diff --git "a/checkpoint-90000/trainer_state.json" "b/checkpoint-90000/trainer_state.json"
--- "a/checkpoint-90000/trainer_state.json"
+++ "b/checkpoint-90000/trainer_state.json"
@@ -1,7 +1,7 @@
 {
-  "best_metric": 3.3056087493896484,
-  "best_model_checkpoint": "/scratch/cl5625/exceptions/models/100M_495/checkpoint-90000",
-  "epoch": 9.6867936712948,
+  "best_metric": 3.301893949508667,
+  "best_model_checkpoint": "/scratch/cl5625/exceptions/models/100M_low_10_495/checkpoint-90000",
+  "epoch": 9.703504043126685,
   "eval_steps": 1000,
   "global_step": 90000,
   "is_hyper_param_search": false,
@@ -9,13418 +9,13418 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.005381552039608223,
-      "grad_norm": 1.0683914422988892,
+      "epoch": 0.005390835579514825,
+      "grad_norm": 2.523895502090454,
       "learning_rate": 0.0003,
-      "loss": 8.6798,
+      "loss": 8.5214,
       "step": 50
     },
     {
-      "epoch": 0.010763104079216447,
-      "grad_norm": 2.859095573425293,
+      "epoch": 0.01078167115902965,
+      "grad_norm": 1.0120031833648682,
       "learning_rate": 0.0006,
-      "loss": 6.9757,
+      "loss": 6.9034,
       "step": 100
     },
     {
-      "epoch": 0.01614465611882467,
-      "grad_norm": 1.489780068397522,
-      "learning_rate": 0.0005996767589699385,
-      "loss": 6.479,
+      "epoch": 0.016172506738544475,
+      "grad_norm": 1.409044623374939,
+      "learning_rate": 0.0005996762007555315,
+      "loss": 6.485,
       "step": 150
     },
     {
-      "epoch": 0.021526208158432893,
-      "grad_norm": 1.3649728298187256,
-      "learning_rate": 0.0005993535179398771,
-      "loss": 6.1969,
+      "epoch": 0.0215633423180593,
+      "grad_norm": 1.1988959312438965,
+      "learning_rate": 0.000599352401511063,
+      "loss": 6.2655,
       "step": 200
     },
     {
-      "epoch": 0.026907760198041114,
-      "grad_norm": 1.613872766494751,
-      "learning_rate": 0.0005990302769098158,
-      "loss": 6.0747,
+      "epoch": 0.026954177897574125,
+      "grad_norm": 0.8952546715736389,
+      "learning_rate": 0.0005990286022665946,
+      "loss": 6.0852,
       "step": 250
     },
     {
-      "epoch": 0.03228931223764934,
-      "grad_norm": 3.6575379371643066,
-      "learning_rate": 0.0005987070358797543,
-      "loss": 5.9726,
+      "epoch": 0.03234501347708895,
+      "grad_norm": 1.8431851863861084,
+      "learning_rate": 0.0005987048030221263,
+      "loss": 5.98,
       "step": 300
     },
     {
-      "epoch": 0.03767086427725756,
-      "grad_norm": 1.7698675394058228,
-      "learning_rate": 0.0005983837948496929,
-      "loss": 5.8947,
+      "epoch": 0.03773584905660377,
+      "grad_norm": 1.0320990085601807,
+      "learning_rate": 0.0005983810037776578,
+      "loss": 5.878,
       "step": 350
     },
     {
-      "epoch": 0.04305241631686579,
-      "grad_norm": 1.8536632061004639,
-      "learning_rate": 0.0005980605538196314,
-      "loss": 5.7945,
+      "epoch": 0.0431266846361186,
+      "grad_norm": 0.8722242712974548,
+      "learning_rate": 0.0005980572045331894,
+      "loss": 5.8147,
       "step": 400
     },
     {
-      "epoch": 0.048433968356474004,
-      "grad_norm": 1.3679734468460083,
-      "learning_rate": 0.0005977373127895701,
-      "loss": 5.721,
+      "epoch": 0.04851752021563342,
+      "grad_norm": 1.0854729413986206,
+      "learning_rate": 0.0005977334052887209,
+      "loss": 5.7348,
       "step": 450
     },
     {
-      "epoch": 0.05381552039608223,
-      "grad_norm": 1.759971022605896,
-      "learning_rate": 0.0005974140717595086,
-      "loss": 5.6392,
+      "epoch": 0.05390835579514825,
+      "grad_norm": 0.9643557071685791,
+      "learning_rate": 0.0005974096060442526,
+      "loss": 5.6805,
       "step": 500
     },
     {
-      "epoch": 0.05919707243569045,
-      "grad_norm": 1.350098729133606,
-      "learning_rate": 0.0005970908307294472,
-      "loss": 5.5834,
+      "epoch": 0.05929919137466307,
+      "grad_norm": 1.469171166419983,
+      "learning_rate": 0.0005970858067997841,
+      "loss": 5.5627,
       "step": 550
     },
     {
-      "epoch": 0.06457862447529868,
-      "grad_norm": 1.1349716186523438,
-      "learning_rate": 0.0005967675896993858,
-      "loss": 5.5375,
+      "epoch": 0.0646900269541779,
+      "grad_norm": 1.1647865772247314,
+      "learning_rate": 0.0005967620075553157,
+      "loss": 5.5272,
       "step": 600
     },
     {
-      "epoch": 0.0699601765149069,
-      "grad_norm": 1.276620626449585,
-      "learning_rate": 0.0005964443486693243,
-      "loss": 5.4386,
+      "epoch": 0.07008086253369272,
+      "grad_norm": 1.41964590549469,
+      "learning_rate": 0.0005964382083108472,
+      "loss": 5.422,
       "step": 650
     },
     {
-      "epoch": 0.07534172855451512,
-      "grad_norm": 0.9349568486213684,
-      "learning_rate": 0.000596121107639263,
-      "loss": 5.389,
+      "epoch": 0.07547169811320754,
+      "grad_norm": 0.929503858089447,
+      "learning_rate": 0.0005961144090663788,
+      "loss": 5.399,
       "step": 700
     },
     {
-      "epoch": 0.08072328059412334,
-      "grad_norm": 1.2208337783813477,
-      "learning_rate": 0.0005957978666092015,
-      "loss": 5.347,
+      "epoch": 0.08086253369272237,
+      "grad_norm": 1.1605098247528076,
+      "learning_rate": 0.0005957906098219104,
+      "loss": 5.3224,
       "step": 750
     },
     {
-      "epoch": 0.08610483263373157,
-      "grad_norm": 0.9988003969192505,
-      "learning_rate": 0.0005954746255791401,
-      "loss": 5.2746,
+      "epoch": 0.0862533692722372,
+      "grad_norm": 1.5475937128067017,
+      "learning_rate": 0.0005954668105774419,
+      "loss": 5.2675,
       "step": 800
     },
     {
-      "epoch": 0.09148638467333979,
-      "grad_norm": 1.0252668857574463,
-      "learning_rate": 0.0005951513845490787,
-      "loss": 5.2421,
+      "epoch": 0.09164420485175202,
+      "grad_norm": 1.2121474742889404,
+      "learning_rate": 0.0005951430113329735,
+      "loss": 5.2399,
       "step": 850
     },
     {
-      "epoch": 0.09686793671294801,
-      "grad_norm": 1.0475060939788818,
-      "learning_rate": 0.0005948281435190174,
-      "loss": 5.1758,
+      "epoch": 0.09703504043126684,
+      "grad_norm": 0.9387132525444031,
+      "learning_rate": 0.0005948192120885051,
+      "loss": 5.1528,
       "step": 900
     },
     {
-      "epoch": 0.10224948875255624,
-      "grad_norm": 1.1841458082199097,
-      "learning_rate": 0.0005945049024889559,
-      "loss": 5.1527,
+      "epoch": 0.10242587601078167,
+      "grad_norm": 1.3440146446228027,
+      "learning_rate": 0.0005944954128440366,
+      "loss": 5.1512,
       "step": 950
     },
     {
-      "epoch": 0.10763104079216446,
-      "grad_norm": 1.3012635707855225,
-      "learning_rate": 0.0005941816614588944,
-      "loss": 5.1178,
+      "epoch": 0.1078167115902965,
+      "grad_norm": 1.029028058052063,
+      "learning_rate": 0.0005941716135995682,
+      "loss": 5.0895,
       "step": 1000
     },
     {
-      "epoch": 0.10763104079216446,
-      "eval_accuracy": 0.2259316197396655,
-      "eval_loss": 5.038772106170654,
-      "eval_runtime": 185.0,
-      "eval_samples_per_second": 97.357,
-      "eval_steps_per_second": 6.086,
+      "epoch": 0.1078167115902965,
+      "eval_accuracy": 0.2268177926926804,
+      "eval_loss": 5.029711723327637,
+      "eval_runtime": 184.0753,
+      "eval_samples_per_second": 97.846,
+      "eval_steps_per_second": 6.117,
       "step": 1000
     },
     {
-      "epoch": 0.11301259283177269,
-      "grad_norm": 1.3463271856307983,
-      "learning_rate": 0.000593858420428833,
-      "loss": 5.0694,
+      "epoch": 0.11320754716981132,
+      "grad_norm": 0.9686921238899231,
+      "learning_rate": 0.0005938478143550997,
+      "loss": 5.0514,
       "step": 1050
     },
     {
-      "epoch": 0.1183941448713809,
-      "grad_norm": 1.3660123348236084,
-      "learning_rate": 0.0005935351793987716,
-      "loss": 5.0413,
+      "epoch": 0.11859838274932614,
+      "grad_norm": 1.5297749042510986,
+      "learning_rate": 0.0005935240151106314,
+      "loss": 5.0125,
       "step": 1100
     },
     {
-      "epoch": 0.12377569691098914,
-      "grad_norm": 1.0985352993011475,
-      "learning_rate": 0.0005932119383687103,
-      "loss": 5.0437,
+      "epoch": 0.12398921832884097,
+      "grad_norm": 1.1212499141693115,
+      "learning_rate": 0.0005932002158661629,
+      "loss": 4.9901,
       "step": 1150
     },
     {
-      "epoch": 0.12915724895059735,
-      "grad_norm": 1.175125241279602,
-      "learning_rate": 0.0005928886973386488,
-      "loss": 4.9567,
+      "epoch": 0.1293800539083558,
+      "grad_norm": 1.0991240739822388,
+      "learning_rate": 0.0005928764166216945,
+      "loss": 4.9656,
       "step": 1200
     },
     {
-      "epoch": 0.13453880099020557,
-      "grad_norm": 1.1065387725830078,
-      "learning_rate": 0.0005925654563085874,
-      "loss": 4.9404,
+      "epoch": 0.1347708894878706,
+      "grad_norm": 0.9258497953414917,
+      "learning_rate": 0.000592552617377226,
+      "loss": 4.9402,
       "step": 1250
     },
     {
-      "epoch": 0.1399203530298138,
-      "grad_norm": 2.134150505065918,
-      "learning_rate": 0.000592242215278526,
-      "loss": 4.8978,
+      "epoch": 0.14016172506738545,
+      "grad_norm": 0.8877195119857788,
+      "learning_rate": 0.0005922288181327577,
+      "loss": 4.8898,
       "step": 1300
     },
     {
-      "epoch": 0.14530190506942203,
-      "grad_norm": 0.826282262802124,
-      "learning_rate": 0.0005919189742484645,
-      "loss": 4.8955,
+      "epoch": 0.14555256064690028,
+      "grad_norm": 1.3482192754745483,
+      "learning_rate": 0.0005919050188882893,
+      "loss": 4.874,
       "step": 1350
     },
     {
-      "epoch": 0.15068345710903025,
-      "grad_norm": 1.1777522563934326,
-      "learning_rate": 0.0005915957332184032,
-      "loss": 4.8565,
+      "epoch": 0.1509433962264151,
+      "grad_norm": 0.9256923198699951,
+      "learning_rate": 0.0005915812196438207,
+      "loss": 4.8448,
       "step": 1400
     },
     {
-      "epoch": 0.15606500914863847,
-      "grad_norm": 1.0407218933105469,
-      "learning_rate": 0.0005912724921883417,
-      "loss": 4.8394,
+      "epoch": 0.15633423180592992,
+      "grad_norm": 1.2883760929107666,
+      "learning_rate": 0.0005912574203993524,
+      "loss": 4.8259,
       "step": 1450
     },
     {
-      "epoch": 0.16144656118824668,
-      "grad_norm": 1.1135185956954956,
-      "learning_rate": 0.0005909492511582803,
-      "loss": 4.8083,
+      "epoch": 0.16172506738544473,
+      "grad_norm": 0.9150322079658508,
+      "learning_rate": 0.0005909336211548839,
+      "loss": 4.8123,
       "step": 1500
     },
     {
-      "epoch": 0.1668281132278549,
-      "grad_norm": 0.9728663563728333,
-      "learning_rate": 0.0005906260101282189,
-      "loss": 4.7959,
+      "epoch": 0.16711590296495957,
+      "grad_norm": 0.8724535703659058,
+      "learning_rate": 0.0005906098219104155,
+      "loss": 4.773,
       "step": 1550
     },
     {
-      "epoch": 0.17220966526746315,
-      "grad_norm": 0.8725541234016418,
-      "learning_rate": 0.0005903027690981575,
-      "loss": 4.7704,
+      "epoch": 0.1725067385444744,
+      "grad_norm": 0.8244979381561279,
+      "learning_rate": 0.000590286022665947,
+      "loss": 4.7843,
       "step": 1600
     },
     {
-      "epoch": 0.17759121730707136,
-      "grad_norm": 0.8491561412811279,
-      "learning_rate": 0.000589979528068096,
-      "loss": 4.7693,
+      "epoch": 0.1778975741239892,
+      "grad_norm": 0.898222804069519,
+      "learning_rate": 0.0005899622234214787,
+      "loss": 4.7642,
       "step": 1650
     },
     {
-      "epoch": 0.18297276934667958,
-      "grad_norm": 1.0023645162582397,
-      "learning_rate": 0.0005896562870380347,
-      "loss": 4.7243,
+      "epoch": 0.18328840970350405,
+      "grad_norm": 1.0301188230514526,
+      "learning_rate": 0.0005896384241770102,
+      "loss": 4.6851,
       "step": 1700
     },
     {
-      "epoch": 0.1883543213862878,
-      "grad_norm": 1.2298816442489624,
-      "learning_rate": 0.0005893330460079732,
-      "loss": 4.7076,
+      "epoch": 0.18867924528301888,
+      "grad_norm": 0.8868728876113892,
+      "learning_rate": 0.0005893146249325418,
+      "loss": 4.6993,
       "step": 1750
     },
     {
-      "epoch": 0.19373587342589602,
-      "grad_norm": 1.063509225845337,
-      "learning_rate": 0.0005890098049779118,
-      "loss": 4.6721,
+      "epoch": 0.1940700808625337,
+      "grad_norm": 1.0902804136276245,
+      "learning_rate": 0.0005889908256880733,
+      "loss": 4.6853,
       "step": 1800
     },
     {
-      "epoch": 0.19911742546550426,
-      "grad_norm": 1.0815796852111816,
-      "learning_rate": 0.0005886865639478504,
-      "loss": 4.6689,
+      "epoch": 0.19946091644204852,
+      "grad_norm": 0.9205830097198486,
+      "learning_rate": 0.0005886670264436049,
+      "loss": 4.6407,
       "step": 1850
     },
     {
-      "epoch": 0.20449897750511248,
-      "grad_norm": 1.0041590929031372,
-      "learning_rate": 0.0005883633229177889,
-      "loss": 4.664,
+      "epoch": 0.20485175202156333,
+      "grad_norm": 0.9026359915733337,
+      "learning_rate": 0.0005883432271991365,
+      "loss": 4.6342,
       "step": 1900
     },
     {
-      "epoch": 0.2098805295447207,
-      "grad_norm": 0.741370439529419,
-      "learning_rate": 0.0005880400818877276,
-      "loss": 4.647,
+      "epoch": 0.21024258760107817,
+      "grad_norm": 0.8825887441635132,
+      "learning_rate": 0.0005880194279546681,
+      "loss": 4.5991,
       "step": 1950
     },
     {
-      "epoch": 0.2152620815843289,
-      "grad_norm": 0.9399243593215942,
-      "learning_rate": 0.0005877168408576662,
-      "loss": 4.5945,
+      "epoch": 0.215633423180593,
+      "grad_norm": 0.9272496104240417,
+      "learning_rate": 0.0005876956287101996,
+      "loss": 4.5777,
       "step": 2000
     },
     {
-      "epoch": 0.2152620815843289,
-      "eval_accuracy": 0.2670903115197812,
-      "eval_loss": 4.538747310638428,
-      "eval_runtime": 182.7156,
-      "eval_samples_per_second": 98.574,
-      "eval_steps_per_second": 6.163,
+      "epoch": 0.215633423180593,
+      "eval_accuracy": 0.2698170643923734,
+      "eval_loss": 4.521856784820557,
+      "eval_runtime": 183.9786,
+      "eval_samples_per_second": 97.897,
+      "eval_steps_per_second": 6.12,
       "step": 2000
     },
     {
-      "epoch": 0.22064363362393713,
-      "grad_norm": 0.8416916131973267,
-      "learning_rate": 0.0005873935998276048,
-      "loss": 4.5951,
+      "epoch": 0.2210242587601078,
+      "grad_norm": 1.0929840803146362,
+      "learning_rate": 0.0005873718294657312,
+      "loss": 4.5928,
       "step": 2050
     },
     {
-      "epoch": 0.22602518566354537,
-      "grad_norm": 1.0621873140335083,
-      "learning_rate": 0.0005870703587975433,
-      "loss": 4.5883,
+      "epoch": 0.22641509433962265,
+      "grad_norm": 1.117784857749939,
+      "learning_rate": 0.0005870480302212628,
+      "loss": 4.5567,
       "step": 2100
     },
     {
-      "epoch": 0.2314067377031536,
-      "grad_norm": 0.958995521068573,
-      "learning_rate": 0.0005867471177674818,
-      "loss": 4.564,
+      "epoch": 0.23180592991913745,
+      "grad_norm": 1.0417054891586304,
+      "learning_rate": 0.0005867242309767943,
+      "loss": 4.5057,
       "step": 2150
     },
     {
-      "epoch": 0.2367882897427618,
-      "grad_norm": 1.0263909101486206,
-      "learning_rate": 0.0005864238767374205,
-      "loss": 4.5361,
+      "epoch": 0.2371967654986523,
+      "grad_norm": 0.8676998615264893,
+      "learning_rate": 0.0005864004317323259,
+      "loss": 4.5406,
       "step": 2200
     },
     {
-      "epoch": 0.24216984178237003,
-      "grad_norm": 0.865015983581543,
-      "learning_rate": 0.0005861006357073591,
-      "loss": 4.5177,
+      "epoch": 0.24258760107816713,
+      "grad_norm": 0.7627687454223633,
+      "learning_rate": 0.0005860766324878575,
+      "loss": 4.475,
       "step": 2250
     },
     {
-      "epoch": 0.24755139382197827,
-      "grad_norm": 0.8319127559661865,
-      "learning_rate": 0.0005857773946772977,
-      "loss": 4.5171,
+      "epoch": 0.24797843665768193,
+      "grad_norm": 0.9335618615150452,
+      "learning_rate": 0.000585752833243389,
+      "loss": 4.4574,
       "step": 2300
     },
     {
-      "epoch": 0.2529329458615865,
-      "grad_norm": 0.8533033132553101,
-      "learning_rate": 0.0005854541536472362,
-      "loss": 4.4716,
+      "epoch": 0.25336927223719674,
+      "grad_norm": 0.922460675239563,
+      "learning_rate": 0.0005854290339989206,
+      "loss": 4.4542,
       "step": 2350
     },
     {
-      "epoch": 0.2583144979011947,
-      "grad_norm": 0.9105812907218933,
-      "learning_rate": 0.0005851309126171749,
-      "loss": 4.4832,
+      "epoch": 0.2587601078167116,
+      "grad_norm": 1.1829935312271118,
+      "learning_rate": 0.0005851052347544521,
+      "loss": 4.4441,
       "step": 2400
     },
     {
-      "epoch": 0.2636960499408029,
-      "grad_norm": 0.9672813415527344,
-      "learning_rate": 0.0005848076715871134,
-      "loss": 4.4371,
+      "epoch": 0.2641509433962264,
+      "grad_norm": 0.8065271377563477,
+      "learning_rate": 0.0005847814355099838,
+      "loss": 4.4474,
       "step": 2450
     },
     {
-      "epoch": 0.26907760198041114,
-      "grad_norm": 0.9679467082023621,
-      "learning_rate": 0.000584484430557052,
-      "loss": 4.4541,
+      "epoch": 0.2695417789757412,
+      "grad_norm": 0.8572820425033569,
+      "learning_rate": 0.0005844576362655154,
+      "loss": 4.4392,
       "step": 2500
     },
     {
-      "epoch": 0.27445915402001936,
-      "grad_norm": 0.8314814567565918,
-      "learning_rate": 0.0005841611895269906,
-      "loss": 4.4395,
+      "epoch": 0.2749326145552561,
+      "grad_norm": 0.8170642256736755,
+      "learning_rate": 0.0005841338370210469,
+      "loss": 4.4017,
       "step": 2550
     },
     {
-      "epoch": 0.2798407060596276,
-      "grad_norm": 0.7759853601455688,
-      "learning_rate": 0.0005838379484969291,
-      "loss": 4.4153,
+      "epoch": 0.2803234501347709,
+      "grad_norm": 0.8466624021530151,
+      "learning_rate": 0.0005838100377765785,
+      "loss": 4.3926,
       "step": 2600
     },
     {
-      "epoch": 0.2852222580992358,
-      "grad_norm": 0.7185021638870239,
-      "learning_rate": 0.0005835147074668678,
-      "loss": 4.3998,
+      "epoch": 0.2857142857142857,
+      "grad_norm": 1.474891185760498,
+      "learning_rate": 0.0005834862385321101,
+      "loss": 4.3916,
       "step": 2650
     },
     {
-      "epoch": 0.29060381013884407,
-      "grad_norm": 0.7233301997184753,
-      "learning_rate": 0.0005831914664368063,
-      "loss": 4.3864,
+      "epoch": 0.29110512129380056,
+      "grad_norm": 1.0594297647476196,
+      "learning_rate": 0.0005831624392876417,
+      "loss": 4.3824,
       "step": 2700
     },
     {
-      "epoch": 0.2959853621784523,
-      "grad_norm": 0.768659234046936,
-      "learning_rate": 0.0005828682254067449,
-      "loss": 4.3805,
+      "epoch": 0.29649595687331537,
+      "grad_norm": 0.8971001505851746,
+      "learning_rate": 0.0005828386400431731,
+      "loss": 4.3443,
       "step": 2750
     },
     {
-      "epoch": 0.3013669142180605,
-      "grad_norm": 0.7125174403190613,
-      "learning_rate": 0.0005825449843766835,
-      "loss": 4.3937,
+      "epoch": 0.3018867924528302,
+      "grad_norm": 0.7338554859161377,
+      "learning_rate": 0.0005825148407987048,
+      "loss": 4.3845,
       "step": 2800
     },
     {
-      "epoch": 0.3067484662576687,
-      "grad_norm": 0.8199774026870728,
-      "learning_rate": 0.0005822217433466221,
-      "loss": 4.366,
+      "epoch": 0.30727762803234504,
+      "grad_norm": 0.9269198775291443,
+      "learning_rate": 0.0005821910415542363,
+      "loss": 4.3515,
       "step": 2850
     },
     {
-      "epoch": 0.31213001829727693,
-      "grad_norm": 0.7131398916244507,
-      "learning_rate": 0.0005818985023165607,
-      "loss": 4.3393,
+      "epoch": 0.31266846361185985,
+      "grad_norm": 0.778174102306366,
+      "learning_rate": 0.0005818672423097679,
+      "loss": 4.3305,
       "step": 2900
     },
     {
-      "epoch": 0.31751157033688515,
-      "grad_norm": 0.8093417286872864,
-      "learning_rate": 0.0005815752612864992,
-      "loss": 4.3397,
+      "epoch": 0.31805929919137466,
+      "grad_norm": 0.8001732230186462,
+      "learning_rate": 0.0005815434430652994,
+      "loss": 4.3308,
       "step": 2950
     },
     {
-      "epoch": 0.32289312237649337,
-      "grad_norm": 0.9575026631355286,
-      "learning_rate": 0.0005812520202564378,
-      "loss": 4.3222,
+      "epoch": 0.32345013477088946,
+      "grad_norm": 0.7768656015396118,
+      "learning_rate": 0.0005812196438208311,
+      "loss": 4.3302,
       "step": 3000
     },
     {
-      "epoch": 0.32289312237649337,
-      "eval_accuracy": 0.297169233717903,
-      "eval_loss": 4.248993873596191,
-      "eval_runtime": 182.6748,
-      "eval_samples_per_second": 98.596,
-      "eval_steps_per_second": 6.164,
+      "epoch": 0.32345013477088946,
+      "eval_accuracy": 0.29869797984945273,
+      "eval_loss": 4.233901500701904,
+      "eval_runtime": 183.8561,
+      "eval_samples_per_second": 97.962,
+      "eval_steps_per_second": 6.124,
       "step": 3000
     },
     {
-      "epoch": 0.3282746744161016,
-      "grad_norm": 0.8685798645019531,
-      "learning_rate": 0.0005809287792263764,
-      "loss": 4.3126,
+      "epoch": 0.3288409703504043,
+      "grad_norm": 0.9780474305152893,
+      "learning_rate": 0.0005808958445763626,
+      "loss": 4.3051,
       "step": 3050
     },
     {
-      "epoch": 0.3336562264557098,
-      "grad_norm": 0.8055946826934814,
-      "learning_rate": 0.0005806055381963151,
-      "loss": 4.303,
+      "epoch": 0.33423180592991913,
+      "grad_norm": 0.7568812966346741,
+      "learning_rate": 0.0005805720453318942,
+      "loss": 4.3013,
       "step": 3100
     },
     {
-      "epoch": 0.3390377784953181,
-      "grad_norm": 0.8136595487594604,
-      "learning_rate": 0.0005802822971662536,
-      "loss": 4.2861,
+      "epoch": 0.33962264150943394,
+      "grad_norm": 0.7448121309280396,
+      "learning_rate": 0.0005802482460874257,
+      "loss": 4.2663,
       "step": 3150
     },
     {
-      "epoch": 0.3444193305349263,
-      "grad_norm": 0.6919870972633362,
-      "learning_rate": 0.0005799590561361922,
-      "loss": 4.3128,
+      "epoch": 0.3450134770889488,
+      "grad_norm": 0.8732165694236755,
+      "learning_rate": 0.0005799244468429573,
+      "loss": 4.2726,
       "step": 3200
     },
     {
-      "epoch": 0.3498008825745345,
-      "grad_norm": 1.4158786535263062,
-      "learning_rate": 0.0005796358151061307,
-      "loss": 4.3008,
+      "epoch": 0.3504043126684636,
+      "grad_norm": 0.6735326647758484,
+      "learning_rate": 0.0005796006475984889,
+      "loss": 4.2478,
       "step": 3250
     },
     {
-      "epoch": 0.35518243461414273,
-      "grad_norm": 0.8841201663017273,
-      "learning_rate": 0.0005793125740760694,
-      "loss": 4.2924,
+      "epoch": 0.3557951482479784,
+      "grad_norm": 0.8929001092910767,
+      "learning_rate": 0.0005792768483540205,
+      "loss": 4.2705,
       "step": 3300
     },
     {
-      "epoch": 0.36056398665375095,
-      "grad_norm": 0.6999953985214233,
-      "learning_rate": 0.0005789893330460079,
-      "loss": 4.2491,
+      "epoch": 0.3611859838274933,
+      "grad_norm": 0.644730806350708,
+      "learning_rate": 0.000578953049109552,
+      "loss": 4.2512,
       "step": 3350
     },
     {
-      "epoch": 0.36594553869335916,
-      "grad_norm": 0.8567237257957458,
-      "learning_rate": 0.0005786660920159465,
-      "loss": 4.2506,
+      "epoch": 0.3665768194070081,
+      "grad_norm": 0.823698103427887,
+      "learning_rate": 0.0005786292498650836,
+      "loss": 4.2416,
       "step": 3400
     },
     {
-      "epoch": 0.3713270907329674,
-      "grad_norm": 0.779323160648346,
-      "learning_rate": 0.0005783428509858851,
-      "loss": 4.2543,
+      "epoch": 0.3719676549865229,
+      "grad_norm": 0.7468538880348206,
+      "learning_rate": 0.0005783054506206152,
+      "loss": 4.2319,
       "step": 3450
     },
     {
-      "epoch": 0.3767086427725756,
-      "grad_norm": 0.7086902260780334,
-      "learning_rate": 0.0005780196099558237,
-      "loss": 4.235,
+      "epoch": 0.37735849056603776,
+      "grad_norm": 0.7215543389320374,
+      "learning_rate": 0.0005779816513761467,
+      "loss": 4.2087,
       "step": 3500
     },
     {
-      "epoch": 0.3820901948121838,
-      "grad_norm": 0.5936145186424255,
-      "learning_rate": 0.0005776963689257623,
-      "loss": 4.2331,
+      "epoch": 0.38274932614555257,
+      "grad_norm": 0.7402747869491577,
+      "learning_rate": 0.0005776578521316782,
+      "loss": 4.2144,
       "step": 3550
     },
     {
-      "epoch": 0.38747174685179203,
-      "grad_norm": 0.632709264755249,
-      "learning_rate": 0.0005773731278957008,
-      "loss": 4.2431,
+      "epoch": 0.3881401617250674,
+      "grad_norm": 0.8125879764556885,
+      "learning_rate": 0.0005773340528872099,
+      "loss": 4.2154,
       "step": 3600
     },
     {
-      "epoch": 0.3928532988914003,
-      "grad_norm": 0.9441436529159546,
-      "learning_rate": 0.0005770498868656394,
-      "loss": 4.2204,
+      "epoch": 0.3935309973045822,
+      "grad_norm": 0.7119957208633423,
+      "learning_rate": 0.0005770102536427414,
+      "loss": 4.2213,
       "step": 3650
     },
     {
-      "epoch": 0.3982348509310085,
-      "grad_norm": 0.8455289602279663,
-      "learning_rate": 0.000576726645835578,
-      "loss": 4.2151,
+      "epoch": 0.39892183288409705,
+      "grad_norm": 0.7396487593650818,
+      "learning_rate": 0.000576686454398273,
+      "loss": 4.1904,
       "step": 3700
     },
     {
-      "epoch": 0.40361640297061674,
-      "grad_norm": 0.6691498756408691,
-      "learning_rate": 0.0005764034048055167,
-      "loss": 4.1907,
+      "epoch": 0.40431266846361186,
+      "grad_norm": 0.8688079714775085,
+      "learning_rate": 0.0005763626551538045,
+      "loss": 4.1874,
       "step": 3750
     },
     {
-      "epoch": 0.40899795501022496,
-      "grad_norm": 0.764274001121521,
-      "learning_rate": 0.0005760801637754552,
-      "loss": 4.1993,
+      "epoch": 0.40970350404312667,
+      "grad_norm": 0.668553352355957,
+      "learning_rate": 0.0005760388559093362,
+      "loss": 4.2061,
       "step": 3800
     },
     {
-      "epoch": 0.4143795070498332,
-      "grad_norm": 0.8627861142158508,
-      "learning_rate": 0.0005757569227453937,
-      "loss": 4.2007,
+      "epoch": 0.41509433962264153,
+      "grad_norm": 0.7356492877006531,
+      "learning_rate": 0.0005757150566648678,
+      "loss": 4.1877,
       "step": 3850
     },
     {
-      "epoch": 0.4197610590894414,
-      "grad_norm": 0.7309780120849609,
-      "learning_rate": 0.0005754336817153324,
-      "loss": 4.1748,
+      "epoch": 0.42048517520215634,
+      "grad_norm": 0.8281533718109131,
+      "learning_rate": 0.0005753912574203993,
+      "loss": 4.1567,
       "step": 3900
     },
     {
-      "epoch": 0.4251426111290496,
-      "grad_norm": 1.0626062154769897,
-      "learning_rate": 0.0005751104406852709,
-      "loss": 4.1841,
+      "epoch": 0.42587601078167114,
+      "grad_norm": 0.6755908131599426,
+      "learning_rate": 0.0005750674581759309,
+      "loss": 4.1564,
       "step": 3950
     },
     {
-      "epoch": 0.4305241631686578,
-      "grad_norm": 0.7397871017456055,
-      "learning_rate": 0.0005747871996552096,
-      "loss": 4.1793,
+      "epoch": 0.431266846361186,
+      "grad_norm": 0.7325392365455627,
+      "learning_rate": 0.0005747436589314624,
+      "loss": 4.1588,
       "step": 4000
     },
     {
-      "epoch": 0.4305241631686578,
-      "eval_accuracy": 0.31092153838146963,
-      "eval_loss": 4.1014580726623535,
-      "eval_runtime": 183.3521,
-      "eval_samples_per_second": 98.232,
-      "eval_steps_per_second": 6.141,
+      "epoch": 0.431266846361186,
+      "eval_accuracy": 0.312624455092186,
+      "eval_loss": 4.089514255523682,
+      "eval_runtime": 183.9561,
+      "eval_samples_per_second": 97.909,
+      "eval_steps_per_second": 6.121,
       "step": 4000
     },
     {
-      "epoch": 0.43590571520826604,
-      "grad_norm": 0.7458042502403259,
-      "learning_rate": 0.0005744639586251481,
-      "loss": 4.1751,
+      "epoch": 0.4366576819407008,
+      "grad_norm": 0.670657753944397,
+      "learning_rate": 0.0005744198596869941,
+      "loss": 4.1574,
       "step": 4050
     },
     {
-      "epoch": 0.44128726724787426,
-      "grad_norm": 0.6242780685424805,
-      "learning_rate": 0.0005741407175950867,
-      "loss": 4.1522,
+      "epoch": 0.4420485175202156,
+      "grad_norm": 0.6704895496368408,
+      "learning_rate": 0.0005740960604425255,
+      "loss": 4.1628,
       "step": 4100
     },
     {
-      "epoch": 0.44666881928748253,
-      "grad_norm": 0.8113278746604919,
-      "learning_rate": 0.0005738174765650253,
-      "loss": 4.1662,
+      "epoch": 0.4474393530997305,
+      "grad_norm": 0.6966278553009033,
+      "learning_rate": 0.0005737722611980572,
+      "loss": 4.1522,
       "step": 4150
     },
     {
-      "epoch": 0.45205037132709075,
-      "grad_norm": 0.7726860046386719,
-      "learning_rate": 0.0005734942355349638,
-      "loss": 4.1576,
+      "epoch": 0.4528301886792453,
+      "grad_norm": 0.647747278213501,
+      "learning_rate": 0.0005734484619535887,
+      "loss": 4.1378,
       "step": 4200
     },
     {
-      "epoch": 0.45743192336669897,
-      "grad_norm": 0.5663923621177673,
-      "learning_rate": 0.0005731709945049025,
-      "loss": 4.1365,
+      "epoch": 0.4582210242587601,
+      "grad_norm": 0.6316284537315369,
+      "learning_rate": 0.0005731246627091203,
+      "loss": 4.1295,
       "step": 4250
     },
     {
-      "epoch": 0.4628134754063072,
-      "grad_norm": 0.9014385938644409,
-      "learning_rate": 0.000572847753474841,
-      "loss": 4.1356,
+      "epoch": 0.4636118598382749,
+      "grad_norm": 0.7107946872711182,
+      "learning_rate": 0.0005728008634646518,
+      "loss": 4.1237,
       "step": 4300
     },
     {
-      "epoch": 0.4681950274459154,
-      "grad_norm": 0.7335566282272339,
-      "learning_rate": 0.0005725245124447796,
-      "loss": 4.1453,
+      "epoch": 0.46900269541778977,
+      "grad_norm": 0.6699416637420654,
+      "learning_rate": 0.0005724770642201835,
+      "loss": 4.1312,
       "step": 4350
     },
     {
-      "epoch": 0.4735765794855236,
-      "grad_norm": 0.779979944229126,
-      "learning_rate": 0.0005722012714147182,
-      "loss": 4.1228,
+      "epoch": 0.4743935309973046,
+      "grad_norm": 0.745444118976593,
+      "learning_rate": 0.000572153264975715,
+      "loss": 4.1147,
       "step": 4400
     },
     {
-      "epoch": 0.47895813152513184,
-      "grad_norm": 0.8542835116386414,
-      "learning_rate": 0.0005718780303846568,
-      "loss": 4.1002,
+      "epoch": 0.4797843665768194,
+      "grad_norm": 0.883878767490387,
+      "learning_rate": 0.0005718294657312466,
+      "loss": 4.1233,
       "step": 4450
     },
     {
-      "epoch": 0.48433968356474005,
-      "grad_norm": 0.6748495697975159,
-      "learning_rate": 0.0005715547893545953,
-      "loss": 4.1213,
+      "epoch": 0.48517520215633425,
+      "grad_norm": 0.6188403367996216,
+      "learning_rate": 0.0005715056664867781,
+      "loss": 4.1024,
       "step": 4500
     },
     {
-      "epoch": 0.48972123560434827,
-      "grad_norm": 0.7007385492324829,
-      "learning_rate": 0.000571231548324534,
-      "loss": 4.1179,
+      "epoch": 0.49056603773584906,
+      "grad_norm": 0.6261562705039978,
+      "learning_rate": 0.0005711818672423097,
+      "loss": 4.0913,
       "step": 4550
     },
     {
-      "epoch": 0.49510278764395654,
-      "grad_norm": 0.6953123807907104,
-      "learning_rate": 0.0005709083072944725,
-      "loss": 4.0866,
+      "epoch": 0.49595687331536387,
+      "grad_norm": 0.6521971225738525,
+      "learning_rate": 0.0005708580679978413,
+      "loss": 4.1008,
       "step": 4600
     },
     {
-      "epoch": 0.5004843396835648,
-      "grad_norm": 0.6920002102851868,
-      "learning_rate": 0.0005705850662644111,
-      "loss": 4.1128,
+      "epoch": 0.5013477088948787,
+      "grad_norm": 0.6763367652893066,
+      "learning_rate": 0.0005705342687533729,
+      "loss": 4.0797,
       "step": 4650
     },
     {
-      "epoch": 0.505865891723173,
-      "grad_norm": 0.6578840613365173,
-      "learning_rate": 0.0005702618252343497,
-      "loss": 4.1059,
+      "epoch": 0.5067385444743935,
+      "grad_norm": 0.7222947478294373,
+      "learning_rate": 0.0005702104695089044,
+      "loss": 4.0854,
       "step": 4700
     },
     {
-      "epoch": 0.5112474437627812,
-      "grad_norm": 0.7448212504386902,
-      "learning_rate": 0.0005699385842042882,
-      "loss": 4.1112,
+      "epoch": 0.5121293800539084,
+      "grad_norm": 0.5783945322036743,
+      "learning_rate": 0.000569886670264436,
+      "loss": 4.0676,
       "step": 4750
     },
     {
-      "epoch": 0.5166289958023894,
-      "grad_norm": 0.6418218612670898,
-      "learning_rate": 0.0005696153431742269,
-      "loss": 4.083,
+      "epoch": 0.5175202156334232,
+      "grad_norm": 0.5462530851364136,
+      "learning_rate": 0.0005695628710199675,
+      "loss": 4.0824,
       "step": 4800
     },
     {
-      "epoch": 0.5220105478419976,
-      "grad_norm": 0.5668521523475647,
-      "learning_rate": 0.0005692921021441655,
-      "loss": 4.092,
+      "epoch": 0.522911051212938,
+      "grad_norm": 0.6322537064552307,
+      "learning_rate": 0.0005692390717754991,
+      "loss": 4.0771,
       "step": 4850
     },
     {
-      "epoch": 0.5273920998816058,
-      "grad_norm": 0.6103580594062805,
-      "learning_rate": 0.0005689688611141041,
-      "loss": 4.0851,
+      "epoch": 0.5283018867924528,
+      "grad_norm": 0.6346331238746643,
+      "learning_rate": 0.0005689152725310306,
+      "loss": 4.0847,
       "step": 4900
     },
     {
-      "epoch": 0.5327736519212141,
-      "grad_norm": 0.691785454750061,
-      "learning_rate": 0.0005686456200840426,
-      "loss": 4.0865,
+      "epoch": 0.5336927223719676,
+      "grad_norm": 0.6653397083282471,
+      "learning_rate": 0.0005685914732865623,
+      "loss": 4.0588,
       "step": 4950
     },
     {
-      "epoch": 0.5381552039608223,
-      "grad_norm": 0.725322425365448,
-      "learning_rate": 0.0005683223790539811,
-      "loss": 4.0646,
+      "epoch": 0.5390835579514824,
+      "grad_norm": 0.594985842704773,
+      "learning_rate": 0.0005682676740420939,
+      "loss": 4.0457,
       "step": 5000
     },
     {
-      "epoch": 0.5381552039608223,
-      "eval_accuracy": 0.3200998824267101,
-      "eval_loss": 4.004289627075195,
-      "eval_runtime": 183.8369,
-      "eval_samples_per_second": 97.973,
-      "eval_steps_per_second": 6.125,
+      "epoch": 0.5390835579514824,
+      "eval_accuracy": 0.32164601302031015,
+      "eval_loss": 3.9885013103485107,
+      "eval_runtime": 183.9828,
+      "eval_samples_per_second": 97.895,
+      "eval_steps_per_second": 6.12,
       "step": 5000
     },
     {
-      "epoch": 0.5435367560004305,
-      "grad_norm": 0.7255427837371826,
-      "learning_rate": 0.0005679991380239198,
-      "loss": 4.0629,
+      "epoch": 0.5444743935309974,
+      "grad_norm": 0.5795463919639587,
+      "learning_rate": 0.0005679438747976254,
+      "loss": 4.0402,
       "step": 5050
     },
     {
-      "epoch": 0.5489183080400387,
-      "grad_norm": 0.8185265064239502,
-      "learning_rate": 0.0005676758969938584,
-      "loss": 4.0571,
+      "epoch": 0.5498652291105122,
+      "grad_norm": 0.8126819133758545,
+      "learning_rate": 0.000567620075553157,
+      "loss": 4.0462,
       "step": 5100
     },
     {
-      "epoch": 0.5542998600796469,
-      "grad_norm": 0.6164122223854065,
-      "learning_rate": 0.000567352655963797,
-      "loss": 4.0833,
+      "epoch": 0.555256064690027,
+      "grad_norm": 0.6497731804847717,
+      "learning_rate": 0.0005672962763086886,
+      "loss": 4.0521,
       "step": 5150
     },
     {
-      "epoch": 0.5596814121192552,
-      "grad_norm": 0.7035121917724609,
-      "learning_rate": 0.0005670294149337355,
-      "loss": 4.0478,
+      "epoch": 0.5606469002695418,
+      "grad_norm": 0.6292665004730225,
+      "learning_rate": 0.0005669724770642202,
+      "loss": 4.0275,
       "step": 5200
     },
     {
-      "epoch": 0.5650629641588634,
-      "grad_norm": 0.6536286473274231,
-      "learning_rate": 0.0005667061739036742,
-      "loss": 4.0478,
+      "epoch": 0.5660377358490566,
+      "grad_norm": 0.6472154259681702,
+      "learning_rate": 0.0005666486778197517,
+      "loss": 4.0289,
       "step": 5250
     },
     {
-      "epoch": 0.5704445161984716,
-      "grad_norm": 0.6543012261390686,
-      "learning_rate": 0.0005663829328736127,
-      "loss": 4.0573,
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.6330726146697998,
+      "learning_rate": 0.0005663248785752833,
+      "loss": 4.0147,
       "step": 5300
     },
     {
-      "epoch": 0.5758260682380799,
-      "grad_norm": 0.7880141139030457,
-      "learning_rate": 0.0005660596918435512,
-      "loss": 4.037,
+      "epoch": 0.5768194070080862,
+      "grad_norm": 0.6442973017692566,
+      "learning_rate": 0.0005660010793308148,
+      "loss": 4.0265,
       "step": 5350
     },
     {
-      "epoch": 0.5812076202776881,
-      "grad_norm": 0.6059675216674805,
-      "learning_rate": 0.0005657364508134899,
-      "loss": 4.0243,
+      "epoch": 0.5822102425876011,
+      "grad_norm": 0.730826199054718,
+      "learning_rate": 0.0005656772800863465,
+      "loss": 4.0211,
       "step": 5400
     },
     {
-      "epoch": 0.5865891723172963,
-      "grad_norm": 0.5372035503387451,
-      "learning_rate": 0.0005654132097834284,
-      "loss": 4.044,
+      "epoch": 0.5876010781671159,
+      "grad_norm": 0.6084363460540771,
+      "learning_rate": 0.0005653534808418779,
+      "loss": 4.027,
       "step": 5450
     },
     {
-      "epoch": 0.5919707243569046,
-      "grad_norm": 0.652932345867157,
-      "learning_rate": 0.0005650899687533671,
-      "loss": 4.0212,
+      "epoch": 0.5929919137466307,
+      "grad_norm": 0.5735021829605103,
+      "learning_rate": 0.0005650296815974096,
+      "loss": 4.0287,
       "step": 5500
     },
     {
-      "epoch": 0.5973522763965128,
-      "grad_norm": 0.827506422996521,
-      "learning_rate": 0.0005647667277233056,
-      "loss": 4.0347,
+      "epoch": 0.5983827493261455,
+      "grad_norm": 0.627565860748291,
+      "learning_rate": 0.0005647058823529411,
+      "loss": 4.0346,
       "step": 5550
     },
     {
-      "epoch": 0.602733828436121,
-      "grad_norm": 0.6992992758750916,
-      "learning_rate": 0.0005644434866932442,
-      "loss": 4.0602,
+      "epoch": 0.6037735849056604,
+      "grad_norm": 0.6399509906768799,
+      "learning_rate": 0.0005643820831084727,
+      "loss": 4.0103,
       "step": 5600
     },
     {
-      "epoch": 0.6081153804757292,
-      "grad_norm": 0.6638625860214233,
-      "learning_rate": 0.0005641202456631828,
-      "loss": 4.0149,
+      "epoch": 0.6091644204851752,
+      "grad_norm": 0.5816676616668701,
+      "learning_rate": 0.0005640582838640042,
+      "loss": 4.0025,
       "step": 5650
     },
     {
-      "epoch": 0.6134969325153374,
-      "grad_norm": 0.6855315566062927,
-      "learning_rate": 0.0005637970046331214,
-      "loss": 3.9967,
+      "epoch": 0.6145552560646901,
+      "grad_norm": 0.6832612752914429,
+      "learning_rate": 0.0005637344846195358,
+      "loss": 4.0081,
       "step": 5700
     },
     {
-      "epoch": 0.6188784845549457,
-      "grad_norm": 0.594383955001831,
-      "learning_rate": 0.00056347376360306,
-      "loss": 4.0143,
+      "epoch": 0.6199460916442049,
+      "grad_norm": 0.596462070941925,
+      "learning_rate": 0.0005634106853750674,
+      "loss": 4.0183,
       "step": 5750
     },
     {
-      "epoch": 0.6242600365945539,
-      "grad_norm": 1.1790200471878052,
-      "learning_rate": 0.0005631505225729985,
-      "loss": 4.0077,
+      "epoch": 0.6253369272237197,
+      "grad_norm": 0.6222463250160217,
+      "learning_rate": 0.000563086886130599,
+      "loss": 4.0067,
       "step": 5800
     },
     {
-      "epoch": 0.6296415886341621,
-      "grad_norm": 0.6079861521720886,
-      "learning_rate": 0.0005628272815429371,
-      "loss": 4.0057,
+      "epoch": 0.6307277628032345,
+      "grad_norm": 0.5949306488037109,
+      "learning_rate": 0.0005627630868861305,
+      "loss": 4.0184,
       "step": 5850
     },
     {
-      "epoch": 0.6350231406737703,
-      "grad_norm": 0.6142423748970032,
-      "learning_rate": 0.0005625040405128757,
-      "loss": 4.0097,
+      "epoch": 0.6361185983827493,
+      "grad_norm": 0.5928850173950195,
+      "learning_rate": 0.0005624392876416621,
+      "loss": 3.9884,
       "step": 5900
     },
     {
-      "epoch": 0.6404046927133785,
-      "grad_norm": 0.6306520700454712,
-      "learning_rate": 0.0005621807994828143,
-      "loss": 4.0023,
+      "epoch": 0.6415094339622641,
+      "grad_norm": 0.6100660562515259,
+      "learning_rate": 0.0005621154883971937,
+      "loss": 3.9837,
       "step": 5950
     },
     {
-      "epoch": 0.6457862447529867,
-      "grad_norm": 0.5871993899345398,
-      "learning_rate": 0.0005618575584527529,
-      "loss": 4.0013,
+      "epoch": 0.6469002695417789,
+      "grad_norm": 0.6185354590415955,
+      "learning_rate": 0.0005617916891527253,
+      "loss": 3.9939,
       "step": 6000
     },
     {
-      "epoch": 0.6457862447529867,
-      "eval_accuracy": 0.327245113635166,
-      "eval_loss": 3.9299139976501465,
-      "eval_runtime": 183.1308,
-      "eval_samples_per_second": 98.35,
-      "eval_steps_per_second": 6.149,
+      "epoch": 0.6469002695417789,
+      "eval_accuracy": 0.32730726308699587,
+      "eval_loss": 3.918417453765869,
+      "eval_runtime": 184.0658,
+      "eval_samples_per_second": 97.851,
+      "eval_steps_per_second": 6.117,
       "step": 6000
     },
     {
-      "epoch": 0.651167796792595,
-      "grad_norm": 0.6969402432441711,
-      "learning_rate": 0.0005615407822432927,
-      "loss": 4.0079,
+      "epoch": 0.6522911051212938,
+      "grad_norm": 0.706950843334198,
+      "learning_rate": 0.0005614678899082568,
+      "loss": 3.9637,
       "step": 6050
     },
     {
-      "epoch": 0.6565493488322032,
-      "grad_norm": 0.6455370187759399,
-      "learning_rate": 0.0005612175412132313,
-      "loss": 3.976,
+      "epoch": 0.6576819407008087,
+      "grad_norm": 0.6735605597496033,
+      "learning_rate": 0.0005611440906637884,
+      "loss": 3.9968,
       "step": 6100
     },
     {
-      "epoch": 0.6619309008718114,
-      "grad_norm": 0.696642279624939,
-      "learning_rate": 0.0005608943001831699,
-      "loss": 3.9809,
+      "epoch": 0.6630727762803235,
+      "grad_norm": 0.6323850750923157,
+      "learning_rate": 0.00056082029141932,
+      "loss": 3.98,
       "step": 6150
     },
     {
-      "epoch": 0.6673124529114196,
-      "grad_norm": 0.6556864380836487,
-      "learning_rate": 0.0005605710591531085,
-      "loss": 3.9824,
+      "epoch": 0.6684636118598383,
+      "grad_norm": 0.6639269590377808,
+      "learning_rate": 0.0005604964921748515,
+      "loss": 3.9597,
       "step": 6200
     },
     {
-      "epoch": 0.6726940049510278,
-      "grad_norm": 0.5885508060455322,
-      "learning_rate": 0.000560247818123047,
-      "loss": 3.9573,
+      "epoch": 0.6738544474393531,
+      "grad_norm": 0.640203058719635,
+      "learning_rate": 0.000560172692930383,
+      "loss": 3.9737,
       "step": 6250
     },
     {
-      "epoch": 0.6780755569906362,
-      "grad_norm": 0.6368624567985535,
-      "learning_rate": 0.0005599245770929855,
-      "loss": 3.9602,
+      "epoch": 0.6792452830188679,
+      "grad_norm": 0.6835509538650513,
+      "learning_rate": 0.0005598488936859147,
+      "loss": 3.9901,
       "step": 6300
     },
     {
-      "epoch": 0.6834571090302444,
-      "grad_norm": 0.6564461588859558,
-      "learning_rate": 0.0005596013360629242,
-      "loss": 3.9735,
+      "epoch": 0.6846361185983828,
+      "grad_norm": 0.5967912673950195,
+      "learning_rate": 0.0005595250944414463,
+      "loss": 3.9536,
       "step": 6350
     },
     {
-      "epoch": 0.6888386610698526,
-      "grad_norm": 0.5797594785690308,
-      "learning_rate": 0.0005592780950328628,
-      "loss": 3.9884,
+      "epoch": 0.6900269541778976,
+      "grad_norm": Infinity,
+      "learning_rate": 0.0005592077711818672,
+      "loss": 3.958,
       "step": 6400
     },
     {
-      "epoch": 0.6942202131094608,
-      "grad_norm": 0.5808694958686829,
-      "learning_rate": 0.0005589548540028014,
-      "loss": 3.981,
+      "epoch": 0.6954177897574124,
+      "grad_norm": 0.6682020425796509,
+      "learning_rate": 0.0005588839719373988,
+      "loss": 3.9431,
       "step": 6450
     },
     {
-      "epoch": 0.699601765149069,
-      "grad_norm": 0.6396571397781372,
-      "learning_rate": 0.0005586316129727399,
-      "loss": 3.9754,
+      "epoch": 0.7008086253369272,
+      "grad_norm": 0.6564862132072449,
+      "learning_rate": 0.0005585601726929303,
+      "loss": 3.9578,
       "step": 6500
     },
     {
-      "epoch": 0.7049833171886772,
-      "grad_norm": 0.5532586574554443,
-      "learning_rate": 0.0005583083719426786,
-      "loss": 3.9448,
+      "epoch": 0.706199460916442,
+      "grad_norm": 0.7015867233276367,
+      "learning_rate": 0.0005582363734484619,
+      "loss": 3.9788,
       "step": 6550
     },
     {
-      "epoch": 0.7103648692282855,
-      "grad_norm": 0.6465068459510803,
-      "learning_rate": 0.0005579851309126171,
-      "loss": 3.9362,
+      "epoch": 0.7115902964959568,
+      "grad_norm": 0.6272884011268616,
+      "learning_rate": 0.0005579125742039935,
+      "loss": 3.9369,
       "step": 6600
     },
     {
-      "epoch": 0.7157464212678937,
-      "grad_norm": 0.6378588080406189,
-      "learning_rate": 0.0005576618898825558,
-      "loss": 3.9604,
+      "epoch": 0.7169811320754716,
+      "grad_norm": 0.619567334651947,
+      "learning_rate": 0.0005575887749595251,
+      "loss": 3.9679,
       "step": 6650
     },
     {
-      "epoch": 0.7211279733075019,
-      "grad_norm": 0.5902277231216431,
-      "learning_rate": 0.0005573386488524943,
-      "loss": 3.9683,
+      "epoch": 0.7223719676549866,
+      "grad_norm": 0.5466760396957397,
+      "learning_rate": 0.0005572649757150566,
+      "loss": 3.9409,
       "step": 6700
     },
     {
-      "epoch": 0.7265095253471101,
-      "grad_norm": 0.6843133568763733,
-      "learning_rate": 0.0005570154078224328,
-      "loss": 3.97,
+      "epoch": 0.7277628032345014,
+      "grad_norm": 0.5688589215278625,
+      "learning_rate": 0.0005569411764705882,
+      "loss": 3.9489,
       "step": 6750
     },
     {
-      "epoch": 0.7318910773867183,
-      "grad_norm": 0.5807896256446838,
-      "learning_rate": 0.0005566921667923715,
-      "loss": 3.9434,
+      "epoch": 0.7331536388140162,
+      "grad_norm": 0.638364851474762,
+      "learning_rate": 0.0005566173772261198,
+      "loss": 3.9358,
       "step": 6800
     },
     {
-      "epoch": 0.7372726294263265,
-      "grad_norm": 0.5920153856277466,
-      "learning_rate": 0.00055636892576231,
-      "loss": 3.9583,
+      "epoch": 0.738544474393531,
+      "grad_norm": 0.6768045425415039,
+      "learning_rate": 0.0005562935779816513,
+      "loss": 3.9168,
       "step": 6850
     },
     {
-      "epoch": 0.7426541814659348,
-      "grad_norm": 0.6305286288261414,
-      "learning_rate": 0.0005560456847322487,
-      "loss": 3.9369,
+      "epoch": 0.7439353099730458,
+      "grad_norm": 0.7640431523323059,
+      "learning_rate": 0.0005559697787371828,
+      "loss": 3.9429,
       "step": 6900
     },
     {
-      "epoch": 0.748035733505543,
-      "grad_norm": 0.6596742868423462,
-      "learning_rate": 0.0005557224437021872,
-      "loss": 3.9349,
+      "epoch": 0.7493261455525606,
+      "grad_norm": 0.6189766526222229,
+      "learning_rate": 0.0005556459794927145,
+      "loss": 3.9399,
       "step": 6950
     },
     {
-      "epoch": 0.7534172855451512,
-      "grad_norm": 0.5761475563049316,
-      "learning_rate": 0.0005553992026721258,
-      "loss": 3.9392,
+      "epoch": 0.7547169811320755,
+      "grad_norm": 0.7579498887062073,
+      "learning_rate": 0.000555322180248246,
+      "loss": 3.927,
       "step": 7000
     },
     {
-      "epoch": 0.7534172855451512,
-      "eval_accuracy": 0.33236201273390115,
-      "eval_loss": 3.871555805206299,
-      "eval_runtime": 182.3203,
-      "eval_samples_per_second": 98.788,
-      "eval_steps_per_second": 6.176,
+      "epoch": 0.7547169811320755,
+      "eval_accuracy": 0.3336846443372668,
+      "eval_loss": 3.8621718883514404,
+      "eval_runtime": 184.0825,
+      "eval_samples_per_second": 97.842,
+      "eval_steps_per_second": 6.117,
       "step": 7000
     },
     {
-      "epoch": 0.7587988375847594,
-      "grad_norm": 0.7002006769180298,
-      "learning_rate": 0.0005550759616420644,
-      "loss": 3.9563,
+      "epoch": 0.7601078167115903,
+      "grad_norm": 0.6436729431152344,
+      "learning_rate": 0.0005549983810037776,
+      "loss": 3.9171,
       "step": 7050
     },
     {
-      "epoch": 0.7641803896243676,
-      "grad_norm": 0.6454929113388062,
-      "learning_rate": 0.000554752720612003,
-      "loss": 3.9207,
+      "epoch": 0.7654986522911051,
+      "grad_norm": 0.6165031790733337,
+      "learning_rate": 0.0005546745817593091,
+      "loss": 3.9291,
       "step": 7100
     },
     {
-      "epoch": 0.7695619416639758,
-      "grad_norm": 0.5613568425178528,
-      "learning_rate": 0.0005544294795819415,
-      "loss": 3.9345,
+      "epoch": 0.77088948787062,
+      "grad_norm": 0.5492625832557678,
+      "learning_rate": 0.0005543507825148408,
+      "loss": 3.9279,
       "step": 7150
     },
     {
-      "epoch": 0.7749434937035841,
-      "grad_norm": 0.575527012348175,
-      "learning_rate": 0.0005541062385518801,
-      "loss": 3.9269,
+      "epoch": 0.7762803234501348,
+      "grad_norm": 0.6611822247505188,
+      "learning_rate": 0.0005540269832703723,
+      "loss": 3.9307,
       "step": 7200
     },
     {
-      "epoch": 0.7803250457431924,
-      "grad_norm": 0.5978026986122131,
-      "learning_rate": 0.0005537829975218188,
-      "loss": 3.9484,
+      "epoch": 0.7816711590296496,
+      "grad_norm": 0.5608403086662292,
+      "learning_rate": 0.0005537031840259039,
+      "loss": 3.8869,
       "step": 7250
     },
     {
-      "epoch": 0.7857065977828006,
-      "grad_norm": 0.5839754939079285,
-      "learning_rate": 0.0005534597564917573,
-      "loss": 3.9268,
+      "epoch": 0.7870619946091644,
+      "grad_norm": 0.6087973713874817,
+      "learning_rate": 0.0005533793847814354,
+      "loss": 3.9247,
       "step": 7300
     },
     {
-      "epoch": 0.7910881498224088,
-      "grad_norm": 0.6187883019447327,
-      "learning_rate": 0.0005531365154616959,
-      "loss": 3.9256,
+      "epoch": 0.7924528301886793,
+      "grad_norm": 0.5851709842681885,
+      "learning_rate": 0.000553055585536967,
+      "loss": 3.9109,
       "step": 7350
     },
     {
-      "epoch": 0.796469701862017,
-      "grad_norm": 0.6549730896949768,
-      "learning_rate": 0.0005528132744316344,
-      "loss": 3.92,
+      "epoch": 0.7978436657681941,
+      "grad_norm": 0.6298819780349731,
+      "learning_rate": 0.0005527317862924987,
+      "loss": 3.9101,
       "step": 7400
     },
     {
-      "epoch": 0.8018512539016253,
-      "grad_norm": 0.5937124490737915,
-      "learning_rate": 0.0005524900334015731,
-      "loss": 3.8838,
+      "epoch": 0.8032345013477089,
+      "grad_norm": 0.5659953355789185,
+      "learning_rate": 0.0005524079870480301,
+      "loss": 3.9013,
       "step": 7450
     },
     {
-      "epoch": 0.8072328059412335,
-      "grad_norm": 0.5783231854438782,
-      "learning_rate": 0.0005521667923715117,
-      "loss": 3.9049,
+      "epoch": 0.8086253369272237,
+      "grad_norm": 0.5907884240150452,
+      "learning_rate": 0.0005520841878035618,
+      "loss": 3.8877,
       "step": 7500
     },
     {
-      "epoch": 0.8126143579808417,
-      "grad_norm": 0.5541702508926392,
-      "learning_rate": 0.0005518435513414502,
-      "loss": 3.9001,
+      "epoch": 0.8140161725067385,
+      "grad_norm": 0.6394366025924683,
+      "learning_rate": 0.0005517603885590933,
+      "loss": 3.9146,
       "step": 7550
     },
     {
-      "epoch": 0.8179959100204499,
-      "grad_norm": 0.5935503244400024,
-      "learning_rate": 0.0005515203103113888,
-      "loss": 3.9101,
+      "epoch": 0.8194070080862533,
+      "grad_norm": 0.5494681000709534,
+      "learning_rate": 0.0005514365893146249,
+      "loss": 3.917,
       "step": 7600
     },
     {
-      "epoch": 0.8233774620600581,
-      "grad_norm": 0.6102548241615295,
-      "learning_rate": 0.0005511970692813274,
-      "loss": 3.9069,
+      "epoch": 0.8247978436657682,
+      "grad_norm": 0.5720885396003723,
+      "learning_rate": 0.0005511127900701564,
+      "loss": 3.892,
       "step": 7650
     },
     {
-      "epoch": 0.8287590140996663,
-      "grad_norm": 0.5769250988960266,
-      "learning_rate": 0.000550873828251266,
-      "loss": 3.9094,
+      "epoch": 0.8301886792452831,
+      "grad_norm": 0.584578812122345,
+      "learning_rate": 0.000550788990825688,
+      "loss": 3.8946,
       "step": 7700
     },
     {
-      "epoch": 0.8341405661392746,
-      "grad_norm": 0.5763354301452637,
-      "learning_rate": 0.0005505505872212045,
-      "loss": 3.9138,
+      "epoch": 0.8355795148247979,
+      "grad_norm": 0.532055139541626,
+      "learning_rate": 0.0005504651915812196,
+      "loss": 3.8789,
       "step": 7750
     },
     {
-      "epoch": 0.8395221181788828,
-      "grad_norm": 0.6475393176078796,
-      "learning_rate": 0.0005502273461911432,
-      "loss": 3.901,
+      "epoch": 0.8409703504043127,
+      "grad_norm": 0.5987346172332764,
+      "learning_rate": 0.0005501413923367512,
+      "loss": 3.9062,
       "step": 7800
     },
     {
-      "epoch": 0.844903670218491,
-      "grad_norm": 0.5787845849990845,
-      "learning_rate": 0.0005499041051610817,
-      "loss": 3.9031,
+      "epoch": 0.8463611859838275,
+      "grad_norm": 0.6259683966636658,
+      "learning_rate": 0.0005498175930922827,
+      "loss": 3.8846,
       "step": 7850
     },
     {
-      "epoch": 0.8502852222580992,
-      "grad_norm": 0.5721645355224609,
-      "learning_rate": 0.0005495808641310204,
-      "loss": 3.8975,
+      "epoch": 0.8517520215633423,
+      "grad_norm": 0.5504730939865112,
+      "learning_rate": 0.0005494937938478143,
+      "loss": 3.8864,
       "step": 7900
     },
     {
-      "epoch": 0.8556667742977074,
-      "grad_norm": 0.6674014925956726,
-      "learning_rate": 0.0005492576231009589,
-      "loss": 3.8725,
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.5990694761276245,
+      "learning_rate": 0.0005491699946033459,
+      "loss": 3.8832,
       "step": 7950
     },
     {
-      "epoch": 0.8610483263373157,
-      "grad_norm": 0.7338675260543823,
-      "learning_rate": 0.0005489343820708974,
-      "loss": 3.8996,
+      "epoch": 0.862533692722372,
+      "grad_norm": 0.5765665769577026,
+      "learning_rate": 0.0005488461953588775,
+      "loss": 3.8714,
       "step": 8000
     },
     {
-      "epoch": 0.8610483263373157,
-      "eval_accuracy": 0.3371352427484791,
-      "eval_loss": 3.8227832317352295,
-      "eval_runtime": 184.0783,
-      "eval_samples_per_second": 97.844,
-      "eval_steps_per_second": 6.117,
+      "epoch": 0.862533692722372,
+      "eval_accuracy": 0.3376904590052111,
+      "eval_loss": 3.8170006275177,
+      "eval_runtime": 183.7756,
+      "eval_samples_per_second": 98.005,
+      "eval_steps_per_second": 6.127,
       "step": 8000
     },
     {
-      "epoch": 0.8664298783769239,
-      "grad_norm": 0.5568352341651917,
-      "learning_rate": 0.0005486111410408361,
-      "loss": 3.8899,
+      "epoch": 0.8679245283018868,
+      "grad_norm": 0.6054647564888,
+      "learning_rate": 0.000548522396114409,
+      "loss": 3.8789,
       "step": 8050
     },
     {
-      "epoch": 0.8718114304165321,
-      "grad_norm": 0.5996348261833191,
-      "learning_rate": 0.0005482879000107746,
-      "loss": 3.8833,
+      "epoch": 0.8733153638814016,
+      "grad_norm": 0.5532026290893555,
+      "learning_rate": 0.0005481985968699406,
+      "loss": 3.8652,
       "step": 8100
     },
     {
-      "epoch": 0.8771929824561403,
-      "grad_norm": 0.6273235082626343,
-      "learning_rate": 0.0005479646589807133,
-      "loss": 3.8805,
+      "epoch": 0.8787061994609164,
+      "grad_norm": 0.5616852641105652,
+      "learning_rate": 0.0005478747976254721,
+      "loss": 3.8763,
       "step": 8150
     },
     {
-      "epoch": 0.8825745344957485,
-      "grad_norm": 0.6420896649360657,
-      "learning_rate": 0.0005476414179506518,
-      "loss": 3.8839,
+      "epoch": 0.8840970350404312,
+      "grad_norm": 0.5442996621131897,
+      "learning_rate": 0.0005475509983810037,
+      "loss": 3.8651,
       "step": 8200
     },
     {
-      "epoch": 0.8879560865353568,
-      "grad_norm": 0.5380816459655762,
-      "learning_rate": 0.0005473246417411916,
-      "loss": 3.8628,
+      "epoch": 0.889487870619946,
+      "grad_norm": 0.6385550498962402,
+      "learning_rate": 0.0005472271991365352,
+      "loss": 3.8525,
       "step": 8250
     },
     {
-      "epoch": 0.8933376385749651,
-      "grad_norm": 0.6761050224304199,
-      "learning_rate": 0.0005470014007111302,
-      "loss": 3.8861,
+      "epoch": 0.894878706199461,
+      "grad_norm": 0.6521883010864258,
+      "learning_rate": 0.0005469033998920669,
+      "loss": 3.8783,
       "step": 8300
     },
     {
-      "epoch": 0.8987191906145733,
-      "grad_norm": 0.6020174622535706,
-      "learning_rate": 0.0005466781596810688,
-      "loss": 3.8795,
+      "epoch": 0.9002695417789758,
+      "grad_norm": 0.6893941760063171,
+      "learning_rate": 0.0005465796006475984,
+      "loss": 3.8598,
       "step": 8350
     },
     {
-      "epoch": 0.9041007426541815,
-      "grad_norm": 0.6668645143508911,
-      "learning_rate": 0.0005463549186510073,
-      "loss": 3.8844,
+      "epoch": 0.9056603773584906,
+      "grad_norm": 0.6244086623191833,
+      "learning_rate": 0.00054625580140313,
+      "loss": 3.8736,
       "step": 8400
     },
     {
-      "epoch": 0.9094822946937897,
-      "grad_norm": 0.5670798420906067,
-      "learning_rate": 0.000546031677620946,
-      "loss": 3.8835,
+      "epoch": 0.9110512129380054,
+      "grad_norm": 0.6285154223442078,
+      "learning_rate": 0.000545938478143551,
+      "loss": 3.8853,
       "step": 8450
     },
     {
-      "epoch": 0.9148638467333979,
-      "grad_norm": 0.5120532512664795,
-      "learning_rate": 0.0005457084365908845,
-      "loss": 3.8562,
+      "epoch": 0.9164420485175202,
+      "grad_norm": 0.5592702627182007,
+      "learning_rate": 0.0005456146788990825,
+      "loss": 3.8774,
       "step": 8500
     },
     {
-      "epoch": 0.9202453987730062,
-      "grad_norm": 0.5259217023849487,
-      "learning_rate": 0.0005453851955608232,
-      "loss": 3.8672,
+      "epoch": 0.921832884097035,
+      "grad_norm": 0.6049821376800537,
+      "learning_rate": 0.000545290879654614,
+      "loss": 3.8566,
       "step": 8550
     },
     {
-      "epoch": 0.9256269508126144,
-      "grad_norm": 0.583608865737915,
-      "learning_rate": 0.0005450619545307617,
-      "loss": 3.8806,
+      "epoch": 0.9272237196765498,
+      "grad_norm": 0.5339543223381042,
+      "learning_rate": 0.0005449670804101457,
+      "loss": 3.8506,
       "step": 8600
     },
     {
-      "epoch": 0.9310085028522226,
-      "grad_norm": 0.64499431848526,
-      "learning_rate": 0.0005447387135007003,
-      "loss": 3.8743,
+      "epoch": 0.9326145552560647,
+      "grad_norm": 0.573306143283844,
+      "learning_rate": 0.0005446432811656773,
+      "loss": 3.8419,
       "step": 8650
     },
     {
-      "epoch": 0.9363900548918308,
-      "grad_norm": 0.5465984344482422,
-      "learning_rate": 0.0005444154724706389,
-      "loss": 3.8684,
+      "epoch": 0.9380053908355795,
+      "grad_norm": 0.5582360029220581,
+      "learning_rate": 0.0005443194819212088,
+      "loss": 3.8489,
       "step": 8700
     },
     {
-      "epoch": 0.941771606931439,
-      "grad_norm": 0.5706291198730469,
-      "learning_rate": 0.0005440922314405775,
-      "loss": 3.8527,
+      "epoch": 0.9433962264150944,
+      "grad_norm": 0.5276901125907898,
+      "learning_rate": 0.0005439956826767404,
+      "loss": 3.8465,
       "step": 8750
     },
     {
-      "epoch": 0.9471531589710472,
-      "grad_norm": 0.5907847285270691,
-      "learning_rate": 0.0005437689904105161,
-      "loss": 3.8579,
+      "epoch": 0.9487870619946092,
+      "grad_norm": 0.5552595853805542,
+      "learning_rate": 0.000543671883432272,
+      "loss": 3.8391,
       "step": 8800
     },
     {
-      "epoch": 0.9525347110106555,
-      "grad_norm": 0.6252136826515198,
-      "learning_rate": 0.0005434457493804546,
-      "loss": 3.8496,
+      "epoch": 0.954177897574124,
+      "grad_norm": 0.5062639117240906,
+      "learning_rate": 0.0005433480841878035,
+      "loss": 3.8313,
       "step": 8850
     },
     {
-      "epoch": 0.9579162630502637,
-      "grad_norm": 0.5910587310791016,
-      "learning_rate": 0.0005431225083503932,
-      "loss": 3.8492,
+      "epoch": 0.9595687331536388,
+      "grad_norm": 0.6003462076187134,
+      "learning_rate": 0.000543024284943335,
+      "loss": 3.8503,
       "step": 8900
     },
     {
-      "epoch": 0.9632978150898719,
-      "grad_norm": 0.5492550730705261,
-      "learning_rate": 0.0005427992673203318,
-      "loss": 3.8486,
+      "epoch": 0.9649595687331537,
+      "grad_norm": 0.6296499967575073,
+      "learning_rate": 0.0005427004856988667,
+      "loss": 3.8522,
       "step": 8950
     },
     {
-      "epoch": 0.9686793671294801,
-      "grad_norm": 0.6138157844543457,
-      "learning_rate": 0.0005424760262902704,
-      "loss": 3.8404,
+      "epoch": 0.9703504043126685,
+      "grad_norm": 0.6041927337646484,
+      "learning_rate": 0.0005423766864543982,
+      "loss": 3.841,
       "step": 9000
     },
     {
-      "epoch": 0.9686793671294801,
-      "eval_accuracy": 0.3408974576419433,
-      "eval_loss": 3.782958984375,
-      "eval_runtime": 183.7954,
-      "eval_samples_per_second": 97.995,
-      "eval_steps_per_second": 6.126,
+      "epoch": 0.9703504043126685,
+      "eval_accuracy": 0.34141247233018396,
+      "eval_loss": 3.777667999267578,
+      "eval_runtime": 184.2135,
+      "eval_samples_per_second": 97.772,
+      "eval_steps_per_second": 6.112,
       "step": 9000
     },
     {
-      "epoch": 0.9740609191690883,
-      "grad_norm": 0.5365199446678162,
-      "learning_rate": 0.000542152785260209,
-      "loss": 3.8344,
+      "epoch": 0.9757412398921833,
+      "grad_norm": 0.5688387751579285,
+      "learning_rate": 0.0005420528872099298,
+      "loss": 3.8361,
       "step": 9050
     },
     {
-      "epoch": 0.9794424712086965,
-      "grad_norm": 0.5951078534126282,
-      "learning_rate": 0.0005418295442301476,
-      "loss": 3.849,
+      "epoch": 0.9811320754716981,
+      "grad_norm": 0.6119299530982971,
+      "learning_rate": 0.0005417290879654613,
+      "loss": 3.8482,
       "step": 9100
     },
     {
-      "epoch": 0.9848240232483048,
-      "grad_norm": 0.586760401725769,
-      "learning_rate": 0.0005415063032000861,
-      "loss": 3.8422,
+      "epoch": 0.9865229110512129,
+      "grad_norm": 0.5677753686904907,
+      "learning_rate": 0.000541405288720993,
+      "loss": 3.8231,
       "step": 9150
     },
     {
-      "epoch": 0.9902055752879131,
-      "grad_norm": 0.5982652902603149,
-      "learning_rate": 0.0005411830621700248,
-      "loss": 3.8481,
+      "epoch": 0.9919137466307277,
+      "grad_norm": 0.5959908366203308,
+      "learning_rate": 0.0005410814894765245,
+      "loss": 3.8191,
       "step": 9200
     },
     {
-      "epoch": 0.9955871273275213,
-      "grad_norm": 0.5718969106674194,
-      "learning_rate": 0.0005408598211399633,
-      "loss": 3.8469,
+      "epoch": 0.9973045822102425,
+      "grad_norm": 0.5968837141990662,
+      "learning_rate": 0.0005407576902320561,
+      "loss": 3.8309,
       "step": 9250
     },
     {
-      "epoch": 1.0009686793671295,
-      "grad_norm": 0.6520361304283142,
-      "learning_rate": 0.0005405365801099019,
-      "loss": 3.8171,
+      "epoch": 1.0026954177897573,
+      "grad_norm": 0.5726041197776794,
+      "learning_rate": 0.0005404338909875876,
+      "loss": 3.8056,
       "step": 9300
     },
     {
-      "epoch": 1.0063502314067376,
-      "grad_norm": 0.5989835262298584,
-      "learning_rate": 0.0005402133390798405,
-      "loss": 3.7906,
+      "epoch": 1.0080862533692723,
+      "grad_norm": 0.565628707408905,
+      "learning_rate": 0.0005401100917431192,
+      "loss": 3.7709,
       "step": 9350
     },
     {
-      "epoch": 1.011731783446346,
-      "grad_norm": 0.5910343527793884,
-      "learning_rate": 0.000539890098049779,
-      "loss": 3.7743,
+      "epoch": 1.013477088948787,
+      "grad_norm": 0.5728825330734253,
+      "learning_rate": 0.0005397862924986508,
+      "loss": 3.7704,
       "step": 9400
     },
     {
-      "epoch": 1.017113335485954,
-      "grad_norm": 0.531580924987793,
-      "learning_rate": 0.0005395668570197177,
-      "loss": 3.7769,
+      "epoch": 1.0188679245283019,
+      "grad_norm": 0.5915510058403015,
+      "learning_rate": 0.0005394624932541824,
+      "loss": 3.7665,
       "step": 9450
     },
     {
-      "epoch": 1.0224948875255624,
-      "grad_norm": 0.5643661022186279,
-      "learning_rate": 0.0005392436159896562,
-      "loss": 3.7784,
+      "epoch": 1.0242587601078168,
+      "grad_norm": 0.6317964196205139,
+      "learning_rate": 0.0005391386940097139,
+      "loss": 3.7653,
       "step": 9500
     },
     {
-      "epoch": 1.0278764395651705,
-      "grad_norm": 0.5594536662101746,
-      "learning_rate": 0.0005389203749595948,
-      "loss": 3.7859,
+      "epoch": 1.0296495956873315,
+      "grad_norm": 0.5165658593177795,
+      "learning_rate": 0.0005388148947652455,
+      "loss": 3.7648,
       "step": 9550
     },
     {
-      "epoch": 1.0332579916047788,
-      "grad_norm": 0.5723885297775269,
-      "learning_rate": 0.0005385971339295334,
-      "loss": 3.7723,
+      "epoch": 1.0350404312668464,
+      "grad_norm": 0.6277357935905457,
+      "learning_rate": 0.000538491095520777,
+      "loss": 3.753,
       "step": 9600
     },
     {
-      "epoch": 1.0386395436443872,
-      "grad_norm": 0.5776793956756592,
-      "learning_rate": 0.000538273892899472,
-      "loss": 3.7622,
+      "epoch": 1.0404312668463611,
+      "grad_norm": 0.6124458312988281,
+      "learning_rate": 0.0005381672962763086,
+      "loss": 3.7933,
       "step": 9650
     },
     {
-      "epoch": 1.0440210956839953,
-      "grad_norm": 1.0429363250732422,
-      "learning_rate": 0.0005379506518694106,
-      "loss": 3.7716,
+      "epoch": 1.045822102425876,
+      "grad_norm": 0.6341213583946228,
+      "learning_rate": 0.0005378434970318403,
+      "loss": 3.779,
       "step": 9700
     },
     {
-      "epoch": 1.0494026477236036,
-      "grad_norm": 0.6570566296577454,
-      "learning_rate": 0.0005376274108393491,
-      "loss": 3.7688,
+      "epoch": 1.0512129380053907,
+      "grad_norm": 0.5159200429916382,
+      "learning_rate": 0.0005375196977873718,
+      "loss": 3.7762,
       "step": 9750
     },
     {
-      "epoch": 1.0547841997632117,
-      "grad_norm": 0.5422590374946594,
-      "learning_rate": 0.0005373041698092877,
-      "loss": 3.775,
+      "epoch": 1.0566037735849056,
+      "grad_norm": 0.5903022289276123,
+      "learning_rate": 0.0005371958985429034,
+      "loss": 3.7842,
       "step": 9800
     },
     {
-      "epoch": 1.06016575180282,
-      "grad_norm": 0.5695186853408813,
-      "learning_rate": 0.0005369809287792263,
-      "loss": 3.7836,
+      "epoch": 1.0619946091644206,
+      "grad_norm": 0.5840915441513062,
+      "learning_rate": 0.0005368720992984349,
+      "loss": 3.7802,
       "step": 9850
     },
     {
-      "epoch": 1.0655473038424281,
-      "grad_norm": 0.5725084543228149,
-      "learning_rate": 0.000536657687749165,
-      "loss": 3.7777,
+      "epoch": 1.0673854447439353,
+      "grad_norm": 0.5262959003448486,
+      "learning_rate": 0.0005365483000539665,
+      "loss": 3.7719,
       "step": 9900
     },
     {
-      "epoch": 1.0709288558820365,
-      "grad_norm": 0.5916622281074524,
-      "learning_rate": 0.0005363344467191035,
-      "loss": 3.7616,
+      "epoch": 1.0727762803234502,
+      "grad_norm": 0.5916352272033691,
+      "learning_rate": 0.0005362245008094981,
+      "loss": 3.7661,
       "step": 9950
     },
     {
-      "epoch": 1.0763104079216446,
-      "grad_norm": 0.6363723278045654,
-      "learning_rate": 0.000536011205689042,
-      "loss": 3.7704,
+      "epoch": 1.0781671159029649,
+      "grad_norm": 0.5976213216781616,
+      "learning_rate": 0.0005359007015650297,
+      "loss": 3.7593,
       "step": 10000
     },
     {
-      "epoch": 1.0763104079216446,
-      "eval_accuracy": 0.34435577040819043,
-      "eval_loss": 3.753837823867798,
-      "eval_runtime": 183.8836,
-      "eval_samples_per_second": 97.948,
+      "epoch": 1.0781671159029649,
+      "eval_accuracy": 0.3447647398779241,
+      "eval_loss": 3.7480316162109375,
+      "eval_runtime": 183.8854,
+      "eval_samples_per_second": 97.947,
       "eval_steps_per_second": 6.123,
       "step": 10000
     },
     {
-      "epoch": 1.081691959961253,
-      "grad_norm": 0.6295488476753235,
-      "learning_rate": 0.0005356879646589807,
-      "loss": 3.7654,
+      "epoch": 1.0835579514824798,
+      "grad_norm": 0.5457878708839417,
+      "learning_rate": 0.0005355769023205612,
+      "loss": 3.7618,
       "step": 10050
     },
     {
-      "epoch": 1.087073512000861,
-      "grad_norm": 0.5673714280128479,
-      "learning_rate": 0.0005353647236289192,
-      "loss": 3.7797,
+      "epoch": 1.0889487870619945,
+      "grad_norm": 0.602408766746521,
+      "learning_rate": 0.0005352531030760928,
+      "loss": 3.7722,
       "step": 10100
     },
     {
-      "epoch": 1.0924550640404693,
-      "grad_norm": 0.5380173325538635,
-      "learning_rate": 0.0005350414825988579,
-      "loss": 3.767,
+      "epoch": 1.0943396226415094,
+      "grad_norm": 0.6541560888290405,
+      "learning_rate": 0.0005349293038316244,
+      "loss": 3.7648,
       "step": 10150
     },
     {
-      "epoch": 1.0978366160800774,
-      "grad_norm": 0.5471936464309692,
-      "learning_rate": 0.0005347182415687964,
-      "loss": 3.7705,
+      "epoch": 1.0997304582210243,
+      "grad_norm": 0.6203254461288452,
+      "learning_rate": 0.0005346055045871559,
+      "loss": 3.766,
       "step": 10200
     },
     {
-      "epoch": 1.1032181681196858,
-      "grad_norm": 0.5841044187545776,
-      "learning_rate": 0.000534395000538735,
-      "loss": 3.7676,
+      "epoch": 1.105121293800539,
+      "grad_norm": 0.5233330130577087,
+      "learning_rate": 0.0005342817053426874,
+      "loss": 3.7736,
       "step": 10250
     },
     {
-      "epoch": 1.1085997201592939,
-      "grad_norm": 0.5935708284378052,
-      "learning_rate": 0.0005340717595086736,
-      "loss": 3.7552,
+      "epoch": 1.110512129380054,
+      "grad_norm": 0.5436099171638489,
+      "learning_rate": 0.0005339579060982191,
+      "loss": 3.7584,
       "step": 10300
     },
     {
-      "epoch": 1.1139812721989022,
-      "grad_norm": 0.5655228495597839,
-      "learning_rate": 0.0005337485184786122,
-      "loss": 3.7727,
+      "epoch": 1.1159029649595686,
+      "grad_norm": 0.8453336358070374,
+      "learning_rate": 0.0005336341068537506,
+      "loss": 3.7473,
       "step": 10350
     },
     {
-      "epoch": 1.1193628242385103,
-      "grad_norm": 0.5181326270103455,
-      "learning_rate": 0.0005334252774485507,
-      "loss": 3.7737,
+      "epoch": 1.1212938005390836,
+      "grad_norm": 0.6195657849311829,
+      "learning_rate": 0.0005333103076092822,
+      "loss": 3.7636,
       "step": 10400
     },
     {
-      "epoch": 1.1247443762781186,
-      "grad_norm": 0.5479722619056702,
-      "learning_rate": 0.0005331085012390905,
-      "loss": 3.7639,
+      "epoch": 1.1266846361185983,
+      "grad_norm": 0.5664616227149963,
+      "learning_rate": 0.0005329929843497031,
+      "loss": 3.7516,
       "step": 10450
     },
     {
-      "epoch": 1.1301259283177267,
-      "grad_norm": 0.6414276361465454,
-      "learning_rate": 0.0005327852602090292,
-      "loss": 3.7678,
+      "epoch": 1.1320754716981132,
+      "grad_norm": 0.6112706065177917,
+      "learning_rate": 0.0005326691851052347,
+      "loss": 3.7407,
       "step": 10500
     },
     {
-      "epoch": 1.135507480357335,
-      "grad_norm": 0.5645982623100281,
-      "learning_rate": 0.0005324620191789678,
-      "loss": 3.781,
+      "epoch": 1.137466307277628,
+      "grad_norm": 0.5114481449127197,
+      "learning_rate": 0.0005323453858607662,
+      "loss": 3.7612,
       "step": 10550
     },
     {
-      "epoch": 1.1408890323969434,
-      "grad_norm": 0.5996174812316895,
-      "learning_rate": 0.0005321387781489063,
-      "loss": 3.7616,
+      "epoch": 1.1428571428571428,
+      "grad_norm": 0.6120179891586304,
+      "learning_rate": 0.0005320215866162979,
+      "loss": 3.7726,
       "step": 10600
     },
     {
-      "epoch": 1.1462705844365515,
-      "grad_norm": 0.5798478126525879,
-      "learning_rate": 0.0005318155371188449,
-      "loss": 3.7509,
+      "epoch": 1.1482479784366577,
+      "grad_norm": 0.6177209615707397,
+      "learning_rate": 0.0005316977873718294,
+      "loss": 3.7496,
       "step": 10650
     },
     {
-      "epoch": 1.1516521364761596,
-      "grad_norm": 0.6145731210708618,
-      "learning_rate": 0.0005314922960887834,
-      "loss": 3.742,
+      "epoch": 1.1536388140161726,
+      "grad_norm": 0.6478197574615479,
+      "learning_rate": 0.000531373988127361,
+      "loss": 3.7485,
       "step": 10700
     },
     {
-      "epoch": 1.157033688515768,
-      "grad_norm": 0.5426759123802185,
-      "learning_rate": 0.0005311690550587221,
-      "loss": 3.7573,
+      "epoch": 1.1590296495956873,
+      "grad_norm": 0.6623829007148743,
+      "learning_rate": 0.0005310501888828925,
+      "loss": 3.7359,
       "step": 10750
     },
     {
-      "epoch": 1.1624152405553763,
-      "grad_norm": 0.6367711424827576,
-      "learning_rate": 0.0005308458140286607,
-      "loss": 3.7447,
+      "epoch": 1.1644204851752022,
+      "grad_norm": 0.5936301350593567,
+      "learning_rate": 0.0005307263896384242,
+      "loss": 3.7265,
       "step": 10800
     },
     {
-      "epoch": 1.1677967925949844,
-      "grad_norm": 0.5617762804031372,
-      "learning_rate": 0.0005305225729985993,
-      "loss": 3.7552,
+      "epoch": 1.169811320754717,
+      "grad_norm": 0.5386740565299988,
+      "learning_rate": 0.0005304025903939556,
+      "loss": 3.7447,
       "step": 10850
     },
     {
-      "epoch": 1.1731783446345927,
-      "grad_norm": 0.5642740726470947,
-      "learning_rate": 0.0005301993319685378,
-      "loss": 3.7717,
+      "epoch": 1.1752021563342319,
+      "grad_norm": 0.5326955318450928,
+      "learning_rate": 0.0005300787911494873,
+      "loss": 3.7473,
       "step": 10900
     },
     {
-      "epoch": 1.1785598966742008,
-      "grad_norm": 0.5389478206634521,
-      "learning_rate": 0.0005298760909384765,
-      "loss": 3.7591,
+      "epoch": 1.1805929919137466,
+      "grad_norm": 0.5992839336395264,
+      "learning_rate": 0.0005297549919050189,
+      "loss": 3.7556,
       "step": 10950
     },
     {
-      "epoch": 1.1839414487138091,
-      "grad_norm": 0.5747219920158386,
-      "learning_rate": 0.0005295593147290163,
-      "loss": 3.7569,
+      "epoch": 1.1859838274932615,
+      "grad_norm": 0.5615512728691101,
+      "learning_rate": 0.0005294311926605504,
+      "loss": 3.7736,
       "step": 11000
     },
     {
-      "epoch": 1.1839414487138091,
-      "eval_accuracy": 0.34640528983103497,
-      "eval_loss": 3.7258360385894775,
-      "eval_runtime": 181.9752,
-      "eval_samples_per_second": 98.975,
-      "eval_steps_per_second": 6.188,
+      "epoch": 1.1859838274932615,
+      "eval_accuracy": 0.3467401580312792,
+      "eval_loss": 3.7230610847473145,
+      "eval_runtime": 183.7379,
+      "eval_samples_per_second": 98.025,
+      "eval_steps_per_second": 6.128,
       "step": 11000
     },
     {
-      "epoch": 1.1893230007534172,
-      "grad_norm": 0.535572350025177,
-      "learning_rate": 0.0005292360736989548,
-      "loss": 3.7503,
+      "epoch": 1.1913746630727764,
+      "grad_norm": 0.5490072965621948,
+      "learning_rate": 0.000529107393416082,
+      "loss": 3.7609,
       "step": 11050
     },
     {
-      "epoch": 1.1947045527930256,
-      "grad_norm": 0.5539134740829468,
-      "learning_rate": 0.0005289128326688933,
-      "loss": 3.7425,
+      "epoch": 1.196765498652291,
+      "grad_norm": 0.5817831158638,
+      "learning_rate": 0.0005287835941716135,
+      "loss": 3.754,
       "step": 11100
     },
     {
-      "epoch": 1.2000861048326337,
-      "grad_norm": 0.5157566070556641,
-      "learning_rate": 0.000528589591638832,
-      "loss": 3.7573,
+      "epoch": 1.202156334231806,
+      "grad_norm": 0.5200773477554321,
+      "learning_rate": 0.0005284597949271452,
+      "loss": 3.7514,
       "step": 11150
     },
     {
-      "epoch": 1.205467656872242,
-      "grad_norm": 0.6375812888145447,
-      "learning_rate": 0.0005282663506087705,
-      "loss": 3.747,
+      "epoch": 1.2075471698113207,
+      "grad_norm": 0.5576964616775513,
+      "learning_rate": 0.0005281359956826767,
+      "loss": 3.7413,
       "step": 11200
     },
     {
-      "epoch": 1.21084920891185,
-      "grad_norm": 0.5762513279914856,
-      "learning_rate": 0.0005279431095787092,
-      "loss": 3.7664,
+      "epoch": 1.2129380053908356,
+      "grad_norm": 0.5839444994926453,
+      "learning_rate": 0.0005278121964382083,
+      "loss": 3.7341,
       "step": 11250
     },
     {
-      "epoch": 1.2162307609514584,
-      "grad_norm": 0.5227396488189697,
-      "learning_rate": 0.0005276198685486477,
-      "loss": 3.7466,
+      "epoch": 1.2183288409703503,
+      "grad_norm": 0.5846340656280518,
+      "learning_rate": 0.0005274883971937398,
+      "loss": 3.7496,
       "step": 11300
     },
     {
-      "epoch": 1.2216123129910665,
-      "grad_norm": 0.5795957446098328,
-      "learning_rate": 0.0005272966275185863,
-      "loss": 3.7608,
+      "epoch": 1.2237196765498652,
+      "grad_norm": 0.6075339317321777,
+      "learning_rate": 0.0005271645979492714,
+      "loss": 3.7346,
       "step": 11350
     },
     {
-      "epoch": 1.2269938650306749,
-      "grad_norm": 0.6271036863327026,
-      "learning_rate": 0.0005269733864885249,
-      "loss": 3.7467,
+      "epoch": 1.2291105121293802,
+      "grad_norm": 0.567361056804657,
+      "learning_rate": 0.000526840798704803,
+      "loss": 3.7509,
       "step": 11400
     },
     {
-      "epoch": 1.232375417070283,
-      "grad_norm": 0.6098865866661072,
-      "learning_rate": 0.0005266501454584636,
-      "loss": 3.7546,
+      "epoch": 1.2345013477088949,
+      "grad_norm": 0.53346848487854,
+      "learning_rate": 0.0005265169994603346,
+      "loss": 3.7406,
       "step": 11450
     },
     {
-      "epoch": 1.2377569691098913,
-      "grad_norm": 0.5986586809158325,
-      "learning_rate": 0.0005263269044284021,
-      "loss": 3.7606,
+      "epoch": 1.2398921832884098,
+      "grad_norm": 0.6361684799194336,
+      "learning_rate": 0.0005261932002158661,
+      "loss": 3.7335,
       "step": 11500
     },
     {
-      "epoch": 1.2431385211494996,
-      "grad_norm": 0.564387857913971,
-      "learning_rate": 0.0005260036633983406,
-      "loss": 3.7539,
+      "epoch": 1.2452830188679245,
+      "grad_norm": 0.5192483067512512,
+      "learning_rate": 0.0005258694009713977,
+      "loss": 3.7439,
       "step": 11550
     },
     {
-      "epoch": 1.2485200731891077,
-      "grad_norm": 0.5610516667366028,
-      "learning_rate": 0.0005256804223682792,
-      "loss": 3.7399,
+      "epoch": 1.2506738544474394,
+      "grad_norm": 0.5366946458816528,
+      "learning_rate": 0.0005255456017269292,
+      "loss": 3.7197,
       "step": 11600
     },
     {
-      "epoch": 1.2539016252287158,
-      "grad_norm": 0.535050094127655,
-      "learning_rate": 0.0005253571813382178,
-      "loss": 3.7442,
+      "epoch": 1.256064690026954,
+      "grad_norm": 0.605530321598053,
+      "learning_rate": 0.0005252218024824608,
+      "loss": 3.7509,
       "step": 11650
     },
     {
-      "epoch": 1.2592831772683242,
-      "grad_norm": 0.600563645362854,
-      "learning_rate": 0.0005250339403081564,
-      "loss": 3.7306,
+      "epoch": 1.261455525606469,
+      "grad_norm": 0.6748290061950684,
+      "learning_rate": 0.0005248980032379924,
+      "loss": 3.7468,
       "step": 11700
     },
     {
-      "epoch": 1.2646647293079325,
-      "grad_norm": 0.6017090678215027,
-      "learning_rate": 0.000524710699278095,
-      "loss": 3.7414,
+      "epoch": 1.266846361185984,
+      "grad_norm": 0.5584892630577087,
+      "learning_rate": 0.000524574203993524,
+      "loss": 3.7387,
       "step": 11750
     },
     {
-      "epoch": 1.2700462813475406,
-      "grad_norm": 0.7243601679801941,
-      "learning_rate": 0.0005243874582480336,
-      "loss": 3.748,
+      "epoch": 1.2722371967654986,
+      "grad_norm": 0.5633678436279297,
+      "learning_rate": 0.0005242504047490555,
+      "loss": 3.731,
       "step": 11800
     },
     {
-      "epoch": 1.275427833387149,
-      "grad_norm": 0.5405592918395996,
-      "learning_rate": 0.0005240642172179722,
-      "loss": 3.7235,
+      "epoch": 1.2776280323450135,
+      "grad_norm": 0.5874388813972473,
+      "learning_rate": 0.0005239266055045871,
+      "loss": 3.7422,
       "step": 11850
     },
     {
-      "epoch": 1.280809385426757,
-      "grad_norm": 0.5994392037391663,
-      "learning_rate": 0.0005237409761879107,
-      "loss": 3.7396,
+      "epoch": 1.2830188679245282,
+      "grad_norm": 0.5580216646194458,
+      "learning_rate": 0.0005236028062601186,
+      "loss": 3.7282,
       "step": 11900
     },
     {
-      "epoch": 1.2861909374663654,
-      "grad_norm": 0.568977415561676,
-      "learning_rate": 0.0005234177351578493,
-      "loss": 3.7417,
+      "epoch": 1.2884097035040432,
+      "grad_norm": 0.6376722455024719,
+      "learning_rate": 0.0005232790070156503,
+      "loss": 3.7408,
       "step": 11950
     },
     {
-      "epoch": 1.2915724895059735,
-      "grad_norm": 0.5830429196357727,
-      "learning_rate": 0.0005230944941277878,
-      "loss": 3.7236,
+      "epoch": 1.2938005390835579,
+      "grad_norm": 0.6610795259475708,
+      "learning_rate": 0.0005229552077711818,
+      "loss": 3.7349,
       "step": 12000
     },
     {
-      "epoch": 1.2915724895059735,
-      "eval_accuracy": 0.3490606577563331,
-      "eval_loss": 3.7004106044769287,
-      "eval_runtime": 181.7656,
-      "eval_samples_per_second": 99.089,
-      "eval_steps_per_second": 6.195,
+      "epoch": 1.2938005390835579,
+      "eval_accuracy": 0.34876555651302893,
+      "eval_loss": 3.6991491317749023,
+      "eval_runtime": 184.2136,
+      "eval_samples_per_second": 97.772,
+      "eval_steps_per_second": 6.112,
       "step": 12000
     },
     {
-      "epoch": 1.2969540415455818,
-      "grad_norm": 0.5280687212944031,
-      "learning_rate": 0.0005227712530977265,
-      "loss": 3.7264,
+      "epoch": 1.2991913746630728,
+      "grad_norm": 0.5748873353004456,
+      "learning_rate": 0.0005226314085267134,
+      "loss": 3.7327,
       "step": 12050
     },
     {
-      "epoch": 1.30233559358519,
-      "grad_norm": 0.5419384837150574,
-      "learning_rate": 0.0005224480120676651,
-      "loss": 3.7493,
+      "epoch": 1.3045822102425877,
+      "grad_norm": 0.585300087928772,
+      "learning_rate": 0.000522307609282245,
+      "loss": 3.7399,
       "step": 12100
     },
     {
-      "epoch": 1.3077171456247982,
-      "grad_norm": 0.5870584845542908,
-      "learning_rate": 0.0005221247710376037,
-      "loss": 3.7321,
+      "epoch": 1.3099730458221024,
+      "grad_norm": 0.5574855208396912,
+      "learning_rate": 0.0005219838100377766,
+      "loss": 3.7566,
       "step": 12150
     },
     {
-      "epoch": 1.3130986976644063,
-      "grad_norm": 0.6159378886222839,
-      "learning_rate": 0.0005218015300075422,
-      "loss": 3.7377,
+      "epoch": 1.3153638814016173,
+      "grad_norm": 0.5814098119735718,
+      "learning_rate": 0.000521660010793308,
+      "loss": 3.7296,
       "step": 12200
     },
     {
-      "epoch": 1.3184802497040147,
-      "grad_norm": 0.6157721281051636,
-      "learning_rate": 0.0005214782889774809,
-      "loss": 3.7259,
+      "epoch": 1.320754716981132,
+      "grad_norm": 0.5733814835548401,
+      "learning_rate": 0.0005213362115488396,
+      "loss": 3.733,
       "step": 12250
     },
     {
-      "epoch": 1.3238618017436228,
-      "grad_norm": 0.6117875576019287,
-      "learning_rate": 0.0005211550479474194,
-      "loss": 3.7355,
+      "epoch": 1.326145552560647,
+      "grad_norm": 0.5674881935119629,
+      "learning_rate": 0.0005210124123043713,
+      "loss": 3.7457,
       "step": 12300
     },
     {
-      "epoch": 1.329243353783231,
-      "grad_norm": 0.6459168791770935,
-      "learning_rate": 0.000520831806917358,
-      "loss": 3.7286,
+      "epoch": 1.3315363881401616,
+      "grad_norm": 0.6036843657493591,
+      "learning_rate": 0.0005206886130599028,
+      "loss": 3.7308,
       "step": 12350
     },
     {
-      "epoch": 1.3346249058228392,
-      "grad_norm": 0.5380271673202515,
-      "learning_rate": 0.0005205085658872966,
-      "loss": 3.7315,
+      "epoch": 1.3369272237196765,
+      "grad_norm": 0.6163053512573242,
+      "learning_rate": 0.0005203648138154344,
+      "loss": 3.7192,
       "step": 12400
     },
     {
-      "epoch": 1.3400064578624475,
-      "grad_norm": 0.5702387094497681,
-      "learning_rate": 0.0005201853248572351,
-      "loss": 3.7275,
+      "epoch": 1.3423180592991915,
+      "grad_norm": 0.601905107498169,
+      "learning_rate": 0.0005200410145709659,
+      "loss": 3.7347,
       "step": 12450
     },
     {
-      "epoch": 1.3453880099020559,
-      "grad_norm": 0.6522827744483948,
-      "learning_rate": 0.0005198620838271738,
-      "loss": 3.731,
+      "epoch": 1.3477088948787062,
+      "grad_norm": 0.748714804649353,
+      "learning_rate": 0.0005197172153264976,
+      "loss": 3.7148,
       "step": 12500
     },
     {
-      "epoch": 1.350769561941664,
-      "grad_norm": 0.6003380417823792,
-      "learning_rate": 0.0005195388427971123,
-      "loss": 3.7168,
+      "epoch": 1.353099730458221,
+      "grad_norm": 0.5848150849342346,
+      "learning_rate": 0.0005193998920669184,
+      "loss": 3.7287,
       "step": 12550
     },
     {
-      "epoch": 1.356151113981272,
-      "grad_norm": 0.5687087178230286,
-      "learning_rate": 0.000519215601767051,
-      "loss": 3.7108,
+      "epoch": 1.3584905660377358,
+      "grad_norm": 0.6040074229240417,
+      "learning_rate": 0.0005190760928224501,
+      "loss": 3.7235,
       "step": 12600
     },
     {
-      "epoch": 1.3615326660208804,
-      "grad_norm": 0.602519690990448,
-      "learning_rate": 0.0005188923607369895,
-      "loss": 3.7322,
+      "epoch": 1.3638814016172507,
+      "grad_norm": 0.5899044275283813,
+      "learning_rate": 0.0005187522935779816,
+      "loss": 3.7114,
       "step": 12650
     },
     {
-      "epoch": 1.3669142180604887,
-      "grad_norm": 0.5567368268966675,
-      "learning_rate": 0.000518569119706928,
-      "loss": 3.7188,
+      "epoch": 1.3692722371967654,
+      "grad_norm": 0.5793396830558777,
+      "learning_rate": 0.0005184284943335132,
+      "loss": 3.7091,
       "step": 12700
     },
     {
-      "epoch": 1.3722957701000968,
-      "grad_norm": 0.6820892095565796,
-      "learning_rate": 0.0005182458786768667,
-      "loss": 3.7278,
+      "epoch": 1.3746630727762803,
+      "grad_norm": 0.6218591332435608,
+      "learning_rate": 0.0005181046950890447,
+      "loss": 3.7277,
       "step": 12750
     },
     {
-      "epoch": 1.3776773221397052,
-      "grad_norm": 0.6009151339530945,
-      "learning_rate": 0.0005179226376468052,
-      "loss": 3.7225,
+      "epoch": 1.3800539083557952,
+      "grad_norm": 0.6013187766075134,
+      "learning_rate": 0.0005177808958445764,
+      "loss": 3.7084,
       "step": 12800
     },
     {
-      "epoch": 1.3830588741793133,
-      "grad_norm": 0.5885601043701172,
-      "learning_rate": 0.0005175993966167438,
-      "loss": 3.7263,
+      "epoch": 1.38544474393531,
+      "grad_norm": 0.6219421625137329,
+      "learning_rate": 0.0005174570966001078,
+      "loss": 3.7217,
       "step": 12850
     },
     {
-      "epoch": 1.3884404262189216,
-      "grad_norm": 0.6636022329330444,
-      "learning_rate": 0.0005172761555866824,
-      "loss": 3.7311,
+      "epoch": 1.3908355795148248,
+      "grad_norm": 0.5010536909103394,
+      "learning_rate": 0.0005171332973556395,
+      "loss": 3.7183,
       "step": 12900
     },
     {
-      "epoch": 1.3938219782585297,
-      "grad_norm": 0.6068388223648071,
-      "learning_rate": 0.0005169529145566211,
-      "loss": 3.7087,
+      "epoch": 1.3962264150943398,
+      "grad_norm": 0.5664235353469849,
+      "learning_rate": 0.000516809498111171,
+      "loss": 3.7112,
       "step": 12950
     },
     {
-      "epoch": 1.399203530298138,
-      "grad_norm": 0.534371018409729,
-      "learning_rate": 0.0005166296735265596,
-      "loss": 3.7062,
+      "epoch": 1.4016172506738545,
+      "grad_norm": 0.5107662081718445,
+      "learning_rate": 0.0005164856988667026,
+      "loss": 3.7068,
       "step": 13000
     },
     {
-      "epoch": 1.399203530298138,
-      "eval_accuracy": 0.35125294707376586,
-      "eval_loss": 3.679492712020874,
-      "eval_runtime": 182.0328,
-      "eval_samples_per_second": 98.944,
-      "eval_steps_per_second": 6.186,
+      "epoch": 1.4016172506738545,
+      "eval_accuracy": 0.35189367315320785,
+      "eval_loss": 3.675448417663574,
+      "eval_runtime": 183.7852,
+      "eval_samples_per_second": 98.0,
+      "eval_steps_per_second": 6.127,
       "step": 13000
     },
     {
-      "epoch": 1.4045850823377461,
-      "grad_norm": 0.5494077801704407,
-      "learning_rate": 0.0005163064324964982,
-      "loss": 3.7203,
+      "epoch": 1.4070080862533692,
+      "grad_norm": 0.5422387719154358,
+      "learning_rate": 0.0005161618996222341,
+      "loss": 3.6854,
       "step": 13050
     },
     {
-      "epoch": 1.4099666343773545,
-      "grad_norm": 0.573820948600769,
-      "learning_rate": 0.0005159831914664367,
-      "loss": 3.6958,
+      "epoch": 1.412398921832884,
+      "grad_norm": 0.5654274821281433,
+      "learning_rate": 0.0005158381003777657,
+      "loss": 3.7081,
       "step": 13100
     },
     {
-      "epoch": 1.4153481864169626,
-      "grad_norm": 0.6078840494155884,
-      "learning_rate": 0.0005156599504363753,
-      "loss": 3.7048,
+      "epoch": 1.417789757412399,
+      "grad_norm": 0.6061548590660095,
+      "learning_rate": 0.0005155143011332973,
+      "loss": 3.7127,
       "step": 13150
     },
     {
-      "epoch": 1.420729738456571,
-      "grad_norm": 0.6435017585754395,
-      "learning_rate": 0.000515336709406314,
-      "loss": 3.719,
+      "epoch": 1.4231805929919137,
+      "grad_norm": 0.5701354742050171,
+      "learning_rate": 0.0005151905018888289,
+      "loss": 3.6938,
       "step": 13200
     },
     {
-      "epoch": 1.426111290496179,
-      "grad_norm": 0.5942484736442566,
-      "learning_rate": 0.0005150134683762525,
-      "loss": 3.724,
+      "epoch": 1.4285714285714286,
+      "grad_norm": 0.5603214502334595,
+      "learning_rate": 0.0005148667026443604,
+      "loss": 3.7115,
       "step": 13250
     },
     {
-      "epoch": 1.4314928425357873,
-      "grad_norm": 0.5379244685173035,
-      "learning_rate": 0.0005146902273461911,
-      "loss": 3.7028,
+      "epoch": 1.4339622641509435,
+      "grad_norm": 0.5795431733131409,
+      "learning_rate": 0.000514542903399892,
+      "loss": 3.7066,
       "step": 13300
     },
     {
-      "epoch": 1.4368743945753955,
-      "grad_norm": 0.5623278617858887,
-      "learning_rate": 0.0005143669863161297,
-      "loss": 3.7117,
+      "epoch": 1.4393530997304582,
+      "grad_norm": 0.5759278535842896,
+      "learning_rate": 0.0005142191041554237,
+      "loss": 3.7127,
       "step": 13350
     },
     {
-      "epoch": 1.4422559466150038,
-      "grad_norm": 0.6037217378616333,
-      "learning_rate": 0.0005140437452860683,
-      "loss": 3.7123,
+      "epoch": 1.444743935309973,
+      "grad_norm": 0.5491745471954346,
+      "learning_rate": 0.0005138953049109552,
+      "loss": 3.7014,
       "step": 13400
     },
     {
-      "epoch": 1.447637498654612,
-      "grad_norm": 0.6184905767440796,
-      "learning_rate": 0.0005137205042560069,
-      "loss": 3.7258,
+      "epoch": 1.4501347708894878,
+      "grad_norm": 0.5556439757347107,
+      "learning_rate": 0.0005135715056664868,
+      "loss": 3.7188,
       "step": 13450
     },
     {
-      "epoch": 1.4530190506942202,
-      "grad_norm": 0.5425773859024048,
-      "learning_rate": 0.0005133972632259455,
-      "loss": 3.6993,
+      "epoch": 1.4555256064690028,
+      "grad_norm": 0.6236615180969238,
+      "learning_rate": 0.0005132477064220183,
+      "loss": 3.7112,
       "step": 13500
     },
     {
-      "epoch": 1.4584006027338283,
-      "grad_norm": 0.6625587344169617,
-      "learning_rate": 0.000513074022195884,
-      "loss": 3.6997,
+      "epoch": 1.4609164420485174,
+      "grad_norm": 0.6058287024497986,
+      "learning_rate": 0.0005129239071775499,
+      "loss": 3.6881,
       "step": 13550
     },
     {
-      "epoch": 1.4637821547734367,
-      "grad_norm": 0.5868763327598572,
-      "learning_rate": 0.0005127507811658226,
-      "loss": 3.7111,
+      "epoch": 1.4663072776280324,
+      "grad_norm": 0.560505211353302,
+      "learning_rate": 0.0005126001079330814,
+      "loss": 3.7167,
       "step": 13600
     },
     {
-      "epoch": 1.469163706813045,
-      "grad_norm": 0.5310501456260681,
-      "learning_rate": 0.0005124275401357612,
-      "loss": 3.7201,
+      "epoch": 1.4716981132075473,
+      "grad_norm": 0.5141763687133789,
+      "learning_rate": 0.000512276308688613,
+      "loss": 3.6977,
       "step": 13650
     },
     {
-      "epoch": 1.474545258852653,
-      "grad_norm": 0.5664429664611816,
-      "learning_rate": 0.0005121042991056997,
-      "loss": 3.689,
+      "epoch": 1.477088948787062,
+      "grad_norm": 0.5746728181838989,
+      "learning_rate": 0.0005119525094441446,
+      "loss": 3.7167,
       "step": 13700
     },
     {
-      "epoch": 1.4799268108922612,
-      "grad_norm": 0.5378715395927429,
-      "learning_rate": 0.0005117810580756384,
-      "loss": 3.7096,
+      "epoch": 1.482479784366577,
+      "grad_norm": 0.5270335078239441,
+      "learning_rate": 0.0005116287101996762,
+      "loss": 3.6876,
       "step": 13750
     },
     {
-      "epoch": 1.4853083629318695,
-      "grad_norm": 0.5423511266708374,
-      "learning_rate": 0.0005114578170455769,
-      "loss": 3.7038,
+      "epoch": 1.4878706199460916,
+      "grad_norm": 0.5588323473930359,
+      "learning_rate": 0.0005113049109552077,
+      "loss": 3.7122,
       "step": 13800
     },
     {
-      "epoch": 1.4906899149714778,
-      "grad_norm": 0.5492138862609863,
-      "learning_rate": 0.0005111345760155156,
-      "loss": 3.7261,
+      "epoch": 1.4932614555256065,
+      "grad_norm": 0.6223024129867554,
+      "learning_rate": 0.0005109811117107393,
+      "loss": 3.6986,
       "step": 13850
     },
     {
-      "epoch": 1.496071467011086,
-      "grad_norm": 0.5724121928215027,
-      "learning_rate": 0.0005108113349854541,
+      "epoch": 1.4986522911051212,
+      "grad_norm": 0.5749328136444092,
+      "learning_rate": 0.0005106573124662708,
       "loss": 3.7067,
       "step": 13900
     },
     {
-      "epoch": 1.501453019050694,
-      "grad_norm": 0.5634517669677734,
-      "learning_rate": 0.0005104880939553926,
-      "loss": 3.6915,
+      "epoch": 1.5040431266846361,
+      "grad_norm": 0.5776667594909668,
+      "learning_rate": 0.0005103335132218025,
+      "loss": 3.6966,
       "step": 13950
     },
     {
-      "epoch": 1.5068345710903024,
-      "grad_norm": 0.6407066583633423,
-      "learning_rate": 0.0005101648529253313,
-      "loss": 3.7181,
+      "epoch": 1.509433962264151,
+      "grad_norm": 0.535430371761322,
+      "learning_rate": 0.000510009713977334,
+      "loss": 3.6959,
       "step": 14000
     },
     {
-      "epoch": 1.5068345710903024,
-      "eval_accuracy": 0.35367851414133633,
-      "eval_loss": 3.6570589542388916,
-      "eval_runtime": 182.1403,
-      "eval_samples_per_second": 98.885,
-      "eval_steps_per_second": 6.182,
+      "epoch": 1.509433962264151,
+      "eval_accuracy": 0.35341481358261057,
+      "eval_loss": 3.656437397003174,
+      "eval_runtime": 184.4166,
+      "eval_samples_per_second": 97.665,
+      "eval_steps_per_second": 6.106,
       "step": 14000
     },
     {
-      "epoch": 1.5122161231299107,
-      "grad_norm": 0.5960928797721863,
-      "learning_rate": 0.0005098416118952699,
-      "loss": 3.6972,
+      "epoch": 1.5148247978436657,
+      "grad_norm": 0.5845943689346313,
+      "learning_rate": 0.0005096859147328656,
+      "loss": 3.6939,
       "step": 14050
     },
     {
-      "epoch": 1.5175976751695188,
-      "grad_norm": 0.5764510035514832,
-      "learning_rate": 0.0005095183708652085,
-      "loss": 3.7012,
+      "epoch": 1.5202156334231804,
+      "grad_norm": 0.6200579404830933,
+      "learning_rate": 0.0005093621154883971,
+      "loss": 3.693,
       "step": 14100
     },
     {
-      "epoch": 1.5229792272091272,
-      "grad_norm": 0.5632539391517639,
-      "learning_rate": 0.000509195129835147,
-      "loss": 3.7139,
+      "epoch": 1.5256064690026954,
+      "grad_norm": 0.551781415939331,
+      "learning_rate": 0.0005090383162439288,
+      "loss": 3.6861,
       "step": 14150
     },
     {
-      "epoch": 1.5283607792487355,
-      "grad_norm": 0.5743679404258728,
-      "learning_rate": 0.0005088718888050856,
-      "loss": 3.6973,
+      "epoch": 1.5309973045822103,
+      "grad_norm": 0.7048491835594177,
+      "learning_rate": 0.0005087145169994602,
+      "loss": 3.6863,
       "step": 14200
     },
     {
-      "epoch": 1.5337423312883436,
-      "grad_norm": 0.5791763663291931,
-      "learning_rate": 0.0005085486477750242,
-      "loss": 3.7127,
+      "epoch": 1.536388140161725,
+      "grad_norm": 0.5546297430992126,
+      "learning_rate": 0.0005083907177549918,
+      "loss": 3.7031,
       "step": 14250
     },
     {
-      "epoch": 1.5391238833279517,
-      "grad_norm": 0.5912649631500244,
-      "learning_rate": 0.0005082254067449629,
-      "loss": 3.6992,
+      "epoch": 1.54177897574124,
+      "grad_norm": 0.5601275563240051,
+      "learning_rate": 0.0005080669185105234,
+      "loss": 3.6996,
       "step": 14300
     },
     {
-      "epoch": 1.54450543536756,
-      "grad_norm": 0.5650004744529724,
-      "learning_rate": 0.0005079021657149014,
-      "loss": 3.7003,
+      "epoch": 1.5471698113207548,
+      "grad_norm": 0.5481005907058716,
+      "learning_rate": 0.000507743119266055,
+      "loss": 3.6938,
       "step": 14350
     },
     {
-      "epoch": 1.5498869874071683,
-      "grad_norm": 0.5783352851867676,
-      "learning_rate": 0.0005075789246848399,
-      "loss": 3.7108,
+      "epoch": 1.5525606469002695,
+      "grad_norm": 0.561043918132782,
+      "learning_rate": 0.0005074193200215865,
+      "loss": 3.7087,
       "step": 14400
     },
     {
-      "epoch": 1.5552685394467765,
-      "grad_norm": 0.6116908192634583,
-      "learning_rate": 0.0005072556836547785,
-      "loss": 3.6884,
+      "epoch": 1.5579514824797842,
+      "grad_norm": 0.6112300753593445,
+      "learning_rate": 0.0005070955207771181,
+      "loss": 3.7007,
       "step": 14450
     },
     {
-      "epoch": 1.5606500914863846,
-      "grad_norm": 0.5558695197105408,
-      "learning_rate": 0.0005069324426247171,
-      "loss": 3.6849,
+      "epoch": 1.5633423180592994,
+      "grad_norm": 0.6545599102973938,
+      "learning_rate": 0.0005067717215326498,
+      "loss": 3.6813,
       "step": 14500
     },
     {
-      "epoch": 1.566031643525993,
-      "grad_norm": 0.6027122735977173,
-      "learning_rate": 0.0005066092015946557,
-      "loss": 3.6764,
+      "epoch": 1.568733153638814,
+      "grad_norm": 0.5471981763839722,
+      "learning_rate": 0.0005064479222881813,
+      "loss": 3.697,
       "step": 14550
     },
     {
-      "epoch": 1.5714131955656012,
-      "grad_norm": 0.6431997418403625,
-      "learning_rate": 0.0005062859605645943,
-      "loss": 3.7126,
+      "epoch": 1.5741239892183287,
+      "grad_norm": 0.6046043038368225,
+      "learning_rate": 0.0005061305990286023,
+      "loss": 3.6834,
       "step": 14600
     },
     {
-      "epoch": 1.5767947476052093,
-      "grad_norm": 0.548612117767334,
-      "learning_rate": 0.0005059627195345329,
-      "loss": 3.6901,
+      "epoch": 1.5795148247978437,
+      "grad_norm": 0.5650051832199097,
+      "learning_rate": 0.0005058067997841338,
+      "loss": 3.6803,
       "step": 14650
     },
     {
-      "epoch": 1.5821762996448174,
-      "grad_norm": 0.6005812883377075,
-      "learning_rate": 0.0005056394785044715,
-      "loss": 3.695,
+      "epoch": 1.5849056603773586,
+      "grad_norm": 0.5594826936721802,
+      "learning_rate": 0.0005054830005396654,
+      "loss": 3.6965,
       "step": 14700
     },
     {
-      "epoch": 1.5875578516844258,
-      "grad_norm": 0.6675664782524109,
-      "learning_rate": 0.00050531623747441,
-      "loss": 3.6855,
+      "epoch": 1.5902964959568733,
+      "grad_norm": 0.5203337669372559,
+      "learning_rate": 0.0005051592012951969,
+      "loss": 3.6953,
       "step": 14750
     },
     {
-      "epoch": 1.592939403724034,
-      "grad_norm": 0.6004554033279419,
-      "learning_rate": 0.0005049929964443486,
-      "loss": 3.6797,
+      "epoch": 1.595687331536388,
+      "grad_norm": 0.5579484105110168,
+      "learning_rate": 0.0005048354020507286,
+      "loss": 3.6674,
       "step": 14800
     },
     {
-      "epoch": 1.5983209557636422,
-      "grad_norm": 0.578523576259613,
-      "learning_rate": 0.0005046697554142871,
-      "loss": 3.6766,
+      "epoch": 1.6010781671159031,
+      "grad_norm": 0.5746201276779175,
+      "learning_rate": 0.00050451160280626,
+      "loss": 3.6882,
       "step": 14850
     },
     {
-      "epoch": 1.6037025078032503,
-      "grad_norm": 0.5477756857872009,
-      "learning_rate": 0.0005043465143842258,
-      "loss": 3.6804,
+      "epoch": 1.6064690026954178,
+      "grad_norm": 0.6085051894187927,
+      "learning_rate": 0.0005041878035617917,
+      "loss": 3.6944,
       "step": 14900
     },
     {
-      "epoch": 1.6090840598428586,
-      "grad_norm": 0.5516514778137207,
-      "learning_rate": 0.0005040232733541644,
-      "loss": 3.6764,
+      "epoch": 1.6118598382749325,
+      "grad_norm": 0.5542673468589783,
+      "learning_rate": 0.0005038640043173232,
+      "loss": 3.6834,
       "step": 14950
     },
     {
-      "epoch": 1.614465611882467,
-      "grad_norm": 0.5752219557762146,
-      "learning_rate": 0.000503700032324103,
-      "loss": 3.6986,
+      "epoch": 1.6172506738544474,
+      "grad_norm": 0.5814019441604614,
+      "learning_rate": 0.0005035402050728548,
+      "loss": 3.6836,
       "step": 15000
     },
     {
-      "epoch": 1.614465611882467,
-      "eval_accuracy": 0.3556933732929681,
-      "eval_loss": 3.6386351585388184,
-      "eval_runtime": 181.8961,
-      "eval_samples_per_second": 99.018,
-      "eval_steps_per_second": 6.19,
+      "epoch": 1.6172506738544474,
+      "eval_accuracy": 0.35548399917814955,
+      "eval_loss": 3.638230085372925,
+      "eval_runtime": 184.35,
+      "eval_samples_per_second": 97.7,
+      "eval_steps_per_second": 6.108,
       "step": 15000
     },
     {
-      "epoch": 1.619847163922075,
-      "grad_norm": 0.5473290085792542,
-      "learning_rate": 0.0005033832561146428,
-      "loss": 3.6819,
+      "epoch": 1.6226415094339623,
+      "grad_norm": 0.5265723466873169,
+      "learning_rate": 0.0005032164058283863,
+      "loss": 3.6749,
       "step": 15050
     },
     {
-      "epoch": 1.6252287159616834,
-      "grad_norm": 0.5905124545097351,
-      "learning_rate": 0.0005030600150845813,
-      "loss": 3.672,
+      "epoch": 1.628032345013477,
+      "grad_norm": 0.5524933934211731,
+      "learning_rate": 0.0005028926065839179,
+      "loss": 3.6688,
       "step": 15100
     },
     {
-      "epoch": 1.6306102680012917,
-      "grad_norm": 0.6535428762435913,
-      "learning_rate": 0.00050273677405452,
-      "loss": 3.6776,
+      "epoch": 1.633423180592992,
+      "grad_norm": 0.5435217022895813,
+      "learning_rate": 0.0005025688073394495,
+      "loss": 3.6679,
       "step": 15150
     },
     {
-      "epoch": 1.6359918200408998,
-      "grad_norm": 0.5464186072349548,
-      "learning_rate": 0.0005024135330244585,
-      "loss": 3.6604,
+      "epoch": 1.6388140161725069,
+      "grad_norm": 0.5551233291625977,
+      "learning_rate": 0.0005022450080949811,
+      "loss": 3.6717,
       "step": 15200
     },
     {
-      "epoch": 1.641373372080508,
-      "grad_norm": 0.5981823801994324,
-      "learning_rate": 0.0005020902919943972,
-      "loss": 3.6632,
+      "epoch": 1.6442048517520216,
+      "grad_norm": 0.5736522078514099,
+      "learning_rate": 0.0005019212088505126,
+      "loss": 3.6661,
       "step": 15250
     },
     {
-      "epoch": 1.6467549241201163,
-      "grad_norm": 0.5628587603569031,
-      "learning_rate": 0.0005017670509643357,
-      "loss": 3.6775,
+      "epoch": 1.6495956873315363,
+      "grad_norm": 0.5549569129943848,
+      "learning_rate": 0.0005015974096060442,
+      "loss": 3.6711,
       "step": 15300
     },
     {
-      "epoch": 1.6521364761597246,
-      "grad_norm": 0.5168464183807373,
-      "learning_rate": 0.0005014438099342743,
-      "loss": 3.6883,
+      "epoch": 1.6549865229110512,
+      "grad_norm": 0.5485915541648865,
+      "learning_rate": 0.0005012736103615758,
+      "loss": 3.6962,
       "step": 15350
     },
     {
-      "epoch": 1.6575180281993327,
-      "grad_norm": 0.5790534019470215,
-      "learning_rate": 0.0005011205689042129,
-      "loss": 3.6784,
+      "epoch": 1.6603773584905661,
+      "grad_norm": 0.6051048636436462,
+      "learning_rate": 0.0005009498111171074,
+      "loss": 3.6853,
       "step": 15400
     },
     {
-      "epoch": 1.6628995802389408,
-      "grad_norm": 0.5653476119041443,
-      "learning_rate": 0.0005007973278741514,
-      "loss": 3.6495,
+      "epoch": 1.6657681940700808,
+      "grad_norm": 0.5809049606323242,
+      "learning_rate": 0.0005006260118726389,
+      "loss": 3.6509,
       "step": 15450
     },
     {
-      "epoch": 1.6682811322785491,
-      "grad_norm": 0.5761940479278564,
-      "learning_rate": 0.00050047408684409,
-      "loss": 3.6852,
+      "epoch": 1.6711590296495957,
+      "grad_norm": 0.6438100934028625,
+      "learning_rate": 0.0005003022126281705,
+      "loss": 3.6689,
       "step": 15500
     },
     {
-      "epoch": 1.6736626843181575,
-      "grad_norm": 0.5976137518882751,
-      "learning_rate": 0.0005001508458140286,
-      "loss": 3.6591,
+      "epoch": 1.6765498652291106,
+      "grad_norm": 0.5837825536727905,
+      "learning_rate": 0.000499978413383702,
+      "loss": 3.6674,
       "step": 15550
     },
     {
-      "epoch": 1.6790442363577656,
-      "grad_norm": 0.5772688388824463,
-      "learning_rate": 0.0004998276047839673,
-      "loss": 3.654,
+      "epoch": 1.6819407008086253,
+      "grad_norm": 0.5168622732162476,
+      "learning_rate": 0.0004996546141392336,
+      "loss": 3.6546,
       "step": 15600
     },
     {
-      "epoch": 1.6844257883973737,
-      "grad_norm": 0.5581541061401367,
-      "learning_rate": 0.0004995043637539058,
-      "loss": 3.6748,
+      "epoch": 1.68733153638814,
+      "grad_norm": 0.5905857682228088,
+      "learning_rate": 0.0004993308148947651,
+      "loss": 3.6717,
       "step": 15650
     },
     {
-      "epoch": 1.689807340436982,
-      "grad_norm": 0.5584582090377808,
-      "learning_rate": 0.0004991811227238443,
-      "loss": 3.666,
+      "epoch": 1.692722371967655,
+      "grad_norm": 0.6183093190193176,
+      "learning_rate": 0.0004990070156502968,
+      "loss": 3.6614,
       "step": 15700
     },
     {
-      "epoch": 1.6951888924765903,
-      "grad_norm": 0.6056560277938843,
-      "learning_rate": 0.0004988578816937829,
-      "loss": 3.656,
+      "epoch": 1.6981132075471699,
+      "grad_norm": 0.6086243391036987,
+      "learning_rate": 0.0004986832164058284,
+      "loss": 3.6868,
       "step": 15750
     },
     {
-      "epoch": 1.7005704445161984,
-      "grad_norm": 0.5387526750564575,
-      "learning_rate": 0.0004985346406637215,
-      "loss": 3.6844,
+      "epoch": 1.7035040431266846,
+      "grad_norm": 0.521761417388916,
+      "learning_rate": 0.0004983594171613599,
+      "loss": 3.681,
       "step": 15800
     },
     {
-      "epoch": 1.7059519965558065,
-      "grad_norm": 0.6060013175010681,
-      "learning_rate": 0.0004982113996336602,
-      "loss": 3.6704,
+      "epoch": 1.7088948787061995,
+      "grad_norm": 0.5507834553718567,
+      "learning_rate": 0.0004980356179168915,
+      "loss": 3.6756,
       "step": 15850
     },
     {
-      "epoch": 1.7113335485954149,
-      "grad_norm": 0.5558328628540039,
-      "learning_rate": 0.0004978881586035987,
-      "loss": 3.6695,
+      "epoch": 1.7142857142857144,
+      "grad_norm": 0.614310085773468,
+      "learning_rate": 0.000497711818672423,
+      "loss": 3.6768,
       "step": 15900
     },
     {
-      "epoch": 1.7167151006350232,
-      "grad_norm": 0.6431989073753357,
-      "learning_rate": 0.0004975649175735373,
-      "loss": 3.6561,
+      "epoch": 1.719676549865229,
+      "grad_norm": 0.5719003677368164,
+      "learning_rate": 0.0004973880194279547,
+      "loss": 3.6592,
       "step": 15950
     },
     {
-      "epoch": 1.7220966526746313,
-      "grad_norm": 0.5722593069076538,
-      "learning_rate": 0.0004972416765434759,
-      "loss": 3.6594,
+      "epoch": 1.7250673854447438,
+      "grad_norm": 0.5983813405036926,
+      "learning_rate": 0.0004970642201834862,
+      "loss": 3.6589,
       "step": 16000
     },
     {
-      "epoch": 1.7220966526746313,
-      "eval_accuracy": 0.35697623793939365,
-      "eval_loss": 3.6226398944854736,
-      "eval_runtime": 182.9771,
-      "eval_samples_per_second": 98.433,
-      "eval_steps_per_second": 6.154,
+      "epoch": 1.7250673854447438,
+      "eval_accuracy": 0.3571446499155061,
+      "eval_loss": 3.620162010192871,
+      "eval_runtime": 184.4759,
+      "eval_samples_per_second": 97.633,
+      "eval_steps_per_second": 6.104,
       "step": 16000
     },
     {
-      "epoch": 1.7274782047142396,
-      "grad_norm": 0.5558207035064697,
-      "learning_rate": 0.0004969184355134145,
-      "loss": 3.6741,
+      "epoch": 1.7304582210242587,
+      "grad_norm": 0.5706288814544678,
+      "learning_rate": 0.0004967404209390178,
+      "loss": 3.6609,
       "step": 16050
     },
     {
-      "epoch": 1.732859756753848,
-      "grad_norm": 0.6220034956932068,
-      "learning_rate": 0.0004965951944833531,
-      "loss": 3.6571,
+      "epoch": 1.7358490566037736,
+      "grad_norm": 0.5533764362335205,
+      "learning_rate": 0.0004964166216945493,
+      "loss": 3.6461,
       "step": 16100
     },
     {
-      "epoch": 1.738241308793456,
-      "grad_norm": 0.5504184365272522,
-      "learning_rate": 0.0004962719534532916,
-      "loss": 3.6812,
+      "epoch": 1.7412398921832883,
+      "grad_norm": 0.6296628713607788,
+      "learning_rate": 0.000496092822450081,
+      "loss": 3.669,
       "step": 16150
     },
     {
-      "epoch": 1.7436228608330642,
-      "grad_norm": 0.6175248622894287,
-      "learning_rate": 0.0004959487124232302,
-      "loss": 3.6989,
+      "epoch": 1.7466307277628033,
+      "grad_norm": 0.5582634806632996,
+      "learning_rate": 0.0004957690232056125,
+      "loss": 3.6814,
       "step": 16200
     },
     {
-      "epoch": 1.7490044128726725,
-      "grad_norm": 0.6186313629150391,
-      "learning_rate": 0.0004956254713931688,
-      "loss": 3.6716,
+      "epoch": 1.7520215633423182,
+      "grad_norm": 0.5720065236091614,
+      "learning_rate": 0.0004954452239611441,
+      "loss": 3.6592,
       "step": 16250
     },
     {
-      "epoch": 1.7543859649122808,
-      "grad_norm": 0.5942351818084717,
-      "learning_rate": 0.0004953022303631074,
-      "loss": 3.6773,
+      "epoch": 1.7574123989218329,
+      "grad_norm": 0.5711783766746521,
+      "learning_rate": 0.0004951214247166756,
+      "loss": 3.6785,
       "step": 16300
     },
     {
-      "epoch": 1.759767516951889,
-      "grad_norm": 0.5420483350753784,
-      "learning_rate": 0.0004949789893330459,
-      "loss": 3.6632,
+      "epoch": 1.7628032345013476,
+      "grad_norm": 0.6209591627120972,
+      "learning_rate": 0.0004947976254722072,
+      "loss": 3.6627,
       "step": 16350
     },
     {
-      "epoch": 1.765149068991497,
-      "grad_norm": 0.5678284764289856,
-      "learning_rate": 0.0004946557483029846,
-      "loss": 3.6603,
+      "epoch": 1.7681940700808625,
+      "grad_norm": 0.5400413274765015,
+      "learning_rate": 0.0004944738262277387,
+      "loss": 3.6526,
       "step": 16400
     },
     {
-      "epoch": 1.7705306210311054,
-      "grad_norm": 0.5654410719871521,
-      "learning_rate": 0.0004943325072729231,
-      "loss": 3.6556,
+      "epoch": 1.7735849056603774,
+      "grad_norm": 0.518601655960083,
+      "learning_rate": 0.0004941500269832703,
+      "loss": 3.659,
       "step": 16450
     },
     {
-      "epoch": 1.7759121730707137,
-      "grad_norm": 0.5799243450164795,
-      "learning_rate": 0.0004940092662428617,
-      "loss": 3.6594,
+      "epoch": 1.778975741239892,
+      "grad_norm": 0.5466724634170532,
+      "learning_rate": 0.0004938262277388019,
+      "loss": 3.6465,
       "step": 16500
     },
     {
-      "epoch": 1.7812937251103218,
-      "grad_norm": 0.5611895322799683,
-      "learning_rate": 0.0004936860252128003,
-      "loss": 3.6558,
+      "epoch": 1.784366576819407,
+      "grad_norm": 0.6133984923362732,
+      "learning_rate": 0.0004935024284943335,
+      "loss": 3.6745,
       "step": 16550
     },
     {
-      "epoch": 1.78667527714993,
-      "grad_norm": 0.6114226579666138,
-      "learning_rate": 0.0004933627841827388,
-      "loss": 3.6482,
+      "epoch": 1.789757412398922,
+      "grad_norm": 0.6439416408538818,
+      "learning_rate": 0.000493178629249865,
+      "loss": 3.6593,
       "step": 16600
     },
     {
-      "epoch": 1.7920568291895382,
-      "grad_norm": 0.5653012990951538,
-      "learning_rate": 0.0004930395431526775,
-      "loss": 3.6621,
+      "epoch": 1.7951482479784366,
+      "grad_norm": 0.5581780076026917,
+      "learning_rate": 0.000492861305990286,
+      "loss": 3.6606,
       "step": 16650
     },
     {
-      "epoch": 1.7974383812291466,
-      "grad_norm": 0.5843732357025146,
-      "learning_rate": 0.0004927163021226161,
-      "loss": 3.6607,
+      "epoch": 1.8005390835579513,
+      "grad_norm": 0.5625990033149719,
+      "learning_rate": 0.0004925375067458175,
+      "loss": 3.6649,
       "step": 16700
     },
     {
-      "epoch": 1.8028199332687547,
-      "grad_norm": 0.622482180595398,
-      "learning_rate": 0.0004923930610925547,
-      "loss": 3.6551,
+      "epoch": 1.8059299191374663,
+      "grad_norm": 0.5239773988723755,
+      "learning_rate": 0.0004922137075013491,
+      "loss": 3.6511,
       "step": 16750
     },
     {
-      "epoch": 1.8082014853083628,
-      "grad_norm": 0.5526164770126343,
-      "learning_rate": 0.0004920698200624932,
-      "loss": 3.6533,
+      "epoch": 1.8113207547169812,
+      "grad_norm": 0.5561639070510864,
+      "learning_rate": 0.0004918899082568807,
+      "loss": 3.6528,
       "step": 16800
     },
     {
-      "epoch": 1.813583037347971,
-      "grad_norm": 0.6223166584968567,
-      "learning_rate": 0.0004917465790324317,
-      "loss": 3.6685,
+      "epoch": 1.8167115902964959,
+      "grad_norm": 0.6131348013877869,
+      "learning_rate": 0.0004915661090124123,
+      "loss": 3.6586,
       "step": 16850
     },
     {
-      "epoch": 1.8189645893875794,
-      "grad_norm": 0.6319237947463989,
-      "learning_rate": 0.0004914233380023704,
-      "loss": 3.6679,
+      "epoch": 1.8221024258760108,
+      "grad_norm": 0.5639000535011292,
+      "learning_rate": 0.0004912487857528331,
+      "loss": 3.6487,
       "step": 16900
     },
     {
-      "epoch": 1.8243461414271875,
-      "grad_norm": 0.5310279726982117,
-      "learning_rate": 0.0004911000969723089,
-      "loss": 3.6702,
+      "epoch": 1.8274932614555257,
+      "grad_norm": 0.5571854114532471,
+      "learning_rate": 0.0004909249865083648,
+      "loss": 3.6724,
       "step": 16950
     },
     {
-      "epoch": 1.8297276934667959,
-      "grad_norm": 0.5771204233169556,
-      "learning_rate": 0.0004907768559422476,
-      "loss": 3.6535,
+      "epoch": 1.8328840970350404,
+      "grad_norm": 0.5779969692230225,
+      "learning_rate": 0.0004906011872638964,
+      "loss": 3.6613,
       "step": 17000
     },
     {
-      "epoch": 1.8297276934667959,
-      "eval_accuracy": 0.3584541345194462,
-      "eval_loss": 3.609053373336792,
-      "eval_runtime": 183.1048,
-      "eval_samples_per_second": 98.364,
-      "eval_steps_per_second": 6.149,
+      "epoch": 1.8328840970350404,
+      "eval_accuracy": 0.3583969831004558,
+      "eval_loss": 3.6074378490448,
+      "eval_runtime": 184.224,
+      "eval_samples_per_second": 97.767,
+      "eval_steps_per_second": 6.112,
       "step": 17000
     },
     {
-      "epoch": 1.8351092455064042,
-      "grad_norm": 0.6009894609451294,
-      "learning_rate": 0.0004904536149121861,
-      "loss": 3.6407,
+      "epoch": 1.838274932614555,
+      "grad_norm": 0.5520243048667908,
+      "learning_rate": 0.0004902773880194279,
+      "loss": 3.6411,
       "step": 17050
     },
     {
-      "epoch": 1.8404907975460123,
-      "grad_norm": 0.5864176750183105,
-      "learning_rate": 0.0004901303738821248,
-      "loss": 3.6577,
+      "epoch": 1.8436657681940702,
+      "grad_norm": 0.5441498756408691,
+      "learning_rate": 0.0004899535887749595,
+      "loss": 3.6552,
       "step": 17100
     },
     {
-      "epoch": 1.8458723495856204,
-      "grad_norm": 0.5710988640785217,
-      "learning_rate": 0.0004898135976726646,
-      "loss": 3.6456,
+      "epoch": 1.849056603773585,
+      "grad_norm": 0.5663284659385681,
+      "learning_rate": 0.000489629789530491,
+      "loss": 3.6513,
       "step": 17150
     },
     {
-      "epoch": 1.8512539016252287,
-      "grad_norm": 0.6018273830413818,
-      "learning_rate": 0.0004894903566426031,
-      "loss": 3.6359,
+      "epoch": 1.8544474393530996,
+      "grad_norm": 0.5562971234321594,
+      "learning_rate": 0.0004893059902860227,
+      "loss": 3.6358,
       "step": 17200
     },
     {
-      "epoch": 1.856635453664837,
-      "grad_norm": 0.613945484161377,
-      "learning_rate": 0.0004891671156125417,
-      "loss": 3.6632,
+      "epoch": 1.8598382749326146,
+      "grad_norm": 0.5903058052062988,
+      "learning_rate": 0.0004889821910415542,
+      "loss": 3.6564,
       "step": 17250
     },
     {
-      "epoch": 1.8620170057044452,
-      "grad_norm": 0.5496900677680969,
-      "learning_rate": 0.0004888438745824803,
-      "loss": 3.6434,
+      "epoch": 1.8652291105121295,
+      "grad_norm": 0.6454324722290039,
+      "learning_rate": 0.0004886583917970858,
+      "loss": 3.6482,
       "step": 17300
     },
     {
-      "epoch": 1.8673985577440533,
-      "grad_norm": 0.5686019659042358,
-      "learning_rate": 0.0004885206335524189,
-      "loss": 3.6718,
+      "epoch": 1.8706199460916442,
+      "grad_norm": 0.6871253848075867,
+      "learning_rate": 0.0004883345925526173,
+      "loss": 3.6511,
       "step": 17350
     },
     {
-      "epoch": 1.8727801097836616,
-      "grad_norm": 0.5641106963157654,
-      "learning_rate": 0.0004881973925223575,
-      "loss": 3.6485,
+      "epoch": 1.8760107816711589,
+      "grad_norm": 0.5943225622177124,
+      "learning_rate": 0.00048801079330814887,
+      "loss": 3.6456,
       "step": 17400
     },
     {
-      "epoch": 1.87816166182327,
-      "grad_norm": 0.5751502513885498,
-      "learning_rate": 0.00048787415149229604,
-      "loss": 3.6619,
+      "epoch": 1.881401617250674,
+      "grad_norm": 0.5526453256607056,
+      "learning_rate": 0.00048768699406368047,
+      "loss": 3.6625,
       "step": 17450
     },
     {
-      "epoch": 1.883543213862878,
-      "grad_norm": 0.5386393666267395,
-      "learning_rate": 0.00048755091046223464,
-      "loss": 3.6627,
+      "epoch": 1.8867924528301887,
+      "grad_norm": 0.5912424921989441,
+      "learning_rate": 0.000487363194819212,
+      "loss": 3.6569,
       "step": 17500
     },
     {
-      "epoch": 1.8889247659024861,
-      "grad_norm": 0.5446299910545349,
-      "learning_rate": 0.0004872276694321732,
-      "loss": 3.6486,
+      "epoch": 1.8921832884097034,
+      "grad_norm": 0.5729259252548218,
+      "learning_rate": 0.0004870393955747436,
+      "loss": 3.6367,
       "step": 17550
     },
     {
-      "epoch": 1.8943063179420945,
-      "grad_norm": 0.5315663814544678,
-      "learning_rate": 0.00048690442840211177,
-      "loss": 3.6565,
+      "epoch": 1.8975741239892183,
+      "grad_norm": 0.5813829898834229,
+      "learning_rate": 0.0004867155963302752,
+      "loss": 3.6454,
       "step": 17600
     },
     {
-      "epoch": 1.8996878699817028,
-      "grad_norm": 0.5348063111305237,
-      "learning_rate": 0.0004865811873720504,
-      "loss": 3.6573,
+      "epoch": 1.9029649595687332,
+      "grad_norm": 0.5631664395332336,
+      "learning_rate": 0.0004863917970858068,
+      "loss": 3.6228,
       "step": 17650
     },
     {
-      "epoch": 1.905069422021311,
-      "grad_norm": 0.5544908046722412,
-      "learning_rate": 0.00048625794634198896,
-      "loss": 3.6473,
+      "epoch": 1.908355795148248,
+      "grad_norm": 0.5724475383758545,
+      "learning_rate": 0.00048606799784133833,
+      "loss": 3.6327,
       "step": 17700
     },
     {
-      "epoch": 1.910450974060919,
-      "grad_norm": 0.5778284668922424,
-      "learning_rate": 0.00048593470531192756,
-      "loss": 3.6594,
+      "epoch": 1.9137466307277629,
+      "grad_norm": 0.6105817556381226,
+      "learning_rate": 0.00048574419859686994,
+      "loss": 3.6354,
       "step": 17750
     },
     {
-      "epoch": 1.9158325261005273,
-      "grad_norm": 0.566500723361969,
-      "learning_rate": 0.00048561146428186615,
-      "loss": 3.6411,
+      "epoch": 1.9191374663072778,
+      "grad_norm": 0.5378417372703552,
+      "learning_rate": 0.0004854203993524015,
+      "loss": 3.6547,
       "step": 17800
     },
     {
-      "epoch": 1.9212140781401357,
-      "grad_norm": 0.5555658340454102,
-      "learning_rate": 0.0004852882232518047,
-      "loss": 3.6261,
+      "epoch": 1.9245283018867925,
+      "grad_norm": 0.571906566619873,
+      "learning_rate": 0.000485096600107933,
+      "loss": 3.6532,
       "step": 17850
     },
     {
-      "epoch": 1.9265956301797438,
-      "grad_norm": 0.5867842435836792,
-      "learning_rate": 0.00048496498222174334,
-      "loss": 3.6291,
+      "epoch": 1.9299191374663072,
+      "grad_norm": 0.5694820284843445,
+      "learning_rate": 0.00048477280086346464,
+      "loss": 3.6608,
       "step": 17900
     },
     {
-      "epoch": 1.931977182219352,
-      "grad_norm": 0.5909476280212402,
-      "learning_rate": 0.00048464174119168193,
-      "loss": 3.6282,
+      "epoch": 1.935309973045822,
+      "grad_norm": 0.5970518589019775,
+      "learning_rate": 0.00048444900161899614,
+      "loss": 3.6495,
       "step": 17950
     },
     {
-      "epoch": 1.9373587342589604,
-      "grad_norm": 0.642209529876709,
-      "learning_rate": 0.0004843185001616205,
-      "loss": 3.6528,
+      "epoch": 1.940700808625337,
+      "grad_norm": 0.5703791975975037,
+      "learning_rate": 0.00048412520237452774,
+      "loss": 3.6366,
       "step": 18000
     },
     {
-      "epoch": 1.9373587342589604,
-      "eval_accuracy": 0.3601386888921219,
-      "eval_loss": 3.5923776626586914,
-      "eval_runtime": 182.6986,
-      "eval_samples_per_second": 98.583,
-      "eval_steps_per_second": 6.163,
+      "epoch": 1.940700808625337,
+      "eval_accuracy": 0.3597993659234773,
+      "eval_loss": 3.5908522605895996,
+      "eval_runtime": 184.0844,
+      "eval_samples_per_second": 97.841,
+      "eval_steps_per_second": 6.117,
       "step": 18000
     },
     {
-      "epoch": 1.9427402862985685,
-      "grad_norm": 0.5417137742042542,
-      "learning_rate": 0.00048399525913155907,
-      "loss": 3.6509,
+      "epoch": 1.9460916442048517,
+      "grad_norm": 0.5802140235900879,
+      "learning_rate": 0.0004838014031300593,
+      "loss": 3.6241,
       "step": 18050
     },
     {
-      "epoch": 1.9481218383381766,
-      "grad_norm": 0.5468534231185913,
-      "learning_rate": 0.0004836720181014976,
-      "loss": 3.6373,
+      "epoch": 1.9514824797843666,
+      "grad_norm": 0.5602383613586426,
+      "learning_rate": 0.0004834776038855909,
+      "loss": 3.6331,
       "step": 18100
     },
     {
-      "epoch": 1.953503390377785,
-      "grad_norm": 0.5604886412620544,
-      "learning_rate": 0.0004833487770714362,
-      "loss": 3.6485,
+      "epoch": 1.9568733153638815,
+      "grad_norm": 0.575650691986084,
+      "learning_rate": 0.00048315380464112245,
+      "loss": 3.6402,
       "step": 18150
     },
     {
-      "epoch": 1.9588849424173933,
-      "grad_norm": 0.5776407718658447,
-      "learning_rate": 0.00048302553604137485,
-      "loss": 3.6288,
+      "epoch": 1.9622641509433962,
+      "grad_norm": 0.5547229647636414,
+      "learning_rate": 0.00048283000539665405,
+      "loss": 3.6414,
       "step": 18200
     },
     {
-      "epoch": 1.9642664944570014,
-      "grad_norm": 0.6095754504203796,
-      "learning_rate": 0.0004827022950113134,
-      "loss": 3.6571,
+      "epoch": 1.967654986522911,
+      "grad_norm": 0.5848639607429504,
+      "learning_rate": 0.0004825062061521856,
+      "loss": 3.6316,
       "step": 18250
     },
     {
-      "epoch": 1.9696480464966095,
-      "grad_norm": 0.5740944147109985,
-      "learning_rate": 0.000482379053981252,
-      "loss": 3.6374,
+      "epoch": 1.9730458221024259,
+      "grad_norm": 0.5338342189788818,
+      "learning_rate": 0.00048218240690771716,
+      "loss": 3.6579,
       "step": 18300
     },
     {
-      "epoch": 1.9750295985362178,
-      "grad_norm": 0.5488967299461365,
-      "learning_rate": 0.0004820558129511906,
-      "loss": 3.6425,
+      "epoch": 1.9784366576819408,
+      "grad_norm": 0.5735607743263245,
+      "learning_rate": 0.00048185860766324876,
+      "loss": 3.6366,
       "step": 18350
     },
     {
-      "epoch": 1.9804111505758262,
-      "grad_norm": 0.5918818712234497,
-      "learning_rate": 0.0004817325719211291,
-      "loss": 3.6471,
+      "epoch": 1.9838274932614555,
+      "grad_norm": 0.5852369666099548,
+      "learning_rate": 0.0004815348084187803,
+      "loss": 3.6405,
       "step": 18400
     },
     {
-      "epoch": 1.9857927026154343,
-      "grad_norm": 0.5412439703941345,
-      "learning_rate": 0.0004814093308910677,
-      "loss": 3.6463,
+      "epoch": 1.9892183288409704,
+      "grad_norm": 0.5547637343406677,
+      "learning_rate": 0.0004812110091743119,
+      "loss": 3.6468,
       "step": 18450
     },
     {
-      "epoch": 1.9911742546550424,
-      "grad_norm": 0.5630549192428589,
-      "learning_rate": 0.00048108608986100637,
-      "loss": 3.6531,
+      "epoch": 1.9946091644204853,
+      "grad_norm": 0.5905916094779968,
+      "learning_rate": 0.00048088720992984347,
+      "loss": 3.6252,
       "step": 18500
     },
     {
-      "epoch": 1.9965558066946507,
-      "grad_norm": 0.5589700937271118,
-      "learning_rate": 0.0004807628488309449,
-      "loss": 3.65,
+      "epoch": 2.0,
+      "grad_norm": 1.1172149181365967,
+      "learning_rate": 0.00048056341068537507,
+      "loss": 3.6283,
       "step": 18550
     },
     {
-      "epoch": 2.001937358734259,
-      "grad_norm": 0.5413640141487122,
-      "learning_rate": 0.0004804396078008835,
-      "loss": 3.6016,
+      "epoch": 2.0053908355795147,
+      "grad_norm": 0.5794585943222046,
+      "learning_rate": 0.00048023961144090657,
+      "loss": 3.5245,
       "step": 18600
     },
     {
-      "epoch": 2.007318910773867,
-      "grad_norm": 0.5609384179115295,
-      "learning_rate": 0.00048011636677082204,
-      "loss": 3.5453,
+      "epoch": 2.01078167115903,
+      "grad_norm": 0.5609965920448303,
+      "learning_rate": 0.00047991581219643817,
+      "loss": 3.5384,
       "step": 18650
     },
     {
-      "epoch": 2.0127004628134753,
-      "grad_norm": 0.6163759827613831,
-      "learning_rate": 0.00047979312574076064,
-      "loss": 3.5521,
+      "epoch": 2.0161725067385445,
+      "grad_norm": 0.5413068532943726,
+      "learning_rate": 0.0004795920129519697,
+      "loss": 3.5363,
       "step": 18700
     },
     {
-      "epoch": 2.018082014853084,
-      "grad_norm": 0.5741806626319885,
-      "learning_rate": 0.0004794698847106992,
-      "loss": 3.5383,
+      "epoch": 2.0215633423180592,
+      "grad_norm": 0.5804000496864319,
+      "learning_rate": 0.0004792682137075013,
+      "loss": 3.5449,
       "step": 18750
     },
     {
-      "epoch": 2.023463566892692,
-      "grad_norm": 0.6155529022216797,
-      "learning_rate": 0.0004791466436806378,
-      "loss": 3.5698,
+      "epoch": 2.026954177897574,
+      "grad_norm": 0.6052377223968506,
+      "learning_rate": 0.0004789444144630329,
+      "loss": 3.5645,
       "step": 18800
     },
     {
-      "epoch": 2.0288451189323,
-      "grad_norm": 0.6185212731361389,
-      "learning_rate": 0.0004788234026505764,
-      "loss": 3.5489,
+      "epoch": 2.032345013477089,
+      "grad_norm": 0.5468222498893738,
+      "learning_rate": 0.00047862061521856443,
+      "loss": 3.5623,
       "step": 18850
     },
     {
-      "epoch": 2.034226670971908,
-      "grad_norm": 0.5259607434272766,
-      "learning_rate": 0.00047850016162051496,
-      "loss": 3.5425,
+      "epoch": 2.0377358490566038,
+      "grad_norm": 0.5067700743675232,
+      "learning_rate": 0.00047829681597409603,
+      "loss": 3.5447,
       "step": 18900
     },
     {
-      "epoch": 2.0396082230115167,
-      "grad_norm": 0.5763450264930725,
-      "learning_rate": 0.00047817692059045356,
-      "loss": 3.5663,
+      "epoch": 2.0431266846361185,
+      "grad_norm": 0.5769860148429871,
+      "learning_rate": 0.0004779730167296276,
+      "loss": 3.5476,
       "step": 18950
     },
     {
-      "epoch": 2.044989775051125,
-      "grad_norm": 0.5624204277992249,
-      "learning_rate": 0.00047785367956039215,
-      "loss": 3.5646,
+      "epoch": 2.0485175202156336,
+      "grad_norm": 0.578884482383728,
+      "learning_rate": 0.0004776492174851592,
+      "loss": 3.5806,
       "step": 19000
     },
     {
-      "epoch": 2.044989775051125,
-      "eval_accuracy": 0.36152814202149347,
-      "eval_loss": 3.5838420391082764,
-      "eval_runtime": 183.3409,
-      "eval_samples_per_second": 98.238,
-      "eval_steps_per_second": 6.142,
+      "epoch": 2.0485175202156336,
+      "eval_accuracy": 0.36154683031819757,
+      "eval_loss": 3.5805275440216064,
+      "eval_runtime": 184.2732,
+      "eval_samples_per_second": 97.741,
+      "eval_steps_per_second": 6.11,
       "step": 19000
     },
     {
-      "epoch": 2.050371327090733,
-      "grad_norm": 0.5626105666160583,
-      "learning_rate": 0.00047753043853033075,
-      "loss": 3.5464,
+      "epoch": 2.0539083557951483,
+      "grad_norm": 0.5976527333259583,
+      "learning_rate": 0.00047732541824069074,
+      "loss": 3.5558,
       "step": 19050
     },
     {
-      "epoch": 2.055752879130341,
-      "grad_norm": 0.5418869853019714,
-      "learning_rate": 0.00047720719750026934,
-      "loss": 3.5534,
+      "epoch": 2.059299191374663,
+      "grad_norm": 0.5781768560409546,
+      "learning_rate": 0.0004770016189962223,
+      "loss": 3.5496,
       "step": 19100
     },
     {
-      "epoch": 2.0611344311699495,
-      "grad_norm": 0.6087026000022888,
-      "learning_rate": 0.00047689042129080915,
-      "loss": 3.5573,
+      "epoch": 2.0646900269541777,
+      "grad_norm": 0.6147288084030151,
+      "learning_rate": 0.0004766778197517539,
+      "loss": 3.5623,
       "step": 19150
     },
     {
-      "epoch": 2.0665159832095576,
-      "grad_norm": 0.5950330495834351,
-      "learning_rate": 0.00047656718026074774,
-      "loss": 3.5528,
+      "epoch": 2.070080862533693,
+      "grad_norm": 0.5968753695487976,
+      "learning_rate": 0.0004763540205072854,
+      "loss": 3.576,
       "step": 19200
     },
     {
-      "epoch": 2.0718975352491658,
-      "grad_norm": 0.5562347769737244,
-      "learning_rate": 0.00047624393923068634,
-      "loss": 3.5703,
+      "epoch": 2.0754716981132075,
+      "grad_norm": 0.539870023727417,
+      "learning_rate": 0.00047603022126281705,
+      "loss": 3.5592,
       "step": 19250
     },
     {
-      "epoch": 2.0772790872887743,
-      "grad_norm": 0.5437284111976624,
-      "learning_rate": 0.0004759206982006249,
-      "loss": 3.5563,
+      "epoch": 2.0808625336927222,
+      "grad_norm": 0.591734766960144,
+      "learning_rate": 0.00047570642201834855,
+      "loss": 3.548,
       "step": 19300
     },
     {
-      "epoch": 2.0826606393283824,
-      "grad_norm": 0.6348011493682861,
-      "learning_rate": 0.00047559745717056347,
-      "loss": 3.5651,
+      "epoch": 2.0862533692722374,
+      "grad_norm": 0.6026017069816589,
+      "learning_rate": 0.00047538262277388015,
+      "loss": 3.5617,
       "step": 19350
     },
     {
-      "epoch": 2.0880421913679905,
-      "grad_norm": 0.5337033271789551,
-      "learning_rate": 0.000475274216140502,
-      "loss": 3.556,
+      "epoch": 2.091644204851752,
+      "grad_norm": 0.5985312461853027,
+      "learning_rate": 0.0004750588235294117,
+      "loss": 3.5408,
       "step": 19400
     },
     {
-      "epoch": 2.0934237434075986,
-      "grad_norm": 0.6107613444328308,
-      "learning_rate": 0.00047495097511044066,
-      "loss": 3.5584,
+      "epoch": 2.0970350404312668,
+      "grad_norm": 0.5874620676040649,
+      "learning_rate": 0.0004747350242849433,
+      "loss": 3.5589,
       "step": 19450
     },
     {
-      "epoch": 2.098805295447207,
-      "grad_norm": 0.6084286570549011,
-      "learning_rate": 0.00047462773408037925,
-      "loss": 3.5613,
+      "epoch": 2.1024258760107815,
+      "grad_norm": 0.588500440120697,
+      "learning_rate": 0.00047441122504047486,
+      "loss": 3.5593,
       "step": 19500
     },
     {
-      "epoch": 2.1041868474868153,
-      "grad_norm": 0.5908589959144592,
-      "learning_rate": 0.0004743044930503178,
-      "loss": 3.5732,
+      "epoch": 2.1078167115902966,
+      "grad_norm": 0.6489098072052002,
+      "learning_rate": 0.0004740874257960064,
+      "loss": 3.5476,
       "step": 19550
     },
     {
-      "epoch": 2.1095683995264234,
-      "grad_norm": 0.5735676884651184,
-      "learning_rate": 0.0004739812520202564,
-      "loss": 3.5512,
+      "epoch": 2.1132075471698113,
+      "grad_norm": 0.5802199244499207,
+      "learning_rate": 0.000473763626551538,
+      "loss": 3.5508,
       "step": 19600
     },
     {
-      "epoch": 2.1149499515660315,
-      "grad_norm": 0.5397831201553345,
-      "learning_rate": 0.000473658010990195,
-      "loss": 3.5579,
+      "epoch": 2.118598382749326,
+      "grad_norm": 0.591067373752594,
+      "learning_rate": 0.00047343982730706956,
+      "loss": 3.5588,
       "step": 19650
     },
     {
-      "epoch": 2.12033150360564,
-      "grad_norm": 0.6003082394599915,
-      "learning_rate": 0.0004733347699601336,
-      "loss": 3.5704,
+      "epoch": 2.123989218328841,
+      "grad_norm": 0.586726725101471,
+      "learning_rate": 0.00047311602806260117,
+      "loss": 3.561,
       "step": 19700
     },
     {
-      "epoch": 2.125713055645248,
-      "grad_norm": 0.6014171242713928,
-      "learning_rate": 0.0004730115289300722,
+      "epoch": 2.129380053908356,
+      "grad_norm": 0.5676229000091553,
+      "learning_rate": 0.0004727922288181327,
       "loss": 3.5688,
       "step": 19750
     },
     {
-      "epoch": 2.1310946076848563,
-      "grad_norm": 0.5529822707176208,
-      "learning_rate": 0.00047268828790001077,
-      "loss": 3.5427,
+      "epoch": 2.1347708894878705,
+      "grad_norm": 0.6371639966964722,
+      "learning_rate": 0.0004724684295736643,
+      "loss": 3.5504,
       "step": 19800
     },
     {
-      "epoch": 2.1364761597244644,
-      "grad_norm": 0.6028132438659668,
-      "learning_rate": 0.0004723650468699493,
-      "loss": 3.5596,
+      "epoch": 2.1401617250673857,
+      "grad_norm": 0.5546467304229736,
+      "learning_rate": 0.0004721446303291959,
+      "loss": 3.5782,
       "step": 19850
     },
     {
-      "epoch": 2.141857711764073,
-      "grad_norm": 0.5742948651313782,
-      "learning_rate": 0.0004720418058398879,
-      "loss": 3.5591,
+      "epoch": 2.1455525606469004,
+      "grad_norm": 0.6736409664154053,
+      "learning_rate": 0.0004718208310847275,
+      "loss": 3.5789,
       "step": 19900
     },
     {
-      "epoch": 2.147239263803681,
-      "grad_norm": 0.6175876259803772,
-      "learning_rate": 0.00047171856480982644,
-      "loss": 3.5649,
+      "epoch": 2.150943396226415,
+      "grad_norm": 0.5970144867897034,
+      "learning_rate": 0.000471497031840259,
+      "loss": 3.5757,
       "step": 19950
     },
     {
-      "epoch": 2.152620815843289,
-      "grad_norm": 0.5835720300674438,
-      "learning_rate": 0.0004713953237797651,
-      "loss": 3.5654,
+      "epoch": 2.1563342318059298,
+      "grad_norm": 0.5611630082130432,
+      "learning_rate": 0.0004711732325957905,
+      "loss": 3.5575,
       "step": 20000
     },
     {
-      "epoch": 2.152620815843289,
-      "eval_accuracy": 0.36219320634780594,
-      "eval_loss": 3.5732574462890625,
-      "eval_runtime": 183.9589,
-      "eval_samples_per_second": 97.908,
-      "eval_steps_per_second": 6.121,
+      "epoch": 2.1563342318059298,
+      "eval_accuracy": 0.3622267800901406,
+      "eval_loss": 3.5747625827789307,
+      "eval_runtime": 183.8989,
+      "eval_samples_per_second": 97.94,
+      "eval_steps_per_second": 6.123,
       "step": 20000
     },
     {
-      "epoch": 2.1580023678828972,
-      "grad_norm": 0.5701449513435364,
-      "learning_rate": 0.0004710720827497037,
-      "loss": 3.5597,
+      "epoch": 2.161725067385445,
+      "grad_norm": 0.6023396253585815,
+      "learning_rate": 0.00047084943335132213,
+      "loss": 3.5588,
       "step": 20050
     },
     {
-      "epoch": 2.163383919922506,
-      "grad_norm": 0.5607191920280457,
-      "learning_rate": 0.00047074884171964223,
-      "loss": 3.5617,
+      "epoch": 2.1671159029649596,
+      "grad_norm": 0.61530601978302,
+      "learning_rate": 0.0004705256341068537,
+      "loss": 3.5634,
       "step": 20100
     },
     {
-      "epoch": 2.168765471962114,
-      "grad_norm": 0.5911417603492737,
-      "learning_rate": 0.0004704256006895808,
-      "loss": 3.5635,
+      "epoch": 2.1725067385444743,
+      "grad_norm": 0.5686957836151123,
+      "learning_rate": 0.0004702018348623853,
+      "loss": 3.5596,
       "step": 20150
     },
     {
-      "epoch": 2.174147024001722,
-      "grad_norm": 0.5466926097869873,
-      "learning_rate": 0.00047010235965951936,
-      "loss": 3.5751,
+      "epoch": 2.177897574123989,
+      "grad_norm": 0.6189788579940796,
+      "learning_rate": 0.00046987803561791684,
+      "loss": 3.5742,
       "step": 20200
     },
     {
-      "epoch": 2.1795285760413305,
-      "grad_norm": 0.58399897813797,
-      "learning_rate": 0.00046977911862945796,
-      "loss": 3.5768,
+      "epoch": 2.183288409703504,
+      "grad_norm": 0.5632470846176147,
+      "learning_rate": 0.00046955423637344844,
+      "loss": 3.5603,
       "step": 20250
     },
     {
-      "epoch": 2.1849101280809387,
-      "grad_norm": 0.6332233548164368,
-      "learning_rate": 0.0004694558775993966,
-      "loss": 3.5679,
+      "epoch": 2.188679245283019,
+      "grad_norm": 0.5766399502754211,
+      "learning_rate": 0.00046923043712898,
+      "loss": 3.568,
       "step": 20300
     },
     {
-      "epoch": 2.1902916801205468,
-      "grad_norm": 0.5964486598968506,
-      "learning_rate": 0.00046913263656933515,
-      "loss": 3.5699,
+      "epoch": 2.1940700808625335,
+      "grad_norm": 0.6770806312561035,
+      "learning_rate": 0.0004689066378845116,
+      "loss": 3.5482,
       "step": 20350
     },
     {
-      "epoch": 2.195673232160155,
-      "grad_norm": 0.5788668394088745,
-      "learning_rate": 0.00046880939553927374,
-      "loss": 3.5586,
+      "epoch": 2.1994609164420487,
+      "grad_norm": 0.6096656918525696,
+      "learning_rate": 0.00046858283864004315,
+      "loss": 3.5546,
       "step": 20400
     },
     {
-      "epoch": 2.2010547841997634,
-      "grad_norm": 0.5830016732215881,
-      "learning_rate": 0.00046848615450921234,
-      "loss": 3.5675,
+      "epoch": 2.2048517520215634,
+      "grad_norm": 0.5745991468429565,
+      "learning_rate": 0.0004682590393955747,
+      "loss": 3.5587,
       "step": 20450
     },
     {
-      "epoch": 2.2064363362393715,
-      "grad_norm": 0.6435354948043823,
-      "learning_rate": 0.0004681629134791509,
-      "loss": 3.572,
+      "epoch": 2.210242587601078,
+      "grad_norm": 0.5892425179481506,
+      "learning_rate": 0.0004679352401511063,
+      "loss": 3.5679,
       "step": 20500
     },
     {
-      "epoch": 2.2118178882789796,
-      "grad_norm": 0.5667049884796143,
-      "learning_rate": 0.00046783967244908947,
-      "loss": 3.5638,
+      "epoch": 2.215633423180593,
+      "grad_norm": 0.5695005655288696,
+      "learning_rate": 0.0004676114409066378,
+      "loss": 3.5644,
       "step": 20550
     },
     {
-      "epoch": 2.2171994403185877,
-      "grad_norm": 0.6753138899803162,
-      "learning_rate": 0.0004675164314190281,
-      "loss": 3.5719,
+      "epoch": 2.221024258760108,
+      "grad_norm": 0.5493165850639343,
+      "learning_rate": 0.00046728764166216946,
+      "loss": 3.5595,
       "step": 20600
     },
     {
-      "epoch": 2.2225809923581963,
-      "grad_norm": 0.6224473118782043,
-      "learning_rate": 0.00046719319038896666,
-      "loss": 3.5729,
+      "epoch": 2.2264150943396226,
+      "grad_norm": 0.5420545935630798,
+      "learning_rate": 0.00046696384241770095,
+      "loss": 3.5603,
       "step": 20650
     },
     {
-      "epoch": 2.2279625443978044,
-      "grad_norm": 0.6385303735733032,
-      "learning_rate": 0.00046686994935890526,
-      "loss": 3.5713,
+      "epoch": 2.2318059299191373,
+      "grad_norm": 0.5981364846229553,
+      "learning_rate": 0.00046664004317323256,
+      "loss": 3.5555,
       "step": 20700
     },
     {
-      "epoch": 2.2333440964374125,
-      "grad_norm": 0.5930754542350769,
-      "learning_rate": 0.0004665467083288438,
-      "loss": 3.5553,
+      "epoch": 2.2371967654986524,
+      "grad_norm": 0.5738416910171509,
+      "learning_rate": 0.0004663162439287641,
+      "loss": 3.5551,
       "step": 20750
     },
     {
-      "epoch": 2.2387256484770206,
-      "grad_norm": 0.5501213669776917,
-      "learning_rate": 0.0004662234672987824,
-      "loss": 3.5775,
+      "epoch": 2.242587601078167,
+      "grad_norm": 0.6114598512649536,
+      "learning_rate": 0.00046599244468429566,
+      "loss": 3.5633,
       "step": 20800
     },
     {
-      "epoch": 2.244107200516629,
-      "grad_norm": 0.5862070322036743,
-      "learning_rate": 0.00046590022626872104,
-      "loss": 3.5752,
+      "epoch": 2.247978436657682,
+      "grad_norm": 0.5876783132553101,
+      "learning_rate": 0.00046566864543982726,
+      "loss": 3.5566,
       "step": 20850
     },
     {
-      "epoch": 2.2494887525562373,
-      "grad_norm": 0.5672935247421265,
-      "learning_rate": 0.0004655769852386596,
-      "loss": 3.5702,
+      "epoch": 2.2533692722371965,
+      "grad_norm": 0.5674042105674744,
+      "learning_rate": 0.0004653448461953588,
+      "loss": 3.5524,
       "step": 20900
     },
     {
-      "epoch": 2.2548703045958454,
-      "grad_norm": 0.564063310623169,
-      "learning_rate": 0.0004652537442085982,
-      "loss": 3.5571,
+      "epoch": 2.2587601078167117,
+      "grad_norm": 0.5803061127662659,
+      "learning_rate": 0.00046502752293577977,
+      "loss": 3.5627,
       "step": 20950
     },
     {
-      "epoch": 2.2602518566354535,
-      "grad_norm": 0.6813990473747253,
-      "learning_rate": 0.00046493050317853677,
-      "loss": 3.5613,
+      "epoch": 2.2641509433962264,
+      "grad_norm": 0.704203188419342,
+      "learning_rate": 0.00046470372369131137,
+      "loss": 3.544,
       "step": 21000
     },
     {
-      "epoch": 2.2602518566354535,
-      "eval_accuracy": 0.36378736151782,
-      "eval_loss": 3.5629472732543945,
-      "eval_runtime": 183.0527,
-      "eval_samples_per_second": 98.392,
-      "eval_steps_per_second": 6.151,
+      "epoch": 2.2641509433962264,
+      "eval_accuracy": 0.3635250734466358,
+      "eval_loss": 3.5621912479400635,
+      "eval_runtime": 184.1964,
+      "eval_samples_per_second": 97.782,
+      "eval_steps_per_second": 6.113,
       "step": 21000
     },
     {
-      "epoch": 2.265633408675062,
-      "grad_norm": 0.5663150548934937,
-      "learning_rate": 0.0004646072621484753,
-      "loss": 3.5268,
+      "epoch": 2.269541778975741,
+      "grad_norm": 0.5122193694114685,
+      "learning_rate": 0.0004643799244468429,
+      "loss": 3.5593,
       "step": 21050
     },
     {
-      "epoch": 2.27101496071467,
-      "grad_norm": 0.5659851431846619,
-      "learning_rate": 0.00046429048593901517,
-      "loss": 3.5753,
+      "epoch": 2.274932614555256,
+      "grad_norm": 0.5831567049026489,
+      "learning_rate": 0.0004640561252023745,
+      "loss": 3.5639,
       "step": 21100
     },
     {
-      "epoch": 2.2763965127542782,
-      "grad_norm": 0.5690809488296509,
-      "learning_rate": 0.0004639672449089537,
-      "loss": 3.5542,
+      "epoch": 2.280323450134771,
+      "grad_norm": 0.6369010210037231,
+      "learning_rate": 0.0004637323259579061,
+      "loss": 3.5753,
       "step": 21150
     },
     {
-      "epoch": 2.281778064793887,
-      "grad_norm": 0.6252328157424927,
-      "learning_rate": 0.0004636440038788923,
-      "loss": 3.565,
+      "epoch": 2.2857142857142856,
+      "grad_norm": 0.5553130507469177,
+      "learning_rate": 0.0004634085267134376,
+      "loss": 3.5288,
       "step": 21200
     },
     {
-      "epoch": 2.287159616833495,
-      "grad_norm": 0.5633666515350342,
-      "learning_rate": 0.00046332076284883095,
-      "loss": 3.5501,
+      "epoch": 2.2911051212938007,
+      "grad_norm": 0.5680311322212219,
+      "learning_rate": 0.00046308472746896923,
+      "loss": 3.5556,
       "step": 21250
     },
     {
-      "epoch": 2.292541168873103,
-      "grad_norm": 0.634242057800293,
-      "learning_rate": 0.0004629975218187695,
-      "loss": 3.5546,
+      "epoch": 2.2964959568733154,
+      "grad_norm": 0.5479399561882019,
+      "learning_rate": 0.00046276092822450073,
+      "loss": 3.5568,
       "step": 21300
     },
     {
-      "epoch": 2.297922720912711,
-      "grad_norm": 0.5788732767105103,
-      "learning_rate": 0.0004626742807887081,
-      "loss": 3.5609,
+      "epoch": 2.30188679245283,
+      "grad_norm": 0.5450858473777771,
+      "learning_rate": 0.00046243712898003233,
+      "loss": 3.55,
       "step": 21350
     },
     {
-      "epoch": 2.303304272952319,
-      "grad_norm": 0.5648044347763062,
-      "learning_rate": 0.00046235103975864663,
-      "loss": 3.546,
+      "epoch": 2.3072776280323453,
+      "grad_norm": 0.5511463284492493,
+      "learning_rate": 0.0004621198057204533,
+      "loss": 3.5647,
       "step": 21400
     },
     {
-      "epoch": 2.3086858249919278,
-      "grad_norm": 0.6103492379188538,
-      "learning_rate": 0.0004620277987285852,
-      "loss": 3.5766,
+      "epoch": 2.31266846361186,
+      "grad_norm": 0.568325936794281,
+      "learning_rate": 0.00046179600647598484,
+      "loss": 3.5618,
       "step": 21450
     },
     {
-      "epoch": 2.314067377031536,
-      "grad_norm": 0.6276610493659973,
-      "learning_rate": 0.0004617045576985239,
-      "loss": 3.5737,
+      "epoch": 2.3180592991913747,
+      "grad_norm": 0.5608011484146118,
+      "learning_rate": 0.0004614722072315164,
+      "loss": 3.5649,
       "step": 21500
     },
     {
-      "epoch": 2.319448929071144,
-      "grad_norm": 0.5594583749771118,
-      "learning_rate": 0.0004613813166684624,
-      "loss": 3.5475,
+      "epoch": 2.3234501347708894,
+      "grad_norm": 0.5979920029640198,
+      "learning_rate": 0.000461148407987048,
+      "loss": 3.5469,
       "step": 21550
     },
     {
-      "epoch": 2.3248304811107525,
-      "grad_norm": 0.5839714407920837,
-      "learning_rate": 0.000461058075638401,
-      "loss": 3.5706,
+      "epoch": 2.3288409703504045,
+      "grad_norm": 0.5907413363456726,
+      "learning_rate": 0.00046082460874257954,
+      "loss": 3.5678,
       "step": 21600
     },
     {
-      "epoch": 2.3302120331503606,
-      "grad_norm": 0.5914338827133179,
-      "learning_rate": 0.00046073483460833955,
-      "loss": 3.5553,
+      "epoch": 2.334231805929919,
+      "grad_norm": 0.5995841026306152,
+      "learning_rate": 0.00046050080949811115,
+      "loss": 3.5616,
       "step": 21650
     },
     {
-      "epoch": 2.3355935851899687,
-      "grad_norm": 0.5473167896270752,
-      "learning_rate": 0.00046041159357827814,
-      "loss": 3.5412,
+      "epoch": 2.339622641509434,
+      "grad_norm": 0.5671567320823669,
+      "learning_rate": 0.0004601770102536427,
+      "loss": 3.5429,
       "step": 21700
     },
     {
-      "epoch": 2.340975137229577,
-      "grad_norm": 0.5862297415733337,
-      "learning_rate": 0.00046008835254821674,
-      "loss": 3.5511,
+      "epoch": 2.3450134770889486,
+      "grad_norm": 0.5616987943649292,
+      "learning_rate": 0.0004598532110091743,
+      "loss": 3.5587,
       "step": 21750
     },
     {
-      "epoch": 2.3463566892691854,
-      "grad_norm": 0.5849961042404175,
-      "learning_rate": 0.0004597651115181554,
-      "loss": 3.5424,
+      "epoch": 2.3504043126684637,
+      "grad_norm": 0.6421756744384766,
+      "learning_rate": 0.00045952941176470585,
+      "loss": 3.5785,
       "step": 21800
     },
     {
-      "epoch": 2.3517382413087935,
-      "grad_norm": 0.6141690015792847,
-      "learning_rate": 0.00045944187048809393,
-      "loss": 3.5617,
+      "epoch": 2.3557951482479784,
+      "grad_norm": 0.5783190131187439,
+      "learning_rate": 0.00045920561252023746,
+      "loss": 3.566,
       "step": 21850
     },
     {
-      "epoch": 2.3571197933484016,
-      "grad_norm": 0.5442442893981934,
-      "learning_rate": 0.0004591186294580325,
-      "loss": 3.5594,
+      "epoch": 2.361185983827493,
+      "grad_norm": 0.6462935209274292,
+      "learning_rate": 0.000458881813275769,
+      "loss": 3.5447,
       "step": 21900
     },
     {
-      "epoch": 2.3625013453880097,
-      "grad_norm": 0.5921179056167603,
-      "learning_rate": 0.00045879538842797106,
-      "loss": 3.5472,
+      "epoch": 2.3665768194070083,
+      "grad_norm": 0.5985819697380066,
+      "learning_rate": 0.0004585580140313005,
+      "loss": 3.5566,
       "step": 21950
     },
     {
-      "epoch": 2.3678828974276183,
-      "grad_norm": 0.5729749798774719,
-      "learning_rate": 0.00045847214739790966,
-      "loss": 3.5588,
+      "epoch": 2.371967654986523,
+      "grad_norm": 0.5702811479568481,
+      "learning_rate": 0.0004582342147868321,
+      "loss": 3.5475,
       "step": 22000
     },
     {
-      "epoch": 2.3678828974276183,
-      "eval_accuracy": 0.36438647354122905,
-      "eval_loss": 3.5525951385498047,
-      "eval_runtime": 183.627,
-      "eval_samples_per_second": 98.085,
-      "eval_steps_per_second": 6.132,
+      "epoch": 2.371967654986523,
+      "eval_accuracy": 0.36468602955293355,
+      "eval_loss": 3.551884174346924,
+      "eval_runtime": 184.1792,
+      "eval_samples_per_second": 97.791,
+      "eval_steps_per_second": 6.114,
       "step": 22000
     },
     {
-      "epoch": 2.3732644494672264,
-      "grad_norm": 0.5790433287620544,
-      "learning_rate": 0.0004581489063678482,
-      "loss": 3.5655,
+      "epoch": 2.3773584905660377,
+      "grad_norm": 0.577484130859375,
+      "learning_rate": 0.00045791041554236366,
+      "loss": 3.5497,
       "step": 22050
     },
     {
-      "epoch": 2.3786460015068345,
-      "grad_norm": 0.5972291827201843,
-      "learning_rate": 0.00045782566533778685,
-      "loss": 3.553,
+      "epoch": 2.382749326145553,
+      "grad_norm": 0.5937749743461609,
+      "learning_rate": 0.00045758661629789526,
+      "loss": 3.5634,
       "step": 22100
     },
     {
-      "epoch": 2.384027553546443,
-      "grad_norm": 0.6700719594955444,
-      "learning_rate": 0.00045750242430772544,
-      "loss": 3.5554,
+      "epoch": 2.3881401617250675,
+      "grad_norm": 0.5452225208282471,
+      "learning_rate": 0.0004572628170534268,
+      "loss": 3.5662,
       "step": 22150
     },
     {
-      "epoch": 2.389409105586051,
-      "grad_norm": 0.5736136436462402,
-      "learning_rate": 0.000457179183277664,
-      "loss": 3.5505,
+      "epoch": 2.393530997304582,
+      "grad_norm": 0.5550395846366882,
+      "learning_rate": 0.0004569390178089584,
+      "loss": 3.549,
       "step": 22200
     },
     {
-      "epoch": 2.3947906576256592,
-      "grad_norm": 0.5969544053077698,
-      "learning_rate": 0.0004568559422476026,
-      "loss": 3.5582,
+      "epoch": 2.398921832884097,
+      "grad_norm": 0.5785404443740845,
+      "learning_rate": 0.00045661521856448997,
+      "loss": 3.5595,
       "step": 22250
     },
     {
-      "epoch": 2.4001722096652673,
-      "grad_norm": 0.5612258911132812,
-      "learning_rate": 0.00045653270121754117,
-      "loss": 3.552,
+      "epoch": 2.404312668463612,
+      "grad_norm": 0.5671253204345703,
+      "learning_rate": 0.0004562914193200216,
+      "loss": 3.545,
       "step": 22300
     },
     {
-      "epoch": 2.4055537617048754,
-      "grad_norm": 0.5993139743804932,
-      "learning_rate": 0.0004562094601874797,
-      "loss": 3.5753,
+      "epoch": 2.4097035040431267,
+      "grad_norm": 0.5955759882926941,
+      "learning_rate": 0.0004559676200755531,
+      "loss": 3.5693,
       "step": 22350
     },
     {
-      "epoch": 2.410935313744484,
-      "grad_norm": 0.5695291757583618,
-      "learning_rate": 0.00045588621915741836,
-      "loss": 3.5636,
+      "epoch": 2.4150943396226414,
+      "grad_norm": 0.5546286702156067,
+      "learning_rate": 0.0004556438208310847,
+      "loss": 3.5524,
       "step": 22400
     },
     {
-      "epoch": 2.416316865784092,
-      "grad_norm": 0.6173685193061829,
-      "learning_rate": 0.00045556297812735696,
-      "loss": 3.5615,
+      "epoch": 2.420485175202156,
+      "grad_norm": 0.5877081751823425,
+      "learning_rate": 0.0004553200215866163,
+      "loss": 3.5545,
       "step": 22450
     },
     {
-      "epoch": 2.4216984178237,
-      "grad_norm": 0.5470847487449646,
-      "learning_rate": 0.0004552397370972955,
-      "loss": 3.5524,
+      "epoch": 2.4258760107816713,
+      "grad_norm": 0.599201500415802,
+      "learning_rate": 0.00045499622234214783,
+      "loss": 3.5632,
       "step": 22500
     },
     {
-      "epoch": 2.4270799698633088,
-      "grad_norm": 0.5483660697937012,
-      "learning_rate": 0.0004549164960672341,
-      "loss": 3.5562,
+      "epoch": 2.431266846361186,
+      "grad_norm": 0.5470482707023621,
+      "learning_rate": 0.00045467242309767944,
+      "loss": 3.5529,
       "step": 22550
     },
     {
-      "epoch": 2.432461521902917,
-      "grad_norm": 0.630761444568634,
-      "learning_rate": 0.00045459325503717263,
-      "loss": 3.564,
+      "epoch": 2.4366576819407006,
+      "grad_norm": 0.6305721402168274,
+      "learning_rate": 0.00045434862385321093,
+      "loss": 3.5827,
       "step": 22600
     },
     {
-      "epoch": 2.437843073942525,
-      "grad_norm": 0.5955401062965393,
-      "learning_rate": 0.0004542700140071113,
-      "loss": 3.5453,
+      "epoch": 2.442048517520216,
+      "grad_norm": 0.6012084484100342,
+      "learning_rate": 0.0004540248246087426,
+      "loss": 3.5438,
       "step": 22650
     },
     {
-      "epoch": 2.443224625982133,
-      "grad_norm": 0.5875324606895447,
-      "learning_rate": 0.0004539467729770499,
-      "loss": 3.5226,
+      "epoch": 2.4474393530997305,
+      "grad_norm": 0.5885149240493774,
+      "learning_rate": 0.0004537010253642741,
+      "loss": 3.5359,
       "step": 22700
     },
     {
-      "epoch": 2.4486061780217416,
-      "grad_norm": 0.5740516185760498,
-      "learning_rate": 0.0004536235319469884,
-      "loss": 3.5704,
+      "epoch": 2.452830188679245,
+      "grad_norm": 0.5837857127189636,
+      "learning_rate": 0.00045337722611980564,
+      "loss": 3.539,
       "step": 22750
     },
     {
-      "epoch": 2.4539877300613497,
-      "grad_norm": 0.5694280862808228,
-      "learning_rate": 0.000453300290916927,
-      "loss": 3.5505,
+      "epoch": 2.4582210242587603,
+      "grad_norm": 0.6058210134506226,
+      "learning_rate": 0.00045305342687533724,
+      "loss": 3.552,
       "step": 22800
     },
     {
-      "epoch": 2.459369282100958,
-      "grad_norm": 0.6105132699012756,
-      "learning_rate": 0.0004529770498868656,
-      "loss": 3.5632,
+      "epoch": 2.463611859838275,
+      "grad_norm": 0.5641347169876099,
+      "learning_rate": 0.0004527296276308688,
+      "loss": 3.5358,
       "step": 22850
     },
     {
-      "epoch": 2.464750834140566,
-      "grad_norm": 0.6120337843894958,
-      "learning_rate": 0.00045265380885680414,
-      "loss": 3.5523,
+      "epoch": 2.4690026954177897,
+      "grad_norm": 0.6286354064941406,
+      "learning_rate": 0.0004524058283864004,
+      "loss": 3.5536,
       "step": 22900
     },
     {
-      "epoch": 2.4701323861801745,
-      "grad_norm": 0.5750212669372559,
-      "learning_rate": 0.0004523305678267428,
-      "loss": 3.5434,
+      "epoch": 2.4743935309973044,
+      "grad_norm": 0.6253069043159485,
+      "learning_rate": 0.00045208202914193195,
+      "loss": 3.5493,
       "step": 22950
     },
     {
-      "epoch": 2.4755139382197826,
-      "grad_norm": 0.5812209248542786,
-      "learning_rate": 0.0004520073267966814,
-      "loss": 3.5626,
+      "epoch": 2.4797843665768196,
+      "grad_norm": 0.6191100478172302,
+      "learning_rate": 0.00045175822989746355,
+      "loss": 3.5419,
       "step": 23000
     },
     {
-      "epoch": 2.4755139382197826,
-      "eval_accuracy": 0.36594868476222564,
-      "eval_loss": 3.5408027172088623,
-      "eval_runtime": 182.9812,
-      "eval_samples_per_second": 98.431,
-      "eval_steps_per_second": 6.154,
+      "epoch": 2.4797843665768196,
+      "eval_accuracy": 0.36554047586270666,
+      "eval_loss": 3.5396900177001953,
+      "eval_runtime": 184.4581,
+      "eval_samples_per_second": 97.643,
+      "eval_steps_per_second": 6.104,
       "step": 23000
     },
     {
-      "epoch": 2.4808954902593907,
-      "grad_norm": 0.6218942403793335,
-      "learning_rate": 0.00045168408576661993,
-      "loss": 3.5572,
+      "epoch": 2.4851752021563343,
+      "grad_norm": 0.5773845911026001,
+      "learning_rate": 0.0004514344306529951,
+      "loss": 3.553,
       "step": 23050
     },
     {
-      "epoch": 2.4862770422989993,
-      "grad_norm": 0.5775257349014282,
-      "learning_rate": 0.0004513608447365585,
-      "loss": 3.5484,
+      "epoch": 2.490566037735849,
+      "grad_norm": 0.6329528093338013,
+      "learning_rate": 0.0004511106314085267,
+      "loss": 3.5569,
       "step": 23100
     },
     {
-      "epoch": 2.4916585943386074,
-      "grad_norm": 0.5762562155723572,
-      "learning_rate": 0.00045103760370649706,
-      "loss": 3.5744,
+      "epoch": 2.4959568733153636,
+      "grad_norm": 0.6235628724098206,
+      "learning_rate": 0.00045078683216405826,
+      "loss": 3.5545,
       "step": 23150
     },
     {
-      "epoch": 2.4970401463782155,
-      "grad_norm": 0.5888144373893738,
-      "learning_rate": 0.00045071436267643566,
-      "loss": 3.5604,
+      "epoch": 2.501347708894879,
+      "grad_norm": 0.5293558239936829,
+      "learning_rate": 0.00045046303291958976,
+      "loss": 3.5504,
       "step": 23200
     },
     {
-      "epoch": 2.5024216984178236,
-      "grad_norm": 0.6638675928115845,
-      "learning_rate": 0.0004503911216463743,
-      "loss": 3.5623,
+      "epoch": 2.5067385444743935,
+      "grad_norm": 0.5963000655174255,
+      "learning_rate": 0.0004501392336751214,
+      "loss": 3.5493,
       "step": 23250
     },
     {
-      "epoch": 2.5078032504574317,
-      "grad_norm": 0.620266318321228,
-      "learning_rate": 0.00045006788061631285,
-      "loss": 3.557,
+      "epoch": 2.512129380053908,
+      "grad_norm": 0.5842360854148865,
+      "learning_rate": 0.0004498154344306529,
+      "loss": 3.534,
       "step": 23300
     },
     {
-      "epoch": 2.5131848024970402,
-      "grad_norm": 0.5920720100402832,
-      "learning_rate": 0.00044974463958625144,
-      "loss": 3.5369,
+      "epoch": 2.5175202156334233,
+      "grad_norm": 0.6223546266555786,
+      "learning_rate": 0.0004494916351861845,
+      "loss": 3.5404,
       "step": 23350
     },
     {
-      "epoch": 2.5185663545366483,
-      "grad_norm": 0.5830405950546265,
-      "learning_rate": 0.00044942139855619004,
-      "loss": 3.5637,
+      "epoch": 2.522911051212938,
+      "grad_norm": 0.6296982765197754,
+      "learning_rate": 0.00044916783594171607,
+      "loss": 3.5378,
       "step": 23400
     },
     {
-      "epoch": 2.5239479065762565,
-      "grad_norm": 0.5792795419692993,
-      "learning_rate": 0.0004490981575261286,
-      "loss": 3.5551,
+      "epoch": 2.5283018867924527,
+      "grad_norm": 0.555363655090332,
+      "learning_rate": 0.00044884403669724767,
+      "loss": 3.5486,
       "step": 23450
     },
     {
-      "epoch": 2.529329458615865,
-      "grad_norm": 0.5703508257865906,
-      "learning_rate": 0.0004487749164960672,
-      "loss": 3.5656,
+      "epoch": 2.533692722371968,
+      "grad_norm": 0.5806699991226196,
+      "learning_rate": 0.0004485202374527792,
+      "loss": 3.5519,
       "step": 23500
     },
     {
-      "epoch": 2.534711010655473,
-      "grad_norm": 0.5803419947624207,
-      "learning_rate": 0.0004484516754660058,
-      "loss": 3.5491,
+      "epoch": 2.5390835579514826,
+      "grad_norm": 0.5958206057548523,
+      "learning_rate": 0.00044819643820831083,
+      "loss": 3.554,
       "step": 23550
     },
     {
-      "epoch": 2.540092562695081,
-      "grad_norm": 0.6119604706764221,
-      "learning_rate": 0.00044812843443594436,
-      "loss": 3.556,
+      "epoch": 2.5444743935309972,
+      "grad_norm": 0.597139835357666,
+      "learning_rate": 0.0004478726389638424,
+      "loss": 3.5571,
       "step": 23600
     },
     {
-      "epoch": 2.5454741147346893,
-      "grad_norm": 0.5884591937065125,
-      "learning_rate": 0.00044780519340588296,
-      "loss": 3.5301,
+      "epoch": 2.5498652291105124,
+      "grad_norm": 0.5986602902412415,
+      "learning_rate": 0.00044754883971937393,
+      "loss": 3.5493,
       "step": 23650
     },
     {
-      "epoch": 2.550855666774298,
-      "grad_norm": 0.5345640778541565,
-      "learning_rate": 0.0004474819523758215,
-      "loss": 3.5249,
+      "epoch": 2.555256064690027,
+      "grad_norm": 0.5681850910186768,
+      "learning_rate": 0.00044722504047490553,
+      "loss": 3.5539,
       "step": 23700
     },
     {
-      "epoch": 2.556237218813906,
-      "grad_norm": 0.5864933729171753,
-      "learning_rate": 0.0004471587113457601,
-      "loss": 3.5468,
+      "epoch": 2.560646900269542,
+      "grad_norm": 0.5716060996055603,
+      "learning_rate": 0.0004469012412304371,
+      "loss": 3.5465,
       "step": 23750
     },
     {
-      "epoch": 2.561618770853514,
-      "grad_norm": 0.5660606622695923,
-      "learning_rate": 0.00044683547031569874,
-      "loss": 3.5576,
+      "epoch": 2.5660377358490565,
+      "grad_norm": 0.6231576204299927,
+      "learning_rate": 0.0004465774419859687,
+      "loss": 3.561,
       "step": 23800
     },
     {
-      "epoch": 2.567000322893122,
-      "grad_norm": 0.5605663061141968,
-      "learning_rate": 0.0004465122292856373,
-      "loss": 3.5404,
+      "epoch": 2.571428571428571,
+      "grad_norm": 0.577888011932373,
+      "learning_rate": 0.00044625364274150024,
+      "loss": 3.564,
       "step": 23850
     },
     {
-      "epoch": 2.5723818749327307,
-      "grad_norm": 0.6553633213043213,
-      "learning_rate": 0.0004461889882555759,
-      "loss": 3.5576,
+      "epoch": 2.5768194070080863,
+      "grad_norm": 0.5581002235412598,
+      "learning_rate": 0.00044592984349703184,
+      "loss": 3.5651,
       "step": 23900
     },
     {
-      "epoch": 2.577763426972339,
-      "grad_norm": 0.6218010783195496,
-      "learning_rate": 0.00044586574722551447,
-      "loss": 3.559,
+      "epoch": 2.582210242587601,
+      "grad_norm": 0.5751173496246338,
+      "learning_rate": 0.00044560604425256334,
+      "loss": 3.5375,
       "step": 23950
     },
     {
-      "epoch": 2.583144979011947,
-      "grad_norm": 0.6036173701286316,
-      "learning_rate": 0.000445542506195453,
-      "loss": 3.5623,
+      "epoch": 2.5876010781671157,
+      "grad_norm": 0.56992107629776,
+      "learning_rate": 0.000445282245008095,
+      "loss": 3.5362,
       "step": 24000
     },
     {
-      "epoch": 2.583144979011947,
-      "eval_accuracy": 0.3662233592626207,
-      "eval_loss": 3.533743381500244,
-      "eval_runtime": 183.448,
-      "eval_samples_per_second": 98.18,
-      "eval_steps_per_second": 6.138,
+      "epoch": 2.5876010781671157,
+      "eval_accuracy": 0.3665120499855437,
+      "eval_loss": 3.53277587890625,
+      "eval_runtime": 184.009,
+      "eval_samples_per_second": 97.881,
+      "eval_steps_per_second": 6.119,
       "step": 24000
     },
     {
-      "epoch": 2.5885265310515555,
-      "grad_norm": 0.6276186108589172,
-      "learning_rate": 0.0004452192651653916,
-      "loss": 3.5581,
+      "epoch": 2.592991913746631,
+      "grad_norm": 0.6139329671859741,
+      "learning_rate": 0.0004449584457636265,
+      "loss": 3.5356,
       "step": 24050
     },
     {
-      "epoch": 2.5939080830911636,
-      "grad_norm": 0.6104943156242371,
-      "learning_rate": 0.00044489602413533025,
-      "loss": 3.5323,
+      "epoch": 2.5983827493261455,
+      "grad_norm": 0.6108660697937012,
+      "learning_rate": 0.00044463464651915805,
+      "loss": 3.536,
       "step": 24100
     },
     {
-      "epoch": 2.5992896351307717,
-      "grad_norm": 0.5991529822349548,
-      "learning_rate": 0.0004445727831052688,
-      "loss": 3.5428,
+      "epoch": 2.6037735849056602,
+      "grad_norm": 0.6183775663375854,
+      "learning_rate": 0.000444317323259579,
+      "loss": 3.5405,
       "step": 24150
     },
     {
-      "epoch": 2.60467118717038,
-      "grad_norm": 0.6116634607315063,
-      "learning_rate": 0.0004442495420752074,
-      "loss": 3.5436,
+      "epoch": 2.6091644204851754,
+      "grad_norm": 0.6223834156990051,
+      "learning_rate": 0.0004439935240151106,
+      "loss": 3.5551,
       "step": 24200
     },
     {
-      "epoch": 2.610052739209988,
-      "grad_norm": 0.637483537197113,
-      "learning_rate": 0.00044392630104514593,
-      "loss": 3.5658,
+      "epoch": 2.61455525606469,
+      "grad_norm": 0.5659055709838867,
+      "learning_rate": 0.00044366972477064215,
+      "loss": 3.557,
       "step": 24250
     },
     {
-      "epoch": 2.6154342912495965,
-      "grad_norm": 0.6104105710983276,
-      "learning_rate": 0.0004436030600150845,
-      "loss": 3.5557,
+      "epoch": 2.6199460916442048,
+      "grad_norm": 0.6240912079811096,
+      "learning_rate": 0.00044334592552617376,
+      "loss": 3.5447,
       "step": 24300
     },
     {
-      "epoch": 2.6208158432892046,
-      "grad_norm": 0.603550910949707,
-      "learning_rate": 0.0004432798189850231,
-      "loss": 3.5442,
+      "epoch": 2.62533692722372,
+      "grad_norm": 0.5938546061515808,
+      "learning_rate": 0.0004430221262817053,
+      "loss": 3.5382,
       "step": 24350
     },
     {
-      "epoch": 2.6261973953288127,
-      "grad_norm": 0.5980435013771057,
-      "learning_rate": 0.0004429565779549617,
-      "loss": 3.5587,
+      "epoch": 2.6307277628032346,
+      "grad_norm": 0.5961259007453918,
+      "learning_rate": 0.00044269832703723686,
+      "loss": 3.5485,
       "step": 24400
     },
     {
-      "epoch": 2.6315789473684212,
-      "grad_norm": 0.556158185005188,
-      "learning_rate": 0.0004426333369249003,
-      "loss": 3.5477,
+      "epoch": 2.6361185983827493,
+      "grad_norm": 0.5874900817871094,
+      "learning_rate": 0.00044237452779276846,
+      "loss": 3.5231,
       "step": 24450
     },
     {
-      "epoch": 2.6369604994080293,
-      "grad_norm": 0.5729323029518127,
-      "learning_rate": 0.0004423100958948389,
-      "loss": 3.5398,
+      "epoch": 2.641509433962264,
+      "grad_norm": 0.6635915040969849,
+      "learning_rate": 0.0004420507285483,
+      "loss": 3.5296,
       "step": 24500
     },
     {
-      "epoch": 2.6423420514476375,
-      "grad_norm": 0.7184017896652222,
-      "learning_rate": 0.00044198685486477744,
-      "loss": 3.5459,
+      "epoch": 2.6469002695417787,
+      "grad_norm": 0.6165565252304077,
+      "learning_rate": 0.0004417269293038316,
+      "loss": 3.5471,
       "step": 24550
     },
     {
-      "epoch": 2.6477236034872456,
-      "grad_norm": 0.6241846680641174,
-      "learning_rate": 0.00044166361383471604,
-      "loss": 3.5393,
+      "epoch": 2.652291105121294,
+      "grad_norm": 0.5879116654396057,
+      "learning_rate": 0.0004414031300593631,
+      "loss": 3.5274,
       "step": 24600
     },
     {
-      "epoch": 2.653105155526854,
-      "grad_norm": 0.625645637512207,
-      "learning_rate": 0.0004413403728046547,
-      "loss": 3.5611,
+      "epoch": 2.6576819407008085,
+      "grad_norm": 0.5458420515060425,
+      "learning_rate": 0.0004410793308148948,
+      "loss": 3.5413,
       "step": 24650
     },
     {
-      "epoch": 2.658486707566462,
-      "grad_norm": 0.5856419205665588,
-      "learning_rate": 0.0004410171317745932,
-      "loss": 3.5536,
+      "epoch": 2.6630727762803232,
+      "grad_norm": 0.6068403124809265,
+      "learning_rate": 0.00044075553157042627,
+      "loss": 3.5423,
       "step": 24700
     },
     {
-      "epoch": 2.6638682596060703,
-      "grad_norm": 0.6091840863227844,
-      "learning_rate": 0.0004406938907445318,
-      "loss": 3.541,
+      "epoch": 2.6684636118598384,
+      "grad_norm": 0.6027436256408691,
+      "learning_rate": 0.0004404317323259579,
+      "loss": 3.5459,
       "step": 24750
     },
     {
-      "epoch": 2.6692498116456784,
-      "grad_norm": 0.6365329027175903,
-      "learning_rate": 0.00044037064971447036,
-      "loss": 3.541,
+      "epoch": 2.673854447439353,
+      "grad_norm": 0.5950092673301697,
+      "learning_rate": 0.0004401079330814894,
+      "loss": 3.5422,
       "step": 24800
     },
     {
-      "epoch": 2.674631363685287,
-      "grad_norm": 0.5892279744148254,
-      "learning_rate": 0.00044004740868440896,
-      "loss": 3.5459,
+      "epoch": 2.6792452830188678,
+      "grad_norm": 0.5921434164047241,
+      "learning_rate": 0.000439784133837021,
+      "loss": 3.5461,
       "step": 24850
     },
     {
-      "epoch": 2.680012915724895,
-      "grad_norm": 0.6100271344184875,
-      "learning_rate": 0.00043972416765434755,
-      "loss": 3.5443,
+      "epoch": 2.684636118598383,
+      "grad_norm": 0.647606611251831,
+      "learning_rate": 0.0004394603345925526,
+      "loss": 3.5624,
       "step": 24900
     },
     {
-      "epoch": 2.685394467764503,
-      "grad_norm": 0.5854329466819763,
-      "learning_rate": 0.00043940092662428615,
-      "loss": 3.5453,
+      "epoch": 2.6900269541778976,
+      "grad_norm": 0.5818614363670349,
+      "learning_rate": 0.00043913653534808413,
+      "loss": 3.5514,
       "step": 24950
     },
     {
-      "epoch": 2.6907760198041117,
-      "grad_norm": 0.5759541392326355,
-      "learning_rate": 0.00043907768559422474,
-      "loss": 3.5571,
+      "epoch": 2.6954177897574123,
+      "grad_norm": 0.5758751630783081,
+      "learning_rate": 0.00043881273610361574,
+      "loss": 3.5453,
       "step": 25000
     },
     {
-      "epoch": 2.6907760198041117,
-      "eval_accuracy": 0.36772407294911424,
-      "eval_loss": 3.526393413543701,
-      "eval_runtime": 183.1728,
-      "eval_samples_per_second": 98.328,
-      "eval_steps_per_second": 6.147,
+      "epoch": 2.6954177897574123,
+      "eval_accuracy": 0.3674399456474794,
+      "eval_loss": 3.5206170082092285,
+      "eval_runtime": 184.8565,
+      "eval_samples_per_second": 97.432,
+      "eval_steps_per_second": 6.091,
       "step": 25000
     },
     {
-      "epoch": 2.69615757184372,
-      "grad_norm": 0.5549103021621704,
-      "learning_rate": 0.00043875444456416334,
-      "loss": 3.5413,
+      "epoch": 2.7008086253369274,
+      "grad_norm": 0.6190797686576843,
+      "learning_rate": 0.0004384889368591473,
+      "loss": 3.5342,
       "step": 25050
     },
     {
-      "epoch": 2.701539123883328,
-      "grad_norm": 0.625041663646698,
-      "learning_rate": 0.0004384312035341019,
-      "loss": 3.5329,
+      "epoch": 2.706199460916442,
+      "grad_norm": 0.5533490777015686,
+      "learning_rate": 0.0004381651376146789,
+      "loss": 3.5463,
       "step": 25100
     },
     {
-      "epoch": 2.706920675922936,
-      "grad_norm": 0.6317231059074402,
-      "learning_rate": 0.0004381144273246417,
-      "loss": 3.5234,
+      "epoch": 2.711590296495957,
+      "grad_norm": 0.6489225029945374,
+      "learning_rate": 0.00043784133837021044,
+      "loss": 3.5213,
       "step": 25150
     },
     {
-      "epoch": 2.712302227962544,
-      "grad_norm": 0.5706122517585754,
-      "learning_rate": 0.0004377911862945803,
-      "loss": 3.5438,
+      "epoch": 2.7169811320754715,
+      "grad_norm": 0.633579432964325,
+      "learning_rate": 0.00043751753912574205,
+      "loss": 3.5434,
       "step": 25200
     },
     {
-      "epoch": 2.7176837800021527,
-      "grad_norm": 0.6089178919792175,
-      "learning_rate": 0.00043746794526451887,
-      "loss": 3.5184,
+      "epoch": 2.7223719676549867,
+      "grad_norm": 0.5992454290390015,
+      "learning_rate": 0.0004371937398812736,
+      "loss": 3.549,
       "step": 25250
     },
     {
-      "epoch": 2.723065332041761,
-      "grad_norm": 0.6188173890113831,
-      "learning_rate": 0.00043714470423445747,
-      "loss": 3.5516,
+      "epoch": 2.7277628032345014,
+      "grad_norm": 0.5443644523620605,
+      "learning_rate": 0.0004368699406368051,
+      "loss": 3.5487,
       "step": 25300
     },
     {
-      "epoch": 2.728446884081369,
-      "grad_norm": 0.5653114914894104,
-      "learning_rate": 0.00043682146320439606,
-      "loss": 3.5304,
+      "epoch": 2.733153638814016,
+      "grad_norm": 0.564338207244873,
+      "learning_rate": 0.0004365461413923367,
+      "loss": 3.5385,
       "step": 25350
     },
     {
-      "epoch": 2.7338284361209775,
-      "grad_norm": 0.6082587838172913,
-      "learning_rate": 0.00043649822217433466,
-      "loss": 3.5538,
+      "epoch": 2.7385444743935308,
+      "grad_norm": 0.678646981716156,
+      "learning_rate": 0.00043622234214786825,
+      "loss": 3.5304,
       "step": 25400
     },
     {
-      "epoch": 2.7392099881605856,
-      "grad_norm": 0.5532857179641724,
-      "learning_rate": 0.0004361749811442732,
-      "loss": 3.5193,
+      "epoch": 2.743935309973046,
+      "grad_norm": 0.5924320220947266,
+      "learning_rate": 0.00043589854290339985,
+      "loss": 3.5338,
       "step": 25450
     },
     {
-      "epoch": 2.7445915402001937,
-      "grad_norm": 0.6214589476585388,
-      "learning_rate": 0.0004358517401142118,
-      "loss": 3.536,
+      "epoch": 2.7493261455525606,
+      "grad_norm": 0.6244056820869446,
+      "learning_rate": 0.0004355747436589314,
+      "loss": 3.5458,
       "step": 25500
     },
     {
-      "epoch": 2.749973092239802,
-      "grad_norm": 0.5975282192230225,
-      "learning_rate": 0.00043552849908415033,
-      "loss": 3.5301,
+      "epoch": 2.7547169811320753,
+      "grad_norm": 0.572502851486206,
+      "learning_rate": 0.000435250944414463,
+      "loss": 3.5406,
       "step": 25550
     },
     {
-      "epoch": 2.7553546442794103,
-      "grad_norm": 0.6577773094177246,
-      "learning_rate": 0.000435205258054089,
-      "loss": 3.5566,
+      "epoch": 2.7601078167115904,
+      "grad_norm": 0.5817327499389648,
+      "learning_rate": 0.00043492714516999456,
+      "loss": 3.5418,
       "step": 25600
     },
     {
-      "epoch": 2.7607361963190185,
-      "grad_norm": 0.5862059593200684,
-      "learning_rate": 0.0004348820170240276,
-      "loss": 3.5315,
+      "epoch": 2.765498652291105,
+      "grad_norm": 0.5926405191421509,
+      "learning_rate": 0.0004346033459255261,
+      "loss": 3.5349,
       "step": 25650
     },
     {
-      "epoch": 2.7661177483586266,
-      "grad_norm": 0.6143049001693726,
-      "learning_rate": 0.0004345587759939661,
-      "loss": 3.5541,
+      "epoch": 2.77088948787062,
+      "grad_norm": 0.6240580081939697,
+      "learning_rate": 0.0004342795466810577,
+      "loss": 3.5396,
       "step": 25700
     },
     {
-      "epoch": 2.7714993003982347,
-      "grad_norm": 0.5826513767242432,
-      "learning_rate": 0.0004342355349639047,
-      "loss": 3.5328,
+      "epoch": 2.776280323450135,
+      "grad_norm": 0.5528091788291931,
+      "learning_rate": 0.00043395574743658927,
+      "loss": 3.5291,
       "step": 25750
     },
     {
-      "epoch": 2.776880852437843,
-      "grad_norm": 0.6236239075660706,
-      "learning_rate": 0.0004339122939338433,
-      "loss": 3.5282,
+      "epoch": 2.7816711590296497,
+      "grad_norm": 0.6151363849639893,
+      "learning_rate": 0.00043363194819212087,
+      "loss": 3.5518,
       "step": 25800
     },
     {
-      "epoch": 2.7822624044774513,
-      "grad_norm": 0.5690028667449951,
-      "learning_rate": 0.00043358905290378184,
-      "loss": 3.5382,
+      "epoch": 2.7870619946091644,
+      "grad_norm": 0.6167548298835754,
+      "learning_rate": 0.0004333081489476524,
+      "loss": 3.5245,
       "step": 25850
     },
     {
-      "epoch": 2.7876439565170594,
-      "grad_norm": 0.630249559879303,
-      "learning_rate": 0.0004332658118737205,
-      "loss": 3.5471,
+      "epoch": 2.7924528301886795,
+      "grad_norm": 0.5798197984695435,
+      "learning_rate": 0.000432984349703184,
+      "loss": 3.5336,
       "step": 25900
     },
     {
-      "epoch": 2.793025508556668,
-      "grad_norm": 0.561437726020813,
-      "learning_rate": 0.0004329425708436591,
-      "loss": 3.5452,
+      "epoch": 2.797843665768194,
+      "grad_norm": 0.6130890250205994,
+      "learning_rate": 0.0004326605504587155,
+      "loss": 3.5323,
       "step": 25950
     },
     {
-      "epoch": 2.798407060596276,
-      "grad_norm": 0.5968010425567627,
-      "learning_rate": 0.00043261932981359763,
-      "loss": 3.5268,
+      "epoch": 2.803234501347709,
+      "grad_norm": 0.546591579914093,
+      "learning_rate": 0.0004323367512142472,
+      "loss": 3.5492,
       "step": 26000
     },
     {
-      "epoch": 2.798407060596276,
-      "eval_accuracy": 0.3683658855574344,
-      "eval_loss": 3.5156972408294678,
-      "eval_runtime": 183.4726,
-      "eval_samples_per_second": 98.167,
-      "eval_steps_per_second": 6.137,
+      "epoch": 2.803234501347709,
+      "eval_accuracy": 0.3684034794566182,
+      "eval_loss": 3.512904167175293,
+      "eval_runtime": 184.0153,
+      "eval_samples_per_second": 97.878,
+      "eval_steps_per_second": 6.119,
       "step": 26000
     },
     {
-      "epoch": 2.803788612635884,
-      "grad_norm": 0.5962703227996826,
-      "learning_rate": 0.0004322960887835362,
-      "loss": 3.5435,
+      "epoch": 2.8086253369272236,
+      "grad_norm": 0.598056972026825,
+      "learning_rate": 0.0004320129519697787,
+      "loss": 3.5518,
       "step": 26050
     },
     {
-      "epoch": 2.8091701646754923,
-      "grad_norm": 0.5704419016838074,
-      "learning_rate": 0.00043197284775347476,
-      "loss": 3.5439,
+      "epoch": 2.8140161725067383,
+      "grad_norm": 0.589799702167511,
+      "learning_rate": 0.00043168915272531023,
+      "loss": 3.5381,
       "step": 26100
     },
     {
-      "epoch": 2.8145517167151004,
-      "grad_norm": 0.6311613917350769,
-      "learning_rate": 0.00043164960672341336,
-      "loss": 3.5348,
+      "epoch": 2.8194070080862534,
+      "grad_norm": 0.5865911245346069,
+      "learning_rate": 0.00043136535348084183,
+      "loss": 3.5325,
       "step": 26150
     },
     {
-      "epoch": 2.819933268754709,
-      "grad_norm": 0.5643249154090881,
-      "learning_rate": 0.00043133283051395317,
-      "loss": 3.5529,
+      "epoch": 2.824797843665768,
+      "grad_norm": 0.5646501183509827,
+      "learning_rate": 0.0004310415542363734,
+      "loss": 3.5463,
       "step": 26200
     },
     {
-      "epoch": 2.825314820794317,
-      "grad_norm": 0.6420338749885559,
-      "learning_rate": 0.0004310095894838918,
-      "loss": 3.545,
+      "epoch": 2.830188679245283,
+      "grad_norm": 0.5588135719299316,
+      "learning_rate": 0.000430717754991905,
+      "loss": 3.5267,
       "step": 26250
     },
     {
-      "epoch": 2.830696372833925,
-      "grad_norm": 0.6328233480453491,
-      "learning_rate": 0.0004306863484538304,
-      "loss": 3.5405,
+      "epoch": 2.835579514824798,
+      "grad_norm": 0.6025974750518799,
+      "learning_rate": 0.00043039395574743654,
+      "loss": 3.5267,
       "step": 26300
     },
     {
-      "epoch": 2.8360779248735337,
-      "grad_norm": 0.6627986431121826,
-      "learning_rate": 0.00043036310742376895,
-      "loss": 3.5228,
+      "epoch": 2.8409703504043127,
+      "grad_norm": 0.6605818271636963,
+      "learning_rate": 0.00043007015650296814,
+      "loss": 3.5478,
       "step": 26350
     },
     {
-      "epoch": 2.841459476913142,
-      "grad_norm": 0.5869179964065552,
-      "learning_rate": 0.00043003986639370754,
-      "loss": 3.5245,
+      "epoch": 2.8463611859838274,
+      "grad_norm": 0.569876492023468,
+      "learning_rate": 0.0004297463572584997,
+      "loss": 3.5241,
       "step": 26400
     },
     {
-      "epoch": 2.84684102895275,
-      "grad_norm": 0.5820196270942688,
-      "learning_rate": 0.0004297166253636461,
-      "loss": 3.5315,
+      "epoch": 2.8517520215633425,
+      "grad_norm": 0.5919357538223267,
+      "learning_rate": 0.0004294225580140313,
+      "loss": 3.543,
       "step": 26450
     },
     {
-      "epoch": 2.852222580992358,
-      "grad_norm": 0.6153207421302795,
-      "learning_rate": 0.0004293933843335847,
-      "loss": 3.5206,
+      "epoch": 2.857142857142857,
+      "grad_norm": 0.5494198799133301,
+      "learning_rate": 0.00042909875876956285,
+      "loss": 3.5505,
       "step": 26500
     },
     {
-      "epoch": 2.857604133031966,
-      "grad_norm": 0.5706645846366882,
-      "learning_rate": 0.00042907014330352333,
-      "loss": 3.5492,
+      "epoch": 2.862533692722372,
+      "grad_norm": 0.6157829761505127,
+      "learning_rate": 0.0004287749595250944,
+      "loss": 3.5316,
       "step": 26550
     },
     {
-      "epoch": 2.8629856850715747,
-      "grad_norm": 0.5671915411949158,
-      "learning_rate": 0.00042874690227346187,
-      "loss": 3.5261,
+      "epoch": 2.867924528301887,
+      "grad_norm": 0.564264178276062,
+      "learning_rate": 0.000428451160280626,
+      "loss": 3.5341,
       "step": 26600
     },
     {
-      "epoch": 2.868367237111183,
-      "grad_norm": 0.6378110647201538,
-      "learning_rate": 0.00042842366124340046,
-      "loss": 3.5554,
+      "epoch": 2.8733153638814017,
+      "grad_norm": 0.5835041999816895,
+      "learning_rate": 0.0004281273610361575,
+      "loss": 3.5176,
       "step": 26650
     },
     {
-      "epoch": 2.873748789150791,
-      "grad_norm": 0.5837578773498535,
-      "learning_rate": 0.00042810042021333906,
-      "loss": 3.5286,
+      "epoch": 2.8787061994609164,
+      "grad_norm": 0.6067073941230774,
+      "learning_rate": 0.0004278035617916891,
+      "loss": 3.5572,
       "step": 26700
     },
     {
-      "epoch": 2.8791303411903995,
-      "grad_norm": 0.5720592737197876,
-      "learning_rate": 0.0004277771791832776,
-      "loss": 3.5193,
+      "epoch": 2.884097035040431,
+      "grad_norm": 0.6081402897834778,
+      "learning_rate": 0.00042747976254722066,
+      "loss": 3.5249,
       "step": 26750
     },
     {
-      "epoch": 2.8845118932300076,
-      "grad_norm": 0.6068984270095825,
-      "learning_rate": 0.0004274539381532162,
-      "loss": 3.5375,
+      "epoch": 2.889487870619946,
+      "grad_norm": 0.5519744157791138,
+      "learning_rate": 0.00042715596330275226,
+      "loss": 3.5073,
       "step": 26800
     },
     {
-      "epoch": 2.8898934452696157,
-      "grad_norm": 0.590977132320404,
-      "learning_rate": 0.00042713069712315484,
-      "loss": 3.515,
+      "epoch": 2.894878706199461,
+      "grad_norm": 0.6072061061859131,
+      "learning_rate": 0.0004268321640582838,
+      "loss": 3.512,
       "step": 26850
     },
     {
-      "epoch": 2.895274997309224,
-      "grad_norm": 0.6040093302726746,
-      "learning_rate": 0.0004268074560930934,
-      "loss": 3.5472,
+      "epoch": 2.9002695417789757,
+      "grad_norm": 0.588100016117096,
+      "learning_rate": 0.0004265083648138154,
+      "loss": 3.5318,
       "step": 26900
     },
     {
-      "epoch": 2.9006565493488323,
-      "grad_norm": 0.5838053822517395,
-      "learning_rate": 0.000426484215063032,
-      "loss": 3.5313,
+      "epoch": 2.9056603773584904,
+      "grad_norm": 0.6119621992111206,
+      "learning_rate": 0.00042618456556934697,
+      "loss": 3.5509,
       "step": 26950
     },
     {
-      "epoch": 2.9060381013884404,
-      "grad_norm": 0.6135042905807495,
-      "learning_rate": 0.0004261609740329705,
-      "loss": 3.5395,
+      "epoch": 2.9110512129380055,
+      "grad_norm": 0.5784225463867188,
+      "learning_rate": 0.0004258607663248785,
+      "loss": 3.5339,
       "step": 27000
     },
     {
-      "epoch": 2.9060381013884404,
-      "eval_accuracy": 0.3692769400217588,
-      "eval_loss": 3.50553035736084,
-      "eval_runtime": 183.0394,
-      "eval_samples_per_second": 98.4,
-      "eval_steps_per_second": 6.152,
+      "epoch": 2.9110512129380055,
+      "eval_accuracy": 0.3697280668119646,
+      "eval_loss": 3.505253553390503,
+      "eval_runtime": 184.5132,
+      "eval_samples_per_second": 97.614,
+      "eval_steps_per_second": 6.103,
       "step": 27000
     },
     {
-      "epoch": 2.9114196534280485,
-      "grad_norm": 0.6675634384155273,
-      "learning_rate": 0.0004258377330029091,
-      "loss": 3.5392,
+      "epoch": 2.91644204851752,
+      "grad_norm": 0.6382215023040771,
+      "learning_rate": 0.0004255369670804101,
+      "loss": 3.5213,
       "step": 27050
     },
     {
-      "epoch": 2.9168012054676566,
-      "grad_norm": 0.5531997680664062,
-      "learning_rate": 0.00042551449197284776,
-      "loss": 3.5078,
+      "epoch": 2.921832884097035,
+      "grad_norm": 0.6182580590248108,
+      "learning_rate": 0.0004252131678359417,
+      "loss": 3.5302,
       "step": 27100
     },
     {
-      "epoch": 2.922182757507265,
-      "grad_norm": 0.6126731634140015,
-      "learning_rate": 0.0004251912509427863,
-      "loss": 3.5329,
+      "epoch": 2.92722371967655,
+      "grad_norm": 0.5992064476013184,
+      "learning_rate": 0.0004248893685914733,
+      "loss": 3.5248,
       "step": 27150
     },
     {
-      "epoch": 2.9275643095468733,
-      "grad_norm": 0.5538141131401062,
-      "learning_rate": 0.0004248680099127249,
-      "loss": 3.5216,
+      "epoch": 2.9326145552560647,
+      "grad_norm": 0.6005051732063293,
+      "learning_rate": 0.00042456556934700483,
+      "loss": 3.5349,
       "step": 27200
     },
     {
-      "epoch": 2.9329458615864814,
-      "grad_norm": 0.5796554088592529,
-      "learning_rate": 0.0004245447688826635,
-      "loss": 3.5247,
+      "epoch": 2.9380053908355794,
+      "grad_norm": 0.57536381483078,
+      "learning_rate": 0.00042424177010253643,
+      "loss": 3.5402,
       "step": 27250
     },
     {
-      "epoch": 2.93832741362609,
-      "grad_norm": 0.5609702467918396,
-      "learning_rate": 0.00042422152785260203,
-      "loss": 3.5229,
+      "epoch": 2.9433962264150946,
+      "grad_norm": 0.611814022064209,
+      "learning_rate": 0.00042391797085806793,
+      "loss": 3.5058,
       "step": 27300
     },
     {
-      "epoch": 2.943708965665698,
-      "grad_norm": 0.6337541937828064,
-      "learning_rate": 0.0004238982868225406,
-      "loss": 3.4924,
+      "epoch": 2.9487870619946093,
+      "grad_norm": 0.5960366725921631,
+      "learning_rate": 0.0004235941716135995,
+      "loss": 3.5244,
       "step": 27350
     },
     {
-      "epoch": 2.949090517705306,
-      "grad_norm": 0.6238342523574829,
-      "learning_rate": 0.0004235750457924793,
-      "loss": 3.5259,
+      "epoch": 2.954177897574124,
+      "grad_norm": 0.634046733379364,
+      "learning_rate": 0.0004232703723691311,
+      "loss": 3.518,
       "step": 27400
     },
     {
-      "epoch": 2.9544720697449143,
-      "grad_norm": 0.707619309425354,
-      "learning_rate": 0.0004232518047624178,
-      "loss": 3.5301,
+      "epoch": 2.9595687331536387,
+      "grad_norm": 0.5948870778083801,
+      "learning_rate": 0.00042294657312466264,
+      "loss": 3.5141,
       "step": 27450
     },
     {
-      "epoch": 2.9598536217845224,
-      "grad_norm": 0.5899302959442139,
-      "learning_rate": 0.0004229285637323564,
-      "loss": 3.5246,
+      "epoch": 2.964959568733154,
+      "grad_norm": 0.5595338940620422,
+      "learning_rate": 0.00042262277388019424,
+      "loss": 3.5382,
       "step": 27500
     },
     {
-      "epoch": 2.965235173824131,
-      "grad_norm": 0.552964985370636,
-      "learning_rate": 0.00042260532270229495,
-      "loss": 3.5266,
+      "epoch": 2.9703504043126685,
+      "grad_norm": 0.6144919395446777,
+      "learning_rate": 0.0004222989746357258,
+      "loss": 3.5399,
       "step": 27550
     },
     {
-      "epoch": 2.970616725863739,
-      "grad_norm": 0.5927809476852417,
-      "learning_rate": 0.00042228208167223354,
-      "loss": 3.5179,
+      "epoch": 2.975741239892183,
+      "grad_norm": 0.5988566875457764,
+      "learning_rate": 0.0004219751753912574,
+      "loss": 3.5205,
       "step": 27600
     },
     {
-      "epoch": 2.975998277903347,
-      "grad_norm": 0.6026394367218018,
-      "learning_rate": 0.00042195884064217214,
-      "loss": 3.531,
+      "epoch": 2.981132075471698,
+      "grad_norm": 0.5763267278671265,
+      "learning_rate": 0.00042165785213167835,
+      "loss": 3.527,
       "step": 27650
     },
     {
-      "epoch": 2.9813798299429557,
-      "grad_norm": 0.6079704165458679,
-      "learning_rate": 0.00042163559961211073,
-      "loss": 3.5447,
+      "epoch": 2.986522911051213,
+      "grad_norm": 0.5751621127128601,
+      "learning_rate": 0.0004213340528872099,
+      "loss": 3.5157,
       "step": 27700
     },
     {
-      "epoch": 2.986761381982564,
-      "grad_norm": 0.5754170417785645,
-      "learning_rate": 0.00042131235858204933,
-      "loss": 3.5106,
+      "epoch": 2.9919137466307277,
+      "grad_norm": 0.6504539251327515,
+      "learning_rate": 0.00042101025364274145,
+      "loss": 3.5171,
       "step": 27750
     },
     {
-      "epoch": 2.992142934022172,
-      "grad_norm": 0.5957307815551758,
-      "learning_rate": 0.0004209891175519879,
-      "loss": 3.5292,
+      "epoch": 2.9973045822102424,
+      "grad_norm": 0.5843790173530579,
+      "learning_rate": 0.00042068645439827305,
+      "loss": 3.5253,
       "step": 27800
     },
     {
-      "epoch": 2.9975244860617805,
-      "grad_norm": 0.5584477782249451,
-      "learning_rate": 0.00042066587652192646,
-      "loss": 3.517,
+      "epoch": 3.0026954177897576,
+      "grad_norm": 0.5703924894332886,
+      "learning_rate": 0.0004203626551538046,
+      "loss": 3.4851,
       "step": 27850
     },
     {
-      "epoch": 3.0029060381013886,
-      "grad_norm": 0.5913463234901428,
-      "learning_rate": 0.00042034263549186506,
-      "loss": 3.4829,
+      "epoch": 3.0080862533692723,
+      "grad_norm": 0.5999839901924133,
+      "learning_rate": 0.0004200388559093362,
+      "loss": 3.4307,
       "step": 27900
     },
     {
-      "epoch": 3.0082875901409967,
-      "grad_norm": 0.5728430151939392,
-      "learning_rate": 0.0004200193944618036,
-      "loss": 3.4424,
+      "epoch": 3.013477088948787,
+      "grad_norm": 0.6257562041282654,
+      "learning_rate": 0.0004197150566648677,
+      "loss": 3.4316,
       "step": 27950
     },
     {
-      "epoch": 3.0136691421806048,
-      "grad_norm": 0.6232100129127502,
-      "learning_rate": 0.00041969615343174225,
-      "loss": 3.4403,
+      "epoch": 3.018867924528302,
+      "grad_norm": 0.5694010257720947,
+      "learning_rate": 0.00041939125742039936,
+      "loss": 3.4321,
       "step": 28000
     },
     {
-      "epoch": 3.0136691421806048,
-      "eval_accuracy": 0.3704998282741108,
-      "eval_loss": 3.501554250717163,
-      "eval_runtime": 181.6654,
-      "eval_samples_per_second": 99.144,
-      "eval_steps_per_second": 6.198,
+      "epoch": 3.018867924528302,
+      "eval_accuracy": 0.3699212516464987,
+      "eval_loss": 3.4992401599884033,
+      "eval_runtime": 184.9672,
+      "eval_samples_per_second": 97.374,
+      "eval_steps_per_second": 6.088,
       "step": 28000
     },
     {
-      "epoch": 3.0190506942202133,
-      "grad_norm": 0.5570206046104431,
-      "learning_rate": 0.00041937291240168084,
-      "loss": 3.436,
+      "epoch": 3.024258760107817,
+      "grad_norm": 0.609790563583374,
+      "learning_rate": 0.00041906745817593086,
+      "loss": 3.4316,
       "step": 28050
     },
     {
-      "epoch": 3.0244322462598214,
-      "grad_norm": 0.6119909882545471,
-      "learning_rate": 0.0004190496713716194,
-      "loss": 3.4374,
+      "epoch": 3.0296495956873315,
+      "grad_norm": 0.6133913993835449,
+      "learning_rate": 0.00041874365893146247,
+      "loss": 3.435,
       "step": 28100
     },
     {
-      "epoch": 3.0298137982994295,
-      "grad_norm": 0.6302096843719482,
-      "learning_rate": 0.000418726430341558,
-      "loss": 3.4448,
+      "epoch": 3.035040431266846,
+      "grad_norm": 0.5948244333267212,
+      "learning_rate": 0.000418419859686994,
+      "loss": 3.4219,
       "step": 28150
     },
     {
-      "epoch": 3.0351953503390376,
-      "grad_norm": 0.5777623653411865,
-      "learning_rate": 0.00041840318931149657,
-      "loss": 3.4356,
+      "epoch": 3.0404312668463613,
+      "grad_norm": 0.5920873284339905,
+      "learning_rate": 0.00041809606044252557,
+      "loss": 3.4391,
       "step": 28200
     },
     {
-      "epoch": 3.040576902378646,
-      "grad_norm": 0.5832511782646179,
-      "learning_rate": 0.00041807994828143517,
-      "loss": 3.4348,
+      "epoch": 3.045822102425876,
+      "grad_norm": 0.5651567578315735,
+      "learning_rate": 0.00041777226119805717,
+      "loss": 3.4298,
       "step": 28250
     },
     {
-      "epoch": 3.0459584544182543,
-      "grad_norm": 0.6186639666557312,
-      "learning_rate": 0.00041775670725137376,
-      "loss": 3.4531,
+      "epoch": 3.0512129380053907,
+      "grad_norm": 0.5934497714042664,
+      "learning_rate": 0.0004174484619535887,
+      "loss": 3.4544,
       "step": 28300
     },
     {
-      "epoch": 3.0513400064578624,
-      "grad_norm": 0.5793587565422058,
-      "learning_rate": 0.00041743346622131236,
-      "loss": 3.4351,
+      "epoch": 3.056603773584906,
+      "grad_norm": 0.5851883888244629,
+      "learning_rate": 0.00041712466270912033,
+      "loss": 3.4393,
       "step": 28350
     },
     {
-      "epoch": 3.0567215584974705,
-      "grad_norm": 0.6191322207450867,
-      "learning_rate": 0.0004171102251912509,
-      "loss": 3.4609,
+      "epoch": 3.0619946091644206,
+      "grad_norm": 0.5736149549484253,
+      "learning_rate": 0.0004168008634646519,
+      "loss": 3.4601,
       "step": 28400
     },
     {
-      "epoch": 3.062103110537079,
-      "grad_norm": 0.5854801535606384,
-      "learning_rate": 0.0004167869841611895,
-      "loss": 3.445,
+      "epoch": 3.0673854447439353,
+      "grad_norm": 0.585663378238678,
+      "learning_rate": 0.0004164770642201835,
+      "loss": 3.4376,
       "step": 28450
     },
     {
-      "epoch": 3.067484662576687,
-      "grad_norm": 0.5954617857933044,
-      "learning_rate": 0.00041646374313112803,
-      "loss": 3.4516,
+      "epoch": 3.07277628032345,
+      "grad_norm": 0.5987605452537537,
+      "learning_rate": 0.00041615326497571503,
+      "loss": 3.4309,
       "step": 28500
     },
     {
-      "epoch": 3.0728662146162953,
-      "grad_norm": 0.6144933104515076,
-      "learning_rate": 0.0004161405021010667,
-      "loss": 3.4306,
+      "epoch": 3.078167115902965,
+      "grad_norm": 0.5901303887367249,
+      "learning_rate": 0.0004158294657312466,
+      "loss": 3.4319,
       "step": 28550
     },
     {
-      "epoch": 3.0782477666559034,
-      "grad_norm": 0.6332607269287109,
-      "learning_rate": 0.0004158172610710053,
-      "loss": 3.4483,
+      "epoch": 3.08355795148248,
+      "grad_norm": 0.623777449131012,
+      "learning_rate": 0.0004155056664867782,
+      "loss": 3.4517,
       "step": 28600
     },
     {
-      "epoch": 3.083629318695512,
-      "grad_norm": 0.6351560354232788,
-      "learning_rate": 0.0004154940200409438,
-      "loss": 3.4597,
+      "epoch": 3.0889487870619945,
+      "grad_norm": 0.653161346912384,
+      "learning_rate": 0.0004151818672423097,
+      "loss": 3.4379,
       "step": 28650
     },
     {
-      "epoch": 3.08901087073512,
-      "grad_norm": 0.580599308013916,
-      "learning_rate": 0.0004151707790108824,
-      "loss": 3.4463,
+      "epoch": 3.0943396226415096,
+      "grad_norm": 0.5761669874191284,
+      "learning_rate": 0.0004148580679978413,
+      "loss": 3.4587,
       "step": 28700
     },
     {
-      "epoch": 3.094392422774728,
-      "grad_norm": 0.636760950088501,
-      "learning_rate": 0.00041484753798082095,
-      "loss": 3.4619,
+      "epoch": 3.0997304582210243,
+      "grad_norm": 0.612464189529419,
+      "learning_rate": 0.00041453426875337284,
+      "loss": 3.424,
       "step": 28750
     },
     {
-      "epoch": 3.0997739748143363,
-      "grad_norm": 0.6538554430007935,
-      "learning_rate": 0.00041452429695075955,
-      "loss": 3.448,
+      "epoch": 3.105121293800539,
+      "grad_norm": 0.5614286661148071,
+      "learning_rate": 0.00041421046950890445,
+      "loss": 3.4539,
       "step": 28800
     },
     {
-      "epoch": 3.105155526853945,
-      "grad_norm": 0.5733194351196289,
-      "learning_rate": 0.0004142010559206982,
-      "loss": 3.4343,
+      "epoch": 3.1105121293800537,
+      "grad_norm": 0.6434953808784485,
+      "learning_rate": 0.000413886670264436,
+      "loss": 3.4426,
       "step": 28850
     },
     {
-      "epoch": 3.110537078893553,
-      "grad_norm": 0.6369328498840332,
-      "learning_rate": 0.00041387781489063673,
-      "loss": 3.4369,
+      "epoch": 3.115902964959569,
+      "grad_norm": 0.5817499160766602,
+      "learning_rate": 0.0004135628710199676,
+      "loss": 3.4565,
       "step": 28900
     },
     {
-      "epoch": 3.115918630933161,
-      "grad_norm": 0.6201010942459106,
-      "learning_rate": 0.00041355457386057533,
-      "loss": 3.453,
+      "epoch": 3.1212938005390836,
+      "grad_norm": 0.6030315160751343,
+      "learning_rate": 0.00041323907177549915,
+      "loss": 3.4534,
       "step": 28950
     },
     {
-      "epoch": 3.121300182972769,
-      "grad_norm": 0.6407283544540405,
-      "learning_rate": 0.0004132313328305139,
-      "loss": 3.4636,
+      "epoch": 3.1266846361185983,
+      "grad_norm": 0.6916279196739197,
+      "learning_rate": 0.0004129152725310307,
+      "loss": 3.4498,
       "step": 29000
     },
     {
-      "epoch": 3.121300182972769,
-      "eval_accuracy": 0.3711315361638642,
-      "eval_loss": 3.497187614440918,
-      "eval_runtime": 181.453,
-      "eval_samples_per_second": 99.26,
-      "eval_steps_per_second": 6.205,
+      "epoch": 3.1266846361185983,
+      "eval_accuracy": 0.3706708479195308,
+      "eval_loss": 3.4956727027893066,
+      "eval_runtime": 185.4027,
+      "eval_samples_per_second": 97.145,
+      "eval_steps_per_second": 6.073,
       "step": 29000
     },
     {
-      "epoch": 3.1266817350123777,
-      "grad_norm": 0.6396225690841675,
-      "learning_rate": 0.00041290809180045246,
-      "loss": 3.4613,
+      "epoch": 3.1320754716981134,
+      "grad_norm": 0.6097031235694885,
+      "learning_rate": 0.0004125914732865623,
+      "loss": 3.455,
       "step": 29050
     },
     {
-      "epoch": 3.132063287051986,
-      "grad_norm": 0.6482574939727783,
-      "learning_rate": 0.0004125848507703911,
-      "loss": 3.4346,
+      "epoch": 3.137466307277628,
+      "grad_norm": 0.5889896750450134,
+      "learning_rate": 0.00041226767404209386,
+      "loss": 3.4526,
       "step": 29100
     },
     {
-      "epoch": 3.137444839091594,
-      "grad_norm": 0.6442139744758606,
-      "learning_rate": 0.0004122616097403297,
-      "loss": 3.4496,
+      "epoch": 3.142857142857143,
+      "grad_norm": 0.6619573831558228,
+      "learning_rate": 0.00041194387479762546,
+      "loss": 3.4564,
       "step": 29150
     },
     {
-      "epoch": 3.1428263911312024,
-      "grad_norm": 0.5868032574653625,
-      "learning_rate": 0.0004119448335308695,
-      "loss": 3.4653,
+      "epoch": 3.1482479784366575,
+      "grad_norm": 0.5755789875984192,
+      "learning_rate": 0.000411620075553157,
+      "loss": 3.4534,
       "step": 29200
     },
     {
-      "epoch": 3.1482079431708105,
-      "grad_norm": 0.8085774779319763,
-      "learning_rate": 0.0004116215925008081,
-      "loss": 3.4591,
+      "epoch": 3.1536388140161726,
+      "grad_norm": 0.5996595025062561,
+      "learning_rate": 0.0004112962763086886,
+      "loss": 3.4721,
       "step": 29250
     },
     {
-      "epoch": 3.1535894952104186,
-      "grad_norm": 0.5951862931251526,
-      "learning_rate": 0.00041129835147074665,
-      "loss": 3.4455,
+      "epoch": 3.1590296495956873,
+      "grad_norm": 0.6542275547981262,
+      "learning_rate": 0.0004109724770642201,
+      "loss": 3.4562,
       "step": 29300
     },
     {
-      "epoch": 3.1589710472500268,
-      "grad_norm": 0.6378307938575745,
-      "learning_rate": 0.00041097511044068524,
-      "loss": 3.4538,
+      "epoch": 3.164420485175202,
+      "grad_norm": 0.654630720615387,
+      "learning_rate": 0.00041064867781975177,
+      "loss": 3.4519,
       "step": 29350
     },
     {
-      "epoch": 3.1643525992896353,
-      "grad_norm": 0.6090945601463318,
-      "learning_rate": 0.0004106518694106238,
-      "loss": 3.448,
+      "epoch": 3.169811320754717,
+      "grad_norm": 0.6278281807899475,
+      "learning_rate": 0.00041032487857528327,
+      "loss": 3.4478,
       "step": 29400
     },
     {
-      "epoch": 3.1697341513292434,
-      "grad_norm": 0.6292023062705994,
-      "learning_rate": 0.0004103286283805624,
-      "loss": 3.4687,
+      "epoch": 3.175202156334232,
+      "grad_norm": 0.5717431306838989,
+      "learning_rate": 0.0004100010793308148,
+      "loss": 3.4575,
       "step": 29450
     },
     {
-      "epoch": 3.1751157033688515,
-      "grad_norm": 0.6239890456199646,
-      "learning_rate": 0.00041000538735050103,
-      "loss": 3.4909,
+      "epoch": 3.1805929919137466,
+      "grad_norm": 0.6188814640045166,
+      "learning_rate": 0.0004096772800863464,
+      "loss": 3.4531,
       "step": 29500
     },
     {
-      "epoch": 3.1804972554084596,
-      "grad_norm": 0.5989189147949219,
-      "learning_rate": 0.00040968214632043957,
-      "loss": 3.4548,
+      "epoch": 3.1859838274932613,
+      "grad_norm": 0.6256982684135437,
+      "learning_rate": 0.000409353480841878,
+      "loss": 3.4639,
       "step": 29550
     },
     {
-      "epoch": 3.185878807448068,
-      "grad_norm": 0.6178577542304993,
-      "learning_rate": 0.00040935890529037816,
-      "loss": 3.4575,
+      "epoch": 3.1913746630727764,
+      "grad_norm": 0.605104923248291,
+      "learning_rate": 0.0004090296815974096,
+      "loss": 3.46,
       "step": 29600
     },
     {
-      "epoch": 3.1912603594876763,
-      "grad_norm": 0.6444182395935059,
-      "learning_rate": 0.00040903566426031676,
-      "loss": 3.4525,
+      "epoch": 3.196765498652291,
+      "grad_norm": 0.6374660730361938,
+      "learning_rate": 0.00040870588235294113,
+      "loss": 3.4476,
       "step": 29650
     },
     {
-      "epoch": 3.1966419115272844,
-      "grad_norm": 0.6180800199508667,
-      "learning_rate": 0.0004087124232302553,
-      "loss": 3.436,
+      "epoch": 3.202156334231806,
+      "grad_norm": 0.5780457258224487,
+      "learning_rate": 0.00040838208310847273,
+      "loss": 3.4792,
       "step": 29700
     },
     {
-      "epoch": 3.2020234635668925,
-      "grad_norm": 0.595629096031189,
-      "learning_rate": 0.0004083891822001939,
-      "loss": 3.4835,
+      "epoch": 3.207547169811321,
+      "grad_norm": 0.6485940217971802,
+      "learning_rate": 0.0004080582838640043,
+      "loss": 3.4503,
       "step": 29750
     },
     {
-      "epoch": 3.207405015606501,
-      "grad_norm": 0.6241080164909363,
-      "learning_rate": 0.00040806594117013254,
-      "loss": 3.454,
+      "epoch": 3.2129380053908356,
+      "grad_norm": 0.6168989539146423,
+      "learning_rate": 0.0004077344846195359,
+      "loss": 3.4598,
       "step": 29800
     },
     {
-      "epoch": 3.212786567646109,
-      "grad_norm": 0.6285580992698669,
-      "learning_rate": 0.0004077427001400711,
-      "loss": 3.4568,
+      "epoch": 3.2183288409703503,
+      "grad_norm": 0.6741005182266235,
+      "learning_rate": 0.00040741068537506744,
+      "loss": 3.4385,
       "step": 29850
     },
     {
-      "epoch": 3.2181681196857173,
-      "grad_norm": 0.648627758026123,
-      "learning_rate": 0.0004074194591100097,
-      "loss": 3.441,
+      "epoch": 3.223719676549865,
+      "grad_norm": 0.616025984287262,
+      "learning_rate": 0.000407086886130599,
+      "loss": 3.4487,
       "step": 29900
     },
     {
-      "epoch": 3.2235496717253254,
-      "grad_norm": 0.64311283826828,
-      "learning_rate": 0.0004070962180799482,
-      "loss": 3.4699,
+      "epoch": 3.22911051212938,
+      "grad_norm": 0.6147024631500244,
+      "learning_rate": 0.0004067630868861306,
+      "loss": 3.4606,
       "step": 29950
     },
     {
-      "epoch": 3.228931223764934,
-      "grad_norm": 0.5918191075325012,
-      "learning_rate": 0.0004067729770498868,
-      "loss": 3.4348,
+      "epoch": 3.234501347708895,
+      "grad_norm": 0.6112721562385559,
+      "learning_rate": 0.0004064392876416621,
+      "loss": 3.465,
       "step": 30000
     },
     {
-      "epoch": 3.228931223764934,
-      "eval_accuracy": 0.37159700513526145,
-      "eval_loss": 3.49194073677063,
-      "eval_runtime": 181.5347,
-      "eval_samples_per_second": 99.215,
-      "eval_steps_per_second": 6.203,
+      "epoch": 3.234501347708895,
+      "eval_accuracy": 0.37162938369583015,
+      "eval_loss": 3.488260507583618,
+      "eval_runtime": 184.1921,
+      "eval_samples_per_second": 97.784,
+      "eval_steps_per_second": 6.113,
       "step": 30000
     },
     {
-      "epoch": 3.234312775804542,
-      "grad_norm": 0.6755573153495789,
-      "learning_rate": 0.00040644973601982546,
-      "loss": 3.4519,
+      "epoch": 3.2398921832884096,
+      "grad_norm": 0.6223701238632202,
+      "learning_rate": 0.0004061154883971937,
+      "loss": 3.4704,
       "step": 30050
     },
     {
-      "epoch": 3.23969432784415,
-      "grad_norm": 0.6835849285125732,
-      "learning_rate": 0.000406126494989764,
-      "loss": 3.4827,
+      "epoch": 3.2452830188679247,
+      "grad_norm": 0.5669355392456055,
+      "learning_rate": 0.00040579168915272525,
+      "loss": 3.4498,
       "step": 30100
     },
     {
-      "epoch": 3.2450758798837587,
-      "grad_norm": 0.6226943135261536,
-      "learning_rate": 0.0004058032539597026,
-      "loss": 3.4645,
+      "epoch": 3.2506738544474394,
+      "grad_norm": 0.6375906467437744,
+      "learning_rate": 0.00040546788990825685,
+      "loss": 3.4531,
       "step": 30150
     },
     {
-      "epoch": 3.250457431923367,
-      "grad_norm": 0.6609615087509155,
-      "learning_rate": 0.00040548001292964114,
-      "loss": 3.4642,
+      "epoch": 3.256064690026954,
+      "grad_norm": 0.5841410756111145,
+      "learning_rate": 0.0004051440906637884,
+      "loss": 3.4557,
       "step": 30200
     },
     {
-      "epoch": 3.255838983962975,
-      "grad_norm": 0.5687957406044006,
-      "learning_rate": 0.00040515677189957973,
-      "loss": 3.4688,
+      "epoch": 3.2614555256064692,
+      "grad_norm": 0.580535888671875,
+      "learning_rate": 0.00040482029141931995,
+      "loss": 3.4355,
       "step": 30250
     },
     {
-      "epoch": 3.261220536002583,
-      "grad_norm": 0.6109760999679565,
-      "learning_rate": 0.0004048335308695183,
-      "loss": 3.4437,
+      "epoch": 3.266846361185984,
+      "grad_norm": 0.5941153168678284,
+      "learning_rate": 0.00040449649217485156,
+      "loss": 3.4692,
       "step": 30300
     },
     {
-      "epoch": 3.2666020880421915,
-      "grad_norm": 0.5959461331367493,
-      "learning_rate": 0.0004045102898394569,
-      "loss": 3.4724,
+      "epoch": 3.2722371967654986,
+      "grad_norm": 0.6014145016670227,
+      "learning_rate": 0.0004041726929303831,
+      "loss": 3.4524,
       "step": 30350
     },
     {
-      "epoch": 3.2719836400817996,
-      "grad_norm": 0.6271927952766418,
-      "learning_rate": 0.0004041870488093955,
-      "loss": 3.4651,
+      "epoch": 3.2776280323450133,
+      "grad_norm": 0.6263529658317566,
+      "learning_rate": 0.0004038488936859147,
+      "loss": 3.4535,
       "step": 30400
     },
     {
-      "epoch": 3.2773651921214078,
-      "grad_norm": 0.6263259649276733,
-      "learning_rate": 0.0004038638077793341,
-      "loss": 3.4471,
+      "epoch": 3.2830188679245285,
+      "grad_norm": 0.6330785751342773,
+      "learning_rate": 0.00040352509444144626,
+      "loss": 3.4558,
       "step": 30450
     },
     {
-      "epoch": 3.282746744161016,
-      "grad_norm": 0.6434895396232605,
-      "learning_rate": 0.00040354056674927265,
-      "loss": 3.4604,
+      "epoch": 3.288409703504043,
+      "grad_norm": 0.6531413793563843,
+      "learning_rate": 0.00040320129519697787,
+      "loss": 3.4831,
       "step": 30500
     },
     {
-      "epoch": 3.2881282962006244,
-      "grad_norm": 0.661169707775116,
-      "learning_rate": 0.00040321732571921124,
-      "loss": 3.4547,
+      "epoch": 3.293800539083558,
+      "grad_norm": 0.6639565825462341,
+      "learning_rate": 0.0004028774959525094,
+      "loss": 3.4557,
       "step": 30550
     },
     {
-      "epoch": 3.2935098482402325,
-      "grad_norm": 0.683679461479187,
-      "learning_rate": 0.0004028940846891498,
-      "loss": 3.4602,
+      "epoch": 3.2991913746630726,
+      "grad_norm": 0.5905919075012207,
+      "learning_rate": 0.000402553696708041,
+      "loss": 3.4628,
       "step": 30600
     },
     {
-      "epoch": 3.2988914002798406,
-      "grad_norm": 0.6047713756561279,
-      "learning_rate": 0.00040257084365908843,
-      "loss": 3.4512,
+      "epoch": 3.3045822102425877,
+      "grad_norm": 0.6184020638465881,
+      "learning_rate": 0.0004022298974635726,
+      "loss": 3.4531,
       "step": 30650
     },
     {
-      "epoch": 3.304272952319449,
-      "grad_norm": 0.6232369542121887,
-      "learning_rate": 0.00040224760262902703,
-      "loss": 3.4713,
+      "epoch": 3.3099730458221024,
+      "grad_norm": 0.6189979910850525,
+      "learning_rate": 0.00040190609821910407,
+      "loss": 3.454,
       "step": 30700
     },
     {
-      "epoch": 3.3096545043590573,
-      "grad_norm": 0.6141099333763123,
-      "learning_rate": 0.00040192436159896557,
-      "loss": 3.4627,
+      "epoch": 3.315363881401617,
+      "grad_norm": 0.6076200008392334,
+      "learning_rate": 0.0004015822989746357,
+      "loss": 3.4433,
       "step": 30750
     },
     {
-      "epoch": 3.3150360563986654,
-      "grad_norm": 0.6429236531257629,
-      "learning_rate": 0.00040160112056890416,
-      "loss": 3.4586,
+      "epoch": 3.3207547169811322,
+      "grad_norm": 0.6227701306343079,
+      "learning_rate": 0.0004012584997301672,
+      "loss": 3.4587,
       "step": 30800
     },
     {
-      "epoch": 3.3204176084382735,
-      "grad_norm": 0.5798470377922058,
-      "learning_rate": 0.00040127787953884276,
-      "loss": 3.4562,
+      "epoch": 3.326145552560647,
+      "grad_norm": 0.650040328502655,
+      "learning_rate": 0.00040093470048569883,
+      "loss": 3.47,
       "step": 30850
     },
     {
-      "epoch": 3.3257991604778816,
-      "grad_norm": 0.6334143280982971,
-      "learning_rate": 0.00040095463850878135,
-      "loss": 3.4408,
+      "epoch": 3.3315363881401616,
+      "grad_norm": 0.7103980183601379,
+      "learning_rate": 0.0004006109012412304,
+      "loss": 3.4603,
       "step": 30900
     },
     {
-      "epoch": 3.33118071251749,
-      "grad_norm": 0.6591746211051941,
-      "learning_rate": 0.00040063139747871995,
-      "loss": 3.4526,
+      "epoch": 3.3369272237196768,
+      "grad_norm": 0.6596335768699646,
+      "learning_rate": 0.000400287101996762,
+      "loss": 3.4639,
       "step": 30950
     },
     {
-      "epoch": 3.3365622645570983,
-      "grad_norm": 0.6119304895401001,
-      "learning_rate": 0.00040030815644865854,
-      "loss": 3.4676,
+      "epoch": 3.3423180592991915,
+      "grad_norm": 0.6008010506629944,
+      "learning_rate": 0.00039996330275229354,
+      "loss": 3.4539,
       "step": 31000
     },
     {
-      "epoch": 3.3365622645570983,
-      "eval_accuracy": 0.3719315473768422,
-      "eval_loss": 3.486844062805176,
-      "eval_runtime": 182.7259,
-      "eval_samples_per_second": 98.568,
-      "eval_steps_per_second": 6.162,
+      "epoch": 3.3423180592991915,
+      "eval_accuracy": 0.3725362006975298,
+      "eval_loss": 3.4830868244171143,
+      "eval_runtime": 184.0288,
+      "eval_samples_per_second": 97.871,
+      "eval_steps_per_second": 6.119,
       "step": 31000
     },
     {
-      "epoch": 3.3419438165967064,
-      "grad_norm": 0.6275706887245178,
-      "learning_rate": 0.0003999849154185971,
-      "loss": 3.4726,
+      "epoch": 3.347708894878706,
+      "grad_norm": 0.6436234712600708,
+      "learning_rate": 0.00039963950350782514,
+      "loss": 3.4644,
       "step": 31050
     },
     {
-      "epoch": 3.347325368636315,
-      "grad_norm": 0.5887126326560974,
-      "learning_rate": 0.0003996616743885357,
-      "loss": 3.4399,
+      "epoch": 3.353099730458221,
+      "grad_norm": 0.6187503933906555,
+      "learning_rate": 0.0003993157042633567,
+      "loss": 3.4525,
       "step": 31100
     },
     {
-      "epoch": 3.352706920675923,
-      "grad_norm": 0.6963056921958923,
-      "learning_rate": 0.0003993384333584742,
-      "loss": 3.4559,
+      "epoch": 3.358490566037736,
+      "grad_norm": 0.6101654171943665,
+      "learning_rate": 0.00039899190501888824,
+      "loss": 3.4555,
       "step": 31150
     },
     {
-      "epoch": 3.358088472715531,
-      "grad_norm": 0.642962634563446,
-      "learning_rate": 0.00039901519232841287,
-      "loss": 3.4906,
+      "epoch": 3.3638814016172507,
+      "grad_norm": 0.5650094747543335,
+      "learning_rate": 0.00039866810577441985,
+      "loss": 3.4643,
       "step": 31200
     },
     {
-      "epoch": 3.3634700247551392,
-      "grad_norm": 0.6062067747116089,
-      "learning_rate": 0.00039869195129835146,
-      "loss": 3.4634,
+      "epoch": 3.3692722371967654,
+      "grad_norm": 0.6570470929145813,
+      "learning_rate": 0.0003983443065299514,
+      "loss": 3.4533,
       "step": 31250
     },
     {
-      "epoch": 3.368851576794748,
-      "grad_norm": 0.6534258723258972,
-      "learning_rate": 0.00039836871026829,
-      "loss": 3.4594,
+      "epoch": 3.37466307277628,
+      "grad_norm": 0.6409232020378113,
+      "learning_rate": 0.00039802698327037235,
+      "loss": 3.459,
       "step": 31300
     },
     {
-      "epoch": 3.374233128834356,
-      "grad_norm": 0.60047847032547,
-      "learning_rate": 0.0003980454692382286,
-      "loss": 3.4708,
+      "epoch": 3.3800539083557952,
+      "grad_norm": 0.6229880452156067,
+      "learning_rate": 0.00039770318402590396,
+      "loss": 3.4636,
       "step": 31350
     },
     {
-      "epoch": 3.379614680873964,
-      "grad_norm": 0.5911248326301575,
-      "learning_rate": 0.0003977222282081672,
-      "loss": 3.4498,
+      "epoch": 3.38544474393531,
+      "grad_norm": 0.6199302077293396,
+      "learning_rate": 0.00039737938478143545,
+      "loss": 3.4731,
       "step": 31400
     },
     {
-      "epoch": 3.384996232913572,
-      "grad_norm": 0.5842762589454651,
-      "learning_rate": 0.00039739898717810573,
-      "loss": 3.472,
+      "epoch": 3.3908355795148246,
+      "grad_norm": 0.6589909195899963,
+      "learning_rate": 0.000397055585536967,
+      "loss": 3.4561,
       "step": 31450
     },
     {
-      "epoch": 3.3903777849531807,
-      "grad_norm": 0.6286013126373291,
-      "learning_rate": 0.0003970757461480444,
-      "loss": 3.4783,
+      "epoch": 3.3962264150943398,
+      "grad_norm": 0.6057537794113159,
+      "learning_rate": 0.0003967317862924986,
+      "loss": 3.4529,
       "step": 31500
     },
     {
-      "epoch": 3.3957593369927888,
-      "grad_norm": 0.6078022718429565,
-      "learning_rate": 0.000396752505117983,
-      "loss": 3.4681,
+      "epoch": 3.4016172506738545,
+      "grad_norm": 0.5773242712020874,
+      "learning_rate": 0.00039640798704803016,
+      "loss": 3.4391,
       "step": 31550
     },
     {
-      "epoch": 3.401140889032397,
-      "grad_norm": 0.6259421706199646,
-      "learning_rate": 0.0003964292640879215,
-      "loss": 3.4482,
+      "epoch": 3.407008086253369,
+      "grad_norm": 0.6419113278388977,
+      "learning_rate": 0.00039608418780356176,
+      "loss": 3.4478,
       "step": 31600
     },
     {
-      "epoch": 3.4065224410720054,
-      "grad_norm": 0.6382803916931152,
-      "learning_rate": 0.0003961060230578601,
-      "loss": 3.4746,
+      "epoch": 3.4123989218328843,
+      "grad_norm": 0.5945045948028564,
+      "learning_rate": 0.0003957603885590933,
+      "loss": 3.4646,
       "step": 31650
     },
     {
-      "epoch": 3.4119039931116135,
-      "grad_norm": 0.6442477107048035,
-      "learning_rate": 0.00039578278202779865,
-      "loss": 3.4505,
+      "epoch": 3.417789757412399,
+      "grad_norm": 0.5897580981254578,
+      "learning_rate": 0.0003954365893146249,
+      "loss": 3.4525,
       "step": 31700
     },
     {
-      "epoch": 3.4172855451512216,
-      "grad_norm": 0.578318178653717,
-      "learning_rate": 0.00039545954099773725,
-      "loss": 3.4637,
+      "epoch": 3.4231805929919137,
+      "grad_norm": 0.6659418940544128,
+      "learning_rate": 0.00039511279007015647,
+      "loss": 3.459,
       "step": 31750
     },
     {
-      "epoch": 3.4226670971908297,
-      "grad_norm": 0.6376295685768127,
-      "learning_rate": 0.0003951362999676759,
-      "loss": 3.4703,
+      "epoch": 3.4285714285714284,
+      "grad_norm": 0.6416268348693848,
+      "learning_rate": 0.00039478899082568807,
+      "loss": 3.4602,
       "step": 31800
     },
     {
-      "epoch": 3.428048649230438,
-      "grad_norm": 0.5918070673942566,
-      "learning_rate": 0.00039481305893761444,
-      "loss": 3.4539,
+      "epoch": 3.4339622641509435,
+      "grad_norm": 0.579162061214447,
+      "learning_rate": 0.0003944651915812196,
+      "loss": 3.4484,
       "step": 31850
     },
     {
-      "epoch": 3.4334302012700464,
-      "grad_norm": 0.6081002354621887,
-      "learning_rate": 0.0003944962827281543,
-      "loss": 3.4505,
+      "epoch": 3.439353099730458,
+      "grad_norm": 0.5872979760169983,
+      "learning_rate": 0.0003941413923367512,
+      "loss": 3.4502,
       "step": 31900
     },
     {
-      "epoch": 3.4388117533096545,
-      "grad_norm": 0.6261876225471497,
-      "learning_rate": 0.00039417304169809284,
-      "loss": 3.4755,
+      "epoch": 3.444743935309973,
+      "grad_norm": 0.588444173336029,
+      "learning_rate": 0.0003938175930922828,
+      "loss": 3.4591,
       "step": 31950
     },
     {
-      "epoch": 3.4441933053492626,
-      "grad_norm": 0.6414044499397278,
-      "learning_rate": 0.00039384980066803143,
-      "loss": 3.4556,
+      "epoch": 3.450134770889488,
+      "grad_norm": 0.6142772436141968,
+      "learning_rate": 0.0003934937938478143,
+      "loss": 3.4622,
       "step": 32000
     },
     {
-      "epoch": 3.4441933053492626,
-      "eval_accuracy": 0.37261671248740036,
-      "eval_loss": 3.481884002685547,
-      "eval_runtime": 181.6828,
-      "eval_samples_per_second": 99.134,
-      "eval_steps_per_second": 6.198,
+      "epoch": 3.450134770889488,
+      "eval_accuracy": 0.37271808563173126,
+      "eval_loss": 3.478914737701416,
+      "eval_runtime": 183.9633,
+      "eval_samples_per_second": 97.905,
+      "eval_steps_per_second": 6.121,
       "step": 32000
     },
     {
-      "epoch": 3.449574857388871,
-      "grad_norm": 0.6231973767280579,
-      "learning_rate": 0.00039352655963796997,
-      "loss": 3.4596,
+      "epoch": 3.4555256064690028,
+      "grad_norm": 0.5936371684074402,
+      "learning_rate": 0.0003931699946033459,
+      "loss": 3.4558,
       "step": 32050
     },
     {
-      "epoch": 3.4549564094284793,
-      "grad_norm": 0.6971520781517029,
-      "learning_rate": 0.00039320331860790857,
-      "loss": 3.4584,
+      "epoch": 3.4609164420485174,
+      "grad_norm": 0.624254047870636,
+      "learning_rate": 0.00039284619535887743,
+      "loss": 3.4586,
       "step": 32100
     },
     {
-      "epoch": 3.4603379614680874,
-      "grad_norm": 0.6719905734062195,
-      "learning_rate": 0.0003928800775778472,
-      "loss": 3.4615,
+      "epoch": 3.466307277628032,
+      "grad_norm": 0.5764742493629456,
+      "learning_rate": 0.00039252239611440904,
+      "loss": 3.4551,
       "step": 32150
     },
     {
-      "epoch": 3.4657195135076955,
-      "grad_norm": 0.7227779626846313,
-      "learning_rate": 0.00039255683654778576,
-      "loss": 3.4486,
+      "epoch": 3.4716981132075473,
+      "grad_norm": 0.6184502243995667,
+      "learning_rate": 0.0003921985968699406,
+      "loss": 3.454,
       "step": 32200
     },
     {
-      "epoch": 3.471101065547304,
-      "grad_norm": 0.6186626553535461,
-      "learning_rate": 0.00039223359551772435,
-      "loss": 3.4744,
+      "epoch": 3.477088948787062,
+      "grad_norm": 0.6165410280227661,
+      "learning_rate": 0.0003918747976254722,
+      "loss": 3.4654,
       "step": 32250
     },
     {
-      "epoch": 3.476482617586912,
-      "grad_norm": 0.6083580851554871,
-      "learning_rate": 0.00039191035448766294,
-      "loss": 3.4683,
+      "epoch": 3.4824797843665767,
+      "grad_norm": 0.6311831474304199,
+      "learning_rate": 0.00039155099838100374,
+      "loss": 3.4728,
       "step": 32300
     },
     {
-      "epoch": 3.4818641696265202,
-      "grad_norm": 0.6660602688789368,
-      "learning_rate": 0.0003915871134576015,
-      "loss": 3.4638,
+      "epoch": 3.487870619946092,
+      "grad_norm": 0.6539682745933533,
+      "learning_rate": 0.0003912271991365353,
+      "loss": 3.4606,
       "step": 32350
     },
     {
-      "epoch": 3.4872457216661283,
-      "grad_norm": 0.6296412944793701,
-      "learning_rate": 0.0003912638724275401,
-      "loss": 3.4557,
+      "epoch": 3.4932614555256065,
+      "grad_norm": 0.61849445104599,
+      "learning_rate": 0.0003909033998920669,
+      "loss": 3.4608,
       "step": 32400
     },
     {
-      "epoch": 3.492627273705737,
-      "grad_norm": 0.6064853072166443,
-      "learning_rate": 0.00039094063139747873,
-      "loss": 3.4495,
+      "epoch": 3.498652291105121,
+      "grad_norm": 0.6489964723587036,
+      "learning_rate": 0.00039057960064759845,
+      "loss": 3.4424,
       "step": 32450
     },
     {
-      "epoch": 3.498008825745345,
-      "grad_norm": 0.6372223496437073,
-      "learning_rate": 0.00039061739036741727,
-      "loss": 3.4593,
+      "epoch": 3.5040431266846364,
+      "grad_norm": 0.6186843514442444,
+      "learning_rate": 0.00039025580140313005,
+      "loss": 3.467,
       "step": 32500
     },
     {
-      "epoch": 3.503390377784953,
-      "grad_norm": 0.6473979949951172,
-      "learning_rate": 0.00039029414933735586,
-      "loss": 3.4618,
+      "epoch": 3.509433962264151,
+      "grad_norm": 0.7145958542823792,
+      "learning_rate": 0.0003899320021586616,
+      "loss": 3.443,
       "step": 32550
     },
     {
-      "epoch": 3.5087719298245617,
-      "grad_norm": 0.5884780883789062,
-      "learning_rate": 0.0003899709083072944,
-      "loss": 3.4682,
+      "epoch": 3.5148247978436657,
+      "grad_norm": 0.6082626581192017,
+      "learning_rate": 0.0003896082029141932,
+      "loss": 3.4647,
       "step": 32600
     },
     {
-      "epoch": 3.5141534818641698,
-      "grad_norm": 0.6261137127876282,
-      "learning_rate": 0.000389647667277233,
-      "loss": 3.4469,
+      "epoch": 3.5202156334231804,
+      "grad_norm": 0.6374468803405762,
+      "learning_rate": 0.00038928440366972476,
+      "loss": 3.4517,
       "step": 32650
     },
     {
-      "epoch": 3.519535033903778,
-      "grad_norm": 0.6143213510513306,
-      "learning_rate": 0.00038932442624717165,
-      "loss": 3.4486,
+      "epoch": 3.525606469002695,
+      "grad_norm": 0.5979472398757935,
+      "learning_rate": 0.00038896060442525636,
+      "loss": 3.4477,
       "step": 32700
     },
     {
-      "epoch": 3.524916585943386,
-      "grad_norm": 0.6216268539428711,
-      "learning_rate": 0.0003890011852171102,
-      "loss": 3.4687,
+      "epoch": 3.5309973045822103,
+      "grad_norm": 0.6199315786361694,
+      "learning_rate": 0.00038863680518078786,
+      "loss": 3.4574,
       "step": 32750
     },
     {
-      "epoch": 3.530298137982994,
-      "grad_norm": 0.6284915208816528,
-      "learning_rate": 0.0003886779441870488,
-      "loss": 3.45,
+      "epoch": 3.536388140161725,
+      "grad_norm": 0.6273955702781677,
+      "learning_rate": 0.0003883130059363194,
+      "loss": 3.4612,
       "step": 32800
     },
     {
-      "epoch": 3.5356796900226026,
-      "grad_norm": 0.7261960506439209,
-      "learning_rate": 0.0003883547031569874,
-      "loss": 3.461,
+      "epoch": 3.5417789757412397,
+      "grad_norm": 0.5842116475105286,
+      "learning_rate": 0.000387989206691851,
+      "loss": 3.4486,
       "step": 32850
     },
     {
-      "epoch": 3.5410612420622107,
-      "grad_norm": 0.6670135855674744,
-      "learning_rate": 0.0003880314621269259,
-      "loss": 3.4522,
+      "epoch": 3.547169811320755,
+      "grad_norm": 0.5716790556907654,
+      "learning_rate": 0.00038766540744738256,
+      "loss": 3.4512,
       "step": 32900
     },
     {
-      "epoch": 3.546442794101819,
-      "grad_norm": 0.6466349959373474,
-      "learning_rate": 0.0003877082210968645,
-      "loss": 3.4644,
+      "epoch": 3.5525606469002695,
+      "grad_norm": 0.6028736233711243,
+      "learning_rate": 0.00038734160820291417,
+      "loss": 3.4706,
       "step": 32950
     },
     {
-      "epoch": 3.5518243461414274,
-      "grad_norm": 0.5869767069816589,
-      "learning_rate": 0.00038738498006680316,
-      "loss": 3.4758,
+      "epoch": 3.557951482479784,
+      "grad_norm": 0.7155103087425232,
+      "learning_rate": 0.0003870178089584457,
+      "loss": 3.4492,
       "step": 33000
     },
     {
-      "epoch": 3.5518243461414274,
-      "eval_accuracy": 0.37266495436958996,
-      "eval_loss": 3.4755313396453857,
-      "eval_runtime": 181.8939,
-      "eval_samples_per_second": 99.019,
-      "eval_steps_per_second": 6.19,
+      "epoch": 3.557951482479784,
+      "eval_accuracy": 0.37338803933799536,
+      "eval_loss": 3.4721028804779053,
+      "eval_runtime": 184.5186,
+      "eval_samples_per_second": 97.611,
+      "eval_steps_per_second": 6.102,
       "step": 33000
     },
     {
-      "epoch": 3.5572058981810355,
-      "grad_norm": 0.6126701831817627,
-      "learning_rate": 0.0003870617390367417,
-      "loss": 3.4703,
+      "epoch": 3.5633423180592994,
+      "grad_norm": 0.6142807006835938,
+      "learning_rate": 0.0003866940097139773,
+      "loss": 3.4518,
       "step": 33050
     },
     {
-      "epoch": 3.5625874502206436,
-      "grad_norm": 0.6517377495765686,
-      "learning_rate": 0.0003867384980066803,
-      "loss": 3.4547,
+      "epoch": 3.568733153638814,
+      "grad_norm": 0.6158440113067627,
+      "learning_rate": 0.0003863702104695089,
+      "loss": 3.473,
       "step": 33100
     },
     {
-      "epoch": 3.5679690022602517,
-      "grad_norm": 0.6763390898704529,
-      "learning_rate": 0.00038641525697661884,
-      "loss": 3.4589,
+      "epoch": 3.5741239892183287,
+      "grad_norm": 0.6133005619049072,
+      "learning_rate": 0.0003860464112250404,
+      "loss": 3.4505,
       "step": 33150
     },
     {
-      "epoch": 3.57335055429986,
-      "grad_norm": 0.6317027807235718,
-      "learning_rate": 0.00038609201594655743,
-      "loss": 3.4612,
+      "epoch": 3.579514824797844,
+      "grad_norm": 0.5994972586631775,
+      "learning_rate": 0.00038572261198057203,
+      "loss": 3.4664,
       "step": 33200
     },
     {
-      "epoch": 3.5787321063394684,
-      "grad_norm": 0.6396004557609558,
-      "learning_rate": 0.000385768774916496,
-      "loss": 3.4317,
+      "epoch": 3.5849056603773586,
+      "grad_norm": 0.5815605521202087,
+      "learning_rate": 0.0003853988127361036,
+      "loss": 3.4632,
       "step": 33250
     },
     {
-      "epoch": 3.5841136583790765,
-      "grad_norm": 0.6065496802330017,
-      "learning_rate": 0.0003854455338864346,
-      "loss": 3.4426,
+      "epoch": 3.5902964959568733,
+      "grad_norm": 0.686196506023407,
+      "learning_rate": 0.0003850750134916352,
+      "loss": 3.4868,
       "step": 33300
     },
     {
-      "epoch": 3.5894952104186846,
-      "grad_norm": 0.6124275326728821,
-      "learning_rate": 0.0003851222928563732,
-      "loss": 3.4479,
+      "epoch": 3.595687331536388,
+      "grad_norm": 0.6251017451286316,
+      "learning_rate": 0.0003847512142471667,
+      "loss": 3.4524,
       "step": 33350
     },
     {
-      "epoch": 3.594876762458293,
-      "grad_norm": 0.6135236620903015,
-      "learning_rate": 0.0003847990518263118,
-      "loss": 3.468,
+      "epoch": 3.601078167115903,
+      "grad_norm": 0.6139613389968872,
+      "learning_rate": 0.0003844274150026983,
+      "loss": 3.4631,
       "step": 33400
     },
     {
-      "epoch": 3.6002583144979012,
-      "grad_norm": 0.6805423498153687,
-      "learning_rate": 0.00038447581079625035,
-      "loss": 3.4838,
+      "epoch": 3.606469002695418,
+      "grad_norm": 0.5957735776901245,
+      "learning_rate": 0.00038410361575822984,
+      "loss": 3.4744,
       "step": 33450
     },
     {
-      "epoch": 3.6056398665375093,
-      "grad_norm": 0.5888933539390564,
-      "learning_rate": 0.00038415256976618895,
-      "loss": 3.4487,
+      "epoch": 3.6118598382749325,
+      "grad_norm": 0.6204478144645691,
+      "learning_rate": 0.00038377981651376144,
+      "loss": 3.4672,
       "step": 33500
     },
     {
-      "epoch": 3.611021418577118,
-      "grad_norm": 0.6695815920829773,
-      "learning_rate": 0.0003838293287361275,
-      "loss": 3.4717,
+      "epoch": 3.617250673854447,
+      "grad_norm": 0.6572640538215637,
+      "learning_rate": 0.000383456017269293,
+      "loss": 3.4535,
       "step": 33550
     },
     {
-      "epoch": 3.616402970616726,
-      "grad_norm": 0.6879743337631226,
-      "learning_rate": 0.00038350608770606613,
-      "loss": 3.4645,
+      "epoch": 3.6226415094339623,
+      "grad_norm": 0.5597906708717346,
+      "learning_rate": 0.00038313221802482454,
+      "loss": 3.4521,
       "step": 33600
     },
     {
-      "epoch": 3.621784522656334,
-      "grad_norm": 0.549407958984375,
-      "learning_rate": 0.00038318284667600473,
-      "loss": 3.4642,
+      "epoch": 3.628032345013477,
+      "grad_norm": 0.6139516830444336,
+      "learning_rate": 0.00038280841878035615,
+      "loss": 3.4436,
       "step": 33650
     },
     {
-      "epoch": 3.627166074695942,
-      "grad_norm": 0.5964043736457825,
-      "learning_rate": 0.00038285960564594327,
-      "loss": 3.4735,
+      "epoch": 3.6334231805929917,
+      "grad_norm": 0.6035452485084534,
+      "learning_rate": 0.0003824846195358877,
+      "loss": 3.4412,
       "step": 33700
     },
     {
-      "epoch": 3.6325476267355503,
-      "grad_norm": 0.6675652265548706,
-      "learning_rate": 0.00038253636461588186,
-      "loss": 3.4586,
+      "epoch": 3.638814016172507,
+      "grad_norm": 0.6189046502113342,
+      "learning_rate": 0.0003821608202914193,
+      "loss": 3.4639,
       "step": 33750
     },
     {
-      "epoch": 3.637929178775159,
-      "grad_norm": 0.6164919137954712,
-      "learning_rate": 0.00038221312358582046,
-      "loss": 3.4589,
+      "epoch": 3.6442048517520216,
+      "grad_norm": 0.6029171943664551,
+      "learning_rate": 0.00038183702104695085,
+      "loss": 3.4417,
       "step": 33800
     },
     {
-      "epoch": 3.643310730814767,
-      "grad_norm": 0.6311922073364258,
-      "learning_rate": 0.00038188988255575905,
-      "loss": 3.4587,
+      "epoch": 3.6495956873315363,
+      "grad_norm": 0.6461770534515381,
+      "learning_rate": 0.00038151322180248246,
+      "loss": 3.4537,
       "step": 33850
     },
     {
-      "epoch": 3.648692282854375,
-      "grad_norm": 0.6482579708099365,
-      "learning_rate": 0.00038156664152569765,
-      "loss": 3.4482,
+      "epoch": 3.6549865229110514,
+      "grad_norm": 0.6269723773002625,
+      "learning_rate": 0.000381189422558014,
+      "loss": 3.4559,
       "step": 33900
     },
     {
-      "epoch": 3.6540738348939836,
-      "grad_norm": 0.7032673358917236,
-      "learning_rate": 0.00038124340049563624,
-      "loss": 3.4705,
+      "epoch": 3.660377358490566,
+      "grad_norm": 0.6271780729293823,
+      "learning_rate": 0.0003808656233135456,
+      "loss": 3.4528,
       "step": 33950
     },
     {
-      "epoch": 3.6594553869335917,
-      "grad_norm": 0.592106819152832,
-      "learning_rate": 0.0003809201594655748,
-      "loss": 3.455,
+      "epoch": 3.665768194070081,
+      "grad_norm": 0.6369091272354126,
+      "learning_rate": 0.00038054182406907716,
+      "loss": 3.4734,
       "step": 34000
     },
     {
-      "epoch": 3.6594553869335917,
-      "eval_accuracy": 0.37426226047335065,
-      "eval_loss": 3.4661221504211426,
-      "eval_runtime": 181.561,
-      "eval_samples_per_second": 99.201,
-      "eval_steps_per_second": 6.202,
+      "epoch": 3.665768194070081,
+      "eval_accuracy": 0.37419109283183216,
+      "eval_loss": 3.4643988609313965,
+      "eval_runtime": 183.71,
+      "eval_samples_per_second": 98.04,
+      "eval_steps_per_second": 6.129,
       "step": 34000
     },
     {
-      "epoch": 3.6648369389732,
-      "grad_norm": 0.5778534412384033,
-      "learning_rate": 0.0003805969184355134,
-      "loss": 3.4494,
+      "epoch": 3.671159029649596,
+      "grad_norm": 0.6655065417289734,
+      "learning_rate": 0.00038021802482460866,
+      "loss": 3.4531,
       "step": 34050
     },
     {
-      "epoch": 3.670218491012808,
-      "grad_norm": 0.5945811867713928,
-      "learning_rate": 0.0003802736774054519,
-      "loss": 3.4473,
+      "epoch": 3.6765498652291106,
+      "grad_norm": 0.5922834277153015,
+      "learning_rate": 0.00037989422558014027,
+      "loss": 3.4615,
       "step": 34100
     },
     {
-      "epoch": 3.675600043052416,
-      "grad_norm": 0.6176246404647827,
-      "learning_rate": 0.00037995043637539057,
-      "loss": 3.455,
+      "epoch": 3.6819407008086253,
+      "grad_norm": 0.6182780265808105,
+      "learning_rate": 0.0003795704263356718,
+      "loss": 3.4697,
       "step": 34150
     },
     {
-      "epoch": 3.6809815950920246,
-      "grad_norm": 0.6745234727859497,
-      "learning_rate": 0.00037962719534532916,
-      "loss": 3.4559,
+      "epoch": 3.68733153638814,
+      "grad_norm": 0.6138507723808289,
+      "learning_rate": 0.0003792466270912034,
+      "loss": 3.4651,
       "step": 34200
     },
     {
-      "epoch": 3.6863631471316327,
-      "grad_norm": 0.5692402124404907,
-      "learning_rate": 0.0003793039543152677,
-      "loss": 3.441,
+      "epoch": 3.6927223719676547,
+      "grad_norm": 0.6309821605682373,
+      "learning_rate": 0.00037892282784673497,
+      "loss": 3.4605,
       "step": 34250
     },
     {
-      "epoch": 3.691744699171241,
-      "grad_norm": 0.6138830184936523,
-      "learning_rate": 0.0003789807132852063,
-      "loss": 3.4481,
+      "epoch": 3.69811320754717,
+      "grad_norm": 0.6199393272399902,
+      "learning_rate": 0.0003785990286022666,
+      "loss": 3.4681,
       "step": 34300
     },
     {
-      "epoch": 3.6971262512108494,
-      "grad_norm": 0.6347447037696838,
-      "learning_rate": 0.0003786574722551449,
-      "loss": 3.4508,
+      "epoch": 3.7035040431266846,
+      "grad_norm": 0.6389825940132141,
+      "learning_rate": 0.00037827522935779813,
+      "loss": 3.476,
       "step": 34350
     },
     {
-      "epoch": 3.7025078032504575,
-      "grad_norm": 0.6304515600204468,
-      "learning_rate": 0.00037833423122508343,
-      "loss": 3.4593,
+      "epoch": 3.7088948787061993,
+      "grad_norm": 0.6566691994667053,
+      "learning_rate": 0.00037795143011332973,
+      "loss": 3.4497,
       "step": 34400
     },
     {
-      "epoch": 3.7078893552900656,
-      "grad_norm": 0.6190226078033447,
-      "learning_rate": 0.0003780109901950221,
-      "loss": 3.4571,
+      "epoch": 3.7142857142857144,
+      "grad_norm": 0.6438661217689514,
+      "learning_rate": 0.0003776276308688613,
+      "loss": 3.435,
       "step": 34450
     },
     {
-      "epoch": 3.713270907329674,
-      "grad_norm": 0.6307307481765747,
-      "learning_rate": 0.0003776877491649607,
-      "loss": 3.4505,
+      "epoch": 3.719676549865229,
+      "grad_norm": 0.66309654712677,
+      "learning_rate": 0.00037730383162439283,
+      "loss": 3.4533,
       "step": 34500
     },
     {
-      "epoch": 3.7186524593692822,
-      "grad_norm": 0.6326085925102234,
-      "learning_rate": 0.0003773645081348992,
-      "loss": 3.4599,
+      "epoch": 3.725067385444744,
+      "grad_norm": 0.6174817681312561,
+      "learning_rate": 0.00037698003237992444,
+      "loss": 3.4729,
       "step": 34550
     },
     {
-      "epoch": 3.7240340114088903,
-      "grad_norm": 0.6394867897033691,
-      "learning_rate": 0.0003770412671048378,
-      "loss": 3.4568,
+      "epoch": 3.730458221024259,
+      "grad_norm": 0.6526914238929749,
+      "learning_rate": 0.000376656233135456,
+      "loss": 3.463,
       "step": 34600
     },
     {
-      "epoch": 3.7294155634484984,
-      "grad_norm": 0.633033037185669,
-      "learning_rate": 0.00037671802607477635,
-      "loss": 3.4429,
+      "epoch": 3.7358490566037736,
+      "grad_norm": 0.6965179443359375,
+      "learning_rate": 0.00037633890987587694,
+      "loss": 3.4368,
       "step": 34650
     },
     {
-      "epoch": 3.7347971154881066,
-      "grad_norm": 0.6172804236412048,
-      "learning_rate": 0.000376394785044715,
-      "loss": 3.4633,
+      "epoch": 3.7412398921832883,
+      "grad_norm": 0.6173874139785767,
+      "learning_rate": 0.00037601511063140855,
+      "loss": 3.4799,
       "step": 34700
     },
     {
-      "epoch": 3.740178667527715,
-      "grad_norm": 0.6102865934371948,
-      "learning_rate": 0.0003760715440146536,
-      "loss": 3.4695,
+      "epoch": 3.7466307277628035,
+      "grad_norm": 0.6308540105819702,
+      "learning_rate": 0.00037569131138694004,
+      "loss": 3.4431,
       "step": 34750
     },
     {
-      "epoch": 3.745560219567323,
-      "grad_norm": 0.6239819526672363,
-      "learning_rate": 0.00037574830298459214,
-      "loss": 3.4567,
+      "epoch": 3.752021563342318,
+      "grad_norm": 0.6584587693214417,
+      "learning_rate": 0.0003753675121424716,
+      "loss": 3.4722,
       "step": 34800
     },
     {
-      "epoch": 3.7509417716069313,
-      "grad_norm": 0.6692067980766296,
-      "learning_rate": 0.00037542506195453073,
-      "loss": 3.4591,
+      "epoch": 3.757412398921833,
+      "grad_norm": 0.6155205368995667,
+      "learning_rate": 0.0003750437128980032,
+      "loss": 3.4869,
       "step": 34850
     },
     {
-      "epoch": 3.75632332364654,
-      "grad_norm": 0.589547336101532,
-      "learning_rate": 0.0003751018209244693,
-      "loss": 3.4723,
+      "epoch": 3.7628032345013476,
+      "grad_norm": 0.6073881983757019,
+      "learning_rate": 0.00037471991365353475,
+      "loss": 3.4715,
       "step": 34900
     },
     {
-      "epoch": 3.761704875686148,
-      "grad_norm": 0.6273283958435059,
-      "learning_rate": 0.00037477857989440787,
-      "loss": 3.4668,
+      "epoch": 3.7681940700808623,
+      "grad_norm": 0.5700336694717407,
+      "learning_rate": 0.00037439611440906635,
+      "loss": 3.4509,
       "step": 34950
     },
     {
-      "epoch": 3.767086427725756,
-      "grad_norm": 0.6687270402908325,
-      "learning_rate": 0.00037446180368494767,
-      "loss": 3.4728,
+      "epoch": 3.7735849056603774,
+      "grad_norm": 0.640873372554779,
+      "learning_rate": 0.0003740723151645979,
+      "loss": 3.4493,
       "step": 35000
     },
     {
-      "epoch": 3.767086427725756,
-      "eval_accuracy": 0.3749564437735974,
-      "eval_loss": 3.4588773250579834,
-      "eval_runtime": 181.7474,
-      "eval_samples_per_second": 99.099,
-      "eval_steps_per_second": 6.195,
+      "epoch": 3.7735849056603774,
+      "eval_accuracy": 0.3748407284480749,
+      "eval_loss": 3.4575767517089844,
+      "eval_runtime": 184.3708,
+      "eval_samples_per_second": 97.689,
+      "eval_steps_per_second": 6.107,
       "step": 35000
     },
     {
-      "epoch": 3.772467979765364,
-      "grad_norm": 0.657584547996521,
-      "learning_rate": 0.00037413856265488627,
-      "loss": 3.4584,
+      "epoch": 3.778975741239892,
+      "grad_norm": 0.6594972610473633,
+      "learning_rate": 0.0003737485159201295,
+      "loss": 3.4722,
       "step": 35050
     },
     {
-      "epoch": 3.7778495318049723,
-      "grad_norm": 0.6070948839187622,
-      "learning_rate": 0.0003738153216248249,
-      "loss": 3.4495,
+      "epoch": 3.784366576819407,
+      "grad_norm": 0.6612352132797241,
+      "learning_rate": 0.00037342471667566106,
+      "loss": 3.4383,
       "step": 35100
     },
     {
-      "epoch": 3.783231083844581,
-      "grad_norm": 0.6411889791488647,
-      "learning_rate": 0.00037349208059476346,
-      "loss": 3.4406,
+      "epoch": 3.789757412398922,
+      "grad_norm": 0.6321669816970825,
+      "learning_rate": 0.00037310091743119266,
+      "loss": 3.4615,
       "step": 35150
     },
     {
-      "epoch": 3.788612635884189,
-      "grad_norm": 0.6415666341781616,
-      "learning_rate": 0.00037316883956470205,
-      "loss": 3.4803,
+      "epoch": 3.7951482479784366,
+      "grad_norm": 0.6197882890701294,
+      "learning_rate": 0.0003727771181867242,
+      "loss": 3.4443,
       "step": 35200
     },
     {
-      "epoch": 3.793994187923797,
-      "grad_norm": 0.6426709890365601,
-      "learning_rate": 0.00037284559853464064,
-      "loss": 3.4618,
+      "epoch": 3.8005390835579513,
+      "grad_norm": 0.6296155452728271,
+      "learning_rate": 0.00037245331894225576,
+      "loss": 3.47,
       "step": 35250
     },
     {
-      "epoch": 3.7993757399634056,
-      "grad_norm": 0.6597058176994324,
-      "learning_rate": 0.0003725223575045792,
-      "loss": 3.4453,
+      "epoch": 3.8059299191374665,
+      "grad_norm": 0.6412382125854492,
+      "learning_rate": 0.00037212951969778737,
+      "loss": 3.4585,
       "step": 35300
     },
     {
-      "epoch": 3.8047572920030137,
-      "grad_norm": 0.6358980536460876,
-      "learning_rate": 0.0003721991164745178,
-      "loss": 3.4715,
+      "epoch": 3.811320754716981,
+      "grad_norm": 0.5623229742050171,
+      "learning_rate": 0.00037180572045331887,
+      "loss": 3.462,
       "step": 35350
     },
     {
-      "epoch": 3.810138844042622,
-      "grad_norm": 0.6410354375839233,
-      "learning_rate": 0.00037187587544445643,
-      "loss": 3.4647,
+      "epoch": 3.816711590296496,
+      "grad_norm": 0.6568892002105713,
+      "learning_rate": 0.00037148192120885047,
+      "loss": 3.4787,
       "step": 35400
     },
     {
-      "epoch": 3.8155203960822304,
-      "grad_norm": 0.6233660578727722,
-      "learning_rate": 0.00037155263441439497,
-      "loss": 3.4698,
+      "epoch": 3.822102425876011,
+      "grad_norm": 0.6241756677627563,
+      "learning_rate": 0.000371158121964382,
+      "loss": 3.4369,
       "step": 35450
     },
     {
-      "epoch": 3.8209019481218385,
-      "grad_norm": 0.628305196762085,
-      "learning_rate": 0.00037122939338433356,
-      "loss": 3.4537,
+      "epoch": 3.8274932614555257,
+      "grad_norm": 0.6201616525650024,
+      "learning_rate": 0.0003708343227199136,
+      "loss": 3.4711,
       "step": 35500
     },
     {
-      "epoch": 3.8262835001614466,
-      "grad_norm": 0.6736071705818176,
-      "learning_rate": 0.0003709061523542721,
-      "loss": 3.4576,
+      "epoch": 3.8328840970350404,
+      "grad_norm": 0.6688598394393921,
+      "learning_rate": 0.0003705105234754452,
+      "loss": 3.4521,
       "step": 35550
     },
     {
-      "epoch": 3.8316650522010547,
-      "grad_norm": 0.6532096266746521,
-      "learning_rate": 0.0003705829113242107,
-      "loss": 3.4494,
+      "epoch": 3.838274932614555,
+      "grad_norm": 0.6160039901733398,
+      "learning_rate": 0.0003701867242309768,
+      "loss": 3.4447,
       "step": 35600
     },
     {
-      "epoch": 3.837046604240663,
-      "grad_norm": 0.5983206629753113,
-      "learning_rate": 0.00037025967029414935,
-      "loss": 3.4538,
+      "epoch": 3.8436657681940702,
+      "grad_norm": 0.6429824829101562,
+      "learning_rate": 0.00036986292498650833,
+      "loss": 3.4559,
       "step": 35650
     },
     {
-      "epoch": 3.8424281562802713,
-      "grad_norm": 0.6398340463638306,
-      "learning_rate": 0.0003699364292640879,
-      "loss": 3.4738,
+      "epoch": 3.849056603773585,
+      "grad_norm": 0.6137093901634216,
+      "learning_rate": 0.0003695391257420399,
+      "loss": 3.4528,
       "step": 35700
     },
     {
-      "epoch": 3.8478097083198795,
-      "grad_norm": 0.5982816219329834,
-      "learning_rate": 0.0003696131882340265,
-      "loss": 3.4687,
+      "epoch": 3.8544474393530996,
+      "grad_norm": 0.6473554968833923,
+      "learning_rate": 0.0003692153264975715,
+      "loss": 3.4329,
       "step": 35750
     },
     {
-      "epoch": 3.8531912603594876,
-      "grad_norm": 0.6313088536262512,
-      "learning_rate": 0.0003692899472039651,
-      "loss": 3.4527,
+      "epoch": 3.8598382749326143,
+      "grad_norm": 0.6077702045440674,
+      "learning_rate": 0.00036889152725310304,
+      "loss": 3.4542,
       "step": 35800
     },
     {
-      "epoch": 3.858572812399096,
-      "grad_norm": 0.6521557569503784,
-      "learning_rate": 0.0003689667061739036,
-      "loss": 3.4547,
+      "epoch": 3.8652291105121295,
+      "grad_norm": 0.6486040353775024,
+      "learning_rate": 0.00036856772800863464,
+      "loss": 3.4603,
       "step": 35850
     },
     {
-      "epoch": 3.863954364438704,
-      "grad_norm": 0.6223726272583008,
-      "learning_rate": 0.0003686434651438422,
-      "loss": 3.4714,
+      "epoch": 3.870619946091644,
+      "grad_norm": 0.5944869518280029,
+      "learning_rate": 0.0003682439287641662,
+      "loss": 3.451,
       "step": 35900
     },
     {
-      "epoch": 3.8693359164783123,
-      "grad_norm": 0.6917114853858948,
-      "learning_rate": 0.00036832022411378086,
-      "loss": 3.4696,
+      "epoch": 3.876010781671159,
+      "grad_norm": 0.5912535786628723,
+      "learning_rate": 0.0003679201295196978,
+      "loss": 3.4528,
       "step": 35950
     },
     {
-      "epoch": 3.8747174685179204,
-      "grad_norm": 0.606867253780365,
-      "learning_rate": 0.0003679969830837194,
-      "loss": 3.4347,
+      "epoch": 3.881401617250674,
+      "grad_norm": 0.6211251020431519,
+      "learning_rate": 0.00036759633027522935,
+      "loss": 3.4348,
       "step": 36000
     },
     {
-      "epoch": 3.8747174685179204,
-      "eval_accuracy": 0.3753462903350757,
-      "eval_loss": 3.4530980587005615,
-      "eval_runtime": 181.4546,
-      "eval_samples_per_second": 99.259,
-      "eval_steps_per_second": 6.205,
+      "epoch": 3.881401617250674,
+      "eval_accuracy": 0.37561661871995816,
+      "eval_loss": 3.451625108718872,
+      "eval_runtime": 183.9535,
+      "eval_samples_per_second": 97.911,
+      "eval_steps_per_second": 6.121,
       "step": 36000
     },
     {
-      "epoch": 3.8800990205575285,
-      "grad_norm": 0.6278716325759888,
-      "learning_rate": 0.000367673742053658,
-      "loss": 3.4351,
+      "epoch": 3.8867924528301887,
+      "grad_norm": 0.6398293972015381,
+      "learning_rate": 0.00036727253103076084,
+      "loss": 3.44,
       "step": 36050
     },
     {
-      "epoch": 3.885480572597137,
-      "grad_norm": 0.625916600227356,
-      "learning_rate": 0.00036735050102359654,
-      "loss": 3.4415,
+      "epoch": 3.8921832884097034,
+      "grad_norm": 0.6175176501274109,
+      "learning_rate": 0.00036694873178629245,
+      "loss": 3.4642,
       "step": 36100
     },
     {
-      "epoch": 3.890862124636745,
-      "grad_norm": 0.6142260432243347,
-      "learning_rate": 0.00036702725999353513,
-      "loss": 3.4539,
+      "epoch": 3.8975741239892185,
+      "grad_norm": 0.6453118324279785,
+      "learning_rate": 0.000366624932541824,
+      "loss": 3.4538,
       "step": 36150
     },
     {
-      "epoch": 3.8962436766763533,
-      "grad_norm": 0.6766964793205261,
-      "learning_rate": 0.0003667040189634737,
-      "loss": 3.4584,
+      "epoch": 3.9029649595687332,
+      "grad_norm": 0.6053382754325867,
+      "learning_rate": 0.0003663011332973556,
+      "loss": 3.4348,
       "step": 36200
     },
     {
-      "epoch": 3.901625228715962,
-      "grad_norm": 0.6694559454917908,
-      "learning_rate": 0.0003663807779334123,
-      "loss": 3.4564,
+      "epoch": 3.908355795148248,
+      "grad_norm": 0.6160609722137451,
+      "learning_rate": 0.00036597733405288715,
+      "loss": 3.4391,
       "step": 36250
     },
     {
-      "epoch": 3.90700678075557,
-      "grad_norm": 0.6292983889579773,
-      "learning_rate": 0.0003660575369033509,
-      "loss": 3.4572,
+      "epoch": 3.913746630727763,
+      "grad_norm": 0.6582005620002747,
+      "learning_rate": 0.00036565353480841876,
+      "loss": 3.4521,
       "step": 36300
     },
     {
-      "epoch": 3.912388332795178,
-      "grad_norm": 0.6223490834236145,
-      "learning_rate": 0.0003657342958732895,
-      "loss": 3.4521,
+      "epoch": 3.9191374663072778,
+      "grad_norm": 0.630403995513916,
+      "learning_rate": 0.0003653297355639503,
+      "loss": 3.4448,
       "step": 36350
     },
     {
-      "epoch": 3.9177698848347866,
-      "grad_norm": 0.6118103861808777,
-      "learning_rate": 0.00036541105484322805,
-      "loss": 3.4535,
+      "epoch": 3.9245283018867925,
+      "grad_norm": 0.630704402923584,
+      "learning_rate": 0.0003650059363194819,
+      "loss": 3.4452,
       "step": 36400
     },
     {
-      "epoch": 3.9231514368743947,
-      "grad_norm": 0.645898163318634,
-      "learning_rate": 0.00036508781381316665,
-      "loss": 3.4518,
+      "epoch": 3.929919137466307,
+      "grad_norm": 0.6400175094604492,
+      "learning_rate": 0.00036468213707501347,
+      "loss": 3.4537,
       "step": 36450
     },
     {
-      "epoch": 3.928532988914003,
-      "grad_norm": 0.6415484547615051,
-      "learning_rate": 0.0003647645727831053,
-      "loss": 3.4585,
+      "epoch": 3.935309973045822,
+      "grad_norm": 0.6246758699417114,
+      "learning_rate": 0.000364358337830545,
+      "loss": 3.4591,
       "step": 36500
     },
     {
-      "epoch": 3.933914540953611,
-      "grad_norm": 0.6804546117782593,
-      "learning_rate": 0.00036444133175304384,
-      "loss": 3.4533,
+      "epoch": 3.940700808625337,
+      "grad_norm": 0.6261600852012634,
+      "learning_rate": 0.0003640345385860766,
+      "loss": 3.4461,
       "step": 36550
     },
     {
-      "epoch": 3.939296092993219,
-      "grad_norm": 0.6136488318443298,
-      "learning_rate": 0.00036411809072298243,
-      "loss": 3.4474,
+      "epoch": 3.9460916442048517,
+      "grad_norm": 0.6653203368186951,
+      "learning_rate": 0.00036371073934160817,
+      "loss": 3.4565,
       "step": 36600
     },
     {
-      "epoch": 3.9446776450328276,
-      "grad_norm": 0.6238126158714294,
-      "learning_rate": 0.00036379484969292097,
-      "loss": 3.458,
+      "epoch": 3.9514824797843664,
+      "grad_norm": 0.5844745635986328,
+      "learning_rate": 0.0003633869400971398,
+      "loss": 3.4513,
       "step": 36650
     },
     {
-      "epoch": 3.9500591970724357,
-      "grad_norm": 0.6238007545471191,
-      "learning_rate": 0.00036347160866285956,
-      "loss": 3.4624,
+      "epoch": 3.9568733153638815,
+      "grad_norm": 0.5904840230941772,
+      "learning_rate": 0.00036306314085267127,
+      "loss": 3.4512,
       "step": 36700
     },
     {
-      "epoch": 3.955440749112044,
-      "grad_norm": 0.6192507743835449,
-      "learning_rate": 0.0003631483676327981,
-      "loss": 3.4544,
+      "epoch": 3.9622641509433962,
+      "grad_norm": 0.5882664322853088,
+      "learning_rate": 0.00036273934160820293,
+      "loss": 3.4395,
       "step": 36750
     },
     {
-      "epoch": 3.9608223011516523,
-      "grad_norm": 0.6187044978141785,
-      "learning_rate": 0.00036282512660273675,
-      "loss": 3.4372,
+      "epoch": 3.967654986522911,
+      "grad_norm": 0.6878264546394348,
+      "learning_rate": 0.00036241554236373443,
+      "loss": 3.4403,
       "step": 36800
     },
     {
-      "epoch": 3.9662038531912605,
-      "grad_norm": 0.6006636619567871,
-      "learning_rate": 0.00036250188557267535,
-      "loss": 3.4588,
+      "epoch": 3.973045822102426,
+      "grad_norm": 0.7116815447807312,
+      "learning_rate": 0.00036209174311926603,
+      "loss": 3.4378,
       "step": 36850
     },
     {
-      "epoch": 3.9715854052308686,
-      "grad_norm": 0.6458033323287964,
-      "learning_rate": 0.0003621786445426139,
-      "loss": 3.4481,
+      "epoch": 3.9784366576819408,
+      "grad_norm": 0.6315473914146423,
+      "learning_rate": 0.0003617679438747976,
+      "loss": 3.4518,
       "step": 36900
     },
     {
-      "epoch": 3.9769669572704767,
-      "grad_norm": 0.6214568614959717,
-      "learning_rate": 0.0003618554035125525,
-      "loss": 3.4495,
+      "epoch": 3.9838274932614555,
+      "grad_norm": 0.6268083453178406,
+      "learning_rate": 0.00036144414463032913,
+      "loss": 3.4638,
       "step": 36950
     },
     {
-      "epoch": 3.9823485093100848,
-      "grad_norm": 0.6304638981819153,
-      "learning_rate": 0.0003615321624824911,
-      "loss": 3.46,
+      "epoch": 3.9892183288409706,
+      "grad_norm": 0.6648722887039185,
+      "learning_rate": 0.00036112034538586074,
+      "loss": 3.4583,
       "step": 37000
     },
     {
-      "epoch": 3.9823485093100848,
-      "eval_accuracy": 0.37597354345642875,
-      "eval_loss": 3.446321725845337,
-      "eval_runtime": 181.7245,
-      "eval_samples_per_second": 99.112,
-      "eval_steps_per_second": 6.196,
+      "epoch": 3.9892183288409706,
+      "eval_accuracy": 0.37630754243357045,
+      "eval_loss": 3.44683575630188,
+      "eval_runtime": 184.2555,
+      "eval_samples_per_second": 97.75,
+      "eval_steps_per_second": 6.111,
       "step": 37000
     },
     {
-      "epoch": 3.9877300613496933,
-      "grad_norm": 0.5896131992340088,
-      "learning_rate": 0.0003612089214524296,
-      "loss": 3.4304,
+      "epoch": 3.9946091644204853,
+      "grad_norm": 0.6127066016197205,
+      "learning_rate": 0.0003607965461413923,
+      "loss": 3.454,
       "step": 37050
     },
     {
-      "epoch": 3.9931116133893014,
-      "grad_norm": 0.6256299614906311,
-      "learning_rate": 0.00036088568042236827,
-      "loss": 3.4539,
+      "epoch": 4.0,
+      "grad_norm": 1.1642690896987915,
+      "learning_rate": 0.0003604727468969239,
+      "loss": 3.4444,
       "step": 37100
     },
     {
-      "epoch": 3.9984931654289095,
-      "grad_norm": 0.6513100862503052,
-      "learning_rate": 0.00036056243939230686,
-      "loss": 3.4543,
+      "epoch": 4.005390835579515,
+      "grad_norm": 0.6003873348236084,
+      "learning_rate": 0.00036014894765245544,
+      "loss": 3.3566,
       "step": 37150
     },
     {
-      "epoch": 4.003874717468518,
-      "grad_norm": 0.6809729337692261,
-      "learning_rate": 0.0003602391983622454,
-      "loss": 3.382,
+      "epoch": 4.010781671159029,
+      "grad_norm": 0.6421666145324707,
+      "learning_rate": 0.00035982514840798705,
+      "loss": 3.3561,
       "step": 37200
     },
     {
-      "epoch": 4.009256269508126,
-      "grad_norm": 0.6216638088226318,
-      "learning_rate": 0.000359915957332184,
-      "loss": 3.3691,
+      "epoch": 4.0161725067385445,
+      "grad_norm": 0.6512959003448486,
+      "learning_rate": 0.0003595013491635186,
+      "loss": 3.3552,
       "step": 37250
     },
     {
-      "epoch": 4.014637821547734,
-      "grad_norm": 0.687582790851593,
-      "learning_rate": 0.00035959271630212254,
-      "loss": 3.3739,
+      "epoch": 4.02156334231806,
+      "grad_norm": 0.669863224029541,
+      "learning_rate": 0.0003591775499190502,
+      "loss": 3.3356,
       "step": 37300
     },
     {
-      "epoch": 4.020019373587343,
-      "grad_norm": 0.6538896560668945,
-      "learning_rate": 0.00035926947527206113,
-      "loss": 3.3553,
+      "epoch": 4.026954177897574,
+      "grad_norm": 0.6304590106010437,
+      "learning_rate": 0.00035885375067458176,
+      "loss": 3.3633,
       "step": 37350
     },
     {
-      "epoch": 4.0254009256269505,
-      "grad_norm": 0.6100491881370544,
-      "learning_rate": 0.0003589462342419998,
-      "loss": 3.3636,
+      "epoch": 4.032345013477089,
+      "grad_norm": 0.6287381052970886,
+      "learning_rate": 0.00035852995143011325,
+      "loss": 3.3679,
       "step": 37400
     },
     {
-      "epoch": 4.030782477666559,
-      "grad_norm": 0.6782578825950623,
-      "learning_rate": 0.0003586229932119383,
-      "loss": 3.3712,
+      "epoch": 4.037735849056604,
+      "grad_norm": 0.6392130255699158,
+      "learning_rate": 0.00035820615218564486,
+      "loss": 3.3555,
       "step": 37450
     },
     {
-      "epoch": 4.036164029706168,
-      "grad_norm": 0.7136054039001465,
-      "learning_rate": 0.0003582997521818769,
-      "loss": 3.3502,
+      "epoch": 4.0431266846361185,
+      "grad_norm": 0.6464784145355225,
+      "learning_rate": 0.0003578823529411764,
+      "loss": 3.3715,
       "step": 37500
     },
     {
-      "epoch": 4.041545581745775,
-      "grad_norm": 1.0063740015029907,
-      "learning_rate": 0.0003579765111518155,
-      "loss": 3.3518,
+      "epoch": 4.048517520215634,
+      "grad_norm": 0.6907263398170471,
+      "learning_rate": 0.000357558553696708,
+      "loss": 3.3617,
       "step": 37550
     },
     {
-      "epoch": 4.046927133785384,
-      "grad_norm": 0.6431665420532227,
-      "learning_rate": 0.00035765327012175405,
-      "loss": 3.3673,
+      "epoch": 4.053908355795148,
+      "grad_norm": 0.6370674967765808,
+      "learning_rate": 0.00035724123043712896,
+      "loss": 3.3571,
       "step": 37600
     },
     {
-      "epoch": 4.0523086858249915,
-      "grad_norm": 0.6726707220077515,
-      "learning_rate": 0.0003573300290916927,
-      "loss": 3.365,
+      "epoch": 4.059299191374663,
+      "grad_norm": 0.6432491540908813,
+      "learning_rate": 0.0003569174311926605,
+      "loss": 3.3712,
       "step": 37650
     },
     {
-      "epoch": 4.0576902378646,
-      "grad_norm": 0.6649792194366455,
-      "learning_rate": 0.0003570067880616313,
-      "loss": 3.373,
+      "epoch": 4.064690026954178,
+      "grad_norm": 0.6911354064941406,
+      "learning_rate": 0.00035659363194819206,
+      "loss": 3.3734,
       "step": 37700
     },
     {
-      "epoch": 4.063071789904209,
-      "grad_norm": 0.6573074460029602,
-      "learning_rate": 0.00035668354703156984,
-      "loss": 3.3614,
+      "epoch": 4.070080862533692,
+      "grad_norm": 0.6535324454307556,
+      "learning_rate": 0.00035626983270372367,
+      "loss": 3.352,
       "step": 37750
     },
     {
-      "epoch": 4.068453341943816,
-      "grad_norm": 0.6372230052947998,
-      "learning_rate": 0.00035636030600150843,
-      "loss": 3.3596,
+      "epoch": 4.0754716981132075,
+      "grad_norm": 0.6952840685844421,
+      "learning_rate": 0.0003559460334592552,
+      "loss": 3.353,
       "step": 37800
     },
     {
-      "epoch": 4.073834893983425,
-      "grad_norm": 0.6461084485054016,
-      "learning_rate": 0.00035603706497144697,
-      "loss": 3.3685,
+      "epoch": 4.080862533692723,
+      "grad_norm": 0.6398538947105408,
+      "learning_rate": 0.0003556222342147868,
+      "loss": 3.3631,
       "step": 37850
     },
     {
-      "epoch": 4.079216446023033,
-      "grad_norm": 0.6296938061714172,
-      "learning_rate": 0.00035571382394138557,
-      "loss": 3.3641,
+      "epoch": 4.086253369272237,
+      "grad_norm": 0.6263781785964966,
+      "learning_rate": 0.0003552984349703184,
+      "loss": 3.3749,
       "step": 37900
     },
     {
-      "epoch": 4.084597998062641,
-      "grad_norm": 0.658966600894928,
-      "learning_rate": 0.0003553905829113242,
-      "loss": 3.3773,
+      "epoch": 4.091644204851752,
+      "grad_norm": 0.6595255136489868,
+      "learning_rate": 0.00035497463572585,
+      "loss": 3.3832,
       "step": 37950
     },
     {
-      "epoch": 4.08997955010225,
-      "grad_norm": 0.6599716544151306,
-      "learning_rate": 0.00035506734188126275,
-      "loss": 3.3683,
+      "epoch": 4.097035040431267,
+      "grad_norm": 0.6543563604354858,
+      "learning_rate": 0.00035465083648138153,
+      "loss": 3.371,
       "step": 38000
     },
     {
-      "epoch": 4.08997955010225,
-      "eval_accuracy": 0.3761422813912046,
-      "eval_loss": 3.44930100440979,
-      "eval_runtime": 181.3188,
-      "eval_samples_per_second": 99.333,
-      "eval_steps_per_second": 6.21,
+      "epoch": 4.097035040431267,
+      "eval_accuracy": 0.3768159292956544,
+      "eval_loss": 3.4490222930908203,
+      "eval_runtime": 183.8828,
+      "eval_samples_per_second": 97.948,
+      "eval_steps_per_second": 6.123,
       "step": 38000
     },
     {
-      "epoch": 4.095361102141858,
-      "grad_norm": 0.6307918429374695,
-      "learning_rate": 0.00035474410085120135,
-      "loss": 3.3854,
+      "epoch": 4.1024258760107815,
+      "grad_norm": 0.6638287305831909,
+      "learning_rate": 0.00035432703723691314,
+      "loss": 3.3648,
       "step": 38050
     },
     {
-      "epoch": 4.100742654181466,
-      "grad_norm": 0.6225454211235046,
-      "learning_rate": 0.00035442085982113994,
-      "loss": 3.3698,
+      "epoch": 4.107816711590297,
+      "grad_norm": 0.6960659027099609,
+      "learning_rate": 0.00035400323799244463,
+      "loss": 3.3746,
       "step": 38100
     },
     {
-      "epoch": 4.106124206221074,
-      "grad_norm": 0.6287394762039185,
-      "learning_rate": 0.0003540976187910785,
-      "loss": 3.3919,
+      "epoch": 4.113207547169812,
+      "grad_norm": 0.6534581780433655,
+      "learning_rate": 0.0003536794387479762,
+      "loss": 3.3655,
       "step": 38150
     },
     {
-      "epoch": 4.111505758260682,
-      "grad_norm": 0.6210476756095886,
-      "learning_rate": 0.0003537743777610171,
-      "loss": 3.3621,
+      "epoch": 4.118598382749326,
+      "grad_norm": 0.6667327284812927,
+      "learning_rate": 0.0003533556395035078,
+      "loss": 3.3894,
       "step": 38200
     },
     {
-      "epoch": 4.1168873103002905,
-      "grad_norm": 0.6711219549179077,
-      "learning_rate": 0.00035345113673095573,
-      "loss": 3.3686,
+      "epoch": 4.123989218328841,
+      "grad_norm": 0.6402404308319092,
+      "learning_rate": 0.00035303184025903934,
+      "loss": 3.3812,
       "step": 38250
     },
     {
-      "epoch": 4.122268862339899,
-      "grad_norm": 0.6280274391174316,
-      "learning_rate": 0.00035312789570089427,
-      "loss": 3.3742,
+      "epoch": 4.129380053908355,
+      "grad_norm": 0.6722440123558044,
+      "learning_rate": 0.00035270804101457094,
+      "loss": 3.383,
       "step": 38300
     },
     {
-      "epoch": 4.127650414379507,
-      "grad_norm": 0.6348408460617065,
-      "learning_rate": 0.00035280465467083286,
-      "loss": 3.3871,
+      "epoch": 4.1347708894878705,
+      "grad_norm": 0.6048206686973572,
+      "learning_rate": 0.0003523842417701025,
+      "loss": 3.3728,
       "step": 38350
     },
     {
-      "epoch": 4.133031966419115,
-      "grad_norm": 0.6895505785942078,
-      "learning_rate": 0.0003524814136407714,
-      "loss": 3.3586,
+      "epoch": 4.140161725067386,
+      "grad_norm": 0.6511948704719543,
+      "learning_rate": 0.0003520604425256341,
+      "loss": 3.368,
       "step": 38400
     },
     {
-      "epoch": 4.138413518458724,
-      "grad_norm": 0.7086821794509888,
-      "learning_rate": 0.00035215817261071,
-      "loss": 3.3695,
+      "epoch": 4.1455525606469,
+      "grad_norm": 0.6331550478935242,
+      "learning_rate": 0.00035173664328116565,
+      "loss": 3.3754,
       "step": 38450
     },
     {
-      "epoch": 4.1437950704983315,
-      "grad_norm": 0.6616617441177368,
-      "learning_rate": 0.00035183493158064865,
-      "loss": 3.382,
+      "epoch": 4.150943396226415,
+      "grad_norm": 0.6639741063117981,
+      "learning_rate": 0.00035141284403669725,
+      "loss": 3.3846,
       "step": 38500
     },
     {
-      "epoch": 4.14917662253794,
-      "grad_norm": 0.6920830607414246,
-      "learning_rate": 0.0003515116905505872,
-      "loss": 3.4122,
+      "epoch": 4.15633423180593,
+      "grad_norm": 0.6906835436820984,
+      "learning_rate": 0.0003510890447922288,
+      "loss": 3.3728,
       "step": 38550
     },
     {
-      "epoch": 4.154558174577549,
-      "grad_norm": 0.7117165923118591,
-      "learning_rate": 0.0003511884495205258,
-      "loss": 3.3802,
+      "epoch": 4.1617250673854445,
+      "grad_norm": 0.6332456469535828,
+      "learning_rate": 0.00035076524554776035,
+      "loss": 3.3699,
       "step": 38600
     },
     {
-      "epoch": 4.159939726617156,
-      "grad_norm": 0.6522443890571594,
-      "learning_rate": 0.0003508652084904644,
-      "loss": 3.3718,
+      "epoch": 4.16711590296496,
+      "grad_norm": 0.676518976688385,
+      "learning_rate": 0.00035044144630329196,
+      "loss": 3.3966,
       "step": 38650
     },
     {
-      "epoch": 4.165321278656765,
-      "grad_norm": 0.7154022455215454,
-      "learning_rate": 0.0003505419674604029,
-      "loss": 3.3929,
+      "epoch": 4.172506738544475,
+      "grad_norm": 0.6861902475357056,
+      "learning_rate": 0.00035011764705882346,
+      "loss": 3.3753,
       "step": 38700
     },
     {
-      "epoch": 4.1707028306963725,
-      "grad_norm": 0.6403151750564575,
-      "learning_rate": 0.0003502187264303415,
-      "loss": 3.3938,
+      "epoch": 4.177897574123989,
+      "grad_norm": 0.652313232421875,
+      "learning_rate": 0.0003497938478143551,
+      "loss": 3.3678,
       "step": 38750
     },
     {
-      "epoch": 4.176084382735981,
-      "grad_norm": 0.5948874354362488,
-      "learning_rate": 0.00034989548540028016,
+      "epoch": 4.183288409703504,
+      "grad_norm": 0.7183229923248291,
+      "learning_rate": 0.0003494700485698866,
       "loss": 3.3894,
       "step": 38800
     },
     {
-      "epoch": 4.18146593477559,
-      "grad_norm": 0.6313942074775696,
-      "learning_rate": 0.0003495722443702187,
-      "loss": 3.384,
+      "epoch": 4.188679245283019,
+      "grad_norm": 0.6372385025024414,
+      "learning_rate": 0.0003491462493254182,
+      "loss": 3.3929,
       "step": 38850
     },
     {
-      "epoch": 4.186847486815197,
-      "grad_norm": 0.6668844223022461,
-      "learning_rate": 0.0003492490033401573,
-      "loss": 3.3911,
+      "epoch": 4.1940700808625335,
+      "grad_norm": 0.6528982520103455,
+      "learning_rate": 0.00034882245008094977,
+      "loss": 3.3923,
       "step": 38900
     },
     {
-      "epoch": 4.192229038854806,
-      "grad_norm": 0.6406861543655396,
-      "learning_rate": 0.00034892576231009584,
-      "loss": 3.3706,
+      "epoch": 4.199460916442049,
+      "grad_norm": 0.6541169285774231,
+      "learning_rate": 0.0003484986508364813,
+      "loss": 3.3841,
       "step": 38950
     },
     {
-      "epoch": 4.197610590894414,
-      "grad_norm": 0.6086711883544922,
-      "learning_rate": 0.00034860252128003443,
-      "loss": 3.3682,
+      "epoch": 4.204851752021563,
+      "grad_norm": 0.6341118812561035,
+      "learning_rate": 0.0003481748515920129,
+      "loss": 3.3802,
       "step": 39000
     },
     {
-      "epoch": 4.197610590894414,
-      "eval_accuracy": 0.3768287503364165,
-      "eval_loss": 3.4479315280914307,
-      "eval_runtime": 181.8639,
-      "eval_samples_per_second": 99.036,
-      "eval_steps_per_second": 6.191,
+      "epoch": 4.204851752021563,
+      "eval_accuracy": 0.3770355167819275,
+      "eval_loss": 3.4421355724334717,
+      "eval_runtime": 184.0152,
+      "eval_samples_per_second": 97.878,
+      "eval_steps_per_second": 6.119,
       "step": 39000
     },
     {
-      "epoch": 4.202992142934022,
-      "grad_norm": 0.6679412722587585,
-      "learning_rate": 0.000348279280249973,
-      "loss": 3.3828,
+      "epoch": 4.210242587601078,
+      "grad_norm": 0.6534522175788879,
+      "learning_rate": 0.00034785105234754447,
+      "loss": 3.3794,
       "step": 39050
     },
     {
-      "epoch": 4.208373694973631,
-      "grad_norm": 0.6079999804496765,
-      "learning_rate": 0.0003479560392199116,
-      "loss": 3.3802,
+      "epoch": 4.215633423180593,
+      "grad_norm": 0.6660813093185425,
+      "learning_rate": 0.0003475272531030761,
+      "loss": 3.3816,
       "step": 39100
     },
     {
-      "epoch": 4.213755247013238,
-      "grad_norm": 0.629065215587616,
-      "learning_rate": 0.00034763926301045137,
-      "loss": 3.3894,
+      "epoch": 4.2210242587601075,
+      "grad_norm": 0.6584950685501099,
+      "learning_rate": 0.00034720345385860763,
+      "loss": 3.3768,
       "step": 39150
     },
     {
-      "epoch": 4.219136799052847,
-      "grad_norm": 0.6328132152557373,
-      "learning_rate": 0.00034731602198039,
-      "loss": 3.3766,
+      "epoch": 4.226415094339623,
+      "grad_norm": 0.601065456867218,
+      "learning_rate": 0.00034687965461413923,
+      "loss": 3.374,
       "step": 39200
     },
     {
-      "epoch": 4.224518351092455,
-      "grad_norm": 0.6607147455215454,
-      "learning_rate": 0.0003469927809503286,
-      "loss": 3.3742,
+      "epoch": 4.231805929919138,
+      "grad_norm": 0.703972339630127,
+      "learning_rate": 0.0003465558553696708,
+      "loss": 3.3717,
       "step": 39250
     },
     {
-      "epoch": 4.229899903132063,
-      "grad_norm": 0.6375463008880615,
-      "learning_rate": 0.00034666953992026716,
-      "loss": 3.3871,
+      "epoch": 4.237196765498652,
+      "grad_norm": 0.6325825452804565,
+      "learning_rate": 0.0003462320561252024,
+      "loss": 3.3852,
       "step": 39300
     },
     {
-      "epoch": 4.2352814551716715,
-      "grad_norm": 0.6818897724151611,
-      "learning_rate": 0.00034634629889020575,
-      "loss": 3.3868,
+      "epoch": 4.242587601078167,
+      "grad_norm": 0.6532464623451233,
+      "learning_rate": 0.00034590825688073394,
+      "loss": 3.384,
       "step": 39350
     },
     {
-      "epoch": 4.24066300721128,
-      "grad_norm": 0.6507583856582642,
-      "learning_rate": 0.00034602305786014435,
-      "loss": 3.3852,
+      "epoch": 4.247978436657682,
+      "grad_norm": 0.6008528470993042,
+      "learning_rate": 0.00034558445763626543,
+      "loss": 3.3898,
       "step": 39400
     },
     {
-      "epoch": 4.246044559250888,
-      "grad_norm": 0.649346113204956,
-      "learning_rate": 0.00034569981683008294,
-      "loss": 3.3816,
+      "epoch": 4.2533692722371965,
+      "grad_norm": 0.6775292158126831,
+      "learning_rate": 0.00034526065839179704,
+      "loss": 3.3776,
       "step": 39450
     },
     {
-      "epoch": 4.251426111290496,
-      "grad_norm": 0.6513179540634155,
-      "learning_rate": 0.00034537657580002154,
-      "loss": 3.3709,
+      "epoch": 4.258760107816712,
+      "grad_norm": 0.6579444408416748,
+      "learning_rate": 0.0003449368591473286,
+      "loss": 3.3708,
       "step": 39500
     },
     {
-      "epoch": 4.256807663330104,
-      "grad_norm": 0.7123865485191345,
-      "learning_rate": 0.00034505333476996013,
-      "loss": 3.3986,
+      "epoch": 4.264150943396227,
+      "grad_norm": 0.688872754573822,
+      "learning_rate": 0.0003446130599028602,
+      "loss": 3.3756,
       "step": 39550
     },
     {
-      "epoch": 4.2621892153697125,
-      "grad_norm": 0.7108580470085144,
-      "learning_rate": 0.00034473009373989867,
-      "loss": 3.3923,
+      "epoch": 4.269541778975741,
+      "grad_norm": 0.6341859698295593,
+      "learning_rate": 0.00034428926065839174,
+      "loss": 3.3747,
       "step": 39600
     },
     {
-      "epoch": 4.267570767409321,
-      "grad_norm": 0.6283742785453796,
-      "learning_rate": 0.00034440685270983727,
-      "loss": 3.401,
+      "epoch": 4.274932614555256,
+      "grad_norm": 0.6867018938064575,
+      "learning_rate": 0.00034396546141392335,
+      "loss": 3.4022,
       "step": 39650
     },
     {
-      "epoch": 4.272952319448929,
-      "grad_norm": 0.6473816633224487,
-      "learning_rate": 0.0003440836116797758,
-      "loss": 3.391,
+      "epoch": 4.280323450134771,
+      "grad_norm": 0.7251633405685425,
+      "learning_rate": 0.0003436416621694549,
+      "loss": 3.3851,
       "step": 39700
     },
     {
-      "epoch": 4.278333871488537,
-      "grad_norm": 0.6747614145278931,
-      "learning_rate": 0.00034376037064971445,
-      "loss": 3.401,
+      "epoch": 4.285714285714286,
+      "grad_norm": 1.0483514070510864,
+      "learning_rate": 0.0003433178629249865,
+      "loss": 3.379,
       "step": 39750
     },
     {
-      "epoch": 4.283715423528146,
-      "grad_norm": 0.6350759863853455,
-      "learning_rate": 0.00034343712961965305,
-      "loss": 3.4075,
+      "epoch": 4.291105121293801,
+      "grad_norm": 0.6018542051315308,
+      "learning_rate": 0.00034299406368051806,
+      "loss": 3.3955,
       "step": 39800
     },
     {
-      "epoch": 4.2890969755677535,
-      "grad_norm": 0.7255776524543762,
-      "learning_rate": 0.0003431138885895916,
-      "loss": 3.3835,
+      "epoch": 4.296495956873315,
+      "grad_norm": 0.6951125860214233,
+      "learning_rate": 0.0003426702644360496,
+      "loss": 3.3968,
       "step": 39850
     },
     {
-      "epoch": 4.294478527607362,
-      "grad_norm": 0.7065731287002563,
-      "learning_rate": 0.00034279711238013145,
-      "loss": 3.396,
+      "epoch": 4.30188679245283,
+      "grad_norm": 0.6492907404899597,
+      "learning_rate": 0.0003423464651915812,
+      "loss": 3.3865,
       "step": 39900
     },
     {
-      "epoch": 4.299860079646971,
-      "grad_norm": 0.67261803150177,
-      "learning_rate": 0.00034247387135007,
-      "loss": 3.3983,
+      "epoch": 4.307277628032345,
+      "grad_norm": 0.6721863150596619,
+      "learning_rate": 0.00034202266594711276,
+      "loss": 3.3901,
       "step": 39950
     },
     {
-      "epoch": 4.305241631686578,
-      "grad_norm": 0.6579726934432983,
-      "learning_rate": 0.0003421506303200086,
-      "loss": 3.3876,
+      "epoch": 4.3126684636118595,
+      "grad_norm": 0.6667438745498657,
+      "learning_rate": 0.00034169886670264437,
+      "loss": 3.3836,
       "step": 40000
     },
     {
-      "epoch": 4.305241631686578,
-      "eval_accuracy": 0.3775528131808122,
-      "eval_loss": 3.441394805908203,
-      "eval_runtime": 181.6119,
-      "eval_samples_per_second": 99.173,
-      "eval_steps_per_second": 6.2,
+      "epoch": 4.3126684636118595,
+      "eval_accuracy": 0.37750283285241754,
+      "eval_loss": 3.4364352226257324,
+      "eval_runtime": 184.1867,
+      "eval_samples_per_second": 97.787,
+      "eval_steps_per_second": 6.113,
       "step": 40000
     },
     {
-      "epoch": 4.310623183726187,
-      "grad_norm": 0.6545993089675903,
-      "learning_rate": 0.0003418273892899471,
-      "loss": 3.382,
+      "epoch": 4.318059299191375,
+      "grad_norm": 0.6284617781639099,
+      "learning_rate": 0.00034137506745817586,
+      "loss": 3.3761,
       "step": 40050
     },
     {
-      "epoch": 4.3160047357657945,
-      "grad_norm": 0.5974786877632141,
-      "learning_rate": 0.0003415041482598858,
-      "loss": 3.3722,
+      "epoch": 4.32345013477089,
+      "grad_norm": 0.6802538633346558,
+      "learning_rate": 0.0003410512682137075,
+      "loss": 3.4042,
       "step": 40100
     },
     {
-      "epoch": 4.321386287805403,
-      "grad_norm": 0.6922400593757629,
-      "learning_rate": 0.00034118090722982437,
-      "loss": 3.3802,
+      "epoch": 4.328840970350404,
+      "grad_norm": 0.6353877186775208,
+      "learning_rate": 0.000340727468969239,
+      "loss": 3.4114,
       "step": 40150
     },
     {
-      "epoch": 4.326767839845012,
-      "grad_norm": 0.6838219165802002,
-      "learning_rate": 0.0003408576661997629,
-      "loss": 3.39,
+      "epoch": 4.334231805929919,
+      "grad_norm": 0.6325104236602783,
+      "learning_rate": 0.0003404036697247706,
+      "loss": 3.38,
       "step": 40200
     },
     {
-      "epoch": 4.332149391884619,
-      "grad_norm": 0.6342960000038147,
-      "learning_rate": 0.0003405344251697015,
-      "loss": 3.3819,
+      "epoch": 4.339622641509434,
+      "grad_norm": 0.6528483629226685,
+      "learning_rate": 0.0003400798704803022,
+      "loss": 3.3855,
       "step": 40250
     },
     {
-      "epoch": 4.337530943924228,
-      "grad_norm": 0.7120860815048218,
-      "learning_rate": 0.0003402111841396401,
-      "loss": 3.374,
+      "epoch": 4.345013477088949,
+      "grad_norm": 0.6801654696464539,
+      "learning_rate": 0.0003397560712358337,
+      "loss": 3.3931,
       "step": 40300
     },
     {
-      "epoch": 4.342912495963836,
-      "grad_norm": 0.648949921131134,
-      "learning_rate": 0.00033988794310957864,
-      "loss": 3.3776,
+      "epoch": 4.350404312668464,
+      "grad_norm": 0.6602048277854919,
+      "learning_rate": 0.00033943227199136533,
+      "loss": 3.3853,
       "step": 40350
     },
     {
-      "epoch": 4.348294048003444,
-      "grad_norm": 0.7276732325553894,
-      "learning_rate": 0.0003395647020795173,
-      "loss": 3.4097,
+      "epoch": 4.355795148247978,
+      "grad_norm": 0.708645761013031,
+      "learning_rate": 0.0003391084727468969,
+      "loss": 3.4001,
       "step": 40400
     },
     {
-      "epoch": 4.3536756000430525,
-      "grad_norm": 0.6344953179359436,
-      "learning_rate": 0.0003392414610494559,
-      "loss": 3.3754,
+      "epoch": 4.361185983827493,
+      "grad_norm": 0.678924024105072,
+      "learning_rate": 0.0003387846735024285,
+      "loss": 3.3711,
       "step": 40450
     },
     {
-      "epoch": 4.359057152082661,
-      "grad_norm": 0.6971585154533386,
-      "learning_rate": 0.0003389182200193944,
-      "loss": 3.4039,
+      "epoch": 4.366576819407008,
+      "grad_norm": 0.6869363784790039,
+      "learning_rate": 0.00033846087425796003,
+      "loss": 3.3749,
       "step": 40500
     },
     {
-      "epoch": 4.364438704122269,
-      "grad_norm": 0.6294052600860596,
-      "learning_rate": 0.000338594978989333,
-      "loss": 3.3791,
+      "epoch": 4.3719676549865225,
+      "grad_norm": 0.6870575547218323,
+      "learning_rate": 0.00033813707501349164,
+      "loss": 3.3985,
       "step": 40550
     },
     {
-      "epoch": 4.369820256161877,
-      "grad_norm": 0.6447839140892029,
-      "learning_rate": 0.00033827173795927156,
-      "loss": 3.3871,
+      "epoch": 4.377358490566038,
+      "grad_norm": 0.6759331226348877,
+      "learning_rate": 0.0003378132757690232,
+      "loss": 3.3822,
       "step": 40600
     },
     {
-      "epoch": 4.375201808201485,
-      "grad_norm": 0.6472920179367065,
-      "learning_rate": 0.00033794849692921015,
-      "loss": 3.3886,
+      "epoch": 4.382749326145553,
+      "grad_norm": 0.6437086462974548,
+      "learning_rate": 0.00033748947652455474,
+      "loss": 3.374,
       "step": 40650
     },
     {
-      "epoch": 4.3805833602410935,
-      "grad_norm": 0.6552985310554504,
-      "learning_rate": 0.0003376252558991488,
-      "loss": 3.3904,
+      "epoch": 4.388140161725067,
+      "grad_norm": 0.6415377855300903,
+      "learning_rate": 0.00033716567728008635,
+      "loss": 3.4025,
       "step": 40700
     },
     {
-      "epoch": 4.385964912280702,
-      "grad_norm": 0.6818690896034241,
-      "learning_rate": 0.00033730201486908734,
-      "loss": 3.3947,
+      "epoch": 4.393530997304582,
+      "grad_norm": 0.6618736386299133,
+      "learning_rate": 0.00033684187803561784,
+      "loss": 3.3988,
       "step": 40750
     },
     {
-      "epoch": 4.39134646432031,
-      "grad_norm": 0.646091639995575,
-      "learning_rate": 0.00033697877383902594,
-      "loss": 3.4024,
+      "epoch": 4.398921832884097,
+      "grad_norm": 0.7001456022262573,
+      "learning_rate": 0.0003365245547760388,
+      "loss": 3.3731,
       "step": 40800
     },
     {
-      "epoch": 4.396728016359918,
-      "grad_norm": 0.6999764442443848,
-      "learning_rate": 0.00033665553280896453,
-      "loss": 3.391,
+      "epoch": 4.404312668463612,
+      "grad_norm": 0.6538615822792053,
+      "learning_rate": 0.0003362007555315704,
+      "loss": 3.3944,
       "step": 40850
     },
     {
-      "epoch": 4.402109568399527,
-      "grad_norm": 0.6619383096694946,
-      "learning_rate": 0.00033633229177890307,
-      "loss": 3.4125,
+      "epoch": 4.409703504043127,
+      "grad_norm": 0.6283511519432068,
+      "learning_rate": 0.00033587695628710195,
+      "loss": 3.3806,
       "step": 40900
     },
     {
-      "epoch": 4.4074911204391345,
-      "grad_norm": 0.8046988844871521,
-      "learning_rate": 0.00033600905074884167,
-      "loss": 3.373,
+      "epoch": 4.415094339622642,
+      "grad_norm": 0.681976318359375,
+      "learning_rate": 0.00033555315704263355,
+      "loss": 3.3863,
       "step": 40950
     },
     {
-      "epoch": 4.412872672478743,
-      "grad_norm": 0.692973256111145,
-      "learning_rate": 0.0003356858097187803,
-      "loss": 3.3813,
+      "epoch": 4.420485175202156,
+      "grad_norm": 0.6590644121170044,
+      "learning_rate": 0.0003352293577981651,
+      "loss": 3.3988,
       "step": 41000
     },
     {
-      "epoch": 4.412872672478743,
-      "eval_accuracy": 0.37769666960427856,
-      "eval_loss": 3.4349679946899414,
-      "eval_runtime": 181.1247,
-      "eval_samples_per_second": 99.44,
-      "eval_steps_per_second": 6.217,
+      "epoch": 4.420485175202156,
+      "eval_accuracy": 0.3779914448889193,
+      "eval_loss": 3.4323019981384277,
+      "eval_runtime": 184.2839,
+      "eval_samples_per_second": 97.735,
+      "eval_steps_per_second": 6.11,
       "step": 41000
     },
     {
-      "epoch": 4.418254224518351,
-      "grad_norm": 0.6343230605125427,
-      "learning_rate": 0.00033536256868871886,
-      "loss": 3.3868,
+      "epoch": 4.425876010781671,
+      "grad_norm": 0.6442763209342957,
+      "learning_rate": 0.00033490555855369665,
+      "loss": 3.3987,
       "step": 41050
     },
     {
-      "epoch": 4.423635776557959,
-      "grad_norm": 0.6828956604003906,
-      "learning_rate": 0.00033503932765865745,
-      "loss": 3.3779,
+      "epoch": 4.431266846361186,
+      "grad_norm": 0.6701889038085938,
+      "learning_rate": 0.00033458175930922826,
+      "loss": 3.3802,
       "step": 41100
     },
     {
-      "epoch": 4.429017328597568,
-      "grad_norm": 0.6604346632957458,
-      "learning_rate": 0.000334716086628596,
-      "loss": 3.3862,
+      "epoch": 4.436657681940701,
+      "grad_norm": 0.6391051411628723,
+      "learning_rate": 0.0003342579600647598,
+      "loss": 3.3623,
       "step": 41150
     },
     {
-      "epoch": 4.4343988806371755,
-      "grad_norm": 0.7162014842033386,
-      "learning_rate": 0.0003343928455985346,
-      "loss": 3.4053,
+      "epoch": 4.442048517520216,
+      "grad_norm": 0.660970151424408,
+      "learning_rate": 0.0003339341608202914,
+      "loss": 3.3969,
       "step": 41200
     },
     {
-      "epoch": 4.439780432676784,
-      "grad_norm": 0.7162100672721863,
-      "learning_rate": 0.00033406960456847324,
-      "loss": 3.397,
+      "epoch": 4.44743935309973,
+      "grad_norm": 0.6699996590614319,
+      "learning_rate": 0.00033361036157582297,
+      "loss": 3.385,
       "step": 41250
     },
     {
-      "epoch": 4.445161984716393,
-      "grad_norm": 0.6890962719917297,
-      "learning_rate": 0.0003337463635384118,
-      "loss": 3.4001,
+      "epoch": 4.452830188679245,
+      "grad_norm": 0.673334002494812,
+      "learning_rate": 0.00033328656233135457,
+      "loss": 3.3826,
       "step": 41300
     },
     {
-      "epoch": 4.450543536756,
-      "grad_norm": 0.714141309261322,
-      "learning_rate": 0.00033342312250835037,
-      "loss": 3.3843,
+      "epoch": 4.45822102425876,
+      "grad_norm": 0.6280801892280579,
+      "learning_rate": 0.0003329627630868861,
+      "loss": 3.3754,
       "step": 41350
     },
     {
-      "epoch": 4.455925088795609,
-      "grad_norm": 0.6847618222236633,
-      "learning_rate": 0.00033309988147828896,
-      "loss": 3.4102,
+      "epoch": 4.463611859838275,
+      "grad_norm": 0.6734782457351685,
+      "learning_rate": 0.0003326389638424177,
+      "loss": 3.3865,
       "step": 41400
     },
     {
-      "epoch": 4.461306640835216,
-      "grad_norm": 0.6703453063964844,
-      "learning_rate": 0.0003327766404482275,
-      "loss": 3.4016,
+      "epoch": 4.46900269541779,
+      "grad_norm": 0.7083989381790161,
+      "learning_rate": 0.0003323151645979492,
+      "loss": 3.3727,
       "step": 41450
     },
     {
-      "epoch": 4.466688192874825,
-      "grad_norm": 0.6432178616523743,
-      "learning_rate": 0.0003324533994181661,
-      "loss": 3.3857,
+      "epoch": 4.474393530997305,
+      "grad_norm": 0.6066727042198181,
+      "learning_rate": 0.00033199136535348077,
+      "loss": 3.3761,
       "step": 41500
     },
     {
-      "epoch": 4.4720697449144335,
-      "grad_norm": 0.6851868033409119,
-      "learning_rate": 0.00033213015838810475,
-      "loss": 3.4059,
+      "epoch": 4.479784366576819,
+      "grad_norm": 0.6598591208457947,
+      "learning_rate": 0.0003316675661090124,
+      "loss": 3.405,
       "step": 41550
     },
     {
-      "epoch": 4.477451296954041,
-      "grad_norm": 0.6780423521995544,
-      "learning_rate": 0.0003318069173580433,
-      "loss": 3.3957,
+      "epoch": 4.485175202156334,
+      "grad_norm": 0.6851210594177246,
+      "learning_rate": 0.00033134376686454393,
+      "loss": 3.3964,
       "step": 41600
     },
     {
-      "epoch": 4.48283284899365,
-      "grad_norm": 0.6518639922142029,
-      "learning_rate": 0.0003314836763279819,
-      "loss": 3.3808,
+      "epoch": 4.490566037735849,
+      "grad_norm": 0.7116537690162659,
+      "learning_rate": 0.00033101996762007553,
+      "loss": 3.3996,
       "step": 41650
     },
     {
-      "epoch": 4.488214401033258,
-      "grad_norm": 0.630031406879425,
-      "learning_rate": 0.0003311604352979204,
-      "loss": 3.3863,
+      "epoch": 4.495956873315364,
+      "grad_norm": 0.6537807583808899,
+      "learning_rate": 0.0003306961683756071,
+      "loss": 3.4022,
       "step": 41700
     },
     {
-      "epoch": 4.493595953072866,
-      "grad_norm": 0.6582158207893372,
-      "learning_rate": 0.000330837194267859,
-      "loss": 3.3722,
+      "epoch": 4.501347708894879,
+      "grad_norm": 0.7407820820808411,
+      "learning_rate": 0.0003303723691311387,
+      "loss": 3.3861,
       "step": 41750
     },
     {
-      "epoch": 4.4989775051124745,
-      "grad_norm": 0.7116129994392395,
-      "learning_rate": 0.0003305139532377976,
-      "loss": 3.389,
+      "epoch": 4.506738544474393,
+      "grad_norm": 0.65674889087677,
+      "learning_rate": 0.00033004856988667024,
+      "loss": 3.3972,
       "step": 41800
     },
     {
-      "epoch": 4.504359057152083,
-      "grad_norm": 0.6760125756263733,
-      "learning_rate": 0.0003301907122077362,
-      "loss": 3.396,
+      "epoch": 4.512129380053908,
+      "grad_norm": 0.6424514651298523,
+      "learning_rate": 0.0003297247706422018,
+      "loss": 3.3755,
       "step": 41850
     },
     {
-      "epoch": 4.509740609191691,
-      "grad_norm": 0.6478492617607117,
-      "learning_rate": 0.0003298674711776748,
-      "loss": 3.401,
+      "epoch": 4.517520215633423,
+      "grad_norm": 0.7046425938606262,
+      "learning_rate": 0.0003294009713977334,
+      "loss": 3.3984,
       "step": 41900
     },
     {
-      "epoch": 4.515122161231299,
-      "grad_norm": 0.6583666205406189,
-      "learning_rate": 0.0003295442301476134,
-      "loss": 3.3866,
+      "epoch": 4.5229110512129385,
+      "grad_norm": 0.7055160999298096,
+      "learning_rate": 0.00032907717215326494,
+      "loss": 3.4006,
       "step": 41950
     },
     {
-      "epoch": 4.520503713270907,
-      "grad_norm": 0.6689998507499695,
-      "learning_rate": 0.00032922098911755194,
-      "loss": 3.3924,
+      "epoch": 4.528301886792453,
+      "grad_norm": 0.6497727632522583,
+      "learning_rate": 0.00032875337290879655,
+      "loss": 3.3984,
       "step": 42000
     },
     {
-      "epoch": 4.520503713270907,
-      "eval_accuracy": 0.37855687451710585,
-      "eval_loss": 3.431107521057129,
-      "eval_runtime": 181.9242,
-      "eval_samples_per_second": 99.003,
-      "eval_steps_per_second": 6.189,
+      "epoch": 4.528301886792453,
+      "eval_accuracy": 0.3786226095142336,
+      "eval_loss": 3.4280104637145996,
+      "eval_runtime": 184.3021,
+      "eval_samples_per_second": 97.725,
+      "eval_steps_per_second": 6.11,
       "step": 42000
     },
     {
-      "epoch": 4.5258852653105155,
-      "grad_norm": 0.655572235584259,
-      "learning_rate": 0.00032889774808749053,
-      "loss": 3.407,
+      "epoch": 4.533692722371968,
+      "grad_norm": 0.7207480669021606,
+      "learning_rate": 0.00032842957366432805,
+      "loss": 3.3754,
       "step": 42050
     },
     {
-      "epoch": 4.531266817350124,
-      "grad_norm": 0.6405171155929565,
-      "learning_rate": 0.0003285745070574292,
-      "loss": 3.4029,
+      "epoch": 4.539083557951482,
+      "grad_norm": 0.7041564583778381,
+      "learning_rate": 0.0003281057744198597,
+      "loss": 3.3845,
       "step": 42100
     },
     {
-      "epoch": 4.536648369389732,
-      "grad_norm": 0.6641229391098022,
-      "learning_rate": 0.0003282512660273677,
-      "loss": 3.3999,
+      "epoch": 4.544474393530997,
+      "grad_norm": 0.7177610397338867,
+      "learning_rate": 0.0003277819751753912,
+      "loss": 3.3904,
       "step": 42150
     },
     {
-      "epoch": 4.54202992142934,
-      "grad_norm": 0.6896253824234009,
-      "learning_rate": 0.0003279280249973063,
-      "loss": 3.4085,
+      "epoch": 4.549865229110512,
+      "grad_norm": 0.7152977585792542,
+      "learning_rate": 0.0003274581759309228,
+      "loss": 3.4014,
       "step": 42200
     },
     {
-      "epoch": 4.547411473468949,
-      "grad_norm": 0.6799988746643066,
-      "learning_rate": 0.00032760478396724486,
-      "loss": 3.4027,
+      "epoch": 4.555256064690027,
+      "grad_norm": 0.673838198184967,
+      "learning_rate": 0.00032713437668645436,
+      "loss": 3.3953,
       "step": 42250
     },
     {
-      "epoch": 4.5527930255085565,
-      "grad_norm": 0.6822940111160278,
-      "learning_rate": 0.00032728154293718345,
-      "loss": 3.3753,
+      "epoch": 4.560646900269542,
+      "grad_norm": 0.6616833209991455,
+      "learning_rate": 0.0003268105774419859,
+      "loss": 3.3859,
       "step": 42300
     },
     {
-      "epoch": 4.558174577548165,
-      "grad_norm": 0.656019389629364,
-      "learning_rate": 0.00032695830190712205,
-      "loss": 3.3807,
+      "epoch": 4.566037735849057,
+      "grad_norm": 0.6552806496620178,
+      "learning_rate": 0.0003264867781975175,
+      "loss": 3.4087,
       "step": 42350
     },
     {
-      "epoch": 4.563556129587774,
-      "grad_norm": 0.650692880153656,
-      "learning_rate": 0.00032663506087706064,
-      "loss": 3.4062,
+      "epoch": 4.571428571428571,
+      "grad_norm": 0.640472412109375,
+      "learning_rate": 0.00032616297895304906,
+      "loss": 3.387,
       "step": 42400
     },
     {
-      "epoch": 4.568937681627381,
-      "grad_norm": 0.6641072630882263,
-      "learning_rate": 0.00032631181984699924,
-      "loss": 3.4034,
+      "epoch": 4.576819407008086,
+      "grad_norm": 0.6295151114463806,
+      "learning_rate": 0.00032583917970858067,
+      "loss": 3.3974,
       "step": 42450
     },
     {
-      "epoch": 4.57431923366699,
-      "grad_norm": 0.6998100280761719,
-      "learning_rate": 0.00032598857881693783,
-      "loss": 3.394,
+      "epoch": 4.5822102425876015,
+      "grad_norm": 0.7369704246520996,
+      "learning_rate": 0.0003255153804641122,
+      "loss": 3.3807,
       "step": 42500
     },
     {
-      "epoch": 4.579700785706597,
-      "grad_norm": 0.7412688732147217,
-      "learning_rate": 0.00032566533778687637,
-      "loss": 3.3865,
+      "epoch": 4.587601078167116,
+      "grad_norm": 0.6615971922874451,
+      "learning_rate": 0.0003251915812196438,
+      "loss": 3.3935,
       "step": 42550
     },
     {
-      "epoch": 4.585082337746206,
-      "grad_norm": 0.6878949999809265,
-      "learning_rate": 0.00032534209675681497,
-      "loss": 3.3875,
+      "epoch": 4.592991913746631,
+      "grad_norm": 0.703863799571991,
+      "learning_rate": 0.00032486778197517537,
+      "loss": 3.3742,
       "step": 42600
     },
     {
-      "epoch": 4.5904638897858145,
-      "grad_norm": 0.718602180480957,
-      "learning_rate": 0.0003250188557267535,
-      "loss": 3.4083,
+      "epoch": 4.598382749326145,
+      "grad_norm": 0.6635894775390625,
+      "learning_rate": 0.000324543982730707,
+      "loss": 3.4012,
       "step": 42650
     },
     {
-      "epoch": 4.595845441825422,
-      "grad_norm": 0.6992116570472717,
-      "learning_rate": 0.00032469561469669215,
-      "loss": 3.3877,
+      "epoch": 4.60377358490566,
+      "grad_norm": 0.6182754635810852,
+      "learning_rate": 0.00032422018348623853,
+      "loss": 3.394,
       "step": 42700
     },
     {
-      "epoch": 4.601226993865031,
-      "grad_norm": 0.6840443015098572,
-      "learning_rate": 0.00032437237366663075,
-      "loss": 3.3814,
+      "epoch": 4.609164420485175,
+      "grad_norm": 0.6388627886772156,
+      "learning_rate": 0.00032389638424177,
+      "loss": 3.3862,
       "step": 42750
     },
     {
-      "epoch": 4.606608545904638,
-      "grad_norm": 0.6726332902908325,
-      "learning_rate": 0.0003240491326365693,
-      "loss": 3.4107,
+      "epoch": 4.6145552560646905,
+      "grad_norm": 0.6418387293815613,
+      "learning_rate": 0.00032357258499730163,
+      "loss": 3.3848,
       "step": 42800
     },
     {
-      "epoch": 4.611990097944247,
-      "grad_norm": 0.7318175435066223,
-      "learning_rate": 0.0003237258916065079,
-      "loss": 3.3882,
+      "epoch": 4.619946091644205,
+      "grad_norm": 0.6612409949302673,
+      "learning_rate": 0.0003232487857528332,
+      "loss": 3.3937,
       "step": 42850
     },
     {
-      "epoch": 4.6173716499838555,
-      "grad_norm": 0.6592293977737427,
-      "learning_rate": 0.0003234026505764465,
-      "loss": 3.4084,
+      "epoch": 4.62533692722372,
+      "grad_norm": 0.6577247381210327,
+      "learning_rate": 0.0003229249865083648,
+      "loss": 3.4042,
       "step": 42900
     },
     {
-      "epoch": 4.622753202023463,
-      "grad_norm": 0.70222407579422,
-      "learning_rate": 0.000323079409546385,
-      "loss": 3.3907,
+      "epoch": 4.630727762803234,
+      "grad_norm": 0.6248788833618164,
+      "learning_rate": 0.00032260118726389634,
+      "loss": 3.4151,
       "step": 42950
     },
     {
-      "epoch": 4.628134754063072,
-      "grad_norm": 0.6887460350990295,
-      "learning_rate": 0.0003227626333369248,
-      "loss": 3.3912,
+      "epoch": 4.636118598382749,
+      "grad_norm": 0.6932069659233093,
+      "learning_rate": 0.00032227738801942794,
+      "loss": 3.3897,
       "step": 43000
     },
     {
-      "epoch": 4.628134754063072,
-      "eval_accuracy": 0.3786048990935198,
-      "eval_loss": 3.426353693008423,
-      "eval_runtime": 183.3187,
-      "eval_samples_per_second": 98.25,
-      "eval_steps_per_second": 6.142,
+      "epoch": 4.636118598382749,
+      "eval_accuracy": 0.3789400932524275,
+      "eval_loss": 3.4228014945983887,
+      "eval_runtime": 183.9109,
+      "eval_samples_per_second": 97.933,
+      "eval_steps_per_second": 6.123,
       "step": 43000
     },
     {
-      "epoch": 4.63351630610268,
-      "grad_norm": 0.6995670795440674,
-      "learning_rate": 0.0003224393923068635,
-      "loss": 3.3847,
+      "epoch": 4.6415094339622645,
+      "grad_norm": 0.6546887159347534,
+      "learning_rate": 0.0003219535887749595,
+      "loss": 3.3913,
       "step": 43050
     },
     {
-      "epoch": 4.638897858142288,
-      "grad_norm": 0.7121273875236511,
-      "learning_rate": 0.00032211615127680207,
-      "loss": 3.3972,
+      "epoch": 4.646900269541779,
+      "grad_norm": 0.670707106590271,
+      "learning_rate": 0.0003216297895304911,
+      "loss": 3.3865,
       "step": 43100
     },
     {
-      "epoch": 4.6442794101818965,
-      "grad_norm": 0.6917558908462524,
-      "learning_rate": 0.0003217929102467406,
-      "loss": 3.4062,
+      "epoch": 4.652291105121294,
+      "grad_norm": 0.6958152055740356,
+      "learning_rate": 0.00032130599028602265,
+      "loss": 3.3945,
       "step": 43150
     },
     {
-      "epoch": 4.649660962221505,
-      "grad_norm": 0.7152304649353027,
-      "learning_rate": 0.0003214696692166792,
-      "loss": 3.3848,
+      "epoch": 4.657681940700809,
+      "grad_norm": 0.7048361897468567,
+      "learning_rate": 0.0003209821910415542,
+      "loss": 3.3928,
       "step": 43200
     },
     {
-      "epoch": 4.655042514261113,
-      "grad_norm": 0.6601977944374084,
-      "learning_rate": 0.0003211464281866178,
-      "loss": 3.3972,
+      "epoch": 4.663072776280323,
+      "grad_norm": 0.6394913196563721,
+      "learning_rate": 0.0003206583917970858,
+      "loss": 3.3815,
       "step": 43250
     },
     {
-      "epoch": 4.660424066300721,
-      "grad_norm": 0.6530590057373047,
-      "learning_rate": 0.00032082318715655634,
-      "loss": 3.388,
+      "epoch": 4.668463611859838,
+      "grad_norm": 0.6621165871620178,
+      "learning_rate": 0.00032033459255261735,
+      "loss": 3.3808,
       "step": 43300
     },
     {
-      "epoch": 4.665805618340329,
-      "grad_norm": 0.6723693013191223,
-      "learning_rate": 0.000320499946126495,
-      "loss": 3.3952,
+      "epoch": 4.6738544474393535,
+      "grad_norm": 0.646432638168335,
+      "learning_rate": 0.00032001079330814896,
+      "loss": 3.38,
       "step": 43350
     },
     {
-      "epoch": 4.6711871703799375,
-      "grad_norm": 0.6972584128379822,
-      "learning_rate": 0.0003201767050964336,
-      "loss": 3.4,
+      "epoch": 4.679245283018868,
+      "grad_norm": 0.6301554441452026,
+      "learning_rate": 0.00031968699406368045,
+      "loss": 3.405,
       "step": 43400
     },
     {
-      "epoch": 4.676568722419546,
-      "grad_norm": 0.692314088344574,
-      "learning_rate": 0.0003198534640663721,
-      "loss": 3.3883,
+      "epoch": 4.684636118598383,
+      "grad_norm": 0.7304412126541138,
+      "learning_rate": 0.0003193631948192121,
+      "loss": 3.3987,
       "step": 43450
     },
     {
-      "epoch": 4.681950274459154,
-      "grad_norm": 0.6952795386314392,
-      "learning_rate": 0.0003195302230363107,
-      "loss": 3.3966,
+      "epoch": 4.690026954177897,
+      "grad_norm": 0.6604732275009155,
+      "learning_rate": 0.0003190393955747436,
+      "loss": 3.4088,
       "step": 43500
     },
     {
-      "epoch": 4.687331826498762,
-      "grad_norm": 0.6832876205444336,
-      "learning_rate": 0.00031920698200624926,
-      "loss": 3.3708,
+      "epoch": 4.695417789757412,
+      "grad_norm": 0.6544510722160339,
+      "learning_rate": 0.00031871559633027516,
+      "loss": 3.3747,
       "step": 43550
     },
     {
-      "epoch": 4.692713378538371,
-      "grad_norm": 0.684529185295105,
-      "learning_rate": 0.00031888374097618785,
-      "loss": 3.3997,
+      "epoch": 4.7008086253369274,
+      "grad_norm": 0.6622135639190674,
+      "learning_rate": 0.00031839179708580676,
+      "loss": 3.3862,
       "step": 43600
     },
     {
-      "epoch": 4.6980949305779784,
-      "grad_norm": 0.6592495441436768,
-      "learning_rate": 0.0003185604999461265,
-      "loss": 3.3893,
+      "epoch": 4.706199460916442,
+      "grad_norm": 0.6810092329978943,
+      "learning_rate": 0.0003180679978413383,
+      "loss": 3.405,
       "step": 43650
     },
     {
-      "epoch": 4.703476482617587,
-      "grad_norm": 0.7246899604797363,
-      "learning_rate": 0.00031823725891606504,
-      "loss": 3.3858,
+      "epoch": 4.711590296495957,
+      "grad_norm": 0.6479467153549194,
+      "learning_rate": 0.0003177441985968699,
+      "loss": 3.3949,
       "step": 43700
     },
     {
-      "epoch": 4.7088580346571955,
-      "grad_norm": 0.7060039043426514,
-      "learning_rate": 0.00031791401788600364,
-      "loss": 3.373,
+      "epoch": 4.716981132075472,
+      "grad_norm": 0.6417392492294312,
+      "learning_rate": 0.00031742039935240147,
+      "loss": 3.3755,
       "step": 43750
     },
     {
-      "epoch": 4.714239586696803,
-      "grad_norm": 0.727345883846283,
-      "learning_rate": 0.00031759077685594223,
-      "loss": 3.3876,
+      "epoch": 4.722371967654986,
+      "grad_norm": 0.6601267457008362,
+      "learning_rate": 0.0003170966001079331,
+      "loss": 3.3851,
       "step": 43800
     },
     {
-      "epoch": 4.719621138736412,
-      "grad_norm": 0.6933861970901489,
-      "learning_rate": 0.00031726753582588077,
-      "loss": 3.4002,
+      "epoch": 4.727762803234501,
+      "grad_norm": 0.6722607612609863,
+      "learning_rate": 0.0003167728008634646,
+      "loss": 3.401,
       "step": 43850
     },
     {
-      "epoch": 4.725002690776019,
-      "grad_norm": 0.7261818647384644,
-      "learning_rate": 0.0003169442947958194,
-      "loss": 3.397,
+      "epoch": 4.7331536388140165,
+      "grad_norm": 0.6935936808586121,
+      "learning_rate": 0.00031644900161899623,
+      "loss": 3.3799,
       "step": 43900
     },
     {
-      "epoch": 4.730384242815628,
-      "grad_norm": 0.6434661149978638,
-      "learning_rate": 0.000316621053765758,
-      "loss": 3.4044,
+      "epoch": 4.738544474393531,
+      "grad_norm": 0.6676217913627625,
+      "learning_rate": 0.0003161252023745278,
+      "loss": 3.39,
       "step": 43950
     },
     {
-      "epoch": 4.7357657948552365,
-      "grad_norm": 0.6868217587471008,
-      "learning_rate": 0.00031629781273569656,
-      "loss": 3.4047,
+      "epoch": 4.743935309973046,
+      "grad_norm": 0.6392797827720642,
+      "learning_rate": 0.00031580140313005933,
+      "loss": 3.3933,
       "step": 44000
     },
     {
-      "epoch": 4.7357657948552365,
-      "eval_accuracy": 0.3793180966491341,
-      "eval_loss": 3.421250581741333,
-      "eval_runtime": 187.5428,
-      "eval_samples_per_second": 96.037,
-      "eval_steps_per_second": 6.004,
+      "epoch": 4.743935309973046,
+      "eval_accuracy": 0.37955876279564316,
+      "eval_loss": 3.418923854827881,
+      "eval_runtime": 183.9112,
+      "eval_samples_per_second": 97.933,
+      "eval_steps_per_second": 6.123,
       "step": 44000
     },
     {
-      "epoch": 4.741147346894844,
-      "grad_norm": 0.6731848120689392,
-      "learning_rate": 0.00031597457170563515,
-      "loss": 3.3793,
+      "epoch": 4.74932614555256,
+      "grad_norm": 0.6438071131706238,
+      "learning_rate": 0.00031547760388559094,
+      "loss": 3.4089,
       "step": 44050
     },
     {
-      "epoch": 4.746528898934453,
-      "grad_norm": 0.6794449687004089,
-      "learning_rate": 0.0003156513306755737,
-      "loss": 3.39,
+      "epoch": 4.754716981132075,
+      "grad_norm": 0.7265536785125732,
+      "learning_rate": 0.00031515380464112243,
+      "loss": 3.3777,
       "step": 44100
     },
     {
-      "epoch": 4.751910450974061,
-      "grad_norm": 0.7639029622077942,
-      "learning_rate": 0.0003153280896455123,
-      "loss": 3.402,
+      "epoch": 4.7601078167115904,
+      "grad_norm": 0.7413805723190308,
+      "learning_rate": 0.00031483000539665404,
+      "loss": 3.388,
       "step": 44150
     },
     {
-      "epoch": 4.757292003013669,
-      "grad_norm": 0.686454176902771,
-      "learning_rate": 0.00031500484861545094,
-      "loss": 3.3893,
+      "epoch": 4.765498652291106,
+      "grad_norm": 0.653020441532135,
+      "learning_rate": 0.0003145062061521856,
+      "loss": 3.3801,
       "step": 44200
     },
     {
-      "epoch": 4.7626735550532775,
-      "grad_norm": 0.7179557085037231,
-      "learning_rate": 0.0003146816075853895,
-      "loss": 3.3874,
+      "epoch": 4.77088948787062,
+      "grad_norm": 0.6411661505699158,
+      "learning_rate": 0.0003141824069077172,
+      "loss": 3.3995,
       "step": 44250
     },
     {
-      "epoch": 4.768055107092886,
-      "grad_norm": 0.6821876168251038,
-      "learning_rate": 0.00031435836655532807,
-      "loss": 3.3928,
+      "epoch": 4.776280323450135,
+      "grad_norm": 0.6204802393913269,
+      "learning_rate": 0.00031385860766324874,
+      "loss": 3.3846,
       "step": 44300
     },
     {
-      "epoch": 4.773436659132494,
-      "grad_norm": 0.6775598526000977,
-      "learning_rate": 0.00031403512552526667,
-      "loss": 3.381,
+      "epoch": 4.781671159029649,
+      "grad_norm": 0.6892203688621521,
+      "learning_rate": 0.00031353480841878035,
+      "loss": 3.3812,
       "step": 44350
     },
     {
-      "epoch": 4.778818211172102,
-      "grad_norm": 0.666349470615387,
-      "learning_rate": 0.0003137118844952052,
-      "loss": 3.4018,
+      "epoch": 4.787061994609164,
+      "grad_norm": 0.6896839141845703,
+      "learning_rate": 0.0003132110091743119,
+      "loss": 3.3901,
       "step": 44400
     },
     {
-      "epoch": 4.78419976321171,
-      "grad_norm": 0.6841625571250916,
-      "learning_rate": 0.0003133886434651438,
-      "loss": 3.3952,
+      "epoch": 4.7924528301886795,
+      "grad_norm": 0.6548848748207092,
+      "learning_rate": 0.00031288720992984345,
+      "loss": 3.4027,
       "step": 44450
     },
     {
-      "epoch": 4.7895813152513185,
-      "grad_norm": 0.6649987697601318,
-      "learning_rate": 0.00031306540243508245,
-      "loss": 3.4116,
+      "epoch": 4.797843665768194,
+      "grad_norm": 0.6836720108985901,
+      "learning_rate": 0.00031256341068537505,
+      "loss": 3.3974,
       "step": 44500
     },
     {
-      "epoch": 4.794962867290927,
-      "grad_norm": 0.6441112756729126,
-      "learning_rate": 0.000312742161405021,
-      "loss": 3.3951,
+      "epoch": 4.803234501347709,
+      "grad_norm": 0.7105928659439087,
+      "learning_rate": 0.0003122396114409066,
+      "loss": 3.3885,
       "step": 44550
     },
     {
-      "epoch": 4.800344419330535,
-      "grad_norm": 0.6589279174804688,
-      "learning_rate": 0.0003124189203749596,
-      "loss": 3.3902,
+      "epoch": 4.808625336927224,
+      "grad_norm": 0.6696828603744507,
+      "learning_rate": 0.0003119158121964382,
+      "loss": 3.3987,
       "step": 44600
     },
     {
-      "epoch": 4.805725971370143,
-      "grad_norm": 0.677620530128479,
-      "learning_rate": 0.0003120956793448981,
-      "loss": 3.3967,
+      "epoch": 4.814016172506738,
+      "grad_norm": 0.6872798800468445,
+      "learning_rate": 0.00031159201295196976,
+      "loss": 3.3948,
       "step": 44650
     },
     {
-      "epoch": 4.811107523409751,
-      "grad_norm": 0.7282724380493164,
-      "learning_rate": 0.0003117724383148367,
-      "loss": 3.386,
+      "epoch": 4.819407008086253,
+      "grad_norm": 0.6887964606285095,
+      "learning_rate": 0.00031126821370750136,
+      "loss": 3.3848,
       "step": 44700
     },
     {
-      "epoch": 4.8164890754493594,
-      "grad_norm": 0.6708424687385559,
-      "learning_rate": 0.00031144919728477526,
-      "loss": 3.3793,
+      "epoch": 4.824797843665769,
+      "grad_norm": 0.6681622266769409,
+      "learning_rate": 0.00031094441446303286,
+      "loss": 3.4018,
       "step": 44750
     },
     {
-      "epoch": 4.821870627488968,
-      "grad_norm": 0.678166925907135,
-      "learning_rate": 0.0003111259562547139,
-      "loss": 3.4066,
+      "epoch": 4.830188679245283,
+      "grad_norm": 0.6492528915405273,
+      "learning_rate": 0.0003106206152185645,
+      "loss": 3.4104,
       "step": 44800
     },
     {
-      "epoch": 4.827252179528576,
-      "grad_norm": 0.6643491387367249,
-      "learning_rate": 0.0003108027152246525,
-      "loss": 3.3997,
+      "epoch": 4.835579514824798,
+      "grad_norm": 0.67716383934021,
+      "learning_rate": 0.00031030329195898536,
+      "loss": 3.3977,
       "step": 44850
     },
     {
-      "epoch": 4.832633731568184,
-      "grad_norm": 0.7268407940864563,
-      "learning_rate": 0.0003104794741945911,
-      "loss": 3.3941,
+      "epoch": 4.840970350404312,
+      "grad_norm": 0.7011221051216125,
+      "learning_rate": 0.00030997949271451697,
+      "loss": 3.405,
       "step": 44900
     },
     {
-      "epoch": 4.838015283607793,
-      "grad_norm": 0.666113555431366,
-      "learning_rate": 0.00031015623316452964,
-      "loss": 3.3933,
+      "epoch": 4.846361185983827,
+      "grad_norm": 0.6942790746688843,
+      "learning_rate": 0.0003096621694549379,
+      "loss": 3.4053,
       "step": 44950
     },
     {
-      "epoch": 4.8433968356474,
-      "grad_norm": 0.6685688495635986,
-      "learning_rate": 0.00030983299213446823,
-      "loss": 3.3776,
+      "epoch": 4.8517520215633425,
+      "grad_norm": 0.8015643358230591,
+      "learning_rate": 0.00030933837021046947,
+      "loss": 3.4001,
       "step": 45000
     },
     {
-      "epoch": 4.8433968356474,
-      "eval_accuracy": 0.37968338765796633,
-      "eval_loss": 3.4163522720336914,
-      "eval_runtime": 202.3061,
-      "eval_samples_per_second": 89.028,
-      "eval_steps_per_second": 5.566,
+      "epoch": 4.8517520215633425,
+      "eval_accuracy": 0.3800734515252203,
+      "eval_loss": 3.4108171463012695,
+      "eval_runtime": 184.1265,
+      "eval_samples_per_second": 97.819,
+      "eval_steps_per_second": 6.115,
       "step": 45000
     },
     {
-      "epoch": 4.848778387687009,
-      "grad_norm": 0.657361626625061,
-      "learning_rate": 0.0003095097511044069,
-      "loss": 3.4067,
+      "epoch": 4.857142857142857,
+      "grad_norm": 0.7075126767158508,
+      "learning_rate": 0.0003090145709660011,
+      "loss": 3.3879,
       "step": 45050
     },
     {
-      "epoch": 4.8541599397266175,
-      "grad_norm": 0.6610834002494812,
-      "learning_rate": 0.0003091865100743454,
-      "loss": 3.4013,
+      "epoch": 4.862533692722372,
+      "grad_norm": 0.6834211349487305,
+      "learning_rate": 0.0003086907717215326,
+      "loss": 3.3926,
       "step": 45100
     },
     {
-      "epoch": 4.859541491766225,
-      "grad_norm": 0.6313742995262146,
-      "learning_rate": 0.00030886973386488523,
-      "loss": 3.4168,
+      "epoch": 4.867924528301887,
+      "grad_norm": 0.6416391730308533,
+      "learning_rate": 0.0003083669724770642,
+      "loss": 3.3974,
       "step": 45150
     },
     {
-      "epoch": 4.864923043805834,
-      "grad_norm": 0.6884166598320007,
-      "learning_rate": 0.0003085464928348238,
-      "loss": 3.3887,
+      "epoch": 4.873315363881401,
+      "grad_norm": 0.7107098698616028,
+      "learning_rate": 0.0003080431732325958,
+      "loss": 3.3776,
       "step": 45200
     },
     {
-      "epoch": 4.870304595845441,
-      "grad_norm": 0.6939536333084106,
-      "learning_rate": 0.0003082232518047624,
-      "loss": 3.3801,
+      "epoch": 4.878706199460916,
+      "grad_norm": 0.6617811918258667,
+      "learning_rate": 0.00030771937398812733,
+      "loss": 3.388,
       "step": 45250
     },
     {
-      "epoch": 4.87568614788505,
-      "grad_norm": 0.6692671775817871,
-      "learning_rate": 0.00030790001077470096,
-      "loss": 3.3979,
+      "epoch": 4.884097035040432,
+      "grad_norm": 0.6666333079338074,
+      "learning_rate": 0.00030739557474365894,
+      "loss": 3.3737,
       "step": 45300
     },
     {
-      "epoch": 4.8810676999246585,
-      "grad_norm": 0.6553664803504944,
-      "learning_rate": 0.00030757676974463955,
-      "loss": 3.3896,
+      "epoch": 4.889487870619946,
+      "grad_norm": 0.6519107818603516,
+      "learning_rate": 0.0003070717754991905,
+      "loss": 3.3884,
       "step": 45350
     },
     {
-      "epoch": 4.886449251964266,
-      "grad_norm": 0.6963851451873779,
-      "learning_rate": 0.0003072535287145781,
-      "loss": 3.4001,
+      "epoch": 4.894878706199461,
+      "grad_norm": 0.6856688857078552,
+      "learning_rate": 0.0003067479762547221,
+      "loss": 3.3933,
       "step": 45400
     },
     {
-      "epoch": 4.891830804003875,
-      "grad_norm": 0.6625418066978455,
-      "learning_rate": 0.00030693028768451674,
-      "loss": 3.3867,
+      "epoch": 4.900269541778976,
+      "grad_norm": 0.6553418636322021,
+      "learning_rate": 0.0003064241770102536,
+      "loss": 3.3924,
       "step": 45450
     },
     {
-      "epoch": 4.897212356043483,
-      "grad_norm": 0.6888773441314697,
-      "learning_rate": 0.00030660704665445534,
-      "loss": 3.3772,
+      "epoch": 4.90566037735849,
+      "grad_norm": 0.7114120125770569,
+      "learning_rate": 0.00030610037776578514,
+      "loss": 3.384,
       "step": 45500
     },
     {
-      "epoch": 4.902593908083091,
-      "grad_norm": 0.657309889793396,
-      "learning_rate": 0.0003062838056243939,
-      "loss": 3.3902,
+      "epoch": 4.9110512129380055,
+      "grad_norm": 0.6625975370407104,
+      "learning_rate": 0.00030577657852131674,
+      "loss": 3.3707,
       "step": 45550
     },
     {
-      "epoch": 4.9079754601226995,
-      "grad_norm": 0.6960524320602417,
-      "learning_rate": 0.00030596056459433247,
-      "loss": 3.3857,
+      "epoch": 4.916442048517521,
+      "grad_norm": 0.653954267501831,
+      "learning_rate": 0.0003054527792768483,
+      "loss": 3.3788,
       "step": 45600
     },
     {
-      "epoch": 4.913357012162308,
-      "grad_norm": 0.6952453255653381,
-      "learning_rate": 0.00030563732356427107,
-      "loss": 3.4173,
+      "epoch": 4.921832884097035,
+      "grad_norm": 0.6749684810638428,
+      "learning_rate": 0.0003051289800323799,
+      "loss": 3.3842,
       "step": 45650
     },
     {
-      "epoch": 4.918738564201916,
-      "grad_norm": 0.6704519987106323,
-      "learning_rate": 0.00030531408253420966,
-      "loss": 3.3778,
+      "epoch": 4.92722371967655,
+      "grad_norm": 0.6280410885810852,
+      "learning_rate": 0.00030480518078791145,
+      "loss": 3.3965,
       "step": 45700
     },
     {
-      "epoch": 4.924120116241524,
-      "grad_norm": 0.7266472578048706,
-      "learning_rate": 0.00030499084150414826,
-      "loss": 3.3823,
+      "epoch": 4.932614555256064,
+      "grad_norm": 0.6328825950622559,
+      "learning_rate": 0.00030448138154344305,
+      "loss": 3.3935,
       "step": 45750
     },
     {
-      "epoch": 4.929501668281132,
-      "grad_norm": 0.6816993355751038,
-      "learning_rate": 0.00030466760047408685,
-      "loss": 3.3877,
+      "epoch": 4.938005390835579,
+      "grad_norm": 0.6992299556732178,
+      "learning_rate": 0.0003041575822989746,
+      "loss": 3.3794,
       "step": 45800
     },
     {
-      "epoch": 4.9348832203207404,
-      "grad_norm": 0.6936447620391846,
-      "learning_rate": 0.0003043443594440254,
-      "loss": 3.3832,
+      "epoch": 4.943396226415095,
+      "grad_norm": 0.7251986861228943,
+      "learning_rate": 0.0003038337830545062,
+      "loss": 3.3851,
       "step": 45850
     },
     {
-      "epoch": 4.940264772360349,
-      "grad_norm": 0.6859702467918396,
-      "learning_rate": 0.000304021118413964,
-      "loss": 3.4126,
+      "epoch": 4.948787061994609,
+      "grad_norm": 0.6330991983413696,
+      "learning_rate": 0.00030350998381003776,
+      "loss": 3.3891,
       "step": 45900
     },
     {
-      "epoch": 4.945646324399957,
-      "grad_norm": 0.6758688688278198,
-      "learning_rate": 0.0003036978773839025,
-      "loss": 3.3839,
+      "epoch": 4.954177897574124,
+      "grad_norm": 0.6726609468460083,
+      "learning_rate": 0.0003031861845655693,
+      "loss": 3.3914,
       "step": 45950
     },
     {
-      "epoch": 4.951027876439565,
-      "grad_norm": 0.6840824484825134,
-      "learning_rate": 0.0003033746363538412,
-      "loss": 3.3902,
+      "epoch": 4.959568733153639,
+      "grad_norm": 0.6512455344200134,
+      "learning_rate": 0.0003028623853211009,
+      "loss": 3.3809,
       "step": 46000
     },
     {
-      "epoch": 4.951027876439565,
-      "eval_accuracy": 0.3802585960460562,
-      "eval_loss": 3.4134035110473633,
-      "eval_runtime": 191.2911,
-      "eval_samples_per_second": 94.155,
-      "eval_steps_per_second": 5.886,
+      "epoch": 4.959568733153639,
+      "eval_accuracy": 0.38065354929326184,
+      "eval_loss": 3.408842086791992,
+      "eval_runtime": 184.1428,
+      "eval_samples_per_second": 97.81,
+      "eval_steps_per_second": 6.115,
       "step": 46000
     },
     {
-      "epoch": 4.956409428479174,
-      "grad_norm": 0.640100359916687,
-      "learning_rate": 0.00030305139532377977,
-      "loss": 3.4115,
+      "epoch": 4.964959568733153,
+      "grad_norm": 0.6779006123542786,
+      "learning_rate": 0.00030253858607663247,
+      "loss": 3.3967,
       "step": 46050
     },
     {
-      "epoch": 4.961790980518781,
-      "grad_norm": 0.6694591641426086,
-      "learning_rate": 0.0003027281542937183,
-      "loss": 3.4089,
+      "epoch": 4.9703504043126685,
+      "grad_norm": 0.6571473479270935,
+      "learning_rate": 0.00030221478683216407,
+      "loss": 3.3794,
       "step": 46100
     },
     {
-      "epoch": 4.96717253255839,
-      "grad_norm": 0.6837192177772522,
-      "learning_rate": 0.0003024049132636569,
-      "loss": 3.4014,
+      "epoch": 4.975741239892184,
+      "grad_norm": 0.6960002183914185,
+      "learning_rate": 0.00030189098758769557,
+      "loss": 3.3976,
       "step": 46150
     },
     {
-      "epoch": 4.9725540845979985,
-      "grad_norm": 0.6926203966140747,
-      "learning_rate": 0.0003020816722335955,
-      "loss": 3.3931,
+      "epoch": 4.981132075471698,
+      "grad_norm": 0.7017962336540222,
+      "learning_rate": 0.00030156718834322717,
+      "loss": 3.3859,
       "step": 46200
     },
     {
-      "epoch": 4.977935636637606,
-      "grad_norm": 0.672174870967865,
-      "learning_rate": 0.00030175843120353404,
-      "loss": 3.3686,
+      "epoch": 4.986522911051213,
+      "grad_norm": 0.6364179849624634,
+      "learning_rate": 0.0003012433890987587,
+      "loss": 3.3941,
       "step": 46250
     },
     {
-      "epoch": 4.983317188677215,
-      "grad_norm": 0.6593598127365112,
-      "learning_rate": 0.0003014351901734727,
-      "loss": 3.4056,
+      "epoch": 4.991913746630727,
+      "grad_norm": 0.6439968347549438,
+      "learning_rate": 0.0003009195898542903,
+      "loss": 3.3979,
       "step": 46300
     },
     {
-      "epoch": 4.988698740716822,
-      "grad_norm": 0.6591873168945312,
-      "learning_rate": 0.0003011119491434113,
-      "loss": 3.4209,
+      "epoch": 4.997304582210242,
+      "grad_norm": 0.6541712284088135,
+      "learning_rate": 0.0003005957906098219,
+      "loss": 3.3811,
       "step": 46350
     },
     {
-      "epoch": 4.994080292756431,
-      "grad_norm": 0.6413924694061279,
-      "learning_rate": 0.0003007887081133498,
-      "loss": 3.3968,
+      "epoch": 5.002695417789758,
+      "grad_norm": 0.6764222383499146,
+      "learning_rate": 0.00030027199136535343,
+      "loss": 3.353,
       "step": 46400
     },
     {
-      "epoch": 4.9994618447960395,
-      "grad_norm": 0.7270100116729736,
-      "learning_rate": 0.0003004654670832884,
-      "loss": 3.4009,
+      "epoch": 5.008086253369272,
+      "grad_norm": 0.6421829462051392,
+      "learning_rate": 0.00029994819212088503,
+      "loss": 3.2867,
       "step": 46450
     },
     {
-      "epoch": 5.004843396835647,
-      "grad_norm": 0.6969584226608276,
-      "learning_rate": 0.00030014222605322696,
-      "loss": 3.3214,
+      "epoch": 5.013477088948787,
+      "grad_norm": 0.6640639901161194,
+      "learning_rate": 0.0002996243928764166,
+      "loss": 3.2883,
       "step": 46500
     },
     {
-      "epoch": 5.010224948875256,
-      "grad_norm": 0.7092729806900024,
-      "learning_rate": 0.00029981898502316555,
-      "loss": 3.3004,
+      "epoch": 5.018867924528302,
+      "grad_norm": 0.6788437366485596,
+      "learning_rate": 0.0002993005936319482,
+      "loss": 3.2858,
       "step": 46550
     },
     {
-      "epoch": 5.015606500914864,
-      "grad_norm": 0.7010934352874756,
-      "learning_rate": 0.00029949574399310415,
-      "loss": 3.3056,
+      "epoch": 5.024258760107816,
+      "grad_norm": 0.7146774530410767,
+      "learning_rate": 0.00029897679438747974,
+      "loss": 3.2966,
       "step": 46600
     },
     {
-      "epoch": 5.020988052954472,
-      "grad_norm": 0.6872923970222473,
-      "learning_rate": 0.00029917250296304274,
-      "loss": 3.3137,
+      "epoch": 5.0296495956873315,
+      "grad_norm": 0.687339723110199,
+      "learning_rate": 0.0002986529951430113,
+      "loss": 3.2941,
       "step": 46650
     },
     {
-      "epoch": 5.0263696049940805,
-      "grad_norm": 0.6743913292884827,
-      "learning_rate": 0.00029884926193298134,
-      "loss": 3.3004,
+      "epoch": 5.035040431266847,
+      "grad_norm": 0.6656438708305359,
+      "learning_rate": 0.0002983291958985429,
+      "loss": 3.3115,
       "step": 46700
     },
     {
-      "epoch": 5.031751157033688,
-      "grad_norm": 0.7134643197059631,
-      "learning_rate": 0.0002985260209029199,
-      "loss": 3.2892,
+      "epoch": 5.040431266846361,
+      "grad_norm": 0.6926629543304443,
+      "learning_rate": 0.00029800539665407444,
+      "loss": 3.3001,
       "step": 46750
     },
     {
-      "epoch": 5.037132709073297,
-      "grad_norm": 0.6986505389213562,
-      "learning_rate": 0.00029820277987285853,
-      "loss": 3.2902,
+      "epoch": 5.045822102425876,
+      "grad_norm": 0.6437568068504333,
+      "learning_rate": 0.000297681597409606,
+      "loss": 3.3164,
       "step": 46800
     },
     {
-      "epoch": 5.042514261112905,
-      "grad_norm": 0.6677493453025818,
-      "learning_rate": 0.00029787953884279707,
-      "loss": 3.3036,
+      "epoch": 5.051212938005391,
+      "grad_norm": 0.7387434840202332,
+      "learning_rate": 0.0002973577981651376,
+      "loss": 3.2965,
       "step": 46850
     },
     {
-      "epoch": 5.047895813152513,
-      "grad_norm": 0.6945018172264099,
-      "learning_rate": 0.00029755629781273566,
-      "loss": 3.3008,
+      "epoch": 5.056603773584905,
+      "grad_norm": 0.6994271278381348,
+      "learning_rate": 0.00029703399892066915,
+      "loss": 3.3078,
       "step": 46900
     },
     {
-      "epoch": 5.0532773651921215,
-      "grad_norm": 0.7299492955207825,
-      "learning_rate": 0.00029723305678267426,
-      "loss": 3.3021,
+      "epoch": 5.061994609164421,
+      "grad_norm": 0.7022091150283813,
+      "learning_rate": 0.00029671019967620076,
+      "loss": 3.318,
       "step": 46950
     },
     {
-      "epoch": 5.05865891723173,
-      "grad_norm": 0.6931959390640259,
-      "learning_rate": 0.00029690981575261285,
-      "loss": 3.2993,
+      "epoch": 5.067385444743936,
+      "grad_norm": 0.7065854072570801,
+      "learning_rate": 0.0002963864004317323,
+      "loss": 3.305,
       "step": 47000
     },
     {
-      "epoch": 5.05865891723173,
-      "eval_accuracy": 0.38065050701240305,
-      "eval_loss": 3.4127960205078125,
-      "eval_runtime": 201.6655,
-      "eval_samples_per_second": 89.311,
-      "eval_steps_per_second": 5.584,
+      "epoch": 5.067385444743936,
+      "eval_accuracy": 0.38065572235101813,
+      "eval_loss": 3.4140918254852295,
+      "eval_runtime": 183.9109,
+      "eval_samples_per_second": 97.933,
+      "eval_steps_per_second": 6.123,
       "step": 47000
     },
     {
-      "epoch": 5.064040469271338,
-      "grad_norm": 0.7374845147132874,
-      "learning_rate": 0.0002965865747225514,
-      "loss": 3.3001,
+      "epoch": 5.07277628032345,
+      "grad_norm": 0.6356472969055176,
+      "learning_rate": 0.00029606260118726386,
+      "loss": 3.3088,
       "step": 47050
     },
     {
-      "epoch": 5.069422021310946,
-      "grad_norm": 0.6965045928955078,
-      "learning_rate": 0.00029626333369249,
-      "loss": 3.3073,
+      "epoch": 5.078167115902965,
+      "grad_norm": 0.7220776081085205,
+      "learning_rate": 0.0002957388019427954,
+      "loss": 3.3118,
       "step": 47100
     },
     {
-      "epoch": 5.074803573350554,
-      "grad_norm": 0.6908303499221802,
-      "learning_rate": 0.0002959400926624286,
-      "loss": 3.3072,
+      "epoch": 5.083557951482479,
+      "grad_norm": 0.7317782640457153,
+      "learning_rate": 0.000295415002698327,
+      "loss": 3.2998,
       "step": 47150
     },
     {
-      "epoch": 5.080185125390162,
-      "grad_norm": 0.6920403242111206,
-      "learning_rate": 0.0002956168516323672,
-      "loss": 3.318,
+      "epoch": 5.0889487870619945,
+      "grad_norm": 0.6589457988739014,
+      "learning_rate": 0.00029509120345385856,
+      "loss": 3.3099,
       "step": 47200
     },
     {
-      "epoch": 5.085566677429771,
-      "grad_norm": 0.7072156071662903,
-      "learning_rate": 0.00029529361060230577,
-      "loss": 3.3103,
+      "epoch": 5.09433962264151,
+      "grad_norm": 0.664459228515625,
+      "learning_rate": 0.00029476740420939017,
+      "loss": 3.2994,
       "step": 47250
     },
     {
-      "epoch": 5.090948229469379,
-      "grad_norm": 0.691167414188385,
-      "learning_rate": 0.0002949703695722443,
-      "loss": 3.3072,
+      "epoch": 5.099730458221024,
+      "grad_norm": 0.689423680305481,
+      "learning_rate": 0.0002944500809498111,
+      "loss": 3.313,
       "step": 47300
     },
     {
-      "epoch": 5.096329781508987,
-      "grad_norm": 0.6893109679222107,
-      "learning_rate": 0.00029464712854218296,
-      "loss": 3.3113,
+      "epoch": 5.105121293800539,
+      "grad_norm": 0.6973190903663635,
+      "learning_rate": 0.00029412628170534267,
+      "loss": 3.3175,
       "step": 47350
     },
     {
-      "epoch": 5.101711333548596,
-      "grad_norm": 0.6975725889205933,
-      "learning_rate": 0.0002943238875121215,
-      "loss": 3.3042,
+      "epoch": 5.110512129380054,
+      "grad_norm": 0.6578836441040039,
+      "learning_rate": 0.0002938024824608742,
+      "loss": 3.3254,
       "step": 47400
     },
     {
-      "epoch": 5.107092885588203,
-      "grad_norm": 0.7136406302452087,
-      "learning_rate": 0.0002940006464820601,
-      "loss": 3.3148,
+      "epoch": 5.115902964959568,
+      "grad_norm": 0.6793028712272644,
+      "learning_rate": 0.00029347868321640577,
+      "loss": 3.3167,
       "step": 47450
     },
     {
-      "epoch": 5.112474437627812,
-      "grad_norm": 0.7137263417243958,
-      "learning_rate": 0.0002936774054519987,
-      "loss": 3.3261,
+      "epoch": 5.121293800539084,
+      "grad_norm": 0.7139934301376343,
+      "learning_rate": 0.0002931548839719374,
+      "loss": 3.3197,
       "step": 47500
     },
     {
-      "epoch": 5.1178559896674205,
-      "grad_norm": 0.6680945754051208,
-      "learning_rate": 0.0002933541644219373,
-      "loss": 3.337,
+      "epoch": 5.126684636118599,
+      "grad_norm": 0.6757574081420898,
+      "learning_rate": 0.0002928310847274689,
+      "loss": 3.3176,
       "step": 47550
     },
     {
-      "epoch": 5.123237541707028,
-      "grad_norm": 0.8087601661682129,
-      "learning_rate": 0.0002930309233918758,
-      "loss": 3.3087,
+      "epoch": 5.132075471698113,
+      "grad_norm": 0.6933146119117737,
+      "learning_rate": 0.00029250728548300053,
+      "loss": 3.3204,
       "step": 47600
     },
     {
-      "epoch": 5.128619093746637,
-      "grad_norm": 0.6731941103935242,
-      "learning_rate": 0.0002927076823618144,
-      "loss": 3.3181,
+      "epoch": 5.137466307277628,
+      "grad_norm": 0.734917163848877,
+      "learning_rate": 0.0002921834862385321,
+      "loss": 3.3085,
       "step": 47650
     },
     {
-      "epoch": 5.134000645786244,
-      "grad_norm": 0.7058475017547607,
-      "learning_rate": 0.000292384441331753,
-      "loss": 3.3101,
+      "epoch": 5.142857142857143,
+      "grad_norm": 0.6720339059829712,
+      "learning_rate": 0.0002918596869940637,
+      "loss": 3.306,
       "step": 47700
     },
     {
-      "epoch": 5.139382197825853,
-      "grad_norm": 0.7096105813980103,
-      "learning_rate": 0.0002920612003016916,
-      "loss": 3.331,
+      "epoch": 5.1482479784366575,
+      "grad_norm": 0.7110573649406433,
+      "learning_rate": 0.0002915358877495952,
+      "loss": 3.3059,
       "step": 47750
     },
     {
-      "epoch": 5.1447637498654615,
-      "grad_norm": 0.7284489274024963,
-      "learning_rate": 0.0002917379592716302,
-      "loss": 3.3443,
+      "epoch": 5.153638814016173,
+      "grad_norm": 0.6773064732551575,
+      "learning_rate": 0.0002912120885051268,
+      "loss": 3.3235,
       "step": 47800
     },
     {
-      "epoch": 5.150145301905069,
-      "grad_norm": 0.6959323883056641,
-      "learning_rate": 0.00029141471824156874,
-      "loss": 3.3013,
+      "epoch": 5.159029649595688,
+      "grad_norm": 0.6786724328994751,
+      "learning_rate": 0.00029088828926065834,
+      "loss": 3.2986,
       "step": 47850
     },
     {
-      "epoch": 5.155526853944678,
-      "grad_norm": 0.7227029204368591,
-      "learning_rate": 0.00029109147721150734,
-      "loss": 3.3017,
+      "epoch": 5.164420485175202,
+      "grad_norm": 0.7131237387657166,
+      "learning_rate": 0.00029056449001618994,
+      "loss": 3.3331,
       "step": 47900
     },
     {
-      "epoch": 5.160908405984286,
-      "grad_norm": 0.6779053807258606,
-      "learning_rate": 0.00029076823618144593,
-      "loss": 3.3166,
+      "epoch": 5.169811320754717,
+      "grad_norm": 0.6635859608650208,
+      "learning_rate": 0.0002902406907717215,
+      "loss": 3.3178,
       "step": 47950
     },
     {
-      "epoch": 5.166289958023894,
-      "grad_norm": 0.7135473489761353,
-      "learning_rate": 0.00029044499515138453,
-      "loss": 3.3255,
+      "epoch": 5.175202156334231,
+      "grad_norm": 0.6670217514038086,
+      "learning_rate": 0.0002899168915272531,
+      "loss": 3.3097,
       "step": 48000
     },
     {
-      "epoch": 5.166289958023894,
-      "eval_accuracy": 0.3803929996682827,
-      "eval_loss": 3.412876844406128,
-      "eval_runtime": 192.717,
-      "eval_samples_per_second": 93.458,
-      "eval_steps_per_second": 5.843,
+      "epoch": 5.175202156334231,
+      "eval_accuracy": 0.38115128817234,
+      "eval_loss": 3.40806245803833,
+      "eval_runtime": 183.8557,
+      "eval_samples_per_second": 97.963,
+      "eval_steps_per_second": 6.124,
       "step": 48000
     },
     {
-      "epoch": 5.1716715100635025,
-      "grad_norm": 0.6599064469337463,
-      "learning_rate": 0.0002901217541213231,
-      "loss": 3.3175,
+      "epoch": 5.180592991913747,
+      "grad_norm": 0.703576922416687,
+      "learning_rate": 0.00028959309228278465,
+      "loss": 3.322,
       "step": 48050
     },
     {
-      "epoch": 5.17705306210311,
-      "grad_norm": 0.7011419534683228,
-      "learning_rate": 0.0002897985130912617,
-      "loss": 3.3167,
+      "epoch": 5.185983827493262,
+      "grad_norm": 0.6757267117500305,
+      "learning_rate": 0.00028926929303831625,
+      "loss": 3.3126,
       "step": 48100
     },
     {
-      "epoch": 5.182434614142719,
-      "grad_norm": 0.7181657552719116,
-      "learning_rate": 0.00028947527206120026,
-      "loss": 3.3397,
+      "epoch": 5.191374663072776,
+      "grad_norm": 0.6691439151763916,
+      "learning_rate": 0.0002889454937938478,
+      "loss": 3.3244,
       "step": 48150
     },
     {
-      "epoch": 5.187816166182327,
-      "grad_norm": 0.6871926784515381,
-      "learning_rate": 0.00028915849585174006,
-      "loss": 3.3162,
+      "epoch": 5.196765498652291,
+      "grad_norm": 0.6777300238609314,
+      "learning_rate": 0.00028862169454937935,
+      "loss": 3.3263,
       "step": 48200
     },
     {
-      "epoch": 5.193197718221935,
-      "grad_norm": 0.7430890202522278,
-      "learning_rate": 0.00028883525482167866,
-      "loss": 3.3273,
+      "epoch": 5.202156334231806,
+      "grad_norm": 0.6776520609855652,
+      "learning_rate": 0.0002882978953049109,
+      "loss": 3.3195,
       "step": 48250
     },
     {
-      "epoch": 5.198579270261543,
-      "grad_norm": 0.7614420056343079,
-      "learning_rate": 0.00028851201379161725,
-      "loss": 3.3204,
+      "epoch": 5.2075471698113205,
+      "grad_norm": 0.7370585799217224,
+      "learning_rate": 0.0002879740960604425,
+      "loss": 3.3199,
       "step": 48300
     },
     {
-      "epoch": 5.203960822301152,
-      "grad_norm": 0.6756367683410645,
-      "learning_rate": 0.00028818877276155585,
-      "loss": 3.327,
+      "epoch": 5.212938005390836,
+      "grad_norm": 0.6429452896118164,
+      "learning_rate": 0.00028765029681597406,
+      "loss": 3.3238,
       "step": 48350
     },
     {
-      "epoch": 5.20934237434076,
-      "grad_norm": 0.7239832282066345,
-      "learning_rate": 0.0002878655317314944,
-      "loss": 3.3298,
+      "epoch": 5.218328840970351,
+      "grad_norm": 0.6375373005867004,
+      "learning_rate": 0.00028732649757150566,
+      "loss": 3.3319,
       "step": 48400
     },
     {
-      "epoch": 5.214723926380368,
-      "grad_norm": 0.7229234576225281,
-      "learning_rate": 0.00028754229070143304,
-      "loss": 3.3227,
+      "epoch": 5.223719676549865,
+      "grad_norm": 0.7334010004997253,
+      "learning_rate": 0.0002870026983270372,
+      "loss": 3.3235,
       "step": 48450
     },
     {
-      "epoch": 5.220105478419977,
-      "grad_norm": 0.715870201587677,
-      "learning_rate": 0.0002872190496713716,
-      "loss": 3.3315,
+      "epoch": 5.22911051212938,
+      "grad_norm": 0.7284398674964905,
+      "learning_rate": 0.00028667889908256877,
+      "loss": 3.3114,
       "step": 48500
     },
     {
-      "epoch": 5.225487030459584,
-      "grad_norm": 0.714283287525177,
-      "learning_rate": 0.0002868958086413102,
-      "loss": 3.3058,
+      "epoch": 5.234501347708895,
+      "grad_norm": 0.6616871953010559,
+      "learning_rate": 0.00028635509983810037,
+      "loss": 3.3215,
       "step": 48550
     },
     {
-      "epoch": 5.230868582499193,
-      "grad_norm": 0.7448010444641113,
-      "learning_rate": 0.00028657256761124877,
-      "loss": 3.3418,
+      "epoch": 5.2398921832884096,
+      "grad_norm": 0.669139564037323,
+      "learning_rate": 0.0002860313005936319,
+      "loss": 3.3438,
       "step": 48600
     },
     {
-      "epoch": 5.236250134538801,
-      "grad_norm": 0.6719698309898376,
-      "learning_rate": 0.00028624932658118736,
-      "loss": 3.338,
+      "epoch": 5.245283018867925,
+      "grad_norm": 0.661720335483551,
+      "learning_rate": 0.00028570750134916347,
+      "loss": 3.3244,
       "step": 48650
     },
     {
-      "epoch": 5.241631686578409,
-      "grad_norm": 0.7021603584289551,
-      "learning_rate": 0.0002859260855511259,
-      "loss": 3.3204,
+      "epoch": 5.250673854447439,
+      "grad_norm": 0.6798533201217651,
+      "learning_rate": 0.0002853837021046951,
+      "loss": 3.327,
       "step": 48700
     },
     {
-      "epoch": 5.247013238618018,
-      "grad_norm": 0.7057817578315735,
-      "learning_rate": 0.0002856028445210645,
-      "loss": 3.3298,
+      "epoch": 5.256064690026954,
+      "grad_norm": 0.6983150839805603,
+      "learning_rate": 0.00028505990286022663,
+      "loss": 3.3337,
       "step": 48750
     },
     {
-      "epoch": 5.252394790657625,
-      "grad_norm": 0.7323166131973267,
-      "learning_rate": 0.0002852796034910031,
-      "loss": 3.3229,
+      "epoch": 5.261455525606469,
+      "grad_norm": 0.690118670463562,
+      "learning_rate": 0.0002847361036157582,
+      "loss": 3.318,
       "step": 48800
     },
     {
-      "epoch": 5.257776342697234,
-      "grad_norm": 0.7022436261177063,
-      "learning_rate": 0.0002849563624609417,
-      "loss": 3.3345,
+      "epoch": 5.2668463611859835,
+      "grad_norm": 0.6665123105049133,
+      "learning_rate": 0.0002844123043712898,
+      "loss": 3.3141,
       "step": 48850
     },
     {
-      "epoch": 5.2631578947368425,
-      "grad_norm": 0.7284606099128723,
-      "learning_rate": 0.0002846331214308803,
-      "loss": 3.3281,
+      "epoch": 5.272237196765499,
+      "grad_norm": 0.7163477540016174,
+      "learning_rate": 0.00028408850512682133,
+      "loss": 3.3283,
       "step": 48900
     },
     {
-      "epoch": 5.26853944677645,
-      "grad_norm": 0.7403408288955688,
-      "learning_rate": 0.0002843098804008188,
-      "loss": 3.3185,
+      "epoch": 5.277628032345014,
+      "grad_norm": 0.7490024566650391,
+      "learning_rate": 0.00028376470588235294,
+      "loss": 3.3285,
       "step": 48950
     },
     {
-      "epoch": 5.273920998816059,
-      "grad_norm": 0.7374369502067566,
-      "learning_rate": 0.00028398663937075747,
-      "loss": 3.3417,
+      "epoch": 5.283018867924528,
+      "grad_norm": 0.7262639403343201,
+      "learning_rate": 0.0002834409066378845,
+      "loss": 3.3345,
       "step": 49000
     },
     {
-      "epoch": 5.273920998816059,
-      "eval_accuracy": 0.3811685639815025,
-      "eval_loss": 3.4100325107574463,
-      "eval_runtime": 186.625,
-      "eval_samples_per_second": 96.509,
-      "eval_steps_per_second": 6.033,
+      "epoch": 5.283018867924528,
+      "eval_accuracy": 0.3816435944070274,
+      "eval_loss": 3.4043190479278564,
+      "eval_runtime": 184.0598,
+      "eval_samples_per_second": 97.854,
+      "eval_steps_per_second": 6.118,
       "step": 49000
     },
     {
-      "epoch": 5.279302550855666,
-      "grad_norm": 0.7417536377906799,
-      "learning_rate": 0.000283663398340696,
-      "loss": 3.3255,
+      "epoch": 5.288409703504043,
+      "grad_norm": 0.7100116014480591,
+      "learning_rate": 0.00028311710739341604,
+      "loss": 3.3319,
       "step": 49050
     },
     {
-      "epoch": 5.284684102895275,
-      "grad_norm": 0.7349126935005188,
-      "learning_rate": 0.0002833401573106346,
-      "loss": 3.3224,
+      "epoch": 5.293800539083558,
+      "grad_norm": 0.7000466585159302,
+      "learning_rate": 0.00028279330814894764,
+      "loss": 3.3455,
       "step": 49100
     },
     {
-      "epoch": 5.2900656549348835,
-      "grad_norm": 0.6606712341308594,
-      "learning_rate": 0.0002830169162805732,
-      "loss": 3.3305,
+      "epoch": 5.2991913746630726,
+      "grad_norm": 0.6913509368896484,
+      "learning_rate": 0.0002824695089044792,
+      "loss": 3.3395,
       "step": 49150
     },
     {
-      "epoch": 5.295447206974491,
-      "grad_norm": 0.6845487952232361,
-      "learning_rate": 0.0002826936752505118,
-      "loss": 3.3289,
+      "epoch": 5.304582210242588,
+      "grad_norm": 0.7095855474472046,
+      "learning_rate": 0.00028214570966001075,
+      "loss": 3.3473,
       "step": 49200
     },
     {
-      "epoch": 5.3008287590141,
-      "grad_norm": 0.7850626111030579,
-      "learning_rate": 0.00028237043422045034,
-      "loss": 3.338,
+      "epoch": 5.309973045822103,
+      "grad_norm": 0.7036840319633484,
+      "learning_rate": 0.00028182191041554235,
+      "loss": 3.3254,
       "step": 49250
     },
     {
-      "epoch": 5.306210311053708,
-      "grad_norm": 0.7478814721107483,
-      "learning_rate": 0.00028204719319038893,
-      "loss": 3.3253,
+      "epoch": 5.315363881401617,
+      "grad_norm": 0.6940468549728394,
+      "learning_rate": 0.0002814981111710739,
+      "loss": 3.3307,
       "step": 49300
     },
     {
-      "epoch": 5.311591863093316,
-      "grad_norm": 0.6879907250404358,
-      "learning_rate": 0.0002817239521603275,
-      "loss": 3.3235,
+      "epoch": 5.320754716981132,
+      "grad_norm": 0.7166381478309631,
+      "learning_rate": 0.0002811743119266055,
+      "loss": 3.3299,
       "step": 49350
     },
     {
-      "epoch": 5.316973415132924,
-      "grad_norm": 0.7131553292274475,
-      "learning_rate": 0.0002814007111302661,
-      "loss": 3.3212,
+      "epoch": 5.3261455525606465,
+      "grad_norm": 0.6922991275787354,
+      "learning_rate": 0.00028085051268213706,
+      "loss": 3.3233,
       "step": 49400
     },
     {
-      "epoch": 5.322354967172533,
-      "grad_norm": 0.6914525032043457,
-      "learning_rate": 0.0002810774701002047,
-      "loss": 3.3049,
+      "epoch": 5.331536388140162,
+      "grad_norm": 0.6697838306427002,
+      "learning_rate": 0.0002805267134376686,
+      "loss": 3.3062,
       "step": 49450
     },
     {
-      "epoch": 5.327736519212141,
-      "grad_norm": 0.7019218802452087,
-      "learning_rate": 0.00028075422907014325,
-      "loss": 3.3435,
+      "epoch": 5.336927223719677,
+      "grad_norm": 0.6904501914978027,
+      "learning_rate": 0.00028020291419320016,
+      "loss": 3.3364,
       "step": 49500
     },
     {
-      "epoch": 5.333118071251749,
-      "grad_norm": 0.7138478755950928,
-      "learning_rate": 0.00028043098804008185,
-      "loss": 3.3277,
+      "epoch": 5.342318059299191,
+      "grad_norm": 0.682349681854248,
+      "learning_rate": 0.00027987911494873176,
+      "loss": 3.3326,
       "step": 49550
     },
     {
-      "epoch": 5.338499623291357,
-      "grad_norm": 0.7029947638511658,
-      "learning_rate": 0.00028010774701002044,
-      "loss": 3.3302,
+      "epoch": 5.347708894878706,
+      "grad_norm": 0.6617938280105591,
+      "learning_rate": 0.0002795553157042633,
+      "loss": 3.3263,
       "step": 49600
     },
     {
-      "epoch": 5.343881175330965,
-      "grad_norm": 0.7279183864593506,
-      "learning_rate": 0.00027978450597995904,
-      "loss": 3.3124,
+      "epoch": 5.353099730458221,
+      "grad_norm": 0.6821538805961609,
+      "learning_rate": 0.0002792315164597949,
+      "loss": 3.3277,
       "step": 49650
     },
     {
-      "epoch": 5.349262727370574,
-      "grad_norm": 0.7827022671699524,
-      "learning_rate": 0.0002794612649498976,
-      "loss": 3.324,
+      "epoch": 5.3584905660377355,
+      "grad_norm": 0.723965585231781,
+      "learning_rate": 0.00027890771721532647,
+      "loss": 3.3133,
       "step": 49700
     },
     {
-      "epoch": 5.354644279410182,
-      "grad_norm": 0.7062620520591736,
-      "learning_rate": 0.00027913802391983623,
-      "loss": 3.3445,
+      "epoch": 5.363881401617251,
+      "grad_norm": 0.6679254174232483,
+      "learning_rate": 0.00027858391797085807,
+      "loss": 3.3514,
       "step": 49750
     },
     {
-      "epoch": 5.36002583144979,
-      "grad_norm": 0.6948559284210205,
-      "learning_rate": 0.00027881478288977477,
-      "loss": 3.347,
+      "epoch": 5.369272237196766,
+      "grad_norm": 0.6869991421699524,
+      "learning_rate": 0.0002782601187263896,
+      "loss": 3.3292,
       "step": 49800
     },
     {
-      "epoch": 5.365407383489399,
-      "grad_norm": 0.7399136424064636,
-      "learning_rate": 0.00027849154185971336,
-      "loss": 3.3452,
+      "epoch": 5.37466307277628,
+      "grad_norm": 0.718062162399292,
+      "learning_rate": 0.0002779363194819212,
+      "loss": 3.3378,
       "step": 49850
     },
     {
-      "epoch": 5.370788935529006,
-      "grad_norm": 0.7397146821022034,
-      "learning_rate": 0.00027816830082965196,
-      "loss": 3.3324,
+      "epoch": 5.380053908355795,
+      "grad_norm": 0.6966448426246643,
+      "learning_rate": 0.0002776125202374527,
+      "loss": 3.3295,
       "step": 49900
     },
     {
-      "epoch": 5.376170487568615,
-      "grad_norm": 0.7319924831390381,
-      "learning_rate": 0.00027784505979959055,
-      "loss": 3.3462,
+      "epoch": 5.38544474393531,
+      "grad_norm": 0.6406621336936951,
+      "learning_rate": 0.00027728872099298433,
+      "loss": 3.3423,
       "step": 49950
     },
     {
-      "epoch": 5.3815520396082235,
-      "grad_norm": 0.7393072843551636,
-      "learning_rate": 0.00027752181876952915,
-      "loss": 3.3513,
+      "epoch": 5.390835579514825,
+      "grad_norm": 0.6367883086204529,
+      "learning_rate": 0.0002769649217485159,
+      "loss": 3.3369,
       "step": 50000
     },
     {
-      "epoch": 5.3815520396082235,
-      "eval_accuracy": 0.38153733188274486,
-      "eval_loss": 3.4052817821502686,
-      "eval_runtime": 202.9452,
-      "eval_samples_per_second": 88.748,
-      "eval_steps_per_second": 5.548,
+      "epoch": 5.390835579514825,
+      "eval_accuracy": 0.38180168435879747,
+      "eval_loss": 3.4020180702209473,
+      "eval_runtime": 184.1627,
+      "eval_samples_per_second": 97.799,
+      "eval_steps_per_second": 6.114,
       "step": 50000
     },
     {
-      "epoch": 5.386933591647831,
-      "grad_norm": 0.6867526173591614,
-      "learning_rate": 0.0002771985777394677,
-      "loss": 3.3115,
+      "epoch": 5.39622641509434,
+      "grad_norm": 0.6698175072669983,
+      "learning_rate": 0.0002766411225040475,
+      "loss": 3.3196,
       "step": 50050
     },
     {
-      "epoch": 5.39231514368744,
-      "grad_norm": 0.6913804411888123,
-      "learning_rate": 0.0002768753367094063,
-      "loss": 3.3469,
+      "epoch": 5.401617250673855,
+      "grad_norm": 0.7395144701004028,
+      "learning_rate": 0.00027631732325957903,
+      "loss": 3.3424,
       "step": 50100
     },
     {
-      "epoch": 5.397696695727047,
-      "grad_norm": 0.7089942097663879,
-      "learning_rate": 0.0002765520956793449,
-      "loss": 3.3267,
+      "epoch": 5.407008086253369,
+      "grad_norm": 0.7581573724746704,
+      "learning_rate": 0.0002759935240151106,
+      "loss": 3.3221,
       "step": 50150
     },
     {
-      "epoch": 5.403078247766656,
-      "grad_norm": 0.6663913726806641,
-      "learning_rate": 0.00027622885464928347,
-      "loss": 3.3485,
+      "epoch": 5.412398921832884,
+      "grad_norm": 0.6988759636878967,
+      "learning_rate": 0.0002756697247706422,
+      "loss": 3.3288,
       "step": 50200
     },
     {
-      "epoch": 5.4084597998062645,
-      "grad_norm": 0.6780911684036255,
-      "learning_rate": 0.000275905613619222,
-      "loss": 3.323,
+      "epoch": 5.4177897574123985,
+      "grad_norm": 0.7205150127410889,
+      "learning_rate": 0.00027534592552617374,
+      "loss": 3.3495,
       "step": 50250
     },
     {
-      "epoch": 5.413841351845872,
-      "grad_norm": 0.6843054294586182,
-      "learning_rate": 0.00027558237258916066,
-      "loss": 3.3323,
+      "epoch": 5.423180592991914,
+      "grad_norm": 0.716349184513092,
+      "learning_rate": 0.00027502212628170535,
+      "loss": 3.3126,
       "step": 50300
     },
     {
-      "epoch": 5.419222903885481,
-      "grad_norm": 0.6963738799095154,
-      "learning_rate": 0.0002752591315590992,
-      "loss": 3.3252,
+      "epoch": 5.428571428571429,
+      "grad_norm": 0.7096773982048035,
+      "learning_rate": 0.0002746983270372369,
+      "loss": 3.3292,
       "step": 50350
     },
     {
-      "epoch": 5.424604455925088,
-      "grad_norm": 0.7334024906158447,
-      "learning_rate": 0.0002749358905290378,
-      "loss": 3.3618,
+      "epoch": 5.433962264150943,
+      "grad_norm": 0.7021868824958801,
+      "learning_rate": 0.00027437452779276845,
+      "loss": 3.3414,
       "step": 50400
     },
     {
-      "epoch": 5.429986007964697,
-      "grad_norm": 0.7072751522064209,
-      "learning_rate": 0.0002746126494989764,
-      "loss": 3.3496,
+      "epoch": 5.439353099730458,
+      "grad_norm": 0.6765516400337219,
+      "learning_rate": 0.00027405072854830005,
+      "loss": 3.3276,
       "step": 50450
     },
     {
-      "epoch": 5.435367560004305,
-      "grad_norm": 0.6565462350845337,
-      "learning_rate": 0.000274289408468915,
-      "loss": 3.3377,
+      "epoch": 5.444743935309973,
+      "grad_norm": 0.687379002571106,
+      "learning_rate": 0.0002737269293038316,
+      "loss": 3.34,
       "step": 50500
     },
     {
-      "epoch": 5.440749112043913,
-      "grad_norm": 0.7185359597206116,
-      "learning_rate": 0.0002739661674388535,
-      "loss": 3.3336,
+      "epoch": 5.450134770889488,
+      "grad_norm": 0.7096045613288879,
+      "learning_rate": 0.00027340313005936315,
+      "loss": 3.3376,
       "step": 50550
     },
     {
-      "epoch": 5.446130664083522,
-      "grad_norm": 0.6985934972763062,
-      "learning_rate": 0.0002736429264087921,
-      "loss": 3.3497,
+      "epoch": 5.455525606469003,
+      "grad_norm": 0.6659510731697083,
+      "learning_rate": 0.00027307933081489476,
+      "loss": 3.3259,
       "step": 50600
     },
     {
-      "epoch": 5.45151221612313,
-      "grad_norm": 0.778202474117279,
-      "learning_rate": 0.0002733196853787307,
-      "loss": 3.332,
+      "epoch": 5.460916442048518,
+      "grad_norm": 0.6847319602966309,
+      "learning_rate": 0.0002727555315704263,
+      "loss": 3.3371,
       "step": 50650
     },
     {
-      "epoch": 5.456893768162738,
-      "grad_norm": 0.7283360958099365,
-      "learning_rate": 0.0002729964443486693,
-      "loss": 3.3343,
+      "epoch": 5.466307277628032,
+      "grad_norm": 0.6880046725273132,
+      "learning_rate": 0.0002724317323259579,
+      "loss": 3.3522,
       "step": 50700
     },
     {
-      "epoch": 5.462275320202346,
-      "grad_norm": 0.7053759098052979,
-      "learning_rate": 0.0002726732033186079,
-      "loss": 3.3314,
+      "epoch": 5.471698113207547,
+      "grad_norm": 0.6896936297416687,
+      "learning_rate": 0.00027210793308148946,
+      "loss": 3.3563,
       "step": 50750
     },
     {
-      "epoch": 5.467656872241955,
-      "grad_norm": 0.7685927748680115,
-      "learning_rate": 0.00027234996228854644,
-      "loss": 3.3358,
+      "epoch": 5.4770889487870615,
+      "grad_norm": 0.6774055361747742,
+      "learning_rate": 0.000271784133837021,
+      "loss": 3.3353,
       "step": 50800
     },
     {
-      "epoch": 5.473038424281563,
-      "grad_norm": 0.6882610321044922,
-      "learning_rate": 0.0002720267212584851,
-      "loss": 3.3294,
+      "epoch": 5.482479784366577,
+      "grad_norm": 0.684891402721405,
+      "learning_rate": 0.00027146033459255256,
+      "loss": 3.3371,
       "step": 50850
     },
     {
-      "epoch": 5.478419976321171,
-      "grad_norm": 0.7401322722434998,
-      "learning_rate": 0.00027170348022842363,
-      "loss": 3.333,
+      "epoch": 5.487870619946092,
+      "grad_norm": 0.6616057753562927,
+      "learning_rate": 0.00027113653534808417,
+      "loss": 3.349,
       "step": 50900
     },
     {
-      "epoch": 5.483801528360779,
-      "grad_norm": 0.7363923192024231,
-      "learning_rate": 0.00027138023919836223,
-      "loss": 3.3233,
+      "epoch": 5.493261455525606,
+      "grad_norm": 0.7112061381340027,
+      "learning_rate": 0.0002708127361036157,
+      "loss": 3.3379,
       "step": 50950
     },
     {
-      "epoch": 5.489183080400387,
-      "grad_norm": 0.6889253258705139,
-      "learning_rate": 0.00027106346298890204,
-      "loss": 3.334,
+      "epoch": 5.498652291105121,
+      "grad_norm": 0.6590116024017334,
+      "learning_rate": 0.0002704889368591473,
+      "loss": 3.3382,
       "step": 51000
     },
     {
-      "epoch": 5.489183080400387,
-      "eval_accuracy": 0.38220174429173037,
-      "eval_loss": 3.4004783630371094,
-      "eval_runtime": 190.6388,
-      "eval_samples_per_second": 94.477,
-      "eval_steps_per_second": 5.906,
+      "epoch": 5.498652291105121,
+      "eval_accuracy": 0.382312678890189,
+      "eval_loss": 3.3979299068450928,
+      "eval_runtime": 184.0489,
+      "eval_samples_per_second": 97.86,
+      "eval_steps_per_second": 6.118,
       "step": 51000
     },
     {
-      "epoch": 5.494564632439996,
-      "grad_norm": 0.7611465454101562,
-      "learning_rate": 0.00027074022195884063,
-      "loss": 3.3499,
+      "epoch": 5.504043126684636,
+      "grad_norm": 0.7828822135925293,
+      "learning_rate": 0.0002701651376146789,
+      "loss": 3.3293,
       "step": 51050
     },
     {
-      "epoch": 5.499946184479604,
-      "grad_norm": 0.6965069770812988,
-      "learning_rate": 0.0002704169809287792,
-      "loss": 3.3126,
+      "epoch": 5.509433962264151,
+      "grad_norm": 0.6920241713523865,
+      "learning_rate": 0.0002698413383702105,
+      "loss": 3.3232,
       "step": 51100
     },
     {
-      "epoch": 5.505327736519212,
-      "grad_norm": 0.6993700861930847,
-      "learning_rate": 0.00027009373989871776,
-      "loss": 3.3497,
+      "epoch": 5.514824797843666,
+      "grad_norm": 0.6921935677528381,
+      "learning_rate": 0.00026951753912574203,
+      "loss": 3.3184,
       "step": 51150
     },
     {
-      "epoch": 5.510709288558821,
-      "grad_norm": 0.7188165187835693,
-      "learning_rate": 0.00026977049886865636,
-      "loss": 3.3468,
+      "epoch": 5.520215633423181,
+      "grad_norm": 0.7086000442504883,
+      "learning_rate": 0.0002691937398812736,
+      "loss": 3.3216,
       "step": 51200
     },
     {
-      "epoch": 5.516090840598428,
-      "grad_norm": 0.7151656746864319,
-      "learning_rate": 0.00026944725783859495,
-      "loss": 3.3301,
+      "epoch": 5.525606469002695,
+      "grad_norm": 0.706422746181488,
+      "learning_rate": 0.00026886994063680513,
+      "loss": 3.3496,
       "step": 51250
     },
     {
-      "epoch": 5.521472392638037,
-      "grad_norm": 0.7025049924850464,
-      "learning_rate": 0.00026912401680853355,
-      "loss": 3.3395,
+      "epoch": 5.53099730458221,
+      "grad_norm": 0.6786699295043945,
+      "learning_rate": 0.00026854614139233674,
+      "loss": 3.3418,
       "step": 51300
     },
     {
-      "epoch": 5.5268539446776455,
-      "grad_norm": 0.7040926814079285,
-      "learning_rate": 0.0002688007757784721,
-      "loss": 3.3292,
+      "epoch": 5.536388140161725,
+      "grad_norm": 0.7025102972984314,
+      "learning_rate": 0.0002682223421478683,
+      "loss": 3.3427,
       "step": 51350
     },
     {
-      "epoch": 5.532235496717253,
-      "grad_norm": 0.7505577802658081,
-      "learning_rate": 0.00026847753474841074,
-      "loss": 3.3488,
+      "epoch": 5.54177897574124,
+      "grad_norm": 0.6521251797676086,
+      "learning_rate": 0.00026790501888828924,
+      "loss": 3.3199,
       "step": 51400
     },
     {
-      "epoch": 5.537617048756862,
-      "grad_norm": 0.7513281106948853,
-      "learning_rate": 0.0002681542937183493,
-      "loss": 3.3521,
+      "epoch": 5.547169811320755,
+      "grad_norm": 0.6589857935905457,
+      "learning_rate": 0.00026758121964382084,
+      "loss": 3.3238,
       "step": 51450
     },
     {
-      "epoch": 5.542998600796469,
-      "grad_norm": 0.790105938911438,
-      "learning_rate": 0.0002678310526882879,
-      "loss": 3.3423,
+      "epoch": 5.55256064690027,
+      "grad_norm": 0.7255028486251831,
+      "learning_rate": 0.00026725742039935234,
+      "loss": 3.3654,
       "step": 51500
     },
     {
-      "epoch": 5.548380152836078,
-      "grad_norm": 0.7354403138160706,
-      "learning_rate": 0.00026750781165822647,
-      "loss": 3.3304,
+      "epoch": 5.557951482479784,
+      "grad_norm": 0.7726871371269226,
+      "learning_rate": 0.00026693362115488394,
+      "loss": 3.3245,
       "step": 51550
     },
     {
-      "epoch": 5.553761704875686,
-      "grad_norm": 0.7015941739082336,
-      "learning_rate": 0.00026718457062816506,
-      "loss": 3.3444,
+      "epoch": 5.563342318059299,
+      "grad_norm": 0.6390557289123535,
+      "learning_rate": 0.0002666098219104155,
+      "loss": 3.3404,
       "step": 51600
     },
     {
-      "epoch": 5.559143256915294,
-      "grad_norm": 0.7292041182518005,
-      "learning_rate": 0.00026686132959810366,
-      "loss": 3.3471,
+      "epoch": 5.568733153638814,
+      "grad_norm": 0.6846970915794373,
+      "learning_rate": 0.0002662860226659471,
+      "loss": 3.3273,
       "step": 51650
     },
     {
-      "epoch": 5.564524808954903,
-      "grad_norm": 0.7539200186729431,
-      "learning_rate": 0.0002665380885680422,
-      "loss": 3.3464,
+      "epoch": 5.574123989218329,
+      "grad_norm": 0.6807509660720825,
+      "learning_rate": 0.00026596222342147865,
+      "loss": 3.3312,
       "step": 51700
     },
     {
-      "epoch": 5.569906360994511,
-      "grad_norm": 0.730979859828949,
-      "learning_rate": 0.0002662148475379808,
-      "loss": 3.3338,
+      "epoch": 5.579514824797844,
+      "grad_norm": 0.6764967441558838,
+      "learning_rate": 0.00026563842417701026,
+      "loss": 3.3441,
       "step": 51750
     },
     {
-      "epoch": 5.575287913034119,
-      "grad_norm": 0.7592946887016296,
-      "learning_rate": 0.0002658916065079194,
-      "loss": 3.3366,
+      "epoch": 5.584905660377358,
+      "grad_norm": 0.7013547420501709,
+      "learning_rate": 0.0002653146249325418,
+      "loss": 3.3156,
       "step": 51800
     },
     {
-      "epoch": 5.580669465073727,
-      "grad_norm": 0.7109083533287048,
-      "learning_rate": 0.000265568365477858,
-      "loss": 3.3529,
+      "epoch": 5.590296495956873,
+      "grad_norm": 0.7346366047859192,
+      "learning_rate": 0.00026499082568807336,
+      "loss": 3.3146,
       "step": 51850
     },
     {
-      "epoch": 5.586051017113336,
-      "grad_norm": 0.7419684529304504,
-      "learning_rate": 0.0002652451244477965,
-      "loss": 3.333,
+      "epoch": 5.595687331536388,
+      "grad_norm": 0.6962804198265076,
+      "learning_rate": 0.00026466702644360496,
+      "loss": 3.3667,
       "step": 51900
     },
     {
-      "epoch": 5.591432569152944,
-      "grad_norm": 0.7088695764541626,
-      "learning_rate": 0.00026492188341773517,
-      "loss": 3.3527,
+      "epoch": 5.601078167115903,
+      "grad_norm": 0.6864445209503174,
+      "learning_rate": 0.0002643432271991365,
+      "loss": 3.3295,
       "step": 51950
     },
     {
-      "epoch": 5.596814121192552,
-      "grad_norm": 0.721508800983429,
-      "learning_rate": 0.0002645986423876737,
-      "loss": 3.334,
+      "epoch": 5.606469002695418,
+      "grad_norm": 0.6670565009117126,
+      "learning_rate": 0.00026401942795466806,
+      "loss": 3.3437,
       "step": 52000
     },
     {
-      "epoch": 5.596814121192552,
-      "eval_accuracy": 0.3824253519348526,
-      "eval_loss": 3.3942997455596924,
-      "eval_runtime": 201.2401,
-      "eval_samples_per_second": 89.5,
-      "eval_steps_per_second": 5.595,
+      "epoch": 5.606469002695418,
+      "eval_accuracy": 0.3829236340783698,
+      "eval_loss": 3.3928582668304443,
+      "eval_runtime": 183.886,
+      "eval_samples_per_second": 97.947,
+      "eval_steps_per_second": 6.123,
       "step": 52000
     },
     {
-      "epoch": 5.60219567323216,
-      "grad_norm": 0.7105787396430969,
-      "learning_rate": 0.0002642754013576123,
-      "loss": 3.3619,
+      "epoch": 5.611859838274933,
+      "grad_norm": 0.6701301336288452,
+      "learning_rate": 0.00026369562871019967,
+      "loss": 3.3233,
       "step": 52050
     },
     {
-      "epoch": 5.607577225271768,
-      "grad_norm": 0.7615808844566345,
-      "learning_rate": 0.0002639521603275509,
-      "loss": 3.333,
+      "epoch": 5.617250673854447,
+      "grad_norm": 0.7099761962890625,
+      "learning_rate": 0.0002633718294657312,
+      "loss": 3.3429,
       "step": 52100
     },
     {
-      "epoch": 5.612958777311377,
-      "grad_norm": 0.7386506199836731,
-      "learning_rate": 0.0002636289192974895,
-      "loss": 3.3353,
+      "epoch": 5.622641509433962,
+      "grad_norm": 0.6761354804039001,
+      "learning_rate": 0.0002630480302212628,
+      "loss": 3.3155,
       "step": 52150
     },
     {
-      "epoch": 5.618340329350985,
-      "grad_norm": 0.7279359102249146,
-      "learning_rate": 0.00026330567826742804,
-      "loss": 3.3381,
+      "epoch": 5.628032345013477,
+      "grad_norm": 0.6866230964660645,
+      "learning_rate": 0.00026272423097679437,
+      "loss": 3.3136,
       "step": 52200
     },
     {
-      "epoch": 5.623721881390593,
-      "grad_norm": 0.7196009755134583,
-      "learning_rate": 0.00026298243723736663,
-      "loss": 3.3369,
+      "epoch": 5.633423180592992,
+      "grad_norm": 0.6745296716690063,
+      "learning_rate": 0.0002624004317323259,
+      "loss": 3.3328,
       "step": 52250
     },
     {
-      "epoch": 5.629103433430201,
-      "grad_norm": 0.7014727592468262,
-      "learning_rate": 0.0002626591962073052,
-      "loss": 3.3391,
+      "epoch": 5.638814016172507,
+      "grad_norm": 0.7361600399017334,
+      "learning_rate": 0.00026207663248785753,
+      "loss": 3.3487,
       "step": 52300
     },
     {
-      "epoch": 5.634484985469809,
-      "grad_norm": 0.7047598958015442,
-      "learning_rate": 0.0002623359551772438,
-      "loss": 3.3345,
+      "epoch": 5.644204851752022,
+      "grad_norm": 0.6899746656417847,
+      "learning_rate": 0.0002617528332433891,
+      "loss": 3.3181,
       "step": 52350
     },
     {
-      "epoch": 5.639866537509418,
-      "grad_norm": 0.764114499092102,
-      "learning_rate": 0.0002620127141471824,
-      "loss": 3.3531,
+      "epoch": 5.649595687331536,
+      "grad_norm": 0.6511830687522888,
+      "learning_rate": 0.00026142903399892063,
+      "loss": 3.314,
       "step": 52400
     },
     {
-      "epoch": 5.645248089549026,
-      "grad_norm": 0.7131911516189575,
-      "learning_rate": 0.00026168947311712095,
-      "loss": 3.3343,
+      "epoch": 5.654986522911051,
+      "grad_norm": 0.785626232624054,
+      "learning_rate": 0.0002611117107393416,
+      "loss": 3.3258,
       "step": 52450
     },
     {
-      "epoch": 5.650629641588634,
-      "grad_norm": 0.7622030973434448,
-      "learning_rate": 0.00026136623208705955,
-      "loss": 3.3465,
+      "epoch": 5.660377358490566,
+      "grad_norm": 0.697786808013916,
+      "learning_rate": 0.00026078791149487313,
+      "loss": 3.3399,
       "step": 52500
     },
     {
-      "epoch": 5.656011193628243,
-      "grad_norm": 0.7329249382019043,
-      "learning_rate": 0.00026104299105699814,
-      "loss": 3.3301,
+      "epoch": 5.665768194070081,
+      "grad_norm": 0.6672894358634949,
+      "learning_rate": 0.00026046411225040474,
+      "loss": 3.3339,
       "step": 52550
     },
     {
-      "epoch": 5.66139274566785,
-      "grad_norm": 0.7030691504478455,
-      "learning_rate": 0.00026071975002693674,
-      "loss": 3.3294,
+      "epoch": 5.671159029649596,
+      "grad_norm": 0.7051243185997009,
+      "learning_rate": 0.0002601403130059363,
+      "loss": 3.3341,
       "step": 52600
     },
     {
-      "epoch": 5.666774297707459,
-      "grad_norm": 0.7377075552940369,
-      "learning_rate": 0.00026039650899687533,
-      "loss": 3.3589,
+      "epoch": 5.67654986522911,
+      "grad_norm": 0.6621816158294678,
+      "learning_rate": 0.0002598165137614679,
+      "loss": 3.3466,
       "step": 52650
     },
     {
-      "epoch": 5.672155849747067,
-      "grad_norm": 0.6939163208007812,
-      "learning_rate": 0.00026007326796681393,
-      "loss": 3.3312,
+      "epoch": 5.681940700808625,
+      "grad_norm": 0.7208759784698486,
+      "learning_rate": 0.00025949271451699944,
+      "loss": 3.3473,
       "step": 52700
     },
     {
-      "epoch": 5.677537401786675,
-      "grad_norm": 0.7312877178192139,
-      "learning_rate": 0.00025975002693675247,
-      "loss": 3.343,
+      "epoch": 5.6873315363881405,
+      "grad_norm": 0.7394353151321411,
+      "learning_rate": 0.000259168915272531,
+      "loss": 3.3419,
       "step": 52750
     },
     {
-      "epoch": 5.682918953826284,
-      "grad_norm": 0.7162089943885803,
-      "learning_rate": 0.00025942678590669106,
-      "loss": 3.3215,
+      "epoch": 5.692722371967655,
+      "grad_norm": 0.6930735111236572,
+      "learning_rate": 0.0002588451160280626,
+      "loss": 3.3213,
       "step": 52800
     },
     {
-      "epoch": 5.688300505865891,
-      "grad_norm": 0.7658792734146118,
-      "learning_rate": 0.00025910354487662966,
-      "loss": 3.3147,
+      "epoch": 5.69811320754717,
+      "grad_norm": 0.7763253450393677,
+      "learning_rate": 0.00025852131678359415,
+      "loss": 3.3258,
       "step": 52850
     },
     {
-      "epoch": 5.6936820579055,
-      "grad_norm": 0.7245572209358215,
-      "learning_rate": 0.00025878030384656825,
-      "loss": 3.3414,
+      "epoch": 5.703504043126685,
+      "grad_norm": 0.7092213034629822,
+      "learning_rate": 0.0002581975175391257,
+      "loss": 3.3478,
       "step": 52900
     },
     {
-      "epoch": 5.699063609945108,
-      "grad_norm": 0.716014564037323,
-      "learning_rate": 0.00025845706281650685,
-      "loss": 3.3339,
+      "epoch": 5.708894878706199,
+      "grad_norm": 0.680250346660614,
+      "learning_rate": 0.0002578737182946573,
+      "loss": 3.3431,
       "step": 52950
     },
     {
-      "epoch": 5.704445161984716,
-      "grad_norm": 0.7805140018463135,
-      "learning_rate": 0.0002581338217864454,
-      "loss": 3.3328,
+      "epoch": 5.714285714285714,
+      "grad_norm": 0.6698858141899109,
+      "learning_rate": 0.00025754991905018885,
+      "loss": 3.337,
       "step": 53000
     },
     {
-      "epoch": 5.704445161984716,
-      "eval_accuracy": 0.3828685470642479,
-      "eval_loss": 3.3905136585235596,
-      "eval_runtime": 201.7044,
-      "eval_samples_per_second": 89.294,
-      "eval_steps_per_second": 5.582,
+      "epoch": 5.714285714285714,
+      "eval_accuracy": 0.3831012815499465,
+      "eval_loss": 3.3877391815185547,
+      "eval_runtime": 184.3299,
+      "eval_samples_per_second": 97.711,
+      "eval_steps_per_second": 6.109,
       "step": 53000
     },
     {
-      "epoch": 5.709826714024325,
-      "grad_norm": 0.748485803604126,
-      "learning_rate": 0.000257810580756384,
-      "loss": 3.3468,
+      "epoch": 5.719676549865229,
+      "grad_norm": 0.6948190927505493,
+      "learning_rate": 0.00025722611980572046,
+      "loss": 3.3217,
       "step": 53050
     },
     {
-      "epoch": 5.715208266063933,
-      "grad_norm": 0.7458595037460327,
-      "learning_rate": 0.0002574873397263226,
-      "loss": 3.3035,
+      "epoch": 5.725067385444744,
+      "grad_norm": 0.7088119387626648,
+      "learning_rate": 0.000256902320561252,
+      "loss": 3.3437,
       "step": 53100
     },
     {
-      "epoch": 5.720589818103541,
-      "grad_norm": 0.6711325645446777,
-      "learning_rate": 0.00025716409869626117,
-      "loss": 3.3394,
+      "epoch": 5.730458221024259,
+      "grad_norm": 0.7612466216087341,
+      "learning_rate": 0.00025657852131678356,
+      "loss": 3.3424,
       "step": 53150
     },
     {
-      "epoch": 5.725971370143149,
-      "grad_norm": 0.7426289319992065,
-      "learning_rate": 0.0002568408576661997,
-      "loss": 3.3419,
+      "epoch": 5.735849056603773,
+      "grad_norm": 0.6875807642936707,
+      "learning_rate": 0.0002562547220723151,
+      "loss": 3.34,
       "step": 53200
     },
     {
-      "epoch": 5.731352922182758,
-      "grad_norm": 0.716453492641449,
-      "learning_rate": 0.00025651761663613836,
-      "loss": 3.3418,
+      "epoch": 5.741239892183288,
+      "grad_norm": 0.766903281211853,
+      "learning_rate": 0.0002559309228278467,
+      "loss": 3.3341,
       "step": 53250
     },
     {
-      "epoch": 5.736734474222366,
-      "grad_norm": 0.7246510982513428,
-      "learning_rate": 0.0002561943756060769,
-      "loss": 3.3546,
+      "epoch": 5.7466307277628035,
+      "grad_norm": 0.6803512573242188,
+      "learning_rate": 0.00025560712358337827,
+      "loss": 3.3541,
       "step": 53300
     },
     {
-      "epoch": 5.742116026261974,
-      "grad_norm": 0.705548882484436,
-      "learning_rate": 0.0002558711345760155,
-      "loss": 3.3553,
+      "epoch": 5.752021563342318,
+      "grad_norm": 0.7405898571014404,
+      "learning_rate": 0.00025528332433890987,
+      "loss": 3.3176,
       "step": 53350
     },
     {
-      "epoch": 5.747497578301582,
-      "grad_norm": 0.7819035053253174,
-      "learning_rate": 0.0002555478935459541,
-      "loss": 3.325,
+      "epoch": 5.757412398921833,
+      "grad_norm": 0.6681659817695618,
+      "learning_rate": 0.0002549595250944414,
+      "loss": 3.339,
       "step": 53400
     },
     {
-      "epoch": 5.75287913034119,
-      "grad_norm": 0.7534633278846741,
-      "learning_rate": 0.00025522465251589263,
-      "loss": 3.3426,
+      "epoch": 5.762803234501348,
+      "grad_norm": 0.7255740761756897,
+      "learning_rate": 0.000254635725849973,
+      "loss": 3.3265,
       "step": 53450
     },
     {
-      "epoch": 5.758260682380799,
-      "grad_norm": 0.7411068081855774,
-      "learning_rate": 0.0002549014114858312,
-      "loss": 3.3346,
+      "epoch": 5.768194070080862,
+      "grad_norm": 0.7309778332710266,
+      "learning_rate": 0.0002543119266055046,
+      "loss": 3.34,
       "step": 53500
     },
     {
-      "epoch": 5.763642234420407,
-      "grad_norm": 0.7141110897064209,
-      "learning_rate": 0.0002545781704557698,
-      "loss": 3.3429,
+      "epoch": 5.773584905660377,
+      "grad_norm": 0.6629784107208252,
+      "learning_rate": 0.00025398812736103613,
+      "loss": 3.3227,
       "step": 53550
     },
     {
-      "epoch": 5.769023786460015,
-      "grad_norm": 0.7201070189476013,
-      "learning_rate": 0.0002542549294257084,
-      "loss": 3.3355,
+      "epoch": 5.7789757412398925,
+      "grad_norm": 0.7654380202293396,
+      "learning_rate": 0.0002536643281165677,
+      "loss": 3.3215,
       "step": 53600
     },
     {
-      "epoch": 5.774405338499624,
-      "grad_norm": 0.774978518486023,
-      "learning_rate": 0.000253931688395647,
-      "loss": 3.3345,
+      "epoch": 5.784366576819407,
+      "grad_norm": 0.6455008387565613,
+      "learning_rate": 0.0002533405288720993,
+      "loss": 3.3171,
       "step": 53650
     },
     {
-      "epoch": 5.779786890539231,
-      "grad_norm": 0.7548540830612183,
-      "learning_rate": 0.0002536084473655856,
-      "loss": 3.3389,
+      "epoch": 5.789757412398922,
+      "grad_norm": 0.6831746697425842,
+      "learning_rate": 0.00025301672962763083,
+      "loss": 3.3296,
       "step": 53700
     },
     {
-      "epoch": 5.78516844257884,
-      "grad_norm": 0.6902544498443604,
-      "learning_rate": 0.00025328520633552415,
-      "loss": 3.3277,
+      "epoch": 5.795148247978437,
+      "grad_norm": 0.681228518486023,
+      "learning_rate": 0.00025269293038316244,
+      "loss": 3.3352,
       "step": 53750
     },
     {
-      "epoch": 5.790549994618448,
-      "grad_norm": 0.7907364964485168,
-      "learning_rate": 0.00025296196530546274,
-      "loss": 3.3482,
+      "epoch": 5.800539083557951,
+      "grad_norm": 0.7498580813407898,
+      "learning_rate": 0.000252369131138694,
+      "loss": 3.3422,
       "step": 53800
     },
     {
-      "epoch": 5.795931546658056,
-      "grad_norm": 0.7004760503768921,
-      "learning_rate": 0.00025263872427540133,
-      "loss": 3.3535,
+      "epoch": 5.8059299191374665,
+      "grad_norm": 0.6587980389595032,
+      "learning_rate": 0.00025204533189422554,
+      "loss": 3.3267,
       "step": 53850
     },
     {
-      "epoch": 5.801313098697665,
-      "grad_norm": 0.7246608138084412,
-      "learning_rate": 0.00025231548324533993,
-      "loss": 3.3375,
+      "epoch": 5.811320754716981,
+      "grad_norm": 0.7188945412635803,
+      "learning_rate": 0.00025172153264975714,
+      "loss": 3.331,
       "step": 53900
     },
     {
-      "epoch": 5.806694650737272,
-      "grad_norm": 0.713356077671051,
-      "learning_rate": 0.0002519922422152785,
-      "loss": 3.3441,
+      "epoch": 5.816711590296496,
+      "grad_norm": 0.6793975234031677,
+      "learning_rate": 0.0002513977334052887,
+      "loss": 3.3298,
       "step": 53950
     },
     {
-      "epoch": 5.812076202776881,
-      "grad_norm": 0.7337074279785156,
-      "learning_rate": 0.00025166900118521706,
-      "loss": 3.3229,
+      "epoch": 5.822102425876011,
+      "grad_norm": 0.6796118021011353,
+      "learning_rate": 0.00025107393416082025,
+      "loss": 3.3258,
       "step": 54000
     },
     {
-      "epoch": 5.812076202776881,
-      "eval_accuracy": 0.3832519831053452,
-      "eval_loss": 3.3878395557403564,
-      "eval_runtime": 188.3897,
-      "eval_samples_per_second": 95.605,
-      "eval_steps_per_second": 5.977,
+      "epoch": 5.822102425876011,
+      "eval_accuracy": 0.3838015494119108,
+      "eval_loss": 3.383310317993164,
+      "eval_runtime": 185.2889,
+      "eval_samples_per_second": 97.205,
+      "eval_steps_per_second": 6.077,
       "step": 54000
     },
     {
-      "epoch": 5.817457754816489,
-      "grad_norm": 0.7938897609710693,
-      "learning_rate": 0.0002513522249757569,
-      "loss": 3.329,
+      "epoch": 5.827493261455525,
+      "grad_norm": 0.668121337890625,
+      "learning_rate": 0.00025075013491635185,
+      "loss": 3.3471,
       "step": 54050
     },
     {
-      "epoch": 5.822839306856097,
-      "grad_norm": 0.707722544670105,
-      "learning_rate": 0.00025102898394569547,
-      "loss": 3.3292,
+      "epoch": 5.83288409703504,
+      "grad_norm": 0.7360820770263672,
+      "learning_rate": 0.0002504263356718834,
+      "loss": 3.3299,
       "step": 54100
     },
     {
-      "epoch": 5.828220858895706,
-      "grad_norm": 0.7328303456306458,
-      "learning_rate": 0.00025070574291563406,
-      "loss": 3.3384,
+      "epoch": 5.8382749326145555,
+      "grad_norm": 0.7131704092025757,
+      "learning_rate": 0.000250102536427415,
+      "loss": 3.3503,
       "step": 54150
     },
     {
-      "epoch": 5.833602410935313,
-      "grad_norm": 0.7096349000930786,
-      "learning_rate": 0.00025038250188557265,
-      "loss": 3.3243,
+      "epoch": 5.84366576819407,
+      "grad_norm": 0.705933153629303,
+      "learning_rate": 0.00024977873718294656,
+      "loss": 3.355,
       "step": 54200
     },
     {
-      "epoch": 5.838983962974922,
-      "grad_norm": 0.7162660360336304,
-      "learning_rate": 0.00025005926085551125,
-      "loss": 3.3405,
+      "epoch": 5.849056603773585,
+      "grad_norm": 0.6799279451370239,
+      "learning_rate": 0.0002494549379384781,
+      "loss": 3.3299,
       "step": 54250
     },
     {
-      "epoch": 5.84436551501453,
-      "grad_norm": 0.7353925108909607,
-      "learning_rate": 0.0002497360198254498,
-      "loss": 3.3328,
+      "epoch": 5.8544474393531,
+      "grad_norm": 0.7277923822402954,
+      "learning_rate": 0.0002491311386940097,
+      "loss": 3.3392,
       "step": 54300
     },
     {
-      "epoch": 5.849747067054138,
-      "grad_norm": 0.6959530115127563,
-      "learning_rate": 0.00024941277879538844,
-      "loss": 3.3515,
+      "epoch": 5.859838274932614,
+      "grad_norm": 0.7259764075279236,
+      "learning_rate": 0.00024880733944954126,
+      "loss": 3.3427,
       "step": 54350
     },
     {
-      "epoch": 5.855128619093747,
-      "grad_norm": 0.7663270235061646,
-      "learning_rate": 0.000249089537765327,
-      "loss": 3.3076,
+      "epoch": 5.8652291105121295,
+      "grad_norm": 0.6502102613449097,
+      "learning_rate": 0.0002484835402050728,
+      "loss": 3.3359,
       "step": 54400
     },
     {
-      "epoch": 5.860510171133355,
-      "grad_norm": 0.6977930665016174,
-      "learning_rate": 0.0002487662967352656,
-      "loss": 3.3244,
+      "epoch": 5.870619946091644,
+      "grad_norm": 0.7053978443145752,
+      "learning_rate": 0.0002481597409606044,
+      "loss": 3.3261,
       "step": 54450
     },
     {
-      "epoch": 5.865891723172963,
-      "grad_norm": 0.7649244666099548,
-      "learning_rate": 0.00024844305570520417,
-      "loss": 3.3401,
+      "epoch": 5.876010781671159,
+      "grad_norm": 0.7245563864707947,
+      "learning_rate": 0.00024783594171613597,
+      "loss": 3.3368,
       "step": 54500
     },
     {
-      "epoch": 5.871273275212571,
-      "grad_norm": 0.7535280585289001,
-      "learning_rate": 0.00024811981467514276,
-      "loss": 3.3485,
+      "epoch": 5.881401617250674,
+      "grad_norm": 0.7368947267532349,
+      "learning_rate": 0.0002475121424716675,
+      "loss": 3.3372,
       "step": 54550
     },
     {
-      "epoch": 5.87665482725218,
-      "grad_norm": 0.6970841884613037,
-      "learning_rate": 0.00024779657364508136,
-      "loss": 3.3303,
+      "epoch": 5.886792452830189,
+      "grad_norm": 0.7119336128234863,
+      "learning_rate": 0.0002471883432271991,
+      "loss": 3.336,
       "step": 54600
     },
     {
-      "epoch": 5.882036379291788,
-      "grad_norm": 0.7283545732498169,
-      "learning_rate": 0.0002474733326150199,
-      "loss": 3.3424,
+      "epoch": 5.892183288409703,
+      "grad_norm": 0.6919927597045898,
+      "learning_rate": 0.0002468645439827307,
+      "loss": 3.344,
       "step": 54650
     },
     {
-      "epoch": 5.887417931331396,
-      "grad_norm": 0.721764326095581,
-      "learning_rate": 0.0002471500915849585,
-      "loss": 3.341,
+      "epoch": 5.8975741239892185,
+      "grad_norm": 0.7063394784927368,
+      "learning_rate": 0.0002465407447382623,
+      "loss": 3.3304,
       "step": 54700
     },
     {
-      "epoch": 5.892799483371004,
-      "grad_norm": 0.687520444393158,
-      "learning_rate": 0.0002468268505548971,
-      "loss": 3.3415,
+      "epoch": 5.902964959568733,
+      "grad_norm": 0.7446856498718262,
+      "learning_rate": 0.00024621694549379383,
+      "loss": 3.3309,
       "step": 54750
     },
     {
-      "epoch": 5.898181035410612,
-      "grad_norm": 0.7138300538063049,
-      "learning_rate": 0.0002465036095248357,
-      "loss": 3.3447,
+      "epoch": 5.908355795148248,
+      "grad_norm": 0.7563053965568542,
+      "learning_rate": 0.00024589314624932543,
+      "loss": 3.3287,
       "step": 54800
     },
     {
-      "epoch": 5.903562587450221,
-      "grad_norm": 0.6954256296157837,
-      "learning_rate": 0.0002461803684947742,
-      "loss": 3.3466,
+      "epoch": 5.913746630727763,
+      "grad_norm": 0.6755961775779724,
+      "learning_rate": 0.00024556934700485693,
+      "loss": 3.333,
       "step": 54850
     },
     {
-      "epoch": 5.9089441394898286,
-      "grad_norm": 0.7353760004043579,
-      "learning_rate": 0.0002458571274647128,
-      "loss": 3.3411,
+      "epoch": 5.919137466307277,
+      "grad_norm": 0.7613006830215454,
+      "learning_rate": 0.00024524554776038853,
+      "loss": 3.3321,
       "step": 54900
     },
     {
-      "epoch": 5.914325691529437,
-      "grad_norm": 0.8013879656791687,
-      "learning_rate": 0.0002455338864346514,
-      "loss": 3.3581,
+      "epoch": 5.9245283018867925,
+      "grad_norm": 0.808539628982544,
+      "learning_rate": 0.0002449282245008095,
+      "loss": 3.3391,
       "step": 54950
     },
     {
-      "epoch": 5.919707243569046,
-      "grad_norm": 0.7387758493423462,
-      "learning_rate": 0.00024521064540459,
-      "loss": 3.3521,
+      "epoch": 5.929919137466308,
+      "grad_norm": 0.7955217361450195,
+      "learning_rate": 0.00024460442525634104,
+      "loss": 3.3469,
       "step": 55000
     },
     {
-      "epoch": 5.919707243569046,
-      "eval_accuracy": 0.3836273788327442,
-      "eval_loss": 3.3835108280181885,
-      "eval_runtime": 197.7765,
-      "eval_samples_per_second": 91.067,
-      "eval_steps_per_second": 5.693,
+      "epoch": 5.929919137466308,
+      "eval_accuracy": 0.38430026616697927,
+      "eval_loss": 3.379293918609619,
+      "eval_runtime": 185.5004,
+      "eval_samples_per_second": 97.094,
+      "eval_steps_per_second": 6.07,
       "step": 55000
     },
     {
-      "epoch": 5.925088795608653,
-      "grad_norm": 0.7418875098228455,
-      "learning_rate": 0.0002448874043745286,
-      "loss": 3.3297,
+      "epoch": 5.935309973045822,
+      "grad_norm": 0.6791992783546448,
+      "learning_rate": 0.00024428062601187264,
+      "loss": 3.3529,
       "step": 55050
     },
     {
-      "epoch": 5.930470347648262,
-      "grad_norm": 0.6959251165390015,
-      "learning_rate": 0.00024456416334446714,
-      "loss": 3.3306,
+      "epoch": 5.940700808625337,
+      "grad_norm": 0.6864568591117859,
+      "learning_rate": 0.0002439568267674042,
+      "loss": 3.3407,
       "step": 55100
     },
     {
-      "epoch": 5.93585189968787,
-      "grad_norm": 0.6997560858726501,
-      "learning_rate": 0.00024424092231440574,
-      "loss": 3.3368,
+      "epoch": 5.946091644204852,
+      "grad_norm": 0.7374833226203918,
+      "learning_rate": 0.00024363302752293574,
+      "loss": 3.3382,
       "step": 55150
     },
     {
-      "epoch": 5.941233451727478,
-      "grad_norm": 0.6774752736091614,
-      "learning_rate": 0.00024391768128434436,
-      "loss": 3.3398,
+      "epoch": 5.951482479784366,
+      "grad_norm": 0.7075588703155518,
+      "learning_rate": 0.00024330922827846732,
+      "loss": 3.3454,
       "step": 55200
     },
     {
-      "epoch": 5.946615003767087,
-      "grad_norm": 0.728191614151001,
-      "learning_rate": 0.00024359444025428293,
-      "loss": 3.3474,
+      "epoch": 5.9568733153638815,
+      "grad_norm": 0.7555676102638245,
+      "learning_rate": 0.0002429854290339989,
+      "loss": 3.351,
       "step": 55250
     },
     {
-      "epoch": 5.951996555806694,
-      "grad_norm": 0.7751122713088989,
-      "learning_rate": 0.0002432711992242215,
-      "loss": 3.3345,
+      "epoch": 5.962264150943396,
+      "grad_norm": 0.7086479067802429,
+      "learning_rate": 0.00024266162978953048,
+      "loss": 3.3152,
       "step": 55300
     },
     {
-      "epoch": 5.957378107846303,
-      "grad_norm": 0.6916903257369995,
-      "learning_rate": 0.0002429479581941601,
-      "loss": 3.3525,
+      "epoch": 5.967654986522911,
+      "grad_norm": 0.73097825050354,
+      "learning_rate": 0.00024233783054506203,
+      "loss": 3.3285,
       "step": 55350
     },
     {
-      "epoch": 5.962759659885911,
-      "grad_norm": 0.704298734664917,
-      "learning_rate": 0.00024262471716409868,
-      "loss": 3.3457,
+      "epoch": 5.973045822102426,
+      "grad_norm": 0.6742330193519592,
+      "learning_rate": 0.0002420140313005936,
+      "loss": 3.328,
       "step": 55400
     },
     {
-      "epoch": 5.968141211925519,
-      "grad_norm": 0.7079257369041443,
-      "learning_rate": 0.00024230147613403728,
-      "loss": 3.3441,
+      "epoch": 5.97843665768194,
+      "grad_norm": 0.7283134460449219,
+      "learning_rate": 0.00024169023205612518,
+      "loss": 3.3211,
       "step": 55450
     },
     {
-      "epoch": 5.973522763965128,
-      "grad_norm": 0.8067532777786255,
-      "learning_rate": 0.00024197823510397584,
-      "loss": 3.3431,
+      "epoch": 5.9838274932614555,
+      "grad_norm": 0.6925774812698364,
+      "learning_rate": 0.00024136643281165676,
+      "loss": 3.3424,
       "step": 55500
     },
     {
-      "epoch": 5.978904316004736,
-      "grad_norm": 0.7205045819282532,
-      "learning_rate": 0.0002416549940739144,
-      "loss": 3.3412,
+      "epoch": 5.989218328840971,
+      "grad_norm": 0.7129718065261841,
+      "learning_rate": 0.00024104263356718834,
+      "loss": 3.3329,
       "step": 55550
     },
     {
-      "epoch": 5.984285868044344,
-      "grad_norm": 0.729873538017273,
-      "learning_rate": 0.00024133175304385303,
-      "loss": 3.3155,
+      "epoch": 5.994609164420485,
+      "grad_norm": 0.7543454766273499,
+      "learning_rate": 0.0002407188343227199,
+      "loss": 3.3168,
       "step": 55600
     },
     {
-      "epoch": 5.989667420083952,
-      "grad_norm": 0.6925444602966309,
-      "learning_rate": 0.0002410085120137916,
-      "loss": 3.3504,
+      "epoch": 6.0,
+      "grad_norm": 1.459039330482483,
+      "learning_rate": 0.00024039503507825147,
+      "loss": 3.33,
       "step": 55650
     },
     {
-      "epoch": 5.995048972123561,
-      "grad_norm": 0.7201747298240662,
-      "learning_rate": 0.00024068527098373017,
-      "loss": 3.3241,
+      "epoch": 6.005390835579515,
+      "grad_norm": 0.731112539768219,
+      "learning_rate": 0.00024007123583378302,
+      "loss": 3.2355,
       "step": 55700
     },
     {
-      "epoch": 6.000430524163169,
-      "grad_norm": 0.7336894273757935,
-      "learning_rate": 0.0002403620299536688,
-      "loss": 3.3297,
+      "epoch": 6.010781671159029,
+      "grad_norm": 0.6690878868103027,
+      "learning_rate": 0.0002397474365893146,
+      "loss": 3.2399,
       "step": 55750
     },
     {
-      "epoch": 6.005812076202777,
-      "grad_norm": 0.6997312307357788,
-      "learning_rate": 0.00024003878892360736,
-      "loss": 3.232,
+      "epoch": 6.0161725067385445,
+      "grad_norm": 0.6819621920585632,
+      "learning_rate": 0.00023942363734484617,
+      "loss": 3.2531,
       "step": 55800
     },
     {
-      "epoch": 6.011193628242385,
-      "grad_norm": 0.7346082329750061,
-      "learning_rate": 0.00023971554789354593,
-      "loss": 3.2338,
+      "epoch": 6.02156334231806,
+      "grad_norm": 0.7173980474472046,
+      "learning_rate": 0.00023909983810037775,
+      "loss": 3.241,
       "step": 55850
     },
     {
-      "epoch": 6.016575180281993,
-      "grad_norm": 0.7229931950569153,
-      "learning_rate": 0.00023939230686348452,
-      "loss": 3.2316,
+      "epoch": 6.026954177897574,
+      "grad_norm": 0.6949658989906311,
+      "learning_rate": 0.00023877603885590933,
+      "loss": 3.2517,
       "step": 55900
     },
     {
-      "epoch": 6.021956732321602,
-      "grad_norm": 0.7411378622055054,
-      "learning_rate": 0.0002390690658334231,
-      "loss": 3.2539,
+      "epoch": 6.032345013477089,
+      "grad_norm": 0.703404426574707,
+      "learning_rate": 0.0002384522396114409,
+      "loss": 3.2445,
       "step": 55950
     },
     {
-      "epoch": 6.0273382843612096,
-      "grad_norm": 0.6895947456359863,
-      "learning_rate": 0.00023874582480336168,
-      "loss": 3.2608,
+      "epoch": 6.037735849056604,
+      "grad_norm": 0.7602894306182861,
+      "learning_rate": 0.00023812844036697248,
+      "loss": 3.2449,
       "step": 56000
     },
     {
-      "epoch": 6.0273382843612096,
-      "eval_accuracy": 0.3839755026853018,
-      "eval_loss": 3.3861966133117676,
-      "eval_runtime": 211.6559,
-      "eval_samples_per_second": 85.096,
-      "eval_steps_per_second": 5.32,
+      "epoch": 6.037735849056604,
+      "eval_accuracy": 0.38426104247447823,
+      "eval_loss": 3.3824236392974854,
+      "eval_runtime": 185.7248,
+      "eval_samples_per_second": 96.977,
+      "eval_steps_per_second": 6.063,
       "step": 56000
     },
     {
-      "epoch": 6.032719836400818,
-      "grad_norm": 0.7139065265655518,
-      "learning_rate": 0.00023842258377330028,
-      "loss": 3.2516,
+      "epoch": 6.0431266846361185,
+      "grad_norm": 0.753330647945404,
+      "learning_rate": 0.000237804641122504,
+      "loss": 3.2452,
       "step": 56050
     },
     {
-      "epoch": 6.038101388440427,
-      "grad_norm": 0.7533993124961853,
-      "learning_rate": 0.00023810580756384006,
-      "loss": 3.2595,
+      "epoch": 6.048517520215634,
+      "grad_norm": 0.6973538994789124,
+      "learning_rate": 0.00023748084187803558,
+      "loss": 3.2656,
       "step": 56100
     },
     {
-      "epoch": 6.043482940480034,
-      "grad_norm": 0.7737258076667786,
-      "learning_rate": 0.00023778256653377868,
-      "loss": 3.2565,
+      "epoch": 6.053908355795148,
+      "grad_norm": 0.7017737030982971,
+      "learning_rate": 0.00023715704263356716,
+      "loss": 3.242,
       "step": 56150
     },
     {
-      "epoch": 6.048864492519643,
-      "grad_norm": 0.709679126739502,
-      "learning_rate": 0.00023745932550371725,
-      "loss": 3.2624,
+      "epoch": 6.059299191374663,
+      "grad_norm": 0.6753481030464172,
+      "learning_rate": 0.00023683324338909874,
+      "loss": 3.2481,
       "step": 56200
     },
     {
-      "epoch": 6.0542460445592505,
-      "grad_norm": 0.7647435069084167,
-      "learning_rate": 0.00023713608447365584,
-      "loss": 3.2578,
+      "epoch": 6.064690026954178,
+      "grad_norm": 0.7410376071929932,
+      "learning_rate": 0.00023650944414463032,
+      "loss": 3.2649,
       "step": 56250
     },
     {
-      "epoch": 6.059627596598859,
-      "grad_norm": 0.7607208490371704,
-      "learning_rate": 0.00023681284344359444,
-      "loss": 3.2503,
+      "epoch": 6.070080862533692,
+      "grad_norm": 0.7393580079078674,
+      "learning_rate": 0.0002361856449001619,
+      "loss": 3.2491,
       "step": 56300
     },
     {
-      "epoch": 6.065009148638468,
-      "grad_norm": 0.76082843542099,
-      "learning_rate": 0.000236489602413533,
-      "loss": 3.2469,
+      "epoch": 6.0754716981132075,
+      "grad_norm": 0.7702028155326843,
+      "learning_rate": 0.00023586184565569347,
+      "loss": 3.2609,
       "step": 56350
     },
     {
-      "epoch": 6.070390700678075,
-      "grad_norm": 0.7211427688598633,
-      "learning_rate": 0.0002361663613834716,
-      "loss": 3.2735,
+      "epoch": 6.080862533692723,
+      "grad_norm": 0.7161694765090942,
+      "learning_rate": 0.00023553804641122502,
+      "loss": 3.2636,
       "step": 56400
     },
     {
-      "epoch": 6.075772252717684,
-      "grad_norm": 0.7562023401260376,
-      "learning_rate": 0.00023584312035341017,
-      "loss": 3.2647,
+      "epoch": 6.086253369272237,
+      "grad_norm": 0.7420124411582947,
+      "learning_rate": 0.00023521424716675657,
+      "loss": 3.2541,
       "step": 56450
     },
     {
-      "epoch": 6.081153804757292,
-      "grad_norm": 0.7383215427398682,
-      "learning_rate": 0.00023551987932334876,
-      "loss": 3.2576,
+      "epoch": 6.091644204851752,
+      "grad_norm": 0.741366446018219,
+      "learning_rate": 0.00023489044792228815,
+      "loss": 3.2627,
       "step": 56500
     },
     {
-      "epoch": 6.0865353567969,
-      "grad_norm": 0.7635093331336975,
-      "learning_rate": 0.00023519663829328735,
-      "loss": 3.2469,
+      "epoch": 6.097035040431267,
+      "grad_norm": 0.7206790447235107,
+      "learning_rate": 0.00023456664867781973,
+      "loss": 3.282,
       "step": 56550
     },
     {
-      "epoch": 6.091916908836509,
-      "grad_norm": 0.740917980670929,
-      "learning_rate": 0.00023487339726322592,
-      "loss": 3.2686,
+      "epoch": 6.1024258760107815,
+      "grad_norm": 0.7371720671653748,
+      "learning_rate": 0.0002342428494333513,
+      "loss": 3.2809,
       "step": 56600
     },
     {
-      "epoch": 6.097298460876116,
-      "grad_norm": 0.8048155307769775,
-      "learning_rate": 0.0002345501562331645,
-      "loss": 3.2547,
+      "epoch": 6.107816711590297,
+      "grad_norm": 0.7509032487869263,
+      "learning_rate": 0.00023391905018888288,
+      "loss": 3.2534,
       "step": 56650
     },
     {
-      "epoch": 6.102680012915725,
-      "grad_norm": 0.7598278522491455,
-      "learning_rate": 0.0002342269152031031,
-      "loss": 3.2496,
+      "epoch": 6.113207547169812,
+      "grad_norm": 0.7074062824249268,
+      "learning_rate": 0.00023359525094441443,
+      "loss": 3.2739,
       "step": 56700
     },
     {
-      "epoch": 6.108061564955333,
-      "grad_norm": 0.7499753832817078,
-      "learning_rate": 0.00023390367417304168,
-      "loss": 3.2653,
+      "epoch": 6.118598382749326,
+      "grad_norm": 0.6730645298957825,
+      "learning_rate": 0.000233271451699946,
+      "loss": 3.2619,
       "step": 56750
     },
     {
-      "epoch": 6.113443116994941,
-      "grad_norm": 0.7181298136711121,
-      "learning_rate": 0.00023358043314298025,
-      "loss": 3.2557,
+      "epoch": 6.123989218328841,
+      "grad_norm": 0.7206562161445618,
+      "learning_rate": 0.0002329476524554776,
+      "loss": 3.2556,
       "step": 56800
     },
     {
-      "epoch": 6.11882466903455,
-      "grad_norm": 0.7646988034248352,
-      "learning_rate": 0.00023325719211291887,
-      "loss": 3.2724,
+      "epoch": 6.129380053908355,
+      "grad_norm": 0.7239189743995667,
+      "learning_rate": 0.00023262385321100917,
+      "loss": 3.2564,
       "step": 56850
     },
     {
-      "epoch": 6.124206221074158,
-      "grad_norm": 0.7216889262199402,
-      "learning_rate": 0.00023293395108285744,
-      "loss": 3.2523,
+      "epoch": 6.1347708894878705,
+      "grad_norm": 0.7226546406745911,
+      "learning_rate": 0.00023230005396654072,
+      "loss": 3.2565,
       "step": 56900
     },
     {
-      "epoch": 6.129587773113766,
-      "grad_norm": 0.7427191734313965,
-      "learning_rate": 0.000232610710052796,
-      "loss": 3.2762,
+      "epoch": 6.140161725067386,
+      "grad_norm": 0.731500506401062,
+      "learning_rate": 0.0002319762547220723,
+      "loss": 3.2693,
       "step": 56950
     },
     {
-      "epoch": 6.134969325153374,
-      "grad_norm": 0.742785632610321,
-      "learning_rate": 0.0002322874690227346,
-      "loss": 3.2733,
+      "epoch": 6.1455525606469,
+      "grad_norm": 0.7238706946372986,
+      "learning_rate": 0.00023165893146249325,
+      "loss": 3.2635,
       "step": 57000
     },
     {
-      "epoch": 6.134969325153374,
-      "eval_accuracy": 0.38427451543256724,
-      "eval_loss": 3.3876760005950928,
-      "eval_runtime": 199.7347,
-      "eval_samples_per_second": 90.175,
-      "eval_steps_per_second": 5.637,
+      "epoch": 6.1455525606469,
+      "eval_accuracy": 0.3841879190809791,
+      "eval_loss": 3.3821895122528076,
+      "eval_runtime": 185.1789,
+      "eval_samples_per_second": 97.263,
+      "eval_steps_per_second": 6.081,
       "step": 57000
     },
     {
-      "epoch": 6.140350877192983,
-      "grad_norm": 0.7947841882705688,
-      "learning_rate": 0.0002319642279926732,
-      "loss": 3.2932,
+      "epoch": 6.150943396226415,
+      "grad_norm": 0.7081495523452759,
+      "learning_rate": 0.0002313351322180248,
+      "loss": 3.2725,
       "step": 57050
     },
     {
-      "epoch": 6.1457324292325906,
-      "grad_norm": 0.7083519101142883,
-      "learning_rate": 0.00023164098696261176,
-      "loss": 3.2631,
+      "epoch": 6.15633423180593,
+      "grad_norm": 0.7516956925392151,
+      "learning_rate": 0.00023101133297355638,
+      "loss": 3.2563,
       "step": 57100
     },
     {
-      "epoch": 6.151113981272199,
-      "grad_norm": 0.7376063466072083,
-      "learning_rate": 0.00023131774593255036,
-      "loss": 3.2754,
+      "epoch": 6.1617250673854445,
+      "grad_norm": 0.7391020059585571,
+      "learning_rate": 0.00023068753372908795,
+      "loss": 3.2662,
       "step": 57150
     },
     {
-      "epoch": 6.156495533311807,
-      "grad_norm": 0.7607095241546631,
-      "learning_rate": 0.00023099450490248892,
-      "loss": 3.282,
+      "epoch": 6.16711590296496,
+      "grad_norm": 0.8158681988716125,
+      "learning_rate": 0.0002303637344846195,
+      "loss": 3.2615,
       "step": 57200
     },
     {
-      "epoch": 6.161877085351415,
-      "grad_norm": 0.7469072341918945,
-      "learning_rate": 0.00023067126387242754,
-      "loss": 3.2629,
+      "epoch": 6.172506738544475,
+      "grad_norm": 0.7187681198120117,
+      "learning_rate": 0.00023003993524015108,
+      "loss": 3.2611,
       "step": 57250
     },
     {
-      "epoch": 6.167258637391024,
-      "grad_norm": 0.7765781283378601,
-      "learning_rate": 0.0002303480228423661,
-      "loss": 3.2591,
+      "epoch": 6.177897574123989,
+      "grad_norm": 0.7636266946792603,
+      "learning_rate": 0.00022971613599568266,
+      "loss": 3.2622,
       "step": 57300
     },
     {
-      "epoch": 6.1726401894306315,
-      "grad_norm": 0.7496424913406372,
-      "learning_rate": 0.00023002478181230468,
-      "loss": 3.2608,
+      "epoch": 6.183288409703504,
+      "grad_norm": 1.2370136976242065,
+      "learning_rate": 0.0002293923367512142,
+      "loss": 3.2723,
       "step": 57350
     },
     {
-      "epoch": 6.17802174147024,
-      "grad_norm": 0.7514758706092834,
-      "learning_rate": 0.00022970154078224327,
-      "loss": 3.2634,
+      "epoch": 6.188679245283019,
+      "grad_norm": 0.7909350991249084,
+      "learning_rate": 0.0002290685375067458,
+      "loss": 3.2762,
       "step": 57400
     },
     {
-      "epoch": 6.183403293509849,
-      "grad_norm": 0.7652694582939148,
-      "learning_rate": 0.00022937829975218187,
-      "loss": 3.2699,
+      "epoch": 6.1940700808625335,
+      "grad_norm": 0.7264539003372192,
+      "learning_rate": 0.00022874473826227736,
+      "loss": 3.2916,
       "step": 57450
     },
     {
-      "epoch": 6.188784845549456,
-      "grad_norm": 0.734973132610321,
-      "learning_rate": 0.00022905505872212044,
-      "loss": 3.2727,
+      "epoch": 6.199460916442049,
+      "grad_norm": 0.7202154397964478,
+      "learning_rate": 0.00022842093901780894,
+      "loss": 3.2426,
       "step": 57500
     },
     {
-      "epoch": 6.194166397589065,
-      "grad_norm": 0.7655386328697205,
-      "learning_rate": 0.00022873181769205903,
-      "loss": 3.2568,
+      "epoch": 6.204851752021563,
+      "grad_norm": 0.786159336566925,
+      "learning_rate": 0.00022809713977334052,
+      "loss": 3.2954,
       "step": 57550
     },
     {
-      "epoch": 6.1995479496286725,
-      "grad_norm": 0.744789719581604,
-      "learning_rate": 0.0002284085766619976,
-      "loss": 3.2845,
+      "epoch": 6.210242587601078,
+      "grad_norm": 0.7010265588760376,
+      "learning_rate": 0.0002277733405288721,
+      "loss": 3.2655,
       "step": 57600
     },
     {
-      "epoch": 6.204929501668281,
-      "grad_norm": 0.7866900563240051,
-      "learning_rate": 0.0002280853356319362,
-      "loss": 3.2839,
+      "epoch": 6.215633423180593,
+      "grad_norm": 0.6955938339233398,
+      "learning_rate": 0.00022744954128440365,
+      "loss": 3.2681,
       "step": 57650
     },
     {
-      "epoch": 6.21031105370789,
-      "grad_norm": 0.753262460231781,
-      "learning_rate": 0.0002277620946018748,
-      "loss": 3.2868,
+      "epoch": 6.2210242587601075,
+      "grad_norm": 0.7094867825508118,
+      "learning_rate": 0.0002271257420399352,
+      "loss": 3.2879,
       "step": 57700
     },
     {
-      "epoch": 6.215692605747497,
-      "grad_norm": 0.7368378639221191,
-      "learning_rate": 0.00022743885357181336,
-      "loss": 3.2648,
+      "epoch": 6.226415094339623,
+      "grad_norm": 0.7032619118690491,
+      "learning_rate": 0.00022680194279546678,
+      "loss": 3.2739,
       "step": 57750
     },
     {
-      "epoch": 6.221074157787106,
-      "grad_norm": 0.7708497643470764,
-      "learning_rate": 0.00022711561254175192,
-      "loss": 3.2733,
+      "epoch": 6.231805929919138,
+      "grad_norm": 0.7983007431030273,
+      "learning_rate": 0.00022647814355099835,
+      "loss": 3.2769,
       "step": 57800
     },
     {
-      "epoch": 6.226455709826714,
-      "grad_norm": 0.7500898241996765,
-      "learning_rate": 0.00022679237151169054,
-      "loss": 3.2719,
+      "epoch": 6.237196765498652,
+      "grad_norm": 0.7237648963928223,
+      "learning_rate": 0.00022615434430652993,
+      "loss": 3.2775,
       "step": 57850
     },
     {
-      "epoch": 6.231837261866322,
-      "grad_norm": 0.7447527647018433,
-      "learning_rate": 0.0002264691304816291,
-      "loss": 3.2606,
+      "epoch": 6.242587601078167,
+      "grad_norm": 0.7277876734733582,
+      "learning_rate": 0.0002258305450620615,
+      "loss": 3.2792,
       "step": 57900
     },
     {
-      "epoch": 6.237218813905931,
-      "grad_norm": 0.7365005016326904,
-      "learning_rate": 0.00022614588945156768,
-      "loss": 3.2832,
+      "epoch": 6.247978436657682,
+      "grad_norm": 0.7612237930297852,
+      "learning_rate": 0.0002255067458175931,
+      "loss": 3.2614,
       "step": 57950
     },
     {
-      "epoch": 6.242600365945538,
-      "grad_norm": 0.7803393006324768,
-      "learning_rate": 0.0002258226484215063,
-      "loss": 3.2897,
+      "epoch": 6.2533692722371965,
+      "grad_norm": 0.7209222316741943,
+      "learning_rate": 0.00022518294657312467,
+      "loss": 3.2828,
       "step": 58000
     },
     {
-      "epoch": 6.242600365945538,
-      "eval_accuracy": 0.38468098588588123,
-      "eval_loss": 3.381647825241089,
-      "eval_runtime": 201.6412,
-      "eval_samples_per_second": 89.322,
-      "eval_steps_per_second": 5.584,
+      "epoch": 6.2533692722371965,
+      "eval_accuracy": 0.3847738841049626,
+      "eval_loss": 3.3785743713378906,
+      "eval_runtime": 185.5,
+      "eval_samples_per_second": 97.094,
+      "eval_steps_per_second": 6.07,
       "step": 58000
     },
     {
-      "epoch": 6.247981917985147,
-      "grad_norm": 0.7322475910186768,
-      "learning_rate": 0.00022549940739144487,
-      "loss": 3.2811,
+      "epoch": 6.258760107816712,
+      "grad_norm": 0.69462651014328,
+      "learning_rate": 0.0002248591473286562,
+      "loss": 3.2711,
       "step": 58050
     },
     {
-      "epoch": 6.253363470024755,
-      "grad_norm": 0.7434014081954956,
-      "learning_rate": 0.00022517616636138344,
-      "loss": 3.2893,
+      "epoch": 6.264150943396227,
+      "grad_norm": 0.7389229536056519,
+      "learning_rate": 0.00022453534808418777,
+      "loss": 3.2776,
       "step": 58100
     },
     {
-      "epoch": 6.258745022064363,
-      "grad_norm": 0.7110326290130615,
-      "learning_rate": 0.00022485292533132203,
-      "loss": 3.2929,
+      "epoch": 6.269541778975741,
+      "grad_norm": 0.728661835193634,
+      "learning_rate": 0.00022421154883971934,
+      "loss": 3.2693,
       "step": 58150
     },
     {
-      "epoch": 6.264126574103972,
-      "grad_norm": 0.7830908298492432,
-      "learning_rate": 0.00022452968430126063,
-      "loss": 3.2784,
+      "epoch": 6.274932614555256,
+      "grad_norm": 0.7259601950645447,
+      "learning_rate": 0.00022388774959525092,
+      "loss": 3.2698,
       "step": 58200
     },
     {
-      "epoch": 6.26950812614358,
-      "grad_norm": 0.7722041606903076,
-      "learning_rate": 0.00022420644327119922,
-      "loss": 3.2607,
+      "epoch": 6.280323450134771,
+      "grad_norm": 0.7676014304161072,
+      "learning_rate": 0.0002235639503507825,
+      "loss": 3.2771,
       "step": 58250
     },
     {
-      "epoch": 6.274889678183188,
-      "grad_norm": 0.7421392202377319,
-      "learning_rate": 0.0002238832022411378,
-      "loss": 3.2634,
+      "epoch": 6.285714285714286,
+      "grad_norm": 0.7163670659065247,
+      "learning_rate": 0.00022324015110631408,
+      "loss": 3.2627,
       "step": 58300
     },
     {
-      "epoch": 6.280271230222796,
-      "grad_norm": 0.7481209635734558,
-      "learning_rate": 0.00022355996121107636,
-      "loss": 3.2692,
+      "epoch": 6.291105121293801,
+      "grad_norm": 0.7317943572998047,
+      "learning_rate": 0.00022291635186184565,
+      "loss": 3.291,
       "step": 58350
     },
     {
-      "epoch": 6.285652782262405,
-      "grad_norm": 0.7708833813667297,
-      "learning_rate": 0.00022323672018101498,
-      "loss": 3.2706,
+      "epoch": 6.296495956873315,
+      "grad_norm": 0.7527638077735901,
+      "learning_rate": 0.0002225925526173772,
+      "loss": 3.2718,
       "step": 58400
     },
     {
-      "epoch": 6.2910343343020125,
-      "grad_norm": 0.8033352494239807,
-      "learning_rate": 0.00022291347915095355,
-      "loss": 3.2784,
+      "epoch": 6.30188679245283,
+      "grad_norm": 0.8808325529098511,
+      "learning_rate": 0.00022226875337290878,
+      "loss": 3.2739,
       "step": 58450
     },
     {
-      "epoch": 6.296415886341621,
-      "grad_norm": 0.7255500555038452,
-      "learning_rate": 0.0002225902381208921,
-      "loss": 3.2862,
+      "epoch": 6.307277628032345,
+      "grad_norm": 0.7042098045349121,
+      "learning_rate": 0.00022194495412844033,
+      "loss": 3.2638,
       "step": 58500
     },
     {
-      "epoch": 6.301797438381229,
-      "grad_norm": 0.7180657386779785,
-      "learning_rate": 0.00022226699709083073,
-      "loss": 3.2808,
+      "epoch": 6.3126684636118595,
+      "grad_norm": 0.7180835008621216,
+      "learning_rate": 0.0002216211548839719,
+      "loss": 3.2755,
       "step": 58550
     },
     {
-      "epoch": 6.307178990420837,
-      "grad_norm": 0.8310112953186035,
-      "learning_rate": 0.0002219437560607693,
-      "loss": 3.2954,
+      "epoch": 6.318059299191375,
+      "grad_norm": 0.715448796749115,
+      "learning_rate": 0.0002212973556395035,
+      "loss": 3.2917,
       "step": 58600
     },
     {
-      "epoch": 6.312560542460446,
-      "grad_norm": 0.7167147994041443,
-      "learning_rate": 0.00022162051503070787,
-      "loss": 3.2825,
+      "epoch": 6.32345013477089,
+      "grad_norm": 0.7007191181182861,
+      "learning_rate": 0.00022097355639503507,
+      "loss": 3.2667,
       "step": 58650
     },
     {
-      "epoch": 6.3179420945000535,
-      "grad_norm": 0.8282204866409302,
-      "learning_rate": 0.00022129727400064646,
-      "loss": 3.2813,
+      "epoch": 6.328840970350404,
+      "grad_norm": 0.7355238795280457,
+      "learning_rate": 0.00022064975715056664,
+      "loss": 3.2805,
       "step": 58700
     },
     {
-      "epoch": 6.323323646539662,
-      "grad_norm": 0.7656278014183044,
-      "learning_rate": 0.00022097403297058506,
-      "loss": 3.282,
+      "epoch": 6.334231805929919,
+      "grad_norm": 0.7063921689987183,
+      "learning_rate": 0.0002203259579060982,
+      "loss": 3.2852,
       "step": 58750
     },
     {
-      "epoch": 6.328705198579271,
-      "grad_norm": 0.729318380355835,
-      "learning_rate": 0.00022065079194052363,
-      "loss": 3.2716,
+      "epoch": 6.339622641509434,
+      "grad_norm": 0.7628762125968933,
+      "learning_rate": 0.00022000215866162977,
+      "loss": 3.2842,
       "step": 58800
     },
     {
-      "epoch": 6.334086750618878,
-      "grad_norm": 0.7712202072143555,
-      "learning_rate": 0.00022032755091046222,
-      "loss": 3.2681,
+      "epoch": 6.345013477088949,
+      "grad_norm": 0.7140092849731445,
+      "learning_rate": 0.00021967835941716135,
+      "loss": 3.2709,
       "step": 58850
     },
     {
-      "epoch": 6.339468302658487,
-      "grad_norm": 0.7325397729873657,
-      "learning_rate": 0.0002200043098804008,
-      "loss": 3.2774,
+      "epoch": 6.350404312668464,
+      "grad_norm": 0.7911513447761536,
+      "learning_rate": 0.0002193545601726929,
+      "loss": 3.2745,
       "step": 58900
     },
     {
-      "epoch": 6.344849854698095,
-      "grad_norm": 0.7850790619850159,
-      "learning_rate": 0.00021968106885033938,
-      "loss": 3.2693,
+      "epoch": 6.355795148247978,
+      "grad_norm": 0.6934738159179688,
+      "learning_rate": 0.00021903076092822448,
+      "loss": 3.2732,
       "step": 58950
     },
     {
-      "epoch": 6.350231406737703,
-      "grad_norm": 0.7271026968955994,
-      "learning_rate": 0.00021935782782027798,
-      "loss": 3.2645,
+      "epoch": 6.361185983827493,
+      "grad_norm": 0.7075552940368652,
+      "learning_rate": 0.00021870696168375606,
+      "loss": 3.267,
       "step": 59000
     },
     {
-      "epoch": 6.350231406737703,
-      "eval_accuracy": 0.384902529124135,
-      "eval_loss": 3.378819704055786,
-      "eval_runtime": 223.2378,
-      "eval_samples_per_second": 80.681,
-      "eval_steps_per_second": 5.044,
+      "epoch": 6.361185983827493,
+      "eval_accuracy": 0.3854025497138572,
+      "eval_loss": 3.374002456665039,
+      "eval_runtime": 183.9817,
+      "eval_samples_per_second": 97.896,
+      "eval_steps_per_second": 6.12,
       "step": 59000
     },
     {
-      "epoch": 6.355612958777312,
-      "grad_norm": 0.7542627453804016,
-      "learning_rate": 0.00021903458679021655,
-      "loss": 3.2872,
+      "epoch": 6.366576819407008,
+      "grad_norm": 0.7140334844589233,
+      "learning_rate": 0.0002183831624392876,
+      "loss": 3.2728,
       "step": 59050
     },
     {
-      "epoch": 6.360994510816919,
-      "grad_norm": 0.7405682802200317,
-      "learning_rate": 0.0002187113457601551,
-      "loss": 3.266,
+      "epoch": 6.3719676549865225,
+      "grad_norm": 0.7283408641815186,
+      "learning_rate": 0.00021805936319481918,
+      "loss": 3.2717,
       "step": 59100
     },
     {
-      "epoch": 6.366376062856528,
-      "grad_norm": 0.7460401654243469,
-      "learning_rate": 0.00021838810473009373,
-      "loss": 3.2759,
+      "epoch": 6.377358490566038,
+      "grad_norm": 0.7204625606536865,
+      "learning_rate": 0.00021773556395035076,
+      "loss": 3.2826,
       "step": 59150
     },
     {
-      "epoch": 6.371757614896136,
-      "grad_norm": 0.7826977372169495,
-      "learning_rate": 0.0002180648637000323,
-      "loss": 3.2753,
+      "epoch": 6.382749326145553,
+      "grad_norm": 0.7599811553955078,
+      "learning_rate": 0.00021741176470588234,
+      "loss": 3.2864,
       "step": 59200
     },
     {
-      "epoch": 6.377139166935744,
-      "grad_norm": 0.7382274270057678,
-      "learning_rate": 0.0002177416226699709,
-      "loss": 3.2868,
+      "epoch": 6.388140161725067,
+      "grad_norm": 0.7962540984153748,
+      "learning_rate": 0.00021708796546141392,
+      "loss": 3.275,
       "step": 59250
     },
     {
-      "epoch": 6.382520718975353,
-      "grad_norm": 0.7684206962585449,
-      "learning_rate": 0.00021741838163990946,
-      "loss": 3.2841,
+      "epoch": 6.393530997304582,
+      "grad_norm": 0.7492397427558899,
+      "learning_rate": 0.0002167641662169455,
+      "loss": 3.2952,
       "step": 59300
     },
     {
-      "epoch": 6.387902271014961,
-      "grad_norm": 0.7826603055000305,
-      "learning_rate": 0.00021709514060984806,
-      "loss": 3.2647,
+      "epoch": 6.398921832884097,
+      "grad_norm": 0.6940380930900574,
+      "learning_rate": 0.00021644036697247702,
+      "loss": 3.2731,
       "step": 59350
     },
     {
-      "epoch": 6.393283823054569,
-      "grad_norm": 0.760877788066864,
-      "learning_rate": 0.00021677189957978665,
-      "loss": 3.2792,
+      "epoch": 6.404312668463612,
+      "grad_norm": 0.6801337003707886,
+      "learning_rate": 0.0002161165677280086,
+      "loss": 3.2718,
       "step": 59400
     },
     {
-      "epoch": 6.398665375094177,
-      "grad_norm": 0.7618247866630554,
-      "learning_rate": 0.00021644865854972522,
-      "loss": 3.2909,
+      "epoch": 6.409703504043127,
+      "grad_norm": 0.7083103060722351,
+      "learning_rate": 0.00021579276848354017,
+      "loss": 3.2852,
       "step": 59450
     },
     {
-      "epoch": 6.404046927133785,
-      "grad_norm": 0.9166077971458435,
-      "learning_rate": 0.0002161254175196638,
-      "loss": 3.2866,
+      "epoch": 6.415094339622642,
+      "grad_norm": 0.7410844564437866,
+      "learning_rate": 0.00021546896923907175,
+      "loss": 3.274,
       "step": 59500
     },
     {
-      "epoch": 6.4094284791733935,
-      "grad_norm": 0.764710009098053,
-      "learning_rate": 0.0002158021764896024,
-      "loss": 3.2599,
+      "epoch": 6.420485175202156,
+      "grad_norm": 0.7300574779510498,
+      "learning_rate": 0.00021514516999460333,
+      "loss": 3.2643,
       "step": 59550
     },
     {
-      "epoch": 6.414810031213002,
-      "grad_norm": 0.7640218734741211,
-      "learning_rate": 0.00021547893545954098,
-      "loss": 3.2676,
+      "epoch": 6.425876010781671,
+      "grad_norm": 0.7533345818519592,
+      "learning_rate": 0.0002148213707501349,
+      "loss": 3.274,
       "step": 59600
     },
     {
-      "epoch": 6.42019158325261,
-      "grad_norm": 0.7581861615180969,
-      "learning_rate": 0.00021515569442947955,
-      "loss": 3.3134,
+      "epoch": 6.431266846361186,
+      "grad_norm": 0.7472398281097412,
+      "learning_rate": 0.00021450404749055586,
+      "loss": 3.2828,
       "step": 59650
     },
     {
-      "epoch": 6.425573135292218,
-      "grad_norm": 0.7885519862174988,
-      "learning_rate": 0.00021483245339941817,
-      "loss": 3.2905,
+      "epoch": 6.436657681940701,
+      "grad_norm": 0.7821720242500305,
+      "learning_rate": 0.00021418024824608738,
+      "loss": 3.2761,
       "step": 59700
     },
     {
-      "epoch": 6.430954687331827,
-      "grad_norm": 0.7909292578697205,
-      "learning_rate": 0.00021450921236935674,
-      "loss": 3.2853,
+      "epoch": 6.442048517520216,
+      "grad_norm": 0.7330186367034912,
+      "learning_rate": 0.00021385644900161896,
+      "loss": 3.2885,
       "step": 59750
     },
     {
-      "epoch": 6.4363362393714345,
-      "grad_norm": 0.7698927521705627,
-      "learning_rate": 0.0002141859713392953,
-      "loss": 3.2799,
+      "epoch": 6.44743935309973,
+      "grad_norm": 0.7098835110664368,
+      "learning_rate": 0.00021353264975715054,
+      "loss": 3.2755,
       "step": 59800
     },
     {
-      "epoch": 6.441717791411043,
-      "grad_norm": 0.7979402542114258,
-      "learning_rate": 0.0002138627303092339,
-      "loss": 3.3018,
+      "epoch": 6.452830188679245,
+      "grad_norm": 0.7708532810211182,
+      "learning_rate": 0.00021320885051268211,
+      "loss": 3.2859,
       "step": 59850
     },
     {
-      "epoch": 6.447099343450651,
-      "grad_norm": 0.7837560772895813,
-      "learning_rate": 0.0002135394892791725,
-      "loss": 3.2739,
+      "epoch": 6.45822102425876,
+      "grad_norm": 0.7420670390129089,
+      "learning_rate": 0.0002128850512682137,
+      "loss": 3.2744,
       "step": 59900
     },
     {
-      "epoch": 6.452480895490259,
-      "grad_norm": 0.7438477873802185,
-      "learning_rate": 0.00021321624824911106,
-      "loss": 3.2845,
+      "epoch": 6.463611859838275,
+      "grad_norm": 0.7334537506103516,
+      "learning_rate": 0.00021256125202374527,
+      "loss": 3.2862,
       "step": 59950
     },
     {
-      "epoch": 6.457862447529868,
-      "grad_norm": 0.7494663596153259,
-      "learning_rate": 0.00021289300721904965,
-      "loss": 3.2801,
+      "epoch": 6.46900269541779,
+      "grad_norm": 0.7944307923316956,
+      "learning_rate": 0.00021223745277927685,
+      "loss": 3.2834,
       "step": 60000
     },
     {
-      "epoch": 6.457862447529868,
-      "eval_accuracy": 0.38531291108141025,
-      "eval_loss": 3.374072313308716,
-      "eval_runtime": 209.5252,
-      "eval_samples_per_second": 85.961,
-      "eval_steps_per_second": 5.374,
+      "epoch": 6.46900269541779,
+      "eval_accuracy": 0.38511244650339255,
+      "eval_loss": 3.37099027633667,
+      "eval_runtime": 184.5482,
+      "eval_samples_per_second": 97.595,
+      "eval_steps_per_second": 6.101,
       "step": 60000
     },
     {
-      "epoch": 6.4632439995694755,
-      "grad_norm": 0.7274054288864136,
-      "learning_rate": 0.00021256976618898822,
-      "loss": 3.2882,
+      "epoch": 6.474393530997305,
+      "grad_norm": 0.6979678273200989,
+      "learning_rate": 0.00021191365353480843,
+      "loss": 3.2911,
       "step": 60050
     },
     {
-      "epoch": 6.468625551609084,
-      "grad_norm": 0.7407465577125549,
-      "learning_rate": 0.00021225298997952806,
-      "loss": 3.2824,
+      "epoch": 6.479784366576819,
+      "grad_norm": 0.762915313243866,
+      "learning_rate": 0.00021158985429033995,
+      "loss": 3.2877,
       "step": 60100
     },
     {
-      "epoch": 6.474007103648693,
-      "grad_norm": 0.7144964933395386,
-      "learning_rate": 0.00021192974894946662,
-      "loss": 3.2957,
+      "epoch": 6.485175202156334,
+      "grad_norm": 0.7652758359909058,
+      "learning_rate": 0.00021126605504587153,
+      "loss": 3.2785,
       "step": 60150
     },
     {
-      "epoch": 6.4793886556883,
-      "grad_norm": 0.7461879849433899,
-      "learning_rate": 0.00021160650791940524,
-      "loss": 3.2838,
+      "epoch": 6.490566037735849,
+      "grad_norm": 0.8516004085540771,
+      "learning_rate": 0.0002109422558014031,
+      "loss": 3.3036,
       "step": 60200
     },
     {
-      "epoch": 6.484770207727909,
-      "grad_norm": 0.755248486995697,
-      "learning_rate": 0.0002112832668893438,
-      "loss": 3.2862,
+      "epoch": 6.495956873315364,
+      "grad_norm": 0.7091172337532043,
+      "learning_rate": 0.00021061845655693468,
+      "loss": 3.277,
       "step": 60250
     },
     {
-      "epoch": 6.490151759767517,
-      "grad_norm": 0.7556350231170654,
-      "learning_rate": 0.00021096002585928238,
-      "loss": 3.2679,
+      "epoch": 6.501347708894879,
+      "grad_norm": 0.70334792137146,
+      "learning_rate": 0.00021029465731246626,
+      "loss": 3.2818,
       "step": 60300
     },
     {
-      "epoch": 6.495533311807125,
-      "grad_norm": 0.7320284843444824,
-      "learning_rate": 0.00021063678482922097,
-      "loss": 3.2899,
+      "epoch": 6.506738544474393,
+      "grad_norm": 0.7396783828735352,
+      "learning_rate": 0.00020997085806799784,
+      "loss": 3.2833,
       "step": 60350
     },
     {
-      "epoch": 6.500914863846734,
-      "grad_norm": 0.7573622465133667,
-      "learning_rate": 0.00021031354379915957,
-      "loss": 3.2869,
+      "epoch": 6.512129380053908,
+      "grad_norm": 0.7701082229614258,
+      "learning_rate": 0.0002096470588235294,
+      "loss": 3.2801,
       "step": 60400
     },
     {
-      "epoch": 6.506296415886341,
-      "grad_norm": 0.7960693836212158,
-      "learning_rate": 0.00020999030276909814,
-      "loss": 3.2784,
+      "epoch": 6.517520215633423,
+      "grad_norm": 0.7087827324867249,
+      "learning_rate": 0.00020932325957906097,
+      "loss": 3.2974,
       "step": 60450
     },
     {
-      "epoch": 6.51167796792595,
-      "grad_norm": 0.8241477608680725,
-      "learning_rate": 0.00020966706173903673,
-      "loss": 3.2832,
+      "epoch": 6.5229110512129385,
+      "grad_norm": 0.727470338344574,
+      "learning_rate": 0.00020899946033459254,
+      "loss": 3.2827,
       "step": 60500
     },
     {
-      "epoch": 6.517059519965558,
-      "grad_norm": 0.773271381855011,
-      "learning_rate": 0.0002093438207089753,
-      "loss": 3.2755,
+      "epoch": 6.528301886792453,
+      "grad_norm": 0.7513377070426941,
+      "learning_rate": 0.0002086756610901241,
+      "loss": 3.3028,
       "step": 60550
     },
     {
-      "epoch": 6.522441072005166,
-      "grad_norm": 0.7589271664619446,
-      "learning_rate": 0.00020902057967891387,
-      "loss": 3.2864,
+      "epoch": 6.533692722371968,
+      "grad_norm": 0.7205342054367065,
+      "learning_rate": 0.00020835186184565567,
+      "loss": 3.2982,
       "step": 60600
     },
     {
-      "epoch": 6.5278226240447745,
-      "grad_norm": 0.735211968421936,
-      "learning_rate": 0.0002086973386488525,
-      "loss": 3.2909,
+      "epoch": 6.539083557951482,
+      "grad_norm": 0.7690211534500122,
+      "learning_rate": 0.00020802806260118725,
+      "loss": 3.2856,
       "step": 60650
     },
     {
-      "epoch": 6.533204176084383,
-      "grad_norm": 0.767329752445221,
-      "learning_rate": 0.00020837409761879106,
-      "loss": 3.2849,
+      "epoch": 6.544474393530997,
+      "grad_norm": 0.807765543460846,
+      "learning_rate": 0.00020770426335671883,
+      "loss": 3.2996,
       "step": 60700
     },
     {
-      "epoch": 6.538585728123991,
-      "grad_norm": 0.7533854842185974,
-      "learning_rate": 0.00020805085658872962,
-      "loss": 3.2829,
+      "epoch": 6.549865229110512,
+      "grad_norm": 0.7733821868896484,
+      "learning_rate": 0.00020738046411225038,
+      "loss": 3.2863,
       "step": 60750
     },
     {
-      "epoch": 6.543967280163599,
-      "grad_norm": 0.8554604649543762,
-      "learning_rate": 0.00020772761555866825,
-      "loss": 3.284,
+      "epoch": 6.555256064690027,
+      "grad_norm": 0.7033819556236267,
+      "learning_rate": 0.00020705666486778196,
+      "loss": 3.2779,
       "step": 60800
     },
     {
-      "epoch": 6.549348832203208,
-      "grad_norm": 0.7716811895370483,
-      "learning_rate": 0.0002074043745286068,
-      "loss": 3.2828,
+      "epoch": 6.560646900269542,
+      "grad_norm": 0.7843565940856934,
+      "learning_rate": 0.00020673286562331353,
+      "loss": 3.2951,
       "step": 60850
     },
     {
-      "epoch": 6.5547303842428155,
-      "grad_norm": 0.7305426597595215,
-      "learning_rate": 0.00020708113349854538,
-      "loss": 3.3054,
+      "epoch": 6.566037735849057,
+      "grad_norm": 0.7759973406791687,
+      "learning_rate": 0.0002064090663788451,
+      "loss": 3.2784,
       "step": 60900
     },
     {
-      "epoch": 6.560111936282424,
-      "grad_norm": 0.8045239448547363,
-      "learning_rate": 0.00020675789246848397,
-      "loss": 3.2869,
+      "epoch": 6.571428571428571,
+      "grad_norm": 0.7242240309715271,
+      "learning_rate": 0.00020608526713437666,
+      "loss": 3.2806,
       "step": 60950
     },
     {
-      "epoch": 6.565493488322032,
-      "grad_norm": 0.7284274697303772,
-      "learning_rate": 0.00020643465143842257,
-      "loss": 3.2836,
+      "epoch": 6.576819407008086,
+      "grad_norm": 0.7385703325271606,
+      "learning_rate": 0.00020576146788990824,
+      "loss": 3.2745,
       "step": 61000
     },
     {
-      "epoch": 6.565493488322032,
-      "eval_accuracy": 0.38564528026523476,
-      "eval_loss": 3.370778799057007,
-      "eval_runtime": 201.2966,
-      "eval_samples_per_second": 89.475,
-      "eval_steps_per_second": 5.594,
+      "epoch": 6.576819407008086,
+      "eval_accuracy": 0.38582933825719246,
+      "eval_loss": 3.3675119876861572,
+      "eval_runtime": 184.8212,
+      "eval_samples_per_second": 97.451,
+      "eval_steps_per_second": 6.092,
       "step": 61000
     },
     {
-      "epoch": 6.57087504036164,
-      "grad_norm": 0.7565922737121582,
-      "learning_rate": 0.00020611141040836116,
-      "loss": 3.285,
+      "epoch": 6.5822102425876015,
+      "grad_norm": 0.7266719937324524,
+      "learning_rate": 0.0002054376686454398,
+      "loss": 3.2732,
       "step": 61050
     },
     {
-      "epoch": 6.576256592401249,
-      "grad_norm": 0.8001964092254639,
-      "learning_rate": 0.00020578816937829973,
-      "loss": 3.2914,
+      "epoch": 6.587601078167116,
+      "grad_norm": 0.726216197013855,
+      "learning_rate": 0.00020511386940097137,
+      "loss": 3.264,
       "step": 61100
     },
     {
-      "epoch": 6.5816381444408565,
-      "grad_norm": 0.7406817078590393,
-      "learning_rate": 0.0002054649283482383,
-      "loss": 3.2947,
+      "epoch": 6.592991913746631,
+      "grad_norm": 0.7396004796028137,
+      "learning_rate": 0.00020479007015650294,
+      "loss": 3.2771,
       "step": 61150
     },
     {
-      "epoch": 6.587019696480465,
-      "grad_norm": 0.7439241409301758,
-      "learning_rate": 0.00020514168731817692,
-      "loss": 3.2742,
+      "epoch": 6.598382749326145,
+      "grad_norm": 0.7365099787712097,
+      "learning_rate": 0.00020446627091203452,
+      "loss": 3.2785,
       "step": 61200
     },
     {
-      "epoch": 6.592401248520073,
-      "grad_norm": 0.7465095520019531,
-      "learning_rate": 0.0002048184462881155,
-      "loss": 3.2787,
+      "epoch": 6.60377358490566,
+      "grad_norm": 0.6806655526161194,
+      "learning_rate": 0.0002041424716675661,
+      "loss": 3.2773,
       "step": 61250
     },
     {
-      "epoch": 6.597782800559681,
-      "grad_norm": 0.7967016100883484,
-      "learning_rate": 0.00020449520525805406,
-      "loss": 3.2923,
+      "epoch": 6.609164420485175,
+      "grad_norm": 0.7373830676078796,
+      "learning_rate": 0.00020381867242309768,
+      "loss": 3.2865,
       "step": 61300
     },
     {
-      "epoch": 6.60316435259929,
-      "grad_norm": 0.7551931142807007,
-      "learning_rate": 0.00020417196422799268,
-      "loss": 3.2902,
+      "epoch": 6.6145552560646905,
+      "grad_norm": 0.7143996953964233,
+      "learning_rate": 0.00020349487317862926,
+      "loss": 3.2832,
       "step": 61350
     },
     {
-      "epoch": 6.608545904638898,
-      "grad_norm": 0.7699846029281616,
-      "learning_rate": 0.00020384872319793125,
-      "loss": 3.2934,
+      "epoch": 6.619946091644205,
+      "grad_norm": 0.7903742790222168,
+      "learning_rate": 0.00020317107393416078,
+      "loss": 3.2696,
       "step": 61400
     },
     {
-      "epoch": 6.613927456678506,
-      "grad_norm": 0.7986684441566467,
-      "learning_rate": 0.0002035254821678698,
-      "loss": 3.2858,
+      "epoch": 6.62533692722372,
+      "grad_norm": 0.7577932476997375,
+      "learning_rate": 0.00020284727468969236,
+      "loss": 3.2969,
       "step": 61450
     },
     {
-      "epoch": 6.619309008718115,
-      "grad_norm": 0.7962992787361145,
-      "learning_rate": 0.0002032022411378084,
-      "loss": 3.2827,
+      "epoch": 6.630727762803234,
+      "grad_norm": 0.7393811941146851,
+      "learning_rate": 0.00020252347544522393,
+      "loss": 3.2759,
       "step": 61500
     },
     {
-      "epoch": 6.624690560757722,
-      "grad_norm": 0.7751727104187012,
-      "learning_rate": 0.000202879000107747,
-      "loss": 3.2885,
+      "epoch": 6.636118598382749,
+      "grad_norm": 0.7777678966522217,
+      "learning_rate": 0.0002021996762007555,
+      "loss": 3.2912,
       "step": 61550
     },
     {
-      "epoch": 6.630072112797331,
-      "grad_norm": 0.7956772446632385,
-      "learning_rate": 0.00020255575907768557,
-      "loss": 3.3027,
+      "epoch": 6.6415094339622645,
+      "grad_norm": 0.7335239052772522,
+      "learning_rate": 0.0002018758769562871,
+      "loss": 3.2766,
       "step": 61600
     },
     {
-      "epoch": 6.635453664836939,
-      "grad_norm": 0.7654341459274292,
-      "learning_rate": 0.00020223251804762416,
-      "loss": 3.2883,
+      "epoch": 6.646900269541779,
+      "grad_norm": 0.7621884346008301,
+      "learning_rate": 0.00020155207771181867,
+      "loss": 3.2724,
       "step": 61650
     },
     {
-      "epoch": 6.640835216876547,
-      "grad_norm": 0.7234878540039062,
-      "learning_rate": 0.00020190927701756273,
-      "loss": 3.3,
+      "epoch": 6.652291105121294,
+      "grad_norm": 0.7772314548492432,
+      "learning_rate": 0.00020122827846735024,
+      "loss": 3.3009,
       "step": 61700
     },
     {
-      "epoch": 6.6462167689161555,
-      "grad_norm": 0.7403562068939209,
-      "learning_rate": 0.00020158603598750133,
-      "loss": 3.2763,
+      "epoch": 6.657681940700809,
+      "grad_norm": 0.7585477828979492,
+      "learning_rate": 0.00020090447922288182,
+      "loss": 3.2896,
       "step": 61750
     },
     {
-      "epoch": 6.651598320955763,
-      "grad_norm": 0.7899567484855652,
-      "learning_rate": 0.00020126279495743992,
-      "loss": 3.2967,
+      "epoch": 6.663072776280323,
+      "grad_norm": 0.7339568138122559,
+      "learning_rate": 0.00020058067997841335,
+      "loss": 3.2802,
       "step": 61800
     },
     {
-      "epoch": 6.656979872995372,
-      "grad_norm": 0.7615472674369812,
-      "learning_rate": 0.0002009395539273785,
-      "loss": 3.2782,
+      "epoch": 6.668463611859838,
+      "grad_norm": 0.7589542865753174,
+      "learning_rate": 0.00020025688073394492,
+      "loss": 3.2881,
       "step": 61850
     },
     {
-      "epoch": 6.66236142503498,
-      "grad_norm": 0.7665838599205017,
-      "learning_rate": 0.00020061631289731706,
-      "loss": 3.2865,
+      "epoch": 6.6738544474393535,
+      "grad_norm": 0.7213110327720642,
+      "learning_rate": 0.0001999330814894765,
+      "loss": 3.2823,
       "step": 61900
     },
     {
-      "epoch": 6.667742977074588,
-      "grad_norm": 0.769633948802948,
-      "learning_rate": 0.00020029307186725568,
-      "loss": 3.3017,
+      "epoch": 6.679245283018868,
+      "grad_norm": 0.7335101366043091,
+      "learning_rate": 0.00019960928224500808,
+      "loss": 3.2817,
       "step": 61950
     },
     {
-      "epoch": 6.6731245291141965,
-      "grad_norm": 0.7775278687477112,
-      "learning_rate": 0.00019996983083719425,
-      "loss": 3.2808,
+      "epoch": 6.684636118598383,
+      "grad_norm": 0.745826005935669,
+      "learning_rate": 0.00019928548300053966,
+      "loss": 3.2936,
       "step": 62000
     },
     {
-      "epoch": 6.6731245291141965,
-      "eval_accuracy": 0.3862611248333672,
-      "eval_loss": 3.3647584915161133,
-      "eval_runtime": 210.4271,
-      "eval_samples_per_second": 85.593,
-      "eval_steps_per_second": 5.351,
+      "epoch": 6.684636118598383,
+      "eval_accuracy": 0.38647299796460544,
+      "eval_loss": 3.3618156909942627,
+      "eval_runtime": 184.5868,
+      "eval_samples_per_second": 97.575,
+      "eval_steps_per_second": 6.1,
       "step": 62000
     },
     {
-      "epoch": 6.678506081153805,
-      "grad_norm": 0.7849041223526001,
-      "learning_rate": 0.00019964658980713284,
-      "loss": 3.2781,
+      "epoch": 6.690026954177897,
+      "grad_norm": 0.7264554500579834,
+      "learning_rate": 0.00019896168375607123,
+      "loss": 3.2809,
       "step": 62050
     },
     {
-      "epoch": 6.683887633193413,
-      "grad_norm": 0.8136065602302551,
-      "learning_rate": 0.00019932334877707144,
-      "loss": 3.3005,
+      "epoch": 6.695417789757412,
+      "grad_norm": 0.6842035055160522,
+      "learning_rate": 0.00019863788451160278,
+      "loss": 3.278,
       "step": 62100
     },
     {
-      "epoch": 6.689269185233021,
-      "grad_norm": 0.7949984073638916,
-      "learning_rate": 0.00019900657256761124,
-      "loss": 3.2789,
+      "epoch": 6.7008086253369274,
+      "grad_norm": 0.8485304117202759,
+      "learning_rate": 0.00019831408526713436,
+      "loss": 3.2842,
       "step": 62150
     },
     {
-      "epoch": 6.69465073727263,
-      "grad_norm": 0.7674379348754883,
-      "learning_rate": 0.0001986833315375498,
-      "loss": 3.2832,
+      "epoch": 6.706199460916442,
+      "grad_norm": 0.7461103796958923,
+      "learning_rate": 0.00019799028602266594,
+      "loss": 3.2869,
       "step": 62200
     },
     {
-      "epoch": 6.7000322893122375,
-      "grad_norm": 0.8072303533554077,
-      "learning_rate": 0.00019836009050748838,
-      "loss": 3.2916,
+      "epoch": 6.711590296495957,
+      "grad_norm": 0.6902636289596558,
+      "learning_rate": 0.0001976664867781975,
+      "loss": 3.2876,
       "step": 62250
     },
     {
-      "epoch": 6.705413841351846,
-      "grad_norm": 0.7814190983772278,
-      "learning_rate": 0.000198036849477427,
-      "loss": 3.3209,
+      "epoch": 6.716981132075472,
+      "grad_norm": 0.7343002557754517,
+      "learning_rate": 0.00019734268753372907,
+      "loss": 3.2796,
       "step": 62300
     },
     {
-      "epoch": 6.710795393391454,
-      "grad_norm": 0.788261890411377,
-      "learning_rate": 0.00019772007326796678,
-      "loss": 3.2981,
+      "epoch": 6.722371967654986,
+      "grad_norm": 0.7256414294242859,
+      "learning_rate": 0.00019701888828926065,
+      "loss": 3.2851,
       "step": 62350
     },
     {
-      "epoch": 6.716176945431062,
-      "grad_norm": 0.7617114186286926,
-      "learning_rate": 0.0001973968322379054,
-      "loss": 3.2803,
+      "epoch": 6.727762803234501,
+      "grad_norm": 0.7835127711296082,
+      "learning_rate": 0.0001966950890447922,
+      "loss": 3.287,
       "step": 62400
     },
     {
-      "epoch": 6.721558497470671,
-      "grad_norm": 0.7262803316116333,
-      "learning_rate": 0.00019707359120784397,
-      "loss": 3.2805,
+      "epoch": 6.7331536388140165,
+      "grad_norm": 0.739169180393219,
+      "learning_rate": 0.00019637776578521315,
+      "loss": 3.2838,
       "step": 62450
     },
     {
-      "epoch": 6.7269400495102785,
-      "grad_norm": 0.7623103260993958,
-      "learning_rate": 0.00019675035017778253,
-      "loss": 3.2771,
+      "epoch": 6.738544474393531,
+      "grad_norm": 0.7858089804649353,
+      "learning_rate": 0.00019605396654074473,
+      "loss": 3.2787,
       "step": 62500
     },
     {
-      "epoch": 6.732321601549887,
-      "grad_norm": 0.7562897801399231,
-      "learning_rate": 0.00019642710914772113,
-      "loss": 3.292,
+      "epoch": 6.743935309973046,
+      "grad_norm": 0.7251155376434326,
+      "learning_rate": 0.0001957301672962763,
+      "loss": 3.2967,
       "step": 62550
     },
     {
-      "epoch": 6.737703153589496,
-      "grad_norm": 0.7772451043128967,
-      "learning_rate": 0.00019610386811765972,
-      "loss": 3.2792,
+      "epoch": 6.74932614555256,
+      "grad_norm": 0.739592969417572,
+      "learning_rate": 0.00019540636805180785,
+      "loss": 3.2745,
       "step": 62600
     },
     {
-      "epoch": 6.743084705629103,
-      "grad_norm": 0.8128804564476013,
-      "learning_rate": 0.00019578062708759832,
-      "loss": 3.3054,
+      "epoch": 6.754716981132075,
+      "grad_norm": 0.7757728695869446,
+      "learning_rate": 0.00019508256880733943,
+      "loss": 3.2818,
       "step": 62650
     },
     {
-      "epoch": 6.748466257668712,
-      "grad_norm": 0.8137348294258118,
-      "learning_rate": 0.00019545738605753689,
-      "loss": 3.2941,
+      "epoch": 6.7601078167115904,
+      "grad_norm": 0.7743464112281799,
+      "learning_rate": 0.000194758769562871,
+      "loss": 3.2802,
       "step": 62700
     },
     {
-      "epoch": 6.75384780970832,
-      "grad_norm": 0.7956899404525757,
-      "learning_rate": 0.00019513414502747545,
-      "loss": 3.2922,
+      "epoch": 6.765498652291106,
+      "grad_norm": 0.7496610283851624,
+      "learning_rate": 0.00019443497031840256,
+      "loss": 3.2758,
       "step": 62750
     },
     {
-      "epoch": 6.759229361747928,
-      "grad_norm": 0.7582367062568665,
-      "learning_rate": 0.00019481090399741408,
-      "loss": 3.2699,
+      "epoch": 6.77088948787062,
+      "grad_norm": 0.7385218143463135,
+      "learning_rate": 0.00019411117107393414,
+      "loss": 3.2788,
       "step": 62800
     },
     {
-      "epoch": 6.7646109137875365,
-      "grad_norm": 1.1021339893341064,
-      "learning_rate": 0.00019448766296735264,
-      "loss": 3.2984,
+      "epoch": 6.776280323450135,
+      "grad_norm": 0.7698500752449036,
+      "learning_rate": 0.00019378737182946572,
+      "loss": 3.2735,
       "step": 62850
     },
     {
-      "epoch": 6.769992465827144,
-      "grad_norm": 0.8446829319000244,
-      "learning_rate": 0.0001941644219372912,
-      "loss": 3.2944,
+      "epoch": 6.781671159029649,
+      "grad_norm": 0.7696132659912109,
+      "learning_rate": 0.0001934635725849973,
+      "loss": 3.2996,
       "step": 62900
     },
     {
-      "epoch": 6.775374017866753,
-      "grad_norm": 0.7830290794372559,
-      "learning_rate": 0.00019384118090722983,
-      "loss": 3.2988,
+      "epoch": 6.787061994609164,
+      "grad_norm": 0.7788563966751099,
+      "learning_rate": 0.00019313977334052887,
+      "loss": 3.2799,
       "step": 62950
     },
     {
-      "epoch": 6.780755569906361,
-      "grad_norm": 0.7751957178115845,
-      "learning_rate": 0.0001935179398771684,
-      "loss": 3.2776,
+      "epoch": 6.7924528301886795,
+      "grad_norm": 0.7606647610664368,
+      "learning_rate": 0.00019281597409606042,
+      "loss": 3.2713,
       "step": 63000
     },
     {
-      "epoch": 6.780755569906361,
-      "eval_accuracy": 0.38648049501386467,
-      "eval_loss": 3.3618505001068115,
-      "eval_runtime": 180.2451,
-      "eval_samples_per_second": 99.925,
-      "eval_steps_per_second": 6.247,
+      "epoch": 6.7924528301886795,
+      "eval_accuracy": 0.3866354340318881,
+      "eval_loss": 3.3592801094055176,
+      "eval_runtime": 184.9177,
+      "eval_samples_per_second": 97.4,
+      "eval_steps_per_second": 6.089,
       "step": 63000
     },
     {
-      "epoch": 6.786137121945969,
-      "grad_norm": 0.8304117321968079,
-      "learning_rate": 0.00019319469884710697,
-      "loss": 3.2927,
+      "epoch": 6.797843665768194,
+      "grad_norm": 0.7435787320137024,
+      "learning_rate": 0.00019249217485159197,
+      "loss": 3.2926,
       "step": 63050
     },
     {
-      "epoch": 6.7915186739855775,
-      "grad_norm": 0.7430444955825806,
-      "learning_rate": 0.00019287145781704556,
-      "loss": 3.2934,
+      "epoch": 6.803234501347709,
+      "grad_norm": 0.7269057035446167,
+      "learning_rate": 0.00019216837560712355,
+      "loss": 3.2883,
       "step": 63100
     },
     {
-      "epoch": 6.796900226025185,
-      "grad_norm": 0.756223738193512,
-      "learning_rate": 0.00019254821678698416,
-      "loss": 3.2953,
+      "epoch": 6.808625336927224,
+      "grad_norm": 0.7587324380874634,
+      "learning_rate": 0.00019184457636265513,
+      "loss": 3.2823,
       "step": 63150
     },
     {
-      "epoch": 6.802281778064794,
-      "grad_norm": 0.7750756144523621,
-      "learning_rate": 0.00019222497575692272,
-      "loss": 3.2911,
+      "epoch": 6.814016172506738,
+      "grad_norm": 0.734873354434967,
+      "learning_rate": 0.0001915207771181867,
+      "loss": 3.2986,
       "step": 63200
     },
     {
-      "epoch": 6.807663330104402,
-      "grad_norm": 0.7659454941749573,
-      "learning_rate": 0.00019190173472686132,
-      "loss": 3.2952,
+      "epoch": 6.819407008086253,
+      "grad_norm": 0.9111540913581848,
+      "learning_rate": 0.00019119697787371828,
+      "loss": 3.2814,
       "step": 63250
     },
     {
-      "epoch": 6.813044882144011,
-      "grad_norm": 0.7663837671279907,
-      "learning_rate": 0.0001915784936967999,
-      "loss": 3.2953,
+      "epoch": 6.824797843665769,
+      "grad_norm": 0.7694358825683594,
+      "learning_rate": 0.00019087317862924986,
+      "loss": 3.284,
       "step": 63300
     },
     {
-      "epoch": 6.8184264341836185,
-      "grad_norm": 0.7659280300140381,
-      "learning_rate": 0.00019125525266673845,
-      "loss": 3.2848,
+      "epoch": 6.830188679245283,
+      "grad_norm": 0.7078369855880737,
+      "learning_rate": 0.00019054937938478144,
+      "loss": 3.2659,
       "step": 63350
     },
     {
-      "epoch": 6.823807986223227,
-      "grad_norm": 0.8383036255836487,
-      "learning_rate": 0.00019093201163667708,
-      "loss": 3.2811,
+      "epoch": 6.835579514824798,
+      "grad_norm": 0.7470163106918335,
+      "learning_rate": 0.00019022558014031302,
+      "loss": 3.2836,
       "step": 63400
     },
     {
-      "epoch": 6.829189538262835,
-      "grad_norm": 0.7594733238220215,
-      "learning_rate": 0.00019060877060661564,
-      "loss": 3.2928,
+      "epoch": 6.840970350404312,
+      "grad_norm": 0.752375602722168,
+      "learning_rate": 0.00018990178089584454,
+      "loss": 3.2735,
       "step": 63450
     },
     {
-      "epoch": 6.834571090302443,
-      "grad_norm": 0.7703410983085632,
-      "learning_rate": 0.0001902855295765542,
-      "loss": 3.2949,
+      "epoch": 6.846361185983827,
+      "grad_norm": 0.7561582326889038,
+      "learning_rate": 0.00018957798165137612,
+      "loss": 3.3003,
       "step": 63500
     },
     {
-      "epoch": 6.839952642342052,
-      "grad_norm": 0.7872480154037476,
-      "learning_rate": 0.00018996228854649283,
-      "loss": 3.2945,
+      "epoch": 6.8517520215633425,
+      "grad_norm": 0.8334784507751465,
+      "learning_rate": 0.00018926065839179707,
+      "loss": 3.2757,
       "step": 63550
     },
     {
-      "epoch": 6.8453341943816595,
-      "grad_norm": 0.7895832061767578,
-      "learning_rate": 0.0001896390475164314,
-      "loss": 3.2881,
+      "epoch": 6.857142857142857,
+      "grad_norm": 0.7926870584487915,
+      "learning_rate": 0.00018893685914732865,
+      "loss": 3.3033,
       "step": 63600
     },
     {
-      "epoch": 6.850715746421268,
-      "grad_norm": 0.7981587052345276,
-      "learning_rate": 0.00018931580648637,
-      "loss": 3.2887,
+      "epoch": 6.862533692722372,
+      "grad_norm": 0.747937023639679,
+      "learning_rate": 0.00018861305990286022,
+      "loss": 3.288,
       "step": 63650
     },
     {
-      "epoch": 6.856097298460876,
-      "grad_norm": 0.7759888172149658,
-      "learning_rate": 0.00018899256545630856,
-      "loss": 3.3,
+      "epoch": 6.867924528301887,
+      "grad_norm": 0.7995949983596802,
+      "learning_rate": 0.0001882892606583918,
+      "loss": 3.2874,
       "step": 63700
     },
     {
-      "epoch": 6.861478850500484,
-      "grad_norm": 0.7870774269104004,
-      "learning_rate": 0.00018866932442624716,
-      "loss": 3.2849,
+      "epoch": 6.873315363881401,
+      "grad_norm": 0.7559580206871033,
+      "learning_rate": 0.00018796546141392333,
+      "loss": 3.3054,
       "step": 63750
     },
     {
-      "epoch": 6.866860402540093,
-      "grad_norm": 0.8005082011222839,
-      "learning_rate": 0.00018834608339618575,
-      "loss": 3.2913,
+      "epoch": 6.878706199460916,
+      "grad_norm": 0.8042141199111938,
+      "learning_rate": 0.0001876416621694549,
+      "loss": 3.2964,
       "step": 63800
     },
     {
-      "epoch": 6.8722419545797,
-      "grad_norm": 0.8050322532653809,
-      "learning_rate": 0.00018802284236612432,
-      "loss": 3.295,
+      "epoch": 6.884097035040432,
+      "grad_norm": 0.7531314492225647,
+      "learning_rate": 0.00018731786292498648,
+      "loss": 3.2854,
       "step": 63850
     },
     {
-      "epoch": 6.877623506619309,
-      "grad_norm": 0.8603694438934326,
-      "learning_rate": 0.0001876996013360629,
-      "loss": 3.2935,
+      "epoch": 6.889487870619946,
+      "grad_norm": 0.7664470672607422,
+      "learning_rate": 0.00018699406368051806,
+      "loss": 3.2785,
       "step": 63900
     },
     {
-      "epoch": 6.8830050586589175,
-      "grad_norm": 0.8038960695266724,
-      "learning_rate": 0.0001873763603060015,
-      "loss": 3.2684,
+      "epoch": 6.894878706199461,
+      "grad_norm": 0.8082917332649231,
+      "learning_rate": 0.00018667026443604964,
+      "loss": 3.2875,
       "step": 63950
     },
     {
-      "epoch": 6.888386610698525,
-      "grad_norm": 0.8121204376220703,
-      "learning_rate": 0.00018705311927594008,
-      "loss": 3.2847,
+      "epoch": 6.900269541778976,
+      "grad_norm": 0.7554585337638855,
+      "learning_rate": 0.00018634646519158121,
+      "loss": 3.2732,
       "step": 64000
     },
     {
-      "epoch": 6.888386610698525,
-      "eval_accuracy": 0.3866242427844432,
-      "eval_loss": 3.361416816711426,
-      "eval_runtime": 180.531,
-      "eval_samples_per_second": 99.767,
-      "eval_steps_per_second": 6.237,
+      "epoch": 6.900269541778976,
+      "eval_accuracy": 0.38727192264870536,
+      "eval_loss": 3.3544528484344482,
+      "eval_runtime": 184.9582,
+      "eval_samples_per_second": 97.379,
+      "eval_steps_per_second": 6.088,
       "step": 64000
     },
     {
-      "epoch": 6.893768162738134,
-      "grad_norm": 0.8602204322814941,
-      "learning_rate": 0.00018672987824587864,
-      "loss": 3.3055,
+      "epoch": 6.90566037735849,
+      "grad_norm": 0.7046903371810913,
+      "learning_rate": 0.0001860226659471128,
+      "loss": 3.2814,
       "step": 64050
     },
     {
-      "epoch": 6.899149714777742,
-      "grad_norm": 0.7669270634651184,
-      "learning_rate": 0.00018640663721581727,
-      "loss": 3.2954,
+      "epoch": 6.9110512129380055,
+      "grad_norm": 0.7323727011680603,
+      "learning_rate": 0.00018569886670264434,
+      "loss": 3.2828,
       "step": 64100
     },
     {
-      "epoch": 6.90453126681735,
-      "grad_norm": 0.7600193619728088,
-      "learning_rate": 0.00018608339618575583,
-      "loss": 3.2762,
+      "epoch": 6.916442048517521,
+      "grad_norm": 0.7641932368278503,
+      "learning_rate": 0.00018537506745817592,
+      "loss": 3.2933,
       "step": 64150
     },
     {
-      "epoch": 6.9099128188569585,
-      "grad_norm": 0.7429041862487793,
-      "learning_rate": 0.0001857601551556944,
-      "loss": 3.3104,
+      "epoch": 6.921832884097035,
+      "grad_norm": 0.771299421787262,
+      "learning_rate": 0.00018505126821370747,
+      "loss": 3.2804,
       "step": 64200
     },
     {
-      "epoch": 6.915294370896566,
-      "grad_norm": 0.8116273283958435,
-      "learning_rate": 0.000185436914125633,
-      "loss": 3.2847,
+      "epoch": 6.92722371967655,
+      "grad_norm": 0.7517485022544861,
+      "learning_rate": 0.00018472746896923905,
+      "loss": 3.2964,
       "step": 64250
     },
     {
-      "epoch": 6.920675922936175,
-      "grad_norm": 0.7676055431365967,
-      "learning_rate": 0.0001851136730955716,
-      "loss": 3.2907,
+      "epoch": 6.932614555256064,
+      "grad_norm": 0.7981433868408203,
+      "learning_rate": 0.00018440366972477063,
+      "loss": 3.2848,
       "step": 64300
     },
     {
-      "epoch": 6.926057474975783,
-      "grad_norm": 0.7990931272506714,
-      "learning_rate": 0.00018479043206551016,
-      "loss": 3.2927,
+      "epoch": 6.938005390835579,
+      "grad_norm": 0.7293975949287415,
+      "learning_rate": 0.0001840798704803022,
+      "loss": 3.2789,
       "step": 64350
     },
     {
-      "epoch": 6.931439027015391,
-      "grad_norm": 0.7669125199317932,
-      "learning_rate": 0.00018446719103544875,
-      "loss": 3.2772,
+      "epoch": 6.943396226415095,
+      "grad_norm": 0.7737505435943604,
+      "learning_rate": 0.00018375607123583378,
+      "loss": 3.2927,
       "step": 64400
     },
     {
-      "epoch": 6.9368205790549995,
-      "grad_norm": 0.8254827260971069,
-      "learning_rate": 0.00018414395000538732,
-      "loss": 3.2951,
+      "epoch": 6.948787061994609,
+      "grad_norm": 0.8144184947013855,
+      "learning_rate": 0.00018343227199136533,
+      "loss": 3.2658,
       "step": 64450
     },
     {
-      "epoch": 6.942202131094608,
-      "grad_norm": 0.7549811601638794,
-      "learning_rate": 0.00018382070897532591,
-      "loss": 3.29,
+      "epoch": 6.954177897574124,
+      "grad_norm": 0.839918851852417,
+      "learning_rate": 0.0001831084727468969,
+      "loss": 3.2983,
       "step": 64500
     },
     {
-      "epoch": 6.947583683134216,
-      "grad_norm": 0.8003718256950378,
-      "learning_rate": 0.0001834974679452645,
-      "loss": 3.2858,
+      "epoch": 6.959568733153639,
+      "grad_norm": 0.7097006440162659,
+      "learning_rate": 0.0001827846735024285,
+      "loss": 3.2876,
       "step": 64550
     },
     {
-      "epoch": 6.952965235173824,
-      "grad_norm": 0.8304224014282227,
-      "learning_rate": 0.00018318069173580434,
-      "loss": 3.2969,
+      "epoch": 6.964959568733153,
+      "grad_norm": 0.742793083190918,
+      "learning_rate": 0.00018246087425796004,
+      "loss": 3.2726,
       "step": 64600
     },
     {
-      "epoch": 6.958346787213433,
-      "grad_norm": 0.7723833322525024,
-      "learning_rate": 0.0001828574507057429,
-      "loss": 3.2879,
+      "epoch": 6.9703504043126685,
+      "grad_norm": 0.7502321600914001,
+      "learning_rate": 0.00018213707501349162,
+      "loss": 3.2865,
       "step": 64650
     },
     {
-      "epoch": 6.9637283392530405,
-      "grad_norm": 0.7792532444000244,
-      "learning_rate": 0.00018253420967568148,
-      "loss": 3.2591,
+      "epoch": 6.975741239892184,
+      "grad_norm": 0.7295121550559998,
+      "learning_rate": 0.0001818132757690232,
+      "loss": 3.2914,
       "step": 64700
     },
     {
-      "epoch": 6.969109891292649,
-      "grad_norm": 0.8019757270812988,
-      "learning_rate": 0.00018221096864562007,
-      "loss": 3.2834,
+      "epoch": 6.981132075471698,
+      "grad_norm": 0.8087372779846191,
+      "learning_rate": 0.00018148947652455474,
+      "loss": 3.2767,
       "step": 64750
     },
     {
-      "epoch": 6.974491443332257,
-      "grad_norm": 0.7884219884872437,
-      "learning_rate": 0.00018188772761555867,
-      "loss": 3.2763,
+      "epoch": 6.986522911051213,
+      "grad_norm": 0.8098896145820618,
+      "learning_rate": 0.00018116567728008632,
+      "loss": 3.2865,
       "step": 64800
     },
     {
-      "epoch": 6.979872995371865,
-      "grad_norm": 0.7742852568626404,
-      "learning_rate": 0.00018156448658549723,
-      "loss": 3.2954,
+      "epoch": 6.991913746630727,
+      "grad_norm": 0.7889067530632019,
+      "learning_rate": 0.0001808418780356179,
+      "loss": 3.2567,
       "step": 64850
     },
     {
-      "epoch": 6.985254547411474,
-      "grad_norm": 0.7695539593696594,
-      "learning_rate": 0.00018124124555543583,
-      "loss": 3.2884,
+      "epoch": 6.997304582210242,
+      "grad_norm": 0.7330875992774963,
+      "learning_rate": 0.00018051807879114948,
+      "loss": 3.3053,
       "step": 64900
     },
     {
-      "epoch": 6.990636099451081,
-      "grad_norm": 0.8002060055732727,
-      "learning_rate": 0.0001809180045253744,
-      "loss": 3.2998,
+      "epoch": 7.002695417789758,
+      "grad_norm": 0.7592718601226807,
+      "learning_rate": 0.00018019427954668105,
+      "loss": 3.2519,
       "step": 64950
     },
     {
-      "epoch": 6.99601765149069,
-      "grad_norm": 0.7580960392951965,
-      "learning_rate": 0.00018059476349531296,
-      "loss": 3.2856,
+      "epoch": 7.008086253369272,
+      "grad_norm": 0.7767481803894043,
+      "learning_rate": 0.00017987048030221263,
+      "loss": 3.203,
       "step": 65000
     },
     {
-      "epoch": 6.99601765149069,
-      "eval_accuracy": 0.3874277308898313,
-      "eval_loss": 3.353395938873291,
-      "eval_runtime": 180.2586,
-      "eval_samples_per_second": 99.918,
-      "eval_steps_per_second": 6.247,
+      "epoch": 7.008086253369272,
+      "eval_accuracy": 0.38746945359875207,
+      "eval_loss": 3.35638165473938,
+      "eval_runtime": 184.1318,
+      "eval_samples_per_second": 97.816,
+      "eval_steps_per_second": 6.115,
       "step": 65000
     },
     {
-      "epoch": 7.0013992035302985,
-      "grad_norm": 0.7535995244979858,
-      "learning_rate": 0.00018027152246525159,
-      "loss": 3.2655,
+      "epoch": 7.013477088948787,
+      "grad_norm": 0.7944175601005554,
+      "learning_rate": 0.00017954668105774416,
+      "loss": 3.206,
       "step": 65050
     },
     {
-      "epoch": 7.006780755569906,
-      "grad_norm": 0.7798009514808655,
-      "learning_rate": 0.00017994828143519015,
-      "loss": 3.2142,
+      "epoch": 7.018867924528302,
+      "grad_norm": 0.7448245882987976,
+      "learning_rate": 0.00017922288181327573,
+      "loss": 3.2075,
       "step": 65100
     },
     {
-      "epoch": 7.012162307609515,
-      "grad_norm": 0.7834371328353882,
-      "learning_rate": 0.00017962504040512872,
-      "loss": 3.1959,
+      "epoch": 7.024258760107816,
+      "grad_norm": 0.7875297665596008,
+      "learning_rate": 0.0001788990825688073,
+      "loss": 3.211,
       "step": 65150
     },
     {
-      "epoch": 7.017543859649122,
-      "grad_norm": 0.7991567850112915,
-      "learning_rate": 0.00017930179937506734,
-      "loss": 3.1903,
+      "epoch": 7.0296495956873315,
+      "grad_norm": 0.7365028858184814,
+      "learning_rate": 0.0001785752833243389,
+      "loss": 3.2182,
       "step": 65200
     },
     {
-      "epoch": 7.022925411688731,
-      "grad_norm": 0.8215781450271606,
-      "learning_rate": 0.0001789785583450059,
-      "loss": 3.2197,
+      "epoch": 7.035040431266847,
+      "grad_norm": 0.7864233255386353,
+      "learning_rate": 0.00017825148407987047,
+      "loss": 3.1938,
       "step": 65250
     },
     {
-      "epoch": 7.0283069637283395,
-      "grad_norm": 0.7673898935317993,
-      "learning_rate": 0.00017865531731494448,
-      "loss": 3.198,
+      "epoch": 7.040431266846361,
+      "grad_norm": 0.7868351340293884,
+      "learning_rate": 0.00017792768483540204,
+      "loss": 3.2094,
       "step": 65300
     },
     {
-      "epoch": 7.033688515767947,
-      "grad_norm": 0.7570087313652039,
-      "learning_rate": 0.00017833207628488307,
-      "loss": 3.1956,
+      "epoch": 7.045822102425876,
+      "grad_norm": 0.7268515825271606,
+      "learning_rate": 0.00017760388559093362,
+      "loss": 3.2043,
       "step": 65350
     },
     {
-      "epoch": 7.039070067807556,
-      "grad_norm": 0.7777502536773682,
-      "learning_rate": 0.00017800883525482167,
-      "loss": 3.2017,
+      "epoch": 7.051212938005391,
+      "grad_norm": 0.748391330242157,
+      "learning_rate": 0.0001772800863464652,
+      "loss": 3.2034,
       "step": 65400
     },
     {
-      "epoch": 7.044451619847164,
-      "grad_norm": 0.8165860176086426,
-      "learning_rate": 0.00017768559422476026,
-      "loss": 3.207,
+      "epoch": 7.056603773584905,
+      "grad_norm": 0.7858370542526245,
+      "learning_rate": 0.00017695628710199672,
+      "loss": 3.1939,
       "step": 65450
     },
     {
-      "epoch": 7.049833171886772,
-      "grad_norm": 0.8510911464691162,
-      "learning_rate": 0.00017736235319469883,
-      "loss": 3.2091,
+      "epoch": 7.061994609164421,
+      "grad_norm": 0.7558709383010864,
+      "learning_rate": 0.0001766324878575283,
+      "loss": 3.207,
       "step": 65500
     },
     {
-      "epoch": 7.0552147239263805,
-      "grad_norm": 0.7689694762229919,
-      "learning_rate": 0.0001770391121646374,
-      "loss": 3.2019,
+      "epoch": 7.067385444743936,
+      "grad_norm": 0.7803695797920227,
+      "learning_rate": 0.00017630868861305988,
+      "loss": 3.1919,
       "step": 65550
     },
     {
-      "epoch": 7.060596275965988,
-      "grad_norm": 0.7681092023849487,
-      "learning_rate": 0.00017671587113457602,
-      "loss": 3.2155,
+      "epoch": 7.07277628032345,
+      "grad_norm": 0.7843615412712097,
+      "learning_rate": 0.00017598488936859146,
+      "loss": 3.1857,
       "step": 65600
     },
     {
-      "epoch": 7.065977828005597,
-      "grad_norm": 0.7637883424758911,
-      "learning_rate": 0.0001763926301045146,
-      "loss": 3.2149,
+      "epoch": 7.078167115902965,
+      "grad_norm": 0.7768192291259766,
+      "learning_rate": 0.00017566109012412303,
+      "loss": 3.2036,
       "step": 65650
     },
     {
-      "epoch": 7.071359380045205,
-      "grad_norm": 0.810635507106781,
-      "learning_rate": 0.00017606938907445315,
-      "loss": 3.2153,
+      "epoch": 7.083557951482479,
+      "grad_norm": 0.7211740016937256,
+      "learning_rate": 0.0001753372908796546,
+      "loss": 3.2304,
       "step": 65700
     },
     {
-      "epoch": 7.076740932084813,
-      "grad_norm": 0.7918748259544373,
-      "learning_rate": 0.00017574614804439178,
-      "loss": 3.2095,
+      "epoch": 7.0889487870619945,
+      "grad_norm": 0.785815417766571,
+      "learning_rate": 0.0001750134916351862,
+      "loss": 3.2147,
       "step": 65750
     },
     {
-      "epoch": 7.0821224841244215,
-      "grad_norm": 0.8016531467437744,
-      "learning_rate": 0.00017542290701433034,
-      "loss": 3.211,
+      "epoch": 7.09433962264151,
+      "grad_norm": 0.7704727649688721,
+      "learning_rate": 0.00017468969239071774,
+      "loss": 3.2049,
       "step": 65800
     },
     {
-      "epoch": 7.08750403616403,
-      "grad_norm": 0.8359585404396057,
-      "learning_rate": 0.0001750996659842689,
-      "loss": 3.2128,
+      "epoch": 7.099730458221024,
+      "grad_norm": 0.7499077916145325,
+      "learning_rate": 0.00017436589314624932,
+      "loss": 3.2178,
       "step": 65850
     },
     {
-      "epoch": 7.092885588203638,
-      "grad_norm": 0.794703483581543,
-      "learning_rate": 0.0001747764249542075,
-      "loss": 3.1945,
+      "epoch": 7.105121293800539,
+      "grad_norm": 0.7801557779312134,
+      "learning_rate": 0.00017404209390178087,
+      "loss": 3.2287,
       "step": 65900
     },
     {
-      "epoch": 7.098267140243246,
-      "grad_norm": 0.7841687202453613,
-      "learning_rate": 0.0001744531839241461,
-      "loss": 3.2197,
+      "epoch": 7.110512129380054,
+      "grad_norm": 0.7626088857650757,
+      "learning_rate": 0.00017371829465731244,
+      "loss": 3.2246,
       "step": 65950
     },
     {
-      "epoch": 7.103648692282855,
-      "grad_norm": 0.7183166742324829,
-      "learning_rate": 0.00017412994289408467,
-      "loss": 3.2254,
+      "epoch": 7.115902964959568,
+      "grad_norm": 0.7502177357673645,
+      "learning_rate": 0.00017339449541284402,
+      "loss": 3.1971,
       "step": 66000
     },
     {
-      "epoch": 7.103648692282855,
-      "eval_accuracy": 0.3870331036012891,
-      "eval_loss": 3.3611578941345215,
-      "eval_runtime": 179.9703,
-      "eval_samples_per_second": 100.078,
-      "eval_steps_per_second": 6.257,
+      "epoch": 7.115902964959568,
+      "eval_accuracy": 0.387290284986746,
+      "eval_loss": 3.3586626052856445,
+      "eval_runtime": 185.0266,
+      "eval_samples_per_second": 97.343,
+      "eval_steps_per_second": 6.086,
       "step": 66000
     },
     {
-      "epoch": 7.109030244322462,
-      "grad_norm": 0.7794191837310791,
-      "learning_rate": 0.00017380670186402326,
-      "loss": 3.223,
+      "epoch": 7.121293800539084,
+      "grad_norm": 0.8061758875846863,
+      "learning_rate": 0.0001730706961683756,
+      "loss": 3.2485,
       "step": 66050
     },
     {
-      "epoch": 7.114411796362071,
-      "grad_norm": 0.8234367966651917,
-      "learning_rate": 0.00017348346083396183,
-      "loss": 3.1997,
+      "epoch": 7.126684636118599,
+      "grad_norm": 0.9009227156639099,
+      "learning_rate": 0.00017274689692390715,
+      "loss": 3.2043,
       "step": 66100
     },
     {
-      "epoch": 7.119793348401679,
-      "grad_norm": 0.8090280890464783,
-      "learning_rate": 0.00017316021980390042,
-      "loss": 3.2169,
+      "epoch": 7.132075471698113,
+      "grad_norm": 0.8046843409538269,
+      "learning_rate": 0.00017242309767943873,
+      "loss": 3.2085,
       "step": 66150
     },
     {
-      "epoch": 7.125174900441287,
-      "grad_norm": 0.8259748816490173,
-      "learning_rate": 0.00017283697877383902,
-      "loss": 3.2235,
+      "epoch": 7.137466307277628,
+      "grad_norm": 0.796795129776001,
+      "learning_rate": 0.0001720992984349703,
+      "loss": 3.2067,
       "step": 66200
     },
     {
-      "epoch": 7.130556452480896,
-      "grad_norm": 0.8138856887817383,
-      "learning_rate": 0.0001725137377437776,
-      "loss": 3.2307,
+      "epoch": 7.142857142857143,
+      "grad_norm": 0.8700855374336243,
+      "learning_rate": 0.00017177549919050188,
+      "loss": 3.2225,
       "step": 66250
     },
     {
-      "epoch": 7.135938004520503,
-      "grad_norm": 0.8415960669517517,
-      "learning_rate": 0.00017219049671371615,
-      "loss": 3.218,
+      "epoch": 7.1482479784366575,
+      "grad_norm": 0.7229040861129761,
+      "learning_rate": 0.00017145169994603346,
+      "loss": 3.2265,
       "step": 66300
     },
     {
-      "epoch": 7.141319556560112,
-      "grad_norm": 0.8021517992019653,
-      "learning_rate": 0.00017186725568365478,
-      "loss": 3.2237,
+      "epoch": 7.153638814016173,
+      "grad_norm": 0.7846745252609253,
+      "learning_rate": 0.000171127900701565,
+      "loss": 3.2216,
       "step": 66350
     },
     {
-      "epoch": 7.1467011085997205,
-      "grad_norm": 0.7997321486473083,
-      "learning_rate": 0.00017154401465359334,
-      "loss": 3.2073,
+      "epoch": 7.159029649595688,
+      "grad_norm": 0.7371578216552734,
+      "learning_rate": 0.0001708041014570966,
+      "loss": 3.2311,
       "step": 66400
     },
     {
-      "epoch": 7.152082660639328,
-      "grad_norm": 0.7626723647117615,
-      "learning_rate": 0.00017122077362353194,
-      "loss": 3.2268,
+      "epoch": 7.164420485175202,
+      "grad_norm": 0.7651979327201843,
+      "learning_rate": 0.00017048030221262814,
+      "loss": 3.2266,
       "step": 66450
     },
     {
-      "epoch": 7.157464212678937,
-      "grad_norm": 0.8034189939498901,
-      "learning_rate": 0.00017089753259347053,
-      "loss": 3.2221,
+      "epoch": 7.169811320754717,
+      "grad_norm": 0.7952542901039124,
+      "learning_rate": 0.00017015650296815972,
+      "loss": 3.2126,
       "step": 66500
     },
     {
-      "epoch": 7.162845764718545,
-      "grad_norm": 0.8207605481147766,
-      "learning_rate": 0.0001705742915634091,
-      "loss": 3.2177,
+      "epoch": 7.175202156334231,
+      "grad_norm": 0.8030452132225037,
+      "learning_rate": 0.0001698327037236913,
+      "loss": 3.2133,
       "step": 66550
     },
     {
-      "epoch": 7.168227316758153,
-      "grad_norm": 0.8149417042732239,
-      "learning_rate": 0.0001702510505333477,
-      "loss": 3.221,
+      "epoch": 7.180592991913747,
+      "grad_norm": 0.7407306432723999,
+      "learning_rate": 0.00016950890447922287,
+      "loss": 3.2166,
       "step": 66600
     },
     {
-      "epoch": 7.1736088687977615,
-      "grad_norm": 0.8007292151451111,
-      "learning_rate": 0.00016992780950328626,
-      "loss": 3.2226,
+      "epoch": 7.185983827493262,
+      "grad_norm": 0.7781436443328857,
+      "learning_rate": 0.00016918510523475445,
+      "loss": 3.2206,
       "step": 66650
     },
     {
-      "epoch": 7.178990420837369,
-      "grad_norm": 0.7681205868721008,
-      "learning_rate": 0.00016960456847322486,
-      "loss": 3.2232,
+      "epoch": 7.191374663072776,
+      "grad_norm": 0.793569028377533,
+      "learning_rate": 0.00016886130599028603,
+      "loss": 3.2167,
       "step": 66700
     },
     {
-      "epoch": 7.184371972876978,
-      "grad_norm": 0.785612940788269,
-      "learning_rate": 0.00016928132744316345,
-      "loss": 3.2141,
+      "epoch": 7.196765498652291,
+      "grad_norm": 0.7684223055839539,
+      "learning_rate": 0.00016853750674581755,
+      "loss": 3.2094,
       "step": 66750
     },
     {
-      "epoch": 7.189753524916586,
-      "grad_norm": 0.7706198692321777,
-      "learning_rate": 0.00016895808641310202,
-      "loss": 3.2163,
+      "epoch": 7.202156334231806,
+      "grad_norm": 0.8211696743965149,
+      "learning_rate": 0.00016821370750134913,
+      "loss": 3.229,
       "step": 66800
     },
     {
-      "epoch": 7.195135076956194,
-      "grad_norm": 0.8889934420585632,
-      "learning_rate": 0.0001686348453830406,
-      "loss": 3.2103,
+      "epoch": 7.2075471698113205,
+      "grad_norm": 0.8771004676818848,
+      "learning_rate": 0.0001678899082568807,
+      "loss": 3.2348,
       "step": 66850
     },
     {
-      "epoch": 7.2005166289958025,
-      "grad_norm": 0.7743992209434509,
-      "learning_rate": 0.0001683116043529792,
-      "loss": 3.2334,
+      "epoch": 7.212938005390836,
+      "grad_norm": 0.7875045537948608,
+      "learning_rate": 0.00016756610901241228,
+      "loss": 3.2158,
       "step": 66900
     },
     {
-      "epoch": 7.205898181035411,
-      "grad_norm": 0.7609925866127014,
-      "learning_rate": 0.00016798836332291778,
-      "loss": 3.2293,
+      "epoch": 7.218328840970351,
+      "grad_norm": 0.753778338432312,
+      "learning_rate": 0.00016724230976794386,
+      "loss": 3.2356,
       "step": 66950
     },
     {
-      "epoch": 7.211279733075019,
-      "grad_norm": 0.778520941734314,
-      "learning_rate": 0.00016766512229285634,
-      "loss": 3.2413,
+      "epoch": 7.223719676549865,
+      "grad_norm": 0.7390396595001221,
+      "learning_rate": 0.00016691851052347544,
+      "loss": 3.2228,
       "step": 67000
     },
     {
-      "epoch": 7.211279733075019,
-      "eval_accuracy": 0.38746108732639034,
-      "eval_loss": 3.359044075012207,
-      "eval_runtime": 180.1842,
-      "eval_samples_per_second": 99.959,
-      "eval_steps_per_second": 6.249,
+      "epoch": 7.223719676549865,
+      "eval_accuracy": 0.3879755587501919,
+      "eval_loss": 3.3560919761657715,
+      "eval_runtime": 184.1802,
+      "eval_samples_per_second": 97.79,
+      "eval_steps_per_second": 6.114,
       "step": 67000
     },
     {
-      "epoch": 7.216661285114627,
-      "grad_norm": 0.7969582080841064,
-      "learning_rate": 0.00016734188126279494,
-      "loss": 3.2454,
+      "epoch": 7.22911051212938,
+      "grad_norm": 0.7987199425697327,
+      "learning_rate": 0.00016659471127900702,
+      "loss": 3.2481,
       "step": 67050
     },
     {
-      "epoch": 7.222042837154235,
-      "grad_norm": 0.7877959609031677,
-      "learning_rate": 0.00016701864023273353,
-      "loss": 3.2199,
+      "epoch": 7.234501347708895,
+      "grad_norm": 0.7510315775871277,
+      "learning_rate": 0.0001662709120345386,
+      "loss": 3.2235,
       "step": 67100
     },
     {
-      "epoch": 7.2274243891938434,
-      "grad_norm": 0.8033312559127808,
-      "learning_rate": 0.0001666953992026721,
-      "loss": 3.2453,
+      "epoch": 7.2398921832884096,
+      "grad_norm": 0.7914296388626099,
+      "learning_rate": 0.00016594711279007015,
+      "loss": 3.2358,
       "step": 67150
     },
     {
-      "epoch": 7.232805941233452,
-      "grad_norm": 0.775666356086731,
-      "learning_rate": 0.0001663721581726107,
-      "loss": 3.2301,
+      "epoch": 7.245283018867925,
+      "grad_norm": 0.7854893803596497,
+      "learning_rate": 0.0001656233135456017,
+      "loss": 3.2215,
       "step": 67200
     },
     {
-      "epoch": 7.23818749327306,
-      "grad_norm": 0.851057767868042,
-      "learning_rate": 0.00016604891714254926,
-      "loss": 3.2248,
+      "epoch": 7.250673854447439,
+      "grad_norm": 0.7990824580192566,
+      "learning_rate": 0.00016529951430113327,
+      "loss": 3.2293,
       "step": 67250
     },
     {
-      "epoch": 7.243569045312668,
-      "grad_norm": 0.8318371176719666,
-      "learning_rate": 0.00016572567611248786,
-      "loss": 3.2132,
+      "epoch": 7.256064690026954,
+      "grad_norm": 0.7979656457901001,
+      "learning_rate": 0.00016497571505666485,
+      "loss": 3.2286,
       "step": 67300
     },
     {
-      "epoch": 7.248950597352277,
-      "grad_norm": 0.817715585231781,
-      "learning_rate": 0.00016540243508242645,
-      "loss": 3.238,
+      "epoch": 7.261455525606469,
+      "grad_norm": 0.7852445244789124,
+      "learning_rate": 0.00016465191581219643,
+      "loss": 3.2231,
       "step": 67350
     },
     {
-      "epoch": 7.254332149391884,
-      "grad_norm": 0.7780318856239319,
-      "learning_rate": 0.00016507919405236502,
-      "loss": 3.2238,
+      "epoch": 7.2668463611859835,
+      "grad_norm": 0.764211118221283,
+      "learning_rate": 0.000164328116567728,
+      "loss": 3.2285,
       "step": 67400
     },
     {
-      "epoch": 7.259713701431493,
-      "grad_norm": 0.7914072275161743,
-      "learning_rate": 0.00016475595302230364,
-      "loss": 3.2442,
+      "epoch": 7.272237196765499,
+      "grad_norm": 0.8192458748817444,
+      "learning_rate": 0.00016400431732325956,
+      "loss": 3.2231,
       "step": 67450
     },
     {
-      "epoch": 7.265095253471101,
-      "grad_norm": 0.7719249129295349,
-      "learning_rate": 0.0001644327119922422,
-      "loss": 3.2172,
+      "epoch": 7.277628032345014,
+      "grad_norm": 0.8220451474189758,
+      "learning_rate": 0.00016368051807879114,
+      "loss": 3.236,
       "step": 67500
     },
     {
-      "epoch": 7.270476805510709,
-      "grad_norm": 0.7739912271499634,
-      "learning_rate": 0.00016410947096218078,
-      "loss": 3.2197,
+      "epoch": 7.283018867924528,
+      "grad_norm": 0.7826405167579651,
+      "learning_rate": 0.0001633567188343227,
+      "loss": 3.2382,
       "step": 67550
     },
     {
-      "epoch": 7.275858357550318,
-      "grad_norm": 0.8210995197296143,
-      "learning_rate": 0.00016378622993211937,
-      "loss": 3.228,
+      "epoch": 7.288409703504043,
+      "grad_norm": 0.7415540218353271,
+      "learning_rate": 0.00016303291958985426,
+      "loss": 3.2105,
       "step": 67600
     },
     {
-      "epoch": 7.281239909589925,
-      "grad_norm": 0.8359017968177795,
-      "learning_rate": 0.00016346298890205797,
-      "loss": 3.2404,
+      "epoch": 7.293800539083558,
+      "grad_norm": 0.76246577501297,
+      "learning_rate": 0.00016270912034538584,
+      "loss": 3.2323,
       "step": 67650
     },
     {
-      "epoch": 7.286621461629534,
-      "grad_norm": 0.8332840204238892,
-      "learning_rate": 0.00016313974787199653,
-      "loss": 3.2456,
+      "epoch": 7.2991913746630726,
+      "grad_norm": 0.7758094072341919,
+      "learning_rate": 0.00016238532110091742,
+      "loss": 3.22,
       "step": 67700
     },
     {
-      "epoch": 7.2920030136691425,
-      "grad_norm": 0.8179646730422974,
-      "learning_rate": 0.00016281650684193513,
-      "loss": 3.2327,
+      "epoch": 7.304582210242588,
+      "grad_norm": 0.7752118706703186,
+      "learning_rate": 0.000162061521856449,
+      "loss": 3.2314,
       "step": 67750
     },
     {
-      "epoch": 7.29738456570875,
-      "grad_norm": 0.7919309139251709,
-      "learning_rate": 0.0001624932658118737,
-      "loss": 3.2263,
+      "epoch": 7.309973045822103,
+      "grad_norm": 0.7858720421791077,
+      "learning_rate": 0.00016173772261198055,
+      "loss": 3.221,
       "step": 67800
     },
     {
-      "epoch": 7.302766117748359,
-      "grad_norm": 0.8252838850021362,
-      "learning_rate": 0.0001621700247818123,
-      "loss": 3.2371,
+      "epoch": 7.315363881401617,
+      "grad_norm": 0.7536090016365051,
+      "learning_rate": 0.00016141392336751212,
+      "loss": 3.2278,
       "step": 67850
     },
     {
-      "epoch": 7.308147669787967,
-      "grad_norm": 0.7692036032676697,
-      "learning_rate": 0.00016184678375175089,
-      "loss": 3.2357,
+      "epoch": 7.320754716981132,
+      "grad_norm": 0.7631314396858215,
+      "learning_rate": 0.0001610901241230437,
+      "loss": 3.2167,
       "step": 67900
     },
     {
-      "epoch": 7.313529221827575,
-      "grad_norm": 0.8547636866569519,
-      "learning_rate": 0.00016153000754229067,
-      "loss": 3.2201,
+      "epoch": 7.3261455525606465,
+      "grad_norm": 0.794683575630188,
+      "learning_rate": 0.00016076632487857528,
+      "loss": 3.2182,
       "step": 67950
     },
     {
-      "epoch": 7.3189107738671835,
-      "grad_norm": 0.8004331588745117,
-      "learning_rate": 0.0001612067665122293,
-      "loss": 3.2095,
+      "epoch": 7.331536388140162,
+      "grad_norm": 0.8233695030212402,
+      "learning_rate": 0.00016044252563410686,
+      "loss": 3.224,
       "step": 68000
     },
     {
-      "epoch": 7.3189107738671835,
-      "eval_accuracy": 0.38778693733694597,
-      "eval_loss": 3.356480598449707,
-      "eval_runtime": 180.04,
-      "eval_samples_per_second": 100.039,
-      "eval_steps_per_second": 6.254,
+      "epoch": 7.331536388140162,
+      "eval_accuracy": 0.38810865853776466,
+      "eval_loss": 3.3514058589935303,
+      "eval_runtime": 184.0263,
+      "eval_samples_per_second": 97.872,
+      "eval_steps_per_second": 6.119,
       "step": 68000
     },
     {
-      "epoch": 7.324292325906791,
-      "grad_norm": 0.811409592628479,
-      "learning_rate": 0.00016088352548216785,
-      "loss": 3.2466,
+      "epoch": 7.336927223719677,
+      "grad_norm": 0.8384366631507874,
+      "learning_rate": 0.0001601187263896384,
+      "loss": 3.2295,
       "step": 68050
     },
     {
-      "epoch": 7.3296738779464,
-      "grad_norm": 0.8523727655410767,
-      "learning_rate": 0.00016056028445210642,
-      "loss": 3.2269,
+      "epoch": 7.342318059299191,
+      "grad_norm": 0.7813538908958435,
+      "learning_rate": 0.00015979492714516996,
+      "loss": 3.258,
       "step": 68100
     },
     {
-      "epoch": 7.335055429986008,
-      "grad_norm": 0.809126615524292,
-      "learning_rate": 0.00016023704342204504,
-      "loss": 3.2503,
+      "epoch": 7.347708894878706,
+      "grad_norm": 0.84609055519104,
+      "learning_rate": 0.00015947112790070154,
+      "loss": 3.2215,
       "step": 68150
     },
     {
-      "epoch": 7.340436982025616,
-      "grad_norm": 0.8085951209068298,
-      "learning_rate": 0.0001599138023919836,
-      "loss": 3.2459,
+      "epoch": 7.353099730458221,
+      "grad_norm": 0.7973068356513977,
+      "learning_rate": 0.00015914732865623311,
+      "loss": 3.2199,
       "step": 68200
     },
     {
-      "epoch": 7.3458185340652244,
-      "grad_norm": 0.7844497561454773,
-      "learning_rate": 0.0001595905613619222,
-      "loss": 3.236,
+      "epoch": 7.3584905660377355,
+      "grad_norm": 0.8102887868881226,
+      "learning_rate": 0.0001588235294117647,
+      "loss": 3.2468,
       "step": 68250
     },
     {
-      "epoch": 7.351200086104833,
-      "grad_norm": 0.8168286681175232,
-      "learning_rate": 0.00015926732033186077,
-      "loss": 3.2236,
+      "epoch": 7.363881401617251,
+      "grad_norm": 0.7900605201721191,
+      "learning_rate": 0.00015849973016729627,
+      "loss": 3.2259,
       "step": 68300
     },
     {
-      "epoch": 7.356581638144441,
-      "grad_norm": 0.8217577338218689,
-      "learning_rate": 0.00015894407930179934,
-      "loss": 3.226,
+      "epoch": 7.369272237196766,
+      "grad_norm": 0.7393832802772522,
+      "learning_rate": 0.00015817593092282785,
+      "loss": 3.2357,
       "step": 68350
     },
     {
-      "epoch": 7.361963190184049,
-      "grad_norm": 0.8433511853218079,
-      "learning_rate": 0.00015862083827173796,
-      "loss": 3.2294,
+      "epoch": 7.37466307277628,
+      "grad_norm": 0.8080168962478638,
+      "learning_rate": 0.00015785860766324877,
+      "loss": 3.2379,
       "step": 68400
     },
     {
-      "epoch": 7.367344742223658,
-      "grad_norm": 0.8781927227973938,
-      "learning_rate": 0.00015829759724167653,
-      "loss": 3.2264,
+      "epoch": 7.380053908355795,
+      "grad_norm": 0.7917481064796448,
+      "learning_rate": 0.00015753480841878032,
+      "loss": 3.2221,
       "step": 68450
     },
     {
-      "epoch": 7.372726294263265,
-      "grad_norm": 0.8104392290115356,
-      "learning_rate": 0.0001579743562116151,
-      "loss": 3.2114,
+      "epoch": 7.38544474393531,
+      "grad_norm": 0.7612196803092957,
+      "learning_rate": 0.0001572110091743119,
+      "loss": 3.209,
       "step": 68500
     },
     {
-      "epoch": 7.378107846302874,
-      "grad_norm": 0.8192629218101501,
-      "learning_rate": 0.00015765111518155372,
-      "loss": 3.2297,
+      "epoch": 7.390835579514825,
+      "grad_norm": 0.7824956774711609,
+      "learning_rate": 0.00015688720992984348,
+      "loss": 3.2334,
       "step": 68550
     },
     {
-      "epoch": 7.383489398342482,
-      "grad_norm": 0.7943025231361389,
-      "learning_rate": 0.0001573278741514923,
-      "loss": 3.2313,
+      "epoch": 7.39622641509434,
+      "grad_norm": 0.7742493748664856,
+      "learning_rate": 0.00015656341068537506,
+      "loss": 3.2152,
       "step": 68600
     },
     {
-      "epoch": 7.38887095038209,
-      "grad_norm": 0.858521580696106,
-      "learning_rate": 0.00015700463312143085,
-      "loss": 3.2268,
+      "epoch": 7.401617250673855,
+      "grad_norm": 0.8115978837013245,
+      "learning_rate": 0.00015623961144090663,
+      "loss": 3.2374,
       "step": 68650
     },
     {
-      "epoch": 7.394252502421699,
-      "grad_norm": 0.804550290107727,
-      "learning_rate": 0.00015668139209136945,
-      "loss": 3.2131,
+      "epoch": 7.407008086253369,
+      "grad_norm": 0.7645772099494934,
+      "learning_rate": 0.0001559158121964382,
+      "loss": 3.2132,
       "step": 68700
     },
     {
-      "epoch": 7.399634054461306,
-      "grad_norm": 0.8339864611625671,
-      "learning_rate": 0.00015635815106130804,
-      "loss": 3.2387,
+      "epoch": 7.412398921832884,
+      "grad_norm": 0.7878434062004089,
+      "learning_rate": 0.0001555920129519698,
+      "loss": 3.2236,
       "step": 68750
     },
     {
-      "epoch": 7.405015606500915,
-      "grad_norm": 0.7839356064796448,
-      "learning_rate": 0.0001560349100312466,
-      "loss": 3.2269,
+      "epoch": 7.4177897574123985,
+      "grad_norm": 0.7802878618240356,
+      "learning_rate": 0.0001552682137075013,
+      "loss": 3.2411,
       "step": 68800
     },
     {
-      "epoch": 7.4103971585405235,
-      "grad_norm": 0.8349343538284302,
-      "learning_rate": 0.0001557116690011852,
-      "loss": 3.2374,
+      "epoch": 7.423180592991914,
+      "grad_norm": 0.7774155735969543,
+      "learning_rate": 0.0001549444144630329,
+      "loss": 3.2333,
       "step": 68850
     },
     {
-      "epoch": 7.415778710580131,
-      "grad_norm": 0.8796477317810059,
-      "learning_rate": 0.00015538842797112377,
-      "loss": 3.2449,
+      "epoch": 7.428571428571429,
+      "grad_norm": 0.8356894850730896,
+      "learning_rate": 0.00015462061521856447,
+      "loss": 3.2405,
       "step": 68900
     },
     {
-      "epoch": 7.42116026261974,
-      "grad_norm": 0.8090130686759949,
-      "learning_rate": 0.00015506518694106237,
-      "loss": 3.2412,
+      "epoch": 7.433962264150943,
+      "grad_norm": 0.807794451713562,
+      "learning_rate": 0.00015429681597409605,
+      "loss": 3.2302,
       "step": 68950
     },
     {
-      "epoch": 7.426541814659347,
-      "grad_norm": 0.8464242219924927,
-      "learning_rate": 0.00015474194591100096,
-      "loss": 3.2314,
+      "epoch": 7.439353099730458,
+      "grad_norm": 0.798193633556366,
+      "learning_rate": 0.00015397301672962762,
+      "loss": 3.2216,
       "step": 69000
     },
     {
-      "epoch": 7.426541814659347,
-      "eval_accuracy": 0.38814625243694845,
-      "eval_loss": 3.3514676094055176,
-      "eval_runtime": 180.0963,
-      "eval_samples_per_second": 100.008,
-      "eval_steps_per_second": 6.252,
+      "epoch": 7.439353099730458,
+      "eval_accuracy": 0.3881358217597183,
+      "eval_loss": 3.347754716873169,
+      "eval_runtime": 183.9951,
+      "eval_samples_per_second": 97.888,
+      "eval_steps_per_second": 6.12,
       "step": 69000
     },
     {
-      "epoch": 7.431923366698956,
-      "grad_norm": 0.848106324672699,
-      "learning_rate": 0.00015441870488093953,
-      "loss": 3.2432,
+      "epoch": 7.444743935309973,
+      "grad_norm": 0.7326217889785767,
+      "learning_rate": 0.0001536492174851592,
+      "loss": 3.2141,
       "step": 69050
     },
     {
-      "epoch": 7.4373049187385645,
-      "grad_norm": 0.8055359125137329,
-      "learning_rate": 0.0001540954638508781,
-      "loss": 3.2373,
+      "epoch": 7.450134770889488,
+      "grad_norm": 0.7936413884162903,
+      "learning_rate": 0.00015332541824069078,
+      "loss": 3.2231,
       "step": 69100
     },
     {
-      "epoch": 7.442686470778172,
-      "grad_norm": 0.8940873742103577,
-      "learning_rate": 0.00015377222282081672,
-      "loss": 3.2389,
+      "epoch": 7.455525606469003,
+      "grad_norm": 0.9106899499893188,
+      "learning_rate": 0.00015300161899622233,
+      "loss": 3.2351,
       "step": 69150
     },
     {
-      "epoch": 7.448068022817781,
-      "grad_norm": 0.7894365787506104,
-      "learning_rate": 0.0001534489817907553,
-      "loss": 3.237,
+      "epoch": 7.460916442048518,
+      "grad_norm": 0.9161011576652527,
+      "learning_rate": 0.00015267781975175388,
+      "loss": 3.2488,
       "step": 69200
     },
     {
-      "epoch": 7.453449574857389,
-      "grad_norm": 0.8256045579910278,
-      "learning_rate": 0.00015312574076069388,
-      "loss": 3.2281,
+      "epoch": 7.466307277628032,
+      "grad_norm": 0.86126708984375,
+      "learning_rate": 0.00015235402050728546,
+      "loss": 3.2249,
       "step": 69250
     },
     {
-      "epoch": 7.458831126896997,
-      "grad_norm": 0.7994895577430725,
-      "learning_rate": 0.00015280249973063248,
-      "loss": 3.2525,
+      "epoch": 7.471698113207547,
+      "grad_norm": 0.8955468535423279,
+      "learning_rate": 0.00015203022126281703,
+      "loss": 3.2533,
       "step": 69300
     },
     {
-      "epoch": 7.4642126789366054,
-      "grad_norm": 0.8331586718559265,
-      "learning_rate": 0.00015247925870057104,
-      "loss": 3.2337,
+      "epoch": 7.4770889487870615,
+      "grad_norm": 0.8683763146400452,
+      "learning_rate": 0.0001517064220183486,
+      "loss": 3.2383,
       "step": 69350
     },
     {
-      "epoch": 7.469594230976213,
-      "grad_norm": 0.7736141085624695,
-      "learning_rate": 0.00015215601767050964,
-      "loss": 3.2536,
+      "epoch": 7.482479784366577,
+      "grad_norm": 0.8334479928016663,
+      "learning_rate": 0.0001513826227738802,
+      "loss": 3.2309,
       "step": 69400
     },
     {
-      "epoch": 7.474975783015822,
-      "grad_norm": 0.8336132764816284,
-      "learning_rate": 0.0001518327766404482,
-      "loss": 3.2434,
+      "epoch": 7.487870619946092,
+      "grad_norm": 0.818920910358429,
+      "learning_rate": 0.00015105882352941177,
+      "loss": 3.2317,
       "step": 69450
     },
     {
-      "epoch": 7.48035733505543,
-      "grad_norm": 0.8452787399291992,
-      "learning_rate": 0.0001515095356103868,
-      "loss": 3.2308,
+      "epoch": 7.493261455525606,
+      "grad_norm": 0.7930528521537781,
+      "learning_rate": 0.00015073502428494332,
+      "loss": 3.2453,
       "step": 69500
     },
     {
-      "epoch": 7.485738887095038,
-      "grad_norm": 0.7975885272026062,
-      "learning_rate": 0.0001511862945803254,
-      "loss": 3.2414,
+      "epoch": 7.498652291105121,
+      "grad_norm": 0.797651469707489,
+      "learning_rate": 0.0001504112250404749,
+      "loss": 3.2219,
       "step": 69550
     },
     {
-      "epoch": 7.491120439134646,
-      "grad_norm": 0.7722816467285156,
-      "learning_rate": 0.00015086305355026396,
-      "loss": 3.2329,
+      "epoch": 7.504043126684636,
+      "grad_norm": 0.796083927154541,
+      "learning_rate": 0.00015008742579600647,
+      "loss": 3.2312,
       "step": 69600
     },
     {
-      "epoch": 7.496501991174255,
-      "grad_norm": 0.7675669193267822,
-      "learning_rate": 0.00015053981252020253,
-      "loss": 3.2375,
+      "epoch": 7.509433962264151,
+      "grad_norm": 1.6214351654052734,
+      "learning_rate": 0.00014976362655153802,
+      "loss": 3.2476,
       "step": 69650
     },
     {
-      "epoch": 7.501883543213863,
-      "grad_norm": 0.7885386347770691,
-      "learning_rate": 0.00015021657149014115,
-      "loss": 3.2363,
+      "epoch": 7.514824797843666,
+      "grad_norm": 0.7855797410011292,
+      "learning_rate": 0.0001494398273070696,
+      "loss": 3.2404,
       "step": 69700
     },
     {
-      "epoch": 7.507265095253471,
-      "grad_norm": 0.792849063873291,
-      "learning_rate": 0.00014989333046007972,
-      "loss": 3.2484,
+      "epoch": 7.520215633423181,
+      "grad_norm": 0.78398597240448,
+      "learning_rate": 0.00014911602806260118,
+      "loss": 3.2591,
       "step": 69750
     },
     {
-      "epoch": 7.51264664729308,
-      "grad_norm": 0.837902843952179,
-      "learning_rate": 0.00014957008943001832,
-      "loss": 3.2335,
+      "epoch": 7.525606469002695,
+      "grad_norm": 0.7878542542457581,
+      "learning_rate": 0.00014879222881813273,
+      "loss": 3.2453,
       "step": 69800
     },
     {
-      "epoch": 7.518028199332687,
-      "grad_norm": 0.8501192331314087,
-      "learning_rate": 0.00014924684839995688,
-      "loss": 3.2218,
+      "epoch": 7.53099730458221,
+      "grad_norm": 0.766707718372345,
+      "learning_rate": 0.0001484684295736643,
+      "loss": 3.2236,
       "step": 69850
     },
     {
-      "epoch": 7.523409751372296,
-      "grad_norm": 0.8147408962249756,
-      "learning_rate": 0.00014892360736989548,
-      "loss": 3.2178,
+      "epoch": 7.536388140161725,
+      "grad_norm": 0.8185433745384216,
+      "learning_rate": 0.00014814463032919589,
+      "loss": 3.2149,
       "step": 69900
     },
     {
-      "epoch": 7.528791303411904,
-      "grad_norm": 0.7939187288284302,
-      "learning_rate": 0.00014860036633983407,
-      "loss": 3.2485,
+      "epoch": 7.54177897574124,
+      "grad_norm": 0.8230583071708679,
+      "learning_rate": 0.00014782083108472744,
+      "loss": 3.2498,
       "step": 69950
     },
     {
-      "epoch": 7.534172855451512,
-      "grad_norm": 0.846481204032898,
-      "learning_rate": 0.00014827712530977264,
-      "loss": 3.2351,
+      "epoch": 7.547169811320755,
+      "grad_norm": 0.8245810270309448,
+      "learning_rate": 0.00014749703184025901,
+      "loss": 3.2303,
       "step": 70000
     },
     {
-      "epoch": 7.534172855451512,
-      "eval_accuracy": 0.38860226860710584,
-      "eval_loss": 3.34824800491333,
-      "eval_runtime": 180.1854,
-      "eval_samples_per_second": 99.958,
-      "eval_steps_per_second": 6.249,
+      "epoch": 7.547169811320755,
+      "eval_accuracy": 0.3886090050861503,
+      "eval_loss": 3.3438832759857178,
+      "eval_runtime": 184.0678,
+      "eval_samples_per_second": 97.85,
+      "eval_steps_per_second": 6.117,
       "step": 70000
     },
     {
-      "epoch": 7.539554407491121,
-      "grad_norm": 0.822512686252594,
-      "learning_rate": 0.00014795388427971123,
-      "loss": 3.2308,
+      "epoch": 7.55256064690027,
+      "grad_norm": 0.7880018353462219,
+      "learning_rate": 0.0001471732325957906,
+      "loss": 3.2221,
       "step": 70050
     },
     {
-      "epoch": 7.544935959530728,
-      "grad_norm": 0.8041254878044128,
-      "learning_rate": 0.0001476306432496498,
-      "loss": 3.2348,
+      "epoch": 7.557951482479784,
+      "grad_norm": 0.7894161939620972,
+      "learning_rate": 0.00014684943335132217,
+      "loss": 3.239,
       "step": 70100
     },
     {
-      "epoch": 7.550317511570337,
-      "grad_norm": 0.8251069784164429,
-      "learning_rate": 0.0001473074022195884,
-      "loss": 3.2205,
+      "epoch": 7.563342318059299,
+      "grad_norm": 0.7724071741104126,
+      "learning_rate": 0.00014652563410685375,
+      "loss": 3.236,
       "step": 70150
     },
     {
-      "epoch": 7.5556990636099455,
-      "grad_norm": 0.7714356184005737,
-      "learning_rate": 0.00014698416118952696,
-      "loss": 3.2205,
+      "epoch": 7.568733153638814,
+      "grad_norm": 0.8679468631744385,
+      "learning_rate": 0.0001462018348623853,
+      "loss": 3.2306,
       "step": 70200
     },
     {
-      "epoch": 7.561080615649553,
-      "grad_norm": 0.8164318203926086,
-      "learning_rate": 0.00014666092015946556,
-      "loss": 3.2301,
+      "epoch": 7.574123989218329,
+      "grad_norm": 0.7930618524551392,
+      "learning_rate": 0.00014587803561791687,
+      "loss": 3.2411,
       "step": 70250
     },
     {
-      "epoch": 7.566462167689162,
-      "grad_norm": 0.8093104958534241,
-      "learning_rate": 0.00014633767912940415,
-      "loss": 3.2324,
+      "epoch": 7.579514824797844,
+      "grad_norm": 0.7628795504570007,
+      "learning_rate": 0.00014555423637344845,
+      "loss": 3.2313,
       "step": 70300
     },
     {
-      "epoch": 7.57184371972877,
-      "grad_norm": 0.822935938835144,
-      "learning_rate": 0.00014601443809934272,
-      "loss": 3.2378,
+      "epoch": 7.584905660377358,
+      "grad_norm": 0.822519838809967,
+      "learning_rate": 0.00014523043712898003,
+      "loss": 3.2355,
       "step": 70350
     },
     {
-      "epoch": 7.577225271768378,
-      "grad_norm": 0.8421322703361511,
-      "learning_rate": 0.00014569119706928132,
-      "loss": 3.2322,
+      "epoch": 7.590296495956873,
+      "grad_norm": 0.7862046360969543,
+      "learning_rate": 0.00014490663788451158,
+      "loss": 3.2449,
       "step": 70400
     },
     {
-      "epoch": 7.5826068238079865,
-      "grad_norm": 0.7766196727752686,
-      "learning_rate": 0.0001453679560392199,
-      "loss": 3.245,
+      "epoch": 7.595687331536388,
+      "grad_norm": 0.8416604399681091,
+      "learning_rate": 0.00014458283864004316,
+      "loss": 3.2476,
       "step": 70450
     },
     {
-      "epoch": 7.587988375847594,
-      "grad_norm": 0.8198171854019165,
-      "learning_rate": 0.00014504471500915848,
-      "loss": 3.2239,
+      "epoch": 7.601078167115903,
+      "grad_norm": 0.7710646986961365,
+      "learning_rate": 0.00014425903939557474,
+      "loss": 3.2381,
       "step": 70500
     },
     {
-      "epoch": 7.593369927887203,
-      "grad_norm": 0.7970831394195557,
-      "learning_rate": 0.00014472147397909707,
-      "loss": 3.2244,
+      "epoch": 7.606469002695418,
+      "grad_norm": 0.7889264225959778,
+      "learning_rate": 0.00014393524015110631,
+      "loss": 3.221,
       "step": 70550
     },
     {
-      "epoch": 7.598751479926811,
-      "grad_norm": 0.8099634051322937,
-      "learning_rate": 0.00014439823294903564,
-      "loss": 3.239,
+      "epoch": 7.611859838274933,
+      "grad_norm": 0.7603769898414612,
+      "learning_rate": 0.00014361144090663786,
+      "loss": 3.2406,
       "step": 70600
     },
     {
-      "epoch": 7.604133031966419,
-      "grad_norm": 0.7905451655387878,
-      "learning_rate": 0.00014407499191897423,
-      "loss": 3.2474,
+      "epoch": 7.617250673854447,
+      "grad_norm": 0.787106990814209,
+      "learning_rate": 0.00014328764166216944,
+      "loss": 3.2182,
       "step": 70650
     },
     {
-      "epoch": 7.609514584006027,
-      "grad_norm": 0.8306140303611755,
-      "learning_rate": 0.0001437517508889128,
-      "loss": 3.2549,
+      "epoch": 7.622641509433962,
+      "grad_norm": 0.8114652037620544,
+      "learning_rate": 0.00014296384241770102,
+      "loss": 3.2206,
       "step": 70700
     },
     {
-      "epoch": 7.614896136045635,
-      "grad_norm": 0.8547305464744568,
-      "learning_rate": 0.0001434285098588514,
-      "loss": 3.245,
+      "epoch": 7.628032345013477,
+      "grad_norm": 0.7583808898925781,
+      "learning_rate": 0.0001426400431732326,
+      "loss": 3.2369,
       "step": 70750
     },
     {
-      "epoch": 7.620277688085244,
-      "grad_norm": 0.8246287107467651,
-      "learning_rate": 0.00014311173364939123,
-      "loss": 3.236,
+      "epoch": 7.633423180592992,
+      "grad_norm": 0.786690890789032,
+      "learning_rate": 0.00014231624392876417,
+      "loss": 3.2429,
       "step": 70800
     },
     {
-      "epoch": 7.625659240124852,
-      "grad_norm": 0.7652173638343811,
-      "learning_rate": 0.0001427884926193298,
-      "loss": 3.2216,
+      "epoch": 7.638814016172507,
+      "grad_norm": 0.7588096261024475,
+      "learning_rate": 0.00014199244468429573,
+      "loss": 3.237,
       "step": 70850
     },
     {
-      "epoch": 7.63104079216446,
-      "grad_norm": 0.8844686150550842,
-      "learning_rate": 0.0001424652515892684,
-      "loss": 3.2323,
+      "epoch": 7.644204851752022,
+      "grad_norm": 0.7805950045585632,
+      "learning_rate": 0.0001416686454398273,
+      "loss": 3.2435,
       "step": 70900
     },
     {
-      "epoch": 7.636422344204068,
-      "grad_norm": 0.8048098087310791,
-      "learning_rate": 0.000142142010559207,
-      "loss": 3.2566,
+      "epoch": 7.649595687331536,
+      "grad_norm": 0.8513138890266418,
+      "learning_rate": 0.00014134484619535888,
+      "loss": 3.2353,
       "step": 70950
     },
     {
-      "epoch": 7.641803896243677,
-      "grad_norm": 0.7921103835105896,
-      "learning_rate": 0.00014181876952914555,
-      "loss": 3.2345,
+      "epoch": 7.654986522911051,
+      "grad_norm": 0.7487355470657349,
+      "learning_rate": 0.00014102104695089043,
+      "loss": 3.2245,
       "step": 71000
     },
     {
-      "epoch": 7.641803896243677,
-      "eval_accuracy": 0.3890942488831298,
-      "eval_loss": 3.343825101852417,
-      "eval_runtime": 180.1847,
-      "eval_samples_per_second": 99.959,
-      "eval_steps_per_second": 6.249,
+      "epoch": 7.654986522911051,
+      "eval_accuracy": 0.3890904460320563,
+      "eval_loss": 3.3406805992126465,
+      "eval_runtime": 184.1375,
+      "eval_samples_per_second": 97.813,
+      "eval_steps_per_second": 6.115,
       "step": 71000
     },
     {
-      "epoch": 7.647185448283285,
-      "grad_norm": 0.799433171749115,
-      "learning_rate": 0.00014149552849908415,
-      "loss": 3.2373,
+      "epoch": 7.660377358490566,
+      "grad_norm": 0.8516783714294434,
+      "learning_rate": 0.000140697247706422,
+      "loss": 3.2481,
       "step": 71050
     },
     {
-      "epoch": 7.652567000322893,
-      "grad_norm": 0.8794057369232178,
-      "learning_rate": 0.00014117228746902272,
-      "loss": 3.2458,
+      "epoch": 7.665768194070081,
+      "grad_norm": 0.7833296656608582,
+      "learning_rate": 0.0001403734484619536,
+      "loss": 3.2407,
       "step": 71100
     },
     {
-      "epoch": 7.657948552362502,
-      "grad_norm": 0.8924712538719177,
-      "learning_rate": 0.0001408490464389613,
-      "loss": 3.2278,
+      "epoch": 7.671159029649596,
+      "grad_norm": 0.7922981977462769,
+      "learning_rate": 0.00014004964921748514,
+      "loss": 3.2462,
       "step": 71150
     },
     {
-      "epoch": 7.663330104402109,
-      "grad_norm": 0.8285935521125793,
-      "learning_rate": 0.00014052580540889988,
-      "loss": 3.2317,
+      "epoch": 7.67654986522911,
+      "grad_norm": 0.760362446308136,
+      "learning_rate": 0.00013972584997301671,
+      "loss": 3.2291,
       "step": 71200
     },
     {
-      "epoch": 7.668711656441718,
-      "grad_norm": 0.8792213201522827,
-      "learning_rate": 0.00014020256437883847,
-      "loss": 3.2425,
+      "epoch": 7.681940700808625,
+      "grad_norm": 0.7816347479820251,
+      "learning_rate": 0.0001394020507285483,
+      "loss": 3.2246,
       "step": 71250
     },
     {
-      "epoch": 7.674093208481326,
-      "grad_norm": 0.8292412161827087,
-      "learning_rate": 0.00013987932334877707,
-      "loss": 3.2234,
+      "epoch": 7.6873315363881405,
+      "grad_norm": 0.8317768573760986,
+      "learning_rate": 0.00013907825148407984,
+      "loss": 3.2305,
       "step": 71300
     },
     {
-      "epoch": 7.679474760520934,
-      "grad_norm": 0.7959591150283813,
-      "learning_rate": 0.00013955608231871564,
-      "loss": 3.2474,
+      "epoch": 7.692722371967655,
+      "grad_norm": 0.8221088647842407,
+      "learning_rate": 0.00013875445223961142,
+      "loss": 3.2173,
       "step": 71350
     },
     {
-      "epoch": 7.684856312560543,
-      "grad_norm": 0.821670651435852,
-      "learning_rate": 0.00013923284128865423,
-      "loss": 3.2452,
+      "epoch": 7.69811320754717,
+      "grad_norm": 0.8052943348884583,
+      "learning_rate": 0.000138430652995143,
+      "loss": 3.2381,
       "step": 71400
     },
     {
-      "epoch": 7.69023786460015,
-      "grad_norm": 0.8048251271247864,
-      "learning_rate": 0.00013890960025859283,
-      "loss": 3.2417,
+      "epoch": 7.703504043126685,
+      "grad_norm": 0.8330031037330627,
+      "learning_rate": 0.00013810685375067455,
+      "loss": 3.2409,
       "step": 71450
     },
     {
-      "epoch": 7.695619416639759,
-      "grad_norm": 0.846571147441864,
-      "learning_rate": 0.0001385863592285314,
-      "loss": 3.2356,
+      "epoch": 7.708894878706199,
+      "grad_norm": 0.8075839281082153,
+      "learning_rate": 0.00013778305450620613,
+      "loss": 3.2238,
       "step": 71500
     },
     {
-      "epoch": 7.7010009686793675,
-      "grad_norm": 0.8968475461006165,
-      "learning_rate": 0.00013826311819847,
-      "loss": 3.2225,
+      "epoch": 7.714285714285714,
+      "grad_norm": 0.8027133345603943,
+      "learning_rate": 0.0001374592552617377,
+      "loss": 3.2536,
       "step": 71550
     },
     {
-      "epoch": 7.706382520718975,
-      "grad_norm": 0.8345704674720764,
-      "learning_rate": 0.00013793987716840858,
-      "loss": 3.2244,
+      "epoch": 7.719676549865229,
+      "grad_norm": 0.8382824063301086,
+      "learning_rate": 0.00013713545601726928,
+      "loss": 3.2313,
       "step": 71600
     },
     {
-      "epoch": 7.711764072758584,
-      "grad_norm": 0.7907124757766724,
-      "learning_rate": 0.00013761663613834715,
-      "loss": 3.2141,
+      "epoch": 7.725067385444744,
+      "grad_norm": 0.7851426005363464,
+      "learning_rate": 0.00013681165677280086,
+      "loss": 3.2151,
       "step": 71650
     },
     {
-      "epoch": 7.717145624798192,
-      "grad_norm": 0.8455408811569214,
-      "learning_rate": 0.00013729339510828572,
-      "loss": 3.2401,
+      "epoch": 7.730458221024259,
+      "grad_norm": 0.8292527198791504,
+      "learning_rate": 0.0001364878575283324,
+      "loss": 3.2434,
       "step": 71700
     },
     {
-      "epoch": 7.7225271768378,
-      "grad_norm": 0.8723199963569641,
-      "learning_rate": 0.0001369701540782243,
-      "loss": 3.241,
+      "epoch": 7.735849056603773,
+      "grad_norm": 0.8183705806732178,
+      "learning_rate": 0.000136164058283864,
+      "loss": 3.2415,
       "step": 71750
     },
     {
-      "epoch": 7.727908728877408,
-      "grad_norm": 0.8454627990722656,
-      "learning_rate": 0.0001366469130481629,
-      "loss": 3.2365,
+      "epoch": 7.741239892183288,
+      "grad_norm": 0.8626248240470886,
+      "learning_rate": 0.00013584025903939557,
+      "loss": 3.2449,
       "step": 71800
     },
     {
-      "epoch": 7.733290280917016,
-      "grad_norm": 0.8864803314208984,
-      "learning_rate": 0.00013632367201810147,
-      "loss": 3.2511,
+      "epoch": 7.7466307277628035,
+      "grad_norm": 0.8289384841918945,
+      "learning_rate": 0.00013551645979492714,
+      "loss": 3.2386,
       "step": 71850
     },
     {
-      "epoch": 7.738671832956625,
-      "grad_norm": 0.8611536026000977,
-      "learning_rate": 0.00013600043098804007,
-      "loss": 3.2308,
+      "epoch": 7.752021563342318,
+      "grad_norm": 0.8308370113372803,
+      "learning_rate": 0.0001351926605504587,
+      "loss": 3.2045,
       "step": 71900
     },
     {
-      "epoch": 7.744053384996233,
-      "grad_norm": 0.8165392875671387,
-      "learning_rate": 0.00013567718995797866,
-      "loss": 3.2312,
+      "epoch": 7.757412398921833,
+      "grad_norm": 0.7915014028549194,
+      "learning_rate": 0.00013486886130599027,
+      "loss": 3.2456,
       "step": 71950
     },
     {
-      "epoch": 7.749434937035841,
-      "grad_norm": 0.8212934732437134,
-      "learning_rate": 0.00013535394892791723,
-      "loss": 3.2414,
+      "epoch": 7.762803234501348,
+      "grad_norm": 0.83826744556427,
+      "learning_rate": 0.00013454506206152185,
+      "loss": 3.2293,
       "step": 72000
     },
     {
-      "epoch": 7.749434937035841,
-      "eval_accuracy": 0.38907805960284547,
-      "eval_loss": 3.3423426151275635,
-      "eval_runtime": 180.2997,
-      "eval_samples_per_second": 99.895,
-      "eval_steps_per_second": 6.245,
+      "epoch": 7.762803234501348,
+      "eval_accuracy": 0.38955678422655604,
+      "eval_loss": 3.3382842540740967,
+      "eval_runtime": 184.0568,
+      "eval_samples_per_second": 97.856,
+      "eval_steps_per_second": 6.118,
       "step": 72000
     },
     {
-      "epoch": 7.754816489075449,
-      "grad_norm": 0.7960253953933716,
-      "learning_rate": 0.00013503070789785583,
-      "loss": 3.2214,
+      "epoch": 7.768194070080862,
+      "grad_norm": 0.7984863519668579,
+      "learning_rate": 0.00013422126281705343,
+      "loss": 3.2529,
       "step": 72050
     },
     {
-      "epoch": 7.760198041115058,
-      "grad_norm": 0.8525424003601074,
-      "learning_rate": 0.00013470746686779442,
-      "loss": 3.2315,
+      "epoch": 7.773584905660377,
+      "grad_norm": 0.8065376281738281,
+      "learning_rate": 0.00013389746357258498,
+      "loss": 3.2351,
       "step": 72100
     },
     {
-      "epoch": 7.765579593154666,
-      "grad_norm": 0.7953471541404724,
-      "learning_rate": 0.000134384225837733,
-      "loss": 3.2473,
+      "epoch": 7.7789757412398925,
+      "grad_norm": 0.8617841005325317,
+      "learning_rate": 0.00013357366432811656,
+      "loss": 3.2283,
       "step": 72150
     },
     {
-      "epoch": 7.770961145194274,
-      "grad_norm": 0.8736206293106079,
-      "learning_rate": 0.00013406098480767158,
-      "loss": 3.241,
+      "epoch": 7.784366576819407,
+      "grad_norm": 0.8340649008750916,
+      "learning_rate": 0.00013324986508364813,
+      "loss": 3.2231,
       "step": 72200
     },
     {
-      "epoch": 7.776342697233883,
-      "grad_norm": 0.794816255569458,
-      "learning_rate": 0.00013373774377761015,
-      "loss": 3.2199,
+      "epoch": 7.789757412398922,
+      "grad_norm": 0.7925496101379395,
+      "learning_rate": 0.0001329260658391797,
+      "loss": 3.2391,
       "step": 72250
     },
     {
-      "epoch": 7.78172424927349,
-      "grad_norm": 0.8048193454742432,
-      "learning_rate": 0.00013341450274754874,
-      "loss": 3.2403,
+      "epoch": 7.795148247978437,
+      "grad_norm": 0.8307934999465942,
+      "learning_rate": 0.00013260226659471126,
+      "loss": 3.2345,
       "step": 72300
     },
     {
-      "epoch": 7.787105801313099,
-      "grad_norm": 0.8669823408126831,
-      "learning_rate": 0.0001330912617174873,
-      "loss": 3.2492,
+      "epoch": 7.800539083557951,
+      "grad_norm": 0.8316272497177124,
+      "learning_rate": 0.00013227846735024284,
+      "loss": 3.2238,
       "step": 72350
     },
     {
-      "epoch": 7.792487353352707,
-      "grad_norm": 0.8714507818222046,
-      "learning_rate": 0.0001327680206874259,
-      "loss": 3.2362,
+      "epoch": 7.8059299191374665,
+      "grad_norm": 0.8381049036979675,
+      "learning_rate": 0.00013195466810577442,
+      "loss": 3.2281,
       "step": 72400
     },
     {
-      "epoch": 7.797868905392315,
-      "grad_norm": 0.7854097485542297,
-      "learning_rate": 0.0001324447796573645,
-      "loss": 3.2506,
+      "epoch": 7.811320754716981,
+      "grad_norm": 0.8335167765617371,
+      "learning_rate": 0.00013163734484619534,
+      "loss": 3.242,
       "step": 72450
     },
     {
-      "epoch": 7.803250457431924,
-      "grad_norm": 0.8088889122009277,
-      "learning_rate": 0.00013212153862730307,
-      "loss": 3.2229,
+      "epoch": 7.816711590296496,
+      "grad_norm": 0.8080608248710632,
+      "learning_rate": 0.00013131354560172692,
+      "loss": 3.2331,
       "step": 72500
     },
     {
-      "epoch": 7.808632009471531,
-      "grad_norm": 0.8250226378440857,
-      "learning_rate": 0.00013179829759724166,
-      "loss": 3.2271,
+      "epoch": 7.822102425876011,
+      "grad_norm": 0.8187773823738098,
+      "learning_rate": 0.0001309897463572585,
+      "loss": 3.224,
       "step": 72550
     },
     {
-      "epoch": 7.81401356151114,
-      "grad_norm": 0.7839990854263306,
-      "learning_rate": 0.00013147505656718026,
-      "loss": 3.2476,
+      "epoch": 7.827493261455525,
+      "grad_norm": 0.8168429136276245,
+      "learning_rate": 0.00013066594711279007,
+      "loss": 3.2175,
       "step": 72600
     },
     {
-      "epoch": 7.819395113550748,
-      "grad_norm": 0.8592643141746521,
-      "learning_rate": 0.00013115181553711883,
-      "loss": 3.2563,
+      "epoch": 7.83288409703504,
+      "grad_norm": 0.7818455696105957,
+      "learning_rate": 0.00013034214786832162,
+      "loss": 3.2186,
       "step": 72650
     },
     {
-      "epoch": 7.824776665590356,
-      "grad_norm": 0.834892749786377,
-      "learning_rate": 0.00013082857450705742,
-      "loss": 3.2486,
+      "epoch": 7.8382749326145555,
+      "grad_norm": 0.8117259740829468,
+      "learning_rate": 0.0001300183486238532,
+      "loss": 3.2438,
       "step": 72700
     },
     {
-      "epoch": 7.830158217629965,
-      "grad_norm": 0.8037497401237488,
-      "learning_rate": 0.00013050533347699602,
-      "loss": 3.2666,
+      "epoch": 7.84366576819407,
+      "grad_norm": 0.821569561958313,
+      "learning_rate": 0.00012969454937938478,
+      "loss": 3.2344,
       "step": 72750
     },
     {
-      "epoch": 7.835539769669572,
-      "grad_norm": 0.7950841188430786,
-      "learning_rate": 0.00013018209244693458,
-      "loss": 3.2504,
+      "epoch": 7.849056603773585,
+      "grad_norm": 0.858958899974823,
+      "learning_rate": 0.00012937075013491636,
+      "loss": 3.2117,
       "step": 72800
     },
     {
-      "epoch": 7.840921321709181,
-      "grad_norm": 0.868694007396698,
-      "learning_rate": 0.00012985885141687318,
-      "loss": 3.2259,
+      "epoch": 7.8544474393531,
+      "grad_norm": 0.8257452249526978,
+      "learning_rate": 0.0001290469508904479,
+      "loss": 3.2328,
       "step": 72850
     },
     {
-      "epoch": 7.846302873748789,
-      "grad_norm": 0.8432836532592773,
-      "learning_rate": 0.00012953561038681175,
-      "loss": 3.2525,
+      "epoch": 7.859838274932614,
+      "grad_norm": 0.8281476497650146,
+      "learning_rate": 0.00012872315164597949,
+      "loss": 3.2316,
       "step": 72900
     },
     {
-      "epoch": 7.851684425788397,
-      "grad_norm": 0.8124203681945801,
-      "learning_rate": 0.00012921236935675034,
-      "loss": 3.2448,
+      "epoch": 7.8652291105121295,
+      "grad_norm": 0.7852951288223267,
+      "learning_rate": 0.00012839935240151106,
+      "loss": 3.2414,
       "step": 72950
     },
     {
-      "epoch": 7.857065977828006,
-      "grad_norm": 0.8090877532958984,
-      "learning_rate": 0.0001288891283266889,
-      "loss": 3.2356,
+      "epoch": 7.870619946091644,
+      "grad_norm": 0.8130508661270142,
+      "learning_rate": 0.00012807555315704261,
+      "loss": 3.2381,
       "step": 73000
     },
     {
-      "epoch": 7.857065977828006,
-      "eval_accuracy": 0.38977637171282914,
-      "eval_loss": 3.336695671081543,
-      "eval_runtime": 180.1915,
-      "eval_samples_per_second": 99.955,
-      "eval_steps_per_second": 6.249,
+      "epoch": 7.870619946091644,
+      "eval_accuracy": 0.3898635113288563,
+      "eval_loss": 3.334094762802124,
+      "eval_runtime": 184.0804,
+      "eval_samples_per_second": 97.843,
+      "eval_steps_per_second": 6.117,
       "step": 73000
     },
     {
-      "epoch": 7.862447529867614,
-      "grad_norm": 0.8222204446792603,
-      "learning_rate": 0.0001285658872966275,
-      "loss": 3.2259,
+      "epoch": 7.876010781671159,
+      "grad_norm": 0.821366548538208,
+      "learning_rate": 0.0001277517539125742,
+      "loss": 3.2282,
       "step": 73050
     },
     {
-      "epoch": 7.867829081907222,
-      "grad_norm": 0.8175197839736938,
-      "learning_rate": 0.0001282426462665661,
-      "loss": 3.2454,
+      "epoch": 7.881401617250674,
+      "grad_norm": 0.8474908471107483,
+      "learning_rate": 0.00012742795466810577,
+      "loss": 3.2316,
       "step": 73100
     },
     {
-      "epoch": 7.87321063394683,
-      "grad_norm": 0.7853533625602722,
-      "learning_rate": 0.00012791940523650466,
-      "loss": 3.2267,
+      "epoch": 7.886792452830189,
+      "grad_norm": 0.8374301195144653,
+      "learning_rate": 0.00012710415542363732,
+      "loss": 3.2432,
       "step": 73150
     },
     {
-      "epoch": 7.878592185986438,
-      "grad_norm": 0.8725240230560303,
-      "learning_rate": 0.00012759616420644326,
-      "loss": 3.2317,
+      "epoch": 7.892183288409703,
+      "grad_norm": 0.8285154700279236,
+      "learning_rate": 0.0001267803561791689,
+      "loss": 3.2094,
       "step": 73200
     },
     {
-      "epoch": 7.883973738026047,
-      "grad_norm": 0.8071550726890564,
-      "learning_rate": 0.00012727292317638185,
-      "loss": 3.2218,
+      "epoch": 7.8975741239892185,
+      "grad_norm": 0.8725809454917908,
+      "learning_rate": 0.00012645655693470048,
+      "loss": 3.2334,
       "step": 73250
     },
     {
-      "epoch": 7.889355290065655,
-      "grad_norm": 0.8475930094718933,
-      "learning_rate": 0.00012694968214632045,
-      "loss": 3.2259,
+      "epoch": 7.902964959568733,
+      "grad_norm": 0.798031747341156,
+      "learning_rate": 0.00012613275769023203,
+      "loss": 3.2256,
       "step": 73300
     },
     {
-      "epoch": 7.894736842105263,
-      "grad_norm": 0.8381208181381226,
-      "learning_rate": 0.00012662644111625902,
-      "loss": 3.2366,
+      "epoch": 7.908355795148248,
+      "grad_norm": 0.7943145632743835,
+      "learning_rate": 0.0001258089584457636,
+      "loss": 3.26,
       "step": 73350
     },
     {
-      "epoch": 7.900118394144871,
-      "grad_norm": 0.809328019618988,
-      "learning_rate": 0.0001263032000861976,
-      "loss": 3.2301,
+      "epoch": 7.913746630727763,
+      "grad_norm": 0.7992303371429443,
+      "learning_rate": 0.00012548515920129518,
+      "loss": 3.2471,
       "step": 73400
     },
     {
-      "epoch": 7.90549994618448,
-      "grad_norm": 0.8766582608222961,
-      "learning_rate": 0.00012597995905613618,
-      "loss": 3.2601,
+      "epoch": 7.919137466307277,
+      "grad_norm": 0.7948381304740906,
+      "learning_rate": 0.00012516135995682676,
+      "loss": 3.245,
       "step": 73450
     },
     {
-      "epoch": 7.910881498224088,
-      "grad_norm": 0.8222476840019226,
-      "learning_rate": 0.00012565671802607477,
-      "loss": 3.2429,
+      "epoch": 7.9245283018867925,
+      "grad_norm": 0.8146044015884399,
+      "learning_rate": 0.0001248375607123583,
+      "loss": 3.2501,
       "step": 73500
     },
     {
-      "epoch": 7.916263050263696,
-      "grad_norm": 0.8451269268989563,
-      "learning_rate": 0.00012533347699601334,
-      "loss": 3.2484,
+      "epoch": 7.929919137466308,
+      "grad_norm": 0.7986351847648621,
+      "learning_rate": 0.0001245137614678899,
+      "loss": 3.2247,
       "step": 73550
     },
     {
-      "epoch": 7.921644602303305,
-      "grad_norm": 0.8155948519706726,
-      "learning_rate": 0.00012501023596595193,
-      "loss": 3.2388,
+      "epoch": 7.935309973045822,
+      "grad_norm": 0.8014178276062012,
+      "learning_rate": 0.00012418996222342147,
+      "loss": 3.2466,
       "step": 73600
     },
     {
-      "epoch": 7.927026154342912,
-      "grad_norm": 0.8491314053535461,
-      "learning_rate": 0.0001246869949358905,
-      "loss": 3.2469,
+      "epoch": 7.940700808625337,
+      "grad_norm": 0.8443597555160522,
+      "learning_rate": 0.00012386616297895304,
+      "loss": 3.2313,
       "step": 73650
     },
     {
-      "epoch": 7.932407706382521,
-      "grad_norm": 0.8263399600982666,
-      "learning_rate": 0.0001243637539058291,
-      "loss": 3.2575,
+      "epoch": 7.946091644204852,
+      "grad_norm": 0.8308108448982239,
+      "learning_rate": 0.0001235423637344846,
+      "loss": 3.224,
       "step": 73700
     },
     {
-      "epoch": 7.937789258422129,
-      "grad_norm": 0.8386456966400146,
-      "learning_rate": 0.0001240405128757677,
-      "loss": 3.2509,
+      "epoch": 7.951482479784366,
+      "grad_norm": 0.8428900241851807,
+      "learning_rate": 0.00012321856449001617,
+      "loss": 3.2326,
       "step": 73750
     },
     {
-      "epoch": 7.943170810461737,
-      "grad_norm": 0.7784790396690369,
-      "learning_rate": 0.00012371727184570629,
-      "loss": 3.2274,
+      "epoch": 7.9568733153638815,
+      "grad_norm": 0.8092119693756104,
+      "learning_rate": 0.00012289476524554775,
+      "loss": 3.2287,
       "step": 73800
     },
     {
-      "epoch": 7.948552362501346,
-      "grad_norm": 0.8531802892684937,
-      "learning_rate": 0.00012339403081564485,
-      "loss": 3.2205,
+      "epoch": 7.962264150943396,
+      "grad_norm": 0.8208738565444946,
+      "learning_rate": 0.00012257096600107933,
+      "loss": 3.23,
       "step": 73850
     },
     {
-      "epoch": 7.953933914540953,
-      "grad_norm": 0.8842747211456299,
-      "learning_rate": 0.00012307078978558345,
-      "loss": 3.2316,
+      "epoch": 7.967654986522911,
+      "grad_norm": 0.7901440262794495,
+      "learning_rate": 0.0001222471667566109,
+      "loss": 3.2279,
       "step": 73900
     },
     {
-      "epoch": 7.959315466580562,
-      "grad_norm": 0.8715245723724365,
-      "learning_rate": 0.00012274754875552202,
-      "loss": 3.2406,
+      "epoch": 7.973045822102426,
+      "grad_norm": 0.7914726734161377,
+      "learning_rate": 0.00012192336751214245,
+      "loss": 3.2462,
       "step": 73950
     },
     {
-      "epoch": 7.96469701862017,
-      "grad_norm": 0.8610970973968506,
-      "learning_rate": 0.0001224243077254606,
-      "loss": 3.2443,
+      "epoch": 7.97843665768194,
+      "grad_norm": 0.8370747566223145,
+      "learning_rate": 0.00012159956826767403,
+      "loss": 3.2493,
       "step": 74000
     },
     {
-      "epoch": 7.96469701862017,
-      "eval_accuracy": 0.39020728906590135,
-      "eval_loss": 3.335205554962158,
-      "eval_runtime": 180.1463,
-      "eval_samples_per_second": 99.98,
-      "eval_steps_per_second": 6.25,
+      "epoch": 7.97843665768194,
+      "eval_accuracy": 0.39020718041301355,
+      "eval_loss": 3.3315134048461914,
+      "eval_runtime": 184.0577,
+      "eval_samples_per_second": 97.855,
+      "eval_steps_per_second": 6.118,
       "step": 74000
     },
     {
-      "epoch": 7.970078570659778,
-      "grad_norm": 0.8783667087554932,
-      "learning_rate": 0.00012210106669539918,
-      "loss": 3.239,
+      "epoch": 7.9838274932614555,
+      "grad_norm": 0.8077449202537537,
+      "learning_rate": 0.00012127576902320561,
+      "loss": 3.2454,
       "step": 74050
     },
     {
-      "epoch": 7.975460122699387,
-      "grad_norm": 0.8203068971633911,
-      "learning_rate": 0.00012177782566533779,
-      "loss": 3.2352,
+      "epoch": 7.989218328840971,
+      "grad_norm": 0.8788904547691345,
+      "learning_rate": 0.00012095196977873717,
+      "loss": 3.2427,
       "step": 74100
     },
     {
-      "epoch": 7.980841674738995,
-      "grad_norm": 0.8619286417961121,
-      "learning_rate": 0.00012145458463527635,
-      "loss": 3.2383,
+      "epoch": 7.994609164420485,
+      "grad_norm": 0.8069302439689636,
+      "learning_rate": 0.00012062817053426874,
+      "loss": 3.2427,
       "step": 74150
     },
     {
-      "epoch": 7.986223226778603,
-      "grad_norm": 0.8609179854393005,
-      "learning_rate": 0.00012113134360521495,
-      "loss": 3.248,
+      "epoch": 8.0,
+      "grad_norm": 1.6512125730514526,
+      "learning_rate": 0.00012030437128980032,
+      "loss": 3.2145,
       "step": 74200
     },
     {
-      "epoch": 7.991604778818211,
-      "grad_norm": 0.8277957439422607,
-      "learning_rate": 0.00012080810257515353,
-      "loss": 3.2689,
+      "epoch": 8.005390835579515,
+      "grad_norm": 0.8190436363220215,
+      "learning_rate": 0.00011998057204533188,
+      "loss": 3.1683,
       "step": 74250
     },
     {
-      "epoch": 7.996986330857819,
-      "grad_norm": 0.8138071894645691,
-      "learning_rate": 0.00012048486154509212,
-      "loss": 3.2372,
+      "epoch": 8.01078167115903,
+      "grad_norm": 0.7968325018882751,
+      "learning_rate": 0.00011965677280086346,
+      "loss": 3.1498,
       "step": 74300
     },
     {
-      "epoch": 8.002367882897428,
-      "grad_norm": 0.8171045184135437,
-      "learning_rate": 0.00012016162051503069,
-      "loss": 3.2023,
+      "epoch": 8.016172506738544,
+      "grad_norm": 0.8100771903991699,
+      "learning_rate": 0.00011933297355639502,
+      "loss": 3.1512,
       "step": 74350
     },
     {
-      "epoch": 8.007749434937036,
-      "grad_norm": 0.884177029132843,
-      "learning_rate": 0.00011983837948496929,
-      "loss": 3.1379,
+      "epoch": 8.021563342318059,
+      "grad_norm": 0.8254095315933228,
+      "learning_rate": 0.00011900917431192659,
+      "loss": 3.1551,
       "step": 74400
     },
     {
-      "epoch": 8.013130986976645,
-      "grad_norm": 0.8275094032287598,
-      "learning_rate": 0.0001195216032755091,
-      "loss": 3.1535,
+      "epoch": 8.026954177897574,
+      "grad_norm": 0.8481687903404236,
+      "learning_rate": 0.00011869185105234754,
+      "loss": 3.1557,
       "step": 74450
     },
     {
-      "epoch": 8.018512539016251,
-      "grad_norm": 0.8046960830688477,
-      "learning_rate": 0.00011919836224544767,
-      "loss": 3.1526,
+      "epoch": 8.032345013477089,
+      "grad_norm": 0.8376723527908325,
+      "learning_rate": 0.0001183680518078791,
+      "loss": 3.1587,
       "step": 74500
     },
     {
-      "epoch": 8.02389409105586,
-      "grad_norm": 0.8428935408592224,
-      "learning_rate": 0.00011887512121538627,
-      "loss": 3.1598,
+      "epoch": 8.037735849056604,
+      "grad_norm": 0.8575873374938965,
+      "learning_rate": 0.00011804425256341068,
+      "loss": 3.1746,
       "step": 74550
     },
     {
-      "epoch": 8.029275643095469,
-      "grad_norm": 0.8413782715797424,
-      "learning_rate": 0.00011855188018532485,
-      "loss": 3.1582,
+      "epoch": 8.04312668463612,
+      "grad_norm": 0.8203369379043579,
+      "learning_rate": 0.00011772045331894224,
+      "loss": 3.1723,
       "step": 74600
     },
     {
-      "epoch": 8.034657195135077,
-      "grad_norm": 0.8058221340179443,
-      "learning_rate": 0.00011822863915526343,
-      "loss": 3.1563,
+      "epoch": 8.048517520215633,
+      "grad_norm": 0.8016600012779236,
+      "learning_rate": 0.00011739665407447382,
+      "loss": 3.1673,
       "step": 74650
     },
     {
-      "epoch": 8.040038747174686,
-      "grad_norm": 0.8850978016853333,
-      "learning_rate": 0.00011790539812520201,
-      "loss": 3.1662,
+      "epoch": 8.053908355795148,
+      "grad_norm": 0.7998136878013611,
+      "learning_rate": 0.00011707285483000539,
+      "loss": 3.1564,
       "step": 74700
     },
     {
-      "epoch": 8.045420299214294,
-      "grad_norm": 0.8446381688117981,
-      "learning_rate": 0.00011758215709514061,
-      "loss": 3.1713,
+      "epoch": 8.059299191374663,
+      "grad_norm": 0.8106847405433655,
+      "learning_rate": 0.00011674905558553695,
+      "loss": 3.1848,
       "step": 74750
     },
     {
-      "epoch": 8.050801851253901,
-      "grad_norm": 0.8259122967720032,
-      "learning_rate": 0.00011725891606507917,
-      "loss": 3.1745,
+      "epoch": 8.064690026954178,
+      "grad_norm": 0.8593471646308899,
+      "learning_rate": 0.00011642525634106853,
+      "loss": 3.1623,
       "step": 74800
     },
     {
-      "epoch": 8.05618340329351,
-      "grad_norm": 0.8464698195457458,
-      "learning_rate": 0.00011693567503501777,
-      "loss": 3.1503,
+      "epoch": 8.070080862533693,
+      "grad_norm": 0.8197831511497498,
+      "learning_rate": 0.0001161014570966001,
+      "loss": 3.1671,
       "step": 74850
     },
     {
-      "epoch": 8.061564955333118,
-      "grad_norm": 0.841551661491394,
-      "learning_rate": 0.00011661243400495635,
-      "loss": 3.1633,
+      "epoch": 8.075471698113208,
+      "grad_norm": 0.8074586987495422,
+      "learning_rate": 0.00011577765785213166,
+      "loss": 3.1612,
       "step": 74900
     },
     {
-      "epoch": 8.066946507372727,
-      "grad_norm": 0.8218307495117188,
-      "learning_rate": 0.00011628919297489493,
-      "loss": 3.1706,
+      "epoch": 8.080862533692722,
+      "grad_norm": 0.8161447644233704,
+      "learning_rate": 0.00011545385860766323,
+      "loss": 3.1644,
       "step": 74950
     },
     {
-      "epoch": 8.072328059412335,
-      "grad_norm": 0.8135493993759155,
-      "learning_rate": 0.00011596595194483351,
-      "loss": 3.1627,
+      "epoch": 8.086253369272237,
+      "grad_norm": 0.8014756441116333,
+      "learning_rate": 0.00011513005936319481,
+      "loss": 3.1598,
       "step": 75000
     },
     {
-      "epoch": 8.072328059412335,
-      "eval_accuracy": 0.3899024090626939,
-      "eval_loss": 3.3387045860290527,
-      "eval_runtime": 180.3237,
-      "eval_samples_per_second": 99.881,
-      "eval_steps_per_second": 6.244,
+      "epoch": 8.086253369272237,
+      "eval_accuracy": 0.3900869016662029,
+      "eval_loss": 3.3359994888305664,
+      "eval_runtime": 184.1513,
+      "eval_samples_per_second": 97.805,
+      "eval_steps_per_second": 6.115,
       "step": 75000
     },
     {
-      "epoch": 8.077709611451942,
-      "grad_norm": 0.8384975790977478,
-      "learning_rate": 0.00011564271091477211,
-      "loss": 3.1987,
+      "epoch": 8.091644204851752,
+      "grad_norm": 0.7970461845397949,
+      "learning_rate": 0.00011480626011872639,
+      "loss": 3.1702,
       "step": 75050
     },
     {
-      "epoch": 8.08309116349155,
-      "grad_norm": 0.8627749681472778,
-      "learning_rate": 0.0001153194698847107,
-      "loss": 3.1645,
+      "epoch": 8.097035040431267,
+      "grad_norm": 0.8236755728721619,
+      "learning_rate": 0.00011448246087425794,
+      "loss": 3.1819,
       "step": 75100
     },
     {
-      "epoch": 8.088472715531159,
-      "grad_norm": 0.8553587794303894,
-      "learning_rate": 0.00011499622885464927,
-      "loss": 3.1827,
+      "epoch": 8.102425876010782,
+      "grad_norm": 0.8050115704536438,
+      "learning_rate": 0.00011415866162978952,
+      "loss": 3.17,
       "step": 75150
     },
     {
-      "epoch": 8.093854267570768,
-      "grad_norm": 0.8322249054908752,
-      "learning_rate": 0.00011467298782458786,
-      "loss": 3.1613,
+      "epoch": 8.107816711590296,
+      "grad_norm": 0.8242062330245972,
+      "learning_rate": 0.0001138348623853211,
+      "loss": 3.1696,
       "step": 75200
     },
     {
-      "epoch": 8.099235819610376,
-      "grad_norm": 0.858957827091217,
-      "learning_rate": 0.00011434974679452645,
-      "loss": 3.156,
+      "epoch": 8.11320754716981,
+      "grad_norm": 0.8975281715393066,
+      "learning_rate": 0.00011351106314085266,
+      "loss": 3.1707,
       "step": 75250
     },
     {
-      "epoch": 8.104617371649983,
-      "grad_norm": 0.8273110389709473,
-      "learning_rate": 0.00011402650576446503,
-      "loss": 3.1704,
+      "epoch": 8.118598382749326,
+      "grad_norm": 0.8202155232429504,
+      "learning_rate": 0.00011318726389638424,
+      "loss": 3.1709,
       "step": 75300
     },
     {
-      "epoch": 8.109998923689592,
-      "grad_norm": 0.829047441482544,
-      "learning_rate": 0.00011370326473440361,
-      "loss": 3.1814,
+      "epoch": 8.123989218328841,
+      "grad_norm": 0.8221680521965027,
+      "learning_rate": 0.0001128634646519158,
+      "loss": 3.1575,
       "step": 75350
     },
     {
-      "epoch": 8.1153804757292,
-      "grad_norm": 0.8540376424789429,
-      "learning_rate": 0.0001133800237043422,
-      "loss": 3.172,
+      "epoch": 8.129380053908356,
+      "grad_norm": 0.8632169961929321,
+      "learning_rate": 0.00011253966540744738,
+      "loss": 3.1677,
       "step": 75400
     },
     {
-      "epoch": 8.120762027768809,
-      "grad_norm": 0.8101136684417725,
-      "learning_rate": 0.00011305678267428077,
-      "loss": 3.17,
+      "epoch": 8.134770889487871,
+      "grad_norm": 0.8233320116996765,
+      "learning_rate": 0.00011221586616297894,
+      "loss": 3.1678,
       "step": 75450
     },
     {
-      "epoch": 8.126143579808417,
-      "grad_norm": 0.8184477686882019,
-      "learning_rate": 0.00011273354164421936,
-      "loss": 3.1818,
+      "epoch": 8.140161725067385,
+      "grad_norm": 0.8470733165740967,
+      "learning_rate": 0.00011189206691851052,
+      "loss": 3.1719,
       "step": 75500
     },
     {
-      "epoch": 8.131525131848026,
-      "grad_norm": 0.8428223133087158,
-      "learning_rate": 0.00011241030061415795,
-      "loss": 3.1718,
+      "epoch": 8.1455525606469,
+      "grad_norm": 0.8300970196723938,
+      "learning_rate": 0.00011156826767404208,
+      "loss": 3.1735,
       "step": 75550
     },
     {
-      "epoch": 8.136906683887632,
-      "grad_norm": 0.8273821473121643,
-      "learning_rate": 0.00011208705958409654,
-      "loss": 3.1799,
+      "epoch": 8.150943396226415,
+      "grad_norm": 0.8340595960617065,
+      "learning_rate": 0.00011124446842957365,
+      "loss": 3.1744,
       "step": 75600
     },
     {
-      "epoch": 8.142288235927241,
-      "grad_norm": 0.8538662195205688,
-      "learning_rate": 0.00011176381855403511,
-      "loss": 3.1665,
+      "epoch": 8.15633423180593,
+      "grad_norm": 0.8598608374595642,
+      "learning_rate": 0.00011092066918510523,
+      "loss": 3.1589,
       "step": 75650
     },
     {
-      "epoch": 8.14766978796685,
-      "grad_norm": 0.8688125014305115,
-      "learning_rate": 0.0001114405775239737,
-      "loss": 3.17,
+      "epoch": 8.161725067385445,
+      "grad_norm": 0.8539202809333801,
+      "learning_rate": 0.0001105968699406368,
+      "loss": 3.1691,
       "step": 75700
     },
     {
-      "epoch": 8.153051340006458,
-      "grad_norm": 0.8661782145500183,
-      "learning_rate": 0.00011111733649391228,
-      "loss": 3.174,
+      "epoch": 8.167115902964959,
+      "grad_norm": 0.8094648122787476,
+      "learning_rate": 0.00011027307069616835,
+      "loss": 3.1875,
       "step": 75750
     },
     {
-      "epoch": 8.158432892046067,
-      "grad_norm": 0.8403778672218323,
-      "learning_rate": 0.00011079409546385086,
-      "loss": 3.1963,
+      "epoch": 8.172506738544474,
+      "grad_norm": 0.8318741917610168,
+      "learning_rate": 0.00010994927145169993,
+      "loss": 3.1929,
       "step": 75800
     },
     {
-      "epoch": 8.163814444085673,
-      "grad_norm": 0.8399979472160339,
-      "learning_rate": 0.00011047085443378945,
-      "loss": 3.1708,
+      "epoch": 8.177897574123989,
+      "grad_norm": 0.8207082748413086,
+      "learning_rate": 0.00010962547220723151,
+      "loss": 3.1792,
       "step": 75850
     },
     {
-      "epoch": 8.169195996125282,
-      "grad_norm": 0.8141344785690308,
-      "learning_rate": 0.00011014761340372804,
-      "loss": 3.1854,
+      "epoch": 8.183288409703504,
+      "grad_norm": 0.9037622213363647,
+      "learning_rate": 0.00010930167296276309,
+      "loss": 3.1638,
       "step": 75900
     },
     {
-      "epoch": 8.17457754816489,
-      "grad_norm": 0.8441236019134521,
-      "learning_rate": 0.00010982437237366661,
-      "loss": 3.1724,
+      "epoch": 8.18867924528302,
+      "grad_norm": 0.8640836477279663,
+      "learning_rate": 0.00010897787371829465,
+      "loss": 3.1712,
       "step": 75950
     },
     {
-      "epoch": 8.1799591002045,
-      "grad_norm": 0.8193464875221252,
-      "learning_rate": 0.0001095011313436052,
-      "loss": 3.1778,
+      "epoch": 8.194070080862534,
+      "grad_norm": 0.8582479953765869,
+      "learning_rate": 0.00010865407447382622,
+      "loss": 3.185,
       "step": 76000
     },
     {
-      "epoch": 8.1799591002045,
-      "eval_accuracy": 0.3904291582628185,
-      "eval_loss": 3.335801839828491,
-      "eval_runtime": 180.1521,
-      "eval_samples_per_second": 99.977,
-      "eval_steps_per_second": 6.25,
+      "epoch": 8.194070080862534,
+      "eval_accuracy": 0.3902142428507215,
+      "eval_loss": 3.336308479309082,
+      "eval_runtime": 184.0376,
+      "eval_samples_per_second": 97.866,
+      "eval_steps_per_second": 6.118,
       "step": 76000
     },
     {
-      "epoch": 8.185340652244108,
-      "grad_norm": 0.8355197906494141,
-      "learning_rate": 0.0001091778903135438,
-      "loss": 3.1686,
+      "epoch": 8.199460916442048,
+      "grad_norm": 0.8695352673530579,
+      "learning_rate": 0.00010833027522935779,
+      "loss": 3.1743,
       "step": 76050
     },
     {
-      "epoch": 8.190722204283716,
-      "grad_norm": 0.8153267502784729,
-      "learning_rate": 0.00010885464928348238,
-      "loss": 3.1838,
+      "epoch": 8.204851752021563,
+      "grad_norm": 0.798423707485199,
+      "learning_rate": 0.00010800647598488936,
+      "loss": 3.1791,
       "step": 76100
     },
     {
-      "epoch": 8.196103756323323,
-      "grad_norm": 0.8740131258964539,
-      "learning_rate": 0.00010853140825342096,
-      "loss": 3.1696,
+      "epoch": 8.210242587601078,
+      "grad_norm": 0.8706480860710144,
+      "learning_rate": 0.00010768267674042093,
+      "loss": 3.1891,
       "step": 76150
     },
     {
-      "epoch": 8.201485308362932,
-      "grad_norm": 0.8934776782989502,
-      "learning_rate": 0.00010820816722335954,
-      "loss": 3.1742,
+      "epoch": 8.215633423180593,
+      "grad_norm": 0.8224576711654663,
+      "learning_rate": 0.0001073588774959525,
+      "loss": 3.1826,
       "step": 76200
     },
     {
-      "epoch": 8.20686686040254,
-      "grad_norm": 0.9179877042770386,
-      "learning_rate": 0.00010788492619329814,
-      "loss": 3.1695,
+      "epoch": 8.221024258760108,
+      "grad_norm": 0.8482567667961121,
+      "learning_rate": 0.00010703507825148406,
+      "loss": 3.1784,
       "step": 76250
     },
     {
-      "epoch": 8.212248412442149,
-      "grad_norm": 0.8545275926589966,
-      "learning_rate": 0.0001075616851632367,
-      "loss": 3.172,
+      "epoch": 8.226415094339623,
+      "grad_norm": 0.8344135880470276,
+      "learning_rate": 0.00010671127900701564,
+      "loss": 3.1754,
       "step": 76300
     },
     {
-      "epoch": 8.217629964481757,
-      "grad_norm": 0.8500077724456787,
-      "learning_rate": 0.0001072384441331753,
-      "loss": 3.1938,
+      "epoch": 8.231805929919137,
+      "grad_norm": 0.870039701461792,
+      "learning_rate": 0.00010638747976254722,
+      "loss": 3.1806,
       "step": 76350
     },
     {
-      "epoch": 8.223011516521364,
-      "grad_norm": 0.8435236811637878,
-      "learning_rate": 0.00010691520310311388,
-      "loss": 3.1834,
+      "epoch": 8.237196765498652,
+      "grad_norm": 0.8409457206726074,
+      "learning_rate": 0.00010606368051807877,
+      "loss": 3.1919,
       "step": 76400
     },
     {
-      "epoch": 8.228393068560973,
-      "grad_norm": 0.8567854762077332,
-      "learning_rate": 0.00010659196207305246,
-      "loss": 3.169,
+      "epoch": 8.242587601078167,
+      "grad_norm": 0.8561326861381531,
+      "learning_rate": 0.00010573988127361035,
+      "loss": 3.1915,
       "step": 76450
     },
     {
-      "epoch": 8.233774620600581,
-      "grad_norm": 0.8797134757041931,
-      "learning_rate": 0.00010626872104299104,
-      "loss": 3.1859,
+      "epoch": 8.247978436657682,
+      "grad_norm": 0.8443933725357056,
+      "learning_rate": 0.00010542255801403128,
+      "loss": 3.1675,
       "step": 76500
     },
     {
-      "epoch": 8.23915617264019,
-      "grad_norm": 0.8667067289352417,
-      "learning_rate": 0.00010594548001292964,
-      "loss": 3.1973,
+      "epoch": 8.253369272237197,
+      "grad_norm": 0.8334992527961731,
+      "learning_rate": 0.00010509875876956286,
+      "loss": 3.1692,
       "step": 76550
     },
     {
-      "epoch": 8.244537724679798,
-      "grad_norm": 0.8613109588623047,
-      "learning_rate": 0.00010562223898286823,
-      "loss": 3.1941,
+      "epoch": 8.25876010781671,
+      "grad_norm": 0.8273249864578247,
+      "learning_rate": 0.00010477495952509443,
+      "loss": 3.1742,
       "step": 76600
     },
     {
-      "epoch": 8.249919276719407,
-      "grad_norm": 0.8120648264884949,
-      "learning_rate": 0.0001052989979528068,
-      "loss": 3.1899,
+      "epoch": 8.264150943396226,
+      "grad_norm": 0.8292849659919739,
+      "learning_rate": 0.000104451160280626,
+      "loss": 3.1838,
       "step": 76650
     },
     {
-      "epoch": 8.255300828759013,
-      "grad_norm": 0.8621121048927307,
-      "learning_rate": 0.00010497575692274539,
-      "loss": 3.1836,
+      "epoch": 8.269541778975741,
+      "grad_norm": 0.8798602819442749,
+      "learning_rate": 0.00010412736103615758,
+      "loss": 3.1755,
       "step": 76700
     },
     {
-      "epoch": 8.260682380798622,
-      "grad_norm": 0.8764901757240295,
-      "learning_rate": 0.00010465251589268397,
-      "loss": 3.2011,
+      "epoch": 8.274932614555256,
+      "grad_norm": 0.8068249821662903,
+      "learning_rate": 0.00010380356179168913,
+      "loss": 3.1704,
       "step": 76750
     },
     {
-      "epoch": 8.26606393283823,
-      "grad_norm": 0.8889468312263489,
-      "learning_rate": 0.00010432927486262254,
-      "loss": 3.1871,
+      "epoch": 8.280323450134771,
+      "grad_norm": 0.8510772585868835,
+      "learning_rate": 0.00010347976254722071,
+      "loss": 3.1885,
       "step": 76800
     },
     {
-      "epoch": 8.27144548487784,
-      "grad_norm": 0.8703168630599976,
-      "learning_rate": 0.00010400603383256114,
-      "loss": 3.1714,
+      "epoch": 8.285714285714286,
+      "grad_norm": 0.8531442880630493,
+      "learning_rate": 0.00010315596330275229,
+      "loss": 3.172,
       "step": 76850
     },
     {
-      "epoch": 8.276827036917448,
-      "grad_norm": 0.8331544399261475,
-      "learning_rate": 0.00010368279280249973,
-      "loss": 3.1679,
+      "epoch": 8.2911051212938,
+      "grad_norm": 0.8510355353355408,
+      "learning_rate": 0.00010283216405828387,
+      "loss": 3.1814,
       "step": 76900
     },
     {
-      "epoch": 8.282208588957054,
-      "grad_norm": 0.8356614708900452,
-      "learning_rate": 0.0001033595517724383,
-      "loss": 3.187,
+      "epoch": 8.296495956873315,
+      "grad_norm": 0.7824259400367737,
+      "learning_rate": 0.00010250836481381542,
+      "loss": 3.1887,
       "step": 76950
     },
     {
-      "epoch": 8.287590140996663,
-      "grad_norm": 0.9109562039375305,
-      "learning_rate": 0.00010303631074237689,
-      "loss": 3.1902,
+      "epoch": 8.30188679245283,
+      "grad_norm": 0.8287332057952881,
+      "learning_rate": 0.000102184565569347,
+      "loss": 3.1816,
       "step": 77000
     },
     {
-      "epoch": 8.287590140996663,
-      "eval_accuracy": 0.3904268765521744,
-      "eval_loss": 3.3348872661590576,
-      "eval_runtime": 180.23,
-      "eval_samples_per_second": 99.933,
-      "eval_steps_per_second": 6.248,
+      "epoch": 8.30188679245283,
+      "eval_accuracy": 0.3907112211595849,
+      "eval_loss": 3.3314242362976074,
+      "eval_runtime": 183.9989,
+      "eval_samples_per_second": 97.886,
+      "eval_steps_per_second": 6.12,
       "step": 77000
     },
     {
-      "epoch": 8.292971693036272,
-      "grad_norm": 0.857496440410614,
-      "learning_rate": 0.00010271306971231547,
-      "loss": 3.1796,
+      "epoch": 8.307277628032345,
+      "grad_norm": 0.84040367603302,
+      "learning_rate": 0.00010186724230976793,
+      "loss": 3.1971,
       "step": 77050
     },
     {
-      "epoch": 8.29835324507588,
-      "grad_norm": 0.8897091746330261,
-      "learning_rate": 0.00010238982868225407,
-      "loss": 3.195,
+      "epoch": 8.31266846361186,
+      "grad_norm": 0.8442682027816772,
+      "learning_rate": 0.0001015434430652995,
+      "loss": 3.1858,
       "step": 77100
     },
     {
-      "epoch": 8.303734797115489,
-      "grad_norm": 0.8627259135246277,
-      "learning_rate": 0.00010206658765219264,
-      "loss": 3.1633,
+      "epoch": 8.318059299191376,
+      "grad_norm": 0.9892132878303528,
+      "learning_rate": 0.00010121964382083107,
+      "loss": 3.167,
       "step": 77150
     },
     {
-      "epoch": 8.309116349155097,
-      "grad_norm": 0.8576809167861938,
-      "learning_rate": 0.00010174334662213123,
+      "epoch": 8.323450134770889,
+      "grad_norm": 0.851172149181366,
+      "learning_rate": 0.00010089584457636265,
       "loss": 3.178,
       "step": 77200
     },
     {
-      "epoch": 8.314497901194704,
-      "grad_norm": 0.913425862789154,
-      "learning_rate": 0.00010142010559206981,
-      "loss": 3.2039,
+      "epoch": 8.328840970350404,
+      "grad_norm": 0.8561208248138428,
+      "learning_rate": 0.00010057204533189423,
+      "loss": 3.1875,
       "step": 77250
     },
     {
-      "epoch": 8.319879453234313,
-      "grad_norm": 0.8628541231155396,
-      "learning_rate": 0.00010109686456200839,
-      "loss": 3.1867,
+      "epoch": 8.33423180592992,
+      "grad_norm": 0.8603373765945435,
+      "learning_rate": 0.00010024824608742578,
+      "loss": 3.1833,
       "step": 77300
     },
     {
-      "epoch": 8.325261005273921,
-      "grad_norm": 0.8340932726860046,
-      "learning_rate": 0.00010077362353194697,
-      "loss": 3.1896,
+      "epoch": 8.339622641509434,
+      "grad_norm": 0.8363519310951233,
+      "learning_rate": 9.992444684295736e-05,
+      "loss": 3.1854,
       "step": 77350
     },
     {
-      "epoch": 8.33064255731353,
-      "grad_norm": 0.8404185175895691,
-      "learning_rate": 0.00010045038250188557,
-      "loss": 3.1841,
+      "epoch": 8.34501347708895,
+      "grad_norm": 0.8520851731300354,
+      "learning_rate": 9.960064759848894e-05,
+      "loss": 3.1809,
       "step": 77400
     },
     {
-      "epoch": 8.336024109353138,
-      "grad_norm": 0.863701343536377,
-      "learning_rate": 0.00010012714147182414,
-      "loss": 3.1791,
+      "epoch": 8.350404312668463,
+      "grad_norm": 0.8456391096115112,
+      "learning_rate": 9.92768483540205e-05,
+      "loss": 3.1793,
       "step": 77450
     },
     {
-      "epoch": 8.341405661392745,
-      "grad_norm": 0.859533965587616,
-      "learning_rate": 9.980390044176273e-05,
-      "loss": 3.1825,
+      "epoch": 8.355795148247978,
+      "grad_norm": 0.8237184882164001,
+      "learning_rate": 9.895304910955206e-05,
+      "loss": 3.1995,
       "step": 77500
     },
     {
-      "epoch": 8.346787213432354,
-      "grad_norm": 0.8829686641693115,
-      "learning_rate": 9.948065941170133e-05,
-      "loss": 3.1839,
+      "epoch": 8.361185983827493,
+      "grad_norm": 0.8380216360092163,
+      "learning_rate": 9.862924986508364e-05,
+      "loss": 3.1862,
       "step": 77550
     },
     {
-      "epoch": 8.352168765471962,
-      "grad_norm": 0.8447333574295044,
-      "learning_rate": 9.91574183816399e-05,
-      "loss": 3.1852,
+      "epoch": 8.366576819407008,
+      "grad_norm": 0.8515947461128235,
+      "learning_rate": 9.83054506206152e-05,
+      "loss": 3.1918,
       "step": 77600
     },
     {
-      "epoch": 8.35755031751157,
-      "grad_norm": 0.8658083081245422,
-      "learning_rate": 9.883417735157849e-05,
-      "loss": 3.1787,
+      "epoch": 8.371967654986523,
+      "grad_norm": 0.8538819551467896,
+      "learning_rate": 9.798165137614678e-05,
+      "loss": 3.172,
       "step": 77650
     },
     {
-      "epoch": 8.36293186955118,
-      "grad_norm": 0.8201994895935059,
-      "learning_rate": 9.851093632151707e-05,
-      "loss": 3.187,
+      "epoch": 8.377358490566039,
+      "grad_norm": 0.8514916896820068,
+      "learning_rate": 9.765785213167835e-05,
+      "loss": 3.1795,
       "step": 77700
     },
     {
-      "epoch": 8.368313421590786,
-      "grad_norm": 0.8592942953109741,
-      "learning_rate": 9.818769529145566e-05,
-      "loss": 3.1745,
+      "epoch": 8.382749326145552,
+      "grad_norm": 0.8395739197731018,
+      "learning_rate": 9.733405288720992e-05,
+      "loss": 3.1694,
       "step": 77750
     },
     {
-      "epoch": 8.373694973630395,
-      "grad_norm": 0.8522955775260925,
-      "learning_rate": 9.786445426139423e-05,
-      "loss": 3.1748,
+      "epoch": 8.388140161725067,
+      "grad_norm": 0.8073175549507141,
+      "learning_rate": 9.701025364274149e-05,
+      "loss": 3.1892,
       "step": 77800
     },
     {
-      "epoch": 8.379076525670003,
-      "grad_norm": 0.8451487421989441,
-      "learning_rate": 9.754121323133283e-05,
-      "loss": 3.1835,
+      "epoch": 8.393530997304582,
+      "grad_norm": 0.8215590119361877,
+      "learning_rate": 9.668645439827307e-05,
+      "loss": 3.1815,
       "step": 77850
     },
     {
-      "epoch": 8.384458077709612,
-      "grad_norm": 0.8497421145439148,
-      "learning_rate": 9.72179722012714e-05,
-      "loss": 3.186,
+      "epoch": 8.398921832884097,
+      "grad_norm": 0.926487922668457,
+      "learning_rate": 9.636265515380464e-05,
+      "loss": 3.169,
       "step": 77900
     },
     {
-      "epoch": 8.38983962974922,
-      "grad_norm": 0.8562092781066895,
-      "learning_rate": 9.689473117120999e-05,
-      "loss": 3.1989,
+      "epoch": 8.404312668463612,
+      "grad_norm": 0.8264053463935852,
+      "learning_rate": 9.60388559093362e-05,
+      "loss": 3.1875,
       "step": 77950
     },
     {
-      "epoch": 8.395221181788829,
-      "grad_norm": 0.8583953380584717,
-      "learning_rate": 9.657149014114857e-05,
-      "loss": 3.1921,
+      "epoch": 8.409703504043126,
+      "grad_norm": 0.8608736991882324,
+      "learning_rate": 9.571505666486777e-05,
+      "loss": 3.1744,
       "step": 78000
     },
     {
-      "epoch": 8.395221181788829,
-      "eval_accuracy": 0.3905158632672945,
-      "eval_loss": 3.331059694290161,
-      "eval_runtime": 180.1209,
-      "eval_samples_per_second": 99.994,
-      "eval_steps_per_second": 6.251,
+      "epoch": 8.409703504043126,
+      "eval_accuracy": 0.39086822458247683,
+      "eval_loss": 3.328752040863037,
+      "eval_runtime": 184.0083,
+      "eval_samples_per_second": 97.881,
+      "eval_steps_per_second": 6.119,
       "step": 78000
     },
     {
-      "epoch": 8.400602733828435,
-      "grad_norm": 0.860522449016571,
-      "learning_rate": 9.624824911108716e-05,
-      "loss": 3.1781,
+      "epoch": 8.415094339622641,
+      "grad_norm": 0.8672065734863281,
+      "learning_rate": 9.539125742039935e-05,
+      "loss": 3.194,
       "step": 78050
     },
     {
-      "epoch": 8.405984285868044,
-      "grad_norm": 0.875423014163971,
-      "learning_rate": 9.592500808102574e-05,
-      "loss": 3.172,
+      "epoch": 8.420485175202156,
+      "grad_norm": 0.9120358228683472,
+      "learning_rate": 9.506745817593093e-05,
+      "loss": 3.179,
       "step": 78100
     },
     {
-      "epoch": 8.411365837907653,
-      "grad_norm": 0.8359448313713074,
-      "learning_rate": 9.560176705096433e-05,
-      "loss": 3.1911,
+      "epoch": 8.425876010781671,
+      "grad_norm": 0.8899123072624207,
+      "learning_rate": 9.474365893146248e-05,
+      "loss": 3.1789,
       "step": 78150
     },
     {
-      "epoch": 8.416747389947261,
-      "grad_norm": 0.8645302653312683,
-      "learning_rate": 9.52785260209029e-05,
-      "loss": 3.1773,
+      "epoch": 8.431266846361186,
+      "grad_norm": 0.8403229713439941,
+      "learning_rate": 9.441985968699406e-05,
+      "loss": 3.189,
       "step": 78200
     },
     {
-      "epoch": 8.42212894198687,
-      "grad_norm": 0.8695240616798401,
-      "learning_rate": 9.49552849908415e-05,
-      "loss": 3.1858,
+      "epoch": 8.436657681940702,
+      "grad_norm": 0.8904227614402771,
+      "learning_rate": 9.409606044252563e-05,
+      "loss": 3.1936,
       "step": 78250
     },
     {
-      "epoch": 8.427510494026476,
-      "grad_norm": 0.8338247537612915,
-      "learning_rate": 9.463204396078007e-05,
-      "loss": 3.1981,
+      "epoch": 8.442048517520215,
+      "grad_norm": 0.8470245003700256,
+      "learning_rate": 9.37722611980572e-05,
+      "loss": 3.1781,
       "step": 78300
     },
     {
-      "epoch": 8.432892046066085,
-      "grad_norm": 0.9113908410072327,
-      "learning_rate": 9.430880293071866e-05,
-      "loss": 3.1838,
+      "epoch": 8.44743935309973,
+      "grad_norm": 0.8289768099784851,
+      "learning_rate": 9.344846195358876e-05,
+      "loss": 3.1957,
       "step": 78350
     },
     {
-      "epoch": 8.438273598105694,
-      "grad_norm": 0.853624165058136,
-      "learning_rate": 9.398556190065726e-05,
-      "loss": 3.1826,
+      "epoch": 8.452830188679245,
+      "grad_norm": 0.8929156064987183,
+      "learning_rate": 9.312466270912034e-05,
+      "loss": 3.1718,
       "step": 78400
     },
     {
-      "epoch": 8.443655150145302,
-      "grad_norm": 0.8597407341003418,
-      "learning_rate": 9.366878569119705e-05,
-      "loss": 3.1836,
+      "epoch": 8.45822102425876,
+      "grad_norm": 0.8774979114532471,
+      "learning_rate": 9.28008634646519e-05,
+      "loss": 3.1644,
       "step": 78450
     },
     {
-      "epoch": 8.44903670218491,
-      "grad_norm": 0.8677853941917419,
-      "learning_rate": 9.334554466113565e-05,
-      "loss": 3.1936,
+      "epoch": 8.463611859838275,
+      "grad_norm": 0.8483632206916809,
+      "learning_rate": 9.247706422018348e-05,
+      "loss": 3.1908,
       "step": 78500
     },
     {
-      "epoch": 8.45441825422452,
-      "grad_norm": 0.938667893409729,
-      "learning_rate": 9.302230363107424e-05,
-      "loss": 3.194,
+      "epoch": 8.46900269541779,
+      "grad_norm": 0.8663318753242493,
+      "learning_rate": 9.215326497571505e-05,
+      "loss": 3.1744,
       "step": 78550
     },
     {
-      "epoch": 8.459799806264126,
-      "grad_norm": 0.8398536443710327,
-      "learning_rate": 9.269906260101281e-05,
-      "loss": 3.1795,
+      "epoch": 8.474393530997304,
+      "grad_norm": 0.8633823394775391,
+      "learning_rate": 9.182946573124661e-05,
+      "loss": 3.1732,
       "step": 78600
     },
     {
-      "epoch": 8.465181358303735,
-      "grad_norm": 0.8896383047103882,
-      "learning_rate": 9.23758215709514e-05,
-      "loss": 3.2054,
+      "epoch": 8.479784366576819,
+      "grad_norm": 0.8195738196372986,
+      "learning_rate": 9.150566648677819e-05,
+      "loss": 3.1827,
       "step": 78650
     },
     {
-      "epoch": 8.470562910343343,
-      "grad_norm": 0.8669857382774353,
-      "learning_rate": 9.205258054088998e-05,
-      "loss": 3.1779,
+      "epoch": 8.485175202156334,
+      "grad_norm": 0.8368211388587952,
+      "learning_rate": 9.118186724230976e-05,
+      "loss": 3.1964,
       "step": 78700
     },
     {
-      "epoch": 8.475944462382952,
-      "grad_norm": 0.8286911249160767,
-      "learning_rate": 9.172933951082856e-05,
-      "loss": 3.1926,
+      "epoch": 8.49056603773585,
+      "grad_norm": 0.8419724702835083,
+      "learning_rate": 9.085806799784134e-05,
+      "loss": 3.2129,
       "step": 78750
     },
     {
-      "epoch": 8.48132601442256,
-      "grad_norm": 0.9176493883132935,
-      "learning_rate": 9.140609848076715e-05,
-      "loss": 3.174,
+      "epoch": 8.495956873315365,
+      "grad_norm": 0.8792768120765686,
+      "learning_rate": 9.053426875337289e-05,
+      "loss": 3.1752,
       "step": 78800
     },
     {
-      "epoch": 8.486707566462167,
-      "grad_norm": 0.8651835322380066,
-      "learning_rate": 9.108285745070574e-05,
-      "loss": 3.1803,
+      "epoch": 8.501347708894878,
+      "grad_norm": 0.8336899280548096,
+      "learning_rate": 9.021046950890447e-05,
+      "loss": 3.1871,
       "step": 78850
     },
     {
-      "epoch": 8.492089118501776,
-      "grad_norm": 0.8289731740951538,
-      "learning_rate": 9.075961642064432e-05,
-      "loss": 3.1996,
+      "epoch": 8.506738544474393,
+      "grad_norm": 0.8057847619056702,
+      "learning_rate": 8.988667026443605e-05,
+      "loss": 3.1891,
       "step": 78900
     },
     {
-      "epoch": 8.497470670541384,
-      "grad_norm": 0.9024319052696228,
-      "learning_rate": 9.04363753905829e-05,
-      "loss": 3.1928,
+      "epoch": 8.512129380053908,
+      "grad_norm": 0.8165305852890015,
+      "learning_rate": 8.956287101996761e-05,
+      "loss": 3.1736,
       "step": 78950
     },
     {
-      "epoch": 8.502852222580993,
-      "grad_norm": 0.8506482243537903,
-      "learning_rate": 9.011313436052148e-05,
-      "loss": 3.1977,
+      "epoch": 8.517520215633423,
+      "grad_norm": 0.8514395952224731,
+      "learning_rate": 8.923907177549918e-05,
+      "loss": 3.1755,
       "step": 79000
     },
     {
-      "epoch": 8.502852222580993,
-      "eval_accuracy": 0.3912251493189474,
-      "eval_loss": 3.3283579349517822,
-      "eval_runtime": 180.3098,
-      "eval_samples_per_second": 99.889,
-      "eval_steps_per_second": 6.245,
+      "epoch": 8.517520215633423,
+      "eval_accuracy": 0.3914293080951508,
+      "eval_loss": 3.325227737426758,
+      "eval_runtime": 184.19,
+      "eval_samples_per_second": 97.785,
+      "eval_steps_per_second": 6.113,
       "step": 79000
     },
     {
-      "epoch": 8.508233774620601,
-      "grad_norm": 0.8655647039413452,
-      "learning_rate": 8.978989333046008e-05,
-      "loss": 3.1843,
+      "epoch": 8.522911051212938,
+      "grad_norm": 0.8737429976463318,
+      "learning_rate": 8.891527253103075e-05,
+      "loss": 3.1657,
       "step": 79050
     },
     {
-      "epoch": 8.513615326660208,
-      "grad_norm": 0.8432492613792419,
-      "learning_rate": 8.946665230039865e-05,
-      "loss": 3.1584,
+      "epoch": 8.528301886792454,
+      "grad_norm": 0.8275548219680786,
+      "learning_rate": 8.859147328656233e-05,
+      "loss": 3.1809,
       "step": 79100
     },
     {
-      "epoch": 8.518996878699816,
-      "grad_norm": 0.817113995552063,
-      "learning_rate": 8.914341127033724e-05,
-      "loss": 3.2022,
+      "epoch": 8.533692722371967,
+      "grad_norm": 0.8466005325317383,
+      "learning_rate": 8.82676740420939e-05,
+      "loss": 3.1837,
       "step": 79150
     },
     {
-      "epoch": 8.524378430739425,
-      "grad_norm": 0.8752052783966064,
-      "learning_rate": 8.882017024027582e-05,
-      "loss": 3.1791,
+      "epoch": 8.539083557951482,
+      "grad_norm": 0.843734085559845,
+      "learning_rate": 8.794387479762546e-05,
+      "loss": 3.1778,
       "step": 79200
     },
     {
-      "epoch": 8.529759982779034,
-      "grad_norm": 0.8376759886741638,
-      "learning_rate": 8.84969292102144e-05,
-      "loss": 3.1689,
+      "epoch": 8.544474393530997,
+      "grad_norm": 0.8223153352737427,
+      "learning_rate": 8.762007555315704e-05,
+      "loss": 3.177,
       "step": 79250
     },
     {
-      "epoch": 8.535141534818642,
-      "grad_norm": 0.8467958569526672,
-      "learning_rate": 8.817368818015298e-05,
-      "loss": 3.1819,
+      "epoch": 8.549865229110512,
+      "grad_norm": 0.8873054385185242,
+      "learning_rate": 8.72962763086886e-05,
+      "loss": 3.1759,
       "step": 79300
     },
     {
-      "epoch": 8.54052308685825,
-      "grad_norm": 0.842562735080719,
-      "learning_rate": 8.785044715009158e-05,
-      "loss": 3.1974,
+      "epoch": 8.555256064690028,
+      "grad_norm": 0.8709816932678223,
+      "learning_rate": 8.697247706422018e-05,
+      "loss": 3.1794,
       "step": 79350
     },
     {
-      "epoch": 8.545904638897857,
-      "grad_norm": 0.9210129976272583,
-      "learning_rate": 8.752720612003017e-05,
-      "loss": 3.191,
+      "epoch": 8.560646900269543,
+      "grad_norm": 0.8853098154067993,
+      "learning_rate": 8.664867781975174e-05,
+      "loss": 3.1844,
       "step": 79400
     },
     {
-      "epoch": 8.551286190937466,
-      "grad_norm": 0.8453426361083984,
-      "learning_rate": 8.720396508996874e-05,
-      "loss": 3.1839,
+      "epoch": 8.566037735849056,
+      "grad_norm": 0.8500816822052002,
+      "learning_rate": 8.632487857528331e-05,
+      "loss": 3.2015,
       "step": 79450
     },
     {
-      "epoch": 8.556667742977075,
-      "grad_norm": 0.8634918928146362,
-      "learning_rate": 8.688072405990734e-05,
-      "loss": 3.1739,
+      "epoch": 8.571428571428571,
+      "grad_norm": 0.8594346046447754,
+      "learning_rate": 8.600107933081489e-05,
+      "loss": 3.1761,
       "step": 79500
     },
     {
-      "epoch": 8.562049295016683,
-      "grad_norm": 0.8293809294700623,
-      "learning_rate": 8.655748302984592e-05,
-      "loss": 3.169,
+      "epoch": 8.576819407008086,
+      "grad_norm": 0.8543336391448975,
+      "learning_rate": 8.567728008634646e-05,
+      "loss": 3.1722,
       "step": 79550
     },
     {
-      "epoch": 8.567430847056292,
-      "grad_norm": 0.8583179116249084,
-      "learning_rate": 8.62342419997845e-05,
-      "loss": 3.1982,
+      "epoch": 8.582210242587601,
+      "grad_norm": 0.8096001148223877,
+      "learning_rate": 8.535348084187804e-05,
+      "loss": 3.1813,
       "step": 79600
     },
     {
-      "epoch": 8.572812399095898,
-      "grad_norm": 0.8895404934883118,
-      "learning_rate": 8.591100096972308e-05,
-      "loss": 3.1812,
+      "epoch": 8.587601078167117,
+      "grad_norm": 0.8324059247970581,
+      "learning_rate": 8.502968159740959e-05,
+      "loss": 3.1788,
       "step": 79650
     },
     {
-      "epoch": 8.578193951135507,
-      "grad_norm": 0.820783257484436,
-      "learning_rate": 8.558775993966167e-05,
-      "loss": 3.1779,
+      "epoch": 8.59299191374663,
+      "grad_norm": 0.8369441628456116,
+      "learning_rate": 8.470588235294117e-05,
+      "loss": 3.1926,
       "step": 79700
     },
     {
-      "epoch": 8.583575503175116,
-      "grad_norm": 0.9021645188331604,
-      "learning_rate": 8.526451890960024e-05,
-      "loss": 3.1975,
+      "epoch": 8.598382749326145,
+      "grad_norm": 0.877312183380127,
+      "learning_rate": 8.438208310847275e-05,
+      "loss": 3.1909,
       "step": 79750
     },
     {
-      "epoch": 8.588957055214724,
-      "grad_norm": 0.8516561985015869,
-      "learning_rate": 8.494127787953884e-05,
-      "loss": 3.1713,
+      "epoch": 8.60377358490566,
+      "grad_norm": 0.8623024225234985,
+      "learning_rate": 8.405828386400431e-05,
+      "loss": 3.2024,
       "step": 79800
     },
     {
-      "epoch": 8.594338607254333,
-      "grad_norm": 0.8891592621803284,
-      "learning_rate": 8.461803684947742e-05,
-      "loss": 3.2062,
+      "epoch": 8.609164420485175,
+      "grad_norm": 0.8372794389724731,
+      "learning_rate": 8.373448461953587e-05,
+      "loss": 3.1818,
       "step": 79850
     },
     {
-      "epoch": 8.599720159293941,
-      "grad_norm": 0.891211211681366,
-      "learning_rate": 8.429479581941601e-05,
-      "loss": 3.1899,
+      "epoch": 8.61455525606469,
+      "grad_norm": 0.8564476370811462,
+      "learning_rate": 8.341068537506745e-05,
+      "loss": 3.1773,
       "step": 79900
     },
     {
-      "epoch": 8.605101711333548,
-      "grad_norm": 0.8360291719436646,
-      "learning_rate": 8.397155478935458e-05,
-      "loss": 3.1905,
+      "epoch": 8.619946091644206,
+      "grad_norm": 0.8738343119621277,
+      "learning_rate": 8.308688613059902e-05,
+      "loss": 3.1957,
       "step": 79950
     },
     {
-      "epoch": 8.610483263373157,
-      "grad_norm": 0.8511280417442322,
-      "learning_rate": 8.364831375929317e-05,
-      "loss": 3.1876,
+      "epoch": 8.625336927223719,
+      "grad_norm": 0.9526636004447937,
+      "learning_rate": 8.27630868861306e-05,
+      "loss": 3.207,
       "step": 80000
     },
     {
-      "epoch": 8.610483263373157,
-      "eval_accuracy": 0.39159695950104856,
-      "eval_loss": 3.3249058723449707,
-      "eval_runtime": 180.4737,
-      "eval_samples_per_second": 99.798,
-      "eval_steps_per_second": 6.239,
+      "epoch": 8.625336927223719,
+      "eval_accuracy": 0.3919737677159892,
+      "eval_loss": 3.32072377204895,
+      "eval_runtime": 184.0424,
+      "eval_samples_per_second": 97.863,
+      "eval_steps_per_second": 6.118,
       "step": 80000
     },
     {
-      "epoch": 8.615864815412765,
-      "grad_norm": 0.9268209934234619,
-      "learning_rate": 8.332507272923177e-05,
-      "loss": 3.2012,
+      "epoch": 8.630727762803234,
+      "grad_norm": 0.9203380942344666,
+      "learning_rate": 8.243928764166216e-05,
+      "loss": 3.1754,
       "step": 80050
     },
     {
-      "epoch": 8.621246367452374,
-      "grad_norm": 0.865637481212616,
-      "learning_rate": 8.300183169917034e-05,
-      "loss": 3.1838,
+      "epoch": 8.63611859838275,
+      "grad_norm": 0.8921791315078735,
+      "learning_rate": 8.211548839719374e-05,
+      "loss": 3.1831,
       "step": 80100
     },
     {
-      "epoch": 8.626627919491982,
-      "grad_norm": 0.8277395367622375,
-      "learning_rate": 8.267859066910892e-05,
-      "loss": 3.1959,
+      "epoch": 8.641509433962264,
+      "grad_norm": 0.8672126531600952,
+      "learning_rate": 8.17916891527253e-05,
+      "loss": 3.1903,
       "step": 80150
     },
     {
-      "epoch": 8.632009471531589,
-      "grad_norm": 0.8716737627983093,
-      "learning_rate": 8.235534963904751e-05,
-      "loss": 3.1944,
+      "epoch": 8.64690026954178,
+      "grad_norm": 0.8813334703445435,
+      "learning_rate": 8.146788990825688e-05,
+      "loss": 3.2037,
       "step": 80200
     },
     {
-      "epoch": 8.637391023571197,
-      "grad_norm": 0.8669558763504028,
-      "learning_rate": 8.203210860898608e-05,
-      "loss": 3.2017,
+      "epoch": 8.652291105121293,
+      "grad_norm": 0.8761472702026367,
+      "learning_rate": 8.114409066378844e-05,
+      "loss": 3.1832,
       "step": 80250
     },
     {
-      "epoch": 8.642772575610806,
-      "grad_norm": 0.842424213886261,
-      "learning_rate": 8.170886757892467e-05,
-      "loss": 3.158,
+      "epoch": 8.657681940700808,
+      "grad_norm": 0.8768702745437622,
+      "learning_rate": 8.082029141932e-05,
+      "loss": 3.1961,
       "step": 80300
     },
     {
-      "epoch": 8.648154127650415,
-      "grad_norm": 0.8678565621376038,
-      "learning_rate": 8.138562654886327e-05,
-      "loss": 3.1759,
+      "epoch": 8.663072776280323,
+      "grad_norm": 0.8847965598106384,
+      "learning_rate": 8.049649217485158e-05,
+      "loss": 3.1832,
       "step": 80350
     },
     {
-      "epoch": 8.653535679690023,
-      "grad_norm": 0.8407251834869385,
-      "learning_rate": 8.106238551880185e-05,
-      "loss": 3.1856,
+      "epoch": 8.668463611859838,
+      "grad_norm": 1.0935829877853394,
+      "learning_rate": 8.017269293038316e-05,
+      "loss": 3.1909,
       "step": 80400
     },
     {
-      "epoch": 8.658917231729632,
-      "grad_norm": 0.8746572732925415,
-      "learning_rate": 8.073914448874043e-05,
-      "loss": 3.1802,
+      "epoch": 8.673854447439354,
+      "grad_norm": 0.8761574029922485,
+      "learning_rate": 7.984889368591474e-05,
+      "loss": 3.1855,
       "step": 80450
     },
     {
-      "epoch": 8.664298783769238,
-      "grad_norm": 0.827447235584259,
-      "learning_rate": 8.042236827928025e-05,
-      "loss": 3.1879,
+      "epoch": 8.679245283018869,
+      "grad_norm": 0.843662440776825,
+      "learning_rate": 7.952509444144629e-05,
+      "loss": 3.1875,
       "step": 80500
     },
     {
-      "epoch": 8.669680335808847,
-      "grad_norm": 0.8395963907241821,
-      "learning_rate": 8.009912724921882e-05,
-      "loss": 3.1832,
+      "epoch": 8.684636118598382,
+      "grad_norm": 0.857285737991333,
+      "learning_rate": 7.920129519697787e-05,
+      "loss": 3.1766,
       "step": 80550
     },
     {
-      "epoch": 8.675061887848456,
-      "grad_norm": 0.8884485363960266,
-      "learning_rate": 7.977588621915741e-05,
-      "loss": 3.1769,
+      "epoch": 8.690026954177897,
+      "grad_norm": 0.8576979637145996,
+      "learning_rate": 7.887749595250945e-05,
+      "loss": 3.1656,
       "step": 80600
     },
     {
-      "epoch": 8.680443439888064,
-      "grad_norm": 0.8452661633491516,
-      "learning_rate": 7.9452645189096e-05,
-      "loss": 3.21,
+      "epoch": 8.695417789757412,
+      "grad_norm": 0.8547276258468628,
+      "learning_rate": 7.855369670804101e-05,
+      "loss": 3.1854,
       "step": 80650
     },
     {
-      "epoch": 8.685824991927673,
-      "grad_norm": 0.9012867212295532,
-      "learning_rate": 7.912940415903459e-05,
-      "loss": 3.1787,
+      "epoch": 8.700808625336927,
+      "grad_norm": 0.9038565158843994,
+      "learning_rate": 7.822989746357257e-05,
+      "loss": 3.1754,
       "step": 80700
     },
     {
-      "epoch": 8.69120654396728,
-      "grad_norm": 0.8785222172737122,
-      "learning_rate": 7.880616312897316e-05,
-      "loss": 3.1909,
+      "epoch": 8.706199460916443,
+      "grad_norm": 0.9170523285865784,
+      "learning_rate": 7.790609821910415e-05,
+      "loss": 3.1767,
       "step": 80750
     },
     {
-      "epoch": 8.696588096006888,
-      "grad_norm": 0.9251953959465027,
-      "learning_rate": 7.848292209891175e-05,
-      "loss": 3.1933,
+      "epoch": 8.711590296495956,
+      "grad_norm": 0.8679968118667603,
+      "learning_rate": 7.758229897463572e-05,
+      "loss": 3.1728,
       "step": 80800
     },
     {
-      "epoch": 8.701969648046497,
-      "grad_norm": 0.8459815382957458,
-      "learning_rate": 7.815968106885033e-05,
-      "loss": 3.1805,
+      "epoch": 8.716981132075471,
+      "grad_norm": 0.8266805410385132,
+      "learning_rate": 7.725849973016729e-05,
+      "loss": 3.1779,
       "step": 80850
     },
     {
-      "epoch": 8.707351200086105,
-      "grad_norm": 0.8944352865219116,
-      "learning_rate": 7.783644003878891e-05,
-      "loss": 3.1875,
+      "epoch": 8.722371967654986,
+      "grad_norm": 0.891463577747345,
+      "learning_rate": 7.693470048569886e-05,
+      "loss": 3.1996,
       "step": 80900
     },
     {
-      "epoch": 8.712732752125714,
-      "grad_norm": 0.8620814681053162,
-      "learning_rate": 7.75131990087275e-05,
-      "loss": 3.1919,
+      "epoch": 8.727762803234501,
+      "grad_norm": 0.9130802750587463,
+      "learning_rate": 7.661090124123042e-05,
+      "loss": 3.1785,
       "step": 80950
     },
     {
-      "epoch": 8.718114304165322,
-      "grad_norm": 0.8819995522499084,
-      "learning_rate": 7.718995797866609e-05,
-      "loss": 3.1825,
+      "epoch": 8.733153638814017,
+      "grad_norm": 0.9209129214286804,
+      "learning_rate": 7.6287101996762e-05,
+      "loss": 3.1605,
       "step": 81000
     },
     {
-      "epoch": 8.718114304165322,
-      "eval_accuracy": 0.3917101758101512,
-      "eval_loss": 3.322167158126831,
-      "eval_runtime": 180.0608,
-      "eval_samples_per_second": 100.027,
-      "eval_steps_per_second": 6.253,
+      "epoch": 8.733153638814017,
+      "eval_accuracy": 0.39192020184229664,
+      "eval_loss": 3.3185806274414062,
+      "eval_runtime": 184.1915,
+      "eval_samples_per_second": 97.784,
+      "eval_steps_per_second": 6.113,
       "step": 81000
     },
     {
-      "epoch": 8.723495856204929,
-      "grad_norm": 0.8823017477989197,
-      "learning_rate": 7.686671694860466e-05,
-      "loss": 3.1884,
+      "epoch": 8.738544474393532,
+      "grad_norm": 0.8487036228179932,
+      "learning_rate": 7.596977873718294e-05,
+      "loss": 3.1816,
       "step": 81050
     },
     {
-      "epoch": 8.728877408244538,
-      "grad_norm": 0.8862807750701904,
-      "learning_rate": 7.654347591854325e-05,
-      "loss": 3.1889,
+      "epoch": 8.743935309973045,
+      "grad_norm": 0.8689419627189636,
+      "learning_rate": 7.564597949271451e-05,
+      "loss": 3.1824,
       "step": 81100
     },
     {
-      "epoch": 8.734258960284146,
-      "grad_norm": 0.9089580774307251,
-      "learning_rate": 7.622023488848185e-05,
-      "loss": 3.1832,
+      "epoch": 8.74932614555256,
+      "grad_norm": 0.8250060081481934,
+      "learning_rate": 7.532218024824608e-05,
+      "loss": 3.1727,
       "step": 81150
     },
     {
-      "epoch": 8.739640512323755,
-      "grad_norm": 0.8399238586425781,
-      "learning_rate": 7.589699385842043e-05,
-      "loss": 3.1812,
+      "epoch": 8.754716981132075,
+      "grad_norm": 0.8402632474899292,
+      "learning_rate": 7.499838100377764e-05,
+      "loss": 3.1955,
       "step": 81200
     },
     {
-      "epoch": 8.745022064363363,
-      "grad_norm": 0.868499755859375,
-      "learning_rate": 7.557375282835901e-05,
-      "loss": 3.1851,
+      "epoch": 8.76010781671159,
+      "grad_norm": 0.8628060817718506,
+      "learning_rate": 7.467458175930922e-05,
+      "loss": 3.182,
       "step": 81250
     },
     {
-      "epoch": 8.75040361640297,
-      "grad_norm": 0.8629152774810791,
-      "learning_rate": 7.525051179829759e-05,
-      "loss": 3.18,
+      "epoch": 8.765498652291106,
+      "grad_norm": 0.8342573642730713,
+      "learning_rate": 7.435078251484078e-05,
+      "loss": 3.1772,
       "step": 81300
     },
     {
-      "epoch": 8.755785168442578,
-      "grad_norm": 0.8513661623001099,
-      "learning_rate": 7.492727076823617e-05,
-      "loss": 3.1762,
+      "epoch": 8.77088948787062,
+      "grad_norm": 0.8328439593315125,
+      "learning_rate": 7.402698327037236e-05,
+      "loss": 3.1754,
       "step": 81350
     },
     {
-      "epoch": 8.761166720482187,
-      "grad_norm": 0.9037156105041504,
-      "learning_rate": 7.460402973817477e-05,
-      "loss": 3.2112,
+      "epoch": 8.776280323450134,
+      "grad_norm": 0.8593884706497192,
+      "learning_rate": 7.370318402590393e-05,
+      "loss": 3.1797,
       "step": 81400
     },
     {
-      "epoch": 8.766548272521796,
-      "grad_norm": 0.8453684449195862,
-      "learning_rate": 7.428078870811335e-05,
-      "loss": 3.182,
+      "epoch": 8.78167115902965,
+      "grad_norm": 0.8676581382751465,
+      "learning_rate": 7.33793847814355e-05,
+      "loss": 3.1887,
       "step": 81450
     },
     {
-      "epoch": 8.771929824561404,
-      "grad_norm": 0.831931471824646,
-      "learning_rate": 7.395754767805193e-05,
-      "loss": 3.1843,
+      "epoch": 8.787061994609164,
+      "grad_norm": 0.8676961064338684,
+      "learning_rate": 7.305558553696707e-05,
+      "loss": 3.2029,
       "step": 81500
     },
     {
-      "epoch": 8.777311376601011,
-      "grad_norm": 0.8770779967308044,
-      "learning_rate": 7.363430664799051e-05,
-      "loss": 3.1861,
+      "epoch": 8.79245283018868,
+      "grad_norm": 0.849511981010437,
+      "learning_rate": 7.273178629249865e-05,
+      "loss": 3.1837,
       "step": 81550
     },
     {
-      "epoch": 8.78269292864062,
-      "grad_norm": 0.8554553985595703,
-      "learning_rate": 7.331753043853033e-05,
-      "loss": 3.1837,
+      "epoch": 8.797843665768195,
+      "grad_norm": 0.8578782081604004,
+      "learning_rate": 7.240798704803021e-05,
+      "loss": 3.1726,
       "step": 81600
     },
     {
-      "epoch": 8.788074480680228,
-      "grad_norm": 0.8347432613372803,
-      "learning_rate": 7.299428940846891e-05,
-      "loss": 3.1902,
+      "epoch": 8.80323450134771,
+      "grad_norm": 0.8497574329376221,
+      "learning_rate": 7.208418780356179e-05,
+      "loss": 3.1865,
       "step": 81650
     },
     {
-      "epoch": 8.793456032719837,
-      "grad_norm": 0.8451250195503235,
-      "learning_rate": 7.267104837840749e-05,
-      "loss": 3.1677,
+      "epoch": 8.808625336927223,
+      "grad_norm": 0.852814257144928,
+      "learning_rate": 7.176038855909337e-05,
+      "loss": 3.1656,
       "step": 81700
     },
     {
-      "epoch": 8.798837584759445,
-      "grad_norm": 0.8608439564704895,
-      "learning_rate": 7.234780734834607e-05,
-      "loss": 3.1794,
+      "epoch": 8.814016172506738,
+      "grad_norm": 0.9635973572731018,
+      "learning_rate": 7.143658931462493e-05,
+      "loss": 3.1735,
       "step": 81750
     },
     {
-      "epoch": 8.804219136799054,
-      "grad_norm": 0.8560290336608887,
-      "learning_rate": 7.202456631828465e-05,
-      "loss": 3.1846,
+      "epoch": 8.819407008086253,
+      "grad_norm": 0.8745135068893433,
+      "learning_rate": 7.11127900701565e-05,
+      "loss": 3.1769,
       "step": 81800
     },
     {
-      "epoch": 8.80960068883866,
-      "grad_norm": 0.8419299721717834,
-      "learning_rate": 7.170132528822325e-05,
-      "loss": 3.1821,
+      "epoch": 8.824797843665769,
+      "grad_norm": 0.8970269560813904,
+      "learning_rate": 7.078899082568807e-05,
+      "loss": 3.1948,
       "step": 81850
     },
     {
-      "epoch": 8.814982240878269,
-      "grad_norm": 0.8612204194068909,
-      "learning_rate": 7.137808425816183e-05,
-      "loss": 3.1991,
+      "epoch": 8.830188679245284,
+      "grad_norm": 0.8449985384941101,
+      "learning_rate": 7.046519158121964e-05,
+      "loss": 3.1824,
       "step": 81900
     },
     {
-      "epoch": 8.820363792917878,
-      "grad_norm": 0.8462232947349548,
-      "learning_rate": 7.105484322810041e-05,
-      "loss": 3.1852,
+      "epoch": 8.835579514824797,
+      "grad_norm": 0.9532569050788879,
+      "learning_rate": 7.014139233675121e-05,
+      "loss": 3.1963,
       "step": 81950
     },
     {
-      "epoch": 8.825745344957486,
-      "grad_norm": 0.8963456153869629,
-      "learning_rate": 7.073160219803899e-05,
-      "loss": 3.2005,
+      "epoch": 8.840970350404312,
+      "grad_norm": 0.8900073766708374,
+      "learning_rate": 6.981759309228278e-05,
+      "loss": 3.177,
       "step": 82000
     },
     {
-      "epoch": 8.825745344957486,
-      "eval_accuracy": 0.3922394240266956,
-      "eval_loss": 3.3186590671539307,
-      "eval_runtime": 180.0835,
-      "eval_samples_per_second": 100.015,
-      "eval_steps_per_second": 6.253,
+      "epoch": 8.840970350404312,
+      "eval_accuracy": 0.3923745882191368,
+      "eval_loss": 3.314103364944458,
+      "eval_runtime": 183.9779,
+      "eval_samples_per_second": 97.898,
+      "eval_steps_per_second": 6.12,
       "step": 82000
     },
     {
-      "epoch": 8.831126896997095,
-      "grad_norm": 0.8609824776649475,
-      "learning_rate": 7.040836116797757e-05,
-      "loss": 3.1799,
+      "epoch": 8.846361185983827,
+      "grad_norm": 0.8444704413414001,
+      "learning_rate": 6.949379384781434e-05,
+      "loss": 3.1792,
       "step": 82050
     },
     {
-      "epoch": 8.836508449036701,
-      "grad_norm": 0.839068591594696,
-      "learning_rate": 7.008512013791617e-05,
-      "loss": 3.1796,
+      "epoch": 8.851752021563343,
+      "grad_norm": 0.8365189433097839,
+      "learning_rate": 6.916999460334592e-05,
+      "loss": 3.1758,
       "step": 82100
     },
     {
-      "epoch": 8.84189000107631,
-      "grad_norm": 0.850020170211792,
-      "learning_rate": 6.976187910785475e-05,
-      "loss": 3.19,
+      "epoch": 8.857142857142858,
+      "grad_norm": 0.8224797248840332,
+      "learning_rate": 6.884619535887748e-05,
+      "loss": 3.1862,
       "step": 82150
     },
     {
-      "epoch": 8.847271553115919,
-      "grad_norm": 0.8698260188102722,
-      "learning_rate": 6.943863807779334e-05,
-      "loss": 3.1946,
+      "epoch": 8.862533692722373,
+      "grad_norm": 0.8906913995742798,
+      "learning_rate": 6.852239611440906e-05,
+      "loss": 3.1829,
       "step": 82200
     },
     {
-      "epoch": 8.852653105155527,
-      "grad_norm": 0.8751007914543152,
-      "learning_rate": 6.911539704773192e-05,
-      "loss": 3.1885,
+      "epoch": 8.867924528301886,
+      "grad_norm": 0.8735101819038391,
+      "learning_rate": 6.819859686994062e-05,
+      "loss": 3.2098,
       "step": 82250
     },
     {
-      "epoch": 8.858034657195136,
-      "grad_norm": 0.8869954943656921,
-      "learning_rate": 6.87921560176705e-05,
-      "loss": 3.1783,
+      "epoch": 8.873315363881401,
+      "grad_norm": 0.8659020066261292,
+      "learning_rate": 6.78747976254722e-05,
+      "loss": 3.1962,
       "step": 82300
     },
     {
-      "epoch": 8.863416209234742,
-      "grad_norm": 0.8722847104072571,
-      "learning_rate": 6.846891498760909e-05,
-      "loss": 3.1916,
+      "epoch": 8.878706199460916,
+      "grad_norm": 0.8740007877349854,
+      "learning_rate": 6.755099838100377e-05,
+      "loss": 3.1672,
       "step": 82350
     },
     {
-      "epoch": 8.868797761274351,
-      "grad_norm": 0.8729155659675598,
-      "learning_rate": 6.814567395754767e-05,
-      "loss": 3.1889,
+      "epoch": 8.884097035040432,
+      "grad_norm": 0.8465805053710938,
+      "learning_rate": 6.722719913653534e-05,
+      "loss": 3.1782,
       "step": 82400
     },
     {
-      "epoch": 8.87417931331396,
-      "grad_norm": 0.8469229936599731,
-      "learning_rate": 6.782243292748626e-05,
-      "loss": 3.188,
+      "epoch": 8.889487870619947,
+      "grad_norm": 0.8515374660491943,
+      "learning_rate": 6.690339989206692e-05,
+      "loss": 3.1868,
       "step": 82450
     },
     {
-      "epoch": 8.879560865353568,
-      "grad_norm": 0.8862374424934387,
-      "learning_rate": 6.749919189742484e-05,
-      "loss": 3.1697,
+      "epoch": 8.89487870619946,
+      "grad_norm": 0.8393348455429077,
+      "learning_rate": 6.657960064759849e-05,
+      "loss": 3.1888,
       "step": 82500
     },
     {
-      "epoch": 8.884942417393177,
-      "grad_norm": 0.8705187439918518,
-      "learning_rate": 6.717595086736342e-05,
-      "loss": 3.1786,
+      "epoch": 8.900269541778975,
+      "grad_norm": 0.8477870225906372,
+      "learning_rate": 6.625580140313006e-05,
+      "loss": 3.1946,
       "step": 82550
     },
     {
-      "epoch": 8.890323969432785,
-      "grad_norm": 0.8562043309211731,
-      "learning_rate": 6.6852709837302e-05,
-      "loss": 3.1752,
+      "epoch": 8.90566037735849,
+      "grad_norm": 0.8809558749198914,
+      "learning_rate": 6.594495412844036e-05,
+      "loss": 3.1771,
       "step": 82600
     },
     {
-      "epoch": 8.895705521472392,
-      "grad_norm": 0.9573428630828857,
-      "learning_rate": 6.652946880724059e-05,
-      "loss": 3.1849,
+      "epoch": 8.911051212938006,
+      "grad_norm": 0.8632633090019226,
+      "learning_rate": 6.562115488397193e-05,
+      "loss": 3.1884,
       "step": 82650
     },
     {
-      "epoch": 8.901087073512,
-      "grad_norm": 0.8421146869659424,
-      "learning_rate": 6.620622777717918e-05,
-      "loss": 3.1855,
+      "epoch": 8.91644204851752,
+      "grad_norm": 0.9357782602310181,
+      "learning_rate": 6.52973556395035e-05,
+      "loss": 3.179,
       "step": 82700
     },
     {
-      "epoch": 8.906468625551609,
-      "grad_norm": 0.8492905497550964,
-      "learning_rate": 6.588298674711776e-05,
-      "loss": 3.1856,
+      "epoch": 8.921832884097036,
+      "grad_norm": 0.8463866114616394,
+      "learning_rate": 6.497355639503507e-05,
+      "loss": 3.1939,
       "step": 82750
     },
     {
-      "epoch": 8.911850177591218,
-      "grad_norm": 0.8871941566467285,
-      "learning_rate": 6.555974571705636e-05,
-      "loss": 3.2015,
+      "epoch": 8.92722371967655,
+      "grad_norm": 0.850902795791626,
+      "learning_rate": 6.464975715056663e-05,
+      "loss": 3.1906,
       "step": 82800
     },
     {
-      "epoch": 8.917231729630826,
-      "grad_norm": 0.8534443378448486,
-      "learning_rate": 6.523650468699494e-05,
-      "loss": 3.168,
+      "epoch": 8.932614555256064,
+      "grad_norm": 0.8806354999542236,
+      "learning_rate": 6.432595790609821e-05,
+      "loss": 3.1934,
       "step": 82850
     },
     {
-      "epoch": 8.922613281670433,
-      "grad_norm": 0.8974738717079163,
-      "learning_rate": 6.491326365693352e-05,
-      "loss": 3.1728,
+      "epoch": 8.93800539083558,
+      "grad_norm": 0.8872904777526855,
+      "learning_rate": 6.400215866162979e-05,
+      "loss": 3.1907,
       "step": 82900
     },
     {
-      "epoch": 8.927994833710041,
-      "grad_norm": 0.8600006103515625,
-      "learning_rate": 6.45900226268721e-05,
-      "loss": 3.1972,
+      "epoch": 8.943396226415095,
+      "grad_norm": 0.84234219789505,
+      "learning_rate": 6.367835941716135e-05,
+      "loss": 3.1692,
       "step": 82950
     },
     {
-      "epoch": 8.93337638574965,
-      "grad_norm": 0.8655703663825989,
-      "learning_rate": 6.426678159681068e-05,
-      "loss": 3.1977,
+      "epoch": 8.94878706199461,
+      "grad_norm": 0.908178985118866,
+      "learning_rate": 6.335456017269293e-05,
+      "loss": 3.1905,
       "step": 83000
     },
     {
-      "epoch": 8.93337638574965,
-      "eval_accuracy": 0.39260102083734216,
-      "eval_loss": 3.314582586288452,
-      "eval_runtime": 180.2494,
-      "eval_samples_per_second": 99.923,
-      "eval_steps_per_second": 6.247,
+      "epoch": 8.94878706199461,
+      "eval_accuracy": 0.39287047999912206,
+      "eval_loss": 3.311105489730835,
+      "eval_runtime": 183.9781,
+      "eval_samples_per_second": 97.898,
+      "eval_steps_per_second": 6.12,
       "step": 83000
     },
     {
-      "epoch": 8.938757937789259,
-      "grad_norm": 0.8925529718399048,
-      "learning_rate": 6.394354056674928e-05,
-      "loss": 3.1958,
+      "epoch": 8.954177897574123,
+      "grad_norm": 0.8998070955276489,
+      "learning_rate": 6.30307609282245e-05,
+      "loss": 3.1686,
       "step": 83050
     },
     {
-      "epoch": 8.944139489828867,
-      "grad_norm": 0.8607410788536072,
-      "learning_rate": 6.362029953668786e-05,
-      "loss": 3.1825,
+      "epoch": 8.959568733153638,
+      "grad_norm": 0.8652181625366211,
+      "learning_rate": 6.270696168375607e-05,
+      "loss": 3.1663,
       "step": 83100
     },
     {
-      "epoch": 8.949521041868476,
-      "grad_norm": 0.8853652477264404,
-      "learning_rate": 6.329705850662644e-05,
-      "loss": 3.1842,
+      "epoch": 8.964959568733153,
+      "grad_norm": 0.9023402333259583,
+      "learning_rate": 6.238316243928764e-05,
+      "loss": 3.1859,
       "step": 83150
     },
     {
-      "epoch": 8.954902593908082,
-      "grad_norm": 0.8844691514968872,
-      "learning_rate": 6.297381747656502e-05,
-      "loss": 3.1719,
+      "epoch": 8.970350404312669,
+      "grad_norm": 0.9153844714164734,
+      "learning_rate": 6.205936319481921e-05,
+      "loss": 3.1969,
       "step": 83200
     },
     {
-      "epoch": 8.960284145947691,
-      "grad_norm": 0.8798843622207642,
-      "learning_rate": 6.26505764465036e-05,
-      "loss": 3.1837,
+      "epoch": 8.975741239892184,
+      "grad_norm": 0.8064547181129456,
+      "learning_rate": 6.173556395035078e-05,
+      "loss": 3.1834,
       "step": 83250
     },
     {
-      "epoch": 8.9656656979873,
-      "grad_norm": 0.8720901608467102,
-      "learning_rate": 6.23273354164422e-05,
-      "loss": 3.1988,
+      "epoch": 8.981132075471699,
+      "grad_norm": 0.9473127722740173,
+      "learning_rate": 6.141176470588236e-05,
+      "loss": 3.1831,
       "step": 83300
     },
     {
-      "epoch": 8.971047250026908,
-      "grad_norm": 0.8754738569259644,
-      "learning_rate": 6.200409438638078e-05,
-      "loss": 3.1926,
+      "epoch": 8.986522911051212,
+      "grad_norm": 0.9589687585830688,
+      "learning_rate": 6.108796546141392e-05,
+      "loss": 3.1675,
       "step": 83350
     },
     {
-      "epoch": 8.976428802066517,
-      "grad_norm": 0.8542868494987488,
-      "learning_rate": 6.168085335631936e-05,
-      "loss": 3.1732,
+      "epoch": 8.991913746630727,
+      "grad_norm": 0.861990213394165,
+      "learning_rate": 6.076416621694549e-05,
+      "loss": 3.1824,
       "step": 83400
     },
     {
-      "epoch": 8.981810354106123,
-      "grad_norm": 0.8611506223678589,
-      "learning_rate": 6.135761232625794e-05,
-      "loss": 3.1916,
+      "epoch": 8.997304582210242,
+      "grad_norm": 0.8862741589546204,
+      "learning_rate": 6.0440366972477055e-05,
+      "loss": 3.1956,
       "step": 83450
     },
     {
-      "epoch": 8.987191906145732,
-      "grad_norm": 0.8679481744766235,
-      "learning_rate": 6.1034371296196526e-05,
-      "loss": 3.1845,
+      "epoch": 9.002695417789758,
+      "grad_norm": 0.9252740144729614,
+      "learning_rate": 6.011656772800863e-05,
+      "loss": 3.1583,
       "step": 83500
     },
     {
-      "epoch": 8.99257345818534,
-      "grad_norm": 0.8732113838195801,
-      "learning_rate": 6.0711130266135114e-05,
-      "loss": 3.1882,
+      "epoch": 9.008086253369273,
+      "grad_norm": 0.8500978946685791,
+      "learning_rate": 5.9792768483540197e-05,
+      "loss": 3.1341,
       "step": 83550
     },
     {
-      "epoch": 8.997955010224949,
-      "grad_norm": 0.8252281546592712,
-      "learning_rate": 6.0387889236073695e-05,
-      "loss": 3.2006,
+      "epoch": 9.013477088948788,
+      "grad_norm": 0.8388237953186035,
+      "learning_rate": 5.946896923907177e-05,
+      "loss": 3.1146,
       "step": 83600
     },
     {
-      "epoch": 9.003336562264558,
-      "grad_norm": 0.8319926857948303,
-      "learning_rate": 6.0064648206012276e-05,
-      "loss": 3.145,
+      "epoch": 9.018867924528301,
+      "grad_norm": 0.8982084393501282,
+      "learning_rate": 5.9145169994603345e-05,
+      "loss": 3.1374,
       "step": 83650
     },
     {
-      "epoch": 9.008718114304166,
-      "grad_norm": 0.8742890357971191,
-      "learning_rate": 5.9741407175950864e-05,
-      "loss": 3.1252,
+      "epoch": 9.024258760107816,
+      "grad_norm": 0.8911091685295105,
+      "learning_rate": 5.882137075013491e-05,
+      "loss": 3.1199,
       "step": 83700
     },
     {
-      "epoch": 9.014099666343773,
-      "grad_norm": 0.8541421294212341,
-      "learning_rate": 5.9418166145889445e-05,
-      "loss": 3.1408,
+      "epoch": 9.029649595687331,
+      "grad_norm": 0.8868625164031982,
+      "learning_rate": 5.849757150566649e-05,
+      "loss": 3.1067,
       "step": 83750
     },
     {
-      "epoch": 9.019481218383381,
-      "grad_norm": 0.9140190482139587,
-      "learning_rate": 5.910138993642926e-05,
-      "loss": 3.1268,
+      "epoch": 9.035040431266847,
+      "grad_norm": 0.8683381080627441,
+      "learning_rate": 5.817377226119805e-05,
+      "loss": 3.1421,
       "step": 83800
     },
     {
-      "epoch": 9.02486277042299,
-      "grad_norm": 0.8530510067939758,
-      "learning_rate": 5.877814890636784e-05,
-      "loss": 3.1121,
+      "epoch": 9.040431266846362,
+      "grad_norm": 0.8653928637504578,
+      "learning_rate": 5.784997301672963e-05,
+      "loss": 3.1353,
       "step": 83850
     },
     {
-      "epoch": 9.030244322462599,
-      "grad_norm": 0.8723753094673157,
-      "learning_rate": 5.845490787630643e-05,
-      "loss": 3.1191,
+      "epoch": 9.045822102425875,
+      "grad_norm": 0.8411943912506104,
+      "learning_rate": 5.752617377226119e-05,
+      "loss": 3.1394,
       "step": 83900
     },
     {
-      "epoch": 9.035625874502207,
-      "grad_norm": 0.8698916435241699,
-      "learning_rate": 5.813166684624501e-05,
-      "loss": 3.1148,
+      "epoch": 9.05121293800539,
+      "grad_norm": 0.865246593952179,
+      "learning_rate": 5.7202374527792764e-05,
+      "loss": 3.1236,
       "step": 83950
     },
     {
-      "epoch": 9.041007426541814,
-      "grad_norm": 0.8469932675361633,
-      "learning_rate": 5.7808425816183596e-05,
-      "loss": 3.1246,
+      "epoch": 9.056603773584905,
+      "grad_norm": 0.8505796790122986,
+      "learning_rate": 5.6878575283324335e-05,
+      "loss": 3.1257,
       "step": 84000
     },
     {
-      "epoch": 9.041007426541814,
-      "eval_accuracy": 0.392553104913816,
-      "eval_loss": 3.317844867706299,
-      "eval_runtime": 180.533,
-      "eval_samples_per_second": 99.766,
-      "eval_steps_per_second": 6.237,
+      "epoch": 9.056603773584905,
+      "eval_accuracy": 0.39273107834405613,
+      "eval_loss": 3.316270589828491,
+      "eval_runtime": 183.9996,
+      "eval_samples_per_second": 97.886,
+      "eval_steps_per_second": 6.12,
       "step": 84000
     },
     {
-      "epoch": 9.046388978581422,
-      "grad_norm": 0.8725888133049011,
-      "learning_rate": 5.748518478612218e-05,
-      "loss": 3.1223,
+      "epoch": 9.06199460916442,
+      "grad_norm": 0.891933023929596,
+      "learning_rate": 5.6554776038855905e-05,
+      "loss": 3.1256,
       "step": 84050
     },
     {
-      "epoch": 9.051770530621031,
-      "grad_norm": 0.8994928598403931,
-      "learning_rate": 5.716194375606076e-05,
-      "loss": 3.1452,
+      "epoch": 9.067385444743936,
+      "grad_norm": 0.8847354054450989,
+      "learning_rate": 5.623097679438747e-05,
+      "loss": 3.1293,
       "step": 84100
     },
     {
-      "epoch": 9.05715208266064,
-      "grad_norm": 0.8656386137008667,
-      "learning_rate": 5.6838702725999346e-05,
-      "loss": 3.1392,
+      "epoch": 9.07277628032345,
+      "grad_norm": 0.9001868963241577,
+      "learning_rate": 5.590717754991905e-05,
+      "loss": 3.1224,
       "step": 84150
     },
     {
-      "epoch": 9.062533634700248,
-      "grad_norm": 0.878341794013977,
-      "learning_rate": 5.651546169593793e-05,
-      "loss": 3.1356,
+      "epoch": 9.078167115902964,
+      "grad_norm": 0.8722281455993652,
+      "learning_rate": 5.558337830545061e-05,
+      "loss": 3.1312,
       "step": 84200
     },
     {
-      "epoch": 9.067915186739857,
-      "grad_norm": 0.8708506226539612,
-      "learning_rate": 5.619222066587652e-05,
-      "loss": 3.1184,
+      "epoch": 9.08355795148248,
+      "grad_norm": 0.8635680079460144,
+      "learning_rate": 5.525957906098219e-05,
+      "loss": 3.1285,
       "step": 84250
     },
     {
-      "epoch": 9.073296738779463,
-      "grad_norm": 0.8545727133750916,
-      "learning_rate": 5.58689796358151e-05,
-      "loss": 3.1142,
+      "epoch": 9.088948787061994,
+      "grad_norm": 0.8610439896583557,
+      "learning_rate": 5.493577981651375e-05,
+      "loss": 3.1161,
       "step": 84300
     },
     {
-      "epoch": 9.078678290819072,
-      "grad_norm": 0.8976442217826843,
-      "learning_rate": 5.554573860575369e-05,
-      "loss": 3.107,
+      "epoch": 9.09433962264151,
+      "grad_norm": 0.8949891924858093,
+      "learning_rate": 5.4611980572045324e-05,
+      "loss": 3.1149,
       "step": 84350
     },
     {
-      "epoch": 9.08405984285868,
-      "grad_norm": 0.8379390239715576,
-      "learning_rate": 5.522249757569227e-05,
-      "loss": 3.1199,
+      "epoch": 9.099730458221025,
+      "grad_norm": 0.8697154521942139,
+      "learning_rate": 5.4288181327576895e-05,
+      "loss": 3.1339,
       "step": 84400
     },
     {
-      "epoch": 9.089441394898289,
-      "grad_norm": 0.8370739221572876,
-      "learning_rate": 5.489925654563085e-05,
-      "loss": 3.1285,
+      "epoch": 9.10512129380054,
+      "grad_norm": 0.8943808078765869,
+      "learning_rate": 5.3964382083108466e-05,
+      "loss": 3.1412,
       "step": 84450
     },
     {
-      "epoch": 9.094822946937898,
-      "grad_norm": 0.8469734191894531,
-      "learning_rate": 5.457601551556944e-05,
-      "loss": 3.1455,
+      "epoch": 9.110512129380053,
+      "grad_norm": 0.8068343997001648,
+      "learning_rate": 5.3640582838640043e-05,
+      "loss": 3.1214,
       "step": 84500
     },
     {
-      "epoch": 9.100204498977504,
-      "grad_norm": 0.8684940934181213,
-      "learning_rate": 5.425277448550802e-05,
-      "loss": 3.1088,
+      "epoch": 9.115902964959568,
+      "grad_norm": 0.8696420788764954,
+      "learning_rate": 5.331678359417161e-05,
+      "loss": 3.1227,
       "step": 84550
     },
     {
-      "epoch": 9.105586051017113,
-      "grad_norm": 0.8705862760543823,
-      "learning_rate": 5.392953345544661e-05,
-      "loss": 3.1245,
+      "epoch": 9.121293800539084,
+      "grad_norm": 0.9139754176139832,
+      "learning_rate": 5.2992984349703185e-05,
+      "loss": 3.146,
       "step": 84600
     },
     {
-      "epoch": 9.110967603056721,
-      "grad_norm": 0.8357287645339966,
-      "learning_rate": 5.360629242538519e-05,
-      "loss": 3.1275,
+      "epoch": 9.126684636118599,
+      "grad_norm": 0.8305987119674683,
+      "learning_rate": 5.266918510523475e-05,
+      "loss": 3.125,
       "step": 84650
     },
     {
-      "epoch": 9.11634915509633,
-      "grad_norm": 0.8421687483787537,
-      "learning_rate": 5.328305139532377e-05,
-      "loss": 3.1243,
+      "epoch": 9.132075471698114,
+      "grad_norm": 0.8564366102218628,
+      "learning_rate": 5.234538586076632e-05,
+      "loss": 3.1225,
       "step": 84700
     },
     {
-      "epoch": 9.121730707135939,
-      "grad_norm": 0.8550165891647339,
-      "learning_rate": 5.295981036526236e-05,
-      "loss": 3.1221,
+      "epoch": 9.137466307277627,
+      "grad_norm": 0.8829382658004761,
+      "learning_rate": 5.202158661629789e-05,
+      "loss": 3.121,
       "step": 84750
     },
     {
-      "epoch": 9.127112259175545,
-      "grad_norm": 0.8496278524398804,
-      "learning_rate": 5.263656933520094e-05,
-      "loss": 3.1369,
+      "epoch": 9.142857142857142,
+      "grad_norm": 0.9046168327331543,
+      "learning_rate": 5.169778737182946e-05,
+      "loss": 3.1215,
       "step": 84800
     },
     {
-      "epoch": 9.132493811215154,
-      "grad_norm": 0.8498865365982056,
-      "learning_rate": 5.231332830513953e-05,
-      "loss": 3.1421,
+      "epoch": 9.148247978436657,
+      "grad_norm": 0.8643724322319031,
+      "learning_rate": 5.1373988127361026e-05,
+      "loss": 3.1284,
       "step": 84850
     },
     {
-      "epoch": 9.137875363254762,
-      "grad_norm": 0.905013382434845,
-      "learning_rate": 5.199008727507811e-05,
-      "loss": 3.1319,
+      "epoch": 9.153638814016173,
+      "grad_norm": 0.8831543326377869,
+      "learning_rate": 5.1050188882892604e-05,
+      "loss": 3.133,
       "step": 84900
     },
     {
-      "epoch": 9.143256915294371,
-      "grad_norm": 0.9200946092605591,
-      "learning_rate": 5.166684624501669e-05,
-      "loss": 3.112,
+      "epoch": 9.159029649595688,
+      "grad_norm": 0.8505508899688721,
+      "learning_rate": 5.072638963842417e-05,
+      "loss": 3.1433,
       "step": 84950
     },
     {
-      "epoch": 9.14863846733398,
-      "grad_norm": 0.847834050655365,
-      "learning_rate": 5.1343605214955286e-05,
-      "loss": 3.1312,
+      "epoch": 9.164420485175203,
+      "grad_norm": 0.8966330885887146,
+      "learning_rate": 5.0402590393955746e-05,
+      "loss": 3.121,
       "step": 85000
     },
     {
-      "epoch": 9.14863846733398,
-      "eval_accuracy": 0.39271803999751836,
-      "eval_loss": 3.315976142883301,
-      "eval_runtime": 179.964,
-      "eval_samples_per_second": 100.081,
-      "eval_steps_per_second": 6.257,
+      "epoch": 9.164420485175203,
+      "eval_accuracy": 0.3930495399582403,
+      "eval_loss": 3.3140501976013184,
+      "eval_runtime": 184.0492,
+      "eval_samples_per_second": 97.86,
+      "eval_steps_per_second": 6.118,
       "step": 85000
     },
     {
-      "epoch": 9.154020019373588,
-      "grad_norm": 0.8495249152183533,
-      "learning_rate": 5.102036418489387e-05,
-      "loss": 3.1226,
+      "epoch": 9.169811320754716,
+      "grad_norm": 0.8646600246429443,
+      "learning_rate": 5.007879114948731e-05,
+      "loss": 3.1231,
       "step": 85050
     },
     {
-      "epoch": 9.159401571413195,
-      "grad_norm": 0.9137347936630249,
-      "learning_rate": 5.0697123154832455e-05,
-      "loss": 3.1358,
+      "epoch": 9.175202156334231,
+      "grad_norm": 0.866457462310791,
+      "learning_rate": 4.975499190501889e-05,
+      "loss": 3.1493,
       "step": 85100
     },
     {
-      "epoch": 9.164783123452803,
-      "grad_norm": 0.8848801851272583,
-      "learning_rate": 5.0373882124771036e-05,
-      "loss": 3.1319,
+      "epoch": 9.180592991913747,
+      "grad_norm": 0.8704493641853333,
+      "learning_rate": 4.943119266055045e-05,
+      "loss": 3.125,
       "step": 85150
     },
     {
-      "epoch": 9.170164675492412,
-      "grad_norm": 0.8463972806930542,
-      "learning_rate": 5.005064109470962e-05,
-      "loss": 3.1284,
+      "epoch": 9.185983827493262,
+      "grad_norm": 0.8545609712600708,
+      "learning_rate": 4.910739341608202e-05,
+      "loss": 3.1436,
       "step": 85200
     },
     {
-      "epoch": 9.17554622753202,
-      "grad_norm": 0.8625044822692871,
-      "learning_rate": 4.9727400064648205e-05,
-      "loss": 3.1187,
+      "epoch": 9.191374663072777,
+      "grad_norm": 0.8586364984512329,
+      "learning_rate": 4.87835941716136e-05,
+      "loss": 3.1261,
       "step": 85250
     },
     {
-      "epoch": 9.180927779571629,
-      "grad_norm": 0.8781099319458008,
-      "learning_rate": 4.9404159034586786e-05,
-      "loss": 3.1446,
+      "epoch": 9.19676549865229,
+      "grad_norm": 0.8791233897209167,
+      "learning_rate": 4.8459794927145164e-05,
+      "loss": 3.1233,
       "step": 85300
     },
     {
-      "epoch": 9.186309331611236,
-      "grad_norm": 0.877238392829895,
-      "learning_rate": 4.9080918004525374e-05,
-      "loss": 3.1407,
+      "epoch": 9.202156334231805,
+      "grad_norm": 0.9160505533218384,
+      "learning_rate": 4.813599568267674e-05,
+      "loss": 3.1356,
       "step": 85350
     },
     {
-      "epoch": 9.191690883650844,
-      "grad_norm": 0.9211892485618591,
-      "learning_rate": 4.8757676974463955e-05,
-      "loss": 3.1393,
+      "epoch": 9.20754716981132,
+      "grad_norm": 0.9123727083206177,
+      "learning_rate": 4.7812196438208306e-05,
+      "loss": 3.11,
       "step": 85400
     },
     {
-      "epoch": 9.197072435690453,
-      "grad_norm": 0.8986982107162476,
-      "learning_rate": 4.8434435944402536e-05,
-      "loss": 3.1425,
+      "epoch": 9.212938005390836,
+      "grad_norm": 0.9233217835426331,
+      "learning_rate": 4.748839719373988e-05,
+      "loss": 3.1294,
       "step": 85450
     },
     {
-      "epoch": 9.202453987730062,
-      "grad_norm": 0.8680441975593567,
-      "learning_rate": 4.8111194914341124e-05,
-      "loss": 3.1271,
+      "epoch": 9.21832884097035,
+      "grad_norm": 0.8824030756950378,
+      "learning_rate": 4.716459794927145e-05,
+      "loss": 3.1502,
       "step": 85500
     },
     {
-      "epoch": 9.20783553976967,
-      "grad_norm": 0.8860616683959961,
-      "learning_rate": 4.7787953884279705e-05,
-      "loss": 3.1492,
+      "epoch": 9.223719676549866,
+      "grad_norm": 0.9348830580711365,
+      "learning_rate": 4.684079870480302e-05,
+      "loss": 3.13,
       "step": 85550
     },
     {
-      "epoch": 9.213217091809279,
-      "grad_norm": 0.895710289478302,
-      "learning_rate": 4.746471285421829e-05,
-      "loss": 3.1334,
+      "epoch": 9.22911051212938,
+      "grad_norm": 0.8730877637863159,
+      "learning_rate": 4.651699946033459e-05,
+      "loss": 3.1421,
       "step": 85600
     },
     {
-      "epoch": 9.218598643848885,
-      "grad_norm": 0.8937450051307678,
-      "learning_rate": 4.7141471824156874e-05,
-      "loss": 3.1275,
+      "epoch": 9.234501347708894,
+      "grad_norm": 0.9030144810676575,
+      "learning_rate": 4.619320021586616e-05,
+      "loss": 3.1409,
       "step": 85650
     },
     {
-      "epoch": 9.223980195888494,
-      "grad_norm": 0.8729709386825562,
-      "learning_rate": 4.6818230794095455e-05,
-      "loss": 3.1199,
+      "epoch": 9.23989218328841,
+      "grad_norm": 0.8537077903747559,
+      "learning_rate": 4.5869400971397725e-05,
+      "loss": 3.1267,
       "step": 85700
     },
     {
-      "epoch": 9.229361747928102,
-      "grad_norm": 0.878231942653656,
-      "learning_rate": 4.649498976403404e-05,
-      "loss": 3.1359,
+      "epoch": 9.245283018867925,
+      "grad_norm": 0.8514314889907837,
+      "learning_rate": 4.55456017269293e-05,
+      "loss": 3.1348,
       "step": 85750
     },
     {
-      "epoch": 9.234743299967711,
-      "grad_norm": 0.8926569223403931,
-      "learning_rate": 4.6171748733972624e-05,
-      "loss": 3.1282,
+      "epoch": 9.25067385444744,
+      "grad_norm": 0.8366344571113586,
+      "learning_rate": 4.5221802482460866e-05,
+      "loss": 3.1373,
       "step": 85800
     },
     {
-      "epoch": 9.24012485200732,
-      "grad_norm": 0.945385754108429,
-      "learning_rate": 4.584850770391122e-05,
-      "loss": 3.1277,
+      "epoch": 9.256064690026955,
+      "grad_norm": 0.8725617527961731,
+      "learning_rate": 4.4898003237992444e-05,
+      "loss": 3.1356,
       "step": 85850
     },
     {
-      "epoch": 9.245506404046926,
-      "grad_norm": 0.8847855925559998,
-      "learning_rate": 4.55252666738498e-05,
-      "loss": 3.1344,
+      "epoch": 9.261455525606468,
+      "grad_norm": 0.9283278584480286,
+      "learning_rate": 4.457420399352401e-05,
+      "loss": 3.1438,
       "step": 85900
     },
     {
-      "epoch": 9.250887956086535,
-      "grad_norm": 0.8700735569000244,
-      "learning_rate": 4.520202564378838e-05,
-      "loss": 3.1481,
+      "epoch": 9.266846361185983,
+      "grad_norm": 0.9082370400428772,
+      "learning_rate": 4.425040474905558e-05,
+      "loss": 3.1311,
       "step": 85950
     },
     {
-      "epoch": 9.256269508126143,
-      "grad_norm": 0.8868721127510071,
-      "learning_rate": 4.487878461372697e-05,
-      "loss": 3.1217,
+      "epoch": 9.272237196765499,
+      "grad_norm": 0.8705192804336548,
+      "learning_rate": 4.392660550458716e-05,
+      "loss": 3.1224,
       "step": 86000
     },
     {
-      "epoch": 9.256269508126143,
-      "eval_accuracy": 0.39288764715539676,
-      "eval_loss": 3.315528392791748,
-      "eval_runtime": 180.0796,
-      "eval_samples_per_second": 100.017,
-      "eval_steps_per_second": 6.253,
+      "epoch": 9.272237196765499,
+      "eval_accuracy": 0.3933823437536161,
+      "eval_loss": 3.3110241889953613,
+      "eval_runtime": 184.1328,
+      "eval_samples_per_second": 97.815,
+      "eval_steps_per_second": 6.115,
       "step": 86000
     },
     {
-      "epoch": 9.261651060165752,
-      "grad_norm": 0.8628836274147034,
-      "learning_rate": 4.455554358366555e-05,
-      "loss": 3.14,
+      "epoch": 9.277628032345014,
+      "grad_norm": 0.8383573889732361,
+      "learning_rate": 4.360280626011872e-05,
+      "loss": 3.1065,
       "step": 86050
     },
     {
-      "epoch": 9.26703261220536,
-      "grad_norm": 0.8455495834350586,
-      "learning_rate": 4.423230255360414e-05,
-      "loss": 3.1434,
+      "epoch": 9.283018867924529,
+      "grad_norm": 0.8710601329803467,
+      "learning_rate": 4.32790070156503e-05,
+      "loss": 3.1207,
       "step": 86100
     },
     {
-      "epoch": 9.272414164244967,
-      "grad_norm": 0.8857659101486206,
-      "learning_rate": 4.390906152354272e-05,
-      "loss": 3.143,
+      "epoch": 9.288409703504042,
+      "grad_norm": 0.8968700170516968,
+      "learning_rate": 4.295520777118186e-05,
+      "loss": 3.1442,
       "step": 86150
     },
     {
-      "epoch": 9.277795716284576,
-      "grad_norm": 0.8970041275024414,
-      "learning_rate": 4.35858204934813e-05,
-      "loss": 3.1309,
+      "epoch": 9.293800539083557,
+      "grad_norm": 0.8627294301986694,
+      "learning_rate": 4.263140852671344e-05,
+      "loss": 3.1209,
       "step": 86200
     },
     {
-      "epoch": 9.283177268324184,
-      "grad_norm": 0.8499646186828613,
-      "learning_rate": 4.326257946341989e-05,
-      "loss": 3.1318,
+      "epoch": 9.299191374663073,
+      "grad_norm": 0.8561044335365295,
+      "learning_rate": 4.2307609282245004e-05,
+      "loss": 3.1491,
       "step": 86250
     },
     {
-      "epoch": 9.288558820363793,
-      "grad_norm": 0.9052905440330505,
-      "learning_rate": 4.293933843335847e-05,
-      "loss": 3.1497,
+      "epoch": 9.304582210242588,
+      "grad_norm": 0.8303085565567017,
+      "learning_rate": 4.1983810037776575e-05,
+      "loss": 3.1314,
       "step": 86300
     },
     {
-      "epoch": 9.293940372403402,
-      "grad_norm": 0.8457171320915222,
-      "learning_rate": 4.261609740329706e-05,
-      "loss": 3.1347,
+      "epoch": 9.309973045822103,
+      "grad_norm": 0.8662682771682739,
+      "learning_rate": 4.1660010793308146e-05,
+      "loss": 3.1228,
       "step": 86350
     },
     {
-      "epoch": 9.29932192444301,
-      "grad_norm": 0.9543759226799011,
-      "learning_rate": 4.229285637323564e-05,
-      "loss": 3.1198,
+      "epoch": 9.315363881401618,
+      "grad_norm": 0.8312433362007141,
+      "learning_rate": 4.133621154883972e-05,
+      "loss": 3.1327,
       "step": 86400
     },
     {
-      "epoch": 9.304703476482617,
-      "grad_norm": 0.8748459219932556,
-      "learning_rate": 4.196961534317422e-05,
-      "loss": 3.1473,
+      "epoch": 9.320754716981131,
+      "grad_norm": 0.8557535409927368,
+      "learning_rate": 4.101241230437128e-05,
+      "loss": 3.1384,
       "step": 86450
     },
     {
-      "epoch": 9.310085028522225,
-      "grad_norm": 0.8755471110343933,
-      "learning_rate": 4.164637431311281e-05,
-      "loss": 3.1387,
+      "epoch": 9.326145552560646,
+      "grad_norm": 0.9027670621871948,
+      "learning_rate": 4.068861305990286e-05,
+      "loss": 3.1343,
       "step": 86500
     },
     {
-      "epoch": 9.315466580561834,
-      "grad_norm": 0.8886383175849915,
-      "learning_rate": 4.132313328305139e-05,
-      "loss": 3.1433,
+      "epoch": 9.331536388140162,
+      "grad_norm": 0.8745648860931396,
+      "learning_rate": 4.036481381543442e-05,
+      "loss": 3.1442,
       "step": 86550
     },
     {
-      "epoch": 9.320848132601443,
-      "grad_norm": 0.9134498238563538,
-      "learning_rate": 4.099989225298998e-05,
-      "loss": 3.1315,
+      "epoch": 9.336927223719677,
+      "grad_norm": 0.8646321296691895,
+      "learning_rate": 4.0041014570966e-05,
+      "loss": 3.1332,
       "step": 86600
     },
     {
-      "epoch": 9.326229684641051,
-      "grad_norm": 0.8910537958145142,
-      "learning_rate": 4.0676651222928564e-05,
-      "loss": 3.1212,
+      "epoch": 9.342318059299192,
+      "grad_norm": 0.884316623210907,
+      "learning_rate": 3.9717215326497565e-05,
+      "loss": 3.1216,
       "step": 86650
     },
     {
-      "epoch": 9.331611236680658,
-      "grad_norm": 0.8579962849617004,
-      "learning_rate": 4.035341019286714e-05,
-      "loss": 3.1215,
+      "epoch": 9.347708894878707,
+      "grad_norm": 0.8479300737380981,
+      "learning_rate": 3.9393416082029136e-05,
+      "loss": 3.1247,
       "step": 86700
     },
     {
-      "epoch": 9.336992788720266,
-      "grad_norm": 0.9000516533851624,
-      "learning_rate": 4.003016916280573e-05,
-      "loss": 3.144,
+      "epoch": 9.35309973045822,
+      "grad_norm": 0.8354265093803406,
+      "learning_rate": 3.9069616837560706e-05,
+      "loss": 3.1316,
       "step": 86750
     },
     {
-      "epoch": 9.342374340759875,
-      "grad_norm": 0.8634475469589233,
-      "learning_rate": 3.9706928132744314e-05,
-      "loss": 3.1411,
+      "epoch": 9.358490566037736,
+      "grad_norm": 0.8406275510787964,
+      "learning_rate": 3.874581759309228e-05,
+      "loss": 3.1421,
       "step": 86800
     },
     {
-      "epoch": 9.347755892799483,
-      "grad_norm": 0.8831536769866943,
-      "learning_rate": 3.93836871026829e-05,
-      "loss": 3.1445,
+      "epoch": 9.36388140161725,
+      "grad_norm": 0.8649464845657349,
+      "learning_rate": 3.8422018348623855e-05,
+      "loss": 3.1479,
       "step": 86850
     },
     {
-      "epoch": 9.353137444839092,
-      "grad_norm": 0.8995107412338257,
-      "learning_rate": 3.906044607262148e-05,
-      "loss": 3.1412,
+      "epoch": 9.369272237196766,
+      "grad_norm": 0.8393669724464417,
+      "learning_rate": 3.809821910415542e-05,
+      "loss": 3.1241,
       "step": 86900
     },
     {
-      "epoch": 9.3585189968787,
-      "grad_norm": 0.8746486306190491,
-      "learning_rate": 3.8737205042560064e-05,
-      "loss": 3.1454,
+      "epoch": 9.374663072776281,
+      "grad_norm": 0.917482852935791,
+      "learning_rate": 3.7780895844576364e-05,
+      "loss": 3.1339,
       "step": 86950
     },
     {
-      "epoch": 9.363900548918307,
-      "grad_norm": 0.8589686155319214,
-      "learning_rate": 3.841396401249865e-05,
-      "loss": 3.1471,
+      "epoch": 9.380053908355794,
+      "grad_norm": 0.8299861550331116,
+      "learning_rate": 3.745709660010793e-05,
+      "loss": 3.1263,
       "step": 87000
     },
     {
-      "epoch": 9.363900548918307,
-      "eval_accuracy": 0.39334594503619824,
-      "eval_loss": 3.312450885772705,
-      "eval_runtime": 180.3192,
-      "eval_samples_per_second": 99.884,
-      "eval_steps_per_second": 6.244,
+      "epoch": 9.380053908355794,
+      "eval_accuracy": 0.3934675276176627,
+      "eval_loss": 3.3092868328094482,
+      "eval_runtime": 184.1599,
+      "eval_samples_per_second": 97.801,
+      "eval_steps_per_second": 6.114,
       "step": 87000
     },
     {
-      "epoch": 9.369282100957916,
-      "grad_norm": 0.8290409445762634,
-      "learning_rate": 3.809072298243723e-05,
-      "loss": 3.1332,
+      "epoch": 9.38544474393531,
+      "grad_norm": 0.857869565486908,
+      "learning_rate": 3.71332973556395e-05,
+      "loss": 3.1252,
       "step": 87050
     },
     {
-      "epoch": 9.374663652997524,
-      "grad_norm": 0.8649762272834778,
-      "learning_rate": 3.776748195237582e-05,
-      "loss": 3.1382,
+      "epoch": 9.390835579514825,
+      "grad_norm": 0.8611093759536743,
+      "learning_rate": 3.680949811117107e-05,
+      "loss": 3.1324,
       "step": 87100
     },
     {
-      "epoch": 9.380045205037133,
-      "grad_norm": 0.9703359603881836,
-      "learning_rate": 3.74442409223144e-05,
-      "loss": 3.1272,
+      "epoch": 9.39622641509434,
+      "grad_norm": 0.894874095916748,
+      "learning_rate": 3.648569886670264e-05,
+      "loss": 3.1358,
       "step": 87150
     },
     {
-      "epoch": 9.385426757076742,
-      "grad_norm": 0.8370105028152466,
-      "learning_rate": 3.712099989225299e-05,
-      "loss": 3.1192,
+      "epoch": 9.401617250673855,
+      "grad_norm": 0.8336328864097595,
+      "learning_rate": 3.616189962223421e-05,
+      "loss": 3.1313,
       "step": 87200
     },
     {
-      "epoch": 9.390808309116348,
-      "grad_norm": 0.8700204491615295,
-      "learning_rate": 3.679775886219157e-05,
-      "loss": 3.144,
+      "epoch": 9.40700808625337,
+      "grad_norm": 0.838655412197113,
+      "learning_rate": 3.583810037776578e-05,
+      "loss": 3.1176,
       "step": 87250
     },
     {
-      "epoch": 9.396189861155957,
-      "grad_norm": 0.9093730449676514,
-      "learning_rate": 3.647451783213015e-05,
-      "loss": 3.1191,
+      "epoch": 9.412398921832883,
+      "grad_norm": 0.8827664256095886,
+      "learning_rate": 3.5514301133297354e-05,
+      "loss": 3.147,
       "step": 87300
     },
     {
-      "epoch": 9.401571413195565,
-      "grad_norm": 0.8856383562088013,
-      "learning_rate": 3.615127680206874e-05,
-      "loss": 3.1242,
+      "epoch": 9.417789757412399,
+      "grad_norm": 0.9007830023765564,
+      "learning_rate": 3.5190501888828925e-05,
+      "loss": 3.1418,
       "step": 87350
     },
     {
-      "epoch": 9.406952965235174,
-      "grad_norm": 0.8729093670845032,
-      "learning_rate": 3.582803577200732e-05,
-      "loss": 3.1429,
+      "epoch": 9.423180592991914,
+      "grad_norm": 0.8785984516143799,
+      "learning_rate": 3.4866702644360496e-05,
+      "loss": 3.1465,
       "step": 87400
     },
     {
-      "epoch": 9.412334517274783,
-      "grad_norm": 0.8521737456321716,
-      "learning_rate": 3.550479474194591e-05,
-      "loss": 3.1177,
+      "epoch": 9.428571428571429,
+      "grad_norm": 0.8638218641281128,
+      "learning_rate": 3.4542903399892067e-05,
+      "loss": 3.1377,
       "step": 87450
     },
     {
-      "epoch": 9.417716069314391,
-      "grad_norm": 0.8820858597755432,
-      "learning_rate": 3.518155371188449e-05,
-      "loss": 3.1373,
+      "epoch": 9.433962264150944,
+      "grad_norm": 0.8835382461547852,
+      "learning_rate": 3.421910415542363e-05,
+      "loss": 3.1389,
       "step": 87500
     },
     {
-      "epoch": 9.423097621353998,
-      "grad_norm": 0.8602292537689209,
-      "learning_rate": 3.485831268182308e-05,
-      "loss": 3.1505,
+      "epoch": 9.439353099730457,
+      "grad_norm": 0.8977878093719482,
+      "learning_rate": 3.38953049109552e-05,
+      "loss": 3.129,
       "step": 87550
     },
     {
-      "epoch": 9.428479173393606,
-      "grad_norm": 0.8709748387336731,
-      "learning_rate": 3.453507165176166e-05,
-      "loss": 3.1302,
+      "epoch": 9.444743935309972,
+      "grad_norm": 0.8864624500274658,
+      "learning_rate": 3.357150566648677e-05,
+      "loss": 3.1159,
       "step": 87600
     },
     {
-      "epoch": 9.433860725433215,
-      "grad_norm": 0.8324317932128906,
-      "learning_rate": 3.421183062170025e-05,
-      "loss": 3.1146,
+      "epoch": 9.450134770889488,
+      "grad_norm": 0.8690477013587952,
+      "learning_rate": 3.324770642201834e-05,
+      "loss": 3.1233,
       "step": 87650
     },
     {
-      "epoch": 9.439242277472824,
-      "grad_norm": 0.8815144300460815,
-      "learning_rate": 3.388858959163883e-05,
-      "loss": 3.1346,
+      "epoch": 9.455525606469003,
+      "grad_norm": 0.8812589645385742,
+      "learning_rate": 3.2923907177549914e-05,
+      "loss": 3.1372,
       "step": 87700
     },
     {
-      "epoch": 9.444623829512432,
-      "grad_norm": 0.866257905960083,
-      "learning_rate": 3.3565348561577416e-05,
-      "loss": 3.1255,
+      "epoch": 9.460916442048518,
+      "grad_norm": 0.9190623760223389,
+      "learning_rate": 3.2600107933081485e-05,
+      "loss": 3.1379,
       "step": 87750
     },
     {
-      "epoch": 9.450005381552039,
-      "grad_norm": 0.8640872240066528,
-      "learning_rate": 3.3242107531516e-05,
-      "loss": 3.1326,
+      "epoch": 9.466307277628033,
+      "grad_norm": 0.8483542799949646,
+      "learning_rate": 3.2276308688613056e-05,
+      "loss": 3.1353,
       "step": 87800
     },
     {
-      "epoch": 9.455386933591647,
-      "grad_norm": 0.876515805721283,
-      "learning_rate": 3.291886650145458e-05,
-      "loss": 3.1376,
+      "epoch": 9.471698113207546,
+      "grad_norm": 0.8880337476730347,
+      "learning_rate": 3.195250944414463e-05,
+      "loss": 3.1372,
       "step": 87850
     },
     {
-      "epoch": 9.460768485631256,
-      "grad_norm": 0.879612147808075,
-      "learning_rate": 3.260209029199439e-05,
-      "loss": 3.1468,
+      "epoch": 9.477088948787062,
+      "grad_norm": 0.9060431718826294,
+      "learning_rate": 3.16287101996762e-05,
+      "loss": 3.1343,
       "step": 87900
     },
     {
-      "epoch": 9.466150037670864,
-      "grad_norm": 0.8799201846122742,
-      "learning_rate": 3.227884926193298e-05,
-      "loss": 3.1336,
+      "epoch": 9.482479784366577,
+      "grad_norm": 0.9178731441497803,
+      "learning_rate": 3.130491095520777e-05,
+      "loss": 3.1319,
       "step": 87950
     },
     {
-      "epoch": 9.471531589710473,
-      "grad_norm": 0.8598930239677429,
-      "learning_rate": 3.195560823187157e-05,
-      "loss": 3.1102,
+      "epoch": 9.487870619946092,
+      "grad_norm": 0.8681443929672241,
+      "learning_rate": 3.098111171073934e-05,
+      "loss": 3.1225,
       "step": 88000
     },
     {
-      "epoch": 9.471531589710473,
-      "eval_accuracy": 0.3935834602489607,
-      "eval_loss": 3.309974431991577,
-      "eval_runtime": 179.9534,
-      "eval_samples_per_second": 100.087,
-      "eval_steps_per_second": 6.257,
+      "epoch": 9.487870619946092,
+      "eval_accuracy": 0.39372568687910986,
+      "eval_loss": 3.306718349456787,
+      "eval_runtime": 184.1331,
+      "eval_samples_per_second": 97.815,
+      "eval_steps_per_second": 6.115,
       "step": 88000
     },
     {
-      "epoch": 9.476913141750082,
-      "grad_norm": 0.8778337836265564,
-      "learning_rate": 3.163236720181014e-05,
-      "loss": 3.1379,
+      "epoch": 9.493261455525607,
+      "grad_norm": 0.8464849591255188,
+      "learning_rate": 3.065731246627091e-05,
+      "loss": 3.1258,
       "step": 88050
     },
     {
-      "epoch": 9.482294693789688,
-      "grad_norm": 0.8822025656700134,
-      "learning_rate": 3.130912617174873e-05,
-      "loss": 3.1384,
+      "epoch": 9.498652291105122,
+      "grad_norm": 0.9113153219223022,
+      "learning_rate": 3.033351322180248e-05,
+      "loss": 3.1226,
       "step": 88100
     },
     {
-      "epoch": 9.487676245829297,
-      "grad_norm": 0.927155077457428,
-      "learning_rate": 3.098588514168732e-05,
-      "loss": 3.1588,
+      "epoch": 9.504043126684635,
+      "grad_norm": 0.8584155440330505,
+      "learning_rate": 3.000971397733405e-05,
+      "loss": 3.1224,
       "step": 88150
     },
     {
-      "epoch": 9.493057797868905,
-      "grad_norm": 0.8520760536193848,
-      "learning_rate": 3.06626441116259e-05,
-      "loss": 3.1472,
+      "epoch": 9.50943396226415,
+      "grad_norm": 0.8799469470977783,
+      "learning_rate": 2.968591473286562e-05,
+      "loss": 3.1317,
       "step": 88200
     },
     {
-      "epoch": 9.498439349908514,
-      "grad_norm": 0.8926303386688232,
-      "learning_rate": 3.0339403081564487e-05,
-      "loss": 3.1344,
+      "epoch": 9.514824797843666,
+      "grad_norm": 0.908055305480957,
+      "learning_rate": 2.936211548839719e-05,
+      "loss": 3.1279,
       "step": 88250
     },
     {
-      "epoch": 9.503820901948123,
-      "grad_norm": 0.8950697779655457,
-      "learning_rate": 3.0016162051503068e-05,
-      "loss": 3.135,
+      "epoch": 9.520215633423181,
+      "grad_norm": 0.9251330494880676,
+      "learning_rate": 2.903831624392876e-05,
+      "loss": 3.1199,
       "step": 88300
     },
     {
-      "epoch": 9.50920245398773,
-      "grad_norm": 0.8820440173149109,
-      "learning_rate": 2.9692921021441652e-05,
-      "loss": 3.1264,
+      "epoch": 9.525606469002696,
+      "grad_norm": 0.8864543437957764,
+      "learning_rate": 2.8714516999460332e-05,
+      "loss": 3.1208,
       "step": 88350
     },
     {
-      "epoch": 9.514584006027338,
-      "grad_norm": 0.8579733371734619,
-      "learning_rate": 2.9369679991380237e-05,
-      "loss": 3.1423,
+      "epoch": 9.530997304582211,
+      "grad_norm": 0.9022347927093506,
+      "learning_rate": 2.83907177549919e-05,
+      "loss": 3.1273,
       "step": 88400
     },
     {
-      "epoch": 9.519965558066946,
-      "grad_norm": 0.8843384385108948,
-      "learning_rate": 2.904643896131882e-05,
-      "loss": 3.1311,
+      "epoch": 9.536388140161725,
+      "grad_norm": 0.8811318874359131,
+      "learning_rate": 2.806691851052347e-05,
+      "loss": 3.1364,
       "step": 88450
     },
     {
-      "epoch": 9.525347110106555,
-      "grad_norm": 0.8826112747192383,
-      "learning_rate": 2.8723197931257406e-05,
-      "loss": 3.1382,
+      "epoch": 9.54177897574124,
+      "grad_norm": 0.857610285282135,
+      "learning_rate": 2.774311926605504e-05,
+      "loss": 3.1315,
       "step": 88500
     },
     {
-      "epoch": 9.530728662146164,
-      "grad_norm": 0.8869061470031738,
-      "learning_rate": 2.8399956901195987e-05,
-      "loss": 3.1228,
+      "epoch": 9.547169811320755,
+      "grad_norm": 0.8949512839317322,
+      "learning_rate": 2.7419320021586613e-05,
+      "loss": 3.1453,
       "step": 88550
     },
     {
-      "epoch": 9.536110214185772,
-      "grad_norm": 0.8704985976219177,
-      "learning_rate": 2.807671587113457e-05,
-      "loss": 3.1219,
+      "epoch": 9.55256064690027,
+      "grad_norm": 0.8894604444503784,
+      "learning_rate": 2.7095520777118187e-05,
+      "loss": 3.1393,
       "step": 88600
     },
     {
-      "epoch": 9.541491766225379,
-      "grad_norm": 0.9074740409851074,
-      "learning_rate": 2.775347484107316e-05,
-      "loss": 3.1322,
+      "epoch": 9.557951482479785,
+      "grad_norm": 0.8723719716072083,
+      "learning_rate": 2.6771721532649758e-05,
+      "loss": 3.1386,
       "step": 88650
     },
     {
-      "epoch": 9.546873318264987,
-      "grad_norm": 0.8920992016792297,
-      "learning_rate": 2.7430233811011744e-05,
-      "loss": 3.1456,
+      "epoch": 9.563342318059298,
+      "grad_norm": 0.874606192111969,
+      "learning_rate": 2.6447922288181325e-05,
+      "loss": 3.133,
       "step": 88700
     },
     {
-      "epoch": 9.552254870304596,
-      "grad_norm": 0.9204345941543579,
-      "learning_rate": 2.7106992780950328e-05,
-      "loss": 3.1497,
+      "epoch": 9.568733153638814,
+      "grad_norm": 0.8843807578086853,
+      "learning_rate": 2.6124123043712896e-05,
+      "loss": 3.1263,
       "step": 88750
     },
     {
-      "epoch": 9.557636422344205,
-      "grad_norm": 0.8399198055267334,
-      "learning_rate": 2.678375175088891e-05,
-      "loss": 3.1334,
+      "epoch": 9.574123989218329,
+      "grad_norm": 0.8979360461235046,
+      "learning_rate": 2.5800323799244467e-05,
+      "loss": 3.1205,
       "step": 88800
     },
     {
-      "epoch": 9.563017974383813,
-      "grad_norm": 0.8751394748687744,
-      "learning_rate": 2.6460510720827494e-05,
-      "loss": 3.1389,
+      "epoch": 9.579514824797844,
+      "grad_norm": 0.8424070477485657,
+      "learning_rate": 2.5476524554776038e-05,
+      "loss": 3.14,
       "step": 88850
     },
     {
-      "epoch": 9.56839952642342,
-      "grad_norm": 0.8489452600479126,
-      "learning_rate": 2.6137269690766078e-05,
-      "loss": 3.1277,
+      "epoch": 9.584905660377359,
+      "grad_norm": 0.8482125401496887,
+      "learning_rate": 2.515272531030761e-05,
+      "loss": 3.1281,
       "step": 88900
     },
     {
-      "epoch": 9.573781078463028,
-      "grad_norm": 0.8947514295578003,
-      "learning_rate": 2.5814028660704663e-05,
-      "loss": 3.1329,
+      "epoch": 9.590296495956874,
+      "grad_norm": 0.8580502867698669,
+      "learning_rate": 2.4828926065839176e-05,
+      "loss": 3.1133,
       "step": 88950
     },
     {
-      "epoch": 9.579162630502637,
-      "grad_norm": 0.8745483160018921,
-      "learning_rate": 2.5490787630643247e-05,
-      "loss": 3.1177,
+      "epoch": 9.595687331536388,
+      "grad_norm": 0.888905942440033,
+      "learning_rate": 2.4505126821370747e-05,
+      "loss": 3.1254,
       "step": 89000
     },
     {
-      "epoch": 9.579162630502637,
-      "eval_accuracy": 0.3937311195235006,
-      "eval_loss": 3.307589054107666,
-      "eval_runtime": 180.1555,
-      "eval_samples_per_second": 99.975,
-      "eval_steps_per_second": 6.25,
+      "epoch": 9.595687331536388,
+      "eval_accuracy": 0.3939789567606054,
+      "eval_loss": 3.3041818141937256,
+      "eval_runtime": 184.0105,
+      "eval_samples_per_second": 97.88,
+      "eval_steps_per_second": 6.119,
       "step": 89000
     },
     {
-      "epoch": 9.584544182542245,
-      "grad_norm": 0.9029579162597656,
-      "learning_rate": 2.5167546600581828e-05,
-      "loss": 3.1263,
+      "epoch": 9.601078167115903,
+      "grad_norm": 0.885810911655426,
+      "learning_rate": 2.4181327576902318e-05,
+      "loss": 3.1368,
       "step": 89050
     },
     {
-      "epoch": 9.589925734581854,
-      "grad_norm": 0.8661422729492188,
-      "learning_rate": 2.4844305570520416e-05,
-      "loss": 3.1395,
+      "epoch": 9.606469002695418,
+      "grad_norm": 0.9069692492485046,
+      "learning_rate": 2.385752833243389e-05,
+      "loss": 3.1263,
       "step": 89100
     },
     {
-      "epoch": 9.59530728662146,
-      "grad_norm": 0.8325968980789185,
-      "learning_rate": 2.4521064540459e-05,
-      "loss": 3.1621,
+      "epoch": 9.611859838274933,
+      "grad_norm": 0.8580989837646484,
+      "learning_rate": 2.353372908796546e-05,
+      "loss": 3.1432,
       "step": 89150
     },
     {
-      "epoch": 9.60068883866107,
-      "grad_norm": 0.8808440566062927,
-      "learning_rate": 2.4197823510397585e-05,
-      "loss": 3.1332,
+      "epoch": 9.617250673854448,
+      "grad_norm": 0.8962355852127075,
+      "learning_rate": 2.3209929843497027e-05,
+      "loss": 3.1306,
       "step": 89200
     },
     {
-      "epoch": 9.606070390700678,
-      "grad_norm": 0.8744810223579407,
-      "learning_rate": 2.387458248033617e-05,
-      "loss": 3.1255,
+      "epoch": 9.622641509433961,
+      "grad_norm": 0.8877291679382324,
+      "learning_rate": 2.2886130599028598e-05,
+      "loss": 3.1432,
       "step": 89250
     },
     {
-      "epoch": 9.611451942740286,
-      "grad_norm": 0.8575035333633423,
-      "learning_rate": 2.355134145027475e-05,
-      "loss": 3.122,
+      "epoch": 9.628032345013477,
+      "grad_norm": 0.846018373966217,
+      "learning_rate": 2.256233135456017e-05,
+      "loss": 3.1269,
       "step": 89300
     },
     {
-      "epoch": 9.616833494779895,
-      "grad_norm": 0.9315617084503174,
-      "learning_rate": 2.3228100420213335e-05,
-      "loss": 3.1325,
+      "epoch": 9.633423180592992,
+      "grad_norm": 0.8430893421173096,
+      "learning_rate": 2.2238532110091743e-05,
+      "loss": 3.1462,
       "step": 89350
     },
     {
-      "epoch": 9.622215046819504,
-      "grad_norm": 0.8789585828781128,
-      "learning_rate": 2.291132421075315e-05,
-      "loss": 3.1261,
+      "epoch": 9.638814016172507,
+      "grad_norm": 0.883787989616394,
+      "learning_rate": 2.1914732865623314e-05,
+      "loss": 3.1353,
       "step": 89400
     },
     {
-      "epoch": 9.62759659885911,
-      "grad_norm": 0.8857566714286804,
-      "learning_rate": 2.2588083180691733e-05,
-      "loss": 3.1192,
+      "epoch": 9.644204851752022,
+      "grad_norm": 0.8752528429031372,
+      "learning_rate": 2.1590933621154885e-05,
+      "loss": 3.1315,
       "step": 89450
     },
     {
-      "epoch": 9.632978150898719,
-      "grad_norm": 0.911800742149353,
-      "learning_rate": 2.2264842150630318e-05,
-      "loss": 3.1527,
+      "epoch": 9.649595687331537,
+      "grad_norm": 0.9006782174110413,
+      "learning_rate": 2.1267134376686453e-05,
+      "loss": 3.132,
       "step": 89500
     },
     {
-      "epoch": 9.638359702938327,
-      "grad_norm": 0.8703020811080933,
-      "learning_rate": 2.1941601120568905e-05,
-      "loss": 3.1369,
+      "epoch": 9.65498652291105,
+      "grad_norm": 0.82303386926651,
+      "learning_rate": 2.0943335132218024e-05,
+      "loss": 3.1338,
       "step": 89550
     },
     {
-      "epoch": 9.643741254977936,
-      "grad_norm": 0.878156840801239,
-      "learning_rate": 2.1618360090507483e-05,
-      "loss": 3.1499,
+      "epoch": 9.660377358490566,
+      "grad_norm": 0.8427075743675232,
+      "learning_rate": 2.0619535887749594e-05,
+      "loss": 3.1121,
       "step": 89600
     },
     {
-      "epoch": 9.649122807017545,
-      "grad_norm": 0.950110912322998,
-      "learning_rate": 2.129511906044607e-05,
-      "loss": 3.1427,
+      "epoch": 9.66576819407008,
+      "grad_norm": 0.8588908910751343,
+      "learning_rate": 2.0295736643281165e-05,
+      "loss": 3.1284,
       "step": 89650
     },
     {
-      "epoch": 9.654504359057151,
-      "grad_norm": 0.8555633425712585,
-      "learning_rate": 2.0971878030384655e-05,
-      "loss": 3.1397,
+      "epoch": 9.671159029649596,
+      "grad_norm": 0.8975538611412048,
+      "learning_rate": 1.9971937398812736e-05,
+      "loss": 3.1318,
       "step": 89700
     },
     {
-      "epoch": 9.65988591109676,
-      "grad_norm": 0.8625248074531555,
-      "learning_rate": 2.064863700032324e-05,
-      "loss": 3.1312,
+      "epoch": 9.676549865229111,
+      "grad_norm": 0.8748254179954529,
+      "learning_rate": 1.9648138154344304e-05,
+      "loss": 3.1282,
       "step": 89750
     },
     {
-      "epoch": 9.665267463136368,
-      "grad_norm": 0.8741613626480103,
-      "learning_rate": 2.0325395970261824e-05,
-      "loss": 3.1403,
+      "epoch": 9.681940700808624,
+      "grad_norm": 0.8568217754364014,
+      "learning_rate": 1.9324338909875875e-05,
+      "loss": 3.1268,
       "step": 89800
     },
     {
-      "epoch": 9.670649015175977,
-      "grad_norm": 0.8547689914703369,
-      "learning_rate": 2.0002154940200406e-05,
-      "loss": 3.1416,
+      "epoch": 9.68733153638814,
+      "grad_norm": 0.903649091720581,
+      "learning_rate": 1.9000539665407446e-05,
+      "loss": 3.1467,
       "step": 89850
     },
     {
-      "epoch": 9.676030567215586,
-      "grad_norm": 0.8787466883659363,
-      "learning_rate": 1.967891391013899e-05,
-      "loss": 3.1432,
+      "epoch": 9.692722371967655,
+      "grad_norm": 0.8876856565475464,
+      "learning_rate": 1.8676740420939016e-05,
+      "loss": 3.1284,
       "step": 89900
     },
     {
-      "epoch": 9.681412119255192,
-      "grad_norm": 0.9100810289382935,
-      "learning_rate": 1.9355672880077575e-05,
-      "loss": 3.1335,
+      "epoch": 9.69811320754717,
+      "grad_norm": 0.8612526655197144,
+      "learning_rate": 1.8352941176470587e-05,
+      "loss": 3.1434,
       "step": 89950
     },
     {
-      "epoch": 9.6867936712948,
-      "grad_norm": 0.8840546607971191,
-      "learning_rate": 1.9032431850016162e-05,
-      "loss": 3.1193,
+      "epoch": 9.703504043126685,
+      "grad_norm": 0.875423789024353,
+      "learning_rate": 1.8029141932002158e-05,
+      "loss": 3.1331,
       "step": 90000
     },
     {
-      "epoch": 9.6867936712948,
-      "eval_accuracy": 0.3940206794695262,
-      "eval_loss": 3.3056087493896484,
-      "eval_runtime": 180.2428,
-      "eval_samples_per_second": 99.926,
-      "eval_steps_per_second": 6.247,
+      "epoch": 9.703504043126685,
+      "eval_accuracy": 0.39423505161718414,
+      "eval_loss": 3.301893949508667,
+      "eval_runtime": 183.9246,
+      "eval_samples_per_second": 97.926,
+      "eval_steps_per_second": 6.122,
       "step": 90000
     }
   ],
   "logging_steps": 50,
-  "max_steps": 92910,
+  "max_steps": 92750,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 10,
   "save_steps": 10000,
@@ -13436,7 +13436,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 7.52488129363968e+17,
+  "total_flos": 7.524622614528e+17,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null