+      "step": 478
+    },
+    {
+      "epoch": 0.6944545125045306,
+      "grad_norm": 1.300202337489008,
+      "learning_rate": 4.4932732322935055e-06,
+      "loss": 1.0083,
+      "step": 479
+    },
+    {
+      "epoch": 0.6959043131569409,
+      "grad_norm": 1.2515417018832822,
+      "learning_rate": 4.45407747625077e-06,
+      "loss": 1.0919,
+      "step": 480
+    },
+    {
+      "epoch": 0.6973541138093512,
+      "grad_norm": 1.2575140895574586,
+      "learning_rate": 4.41500438496937e-06,
+      "loss": 0.9989,
+      "step": 481
+    },
+    {
+      "epoch": 0.6988039144617615,
+      "grad_norm": 1.257846489777046,
+      "learning_rate": 4.376054822668262e-06,
+      "loss": 0.9638,
+      "step": 482
+    },
+    {
+      "epoch": 0.7002537151141718,
+      "grad_norm": 1.3450038374789832,
+      "learning_rate": 4.3372296508342025e-06,
+      "loss": 1.0569,
+      "step": 483
+    },
+    {
+      "epoch": 0.7017035157665821,
+      "grad_norm": 1.3064730812529743,
+      "learning_rate": 4.298529728202661e-06,
+      "loss": 1.0499,
+      "step": 484
+    },
+    {
+      "epoch": 0.7031533164189924,
+      "grad_norm": 1.216060179680735,
+      "learning_rate": 4.2599559107388645e-06,
+      "loss": 0.9989,
+      "step": 485
+    },
+    {
+      "epoch": 0.7046031170714027,
+      "grad_norm": 1.403877454190549,
+      "learning_rate": 4.221509051618827e-06,
+      "loss": 1.0668,
+      "step": 486
+    },
+    {
+      "epoch": 0.706052917723813,
+      "grad_norm": 1.2471637872272192,
+      "learning_rate": 4.183190001210511e-06,
+      "loss": 0.9126,
+      "step": 487
+    },
+    {
+      "epoch": 0.7075027183762232,
+      "grad_norm": 1.5380437022811593,
+      "learning_rate": 4.1449996070549895e-06,
+      "loss": 1.0407,
+      "step": 488
+    },
+    {
+      "epoch": 0.7089525190286335,
+      "grad_norm": 1.3287415259209572,
+      "learning_rate": 4.10693871384773e-06,
+      "loss": 0.9752,
+      "step": 489
+    },
+    {
+      "epoch": 0.7104023196810438,
+      "grad_norm": 1.3816130867909022,
+      "learning_rate": 4.0690081634198865e-06,
+      "loss": 0.9326,
+      "step": 490
+    },
+    {
+      "epoch": 0.7118521203334541,
+      "grad_norm": 1.392882693605874,
+      "learning_rate": 4.031208794719695e-06,
+      "loss": 1.005,
+      "step": 491
+    },
+    {
+      "epoch": 0.7133019209858644,
+      "grad_norm": 1.1610226400724326,
+      "learning_rate": 3.993541443793914e-06,
+      "loss": 0.9714,
+      "step": 492
+    },
+    {
+      "epoch": 0.7147517216382747,
+      "grad_norm": 1.3366104155407779,
+      "learning_rate": 3.956006943769331e-06,
+      "loss": 0.9843,
+      "step": 493
+    },
+    {
+      "epoch": 0.716201522290685,
+      "grad_norm": 1.257302542172201,
+      "learning_rate": 3.918606124834337e-06,
+      "loss": 0.9814,
+      "step": 494
+    },
+    {
+      "epoch": 0.7176513229430953,
+      "grad_norm": 1.3101547776782458,
+      "learning_rate": 3.881339814220555e-06,
+      "loss": 1.0216,
+      "step": 495
+    },
+    {
+      "epoch": 0.7191011235955056,
+      "grad_norm": 1.3062561100861712,
+      "learning_rate": 3.844208836184567e-06,
+      "loss": 0.9686,
+      "step": 496
+    },
+    {
+      "epoch": 0.7205509242479159,
+      "grad_norm": 1.467540203840732,
+      "learning_rate": 3.8072140119896507e-06,
+      "loss": 1.046,
+      "step": 497
+    },
+    {
+      "epoch": 0.7220007249003262,
+      "grad_norm": 1.5199362784717665,
+      "learning_rate": 3.770356159887646e-06,
+      "loss": 1.0168,
+      "step": 498
+    },
+    {
+      "epoch": 0.7234505255527365,
+      "grad_norm": 1.3782512083139193,
+      "learning_rate": 3.733636095100838e-06,
+      "loss": 1.0853,
+      "step": 499
+    },
+    {
+      "epoch": 0.7249003262051468,
+      "grad_norm": 1.3445709367931769,
+      "learning_rate": 3.697054629803931e-06,
+      "loss": 1.0435,
+      "step": 500
+    },
+    {
+      "epoch": 0.7263501268575571,
+      "grad_norm": 1.2850228283494465,
+      "learning_rate": 3.6606125731060815e-06,
+      "loss": 1.047,
+      "step": 501
+    },
+    {
+      "epoch": 0.7277999275099674,
+      "grad_norm": 1.2178994330288275,
+      "learning_rate": 3.6243107310330117e-06,
+      "loss": 0.9586,
+      "step": 502
+    },
+    {
+      "epoch": 0.7292497281623777,
+      "grad_norm": 1.4308065363424847,
+      "learning_rate": 3.5881499065091663e-06,
+      "loss": 1.0557,
+      "step": 503
+    },
+    {
+      "epoch": 0.730699528814788,
+      "grad_norm": 1.270257919604597,
+      "learning_rate": 3.552130899339975e-06,
+      "loss": 0.9841,
+      "step": 504
+    },
+    {
+      "epoch": 0.7321493294671982,
+      "grad_norm": 1.267583121399084,
+      "learning_rate": 3.5162545061941332e-06,
+      "loss": 1.0217,
+      "step": 505
+    },
+    {
+      "epoch": 0.7335991301196085,
+      "grad_norm": 1.3938936411880694,
+      "learning_rate": 3.480521520586021e-06,
+      "loss": 1.0807,
+      "step": 506
+    },
+    {
+      "epoch": 0.7350489307720188,
+      "grad_norm": 1.299027061873236,
+      "learning_rate": 3.4449327328581075e-06,
+      "loss": 1.0381,
+      "step": 507
+    },
+    {
+      "epoch": 0.7364987314244291,
+      "grad_norm": 1.2442243095224228,
+      "learning_rate": 3.4094889301635036e-06,
+      "loss": 0.9005,
+      "step": 508
+    },
+    {
+      "epoch": 0.7379485320768394,
+      "grad_norm": 1.4602133625334406,
+      "learning_rate": 3.3741908964485415e-06,
+      "loss": 0.944,
+      "step": 509
+    },
+    {
+      "epoch": 0.7393983327292497,
+      "grad_norm": 1.317976015127126,
+      "learning_rate": 3.339039412435422e-06,
+      "loss": 1.118,
+      "step": 510
+    },
+    {
+      "epoch": 0.74084813338166,
+      "grad_norm": 1.2816895441611689,
+      "learning_rate": 3.304035255604974e-06,
+      "loss": 1.0406,
+      "step": 511
+    },
+    {
+      "epoch": 0.7422979340340703,
+      "grad_norm": 1.309820230684799,
+      "learning_rate": 3.2691792001794244e-06,
+      "loss": 1.0436,
+      "step": 512
+    },
+    {
+      "epoch": 0.7437477346864806,
+      "grad_norm": 1.3356269254644924,
+      "learning_rate": 3.2344720171053134e-06,
+      "loss": 1.0924,
+      "step": 513
+    },
+    {
+      "epoch": 0.7451975353388909,
+      "grad_norm": 1.579391805960199,
+      "learning_rate": 3.199914474036403e-06,
+      "loss": 0.994,
+      "step": 514
+    },
+    {
+      "epoch": 0.7466473359913012,
+      "grad_norm": 1.4558739786151553,
+      "learning_rate": 3.1655073353167285e-06,
+      "loss": 1.0675,
+      "step": 515
+    },
+    {
+      "epoch": 0.7480971366437115,
+      "grad_norm": 1.380168143694583,
+      "learning_rate": 3.1312513619636686e-06,
+      "loss": 1.1076,
+      "step": 516
+    },
+    {
+      "epoch": 0.7495469372961218,
+      "grad_norm": 1.5666416634993225,
+      "learning_rate": 3.09714731165114e-06,
+      "loss": 1.0345,
+      "step": 517
+    },
+    {
+      "epoch": 0.7509967379485321,
+      "grad_norm": 1.275179793489328,
+      "learning_rate": 3.0631959386928078e-06,
+      "loss": 1.0408,
+      "step": 518
+    },
+    {
+      "epoch": 0.7524465386009423,
+      "grad_norm": 1.3306512216979933,
+      "learning_rate": 3.0293979940254316e-06,
+      "loss": 1.0166,
+      "step": 519
+    },
+    {
+      "epoch": 0.7524465386009423,
+      "eval_loss": 1.7561454772949219,
+      "eval_runtime": 337.3172,
+      "eval_samples_per_second": 52.336,
+      "eval_steps_per_second": 0.818,
+      "step": 519
+    },
+    {
+      "epoch": 0.7538963392533526,
+      "grad_norm": 1.2133996732356642,
+      "learning_rate": 2.9957542251922354e-06,
+      "loss": 0.944,
+      "step": 520
+    },
+    {
+      "epoch": 0.7553461399057629,
+      "grad_norm": 1.3015002062584782,
+      "learning_rate": 2.9622653763263874e-06,
+      "loss": 0.9881,
+      "step": 521
+    },
+    {
+      "epoch": 0.7567959405581732,
+      "grad_norm": 1.3844376838396761,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 1.066,
+      "step": 522
+    },
+    {
+      "epoch": 0.7582457412105835,
+      "grad_norm": 1.414197971905587,
+      "learning_rate": 2.8957553978803897e-06,
+      "loss": 1.0488,
+      "step": 523
+    },
+    {
+      "epoch": 0.7596955418629938,
+      "grad_norm": 1.3136610736001848,
+      "learning_rate": 2.8627357393685117e-06,
+      "loss": 0.9604,
+      "step": 524
+    },
+    {
+      "epoch": 0.7611453425154041,
+      "grad_norm": 1.3365626607360204,
+      "learning_rate": 2.829873942927971e-06,
+      "loss": 1.0572,
+      "step": 525
+    },
+    {
+      "epoch": 0.7625951431678144,
+      "grad_norm": 1.2680429239871274,
+      "learning_rate": 2.797170735396263e-06,
+      "loss": 1.001,
+      "step": 526
+    },
+    {
+      "epoch": 0.7640449438202247,
+      "grad_norm": 1.4573051984832146,
+      "learning_rate": 2.7646268401032086e-06,
+      "loss": 1.1173,
+      "step": 527
+    },
+    {
+      "epoch": 0.765494744472635,
+      "grad_norm": 1.227615026200606,
+      "learning_rate": 2.732242976854962e-06,
+      "loss": 0.9341,
+      "step": 528
+    },
+    {
+      "epoch": 0.7669445451250453,
+      "grad_norm": 1.3846041663038822,
+      "learning_rate": 2.7000198619180795e-06,
+      "loss": 1.0296,
+      "step": 529
+    },
+    {
+      "epoch": 0.7683943457774556,
+      "grad_norm": 1.3023234466104958,
+      "learning_rate": 2.6679582080036957e-06,
+      "loss": 1.0227,
+      "step": 530
+    },
+    {
+      "epoch": 0.7698441464298659,
+      "grad_norm": 1.3565639751939609,
+      "learning_rate": 2.636058724251739e-06,
+      "loss": 1.0443,
+      "step": 531
+    },
+    {
+      "epoch": 0.7712939470822762,
+      "grad_norm": 1.2612126435451934,
+      "learning_rate": 2.6043221162152632e-06,
+      "loss": 0.9805,
+      "step": 532
+    },
+    {
+      "epoch": 0.7727437477346865,
+      "grad_norm": 1.3120593238081977,
+      "learning_rate": 2.572749085844829e-06,
+      "loss": 1.0482,
+      "step": 533
+    },
+    {
+      "epoch": 0.7741935483870968,
+      "grad_norm": 1.1853293814882375,
+      "learning_rate": 2.541340331472998e-06,
+      "loss": 1.074,
+      "step": 534
+    },
+    {
+      "epoch": 0.775643349039507,
+      "grad_norm": 1.3028595208984797,
+      "learning_rate": 2.510096547798857e-06,
+      "loss": 0.9442,
+      "step": 535
+    },
+    {
+      "epoch": 0.7770931496919173,
+      "grad_norm": 1.3029031259349746,
+      "learning_rate": 2.479018425872686e-06,
+      "loss": 0.958,
+      "step": 536
+    },
+    {
+      "epoch": 0.7785429503443276,
+      "grad_norm": 1.3678907780637193,
+      "learning_rate": 2.44810665308064e-06,
+      "loss": 1.0235,
+      "step": 537
+    },
+    {
+      "epoch": 0.7799927509967379,
+      "grad_norm": 1.2574446262912995,
+      "learning_rate": 2.4173619131295754e-06,
+      "loss": 1.1021,
+      "step": 538
+    },
+    {
+      "epoch": 0.7814425516491482,
+      "grad_norm": 1.2829828073406941,
+      "learning_rate": 2.3867848860319143e-06,
+      "loss": 1.0582,
+      "step": 539
+    },
+    {
+      "epoch": 0.7828923523015585,
+      "grad_norm": 1.375052207034499,
+      "learning_rate": 2.3563762480905917e-06,
+      "loss": 1.0744,
+      "step": 540
+    },
+    {
+      "epoch": 0.7843421529539688,
+      "grad_norm": 1.2297319430181648,
+      "learning_rate": 2.3261366718841307e-06,
+      "loss": 0.9738,
+      "step": 541
+    },
+    {
+      "epoch": 0.7857919536063791,
+      "grad_norm": 1.267540302473847,
+      "learning_rate": 2.296066826251725e-06,
+      "loss": 0.9741,
+      "step": 542
+    },
+    {
+      "epoch": 0.7872417542587894,
+      "grad_norm": 1.289236253971893,
+      "learning_rate": 2.266167376278481e-06,
+      "loss": 1.0434,
+      "step": 543
+    },
+    {
+      "epoch": 0.7886915549111997,
+      "grad_norm": 1.3632774164309054,
+      "learning_rate": 2.23643898328068e-06,
+      "loss": 1.0244,
+      "step": 544
+    },
+    {
+      "epoch": 0.79014135556361,
+      "grad_norm": 1.4168209885158227,
+      "learning_rate": 2.206882304791176e-06,
+      "loss": 1.0258,
+      "step": 545
+    },
+    {
+      "epoch": 0.7915911562160203,
+      "grad_norm": 1.226758120955334,
+      "learning_rate": 2.177497994544828e-06,
+      "loss": 1.0196,
+      "step": 546
+    },
+    {
+      "epoch": 0.7930409568684306,
+      "grad_norm": 1.3653110992205433,
+      "learning_rate": 2.148286702464063e-06,
+      "loss": 1.0012,
+      "step": 547
+    },
+    {
+      "epoch": 0.7944907575208409,
+      "grad_norm": 1.2727224041586365,
+      "learning_rate": 2.119249074644485e-06,
+      "loss": 0.9762,
+      "step": 548
+    },
+    {
+      "epoch": 0.7959405581732512,
+      "grad_norm": 1.4561299937732688,
+      "learning_rate": 2.090385753340596e-06,
+      "loss": 0.9852,
+      "step": 549
+    },
+    {
+      "epoch": 0.7973903588256614,
+      "grad_norm": 1.2785544497433292,
+      "learning_rate": 2.0616973769515737e-06,
+      "loss": 1.0714,
+      "step": 550
+    },
+    {
+      "epoch": 0.7988401594780717,
+      "grad_norm": 1.2887530788526531,
+      "learning_rate": 2.033184580007177e-06,
+      "loss": 0.9725,
+      "step": 551
+    },
+    {
+      "epoch": 0.800289960130482,
+      "grad_norm": 1.2819657019965556,
+      "learning_rate": 2.0048479931536833e-06,
+      "loss": 1.0229,
+      "step": 552
+    },
+    {
+      "epoch": 0.8017397607828923,
+      "grad_norm": 1.292792340896605,
+      "learning_rate": 1.976688243139965e-06,
+      "loss": 0.9534,
+      "step": 553
+    },
+    {
+      "epoch": 0.8031895614353026,
+      "grad_norm": 1.2779073467139206,
+      "learning_rate": 1.94870595280361e-06,
+      "loss": 0.9923,
+      "step": 554
+    },
+    {
+      "epoch": 0.8046393620877129,
+      "grad_norm": 1.4018766648787333,
+      "learning_rate": 1.920901741057153e-06,
+      "loss": 1.038,
+      "step": 555
+    },
+    {
+      "epoch": 0.8060891627401232,
+      "grad_norm": 1.336939551471127,
+      "learning_rate": 1.8932762228743884e-06,
+      "loss": 0.9882,
+      "step": 556
+    },
+    {
+      "epoch": 0.8075389633925335,
+      "grad_norm": 1.4215389334791384,
+      "learning_rate": 1.8658300092767546e-06,
+      "loss": 1.0968,
+      "step": 557
+    },
+    {
+      "epoch": 0.8089887640449438,
+      "grad_norm": 1.2231572393228993,
+      "learning_rate": 1.8385637073198426e-06,
+      "loss": 1.1011,
+      "step": 558
+    },
+    {
+      "epoch": 0.8104385646973541,
+      "grad_norm": 1.453187575428072,
+      "learning_rate": 1.8114779200799437e-06,
+      "loss": 1.0886,
+      "step": 559
+    },
+    {
+      "epoch": 0.8118883653497644,
+      "grad_norm": 1.3033205781286596,
+      "learning_rate": 1.7845732466407361e-06,
+      "loss": 1.0132,
+      "step": 560
+    },
+    {
+      "epoch": 0.8133381660021747,
+      "grad_norm": 1.23922416489567,
+      "learning_rate": 1.7578502820800048e-06,
+      "loss": 0.9818,
+      "step": 561
+    },
+    {
+      "epoch": 0.814787966654585,
+      "grad_norm": 1.2725050359585715,
+      "learning_rate": 1.731309617456518e-06,
+      "loss": 0.9207,
+      "step": 562
+    },
+    {
+      "epoch": 0.8162377673069953,
+      "grad_norm": 1.27403371132038,
+      "learning_rate": 1.7049518397969135e-06,
+      "loss": 1.0561,
+      "step": 563
+    },
+    {
+      "epoch": 0.8176875679594056,
+      "grad_norm": 1.2385826340741537,
+      "learning_rate": 1.6787775320827438e-06,
+      "loss": 1.0883,
+      "step": 564
+    },
+    {
+      "epoch": 0.8191373686118159,
+      "grad_norm": 1.3416806711685951,
+      "learning_rate": 1.652787273237565e-06,
+      "loss": 0.9715,
+      "step": 565
+    },
+    {
+      "epoch": 0.8205871692642261,
+      "grad_norm": 1.4616706640037913,
+      "learning_rate": 1.6269816381141446e-06,
+      "loss": 0.915,
+      "step": 566
+    },
+    {
+      "epoch": 0.8220369699166364,
+      "grad_norm": 1.429083031259375,
+      "learning_rate": 1.601361197481741e-06,
+      "loss": 1.0598,
+      "step": 567
+    },
+    {
+      "epoch": 0.8234867705690467,
+      "grad_norm": 1.4529881489877399,
+      "learning_rate": 1.5759265180134709e-06,
+      "loss": 1.0605,
+      "step": 568
+    },
+    {
+      "epoch": 0.824936571221457,
+      "grad_norm": 1.2488243220566153,
+      "learning_rate": 1.5506781622737943e-06,
+      "loss": 0.9953,
+      "step": 569
+    },
+    {
+      "epoch": 0.8263863718738673,
+      "grad_norm": 1.311200219775698,
+      "learning_rate": 1.5256166887060575e-06,
+      "loss": 1.0748,
+      "step": 570
+    },
+    {
+      "epoch": 0.8278361725262776,
+      "grad_norm": 1.30529044900903,
+      "learning_rate": 1.5007426516201451e-06,
+      "loss": 0.9651,
+      "step": 571
+    },
+    {
+      "epoch": 0.8292859731786879,
+      "grad_norm": 1.2574395772932054,
+      "learning_rate": 1.4760566011802168e-06,
+      "loss": 1.0695,
+      "step": 572
+    },
+    {
+      "epoch": 0.8307357738310982,
+      "grad_norm": 1.4195989361487449,
+      "learning_rate": 1.4515590833925508e-06,
+      "loss": 1.0078,
+      "step": 573
+    },
+    {
+      "epoch": 0.8321855744835085,
+      "grad_norm": 1.3054162868722192,
+      "learning_rate": 1.4272506400934461e-06,
+      "loss": 0.9843,
+      "step": 574
+    },
+    {
+      "epoch": 0.8336353751359188,
+      "grad_norm": 1.24663689810913,
+      "learning_rate": 1.4031318089372614e-06,
+      "loss": 0.9124,
+      "step": 575
+    },
+    {
+      "epoch": 0.8350851757883291,
+      "grad_norm": 1.2912554144001227,
+      "learning_rate": 1.379203123384507e-06,
+      "loss": 1.0028,
+      "step": 576
+    },
+    {
+      "epoch": 0.8365349764407394,
+      "grad_norm": 1.2528284513274668,
+      "learning_rate": 1.3554651126900564e-06,
+      "loss": 1.0523,
+      "step": 577
+    },
+    {
+      "epoch": 0.8379847770931497,
+      "grad_norm": 1.338480421694086,
+      "learning_rate": 1.331918301891425e-06,
+      "loss": 1.0918,
+      "step": 578
+    },
+    {
+      "epoch": 0.83943457774556,
+      "grad_norm": 1.2916114320863261,
+      "learning_rate": 1.3085632117971792e-06,
+      "loss": 0.9343,
+      "step": 579
+    },
+    {
+      "epoch": 0.8408843783979703,
+      "grad_norm": 1.3931965545189373,
+      "learning_rate": 1.285400358975395e-06,
+      "loss": 0.9977,
+      "step": 580
+    },
+    {
+      "epoch": 0.8423341790503805,
+      "grad_norm": 1.3455117902801017,
+      "learning_rate": 1.2624302557422475e-06,
+      "loss": 0.9954,
+      "step": 581
+    },
+    {
+      "epoch": 0.8437839797027908,
+      "grad_norm": 1.3257455309212303,
+      "learning_rate": 1.2396534101506774e-06,
+      "loss": 1.0259,
+      "step": 582
+    },
+    {
+      "epoch": 0.8452337803552011,
+      "grad_norm": 1.3098326535180438,
+      "learning_rate": 1.2170703259791472e-06,
+      "loss": 1.0041,
+      "step": 583
+    },
+    {
+      "epoch": 0.8466835810076114,
+      "grad_norm": 1.2527275441261705,
+      "learning_rate": 1.1946815027205027e-06,
+      "loss": 1.0391,
+      "step": 584
+    },
+    {
+      "epoch": 0.8481333816600217,
+      "grad_norm": 1.2514869611463302,
+      "learning_rate": 1.1724874355709259e-06,
+      "loss": 0.9788,
+      "step": 585
+    },
+    {
+      "epoch": 0.849583182312432,
+      "grad_norm": 1.2687402433604176,
+      "learning_rate": 1.1504886154189854e-06,
+      "loss": 0.9848,
+      "step": 586
+    },
+    {
+      "epoch": 0.8510329829648423,
+      "grad_norm": 1.1689013548706937,
+      "learning_rate": 1.1286855288347676e-06,
+      "loss": 1.0284,
+      "step": 587
+    },
+    {
+      "epoch": 0.8524827836172526,
+      "grad_norm": 1.1869466142011291,
+      "learning_rate": 1.1070786580591287e-06,
+      "loss": 1.0732,
+      "step": 588
+    },
+    {
+      "epoch": 0.8539325842696629,
+      "grad_norm": 1.235135055385479,
+      "learning_rate": 1.085668480993015e-06,
+      "loss": 0.9619,
+      "step": 589
+    },
+    {
+      "epoch": 0.8553823849220732,
+      "grad_norm": 1.6003564470955078,
+      "learning_rate": 1.0644554711869115e-06,
+      "loss": 1.0173,
+      "step": 590
+    },
+    {
+      "epoch": 0.8568321855744835,
+      "grad_norm": 1.3458888931211694,
+      "learning_rate": 1.0434400978303427e-06,
+      "loss": 0.9524,
+      "step": 591
+    },
+    {
+      "epoch": 0.8582819862268938,
+      "grad_norm": 1.40322828532142,
+      "learning_rate": 1.0226228257415183e-06,
+      "loss": 0.9223,
+      "step": 592
+    },
+    {
+      "epoch": 0.8597317868793041,
+      "grad_norm": 1.4316768182434179,
+      "learning_rate": 1.0020041153570349e-06,
+      "loss": 1.0509,
+      "step": 593
+    },
+    {
+      "epoch": 0.8611815875317144,
+      "grad_norm": 1.2138180519210295,
+      "learning_rate": 9.815844227217065e-07,
+      "loss": 0.9161,
+      "step": 594
+    },
+    {
+      "epoch": 0.8626313881841247,
+      "grad_norm": 1.249932808809491,
+      "learning_rate": 9.61364199478465e-07,
+      "loss": 0.9717,
+      "step": 595
+    },
+    {
+      "epoch": 0.864081188836535,
+      "grad_norm": 1.341917882897787,
+      "learning_rate": 9.413438928583807e-07,
+      "loss": 1.065,
+      "step": 596
+    },
+    {
+      "epoch": 0.8655309894889452,
+      "grad_norm": 1.3781540382948174,
+      "learning_rate": 9.215239456707636e-07,
+      "loss": 0.9665,
+      "step": 597
+    },
+    {
+      "epoch": 0.8669807901413555,
+      "grad_norm": 1.3241089475936643,
+      "learning_rate": 9.01904796293378e-07,
+      "loss": 1.0077,
+      "step": 598
+    },
+    {
+      "epoch": 0.8684305907937658,
+      "grad_norm": 1.3560107396518426,
+      "learning_rate": 8.824868786627305e-07,
+      "loss": 1.0072,
+      "step": 599
+    },
+    {
+      "epoch": 0.8698803914461761,
+      "grad_norm": 1.290694107049599,
+      "learning_rate": 8.632706222644893e-07,
+      "loss": 0.9804,
+      "step": 600
+    },
+    {
+      "epoch": 0.8713301920985864,
+      "grad_norm": 1.440724451257882,
+      "learning_rate": 8.442564521239782e-07,
+      "loss": 0.9833,
+      "step": 601
+    },
+    {
+      "epoch": 0.8727799927509967,
+      "grad_norm": 1.291754928966789,
+      "learning_rate": 8.254447887967698e-07,
+      "loss": 0.9356,
+      "step": 602
+    },
+    {
+      "epoch": 0.874229793403407,
+      "grad_norm": 1.487443781949088,
+      "learning_rate": 8.06836048359394e-07,
+      "loss": 1.0506,
+      "step": 603
+    },
+    {
+      "epoch": 0.8756795940558173,
+      "grad_norm": 1.3380341740989574,
+      "learning_rate": 7.884306424001309e-07,
+      "loss": 0.9808,
+      "step": 604
+    },
+    {
+      "epoch": 0.8771293947082276,
+      "grad_norm": 1.321192697297726,
+      "learning_rate": 7.70228978009907e-07,
+      "loss": 0.9753,
+      "step": 605
+    },
+    {
+      "epoch": 0.8785791953606379,
+      "grad_norm": 1.3760212072745275,
+      "learning_rate": 7.522314577732836e-07,
+      "loss": 0.9608,
+      "step": 606
+    },
+    {
+      "epoch": 0.8800289960130482,
+      "grad_norm": 1.3126721788742706,
+      "learning_rate": 7.344384797595717e-07,
+      "loss": 1.0463,
+      "step": 607
+    },
+    {
+      "epoch": 0.8814787966654585,
+      "grad_norm": 1.3003446374494485,
+      "learning_rate": 7.16850437514004e-07,
+      "loss": 0.9371,
+      "step": 608
+    },
+    {
+      "epoch": 0.8829285973178688,
+      "grad_norm": 1.2967704133219122,
+      "learning_rate": 6.994677200490508e-07,
+      "loss": 0.968,
+      "step": 609
+    },
+    {
+      "epoch": 0.8843783979702791,
+      "grad_norm": 1.338855766066182,
+      "learning_rate": 6.822907118358047e-07,
+      "loss": 0.9496,
+      "step": 610
+    },
+    {
+      "epoch": 0.8858281986226894,
+      "grad_norm": 1.3005852982857657,
+      "learning_rate": 6.65319792795479e-07,
+      "loss": 0.9956,
+      "step": 611
+    },
+    {
+      "epoch": 0.8872779992750996,
+      "grad_norm": 1.3675565314367832,
+      "learning_rate": 6.485553382910026e-07,
+      "loss": 1.0487,
+      "step": 612
+    },
+    {
+      "epoch": 0.8887277999275099,
+      "grad_norm": 1.4304253140048604,
+      "learning_rate": 6.319977191187232e-07,
+      "loss": 0.9604,
+      "step": 613
+    },
+    {
+      "epoch": 0.8901776005799202,
+      "grad_norm": 1.2814463914515632,
+      "learning_rate": 6.15647301500203e-07,
+      "loss": 0.9436,
+      "step": 614
+    },
+    {
+      "epoch": 0.8916274012323305,
+      "grad_norm": 1.291739796206516,
+      "learning_rate": 5.995044470741151e-07,
+      "loss": 0.9268,
+      "step": 615
+    },
+    {
+      "epoch": 0.8930772018847408,
+      "grad_norm": 1.4434286879926685,
+      "learning_rate": 5.835695128882512e-07,
+      "loss": 1.0615,
+      "step": 616
+    },
+    {
+      "epoch": 0.8945270025371511,
+      "grad_norm": 1.1516402516210897,
+      "learning_rate": 5.678428513916212e-07,
+      "loss": 0.9621,
+      "step": 617
+    },
+    {
+      "epoch": 0.8959768031895614,
+      "grad_norm": 1.5301589718837776,
+      "learning_rate": 5.523248104266577e-07,
+      "loss": 0.993,
+      "step": 618
+    },
+    {
+      "epoch": 0.8974266038419717,
+      "grad_norm": 1.3758002508578118,
+      "learning_rate": 5.370157332215209e-07,
+      "loss": 1.013,
+      "step": 619
+    },
+    {
+      "epoch": 0.898876404494382,
+      "grad_norm": 1.4523008877332442,
+      "learning_rate": 5.219159583825107e-07,
+      "loss": 0.969,
+      "step": 620
+    },
+    {
+      "epoch": 0.9003262051467923,
+      "grad_norm": 1.2802568056864034,
+      "learning_rate": 5.070258198865741e-07,
+      "loss": 0.969,
+      "step": 621
+    },
+    {
+      "epoch": 0.9017760057992026,
+      "grad_norm": 1.2935410612495564,
+      "learning_rate": 4.923456470739219e-07,
+      "loss": 0.9694,
+      "step": 622
+    },
+    {
+      "epoch": 0.9032258064516129,
+      "grad_norm": 1.3388952095300914,
+      "learning_rate": 4.778757646407361e-07,
+      "loss": 0.9951,
+      "step": 623
+    },
+    {
+      "epoch": 0.9046756071040232,
+      "grad_norm": 1.2740474678923634,
+      "learning_rate": 4.636164926320064e-07,
+      "loss": 1.0014,
+      "step": 624
+    },
+    {
+      "epoch": 0.9061254077564335,
+      "grad_norm": 1.2302959597531815,
+      "learning_rate": 4.495681464344259e-07,
+      "loss": 0.9721,
+      "step": 625
+    },
+    {
+      "epoch": 0.9075752084088438,
+      "grad_norm": 1.441624990440475,
+      "learning_rate": 4.3573103676943784e-07,
+      "loss": 1.0521,
+      "step": 626
+    },
+    {
+      "epoch": 0.909025009061254,
+      "grad_norm": 1.2339680834410676,
+      "learning_rate": 4.2210546968634803e-07,
+      "loss": 1.0058,
+      "step": 627
+    },
+    {
+      "epoch": 0.9104748097136643,
+      "grad_norm": 1.4102170412140371,
+      "learning_rate": 4.086917465555662e-07,
+      "loss": 1.0479,
+      "step": 628
+    },
+    {
+      "epoch": 0.9119246103660746,
+      "grad_norm": 1.3484888198197635,
+      "learning_rate": 3.9549016406193686e-07,
+      "loss": 1.0309,
+      "step": 629
+    },
+    {
+      "epoch": 0.9133744110184849,
+      "grad_norm": 1.3462836308466875,
+      "learning_rate": 3.8250101419816777e-07,
+      "loss": 1.0239,
+      "step": 630
+    },
+    {
+      "epoch": 0.9148242116708952,
+      "grad_norm": 1.3902756002467385,
+      "learning_rate": 3.6972458425838963e-07,
+      "loss": 0.9894,
+      "step": 631
+    },
+    {
+      "epoch": 0.9162740123233055,
+      "grad_norm": 1.3618197476543459,
+      "learning_rate": 3.5716115683178563e-07,
+      "loss": 1.071,
+      "step": 632
+    },
+    {
+      "epoch": 0.9177238129757158,
+      "grad_norm": 1.3494479326684419,
+      "learning_rate": 3.4481100979635304e-07,
+      "loss": 1.0439,
+      "step": 633
+    },
+    {
+      "epoch": 0.9191736136281261,
+      "grad_norm": 1.2315875658107498,
+      "learning_rate": 3.32674416312746e-07,
+      "loss": 1.0079,
+      "step": 634
+    },
+    {
+      "epoch": 0.9206234142805364,
+      "grad_norm": 1.2526446620227205,
+      "learning_rate": 3.207516448182435e-07,
+      "loss": 1.0339,
+      "step": 635
+    },
+    {
+      "epoch": 0.9220732149329467,
+      "grad_norm": 1.2839920262552493,
+      "learning_rate": 3.090429590208033e-07,
+      "loss": 0.9828,
+      "step": 636
+    },
+    {
+      "epoch": 0.923523015585357,
+      "grad_norm": 1.314916482296673,
+      "learning_rate": 2.9754861789324074e-07,
+      "loss": 1.0685,
+      "step": 637
+    },
+    {
+      "epoch": 0.9249728162377673,
+      "grad_norm": 1.3189576875199625,
+      "learning_rate": 2.862688756674881e-07,
+      "loss": 0.9894,
+      "step": 638
+    },
+    {
+      "epoch": 0.9264226168901776,
+      "grad_norm": 1.2581079363465202,
+      "learning_rate": 2.752039818289809e-07,
+      "loss": 1.0005,
+      "step": 639
+    },
+    {
+      "epoch": 0.9278724175425879,
+      "grad_norm": 1.2714615560656808,
+      "learning_rate": 2.6435418111113277e-07,
+      "loss": 1.0642,
+      "step": 640
+    },
+    {
+      "epoch": 0.9293222181949982,
+      "grad_norm": 1.2602010713105205,
+      "learning_rate": 2.537197134899294e-07,
+      "loss": 0.9685,
+      "step": 641
+    },
+    {
+      "epoch": 0.9307720188474085,
+      "grad_norm": 1.2878154181983021,
+      "learning_rate": 2.4330081417861527e-07,
+      "loss": 1.0127,
+      "step": 642
+    },
+    {
+      "epoch": 0.9322218194998187,
+      "grad_norm": 1.3249530773675078,
+      "learning_rate": 2.3309771362249323e-07,
+      "loss": 1.0364,
+      "step": 643
+    },
+    {
+      "epoch": 0.933671620152229,
+      "grad_norm": 1.3258719685665514,
+      "learning_rate": 2.2311063749382743e-07,
+      "loss": 0.9507,
+      "step": 644
+    },
+    {
+      "epoch": 0.9351214208046393,
+      "grad_norm": 1.366993975610368,
+      "learning_rate": 2.1333980668685416e-07,
+      "loss": 1.0465,
+      "step": 645
+    },
+    {
+      "epoch": 0.9365712214570496,
+      "grad_norm": 1.2690733979571618,
+      "learning_rate": 2.037854373128889e-07,
+      "loss": 1.0369,
+      "step": 646
+    },
+    {
+      "epoch": 0.9380210221094599,
+      "grad_norm": 1.321695680088303,
+      "learning_rate": 1.9444774069555694e-07,
+      "loss": 1.0672,
+      "step": 647
+    },
+    {
+      "epoch": 0.9394708227618702,
+      "grad_norm": 1.351855604372839,
+      "learning_rate": 1.8532692336611037e-07,
+      "loss": 1.032,
+      "step": 648
+    },
+    {
+      "epoch": 0.9409206234142805,
+      "grad_norm": 1.2034781570248334,
+      "learning_rate": 1.7642318705886284e-07,
+      "loss": 1.0574,
+      "step": 649
+    },
+    {
+      "epoch": 0.9423704240666908,
+      "grad_norm": 1.3298345225202324,
+      "learning_rate": 1.677367287067322e-07,
+      "loss": 1.0341,
+      "step": 650
+    },
+    {
+      "epoch": 0.9438202247191011,
+      "grad_norm": 1.3202581112735672,
+      "learning_rate": 1.5926774043687366e-07,
+      "loss": 0.9583,
+      "step": 651
+    },
+    {
+      "epoch": 0.9452700253715114,
+      "grad_norm": 1.3278150427538842,
+      "learning_rate": 1.510164095664457e-07,
+      "loss": 1.0301,
+      "step": 652
+    },
+    {
+      "epoch": 0.9467198260239217,
+      "grad_norm": 1.3039826590369827,
+      "learning_rate": 1.4298291859845215e-07,
+      "loss": 1.0576,
+      "step": 653
+    },
+    {
+      "epoch": 0.948169626676332,
+      "grad_norm": 1.3672346616948794,
+      "learning_rate": 1.351674452177143e-07,
+      "loss": 1.0586,
+      "step": 654
+    },
+    {
+      "epoch": 0.9496194273287423,
+      "grad_norm": 1.392892122837727,
+      "learning_rate": 1.2757016228693964e-07,
+      "loss": 1.0308,
+      "step": 655
+    },
+    {
+      "epoch": 0.9510692279811526,
+      "grad_norm": 1.2634391727890335,
+      "learning_rate": 1.2019123784289488e-07,
+      "loss": 1.0131,
+      "step": 656
+    },
+    {
+      "epoch": 0.9525190286335629,
+      "grad_norm": 1.3498344868334864,
+      "learning_rate": 1.1303083509269452e-07,
+      "loss": 0.9741,
+      "step": 657
+    },
+    {
+      "epoch": 0.9539688292859732,
+      "grad_norm": 1.1841792686541104,
+      "learning_rate": 1.0608911241018594e-07,
+      "loss": 0.9441,
+      "step": 658
+    },
+    {
+      "epoch": 0.9554186299383834,
+      "grad_norm": 1.3620448200945166,
+      "learning_rate": 9.936622333245105e-08,
+      "loss": 1.0612,
+      "step": 659
+    },
+    {
+      "epoch": 0.9568684305907937,
+      "grad_norm": 1.2715804394312147,
+      "learning_rate": 9.286231655640799e-08,
+      "loss": 1.0072,
+      "step": 660
+    },
+    {
+      "epoch": 0.958318231243204,
+      "grad_norm": 1.2862912446008476,
+      "learning_rate": 8.657753593552142e-08,
+      "loss": 0.909,
+      "step": 661
+    },
+    {
+      "epoch": 0.9597680318956143,
+      "grad_norm": 1.3383879582112728,
+      "learning_rate": 8.051202047662187e-08,
+      "loss": 1.0296,
+      "step": 662
+    },
+    {
+      "epoch": 0.9612178325480246,
+      "grad_norm": 1.306492588141374,
+      "learning_rate": 7.466590433683252e-08,
+      "loss": 0.9622,
+      "step": 663
+    },
+    {
+      "epoch": 0.9626676332004349,
+      "grad_norm": 1.1852100687872238,
+      "learning_rate": 6.903931682059828e-08,
+      "loss": 0.9414,
+      "step": 664
+    },
+    {
+      "epoch": 0.9641174338528452,
+      "grad_norm": 1.2807170789968838,
+      "learning_rate": 6.363238237683034e-08,
+      "loss": 0.9236,
+      "step": 665
+    },
+    {
+      "epoch": 0.9655672345052555,
+      "grad_norm": 1.3780961822912705,
+      "learning_rate": 5.8445220596149434e-08,
+      "loss": 0.9631,
+      "step": 666
+    },
+    {
+      "epoch": 0.9670170351576658,
+      "grad_norm": 1.2602173856785095,
+      "learning_rate": 5.347794620824576e-08,
+      "loss": 0.8835,
+      "step": 667
+    },
+    {
+      "epoch": 0.9684668358100761,
+      "grad_norm": 1.4125556378040163,
+      "learning_rate": 4.8730669079335434e-08,
+      "loss": 1.0597,
+      "step": 668
+    },
+    {
+      "epoch": 0.9699166364624864,
+      "grad_norm": 1.2271306588718005,
+      "learning_rate": 4.420349420973358e-08,
+      "loss": 0.9816,
+      "step": 669
+    },
+    {
+      "epoch": 0.9713664371148967,
+      "grad_norm": 1.3710121474159438,
+      "learning_rate": 3.98965217315328e-08,
+      "loss": 0.9487,
+      "step": 670
+    },
+    {
+      "epoch": 0.972816237767307,
+      "grad_norm": 1.1912286976372013,
+      "learning_rate": 3.580984690638611e-08,
+      "loss": 0.9481,
+      "step": 671
+    },
+    {
+      "epoch": 0.9742660384197173,
+      "grad_norm": 1.3065455236813233,
+      "learning_rate": 3.194356012340305e-08,
+      "loss": 0.9503,
+      "step": 672
+    },
+    {
+      "epoch": 0.9757158390721276,
+      "grad_norm": 1.4233806827884858,
+      "learning_rate": 2.8297746897146816e-08,
+      "loss": 0.9955,
+      "step": 673
+    },
+    {
+      "epoch": 0.9771656397245378,
+      "grad_norm": 1.3606744616372923,
+      "learning_rate": 2.4872487865744698e-08,
+      "loss": 1.0685,
+      "step": 674
+    },
+    {
+      "epoch": 0.9786154403769481,
+      "grad_norm": 1.2766562729270208,
+      "learning_rate": 2.1667858789105046e-08,
+      "loss": 0.9608,
+      "step": 675
+    },
+    {
+      "epoch": 0.9800652410293584,
+      "grad_norm": 1.3573940191596772,
+      "learning_rate": 1.8683930547243044e-08,
+      "loss": 1.0457,
+      "step": 676
+    },
+    {
+      "epoch": 0.9815150416817687,
+      "grad_norm": 1.3473000817976855,
+      "learning_rate": 1.592076913870644e-08,
+      "loss": 1.0004,
+      "step": 677
+    },
+    {
+      "epoch": 0.982964842334179,
+      "grad_norm": 1.256152796615069,
+      "learning_rate": 1.3378435679122225e-08,
+      "loss": 0.9829,
+      "step": 678
+    },
+    {
+      "epoch": 0.9844146429865893,
+      "grad_norm": 1.2749954489768385,
+      "learning_rate": 1.1056986399845537e-08,
+      "loss": 0.9726,
+      "step": 679
+    },
+    {
+      "epoch": 0.9858644436389996,
+      "grad_norm": 1.3684579466565545,
+      "learning_rate": 8.956472646710624e-09,
+      "loss": 1.0443,
+      "step": 680
+    },
+    {
+      "epoch": 0.9873142442914099,
+      "grad_norm": 1.3570843541382862,
+      "learning_rate": 7.07694087889621e-09,
+      "loss": 1.025,
+      "step": 681
+    },
+    {
+      "epoch": 0.9887640449438202,
+      "grad_norm": 1.348152036520672,
+      "learning_rate": 5.418432667905204e-09,
+      "loss": 0.9878,
+      "step": 682
+    },
+    {
+      "epoch": 0.9902138455962305,
+      "grad_norm": 1.4511295025437614,
+      "learning_rate": 3.9809846966343265e-09,
+      "loss": 1.0356,
+      "step": 683
+    },
+    {
+      "epoch": 0.9916636462486408,
+      "grad_norm": 1.2920520955685484,
+      "learning_rate": 2.764628758570309e-09,
+      "loss": 0.9853,
+      "step": 684
+    },
+    {
+      "epoch": 0.9931134469010511,
+      "grad_norm": 1.2697185635981878,
+      "learning_rate": 1.7693917570837938e-09,
+      "loss": 0.979,
+      "step": 685
+    },
+    {
+      "epoch": 0.9945632475534614,
+      "grad_norm": 1.513387887606993,
+      "learning_rate": 9.95295704835364e-10,
+      "loss": 0.9742,
+      "step": 686
+    },
+    {
+      "epoch": 0.9960130482058717,
+      "grad_norm": 1.1678402119964024,
+      "learning_rate": 4.4235772328815594e-10,
+      "loss": 0.9595,
+      "step": 687
+    },
+    {
+      "epoch": 0.997462848858282,
+      "grad_norm": 1.2979218458306572,
+      "learning_rate": 1.1059004233038296e-10,
+      "loss": 0.9947,
+      "step": 688
+    },
+    {
+      "epoch": 0.9989126495106923,
+      "grad_norm": 1.3440968983834052,
+      "learning_rate": 0.0,
+      "loss": 1.0024,
+      "step": 689
+    },
+    {
+      "epoch": 0.9989126495106923,
+      "step": 689,
+      "total_flos": 6993087226183680.0,
+      "train_loss": 1.0723237899633555,
+      "train_runtime": 12822.114,
+      "train_samples_per_second": 13.769,
+      "train_steps_per_second": 0.054
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 689,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 173,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6993087226183680.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}