diff --git "a/checkpoint-70000/trainer_state.json" "b/checkpoint-70000/trainer_state.json"
--- "a/checkpoint-70000/trainer_state.json"
+++ "b/checkpoint-70000/trainer_state.json"
@@ -1,7 +1,7 @@
 {
-  "best_metric": 3.343346357345581,
-  "best_model_checkpoint": "/scratch/cl5625/exceptions/models/100M_6910/checkpoint-70000",
-  "epoch": 7.534172855451512,
+  "best_metric": 3.342951774597168,
+  "best_model_checkpoint": "/scratch/cl5625/exceptions/models/100M_low_10_6910/checkpoint-70000",
+  "epoch": 7.547169811320755,
   "eval_steps": 1000,
   "global_step": 70000,
   "is_hyper_param_search": false,
@@ -9,10438 +9,10438 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.005381552039608223,
-      "grad_norm": 3.0505740642547607,
-      "learning_rate": 0.00027,
-      "loss": 8.7845,
+      "epoch": 0.005390835579514825,
+      "grad_norm": 2.2910852432250977,
+      "learning_rate": 0.000276,
+      "loss": 9.0222,
       "step": 50
     },
     {
-      "epoch": 0.010763104079216447,
-      "grad_norm": 2.040311574935913,
-      "learning_rate": 0.00057,
-      "loss": 7.0518,
+      "epoch": 0.01078167115902965,
+      "grad_norm": 2.704249382019043,
+      "learning_rate": 0.0005759999999999999,
+      "loss": 6.9445,
       "step": 100
     },
     {
-      "epoch": 0.01614465611882467,
-      "grad_norm": 0.9394263029098511,
-      "learning_rate": 0.0005997090830729447,
-      "loss": 6.5799,
+      "epoch": 0.016172506738544475,
+      "grad_norm": 1.139755129814148,
+      "learning_rate": 0.000599702104695089,
+      "loss": 6.5025,
       "step": 150
     },
     {
-      "epoch": 0.021526208158432893,
-      "grad_norm": 0.9099545478820801,
-      "learning_rate": 0.0005993858420428833,
-      "loss": 6.2625,
+      "epoch": 0.0215633423180593,
+      "grad_norm": 0.7348619699478149,
+      "learning_rate": 0.0005993783054506205,
+      "loss": 6.2267,
       "step": 200
     },
     {
-      "epoch": 0.026907760198041114,
-      "grad_norm": 1.243497610092163,
-      "learning_rate": 0.0005990626010128218,
-      "loss": 6.1281,
+      "epoch": 0.026954177897574125,
+      "grad_norm": 1.5639214515686035,
+      "learning_rate": 0.0005990545062061521,
+      "loss": 6.0776,
       "step": 250
     },
     {
-      "epoch": 0.03228931223764934,
-      "grad_norm": 1.1881513595581055,
-      "learning_rate": 0.0005987393599827604,
-      "loss": 6.0039,
+      "epoch": 0.03234501347708895,
+      "grad_norm": 2.331084966659546,
+      "learning_rate": 0.0005987307069616836,
+      "loss": 5.9368,
       "step": 300
     },
     {
-      "epoch": 0.03767086427725756,
-      "grad_norm": 1.3128442764282227,
-      "learning_rate": 0.000598416118952699,
-      "loss": 5.8946,
+      "epoch": 0.03773584905660377,
+      "grad_norm": 1.5713136196136475,
+      "learning_rate": 0.0005984069077172153,
+      "loss": 5.8663,
       "step": 350
     },
     {
-      "epoch": 0.04305241631686579,
-      "grad_norm": 1.8052157163619995,
-      "learning_rate": 0.0005980928779226376,
-      "loss": 5.8332,
+      "epoch": 0.0431266846361186,
+      "grad_norm": 1.3305473327636719,
+      "learning_rate": 0.0005980831084727469,
+      "loss": 5.8001,
       "step": 400
     },
     {
-      "epoch": 0.048433968356474004,
-      "grad_norm": 1.1714507341384888,
-      "learning_rate": 0.0005977696368925762,
-      "loss": 5.7612,
+      "epoch": 0.04851752021563342,
+      "grad_norm": 0.9089034795761108,
+      "learning_rate": 0.0005977593092282784,
+      "loss": 5.7367,
       "step": 450
     },
     {
-      "epoch": 0.05381552039608223,
-      "grad_norm": 1.0833935737609863,
-      "learning_rate": 0.0005974463958625148,
-      "loss": 5.6894,
+      "epoch": 0.05390835579514825,
+      "grad_norm": 0.9432206749916077,
+      "learning_rate": 0.00059743550998381,
+      "loss": 5.6294,
       "step": 500
     },
     {
-      "epoch": 0.05919707243569045,
-      "grad_norm": 1.421362280845642,
-      "learning_rate": 0.0005971231548324534,
-      "loss": 5.6037,
+      "epoch": 0.05929919137466307,
+      "grad_norm": 1.129724383354187,
+      "learning_rate": 0.0005971117107393416,
+      "loss": 5.5819,
       "step": 550
     },
     {
-      "epoch": 0.06457862447529868,
-      "grad_norm": 1.5539004802703857,
-      "learning_rate": 0.0005967999138023919,
-      "loss": 5.5302,
+      "epoch": 0.0646900269541779,
+      "grad_norm": 1.60132896900177,
+      "learning_rate": 0.0005967879114948732,
+      "loss": 5.5109,
       "step": 600
     },
     {
-      "epoch": 0.0699601765149069,
-      "grad_norm": 1.172881841659546,
-      "learning_rate": 0.0005964766727723304,
-      "loss": 5.4609,
+      "epoch": 0.07008086253369272,
+      "grad_norm": 1.1178468465805054,
+      "learning_rate": 0.0005964641122504047,
+      "loss": 5.446,
       "step": 650
     },
     {
-      "epoch": 0.07534172855451512,
-      "grad_norm": 0.9414631724357605,
-      "learning_rate": 0.0005961534317422691,
-      "loss": 5.419,
+      "epoch": 0.07547169811320754,
+      "grad_norm": 1.2521672248840332,
+      "learning_rate": 0.0005961403130059363,
+      "loss": 5.3701,
       "step": 700
     },
     {
-      "epoch": 0.08072328059412334,
-      "grad_norm": 1.2072681188583374,
-      "learning_rate": 0.0005958301907122077,
-      "loss": 5.335,
+      "epoch": 0.08086253369272237,
+      "grad_norm": 1.2022509574890137,
+      "learning_rate": 0.0005958165137614678,
+      "loss": 5.3038,
       "step": 750
     },
     {
-      "epoch": 0.08610483263373157,
-      "grad_norm": 1.4625303745269775,
-      "learning_rate": 0.0005955069496821463,
-      "loss": 5.289,
+      "epoch": 0.0862533692722372,
+      "grad_norm": 1.061596393585205,
+      "learning_rate": 0.0005954927145169995,
+      "loss": 5.2369,
       "step": 800
     },
     {
-      "epoch": 0.09148638467333979,
-      "grad_norm": 1.2793816328048706,
-      "learning_rate": 0.0005951837086520848,
-      "loss": 5.2324,
+      "epoch": 0.09164420485175202,
+      "grad_norm": 1.0844577550888062,
+      "learning_rate": 0.0005951689152725309,
+      "loss": 5.2157,
       "step": 850
     },
     {
-      "epoch": 0.09686793671294801,
-      "grad_norm": 1.0962399244308472,
-      "learning_rate": 0.0005948604676220235,
-      "loss": 5.2125,
+      "epoch": 0.09703504043126684,
+      "grad_norm": 1.3673590421676636,
+      "learning_rate": 0.0005948451160280626,
+      "loss": 5.1597,
       "step": 900
     },
     {
-      "epoch": 0.10224948875255624,
-      "grad_norm": 1.6166024208068848,
-      "learning_rate": 0.000594537226591962,
-      "loss": 5.131,
+      "epoch": 0.10242587601078167,
+      "grad_norm": 0.8875324130058289,
+      "learning_rate": 0.0005945213167835941,
+      "loss": 5.102,
       "step": 950
     },
     {
-      "epoch": 0.10763104079216446,
-      "grad_norm": 0.9716492891311646,
-      "learning_rate": 0.0005942139855619007,
-      "loss": 5.102,
+      "epoch": 0.1078167115902965,
+      "grad_norm": 1.167039155960083,
+      "learning_rate": 0.0005941975175391257,
+      "loss": 5.1062,
       "step": 1000
     },
     {
-      "epoch": 0.10763104079216446,
-      "eval_accuracy": 0.22419263027019473,
-      "eval_loss": 5.052553653717041,
-      "eval_runtime": 184.3461,
-      "eval_samples_per_second": 97.702,
-      "eval_steps_per_second": 6.108,
+      "epoch": 0.1078167115902965,
+      "eval_accuracy": 0.22781881174811522,
+      "eval_loss": 5.014029026031494,
+      "eval_runtime": 184.2858,
+      "eval_samples_per_second": 97.734,
+      "eval_steps_per_second": 6.11,
       "step": 1000
     },
     {
-      "epoch": 0.11301259283177269,
-      "grad_norm": 1.4451881647109985,
-      "learning_rate": 0.0005938907445318392,
-      "loss": 5.0886,
+      "epoch": 0.11320754716981132,
+      "grad_norm": 1.1135107278823853,
+      "learning_rate": 0.0005938737182946572,
+      "loss": 5.0362,
       "step": 1050
     },
     {
-      "epoch": 0.1183941448713809,
-      "grad_norm": 1.0037076473236084,
-      "learning_rate": 0.0005935675035017777,
-      "loss": 5.0381,
+      "epoch": 0.11859838274932614,
+      "grad_norm": 1.4026278257369995,
+      "learning_rate": 0.0005935499190501888,
+      "loss": 4.9994,
       "step": 1100
     },
     {
-      "epoch": 0.12377569691098914,
-      "grad_norm": 1.0812281370162964,
-      "learning_rate": 0.0005932442624717164,
-      "loss": 5.0122,
+      "epoch": 0.12398921832884097,
+      "grad_norm": 1.4854403734207153,
+      "learning_rate": 0.0005932261198057204,
+      "loss": 5.0013,
       "step": 1150
     },
     {
-      "epoch": 0.12915724895059735,
-      "grad_norm": 1.2637274265289307,
-      "learning_rate": 0.0005929210214416549,
-      "loss": 4.989,
+      "epoch": 0.1293800539083558,
+      "grad_norm": 1.0926762819290161,
+      "learning_rate": 0.000592902320561252,
+      "loss": 4.9591,
       "step": 1200
     },
     {
-      "epoch": 0.13453880099020557,
-      "grad_norm": 0.7775391340255737,
-      "learning_rate": 0.0005925977804115936,
-      "loss": 4.9312,
+      "epoch": 0.1347708894878706,
+      "grad_norm": 1.0897889137268066,
+      "learning_rate": 0.0005925785213167835,
+      "loss": 4.9362,
       "step": 1250
     },
     {
-      "epoch": 0.1399203530298138,
-      "grad_norm": 1.0451130867004395,
-      "learning_rate": 0.0005922745393815321,
-      "loss": 4.9263,
+      "epoch": 0.14016172506738545,
+      "grad_norm": 1.180755853652954,
+      "learning_rate": 0.0005922547220723151,
+      "loss": 4.8944,
       "step": 1300
     },
     {
-      "epoch": 0.14530190506942203,
-      "grad_norm": 0.9460002779960632,
-      "learning_rate": 0.0005919512983514707,
-      "loss": 4.8936,
+      "epoch": 0.14555256064690028,
+      "grad_norm": 0.8999577760696411,
+      "learning_rate": 0.0005919309228278468,
+      "loss": 4.8439,
       "step": 1350
     },
     {
-      "epoch": 0.15068345710903025,
-      "grad_norm": 1.093346118927002,
-      "learning_rate": 0.0005916280573214093,
-      "loss": 4.8726,
+      "epoch": 0.1509433962264151,
+      "grad_norm": 1.0962032079696655,
+      "learning_rate": 0.0005916071235833783,
+      "loss": 4.8462,
       "step": 1400
     },
     {
-      "epoch": 0.15606500914863847,
-      "grad_norm": 1.0611810684204102,
-      "learning_rate": 0.0005913048162913478,
-      "loss": 4.855,
+      "epoch": 0.15633423180592992,
+      "grad_norm": 0.8688578009605408,
+      "learning_rate": 0.0005912833243389097,
+      "loss": 4.8058,
       "step": 1450
     },
     {
-      "epoch": 0.16144656118824668,
-      "grad_norm": 1.073016881942749,
-      "learning_rate": 0.0005909815752612864,
-      "loss": 4.8163,
+      "epoch": 0.16172506738544473,
+      "grad_norm": 0.7583906054496765,
+      "learning_rate": 0.0005909595250944414,
+      "loss": 4.8029,
       "step": 1500
     },
     {
-      "epoch": 0.1668281132278549,
-      "grad_norm": 1.0239862203598022,
-      "learning_rate": 0.000590658334231225,
-      "loss": 4.8212,
+      "epoch": 0.16711590296495957,
+      "grad_norm": 1.1127862930297852,
+      "learning_rate": 0.000590635725849973,
+      "loss": 4.7834,
       "step": 1550
     },
     {
-      "epoch": 0.17220966526746315,
-      "grad_norm": 0.8779550790786743,
-      "learning_rate": 0.0005903350932011637,
-      "loss": 4.7843,
+      "epoch": 0.1725067385444744,
+      "grad_norm": 1.0938029289245605,
+      "learning_rate": 0.0005903119266055045,
+      "loss": 4.7643,
       "step": 1600
     },
     {
-      "epoch": 0.17759121730707136,
-      "grad_norm": 1.2513017654418945,
-      "learning_rate": 0.0005900118521711022,
-      "loss": 4.7629,
+      "epoch": 0.1778975741239892,
+      "grad_norm": 1.2191473245620728,
+      "learning_rate": 0.0005899881273610361,
+      "loss": 4.7331,
       "step": 1650
     },
     {
-      "epoch": 0.18297276934667958,
-      "grad_norm": 1.0985004901885986,
-      "learning_rate": 0.0005896886111410408,
-      "loss": 4.7096,
+      "epoch": 0.18328840970350405,
+      "grad_norm": 1.192352056503296,
+      "learning_rate": 0.0005896643281165677,
+      "loss": 4.7143,
       "step": 1700
     },
     {
-      "epoch": 0.1883543213862878,
-      "grad_norm": 0.8831003904342651,
-      "learning_rate": 0.0005893653701109793,
-      "loss": 4.7202,
+      "epoch": 0.18867924528301888,
+      "grad_norm": 0.7933173179626465,
+      "learning_rate": 0.0005893405288720993,
+      "loss": 4.6626,
       "step": 1750
     },
     {
-      "epoch": 0.19373587342589602,
-      "grad_norm": 0.9399479031562805,
-      "learning_rate": 0.000589042129080918,
-      "loss": 4.7024,
+      "epoch": 0.1940700808625337,
+      "grad_norm": 0.7938002347946167,
+      "learning_rate": 0.0005890167296276308,
+      "loss": 4.6891,
       "step": 1800
     },
     {
-      "epoch": 0.19911742546550426,
-      "grad_norm": 0.9532297253608704,
-      "learning_rate": 0.0005887188880508566,
-      "loss": 4.6769,
+      "epoch": 0.19946091644204852,
+      "grad_norm": 1.3926692008972168,
+      "learning_rate": 0.0005886929303831624,
+      "loss": 4.6364,
       "step": 1850
     },
     {
-      "epoch": 0.20449897750511248,
-      "grad_norm": 1.072829008102417,
-      "learning_rate": 0.0005883956470207951,
-      "loss": 4.6437,
+      "epoch": 0.20485175202156333,
+      "grad_norm": 0.6863223314285278,
+      "learning_rate": 0.0005883691311386939,
+      "loss": 4.6131,
       "step": 1900
     },
     {
-      "epoch": 0.2098805295447207,
-      "grad_norm": 0.9652490615844727,
-      "learning_rate": 0.0005880724059907337,
-      "loss": 4.6334,
+      "epoch": 0.21024258760107817,
+      "grad_norm": 0.6248738765716553,
+      "learning_rate": 0.0005880453318942256,
+      "loss": 4.5923,
       "step": 1950
     },
     {
-      "epoch": 0.2152620815843289,
-      "grad_norm": 0.8171770572662354,
-      "learning_rate": 0.0005877491649606723,
-      "loss": 4.6223,
+      "epoch": 0.215633423180593,
+      "grad_norm": 1.0679149627685547,
+      "learning_rate": 0.0005877215326497571,
+      "loss": 4.576,
       "step": 2000
     },
     {
-      "epoch": 0.2152620815843289,
-      "eval_accuracy": 0.266313226066132,
-      "eval_loss": 4.544240951538086,
-      "eval_runtime": 183.7091,
-      "eval_samples_per_second": 98.041,
-      "eval_steps_per_second": 6.129,
+      "epoch": 0.215633423180593,
+      "eval_accuracy": 0.2711211163519228,
+      "eval_loss": 4.5000410079956055,
+      "eval_runtime": 183.8027,
+      "eval_samples_per_second": 97.991,
+      "eval_steps_per_second": 6.126,
       "step": 2000
     },
     {
-      "epoch": 0.22064363362393713,
-      "grad_norm": 1.0544344186782837,
-      "learning_rate": 0.0005874259239306109,
-      "loss": 4.5787,
+      "epoch": 0.2210242587601078,
+      "grad_norm": 0.7934781908988953,
+      "learning_rate": 0.0005873977334052887,
+      "loss": 4.559,
       "step": 2050
     },
     {
-      "epoch": 0.22602518566354537,
-      "grad_norm": 1.0536537170410156,
-      "learning_rate": 0.0005871026829005494,
-      "loss": 4.5656,
+      "epoch": 0.22641509433962265,
+      "grad_norm": 1.0624268054962158,
+      "learning_rate": 0.0005870739341608202,
+      "loss": 4.5441,
       "step": 2100
     },
     {
-      "epoch": 0.2314067377031536,
-      "grad_norm": 0.8931703567504883,
-      "learning_rate": 0.0005867794418704881,
-      "loss": 4.5716,
+      "epoch": 0.23180592991913745,
+      "grad_norm": 0.8633057475090027,
+      "learning_rate": 0.0005867501349163519,
+      "loss": 4.5235,
       "step": 2150
     },
     {
-      "epoch": 0.2367882897427618,
-      "grad_norm": 1.0160928964614868,
-      "learning_rate": 0.0005864562008404266,
-      "loss": 4.5331,
+      "epoch": 0.2371967654986523,
+      "grad_norm": 0.9386199712753296,
+      "learning_rate": 0.0005864263356718833,
+      "loss": 4.5135,
       "step": 2200
     },
     {
-      "epoch": 0.24216984178237003,
-      "grad_norm": 0.7671811580657959,
-      "learning_rate": 0.0005861329598103653,
-      "loss": 4.5273,
+      "epoch": 0.24258760107816713,
+      "grad_norm": 0.9182597994804382,
+      "learning_rate": 0.000586102536427415,
+      "loss": 4.4893,
       "step": 2250
     },
     {
-      "epoch": 0.24755139382197827,
-      "grad_norm": 0.8952691555023193,
-      "learning_rate": 0.0005858097187803038,
-      "loss": 4.5076,
+      "epoch": 0.24797843665768193,
+      "grad_norm": 0.7566362619400024,
+      "learning_rate": 0.0005857787371829465,
+      "loss": 4.4831,
       "step": 2300
     },
     {
-      "epoch": 0.2529329458615865,
-      "grad_norm": 0.9314284920692444,
-      "learning_rate": 0.0005854864777502423,
-      "loss": 4.464,
+      "epoch": 0.25336927223719674,
+      "grad_norm": 1.053139090538025,
+      "learning_rate": 0.0005854549379384781,
+      "loss": 4.4653,
       "step": 2350
     },
     {
-      "epoch": 0.2583144979011947,
-      "grad_norm": 1.0489318370819092,
-      "learning_rate": 0.000585163236720181,
-      "loss": 4.4852,
+      "epoch": 0.2587601078167116,
+      "grad_norm": 0.931647002696991,
+      "learning_rate": 0.0005851311386940096,
+      "loss": 4.4438,
       "step": 2400
     },
     {
-      "epoch": 0.2636960499408029,
-      "grad_norm": 0.8147399425506592,
-      "learning_rate": 0.0005848399956901195,
-      "loss": 4.4577,
+      "epoch": 0.2641509433962264,
+      "grad_norm": 0.8096271753311157,
+      "learning_rate": 0.0005848073394495412,
+      "loss": 4.432,
       "step": 2450
     },
     {
-      "epoch": 0.26907760198041114,
-      "grad_norm": 0.7356587052345276,
-      "learning_rate": 0.0005845167546600582,
-      "loss": 4.4486,
+      "epoch": 0.2695417789757412,
+      "grad_norm": 0.8853194117546082,
+      "learning_rate": 0.0005844835402050728,
+      "loss": 4.4137,
       "step": 2500
     },
     {
-      "epoch": 0.27445915402001936,
-      "grad_norm": 0.8844388127326965,
-      "learning_rate": 0.0005841935136299967,
-      "loss": 4.4172,
+      "epoch": 0.2749326145552561,
+      "grad_norm": 0.7149171829223633,
+      "learning_rate": 0.0005841597409606044,
+      "loss": 4.382,
       "step": 2550
     },
     {
-      "epoch": 0.2798407060596276,
-      "grad_norm": 0.8647958636283875,
-      "learning_rate": 0.0005838702725999353,
-      "loss": 4.4164,
+      "epoch": 0.2803234501347709,
+      "grad_norm": 0.9149443507194519,
+      "learning_rate": 0.000583835941716136,
+      "loss": 4.3951,
       "step": 2600
     },
     {
-      "epoch": 0.2852222580992358,
-      "grad_norm": 0.8443241119384766,
-      "learning_rate": 0.0005835470315698739,
-      "loss": 4.4178,
+      "epoch": 0.2857142857142857,
+      "grad_norm": 0.7679409980773926,
+      "learning_rate": 0.0005835121424716675,
+      "loss": 4.3752,
       "step": 2650
     },
     {
-      "epoch": 0.29060381013884407,
-      "grad_norm": 0.8517235517501831,
-      "learning_rate": 0.0005832237905398125,
-      "loss": 4.3873,
+      "epoch": 0.29110512129380056,
+      "grad_norm": 0.9908820390701294,
+      "learning_rate": 0.0005831883432271992,
+      "loss": 4.3636,
       "step": 2700
     },
     {
-      "epoch": 0.2959853621784523,
-      "grad_norm": 0.8355386853218079,
-      "learning_rate": 0.0005829005495097511,
-      "loss": 4.3637,
+      "epoch": 0.29649595687331537,
+      "grad_norm": 0.911641538143158,
+      "learning_rate": 0.0005828645439827307,
+      "loss": 4.3599,
       "step": 2750
     },
     {
-      "epoch": 0.3013669142180605,
-      "grad_norm": 0.8551438450813293,
-      "learning_rate": 0.0005825773084796896,
-      "loss": 4.38,
+      "epoch": 0.3018867924528302,
+      "grad_norm": 0.7766907811164856,
+      "learning_rate": 0.0005825407447382622,
+      "loss": 4.3141,
       "step": 2800
     },
     {
-      "epoch": 0.3067484662576687,
-      "grad_norm": 0.7978211045265198,
-      "learning_rate": 0.0005822540674496282,
-      "loss": 4.3506,
+      "epoch": 0.30727762803234504,
+      "grad_norm": 0.873397707939148,
+      "learning_rate": 0.0005822169454937938,
+      "loss": 4.3448,
       "step": 2850
     },
     {
-      "epoch": 0.31213001829727693,
-      "grad_norm": 0.9600444436073303,
-      "learning_rate": 0.0005819308264195668,
-      "loss": 4.3455,
+      "epoch": 0.31266846361185985,
+      "grad_norm": 0.7710452079772949,
+      "learning_rate": 0.0005818931462493254,
+      "loss": 4.3085,
       "step": 2900
     },
     {
-      "epoch": 0.31751157033688515,
-      "grad_norm": 0.7661147713661194,
-      "learning_rate": 0.0005816075853895054,
-      "loss": 4.3498,
+      "epoch": 0.31805929919137466,
+      "grad_norm": 0.7592593431472778,
+      "learning_rate": 0.0005815693470048569,
+      "loss": 4.317,
       "step": 2950
     },
     {
-      "epoch": 0.32289312237649337,
-      "grad_norm": 0.7429329752922058,
-      "learning_rate": 0.000581284344359444,
-      "loss": 4.323,
+      "epoch": 0.32345013477088946,
+      "grad_norm": 0.761400580406189,
+      "learning_rate": 0.0005812455477603885,
+      "loss": 4.3061,
       "step": 3000
     },
     {
-      "epoch": 0.32289312237649337,
-      "eval_accuracy": 0.29698419784995495,
-      "eval_loss": 4.252286911010742,
-      "eval_runtime": 183.6193,
-      "eval_samples_per_second": 98.089,
-      "eval_steps_per_second": 6.132,
+      "epoch": 0.32345013477088946,
+      "eval_accuracy": 0.2992687334691422,
+      "eval_loss": 4.229072093963623,
+      "eval_runtime": 183.7234,
+      "eval_samples_per_second": 98.033,
+      "eval_steps_per_second": 6.129,
       "step": 3000
     },
     {
-      "epoch": 0.3282746744161016,
-      "grad_norm": 0.6843370795249939,
-      "learning_rate": 0.0005809611033293825,
-      "loss": 4.2999,
+      "epoch": 0.3288409703504043,
+      "grad_norm": 0.6839197874069214,
+      "learning_rate": 0.0005809217485159201,
+      "loss": 4.3048,
       "step": 3050
     },
     {
-      "epoch": 0.3336562264557098,
-      "grad_norm": 0.9368823170661926,
-      "learning_rate": 0.0005806378622993211,
-      "loss": 4.2817,
+      "epoch": 0.33423180592991913,
+      "grad_norm": 0.8009073138237,
+      "learning_rate": 0.0005805979492714517,
+      "loss": 4.2831,
       "step": 3100
     },
     {
-      "epoch": 0.3390377784953181,
-      "grad_norm": 0.6842812299728394,
-      "learning_rate": 0.0005803146212692597,
-      "loss": 4.3011,
+      "epoch": 0.33962264150943394,
+      "grad_norm": 0.8542678952217102,
+      "learning_rate": 0.0005802741500269832,
+      "loss": 4.2646,
       "step": 3150
     },
     {
-      "epoch": 0.3444193305349263,
-      "grad_norm": 0.7644530534744263,
-      "learning_rate": 0.0005799913802391983,
-      "loss": 4.2858,
+      "epoch": 0.3450134770889488,
+      "grad_norm": 0.6708468794822693,
+      "learning_rate": 0.0005799503507825148,
+      "loss": 4.2665,
       "step": 3200
     },
     {
-      "epoch": 0.3498008825745345,
-      "grad_norm": 0.8856046795845032,
-      "learning_rate": 0.0005796681392091369,
-      "loss": 4.2867,
+      "epoch": 0.3504043126684636,
+      "grad_norm": 0.8015856742858887,
+      "learning_rate": 0.0005796265515380463,
+      "loss": 4.2619,
       "step": 3250
     },
     {
-      "epoch": 0.35518243461414273,
-      "grad_norm": 0.7749567031860352,
-      "learning_rate": 0.0005793448981790755,
-      "loss": 4.2697,
+      "epoch": 0.3557951482479784,
+      "grad_norm": 0.8879486322402954,
+      "learning_rate": 0.000579302752293578,
+      "loss": 4.2621,
       "step": 3300
     },
     {
-      "epoch": 0.36056398665375095,
-      "grad_norm": 0.8580338358879089,
-      "learning_rate": 0.0005790216571490141,
-      "loss": 4.279,
+      "epoch": 0.3611859838274933,
+      "grad_norm": 0.8756187558174133,
+      "learning_rate": 0.0005789789530491095,
+      "loss": 4.2342,
       "step": 3350
     },
     {
-      "epoch": 0.36594553869335916,
-      "grad_norm": 0.7386928200721741,
-      "learning_rate": 0.0005786984161189527,
-      "loss": 4.277,
+      "epoch": 0.3665768194070081,
+      "grad_norm": 0.8013753890991211,
+      "learning_rate": 0.0005786551538046411,
+      "loss": 4.2267,
       "step": 3400
     },
     {
-      "epoch": 0.3713270907329674,
-      "grad_norm": 0.8982140421867371,
-      "learning_rate": 0.0005783751750888912,
-      "loss": 4.2585,
+      "epoch": 0.3719676549865229,
+      "grad_norm": 0.6387489438056946,
+      "learning_rate": 0.0005783313545601726,
+      "loss": 4.2194,
       "step": 3450
     },
     {
-      "epoch": 0.3767086427725756,
-      "grad_norm": 0.8601508140563965,
-      "learning_rate": 0.0005780519340588297,
-      "loss": 4.2415,
+      "epoch": 0.37735849056603776,
+      "grad_norm": 1.016513705253601,
+      "learning_rate": 0.0005780075553157043,
+      "loss": 4.2248,
       "step": 3500
     },
     {
-      "epoch": 0.3820901948121838,
-      "grad_norm": 0.8936505913734436,
-      "learning_rate": 0.0005777286930287684,
-      "loss": 4.2185,
+      "epoch": 0.38274932614555257,
+      "grad_norm": 0.75726318359375,
+      "learning_rate": 0.0005776837560712357,
+      "loss": 4.216,
       "step": 3550
     },
     {
-      "epoch": 0.38747174685179203,
-      "grad_norm": 0.7999247312545776,
-      "learning_rate": 0.000577405451998707,
-      "loss": 4.2148,
+      "epoch": 0.3881401617250674,
+      "grad_norm": 0.9124229550361633,
+      "learning_rate": 0.0005773599568267673,
+      "loss": 4.206,
       "step": 3600
     },
     {
-      "epoch": 0.3928532988914003,
-      "grad_norm": 0.75700843334198,
-      "learning_rate": 0.0005770822109686456,
-      "loss": 4.2042,
+      "epoch": 0.3935309973045822,
+      "grad_norm": 0.7262310981750488,
+      "learning_rate": 0.0005770361575822989,
+      "loss": 4.1951,
       "step": 3650
     },
     {
-      "epoch": 0.3982348509310085,
-      "grad_norm": 0.712009608745575,
-      "learning_rate": 0.0005767589699385841,
-      "loss": 4.1869,
+      "epoch": 0.39892183288409705,
+      "grad_norm": 0.7197564840316772,
+      "learning_rate": 0.0005767123583378305,
+      "loss": 4.1822,
       "step": 3700
     },
     {
-      "epoch": 0.40361640297061674,
-      "grad_norm": 0.758716344833374,
-      "learning_rate": 0.0005764357289085228,
-      "loss": 4.1999,
+      "epoch": 0.40431266846361186,
+      "grad_norm": 0.7607842087745667,
+      "learning_rate": 0.000576388559093362,
+      "loss": 4.1864,
       "step": 3750
     },
     {
-      "epoch": 0.40899795501022496,
-      "grad_norm": 0.7437872886657715,
-      "learning_rate": 0.0005761124878784613,
-      "loss": 4.2026,
+      "epoch": 0.40970350404312667,
+      "grad_norm": 0.6156255602836609,
+      "learning_rate": 0.0005760647598488936,
+      "loss": 4.1517,
       "step": 3800
     },
     {
-      "epoch": 0.4143795070498332,
-      "grad_norm": 0.6111822128295898,
-      "learning_rate": 0.0005757892468484,
-      "loss": 4.1922,
+      "epoch": 0.41509433962264153,
+      "grad_norm": 0.5746268630027771,
+      "learning_rate": 0.0005757409606044253,
+      "loss": 4.1628,
       "step": 3850
     },
     {
-      "epoch": 0.4197610590894414,
-      "grad_norm": 0.6579766273498535,
-      "learning_rate": 0.0005754660058183385,
-      "loss": 4.1893,
+      "epoch": 0.42048517520215634,
+      "grad_norm": 0.7341263294219971,
+      "learning_rate": 0.0005754171613599568,
+      "loss": 4.1786,
       "step": 3900
     },
     {
-      "epoch": 0.4251426111290496,
-      "grad_norm": 1.0287907123565674,
-      "learning_rate": 0.000575142764788277,
-      "loss": 4.1699,
+      "epoch": 0.42587601078167114,
+      "grad_norm": 1.2203543186187744,
+      "learning_rate": 0.0005750933621154884,
+      "loss": 4.1469,
       "step": 3950
     },
     {
-      "epoch": 0.4305241631686578,
-      "grad_norm": 0.7824015617370605,
-      "learning_rate": 0.0005748195237582157,
-      "loss": 4.1732,
+      "epoch": 0.431266846361186,
+      "grad_norm": 0.7009409070014954,
+      "learning_rate": 0.0005747695628710199,
+      "loss": 4.1509,
       "step": 4000
     },
     {
-      "epoch": 0.4305241631686578,
-      "eval_accuracy": 0.31110440119166144,
-      "eval_loss": 4.102444648742676,
-      "eval_runtime": 183.6988,
-      "eval_samples_per_second": 98.046,
-      "eval_steps_per_second": 6.13,
+      "epoch": 0.431266846361186,
+      "eval_accuracy": 0.3124032378125957,
+      "eval_loss": 4.086333751678467,
+      "eval_runtime": 183.7135,
+      "eval_samples_per_second": 98.039,
+      "eval_steps_per_second": 6.129,
       "step": 4000
     },
     {
-      "epoch": 0.43590571520826604,
-      "grad_norm": 0.7554957866668701,
-      "learning_rate": 0.0005744962827281542,
-      "loss": 4.1376,
+      "epoch": 0.4366576819407008,
+      "grad_norm": 0.7461296319961548,
+      "learning_rate": 0.0005744457636265515,
+      "loss": 4.1549,
       "step": 4050
     },
     {
-      "epoch": 0.44128726724787426,
-      "grad_norm": 0.8211575746536255,
-      "learning_rate": 0.0005741730416980928,
-      "loss": 4.1503,
+      "epoch": 0.4420485175202156,
+      "grad_norm": 0.7512764930725098,
+      "learning_rate": 0.0005741219643820831,
+      "loss": 4.1532,
       "step": 4100
     },
     {
-      "epoch": 0.44666881928748253,
-      "grad_norm": 0.697628915309906,
-      "learning_rate": 0.0005738498006680314,
-      "loss": 4.1588,
+      "epoch": 0.4474393530997305,
+      "grad_norm": 0.6471153497695923,
+      "learning_rate": 0.0005737981651376146,
+      "loss": 4.1336,
       "step": 4150
     },
     {
-      "epoch": 0.45205037132709075,
-      "grad_norm": 0.7968320250511169,
-      "learning_rate": 0.00057352655963797,
-      "loss": 4.1596,
+      "epoch": 0.4528301886792453,
+      "grad_norm": 0.7613458037376404,
+      "learning_rate": 0.0005734743658931462,
+      "loss": 4.1392,
       "step": 4200
     },
     {
-      "epoch": 0.45743192336669897,
-      "grad_norm": 0.7107890844345093,
-      "learning_rate": 0.0005732033186079086,
-      "loss": 4.1658,
+      "epoch": 0.4582210242587601,
+      "grad_norm": 0.5763305425643921,
+      "learning_rate": 0.0005731505666486778,
+      "loss": 4.1258,
       "step": 4250
     },
     {
-      "epoch": 0.4628134754063072,
-      "grad_norm": 0.7855952978134155,
-      "learning_rate": 0.0005728800775778471,
-      "loss": 4.1222,
+      "epoch": 0.4636118598382749,
+      "grad_norm": 0.8002066612243652,
+      "learning_rate": 0.0005728267674042093,
+      "loss": 4.1313,
       "step": 4300
     },
     {
-      "epoch": 0.4681950274459154,
-      "grad_norm": 0.5928748250007629,
-      "learning_rate": 0.0005725568365477857,
-      "loss": 4.1282,
+      "epoch": 0.46900269541778977,
+      "grad_norm": 0.7219645977020264,
+      "learning_rate": 0.0005725029681597409,
+      "loss": 4.1181,
       "step": 4350
     },
     {
-      "epoch": 0.4735765794855236,
-      "grad_norm": 0.7036423087120056,
-      "learning_rate": 0.0005722335955177243,
-      "loss": 4.1439,
+      "epoch": 0.4743935309973046,
+      "grad_norm": 0.6241794228553772,
+      "learning_rate": 0.0005721791689152725,
+      "loss": 4.1022,
       "step": 4400
     },
     {
-      "epoch": 0.47895813152513184,
-      "grad_norm": 0.8669459223747253,
-      "learning_rate": 0.000571910354487663,
-      "loss": 4.1135,
+      "epoch": 0.4797843665768194,
+      "grad_norm": 0.659267783164978,
+      "learning_rate": 0.0005718553696708041,
+      "loss": 4.1075,
       "step": 4450
     },
     {
-      "epoch": 0.48433968356474005,
-      "grad_norm": 0.8780127763748169,
-      "learning_rate": 0.0005715871134576015,
-      "loss": 4.1044,
+      "epoch": 0.48517520215633425,
+      "grad_norm": 0.6272718906402588,
+      "learning_rate": 0.0005715315704263356,
+      "loss": 4.104,
       "step": 4500
     },
     {
-      "epoch": 0.48972123560434827,
-      "grad_norm": 0.6267021298408508,
-      "learning_rate": 0.0005712638724275401,
-      "loss": 4.1107,
+      "epoch": 0.49056603773584906,
+      "grad_norm": 0.7427578568458557,
+      "learning_rate": 0.0005712077711818672,
+      "loss": 4.1125,
       "step": 4550
     },
     {
-      "epoch": 0.49510278764395654,
-      "grad_norm": 0.5800017714500427,
-      "learning_rate": 0.0005709406313974786,
-      "loss": 4.095,
+      "epoch": 0.49595687331536387,
+      "grad_norm": 0.8200653791427612,
+      "learning_rate": 0.0005708839719373987,
+      "loss": 4.1044,
       "step": 4600
     },
     {
-      "epoch": 0.5004843396835648,
-      "grad_norm": 0.7420774698257446,
-      "learning_rate": 0.0005706173903674173,
-      "loss": 4.1085,
+      "epoch": 0.5013477088948787,
+      "grad_norm": 0.622901201248169,
+      "learning_rate": 0.0005705601726929304,
+      "loss": 4.0639,
       "step": 4650
     },
     {
-      "epoch": 0.505865891723173,
-      "grad_norm": 0.600246250629425,
-      "learning_rate": 0.0005702941493373559,
-      "loss": 4.1027,
+      "epoch": 0.5067385444743935,
+      "grad_norm": 0.6671782732009888,
+      "learning_rate": 0.0005702363734484619,
+      "loss": 4.0661,
       "step": 4700
     },
     {
-      "epoch": 0.5112474437627812,
-      "grad_norm": 0.5375741124153137,
-      "learning_rate": 0.0005699709083072944,
-      "loss": 4.1006,
+      "epoch": 0.5121293800539084,
+      "grad_norm": 0.5631647109985352,
+      "learning_rate": 0.0005699125742039935,
+      "loss": 4.0739,
       "step": 4750
     },
     {
-      "epoch": 0.5166289958023894,
-      "grad_norm": 0.8641989231109619,
-      "learning_rate": 0.000569647667277233,
-      "loss": 4.1112,
+      "epoch": 0.5175202156334232,
+      "grad_norm": 0.6161767840385437,
+      "learning_rate": 0.000569588774959525,
+      "loss": 4.0669,
       "step": 4800
     },
     {
-      "epoch": 0.5220105478419976,
-      "grad_norm": 0.6819212436676025,
-      "learning_rate": 0.0005693244262471716,
-      "loss": 4.0768,
+      "epoch": 0.522911051212938,
+      "grad_norm": 0.7167460918426514,
+      "learning_rate": 0.0005692649757150567,
+      "loss": 4.0629,
       "step": 4850
     },
     {
-      "epoch": 0.5273920998816058,
-      "grad_norm": 0.683928906917572,
-      "learning_rate": 0.0005690011852171102,
-      "loss": 4.0565,
+      "epoch": 0.5283018867924528,
+      "grad_norm": 0.6978079080581665,
+      "learning_rate": 0.0005689411764705881,
+      "loss": 4.0651,
       "step": 4900
     },
     {
-      "epoch": 0.5327736519212141,
-      "grad_norm": 0.6949369311332703,
-      "learning_rate": 0.0005686779441870487,
-      "loss": 4.0832,
+      "epoch": 0.5336927223719676,
+      "grad_norm": 0.671887993812561,
+      "learning_rate": 0.0005686173772261197,
+      "loss": 4.0554,
       "step": 4950
     },
     {
-      "epoch": 0.5381552039608223,
-      "grad_norm": 0.694450855255127,
-      "learning_rate": 0.0005683547031569874,
-      "loss": 4.0834,
+      "epoch": 0.5390835579514824,
+      "grad_norm": 0.7381294965744019,
+      "learning_rate": 0.0005682935779816514,
+      "loss": 4.0543,
       "step": 5000
     },
     {
-      "epoch": 0.5381552039608223,
-      "eval_accuracy": 0.32085958341830895,
-      "eval_loss": 3.9946889877319336,
-      "eval_runtime": 183.6781,
-      "eval_samples_per_second": 98.057,
-      "eval_steps_per_second": 6.13,
+      "epoch": 0.5390835579514824,
+      "eval_accuracy": 0.3217498851810608,
+      "eval_loss": 3.9867610931396484,
+      "eval_runtime": 183.7028,
+      "eval_samples_per_second": 98.044,
+      "eval_steps_per_second": 6.129,
       "step": 5000
     },
     {
-      "epoch": 0.5435367560004305,
-      "grad_norm": 0.6287670135498047,
-      "learning_rate": 0.0005680314621269259,
-      "loss": 4.0635,
+      "epoch": 0.5444743935309974,
+      "grad_norm": 0.6387858390808105,
+      "learning_rate": 0.0005679697787371829,
+      "loss": 4.0612,
       "step": 5050
     },
     {
-      "epoch": 0.5489183080400387,
-      "grad_norm": 0.6268389821052551,
-      "learning_rate": 0.0005677082210968646,
-      "loss": 4.0653,
+      "epoch": 0.5498652291105122,
+      "grad_norm": 0.5617968440055847,
+      "learning_rate": 0.0005676459794927145,
+      "loss": 4.046,
       "step": 5100
     },
     {
-      "epoch": 0.5542998600796469,
-      "grad_norm": 0.5669043660163879,
-      "learning_rate": 0.0005673849800668031,
-      "loss": 4.0548,
+      "epoch": 0.555256064690027,
+      "grad_norm": 0.6449678540229797,
+      "learning_rate": 0.000567322180248246,
+      "loss": 4.041,
       "step": 5150
     },
     {
-      "epoch": 0.5596814121192552,
-      "grad_norm": 0.6583690047264099,
-      "learning_rate": 0.0005670617390367416,
-      "loss": 4.0645,
+      "epoch": 0.5606469002695418,
+      "grad_norm": 0.672944962978363,
+      "learning_rate": 0.0005669983810037777,
+      "loss": 4.05,
       "step": 5200
     },
     {
-      "epoch": 0.5650629641588634,
-      "grad_norm": 0.7188791036605835,
-      "learning_rate": 0.0005667384980066803,
-      "loss": 4.0636,
+      "epoch": 0.5660377358490566,
+      "grad_norm": 0.6287308931350708,
+      "learning_rate": 0.0005666745817593092,
+      "loss": 4.026,
       "step": 5250
     },
     {
-      "epoch": 0.5704445161984716,
-      "grad_norm": 0.6912909746170044,
-      "learning_rate": 0.0005664152569766188,
-      "loss": 4.0563,
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.5841106176376343,
+      "learning_rate": 0.0005663507825148408,
+      "loss": 4.0254,
       "step": 5300
     },
     {
-      "epoch": 0.5758260682380799,
-      "grad_norm": 0.6811025738716125,
-      "learning_rate": 0.0005660920159465575,
-      "loss": 4.0438,
+      "epoch": 0.5768194070080862,
+      "grad_norm": 0.7563546895980835,
+      "learning_rate": 0.0005660269832703723,
+      "loss": 4.0185,
       "step": 5350
     },
     {
-      "epoch": 0.5812076202776881,
-      "grad_norm": 0.7273007035255432,
-      "learning_rate": 0.000565768774916496,
-      "loss": 4.0386,
+      "epoch": 0.5822102425876011,
+      "grad_norm": 0.6202201247215271,
+      "learning_rate": 0.0005657031840259039,
+      "loss": 4.0234,
       "step": 5400
     },
     {
-      "epoch": 0.5865891723172963,
-      "grad_norm": 0.7094576358795166,
-      "learning_rate": 0.0005654455338864346,
-      "loss": 4.0236,
+      "epoch": 0.5876010781671159,
+      "grad_norm": 0.7344704270362854,
+      "learning_rate": 0.0005653793847814355,
+      "loss": 4.0227,
       "step": 5450
     },
     {
-      "epoch": 0.5919707243569046,
-      "grad_norm": 0.6505335569381714,
-      "learning_rate": 0.0005651222928563732,
-      "loss": 4.0119,
+      "epoch": 0.5929919137466307,
+      "grad_norm": 0.6420839428901672,
+      "learning_rate": 0.000565055585536967,
+      "loss": 3.9993,
       "step": 5500
     },
     {
-      "epoch": 0.5973522763965128,
-      "grad_norm": 0.5709679126739502,
-      "learning_rate": 0.0005647990518263118,
-      "loss": 4.0309,
+      "epoch": 0.5983827493261455,
+      "grad_norm": 0.6762014031410217,
+      "learning_rate": 0.0005647317862924986,
+      "loss": 4.026,
       "step": 5550
     },
     {
-      "epoch": 0.602733828436121,
-      "grad_norm": 0.6012085676193237,
-      "learning_rate": 0.0005644758107962504,
-      "loss": 3.9989,
+      "epoch": 0.6037735849056604,
+      "grad_norm": 0.5997875928878784,
+      "learning_rate": 0.0005644079870480302,
+      "loss": 3.9718,
       "step": 5600
     },
     {
-      "epoch": 0.6081153804757292,
-      "grad_norm": 0.5513839721679688,
-      "learning_rate": 0.0005641525697661889,
-      "loss": 4.0163,
+      "epoch": 0.6091644204851752,
+      "grad_norm": 0.6339121460914612,
+      "learning_rate": 0.0005640841878035617,
+      "loss": 4.0075,
       "step": 5650
     },
     {
-      "epoch": 0.6134969325153374,
-      "grad_norm": 0.80906742811203,
-      "learning_rate": 0.0005638293287361275,
-      "loss": 4.0056,
+      "epoch": 0.6145552560646901,
+      "grad_norm": 0.665338397026062,
+      "learning_rate": 0.0005637603885590933,
+      "loss": 4.0065,
       "step": 5700
     },
     {
-      "epoch": 0.6188784845549457,
-      "grad_norm": 0.6083957552909851,
-      "learning_rate": 0.0005635060877060661,
-      "loss": 4.0,
+      "epoch": 0.6199460916442049,
+      "grad_norm": 0.6611903309822083,
+      "learning_rate": 0.0005634365893146248,
+      "loss": 4.0088,
       "step": 5750
     },
     {
-      "epoch": 0.6242600365945539,
-      "grad_norm": 0.6568366885185242,
-      "learning_rate": 0.0005631828466760047,
-      "loss": 3.9893,
+      "epoch": 0.6253369272237197,
+      "grad_norm": 0.6619381308555603,
+      "learning_rate": 0.0005631127900701565,
+      "loss": 3.9978,
       "step": 5800
     },
     {
-      "epoch": 0.6296415886341621,
-      "grad_norm": 0.7716871500015259,
-      "learning_rate": 0.0005628596056459433,
-      "loss": 4.0293,
+      "epoch": 0.6307277628032345,
+      "grad_norm": 0.6455600261688232,
+      "learning_rate": 0.000562788990825688,
+      "loss": 3.9697,
       "step": 5850
     },
     {
-      "epoch": 0.6350231406737703,
-      "grad_norm": 0.7614040374755859,
-      "learning_rate": 0.0005625363646158818,
-      "loss": 4.0103,
+      "epoch": 0.6361185983827493,
+      "grad_norm": 0.5724530220031738,
+      "learning_rate": 0.0005624651915812196,
+      "loss": 3.9838,
       "step": 5900
     },
     {
-      "epoch": 0.6404046927133785,
-      "grad_norm": 0.5670299530029297,
-      "learning_rate": 0.0005622131235858205,
-      "loss": 3.9907,
+      "epoch": 0.6415094339622641,
+      "grad_norm": 0.7462118864059448,
+      "learning_rate": 0.0005621413923367511,
+      "loss": 3.9893,
       "step": 5950
     },
     {
-      "epoch": 0.6457862447529867,
-      "grad_norm": 0.6070712804794312,
-      "learning_rate": 0.000561889882555759,
-      "loss": 3.9779,
+      "epoch": 0.6469002695417789,
+      "grad_norm": 0.5405612587928772,
+      "learning_rate": 0.0005618175930922828,
+      "loss": 3.9968,
       "step": 6000
     },
     {
-      "epoch": 0.6457862447529867,
-      "eval_accuracy": 0.3274067891322339,
-      "eval_loss": 3.9255194664001465,
-      "eval_runtime": 183.4556,
-      "eval_samples_per_second": 98.176,
-      "eval_steps_per_second": 6.138,
+      "epoch": 0.6469002695417789,
+      "eval_accuracy": 0.32842356285640184,
+      "eval_loss": 3.9113423824310303,
+      "eval_runtime": 183.7488,
+      "eval_samples_per_second": 98.02,
+      "eval_steps_per_second": 6.128,
       "step": 6000
     },
     {
-      "epoch": 0.651167796792595,
-      "grad_norm": 0.7539680600166321,
-      "learning_rate": 0.0005615666415256976,
-      "loss": 4.0013,
+      "epoch": 0.6522911051212938,
+      "grad_norm": 0.5777209997177124,
+      "learning_rate": 0.0005614937938478143,
+      "loss": 3.9866,
       "step": 6050
     },
     {
-      "epoch": 0.6565493488322032,
-      "grad_norm": 0.596575140953064,
-      "learning_rate": 0.0005612434004956361,
-      "loss": 3.9909,
+      "epoch": 0.6576819407008087,
+      "grad_norm": 0.6456552743911743,
+      "learning_rate": 0.0005611699946033459,
+      "loss": 3.9793,
       "step": 6100
     },
     {
-      "epoch": 0.6619309008718114,
-      "grad_norm": 0.6477878093719482,
-      "learning_rate": 0.0005609201594655748,
-      "loss": 3.9756,
+      "epoch": 0.6630727762803235,
+      "grad_norm": 0.5341396927833557,
+      "learning_rate": 0.0005608461953588774,
+      "loss": 3.9731,
       "step": 6150
     },
     {
-      "epoch": 0.6673124529114196,
-      "grad_norm": 0.5999401211738586,
-      "learning_rate": 0.0005605969184355134,
-      "loss": 3.9741,
+      "epoch": 0.6684636118598383,
+      "grad_norm": 0.5835209488868713,
+      "learning_rate": 0.000560522396114409,
+      "loss": 3.9675,
       "step": 6200
     },
     {
-      "epoch": 0.6726940049510278,
-      "grad_norm": 0.6590232849121094,
-      "learning_rate": 0.000560273677405452,
-      "loss": 3.9687,
+      "epoch": 0.6738544474393531,
+      "grad_norm": 0.6508961319923401,
+      "learning_rate": 0.0005601985968699405,
+      "loss": 3.9502,
       "step": 6250
     },
     {
-      "epoch": 0.6780755569906362,
-      "grad_norm": 0.6066744923591614,
-      "learning_rate": 0.0005599504363753905,
-      "loss": 3.9584,
+      "epoch": 0.6792452830188679,
+      "grad_norm": 0.6358301639556885,
+      "learning_rate": 0.0005598747976254721,
+      "loss": 3.9653,
       "step": 6300
     },
     {
-      "epoch": 0.6834571090302444,
-      "grad_norm": 0.6018451452255249,
-      "learning_rate": 0.0005596271953453291,
-      "loss": 3.9905,
+      "epoch": 0.6846361185983828,
+      "grad_norm": 0.561137855052948,
+      "learning_rate": 0.0005595509983810038,
+      "loss": 3.937,
       "step": 6350
     },
     {
-      "epoch": 0.6888386610698526,
-      "grad_norm": 0.6121416687965393,
-      "learning_rate": 0.0005593039543152677,
-      "loss": 3.9772,
+      "epoch": 0.6900269541778976,
+      "grad_norm": 0.6251780390739441,
+      "learning_rate": 0.0005592271991365353,
+      "loss": 3.9524,
       "step": 6400
     },
     {
-      "epoch": 0.6942202131094608,
-      "grad_norm": 0.551299512386322,
-      "learning_rate": 0.0005589807132852063,
-      "loss": 3.9937,
+      "epoch": 0.6954177897574124,
+      "grad_norm": 0.6292946338653564,
+      "learning_rate": 0.0005589033998920669,
+      "loss": 3.9657,
       "step": 6450
     },
     {
-      "epoch": 0.699601765149069,
-      "grad_norm": 0.5864540338516235,
-      "learning_rate": 0.0005586574722551449,
-      "loss": 3.9604,
+      "epoch": 0.7008086253369272,
+      "grad_norm": 0.6746073961257935,
+      "learning_rate": 0.0005585796006475984,
+      "loss": 3.9375,
       "step": 6500
     },
     {
-      "epoch": 0.7049833171886772,
-      "grad_norm": 0.7134374380111694,
-      "learning_rate": 0.0005583342312250834,
-      "loss": 3.9516,
+      "epoch": 0.706199460916442,
+      "grad_norm": 0.6384598016738892,
+      "learning_rate": 0.0005582558014031301,
+      "loss": 3.9633,
       "step": 6550
     },
     {
-      "epoch": 0.7103648692282855,
-      "grad_norm": 0.5689718127250671,
-      "learning_rate": 0.0005580109901950221,
-      "loss": 3.9513,
+      "epoch": 0.7115902964959568,
+      "grad_norm": 0.5726682543754578,
+      "learning_rate": 0.0005579320021586616,
+      "loss": 3.9444,
       "step": 6600
     },
     {
-      "epoch": 0.7157464212678937,
-      "grad_norm": 0.6758838295936584,
-      "learning_rate": 0.0005576877491649606,
-      "loss": 3.9546,
+      "epoch": 0.7169811320754716,
+      "grad_norm": 0.6135655641555786,
+      "learning_rate": 0.0005576082029141932,
+      "loss": 3.9389,
       "step": 6650
     },
     {
-      "epoch": 0.7211279733075019,
-      "grad_norm": 0.5809199810028076,
-      "learning_rate": 0.0005573645081348993,
-      "loss": 3.9451,
+      "epoch": 0.7223719676549866,
+      "grad_norm": 0.5083711743354797,
+      "learning_rate": 0.0005572844036697247,
+      "loss": 3.9261,
       "step": 6700
     },
     {
-      "epoch": 0.7265095253471101,
-      "grad_norm": 0.6426367163658142,
-      "learning_rate": 0.0005570412671048378,
-      "loss": 3.9481,
+      "epoch": 0.7277628032345014,
+      "grad_norm": 0.5756128430366516,
+      "learning_rate": 0.0005569606044252563,
+      "loss": 3.9306,
       "step": 6750
     },
     {
-      "epoch": 0.7318910773867183,
-      "grad_norm": 0.6130222678184509,
-      "learning_rate": 0.0005567180260747763,
-      "loss": 3.9252,
+      "epoch": 0.7331536388140162,
+      "grad_norm": 0.8808412551879883,
+      "learning_rate": 0.0005566368051807879,
+      "loss": 3.9502,
       "step": 6800
     },
     {
-      "epoch": 0.7372726294263265,
-      "grad_norm": 0.57786625623703,
-      "learning_rate": 0.000556394785044715,
-      "loss": 3.9364,
+      "epoch": 0.738544474393531,
+      "grad_norm": 0.6494237780570984,
+      "learning_rate": 0.0005563130059363194,
+      "loss": 3.9371,
       "step": 6850
     },
     {
-      "epoch": 0.7426541814659348,
-      "grad_norm": 0.5700427889823914,
-      "learning_rate": 0.0005560715440146535,
-      "loss": 3.9278,
+      "epoch": 0.7439353099730458,
+      "grad_norm": 0.531914234161377,
+      "learning_rate": 0.000555989206691851,
+      "loss": 3.9346,
       "step": 6900
     },
     {
-      "epoch": 0.748035733505543,
-      "grad_norm": 0.6527670621871948,
-      "learning_rate": 0.0005557483029845921,
-      "loss": 3.9457,
+      "epoch": 0.7493261455525606,
+      "grad_norm": 0.6347932815551758,
+      "learning_rate": 0.0005556654074473826,
+      "loss": 3.939,
       "step": 6950
     },
     {
-      "epoch": 0.7534172855451512,
-      "grad_norm": 0.5725962519645691,
-      "learning_rate": 0.0005554250619545307,
-      "loss": 3.9308,
+      "epoch": 0.7547169811320755,
+      "grad_norm": 0.6388944387435913,
+      "learning_rate": 0.0005553416082029141,
+      "loss": 3.9218,
       "step": 7000
     },
     {
-      "epoch": 0.7534172855451512,
-      "eval_accuracy": 0.3330480470675618,
-      "eval_loss": 3.8677163124084473,
-      "eval_runtime": 183.9066,
-      "eval_samples_per_second": 97.936,
-      "eval_steps_per_second": 6.123,
+      "epoch": 0.7547169811320755,
+      "eval_accuracy": 0.3334278975633612,
+      "eval_loss": 3.857858896255493,
+      "eval_runtime": 184.2675,
+      "eval_samples_per_second": 97.744,
+      "eval_steps_per_second": 6.111,
       "step": 7000
     },
     {
-      "epoch": 0.7587988375847594,
-      "grad_norm": 0.6234815120697021,
-      "learning_rate": 0.0005551018209244694,
-      "loss": 3.9224,
+      "epoch": 0.7601078167115903,
+      "grad_norm": 0.6419737339019775,
+      "learning_rate": 0.0005550178089584457,
+      "loss": 3.9141,
       "step": 7050
     },
     {
-      "epoch": 0.7641803896243676,
-      "grad_norm": 0.5600576400756836,
-      "learning_rate": 0.0005547785798944079,
-      "loss": 3.9266,
+      "epoch": 0.7654986522911051,
+      "grad_norm": 0.5612533092498779,
+      "learning_rate": 0.0005546940097139772,
+      "loss": 3.9144,
       "step": 7100
     },
     {
-      "epoch": 0.7695619416639758,
-      "grad_norm": 0.6040908694267273,
-      "learning_rate": 0.0005544553388643464,
-      "loss": 3.9326,
+      "epoch": 0.77088948787062,
+      "grad_norm": 0.626680850982666,
+      "learning_rate": 0.0005543702104695089,
+      "loss": 3.9178,
       "step": 7150
     },
     {
-      "epoch": 0.7749434937035841,
-      "grad_norm": 0.5893228650093079,
-      "learning_rate": 0.000554132097834285,
-      "loss": 3.943,
+      "epoch": 0.7762803234501348,
+      "grad_norm": 0.5791966319084167,
+      "learning_rate": 0.0005540464112250404,
+      "loss": 3.9151,
       "step": 7200
     },
     {
-      "epoch": 0.7803250457431924,
-      "grad_norm": 0.5626388192176819,
-      "learning_rate": 0.0005538088568042236,
-      "loss": 3.9363,
+      "epoch": 0.7816711590296496,
+      "grad_norm": 0.6918361783027649,
+      "learning_rate": 0.000553722611980572,
+      "loss": 3.9269,
       "step": 7250
     },
     {
-      "epoch": 0.7857065977828006,
-      "grad_norm": 0.5734874606132507,
-      "learning_rate": 0.0005534856157741623,
-      "loss": 3.9233,
+      "epoch": 0.7870619946091644,
+      "grad_norm": 0.623336136341095,
+      "learning_rate": 0.0005533988127361035,
+      "loss": 3.9116,
       "step": 7300
     },
     {
-      "epoch": 0.7910881498224088,
-      "grad_norm": 0.6031030416488647,
-      "learning_rate": 0.0005531623747441008,
-      "loss": 3.9101,
+      "epoch": 0.7924528301886793,
+      "grad_norm": 0.5773249268531799,
+      "learning_rate": 0.0005530750134916352,
+      "loss": 3.9271,
       "step": 7350
     },
     {
-      "epoch": 0.796469701862017,
-      "grad_norm": 0.7114644646644592,
-      "learning_rate": 0.0005528391337140394,
-      "loss": 3.9138,
+      "epoch": 0.7978436657681941,
+      "grad_norm": 0.6273781657218933,
+      "learning_rate": 0.0005527512142471668,
+      "loss": 3.8999,
       "step": 7400
     },
     {
-      "epoch": 0.8018512539016253,
-      "grad_norm": 0.6150217056274414,
-      "learning_rate": 0.0005525158926839779,
-      "loss": 3.9134,
+      "epoch": 0.8032345013477089,
+      "grad_norm": 0.6004655957221985,
+      "learning_rate": 0.0005524274150026982,
+      "loss": 3.9011,
       "step": 7450
     },
     {
-      "epoch": 0.8072328059412335,
-      "grad_norm": 0.6344792246818542,
-      "learning_rate": 0.0005521926516539166,
-      "loss": 3.9108,
+      "epoch": 0.8086253369272237,
+      "grad_norm": 0.5613580346107483,
+      "learning_rate": 0.0005521036157582299,
+      "loss": 3.8929,
       "step": 7500
     },
     {
-      "epoch": 0.8126143579808417,
-      "grad_norm": 0.5256510376930237,
-      "learning_rate": 0.0005518694106238552,
-      "loss": 3.8883,
+      "epoch": 0.8140161725067385,
+      "grad_norm": 0.7210002541542053,
+      "learning_rate": 0.0005517798165137614,
+      "loss": 3.897,
       "step": 7550
     },
     {
-      "epoch": 0.8179959100204499,
-      "grad_norm": 0.6160898208618164,
-      "learning_rate": 0.0005515461695937937,
-      "loss": 3.913,
+      "epoch": 0.8194070080862533,
+      "grad_norm": 0.6069796681404114,
+      "learning_rate": 0.000551456017269293,
+      "loss": 3.8717,
       "step": 7600
     },
     {
-      "epoch": 0.8233774620600581,
-      "grad_norm": 0.5938844084739685,
-      "learning_rate": 0.0005512229285637323,
-      "loss": 3.8897,
+      "epoch": 0.8247978436657682,
+      "grad_norm": 0.5973781943321228,
+      "learning_rate": 0.0005511322180248245,
+      "loss": 3.8905,
       "step": 7650
     },
     {
-      "epoch": 0.8287590140996663,
-      "grad_norm": 0.5758791565895081,
-      "learning_rate": 0.0005508996875336709,
-      "loss": 3.8806,
+      "epoch": 0.8301886792452831,
+      "grad_norm": 0.619523823261261,
+      "learning_rate": 0.0005508084187803562,
+      "loss": 3.8992,
       "step": 7700
     },
     {
-      "epoch": 0.8341405661392746,
-      "grad_norm": 0.650128185749054,
-      "learning_rate": 0.0005505764465036095,
-      "loss": 3.9052,
+      "epoch": 0.8355795148247979,
+      "grad_norm": 0.5434247851371765,
+      "learning_rate": 0.0005504846195358877,
+      "loss": 3.8957,
       "step": 7750
     },
     {
-      "epoch": 0.8395221181788828,
-      "grad_norm": 0.6153085231781006,
-      "learning_rate": 0.000550253205473548,
-      "loss": 3.8912,
+      "epoch": 0.8409703504043127,
+      "grad_norm": 0.5716760158538818,
+      "learning_rate": 0.0005501608202914193,
+      "loss": 3.8843,
       "step": 7800
     },
     {
-      "epoch": 0.844903670218491,
-      "grad_norm": 0.6277813911437988,
-      "learning_rate": 0.0005499299644434867,
-      "loss": 3.8741,
+      "epoch": 0.8463611859838275,
+      "grad_norm": 0.6962743997573853,
+      "learning_rate": 0.0005498370210469508,
+      "loss": 3.8875,
       "step": 7850
     },
     {
-      "epoch": 0.8502852222580992,
-      "grad_norm": 0.5605838298797607,
-      "learning_rate": 0.0005496067234134252,
-      "loss": 3.8952,
+      "epoch": 0.8517520215633423,
+      "grad_norm": 0.5928980112075806,
+      "learning_rate": 0.0005495132218024824,
+      "loss": 3.8718,
       "step": 7900
     },
     {
-      "epoch": 0.8556667742977074,
-      "grad_norm": 0.6682559251785278,
-      "learning_rate": 0.0005492834823833639,
-      "loss": 3.8891,
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.5467910170555115,
+      "learning_rate": 0.000549189422558014,
+      "loss": 3.8822,
       "step": 7950
     },
     {
-      "epoch": 0.8610483263373157,
-      "grad_norm": 0.6065753102302551,
-      "learning_rate": 0.0005489602413533024,
-      "loss": 3.8613,
+      "epoch": 0.862533692722372,
+      "grad_norm": 0.5513741970062256,
+      "learning_rate": 0.0005488656233135456,
+      "loss": 3.878,
       "step": 8000
     },
     {
-      "epoch": 0.8610483263373157,
-      "eval_accuracy": 0.33730810949299195,
-      "eval_loss": 3.8242199420928955,
-      "eval_runtime": 183.8039,
-      "eval_samples_per_second": 97.99,
-      "eval_steps_per_second": 6.126,
+      "epoch": 0.862533692722372,
+      "eval_accuracy": 0.33798110548011484,
+      "eval_loss": 3.8124523162841797,
+      "eval_runtime": 183.5173,
+      "eval_samples_per_second": 98.143,
+      "eval_steps_per_second": 6.136,
       "step": 8000
     },
     {
-      "epoch": 0.8664298783769239,
-      "grad_norm": 0.7891616821289062,
-      "learning_rate": 0.0005486370003232409,
-      "loss": 3.8985,
+      "epoch": 0.8679245283018868,
+      "grad_norm": 0.5857316851615906,
+      "learning_rate": 0.0005485418240690771,
+      "loss": 3.8602,
       "step": 8050
     },
     {
-      "epoch": 0.8718114304165321,
-      "grad_norm": 0.6006017923355103,
-      "learning_rate": 0.0005483137592931796,
-      "loss": 3.8671,
+      "epoch": 0.8733153638814016,
+      "grad_norm": 0.6293712854385376,
+      "learning_rate": 0.0005482180248246087,
+      "loss": 3.8769,
       "step": 8100
     },
     {
-      "epoch": 0.8771929824561403,
-      "grad_norm": 0.6902142763137817,
-      "learning_rate": 0.0005479905182631181,
-      "loss": 3.8979,
+      "epoch": 0.8787061994609164,
+      "grad_norm": 0.6371458768844604,
+      "learning_rate": 0.0005478942255801403,
+      "loss": 3.8726,
       "step": 8150
     },
     {
-      "epoch": 0.8825745344957485,
-      "grad_norm": 0.5865663290023804,
-      "learning_rate": 0.0005476672772330568,
-      "loss": 3.8725,
+      "epoch": 0.8840970350404312,
+      "grad_norm": 0.5665318369865417,
+      "learning_rate": 0.0005475704263356718,
+      "loss": 3.8611,
       "step": 8200
     },
     {
-      "epoch": 0.8879560865353568,
-      "grad_norm": 0.5427098870277405,
-      "learning_rate": 0.0005473440362029953,
-      "loss": 3.8684,
+      "epoch": 0.889487870619946,
+      "grad_norm": 0.5692288279533386,
+      "learning_rate": 0.0005472466270912034,
+      "loss": 3.8725,
       "step": 8250
     },
     {
-      "epoch": 0.8933376385749651,
-      "grad_norm": 0.6365987062454224,
-      "learning_rate": 0.0005470207951729339,
-      "loss": 3.864,
+      "epoch": 0.894878706199461,
+      "grad_norm": 0.6609965562820435,
+      "learning_rate": 0.000546922827846735,
+      "loss": 3.8591,
       "step": 8300
     },
     {
-      "epoch": 0.8987191906145733,
-      "grad_norm": 0.5786712169647217,
-      "learning_rate": 0.0005466975541428725,
-      "loss": 3.8644,
+      "epoch": 0.9002695417789758,
+      "grad_norm": 0.5426245331764221,
+      "learning_rate": 0.0005465990286022665,
+      "loss": 3.8598,
       "step": 8350
     },
     {
-      "epoch": 0.9041007426541815,
-      "grad_norm": 0.589103639125824,
-      "learning_rate": 0.0005463743131128111,
-      "loss": 3.8658,
+      "epoch": 0.9056603773584906,
+      "grad_norm": 0.5615479946136475,
+      "learning_rate": 0.0005462752293577981,
+      "loss": 3.8773,
       "step": 8400
     },
     {
-      "epoch": 0.9094822946937897,
-      "grad_norm": 0.6997091770172119,
-      "learning_rate": 0.0005460510720827497,
-      "loss": 3.874,
+      "epoch": 0.9110512129380054,
+      "grad_norm": 0.5803127884864807,
+      "learning_rate": 0.0005459514301133296,
+      "loss": 3.8718,
       "step": 8450
     },
     {
-      "epoch": 0.9148638467333979,
-      "grad_norm": 0.5358573794364929,
-      "learning_rate": 0.0005457278310526882,
-      "loss": 3.8777,
+      "epoch": 0.9164420485175202,
+      "grad_norm": 0.5748214721679688,
+      "learning_rate": 0.0005456276308688613,
+      "loss": 3.8677,
       "step": 8500
     },
     {
-      "epoch": 0.9202453987730062,
-      "grad_norm": 0.7259039878845215,
-      "learning_rate": 0.0005454045900226268,
-      "loss": 3.8575,
+      "epoch": 0.921832884097035,
+      "grad_norm": 0.6031325459480286,
+      "learning_rate": 0.0005453038316243929,
+      "loss": 3.8572,
       "step": 8550
     },
     {
-      "epoch": 0.9256269508126144,
-      "grad_norm": 0.5783784985542297,
-      "learning_rate": 0.0005450813489925654,
-      "loss": 3.8476,
+      "epoch": 0.9272237196765498,
+      "grad_norm": 0.5571451783180237,
+      "learning_rate": 0.0005449800323799244,
+      "loss": 3.8675,
       "step": 8600
     },
     {
-      "epoch": 0.9310085028522226,
-      "grad_norm": 0.5237821340560913,
-      "learning_rate": 0.000544758107962504,
-      "loss": 3.8516,
+      "epoch": 0.9326145552560647,
+      "grad_norm": 0.5379493236541748,
+      "learning_rate": 0.000544656233135456,
+      "loss": 3.8486,
       "step": 8650
     },
     {
-      "epoch": 0.9363900548918308,
-      "grad_norm": 0.5917977094650269,
-      "learning_rate": 0.0005444348669324426,
-      "loss": 3.8578,
+      "epoch": 0.9380053908355795,
+      "grad_norm": 0.6122162938117981,
+      "learning_rate": 0.0005443324338909875,
+      "loss": 3.8542,
       "step": 8700
     },
     {
-      "epoch": 0.941771606931439,
-      "grad_norm": 0.5565261840820312,
-      "learning_rate": 0.0005441116259023811,
-      "loss": 3.8558,
+      "epoch": 0.9433962264150944,
+      "grad_norm": 0.599349319934845,
+      "learning_rate": 0.0005440086346465192,
+      "loss": 3.8629,
       "step": 8750
     },
     {
-      "epoch": 0.9471531589710472,
-      "grad_norm": 0.5900228023529053,
-      "learning_rate": 0.0005437883848723198,
-      "loss": 3.8642,
+      "epoch": 0.9487870619946092,
+      "grad_norm": 0.5931590795516968,
+      "learning_rate": 0.0005436848354020506,
+      "loss": 3.846,
       "step": 8800
     },
     {
-      "epoch": 0.9525347110106555,
-      "grad_norm": 0.6702139377593994,
-      "learning_rate": 0.0005434651438422583,
-      "loss": 3.8506,
+      "epoch": 0.954177897574124,
+      "grad_norm": 0.5675060153007507,
+      "learning_rate": 0.0005433610361575823,
+      "loss": 3.8524,
       "step": 8850
     },
     {
-      "epoch": 0.9579162630502637,
-      "grad_norm": 0.5468742251396179,
-      "learning_rate": 0.0005431419028121969,
-      "loss": 3.8481,
+      "epoch": 0.9595687331536388,
+      "grad_norm": 0.5403425097465515,
+      "learning_rate": 0.0005430372369131138,
+      "loss": 3.835,
       "step": 8900
     },
     {
-      "epoch": 0.9632978150898719,
-      "grad_norm": 0.5315669775009155,
-      "learning_rate": 0.0005428186617821354,
-      "loss": 3.8539,
+      "epoch": 0.9649595687331537,
+      "grad_norm": 0.5725551843643188,
+      "learning_rate": 0.0005427134376686454,
+      "loss": 3.8487,
       "step": 8950
     },
     {
-      "epoch": 0.9686793671294801,
-      "grad_norm": 0.586371898651123,
-      "learning_rate": 0.0005424954207520741,
-      "loss": 3.8433,
+      "epoch": 0.9703504043126685,
+      "grad_norm": 0.5887113213539124,
+      "learning_rate": 0.0005423896384241769,
+      "loss": 3.8355,
       "step": 9000
     },
     {
-      "epoch": 0.9686793671294801,
-      "eval_accuracy": 0.3411381237884524,
-      "eval_loss": 3.781165361404419,
-      "eval_runtime": 183.5101,
-      "eval_samples_per_second": 98.147,
-      "eval_steps_per_second": 6.136,
+      "epoch": 0.9703504043126685,
+      "eval_accuracy": 0.3413660775470872,
+      "eval_loss": 3.7759971618652344,
+      "eval_runtime": 184.0381,
+      "eval_samples_per_second": 97.866,
+      "eval_steps_per_second": 6.118,
       "step": 9000
     },
     {
-      "epoch": 0.9740609191690883,
-      "grad_norm": 0.5976994633674622,
-      "learning_rate": 0.0005421721797220127,
-      "loss": 3.8478,
+      "epoch": 0.9757412398921833,
+      "grad_norm": 0.6292560696601868,
+      "learning_rate": 0.0005420658391797086,
+      "loss": 3.8289,
       "step": 9050
     },
     {
-      "epoch": 0.9794424712086965,
-      "grad_norm": 0.5606995224952698,
-      "learning_rate": 0.0005418489386919513,
-      "loss": 3.8271,
+      "epoch": 0.9811320754716981,
+      "grad_norm": 0.580018162727356,
+      "learning_rate": 0.0005417420399352401,
+      "loss": 3.8412,
       "step": 9100
     },
     {
-      "epoch": 0.9848240232483048,
-      "grad_norm": 0.5494250059127808,
-      "learning_rate": 0.0005415256976618898,
-      "loss": 3.8593,
+      "epoch": 0.9865229110512129,
+      "grad_norm": 0.6424311399459839,
+      "learning_rate": 0.0005414182406907717,
+      "loss": 3.8351,
       "step": 9150
     },
     {
-      "epoch": 0.9902055752879131,
-      "grad_norm": 0.5214250087738037,
-      "learning_rate": 0.0005412024566318284,
-      "loss": 3.844,
+      "epoch": 0.9919137466307277,
+      "grad_norm": 0.6523851156234741,
+      "learning_rate": 0.0005410944414463032,
+      "loss": 3.8493,
       "step": 9200
     },
     {
-      "epoch": 0.9955871273275213,
-      "grad_norm": 0.5299152731895447,
-      "learning_rate": 0.000540879215601767,
-      "loss": 3.8402,
+      "epoch": 0.9973045822102425,
+      "grad_norm": 0.6284000873565674,
+      "learning_rate": 0.0005407706422018348,
+      "loss": 3.847,
       "step": 9250
     },
     {
-      "epoch": 1.0009686793671295,
-      "grad_norm": 0.6352019906044006,
-      "learning_rate": 0.0005405559745717056,
-      "loss": 3.8342,
+      "epoch": 1.0026954177897573,
+      "grad_norm": 0.5991977453231812,
+      "learning_rate": 0.0005404468429573664,
+      "loss": 3.795,
       "step": 9300
     },
     {
-      "epoch": 1.0063502314067376,
-      "grad_norm": 0.6200937628746033,
-      "learning_rate": 0.0005402327335416442,
-      "loss": 3.7525,
+      "epoch": 1.0080862533692723,
+      "grad_norm": 0.6261839270591736,
+      "learning_rate": 0.000540123043712898,
+      "loss": 3.7673,
       "step": 9350
     },
     {
-      "epoch": 1.011731783446346,
-      "grad_norm": 0.5944552421569824,
-      "learning_rate": 0.0005399094925115827,
-      "loss": 3.7798,
+      "epoch": 1.013477088948787,
+      "grad_norm": 0.6294857859611511,
+      "learning_rate": 0.0005397992444684295,
+      "loss": 3.7483,
       "step": 9400
     },
     {
-      "epoch": 1.017113335485954,
-      "grad_norm": 0.6369380950927734,
-      "learning_rate": 0.0005395862514815214,
-      "loss": 3.7553,
+      "epoch": 1.0188679245283019,
+      "grad_norm": 0.6605191230773926,
+      "learning_rate": 0.0005394754452239611,
+      "loss": 3.761,
       "step": 9450
     },
     {
-      "epoch": 1.0224948875255624,
-      "grad_norm": 0.6259318590164185,
-      "learning_rate": 0.0005392630104514599,
-      "loss": 3.7643,
+      "epoch": 1.0242587601078168,
+      "grad_norm": 0.540192186832428,
+      "learning_rate": 0.0005391516459794927,
+      "loss": 3.7716,
       "step": 9500
     },
     {
-      "epoch": 1.0278764395651705,
-      "grad_norm": 0.5792639851570129,
-      "learning_rate": 0.0005389397694213986,
-      "loss": 3.7748,
+      "epoch": 1.0296495956873315,
+      "grad_norm": 0.5671611428260803,
+      "learning_rate": 0.0005388278467350242,
+      "loss": 3.7805,
       "step": 9550
     },
     {
-      "epoch": 1.0332579916047788,
-      "grad_norm": 0.6107578277587891,
-      "learning_rate": 0.0005386165283913371,
-      "loss": 3.7812,
+      "epoch": 1.0350404312668464,
+      "grad_norm": 0.5405906438827515,
+      "learning_rate": 0.0005385040474905557,
+      "loss": 3.7731,
       "step": 9600
     },
     {
-      "epoch": 1.0386395436443872,
-      "grad_norm": 0.6191133856773376,
-      "learning_rate": 0.0005382932873612756,
-      "loss": 3.781,
+      "epoch": 1.0404312668463611,
+      "grad_norm": 0.5329062938690186,
+      "learning_rate": 0.0005381802482460874,
+      "loss": 3.7825,
       "step": 9650
     },
     {
-      "epoch": 1.0440210956839953,
-      "grad_norm": 0.5727741718292236,
-      "learning_rate": 0.0005379700463312143,
-      "loss": 3.7702,
+      "epoch": 1.045822102425876,
+      "grad_norm": 0.5355062484741211,
+      "learning_rate": 0.000537856449001619,
+      "loss": 3.7827,
       "step": 9700
     },
     {
-      "epoch": 1.0494026477236036,
-      "grad_norm": 0.5681310296058655,
-      "learning_rate": 0.0005376468053011528,
-      "loss": 3.7599,
+      "epoch": 1.0512129380053907,
+      "grad_norm": 0.5401003360748291,
+      "learning_rate": 0.0005375326497571505,
+      "loss": 3.7522,
       "step": 9750
     },
     {
-      "epoch": 1.0547841997632117,
-      "grad_norm": 0.5764620304107666,
-      "learning_rate": 0.0005373235642710914,
-      "loss": 3.7743,
+      "epoch": 1.0566037735849056,
+      "grad_norm": 0.5838435292243958,
+      "learning_rate": 0.000537208850512682,
+      "loss": 3.7609,
       "step": 9800
     },
     {
-      "epoch": 1.06016575180282,
-      "grad_norm": 0.5260862708091736,
-      "learning_rate": 0.00053700032324103,
-      "loss": 3.7627,
+      "epoch": 1.0619946091644206,
+      "grad_norm": 0.5370477437973022,
+      "learning_rate": 0.0005368850512682137,
+      "loss": 3.7586,
       "step": 9850
     },
     {
-      "epoch": 1.0655473038424281,
-      "grad_norm": 0.5996755361557007,
-      "learning_rate": 0.0005366770822109687,
-      "loss": 3.7853,
+      "epoch": 1.0673854447439353,
+      "grad_norm": 0.5118876695632935,
+      "learning_rate": 0.0005365612520237453,
+      "loss": 3.7702,
       "step": 9900
     },
     {
-      "epoch": 1.0709288558820365,
-      "grad_norm": 0.596442699432373,
-      "learning_rate": 0.0005363538411809072,
-      "loss": 3.745,
+      "epoch": 1.0727762803234502,
+      "grad_norm": 0.5664620995521545,
+      "learning_rate": 0.0005362374527792768,
+      "loss": 3.7714,
       "step": 9950
     },
     {
-      "epoch": 1.0763104079216446,
-      "grad_norm": 0.6527745723724365,
-      "learning_rate": 0.0005360306001508457,
-      "loss": 3.75,
+      "epoch": 1.0781671159029649,
+      "grad_norm": 0.6131268739700317,
+      "learning_rate": 0.0005359136535348084,
+      "loss": 3.7731,
       "step": 10000
     },
     {
-      "epoch": 1.0763104079216446,
-      "eval_accuracy": 0.3444462782637399,
-      "eval_loss": 3.751065254211426,
-      "eval_runtime": 183.9006,
-      "eval_samples_per_second": 97.939,
-      "eval_steps_per_second": 6.123,
+      "epoch": 1.0781671159029649,
+      "eval_accuracy": 0.34464945916395295,
+      "eval_loss": 3.7502784729003906,
+      "eval_runtime": 183.8229,
+      "eval_samples_per_second": 97.98,
+      "eval_steps_per_second": 6.125,
       "step": 10000
     },
     {
-      "epoch": 1.081691959961253,
-      "grad_norm": 0.6665418148040771,
-      "learning_rate": 0.0005357073591207843,
-      "loss": 3.7483,
+      "epoch": 1.0835579514824798,
+      "grad_norm": 0.5173006057739258,
+      "learning_rate": 0.0005355898542903399,
+      "loss": 3.7492,
       "step": 10050
     },
     {
-      "epoch": 1.087073512000861,
-      "grad_norm": 0.5916799306869507,
-      "learning_rate": 0.0005353841180907229,
-      "loss": 3.7433,
+      "epoch": 1.0889487870619945,
+      "grad_norm": 0.567226231098175,
+      "learning_rate": 0.0005352660550458716,
+      "loss": 3.7538,
       "step": 10100
     },
     {
-      "epoch": 1.0924550640404693,
-      "grad_norm": 0.6221281290054321,
-      "learning_rate": 0.0005350608770606616,
-      "loss": 3.7746,
+      "epoch": 1.0943396226415094,
+      "grad_norm": 0.614502489566803,
+      "learning_rate": 0.000534942255801403,
+      "loss": 3.7424,
       "step": 10150
     },
     {
-      "epoch": 1.0978366160800774,
-      "grad_norm": 0.5531126260757446,
-      "learning_rate": 0.0005347376360306001,
-      "loss": 3.7726,
+      "epoch": 1.0997304582210243,
+      "grad_norm": 0.5912910103797913,
+      "learning_rate": 0.0005346184565569347,
+      "loss": 3.7582,
       "step": 10200
     },
     {
-      "epoch": 1.1032181681196858,
-      "grad_norm": 0.5145325660705566,
-      "learning_rate": 0.0005344143950005387,
-      "loss": 3.7758,
+      "epoch": 1.105121293800539,
+      "grad_norm": 0.5250498056411743,
+      "learning_rate": 0.0005342946573124662,
+      "loss": 3.7766,
       "step": 10250
     },
     {
-      "epoch": 1.1085997201592939,
-      "grad_norm": 0.6110062003135681,
-      "learning_rate": 0.0005340911539704773,
-      "loss": 3.7719,
+      "epoch": 1.110512129380054,
+      "grad_norm": 0.6503591537475586,
+      "learning_rate": 0.0005339708580679978,
+      "loss": 3.7764,
       "step": 10300
     },
     {
-      "epoch": 1.1139812721989022,
-      "grad_norm": 0.5908063650131226,
-      "learning_rate": 0.0005337679129404159,
-      "loss": 3.7502,
+      "epoch": 1.1159029649595686,
+      "grad_norm": 0.5180767178535461,
+      "learning_rate": 0.0005336470588235293,
+      "loss": 3.7537,
       "step": 10350
     },
     {
-      "epoch": 1.1193628242385103,
-      "grad_norm": 0.6414334774017334,
-      "learning_rate": 0.0005334446719103545,
-      "loss": 3.7641,
+      "epoch": 1.1212938005390836,
+      "grad_norm": 0.5671932697296143,
+      "learning_rate": 0.000533323259579061,
+      "loss": 3.7577,
       "step": 10400
     },
     {
-      "epoch": 1.1247443762781186,
-      "grad_norm": 0.5623446106910706,
-      "learning_rate": 0.000533121430880293,
-      "loss": 3.7611,
+      "epoch": 1.1266846361185983,
+      "grad_norm": 0.5335789322853088,
+      "learning_rate": 0.0005329994603345925,
+      "loss": 3.753,
       "step": 10450
     },
     {
-      "epoch": 1.1301259283177267,
-      "grad_norm": 0.5651029944419861,
-      "learning_rate": 0.0005327981898502316,
-      "loss": 3.7515,
+      "epoch": 1.1320754716981132,
+      "grad_norm": 0.5631784200668335,
+      "learning_rate": 0.0005326756610901241,
+      "loss": 3.7534,
       "step": 10500
     },
     {
-      "epoch": 1.135507480357335,
-      "grad_norm": 0.601318895816803,
-      "learning_rate": 0.0005324749488201702,
-      "loss": 3.7561,
+      "epoch": 1.137466307277628,
+      "grad_norm": 0.5201789140701294,
+      "learning_rate": 0.0005323518618456556,
+      "loss": 3.7522,
       "step": 10550
     },
     {
-      "epoch": 1.1408890323969434,
-      "grad_norm": 0.6262565851211548,
-      "learning_rate": 0.0005321517077901088,
-      "loss": 3.76,
+      "epoch": 1.1428571428571428,
+      "grad_norm": 0.552317202091217,
+      "learning_rate": 0.0005320280626011872,
+      "loss": 3.7637,
       "step": 10600
     },
     {
-      "epoch": 1.1462705844365515,
-      "grad_norm": 0.5618316531181335,
-      "learning_rate": 0.0005318284667600473,
-      "loss": 3.7599,
+      "epoch": 1.1482479784366577,
+      "grad_norm": 0.5951991677284241,
+      "learning_rate": 0.0005317042633567188,
+      "loss": 3.7519,
       "step": 10650
     },
     {
-      "epoch": 1.1516521364761596,
-      "grad_norm": 0.5315815210342407,
-      "learning_rate": 0.000531505225729986,
-      "loss": 3.752,
+      "epoch": 1.1536388140161726,
+      "grad_norm": 0.6047458052635193,
+      "learning_rate": 0.0005313804641122504,
+      "loss": 3.7564,
       "step": 10700
     },
     {
-      "epoch": 1.157033688515768,
-      "grad_norm": 0.5935831069946289,
-      "learning_rate": 0.0005311819846999245,
-      "loss": 3.7714,
+      "epoch": 1.1590296495956873,
+      "grad_norm": 0.5586419105529785,
+      "learning_rate": 0.0005310566648677819,
+      "loss": 3.7513,
       "step": 10750
     },
     {
-      "epoch": 1.1624152405553763,
-      "grad_norm": 0.6924194693565369,
-      "learning_rate": 0.0005308587436698631,
-      "loss": 3.7642,
+      "epoch": 1.1644204851752022,
+      "grad_norm": 0.5642745494842529,
+      "learning_rate": 0.0005307328656233135,
+      "loss": 3.742,
       "step": 10800
     },
     {
-      "epoch": 1.1677967925949844,
-      "grad_norm": 0.574368417263031,
-      "learning_rate": 0.0005305355026398017,
-      "loss": 3.7221,
+      "epoch": 1.169811320754717,
+      "grad_norm": 0.5807494521141052,
+      "learning_rate": 0.000530409066378845,
+      "loss": 3.7399,
       "step": 10850
     },
     {
-      "epoch": 1.1731783446345927,
-      "grad_norm": 0.5658952593803406,
-      "learning_rate": 0.0005302122616097402,
-      "loss": 3.7458,
+      "epoch": 1.1752021563342319,
+      "grad_norm": 1.1655031442642212,
+      "learning_rate": 0.0005300852671343766,
+      "loss": 3.7502,
       "step": 10900
     },
     {
-      "epoch": 1.1785598966742008,
-      "grad_norm": 0.5859235525131226,
-      "learning_rate": 0.0005298890205796789,
-      "loss": 3.7476,
+      "epoch": 1.1805929919137466,
+      "grad_norm": 0.720877468585968,
+      "learning_rate": 0.0005297614678899081,
+      "loss": 3.7432,
       "step": 10950
     },
     {
-      "epoch": 1.1839414487138091,
-      "grad_norm": 0.5732556581497192,
-      "learning_rate": 0.0005295657795496175,
-      "loss": 3.7672,
+      "epoch": 1.1859838274932615,
+      "grad_norm": 0.5574724078178406,
+      "learning_rate": 0.0005294376686454398,
+      "loss": 3.7336,
       "step": 11000
     },
     {
-      "epoch": 1.1839414487138091,
-      "eval_accuracy": 0.3474046790931526,
-      "eval_loss": 3.7189364433288574,
-      "eval_runtime": 183.8795,
-      "eval_samples_per_second": 97.95,
-      "eval_steps_per_second": 6.124,
+      "epoch": 1.1859838274932615,
+      "eval_accuracy": 0.3475620084747079,
+      "eval_loss": 3.7182576656341553,
+      "eval_runtime": 183.8299,
+      "eval_samples_per_second": 97.976,
+      "eval_steps_per_second": 6.125,
       "step": 11000
     },
     {
-      "epoch": 1.1893230007534172,
-      "grad_norm": 0.6487847566604614,
-      "learning_rate": 0.0005292425385195561,
-      "loss": 3.7339,
+      "epoch": 1.1913746630727764,
+      "grad_norm": 0.573558509349823,
+      "learning_rate": 0.0005291138694009714,
+      "loss": 3.7453,
       "step": 11050
     },
     {
-      "epoch": 1.1947045527930256,
-      "grad_norm": 0.5545188188552856,
-      "learning_rate": 0.0005289192974894946,
-      "loss": 3.7522,
+      "epoch": 1.196765498652291,
+      "grad_norm": 0.6689596176147461,
+      "learning_rate": 0.0005287900701565029,
+      "loss": 3.7372,
       "step": 11100
     },
     {
-      "epoch": 1.2000861048326337,
-      "grad_norm": 0.5902015566825867,
-      "learning_rate": 0.0005285960564594331,
-      "loss": 3.7513,
+      "epoch": 1.202156334231806,
+      "grad_norm": 0.6200105547904968,
+      "learning_rate": 0.0005284662709120345,
+      "loss": 3.7373,
       "step": 11150
     },
     {
-      "epoch": 1.205467656872242,
-      "grad_norm": 0.5698803663253784,
-      "learning_rate": 0.0005282728154293718,
-      "loss": 3.7605,
+      "epoch": 1.2075471698113207,
+      "grad_norm": 0.7147562503814697,
+      "learning_rate": 0.0005281424716675661,
+      "loss": 3.7346,
       "step": 11200
     },
     {
-      "epoch": 1.21084920891185,
-      "grad_norm": 0.5888674855232239,
-      "learning_rate": 0.0005279495743993104,
-      "loss": 3.7523,
+      "epoch": 1.2129380053908356,
+      "grad_norm": 0.6362359523773193,
+      "learning_rate": 0.0005278186724230977,
+      "loss": 3.7283,
       "step": 11250
     },
     {
-      "epoch": 1.2162307609514584,
-      "grad_norm": 0.6065134406089783,
-      "learning_rate": 0.000527626333369249,
-      "loss": 3.7479,
+      "epoch": 1.2183288409703503,
+      "grad_norm": 0.5415824055671692,
+      "learning_rate": 0.0005274948731786292,
+      "loss": 3.7276,
       "step": 11300
     },
     {
-      "epoch": 1.2216123129910665,
-      "grad_norm": 0.6043328046798706,
-      "learning_rate": 0.0005273030923391875,
-      "loss": 3.7426,
+      "epoch": 1.2237196765498652,
+      "grad_norm": 0.6271039247512817,
+      "learning_rate": 0.0005271710739341608,
+      "loss": 3.7457,
       "step": 11350
     },
     {
-      "epoch": 1.2269938650306749,
-      "grad_norm": 0.6097142696380615,
-      "learning_rate": 0.0005269798513091261,
-      "loss": 3.7385,
+      "epoch": 1.2291105121293802,
+      "grad_norm": 0.5612584352493286,
+      "learning_rate": 0.0005268472746896923,
+      "loss": 3.7387,
       "step": 11400
     },
     {
-      "epoch": 1.232375417070283,
-      "grad_norm": 0.5741665959358215,
-      "learning_rate": 0.0005266566102790647,
-      "loss": 3.7502,
+      "epoch": 1.2345013477088949,
+      "grad_norm": 0.6260232329368591,
+      "learning_rate": 0.000526523475445224,
+      "loss": 3.7283,
       "step": 11450
     },
     {
-      "epoch": 1.2377569691098913,
-      "grad_norm": 0.5567603707313538,
-      "learning_rate": 0.0005263333692490033,
-      "loss": 3.7448,
+      "epoch": 1.2398921832884098,
+      "grad_norm": 0.5288992524147034,
+      "learning_rate": 0.0005261996762007554,
+      "loss": 3.7155,
       "step": 11500
     },
     {
-      "epoch": 1.2431385211494996,
-      "grad_norm": 0.5684623122215271,
-      "learning_rate": 0.0005260101282189419,
-      "loss": 3.7307,
+      "epoch": 1.2452830188679245,
+      "grad_norm": 0.5613659024238586,
+      "learning_rate": 0.0005258758769562871,
+      "loss": 3.7368,
       "step": 11550
     },
     {
-      "epoch": 1.2485200731891077,
-      "grad_norm": 0.5747365355491638,
-      "learning_rate": 0.0005256868871888804,
-      "loss": 3.7506,
+      "epoch": 1.2506738544474394,
+      "grad_norm": 0.621526837348938,
+      "learning_rate": 0.0005255520777118186,
+      "loss": 3.7384,
       "step": 11600
     },
     {
-      "epoch": 1.2539016252287158,
-      "grad_norm": 0.5683662295341492,
-      "learning_rate": 0.0005253636461588191,
-      "loss": 3.7345,
+      "epoch": 1.256064690026954,
+      "grad_norm": 0.5767836570739746,
+      "learning_rate": 0.0005252282784673502,
+      "loss": 3.7306,
       "step": 11650
     },
     {
-      "epoch": 1.2592831772683242,
-      "grad_norm": 0.5467516183853149,
-      "learning_rate": 0.0005250404051287576,
-      "loss": 3.7428,
+      "epoch": 1.261455525606469,
+      "grad_norm": 0.5777313113212585,
+      "learning_rate": 0.0005249044792228817,
+      "loss": 3.7324,
       "step": 11700
     },
     {
-      "epoch": 1.2646647293079325,
-      "grad_norm": 0.5645882487297058,
-      "learning_rate": 0.0005247171640986962,
-      "loss": 3.7222,
+      "epoch": 1.266846361185984,
+      "grad_norm": 0.5307707786560059,
+      "learning_rate": 0.0005245806799784133,
+      "loss": 3.726,
       "step": 11750
     },
     {
-      "epoch": 1.2700462813475406,
-      "grad_norm": 0.5676699280738831,
-      "learning_rate": 0.0005243939230686347,
-      "loss": 3.7351,
+      "epoch": 1.2722371967654986,
+      "grad_norm": 0.6157434582710266,
+      "learning_rate": 0.0005242568807339449,
+      "loss": 3.7227,
       "step": 11800
     },
     {
-      "epoch": 1.275427833387149,
-      "grad_norm": 0.5621583461761475,
-      "learning_rate": 0.0005240706820385734,
-      "loss": 3.744,
+      "epoch": 1.2776280323450135,
+      "grad_norm": 0.5853699445724487,
+      "learning_rate": 0.0005239330814894765,
+      "loss": 3.7163,
       "step": 11850
     },
     {
-      "epoch": 1.280809385426757,
-      "grad_norm": 0.6436269879341125,
-      "learning_rate": 0.000523747441008512,
-      "loss": 3.7413,
+      "epoch": 1.2830188679245282,
+      "grad_norm": 0.5218607783317566,
+      "learning_rate": 0.000523609282245008,
+      "loss": 3.7381,
       "step": 11900
     },
     {
-      "epoch": 1.2861909374663654,
-      "grad_norm": 0.5486816167831421,
-      "learning_rate": 0.0005234241999784506,
-      "loss": 3.7275,
+      "epoch": 1.2884097035040432,
+      "grad_norm": 0.5536827445030212,
+      "learning_rate": 0.0005232854830005396,
+      "loss": 3.7357,
       "step": 11950
     },
     {
-      "epoch": 1.2915724895059735,
-      "grad_norm": 0.5652892589569092,
-      "learning_rate": 0.0005231009589483891,
-      "loss": 3.7053,
+      "epoch": 1.2938005390835579,
+      "grad_norm": 0.5603932738304138,
+      "learning_rate": 0.0005229616837560712,
+      "loss": 3.7301,
       "step": 12000
     },
     {
-      "epoch": 1.2915724895059735,
-      "eval_accuracy": 0.34900806975863086,
-      "eval_loss": 3.6980960369110107,
-      "eval_runtime": 183.6237,
-      "eval_samples_per_second": 98.086,
-      "eval_steps_per_second": 6.132,
+      "epoch": 1.2938005390835579,
+      "eval_accuracy": 0.34970399150508263,
+      "eval_loss": 3.697612762451172,
+      "eval_runtime": 184.1246,
+      "eval_samples_per_second": 97.82,
+      "eval_steps_per_second": 6.115,
       "step": 12000
     },
     {
-      "epoch": 1.2969540415455818,
-      "grad_norm": 0.5666853785514832,
-      "learning_rate": 0.0005227777179183277,
-      "loss": 3.7347,
+      "epoch": 1.2991913746630728,
+      "grad_norm": 0.5378450751304626,
+      "learning_rate": 0.0005226378845116028,
+      "loss": 3.7119,
       "step": 12050
     },
     {
-      "epoch": 1.30233559358519,
-      "grad_norm": 0.6039916276931763,
-      "learning_rate": 0.0005224544768882663,
-      "loss": 3.7439,
+      "epoch": 1.3045822102425877,
+      "grad_norm": 0.593679666519165,
+      "learning_rate": 0.0005223140852671344,
+      "loss": 3.7245,
       "step": 12100
     },
     {
-      "epoch": 1.3077171456247982,
-      "grad_norm": 0.5985672473907471,
-      "learning_rate": 0.0005221312358582049,
-      "loss": 3.7142,
+      "epoch": 1.3099730458221024,
+      "grad_norm": 0.6228883266448975,
+      "learning_rate": 0.0005219902860226659,
+      "loss": 3.7351,
       "step": 12150
     },
     {
-      "epoch": 1.3130986976644063,
-      "grad_norm": 0.5667858123779297,
-      "learning_rate": 0.0005218079948281435,
-      "loss": 3.7405,
+      "epoch": 1.3153638814016173,
+      "grad_norm": 0.582563042640686,
+      "learning_rate": 0.0005216664867781975,
+      "loss": 3.7346,
       "step": 12200
     },
     {
-      "epoch": 1.3184802497040147,
-      "grad_norm": 0.5678979754447937,
-      "learning_rate": 0.000521484753798082,
-      "loss": 3.737,
+      "epoch": 1.320754716981132,
+      "grad_norm": 0.6171168684959412,
+      "learning_rate": 0.000521342687533729,
+      "loss": 3.7159,
       "step": 12250
     },
     {
-      "epoch": 1.3238618017436228,
-      "grad_norm": 0.5826097130775452,
-      "learning_rate": 0.0005211615127680207,
-      "loss": 3.7052,
+      "epoch": 1.326145552560647,
+      "grad_norm": 0.5984110832214355,
+      "learning_rate": 0.0005210188882892606,
+      "loss": 3.725,
       "step": 12300
     },
     {
-      "epoch": 1.329243353783231,
-      "grad_norm": 0.5644336938858032,
-      "learning_rate": 0.0005208382717379592,
-      "loss": 3.7091,
+      "epoch": 1.3315363881401616,
+      "grad_norm": 0.5445953011512756,
+      "learning_rate": 0.0005206950890447922,
+      "loss": 3.7152,
       "step": 12350
     },
     {
-      "epoch": 1.3346249058228392,
-      "grad_norm": 0.5295429229736328,
-      "learning_rate": 0.0005205150307078979,
-      "loss": 3.7178,
+      "epoch": 1.3369272237196765,
+      "grad_norm": 0.519636869430542,
+      "learning_rate": 0.0005203712898003238,
+      "loss": 3.7301,
       "step": 12400
     },
     {
-      "epoch": 1.3400064578624475,
-      "grad_norm": 0.6082366704940796,
-      "learning_rate": 0.0005201917896778364,
-      "loss": 3.7232,
+      "epoch": 1.3423180592991915,
+      "grad_norm": 0.6377416253089905,
+      "learning_rate": 0.0005200474905558553,
+      "loss": 3.7191,
       "step": 12450
     },
     {
-      "epoch": 1.3453880099020559,
-      "grad_norm": 0.5737263560295105,
-      "learning_rate": 0.0005198685486477749,
-      "loss": 3.7125,
+      "epoch": 1.3477088948787062,
+      "grad_norm": 0.6142134666442871,
+      "learning_rate": 0.0005197236913113869,
+      "loss": 3.7213,
       "step": 12500
     },
     {
-      "epoch": 1.350769561941664,
-      "grad_norm": 0.6570215225219727,
-      "learning_rate": 0.0005195453076177136,
-      "loss": 3.7108,
+      "epoch": 1.353099730458221,
+      "grad_norm": 0.5391523838043213,
+      "learning_rate": 0.0005193998920669184,
+      "loss": 3.7277,
       "step": 12550
     },
     {
-      "epoch": 1.356151113981272,
-      "grad_norm": 0.5982895493507385,
-      "learning_rate": 0.0005192220665876521,
-      "loss": 3.7327,
+      "epoch": 1.3584905660377358,
+      "grad_norm": 0.5840072631835938,
+      "learning_rate": 0.0005190760928224501,
+      "loss": 3.7111,
       "step": 12600
     },
     {
-      "epoch": 1.3615326660208804,
-      "grad_norm": 0.563357412815094,
-      "learning_rate": 0.0005188988255575907,
-      "loss": 3.7202,
+      "epoch": 1.3638814016172507,
+      "grad_norm": 0.550019383430481,
+      "learning_rate": 0.0005187522935779816,
+      "loss": 3.7188,
       "step": 12650
     },
     {
-      "epoch": 1.3669142180604887,
-      "grad_norm": 0.5622259378433228,
-      "learning_rate": 0.0005185755845275293,
-      "loss": 3.7272,
+      "epoch": 1.3692722371967654,
+      "grad_norm": 0.5857395529747009,
+      "learning_rate": 0.0005184284943335132,
+      "loss": 3.7328,
       "step": 12700
     },
     {
-      "epoch": 1.3722957701000968,
-      "grad_norm": 0.6199678778648376,
-      "learning_rate": 0.000518252343497468,
-      "loss": 3.715,
+      "epoch": 1.3746630727762803,
+      "grad_norm": 0.586082935333252,
+      "learning_rate": 0.0005181046950890447,
+      "loss": 3.7182,
       "step": 12750
     },
     {
-      "epoch": 1.3776773221397052,
-      "grad_norm": 0.5978006720542908,
-      "learning_rate": 0.0005179291024674065,
-      "loss": 3.7347,
+      "epoch": 1.3800539083557952,
+      "grad_norm": 0.5567058324813843,
+      "learning_rate": 0.0005177808958445764,
+      "loss": 3.7106,
       "step": 12800
     },
     {
-      "epoch": 1.3830588741793133,
-      "grad_norm": 0.5449814796447754,
-      "learning_rate": 0.000517605861437345,
-      "loss": 3.7236,
+      "epoch": 1.38544474393531,
+      "grad_norm": 0.6352502703666687,
+      "learning_rate": 0.0005174570966001078,
+      "loss": 3.7053,
       "step": 12850
     },
     {
-      "epoch": 1.3884404262189216,
-      "grad_norm": 0.6005086898803711,
-      "learning_rate": 0.0005172826204072836,
-      "loss": 3.7295,
+      "epoch": 1.3908355795148248,
+      "grad_norm": 0.5695269703865051,
+      "learning_rate": 0.0005171332973556395,
+      "loss": 3.7114,
       "step": 12900
     },
     {
-      "epoch": 1.3938219782585297,
-      "grad_norm": 0.5988556146621704,
-      "learning_rate": 0.0005169593793772222,
-      "loss": 3.7082,
+      "epoch": 1.3962264150943398,
+      "grad_norm": 0.5709618926048279,
+      "learning_rate": 0.000516809498111171,
+      "loss": 3.7107,
       "step": 12950
     },
     {
-      "epoch": 1.399203530298138,
-      "grad_norm": 0.5763736367225647,
-      "learning_rate": 0.0005166361383471609,
-      "loss": 3.7077,
+      "epoch": 1.4016172506738545,
+      "grad_norm": 0.5886682271957397,
+      "learning_rate": 0.0005164856988667026,
+      "loss": 3.7061,
       "step": 13000
     },
     {
-      "epoch": 1.399203530298138,
-      "eval_accuracy": 0.35155250308547037,
-      "eval_loss": 3.675130844116211,
-      "eval_runtime": 183.856,
-      "eval_samples_per_second": 97.963,
-      "eval_steps_per_second": 6.124,
+      "epoch": 1.4016172506738545,
+      "eval_accuracy": 0.3515271869626096,
+      "eval_loss": 3.675485134124756,
+      "eval_runtime": 183.8508,
+      "eval_samples_per_second": 97.965,
+      "eval_steps_per_second": 6.125,
       "step": 13000
     },
     {
-      "epoch": 1.4045850823377461,
-      "grad_norm": 0.5773831009864807,
-      "learning_rate": 0.0005163128973170994,
-      "loss": 3.6965,
+      "epoch": 1.4070080862533692,
+      "grad_norm": 0.5804755687713623,
+      "learning_rate": 0.0005161618996222341,
+      "loss": 3.7183,
       "step": 13050
     },
     {
-      "epoch": 1.4099666343773545,
-      "grad_norm": 0.5801844000816345,
-      "learning_rate": 0.000515989656287038,
-      "loss": 3.7124,
+      "epoch": 1.412398921832884,
+      "grad_norm": 0.5670456290245056,
+      "learning_rate": 0.0005158381003777657,
+      "loss": 3.7274,
       "step": 13100
     },
     {
-      "epoch": 1.4153481864169626,
-      "grad_norm": 0.5833303332328796,
-      "learning_rate": 0.0005156664152569766,
-      "loss": 3.7019,
+      "epoch": 1.417789757412399,
+      "grad_norm": 0.5603575706481934,
+      "learning_rate": 0.0005155143011332973,
+      "loss": 3.7077,
       "step": 13150
     },
     {
-      "epoch": 1.420729738456571,
-      "grad_norm": 0.5930039882659912,
-      "learning_rate": 0.0005153431742269152,
-      "loss": 3.708,
+      "epoch": 1.4231805929919137,
+      "grad_norm": 0.5187303423881531,
+      "learning_rate": 0.0005151905018888289,
+      "loss": 3.7094,
       "step": 13200
     },
     {
-      "epoch": 1.426111290496179,
-      "grad_norm": 0.5815587639808655,
-      "learning_rate": 0.0005150199331968538,
-      "loss": 3.7094,
+      "epoch": 1.4285714285714286,
+      "grad_norm": 0.5651947259902954,
+      "learning_rate": 0.0005148731786292498,
+      "loss": 3.704,
       "step": 13250
     },
     {
-      "epoch": 1.4314928425357873,
-      "grad_norm": 0.5351225137710571,
-      "learning_rate": 0.0005146966921667923,
-      "loss": 3.7122,
+      "epoch": 1.4339622641509435,
+      "grad_norm": 0.5489131808280945,
+      "learning_rate": 0.0005145493793847814,
+      "loss": 3.7031,
       "step": 13300
     },
     {
-      "epoch": 1.4368743945753955,
-      "grad_norm": 0.5802038908004761,
-      "learning_rate": 0.0005143734511367309,
-      "loss": 3.7082,
+      "epoch": 1.4393530997304582,
+      "grad_norm": 0.5978963971138,
+      "learning_rate": 0.000514225580140313,
+      "loss": 3.689,
       "step": 13350
     },
     {
-      "epoch": 1.4422559466150038,
-      "grad_norm": 0.583487868309021,
-      "learning_rate": 0.0005140502101066695,
-      "loss": 3.6901,
+      "epoch": 1.444743935309973,
+      "grad_norm": 0.5894157290458679,
+      "learning_rate": 0.0005139017808958445,
+      "loss": 3.7002,
       "step": 13400
     },
     {
-      "epoch": 1.447637498654612,
-      "grad_norm": 0.5486982464790344,
-      "learning_rate": 0.0005137269690766081,
-      "loss": 3.7178,
+      "epoch": 1.4501347708894878,
+      "grad_norm": 0.5639384984970093,
+      "learning_rate": 0.0005135779816513762,
+      "loss": 3.7096,
       "step": 13450
     },
     {
-      "epoch": 1.4530190506942202,
-      "grad_norm": 0.6117085218429565,
-      "learning_rate": 0.0005134037280465466,
-      "loss": 3.726,
+      "epoch": 1.4555256064690028,
+      "grad_norm": 0.5487292408943176,
+      "learning_rate": 0.0005132541824069076,
+      "loss": 3.6927,
       "step": 13500
     },
     {
-      "epoch": 1.4584006027338283,
-      "grad_norm": 0.6490427851676941,
-      "learning_rate": 0.0005130804870164853,
-      "loss": 3.7114,
+      "epoch": 1.4609164420485174,
+      "grad_norm": 0.5995162725448608,
+      "learning_rate": 0.0005129303831624393,
+      "loss": 3.7069,
       "step": 13550
     },
     {
-      "epoch": 1.4637821547734367,
-      "grad_norm": 0.5825288891792297,
-      "learning_rate": 0.0005127572459864238,
-      "loss": 3.6955,
+      "epoch": 1.4663072776280324,
+      "grad_norm": 0.5789255499839783,
+      "learning_rate": 0.0005126065839179708,
+      "loss": 3.7079,
       "step": 13600
     },
     {
-      "epoch": 1.469163706813045,
-      "grad_norm": 0.6085303425788879,
-      "learning_rate": 0.0005124340049563624,
-      "loss": 3.6936,
+      "epoch": 1.4716981132075473,
+      "grad_norm": 0.5981630682945251,
+      "learning_rate": 0.0005122827846735024,
+      "loss": 3.6824,
       "step": 13650
     },
     {
-      "epoch": 1.474545258852653,
-      "grad_norm": 0.585913896560669,
-      "learning_rate": 0.000512110763926301,
-      "loss": 3.7044,
+      "epoch": 1.477088948787062,
+      "grad_norm": 0.5376117825508118,
+      "learning_rate": 0.0005119589854290339,
+      "loss": 3.6785,
       "step": 13700
     },
     {
-      "epoch": 1.4799268108922612,
-      "grad_norm": 0.5770050287246704,
-      "learning_rate": 0.0005117875228962395,
-      "loss": 3.7319,
+      "epoch": 1.482479784366577,
+      "grad_norm": 0.5373963713645935,
+      "learning_rate": 0.0005116351861845655,
+      "loss": 3.6994,
       "step": 13750
     },
     {
-      "epoch": 1.4853083629318695,
-      "grad_norm": 0.6257504224777222,
-      "learning_rate": 0.0005114642818661782,
-      "loss": 3.685,
+      "epoch": 1.4878706199460916,
+      "grad_norm": 0.624566912651062,
+      "learning_rate": 0.0005113113869400971,
+      "loss": 3.7128,
       "step": 13800
     },
     {
-      "epoch": 1.4906899149714778,
-      "grad_norm": 0.5979099273681641,
-      "learning_rate": 0.0005111410408361168,
-      "loss": 3.7018,
+      "epoch": 1.4932614555256065,
+      "grad_norm": 0.6030147075653076,
+      "learning_rate": 0.0005109875876956287,
+      "loss": 3.7068,
       "step": 13850
     },
     {
-      "epoch": 1.496071467011086,
-      "grad_norm": 0.5666427612304688,
-      "learning_rate": 0.0005108177998060554,
-      "loss": 3.6985,
+      "epoch": 1.4986522911051212,
+      "grad_norm": 0.6407404541969299,
+      "learning_rate": 0.0005106637884511602,
+      "loss": 3.7212,
       "step": 13900
     },
     {
-      "epoch": 1.501453019050694,
-      "grad_norm": 0.5651612281799316,
-      "learning_rate": 0.0005104945587759939,
-      "loss": 3.6947,
+      "epoch": 1.5040431266846361,
+      "grad_norm": 0.5542694330215454,
+      "learning_rate": 0.0005103399892066918,
+      "loss": 3.6961,
       "step": 13950
     },
     {
-      "epoch": 1.5068345710903024,
-      "grad_norm": 0.5814973711967468,
-      "learning_rate": 0.0005101713177459324,
-      "loss": 3.6749,
+      "epoch": 1.509433962264151,
+      "grad_norm": 0.6078199148178101,
+      "learning_rate": 0.0005100161899622234,
+      "loss": 3.6945,
       "step": 14000
     },
     {
-      "epoch": 1.5068345710903024,
-      "eval_accuracy": 0.353589092814665,
-      "eval_loss": 3.654818534851074,
-      "eval_runtime": 183.9334,
-      "eval_samples_per_second": 97.921,
-      "eval_steps_per_second": 6.122,
+      "epoch": 1.509433962264151,
+      "eval_accuracy": 0.35363820391995715,
+      "eval_loss": 3.6544554233551025,
+      "eval_runtime": 183.8873,
+      "eval_samples_per_second": 97.946,
+      "eval_steps_per_second": 6.123,
       "step": 14000
     },
     {
-      "epoch": 1.5122161231299107,
-      "grad_norm": 0.6138876080513,
-      "learning_rate": 0.0005098480767158711,
-      "loss": 3.6879,
+      "epoch": 1.5148247978436657,
+      "grad_norm": 0.5554205179214478,
+      "learning_rate": 0.000509692390717755,
+      "loss": 3.6866,
       "step": 14050
     },
     {
-      "epoch": 1.5175976751695188,
-      "grad_norm": 0.5884425640106201,
-      "learning_rate": 0.0005095248356858097,
-      "loss": 3.6825,
+      "epoch": 1.5202156334231804,
+      "grad_norm": 0.6609663367271423,
+      "learning_rate": 0.0005093685914732865,
+      "loss": 3.6831,
       "step": 14100
     },
     {
-      "epoch": 1.5229792272091272,
-      "grad_norm": 0.5270723700523376,
-      "learning_rate": 0.0005092015946557483,
-      "loss": 3.691,
+      "epoch": 1.5256064690026954,
+      "grad_norm": 0.5601198673248291,
+      "learning_rate": 0.0005090447922288181,
+      "loss": 3.6817,
       "step": 14150
     },
     {
-      "epoch": 1.5283607792487355,
-      "grad_norm": 0.5885605216026306,
-      "learning_rate": 0.0005088783536256868,
-      "loss": 3.6815,
+      "epoch": 1.5309973045822103,
+      "grad_norm": 0.5430298447608948,
+      "learning_rate": 0.0005087209929843496,
+      "loss": 3.7151,
       "step": 14200
     },
     {
-      "epoch": 1.5337423312883436,
-      "grad_norm": 0.5387929081916809,
-      "learning_rate": 0.0005085551125956255,
-      "loss": 3.6705,
+      "epoch": 1.536388140161725,
+      "grad_norm": 0.5450032949447632,
+      "learning_rate": 0.0005083971937398812,
+      "loss": 3.6871,
       "step": 14250
     },
     {
-      "epoch": 1.5391238833279517,
-      "grad_norm": 0.5725566744804382,
-      "learning_rate": 0.000508231871565564,
-      "loss": 3.6678,
+      "epoch": 1.54177897574124,
+      "grad_norm": 0.6001717448234558,
+      "learning_rate": 0.0005080733944954127,
+      "loss": 3.6907,
       "step": 14300
     },
     {
-      "epoch": 1.54450543536756,
-      "grad_norm": 0.6244074106216431,
-      "learning_rate": 0.0005079086305355026,
-      "loss": 3.6958,
+      "epoch": 1.5471698113207548,
+      "grad_norm": 0.5790493488311768,
+      "learning_rate": 0.0005077495952509444,
+      "loss": 3.6737,
       "step": 14350
     },
     {
-      "epoch": 1.5498869874071683,
-      "grad_norm": 0.5698248147964478,
-      "learning_rate": 0.0005075853895054412,
-      "loss": 3.6911,
+      "epoch": 1.5525606469002695,
+      "grad_norm": 0.5388293862342834,
+      "learning_rate": 0.0005074257960064759,
+      "loss": 3.6859,
       "step": 14400
     },
     {
-      "epoch": 1.5552685394467765,
-      "grad_norm": 0.6287437081336975,
-      "learning_rate": 0.0005072621484753797,
-      "loss": 3.6956,
+      "epoch": 1.5579514824797842,
+      "grad_norm": 0.5991535782814026,
+      "learning_rate": 0.0005071019967620075,
+      "loss": 3.6928,
       "step": 14450
     },
     {
-      "epoch": 1.5606500914863846,
-      "grad_norm": 0.5749585628509521,
-      "learning_rate": 0.0005069389074453184,
-      "loss": 3.6836,
+      "epoch": 1.5633423180592994,
+      "grad_norm": 0.532653272151947,
+      "learning_rate": 0.000506778197517539,
+      "loss": 3.6826,
       "step": 14500
     },
     {
-      "epoch": 1.566031643525993,
-      "grad_norm": 0.8090015053749084,
-      "learning_rate": 0.0005066156664152569,
-      "loss": 3.694,
+      "epoch": 1.568733153638814,
+      "grad_norm": 0.5869577527046204,
+      "learning_rate": 0.0005064543982730707,
+      "loss": 3.6943,
       "step": 14550
     },
     {
-      "epoch": 1.5714131955656012,
-      "grad_norm": 0.5545092821121216,
-      "learning_rate": 0.0005062924253851955,
-      "loss": 3.6698,
+      "epoch": 1.5741239892183287,
+      "grad_norm": 0.6177583932876587,
+      "learning_rate": 0.0005061305990286023,
+      "loss": 3.6851,
       "step": 14600
     },
     {
-      "epoch": 1.5767947476052093,
-      "grad_norm": 0.5888755321502686,
-      "learning_rate": 0.0005059691843551341,
-      "loss": 3.665,
+      "epoch": 1.5795148247978437,
+      "grad_norm": 0.6138556003570557,
+      "learning_rate": 0.0005058067997841338,
+      "loss": 3.6947,
       "step": 14650
     },
     {
-      "epoch": 1.5821762996448174,
-      "grad_norm": 0.5493360161781311,
-      "learning_rate": 0.0005056459433250727,
-      "loss": 3.681,
+      "epoch": 1.5849056603773586,
+      "grad_norm": 0.5741438865661621,
+      "learning_rate": 0.0005054830005396654,
+      "loss": 3.6807,
       "step": 14700
     },
     {
-      "epoch": 1.5875578516844258,
-      "grad_norm": 0.5884751081466675,
-      "learning_rate": 0.0005053227022950113,
-      "loss": 3.6865,
+      "epoch": 1.5902964959568733,
+      "grad_norm": 0.6064146161079407,
+      "learning_rate": 0.0005051592012951969,
+      "loss": 3.6746,
       "step": 14750
     },
     {
-      "epoch": 1.592939403724034,
-      "grad_norm": 0.5413837432861328,
-      "learning_rate": 0.0005049994612649499,
-      "loss": 3.6916,
+      "epoch": 1.595687331536388,
+      "grad_norm": 0.5631973147392273,
+      "learning_rate": 0.0005048354020507286,
+      "loss": 3.6921,
       "step": 14800
     },
     {
-      "epoch": 1.5983209557636422,
-      "grad_norm": 0.53194260597229,
-      "learning_rate": 0.0005046762202348884,
-      "loss": 3.6894,
+      "epoch": 1.6010781671159031,
+      "grad_norm": 0.5781230926513672,
+      "learning_rate": 0.00050451160280626,
+      "loss": 3.6949,
       "step": 14850
     },
     {
-      "epoch": 1.6037025078032503,
-      "grad_norm": 0.5218849182128906,
-      "learning_rate": 0.000504352979204827,
-      "loss": 3.6831,
+      "epoch": 1.6064690026954178,
+      "grad_norm": 0.5288845896720886,
+      "learning_rate": 0.0005041878035617917,
+      "loss": 3.6835,
       "step": 14900
     },
     {
-      "epoch": 1.6090840598428586,
-      "grad_norm": 0.5776587128639221,
-      "learning_rate": 0.0005040297381747656,
-      "loss": 3.6722,
+      "epoch": 1.6118598382749325,
+      "grad_norm": 0.5479040145874023,
+      "learning_rate": 0.0005038640043173232,
+      "loss": 3.679,
       "step": 14950
     },
     {
-      "epoch": 1.614465611882467,
-      "grad_norm": 0.5039575099945068,
-      "learning_rate": 0.0005037064971447042,
-      "loss": 3.6631,
+      "epoch": 1.6172506738544474,
+      "grad_norm": 0.6492672562599182,
+      "learning_rate": 0.0005035402050728548,
+      "loss": 3.674,
       "step": 15000
     },
     {
-      "epoch": 1.614465611882467,
-      "eval_accuracy": 0.3553025488554994,
-      "eval_loss": 3.6372177600860596,
-      "eval_runtime": 183.9276,
-      "eval_samples_per_second": 97.924,
-      "eval_steps_per_second": 6.122,
+      "epoch": 1.6172506738544474,
+      "eval_accuracy": 0.3557005443835638,
+      "eval_loss": 3.63702130317688,
+      "eval_runtime": 183.8633,
+      "eval_samples_per_second": 97.959,
+      "eval_steps_per_second": 6.124,
       "step": 15000
     },
     {
-      "epoch": 1.619847163922075,
-      "grad_norm": 0.6165758967399597,
-      "learning_rate": 0.0005033832561146428,
-      "loss": 3.6742,
+      "epoch": 1.6226415094339623,
+      "grad_norm": 0.5320624113082886,
+      "learning_rate": 0.0005032164058283863,
+      "loss": 3.6766,
       "step": 15050
     },
     {
-      "epoch": 1.6252287159616834,
-      "grad_norm": 0.6052930951118469,
-      "learning_rate": 0.0005030600150845813,
-      "loss": 3.6831,
+      "epoch": 1.628032345013477,
+      "grad_norm": 0.5401520133018494,
+      "learning_rate": 0.0005028926065839179,
+      "loss": 3.6719,
       "step": 15100
     },
     {
-      "epoch": 1.6306102680012917,
-      "grad_norm": 0.561308741569519,
-      "learning_rate": 0.00050273677405452,
-      "loss": 3.6891,
+      "epoch": 1.633423180592992,
+      "grad_norm": 0.6083766222000122,
+      "learning_rate": 0.0005025688073394495,
+      "loss": 3.6746,
       "step": 15150
     },
     {
-      "epoch": 1.6359918200408998,
-      "grad_norm": 0.5598635077476501,
-      "learning_rate": 0.0005024135330244585,
-      "loss": 3.6678,
+      "epoch": 1.6388140161725069,
+      "grad_norm": 0.6010870337486267,
+      "learning_rate": 0.0005022450080949811,
+      "loss": 3.6791,
       "step": 15200
     },
     {
-      "epoch": 1.641373372080508,
-      "grad_norm": 0.6275472044944763,
-      "learning_rate": 0.0005020902919943972,
-      "loss": 3.6804,
+      "epoch": 1.6442048517520216,
+      "grad_norm": 0.6043290495872498,
+      "learning_rate": 0.0005019212088505126,
+      "loss": 3.681,
       "step": 15250
     },
     {
-      "epoch": 1.6467549241201163,
-      "grad_norm": 0.6472123265266418,
-      "learning_rate": 0.0005017670509643357,
-      "loss": 3.661,
+      "epoch": 1.6495956873315363,
+      "grad_norm": 0.5123810172080994,
+      "learning_rate": 0.0005015974096060442,
+      "loss": 3.6828,
       "step": 15300
     },
     {
-      "epoch": 1.6521364761597246,
-      "grad_norm": 0.5375542044639587,
-      "learning_rate": 0.0005014438099342743,
-      "loss": 3.6656,
+      "epoch": 1.6549865229110512,
+      "grad_norm": 0.5891865491867065,
+      "learning_rate": 0.0005012736103615758,
+      "loss": 3.6674,
       "step": 15350
     },
     {
-      "epoch": 1.6575180281993327,
-      "grad_norm": 0.5594198107719421,
-      "learning_rate": 0.0005011205689042129,
-      "loss": 3.6852,
+      "epoch": 1.6603773584905661,
+      "grad_norm": 0.5856804847717285,
+      "learning_rate": 0.0005009498111171074,
+      "loss": 3.671,
       "step": 15400
     },
     {
-      "epoch": 1.6628995802389408,
-      "grad_norm": 0.5270388722419739,
-      "learning_rate": 0.0005007973278741514,
-      "loss": 3.6723,
+      "epoch": 1.6657681940700808,
+      "grad_norm": 0.5844360589981079,
+      "learning_rate": 0.0005006260118726389,
+      "loss": 3.6715,
       "step": 15450
     },
     {
-      "epoch": 1.6682811322785491,
-      "grad_norm": 0.5340932607650757,
-      "learning_rate": 0.00050047408684409,
-      "loss": 3.6879,
+      "epoch": 1.6711590296495957,
+      "grad_norm": 0.5972587466239929,
+      "learning_rate": 0.0005003022126281705,
+      "loss": 3.6718,
       "step": 15500
     },
     {
-      "epoch": 1.6736626843181575,
-      "grad_norm": 0.5946788787841797,
-      "learning_rate": 0.0005001508458140286,
-      "loss": 3.6621,
+      "epoch": 1.6765498652291106,
+      "grad_norm": 0.6326434016227722,
+      "learning_rate": 0.000499978413383702,
+      "loss": 3.6659,
       "step": 15550
     },
     {
-      "epoch": 1.6790442363577656,
-      "grad_norm": 0.5601723790168762,
-      "learning_rate": 0.0004998276047839673,
-      "loss": 3.6689,
+      "epoch": 1.6819407008086253,
+      "grad_norm": 0.5700421333312988,
+      "learning_rate": 0.0004996546141392336,
+      "loss": 3.6612,
       "step": 15600
     },
     {
-      "epoch": 1.6844257883973737,
-      "grad_norm": 0.5660136342048645,
-      "learning_rate": 0.0004995043637539058,
-      "loss": 3.6798,
+      "epoch": 1.68733153638814,
+      "grad_norm": 0.5672731399536133,
+      "learning_rate": 0.0004993308148947651,
+      "loss": 3.671,
       "step": 15650
     },
     {
-      "epoch": 1.689807340436982,
-      "grad_norm": 0.5419800281524658,
-      "learning_rate": 0.0004991811227238443,
-      "loss": 3.6538,
+      "epoch": 1.692722371967655,
+      "grad_norm": 0.4939505159854889,
+      "learning_rate": 0.0004990070156502968,
+      "loss": 3.6605,
       "step": 15700
     },
     {
-      "epoch": 1.6951888924765903,
-      "grad_norm": 0.6684157252311707,
-      "learning_rate": 0.0004988578816937829,
-      "loss": 3.6607,
+      "epoch": 1.6981132075471699,
+      "grad_norm": 0.5950140953063965,
+      "learning_rate": 0.0004986832164058284,
+      "loss": 3.6563,
       "step": 15750
     },
     {
-      "epoch": 1.7005704445161984,
-      "grad_norm": 0.5901089310646057,
-      "learning_rate": 0.0004985346406637215,
-      "loss": 3.667,
+      "epoch": 1.7035040431266846,
+      "grad_norm": 0.5389714241027832,
+      "learning_rate": 0.0004983594171613599,
+      "loss": 3.6703,
       "step": 15800
     },
     {
-      "epoch": 1.7059519965558065,
-      "grad_norm": 0.644408106803894,
-      "learning_rate": 0.0004982113996336602,
-      "loss": 3.6735,
+      "epoch": 1.7088948787061995,
+      "grad_norm": 0.566318929195404,
+      "learning_rate": 0.0004980356179168915,
+      "loss": 3.678,
       "step": 15850
     },
     {
-      "epoch": 1.7113335485954149,
-      "grad_norm": 0.5586075186729431,
-      "learning_rate": 0.0004978881586035987,
-      "loss": 3.6414,
+      "epoch": 1.7142857142857144,
+      "grad_norm": 0.566290557384491,
+      "learning_rate": 0.000497711818672423,
+      "loss": 3.6991,
       "step": 15900
     },
     {
-      "epoch": 1.7167151006350232,
-      "grad_norm": 0.5893074870109558,
-      "learning_rate": 0.0004975649175735373,
-      "loss": 3.6631,
+      "epoch": 1.719676549865229,
+      "grad_norm": 0.5044989585876465,
+      "learning_rate": 0.0004973880194279547,
+      "loss": 3.665,
       "step": 15950
     },
     {
-      "epoch": 1.7220966526746313,
-      "grad_norm": 0.5630667805671692,
-      "learning_rate": 0.0004972416765434759,
-      "loss": 3.6705,
+      "epoch": 1.7250673854447438,
+      "grad_norm": 0.5576607584953308,
+      "learning_rate": 0.0004970642201834862,
+      "loss": 3.6723,
       "step": 16000
     },
     {
-      "epoch": 1.7220966526746313,
-      "eval_accuracy": 0.3572014753758331,
-      "eval_loss": 3.6194918155670166,
-      "eval_runtime": 184.0585,
-      "eval_samples_per_second": 97.855,
-      "eval_steps_per_second": 6.118,
+      "epoch": 1.7250673854447438,
+      "eval_accuracy": 0.35732599158526845,
+      "eval_loss": 3.6155223846435547,
+      "eval_runtime": 183.6856,
+      "eval_samples_per_second": 98.053,
+      "eval_steps_per_second": 6.13,
       "step": 16000
     },
     {
-      "epoch": 1.7274782047142396,
-      "grad_norm": 0.6027358174324036,
-      "learning_rate": 0.0004969249003340157,
-      "loss": 3.6473,
+      "epoch": 1.7304582210242587,
+      "grad_norm": 0.5495242476463318,
+      "learning_rate": 0.0004967404209390178,
+      "loss": 3.6426,
       "step": 16050
     },
     {
-      "epoch": 1.732859756753848,
-      "grad_norm": 0.558164656162262,
-      "learning_rate": 0.0004966016593039543,
-      "loss": 3.6848,
+      "epoch": 1.7358490566037736,
+      "grad_norm": 0.5208356976509094,
+      "learning_rate": 0.0004964166216945493,
+      "loss": 3.65,
       "step": 16100
     },
     {
-      "epoch": 1.738241308793456,
-      "grad_norm": 0.6493954658508301,
-      "learning_rate": 0.0004962784182738928,
-      "loss": 3.6784,
+      "epoch": 1.7412398921832883,
+      "grad_norm": 0.539473831653595,
+      "learning_rate": 0.000496092822450081,
+      "loss": 3.6633,
       "step": 16150
     },
     {
-      "epoch": 1.7436228608330642,
-      "grad_norm": 0.5846709609031677,
-      "learning_rate": 0.0004959551772438314,
-      "loss": 3.6728,
+      "epoch": 1.7466307277628033,
+      "grad_norm": 0.5842383503913879,
+      "learning_rate": 0.0004957690232056125,
+      "loss": 3.6548,
       "step": 16200
     },
     {
-      "epoch": 1.7490044128726725,
-      "grad_norm": 0.5477911233901978,
-      "learning_rate": 0.0004956319362137701,
-      "loss": 3.6677,
+      "epoch": 1.7520215633423182,
+      "grad_norm": 0.5521345734596252,
+      "learning_rate": 0.0004954452239611441,
+      "loss": 3.6612,
       "step": 16250
     },
     {
-      "epoch": 1.7543859649122808,
-      "grad_norm": 0.6721983551979065,
-      "learning_rate": 0.0004953086951837086,
-      "loss": 3.6634,
+      "epoch": 1.7574123989218329,
+      "grad_norm": 0.5061966180801392,
+      "learning_rate": 0.0004951214247166756,
+      "loss": 3.6518,
       "step": 16300
     },
     {
-      "epoch": 1.759767516951889,
-      "grad_norm": 0.6212129592895508,
-      "learning_rate": 0.0004949854541536472,
-      "loss": 3.6841,
+      "epoch": 1.7628032345013476,
+      "grad_norm": 0.5491743087768555,
+      "learning_rate": 0.0004947976254722072,
+      "loss": 3.6473,
       "step": 16350
     },
     {
-      "epoch": 1.765149068991497,
-      "grad_norm": 0.5700268149375916,
-      "learning_rate": 0.0004946622131235857,
-      "loss": 3.6504,
+      "epoch": 1.7681940700808625,
+      "grad_norm": 0.5807679295539856,
+      "learning_rate": 0.0004944738262277387,
+      "loss": 3.6619,
       "step": 16400
     },
     {
-      "epoch": 1.7705306210311054,
-      "grad_norm": 0.587509036064148,
-      "learning_rate": 0.0004943389720935244,
-      "loss": 3.6615,
+      "epoch": 1.7735849056603774,
+      "grad_norm": 0.5847117900848389,
+      "learning_rate": 0.0004941500269832703,
+      "loss": 3.6612,
       "step": 16450
     },
     {
-      "epoch": 1.7759121730707137,
-      "grad_norm": 0.5693418979644775,
-      "learning_rate": 0.000494015731063463,
-      "loss": 3.6447,
+      "epoch": 1.778975741239892,
+      "grad_norm": 0.5234289169311523,
+      "learning_rate": 0.0004938262277388019,
+      "loss": 3.6456,
       "step": 16500
     },
     {
-      "epoch": 1.7812937251103218,
-      "grad_norm": 0.5522615313529968,
-      "learning_rate": 0.0004936924900334016,
-      "loss": 3.6496,
+      "epoch": 1.784366576819407,
+      "grad_norm": 0.5795573592185974,
+      "learning_rate": 0.0004935024284943335,
+      "loss": 3.6425,
       "step": 16550
     },
     {
-      "epoch": 1.78667527714993,
-      "grad_norm": 0.5877909064292908,
-      "learning_rate": 0.0004933692490033401,
+      "epoch": 1.789757412398922,
+      "grad_norm": 0.5852517485618591,
+      "learning_rate": 0.000493178629249865,
       "loss": 3.6469,
       "step": 16600
     },
     {
-      "epoch": 1.7920568291895382,
-      "grad_norm": 0.5631406903266907,
-      "learning_rate": 0.0004930460079732786,
-      "loss": 3.6549,
+      "epoch": 1.7951482479784366,
+      "grad_norm": 0.550918698310852,
+      "learning_rate": 0.0004928548300053966,
+      "loss": 3.6576,
       "step": 16650
     },
     {
-      "epoch": 1.7974383812291466,
-      "grad_norm": 0.5792685151100159,
-      "learning_rate": 0.0004927227669432173,
-      "loss": 3.6396,
+      "epoch": 1.8005390835579513,
+      "grad_norm": 0.5831260681152344,
+      "learning_rate": 0.0004925310307609282,
+      "loss": 3.6495,
       "step": 16700
     },
     {
-      "epoch": 1.8028199332687547,
-      "grad_norm": 0.5577995777130127,
-      "learning_rate": 0.0004923995259131558,
-      "loss": 3.6507,
+      "epoch": 1.8059299191374663,
+      "grad_norm": 0.5384604334831238,
+      "learning_rate": 0.0004922072315164598,
+      "loss": 3.6432,
       "step": 16750
     },
     {
-      "epoch": 1.8082014853083628,
-      "grad_norm": 0.6078580021858215,
-      "learning_rate": 0.0004920762848830945,
-      "loss": 3.6691,
+      "epoch": 1.8113207547169812,
+      "grad_norm": 0.5421329736709595,
+      "learning_rate": 0.0004918834322719913,
+      "loss": 3.6478,
       "step": 16800
     },
     {
-      "epoch": 1.813583037347971,
-      "grad_norm": 0.540854275226593,
-      "learning_rate": 0.000491753043853033,
-      "loss": 3.6818,
+      "epoch": 1.8167115902964959,
+      "grad_norm": 0.5571365356445312,
+      "learning_rate": 0.0004915596330275229,
+      "loss": 3.6384,
       "step": 16850
     },
     {
-      "epoch": 1.8189645893875794,
-      "grad_norm": 0.566217839717865,
-      "learning_rate": 0.0004914298028229717,
-      "loss": 3.6256,
+      "epoch": 1.8221024258760108,
+      "grad_norm": 0.6123501658439636,
+      "learning_rate": 0.0004912358337830544,
+      "loss": 3.644,
       "step": 16900
     },
     {
-      "epoch": 1.8243461414271875,
-      "grad_norm": 0.5962033867835999,
-      "learning_rate": 0.0004911065617929102,
-      "loss": 3.6498,
+      "epoch": 1.8274932614555257,
+      "grad_norm": 0.5622369647026062,
+      "learning_rate": 0.000490912034538586,
+      "loss": 3.6504,
       "step": 16950
     },
     {
-      "epoch": 1.8297276934667959,
-      "grad_norm": 0.5795658230781555,
-      "learning_rate": 0.0004907833207628487,
-      "loss": 3.6589,
+      "epoch": 1.8328840970350404,
+      "grad_norm": 0.5058190822601318,
+      "learning_rate": 0.0004905882352941175,
+      "loss": 3.658,
       "step": 17000
     },
     {
-      "epoch": 1.8297276934667959,
-      "eval_accuracy": 0.3590019623798068,
-      "eval_loss": 3.603217840194702,
-      "eval_runtime": 184.2298,
-      "eval_samples_per_second": 97.764,
-      "eval_steps_per_second": 6.112,
+      "epoch": 1.8328840970350404,
+      "eval_accuracy": 0.35853345112755075,
+      "eval_loss": 3.6034040451049805,
+      "eval_runtime": 184.0647,
+      "eval_samples_per_second": 97.851,
+      "eval_steps_per_second": 6.117,
       "step": 17000
     },
     {
-      "epoch": 1.8351092455064042,
-      "grad_norm": 0.5930823087692261,
-      "learning_rate": 0.0004904600797327874,
-      "loss": 3.6381,
+      "epoch": 1.838274932614555,
+      "grad_norm": 0.5588904023170471,
+      "learning_rate": 0.0004902644360496492,
+      "loss": 3.6477,
       "step": 17050
     },
     {
-      "epoch": 1.8404907975460123,
-      "grad_norm": 0.5686311721801758,
-      "learning_rate": 0.0004901368387027259,
-      "loss": 3.662,
+      "epoch": 1.8436657681940702,
+      "grad_norm": 0.5480935573577881,
+      "learning_rate": 0.0004899406368051808,
+      "loss": 3.6383,
       "step": 17100
     },
     {
-      "epoch": 1.8458723495856204,
-      "grad_norm": 0.662773609161377,
-      "learning_rate": 0.0004898135976726646,
-      "loss": 3.6539,
+      "epoch": 1.849056603773585,
+      "grad_norm": 0.5919249653816223,
+      "learning_rate": 0.0004896168375607123,
+      "loss": 3.6386,
       "step": 17150
     },
     {
-      "epoch": 1.8512539016252287,
-      "grad_norm": 0.572503387928009,
-      "learning_rate": 0.0004894903566426031,
-      "loss": 3.6542,
+      "epoch": 1.8544474393530996,
+      "grad_norm": 0.6303591132164001,
+      "learning_rate": 0.0004892930383162439,
+      "loss": 3.6361,
       "step": 17200
     },
     {
-      "epoch": 1.856635453664837,
-      "grad_norm": 0.5672059059143066,
-      "learning_rate": 0.0004891671156125417,
-      "loss": 3.6453,
+      "epoch": 1.8598382749326146,
+      "grad_norm": 0.624796986579895,
+      "learning_rate": 0.0004889757150566648,
+      "loss": 3.6354,
       "step": 17250
     },
     {
-      "epoch": 1.8620170057044452,
-      "grad_norm": 0.6035781502723694,
-      "learning_rate": 0.0004888438745824803,
-      "loss": 3.655,
+      "epoch": 1.8652291105121295,
+      "grad_norm": 0.5499162673950195,
+      "learning_rate": 0.0004886519158121964,
+      "loss": 3.66,
       "step": 17300
     },
     {
-      "epoch": 1.8673985577440533,
-      "grad_norm": 0.5393502116203308,
-      "learning_rate": 0.0004885206335524189,
-      "loss": 3.6383,
+      "epoch": 1.8706199460916442,
+      "grad_norm": 0.6029002070426941,
+      "learning_rate": 0.000488328116567728,
+      "loss": 3.6619,
       "step": 17350
     },
     {
-      "epoch": 1.8727801097836616,
-      "grad_norm": 0.6165359616279602,
-      "learning_rate": 0.0004881973925223575,
-      "loss": 3.6602,
+      "epoch": 1.8760107816711589,
+      "grad_norm": 0.5472336411476135,
+      "learning_rate": 0.0004880043173232595,
+      "loss": 3.6411,
       "step": 17400
     },
     {
-      "epoch": 1.87816166182327,
-      "grad_norm": 0.5423830151557922,
-      "learning_rate": 0.00048787415149229604,
-      "loss": 3.6386,
+      "epoch": 1.881401617250674,
+      "grad_norm": 0.527107298374176,
+      "learning_rate": 0.0004876805180787911,
+      "loss": 3.6508,
       "step": 17450
     },
     {
-      "epoch": 1.883543213862878,
-      "grad_norm": 0.6336809992790222,
-      "learning_rate": 0.00048755091046223464,
-      "loss": 3.6604,
+      "epoch": 1.8867924528301887,
+      "grad_norm": 0.5641130208969116,
+      "learning_rate": 0.0004873567188343227,
+      "loss": 3.635,
       "step": 17500
     },
     {
-      "epoch": 1.8889247659024861,
-      "grad_norm": 0.5486282706260681,
-      "learning_rate": 0.0004872276694321732,
-      "loss": 3.6473,
+      "epoch": 1.8921832884097034,
+      "grad_norm": 0.578941285610199,
+      "learning_rate": 0.0004870329195898542,
+      "loss": 3.6486,
       "step": 17550
     },
     {
-      "epoch": 1.8943063179420945,
-      "grad_norm": 0.5997084975242615,
-      "learning_rate": 0.00048690442840211177,
-      "loss": 3.6456,
+      "epoch": 1.8975741239892183,
+      "grad_norm": 0.5452243089675903,
+      "learning_rate": 0.00048670912034538583,
+      "loss": 3.6368,
       "step": 17600
     },
     {
-      "epoch": 1.8996878699817028,
-      "grad_norm": 0.5752872824668884,
-      "learning_rate": 0.0004865811873720504,
-      "loss": 3.6371,
+      "epoch": 1.9029649595687332,
+      "grad_norm": 0.5425637364387512,
+      "learning_rate": 0.0004863853211009174,
+      "loss": 3.6341,
       "step": 17650
     },
     {
-      "epoch": 1.905069422021311,
-      "grad_norm": 0.5607645511627197,
-      "learning_rate": 0.00048625794634198896,
-      "loss": 3.6451,
+      "epoch": 1.908355795148248,
+      "grad_norm": 0.5863619446754456,
+      "learning_rate": 0.000486061521856449,
+      "loss": 3.6463,
       "step": 17700
     },
     {
-      "epoch": 1.910450974060919,
-      "grad_norm": 0.5504919290542603,
-      "learning_rate": 0.00048593470531192756,
-      "loss": 3.65,
+      "epoch": 1.9137466307277629,
+      "grad_norm": 0.5500699281692505,
+      "learning_rate": 0.00048573772261198054,
+      "loss": 3.6371,
       "step": 17750
     },
     {
-      "epoch": 1.9158325261005273,
-      "grad_norm": 0.5980883240699768,
-      "learning_rate": 0.00048561146428186615,
-      "loss": 3.633,
+      "epoch": 1.9191374663072778,
+      "grad_norm": 0.5982254147529602,
+      "learning_rate": 0.00048541392336751214,
+      "loss": 3.6519,
       "step": 17800
     },
     {
-      "epoch": 1.9212140781401357,
-      "grad_norm": 0.5818303823471069,
-      "learning_rate": 0.0004852882232518047,
-      "loss": 3.6397,
+      "epoch": 1.9245283018867925,
+      "grad_norm": 0.5623408555984497,
+      "learning_rate": 0.0004850901241230437,
+      "loss": 3.6572,
       "step": 17850
     },
     {
-      "epoch": 1.9265956301797438,
-      "grad_norm": 0.5449066162109375,
-      "learning_rate": 0.00048496498222174334,
-      "loss": 3.6361,
+      "epoch": 1.9299191374663072,
+      "grad_norm": 0.5582827925682068,
+      "learning_rate": 0.0004847663248785753,
+      "loss": 3.6323,
       "step": 17900
     },
     {
-      "epoch": 1.931977182219352,
-      "grad_norm": 0.5642189383506775,
-      "learning_rate": 0.00048464174119168193,
-      "loss": 3.6364,
+      "epoch": 1.935309973045822,
+      "grad_norm": 0.5745336413383484,
+      "learning_rate": 0.0004844425256341068,
+      "loss": 3.6433,
       "step": 17950
     },
     {
-      "epoch": 1.9373587342589604,
-      "grad_norm": 0.5773710608482361,
-      "learning_rate": 0.0004843185001616205,
-      "loss": 3.6504,
+      "epoch": 1.940700808625337,
+      "grad_norm": 0.5856007933616638,
+      "learning_rate": 0.00048411872638963834,
+      "loss": 3.6391,
       "step": 18000
     },
     {
-      "epoch": 1.9373587342589604,
-      "eval_accuracy": 0.3603285054871338,
-      "eval_loss": 3.591597080230713,
-      "eval_runtime": 183.9007,
-      "eval_samples_per_second": 97.939,
-      "eval_steps_per_second": 6.123,
+      "epoch": 1.940700808625337,
+      "eval_accuracy": 0.3603370890652712,
+      "eval_loss": 3.588714122772217,
+      "eval_runtime": 183.5899,
+      "eval_samples_per_second": 98.105,
+      "eval_steps_per_second": 6.133,
       "step": 18000
     },
     {
-      "epoch": 1.9427402862985685,
-      "grad_norm": 0.6188564300537109,
-      "learning_rate": 0.0004840017239521603,
-      "loss": 3.6402,
+      "epoch": 1.9460916442048517,
+      "grad_norm": 0.5621957182884216,
+      "learning_rate": 0.00048379492714516995,
+      "loss": 3.6223,
       "step": 18050
     },
     {
-      "epoch": 1.9481218383381766,
-      "grad_norm": 0.5938689112663269,
-      "learning_rate": 0.0004836784829220989,
-      "loss": 3.6292,
+      "epoch": 1.9514824797843666,
+      "grad_norm": 0.5352705717086792,
+      "learning_rate": 0.0004834711279007015,
+      "loss": 3.6519,
       "step": 18100
     },
     {
-      "epoch": 1.953503390377785,
-      "grad_norm": 0.5374168753623962,
-      "learning_rate": 0.00048335524189203747,
-      "loss": 3.6388,
+      "epoch": 1.9568733153638815,
+      "grad_norm": 0.5720267295837402,
+      "learning_rate": 0.0004831473286562331,
+      "loss": 3.653,
       "step": 18150
     },
     {
-      "epoch": 1.9588849424173933,
-      "grad_norm": 0.6024685502052307,
-      "learning_rate": 0.000483032000861976,
-      "loss": 3.6345,
+      "epoch": 1.9622641509433962,
+      "grad_norm": 0.6092367172241211,
+      "learning_rate": 0.00048282352941176465,
+      "loss": 3.6395,
       "step": 18200
     },
     {
-      "epoch": 1.9642664944570014,
-      "grad_norm": 0.5500667095184326,
-      "learning_rate": 0.0004827152246525158,
-      "loss": 3.6604,
+      "epoch": 1.967654986522911,
+      "grad_norm": 0.6039474606513977,
+      "learning_rate": 0.00048249973016729626,
+      "loss": 3.6432,
       "step": 18250
     },
     {
-      "epoch": 1.9696480464966095,
-      "grad_norm": 0.5466398000717163,
-      "learning_rate": 0.0004823919836224544,
-      "loss": 3.6515,
+      "epoch": 1.9730458221024259,
+      "grad_norm": 0.5490421652793884,
+      "learning_rate": 0.0004821759309228278,
+      "loss": 3.6217,
       "step": 18300
     },
     {
-      "epoch": 1.9750295985362178,
-      "grad_norm": 0.5468305349349976,
-      "learning_rate": 0.000482068742592393,
-      "loss": 3.6272,
+      "epoch": 1.9784366576819408,
+      "grad_norm": 0.6812778115272522,
+      "learning_rate": 0.00048185213167835936,
+      "loss": 3.6456,
       "step": 18350
     },
     {
-      "epoch": 1.9804111505758262,
-      "grad_norm": 0.5989289283752441,
-      "learning_rate": 0.0004817455015623316,
-      "loss": 3.6305,
+      "epoch": 1.9838274932614555,
+      "grad_norm": 0.5026232600212097,
+      "learning_rate": 0.00048152833243389096,
+      "loss": 3.6235,
       "step": 18400
     },
     {
-      "epoch": 1.9857927026154343,
-      "grad_norm": 0.5450242161750793,
-      "learning_rate": 0.0004814222605322702,
-      "loss": 3.6303,
+      "epoch": 1.9892183288409704,
+      "grad_norm": 0.5733034014701843,
+      "learning_rate": 0.0004812045331894225,
+      "loss": 3.6258,
       "step": 18450
     },
     {
-      "epoch": 1.9911742546550424,
-      "grad_norm": 0.5922088623046875,
-      "learning_rate": 0.0004810990195022088,
-      "loss": 3.6453,
+      "epoch": 1.9946091644204853,
+      "grad_norm": 0.5714686512947083,
+      "learning_rate": 0.0004808807339449541,
+      "loss": 3.6177,
       "step": 18500
     },
     {
-      "epoch": 1.9965558066946507,
-      "grad_norm": 0.5734391212463379,
-      "learning_rate": 0.00048077577847214733,
-      "loss": 3.6382,
+      "epoch": 2.0,
+      "grad_norm": 1.0521697998046875,
+      "learning_rate": 0.0004805569347004856,
+      "loss": 3.633,
       "step": 18550
     },
     {
-      "epoch": 2.001937358734259,
-      "grad_norm": 0.5578837990760803,
-      "learning_rate": 0.0004804525374420859,
-      "loss": 3.605,
+      "epoch": 2.0053908355795147,
+      "grad_norm": 0.629177987575531,
+      "learning_rate": 0.0004802331354560173,
+      "loss": 3.5478,
       "step": 18600
     },
     {
-      "epoch": 2.007318910773867,
-      "grad_norm": 0.6076123118400574,
-      "learning_rate": 0.0004801292964120246,
-      "loss": 3.5479,
+      "epoch": 2.01078167115903,
+      "grad_norm": 0.5587250590324402,
+      "learning_rate": 0.00047990933621154877,
+      "loss": 3.5391,
       "step": 18650
     },
     {
-      "epoch": 2.0127004628134753,
-      "grad_norm": 0.5363115668296814,
-      "learning_rate": 0.0004798060553819631,
-      "loss": 3.524,
+      "epoch": 2.0161725067385445,
+      "grad_norm": 0.5593814253807068,
+      "learning_rate": 0.0004795855369670804,
+      "loss": 3.5466,
       "step": 18700
     },
     {
-      "epoch": 2.018082014853084,
-      "grad_norm": 0.6044074296951294,
-      "learning_rate": 0.0004794828143519017,
-      "loss": 3.5659,
+      "epoch": 2.0215633423180592,
+      "grad_norm": 0.6031098961830139,
+      "learning_rate": 0.0004792617377226119,
+      "loss": 3.5394,
       "step": 18750
     },
     {
-      "epoch": 2.023463566892692,
-      "grad_norm": 0.5858488082885742,
-      "learning_rate": 0.00047915957332184025,
-      "loss": 3.5431,
+      "epoch": 2.026954177897574,
+      "grad_norm": 0.6028936505317688,
+      "learning_rate": 0.0004789379384781435,
+      "loss": 3.5464,
       "step": 18800
     },
     {
-      "epoch": 2.0288451189323,
-      "grad_norm": 0.5654283761978149,
-      "learning_rate": 0.00047883633229177884,
-      "loss": 3.541,
+      "epoch": 2.032345013477089,
+      "grad_norm": 0.530657947063446,
+      "learning_rate": 0.0004786141392336751,
+      "loss": 3.5433,
       "step": 18850
     },
     {
-      "epoch": 2.034226670971908,
-      "grad_norm": 0.5618985891342163,
-      "learning_rate": 0.00047851309126171744,
-      "loss": 3.5387,
+      "epoch": 2.0377358490566038,
+      "grad_norm": 0.5888926386833191,
+      "learning_rate": 0.00047829033998920663,
+      "loss": 3.5579,
       "step": 18900
     },
     {
-      "epoch": 2.0396082230115167,
-      "grad_norm": 0.582912802696228,
-      "learning_rate": 0.00047818985023165603,
-      "loss": 3.5436,
+      "epoch": 2.0431266846361185,
+      "grad_norm": 0.576383650302887,
+      "learning_rate": 0.00047796654074473824,
+      "loss": 3.5521,
       "step": 18950
     },
     {
-      "epoch": 2.044989775051125,
-      "grad_norm": 0.5865135192871094,
-      "learning_rate": 0.00047786660920159463,
-      "loss": 3.5301,
+      "epoch": 2.0485175202156336,
+      "grad_norm": 0.5961529612541199,
+      "learning_rate": 0.0004776427415002698,
+      "loss": 3.5531,
       "step": 19000
     },
     {
-      "epoch": 2.044989775051125,
-      "eval_accuracy": 0.3611175427584426,
-      "eval_loss": 3.582451581954956,
-      "eval_runtime": 183.4154,
-      "eval_samples_per_second": 98.198,
-      "eval_steps_per_second": 6.139,
+      "epoch": 2.0485175202156336,
+      "eval_accuracy": 0.36173795074786325,
+      "eval_loss": 3.578604221343994,
+      "eval_runtime": 183.878,
+      "eval_samples_per_second": 97.951,
+      "eval_steps_per_second": 6.124,
       "step": 19000
     },
     {
-      "epoch": 2.050371327090733,
-      "grad_norm": 0.6071602702140808,
-      "learning_rate": 0.0004775433681715332,
-      "loss": 3.5575,
+      "epoch": 2.0539083557951483,
+      "grad_norm": 0.529506504535675,
+      "learning_rate": 0.0004773189422558014,
+      "loss": 3.5622,
       "step": 19050
     },
     {
-      "epoch": 2.055752879130341,
-      "grad_norm": 0.580239474773407,
-      "learning_rate": 0.00047722012714147176,
-      "loss": 3.5628,
+      "epoch": 2.059299191374663,
+      "grad_norm": 0.569199800491333,
+      "learning_rate": 0.00047699514301133294,
+      "loss": 3.554,
       "step": 19100
     },
     {
-      "epoch": 2.0611344311699495,
-      "grad_norm": 0.6349200010299683,
-      "learning_rate": 0.00047689688611141036,
-      "loss": 3.5564,
+      "epoch": 2.0646900269541777,
+      "grad_norm": 0.5616893768310547,
+      "learning_rate": 0.00047667134376686455,
+      "loss": 3.5492,
       "step": 19150
     },
     {
-      "epoch": 2.0665159832095576,
-      "grad_norm": 0.6028494238853455,
-      "learning_rate": 0.0004765736450813489,
-      "loss": 3.5471,
+      "epoch": 2.070080862533693,
+      "grad_norm": 0.5315808653831482,
+      "learning_rate": 0.0004763475445223961,
+      "loss": 3.5705,
       "step": 19200
     },
     {
-      "epoch": 2.0718975352491658,
-      "grad_norm": 0.5523431897163391,
-      "learning_rate": 0.00047625040405128755,
-      "loss": 3.5617,
+      "epoch": 2.0754716981132075,
+      "grad_norm": 0.5792466998100281,
+      "learning_rate": 0.00047603022126281705,
+      "loss": 3.5619,
       "step": 19250
     },
     {
-      "epoch": 2.0772790872887743,
-      "grad_norm": 0.6101455092430115,
-      "learning_rate": 0.00047592716302122614,
-      "loss": 3.5496,
+      "epoch": 2.0808625336927222,
+      "grad_norm": 0.5894599556922913,
+      "learning_rate": 0.00047570642201834855,
+      "loss": 3.5514,
       "step": 19300
     },
     {
-      "epoch": 2.0826606393283824,
-      "grad_norm": 0.6239585280418396,
-      "learning_rate": 0.0004756039219911647,
-      "loss": 3.5715,
+      "epoch": 2.0862533692722374,
+      "grad_norm": 0.5664039254188538,
+      "learning_rate": 0.00047538262277388015,
+      "loss": 3.5679,
       "step": 19350
     },
     {
-      "epoch": 2.0880421913679905,
-      "grad_norm": 0.5511799454689026,
-      "learning_rate": 0.0004752806809611033,
-      "loss": 3.5706,
+      "epoch": 2.091644204851752,
+      "grad_norm": 0.5353388786315918,
+      "learning_rate": 0.0004750588235294117,
+      "loss": 3.558,
       "step": 19400
     },
     {
-      "epoch": 2.0934237434075986,
-      "grad_norm": 0.6214755773544312,
-      "learning_rate": 0.00047495743993104187,
-      "loss": 3.5592,
+      "epoch": 2.0970350404312668,
+      "grad_norm": 0.6546982526779175,
+      "learning_rate": 0.0004747350242849433,
+      "loss": 3.5405,
       "step": 19450
     },
     {
-      "epoch": 2.098805295447207,
-      "grad_norm": 0.5760681629180908,
-      "learning_rate": 0.00047463419890098047,
-      "loss": 3.5538,
+      "epoch": 2.1024258760107815,
+      "grad_norm": 0.5824753046035767,
+      "learning_rate": 0.00047441122504047486,
+      "loss": 3.565,
       "step": 19500
     },
     {
-      "epoch": 2.1041868474868153,
-      "grad_norm": 0.6076908707618713,
-      "learning_rate": 0.00047431095787091906,
-      "loss": 3.5466,
+      "epoch": 2.1078167115902966,
+      "grad_norm": 0.5634300708770752,
+      "learning_rate": 0.0004740874257960064,
+      "loss": 3.5563,
       "step": 19550
     },
     {
-      "epoch": 2.1095683995264234,
-      "grad_norm": 0.5608106255531311,
-      "learning_rate": 0.00047398771684085766,
-      "loss": 3.5495,
+      "epoch": 2.1132075471698113,
+      "grad_norm": 0.6132933497428894,
+      "learning_rate": 0.000473763626551538,
+      "loss": 3.5536,
       "step": 19600
     },
     {
-      "epoch": 2.1149499515660315,
-      "grad_norm": 0.5763227939605713,
-      "learning_rate": 0.0004736644758107962,
-      "loss": 3.5705,
+      "epoch": 2.118598382749326,
+      "grad_norm": 0.5740868449211121,
+      "learning_rate": 0.00047343982730706956,
+      "loss": 3.5564,
       "step": 19650
     },
     {
-      "epoch": 2.12033150360564,
-      "grad_norm": 0.566933810710907,
-      "learning_rate": 0.0004733412347807348,
-      "loss": 3.553,
+      "epoch": 2.123989218328841,
+      "grad_norm": 0.5281256437301636,
+      "learning_rate": 0.00047311602806260117,
+      "loss": 3.5488,
       "step": 19700
     },
     {
-      "epoch": 2.125713055645248,
-      "grad_norm": 0.5792555212974548,
-      "learning_rate": 0.00047301799375067333,
-      "loss": 3.5502,
+      "epoch": 2.129380053908356,
+      "grad_norm": 0.6066663861274719,
+      "learning_rate": 0.0004727922288181327,
+      "loss": 3.5384,
       "step": 19750
     },
     {
-      "epoch": 2.1310946076848563,
-      "grad_norm": 0.6911931037902832,
-      "learning_rate": 0.000472694752720612,
-      "loss": 3.5562,
+      "epoch": 2.1347708894878705,
+      "grad_norm": 0.5933831334114075,
+      "learning_rate": 0.0004724684295736643,
+      "loss": 3.5605,
       "step": 19800
     },
     {
-      "epoch": 2.1364761597244644,
-      "grad_norm": 0.5647987723350525,
-      "learning_rate": 0.0004723715116905506,
-      "loss": 3.5694,
+      "epoch": 2.1401617250673857,
+      "grad_norm": 0.5494194030761719,
+      "learning_rate": 0.0004721446303291959,
+      "loss": 3.5522,
       "step": 19850
     },
     {
-      "epoch": 2.141857711764073,
-      "grad_norm": 0.6511532664299011,
-      "learning_rate": 0.0004720482706604891,
-      "loss": 3.5709,
+      "epoch": 2.1455525606469004,
+      "grad_norm": 0.5679726600646973,
+      "learning_rate": 0.0004718208310847275,
+      "loss": 3.5555,
       "step": 19900
     },
     {
-      "epoch": 2.147239263803681,
-      "grad_norm": 0.6013280749320984,
-      "learning_rate": 0.0004717250296304277,
-      "loss": 3.5529,
+      "epoch": 2.150943396226415,
+      "grad_norm": 0.5585212707519531,
+      "learning_rate": 0.000471497031840259,
+      "loss": 3.5489,
       "step": 19950
     },
     {
-      "epoch": 2.152620815843289,
-      "grad_norm": 0.5992707014083862,
-      "learning_rate": 0.0004714017886003663,
-      "loss": 3.5617,
+      "epoch": 2.1563342318059298,
+      "grad_norm": 0.615452229976654,
+      "learning_rate": 0.0004711732325957905,
+      "loss": 3.551,
       "step": 20000
     },
     {
-      "epoch": 2.152620815843289,
-      "eval_accuracy": 0.3623941055373749,
-      "eval_loss": 3.572652578353882,
-      "eval_runtime": 184.1144,
-      "eval_samples_per_second": 97.825,
-      "eval_steps_per_second": 6.116,
+      "epoch": 2.1563342318059298,
+      "eval_accuracy": 0.36307796681327925,
+      "eval_loss": 3.5685019493103027,
+      "eval_runtime": 183.5714,
+      "eval_samples_per_second": 98.114,
+      "eval_steps_per_second": 6.134,
       "step": 20000
     },
     {
-      "epoch": 2.1580023678828972,
-      "grad_norm": 0.6025944948196411,
-      "learning_rate": 0.00047107854757030485,
-      "loss": 3.559,
+      "epoch": 2.161725067385445,
+      "grad_norm": 0.5647128224372864,
+      "learning_rate": 0.00047084943335132213,
+      "loss": 3.5557,
       "step": 20050
     },
     {
-      "epoch": 2.163383919922506,
-      "grad_norm": 0.6153149008750916,
-      "learning_rate": 0.0004707553065402435,
-      "loss": 3.5597,
+      "epoch": 2.1671159029649596,
+      "grad_norm": 0.5806307196617126,
+      "learning_rate": 0.0004705256341068537,
+      "loss": 3.5516,
       "step": 20100
     },
     {
-      "epoch": 2.168765471962114,
-      "grad_norm": 0.5391000509262085,
-      "learning_rate": 0.0004704320655101821,
-      "loss": 3.5528,
+      "epoch": 2.1725067385444743,
+      "grad_norm": 0.5652997493743896,
+      "learning_rate": 0.0004702018348623853,
+      "loss": 3.562,
       "step": 20150
     },
     {
-      "epoch": 2.174147024001722,
-      "grad_norm": 0.564768373966217,
-      "learning_rate": 0.00047010882448012063,
-      "loss": 3.5711,
+      "epoch": 2.177897574123989,
+      "grad_norm": 0.5413397550582886,
+      "learning_rate": 0.00046987803561791684,
+      "loss": 3.5631,
       "step": 20200
     },
     {
-      "epoch": 2.1795285760413305,
-      "grad_norm": 0.5776879191398621,
-      "learning_rate": 0.0004697855834500592,
-      "loss": 3.5586,
+      "epoch": 2.183288409703504,
+      "grad_norm": 0.5363370776176453,
+      "learning_rate": 0.00046955423637344844,
+      "loss": 3.5669,
       "step": 20250
     },
     {
-      "epoch": 2.1849101280809387,
-      "grad_norm": 0.5807245969772339,
-      "learning_rate": 0.00046946234241999776,
-      "loss": 3.5491,
+      "epoch": 2.188679245283019,
+      "grad_norm": 0.6292061805725098,
+      "learning_rate": 0.00046923043712898,
+      "loss": 3.548,
       "step": 20300
     },
     {
-      "epoch": 2.1902916801205468,
-      "grad_norm": 0.548610508441925,
-      "learning_rate": 0.00046913910138993636,
-      "loss": 3.5428,
+      "epoch": 2.1940700808625335,
+      "grad_norm": 0.5715665221214294,
+      "learning_rate": 0.0004689066378845116,
+      "loss": 3.5434,
       "step": 20350
     },
     {
-      "epoch": 2.195673232160155,
-      "grad_norm": 0.5948867797851562,
-      "learning_rate": 0.000468815860359875,
-      "loss": 3.5504,
+      "epoch": 2.1994609164420487,
+      "grad_norm": 0.58626389503479,
+      "learning_rate": 0.00046858283864004315,
+      "loss": 3.5582,
       "step": 20400
     },
     {
-      "epoch": 2.2010547841997634,
-      "grad_norm": 0.570107638835907,
-      "learning_rate": 0.00046849261932981355,
-      "loss": 3.5584,
+      "epoch": 2.2048517520215634,
+      "grad_norm": 0.8618109226226807,
+      "learning_rate": 0.0004682590393955747,
+      "loss": 3.5514,
       "step": 20450
     },
     {
-      "epoch": 2.2064363362393715,
-      "grad_norm": 0.6191420555114746,
-      "learning_rate": 0.00046816937829975214,
-      "loss": 3.5501,
+      "epoch": 2.210242587601078,
+      "grad_norm": 0.5103957056999207,
+      "learning_rate": 0.0004679352401511063,
+      "loss": 3.568,
       "step": 20500
     },
     {
-      "epoch": 2.2118178882789796,
-      "grad_norm": 0.5583657026290894,
-      "learning_rate": 0.00046784613726969074,
-      "loss": 3.5461,
+      "epoch": 2.215633423180593,
+      "grad_norm": 0.5664064884185791,
+      "learning_rate": 0.0004676114409066378,
+      "loss": 3.5626,
       "step": 20550
     },
     {
-      "epoch": 2.2171994403185877,
-      "grad_norm": 0.6070858836174011,
-      "learning_rate": 0.0004675228962396293,
-      "loss": 3.5564,
+      "epoch": 2.221024258760108,
+      "grad_norm": 0.6470500826835632,
+      "learning_rate": 0.00046728764166216946,
+      "loss": 3.5591,
       "step": 20600
     },
     {
-      "epoch": 2.2225809923581963,
-      "grad_norm": 0.5971397161483765,
-      "learning_rate": 0.0004671996552095679,
-      "loss": 3.5653,
+      "epoch": 2.2264150943396226,
+      "grad_norm": 0.6026623845100403,
+      "learning_rate": 0.00046696384241770095,
+      "loss": 3.5675,
       "step": 20650
     },
     {
-      "epoch": 2.2279625443978044,
-      "grad_norm": 0.5894590020179749,
-      "learning_rate": 0.0004668764141795065,
-      "loss": 3.5568,
+      "epoch": 2.2318059299191373,
+      "grad_norm": 0.6435901522636414,
+      "learning_rate": 0.00046664004317323256,
+      "loss": 3.5509,
       "step": 20700
     },
     {
-      "epoch": 2.2333440964374125,
-      "grad_norm": 0.5683594346046448,
-      "learning_rate": 0.00046655317314944506,
-      "loss": 3.5531,
+      "epoch": 2.2371967654986524,
+      "grad_norm": 0.5778759121894836,
+      "learning_rate": 0.0004663162439287641,
+      "loss": 3.5691,
       "step": 20750
     },
     {
-      "epoch": 2.2387256484770206,
-      "grad_norm": 0.5684804320335388,
-      "learning_rate": 0.00046622993211938366,
-      "loss": 3.5613,
+      "epoch": 2.242587601078167,
+      "grad_norm": 0.585109531879425,
+      "learning_rate": 0.00046599244468429566,
+      "loss": 3.5645,
       "step": 20800
     },
     {
-      "epoch": 2.244107200516629,
-      "grad_norm": 0.5961329340934753,
-      "learning_rate": 0.0004659066910893222,
-      "loss": 3.5757,
+      "epoch": 2.247978436657682,
+      "grad_norm": 0.5531785488128662,
+      "learning_rate": 0.00046566864543982726,
+      "loss": 3.5658,
       "step": 20850
     },
     {
-      "epoch": 2.2494887525562373,
-      "grad_norm": 0.583321750164032,
-      "learning_rate": 0.0004655834500592608,
-      "loss": 3.551,
+      "epoch": 2.2533692722371965,
+      "grad_norm": 0.5610100030899048,
+      "learning_rate": 0.0004653448461953588,
+      "loss": 3.5545,
       "step": 20900
     },
     {
-      "epoch": 2.2548703045958454,
-      "grad_norm": 0.607089638710022,
-      "learning_rate": 0.00046526020902919944,
-      "loss": 3.5526,
+      "epoch": 2.2587601078167117,
+      "grad_norm": 0.5870633125305176,
+      "learning_rate": 0.0004650210469508904,
+      "loss": 3.5535,
       "step": 20950
     },
     {
-      "epoch": 2.2602518566354535,
-      "grad_norm": 0.5979416966438293,
-      "learning_rate": 0.000464936967999138,
-      "loss": 3.5523,
+      "epoch": 2.2641509433962264,
+      "grad_norm": 0.5517866611480713,
+      "learning_rate": 0.00046469724770642197,
+      "loss": 3.5532,
       "step": 21000
     },
     {
-      "epoch": 2.2602518566354535,
-      "eval_accuracy": 0.3637333610325762,
-      "eval_loss": 3.560683488845825,
-      "eval_runtime": 183.5693,
-      "eval_samples_per_second": 98.116,
-      "eval_steps_per_second": 6.134,
+      "epoch": 2.2641509433962264,
+      "eval_accuracy": 0.36412092588341044,
+      "eval_loss": 3.5580224990844727,
+      "eval_runtime": 183.8151,
+      "eval_samples_per_second": 97.984,
+      "eval_steps_per_second": 6.126,
       "step": 21000
     },
     {
-      "epoch": 2.265633408675062,
-      "grad_norm": 0.5803635716438293,
-      "learning_rate": 0.0004646137269690766,
-      "loss": 3.5575,
+      "epoch": 2.269541778975741,
+      "grad_norm": 0.6098730564117432,
+      "learning_rate": 0.0004643734484619536,
+      "loss": 3.5498,
       "step": 21050
     },
     {
-      "epoch": 2.27101496071467,
-      "grad_norm": 0.5625525712966919,
-      "learning_rate": 0.00046429048593901517,
-      "loss": 3.5688,
+      "epoch": 2.274932614555256,
+      "grad_norm": 0.6386333703994751,
+      "learning_rate": 0.0004640496492174851,
+      "loss": 3.5404,
       "step": 21100
     },
     {
-      "epoch": 2.2763965127542782,
-      "grad_norm": 0.6204305291175842,
-      "learning_rate": 0.0004639672449089537,
-      "loss": 3.561,
+      "epoch": 2.280323450134771,
+      "grad_norm": 0.5949533581733704,
+      "learning_rate": 0.00046372584997301673,
+      "loss": 3.5695,
       "step": 21150
     },
     {
-      "epoch": 2.281778064793887,
-      "grad_norm": 0.6559270620346069,
-      "learning_rate": 0.0004636440038788923,
-      "loss": 3.564,
+      "epoch": 2.2857142857142856,
+      "grad_norm": 0.5580140948295593,
+      "learning_rate": 0.0004634020507285483,
+      "loss": 3.5476,
       "step": 21200
     },
     {
-      "epoch": 2.287159616833495,
-      "grad_norm": 0.6153460144996643,
-      "learning_rate": 0.00046332076284883095,
-      "loss": 3.549,
+      "epoch": 2.2911051212938007,
+      "grad_norm": 0.566337525844574,
+      "learning_rate": 0.00046308472746896923,
+      "loss": 3.5459,
       "step": 21250
     },
     {
-      "epoch": 2.292541168873103,
-      "grad_norm": 0.602618396282196,
-      "learning_rate": 0.0004629975218187695,
-      "loss": 3.5636,
+      "epoch": 2.2964959568733154,
+      "grad_norm": 0.5436205267906189,
+      "learning_rate": 0.00046276092822450073,
+      "loss": 3.5552,
       "step": 21300
     },
     {
-      "epoch": 2.297922720912711,
-      "grad_norm": 0.5812705755233765,
-      "learning_rate": 0.0004626742807887081,
-      "loss": 3.5533,
+      "epoch": 2.30188679245283,
+      "grad_norm": 0.5451183915138245,
+      "learning_rate": 0.00046243712898003233,
+      "loss": 3.5339,
       "step": 21350
     },
     {
-      "epoch": 2.303304272952319,
-      "grad_norm": 0.5544642806053162,
-      "learning_rate": 0.00046235103975864663,
-      "loss": 3.543,
+      "epoch": 2.3072776280323453,
+      "grad_norm": 0.5662717819213867,
+      "learning_rate": 0.0004621133297355639,
+      "loss": 3.5704,
       "step": 21400
     },
     {
-      "epoch": 2.3086858249919278,
-      "grad_norm": 0.5726743936538696,
-      "learning_rate": 0.0004620277987285852,
-      "loss": 3.563,
+      "epoch": 2.31266846361186,
+      "grad_norm": 0.5711853504180908,
+      "learning_rate": 0.0004617895304910955,
+      "loss": 3.5297,
       "step": 21450
     },
     {
-      "epoch": 2.314067377031536,
-      "grad_norm": 0.5752015113830566,
-      "learning_rate": 0.0004617045576985239,
-      "loss": 3.5556,
+      "epoch": 2.3180592991913747,
+      "grad_norm": 0.5936728715896606,
+      "learning_rate": 0.00046146573124662704,
+      "loss": 3.5607,
       "step": 21500
     },
     {
-      "epoch": 2.319448929071144,
-      "grad_norm": 0.5700117349624634,
-      "learning_rate": 0.0004613813166684624,
-      "loss": 3.5522,
+      "epoch": 2.3234501347708894,
+      "grad_norm": 0.5846163034439087,
+      "learning_rate": 0.00046114193200215864,
+      "loss": 3.5591,
       "step": 21550
     },
     {
-      "epoch": 2.3248304811107525,
-      "grad_norm": 0.5672240853309631,
-      "learning_rate": 0.000461058075638401,
-      "loss": 3.5553,
+      "epoch": 2.3288409703504045,
+      "grad_norm": 0.5686417818069458,
+      "learning_rate": 0.0004608181327576902,
+      "loss": 3.5332,
       "step": 21600
     },
     {
-      "epoch": 2.3302120331503606,
-      "grad_norm": 0.6115524768829346,
-      "learning_rate": 0.00046073483460833955,
-      "loss": 3.5433,
+      "epoch": 2.334231805929919,
+      "grad_norm": 0.6202867031097412,
+      "learning_rate": 0.00046049433351322175,
+      "loss": 3.5547,
       "step": 21650
     },
     {
-      "epoch": 2.3355935851899687,
-      "grad_norm": 0.5477917790412903,
-      "learning_rate": 0.00046041159357827814,
-      "loss": 3.5524,
+      "epoch": 2.339622641509434,
+      "grad_norm": 0.5648440718650818,
+      "learning_rate": 0.00046017053426875335,
+      "loss": 3.5678,
       "step": 21700
     },
     {
-      "epoch": 2.340975137229577,
-      "grad_norm": 0.5689560770988464,
-      "learning_rate": 0.00046008835254821674,
-      "loss": 3.5692,
+      "epoch": 2.3450134770889486,
+      "grad_norm": 0.5897420048713684,
+      "learning_rate": 0.0004598467350242849,
+      "loss": 3.5483,
       "step": 21750
     },
     {
-      "epoch": 2.3463566892691854,
-      "grad_norm": 0.5863248109817505,
-      "learning_rate": 0.0004597651115181554,
-      "loss": 3.5527,
+      "epoch": 2.3504043126684637,
+      "grad_norm": 0.5510711073875427,
+      "learning_rate": 0.0004595229357798165,
+      "loss": 3.56,
       "step": 21800
     },
     {
-      "epoch": 2.3517382413087935,
-      "grad_norm": 0.5983529090881348,
-      "learning_rate": 0.00045944187048809393,
-      "loss": 3.5373,
+      "epoch": 2.3557951482479784,
+      "grad_norm": 0.6083936095237732,
+      "learning_rate": 0.00045919913653534806,
+      "loss": 3.5596,
       "step": 21850
     },
     {
-      "epoch": 2.3571197933484016,
-      "grad_norm": 0.5777842402458191,
-      "learning_rate": 0.0004591186294580325,
-      "loss": 3.5481,
+      "epoch": 2.361185983827493,
+      "grad_norm": 0.5641026496887207,
+      "learning_rate": 0.00045887533729087966,
+      "loss": 3.5529,
       "step": 21900
     },
     {
-      "epoch": 2.3625013453880097,
-      "grad_norm": 0.6365346312522888,
-      "learning_rate": 0.00045879538842797106,
-      "loss": 3.5666,
+      "epoch": 2.3665768194070083,
+      "grad_norm": 0.5562753677368164,
+      "learning_rate": 0.00045855153804641116,
+      "loss": 3.5434,
       "step": 21950
     },
     {
-      "epoch": 2.3678828974276183,
-      "grad_norm": 0.598414957523346,
-      "learning_rate": 0.00045847214739790966,
-      "loss": 3.554,
+      "epoch": 2.371967654986523,
+      "grad_norm": 0.5417619943618774,
+      "learning_rate": 0.0004582277388019427,
+      "loss": 3.5392,
       "step": 22000
     },
     {
-      "epoch": 2.3678828974276183,
-      "eval_accuracy": 0.36512639970724564,
-      "eval_loss": 3.5473287105560303,
-      "eval_runtime": 183.9831,
-      "eval_samples_per_second": 97.895,
-      "eval_steps_per_second": 6.12,
+      "epoch": 2.371967654986523,
+      "eval_accuracy": 0.3649375609882241,
+      "eval_loss": 3.547846555709839,
+      "eval_runtime": 183.6784,
+      "eval_samples_per_second": 98.057,
+      "eval_steps_per_second": 6.13,
       "step": 22000
     },
     {
-      "epoch": 2.3732644494672264,
-      "grad_norm": 0.5717033743858337,
-      "learning_rate": 0.0004581489063678482,
-      "loss": 3.5563,
+      "epoch": 2.3773584905660377,
+      "grad_norm": 0.5873769521713257,
+      "learning_rate": 0.0004579039395574743,
+      "loss": 3.5399,
       "step": 22050
     },
     {
-      "epoch": 2.3786460015068345,
-      "grad_norm": 0.6341887712478638,
-      "learning_rate": 0.00045782566533778685,
-      "loss": 3.5697,
+      "epoch": 2.382749326145553,
+      "grad_norm": 0.5401739478111267,
+      "learning_rate": 0.00045758014031300586,
+      "loss": 3.5597,
       "step": 22100
     },
     {
-      "epoch": 2.384027553546443,
-      "grad_norm": 0.5593985319137573,
-      "learning_rate": 0.00045750242430772544,
-      "loss": 3.5573,
+      "epoch": 2.3881401617250675,
+      "grad_norm": 0.5652412176132202,
+      "learning_rate": 0.00045725634106853747,
+      "loss": 3.5511,
       "step": 22150
     },
     {
-      "epoch": 2.389409105586051,
-      "grad_norm": 0.6116588115692139,
-      "learning_rate": 0.000457179183277664,
-      "loss": 3.5413,
+      "epoch": 2.393530997304582,
+      "grad_norm": 0.6286122798919678,
+      "learning_rate": 0.000456932541824069,
+      "loss": 3.5383,
       "step": 22200
     },
     {
-      "epoch": 2.3947906576256592,
-      "grad_norm": 0.5428621172904968,
-      "learning_rate": 0.00045686240706820384,
-      "loss": 3.551,
+      "epoch": 2.398921832884097,
+      "grad_norm": 0.8063929080963135,
+      "learning_rate": 0.0004566087425796006,
+      "loss": 3.5307,
       "step": 22250
     },
     {
-      "epoch": 2.4001722096652673,
-      "grad_norm": 0.611487090587616,
-      "learning_rate": 0.0004565391660381424,
-      "loss": 3.5515,
+      "epoch": 2.404312668463612,
+      "grad_norm": 0.6388524174690247,
+      "learning_rate": 0.0004562849433351322,
+      "loss": 3.5602,
       "step": 22300
     },
     {
-      "epoch": 2.4055537617048754,
-      "grad_norm": 0.5635246634483337,
-      "learning_rate": 0.000456215925008081,
-      "loss": 3.5428,
+      "epoch": 2.4097035040431267,
+      "grad_norm": 0.6252179145812988,
+      "learning_rate": 0.0004559611440906638,
+      "loss": 3.5697,
       "step": 22350
     },
     {
-      "epoch": 2.410935313744484,
-      "grad_norm": 0.5477227568626404,
-      "learning_rate": 0.00045589268397801957,
-      "loss": 3.5668,
+      "epoch": 2.4150943396226414,
+      "grad_norm": 0.6091659069061279,
+      "learning_rate": 0.00045563734484619533,
+      "loss": 3.5544,
       "step": 22400
     },
     {
-      "epoch": 2.416316865784092,
-      "grad_norm": 0.6393861174583435,
-      "learning_rate": 0.00045556944294795817,
-      "loss": 3.5566,
+      "epoch": 2.420485175202156,
+      "grad_norm": 0.5851937532424927,
+      "learning_rate": 0.0004553135456017269,
+      "loss": 3.5591,
       "step": 22450
     },
     {
-      "epoch": 2.4216984178237,
-      "grad_norm": 0.5626586675643921,
-      "learning_rate": 0.00045524620191789676,
-      "loss": 3.5427,
+      "epoch": 2.4258760107816713,
+      "grad_norm": 0.6141412854194641,
+      "learning_rate": 0.0004549897463572585,
+      "loss": 3.5584,
       "step": 22500
     },
     {
-      "epoch": 2.4270799698633088,
-      "grad_norm": 0.589106023311615,
-      "learning_rate": 0.00045492296088783536,
-      "loss": 3.5564,
+      "epoch": 2.431266846361186,
+      "grad_norm": 0.5884567499160767,
+      "learning_rate": 0.00045466594711279,
+      "loss": 3.5366,
       "step": 22550
     },
     {
-      "epoch": 2.432461521902917,
-      "grad_norm": 0.5803443193435669,
-      "learning_rate": 0.0004545997198577739,
-      "loss": 3.536,
+      "epoch": 2.4366576819407006,
+      "grad_norm": 0.6018183827400208,
+      "learning_rate": 0.00045434214786832164,
+      "loss": 3.5546,
       "step": 22600
     },
     {
-      "epoch": 2.437843073942525,
-      "grad_norm": 0.61843341588974,
-      "learning_rate": 0.0004542764788277125,
-      "loss": 3.5557,
+      "epoch": 2.442048517520216,
+      "grad_norm": 0.5768996477127075,
+      "learning_rate": 0.00045401834862385314,
+      "loss": 3.5486,
       "step": 22650
     },
     {
-      "epoch": 2.443224625982133,
-      "grad_norm": 0.6228456497192383,
-      "learning_rate": 0.00045395323779765103,
-      "loss": 3.5369,
+      "epoch": 2.4474393530997305,
+      "grad_norm": 0.6336095333099365,
+      "learning_rate": 0.00045369454937938474,
+      "loss": 3.557,
       "step": 22700
     },
     {
-      "epoch": 2.4486061780217416,
-      "grad_norm": 0.5836499333381653,
-      "learning_rate": 0.0004536299967675897,
-      "loss": 3.5434,
+      "epoch": 2.452830188679245,
+      "grad_norm": 0.5800439715385437,
+      "learning_rate": 0.0004533707501349163,
+      "loss": 3.5454,
       "step": 22750
     },
     {
-      "epoch": 2.4539877300613497,
-      "grad_norm": 0.6299383044242859,
-      "learning_rate": 0.0004533067557375283,
-      "loss": 3.5723,
+      "epoch": 2.4582210242587603,
+      "grad_norm": 0.5589069724082947,
+      "learning_rate": 0.0004530469508904479,
+      "loss": 3.5432,
       "step": 22800
     },
     {
-      "epoch": 2.459369282100958,
-      "grad_norm": 0.6288545727729797,
-      "learning_rate": 0.0004529835147074668,
-      "loss": 3.5338,
+      "epoch": 2.463611859838275,
+      "grad_norm": 0.5995154976844788,
+      "learning_rate": 0.00045272315164597945,
+      "loss": 3.5327,
       "step": 22850
     },
     {
-      "epoch": 2.464750834140566,
-      "grad_norm": 0.592365026473999,
-      "learning_rate": 0.0004526602736774054,
-      "loss": 3.5546,
+      "epoch": 2.4690026954177897,
+      "grad_norm": 0.5498743057250977,
+      "learning_rate": 0.000452399352401511,
+      "loss": 3.5464,
       "step": 22900
     },
     {
-      "epoch": 2.4701323861801745,
-      "grad_norm": 0.6130805611610413,
-      "learning_rate": 0.00045233703264734395,
-      "loss": 3.5488,
+      "epoch": 2.4743935309973044,
+      "grad_norm": 0.6081226468086243,
+      "learning_rate": 0.0004520755531570426,
+      "loss": 3.5324,
       "step": 22950
     },
     {
-      "epoch": 2.4755139382197826,
-      "grad_norm": 0.6213376522064209,
-      "learning_rate": 0.00045201379161728255,
-      "loss": 3.5467,
+      "epoch": 2.4797843665768196,
+      "grad_norm": 0.5983739495277405,
+      "learning_rate": 0.00045175175391257415,
+      "loss": 3.5478,
       "step": 23000
     },
     {
-      "epoch": 2.4755139382197826,
-      "eval_accuracy": 0.36592825801931655,
-      "eval_loss": 3.537663221359253,
-      "eval_runtime": 183.729,
-      "eval_samples_per_second": 98.03,
-      "eval_steps_per_second": 6.129,
+      "epoch": 2.4797843665768196,
+      "eval_accuracy": 0.365551993068815,
+      "eval_loss": 3.5384716987609863,
+      "eval_runtime": 183.9573,
+      "eval_samples_per_second": 97.909,
+      "eval_steps_per_second": 6.121,
       "step": 23000
     },
     {
-      "epoch": 2.4808954902593907,
-      "grad_norm": 0.5639161467552185,
-      "learning_rate": 0.0004516905505872212,
-      "loss": 3.5533,
+      "epoch": 2.4851752021563343,
+      "grad_norm": 0.6446165442466736,
+      "learning_rate": 0.00045142795466810576,
+      "loss": 3.555,
       "step": 23050
     },
     {
-      "epoch": 2.4862770422989993,
-      "grad_norm": 0.5921719074249268,
-      "learning_rate": 0.00045136730955715973,
-      "loss": 3.545,
+      "epoch": 2.490566037735849,
+      "grad_norm": 0.5573102831840515,
+      "learning_rate": 0.0004511041554236373,
+      "loss": 3.5531,
       "step": 23100
     },
     {
-      "epoch": 2.4916585943386074,
-      "grad_norm": 0.5608890056610107,
-      "learning_rate": 0.00045104406852709833,
-      "loss": 3.5461,
+      "epoch": 2.4959568733153636,
+      "grad_norm": 0.5586193203926086,
+      "learning_rate": 0.0004507803561791689,
+      "loss": 3.5636,
       "step": 23150
     },
     {
-      "epoch": 2.4970401463782155,
-      "grad_norm": 0.5927132964134216,
-      "learning_rate": 0.0004507208274970369,
-      "loss": 3.5691,
+      "epoch": 2.501347708894879,
+      "grad_norm": 0.6083266735076904,
+      "learning_rate": 0.00045045655693470046,
+      "loss": 3.5558,
       "step": 23200
     },
     {
-      "epoch": 2.5024216984178236,
-      "grad_norm": 0.5782831907272339,
-      "learning_rate": 0.00045039758646697546,
-      "loss": 3.5516,
+      "epoch": 2.5067385444743935,
+      "grad_norm": 0.5813544988632202,
+      "learning_rate": 0.00045013275769023207,
+      "loss": 3.5534,
       "step": 23250
     },
     {
-      "epoch": 2.5078032504574317,
-      "grad_norm": 0.5783706307411194,
-      "learning_rate": 0.0004500743454369141,
-      "loss": 3.5644,
+      "epoch": 2.512129380053908,
+      "grad_norm": 0.5284630656242371,
+      "learning_rate": 0.0004498154344306529,
+      "loss": 3.5499,
       "step": 23300
     },
     {
-      "epoch": 2.5131848024970402,
-      "grad_norm": 0.66034334897995,
-      "learning_rate": 0.0004497511044068527,
-      "loss": 3.5407,
+      "epoch": 2.5175202156334233,
+      "grad_norm": 0.5972829461097717,
+      "learning_rate": 0.0004494916351861845,
+      "loss": 3.5508,
       "step": 23350
     },
     {
-      "epoch": 2.5185663545366483,
-      "grad_norm": 0.6267566084861755,
-      "learning_rate": 0.00044942786337679125,
-      "loss": 3.5545,
+      "epoch": 2.522911051212938,
+      "grad_norm": 0.5728029608726501,
+      "learning_rate": 0.00044916783594171607,
+      "loss": 3.5279,
       "step": 23400
     },
     {
-      "epoch": 2.5239479065762565,
-      "grad_norm": 0.5771489143371582,
-      "learning_rate": 0.00044910462234672984,
-      "loss": 3.5338,
+      "epoch": 2.5283018867924527,
+      "grad_norm": 0.6097074151039124,
+      "learning_rate": 0.00044884403669724767,
+      "loss": 3.5448,
       "step": 23450
     },
     {
-      "epoch": 2.529329458615865,
-      "grad_norm": 0.5941733121871948,
-      "learning_rate": 0.0004487813813166684,
-      "loss": 3.5503,
+      "epoch": 2.533692722371968,
+      "grad_norm": 0.5804665684700012,
+      "learning_rate": 0.0004485202374527792,
+      "loss": 3.5404,
       "step": 23500
     },
     {
-      "epoch": 2.534711010655473,
-      "grad_norm": 0.5713851451873779,
-      "learning_rate": 0.000448458140286607,
-      "loss": 3.5481,
+      "epoch": 2.5390835579514826,
+      "grad_norm": 0.5701937675476074,
+      "learning_rate": 0.00044819643820831083,
+      "loss": 3.5421,
       "step": 23550
     },
     {
-      "epoch": 2.540092562695081,
-      "grad_norm": 0.6372710466384888,
-      "learning_rate": 0.00044813489925654563,
-      "loss": 3.5456,
+      "epoch": 2.5444743935309972,
+      "grad_norm": 0.5812232494354248,
+      "learning_rate": 0.0004478726389638424,
+      "loss": 3.5473,
       "step": 23600
     },
     {
-      "epoch": 2.5454741147346893,
-      "grad_norm": 0.6033260226249695,
-      "learning_rate": 0.00044781165822648417,
-      "loss": 3.5446,
+      "epoch": 2.5498652291105124,
+      "grad_norm": 0.5841829776763916,
+      "learning_rate": 0.00044754883971937393,
+      "loss": 3.5346,
       "step": 23650
     },
     {
-      "epoch": 2.550855666774298,
-      "grad_norm": 0.5545850992202759,
-      "learning_rate": 0.00044748841719642276,
-      "loss": 3.5474,
+      "epoch": 2.555256064690027,
+      "grad_norm": 0.5946136713027954,
+      "learning_rate": 0.00044722504047490553,
+      "loss": 3.5229,
       "step": 23700
     },
     {
-      "epoch": 2.556237218813906,
-      "grad_norm": 0.595909595489502,
-      "learning_rate": 0.00044716517616636136,
-      "loss": 3.5294,
+      "epoch": 2.560646900269542,
+      "grad_norm": 0.6016232371330261,
+      "learning_rate": 0.0004469012412304371,
+      "loss": 3.5331,
       "step": 23750
     },
     {
-      "epoch": 2.561618770853514,
-      "grad_norm": 0.6029964685440063,
-      "learning_rate": 0.0004468419351362999,
-      "loss": 3.5316,
+      "epoch": 2.5660377358490565,
+      "grad_norm": 0.5311474800109863,
+      "learning_rate": 0.0004465774419859687,
+      "loss": 3.5504,
       "step": 23800
     },
     {
-      "epoch": 2.567000322893122,
-      "grad_norm": 0.6632453203201294,
-      "learning_rate": 0.0004465186941062385,
-      "loss": 3.5433,
+      "epoch": 2.571428571428571,
+      "grad_norm": 0.5677634477615356,
+      "learning_rate": 0.00044625364274150024,
+      "loss": 3.5389,
       "step": 23850
     },
     {
-      "epoch": 2.5723818749327307,
-      "grad_norm": 0.5608298778533936,
-      "learning_rate": 0.00044619545307617714,
-      "loss": 3.5577,
+      "epoch": 2.5768194070080863,
+      "grad_norm": 0.6015117168426514,
+      "learning_rate": 0.00044592984349703184,
+      "loss": 3.5387,
       "step": 23900
     },
     {
-      "epoch": 2.577763426972339,
-      "grad_norm": 0.6114339828491211,
-      "learning_rate": 0.0004458722120461157,
-      "loss": 3.5397,
+      "epoch": 2.582210242587601,
+      "grad_norm": 0.6096858978271484,
+      "learning_rate": 0.00044560604425256334,
+      "loss": 3.5308,
       "step": 23950
     },
     {
-      "epoch": 2.583144979011947,
-      "grad_norm": 0.573918342590332,
-      "learning_rate": 0.0004455489710160543,
-      "loss": 3.5295,
+      "epoch": 2.5876010781671157,
+      "grad_norm": 0.5935425162315369,
+      "learning_rate": 0.000445282245008095,
+      "loss": 3.5356,
       "step": 24000
     },
     {
-      "epoch": 2.583144979011947,
-      "eval_accuracy": 0.3665664850823388,
-      "eval_loss": 3.5296592712402344,
-      "eval_runtime": 184.218,
-      "eval_samples_per_second": 97.77,
-      "eval_steps_per_second": 6.112,
+      "epoch": 2.5876010781671157,
+      "eval_accuracy": 0.3669573095198075,
+      "eval_loss": 3.5286059379577637,
+      "eval_runtime": 183.6483,
+      "eval_samples_per_second": 98.073,
+      "eval_steps_per_second": 6.131,
       "step": 24000
     },
     {
-      "epoch": 2.5885265310515555,
-      "grad_norm": 0.5977112054824829,
-      "learning_rate": 0.0004452257299859928,
-      "loss": 3.5388,
+      "epoch": 2.592991913746631,
+      "grad_norm": 0.6075648665428162,
+      "learning_rate": 0.0004449584457636265,
+      "loss": 3.5602,
       "step": 24050
     },
     {
-      "epoch": 2.5939080830911636,
-      "grad_norm": 0.6366156935691833,
-      "learning_rate": 0.0004449024889559314,
-      "loss": 3.525,
+      "epoch": 2.5983827493261455,
+      "grad_norm": 0.6200929284095764,
+      "learning_rate": 0.00044463464651915805,
+      "loss": 3.5351,
       "step": 24100
     },
     {
-      "epoch": 2.5992896351307717,
-      "grad_norm": 0.5739189982414246,
-      "learning_rate": 0.00044457924792587,
-      "loss": 3.5505,
+      "epoch": 2.6037735849056602,
+      "grad_norm": 0.5483993291854858,
+      "learning_rate": 0.00044431084727468965,
+      "loss": 3.548,
       "step": 24150
     },
     {
-      "epoch": 2.60467118717038,
-      "grad_norm": 0.6196433305740356,
-      "learning_rate": 0.0004442560068958086,
-      "loss": 3.5391,
+      "epoch": 2.6091644204851754,
+      "grad_norm": 0.6451470851898193,
+      "learning_rate": 0.0004439870480302212,
+      "loss": 3.5318,
       "step": 24200
     },
     {
-      "epoch": 2.610052739209988,
-      "grad_norm": 0.6473907828330994,
-      "learning_rate": 0.0004439327658657472,
-      "loss": 3.5327,
+      "epoch": 2.61455525606469,
+      "grad_norm": 0.6065393090248108,
+      "learning_rate": 0.0004436632487857528,
+      "loss": 3.5511,
       "step": 24250
     },
     {
-      "epoch": 2.6154342912495965,
-      "grad_norm": 0.5651465058326721,
-      "learning_rate": 0.000443615989656287,
-      "loss": 3.5296,
+      "epoch": 2.6199460916442048,
+      "grad_norm": 0.5803229808807373,
+      "learning_rate": 0.00044333944954128436,
+      "loss": 3.5437,
       "step": 24300
     },
     {
-      "epoch": 2.6208158432892046,
-      "grad_norm": 0.6047065258026123,
-      "learning_rate": 0.0004432927486262256,
-      "loss": 3.5455,
+      "epoch": 2.62533692722372,
+      "grad_norm": 0.5576533079147339,
+      "learning_rate": 0.00044301565029681596,
+      "loss": 3.536,
       "step": 24350
     },
     {
-      "epoch": 2.6261973953288127,
-      "grad_norm": 0.5989174246788025,
-      "learning_rate": 0.00044296950759616414,
-      "loss": 3.5358,
+      "epoch": 2.6307277628032346,
+      "grad_norm": 0.6252774000167847,
+      "learning_rate": 0.0004426918510523475,
+      "loss": 3.5387,
       "step": 24400
     },
     {
-      "epoch": 2.6315789473684212,
-      "grad_norm": 0.5492793321609497,
-      "learning_rate": 0.00044264626656610273,
-      "loss": 3.5434,
+      "epoch": 2.6361185983827493,
+      "grad_norm": 0.5675030946731567,
+      "learning_rate": 0.0004423680518078791,
+      "loss": 3.5379,
       "step": 24450
     },
     {
-      "epoch": 2.6369604994080293,
-      "grad_norm": 0.6315013766288757,
-      "learning_rate": 0.0004423230255360413,
-      "loss": 3.5234,
+      "epoch": 2.641509433962264,
+      "grad_norm": 0.5933482646942139,
+      "learning_rate": 0.00044204425256341067,
+      "loss": 3.5467,
       "step": 24500
     },
     {
-      "epoch": 2.6423420514476375,
-      "grad_norm": 0.5565677285194397,
-      "learning_rate": 0.00044199978450598,
-      "loss": 3.5216,
+      "epoch": 2.6469002695417787,
+      "grad_norm": 0.6065881848335266,
+      "learning_rate": 0.0004417204533189422,
+      "loss": 3.5304,
       "step": 24550
     },
     {
-      "epoch": 2.6477236034872456,
-      "grad_norm": 0.5790075063705444,
-      "learning_rate": 0.0004416765434759185,
-      "loss": 3.5266,
+      "epoch": 2.652291105121294,
+      "grad_norm": 0.616068959236145,
+      "learning_rate": 0.0004413966540744738,
+      "loss": 3.5389,
       "step": 24600
     },
     {
-      "epoch": 2.653105155526854,
-      "grad_norm": 0.8480896949768066,
-      "learning_rate": 0.0004413533024458571,
-      "loss": 3.5287,
+      "epoch": 2.6576819407008085,
+      "grad_norm": 0.5536131858825684,
+      "learning_rate": 0.0004410728548300053,
+      "loss": 3.5346,
       "step": 24650
     },
     {
-      "epoch": 2.658486707566462,
-      "grad_norm": 0.5691506266593933,
-      "learning_rate": 0.00044103006141579565,
-      "loss": 3.5434,
+      "epoch": 2.6630727762803232,
+      "grad_norm": 0.6090572476387024,
+      "learning_rate": 0.0004407490555855369,
+      "loss": 3.5411,
       "step": 24700
     },
     {
-      "epoch": 2.6638682596060703,
-      "grad_norm": 0.6176285147666931,
-      "learning_rate": 0.0004407132852063355,
-      "loss": 3.5372,
+      "epoch": 2.6684636118598384,
+      "grad_norm": 0.584028422832489,
+      "learning_rate": 0.0004404252563410685,
+      "loss": 3.5584,
       "step": 24750
     },
     {
-      "epoch": 2.6692498116456784,
-      "grad_norm": 0.6177787780761719,
-      "learning_rate": 0.00044039004417627405,
-      "loss": 3.5313,
+      "epoch": 2.673854447439353,
+      "grad_norm": 0.5654086470603943,
+      "learning_rate": 0.0004401014570966001,
+      "loss": 3.5504,
       "step": 24800
     },
     {
-      "epoch": 2.674631363685287,
-      "grad_norm": 0.6155895590782166,
-      "learning_rate": 0.00044006680314621265,
-      "loss": 3.5309,
+      "epoch": 2.6792452830188678,
+      "grad_norm": 0.5463288426399231,
+      "learning_rate": 0.00043977765785213163,
+      "loss": 3.5503,
       "step": 24850
     },
     {
-      "epoch": 2.680012915724895,
-      "grad_norm": 0.6472774147987366,
-      "learning_rate": 0.0004397435621161513,
-      "loss": 3.5388,
+      "epoch": 2.684636118598383,
+      "grad_norm": 0.58181232213974,
+      "learning_rate": 0.0004394538586076632,
+      "loss": 3.5224,
       "step": 24900
     },
     {
-      "epoch": 2.685394467764503,
-      "grad_norm": 0.5491383075714111,
-      "learning_rate": 0.00043942032108608984,
-      "loss": 3.546,
+      "epoch": 2.6900269541778976,
+      "grad_norm": 0.5880516767501831,
+      "learning_rate": 0.0004391300593631948,
+      "loss": 3.5417,
       "step": 24950
     },
     {
-      "epoch": 2.6907760198041117,
-      "grad_norm": 0.6182669401168823,
-      "learning_rate": 0.00043909708005602843,
-      "loss": 3.542,
+      "epoch": 2.6954177897574123,
+      "grad_norm": 0.5604709982872009,
+      "learning_rate": 0.00043880626011872634,
+      "loss": 3.5206,
       "step": 25000
     },
     {
-      "epoch": 2.6907760198041117,
-      "eval_accuracy": 0.367936815303455,
-      "eval_loss": 3.519559383392334,
-      "eval_runtime": 183.8108,
-      "eval_samples_per_second": 97.987,
-      "eval_steps_per_second": 6.126,
+      "epoch": 2.6954177897574123,
+      "eval_accuracy": 0.368059810372461,
+      "eval_loss": 3.5181727409362793,
+      "eval_runtime": 183.8726,
+      "eval_samples_per_second": 97.954,
+      "eval_steps_per_second": 6.124,
       "step": 25000
     },
     {
-      "epoch": 2.69615757184372,
-      "grad_norm": 0.6307483315467834,
-      "learning_rate": 0.00043877383902596697,
-      "loss": 3.5389,
+      "epoch": 2.7008086253369274,
+      "grad_norm": 0.5966419577598572,
+      "learning_rate": 0.00043848246087425794,
+      "loss": 3.5336,
       "step": 25050
     },
     {
-      "epoch": 2.701539123883328,
-      "grad_norm": 0.5816606879234314,
-      "learning_rate": 0.00043845059799590557,
-      "loss": 3.5345,
+      "epoch": 2.706199460916442,
+      "grad_norm": 0.5811107754707336,
+      "learning_rate": 0.0004381586616297895,
+      "loss": 3.5356,
       "step": 25100
     },
     {
-      "epoch": 2.706920675922936,
-      "grad_norm": 0.6173508763313293,
-      "learning_rate": 0.00043812735696584416,
-      "loss": 3.547,
+      "epoch": 2.711590296495957,
+      "grad_norm": 0.5727028846740723,
+      "learning_rate": 0.0004378348623853211,
+      "loss": 3.5371,
       "step": 25150
     },
     {
-      "epoch": 2.712302227962544,
-      "grad_norm": 0.5916981101036072,
-      "learning_rate": 0.00043780411593578275,
-      "loss": 3.545,
+      "epoch": 2.7169811320754715,
+      "grad_norm": 0.6205406785011292,
+      "learning_rate": 0.00043751106314085265,
+      "loss": 3.5372,
       "step": 25200
     },
     {
-      "epoch": 2.7176837800021527,
-      "grad_norm": 0.6386535167694092,
-      "learning_rate": 0.00043748087490572135,
-      "loss": 3.5267,
+      "epoch": 2.7223719676549867,
+      "grad_norm": 0.6361661553382874,
+      "learning_rate": 0.00043718726389638425,
+      "loss": 3.5299,
       "step": 25250
     },
     {
-      "epoch": 2.723065332041761,
-      "grad_norm": 0.600104808807373,
-      "learning_rate": 0.00043715763387565994,
-      "loss": 3.5417,
+      "epoch": 2.7277628032345014,
+      "grad_norm": 0.6152952313423157,
+      "learning_rate": 0.0004368699406368051,
+      "loss": 3.537,
       "step": 25300
     },
     {
-      "epoch": 2.728446884081369,
-      "grad_norm": 0.5657129287719727,
-      "learning_rate": 0.0004368343928455985,
-      "loss": 3.5277,
+      "epoch": 2.733153638814016,
+      "grad_norm": 0.628601610660553,
+      "learning_rate": 0.0004365461413923367,
+      "loss": 3.5323,
       "step": 25350
     },
     {
-      "epoch": 2.7338284361209775,
-      "grad_norm": 0.6129351258277893,
-      "learning_rate": 0.0004365111518155371,
-      "loss": 3.5314,
+      "epoch": 2.7385444743935308,
+      "grad_norm": 0.5765767693519592,
+      "learning_rate": 0.00043622234214786825,
+      "loss": 3.5279,
       "step": 25400
     },
     {
-      "epoch": 2.7392099881605856,
-      "grad_norm": 0.5998933911323547,
-      "learning_rate": 0.0004361879107854756,
-      "loss": 3.5334,
+      "epoch": 2.743935309973046,
+      "grad_norm": 0.5758444666862488,
+      "learning_rate": 0.00043589854290339985,
+      "loss": 3.5427,
       "step": 25450
     },
     {
-      "epoch": 2.7445915402001937,
-      "grad_norm": 0.5449445843696594,
-      "learning_rate": 0.00043586466975541427,
-      "loss": 3.5356,
+      "epoch": 2.7493261455525606,
+      "grad_norm": 0.5565394759178162,
+      "learning_rate": 0.0004355747436589314,
+      "loss": 3.5207,
       "step": 25500
     },
     {
-      "epoch": 2.749973092239802,
-      "grad_norm": 0.6112343072891235,
-      "learning_rate": 0.00043554142872535286,
-      "loss": 3.5306,
+      "epoch": 2.7547169811320753,
+      "grad_norm": 0.576356053352356,
+      "learning_rate": 0.000435250944414463,
+      "loss": 3.5136,
       "step": 25550
     },
     {
-      "epoch": 2.7553546442794103,
-      "grad_norm": 0.6336021423339844,
-      "learning_rate": 0.0004352181876952914,
-      "loss": 3.5165,
+      "epoch": 2.7601078167115904,
+      "grad_norm": 0.5538831353187561,
+      "learning_rate": 0.00043492714516999456,
+      "loss": 3.5184,
       "step": 25600
     },
     {
-      "epoch": 2.7607361963190185,
-      "grad_norm": 0.6483476161956787,
-      "learning_rate": 0.00043489494666523,
-      "loss": 3.5418,
+      "epoch": 2.765498652291105,
+      "grad_norm": 0.6034234762191772,
+      "learning_rate": 0.0004346033459255261,
+      "loss": 3.5247,
       "step": 25650
     },
     {
-      "epoch": 2.7661177483586266,
-      "grad_norm": 0.588033139705658,
-      "learning_rate": 0.00043457170563516854,
-      "loss": 3.5375,
+      "epoch": 2.77088948787062,
+      "grad_norm": 0.6086429953575134,
+      "learning_rate": 0.0004342795466810577,
+      "loss": 3.5128,
       "step": 25700
     },
     {
-      "epoch": 2.7714993003982347,
-      "grad_norm": 0.5726504921913147,
-      "learning_rate": 0.00043424846460510713,
-      "loss": 3.5235,
+      "epoch": 2.776280323450135,
+      "grad_norm": 0.5929721593856812,
+      "learning_rate": 0.00043396222342147867,
+      "loss": 3.5175,
       "step": 25750
     },
     {
-      "epoch": 2.776880852437843,
-      "grad_norm": 0.6130396127700806,
-      "learning_rate": 0.0004339252235750458,
-      "loss": 3.5322,
+      "epoch": 2.7816711590296497,
+      "grad_norm": 0.5821933150291443,
+      "learning_rate": 0.0004336384241770102,
+      "loss": 3.5297,
       "step": 25800
     },
     {
-      "epoch": 2.7822624044774513,
-      "grad_norm": 0.6128621697425842,
-      "learning_rate": 0.0004336019825449843,
-      "loss": 3.5161,
+      "epoch": 2.7870619946091644,
+      "grad_norm": 0.5622891187667847,
+      "learning_rate": 0.0004333146249325418,
+      "loss": 3.5163,
       "step": 25850
     },
     {
-      "epoch": 2.7876439565170594,
-      "grad_norm": 0.5874921679496765,
-      "learning_rate": 0.0004332787415149229,
-      "loss": 3.5316,
+      "epoch": 2.7924528301886795,
+      "grad_norm": 0.5431972146034241,
+      "learning_rate": 0.0004329908256880734,
+      "loss": 3.5208,
       "step": 25900
     },
     {
-      "epoch": 2.793025508556668,
-      "grad_norm": 0.609630823135376,
-      "learning_rate": 0.0004329555004848615,
-      "loss": 3.5424,
+      "epoch": 2.797843665768194,
+      "grad_norm": 0.5647171139717102,
+      "learning_rate": 0.000432667026443605,
+      "loss": 3.5288,
       "step": 25950
     },
     {
-      "epoch": 2.798407060596276,
-      "grad_norm": 0.621626615524292,
-      "learning_rate": 0.00043263225945480005,
-      "loss": 3.5277,
+      "epoch": 2.803234501347709,
+      "grad_norm": 0.5956171154975891,
+      "learning_rate": 0.0004323432271991365,
+      "loss": 3.5331,
       "step": 26000
     },
     {
-      "epoch": 2.798407060596276,
-      "eval_accuracy": 0.3684354234056357,
-      "eval_loss": 3.5107884407043457,
-      "eval_runtime": 183.6656,
-      "eval_samples_per_second": 98.064,
+      "epoch": 2.803234501347709,
+      "eval_accuracy": 0.3691224356152866,
+      "eval_loss": 3.5090320110321045,
+      "eval_runtime": 183.655,
+      "eval_samples_per_second": 98.07,
       "eval_steps_per_second": 6.131,
       "step": 26000
     },
     {
-      "epoch": 2.803788612635884,
-      "grad_norm": 0.6718447804450989,
-      "learning_rate": 0.0004323090184247387,
-      "loss": 3.5398,
+      "epoch": 2.8086253369272236,
+      "grad_norm": 0.60736483335495,
+      "learning_rate": 0.000432019427954668,
+      "loss": 3.5338,
       "step": 26050
     },
     {
-      "epoch": 2.8091701646754923,
-      "grad_norm": 0.638828694820404,
-      "learning_rate": 0.0004319857773946773,
-      "loss": 3.5499,
+      "epoch": 2.8140161725067383,
+      "grad_norm": 0.6084789037704468,
+      "learning_rate": 0.00043169562871019963,
+      "loss": 3.5146,
       "step": 26100
     },
     {
-      "epoch": 2.8145517167151004,
-      "grad_norm": 0.5668681859970093,
-      "learning_rate": 0.00043166253636461584,
-      "loss": 3.5197,
+      "epoch": 2.8194070080862534,
+      "grad_norm": 0.588055431842804,
+      "learning_rate": 0.0004313718294657312,
+      "loss": 3.5224,
       "step": 26150
     },
     {
-      "epoch": 2.819933268754709,
-      "grad_norm": 0.5784147381782532,
-      "learning_rate": 0.00043133929533455443,
-      "loss": 3.5241,
+      "epoch": 2.824797843665768,
+      "grad_norm": 0.5956451892852783,
+      "learning_rate": 0.0004310480302212628,
+      "loss": 3.5207,
       "step": 26200
     },
     {
-      "epoch": 2.825314820794317,
-      "grad_norm": 0.5837064981460571,
-      "learning_rate": 0.00043101605430449297,
-      "loss": 3.4931,
+      "epoch": 2.830188679245283,
+      "grad_norm": 0.574309229850769,
+      "learning_rate": 0.00043072423097679434,
+      "loss": 3.5155,
       "step": 26250
     },
     {
-      "epoch": 2.830696372833925,
-      "grad_norm": 0.5364549160003662,
-      "learning_rate": 0.00043069281327443157,
-      "loss": 3.5163,
+      "epoch": 2.835579514824798,
+      "grad_norm": 0.5954546332359314,
+      "learning_rate": 0.00043040043173232594,
+      "loss": 3.5306,
       "step": 26300
     },
     {
-      "epoch": 2.8360779248735337,
-      "grad_norm": 0.637577474117279,
-      "learning_rate": 0.0004303695722443702,
-      "loss": 3.5169,
+      "epoch": 2.8409703504043127,
+      "grad_norm": 0.5666508674621582,
+      "learning_rate": 0.0004300766324878575,
+      "loss": 3.5346,
       "step": 26350
     },
     {
-      "epoch": 2.841459476913142,
-      "grad_norm": 0.6344022154808044,
-      "learning_rate": 0.00043004633121430876,
-      "loss": 3.5156,
+      "epoch": 2.8463611859838274,
+      "grad_norm": 0.6565783619880676,
+      "learning_rate": 0.0004297528332433891,
+      "loss": 3.5223,
       "step": 26400
     },
     {
-      "epoch": 2.84684102895275,
-      "grad_norm": 0.5951647758483887,
-      "learning_rate": 0.00042972309018424735,
-      "loss": 3.5126,
+      "epoch": 2.8517520215633425,
+      "grad_norm": 0.5818801522254944,
+      "learning_rate": 0.00042942903399892065,
+      "loss": 3.5229,
       "step": 26450
     },
     {
-      "epoch": 2.852222580992358,
-      "grad_norm": 0.6024564504623413,
-      "learning_rate": 0.00042939984915418594,
-      "loss": 3.5264,
+      "epoch": 2.857142857142857,
+      "grad_norm": 0.5973237752914429,
+      "learning_rate": 0.0004291052347544522,
+      "loss": 3.5203,
       "step": 26500
     },
     {
-      "epoch": 2.857604133031966,
-      "grad_norm": 0.5702798962593079,
-      "learning_rate": 0.0004290766081241245,
-      "loss": 3.5238,
+      "epoch": 2.862533692722372,
+      "grad_norm": 0.6071364283561707,
+      "learning_rate": 0.0004287814355099838,
+      "loss": 3.5116,
       "step": 26550
     },
     {
-      "epoch": 2.8629856850715747,
-      "grad_norm": 0.5641034245491028,
-      "learning_rate": 0.0004287533670940631,
-      "loss": 3.5011,
+      "epoch": 2.867924528301887,
+      "grad_norm": 0.621721088886261,
+      "learning_rate": 0.0004284576362655153,
+      "loss": 3.5221,
       "step": 26600
     },
     {
-      "epoch": 2.868367237111183,
-      "grad_norm": 0.6096258163452148,
-      "learning_rate": 0.00042843012606400173,
-      "loss": 3.5268,
+      "epoch": 2.8733153638814017,
+      "grad_norm": 0.5732737183570862,
+      "learning_rate": 0.00042813383702104696,
+      "loss": 3.52,
       "step": 26650
     },
     {
-      "epoch": 2.873748789150791,
-      "grad_norm": 0.5629540085792542,
-      "learning_rate": 0.00042810688503394027,
-      "loss": 3.5278,
+      "epoch": 2.8787061994609164,
+      "grad_norm": 0.5745946168899536,
+      "learning_rate": 0.00042781003777657845,
+      "loss": 3.5173,
       "step": 26700
     },
     {
-      "epoch": 2.8791303411903995,
-      "grad_norm": 0.5838139653205872,
-      "learning_rate": 0.00042778364400387886,
-      "loss": 3.5462,
+      "epoch": 2.884097035040431,
+      "grad_norm": 0.5461169481277466,
+      "learning_rate": 0.00042748623853211006,
+      "loss": 3.5357,
       "step": 26750
     },
     {
-      "epoch": 2.8845118932300076,
-      "grad_norm": 0.5901974439620972,
-      "learning_rate": 0.0004274604029738174,
-      "loss": 3.5334,
+      "epoch": 2.889487870619946,
+      "grad_norm": 0.6288904547691345,
+      "learning_rate": 0.0004271624392876416,
+      "loss": 3.5216,
       "step": 26800
     },
     {
-      "epoch": 2.8898934452696157,
-      "grad_norm": 0.7880048751831055,
-      "learning_rate": 0.000427137161943756,
-      "loss": 3.5302,
+      "epoch": 2.894878706199461,
+      "grad_norm": 0.5487993359565735,
+      "learning_rate": 0.00042683864004317316,
+      "loss": 3.5164,
       "step": 26850
     },
     {
-      "epoch": 2.895274997309224,
-      "grad_norm": 0.6115409135818481,
-      "learning_rate": 0.00042681392091369465,
-      "loss": 3.5149,
+      "epoch": 2.9002695417789757,
+      "grad_norm": 0.6259924173355103,
+      "learning_rate": 0.00042651484079870476,
+      "loss": 3.5397,
       "step": 26900
     },
     {
-      "epoch": 2.9006565493488323,
-      "grad_norm": 0.6812686324119568,
-      "learning_rate": 0.0004264906798836332,
-      "loss": 3.5029,
+      "epoch": 2.9056603773584904,
+      "grad_norm": 0.5501232743263245,
+      "learning_rate": 0.0004261910415542363,
+      "loss": 3.5312,
       "step": 26950
     },
     {
-      "epoch": 2.9060381013884404,
-      "grad_norm": 0.5887460708618164,
-      "learning_rate": 0.0004261674388535718,
-      "loss": 3.5215,
+      "epoch": 2.9110512129380055,
+      "grad_norm": 0.5408157110214233,
+      "learning_rate": 0.0004258672423097679,
+      "loss": 3.5173,
       "step": 27000
     },
     {
-      "epoch": 2.9060381013884404,
-      "eval_accuracy": 0.36936288445602006,
-      "eval_loss": 3.5005524158477783,
-      "eval_runtime": 183.9921,
-      "eval_samples_per_second": 97.89,
-      "eval_steps_per_second": 6.12,
+      "epoch": 2.9110512129380055,
+      "eval_accuracy": 0.3698140112462258,
+      "eval_loss": 3.5014569759368896,
+      "eval_runtime": 183.9248,
+      "eval_samples_per_second": 97.926,
+      "eval_steps_per_second": 6.122,
       "step": 27000
     },
     {
-      "epoch": 2.9114196534280485,
-      "grad_norm": 0.6023585796356201,
-      "learning_rate": 0.0004258441978235104,
-      "loss": 3.5278,
+      "epoch": 2.91644204851752,
+      "grad_norm": 0.6212002635002136,
+      "learning_rate": 0.00042554344306529947,
+      "loss": 3.5328,
       "step": 27050
     },
     {
-      "epoch": 2.9168012054676566,
-      "grad_norm": 0.5896843671798706,
-      "learning_rate": 0.0004255209567934489,
-      "loss": 3.5184,
+      "epoch": 2.921832884097035,
+      "grad_norm": 0.5572149157524109,
+      "learning_rate": 0.0004252196438208311,
+      "loss": 3.539,
       "step": 27100
     },
     {
-      "epoch": 2.922182757507265,
-      "grad_norm": 0.6131631135940552,
-      "learning_rate": 0.0004251977157633875,
-      "loss": 3.5262,
+      "epoch": 2.92722371967655,
+      "grad_norm": 0.6017417311668396,
+      "learning_rate": 0.0004248958445763626,
+      "loss": 3.5234,
       "step": 27150
     },
     {
-      "epoch": 2.9275643095468733,
-      "grad_norm": 0.6602053046226501,
-      "learning_rate": 0.00042487447473332616,
-      "loss": 3.5297,
+      "epoch": 2.9326145552560647,
+      "grad_norm": 0.6045824289321899,
+      "learning_rate": 0.00042457204533189423,
+      "loss": 3.5132,
       "step": 27200
     },
     {
-      "epoch": 2.9329458615864814,
-      "grad_norm": 0.6800670027732849,
-      "learning_rate": 0.0004245512337032647,
-      "loss": 3.5112,
+      "epoch": 2.9380053908355794,
+      "grad_norm": 0.5487478971481323,
+      "learning_rate": 0.0004242482460874258,
+      "loss": 3.5267,
       "step": 27250
     },
     {
-      "epoch": 2.93832741362609,
-      "grad_norm": 0.6654891967773438,
-      "learning_rate": 0.00042423445749380456,
-      "loss": 3.5178,
+      "epoch": 2.9433962264150946,
+      "grad_norm": 0.5796363949775696,
+      "learning_rate": 0.0004239244468429573,
+      "loss": 3.5224,
       "step": 27300
     },
     {
-      "epoch": 2.943708965665698,
-      "grad_norm": 0.6592779159545898,
-      "learning_rate": 0.0004239112164637431,
-      "loss": 3.5272,
+      "epoch": 2.9487870619946093,
+      "grad_norm": 0.6089639663696289,
+      "learning_rate": 0.0004236006475984889,
+      "loss": 3.5161,
       "step": 27350
     },
     {
-      "epoch": 2.949090517705306,
-      "grad_norm": 0.5880259871482849,
-      "learning_rate": 0.0004235879754336817,
-      "loss": 3.5181,
+      "epoch": 2.954177897574124,
+      "grad_norm": 0.5658247470855713,
+      "learning_rate": 0.00042327684835402043,
+      "loss": 3.5189,
       "step": 27400
     },
     {
-      "epoch": 2.9544720697449143,
-      "grad_norm": 0.591659426689148,
-      "learning_rate": 0.00042326473440362024,
-      "loss": 3.5223,
+      "epoch": 2.9595687331536387,
+      "grad_norm": 0.5259602665901184,
+      "learning_rate": 0.00042295304910955204,
+      "loss": 3.5155,
       "step": 27450
     },
     {
-      "epoch": 2.9598536217845224,
-      "grad_norm": 0.562332272529602,
-      "learning_rate": 0.00042294149337355883,
-      "loss": 3.5198,
+      "epoch": 2.964959568733154,
+      "grad_norm": 0.5658301711082458,
+      "learning_rate": 0.0004226292498650836,
+      "loss": 3.5213,
       "step": 27500
     },
     {
-      "epoch": 2.965235173824131,
-      "grad_norm": 0.5543864369392395,
-      "learning_rate": 0.0004226182523434974,
-      "loss": 3.5125,
+      "epoch": 2.9703504043126685,
+      "grad_norm": 0.5679548978805542,
+      "learning_rate": 0.0004223054506206152,
+      "loss": 3.5261,
       "step": 27550
     },
     {
-      "epoch": 2.970616725863739,
-      "grad_norm": 0.6262898445129395,
-      "learning_rate": 0.000422295011313436,
-      "loss": 3.5239,
+      "epoch": 2.975741239892183,
+      "grad_norm": 0.568275511264801,
+      "learning_rate": 0.00042198165137614674,
+      "loss": 3.5253,
       "step": 27600
     },
     {
-      "epoch": 2.975998277903347,
-      "grad_norm": 0.6862922310829163,
-      "learning_rate": 0.0004219717702833746,
-      "loss": 3.5356,
+      "epoch": 2.981132075471698,
+      "grad_norm": 0.5689859986305237,
+      "learning_rate": 0.00042165785213167835,
+      "loss": 3.5271,
       "step": 27650
     },
     {
-      "epoch": 2.9813798299429557,
-      "grad_norm": 0.6158599853515625,
-      "learning_rate": 0.00042164852925331316,
-      "loss": 3.5296,
+      "epoch": 2.986522911051213,
+      "grad_norm": 0.6435585618019104,
+      "learning_rate": 0.0004213340528872099,
+      "loss": 3.519,
       "step": 27700
     },
     {
-      "epoch": 2.986761381982564,
-      "grad_norm": 0.5659825801849365,
-      "learning_rate": 0.00042132528822325175,
-      "loss": 3.5115,
+      "epoch": 2.9919137466307277,
+      "grad_norm": 0.5717616677284241,
+      "learning_rate": 0.00042101025364274145,
+      "loss": 3.5292,
       "step": 27750
     },
     {
-      "epoch": 2.992142934022172,
-      "grad_norm": 0.5528464317321777,
-      "learning_rate": 0.00042100204719319035,
-      "loss": 3.5091,
+      "epoch": 2.9973045822102424,
+      "grad_norm": 0.6246117949485779,
+      "learning_rate": 0.00042068645439827305,
+      "loss": 3.5258,
       "step": 27800
     },
     {
-      "epoch": 2.9975244860617805,
-      "grad_norm": 0.5595229864120483,
-      "learning_rate": 0.00042067880616312894,
-      "loss": 3.5031,
+      "epoch": 3.0026954177897576,
+      "grad_norm": 0.6157830953598022,
+      "learning_rate": 0.0004203626551538046,
+      "loss": 3.4716,
       "step": 27850
     },
     {
-      "epoch": 3.0029060381013886,
-      "grad_norm": 0.5878441333770752,
-      "learning_rate": 0.00042035556513306754,
-      "loss": 3.4675,
+      "epoch": 3.0080862533692723,
+      "grad_norm": 0.6134538054466248,
+      "learning_rate": 0.0004200388559093362,
+      "loss": 3.4209,
       "step": 27900
     },
     {
-      "epoch": 3.0082875901409967,
-      "grad_norm": 0.6017246842384338,
-      "learning_rate": 0.00042003232410300613,
-      "loss": 3.4328,
+      "epoch": 3.013477088948787,
+      "grad_norm": 0.6604688763618469,
+      "learning_rate": 0.0004197150566648677,
+      "loss": 3.4191,
       "step": 27950
     },
     {
-      "epoch": 3.0136691421806048,
-      "grad_norm": 0.6545405387878418,
-      "learning_rate": 0.00041970908307294467,
-      "loss": 3.426,
+      "epoch": 3.018867924528302,
+      "grad_norm": 0.5869320034980774,
+      "learning_rate": 0.00041939125742039936,
+      "loss": 3.4408,
       "step": 28000
     },
     {
-      "epoch": 3.0136691421806048,
-      "eval_accuracy": 0.37061956375648236,
-      "eval_loss": 3.496546983718872,
-      "eval_runtime": 184.0749,
-      "eval_samples_per_second": 97.846,
-      "eval_steps_per_second": 6.117,
+      "epoch": 3.018867924528302,
+      "eval_accuracy": 0.3707334319829119,
+      "eval_loss": 3.496290683746338,
+      "eval_runtime": 183.5001,
+      "eval_samples_per_second": 98.153,
+      "eval_steps_per_second": 6.136,
       "step": 28000
     },
     {
-      "epoch": 3.0190506942202133,
-      "grad_norm": 0.5815675854682922,
-      "learning_rate": 0.00041938584204288327,
-      "loss": 3.406,
+      "epoch": 3.024258760107817,
+      "grad_norm": 0.5957772135734558,
+      "learning_rate": 0.00041906745817593086,
+      "loss": 3.4343,
       "step": 28050
     },
     {
-      "epoch": 3.0244322462598214,
-      "grad_norm": 0.5926801562309265,
-      "learning_rate": 0.0004190626010128218,
-      "loss": 3.4401,
+      "epoch": 3.0296495956873315,
+      "grad_norm": 0.5896233320236206,
+      "learning_rate": 0.00041874365893146247,
+      "loss": 3.4408,
       "step": 28100
     },
     {
-      "epoch": 3.0298137982994295,
-      "grad_norm": 0.578240692615509,
-      "learning_rate": 0.00041873935998276046,
-      "loss": 3.445,
+      "epoch": 3.035040431266846,
+      "grad_norm": 0.6093875765800476,
+      "learning_rate": 0.000418419859686994,
+      "loss": 3.4303,
       "step": 28150
     },
     {
-      "epoch": 3.0351953503390376,
-      "grad_norm": 0.5973519682884216,
-      "learning_rate": 0.00041841611895269905,
-      "loss": 3.4291,
+      "epoch": 3.0404312668463613,
+      "grad_norm": 0.6102927923202515,
+      "learning_rate": 0.00041809606044252557,
+      "loss": 3.434,
       "step": 28200
     },
     {
-      "epoch": 3.040576902378646,
-      "grad_norm": 0.6175142526626587,
-      "learning_rate": 0.0004180928779226376,
-      "loss": 3.4451,
+      "epoch": 3.045822102425876,
+      "grad_norm": 0.5978097915649414,
+      "learning_rate": 0.00041777226119805717,
+      "loss": 3.4171,
       "step": 28250
     },
     {
-      "epoch": 3.0459584544182543,
-      "grad_norm": 0.6148022413253784,
-      "learning_rate": 0.0004177696368925762,
-      "loss": 3.4522,
+      "epoch": 3.0512129380053907,
+      "grad_norm": 0.5706090927124023,
+      "learning_rate": 0.0004174484619535887,
+      "loss": 3.4384,
       "step": 28300
     },
     {
-      "epoch": 3.0513400064578624,
-      "grad_norm": 0.6846833825111389,
-      "learning_rate": 0.0004174463958625148,
-      "loss": 3.4641,
+      "epoch": 3.056603773584906,
+      "grad_norm": 0.592136561870575,
+      "learning_rate": 0.00041712466270912033,
+      "loss": 3.4125,
       "step": 28350
     },
     {
-      "epoch": 3.0567215584974705,
-      "grad_norm": 0.5982133746147156,
-      "learning_rate": 0.0004171231548324533,
-      "loss": 3.421,
+      "epoch": 3.0619946091644206,
+      "grad_norm": 0.5772967338562012,
+      "learning_rate": 0.0004168008634646519,
+      "loss": 3.4413,
       "step": 28400
     },
     {
-      "epoch": 3.062103110537079,
-      "grad_norm": 0.5601935982704163,
-      "learning_rate": 0.00041679991380239197,
-      "loss": 3.4459,
+      "epoch": 3.0673854447439353,
+      "grad_norm": 0.6145889163017273,
+      "learning_rate": 0.0004164770642201835,
+      "loss": 3.4119,
       "step": 28450
     },
     {
-      "epoch": 3.067484662576687,
-      "grad_norm": 0.5927339196205139,
-      "learning_rate": 0.00041647667277233056,
-      "loss": 3.445,
+      "epoch": 3.07277628032345,
+      "grad_norm": 0.6122552156448364,
+      "learning_rate": 0.00041615326497571503,
+      "loss": 3.4363,
       "step": 28500
     },
     {
-      "epoch": 3.0728662146162953,
-      "grad_norm": 0.5949593782424927,
-      "learning_rate": 0.0004161534317422691,
-      "loss": 3.4354,
+      "epoch": 3.078167115902965,
+      "grad_norm": 0.6149709224700928,
+      "learning_rate": 0.0004158294657312466,
+      "loss": 3.4481,
       "step": 28550
     },
     {
-      "epoch": 3.0782477666559034,
-      "grad_norm": 0.6130343675613403,
-      "learning_rate": 0.0004158301907122077,
-      "loss": 3.4323,
+      "epoch": 3.08355795148248,
+      "grad_norm": 0.6702295541763306,
+      "learning_rate": 0.0004155056664867782,
+      "loss": 3.4345,
       "step": 28600
     },
     {
-      "epoch": 3.083629318695512,
-      "grad_norm": 0.6153140664100647,
-      "learning_rate": 0.00041550694968214624,
-      "loss": 3.4363,
+      "epoch": 3.0889487870619945,
+      "grad_norm": 0.6043539047241211,
+      "learning_rate": 0.0004151818672423097,
+      "loss": 3.4439,
       "step": 28650
     },
     {
-      "epoch": 3.08901087073512,
-      "grad_norm": 0.6141806840896606,
-      "learning_rate": 0.0004151837086520849,
-      "loss": 3.4455,
+      "epoch": 3.0943396226415096,
+      "grad_norm": 0.5806813836097717,
+      "learning_rate": 0.0004148580679978413,
+      "loss": 3.4421,
       "step": 28700
     },
     {
-      "epoch": 3.094392422774728,
-      "grad_norm": 0.5657984018325806,
-      "learning_rate": 0.0004148604676220235,
-      "loss": 3.466,
+      "epoch": 3.0997304582210243,
+      "grad_norm": 0.6109824776649475,
+      "learning_rate": 0.00041453426875337284,
+      "loss": 3.467,
       "step": 28750
     },
     {
-      "epoch": 3.0997739748143363,
-      "grad_norm": 0.5603064298629761,
-      "learning_rate": 0.000414537226591962,
-      "loss": 3.4407,
+      "epoch": 3.105121293800539,
+      "grad_norm": 0.6003828644752502,
+      "learning_rate": 0.00041421046950890445,
+      "loss": 3.4269,
       "step": 28800
     },
     {
-      "epoch": 3.105155526853945,
-      "grad_norm": 0.642330527305603,
-      "learning_rate": 0.0004142139855619006,
-      "loss": 3.4445,
+      "epoch": 3.1105121293800537,
+      "grad_norm": 0.6085829734802246,
+      "learning_rate": 0.000413886670264436,
+      "loss": 3.4366,
       "step": 28850
     },
     {
-      "epoch": 3.110537078893553,
-      "grad_norm": 0.5735533833503723,
-      "learning_rate": 0.0004138907445318392,
-      "loss": 3.4107,
+      "epoch": 3.115902964959569,
+      "grad_norm": 0.6458422541618347,
+      "learning_rate": 0.0004135628710199676,
+      "loss": 3.4318,
       "step": 28900
     },
     {
-      "epoch": 3.115918630933161,
-      "grad_norm": 0.588525652885437,
-      "learning_rate": 0.00041356750350177775,
-      "loss": 3.4443,
+      "epoch": 3.1212938005390836,
+      "grad_norm": 0.6379101276397705,
+      "learning_rate": 0.00041323907177549915,
+      "loss": 3.4625,
       "step": 28950
     },
     {
-      "epoch": 3.121300182972769,
-      "grad_norm": 0.6355871558189392,
-      "learning_rate": 0.0004132442624717164,
-      "loss": 3.449,
+      "epoch": 3.1266846361185983,
+      "grad_norm": 0.5903958678245544,
+      "learning_rate": 0.0004129152725310307,
+      "loss": 3.434,
       "step": 29000
     },
     {
-      "epoch": 3.121300182972769,
-      "eval_accuracy": 0.3712649619101004,
-      "eval_loss": 3.4927072525024414,
-      "eval_runtime": 184.1671,
-      "eval_samples_per_second": 97.797,
-      "eval_steps_per_second": 6.114,
+      "epoch": 3.1266846361185983,
+      "eval_accuracy": 0.3708619683491965,
+      "eval_loss": 3.4925551414489746,
+      "eval_runtime": 183.7537,
+      "eval_samples_per_second": 98.017,
+      "eval_steps_per_second": 6.128,
       "step": 29000
     },
     {
-      "epoch": 3.1266817350123777,
-      "grad_norm": 0.6104164719581604,
-      "learning_rate": 0.000412921021441655,
-      "loss": 3.4318,
+      "epoch": 3.1320754716981134,
+      "grad_norm": 0.6094850301742554,
+      "learning_rate": 0.0004125914732865623,
+      "loss": 3.4501,
       "step": 29050
     },
     {
-      "epoch": 3.132063287051986,
-      "grad_norm": 0.5897156596183777,
-      "learning_rate": 0.00041259778041159354,
-      "loss": 3.4409,
+      "epoch": 3.137466307277628,
+      "grad_norm": 0.5834479331970215,
+      "learning_rate": 0.00041226767404209386,
+      "loss": 3.4516,
       "step": 29100
     },
     {
-      "epoch": 3.137444839091594,
-      "grad_norm": 0.5589237809181213,
-      "learning_rate": 0.00041227453938153213,
-      "loss": 3.4443,
+      "epoch": 3.142857142857143,
+      "grad_norm": 0.6338595151901245,
+      "learning_rate": 0.00041194387479762546,
+      "loss": 3.444,
       "step": 29150
     },
     {
-      "epoch": 3.1428263911312024,
-      "grad_norm": 0.680891752243042,
-      "learning_rate": 0.00041195129835147067,
-      "loss": 3.4522,
+      "epoch": 3.1482479784366575,
+      "grad_norm": 0.5886125564575195,
+      "learning_rate": 0.000411620075553157,
+      "loss": 3.4538,
       "step": 29200
     },
     {
-      "epoch": 3.1482079431708105,
-      "grad_norm": 0.6567697525024414,
-      "learning_rate": 0.00041162805732140927,
-      "loss": 3.459,
+      "epoch": 3.1536388140161726,
+      "grad_norm": 0.5797938704490662,
+      "learning_rate": 0.0004112962763086886,
+      "loss": 3.4554,
       "step": 29250
     },
     {
-      "epoch": 3.1535894952104186,
-      "grad_norm": 0.6201472282409668,
-      "learning_rate": 0.0004113048162913479,
-      "loss": 3.436,
+      "epoch": 3.1590296495956873,
+      "grad_norm": 0.5943017601966858,
+      "learning_rate": 0.0004109724770642201,
+      "loss": 3.4576,
       "step": 29300
     },
     {
-      "epoch": 3.1589710472500268,
-      "grad_norm": 0.592890202999115,
-      "learning_rate": 0.00041098157526128646,
-      "loss": 3.4473,
+      "epoch": 3.164420485175202,
+      "grad_norm": 0.6805914044380188,
+      "learning_rate": 0.00041064867781975177,
+      "loss": 3.4509,
       "step": 29350
     },
     {
-      "epoch": 3.1643525992896353,
-      "grad_norm": 0.6381814479827881,
-      "learning_rate": 0.00041065833423122505,
-      "loss": 3.4491,
+      "epoch": 3.169811320754717,
+      "grad_norm": 0.6030518412590027,
+      "learning_rate": 0.00041032487857528327,
+      "loss": 3.4417,
       "step": 29400
     },
     {
-      "epoch": 3.1697341513292434,
-      "grad_norm": 0.6653668284416199,
-      "learning_rate": 0.00041033509320116365,
-      "loss": 3.4462,
+      "epoch": 3.175202156334232,
+      "grad_norm": 0.5681232810020447,
+      "learning_rate": 0.0004100010793308148,
+      "loss": 3.4272,
       "step": 29450
     },
     {
-      "epoch": 3.1751157033688515,
-      "grad_norm": 0.6219462752342224,
-      "learning_rate": 0.0004100118521711022,
-      "loss": 3.4496,
+      "epoch": 3.1805929919137466,
+      "grad_norm": 0.5617977976799011,
+      "learning_rate": 0.0004096772800863464,
+      "loss": 3.4525,
       "step": 29500
     },
     {
-      "epoch": 3.1804972554084596,
-      "grad_norm": 0.6632649898529053,
-      "learning_rate": 0.0004096886111410408,
-      "loss": 3.432,
+      "epoch": 3.1859838274932613,
+      "grad_norm": 0.6035775542259216,
+      "learning_rate": 0.000409353480841878,
+      "loss": 3.4388,
       "step": 29550
     },
     {
-      "epoch": 3.185878807448068,
-      "grad_norm": 0.6093735694885254,
-      "learning_rate": 0.00040936537011097943,
-      "loss": 3.4529,
+      "epoch": 3.1913746630727764,
+      "grad_norm": 0.5969619154930115,
+      "learning_rate": 0.0004090296815974096,
+      "loss": 3.4451,
       "step": 29600
     },
     {
-      "epoch": 3.1912603594876763,
-      "grad_norm": 0.6073389649391174,
-      "learning_rate": 0.00040904212908091797,
-      "loss": 3.4418,
+      "epoch": 3.196765498652291,
+      "grad_norm": 0.6183298826217651,
+      "learning_rate": 0.00040870588235294113,
+      "loss": 3.4624,
       "step": 29650
     },
     {
-      "epoch": 3.1966419115272844,
-      "grad_norm": 0.6150850653648376,
-      "learning_rate": 0.00040871888805085656,
-      "loss": 3.4603,
+      "epoch": 3.202156334231806,
+      "grad_norm": 0.6094223856925964,
+      "learning_rate": 0.00040838208310847273,
+      "loss": 3.4457,
       "step": 29700
     },
     {
-      "epoch": 3.2020234635668925,
-      "grad_norm": 0.5961862802505493,
-      "learning_rate": 0.0004083956470207951,
-      "loss": 3.4617,
+      "epoch": 3.207547169811321,
+      "grad_norm": 0.5874677300453186,
+      "learning_rate": 0.0004080582838640043,
+      "loss": 3.4551,
       "step": 29750
     },
     {
-      "epoch": 3.207405015606501,
-      "grad_norm": 0.5796571969985962,
-      "learning_rate": 0.0004080724059907337,
-      "loss": 3.443,
+      "epoch": 3.2129380053908356,
+      "grad_norm": 0.6006694436073303,
+      "learning_rate": 0.00040774096060442524,
+      "loss": 3.4634,
       "step": 29800
     },
     {
-      "epoch": 3.212786567646109,
-      "grad_norm": 0.6276682615280151,
-      "learning_rate": 0.00040774916496067235,
-      "loss": 3.4667,
+      "epoch": 3.2183288409703503,
+      "grad_norm": 0.5807769894599915,
+      "learning_rate": 0.0004074171613599568,
+      "loss": 3.4686,
       "step": 29850
     },
     {
-      "epoch": 3.2181681196857173,
-      "grad_norm": 0.5951356291770935,
-      "learning_rate": 0.0004074259239306109,
-      "loss": 3.4557,
+      "epoch": 3.223719676549865,
+      "grad_norm": 0.6024731397628784,
+      "learning_rate": 0.0004070933621154884,
+      "loss": 3.4362,
       "step": 29900
     },
     {
-      "epoch": 3.2235496717253254,
-      "grad_norm": 0.6082077622413635,
-      "learning_rate": 0.0004071026829005495,
-      "loss": 3.452,
+      "epoch": 3.22911051212938,
+      "grad_norm": 0.5823608040809631,
+      "learning_rate": 0.00040676956287101994,
+      "loss": 3.4381,
       "step": 29950
     },
     {
-      "epoch": 3.228931223764934,
-      "grad_norm": 0.6054561734199524,
-      "learning_rate": 0.0004067794418704881,
-      "loss": 3.4475,
+      "epoch": 3.234501347708895,
+      "grad_norm": 0.5897881984710693,
+      "learning_rate": 0.00040644576362655155,
+      "loss": 3.4626,
       "step": 30000
     },
     {
-      "epoch": 3.228931223764934,
-      "eval_accuracy": 0.37177812949924816,
-      "eval_loss": 3.4881415367126465,
-      "eval_runtime": 183.6772,
-      "eval_samples_per_second": 98.058,
-      "eval_steps_per_second": 6.13,
+      "epoch": 3.234501347708895,
+      "eval_accuracy": 0.372028900364324,
+      "eval_loss": 3.4859824180603027,
+      "eval_runtime": 183.8563,
+      "eval_samples_per_second": 97.962,
+      "eval_steps_per_second": 6.124,
       "step": 30000
     },
     {
-      "epoch": 3.234312775804542,
-      "grad_norm": 0.610119640827179,
-      "learning_rate": 0.0004064562008404266,
-      "loss": 3.4477,
+      "epoch": 3.2398921832884096,
+      "grad_norm": 0.625234842300415,
+      "learning_rate": 0.00040612196438208304,
+      "loss": 3.4502,
       "step": 30050
     },
     {
-      "epoch": 3.23969432784415,
-      "grad_norm": 0.6288905143737793,
-      "learning_rate": 0.0004061329598103652,
-      "loss": 3.4563,
+      "epoch": 3.2452830188679247,
+      "grad_norm": 0.5733723640441895,
+      "learning_rate": 0.00040579816513761465,
+      "loss": 3.4553,
       "step": 30100
     },
     {
-      "epoch": 3.2450758798837587,
-      "grad_norm": 0.6015350818634033,
-      "learning_rate": 0.00040580971878030386,
+      "epoch": 3.2506738544474394,
+      "grad_norm": 0.6071932911872864,
+      "learning_rate": 0.0004054743658931462,
       "loss": 3.462,
       "step": 30150
     },
     {
-      "epoch": 3.250457431923367,
-      "grad_norm": 0.5772379636764526,
-      "learning_rate": 0.0004054864777502424,
-      "loss": 3.463,
+      "epoch": 3.256064690026954,
+      "grad_norm": 0.6102959513664246,
+      "learning_rate": 0.00040515056664867775,
+      "loss": 3.4452,
       "step": 30200
     },
     {
-      "epoch": 3.255838983962975,
-      "grad_norm": 0.6234760880470276,
-      "learning_rate": 0.000405163236720181,
-      "loss": 3.4452,
+      "epoch": 3.2614555256064692,
+      "grad_norm": 0.6284629702568054,
+      "learning_rate": 0.00040482676740420935,
+      "loss": 3.4501,
       "step": 30250
     },
     {
-      "epoch": 3.261220536002583,
-      "grad_norm": 0.6109282374382019,
-      "learning_rate": 0.00040483999569011954,
-      "loss": 3.4702,
+      "epoch": 3.266846361185984,
+      "grad_norm": 0.6206251382827759,
+      "learning_rate": 0.0004045029681597409,
+      "loss": 3.4372,
       "step": 30300
     },
     {
-      "epoch": 3.2666020880421915,
-      "grad_norm": 0.635185956954956,
-      "learning_rate": 0.00040451675466005813,
-      "loss": 3.4587,
+      "epoch": 3.2722371967654986,
+      "grad_norm": 0.6287161707878113,
+      "learning_rate": 0.0004041791689152725,
+      "loss": 3.4516,
       "step": 30350
     },
     {
-      "epoch": 3.2719836400817996,
-      "grad_norm": 0.6079793572425842,
-      "learning_rate": 0.0004041935136299967,
-      "loss": 3.4737,
+      "epoch": 3.2776280323450133,
+      "grad_norm": 0.5700094103813171,
+      "learning_rate": 0.00040385536967080406,
+      "loss": 3.4621,
       "step": 30400
     },
     {
-      "epoch": 3.2773651921214078,
-      "grad_norm": 0.6278197765350342,
-      "learning_rate": 0.0004038702725999353,
-      "loss": 3.4601,
+      "epoch": 3.2830188679245285,
+      "grad_norm": 0.6157817244529724,
+      "learning_rate": 0.00040353157042633567,
+      "loss": 3.451,
       "step": 30450
     },
     {
-      "epoch": 3.282746744161016,
-      "grad_norm": 0.6328572034835815,
-      "learning_rate": 0.0004035470315698739,
-      "loss": 3.4594,
+      "epoch": 3.288409703504043,
+      "grad_norm": 0.6491467356681824,
+      "learning_rate": 0.0004032077711818672,
+      "loss": 3.4718,
       "step": 30500
     },
     {
-      "epoch": 3.2881282962006244,
-      "grad_norm": 0.6899449825286865,
-      "learning_rate": 0.0004032237905398125,
-      "loss": 3.4512,
+      "epoch": 3.293800539083558,
+      "grad_norm": 0.5887497663497925,
+      "learning_rate": 0.0004028839719373988,
+      "loss": 3.4614,
       "step": 30550
     },
     {
-      "epoch": 3.2935098482402325,
-      "grad_norm": 0.643539547920227,
-      "learning_rate": 0.00040290054950975105,
-      "loss": 3.4587,
+      "epoch": 3.2991913746630726,
+      "grad_norm": 0.5873361825942993,
+      "learning_rate": 0.00040256017269293037,
+      "loss": 3.4561,
       "step": 30600
     },
     {
-      "epoch": 3.2988914002798406,
-      "grad_norm": 0.5721180438995361,
-      "learning_rate": 0.00040257730847968965,
-      "loss": 3.4466,
+      "epoch": 3.3045822102425877,
+      "grad_norm": 0.6074446439743042,
+      "learning_rate": 0.00040223637344846187,
+      "loss": 3.4546,
       "step": 30650
     },
     {
-      "epoch": 3.304272952319449,
-      "grad_norm": 0.6796326637268066,
-      "learning_rate": 0.0004022540674496283,
-      "loss": 3.4683,
+      "epoch": 3.3099730458221024,
+      "grad_norm": 0.6311209201812744,
+      "learning_rate": 0.00040191257420399347,
+      "loss": 3.446,
       "step": 30700
     },
     {
-      "epoch": 3.3096545043590573,
-      "grad_norm": 0.570809006690979,
-      "learning_rate": 0.00040193082641956684,
-      "loss": 3.4426,
+      "epoch": 3.315363881401617,
+      "grad_norm": 0.5985173583030701,
+      "learning_rate": 0.000401588774959525,
+      "loss": 3.4513,
       "step": 30750
     },
     {
-      "epoch": 3.3150360563986654,
-      "grad_norm": 0.6097874641418457,
-      "learning_rate": 0.00040160758538950543,
-      "loss": 3.4536,
+      "epoch": 3.3207547169811322,
+      "grad_norm": 0.5867009162902832,
+      "learning_rate": 0.00040126497571505663,
+      "loss": 3.4544,
       "step": 30800
     },
     {
-      "epoch": 3.3204176084382735,
-      "grad_norm": 0.6322263479232788,
-      "learning_rate": 0.00040128434435944397,
-      "loss": 3.4515,
+      "epoch": 3.326145552560647,
+      "grad_norm": 0.5650242567062378,
+      "learning_rate": 0.0004009411764705882,
+      "loss": 3.4455,
       "step": 30850
     },
     {
-      "epoch": 3.3257991604778816,
-      "grad_norm": 0.5945123434066772,
-      "learning_rate": 0.00040096110332938257,
-      "loss": 3.4598,
+      "epoch": 3.3315363881401616,
+      "grad_norm": 0.6401687264442444,
+      "learning_rate": 0.0004006173772261198,
+      "loss": 3.4682,
       "step": 30900
     },
     {
-      "epoch": 3.33118071251749,
-      "grad_norm": 0.5576284527778625,
-      "learning_rate": 0.00040063786229932116,
-      "loss": 3.4422,
+      "epoch": 3.3369272237196768,
+      "grad_norm": 0.6049325466156006,
+      "learning_rate": 0.00040029357798165133,
+      "loss": 3.4527,
       "step": 30950
     },
     {
-      "epoch": 3.3365622645570983,
-      "grad_norm": 0.6041018962860107,
-      "learning_rate": 0.00040031462126925975,
-      "loss": 3.4428,
+      "epoch": 3.3423180592991915,
+      "grad_norm": 0.5639699697494507,
+      "learning_rate": 0.00039996977873718294,
+      "loss": 3.4682,
       "step": 31000
     },
     {
-      "epoch": 3.3365622645570983,
-      "eval_accuracy": 0.3725362006975298,
-      "eval_loss": 3.48136568069458,
-      "eval_runtime": 184.0483,
-      "eval_samples_per_second": 97.86,
-      "eval_steps_per_second": 6.118,
+      "epoch": 3.3423180592991915,
+      "eval_accuracy": 0.3728585738156754,
+      "eval_loss": 3.4800660610198975,
+      "eval_runtime": 183.7942,
+      "eval_samples_per_second": 97.995,
+      "eval_steps_per_second": 6.126,
       "step": 31000
     },
     {
-      "epoch": 3.3419438165967064,
-      "grad_norm": 0.6239970326423645,
-      "learning_rate": 0.00039999138023919835,
-      "loss": 3.4362,
+      "epoch": 3.347708894878706,
+      "grad_norm": 0.5923236608505249,
+      "learning_rate": 0.0003996459794927145,
+      "loss": 3.444,
       "step": 31050
     },
     {
-      "epoch": 3.347325368636315,
-      "grad_norm": 0.6585947275161743,
-      "learning_rate": 0.00039966813920913694,
-      "loss": 3.4509,
+      "epoch": 3.353099730458221,
+      "grad_norm": 0.5859229564666748,
+      "learning_rate": 0.00039932218024824604,
+      "loss": 3.4652,
       "step": 31100
     },
     {
-      "epoch": 3.352706920675923,
-      "grad_norm": 0.6816777586936951,
-      "learning_rate": 0.0003993448981790755,
-      "loss": 3.4429,
+      "epoch": 3.358490566037736,
+      "grad_norm": 0.6009340286254883,
+      "learning_rate": 0.00039899838100377764,
+      "loss": 3.4424,
       "step": 31150
     },
     {
-      "epoch": 3.358088472715531,
-      "grad_norm": 0.632325291633606,
-      "learning_rate": 0.0003990216571490141,
-      "loss": 3.4549,
+      "epoch": 3.3638814016172507,
+      "grad_norm": 0.5966060757637024,
+      "learning_rate": 0.0003986745817593092,
+      "loss": 3.4464,
       "step": 31200
     },
     {
-      "epoch": 3.3634700247551392,
-      "grad_norm": 0.5963637828826904,
-      "learning_rate": 0.0003986984161189526,
-      "loss": 3.4615,
+      "epoch": 3.3692722371967654,
+      "grad_norm": 0.6093798875808716,
+      "learning_rate": 0.0003983507825148408,
+      "loss": 3.4616,
       "step": 31250
     },
     {
-      "epoch": 3.368851576794748,
-      "grad_norm": 0.6046538352966309,
-      "learning_rate": 0.0003983816399094925,
-      "loss": 3.4458,
+      "epoch": 3.37466307277628,
+      "grad_norm": 0.587139904499054,
+      "learning_rate": 0.00039802698327037235,
+      "loss": 3.4474,
       "step": 31300
     },
     {
-      "epoch": 3.374233128834356,
-      "grad_norm": 0.6418168544769287,
-      "learning_rate": 0.000398058398879431,
-      "loss": 3.4396,
+      "epoch": 3.3800539083557952,
+      "grad_norm": 0.6301162838935852,
+      "learning_rate": 0.00039770318402590396,
+      "loss": 3.4351,
       "step": 31350
     },
     {
-      "epoch": 3.379614680873964,
-      "grad_norm": 0.5885481834411621,
-      "learning_rate": 0.00039773515784936967,
-      "loss": 3.4549,
+      "epoch": 3.38544474393531,
+      "grad_norm": 0.6002510190010071,
+      "learning_rate": 0.00039737938478143545,
+      "loss": 3.4439,
       "step": 31400
     },
     {
-      "epoch": 3.384996232913572,
-      "grad_norm": 0.7410529255867004,
-      "learning_rate": 0.00039741191681930826,
-      "loss": 3.4509,
+      "epoch": 3.3908355795148246,
+      "grad_norm": 0.5881652235984802,
+      "learning_rate": 0.000397055585536967,
+      "loss": 3.4618,
       "step": 31450
     },
     {
-      "epoch": 3.3903777849531807,
-      "grad_norm": 0.6753930449485779,
-      "learning_rate": 0.0003970886757892468,
-      "loss": 3.446,
+      "epoch": 3.3962264150943398,
+      "grad_norm": 0.584099292755127,
+      "learning_rate": 0.0003967317862924986,
+      "loss": 3.4633,
       "step": 31500
     },
     {
-      "epoch": 3.3957593369927888,
-      "grad_norm": 0.6014970541000366,
-      "learning_rate": 0.0003967718995797866,
-      "loss": 3.4746,
+      "epoch": 3.4016172506738545,
+      "grad_norm": 0.6634021997451782,
+      "learning_rate": 0.00039640798704803016,
+      "loss": 3.4486,
       "step": 31550
     },
     {
-      "epoch": 3.401140889032397,
-      "grad_norm": 0.6676061749458313,
-      "learning_rate": 0.0003964486585497252,
-      "loss": 3.4692,
+      "epoch": 3.407008086253369,
+      "grad_norm": 0.6125302314758301,
+      "learning_rate": 0.00039608418780356176,
+      "loss": 3.4371,
       "step": 31600
     },
     {
-      "epoch": 3.4065224410720054,
-      "grad_norm": 0.6010334491729736,
-      "learning_rate": 0.0003961254175196638,
-      "loss": 3.4396,
+      "epoch": 3.4123989218328843,
+      "grad_norm": 0.6418920159339905,
+      "learning_rate": 0.0003957603885590933,
+      "loss": 3.4565,
       "step": 31650
     },
     {
-      "epoch": 3.4119039931116135,
-      "grad_norm": 0.588824987411499,
-      "learning_rate": 0.00039580217648960234,
-      "loss": 3.4295,
+      "epoch": 3.417789757412399,
+      "grad_norm": 0.6478168964385986,
+      "learning_rate": 0.0003954365893146249,
+      "loss": 3.462,
       "step": 31700
     },
     {
-      "epoch": 3.4172855451512216,
-      "grad_norm": 0.6397528648376465,
-      "learning_rate": 0.000395478935459541,
-      "loss": 3.4499,
+      "epoch": 3.4231805929919137,
+      "grad_norm": 0.5884379148483276,
+      "learning_rate": 0.00039511279007015647,
+      "loss": 3.4599,
       "step": 31750
     },
     {
-      "epoch": 3.4226670971908297,
-      "grad_norm": 0.608925998210907,
-      "learning_rate": 0.0003951556944294796,
-      "loss": 3.4489,
+      "epoch": 3.4285714285714284,
+      "grad_norm": 0.635249137878418,
+      "learning_rate": 0.0003947954668105774,
+      "loss": 3.4457,
       "step": 31800
     },
     {
-      "epoch": 3.428048649230438,
-      "grad_norm": 0.6103501319885254,
-      "learning_rate": 0.0003948324533994181,
-      "loss": 3.4534,
+      "epoch": 3.4339622641509435,
+      "grad_norm": 0.5733765363693237,
+      "learning_rate": 0.00039447166756610897,
+      "loss": 3.4488,
       "step": 31850
     },
     {
-      "epoch": 3.4334302012700464,
-      "grad_norm": 0.619126558303833,
-      "learning_rate": 0.0003945092123693567,
-      "loss": 3.442,
+      "epoch": 3.439353099730458,
+      "grad_norm": 0.6273685097694397,
+      "learning_rate": 0.0003941478683216406,
+      "loss": 3.4605,
       "step": 31900
     },
     {
-      "epoch": 3.4388117533096545,
-      "grad_norm": 0.6572946906089783,
-      "learning_rate": 0.00039418597133929526,
-      "loss": 3.449,
+      "epoch": 3.444743935309973,
+      "grad_norm": 0.5997422933578491,
+      "learning_rate": 0.0003938240690771721,
+      "loss": 3.4442,
       "step": 31950
     },
     {
-      "epoch": 3.4441933053492626,
-      "grad_norm": 0.6258237361907959,
-      "learning_rate": 0.00039386273030923385,
-      "loss": 3.4368,
+      "epoch": 3.450134770889488,
+      "grad_norm": 0.5667937397956848,
+      "learning_rate": 0.00039350026983270373,
+      "loss": 3.4562,
       "step": 32000
     },
     {
-      "epoch": 3.4441933053492626,
-      "eval_accuracy": 0.3733878220322197,
-      "eval_loss": 3.4752941131591797,
-      "eval_runtime": 184.2182,
-      "eval_samples_per_second": 97.77,
-      "eval_steps_per_second": 6.112,
+      "epoch": 3.450134770889488,
+      "eval_accuracy": 0.37337087218172066,
+      "eval_loss": 3.4724786281585693,
+      "eval_runtime": 183.8966,
+      "eval_samples_per_second": 97.941,
+      "eval_steps_per_second": 6.123,
       "step": 32000
     },
     {
-      "epoch": 3.449574857388871,
-      "grad_norm": 0.6714693903923035,
-      "learning_rate": 0.0003935394892791725,
-      "loss": 3.4483,
+      "epoch": 3.4555256064690028,
+      "grad_norm": 0.5887783765792847,
+      "learning_rate": 0.00039317647058823523,
+      "loss": 3.4795,
       "step": 32050
     },
     {
-      "epoch": 3.4549564094284793,
-      "grad_norm": 0.6188742518424988,
-      "learning_rate": 0.00039321624824911104,
-      "loss": 3.4352,
+      "epoch": 3.4609164420485174,
+      "grad_norm": 0.6130501627922058,
+      "learning_rate": 0.00039285267134376683,
+      "loss": 3.4697,
       "step": 32100
     },
     {
-      "epoch": 3.4603379614680874,
-      "grad_norm": 0.5907621383666992,
-      "learning_rate": 0.00039289300721904964,
-      "loss": 3.4593,
+      "epoch": 3.466307277628032,
+      "grad_norm": 0.6132586002349854,
+      "learning_rate": 0.0003925288720992984,
+      "loss": 3.4405,
       "step": 32150
     },
     {
-      "epoch": 3.4657195135076955,
-      "grad_norm": 0.5836877822875977,
-      "learning_rate": 0.00039256976618898823,
-      "loss": 3.4585,
+      "epoch": 3.4716981132075473,
+      "grad_norm": 0.5899814963340759,
+      "learning_rate": 0.00039220507285482993,
+      "loss": 3.4555,
       "step": 32200
     },
     {
-      "epoch": 3.471101065547304,
-      "grad_norm": 0.6085897088050842,
-      "learning_rate": 0.0003922465251589268,
-      "loss": 3.4533,
+      "epoch": 3.477088948787062,
+      "grad_norm": 0.5846422910690308,
+      "learning_rate": 0.00039188127361036154,
+      "loss": 3.4453,
       "step": 32250
     },
     {
-      "epoch": 3.476482617586912,
-      "grad_norm": 0.5915929079055786,
-      "learning_rate": 0.0003919232841288654,
-      "loss": 3.4515,
+      "epoch": 3.4824797843665767,
+      "grad_norm": 0.6742145419120789,
+      "learning_rate": 0.0003915574743658931,
+      "loss": 3.449,
       "step": 32300
     },
     {
-      "epoch": 3.4818641696265202,
-      "grad_norm": 0.6591131687164307,
-      "learning_rate": 0.000391600043098804,
-      "loss": 3.434,
+      "epoch": 3.487870619946092,
+      "grad_norm": 0.6021944284439087,
+      "learning_rate": 0.0003912336751214247,
+      "loss": 3.442,
       "step": 32350
     },
     {
-      "epoch": 3.4872457216661283,
-      "grad_norm": 0.6426652669906616,
-      "learning_rate": 0.00039127680206874256,
-      "loss": 3.4555,
+      "epoch": 3.4932614555256065,
+      "grad_norm": 0.6250495314598083,
+      "learning_rate": 0.00039090987587695624,
+      "loss": 3.4462,
       "step": 32400
     },
     {
-      "epoch": 3.492627273705737,
-      "grad_norm": 0.6139284372329712,
-      "learning_rate": 0.00039095356103868115,
-      "loss": 3.4575,
+      "epoch": 3.498652291105121,
+      "grad_norm": 0.5637419819831848,
+      "learning_rate": 0.00039058607663248785,
+      "loss": 3.4564,
       "step": 32450
     },
     {
-      "epoch": 3.498008825745345,
-      "grad_norm": 0.6101264357566833,
-      "learning_rate": 0.0003906303200086197,
-      "loss": 3.4436,
+      "epoch": 3.5040431266846364,
+      "grad_norm": 0.6372784376144409,
+      "learning_rate": 0.0003902622773880194,
+      "loss": 3.4624,
       "step": 32500
     },
     {
-      "epoch": 3.503390377784953,
-      "grad_norm": 0.6180574297904968,
-      "learning_rate": 0.0003903070789785583,
-      "loss": 3.447,
+      "epoch": 3.509433962264151,
+      "grad_norm": 0.6097209453582764,
+      "learning_rate": 0.000389938478143551,
+      "loss": 3.4625,
       "step": 32550
     },
     {
-      "epoch": 3.5087719298245617,
-      "grad_norm": 0.6128636002540588,
-      "learning_rate": 0.00038998383794849694,
-      "loss": 3.4593,
+      "epoch": 3.5148247978436657,
+      "grad_norm": 0.6254315972328186,
+      "learning_rate": 0.00038961467889908255,
+      "loss": 3.4506,
       "step": 32600
     },
     {
-      "epoch": 3.5141534818641698,
-      "grad_norm": 0.644425094127655,
-      "learning_rate": 0.0003896605969184355,
-      "loss": 3.4465,
+      "epoch": 3.5202156334231804,
+      "grad_norm": 0.595946729183197,
+      "learning_rate": 0.00038929087965461405,
+      "loss": 3.4544,
       "step": 32650
     },
     {
-      "epoch": 3.519535033903778,
-      "grad_norm": 0.6270672678947449,
-      "learning_rate": 0.00038933735588837407,
-      "loss": 3.4493,
+      "epoch": 3.525606469002695,
+      "grad_norm": 0.5904536843299866,
+      "learning_rate": 0.00038896708041014566,
+      "loss": 3.471,
       "step": 32700
     },
     {
-      "epoch": 3.524916585943386,
-      "grad_norm": 0.6661702394485474,
-      "learning_rate": 0.00038901411485831267,
-      "loss": 3.4555,
+      "epoch": 3.5309973045822103,
+      "grad_norm": 0.6004275679588318,
+      "learning_rate": 0.0003886432811656772,
+      "loss": 3.4445,
       "step": 32750
     },
     {
-      "epoch": 3.530298137982994,
-      "grad_norm": 0.6290133595466614,
-      "learning_rate": 0.0003886908738282512,
-      "loss": 3.4603,
+      "epoch": 3.536388140161725,
+      "grad_norm": 0.6042011976242065,
+      "learning_rate": 0.0003883194819212088,
+      "loss": 3.4722,
       "step": 32800
     },
     {
-      "epoch": 3.5356796900226026,
-      "grad_norm": 0.6530129313468933,
-      "learning_rate": 0.0003883676327981898,
-      "loss": 3.4368,
+      "epoch": 3.5417789757412397,
+      "grad_norm": 0.5697953701019287,
+      "learning_rate": 0.00038799568267674036,
+      "loss": 3.4711,
       "step": 32850
     },
     {
-      "epoch": 3.5410612420622107,
-      "grad_norm": 0.6154062151908875,
-      "learning_rate": 0.00038804439176812845,
-      "loss": 3.4257,
+      "epoch": 3.547169811320755,
+      "grad_norm": 0.5944839119911194,
+      "learning_rate": 0.00038767188343227197,
+      "loss": 3.4659,
       "step": 32900
     },
     {
-      "epoch": 3.546442794101819,
-      "grad_norm": 0.5952048897743225,
-      "learning_rate": 0.000387721150738067,
-      "loss": 3.4476,
+      "epoch": 3.5525606469002695,
+      "grad_norm": 0.6435691714286804,
+      "learning_rate": 0.0003873480841878035,
+      "loss": 3.4531,
       "step": 32950
     },
     {
-      "epoch": 3.5518243461414274,
-      "grad_norm": 0.607017457485199,
-      "learning_rate": 0.0003873979097080056,
-      "loss": 3.4549,
+      "epoch": 3.557951482479784,
+      "grad_norm": 0.6083458065986633,
+      "learning_rate": 0.0003870242849433351,
+      "loss": 3.4489,
       "step": 33000
     },
     {
-      "epoch": 3.5518243461414274,
-      "eval_accuracy": 0.3740801582333736,
-      "eval_loss": 3.467834234237671,
-      "eval_runtime": 183.7772,
-      "eval_samples_per_second": 98.005,
-      "eval_steps_per_second": 6.127,
+      "epoch": 3.557951482479784,
+      "eval_accuracy": 0.3737929886508799,
+      "eval_loss": 3.4655747413635254,
+      "eval_runtime": 184.0837,
+      "eval_samples_per_second": 97.841,
+      "eval_steps_per_second": 6.117,
       "step": 33000
     },
     {
-      "epoch": 3.5572058981810355,
-      "grad_norm": 0.6238864064216614,
-      "learning_rate": 0.0003870746686779441,
-      "loss": 3.4765,
+      "epoch": 3.5633423180592994,
+      "grad_norm": 0.6135107278823853,
+      "learning_rate": 0.00038670048569886667,
+      "loss": 3.4598,
       "step": 33050
     },
     {
-      "epoch": 3.5625874502206436,
-      "grad_norm": 0.6304193735122681,
-      "learning_rate": 0.0003867514276478827,
-      "loss": 3.4581,
+      "epoch": 3.568733153638814,
+      "grad_norm": 0.6154115200042725,
+      "learning_rate": 0.0003863766864543982,
+      "loss": 3.462,
       "step": 33100
     },
     {
-      "epoch": 3.5679690022602517,
-      "grad_norm": 0.6388722062110901,
-      "learning_rate": 0.0003864281866178213,
-      "loss": 3.4424,
+      "epoch": 3.5741239892183287,
+      "grad_norm": 0.6276868581771851,
+      "learning_rate": 0.00038605288720992983,
+      "loss": 3.4638,
       "step": 33150
     },
     {
-      "epoch": 3.57335055429986,
-      "grad_norm": 0.644290030002594,
-      "learning_rate": 0.0003861049455877599,
-      "loss": 3.4588,
+      "epoch": 3.579514824797844,
+      "grad_norm": 0.5951337814331055,
+      "learning_rate": 0.0003857290879654614,
+      "loss": 3.46,
       "step": 33200
     },
     {
-      "epoch": 3.5787321063394684,
-      "grad_norm": 0.6192201375961304,
-      "learning_rate": 0.0003857817045576985,
-      "loss": 3.4581,
+      "epoch": 3.5849056603773586,
+      "grad_norm": 0.5743973851203918,
+      "learning_rate": 0.000385405288720993,
+      "loss": 3.4431,
       "step": 33250
     },
     {
-      "epoch": 3.5841136583790765,
-      "grad_norm": 0.7274168133735657,
-      "learning_rate": 0.0003854584635276371,
-      "loss": 3.4465,
+      "epoch": 3.5902964959568733,
+      "grad_norm": 0.6152252554893494,
+      "learning_rate": 0.00038508148947652453,
+      "loss": 3.4705,
       "step": 33300
     },
     {
-      "epoch": 3.5894952104186846,
-      "grad_norm": 0.6687127351760864,
-      "learning_rate": 0.00038513522249757564,
-      "loss": 3.452,
+      "epoch": 3.595687331536388,
+      "grad_norm": 0.639670193195343,
+      "learning_rate": 0.00038475769023205614,
+      "loss": 3.4578,
       "step": 33350
     },
     {
-      "epoch": 3.594876762458293,
-      "grad_norm": 0.6331846117973328,
-      "learning_rate": 0.00038481198146751423,
-      "loss": 3.4341,
+      "epoch": 3.601078167115903,
+      "grad_norm": 0.6149652004241943,
+      "learning_rate": 0.00038443389098758763,
+      "loss": 3.4546,
       "step": 33400
     },
     {
-      "epoch": 3.6002583144979012,
-      "grad_norm": 0.5981317162513733,
-      "learning_rate": 0.0003844887404374529,
-      "loss": 3.4398,
+      "epoch": 3.606469002695418,
+      "grad_norm": 0.6508069634437561,
+      "learning_rate": 0.00038411009174311924,
+      "loss": 3.4698,
       "step": 33450
     },
     {
-      "epoch": 3.6056398665375093,
-      "grad_norm": 0.6762210726737976,
-      "learning_rate": 0.0003841654994073914,
-      "loss": 3.4495,
+      "epoch": 3.6118598382749325,
+      "grad_norm": 0.5838162302970886,
+      "learning_rate": 0.0003837862924986508,
+      "loss": 3.4474,
       "step": 33500
     },
     {
-      "epoch": 3.611021418577118,
-      "grad_norm": 0.6356139183044434,
-      "learning_rate": 0.00038384225837733,
-      "loss": 3.4497,
+      "epoch": 3.617250673854447,
+      "grad_norm": 0.6133880019187927,
+      "learning_rate": 0.00038346249325418234,
+      "loss": 3.4456,
       "step": 33550
     },
     {
-      "epoch": 3.616402970616726,
-      "grad_norm": 0.6036195158958435,
-      "learning_rate": 0.00038351901734726856,
-      "loss": 3.4357,
+      "epoch": 3.6226415094339623,
+      "grad_norm": 0.6268334984779358,
+      "learning_rate": 0.00038313869400971395,
+      "loss": 3.4414,
       "step": 33600
     },
     {
-      "epoch": 3.621784522656334,
-      "grad_norm": 0.6225637793540955,
-      "learning_rate": 0.00038319577631720715,
-      "loss": 3.4414,
+      "epoch": 3.628032345013477,
+      "grad_norm": 0.6237933039665222,
+      "learning_rate": 0.0003828148947652455,
+      "loss": 3.458,
       "step": 33650
     },
     {
-      "epoch": 3.627166074695942,
-      "grad_norm": 0.6315037608146667,
-      "learning_rate": 0.00038287253528714575,
-      "loss": 3.4592,
+      "epoch": 3.6334231805929917,
+      "grad_norm": 0.618687093257904,
+      "learning_rate": 0.0003824910955207771,
+      "loss": 3.4619,
       "step": 33700
     },
     {
-      "epoch": 3.6325476267355503,
-      "grad_norm": 0.6238998770713806,
-      "learning_rate": 0.00038255575907768555,
-      "loss": 3.4327,
+      "epoch": 3.638814016172507,
+      "grad_norm": 0.6351138949394226,
+      "learning_rate": 0.00038216729627630865,
+      "loss": 3.4415,
       "step": 33750
     },
     {
-      "epoch": 3.637929178775159,
-      "grad_norm": 0.676425039768219,
-      "learning_rate": 0.0003822325180476241,
-      "loss": 3.4338,
+      "epoch": 3.6442048517520216,
+      "grad_norm": 0.617386519908905,
+      "learning_rate": 0.00038184349703184026,
+      "loss": 3.4447,
       "step": 33800
     },
     {
-      "epoch": 3.643310730814767,
-      "grad_norm": 0.6187206506729126,
-      "learning_rate": 0.00038190927701756274,
-      "loss": 3.4472,
+      "epoch": 3.6495956873315363,
+      "grad_norm": 0.5832499861717224,
+      "learning_rate": 0.00038152617377226115,
+      "loss": 3.4326,
       "step": 33850
     },
     {
-      "epoch": 3.648692282854375,
-      "grad_norm": 0.5785842537879944,
-      "learning_rate": 0.00038158603598750134,
-      "loss": 3.4385,
+      "epoch": 3.6549865229110514,
+      "grad_norm": 0.6513313055038452,
+      "learning_rate": 0.00038120237452779276,
+      "loss": 3.423,
       "step": 33900
     },
     {
-      "epoch": 3.6540738348939836,
-      "grad_norm": 0.635712742805481,
-      "learning_rate": 0.0003812627949574399,
-      "loss": 3.4416,
+      "epoch": 3.660377358490566,
+      "grad_norm": 0.6100399494171143,
+      "learning_rate": 0.0003808785752833243,
+      "loss": 3.4499,
       "step": 33950
     },
     {
-      "epoch": 3.6594553869335917,
-      "grad_norm": 0.6430516242980957,
-      "learning_rate": 0.0003809395539273785,
-      "loss": 3.4528,
+      "epoch": 3.665768194070081,
+      "grad_norm": 0.6160393953323364,
+      "learning_rate": 0.0003805547760388559,
+      "loss": 3.4475,
       "step": 34000
     },
     {
-      "epoch": 3.6594553869335917,
-      "eval_accuracy": 0.37433277619754224,
-      "eval_loss": 3.460512399673462,
-      "eval_runtime": 183.9245,
-      "eval_samples_per_second": 97.926,
-      "eval_steps_per_second": 6.122,
+      "epoch": 3.665768194070081,
+      "eval_accuracy": 0.3747759713269375,
+      "eval_loss": 3.46094012260437,
+      "eval_runtime": 183.7639,
+      "eval_samples_per_second": 98.012,
+      "eval_steps_per_second": 6.127,
       "step": 34000
     },
     {
-      "epoch": 3.6648369389732,
-      "grad_norm": 0.5994505286216736,
-      "learning_rate": 0.00038061631289731707,
-      "loss": 3.4486,
+      "epoch": 3.671159029649596,
+      "grad_norm": 0.6878004670143127,
+      "learning_rate": 0.0003802309767943874,
+      "loss": 3.4495,
       "step": 34050
     },
     {
-      "epoch": 3.670218491012808,
-      "grad_norm": 0.6591764092445374,
-      "learning_rate": 0.00038029307186725566,
-      "loss": 3.4429,
+      "epoch": 3.6765498652291106,
+      "grad_norm": 0.634497880935669,
+      "learning_rate": 0.000379907177549919,
+      "loss": 3.4485,
       "step": 34100
     },
     {
-      "epoch": 3.675600043052416,
-      "grad_norm": 0.6138482689857483,
-      "learning_rate": 0.00037996983083719426,
-      "loss": 3.4531,
+      "epoch": 3.6819407008086253,
+      "grad_norm": 0.6088259220123291,
+      "learning_rate": 0.00037958337830545057,
+      "loss": 3.445,
       "step": 34150
     },
     {
-      "epoch": 3.6809815950920246,
-      "grad_norm": 0.6182659268379211,
-      "learning_rate": 0.00037964658980713285,
-      "loss": 3.4526,
+      "epoch": 3.68733153638814,
+      "grad_norm": 0.7003622055053711,
+      "learning_rate": 0.00037925957906098217,
+      "loss": 3.4542,
       "step": 34200
     },
     {
-      "epoch": 3.6863631471316327,
-      "grad_norm": 0.5865574479103088,
-      "learning_rate": 0.0003793233487770714,
-      "loss": 3.45,
+      "epoch": 3.6927223719676547,
+      "grad_norm": 0.6519274115562439,
+      "learning_rate": 0.0003789357798165137,
+      "loss": 3.4431,
       "step": 34250
     },
     {
-      "epoch": 3.691744699171241,
-      "grad_norm": 0.6625626087188721,
-      "learning_rate": 0.00037900010774701,
-      "loss": 3.4502,
+      "epoch": 3.69811320754717,
+      "grad_norm": 0.6134795546531677,
+      "learning_rate": 0.00037861198057204527,
+      "loss": 3.4494,
       "step": 34300
     },
     {
-      "epoch": 3.6971262512108494,
-      "grad_norm": 0.652292788028717,
-      "learning_rate": 0.00037867686671694853,
-      "loss": 3.4453,
+      "epoch": 3.7035040431266846,
+      "grad_norm": 0.6110228896141052,
+      "learning_rate": 0.0003782881813275769,
+      "loss": 3.4472,
       "step": 34350
     },
     {
-      "epoch": 3.7025078032504575,
-      "grad_norm": 0.6206428408622742,
-      "learning_rate": 0.0003783536256868872,
-      "loss": 3.4455,
+      "epoch": 3.7088948787061993,
+      "grad_norm": 0.5986830592155457,
+      "learning_rate": 0.0003779643820831084,
+      "loss": 3.4567,
       "step": 34400
     },
     {
-      "epoch": 3.7078893552900656,
-      "grad_norm": 0.6362066268920898,
-      "learning_rate": 0.00037803038465682577,
-      "loss": 3.4452,
+      "epoch": 3.7142857142857144,
+      "grad_norm": 0.6216215491294861,
+      "learning_rate": 0.00037764058283864003,
+      "loss": 3.425,
       "step": 34450
     },
     {
-      "epoch": 3.713270907329674,
-      "grad_norm": 0.6381617784500122,
-      "learning_rate": 0.0003777071436267643,
-      "loss": 3.4306,
+      "epoch": 3.719676549865229,
+      "grad_norm": 0.6370998024940491,
+      "learning_rate": 0.0003773167835941716,
+      "loss": 3.4673,
       "step": 34500
     },
     {
-      "epoch": 3.7186524593692822,
-      "grad_norm": 0.6452015042304993,
-      "learning_rate": 0.0003773839025967029,
-      "loss": 3.4485,
+      "epoch": 3.725067385444744,
+      "grad_norm": 0.6130332350730896,
+      "learning_rate": 0.0003769929843497032,
+      "loss": 3.4458,
       "step": 34550
     },
     {
-      "epoch": 3.7240340114088903,
-      "grad_norm": 0.6497765779495239,
-      "learning_rate": 0.0003770606615666415,
-      "loss": 3.4464,
+      "epoch": 3.730458221024259,
+      "grad_norm": 0.5908277630805969,
+      "learning_rate": 0.00037666918510523474,
+      "loss": 3.4493,
       "step": 34600
     },
     {
-      "epoch": 3.7294155634484984,
-      "grad_norm": 0.6384848952293396,
-      "learning_rate": 0.00037673742053658004,
-      "loss": 3.4458,
+      "epoch": 3.7358490566037736,
+      "grad_norm": 0.6192821264266968,
+      "learning_rate": 0.00037634538586076634,
+      "loss": 3.4547,
       "step": 34650
     },
     {
-      "epoch": 3.7347971154881066,
-      "grad_norm": 0.6216375231742859,
-      "learning_rate": 0.0003764141795065187,
-      "loss": 3.448,
+      "epoch": 3.7412398921832883,
+      "grad_norm": 0.6178766489028931,
+      "learning_rate": 0.00037602158661629784,
+      "loss": 3.4597,
       "step": 34700
     },
     {
-      "epoch": 3.740178667527715,
-      "grad_norm": 0.6330543160438538,
-      "learning_rate": 0.0003760909384764573,
-      "loss": 3.4331,
+      "epoch": 3.7466307277628035,
+      "grad_norm": 0.599069356918335,
+      "learning_rate": 0.0003756977873718294,
+      "loss": 3.4641,
       "step": 34750
     },
     {
-      "epoch": 3.745560219567323,
-      "grad_norm": 0.6534463763237,
-      "learning_rate": 0.0003757676974463958,
-      "loss": 3.4593,
+      "epoch": 3.752021563342318,
+      "grad_norm": 0.6087269186973572,
+      "learning_rate": 0.000375373988127361,
+      "loss": 3.4476,
       "step": 34800
     },
     {
-      "epoch": 3.7509417716069313,
-      "grad_norm": 0.6447902917861938,
-      "learning_rate": 0.0003754444564163344,
-      "loss": 3.451,
+      "epoch": 3.757412398921833,
+      "grad_norm": 0.6363670229911804,
+      "learning_rate": 0.00037505018888289254,
+      "loss": 3.4571,
       "step": 34850
     },
     {
-      "epoch": 3.75632332364654,
-      "grad_norm": 0.5719825029373169,
-      "learning_rate": 0.00037512121538627296,
-      "loss": 3.4552,
+      "epoch": 3.7628032345013476,
+      "grad_norm": 0.6591955423355103,
+      "learning_rate": 0.00037472638963842415,
+      "loss": 3.4474,
       "step": 34900
     },
     {
-      "epoch": 3.761704875686148,
-      "grad_norm": 0.7099481821060181,
-      "learning_rate": 0.00037479797435621155,
-      "loss": 3.4497,
+      "epoch": 3.7681940700808623,
+      "grad_norm": 0.6446316242218018,
+      "learning_rate": 0.0003744025903939557,
+      "loss": 3.44,
       "step": 34950
     },
     {
-      "epoch": 3.767086427725756,
-      "grad_norm": 0.5851026177406311,
-      "learning_rate": 0.0003744747333261502,
-      "loss": 3.4482,
+      "epoch": 3.7735849056603774,
+      "grad_norm": 0.6901982426643372,
+      "learning_rate": 0.0003740787911494873,
+      "loss": 3.4642,
       "step": 35000
     },
     {
-      "epoch": 3.767086427725756,
-      "eval_accuracy": 0.37522459910072353,
-      "eval_loss": 3.4531424045562744,
-      "eval_runtime": 183.9235,
-      "eval_samples_per_second": 97.927,
-      "eval_steps_per_second": 6.122,
+      "epoch": 3.7735849056603774,
+      "eval_accuracy": 0.3754713498089502,
+      "eval_loss": 3.4537289142608643,
+      "eval_runtime": 184.0286,
+      "eval_samples_per_second": 97.871,
+      "eval_steps_per_second": 6.119,
       "step": 35000
     },
     {
-      "epoch": 3.772467979765364,
-      "grad_norm": 0.5889676213264465,
-      "learning_rate": 0.00037415149229608874,
-      "loss": 3.4361,
+      "epoch": 3.778975741239892,
+      "grad_norm": 0.6182703971862793,
+      "learning_rate": 0.00037375499190501885,
+      "loss": 3.4501,
       "step": 35050
     },
     {
-      "epoch": 3.7778495318049723,
-      "grad_norm": 0.6418570280075073,
-      "learning_rate": 0.00037382825126602734,
-      "loss": 3.4714,
+      "epoch": 3.784366576819407,
+      "grad_norm": 0.6139285564422607,
+      "learning_rate": 0.0003734311926605504,
+      "loss": 3.4526,
       "step": 35100
     },
     {
-      "epoch": 3.783231083844581,
-      "grad_norm": 0.6317884922027588,
-      "learning_rate": 0.00037350501023596593,
-      "loss": 3.4764,
+      "epoch": 3.789757412398922,
+      "grad_norm": 0.6648865938186646,
+      "learning_rate": 0.000373107393416082,
+      "loss": 3.4371,
       "step": 35150
     },
     {
-      "epoch": 3.788612635884189,
-      "grad_norm": 0.6682846546173096,
-      "learning_rate": 0.0003731817692059045,
-      "loss": 3.4493,
+      "epoch": 3.7951482479784366,
+      "grad_norm": 0.5710464119911194,
+      "learning_rate": 0.00037278359417161356,
+      "loss": 3.4416,
       "step": 35200
     },
     {
-      "epoch": 3.793994187923797,
-      "grad_norm": 0.5991795063018799,
-      "learning_rate": 0.0003728585281758431,
-      "loss": 3.4598,
+      "epoch": 3.8005390835579513,
+      "grad_norm": 0.6638919115066528,
+      "learning_rate": 0.00037245979492714517,
+      "loss": 3.4506,
       "step": 35250
     },
     {
-      "epoch": 3.7993757399634056,
-      "grad_norm": 0.591937780380249,
-      "learning_rate": 0.0003725352871457817,
-      "loss": 3.4302,
+      "epoch": 3.8059299191374665,
+      "grad_norm": 0.6188521981239319,
+      "learning_rate": 0.0003721359956826767,
+      "loss": 3.4423,
       "step": 35300
     },
     {
-      "epoch": 3.8047572920030137,
-      "grad_norm": 0.7012965083122253,
-      "learning_rate": 0.00037221204611572026,
-      "loss": 3.4373,
+      "epoch": 3.811320754716981,
+      "grad_norm": 0.6108767986297607,
+      "learning_rate": 0.0003718121964382083,
+      "loss": 3.463,
       "step": 35350
     },
     {
-      "epoch": 3.810138844042622,
-      "grad_norm": 0.6041591763496399,
-      "learning_rate": 0.00037188880508565885,
-      "loss": 3.44,
+      "epoch": 3.816711590296496,
+      "grad_norm": 0.6063377261161804,
+      "learning_rate": 0.0003714883971937398,
+      "loss": 3.4446,
       "step": 35400
     },
     {
-      "epoch": 3.8155203960822304,
-      "grad_norm": 0.6080672144889832,
-      "learning_rate": 0.0003715655640555974,
-      "loss": 3.4442,
+      "epoch": 3.822102425876011,
+      "grad_norm": 0.5719074010848999,
+      "learning_rate": 0.0003711645979492714,
+      "loss": 3.4502,
       "step": 35450
     },
     {
-      "epoch": 3.8209019481218385,
-      "grad_norm": 0.5807365775108337,
-      "learning_rate": 0.000371242323025536,
-      "loss": 3.4344,
+      "epoch": 3.8274932614555257,
+      "grad_norm": 0.6371691823005676,
+      "learning_rate": 0.00037084079870480297,
+      "loss": 3.4405,
       "step": 35500
     },
     {
-      "epoch": 3.8262835001614466,
-      "grad_norm": 0.6241209506988525,
-      "learning_rate": 0.00037091908199547464,
-      "loss": 3.4506,
+      "epoch": 3.8328840970350404,
+      "grad_norm": 0.6494392156600952,
+      "learning_rate": 0.0003705169994603345,
+      "loss": 3.4449,
       "step": 35550
     },
     {
-      "epoch": 3.8316650522010547,
-      "grad_norm": 0.6381786465644836,
-      "learning_rate": 0.0003705958409654132,
-      "loss": 3.4572,
+      "epoch": 3.838274932614555,
+      "grad_norm": 0.6301707029342651,
+      "learning_rate": 0.00037019320021586613,
+      "loss": 3.4547,
       "step": 35600
     },
     {
-      "epoch": 3.837046604240663,
-      "grad_norm": 0.6080597043037415,
-      "learning_rate": 0.00037027259993535177,
-      "loss": 3.4516,
+      "epoch": 3.8436657681940702,
+      "grad_norm": 0.6397441625595093,
+      "learning_rate": 0.0003698694009713977,
+      "loss": 3.4399,
       "step": 35650
     },
     {
-      "epoch": 3.8424281562802713,
-      "grad_norm": 0.6270143389701843,
-      "learning_rate": 0.0003699493589052903,
-      "loss": 3.4433,
+      "epoch": 3.849056603773585,
+      "grad_norm": 0.6234208941459656,
+      "learning_rate": 0.0003695456017269293,
+      "loss": 3.4411,
       "step": 35700
     },
     {
-      "epoch": 3.8478097083198795,
-      "grad_norm": 0.6298555135726929,
-      "learning_rate": 0.0003696261178752289,
-      "loss": 3.4448,
+      "epoch": 3.8544474393530996,
+      "grad_norm": 0.603117823600769,
+      "learning_rate": 0.00036922180248246083,
+      "loss": 3.4506,
       "step": 35750
     },
     {
-      "epoch": 3.8531912603594876,
-      "grad_norm": 0.6021280288696289,
-      "learning_rate": 0.0003693093416657687,
-      "loss": 3.4381,
+      "epoch": 3.8598382749326143,
+      "grad_norm": 0.616421103477478,
+      "learning_rate": 0.00036889800323799244,
+      "loss": 3.4384,
       "step": 35800
     },
     {
-      "epoch": 3.858572812399096,
-      "grad_norm": 0.6799759268760681,
-      "learning_rate": 0.0003689861006357073,
-      "loss": 3.4548,
+      "epoch": 3.8652291105121295,
+      "grad_norm": 0.6407124400138855,
+      "learning_rate": 0.0003685806799784134,
+      "loss": 3.4472,
       "step": 35850
     },
     {
-      "epoch": 3.863954364438704,
-      "grad_norm": 0.6433078050613403,
-      "learning_rate": 0.00036866285960564596,
-      "loss": 3.4524,
+      "epoch": 3.870619946091644,
+      "grad_norm": 0.6404979825019836,
+      "learning_rate": 0.00036825688073394494,
+      "loss": 3.4602,
       "step": 35900
     },
     {
-      "epoch": 3.8693359164783123,
-      "grad_norm": 0.6047008633613586,
-      "learning_rate": 0.0003683396185755845,
-      "loss": 3.4313,
+      "epoch": 3.876010781671159,
+      "grad_norm": 0.6027001738548279,
+      "learning_rate": 0.0003679330814894765,
+      "loss": 3.4606,
       "step": 35950
     },
     {
-      "epoch": 3.8747174685179204,
-      "grad_norm": 0.6417443752288818,
-      "learning_rate": 0.0003680163775455231,
-      "loss": 3.4616,
+      "epoch": 3.881401617250674,
+      "grad_norm": 0.606300950050354,
+      "learning_rate": 0.0003676092822450081,
+      "loss": 3.4313,
       "step": 36000
     },
     {
-      "epoch": 3.8747174685179204,
-      "eval_accuracy": 0.3758969431705195,
-      "eval_loss": 3.4463951587677,
-      "eval_runtime": 184.1368,
-      "eval_samples_per_second": 97.813,
-      "eval_steps_per_second": 6.115,
+      "epoch": 3.881401617250674,
+      "eval_accuracy": 0.375837510040885,
+      "eval_loss": 3.4485042095184326,
+      "eval_runtime": 183.7341,
+      "eval_samples_per_second": 98.028,
+      "eval_steps_per_second": 6.128,
       "step": 36000
     },
     {
-      "epoch": 3.8800990205575285,
-      "grad_norm": 0.6225312948226929,
-      "learning_rate": 0.0003676931365154617,
-      "loss": 3.4509,
+      "epoch": 3.8867924528301887,
+      "grad_norm": 0.5973747968673706,
+      "learning_rate": 0.0003672854830005396,
+      "loss": 3.4663,
       "step": 36050
     },
     {
-      "epoch": 3.885480572597137,
-      "grad_norm": 0.6264508366584778,
-      "learning_rate": 0.0003673698954854002,
-      "loss": 3.4698,
+      "epoch": 3.8921832884097034,
+      "grad_norm": 0.622265636920929,
+      "learning_rate": 0.0003669616837560712,
+      "loss": 3.4324,
       "step": 36100
     },
     {
-      "epoch": 3.890862124636745,
-      "grad_norm": 0.6349591016769409,
-      "learning_rate": 0.0003670466544553388,
-      "loss": 3.4474,
+      "epoch": 3.8975741239892185,
+      "grad_norm": 0.6492241024971008,
+      "learning_rate": 0.00036663788451160275,
+      "loss": 3.4388,
       "step": 36150
     },
     {
-      "epoch": 3.8962436766763533,
-      "grad_norm": 0.6341753005981445,
-      "learning_rate": 0.00036672341342527747,
-      "loss": 3.4209,
+      "epoch": 3.9029649595687332,
+      "grad_norm": 0.5774116516113281,
+      "learning_rate": 0.00036631408526713435,
+      "loss": 3.4451,
       "step": 36200
     },
     {
-      "epoch": 3.901625228715962,
-      "grad_norm": 0.6438577771186829,
-      "learning_rate": 0.000366400172395216,
-      "loss": 3.4468,
+      "epoch": 3.908355795148248,
+      "grad_norm": 0.5866903066635132,
+      "learning_rate": 0.0003659902860226659,
+      "loss": 3.4514,
       "step": 36250
     },
     {
-      "epoch": 3.90700678075557,
-      "grad_norm": 0.6473516821861267,
-      "learning_rate": 0.0003660769313651546,
-      "loss": 3.4544,
+      "epoch": 3.913746630727763,
+      "grad_norm": 0.5962615609169006,
+      "learning_rate": 0.00036566648677819745,
+      "loss": 3.4538,
       "step": 36300
     },
     {
-      "epoch": 3.912388332795178,
-      "grad_norm": 0.6102363467216492,
-      "learning_rate": 0.00036575369033509315,
-      "loss": 3.4431,
+      "epoch": 3.9191374663072778,
+      "grad_norm": 0.5820170640945435,
+      "learning_rate": 0.00036534268753372906,
+      "loss": 3.465,
       "step": 36350
     },
     {
-      "epoch": 3.9177698848347866,
-      "grad_norm": 0.6443660259246826,
-      "learning_rate": 0.00036543044930503174,
-      "loss": 3.4383,
+      "epoch": 3.9245283018867925,
+      "grad_norm": 0.5907723903656006,
+      "learning_rate": 0.0003650188882892606,
+      "loss": 3.4339,
       "step": 36400
     },
     {
-      "epoch": 3.9231514368743947,
-      "grad_norm": 0.5729968547821045,
-      "learning_rate": 0.00036510720827497034,
-      "loss": 3.4486,
+      "epoch": 3.929919137466307,
+      "grad_norm": 0.6356750130653381,
+      "learning_rate": 0.0003646950890447922,
+      "loss": 3.4399,
       "step": 36450
     },
     {
-      "epoch": 3.928532988914003,
-      "grad_norm": 0.656282365322113,
-      "learning_rate": 0.00036478396724490893,
-      "loss": 3.4548,
+      "epoch": 3.935309973045822,
+      "grad_norm": 0.6064892411231995,
+      "learning_rate": 0.00036437128980032376,
+      "loss": 3.454,
       "step": 36500
     },
     {
-      "epoch": 3.933914540953611,
-      "grad_norm": 0.6700734496116638,
-      "learning_rate": 0.0003644607262148475,
-      "loss": 3.4411,
+      "epoch": 3.940700808625337,
+      "grad_norm": 0.5978058576583862,
+      "learning_rate": 0.00036404749055585537,
+      "loss": 3.4661,
       "step": 36550
     },
     {
-      "epoch": 3.939296092993219,
-      "grad_norm": 0.5883575081825256,
-      "learning_rate": 0.0003641374851847861,
-      "loss": 3.4402,
+      "epoch": 3.9460916442048517,
+      "grad_norm": 0.6196339130401611,
+      "learning_rate": 0.0003637236913113869,
+      "loss": 3.4193,
       "step": 36600
     },
     {
-      "epoch": 3.9446776450328276,
-      "grad_norm": 0.6195216774940491,
-      "learning_rate": 0.00036381424415472466,
-      "loss": 3.428,
+      "epoch": 3.9514824797843664,
+      "grad_norm": 0.6410916447639465,
+      "learning_rate": 0.0003633998920669185,
+      "loss": 3.4378,
       "step": 36650
     },
     {
-      "epoch": 3.9500591970724357,
-      "grad_norm": 0.6570115089416504,
-      "learning_rate": 0.00036349100312466325,
-      "loss": 3.4348,
+      "epoch": 3.9568733153638815,
+      "grad_norm": 0.6687703728675842,
+      "learning_rate": 0.0003630760928224501,
+      "loss": 3.4431,
       "step": 36700
     },
     {
-      "epoch": 3.955440749112044,
-      "grad_norm": 0.6123061180114746,
-      "learning_rate": 0.0003631677620946018,
-      "loss": 3.4504,
+      "epoch": 3.9622641509433962,
+      "grad_norm": 0.602182924747467,
+      "learning_rate": 0.00036275229357798157,
+      "loss": 3.4239,
       "step": 36750
     },
     {
-      "epoch": 3.9608223011516523,
-      "grad_norm": 0.6113377809524536,
-      "learning_rate": 0.00036284452106454044,
-      "loss": 3.4495,
+      "epoch": 3.967654986522911,
+      "grad_norm": 0.5629474520683289,
+      "learning_rate": 0.0003624284943335132,
+      "loss": 3.4362,
       "step": 36800
     },
     {
-      "epoch": 3.9662038531912605,
-      "grad_norm": 0.6509117484092712,
-      "learning_rate": 0.00036252128003447904,
-      "loss": 3.4472,
+      "epoch": 3.973045822102426,
+      "grad_norm": 0.6088916063308716,
+      "learning_rate": 0.00036210469508904473,
+      "loss": 3.4476,
       "step": 36850
     },
     {
-      "epoch": 3.9715854052308686,
-      "grad_norm": 0.6293677687644958,
-      "learning_rate": 0.0003621980390044176,
-      "loss": 3.4501,
+      "epoch": 3.9784366576819408,
+      "grad_norm": 0.6233236193656921,
+      "learning_rate": 0.00036178089584457633,
+      "loss": 3.4449,
       "step": 36900
     },
     {
-      "epoch": 3.9769669572704767,
-      "grad_norm": 0.6180201172828674,
-      "learning_rate": 0.0003618747979743562,
-      "loss": 3.4335,
+      "epoch": 3.9838274932614555,
+      "grad_norm": 0.6254053711891174,
+      "learning_rate": 0.0003614570966001079,
+      "loss": 3.4227,
       "step": 36950
     },
     {
-      "epoch": 3.9823485093100848,
-      "grad_norm": 0.6613346338272095,
-      "learning_rate": 0.0003615515569442947,
-      "loss": 3.432,
+      "epoch": 3.9892183288409706,
+      "grad_norm": 0.6487621665000916,
+      "learning_rate": 0.0003611332973556395,
+      "loss": 3.4118,
       "step": 37000
     },
     {
-      "epoch": 3.9823485093100848,
-      "eval_accuracy": 0.3764660669968918,
-      "eval_loss": 3.440922737121582,
-      "eval_runtime": 183.7763,
-      "eval_samples_per_second": 98.005,
-      "eval_steps_per_second": 6.127,
+      "epoch": 3.9892183288409706,
+      "eval_accuracy": 0.37649953208633863,
+      "eval_loss": 3.4413223266601562,
+      "eval_runtime": 184.0756,
+      "eval_samples_per_second": 97.846,
+      "eval_steps_per_second": 6.117,
       "step": 37000
     },
     {
-      "epoch": 3.9877300613496933,
-      "grad_norm": 0.6783581972122192,
-      "learning_rate": 0.00036122831591423336,
-      "loss": 3.4577,
+      "epoch": 3.9946091644204853,
+      "grad_norm": 0.6213529706001282,
+      "learning_rate": 0.00036080949811117104,
+      "loss": 3.4638,
       "step": 37050
     },
     {
-      "epoch": 3.9931116133893014,
-      "grad_norm": 0.6816239356994629,
-      "learning_rate": 0.00036090507488417196,
-      "loss": 3.4461,
+      "epoch": 4.0,
+      "grad_norm": 1.2671101093292236,
+      "learning_rate": 0.00036048569886670264,
+      "loss": 3.4273,
       "step": 37100
     },
     {
-      "epoch": 3.9984931654289095,
-      "grad_norm": 0.6532045602798462,
-      "learning_rate": 0.0003605818338541105,
-      "loss": 3.4411,
+      "epoch": 4.005390835579515,
+      "grad_norm": 0.5932519435882568,
+      "learning_rate": 0.0003601618996222342,
+      "loss": 3.3586,
       "step": 37150
     },
     {
-      "epoch": 4.003874717468518,
-      "grad_norm": 0.6157842874526978,
-      "learning_rate": 0.0003602585928240491,
-      "loss": 3.3606,
+      "epoch": 4.010781671159029,
+      "grad_norm": 0.5957099199295044,
+      "learning_rate": 0.00035983810037776574,
+      "loss": 3.3539,
       "step": 37200
     },
     {
-      "epoch": 4.009256269508126,
-      "grad_norm": 0.6219887137413025,
-      "learning_rate": 0.0003599353517939877,
-      "loss": 3.3417,
+      "epoch": 4.0161725067385445,
+      "grad_norm": 0.6529663801193237,
+      "learning_rate": 0.00035951430113329735,
+      "loss": 3.3491,
       "step": 37250
     },
     {
-      "epoch": 4.014637821547734,
-      "grad_norm": 0.668585479259491,
-      "learning_rate": 0.00035961211076392623,
-      "loss": 3.3517,
+      "epoch": 4.02156334231806,
+      "grad_norm": 0.6074540615081787,
+      "learning_rate": 0.0003591969778737183,
+      "loss": 3.354,
       "step": 37300
     },
     {
-      "epoch": 4.020019373587343,
-      "grad_norm": 0.6259785294532776,
-      "learning_rate": 0.0003592888697338649,
-      "loss": 3.3491,
+      "epoch": 4.026954177897574,
+      "grad_norm": 0.615213930606842,
+      "learning_rate": 0.00035887317862924985,
+      "loss": 3.3358,
       "step": 37350
     },
     {
-      "epoch": 4.0254009256269505,
-      "grad_norm": 0.5741066932678223,
-      "learning_rate": 0.00035896562870380347,
-      "loss": 3.3637,
+      "epoch": 4.032345013477089,
+      "grad_norm": 0.6170495748519897,
+      "learning_rate": 0.00035854937938478146,
+      "loss": 3.3403,
       "step": 37400
     },
     {
-      "epoch": 4.030782477666559,
-      "grad_norm": 0.5968427658081055,
-      "learning_rate": 0.000358642387673742,
-      "loss": 3.3617,
+      "epoch": 4.037735849056604,
+      "grad_norm": 0.6581596732139587,
+      "learning_rate": 0.00035822558014031295,
+      "loss": 3.3488,
       "step": 37450
     },
     {
-      "epoch": 4.036164029706168,
-      "grad_norm": 0.6598249673843384,
-      "learning_rate": 0.0003583191466436806,
-      "loss": 3.354,
+      "epoch": 4.0431266846361185,
+      "grad_norm": 0.6761884689331055,
+      "learning_rate": 0.0003579017808958445,
+      "loss": 3.3478,
       "step": 37500
     },
     {
-      "epoch": 4.041545581745775,
-      "grad_norm": 0.6724773645401001,
-      "learning_rate": 0.00035799590561361915,
-      "loss": 3.3533,
+      "epoch": 4.048517520215634,
+      "grad_norm": 0.6621485352516174,
+      "learning_rate": 0.0003575779816513761,
+      "loss": 3.3599,
       "step": 37550
     },
     {
-      "epoch": 4.046927133785384,
-      "grad_norm": 0.7567223310470581,
-      "learning_rate": 0.00035767266458355774,
-      "loss": 3.3443,
+      "epoch": 4.053908355795148,
+      "grad_norm": 0.6355655193328857,
+      "learning_rate": 0.00035725418240690766,
+      "loss": 3.3524,
       "step": 37600
     },
     {
-      "epoch": 4.0523086858249915,
-      "grad_norm": 0.6546953916549683,
-      "learning_rate": 0.0003573494235534964,
-      "loss": 3.3594,
+      "epoch": 4.059299191374663,
+      "grad_norm": 0.6401028633117676,
+      "learning_rate": 0.00035693038316243926,
+      "loss": 3.3601,
       "step": 37650
     },
     {
-      "epoch": 4.0576902378646,
-      "grad_norm": 0.6540945768356323,
-      "learning_rate": 0.00035702618252343493,
-      "loss": 3.3593,
+      "epoch": 4.064690026954178,
+      "grad_norm": 0.5923633575439453,
+      "learning_rate": 0.0003566065839179708,
+      "loss": 3.3616,
       "step": 37700
     },
     {
-      "epoch": 4.063071789904209,
-      "grad_norm": 0.6679083108901978,
-      "learning_rate": 0.0003567029414933735,
-      "loss": 3.3574,
+      "epoch": 4.070080862533692,
+      "grad_norm": 0.5979942679405212,
+      "learning_rate": 0.0003562827846735024,
+      "loss": 3.3675,
       "step": 37750
     },
     {
-      "epoch": 4.068453341943816,
-      "grad_norm": 0.6649799942970276,
-      "learning_rate": 0.0003563797004633121,
-      "loss": 3.3549,
+      "epoch": 4.0754716981132075,
+      "grad_norm": 0.6190664172172546,
+      "learning_rate": 0.00035595898542903397,
+      "loss": 3.3607,
       "step": 37800
     },
     {
-      "epoch": 4.073834893983425,
-      "grad_norm": 0.6234748363494873,
-      "learning_rate": 0.00035605645943325066,
-      "loss": 3.3514,
+      "epoch": 4.080862533692723,
+      "grad_norm": 0.649185836315155,
+      "learning_rate": 0.0003556351861845656,
+      "loss": 3.3657,
       "step": 37850
     },
     {
-      "epoch": 4.079216446023033,
-      "grad_norm": 0.6429659128189087,
-      "learning_rate": 0.0003557332184031893,
-      "loss": 3.367,
+      "epoch": 4.086253369272237,
+      "grad_norm": 0.6081229448318481,
+      "learning_rate": 0.0003553113869400971,
+      "loss": 3.3736,
       "step": 37900
     },
     {
-      "epoch": 4.084597998062641,
-      "grad_norm": 0.598976194858551,
-      "learning_rate": 0.0003554099773731279,
-      "loss": 3.3676,
+      "epoch": 4.091644204851752,
+      "grad_norm": 0.6000233888626099,
+      "learning_rate": 0.0003549875876956287,
+      "loss": 3.3716,
       "step": 37950
     },
     {
-      "epoch": 4.08997955010225,
-      "grad_norm": 0.645483672618866,
-      "learning_rate": 0.00035508673634306644,
-      "loss": 3.3552,
+      "epoch": 4.097035040431267,
+      "grad_norm": 0.6363604664802551,
+      "learning_rate": 0.0003546637884511603,
+      "loss": 3.3556,
       "step": 38000
     },
     {
-      "epoch": 4.08997955010225,
-      "eval_accuracy": 0.3766763103348128,
-      "eval_loss": 3.44470477104187,
-      "eval_runtime": 183.8588,
-      "eval_samples_per_second": 97.961,
-      "eval_steps_per_second": 6.124,
+      "epoch": 4.097035040431267,
+      "eval_accuracy": 0.37682255712181106,
+      "eval_loss": 3.445936441421509,
+      "eval_runtime": 183.7808,
+      "eval_samples_per_second": 98.003,
+      "eval_steps_per_second": 6.127,
       "step": 38000
     },
     {
-      "epoch": 4.095361102141858,
-      "grad_norm": 0.6297438740730286,
-      "learning_rate": 0.00035476349531300504,
-      "loss": 3.3672,
+      "epoch": 4.1024258760107815,
+      "grad_norm": 0.6588642001152039,
+      "learning_rate": 0.0003543399892066918,
+      "loss": 3.3708,
       "step": 38050
     },
     {
-      "epoch": 4.100742654181466,
-      "grad_norm": 0.640561044216156,
-      "learning_rate": 0.0003544402542829436,
-      "loss": 3.3748,
+      "epoch": 4.107816711590297,
+      "grad_norm": 0.6509077548980713,
+      "learning_rate": 0.0003540161899622234,
+      "loss": 3.3555,
       "step": 38100
     },
     {
-      "epoch": 4.106124206221074,
-      "grad_norm": 0.6440055966377258,
-      "learning_rate": 0.0003541170132528822,
-      "loss": 3.3435,
+      "epoch": 4.113207547169812,
+      "grad_norm": 0.6492730379104614,
+      "learning_rate": 0.00035369239071775493,
+      "loss": 3.3716,
       "step": 38150
     },
     {
-      "epoch": 4.111505758260682,
-      "grad_norm": 0.6432259678840637,
-      "learning_rate": 0.0003537937722228208,
-      "loss": 3.365,
+      "epoch": 4.118598382749326,
+      "grad_norm": 0.665730893611908,
+      "learning_rate": 0.00035336859147328654,
+      "loss": 3.3745,
       "step": 38200
     },
     {
-      "epoch": 4.1168873103002905,
-      "grad_norm": 0.6475505828857422,
-      "learning_rate": 0.00035347053119275936,
-      "loss": 3.3665,
+      "epoch": 4.123989218328841,
+      "grad_norm": 0.639869749546051,
+      "learning_rate": 0.0003530447922288181,
+      "loss": 3.3625,
       "step": 38250
     },
     {
-      "epoch": 4.122268862339899,
-      "grad_norm": 0.6022783517837524,
-      "learning_rate": 0.00035314729016269796,
-      "loss": 3.3706,
+      "epoch": 4.129380053908355,
+      "grad_norm": 0.6096779108047485,
+      "learning_rate": 0.0003527209929843497,
+      "loss": 3.3665,
       "step": 38300
     },
     {
-      "epoch": 4.127650414379507,
-      "grad_norm": 0.6375998258590698,
-      "learning_rate": 0.00035282404913263655,
-      "loss": 3.3683,
+      "epoch": 4.1347708894878705,
+      "grad_norm": 0.6341913938522339,
+      "learning_rate": 0.00035239719373988124,
+      "loss": 3.351,
       "step": 38350
     },
     {
-      "epoch": 4.133031966419115,
-      "grad_norm": 0.6592504978179932,
-      "learning_rate": 0.0003525008081025751,
-      "loss": 3.3407,
+      "epoch": 4.140161725067386,
+      "grad_norm": 0.634777843952179,
+      "learning_rate": 0.0003520733944954128,
+      "loss": 3.3723,
       "step": 38400
     },
     {
-      "epoch": 4.138413518458724,
-      "grad_norm": 0.6265368461608887,
-      "learning_rate": 0.0003521775670725137,
-      "loss": 3.3748,
+      "epoch": 4.1455525606469,
+      "grad_norm": 0.6281129717826843,
+      "learning_rate": 0.0003517495952509444,
+      "loss": 3.3679,
       "step": 38450
     },
     {
-      "epoch": 4.1437950704983315,
-      "grad_norm": 0.6979097723960876,
-      "learning_rate": 0.00035185432604245234,
-      "loss": 3.3859,
+      "epoch": 4.150943396226415,
+      "grad_norm": 0.6115019917488098,
+      "learning_rate": 0.00035142579600647595,
+      "loss": 3.3631,
       "step": 38500
     },
     {
-      "epoch": 4.14917662253794,
-      "grad_norm": 0.6460967063903809,
-      "learning_rate": 0.0003515310850123909,
-      "loss": 3.3645,
+      "epoch": 4.15633423180593,
+      "grad_norm": 0.5915253162384033,
+      "learning_rate": 0.00035110199676200755,
+      "loss": 3.3618,
       "step": 38550
     },
     {
-      "epoch": 4.154558174577549,
-      "grad_norm": 0.6688167452812195,
-      "learning_rate": 0.00035120784398232947,
-      "loss": 3.3698,
+      "epoch": 4.1617250673854445,
+      "grad_norm": 0.6483709216117859,
+      "learning_rate": 0.0003507781975175391,
+      "loss": 3.36,
       "step": 38600
     },
     {
-      "epoch": 4.159939726617156,
-      "grad_norm": 0.6723623871803284,
-      "learning_rate": 0.000350884602952268,
-      "loss": 3.3629,
+      "epoch": 4.16711590296496,
+      "grad_norm": 0.68476402759552,
+      "learning_rate": 0.0003504543982730707,
+      "loss": 3.365,
       "step": 38650
     },
     {
-      "epoch": 4.165321278656765,
-      "grad_norm": 0.5828025341033936,
-      "learning_rate": 0.0003505613619222066,
-      "loss": 3.3532,
+      "epoch": 4.172506738544475,
+      "grad_norm": 0.6686002016067505,
+      "learning_rate": 0.00035013059902860226,
+      "loss": 3.382,
       "step": 38700
     },
     {
-      "epoch": 4.1707028306963725,
-      "grad_norm": 0.6767308115959167,
-      "learning_rate": 0.0003502381208921452,
-      "loss": 3.366,
+      "epoch": 4.177897574123989,
+      "grad_norm": 0.6278947591781616,
+      "learning_rate": 0.00034980679978413375,
+      "loss": 3.3561,
       "step": 38750
     },
     {
-      "epoch": 4.176084382735981,
-      "grad_norm": 0.6394909620285034,
-      "learning_rate": 0.0003499148798620838,
-      "loss": 3.3649,
+      "epoch": 4.183288409703504,
+      "grad_norm": 0.613915741443634,
+      "learning_rate": 0.00034948300053966536,
+      "loss": 3.3725,
       "step": 38800
     },
     {
-      "epoch": 4.18146593477559,
-      "grad_norm": 0.6038942933082581,
-      "learning_rate": 0.0003495916388320224,
-      "loss": 3.3777,
+      "epoch": 4.188679245283019,
+      "grad_norm": 0.6106895208358765,
+      "learning_rate": 0.0003491592012951969,
+      "loss": 3.3815,
       "step": 38850
     },
     {
-      "epoch": 4.186847486815197,
-      "grad_norm": 0.6634101271629333,
-      "learning_rate": 0.000349268397801961,
-      "loss": 3.353,
+      "epoch": 4.1940700808625335,
+      "grad_norm": 0.6201578974723816,
+      "learning_rate": 0.0003488354020507285,
+      "loss": 3.3992,
       "step": 38900
     },
     {
-      "epoch": 4.192229038854806,
-      "grad_norm": 0.6209967732429504,
-      "learning_rate": 0.0003489451567718995,
-      "loss": 3.3622,
+      "epoch": 4.199460916442049,
+      "grad_norm": 0.6516886353492737,
+      "learning_rate": 0.00034851160280626007,
+      "loss": 3.3625,
       "step": 38950
     },
     {
-      "epoch": 4.197610590894414,
-      "grad_norm": 0.6100993752479553,
-      "learning_rate": 0.0003486219157418381,
-      "loss": 3.3569,
+      "epoch": 4.204851752021563,
+      "grad_norm": 0.6607277393341064,
+      "learning_rate": 0.00034818780356179167,
+      "loss": 3.3789,
       "step": 39000
     },
     {
-      "epoch": 4.197610590894414,
-      "eval_accuracy": 0.3773208392653283,
-      "eval_loss": 3.439364194869995,
-      "eval_runtime": 183.8535,
-      "eval_samples_per_second": 97.964,
-      "eval_steps_per_second": 6.124,
+      "epoch": 4.204851752021563,
+      "eval_accuracy": 0.3773035634561658,
+      "eval_loss": 3.4433228969573975,
+      "eval_runtime": 183.9605,
+      "eval_samples_per_second": 97.907,
+      "eval_steps_per_second": 6.121,
       "step": 39000
     },
     {
-      "epoch": 4.202992142934022,
-      "grad_norm": 0.6557092666625977,
-      "learning_rate": 0.00034829867471177677,
-      "loss": 3.3859,
+      "epoch": 4.210242587601078,
+      "grad_norm": 0.6869006156921387,
+      "learning_rate": 0.0003478640043173232,
+      "loss": 3.3845,
       "step": 39050
     },
     {
-      "epoch": 4.208373694973631,
-      "grad_norm": 0.6308462023735046,
-      "learning_rate": 0.0003479754336817153,
-      "loss": 3.3624,
+      "epoch": 4.215633423180593,
+      "grad_norm": 0.6801799535751343,
+      "learning_rate": 0.0003475402050728548,
+      "loss": 3.3865,
       "step": 39100
     },
     {
-      "epoch": 4.213755247013238,
-      "grad_norm": 0.6929288506507874,
-      "learning_rate": 0.0003476521926516539,
-      "loss": 3.3944,
+      "epoch": 4.2210242587601075,
+      "grad_norm": 0.6092689037322998,
+      "learning_rate": 0.0003472164058283864,
+      "loss": 3.374,
       "step": 39150
     },
     {
-      "epoch": 4.219136799052847,
-      "grad_norm": 0.6218537092208862,
-      "learning_rate": 0.00034732895162159245,
-      "loss": 3.3553,
+      "epoch": 4.226415094339623,
+      "grad_norm": 0.6734305024147034,
+      "learning_rate": 0.0003468926065839179,
+      "loss": 3.3649,
       "step": 39200
     },
     {
-      "epoch": 4.224518351092455,
-      "grad_norm": 0.6832320690155029,
-      "learning_rate": 0.00034700571059153104,
-      "loss": 3.3719,
+      "epoch": 4.231805929919138,
+      "grad_norm": 0.6141255497932434,
+      "learning_rate": 0.00034656880733944953,
+      "loss": 3.368,
       "step": 39250
     },
     {
-      "epoch": 4.229899903132063,
-      "grad_norm": 0.6525055766105652,
-      "learning_rate": 0.00034668246956146963,
-      "loss": 3.3731,
+      "epoch": 4.237196765498652,
+      "grad_norm": 0.6386260390281677,
+      "learning_rate": 0.0003462450080949811,
+      "loss": 3.3712,
       "step": 39300
     },
     {
-      "epoch": 4.2352814551716715,
-      "grad_norm": 0.6206095218658447,
-      "learning_rate": 0.00034635922853140823,
-      "loss": 3.3895,
+      "epoch": 4.242587601078167,
+      "grad_norm": 0.6212135553359985,
+      "learning_rate": 0.0003459212088505127,
+      "loss": 3.3832,
       "step": 39350
     },
     {
-      "epoch": 4.24066300721128,
-      "grad_norm": 0.6386441588401794,
-      "learning_rate": 0.0003460359875013468,
-      "loss": 3.3934,
+      "epoch": 4.247978436657682,
+      "grad_norm": 0.6588135957717896,
+      "learning_rate": 0.0003455974096060442,
+      "loss": 3.3776,
       "step": 39400
     },
     {
-      "epoch": 4.246044559250888,
-      "grad_norm": 0.6195468902587891,
-      "learning_rate": 0.0003457127464712854,
-      "loss": 3.3748,
+      "epoch": 4.2533692722371965,
+      "grad_norm": 0.6394897699356079,
+      "learning_rate": 0.0003452736103615758,
+      "loss": 3.3862,
       "step": 39450
     },
     {
-      "epoch": 4.251426111290496,
-      "grad_norm": 0.6394476890563965,
-      "learning_rate": 0.00034538950544122396,
-      "loss": 3.3844,
+      "epoch": 4.258760107816712,
+      "grad_norm": 0.6298036575317383,
+      "learning_rate": 0.00034494981111710734,
+      "loss": 3.3742,
       "step": 39500
     },
     {
-      "epoch": 4.256807663330104,
-      "grad_norm": 0.6101006865501404,
-      "learning_rate": 0.00034506626441116255,
-      "loss": 3.3612,
+      "epoch": 4.264150943396227,
+      "grad_norm": 0.6218891739845276,
+      "learning_rate": 0.00034462601187263894,
+      "loss": 3.3619,
       "step": 39550
     },
     {
-      "epoch": 4.2621892153697125,
-      "grad_norm": 0.655703067779541,
-      "learning_rate": 0.0003447430233811011,
-      "loss": 3.3767,
+      "epoch": 4.269541778975741,
+      "grad_norm": 0.6185681819915771,
+      "learning_rate": 0.0003443022126281705,
+      "loss": 3.3745,
       "step": 39600
     },
     {
-      "epoch": 4.267570767409321,
-      "grad_norm": 0.6637636423110962,
-      "learning_rate": 0.00034441978235103974,
-      "loss": 3.3774,
+      "epoch": 4.274932614555256,
+      "grad_norm": 0.6114700436592102,
+      "learning_rate": 0.00034397841338370204,
+      "loss": 3.386,
       "step": 39650
     },
     {
-      "epoch": 4.272952319448929,
-      "grad_norm": 0.6590506434440613,
-      "learning_rate": 0.00034409654132097834,
-      "loss": 3.3743,
+      "epoch": 4.280323450134771,
+      "grad_norm": 0.670347273349762,
+      "learning_rate": 0.00034365461413923365,
+      "loss": 3.3683,
       "step": 39700
     },
     {
-      "epoch": 4.278333871488537,
-      "grad_norm": 0.704430103302002,
-      "learning_rate": 0.0003437733002909169,
-      "loss": 3.3713,
+      "epoch": 4.285714285714286,
+      "grad_norm": 0.6444869041442871,
+      "learning_rate": 0.0003433308148947652,
+      "loss": 3.3911,
       "step": 39750
     },
     {
-      "epoch": 4.283715423528146,
-      "grad_norm": 0.6607556939125061,
-      "learning_rate": 0.00034345652408145674,
-      "loss": 3.3631,
+      "epoch": 4.291105121293801,
+      "grad_norm": 0.6175254583358765,
+      "learning_rate": 0.0003430070156502968,
+      "loss": 3.378,
       "step": 39800
     },
     {
-      "epoch": 4.2890969755677535,
-      "grad_norm": 0.6469069123268127,
-      "learning_rate": 0.0003431332830513953,
-      "loss": 3.3846,
+      "epoch": 4.296495956873315,
+      "grad_norm": 0.6191949248313904,
+      "learning_rate": 0.00034268321640582835,
+      "loss": 3.3791,
       "step": 39850
     },
     {
-      "epoch": 4.294478527607362,
-      "grad_norm": 0.6928568482398987,
-      "learning_rate": 0.0003428100420213339,
-      "loss": 3.3829,
+      "epoch": 4.30188679245283,
+      "grad_norm": 0.6170054078102112,
+      "learning_rate": 0.00034235941716135996,
+      "loss": 3.3827,
       "step": 39900
     },
     {
-      "epoch": 4.299860079646971,
-      "grad_norm": 0.639872133731842,
-      "learning_rate": 0.0003424868009912724,
-      "loss": 3.3776,
+      "epoch": 4.307277628032345,
+      "grad_norm": 0.7263216972351074,
+      "learning_rate": 0.0003420356179168915,
+      "loss": 3.3888,
       "step": 39950
     },
     {
-      "epoch": 4.305241631686578,
-      "grad_norm": 0.6535642147064209,
-      "learning_rate": 0.00034216355996121106,
-      "loss": 3.3986,
+      "epoch": 4.3126684636118595,
+      "grad_norm": 0.6494803428649902,
+      "learning_rate": 0.0003417118186724231,
+      "loss": 3.3939,
       "step": 40000
     },
     {
-      "epoch": 4.305241631686578,
-      "eval_accuracy": 0.3778865948521783,
-      "eval_loss": 3.4329850673675537,
-      "eval_runtime": 184.1221,
-      "eval_samples_per_second": 97.821,
-      "eval_steps_per_second": 6.116,
+      "epoch": 4.3126684636118595,
+      "eval_accuracy": 0.37754651131331896,
+      "eval_loss": 3.4369258880615234,
+      "eval_runtime": 183.7951,
+      "eval_samples_per_second": 97.995,
+      "eval_steps_per_second": 6.126,
       "step": 40000
     },
     {
-      "epoch": 4.310623183726187,
-      "grad_norm": 0.6626715660095215,
-      "learning_rate": 0.00034184031893114966,
-      "loss": 3.3762,
+      "epoch": 4.318059299191375,
+      "grad_norm": 0.625522792339325,
+      "learning_rate": 0.00034138801942795467,
+      "loss": 3.3828,
       "step": 40050
     },
     {
-      "epoch": 4.3160047357657945,
-      "grad_norm": 0.6559535264968872,
-      "learning_rate": 0.0003415170779010882,
-      "loss": 3.3912,
+      "epoch": 4.32345013477089,
+      "grad_norm": 0.6647518873214722,
+      "learning_rate": 0.00034106422018348616,
+      "loss": 3.3883,
       "step": 40100
     },
     {
-      "epoch": 4.321386287805403,
-      "grad_norm": 0.6913949847221375,
-      "learning_rate": 0.0003411938368710268,
-      "loss": 3.3687,
+      "epoch": 4.328840970350404,
+      "grad_norm": 0.6370512843132019,
+      "learning_rate": 0.00034074042093901777,
+      "loss": 3.3695,
       "step": 40150
     },
     {
-      "epoch": 4.326767839845012,
-      "grad_norm": 0.6739501357078552,
-      "learning_rate": 0.0003408705958409654,
-      "loss": 3.3712,
+      "epoch": 4.334231805929919,
+      "grad_norm": 0.6868136525154114,
+      "learning_rate": 0.0003404166216945493,
+      "loss": 3.4028,
       "step": 40200
     },
     {
-      "epoch": 4.332149391884619,
-      "grad_norm": 0.6367455124855042,
-      "learning_rate": 0.00034054735481090393,
-      "loss": 3.3738,
+      "epoch": 4.339622641509434,
+      "grad_norm": 0.7005695104598999,
+      "learning_rate": 0.0003400928224500809,
+      "loss": 3.4074,
       "step": 40250
     },
     {
-      "epoch": 4.337530943924228,
-      "grad_norm": 0.6584734916687012,
-      "learning_rate": 0.0003402241137808426,
-      "loss": 3.3804,
+      "epoch": 4.345013477088949,
+      "grad_norm": 0.6485339999198914,
+      "learning_rate": 0.00033976902320561247,
+      "loss": 3.377,
       "step": 40300
     },
     {
-      "epoch": 4.342912495963836,
-      "grad_norm": 0.6525224447250366,
-      "learning_rate": 0.00033990087275078117,
-      "loss": 3.3891,
+      "epoch": 4.350404312668464,
+      "grad_norm": 0.6482576727867126,
+      "learning_rate": 0.0003394452239611441,
+      "loss": 3.3882,
       "step": 40350
     },
     {
-      "epoch": 4.348294048003444,
-      "grad_norm": 0.6361015439033508,
-      "learning_rate": 0.0003395776317207197,
-      "loss": 3.3817,
+      "epoch": 4.355795148247978,
+      "grad_norm": 0.6321107745170593,
+      "learning_rate": 0.00033912142471667563,
+      "loss": 3.3844,
       "step": 40400
     },
     {
-      "epoch": 4.3536756000430525,
-      "grad_norm": 0.7042331695556641,
-      "learning_rate": 0.0003392543906906583,
-      "loss": 3.38,
+      "epoch": 4.361185983827493,
+      "grad_norm": 0.6567569971084595,
+      "learning_rate": 0.00033879762547220723,
+      "loss": 3.3851,
       "step": 40450
     },
     {
-      "epoch": 4.359057152082661,
-      "grad_norm": 0.6377615332603455,
-      "learning_rate": 0.00033893114966059685,
-      "loss": 3.3748,
+      "epoch": 4.366576819407008,
+      "grad_norm": 0.6365229487419128,
+      "learning_rate": 0.0003384738262277388,
+      "loss": 3.3778,
       "step": 40500
     },
     {
-      "epoch": 4.364438704122269,
-      "grad_norm": 0.6654162406921387,
-      "learning_rate": 0.00033860790863053544,
-      "loss": 3.3691,
+      "epoch": 4.3719676549865225,
+      "grad_norm": 0.6746280193328857,
+      "learning_rate": 0.00033815002698327033,
+      "loss": 3.3856,
       "step": 40550
     },
     {
-      "epoch": 4.369820256161877,
-      "grad_norm": 0.7069597840309143,
-      "learning_rate": 0.0003382846676004741,
-      "loss": 3.3695,
+      "epoch": 4.377358490566038,
+      "grad_norm": 0.6441705822944641,
+      "learning_rate": 0.00033782622773880194,
+      "loss": 3.3913,
       "step": 40600
     },
     {
-      "epoch": 4.375201808201485,
-      "grad_norm": 0.7014625668525696,
-      "learning_rate": 0.00033796142657041263,
-      "loss": 3.3661,
+      "epoch": 4.382749326145553,
+      "grad_norm": 0.6651356816291809,
+      "learning_rate": 0.0003375024284943335,
+      "loss": 3.3817,
       "step": 40650
     },
     {
-      "epoch": 4.3805833602410935,
-      "grad_norm": 0.6131076216697693,
-      "learning_rate": 0.0003376381855403512,
-      "loss": 3.3891,
+      "epoch": 4.388140161725067,
+      "grad_norm": 0.6996265649795532,
+      "learning_rate": 0.0003371786292498651,
+      "loss": 3.3963,
       "step": 40700
     },
     {
-      "epoch": 4.385964912280702,
-      "grad_norm": 0.6640186309814453,
-      "learning_rate": 0.0003373149445102898,
-      "loss": 3.3761,
+      "epoch": 4.393530997304582,
+      "grad_norm": 0.6313248872756958,
+      "learning_rate": 0.0003368548300053966,
+      "loss": 3.398,
       "step": 40750
     },
     {
-      "epoch": 4.39134646432031,
-      "grad_norm": 0.6470939517021179,
-      "learning_rate": 0.00033699170348022836,
-      "loss": 3.3901,
+      "epoch": 4.398921832884097,
+      "grad_norm": 0.6112807989120483,
+      "learning_rate": 0.0003365310307609282,
+      "loss": 3.3863,
       "step": 40800
     },
     {
-      "epoch": 4.396728016359918,
-      "grad_norm": 0.6494837403297424,
-      "learning_rate": 0.000336668462450167,
-      "loss": 3.3881,
+      "epoch": 4.404312668463612,
+      "grad_norm": 0.6694891452789307,
+      "learning_rate": 0.00033620723151645975,
+      "loss": 3.3773,
       "step": 40850
     },
     {
-      "epoch": 4.402109568399527,
-      "grad_norm": 0.6405217051506042,
-      "learning_rate": 0.0003363452214201056,
-      "loss": 3.3895,
+      "epoch": 4.409703504043127,
+      "grad_norm": 0.6523438692092896,
+      "learning_rate": 0.0003358834322719913,
+      "loss": 3.3788,
       "step": 40900
     },
     {
-      "epoch": 4.4074911204391345,
-      "grad_norm": 0.6740948557853699,
-      "learning_rate": 0.00033602198039004415,
-      "loss": 3.3668,
+      "epoch": 4.415094339622642,
+      "grad_norm": 1.0051461458206177,
+      "learning_rate": 0.0003355596330275229,
+      "loss": 3.3738,
       "step": 40950
     },
     {
-      "epoch": 4.412872672478743,
-      "grad_norm": 0.6452889442443848,
-      "learning_rate": 0.00033569873935998274,
-      "loss": 3.3827,
+      "epoch": 4.420485175202156,
+      "grad_norm": 0.696164608001709,
+      "learning_rate": 0.00033523583378305445,
+      "loss": 3.409,
       "step": 41000
     },
     {
-      "epoch": 4.412872672478743,
-      "eval_accuracy": 0.3785802348879859,
-      "eval_loss": 3.429490327835083,
-      "eval_runtime": 183.8858,
-      "eval_samples_per_second": 97.947,
-      "eval_steps_per_second": 6.123,
+      "epoch": 4.420485175202156,
+      "eval_accuracy": 0.3786097884734715,
+      "eval_loss": 3.4320640563964844,
+      "eval_runtime": 184.0465,
+      "eval_samples_per_second": 97.861,
+      "eval_steps_per_second": 6.118,
       "step": 41000
     },
     {
-      "epoch": 4.418254224518351,
-      "grad_norm": 0.632122814655304,
-      "learning_rate": 0.0003353754983299213,
-      "loss": 3.38,
+      "epoch": 4.425876010781671,
+      "grad_norm": 0.6436068415641785,
+      "learning_rate": 0.00033491203453858606,
+      "loss": 3.386,
       "step": 41050
     },
     {
-      "epoch": 4.423635776557959,
-      "grad_norm": 0.6197381019592285,
-      "learning_rate": 0.0003350522572998599,
-      "loss": 3.4,
+      "epoch": 4.431266846361186,
+      "grad_norm": 0.6470723152160645,
+      "learning_rate": 0.0003345882352941176,
+      "loss": 3.3928,
       "step": 41100
     },
     {
-      "epoch": 4.429017328597568,
-      "grad_norm": 0.6479023098945618,
-      "learning_rate": 0.0003347290162697985,
-      "loss": 3.3803,
+      "epoch": 4.436657681940701,
+      "grad_norm": 0.6608076691627502,
+      "learning_rate": 0.0003342644360496492,
+      "loss": 3.3998,
       "step": 41150
     },
     {
-      "epoch": 4.4343988806371755,
-      "grad_norm": 0.7072307467460632,
-      "learning_rate": 0.00033440577523973706,
-      "loss": 3.3672,
+      "epoch": 4.442048517520216,
+      "grad_norm": 0.5988414883613586,
+      "learning_rate": 0.00033394063680518076,
+      "loss": 3.3837,
       "step": 41200
     },
     {
-      "epoch": 4.439780432676784,
-      "grad_norm": 0.7561652660369873,
-      "learning_rate": 0.00033408253420967566,
-      "loss": 3.3668,
+      "epoch": 4.44743935309973,
+      "grad_norm": 0.6084816455841064,
+      "learning_rate": 0.00033361683756071237,
+      "loss": 3.4013,
       "step": 41250
     },
     {
-      "epoch": 4.445161984716393,
-      "grad_norm": 0.641432523727417,
-      "learning_rate": 0.00033375929317961425,
-      "loss": 3.381,
+      "epoch": 4.452830188679245,
+      "grad_norm": 0.6011629104614258,
+      "learning_rate": 0.00033329951430113326,
+      "loss": 3.373,
       "step": 41300
     },
     {
-      "epoch": 4.450543536756,
-      "grad_norm": 0.6689733862876892,
-      "learning_rate": 0.0003334360521495528,
-      "loss": 3.3774,
+      "epoch": 4.45822102425876,
+      "grad_norm": 0.670461118221283,
+      "learning_rate": 0.00033297571505666487,
+      "loss": 3.3782,
       "step": 41350
     },
     {
-      "epoch": 4.455925088795609,
-      "grad_norm": 0.6350441575050354,
-      "learning_rate": 0.0003331128111194914,
-      "loss": 3.3805,
+      "epoch": 4.463611859838275,
+      "grad_norm": 0.6132466793060303,
+      "learning_rate": 0.00033265191581219637,
+      "loss": 3.4029,
       "step": 41400
     },
     {
-      "epoch": 4.461306640835216,
-      "grad_norm": 0.6296270489692688,
-      "learning_rate": 0.00033278957008943004,
-      "loss": 3.3602,
+      "epoch": 4.46900269541779,
+      "grad_norm": 0.6491686701774597,
+      "learning_rate": 0.0003323345925526173,
+      "loss": 3.377,
       "step": 41450
     },
     {
-      "epoch": 4.466688192874825,
-      "grad_norm": 0.6392343044281006,
-      "learning_rate": 0.0003324663290593686,
-      "loss": 3.391,
+      "epoch": 4.474393530997305,
+      "grad_norm": 0.9079296588897705,
+      "learning_rate": 0.0003320107933081489,
+      "loss": 3.3898,
       "step": 41500
     },
     {
-      "epoch": 4.4720697449144335,
-      "grad_norm": 0.6158018708229065,
-      "learning_rate": 0.00033214308802930717,
+      "epoch": 4.479784366576819,
+      "grad_norm": 0.6514104008674622,
+      "learning_rate": 0.0003316869940636805,
       "loss": 3.384,
       "step": 41550
     },
     {
-      "epoch": 4.477451296954041,
-      "grad_norm": 0.6282882690429688,
-      "learning_rate": 0.0003318198469992457,
-      "loss": 3.3849,
+      "epoch": 4.485175202156334,
+      "grad_norm": 0.6291401982307434,
+      "learning_rate": 0.000331363194819212,
+      "loss": 3.3934,
       "step": 41600
     },
     {
-      "epoch": 4.48283284899365,
-      "grad_norm": 0.804178774356842,
-      "learning_rate": 0.0003314966059691843,
-      "loss": 3.3743,
+      "epoch": 4.490566037735849,
+      "grad_norm": 0.6174262762069702,
+      "learning_rate": 0.00033103939557474363,
+      "loss": 3.3776,
       "step": 41650
     },
     {
-      "epoch": 4.488214401033258,
-      "grad_norm": 0.7374902367591858,
-      "learning_rate": 0.00033117336493912296,
-      "loss": 3.3817,
+      "epoch": 4.495956873315364,
+      "grad_norm": 0.6103610992431641,
+      "learning_rate": 0.0003307155963302752,
+      "loss": 3.4008,
       "step": 41700
     },
     {
-      "epoch": 4.493595953072866,
-      "grad_norm": 0.6747592687606812,
-      "learning_rate": 0.0003308501239090615,
-      "loss": 3.3905,
+      "epoch": 4.501347708894879,
+      "grad_norm": 0.6482797265052795,
+      "learning_rate": 0.0003303917970858068,
+      "loss": 3.383,
       "step": 41750
     },
     {
-      "epoch": 4.4989775051124745,
-      "grad_norm": 0.6296213269233704,
-      "learning_rate": 0.00033053334769960136,
-      "loss": 3.3789,
+      "epoch": 4.506738544474393,
+      "grad_norm": 0.6536639332771301,
+      "learning_rate": 0.00033006799784133833,
+      "loss": 3.3882,
       "step": 41800
     },
     {
-      "epoch": 4.504359057152083,
-      "grad_norm": 0.6893779039382935,
-      "learning_rate": 0.0003302101066695399,
-      "loss": 3.3551,
+      "epoch": 4.512129380053908,
+      "grad_norm": 0.6389809846878052,
+      "learning_rate": 0.00032974419859686994,
+      "loss": 3.3769,
       "step": 41850
     },
     {
-      "epoch": 4.509740609191691,
-      "grad_norm": 0.6661280989646912,
-      "learning_rate": 0.0003298868656394785,
-      "loss": 3.3928,
+      "epoch": 4.517520215633423,
+      "grad_norm": 0.6646363735198975,
+      "learning_rate": 0.0003294203993524015,
+      "loss": 3.3878,
       "step": 41900
     },
     {
-      "epoch": 4.515122161231299,
-      "grad_norm": 0.6617748737335205,
-      "learning_rate": 0.00032956362460941703,
-      "loss": 3.3853,
+      "epoch": 4.5229110512129385,
+      "grad_norm": 0.6328353881835938,
+      "learning_rate": 0.0003290966001079331,
+      "loss": 3.3711,
       "step": 41950
     },
     {
-      "epoch": 4.520503713270907,
-      "grad_norm": 0.6845855116844177,
-      "learning_rate": 0.00032924038357935563,
-      "loss": 3.3957,
+      "epoch": 4.528301886792453,
+      "grad_norm": 0.604308545589447,
+      "learning_rate": 0.00032877280086346465,
+      "loss": 3.3812,
       "step": 42000
     },
     {
-      "epoch": 4.520503713270907,
-      "eval_accuracy": 0.3783456533031945,
-      "eval_loss": 3.425133466720581,
-      "eval_runtime": 184.0009,
-      "eval_samples_per_second": 97.885,
-      "eval_steps_per_second": 6.12,
+      "epoch": 4.528301886792453,
+      "eval_accuracy": 0.37897214585433275,
+      "eval_loss": 3.4261536598205566,
+      "eval_runtime": 184.1249,
+      "eval_samples_per_second": 97.819,
+      "eval_steps_per_second": 6.115,
       "step": 42000
     },
     {
-      "epoch": 4.5258852653105155,
-      "grad_norm": 0.6436318159103394,
-      "learning_rate": 0.0003289171425492942,
-      "loss": 3.3901,
+      "epoch": 4.533692722371968,
+      "grad_norm": 0.6494730114936829,
+      "learning_rate": 0.00032844900161899614,
+      "loss": 3.3792,
       "step": 42050
     },
     {
-      "epoch": 4.531266817350124,
-      "grad_norm": 0.6274494528770447,
-      "learning_rate": 0.0003285939015192328,
-      "loss": 3.3886,
+      "epoch": 4.539083557951482,
+      "grad_norm": 0.6747742295265198,
+      "learning_rate": 0.0003281252023745278,
+      "loss": 3.3877,
       "step": 42100
     },
     {
-      "epoch": 4.536648369389732,
-      "grad_norm": 0.6229202747344971,
-      "learning_rate": 0.0003282706604891714,
-      "loss": 3.3847,
+      "epoch": 4.544474393530997,
+      "grad_norm": 0.6609525084495544,
+      "learning_rate": 0.0003278014031300593,
+      "loss": 3.4154,
       "step": 42150
     },
     {
-      "epoch": 4.54202992142934,
-      "grad_norm": 0.663155198097229,
-      "learning_rate": 0.00032794741945911,
-      "loss": 3.3863,
+      "epoch": 4.549865229110512,
+      "grad_norm": 0.6202364563941956,
+      "learning_rate": 0.0003274776038855909,
+      "loss": 3.3821,
       "step": 42200
     },
     {
-      "epoch": 4.547411473468949,
-      "grad_norm": 0.6832782626152039,
-      "learning_rate": 0.00032762417842904855,
-      "loss": 3.392,
+      "epoch": 4.555256064690027,
+      "grad_norm": 0.6902115941047668,
+      "learning_rate": 0.00032715380464112245,
+      "loss": 3.3813,
       "step": 42250
     },
     {
-      "epoch": 4.5527930255085565,
-      "grad_norm": 0.6422198414802551,
-      "learning_rate": 0.00032730093739898714,
-      "loss": 3.401,
+      "epoch": 4.560646900269542,
+      "grad_norm": 0.6965022683143616,
+      "learning_rate": 0.00032683000539665406,
+      "loss": 3.3855,
       "step": 42300
     },
     {
-      "epoch": 4.558174577548165,
-      "grad_norm": 0.702488124370575,
-      "learning_rate": 0.0003269776963689257,
-      "loss": 3.3698,
+      "epoch": 4.566037735849057,
+      "grad_norm": 0.6242098212242126,
+      "learning_rate": 0.0003265062061521856,
+      "loss": 3.386,
       "step": 42350
     },
     {
-      "epoch": 4.563556129587774,
-      "grad_norm": 0.658799409866333,
-      "learning_rate": 0.00032665445533886433,
-      "loss": 3.3685,
+      "epoch": 4.571428571428571,
+      "grad_norm": 0.6442424058914185,
+      "learning_rate": 0.0003261824069077172,
+      "loss": 3.3753,
       "step": 42400
     },
     {
-      "epoch": 4.568937681627381,
-      "grad_norm": 0.6828705668449402,
-      "learning_rate": 0.0003263312143088029,
-      "loss": 3.3899,
+      "epoch": 4.576819407008086,
+      "grad_norm": 0.6546700596809387,
+      "learning_rate": 0.00032585860766324876,
+      "loss": 3.388,
       "step": 42450
     },
     {
-      "epoch": 4.57431923366699,
-      "grad_norm": 0.6380706429481506,
-      "learning_rate": 0.00032600797327874147,
-      "loss": 3.3639,
+      "epoch": 4.5822102425876015,
+      "grad_norm": 0.620097279548645,
+      "learning_rate": 0.0003255348084187803,
+      "loss": 3.3852,
       "step": 42500
     },
     {
-      "epoch": 4.579700785706597,
-      "grad_norm": 0.6615903973579407,
-      "learning_rate": 0.00032568473224868006,
-      "loss": 3.4064,
+      "epoch": 4.587601078167116,
+      "grad_norm": 0.6199333667755127,
+      "learning_rate": 0.0003252110091743119,
+      "loss": 3.3851,
       "step": 42550
     },
     {
-      "epoch": 4.585082337746206,
-      "grad_norm": 0.6431669592857361,
-      "learning_rate": 0.00032536149121861866,
-      "loss": 3.3986,
+      "epoch": 4.592991913746631,
+      "grad_norm": 0.6354528665542603,
+      "learning_rate": 0.00032488720992984347,
+      "loss": 3.3688,
       "step": 42600
     },
     {
-      "epoch": 4.5904638897858145,
-      "grad_norm": 0.6949295997619629,
-      "learning_rate": 0.00032503825018855725,
-      "loss": 3.3908,
+      "epoch": 4.598382749326145,
+      "grad_norm": 0.7361127138137817,
+      "learning_rate": 0.0003245634106853751,
+      "loss": 3.3691,
       "step": 42650
     },
     {
-      "epoch": 4.595845441825422,
-      "grad_norm": 0.6859197020530701,
-      "learning_rate": 0.00032471500915849584,
-      "loss": 3.3957,
+      "epoch": 4.60377358490566,
+      "grad_norm": 0.6726773381233215,
+      "learning_rate": 0.0003242396114409066,
+      "loss": 3.3806,
       "step": 42700
     },
     {
-      "epoch": 4.601226993865031,
-      "grad_norm": 0.6373593807220459,
-      "learning_rate": 0.00032439176812843444,
-      "loss": 3.3912,
+      "epoch": 4.609164420485175,
+      "grad_norm": 0.650661289691925,
+      "learning_rate": 0.00032391581219643823,
+      "loss": 3.3981,
       "step": 42750
     },
     {
-      "epoch": 4.606608545904638,
-      "grad_norm": 0.6689690947532654,
-      "learning_rate": 0.000324068527098373,
-      "loss": 3.3827,
+      "epoch": 4.6145552560646905,
+      "grad_norm": 0.6549949049949646,
+      "learning_rate": 0.0003235920129519697,
+      "loss": 3.3685,
       "step": 42800
     },
     {
-      "epoch": 4.611990097944247,
-      "grad_norm": 0.6162630915641785,
-      "learning_rate": 0.0003237452860683116,
-      "loss": 3.4155,
+      "epoch": 4.619946091644205,
+      "grad_norm": 0.6273858547210693,
+      "learning_rate": 0.0003232682137075013,
+      "loss": 3.376,
       "step": 42850
     },
     {
-      "epoch": 4.6173716499838555,
-      "grad_norm": 0.6874521374702454,
-      "learning_rate": 0.0003234220450382501,
-      "loss": 3.3717,
+      "epoch": 4.62533692722372,
+      "grad_norm": 0.6732528209686279,
+      "learning_rate": 0.0003229444144630329,
+      "loss": 3.3738,
       "step": 42900
     },
     {
-      "epoch": 4.622753202023463,
-      "grad_norm": 0.64226233959198,
-      "learning_rate": 0.00032309880400818876,
-      "loss": 3.375,
+      "epoch": 4.630727762803234,
+      "grad_norm": 0.6907714009284973,
+      "learning_rate": 0.00032262061521856443,
+      "loss": 3.3966,
       "step": 42950
     },
     {
-      "epoch": 4.628134754063072,
-      "grad_norm": 0.6314151287078857,
-      "learning_rate": 0.00032277556297812736,
-      "loss": 3.388,
+      "epoch": 4.636118598382749,
+      "grad_norm": 0.6889829635620117,
+      "learning_rate": 0.00032229681597409604,
+      "loss": 3.4015,
       "step": 43000
     },
     {
-      "epoch": 4.628134754063072,
-      "eval_accuracy": 0.37912784544257094,
-      "eval_loss": 3.4221925735473633,
-      "eval_runtime": 183.8944,
-      "eval_samples_per_second": 97.942,
-      "eval_steps_per_second": 6.123,
+      "epoch": 4.636118598382749,
+      "eval_accuracy": 0.3794535868002387,
+      "eval_loss": 3.4208765029907227,
+      "eval_runtime": 183.4902,
+      "eval_samples_per_second": 98.158,
+      "eval_steps_per_second": 6.137,
       "step": 43000
     },
     {
-      "epoch": 4.63351630610268,
-      "grad_norm": 0.6170079112052917,
-      "learning_rate": 0.0003224523219480659,
-      "loss": 3.3661,
+      "epoch": 4.6415094339622645,
+      "grad_norm": 0.6502795219421387,
+      "learning_rate": 0.0003219730167296276,
+      "loss": 3.3861,
       "step": 43050
     },
     {
-      "epoch": 4.638897858142288,
-      "grad_norm": 0.6966348886489868,
-      "learning_rate": 0.0003221290809180045,
-      "loss": 3.3799,
+      "epoch": 4.646900269541779,
+      "grad_norm": 0.6267656683921814,
+      "learning_rate": 0.0003216492174851592,
+      "loss": 3.3928,
       "step": 43100
     },
     {
-      "epoch": 4.6442794101818965,
-      "grad_norm": 0.6852479577064514,
-      "learning_rate": 0.0003218058398879431,
-      "loss": 3.4044,
+      "epoch": 4.652291105121294,
+      "grad_norm": 0.6557515859603882,
+      "learning_rate": 0.00032132541824069074,
+      "loss": 3.4049,
       "step": 43150
     },
     {
-      "epoch": 4.649660962221505,
-      "grad_norm": 0.6410502195358276,
-      "learning_rate": 0.00032148259885788163,
-      "loss": 3.3806,
+      "epoch": 4.657681940700809,
+      "grad_norm": 0.6499486565589905,
+      "learning_rate": 0.00032100161899622235,
+      "loss": 3.3779,
       "step": 43200
     },
     {
-      "epoch": 4.655042514261113,
-      "grad_norm": 0.6484990119934082,
-      "learning_rate": 0.0003211593578278203,
-      "loss": 3.3897,
+      "epoch": 4.663072776280323,
+      "grad_norm": 0.6578323245048523,
+      "learning_rate": 0.0003206778197517539,
+      "loss": 3.3759,
       "step": 43250
     },
     {
-      "epoch": 4.660424066300721,
-      "grad_norm": 0.6358705759048462,
-      "learning_rate": 0.00032083611679775887,
-      "loss": 3.381,
+      "epoch": 4.668463611859838,
+      "grad_norm": 0.6243415474891663,
+      "learning_rate": 0.00032035402050728545,
+      "loss": 3.3877,
       "step": 43300
     },
     {
-      "epoch": 4.665805618340329,
-      "grad_norm": 0.6190937757492065,
-      "learning_rate": 0.0003205128757676974,
-      "loss": 3.3935,
+      "epoch": 4.6738544474393535,
+      "grad_norm": 0.6552394032478333,
+      "learning_rate": 0.00032003022126281705,
+      "loss": 3.3874,
       "step": 43350
     },
     {
-      "epoch": 4.6711871703799375,
-      "grad_norm": 0.6699424982070923,
-      "learning_rate": 0.000320189634737636,
-      "loss": 3.3785,
+      "epoch": 4.679245283018868,
+      "grad_norm": 0.6453246474266052,
+      "learning_rate": 0.00031970642201834855,
+      "loss": 3.3983,
       "step": 43400
     },
     {
-      "epoch": 4.676568722419546,
-      "grad_norm": 0.6294956803321838,
-      "learning_rate": 0.00031986639370757455,
-      "loss": 3.3871,
+      "epoch": 4.684636118598383,
+      "grad_norm": 0.6403689980506897,
+      "learning_rate": 0.0003193826227738802,
+      "loss": 3.3845,
       "step": 43450
     },
     {
-      "epoch": 4.681950274459154,
-      "grad_norm": 0.6462880373001099,
-      "learning_rate": 0.0003195431526775132,
-      "loss": 3.3706,
+      "epoch": 4.690026954177897,
+      "grad_norm": 0.6712886095046997,
+      "learning_rate": 0.0003190588235294117,
+      "loss": 3.3987,
       "step": 43500
     },
     {
-      "epoch": 4.687331826498762,
-      "grad_norm": 0.6510257720947266,
-      "learning_rate": 0.0003192199116474518,
-      "loss": 3.3854,
+      "epoch": 4.695417789757412,
+      "grad_norm": 0.6037420034408569,
+      "learning_rate": 0.0003187350242849433,
+      "loss": 3.3844,
       "step": 43550
     },
     {
-      "epoch": 4.692713378538371,
-      "grad_norm": 0.6552780270576477,
-      "learning_rate": 0.00031889667061739033,
-      "loss": 3.3763,
+      "epoch": 4.7008086253369274,
+      "grad_norm": 0.64529949426651,
+      "learning_rate": 0.00031841122504047486,
+      "loss": 3.372,
       "step": 43600
     },
     {
-      "epoch": 4.6980949305779784,
-      "grad_norm": 0.6381099224090576,
-      "learning_rate": 0.0003185734295873289,
-      "loss": 3.3775,
+      "epoch": 4.706199460916442,
+      "grad_norm": 0.6656496524810791,
+      "learning_rate": 0.00031808742579600646,
+      "loss": 3.389,
       "step": 43650
     },
     {
-      "epoch": 4.703476482617587,
-      "grad_norm": 0.6657686829566956,
-      "learning_rate": 0.00031825665337786873,
-      "loss": 3.3893,
+      "epoch": 4.711590296495957,
+      "grad_norm": 0.6482957601547241,
+      "learning_rate": 0.000317763626551538,
+      "loss": 3.3717,
       "step": 43700
     },
     {
-      "epoch": 4.7088580346571955,
-      "grad_norm": 0.6148411631584167,
-      "learning_rate": 0.00031793341234780733,
-      "loss": 3.398,
+      "epoch": 4.716981132075472,
+      "grad_norm": 0.6282253265380859,
+      "learning_rate": 0.00031743982730706957,
+      "loss": 3.3914,
       "step": 43750
     },
     {
-      "epoch": 4.714239586696803,
-      "grad_norm": 0.6337029933929443,
-      "learning_rate": 0.00031761017131774587,
-      "loss": 3.3904,
+      "epoch": 4.722371967654986,
+      "grad_norm": 0.6619946956634521,
+      "learning_rate": 0.00031711602806260117,
+      "loss": 3.3879,
       "step": 43800
     },
     {
-      "epoch": 4.719621138736412,
-      "grad_norm": 0.6424337029457092,
-      "learning_rate": 0.00031728693028768446,
-      "loss": 3.3719,
+      "epoch": 4.727762803234501,
+      "grad_norm": 0.6631132364273071,
+      "learning_rate": 0.0003167922288181327,
+      "loss": 3.3882,
       "step": 43850
     },
     {
-      "epoch": 4.725002690776019,
-      "grad_norm": 0.6574138402938843,
-      "learning_rate": 0.0003169636892576231,
-      "loss": 3.39,
+      "epoch": 4.7331536388140165,
+      "grad_norm": 0.60288405418396,
+      "learning_rate": 0.0003164684295736643,
+      "loss": 3.3916,
       "step": 43900
     },
     {
-      "epoch": 4.730384242815628,
-      "grad_norm": 0.6762685179710388,
-      "learning_rate": 0.00031664044822756165,
-      "loss": 3.3763,
+      "epoch": 4.738544474393531,
+      "grad_norm": 0.6539262533187866,
+      "learning_rate": 0.0003161446303291959,
+      "loss": 3.3823,
       "step": 43950
     },
     {
-      "epoch": 4.7357657948552365,
-      "grad_norm": 0.6850994825363159,
-      "learning_rate": 0.00031631720719750025,
-      "loss": 3.3817,
+      "epoch": 4.743935309973046,
+      "grad_norm": 0.6632742881774902,
+      "learning_rate": 0.0003158208310847275,
+      "loss": 3.3866,
       "step": 44000
     },
     {
-      "epoch": 4.7357657948552365,
-      "eval_accuracy": 0.3798224633543689,
-      "eval_loss": 3.4146735668182373,
-      "eval_runtime": 191.5678,
-      "eval_samples_per_second": 94.019,
-      "eval_steps_per_second": 5.878,
+      "epoch": 4.743935309973046,
+      "eval_accuracy": 0.3797525995475042,
+      "eval_loss": 3.4176528453826904,
+      "eval_runtime": 183.9083,
+      "eval_samples_per_second": 97.935,
+      "eval_steps_per_second": 6.123,
       "step": 44000
     },
     {
-      "epoch": 4.741147346894844,
-      "grad_norm": 0.689365029335022,
-      "learning_rate": 0.00031599396616743884,
-      "loss": 3.3669,
+      "epoch": 4.74932614555256,
+      "grad_norm": 0.6748712062835693,
+      "learning_rate": 0.00031549703184025903,
+      "loss": 3.3962,
       "step": 44050
     },
     {
-      "epoch": 4.746528898934453,
-      "grad_norm": 0.7000561952590942,
-      "learning_rate": 0.0003156707251373774,
-      "loss": 3.3842,
+      "epoch": 4.754716981132075,
+      "grad_norm": 0.6657115817070007,
+      "learning_rate": 0.00031517323259579064,
+      "loss": 3.3917,
       "step": 44100
     },
     {
-      "epoch": 4.751910450974061,
-      "grad_norm": 0.6852726340293884,
-      "learning_rate": 0.000315347484107316,
-      "loss": 3.3766,
+      "epoch": 4.7601078167115904,
+      "grad_norm": 0.6750602126121521,
+      "learning_rate": 0.00031484943335132213,
+      "loss": 3.4017,
       "step": 44150
     },
     {
-      "epoch": 4.757292003013669,
-      "grad_norm": 0.7094383835792542,
-      "learning_rate": 0.0003150242430772546,
-      "loss": 3.3928,
+      "epoch": 4.765498652291106,
+      "grad_norm": 0.6329632997512817,
+      "learning_rate": 0.0003145256341068537,
+      "loss": 3.3982,
       "step": 44200
     },
     {
-      "epoch": 4.7626735550532775,
-      "grad_norm": 0.7322360873222351,
-      "learning_rate": 0.00031470100204719317,
-      "loss": 3.3771,
+      "epoch": 4.77088948787062,
+      "grad_norm": 0.6389133930206299,
+      "learning_rate": 0.0003142018348623853,
+      "loss": 3.358,
       "step": 44250
     },
     {
-      "epoch": 4.768055107092886,
-      "grad_norm": 0.6819632649421692,
-      "learning_rate": 0.00031437776101713176,
-      "loss": 3.4,
+      "epoch": 4.776280323450135,
+      "grad_norm": 0.6795648336410522,
+      "learning_rate": 0.00031387803561791684,
+      "loss": 3.3852,
       "step": 44300
     },
     {
-      "epoch": 4.773436659132494,
-      "grad_norm": 0.7061000466346741,
-      "learning_rate": 0.0003140545199870703,
-      "loss": 3.3975,
+      "epoch": 4.781671159029649,
+      "grad_norm": 0.6384751796722412,
+      "learning_rate": 0.00031355423637344844,
+      "loss": 3.3896,
       "step": 44350
     },
     {
-      "epoch": 4.778818211172102,
-      "grad_norm": 0.6955307126045227,
-      "learning_rate": 0.0003137312789570089,
-      "loss": 3.384,
+      "epoch": 4.787061994609164,
+      "grad_norm": 0.6184129118919373,
+      "learning_rate": 0.00031323043712898,
+      "loss": 3.3736,
       "step": 44400
     },
     {
-      "epoch": 4.78419976321171,
-      "grad_norm": 0.659832239151001,
-      "learning_rate": 0.00031340803792694754,
-      "loss": 3.3893,
+      "epoch": 4.7924528301886795,
+      "grad_norm": 0.6460205316543579,
+      "learning_rate": 0.0003129066378845116,
+      "loss": 3.3961,
       "step": 44450
     },
     {
-      "epoch": 4.7895813152513185,
-      "grad_norm": 0.6707594394683838,
-      "learning_rate": 0.0003130847968968861,
-      "loss": 3.3847,
+      "epoch": 4.797843665768194,
+      "grad_norm": 0.6743592023849487,
+      "learning_rate": 0.00031258283864004315,
+      "loss": 3.3942,
       "step": 44500
     },
     {
-      "epoch": 4.794962867290927,
-      "grad_norm": 0.6737433671951294,
-      "learning_rate": 0.0003127615558668247,
-      "loss": 3.3815,
+      "epoch": 4.803234501347709,
+      "grad_norm": 0.6539885401725769,
+      "learning_rate": 0.0003122590393955747,
+      "loss": 3.3671,
       "step": 44550
     },
     {
-      "epoch": 4.800344419330535,
-      "grad_norm": 0.6684089303016663,
-      "learning_rate": 0.0003124383148367633,
-      "loss": 3.3777,
+      "epoch": 4.808625336927224,
+      "grad_norm": 0.6482704281806946,
+      "learning_rate": 0.0003119352401511063,
+      "loss": 3.3842,
       "step": 44600
     },
     {
-      "epoch": 4.805725971370143,
-      "grad_norm": 0.6295944452285767,
-      "learning_rate": 0.0003121150738067018,
-      "loss": 3.3759,
+      "epoch": 4.814016172506738,
+      "grad_norm": 0.6400067806243896,
+      "learning_rate": 0.00031161144090663786,
+      "loss": 3.3818,
       "step": 44650
     },
     {
-      "epoch": 4.811107523409751,
-      "grad_norm": 0.6227767467498779,
-      "learning_rate": 0.0003117918327766404,
-      "loss": 3.3892,
+      "epoch": 4.819407008086253,
+      "grad_norm": 0.6543363928794861,
+      "learning_rate": 0.00031128764166216946,
+      "loss": 3.3819,
       "step": 44700
     },
     {
-      "epoch": 4.8164890754493594,
-      "grad_norm": 0.7158105969429016,
-      "learning_rate": 0.00031146859174657906,
-      "loss": 3.3923,
+      "epoch": 4.824797843665769,
+      "grad_norm": 0.6683675050735474,
+      "learning_rate": 0.00031096384241770096,
+      "loss": 3.397,
       "step": 44750
     },
     {
-      "epoch": 4.821870627488968,
-      "grad_norm": 0.6545218229293823,
-      "learning_rate": 0.0003111453507165176,
-      "loss": 3.3871,
+      "epoch": 4.830188679245283,
+      "grad_norm": 0.7176012992858887,
+      "learning_rate": 0.0003106400431732326,
+      "loss": 3.3673,
       "step": 44800
     },
     {
-      "epoch": 4.827252179528576,
-      "grad_norm": 0.6520293951034546,
-      "learning_rate": 0.0003108221096864562,
-      "loss": 3.3819,
+      "epoch": 4.835579514824798,
+      "grad_norm": 0.6655737161636353,
+      "learning_rate": 0.0003103227199136535,
+      "loss": 3.382,
       "step": 44850
     },
     {
-      "epoch": 4.832633731568184,
-      "grad_norm": 0.6334442496299744,
-      "learning_rate": 0.00031049886865639473,
-      "loss": 3.375,
+      "epoch": 4.840970350404312,
+      "grad_norm": 0.6277079582214355,
+      "learning_rate": 0.00030999892066918506,
+      "loss": 3.359,
       "step": 44900
     },
     {
-      "epoch": 4.838015283607793,
-      "grad_norm": 0.6272345185279846,
-      "learning_rate": 0.00031017562762633333,
-      "loss": 3.3853,
+      "epoch": 4.846361185983827,
+      "grad_norm": 0.6509829759597778,
+      "learning_rate": 0.0003096751214247166,
+      "loss": 3.3982,
       "step": 44950
     },
     {
-      "epoch": 4.8433968356474,
-      "grad_norm": 0.6148518919944763,
-      "learning_rate": 0.00030985238659627187,
-      "loss": 3.3921,
+      "epoch": 4.8517520215633425,
+      "grad_norm": 0.6577560305595398,
+      "learning_rate": 0.0003093513221802482,
+      "loss": 3.3838,
       "step": 45000
     },
     {
-      "epoch": 4.8433968356474,
-      "eval_accuracy": 0.3802495778563676,
-      "eval_loss": 3.4112040996551514,
-      "eval_runtime": 197.6155,
-      "eval_samples_per_second": 91.142,
-      "eval_steps_per_second": 5.698,
+      "epoch": 4.8517520215633425,
+      "eval_accuracy": 0.3807657877263742,
+      "eval_loss": 3.4097847938537598,
+      "eval_runtime": 183.7161,
+      "eval_samples_per_second": 98.037,
+      "eval_steps_per_second": 6.129,
       "step": 45000
     },
     {
-      "epoch": 4.848778387687009,
-      "grad_norm": 0.6464148163795471,
-      "learning_rate": 0.0003095291455662105,
-      "loss": 3.3751,
+      "epoch": 4.857142857142857,
+      "grad_norm": 0.6641700863838196,
+      "learning_rate": 0.00030902752293577977,
+      "loss": 3.3937,
       "step": 45050
     },
     {
-      "epoch": 4.8541599397266175,
-      "grad_norm": 0.6594602465629578,
-      "learning_rate": 0.0003092059045361491,
-      "loss": 3.3842,
+      "epoch": 4.862533692722372,
+      "grad_norm": 0.6280818581581116,
+      "learning_rate": 0.0003087037236913114,
+      "loss": 3.3982,
       "step": 45100
     },
     {
-      "epoch": 4.859541491766225,
-      "grad_norm": 0.7254953384399414,
-      "learning_rate": 0.00030888266350608765,
-      "loss": 3.394,
+      "epoch": 4.867924528301887,
+      "grad_norm": 0.6783115863800049,
+      "learning_rate": 0.0003083799244468429,
+      "loss": 3.3668,
       "step": 45150
     },
     {
-      "epoch": 4.864923043805834,
-      "grad_norm": 0.6902217864990234,
-      "learning_rate": 0.00030855942247602625,
-      "loss": 3.4016,
+      "epoch": 4.873315363881401,
+      "grad_norm": 0.6354468464851379,
+      "learning_rate": 0.00030805612520237453,
+      "loss": 3.3758,
       "step": 45200
     },
     {
-      "epoch": 4.870304595845441,
-      "grad_norm": 0.6273921132087708,
-      "learning_rate": 0.00030823618144596484,
-      "loss": 3.3622,
+      "epoch": 4.878706199460916,
+      "grad_norm": 0.6852909922599792,
+      "learning_rate": 0.0003077323259579061,
+      "loss": 3.3916,
       "step": 45250
     },
     {
-      "epoch": 4.87568614788505,
-      "grad_norm": 0.6722912192344666,
-      "learning_rate": 0.0003079129404159035,
-      "loss": 3.4006,
+      "epoch": 4.884097035040432,
+      "grad_norm": 0.6476108431816101,
+      "learning_rate": 0.00030740852671343763,
+      "loss": 3.4026,
       "step": 45300
     },
     {
-      "epoch": 4.8810676999246585,
-      "grad_norm": 0.6417133808135986,
-      "learning_rate": 0.00030758969938584203,
-      "loss": 3.4034,
+      "epoch": 4.889487870619946,
+      "grad_norm": 0.6892932653427124,
+      "learning_rate": 0.00030708472746896924,
+      "loss": 3.3816,
       "step": 45350
     },
     {
-      "epoch": 4.886449251964266,
-      "grad_norm": 0.688738226890564,
-      "learning_rate": 0.0003072664583557806,
-      "loss": 3.389,
+      "epoch": 4.894878706199461,
+      "grad_norm": 0.6625335216522217,
+      "learning_rate": 0.00030676092822450073,
+      "loss": 3.3913,
       "step": 45400
     },
     {
-      "epoch": 4.891830804003875,
-      "grad_norm": 0.6518829464912415,
-      "learning_rate": 0.00030694321732571917,
-      "loss": 3.3659,
+      "epoch": 4.900269541778976,
+      "grad_norm": 0.6800249218940735,
+      "learning_rate": 0.0003064371289800324,
+      "loss": 3.3865,
       "step": 45450
     },
     {
-      "epoch": 4.897212356043483,
-      "grad_norm": 0.6703770160675049,
-      "learning_rate": 0.00030661997629565776,
-      "loss": 3.3799,
+      "epoch": 4.90566037735849,
+      "grad_norm": 0.6753106713294983,
+      "learning_rate": 0.0003061133297355639,
+      "loss": 3.3862,
       "step": 45500
     },
     {
-      "epoch": 4.902593908083091,
-      "grad_norm": 0.6515799760818481,
-      "learning_rate": 0.0003062967352655963,
-      "loss": 3.3737,
+      "epoch": 4.9110512129380055,
+      "grad_norm": 0.6245962977409363,
+      "learning_rate": 0.0003057895304910955,
+      "loss": 3.3872,
       "step": 45550
     },
     {
-      "epoch": 4.9079754601226995,
-      "grad_norm": 0.6656754016876221,
-      "learning_rate": 0.00030597349423553495,
-      "loss": 3.3903,
+      "epoch": 4.916442048517521,
+      "grad_norm": 0.6651871204376221,
+      "learning_rate": 0.00030546573124662704,
+      "loss": 3.3808,
       "step": 45600
     },
     {
-      "epoch": 4.913357012162308,
-      "grad_norm": 0.6744281053543091,
-      "learning_rate": 0.00030565025320547355,
-      "loss": 3.3796,
+      "epoch": 4.921832884097035,
+      "grad_norm": 0.6639772057533264,
+      "learning_rate": 0.00030514193200215865,
+      "loss": 3.3856,
       "step": 45650
     },
     {
-      "epoch": 4.918738564201916,
-      "grad_norm": 0.6151371002197266,
-      "learning_rate": 0.0003053270121754121,
-      "loss": 3.3471,
+      "epoch": 4.92722371967655,
+      "grad_norm": 0.6940948367118835,
+      "learning_rate": 0.0003048181327576902,
+      "loss": 3.3611,
       "step": 45700
     },
     {
-      "epoch": 4.924120116241524,
-      "grad_norm": 0.6114195585250854,
-      "learning_rate": 0.0003050037711453507,
-      "loss": 3.3809,
+      "epoch": 4.932614555256064,
+      "grad_norm": 0.6986059546470642,
+      "learning_rate": 0.00030449433351322175,
+      "loss": 3.3792,
       "step": 45750
     },
     {
-      "epoch": 4.929501668281132,
-      "grad_norm": 0.6933553814888,
-      "learning_rate": 0.0003046805301152893,
-      "loss": 3.3861,
+      "epoch": 4.938005390835579,
+      "grad_norm": 0.7022907137870789,
+      "learning_rate": 0.00030417053426875335,
+      "loss": 3.3989,
       "step": 45800
     },
     {
-      "epoch": 4.9348832203207404,
-      "grad_norm": 0.6847333312034607,
-      "learning_rate": 0.0003043572890852278,
-      "loss": 3.383,
+      "epoch": 4.943396226415095,
+      "grad_norm": 0.7218758463859558,
+      "learning_rate": 0.0003038467350242849,
+      "loss": 3.3752,
       "step": 45850
     },
     {
-      "epoch": 4.940264772360349,
-      "grad_norm": 0.6658839583396912,
-      "learning_rate": 0.0003040405128757677,
-      "loss": 3.3774,
+      "epoch": 4.948787061994609,
+      "grad_norm": 0.6828824877738953,
+      "learning_rate": 0.0003035229357798165,
+      "loss": 3.3848,
       "step": 45900
     },
     {
-      "epoch": 4.945646324399957,
-      "grad_norm": 0.6456676721572876,
-      "learning_rate": 0.0003037172718457062,
-      "loss": 3.393,
+      "epoch": 4.954177897574124,
+      "grad_norm": 0.6036968231201172,
+      "learning_rate": 0.00030319913653534806,
+      "loss": 3.3755,
       "step": 45950
     },
     {
-      "epoch": 4.951027876439565,
-      "grad_norm": 0.6581215262413025,
-      "learning_rate": 0.00030339403081564487,
-      "loss": 3.3767,
+      "epoch": 4.959568733153639,
+      "grad_norm": 0.6400105953216553,
+      "learning_rate": 0.00030287533729087966,
+      "loss": 3.3829,
       "step": 46000
     },
     {
-      "epoch": 4.951027876439565,
-      "eval_accuracy": 0.3806143256007608,
-      "eval_loss": 3.4063382148742676,
-      "eval_runtime": 198.5903,
-      "eval_samples_per_second": 90.694,
-      "eval_steps_per_second": 5.67,
+      "epoch": 4.959568733153639,
+      "eval_accuracy": 0.3808671608707051,
+      "eval_loss": 3.409538984298706,
+      "eval_runtime": 184.193,
+      "eval_samples_per_second": 97.783,
+      "eval_steps_per_second": 6.113,
       "step": 46000
     },
     {
-      "epoch": 4.956409428479174,
-      "grad_norm": 0.6951839327812195,
-      "learning_rate": 0.00030307078978558346,
-      "loss": 3.3863,
+      "epoch": 4.964959568733153,
+      "grad_norm": 0.6539743542671204,
+      "learning_rate": 0.0003025515380464112,
+      "loss": 3.3932,
       "step": 46050
     },
     {
-      "epoch": 4.961790980518781,
-      "grad_norm": 0.6826249957084656,
-      "learning_rate": 0.000302747548755522,
-      "loss": 3.3802,
+      "epoch": 4.9703504043126685,
+      "grad_norm": 0.6177482604980469,
+      "learning_rate": 0.0003022277388019428,
+      "loss": 3.3622,
       "step": 46100
     },
     {
-      "epoch": 4.96717253255839,
-      "grad_norm": 0.7239309549331665,
-      "learning_rate": 0.0003024243077254606,
-      "loss": 3.3635,
+      "epoch": 4.975741239892184,
+      "grad_norm": 0.6319087147712708,
+      "learning_rate": 0.0003019039395574743,
+      "loss": 3.3702,
       "step": 46150
     },
     {
-      "epoch": 4.9725540845979985,
-      "grad_norm": 0.7058936357498169,
-      "learning_rate": 0.00030210106669539914,
-      "loss": 3.3733,
+      "epoch": 4.981132075471698,
+      "grad_norm": 0.6719312071800232,
+      "learning_rate": 0.00030158014031300587,
+      "loss": 3.3815,
       "step": 46200
     },
     {
-      "epoch": 4.977935636637606,
-      "grad_norm": 0.6981608271598816,
-      "learning_rate": 0.0003017778256653378,
-      "loss": 3.3847,
+      "epoch": 4.986522911051213,
+      "grad_norm": 0.6484641432762146,
+      "learning_rate": 0.00030125634106853747,
+      "loss": 3.3979,
       "step": 46250
     },
     {
-      "epoch": 4.983317188677215,
-      "grad_norm": 0.6583619117736816,
-      "learning_rate": 0.0003014545846352764,
-      "loss": 3.3835,
+      "epoch": 4.991913746630727,
+      "grad_norm": 0.6167640089988708,
+      "learning_rate": 0.000300932541824069,
+      "loss": 3.3947,
       "step": 46300
     },
     {
-      "epoch": 4.988698740716822,
-      "grad_norm": 0.701775848865509,
-      "learning_rate": 0.0003011313436052149,
-      "loss": 3.3868,
+      "epoch": 4.997304582210242,
+      "grad_norm": 0.6704341769218445,
+      "learning_rate": 0.0003006087425796006,
+      "loss": 3.3725,
       "step": 46350
     },
     {
-      "epoch": 4.994080292756431,
-      "grad_norm": 0.6978148818016052,
-      "learning_rate": 0.0003008081025751535,
-      "loss": 3.3632,
+      "epoch": 5.002695417789758,
+      "grad_norm": 0.6638703942298889,
+      "learning_rate": 0.0003002849433351322,
+      "loss": 3.3222,
       "step": 46400
     },
     {
-      "epoch": 4.9994618447960395,
-      "grad_norm": 0.6526729464530945,
-      "learning_rate": 0.00030048486154509205,
-      "loss": 3.3829,
+      "epoch": 5.008086253369272,
+      "grad_norm": 0.7498042583465576,
+      "learning_rate": 0.00029996114409066373,
+      "loss": 3.2995,
       "step": 46450
     },
     {
-      "epoch": 5.004843396835647,
-      "grad_norm": 0.7245412468910217,
-      "learning_rate": 0.00030016162051503065,
+      "epoch": 5.013477088948787,
+      "grad_norm": 0.6590642333030701,
+      "learning_rate": 0.00029963734484619533,
       "loss": 3.2992,
       "step": 46500
     },
     {
-      "epoch": 5.010224948875256,
-      "grad_norm": 0.6889564394950867,
-      "learning_rate": 0.00029983837948496924,
-      "loss": 3.2678,
+      "epoch": 5.018867924528302,
+      "grad_norm": 0.6739321947097778,
+      "learning_rate": 0.0002993135456017269,
+      "loss": 3.2998,
       "step": 46550
     },
     {
-      "epoch": 5.015606500914864,
-      "grad_norm": 0.6887182593345642,
-      "learning_rate": 0.0002995151384549079,
-      "loss": 3.3038,
+      "epoch": 5.024258760107816,
+      "grad_norm": 0.6751723289489746,
+      "learning_rate": 0.0002989897463572585,
+      "loss": 3.2851,
       "step": 46600
     },
     {
-      "epoch": 5.020988052954472,
-      "grad_norm": 0.6541692614555359,
-      "learning_rate": 0.00029919189742484643,
-      "loss": 3.2812,
+      "epoch": 5.0296495956873315,
+      "grad_norm": 0.7554724812507629,
+      "learning_rate": 0.00029866594711279004,
+      "loss": 3.2914,
       "step": 46650
     },
     {
-      "epoch": 5.0263696049940805,
-      "grad_norm": 0.7237575650215149,
-      "learning_rate": 0.00029886865639478503,
-      "loss": 3.2945,
+      "epoch": 5.035040431266847,
+      "grad_norm": 0.6768006086349487,
+      "learning_rate": 0.00029834214786832164,
+      "loss": 3.2929,
       "step": 46700
     },
     {
-      "epoch": 5.031751157033688,
-      "grad_norm": 0.7217345833778381,
-      "learning_rate": 0.0002985454153647236,
-      "loss": 3.2973,
+      "epoch": 5.040431266846361,
+      "grad_norm": 0.699368953704834,
+      "learning_rate": 0.0002980183486238532,
+      "loss": 3.2966,
       "step": 46750
     },
     {
-      "epoch": 5.037132709073297,
-      "grad_norm": 0.7119247913360596,
-      "learning_rate": 0.00029822217433466216,
-      "loss": 3.2822,
+      "epoch": 5.045822102425876,
+      "grad_norm": 0.6774499416351318,
+      "learning_rate": 0.00029769454937938474,
+      "loss": 3.3041,
       "step": 46800
     },
     {
-      "epoch": 5.042514261112905,
-      "grad_norm": 0.715711772441864,
-      "learning_rate": 0.00029789893330460076,
-      "loss": 3.2914,
+      "epoch": 5.051212938005391,
+      "grad_norm": 0.6650752425193787,
+      "learning_rate": 0.0002973707501349163,
+      "loss": 3.3062,
       "step": 46850
     },
     {
-      "epoch": 5.047895813152513,
-      "grad_norm": 0.735707700252533,
-      "learning_rate": 0.00029757569227453935,
-      "loss": 3.2931,
+      "epoch": 5.056603773584905,
+      "grad_norm": 0.662289023399353,
+      "learning_rate": 0.0002970469508904479,
+      "loss": 3.3009,
       "step": 46900
     },
     {
-      "epoch": 5.0532773651921215,
-      "grad_norm": 0.6921928524971008,
-      "learning_rate": 0.00029725245124447795,
-      "loss": 3.3198,
+      "epoch": 5.061994609164421,
+      "grad_norm": 0.7000959515571594,
+      "learning_rate": 0.00029672315164597945,
+      "loss": 3.2971,
       "step": 46950
     },
     {
-      "epoch": 5.05865891723173,
-      "grad_norm": 0.6644381880760193,
-      "learning_rate": 0.0002969292102144165,
-      "loss": 3.303,
+      "epoch": 5.067385444743936,
+      "grad_norm": 0.7076292634010315,
+      "learning_rate": 0.00029639935240151105,
+      "loss": 3.3085,
       "step": 47000
     },
     {
-      "epoch": 5.05865891723173,
-      "eval_accuracy": 0.3808069671708559,
-      "eval_loss": 3.408489942550659,
-      "eval_runtime": 186.9297,
-      "eval_samples_per_second": 96.352,
-      "eval_steps_per_second": 6.024,
+      "epoch": 5.067385444743936,
+      "eval_accuracy": 0.38101362496347907,
+      "eval_loss": 3.40975284576416,
+      "eval_runtime": 183.5541,
+      "eval_samples_per_second": 98.124,
+      "eval_steps_per_second": 6.134,
       "step": 47000
     },
     {
-      "epoch": 5.064040469271338,
-      "grad_norm": 0.7082266211509705,
-      "learning_rate": 0.00029660596918435514,
-      "loss": 3.2757,
+      "epoch": 5.07277628032345,
+      "grad_norm": 0.7516024708747864,
+      "learning_rate": 0.0002960755531570426,
+      "loss": 3.3012,
       "step": 47050
     },
     {
-      "epoch": 5.069422021310946,
-      "grad_norm": 0.6677634119987488,
-      "learning_rate": 0.0002962827281542937,
-      "loss": 3.2786,
+      "epoch": 5.078167115902965,
+      "grad_norm": 0.6497899293899536,
+      "learning_rate": 0.0002957517539125742,
+      "loss": 3.296,
       "step": 47100
     },
     {
-      "epoch": 5.074803573350554,
-      "grad_norm": 0.6854336261749268,
-      "learning_rate": 0.00029595948712423227,
-      "loss": 3.3234,
+      "epoch": 5.083557951482479,
+      "grad_norm": 0.6259191036224365,
+      "learning_rate": 0.00029542795466810576,
+      "loss": 3.3085,
       "step": 47150
     },
     {
-      "epoch": 5.080185125390162,
-      "grad_norm": 0.720350444316864,
-      "learning_rate": 0.00029563624609417087,
-      "loss": 3.2982,
+      "epoch": 5.0889487870619945,
+      "grad_norm": 0.6975776553153992,
+      "learning_rate": 0.0002951041554236373,
+      "loss": 3.308,
       "step": 47200
     },
     {
-      "epoch": 5.085566677429771,
-      "grad_norm": 0.6741048693656921,
-      "learning_rate": 0.00029531300506410946,
-      "loss": 3.3021,
+      "epoch": 5.09433962264151,
+      "grad_norm": 0.6593020558357239,
+      "learning_rate": 0.00029478035617916886,
+      "loss": 3.2919,
       "step": 47250
     },
     {
-      "epoch": 5.090948229469379,
-      "grad_norm": 0.6871724128723145,
-      "learning_rate": 0.000294989764034048,
-      "loss": 3.3146,
+      "epoch": 5.099730458221024,
+      "grad_norm": 0.7125478982925415,
+      "learning_rate": 0.00029445655693470047,
+      "loss": 3.316,
       "step": 47300
     },
     {
-      "epoch": 5.096329781508987,
-      "grad_norm": 0.6869731545448303,
-      "learning_rate": 0.0002946665230039866,
-      "loss": 3.3125,
+      "epoch": 5.105121293800539,
+      "grad_norm": 0.6814964413642883,
+      "learning_rate": 0.000294132757690232,
+      "loss": 3.3162,
       "step": 47350
     },
     {
-      "epoch": 5.101711333548596,
-      "grad_norm": 0.6518433690071106,
-      "learning_rate": 0.0002943432819739252,
-      "loss": 3.3169,
+      "epoch": 5.110512129380054,
+      "grad_norm": 0.682304322719574,
+      "learning_rate": 0.0002938089584457636,
+      "loss": 3.297,
       "step": 47400
     },
     {
-      "epoch": 5.107092885588203,
-      "grad_norm": 0.6880452632904053,
-      "learning_rate": 0.0002940200409438638,
-      "loss": 3.3051,
+      "epoch": 5.115902964959568,
+      "grad_norm": 0.680831253528595,
+      "learning_rate": 0.00029348515920129517,
+      "loss": 3.302,
       "step": 47450
     },
     {
-      "epoch": 5.112474437627812,
-      "grad_norm": 0.6545599102973938,
-      "learning_rate": 0.0002936967999138024,
-      "loss": 3.3126,
+      "epoch": 5.121293800539084,
+      "grad_norm": 0.706608772277832,
+      "learning_rate": 0.0002931613599568267,
+      "loss": 3.2928,
       "step": 47500
     },
     {
-      "epoch": 5.1178559896674205,
-      "grad_norm": 0.6959748864173889,
-      "learning_rate": 0.0002933735588837409,
-      "loss": 3.3036,
+      "epoch": 5.126684636118599,
+      "grad_norm": 0.6527232527732849,
+      "learning_rate": 0.00029283756071235833,
+      "loss": 3.2992,
       "step": 47550
     },
     {
-      "epoch": 5.123237541707028,
-      "grad_norm": 0.6538510918617249,
-      "learning_rate": 0.00029305031785367957,
-      "loss": 3.2978,
+      "epoch": 5.132075471698113,
+      "grad_norm": 0.6953222155570984,
+      "learning_rate": 0.0002925137614678899,
+      "loss": 3.2894,
       "step": 47600
     },
     {
-      "epoch": 5.128619093746637,
-      "grad_norm": 0.6865313649177551,
-      "learning_rate": 0.0002927270768236181,
-      "loss": 3.3145,
+      "epoch": 5.137466307277628,
+      "grad_norm": 0.6943708658218384,
+      "learning_rate": 0.00029218996222342143,
+      "loss": 3.3042,
       "step": 47650
     },
     {
-      "epoch": 5.134000645786244,
-      "grad_norm": 0.7012115120887756,
-      "learning_rate": 0.0002924038357935567,
-      "loss": 3.2994,
+      "epoch": 5.142857142857143,
+      "grad_norm": 0.670811116695404,
+      "learning_rate": 0.00029186616297895303,
+      "loss": 3.319,
       "step": 47700
     },
     {
-      "epoch": 5.139382197825853,
-      "grad_norm": 0.6826618909835815,
-      "learning_rate": 0.0002920805947634953,
-      "loss": 3.3151,
+      "epoch": 5.1482479784366575,
+      "grad_norm": 0.6703544855117798,
+      "learning_rate": 0.0002915423637344846,
+      "loss": 3.3078,
       "step": 47750
     },
     {
-      "epoch": 5.1447637498654615,
-      "grad_norm": 0.6896260976791382,
-      "learning_rate": 0.0002917573537334339,
-      "loss": 3.3182,
+      "epoch": 5.153638814016173,
+      "grad_norm": 0.6404094099998474,
+      "learning_rate": 0.00029121856449001613,
+      "loss": 3.2941,
       "step": 47800
     },
     {
-      "epoch": 5.150145301905069,
-      "grad_norm": 0.6824194192886353,
-      "learning_rate": 0.00029143411270337243,
-      "loss": 3.3107,
+      "epoch": 5.159029649595688,
+      "grad_norm": 0.6793299317359924,
+      "learning_rate": 0.00029089476524554774,
+      "loss": 3.3025,
       "step": 47850
     },
     {
-      "epoch": 5.155526853944678,
-      "grad_norm": 0.7328106760978699,
-      "learning_rate": 0.00029111087167331103,
-      "loss": 3.3068,
+      "epoch": 5.164420485175202,
+      "grad_norm": 0.7005292177200317,
+      "learning_rate": 0.0002905709660010793,
+      "loss": 3.3268,
       "step": 47900
     },
     {
-      "epoch": 5.160908405984286,
-      "grad_norm": 0.7673147320747375,
-      "learning_rate": 0.0002907876306432496,
-      "loss": 3.3058,
+      "epoch": 5.169811320754717,
+      "grad_norm": 0.6688427329063416,
+      "learning_rate": 0.0002902471667566109,
+      "loss": 3.3082,
       "step": 47950
     },
     {
-      "epoch": 5.166289958023894,
-      "grad_norm": 0.675337016582489,
-      "learning_rate": 0.00029047085443378943,
-      "loss": 3.3034,
+      "epoch": 5.175202156334231,
+      "grad_norm": 0.6847968101501465,
+      "learning_rate": 0.00028992336751214245,
+      "loss": 3.302,
       "step": 48000
     },
     {
-      "epoch": 5.166289958023894,
-      "eval_accuracy": 0.38097027246124104,
-      "eval_loss": 3.4084055423736572,
-      "eval_runtime": 188.5486,
-      "eval_samples_per_second": 95.524,
-      "eval_steps_per_second": 5.972,
+      "epoch": 5.175202156334231,
+      "eval_accuracy": 0.38124233929232854,
+      "eval_loss": 3.4091999530792236,
+      "eval_runtime": 183.6086,
+      "eval_samples_per_second": 98.095,
+      "eval_steps_per_second": 6.133,
       "step": 48000
     },
     {
-      "epoch": 5.1716715100635025,
-      "grad_norm": 0.7029422521591187,
-      "learning_rate": 0.000290147613403728,
-      "loss": 3.3148,
+      "epoch": 5.180592991913747,
+      "grad_norm": 0.6164307594299316,
+      "learning_rate": 0.00028959956826767405,
+      "loss": 3.308,
       "step": 48050
     },
     {
-      "epoch": 5.17705306210311,
-      "grad_norm": 0.6556596159934998,
-      "learning_rate": 0.00028982437237366656,
-      "loss": 3.3102,
+      "epoch": 5.185983827493262,
+      "grad_norm": 0.66629558801651,
+      "learning_rate": 0.00028927576902320555,
+      "loss": 3.2981,
       "step": 48100
     },
     {
-      "epoch": 5.182434614142719,
-      "grad_norm": 0.7232239246368408,
-      "learning_rate": 0.0002895011313436052,
-      "loss": 3.3241,
+      "epoch": 5.191374663072776,
+      "grad_norm": 0.7151822447776794,
+      "learning_rate": 0.00028895196977873715,
+      "loss": 3.3072,
       "step": 48150
     },
     {
-      "epoch": 5.187816166182327,
-      "grad_norm": 0.7116256356239319,
-      "learning_rate": 0.00028917789031354375,
-      "loss": 3.3023,
+      "epoch": 5.196765498652291,
+      "grad_norm": 0.6297134160995483,
+      "learning_rate": 0.0002886281705342687,
+      "loss": 3.321,
       "step": 48200
     },
     {
-      "epoch": 5.193197718221935,
-      "grad_norm": 0.6975775361061096,
-      "learning_rate": 0.00028885464928348235,
-      "loss": 3.3159,
+      "epoch": 5.202156334231806,
+      "grad_norm": 0.7029529809951782,
+      "learning_rate": 0.0002883043712898003,
+      "loss": 3.314,
       "step": 48250
     },
     {
-      "epoch": 5.198579270261543,
-      "grad_norm": 0.6999067664146423,
-      "learning_rate": 0.00028853140825342094,
-      "loss": 3.3139,
+      "epoch": 5.2075471698113205,
+      "grad_norm": 0.7096197605133057,
+      "learning_rate": 0.00028798057204533186,
+      "loss": 3.3148,
       "step": 48300
     },
     {
-      "epoch": 5.203960822301152,
-      "grad_norm": 0.6723634004592896,
-      "learning_rate": 0.00028820816722335954,
-      "loss": 3.3282,
+      "epoch": 5.212938005390836,
+      "grad_norm": 0.7064436674118042,
+      "learning_rate": 0.00028765677280086346,
+      "loss": 3.3236,
       "step": 48350
     },
     {
-      "epoch": 5.20934237434076,
-      "grad_norm": 0.6788647174835205,
-      "learning_rate": 0.00028788492619329813,
-      "loss": 3.3046,
+      "epoch": 5.218328840970351,
+      "grad_norm": 0.7410140037536621,
+      "learning_rate": 0.000287332973556395,
+      "loss": 3.3166,
       "step": 48400
     },
     {
-      "epoch": 5.214723926380368,
-      "grad_norm": 0.6587629318237305,
-      "learning_rate": 0.0002875616851632367,
-      "loss": 3.2944,
+      "epoch": 5.223719676549865,
+      "grad_norm": 0.6449042558670044,
+      "learning_rate": 0.0002870091743119266,
+      "loss": 3.3069,
       "step": 48450
     },
     {
-      "epoch": 5.220105478419977,
-      "grad_norm": 0.7028741240501404,
-      "learning_rate": 0.00028723844413317527,
-      "loss": 3.313,
+      "epoch": 5.22911051212938,
+      "grad_norm": 0.6682385206222534,
+      "learning_rate": 0.00028668537506745817,
+      "loss": 3.3408,
       "step": 48500
     },
     {
-      "epoch": 5.225487030459584,
-      "grad_norm": 0.6964418292045593,
-      "learning_rate": 0.00028691520310311386,
-      "loss": 3.3077,
+      "epoch": 5.234501347708895,
+      "grad_norm": 0.6333697438240051,
+      "learning_rate": 0.0002863615758229897,
+      "loss": 3.3153,
       "step": 48550
     },
     {
-      "epoch": 5.230868582499193,
-      "grad_norm": 0.6499736905097961,
-      "learning_rate": 0.00028659196207305246,
-      "loss": 3.3179,
+      "epoch": 5.2398921832884096,
+      "grad_norm": 0.6397345066070557,
+      "learning_rate": 0.00028603777657852127,
+      "loss": 3.3148,
       "step": 48600
     },
     {
-      "epoch": 5.236250134538801,
-      "grad_norm": 0.7362401485443115,
-      "learning_rate": 0.000286268721042991,
-      "loss": 3.3096,
+      "epoch": 5.245283018867925,
+      "grad_norm": 0.6702921986579895,
+      "learning_rate": 0.0002857139773340529,
+      "loss": 3.3136,
       "step": 48650
     },
     {
-      "epoch": 5.241631686578409,
-      "grad_norm": 0.6766545176506042,
-      "learning_rate": 0.00028594548001292965,
-      "loss": 3.3152,
+      "epoch": 5.250673854447439,
+      "grad_norm": 0.7235493659973145,
+      "learning_rate": 0.0002853901780895844,
+      "loss": 3.3155,
       "step": 48700
     },
     {
-      "epoch": 5.247013238618018,
-      "grad_norm": 0.6662083268165588,
-      "learning_rate": 0.0002856222389828682,
-      "loss": 3.3175,
+      "epoch": 5.256064690026954,
+      "grad_norm": 0.6859371662139893,
+      "learning_rate": 0.00028506637884511603,
+      "loss": 3.3356,
       "step": 48750
     },
     {
-      "epoch": 5.252394790657625,
-      "grad_norm": 0.6704933643341064,
-      "learning_rate": 0.0002852989979528068,
-      "loss": 3.3085,
+      "epoch": 5.261455525606469,
+      "grad_norm": 0.6729069948196411,
+      "learning_rate": 0.0002847425796006476,
+      "loss": 3.3336,
       "step": 48800
     },
     {
-      "epoch": 5.257776342697234,
-      "grad_norm": 0.6517661213874817,
-      "learning_rate": 0.0002849757569227454,
-      "loss": 3.2974,
+      "epoch": 5.2668463611859835,
+      "grad_norm": 0.6778712868690491,
+      "learning_rate": 0.0002844252563410685,
+      "loss": 3.3111,
       "step": 48850
     },
     {
-      "epoch": 5.2631578947368425,
-      "grad_norm": 0.8727652430534363,
-      "learning_rate": 0.00028465251589268397,
-      "loss": 3.3257,
+      "epoch": 5.272237196765499,
+      "grad_norm": 0.7014955282211304,
+      "learning_rate": 0.0002841014570966001,
+      "loss": 3.311,
       "step": 48900
     },
     {
-      "epoch": 5.26853944677645,
-      "grad_norm": 0.6727677583694458,
-      "learning_rate": 0.0002843292748626225,
-      "loss": 3.3271,
+      "epoch": 5.277628032345014,
+      "grad_norm": 0.6397571563720703,
+      "learning_rate": 0.00028377765785213163,
+      "loss": 3.3331,
       "step": 48950
     },
     {
-      "epoch": 5.273920998816059,
-      "grad_norm": 0.6924176216125488,
-      "learning_rate": 0.0002840060338325611,
-      "loss": 3.32,
+      "epoch": 5.283018867924528,
+      "grad_norm": 0.6588596105575562,
+      "learning_rate": 0.00028345385860766324,
+      "loss": 3.3225,
       "step": 49000
     },
     {
-      "epoch": 5.273920998816059,
-      "eval_accuracy": 0.3817201946929366,
-      "eval_loss": 3.4039976596832275,
-      "eval_runtime": 190.227,
-      "eval_samples_per_second": 94.682,
-      "eval_steps_per_second": 5.919,
+      "epoch": 5.283018867924528,
+      "eval_accuracy": 0.3817110678503602,
+      "eval_loss": 3.4049184322357178,
+      "eval_runtime": 183.6448,
+      "eval_samples_per_second": 98.075,
+      "eval_steps_per_second": 6.131,
       "step": 49000
     },
     {
-      "epoch": 5.279302550855666,
-      "grad_norm": 0.6958600282669067,
-      "learning_rate": 0.0002836827928024997,
-      "loss": 3.3028,
+      "epoch": 5.288409703504043,
+      "grad_norm": 0.6483611464500427,
+      "learning_rate": 0.0002831300593631948,
+      "loss": 3.3306,
       "step": 49050
     },
     {
-      "epoch": 5.284684102895275,
-      "grad_norm": 0.7136051654815674,
-      "learning_rate": 0.0002833660165930395,
-      "loss": 3.3148,
+      "epoch": 5.293800539083558,
+      "grad_norm": 0.6639255285263062,
+      "learning_rate": 0.0002828062601187264,
+      "loss": 3.3269,
       "step": 49100
     },
     {
-      "epoch": 5.2900656549348835,
-      "grad_norm": 0.7401239275932312,
-      "learning_rate": 0.0002830427755629781,
-      "loss": 3.3216,
+      "epoch": 5.2991913746630726,
+      "grad_norm": 0.6533463001251221,
+      "learning_rate": 0.00028248246087425794,
+      "loss": 3.3255,
       "step": 49150
     },
     {
-      "epoch": 5.295447206974491,
-      "grad_norm": 0.7072134017944336,
-      "learning_rate": 0.0002827195345329167,
-      "loss": 3.3095,
+      "epoch": 5.304582210242588,
+      "grad_norm": 0.7285601496696472,
+      "learning_rate": 0.0002821586616297895,
+      "loss": 3.3342,
       "step": 49200
     },
     {
-      "epoch": 5.3008287590141,
-      "grad_norm": 0.6892372369766235,
-      "learning_rate": 0.0002823962935028553,
-      "loss": 3.3046,
+      "epoch": 5.309973045822103,
+      "grad_norm": 0.6505836248397827,
+      "learning_rate": 0.0002818348623853211,
+      "loss": 3.3224,
       "step": 49250
     },
     {
-      "epoch": 5.306210311053708,
-      "grad_norm": 0.6895684599876404,
-      "learning_rate": 0.00028207305247279383,
-      "loss": 3.3206,
+      "epoch": 5.315363881401617,
+      "grad_norm": 0.6910222172737122,
+      "learning_rate": 0.00028151106314085265,
+      "loss": 3.3221,
       "step": 49300
     },
     {
-      "epoch": 5.311591863093316,
-      "grad_norm": 0.6668702960014343,
-      "learning_rate": 0.0002817498114427325,
-      "loss": 3.3246,
+      "epoch": 5.320754716981132,
+      "grad_norm": 0.7005528807640076,
+      "learning_rate": 0.0002811872638963842,
+      "loss": 3.3316,
       "step": 49350
     },
     {
-      "epoch": 5.316973415132924,
-      "grad_norm": 0.7270509600639343,
-      "learning_rate": 0.000281426570412671,
-      "loss": 3.3146,
+      "epoch": 5.3261455525606465,
+      "grad_norm": 0.7789005637168884,
+      "learning_rate": 0.0002808634646519158,
+      "loss": 3.3204,
       "step": 49400
     },
     {
-      "epoch": 5.322354967172533,
-      "grad_norm": 0.7968111634254456,
-      "learning_rate": 0.0002811033293826096,
-      "loss": 3.3385,
+      "epoch": 5.331536388140162,
+      "grad_norm": 0.6945908069610596,
+      "learning_rate": 0.00028053966540744736,
+      "loss": 3.3113,
       "step": 49450
     },
     {
-      "epoch": 5.327736519212141,
-      "grad_norm": 0.6827276349067688,
-      "learning_rate": 0.0002807800883525482,
-      "loss": 3.3272,
+      "epoch": 5.336927223719677,
+      "grad_norm": 0.6795584559440613,
+      "learning_rate": 0.0002802158661629789,
+      "loss": 3.3384,
       "step": 49500
     },
     {
-      "epoch": 5.333118071251749,
-      "grad_norm": 0.8143443465232849,
-      "learning_rate": 0.00028045684732248675,
-      "loss": 3.3041,
+      "epoch": 5.342318059299191,
+      "grad_norm": 0.6610580086708069,
+      "learning_rate": 0.0002798920669185105,
+      "loss": 3.3105,
       "step": 49550
     },
     {
-      "epoch": 5.338499623291357,
-      "grad_norm": 0.7310146689414978,
-      "learning_rate": 0.00028013360629242535,
-      "loss": 3.3331,
+      "epoch": 5.347708894878706,
+      "grad_norm": 0.7066441178321838,
+      "learning_rate": 0.00027956826767404206,
+      "loss": 3.3288,
       "step": 49600
     },
     {
-      "epoch": 5.343881175330965,
-      "grad_norm": 0.6941249370574951,
-      "learning_rate": 0.00027981036526236394,
-      "loss": 3.3381,
+      "epoch": 5.353099730458221,
+      "grad_norm": 0.7495715022087097,
+      "learning_rate": 0.00027924446842957367,
+      "loss": 3.3264,
       "step": 49650
     },
     {
-      "epoch": 5.349262727370574,
-      "grad_norm": 0.6752241253852844,
-      "learning_rate": 0.00027948712423230253,
-      "loss": 3.3184,
+      "epoch": 5.3584905660377355,
+      "grad_norm": 0.6596075296401978,
+      "learning_rate": 0.0002789206691851052,
+      "loss": 3.332,
       "step": 49700
     },
     {
-      "epoch": 5.354644279410182,
-      "grad_norm": 0.6993463635444641,
-      "learning_rate": 0.0002791638832022411,
-      "loss": 3.334,
+      "epoch": 5.363881401617251,
+      "grad_norm": 0.6785242557525635,
+      "learning_rate": 0.00027859686994063677,
+      "loss": 3.3167,
       "step": 49750
     },
     {
-      "epoch": 5.36002583144979,
-      "grad_norm": 0.6740778088569641,
-      "learning_rate": 0.0002788406421721797,
-      "loss": 3.3282,
+      "epoch": 5.369272237196766,
+      "grad_norm": 0.6717910170555115,
+      "learning_rate": 0.0002782730706961683,
+      "loss": 3.3199,
       "step": 49800
     },
     {
-      "epoch": 5.365407383489399,
-      "grad_norm": 0.6712247729301453,
-      "learning_rate": 0.00027851740114211826,
-      "loss": 3.3199,
+      "epoch": 5.37466307277628,
+      "grad_norm": 0.6411001086235046,
+      "learning_rate": 0.0002779492714516999,
+      "loss": 3.3356,
       "step": 49850
     },
     {
-      "epoch": 5.370788935529006,
-      "grad_norm": 0.7134065628051758,
-      "learning_rate": 0.00027819416011205686,
-      "loss": 3.318,
+      "epoch": 5.380053908355795,
+      "grad_norm": NaN,
+      "learning_rate": 0.0002776319481921209,
+      "loss": 3.3302,
       "step": 49900
     },
     {
-      "epoch": 5.376170487568615,
-      "grad_norm": 0.6927216649055481,
-      "learning_rate": 0.00027787091908199545,
-      "loss": 3.3344,
+      "epoch": 5.38544474393531,
+      "grad_norm": 0.6625006198883057,
+      "learning_rate": 0.0002773081489476524,
+      "loss": 3.3146,
       "step": 49950
     },
     {
-      "epoch": 5.3815520396082235,
-      "grad_norm": 0.6422109007835388,
-      "learning_rate": 0.00027754767805193405,
-      "loss": 3.3174,
+      "epoch": 5.390835579514825,
+      "grad_norm": 0.6475867033004761,
+      "learning_rate": 0.00027698434970318403,
+      "loss": 3.3354,
       "step": 50000
     },
     {
-      "epoch": 5.3815520396082235,
-      "eval_accuracy": 0.38204441491017505,
-      "eval_loss": 3.400477170944214,
-      "eval_runtime": 190.5095,
-      "eval_samples_per_second": 94.541,
-      "eval_steps_per_second": 5.91,
+      "epoch": 5.390835579514825,
+      "eval_accuracy": 0.3823062683698079,
+      "eval_loss": 3.396789312362671,
+      "eval_runtime": 183.9015,
+      "eval_samples_per_second": 97.938,
+      "eval_steps_per_second": 6.123,
       "step": 50000
     },
     {
-      "epoch": 5.386933591647831,
-      "grad_norm": 0.7164682745933533,
-      "learning_rate": 0.0002772244370218726,
-      "loss": 3.3089,
+      "epoch": 5.39622641509434,
+      "grad_norm": 0.6825270652770996,
+      "learning_rate": 0.0002766605504587156,
+      "loss": 3.346,
       "step": 50050
     },
     {
-      "epoch": 5.39231514368744,
-      "grad_norm": 0.7046695351600647,
-      "learning_rate": 0.0002769011959918112,
-      "loss": 3.3237,
+      "epoch": 5.401617250673855,
+      "grad_norm": 0.6633207201957703,
+      "learning_rate": 0.00027633675121424713,
+      "loss": 3.335,
       "step": 50100
     },
     {
-      "epoch": 5.397696695727047,
-      "grad_norm": 0.6866887211799622,
-      "learning_rate": 0.0002765779549617498,
-      "loss": 3.3355,
+      "epoch": 5.407008086253369,
+      "grad_norm": 0.6778284311294556,
+      "learning_rate": 0.0002760129519697787,
+      "loss": 3.327,
       "step": 50150
     },
     {
-      "epoch": 5.403078247766656,
-      "grad_norm": 0.6720688343048096,
-      "learning_rate": 0.0002762547139316884,
-      "loss": 3.3342,
+      "epoch": 5.412398921832884,
+      "grad_norm": 0.6515063643455505,
+      "learning_rate": 0.0002756891527253103,
+      "loss": 3.3105,
       "step": 50200
     },
     {
-      "epoch": 5.4084597998062645,
-      "grad_norm": 0.6479718685150146,
-      "learning_rate": 0.00027593147290162697,
-      "loss": 3.3208,
+      "epoch": 5.4177897574123985,
+      "grad_norm": 0.6487166285514832,
+      "learning_rate": 0.00027536535348084184,
+      "loss": 3.3413,
       "step": 50250
     },
     {
-      "epoch": 5.413841351845872,
-      "grad_norm": 0.7165466547012329,
-      "learning_rate": 0.0002756082318715655,
-      "loss": 3.3086,
+      "epoch": 5.423180592991914,
+      "grad_norm": 0.651813805103302,
+      "learning_rate": 0.00027504155423637344,
+      "loss": 3.333,
       "step": 50300
     },
     {
-      "epoch": 5.419222903885481,
-      "grad_norm": 0.6545393466949463,
-      "learning_rate": 0.00027528499084150416,
-      "loss": 3.3368,
+      "epoch": 5.428571428571429,
+      "grad_norm": 0.679421603679657,
+      "learning_rate": 0.000274717754991905,
+      "loss": 3.3219,
       "step": 50350
     },
     {
-      "epoch": 5.424604455925088,
-      "grad_norm": 0.7240278720855713,
-      "learning_rate": 0.0002749617498114427,
-      "loss": 3.3179,
+      "epoch": 5.433962264150943,
+      "grad_norm": 0.7000804543495178,
+      "learning_rate": 0.0002743939557474366,
+      "loss": 3.3415,
       "step": 50400
     },
     {
-      "epoch": 5.429986007964697,
-      "grad_norm": 0.6702057719230652,
-      "learning_rate": 0.0002746385087813813,
-      "loss": 3.3258,
+      "epoch": 5.439353099730458,
+      "grad_norm": 0.7223173379898071,
+      "learning_rate": 0.00027407015650296815,
+      "loss": 3.3387,
       "step": 50450
     },
     {
-      "epoch": 5.435367560004305,
-      "grad_norm": 0.8188278675079346,
-      "learning_rate": 0.0002743152677513199,
-      "loss": 3.3275,
+      "epoch": 5.444743935309973,
+      "grad_norm": 0.6525177955627441,
+      "learning_rate": 0.0002737463572584997,
+      "loss": 3.3237,
       "step": 50500
     },
     {
-      "epoch": 5.440749112043913,
-      "grad_norm": 0.6691777110099792,
-      "learning_rate": 0.0002739920267212585,
-      "loss": 3.3066,
+      "epoch": 5.450134770889488,
+      "grad_norm": 0.6648662686347961,
+      "learning_rate": 0.00027342255801403125,
+      "loss": 3.3207,
       "step": 50550
     },
     {
-      "epoch": 5.446130664083522,
-      "grad_norm": 0.7455892562866211,
-      "learning_rate": 0.000273668785691197,
-      "loss": 3.315,
+      "epoch": 5.455525606469003,
+      "grad_norm": 0.7127551436424255,
+      "learning_rate": 0.00027309875876956285,
+      "loss": 3.3364,
       "step": 50600
     },
     {
-      "epoch": 5.45151221612313,
-      "grad_norm": 0.7148830890655518,
-      "learning_rate": 0.0002733455446611356,
-      "loss": 3.3265,
+      "epoch": 5.460916442048518,
+      "grad_norm": 0.828137218952179,
+      "learning_rate": 0.0002727749595250944,
+      "loss": 3.3301,
       "step": 50650
     },
     {
-      "epoch": 5.456893768162738,
-      "grad_norm": 0.7596555352210999,
-      "learning_rate": 0.0002730223036310742,
-      "loss": 3.3433,
+      "epoch": 5.466307277628032,
+      "grad_norm": 0.7188193798065186,
+      "learning_rate": 0.000272451160280626,
+      "loss": 3.3211,
       "step": 50700
     },
     {
-      "epoch": 5.462275320202346,
-      "grad_norm": 0.7365542054176331,
-      "learning_rate": 0.0002726990626010128,
-      "loss": 3.3319,
+      "epoch": 5.471698113207547,
+      "grad_norm": 0.6920650005340576,
+      "learning_rate": 0.00027212736103615756,
+      "loss": 3.3072,
       "step": 50750
     },
     {
-      "epoch": 5.467656872241955,
-      "grad_norm": 0.7099530696868896,
-      "learning_rate": 0.0002723758215709514,
-      "loss": 3.3228,
+      "epoch": 5.4770889487870615,
+      "grad_norm": 0.6851574778556824,
+      "learning_rate": 0.00027180356179168916,
+      "loss": 3.3367,
       "step": 50800
     },
     {
-      "epoch": 5.473038424281563,
-      "grad_norm": 0.7575094103813171,
-      "learning_rate": 0.00027205258054088994,
-      "loss": 3.3276,
+      "epoch": 5.482479784366577,
+      "grad_norm": 0.7109363079071045,
+      "learning_rate": 0.0002714797625472207,
+      "loss": 3.3203,
       "step": 50850
     },
     {
-      "epoch": 5.478419976321171,
-      "grad_norm": 0.7478823065757751,
-      "learning_rate": 0.00027172933951082854,
-      "loss": 3.3282,
+      "epoch": 5.487870619946092,
+      "grad_norm": 0.6401546597480774,
+      "learning_rate": 0.00027115596330275226,
+      "loss": 3.3299,
       "step": 50900
     },
     {
-      "epoch": 5.483801528360779,
-      "grad_norm": 0.7028821110725403,
-      "learning_rate": 0.00027140609848076713,
-      "loss": 3.3204,
+      "epoch": 5.493261455525606,
+      "grad_norm": 0.6815487146377563,
+      "learning_rate": 0.0002708321640582838,
+      "loss": 3.3462,
       "step": 50950
     },
     {
-      "epoch": 5.489183080400387,
-      "grad_norm": 0.6952551007270813,
-      "learning_rate": 0.0002710828574507057,
-      "loss": 3.3082,
+      "epoch": 5.498652291105121,
+      "grad_norm": 0.6761698126792908,
+      "learning_rate": 0.0002705083648138154,
+      "loss": 3.3108,
       "step": 51000
     },
     {
-      "epoch": 5.489183080400387,
-      "eval_accuracy": 0.3827374030286558,
-      "eval_loss": 3.3939788341522217,
-      "eval_runtime": 187.668,
-      "eval_samples_per_second": 95.973,
-      "eval_steps_per_second": 6.0,
+      "epoch": 5.498652291105121,
+      "eval_accuracy": 0.38252563855030536,
+      "eval_loss": 3.396152973175049,
+      "eval_runtime": 183.5373,
+      "eval_samples_per_second": 98.133,
+      "eval_steps_per_second": 6.135,
       "step": 51000
     },
     {
-      "epoch": 5.494564632439996,
-      "grad_norm": 0.68047696352005,
-      "learning_rate": 0.00027075961642064427,
-      "loss": 3.3143,
+      "epoch": 5.504043126684636,
+      "grad_norm": 0.6581487655639648,
+      "learning_rate": 0.00027018456556934697,
+      "loss": 3.318,
       "step": 51050
     },
     {
-      "epoch": 5.499946184479604,
-      "grad_norm": 0.6872037053108215,
-      "learning_rate": 0.0002704363753905829,
-      "loss": 3.3336,
+      "epoch": 5.509433962264151,
+      "grad_norm": 0.6861607432365417,
+      "learning_rate": 0.0002698607663248786,
+      "loss": 3.328,
       "step": 51100
     },
     {
-      "epoch": 5.505327736519212,
-      "grad_norm": 0.6947546601295471,
-      "learning_rate": 0.00027011313436052145,
-      "loss": 3.3335,
+      "epoch": 5.514824797843666,
+      "grad_norm": 0.6859360933303833,
+      "learning_rate": 0.0002695369670804101,
+      "loss": 3.3452,
       "step": 51150
     },
     {
-      "epoch": 5.510709288558821,
-      "grad_norm": 0.7146032452583313,
-      "learning_rate": 0.00026978989333046005,
-      "loss": 3.344,
+      "epoch": 5.520215633423181,
+      "grad_norm": 0.6749665141105652,
+      "learning_rate": 0.0002692131678359417,
+      "loss": 3.3199,
       "step": 51200
     },
     {
-      "epoch": 5.516090840598428,
-      "grad_norm": 1.1656943559646606,
-      "learning_rate": 0.00026946665230039864,
-      "loss": 3.3217,
+      "epoch": 5.525606469002695,
+      "grad_norm": 0.7556314468383789,
+      "learning_rate": 0.0002688893685914733,
+      "loss": 3.3256,
       "step": 51250
     },
     {
-      "epoch": 5.521472392638037,
-      "grad_norm": 0.7083805203437805,
-      "learning_rate": 0.00026914341127033724,
-      "loss": 3.3247,
+      "epoch": 5.53099730458221,
+      "grad_norm": 0.6610503196716309,
+      "learning_rate": 0.00026856556934700483,
+      "loss": 3.3375,
       "step": 51300
     },
     {
-      "epoch": 5.5268539446776455,
-      "grad_norm": 0.7227205038070679,
-      "learning_rate": 0.00026882017024027583,
-      "loss": 3.3338,
+      "epoch": 5.536388140161725,
+      "grad_norm": 0.707281231880188,
+      "learning_rate": 0.0002682417701025364,
+      "loss": 3.3173,
       "step": 51350
     },
     {
-      "epoch": 5.532235496717253,
-      "grad_norm": 0.7416982650756836,
-      "learning_rate": 0.0002684969292102144,
-      "loss": 3.3197,
+      "epoch": 5.54177897574124,
+      "grad_norm": 0.6805925369262695,
+      "learning_rate": 0.000267917970858068,
+      "loss": 3.3251,
       "step": 51400
     },
     {
-      "epoch": 5.537617048756862,
-      "grad_norm": 0.7051200270652771,
-      "learning_rate": 0.00026817368818015297,
-      "loss": 3.303,
+      "epoch": 5.547169811320755,
+      "grad_norm": 0.7159254550933838,
+      "learning_rate": 0.00026759417161359954,
+      "loss": 3.3427,
       "step": 51450
     },
     {
-      "epoch": 5.542998600796469,
-      "grad_norm": 0.6784094572067261,
-      "learning_rate": 0.00026785044715009156,
-      "loss": 3.3179,
+      "epoch": 5.55256064690027,
+      "grad_norm": 0.6849930286407471,
+      "learning_rate": 0.0002672703723691311,
+      "loss": 3.3285,
       "step": 51500
     },
     {
-      "epoch": 5.548380152836078,
-      "grad_norm": 0.974920392036438,
-      "learning_rate": 0.00026752720612003016,
-      "loss": 3.3245,
+      "epoch": 5.557951482479784,
+      "grad_norm": 0.683801531791687,
+      "learning_rate": 0.0002669465731246627,
+      "loss": 3.3213,
       "step": 51550
     },
     {
-      "epoch": 5.553761704875686,
-      "grad_norm": 0.671526312828064,
-      "learning_rate": 0.0002672039650899687,
-      "loss": 3.3298,
+      "epoch": 5.563342318059299,
+      "grad_norm": 0.7315948605537415,
+      "learning_rate": 0.00026662277388019424,
+      "loss": 3.3503,
       "step": 51600
     },
     {
-      "epoch": 5.559143256915294,
-      "grad_norm": 0.7471204400062561,
-      "learning_rate": 0.00026688072405990735,
-      "loss": 3.33,
+      "epoch": 5.568733153638814,
+      "grad_norm": 0.649257481098175,
+      "learning_rate": 0.00026629897463572585,
+      "loss": 3.344,
       "step": 51650
     },
     {
-      "epoch": 5.564524808954903,
-      "grad_norm": 0.6865817904472351,
-      "learning_rate": 0.0002665574830298459,
-      "loss": 3.3178,
+      "epoch": 5.574123989218329,
+      "grad_norm": 0.7025150060653687,
+      "learning_rate": 0.0002659751753912574,
+      "loss": 3.3258,
       "step": 51700
     },
     {
-      "epoch": 5.569906360994511,
-      "grad_norm": 0.7355602979660034,
-      "learning_rate": 0.0002662342419997845,
-      "loss": 3.3154,
+      "epoch": 5.579514824797844,
+      "grad_norm": 0.6589361429214478,
+      "learning_rate": 0.00026565137614678895,
+      "loss": 3.3222,
       "step": 51750
     },
     {
-      "epoch": 5.575287913034119,
-      "grad_norm": 0.7200061082839966,
-      "learning_rate": 0.0002659110009697231,
-      "loss": 3.3287,
+      "epoch": 5.584905660377358,
+      "grad_norm": 0.7581202387809753,
+      "learning_rate": 0.0002653275769023205,
+      "loss": 3.335,
       "step": 51800
     },
     {
-      "epoch": 5.580669465073727,
-      "grad_norm": 0.7179515957832336,
-      "learning_rate": 0.00026558775993966167,
-      "loss": 3.3307,
+      "epoch": 5.590296495956873,
+      "grad_norm": 0.7484220266342163,
+      "learning_rate": 0.0002650037776578521,
+      "loss": 3.3308,
       "step": 51850
     },
     {
-      "epoch": 5.586051017113336,
-      "grad_norm": 0.6826567053794861,
-      "learning_rate": 0.0002652645189096002,
-      "loss": 3.3222,
+      "epoch": 5.595687331536388,
+      "grad_norm": 0.6836903095245361,
+      "learning_rate": 0.00026467997841338366,
+      "loss": 3.3376,
       "step": 51900
     },
     {
-      "epoch": 5.591432569152944,
-      "grad_norm": 0.6857197880744934,
-      "learning_rate": 0.0002649412778795388,
-      "loss": 3.313,
+      "epoch": 5.601078167115903,
+      "grad_norm": 0.6784467697143555,
+      "learning_rate": 0.00026435617916891526,
+      "loss": 3.3272,
       "step": 51950
     },
     {
-      "epoch": 5.596814121192552,
-      "grad_norm": 0.7227532267570496,
-      "learning_rate": 0.0002646180368494774,
-      "loss": 3.3311,
+      "epoch": 5.606469002695418,
+      "grad_norm": 0.6856781840324402,
+      "learning_rate": 0.0002640323799244468,
+      "loss": 3.3215,
       "step": 52000
     },
     {
-      "epoch": 5.596814121192552,
-      "eval_accuracy": 0.38301566307434864,
-      "eval_loss": 3.389040470123291,
-      "eval_runtime": 187.7263,
-      "eval_samples_per_second": 95.943,
-      "eval_steps_per_second": 5.998,
+      "epoch": 5.606469002695418,
+      "eval_accuracy": 0.38309497968245326,
+      "eval_loss": 3.3896756172180176,
+      "eval_runtime": 183.8027,
+      "eval_samples_per_second": 97.991,
+      "eval_steps_per_second": 6.126,
       "step": 52000
     },
     {
-      "epoch": 5.60219567323216,
-      "grad_norm": 0.7398081421852112,
-      "learning_rate": 0.000264294795819416,
-      "loss": 3.332,
+      "epoch": 5.611859838274933,
+      "grad_norm": 0.6830776929855347,
+      "learning_rate": 0.0002637085806799784,
+      "loss": 3.346,
       "step": 52050
     },
     {
-      "epoch": 5.607577225271768,
-      "grad_norm": 0.7204583883285522,
-      "learning_rate": 0.0002639715547893546,
-      "loss": 3.3236,
+      "epoch": 5.617250673854447,
+      "grad_norm": 0.7184668779373169,
+      "learning_rate": 0.00026338478143550997,
+      "loss": 3.3239,
       "step": 52100
     },
     {
-      "epoch": 5.612958777311377,
-      "grad_norm": 0.7188034057617188,
-      "learning_rate": 0.00026364831375929313,
-      "loss": 3.3372,
+      "epoch": 5.622641509433962,
+      "grad_norm": 0.7460675835609436,
+      "learning_rate": 0.00026306098219104157,
+      "loss": 3.3267,
       "step": 52150
     },
     {
-      "epoch": 5.618340329350985,
-      "grad_norm": 0.7495497465133667,
-      "learning_rate": 0.0002633250727292318,
-      "loss": 3.3213,
+      "epoch": 5.628032345013477,
+      "grad_norm": 0.709586501121521,
+      "learning_rate": 0.00026273718294657307,
+      "loss": 3.3231,
       "step": 52200
     },
     {
-      "epoch": 5.623721881390593,
-      "grad_norm": 0.7071952223777771,
-      "learning_rate": 0.0002630018316991703,
-      "loss": 3.3375,
+      "epoch": 5.633423180592992,
+      "grad_norm": 0.6780404448509216,
+      "learning_rate": 0.00026241338370210467,
+      "loss": 3.3244,
       "step": 52250
     },
     {
-      "epoch": 5.629103433430201,
-      "grad_norm": 0.7216173410415649,
-      "learning_rate": 0.0002626785906691089,
-      "loss": 3.3223,
+      "epoch": 5.638814016172507,
+      "grad_norm": 0.6592050790786743,
+      "learning_rate": 0.0002620895844576362,
+      "loss": 3.3275,
       "step": 52300
     },
     {
-      "epoch": 5.634484985469809,
-      "grad_norm": 0.6621602773666382,
-      "learning_rate": 0.0002623553496390475,
-      "loss": 3.3234,
+      "epoch": 5.644204851752022,
+      "grad_norm": 0.6790298819541931,
+      "learning_rate": 0.00026176578521316783,
+      "loss": 3.3332,
       "step": 52350
     },
     {
-      "epoch": 5.639866537509418,
-      "grad_norm": 0.7445098161697388,
-      "learning_rate": 0.0002620321086089861,
-      "loss": 3.3112,
+      "epoch": 5.649595687331536,
+      "grad_norm": 0.6760024428367615,
+      "learning_rate": 0.0002614419859686994,
+      "loss": 3.3223,
       "step": 52400
     },
     {
-      "epoch": 5.645248089549026,
-      "grad_norm": 0.6725427508354187,
-      "learning_rate": 0.00026170886757892464,
-      "loss": 3.3233,
+      "epoch": 5.654986522911051,
+      "grad_norm": 0.6775163412094116,
+      "learning_rate": 0.000261118186724231,
+      "loss": 3.3504,
       "step": 52450
     },
     {
-      "epoch": 5.650629641588634,
-      "grad_norm": 0.7050355672836304,
-      "learning_rate": 0.00026138562654886324,
-      "loss": 3.3311,
+      "epoch": 5.660377358490566,
+      "grad_norm": 0.7145818471908569,
+      "learning_rate": 0.00026079438747976253,
+      "loss": 3.3267,
       "step": 52500
     },
     {
-      "epoch": 5.656011193628243,
-      "grad_norm": 0.681422770023346,
-      "learning_rate": 0.00026106238551880183,
-      "loss": 3.3485,
+      "epoch": 5.665768194070081,
+      "grad_norm": 0.7327495217323303,
+      "learning_rate": 0.0002604705882352941,
+      "loss": 3.3419,
       "step": 52550
     },
     {
-      "epoch": 5.66139274566785,
-      "grad_norm": 0.688869059085846,
-      "learning_rate": 0.00026073914448874043,
-      "loss": 3.322,
+      "epoch": 5.671159029649596,
+      "grad_norm": 0.6712895631790161,
+      "learning_rate": 0.00026014678899082563,
+      "loss": 3.3549,
       "step": 52600
     },
     {
-      "epoch": 5.666774297707459,
-      "grad_norm": 0.7082963585853577,
-      "learning_rate": 0.000260415903458679,
-      "loss": 3.3165,
+      "epoch": 5.67654986522911,
+      "grad_norm": 0.7261403799057007,
+      "learning_rate": 0.00025982298974635724,
+      "loss": 3.3273,
       "step": 52650
     },
     {
-      "epoch": 5.672155849747067,
-      "grad_norm": 0.6897167563438416,
-      "learning_rate": 0.00026009266242861756,
-      "loss": 3.3172,
+      "epoch": 5.681940700808625,
+      "grad_norm": 0.7122320532798767,
+      "learning_rate": 0.0002594991905018888,
+      "loss": 3.3186,
       "step": 52700
     },
     {
-      "epoch": 5.677537401786675,
-      "grad_norm": 0.7198052406311035,
-      "learning_rate": 0.00025976942139855616,
-      "loss": 3.3216,
+      "epoch": 5.6873315363881405,
+      "grad_norm": 0.6834776997566223,
+      "learning_rate": 0.0002591753912574204,
+      "loss": 3.3341,
       "step": 52750
     },
     {
-      "epoch": 5.682918953826284,
-      "grad_norm": 0.738345205783844,
-      "learning_rate": 0.00025944618036849475,
-      "loss": 3.3159,
+      "epoch": 5.692722371967655,
+      "grad_norm": 0.6893863081932068,
+      "learning_rate": 0.00025885159201295195,
+      "loss": 3.3168,
       "step": 52800
     },
     {
-      "epoch": 5.688300505865891,
-      "grad_norm": 0.6863308548927307,
-      "learning_rate": 0.00025912293933843335,
-      "loss": 3.3396,
+      "epoch": 5.69811320754717,
+      "grad_norm": 0.6849216222763062,
+      "learning_rate": 0.0002585277927684835,
+      "loss": 3.3133,
       "step": 52850
     },
     {
-      "epoch": 5.6936820579055,
-      "grad_norm": 0.725994348526001,
-      "learning_rate": 0.0002587996983083719,
-      "loss": 3.3247,
+      "epoch": 5.703504043126685,
+      "grad_norm": 0.6997511386871338,
+      "learning_rate": 0.0002582039935240151,
+      "loss": 3.318,
       "step": 52900
     },
     {
-      "epoch": 5.699063609945108,
-      "grad_norm": 0.8310593366622925,
-      "learning_rate": 0.00025847645727831054,
-      "loss": 3.3456,
+      "epoch": 5.708894878706199,
+      "grad_norm": 0.7135558724403381,
+      "learning_rate": 0.000257886670264436,
+      "loss": 3.3224,
       "step": 52950
     },
     {
-      "epoch": 5.704445161984716,
-      "grad_norm": 0.7754035592079163,
-      "learning_rate": 0.0002581532162482491,
-      "loss": 3.3328,
+      "epoch": 5.714285714285714,
+      "grad_norm": 0.7129635810852051,
+      "learning_rate": 0.0002575628710199676,
+      "loss": 3.3343,
       "step": 53000
     },
     {
-      "epoch": 5.704445161984716,
-      "eval_accuracy": 0.3833655253731113,
-      "eval_loss": 3.386369466781616,
-      "eval_runtime": 200.806,
-      "eval_samples_per_second": 89.694,
-      "eval_steps_per_second": 5.607,
+      "epoch": 5.714285714285714,
+      "eval_accuracy": 0.38362455385766103,
+      "eval_loss": 3.3855338096618652,
+      "eval_runtime": 184.0916,
+      "eval_samples_per_second": 97.837,
+      "eval_steps_per_second": 6.117,
       "step": 53000
     },
     {
-      "epoch": 5.709826714024325,
-      "grad_norm": 0.7284052968025208,
-      "learning_rate": 0.00025782997521818767,
-      "loss": 3.3356,
+      "epoch": 5.719676549865229,
+      "grad_norm": 0.6876802444458008,
+      "learning_rate": 0.00025723907177549915,
+      "loss": 3.3241,
       "step": 53050
     },
     {
-      "epoch": 5.715208266063933,
-      "grad_norm": 0.6969070434570312,
-      "learning_rate": 0.0002575131990087275,
-      "loss": 3.3319,
+      "epoch": 5.725067385444744,
+      "grad_norm": 0.7315533757209778,
+      "learning_rate": 0.00025691527253103076,
+      "loss": 3.335,
       "step": 53100
     },
     {
-      "epoch": 5.720589818103541,
-      "grad_norm": 0.7205783724784851,
-      "learning_rate": 0.0002571899579786661,
-      "loss": 3.3317,
+      "epoch": 5.730458221024259,
+      "grad_norm": 0.6807094812393188,
+      "learning_rate": 0.0002565914732865623,
+      "loss": 3.3275,
       "step": 53150
     },
     {
-      "epoch": 5.725971370143149,
-      "grad_norm": 0.7033756375312805,
-      "learning_rate": 0.00025686671694860467,
-      "loss": 3.326,
+      "epoch": 5.735849056603773,
+      "grad_norm": 0.7097885012626648,
+      "learning_rate": 0.00025626767404209386,
+      "loss": 3.3395,
       "step": 53200
     },
     {
-      "epoch": 5.731352922182758,
-      "grad_norm": 0.7156730890274048,
-      "learning_rate": 0.0002565434759185432,
-      "loss": 3.3327,
+      "epoch": 5.741239892183288,
+      "grad_norm": 0.6894499063491821,
+      "learning_rate": 0.00025594387479762546,
+      "loss": 3.3106,
       "step": 53250
     },
     {
-      "epoch": 5.736734474222366,
-      "grad_norm": 0.7606733441352844,
-      "learning_rate": 0.00025622023488848186,
-      "loss": 3.3226,
+      "epoch": 5.7466307277628035,
+      "grad_norm": 0.7534905672073364,
+      "learning_rate": 0.000255620075553157,
+      "loss": 3.353,
       "step": 53300
     },
     {
-      "epoch": 5.742116026261974,
-      "grad_norm": 0.7340618968009949,
-      "learning_rate": 0.0002558969938584204,
-      "loss": 3.3345,
+      "epoch": 5.752021563342318,
+      "grad_norm": 0.6421119570732117,
+      "learning_rate": 0.00025529627630868857,
+      "loss": 3.3371,
       "step": 53350
     },
     {
-      "epoch": 5.747497578301582,
-      "grad_norm": 0.7599658966064453,
-      "learning_rate": 0.000255573752828359,
-      "loss": 3.3379,
+      "epoch": 5.757412398921833,
+      "grad_norm": 0.7399626970291138,
+      "learning_rate": 0.00025497247706422017,
+      "loss": 3.3406,
       "step": 53400
     },
     {
-      "epoch": 5.75287913034119,
-      "grad_norm": 0.7094115614891052,
-      "learning_rate": 0.0002552505117982976,
-      "loss": 3.3175,
+      "epoch": 5.762803234501348,
+      "grad_norm": 0.651099681854248,
+      "learning_rate": 0.0002546486778197517,
+      "loss": 3.3378,
       "step": 53450
     },
     {
-      "epoch": 5.758260682380799,
-      "grad_norm": 0.7160277962684631,
-      "learning_rate": 0.0002549272707682362,
-      "loss": 3.3234,
+      "epoch": 5.768194070080862,
+      "grad_norm": 0.7169281840324402,
+      "learning_rate": 0.00025432487857528327,
+      "loss": 3.3194,
       "step": 53500
     },
     {
-      "epoch": 5.763642234420407,
-      "grad_norm": 0.7079792618751526,
-      "learning_rate": 0.0002546040297381747,
-      "loss": 3.3118,
+      "epoch": 5.773584905660377,
+      "grad_norm": 0.688315212726593,
+      "learning_rate": 0.0002540010793308149,
+      "loss": 3.3187,
       "step": 53550
     },
     {
-      "epoch": 5.769023786460015,
-      "grad_norm": 0.7237218022346497,
-      "learning_rate": 0.0002542807887081133,
-      "loss": 3.3186,
+      "epoch": 5.7789757412398925,
+      "grad_norm": 0.7857226729393005,
+      "learning_rate": 0.0002536772800863464,
+      "loss": 3.3393,
       "step": 53600
     },
     {
-      "epoch": 5.774405338499624,
-      "grad_norm": 0.6792340874671936,
-      "learning_rate": 0.0002539575476780519,
-      "loss": 3.3267,
+      "epoch": 5.784366576819407,
+      "grad_norm": 0.6583890318870544,
+      "learning_rate": 0.00025335348084187803,
+      "loss": 3.3207,
       "step": 53650
     },
     {
-      "epoch": 5.779786890539231,
-      "grad_norm": 0.7164262533187866,
-      "learning_rate": 0.0002536343066479905,
-      "loss": 3.3334,
+      "epoch": 5.789757412398922,
+      "grad_norm": 0.6898304224014282,
+      "learning_rate": 0.0002530296815974096,
+      "loss": 3.3432,
       "step": 53700
     },
     {
-      "epoch": 5.78516844257884,
-      "grad_norm": 0.6917945742607117,
-      "learning_rate": 0.0002533110656179291,
-      "loss": 3.3347,
+      "epoch": 5.795148247978437,
+      "grad_norm": 0.6822751760482788,
+      "learning_rate": 0.0002527058823529412,
+      "loss": 3.3302,
       "step": 53750
     },
     {
-      "epoch": 5.790549994618448,
-      "grad_norm": 0.6914453506469727,
-      "learning_rate": 0.00025298782458786764,
-      "loss": 3.3368,
+      "epoch": 5.800539083557951,
+      "grad_norm": 0.6696268916130066,
+      "learning_rate": 0.0002523820831084727,
+      "loss": 3.3208,
       "step": 53800
     },
     {
-      "epoch": 5.795931546658056,
-      "grad_norm": 0.7576737403869629,
-      "learning_rate": 0.00025266458355780624,
-      "loss": 3.3418,
+      "epoch": 5.8059299191374665,
+      "grad_norm": 0.7353618741035461,
+      "learning_rate": 0.0002520582838640043,
+      "loss": 3.3235,
       "step": 53850
     },
     {
-      "epoch": 5.801313098697665,
-      "grad_norm": 0.7011058926582336,
-      "learning_rate": 0.00025234134252774483,
-      "loss": 3.3221,
+      "epoch": 5.811320754716981,
+      "grad_norm": 0.7432413101196289,
+      "learning_rate": 0.00025173448461953584,
+      "loss": 3.3253,
       "step": 53900
     },
     {
-      "epoch": 5.806694650737272,
-      "grad_norm": 0.7232387065887451,
-      "learning_rate": 0.0002520181014976834,
-      "loss": 3.3471,
+      "epoch": 5.816711590296496,
+      "grad_norm": 0.7125387787818909,
+      "learning_rate": 0.00025141068537506744,
+      "loss": 3.3226,
       "step": 53950
     },
     {
-      "epoch": 5.812076202776881,
-      "grad_norm": 0.7181257009506226,
-      "learning_rate": 0.000251694860467622,
-      "loss": 3.3309,
+      "epoch": 5.822102425876011,
+      "grad_norm": 0.7176763415336609,
+      "learning_rate": 0.000251086886130599,
+      "loss": 3.3191,
       "step": 54000
     },
     {
-      "epoch": 5.812076202776881,
-      "eval_accuracy": 0.3835403478696048,
-      "eval_loss": 3.381748914718628,
-      "eval_runtime": 200.1304,
-      "eval_samples_per_second": 89.996,
-      "eval_steps_per_second": 5.626,
+      "epoch": 5.822102425876011,
+      "eval_accuracy": 0.384046235715269,
+      "eval_loss": 3.3808350563049316,
+      "eval_runtime": 185.0871,
+      "eval_samples_per_second": 97.311,
+      "eval_steps_per_second": 6.084,
       "step": 54000
     },
     {
-      "epoch": 5.817457754816489,
-      "grad_norm": 0.7484964728355408,
-      "learning_rate": 0.0002513716194375606,
-      "loss": 3.343,
+      "epoch": 5.827493261455525,
+      "grad_norm": 0.6629961133003235,
+      "learning_rate": 0.0002507630868861306,
+      "loss": 3.3285,
       "step": 54050
     },
     {
-      "epoch": 5.822839306856097,
-      "grad_norm": 0.7453248500823975,
-      "learning_rate": 0.00025104837840749916,
-      "loss": 3.3381,
+      "epoch": 5.83288409703504,
+      "grad_norm": 0.6955593824386597,
+      "learning_rate": 0.00025043928764166215,
+      "loss": 3.3169,
       "step": 54100
     },
     {
-      "epoch": 5.828220858895706,
-      "grad_norm": 0.721716582775116,
-      "learning_rate": 0.00025072513737743775,
-      "loss": 3.3265,
+      "epoch": 5.8382749326145555,
+      "grad_norm": 0.7469235062599182,
+      "learning_rate": 0.00025011548839719375,
+      "loss": 3.3331,
       "step": 54150
     },
     {
-      "epoch": 5.833602410935313,
-      "grad_norm": 0.7338887453079224,
-      "learning_rate": 0.00025040189634737634,
-      "loss": 3.3275,
+      "epoch": 5.84366576819407,
+      "grad_norm": 0.721454381942749,
+      "learning_rate": 0.0002497916891527253,
+      "loss": 3.3211,
       "step": 54200
     },
     {
-      "epoch": 5.838983962974922,
-      "grad_norm": 0.797236979007721,
-      "learning_rate": 0.00025007865531731494,
-      "loss": 3.3308,
+      "epoch": 5.849056603773585,
+      "grad_norm": 0.6820123195648193,
+      "learning_rate": 0.00024946788990825686,
+      "loss": 3.3228,
       "step": 54250
     },
     {
-      "epoch": 5.84436551501453,
-      "grad_norm": 0.7230899333953857,
-      "learning_rate": 0.00024975541428725353,
-      "loss": 3.3089,
+      "epoch": 5.8544474393531,
+      "grad_norm": 0.7771530151367188,
+      "learning_rate": 0.0002491440906637884,
+      "loss": 3.3229,
       "step": 54300
     },
     {
-      "epoch": 5.849747067054138,
-      "grad_norm": 0.6964775323867798,
-      "learning_rate": 0.0002494321732571921,
-      "loss": 3.323,
+      "epoch": 5.859838274932614,
+      "grad_norm": 0.725572943687439,
+      "learning_rate": 0.00024882029141932,
+      "loss": 3.3351,
       "step": 54350
     },
     {
-      "epoch": 5.855128619093747,
-      "grad_norm": 0.7054535746574402,
-      "learning_rate": 0.00024910893222713067,
-      "loss": 3.3238,
+      "epoch": 5.8652291105121295,
+      "grad_norm": 0.7238464951515198,
+      "learning_rate": 0.00024849649217485156,
+      "loss": 3.3189,
       "step": 54400
     },
     {
-      "epoch": 5.860510171133355,
-      "grad_norm": 0.7256869673728943,
-      "learning_rate": 0.00024878569119706926,
-      "loss": 3.3511,
+      "epoch": 5.870619946091644,
+      "grad_norm": 0.7001297473907471,
+      "learning_rate": 0.00024817269293038317,
+      "loss": 3.3365,
       "step": 54450
     },
     {
-      "epoch": 5.865891723172963,
-      "grad_norm": 0.7951526045799255,
-      "learning_rate": 0.00024846245016700786,
-      "loss": 3.341,
+      "epoch": 5.876010781671159,
+      "grad_norm": 0.6613125205039978,
+      "learning_rate": 0.0002478488936859147,
+      "loss": 3.3376,
       "step": 54500
     },
     {
-      "epoch": 5.871273275212571,
-      "grad_norm": 0.742369532585144,
-      "learning_rate": 0.0002481392091369464,
-      "loss": 3.3285,
+      "epoch": 5.881401617250674,
+      "grad_norm": 0.7346794009208679,
+      "learning_rate": 0.00024752509444144627,
+      "loss": 3.3123,
       "step": 54550
     },
     {
-      "epoch": 5.87665482725218,
-      "grad_norm": 0.7485157251358032,
-      "learning_rate": 0.00024781596810688505,
-      "loss": 3.3187,
+      "epoch": 5.886792452830189,
+      "grad_norm": 0.6775569915771484,
+      "learning_rate": 0.00024720129519697787,
+      "loss": 3.3376,
       "step": 54600
     },
     {
-      "epoch": 5.882036379291788,
-      "grad_norm": 0.7049429416656494,
-      "learning_rate": 0.0002474927270768236,
-      "loss": 3.3259,
+      "epoch": 5.892183288409703,
+      "grad_norm": 0.6779545545578003,
+      "learning_rate": 0.0002468774959525094,
+      "loss": 3.3189,
       "step": 54650
     },
     {
-      "epoch": 5.887417931331396,
-      "grad_norm": 0.7091401815414429,
-      "learning_rate": 0.0002471694860467622,
-      "loss": 3.3246,
+      "epoch": 5.8975741239892185,
+      "grad_norm": 0.6905809044837952,
+      "learning_rate": 0.00024655369670804097,
+      "loss": 3.3262,
       "step": 54700
     },
     {
-      "epoch": 5.892799483371004,
-      "grad_norm": 0.727120578289032,
-      "learning_rate": 0.0002468462450167008,
-      "loss": 3.3301,
+      "epoch": 5.902964959568733,
+      "grad_norm": 0.6714254021644592,
+      "learning_rate": 0.0002462298974635726,
+      "loss": 3.3351,
       "step": 54750
     },
     {
-      "epoch": 5.898181035410612,
-      "grad_norm": 0.7163100242614746,
-      "learning_rate": 0.0002465230039866393,
-      "loss": 3.3124,
+      "epoch": 5.908355795148248,
+      "grad_norm": 0.6869944930076599,
+      "learning_rate": 0.00024590609821910413,
+      "loss": 3.3285,
       "step": 54800
     },
     {
-      "epoch": 5.903562587450221,
-      "grad_norm": 0.6951466202735901,
-      "learning_rate": 0.0002461997629565779,
-      "loss": 3.3158,
+      "epoch": 5.913746630727763,
+      "grad_norm": 0.6963022947311401,
+      "learning_rate": 0.0002455822989746357,
+      "loss": 3.3355,
       "step": 54850
     },
     {
-      "epoch": 5.9089441394898286,
-      "grad_norm": 0.7021745443344116,
-      "learning_rate": 0.0002458765219265165,
-      "loss": 3.3261,
+      "epoch": 5.919137466307277,
+      "grad_norm": 0.7132091522216797,
+      "learning_rate": 0.0002452584997301673,
+      "loss": 3.3355,
       "step": 54900
     },
     {
-      "epoch": 5.914325691529437,
-      "grad_norm": 0.7760040760040283,
-      "learning_rate": 0.0002455532808964551,
-      "loss": 3.321,
+      "epoch": 5.9245283018867925,
+      "grad_norm": 0.7264422178268433,
+      "learning_rate": 0.00024493470048569883,
+      "loss": 3.338,
       "step": 54950
     },
     {
-      "epoch": 5.919707243569046,
-      "grad_norm": 0.7871654033660889,
-      "learning_rate": 0.0002452300398663937,
-      "loss": 3.3271,
+      "epoch": 5.929919137466308,
+      "grad_norm": 0.7266690135002136,
+      "learning_rate": 0.00024461090124123044,
+      "loss": 3.3208,
       "step": 55000
     },
     {
-      "epoch": 5.919707243569046,
-      "eval_accuracy": 0.38421606017892307,
-      "eval_loss": 3.3762869834899902,
-      "eval_runtime": 190.6299,
-      "eval_samples_per_second": 94.482,
-      "eval_steps_per_second": 5.907,
+      "epoch": 5.929919137466308,
+      "eval_accuracy": 0.3844975798112504,
+      "eval_loss": 3.3767948150634766,
+      "eval_runtime": 185.2244,
+      "eval_samples_per_second": 97.239,
+      "eval_steps_per_second": 6.079,
       "step": 55000
     },
     {
-      "epoch": 5.925088795608653,
-      "grad_norm": 0.7077029347419739,
-      "learning_rate": 0.0002449067988363323,
-      "loss": 3.3386,
+      "epoch": 5.935309973045822,
+      "grad_norm": 0.6925628781318665,
+      "learning_rate": 0.000244287101996762,
+      "loss": 3.3498,
       "step": 55050
     },
     {
-      "epoch": 5.930470347648262,
-      "grad_norm": 0.7377938032150269,
-      "learning_rate": 0.00024458355780627083,
-      "loss": 3.3179,
+      "epoch": 5.940700808625337,
+      "grad_norm": 0.6779472827911377,
+      "learning_rate": 0.00024396330275229354,
+      "loss": 3.3142,
       "step": 55100
     },
     {
-      "epoch": 5.93585189968787,
-      "grad_norm": 0.7463526725769043,
-      "learning_rate": 0.0002442603167762094,
-      "loss": 3.3256,
+      "epoch": 5.946091644204852,
+      "grad_norm": 0.6793407797813416,
+      "learning_rate": 0.00024363950350782512,
+      "loss": 3.3226,
       "step": 55150
     },
     {
-      "epoch": 5.941233451727478,
-      "grad_norm": 0.7117036581039429,
-      "learning_rate": 0.00024393707574614802,
-      "loss": 3.3449,
+      "epoch": 5.951482479784366,
+      "grad_norm": 0.7185749411582947,
+      "learning_rate": 0.0002433157042633567,
+      "loss": 3.3188,
       "step": 55200
     },
     {
-      "epoch": 5.946615003767087,
-      "grad_norm": 0.7105012536048889,
-      "learning_rate": 0.0002436138347160866,
-      "loss": 3.3167,
+      "epoch": 5.9568733153638815,
+      "grad_norm": 0.7417930364608765,
+      "learning_rate": 0.00024299190501888827,
+      "loss": 3.3229,
       "step": 55250
     },
     {
-      "epoch": 5.951996555806694,
-      "grad_norm": 0.7295982837677002,
-      "learning_rate": 0.0002432905936860252,
-      "loss": 3.3308,
+      "epoch": 5.962264150943396,
+      "grad_norm": 0.6613512635231018,
+      "learning_rate": 0.00024266810577441985,
+      "loss": 3.3173,
       "step": 55300
     },
     {
-      "epoch": 5.957378107846303,
-      "grad_norm": 0.7493639588356018,
-      "learning_rate": 0.00024296735265596378,
-      "loss": 3.3332,
+      "epoch": 5.967654986522911,
+      "grad_norm": 0.7257469892501831,
+      "learning_rate": 0.00024234430652995143,
+      "loss": 3.319,
       "step": 55350
     },
     {
-      "epoch": 5.962759659885911,
-      "grad_norm": 0.7515854239463806,
-      "learning_rate": 0.00024264411162590235,
-      "loss": 3.3248,
+      "epoch": 5.973045822102426,
+      "grad_norm": 0.7043033242225647,
+      "learning_rate": 0.00024202050728548298,
+      "loss": 3.3218,
       "step": 55400
     },
     {
-      "epoch": 5.968141211925519,
-      "grad_norm": 0.7299107909202576,
-      "learning_rate": 0.00024232087059584097,
-      "loss": 3.3305,
+      "epoch": 5.97843665768194,
+      "grad_norm": 0.6857475638389587,
+      "learning_rate": 0.00024169670804101456,
+      "loss": 3.3423,
       "step": 55450
     },
     {
-      "epoch": 5.973522763965128,
-      "grad_norm": 0.7446507215499878,
-      "learning_rate": 0.00024199762956577953,
-      "loss": 3.3299,
+      "epoch": 5.9838274932614555,
+      "grad_norm": 0.8181941509246826,
+      "learning_rate": 0.0002413729087965461,
+      "loss": 3.3222,
       "step": 55500
     },
     {
-      "epoch": 5.978904316004736,
-      "grad_norm": 0.6979471445083618,
-      "learning_rate": 0.0002416743885357181,
-      "loss": 3.3072,
+      "epoch": 5.989218328840971,
+      "grad_norm": 0.7062280774116516,
+      "learning_rate": 0.00024104910955207768,
+      "loss": 3.3274,
       "step": 55550
     },
     {
-      "epoch": 5.984285868044344,
-      "grad_norm": 0.6954522132873535,
-      "learning_rate": 0.0002413511475056567,
-      "loss": 3.3296,
+      "epoch": 5.994609164420485,
+      "grad_norm": 0.7362241148948669,
+      "learning_rate": 0.00024072531030760926,
+      "loss": 3.3441,
       "step": 55600
     },
     {
-      "epoch": 5.989667420083952,
-      "grad_norm": 0.7171802520751953,
-      "learning_rate": 0.0002410279064755953,
-      "loss": 3.3144,
+      "epoch": 6.0,
+      "grad_norm": 1.553808569908142,
+      "learning_rate": 0.00024040151106314084,
+      "loss": 3.3417,
       "step": 55650
     },
     {
-      "epoch": 5.995048972123561,
-      "grad_norm": 0.6994540095329285,
-      "learning_rate": 0.00024070466544553386,
-      "loss": 3.3094,
+      "epoch": 6.005390835579515,
+      "grad_norm": 0.7019472718238831,
+      "learning_rate": 0.0002400777118186724,
+      "loss": 3.2374,
       "step": 55700
     },
     {
-      "epoch": 6.000430524163169,
-      "grad_norm": 0.7464019060134888,
-      "learning_rate": 0.00024038142441547245,
-      "loss": 3.3063,
+      "epoch": 6.010781671159029,
+      "grad_norm": 0.7424504160881042,
+      "learning_rate": 0.00023975391257420397,
+      "loss": 3.2381,
       "step": 55750
     },
     {
-      "epoch": 6.005812076202777,
-      "grad_norm": 0.7907078266143799,
-      "learning_rate": 0.00024005818338541102,
-      "loss": 3.2375,
+      "epoch": 6.0161725067385445,
+      "grad_norm": 0.7570995092391968,
+      "learning_rate": 0.00023943011332973555,
+      "loss": 3.2323,
       "step": 55800
     },
     {
-      "epoch": 6.011193628242385,
-      "grad_norm": 0.6996508836746216,
-      "learning_rate": 0.00023974140717595085,
-      "loss": 3.2103,
+      "epoch": 6.02156334231806,
+      "grad_norm": 0.719301164150238,
+      "learning_rate": 0.00023910631408526712,
+      "loss": 3.239,
       "step": 55850
     },
     {
-      "epoch": 6.016575180281993,
-      "grad_norm": 0.7290385961532593,
-      "learning_rate": 0.00023941816614588942,
-      "loss": 3.2373,
+      "epoch": 6.026954177897574,
+      "grad_norm": 0.7143483757972717,
+      "learning_rate": 0.0002387825148407987,
+      "loss": 3.2396,
       "step": 55900
     },
     {
-      "epoch": 6.021956732321602,
-      "grad_norm": 0.7101258635520935,
-      "learning_rate": 0.00023909492511582802,
-      "loss": 3.2352,
+      "epoch": 6.032345013477089,
+      "grad_norm": 0.7269814014434814,
+      "learning_rate": 0.00023845871559633025,
+      "loss": 3.2359,
       "step": 55950
     },
     {
-      "epoch": 6.0273382843612096,
-      "grad_norm": 0.7947726249694824,
-      "learning_rate": 0.0002387716840857666,
-      "loss": 3.2538,
+      "epoch": 6.037735849056604,
+      "grad_norm": 0.7131744623184204,
+      "learning_rate": 0.0002381349163518618,
+      "loss": 3.2305,
       "step": 56000
     },
     {
-      "epoch": 6.0273382843612096,
-      "eval_accuracy": 0.3846035163768695,
-      "eval_loss": 3.381119728088379,
-      "eval_runtime": 207.8084,
-      "eval_samples_per_second": 86.671,
-      "eval_steps_per_second": 5.418,
+      "epoch": 6.037735849056604,
+      "eval_accuracy": 0.38423692153338346,
+      "eval_loss": 3.3811709880828857,
+      "eval_runtime": 185.1649,
+      "eval_samples_per_second": 97.27,
+      "eval_steps_per_second": 6.081,
       "step": 56000
     },
     {
-      "epoch": 6.032719836400818,
-      "grad_norm": 0.7130570411682129,
-      "learning_rate": 0.00023844844305570518,
-      "loss": 3.2478,
+      "epoch": 6.0431266846361185,
+      "grad_norm": 0.7208868265151978,
+      "learning_rate": 0.00023781111710739338,
+      "loss": 3.2402,
       "step": 56050
     },
     {
-      "epoch": 6.038101388440427,
-      "grad_norm": 0.7319710850715637,
-      "learning_rate": 0.00023812520202564377,
-      "loss": 3.2264,
+      "epoch": 6.048517520215634,
+      "grad_norm": 0.7629277110099792,
+      "learning_rate": 0.00023749379384781433,
+      "loss": 3.2381,
       "step": 56100
     },
     {
-      "epoch": 6.043482940480034,
-      "grad_norm": 0.7575207352638245,
-      "learning_rate": 0.00023780196099558234,
-      "loss": 3.2426,
+      "epoch": 6.053908355795148,
+      "grad_norm": 0.7236301898956299,
+      "learning_rate": 0.0002371699946033459,
+      "loss": 3.2533,
       "step": 56150
     },
     {
-      "epoch": 6.048864492519643,
-      "grad_norm": 0.7339271903038025,
-      "learning_rate": 0.00023747871996552094,
-      "loss": 3.2588,
+      "epoch": 6.059299191374663,
+      "grad_norm": 0.742306649684906,
+      "learning_rate": 0.0002368461953588775,
+      "loss": 3.2363,
       "step": 56200
     },
     {
-      "epoch": 6.0542460445592505,
-      "grad_norm": 0.7257949709892273,
-      "learning_rate": 0.00023715547893545953,
-      "loss": 3.2364,
+      "epoch": 6.064690026954178,
+      "grad_norm": 0.7459393739700317,
+      "learning_rate": 0.00023652239611440904,
+      "loss": 3.2427,
       "step": 56250
     },
     {
-      "epoch": 6.059627596598859,
-      "grad_norm": 0.7105091214179993,
-      "learning_rate": 0.0002368322379053981,
-      "loss": 3.2404,
+      "epoch": 6.070080862533692,
+      "grad_norm": 0.6891180872917175,
+      "learning_rate": 0.00023619859686994062,
+      "loss": 3.2508,
       "step": 56300
     },
     {
-      "epoch": 6.065009148638468,
-      "grad_norm": 0.7371712327003479,
-      "learning_rate": 0.00023650899687533667,
-      "loss": 3.2625,
+      "epoch": 6.0754716981132075,
+      "grad_norm": 0.7101725339889526,
+      "learning_rate": 0.00023587479762547217,
+      "loss": 3.2488,
       "step": 56350
     },
     {
-      "epoch": 6.070390700678075,
-      "grad_norm": 0.7497949004173279,
-      "learning_rate": 0.0002361857558452753,
-      "loss": 3.2429,
+      "epoch": 6.080862533692723,
+      "grad_norm": 0.714775800704956,
+      "learning_rate": 0.00023555099838100374,
+      "loss": 3.2692,
       "step": 56400
     },
     {
-      "epoch": 6.075772252717684,
-      "grad_norm": 0.7092165946960449,
-      "learning_rate": 0.00023586251481521386,
-      "loss": 3.2507,
+      "epoch": 6.086253369272237,
+      "grad_norm": 0.70735102891922,
+      "learning_rate": 0.00023522719913653532,
+      "loss": 3.2572,
       "step": 56450
     },
     {
-      "epoch": 6.081153804757292,
-      "grad_norm": 0.8313487768173218,
-      "learning_rate": 0.00023553927378515242,
-      "loss": 3.2328,
+      "epoch": 6.091644204851752,
+      "grad_norm": 0.7020072340965271,
+      "learning_rate": 0.0002349033998920669,
+      "loss": 3.2382,
       "step": 56500
     },
     {
-      "epoch": 6.0865353567969,
-      "grad_norm": 0.7452508211135864,
-      "learning_rate": 0.00023521603275509104,
-      "loss": 3.2596,
+      "epoch": 6.097035040431267,
+      "grad_norm": 0.6657240390777588,
+      "learning_rate": 0.00023457960064759848,
+      "loss": 3.27,
       "step": 56550
     },
     {
-      "epoch": 6.091916908836509,
-      "grad_norm": 0.7973877787590027,
-      "learning_rate": 0.0002348927917250296,
-      "loss": 3.2535,
+      "epoch": 6.1024258760107815,
+      "grad_norm": 0.7259310483932495,
+      "learning_rate": 0.00023425580140313005,
+      "loss": 3.2525,
       "step": 56600
     },
     {
-      "epoch": 6.097298460876116,
-      "grad_norm": 0.7103460431098938,
-      "learning_rate": 0.00023456955069496818,
-      "loss": 3.2346,
+      "epoch": 6.107816711590297,
+      "grad_norm": 0.7217092514038086,
+      "learning_rate": 0.00023393200215866163,
+      "loss": 3.2364,
       "step": 56650
     },
     {
-      "epoch": 6.102680012915725,
-      "grad_norm": 0.7472631335258484,
-      "learning_rate": 0.00023424630966490677,
-      "loss": 3.2521,
+      "epoch": 6.113207547169812,
+      "grad_norm": 0.7307034134864807,
+      "learning_rate": 0.00023360820291419316,
+      "loss": 3.2675,
       "step": 56700
     },
     {
-      "epoch": 6.108061564955333,
-      "grad_norm": 0.7677678465843201,
-      "learning_rate": 0.00023392306863484537,
-      "loss": 3.2521,
+      "epoch": 6.118598382749326,
+      "grad_norm": 0.7697921991348267,
+      "learning_rate": 0.00023328440366972473,
+      "loss": 3.2735,
       "step": 56750
     },
     {
-      "epoch": 6.113443116994941,
-      "grad_norm": 0.7261399030685425,
-      "learning_rate": 0.00023359982760478396,
-      "loss": 3.2393,
+      "epoch": 6.123989218328841,
+      "grad_norm": 0.7372835278511047,
+      "learning_rate": 0.0002329606044252563,
+      "loss": 3.2638,
       "step": 56800
     },
     {
-      "epoch": 6.11882466903455,
-      "grad_norm": 0.7391446232795715,
-      "learning_rate": 0.00023327658657472253,
-      "loss": 3.2478,
+      "epoch": 6.129380053908355,
+      "grad_norm": 0.7402812242507935,
+      "learning_rate": 0.0002326368051807879,
+      "loss": 3.2475,
       "step": 56850
     },
     {
-      "epoch": 6.124206221074158,
-      "grad_norm": 0.7893763184547424,
-      "learning_rate": 0.0002329533455446611,
-      "loss": 3.2512,
+      "epoch": 6.1347708894878705,
+      "grad_norm": 0.6876668334007263,
+      "learning_rate": 0.00023231300593631947,
+      "loss": 3.2637,
       "step": 56900
     },
     {
-      "epoch": 6.129587773113766,
-      "grad_norm": 0.7050673365592957,
-      "learning_rate": 0.00023263010451459972,
-      "loss": 3.2686,
+      "epoch": 6.140161725067386,
+      "grad_norm": 0.7216595411300659,
+      "learning_rate": 0.00023198920669185104,
+      "loss": 3.2467,
       "step": 56950
     },
     {
-      "epoch": 6.134969325153374,
-      "grad_norm": 0.7486889958381653,
-      "learning_rate": 0.0002323068634845383,
-      "loss": 3.258,
+      "epoch": 6.1455525606469,
+      "grad_norm": 0.7456094622612,
+      "learning_rate": 0.00023166540744738262,
+      "loss": 3.2622,
       "step": 57000
     },
     {
-      "epoch": 6.134969325153374,
-      "eval_accuracy": 0.3847839888235293,
-      "eval_loss": 3.378803014755249,
-      "eval_runtime": 201.1191,
-      "eval_samples_per_second": 89.554,
-      "eval_steps_per_second": 5.599,
+      "epoch": 6.1455525606469,
+      "eval_accuracy": 0.3843653492467802,
+      "eval_loss": 3.3819985389709473,
+      "eval_runtime": 184.8965,
+      "eval_samples_per_second": 97.411,
+      "eval_steps_per_second": 6.09,
       "step": 57000
     },
     {
-      "epoch": 6.140350877192983,
-      "grad_norm": 0.7525475025177002,
-      "learning_rate": 0.00023198362245447686,
-      "loss": 3.2444,
+      "epoch": 6.150943396226415,
+      "grad_norm": 0.7461252808570862,
+      "learning_rate": 0.0002313416082029142,
+      "loss": 3.2524,
       "step": 57050
     },
     {
-      "epoch": 6.1457324292325906,
-      "grad_norm": 0.7189371585845947,
-      "learning_rate": 0.00023166038142441548,
-      "loss": 3.2477,
+      "epoch": 6.15633423180593,
+      "grad_norm": 0.7325547337532043,
+      "learning_rate": 0.00023101780895844572,
+      "loss": 3.2621,
       "step": 57100
     },
     {
-      "epoch": 6.151113981272199,
-      "grad_norm": 0.7094836235046387,
-      "learning_rate": 0.00023133714039435404,
-      "loss": 3.2509,
+      "epoch": 6.1617250673854445,
+      "grad_norm": 0.6908353567123413,
+      "learning_rate": 0.0002306940097139773,
+      "loss": 3.2767,
       "step": 57150
     },
     {
-      "epoch": 6.156495533311807,
-      "grad_norm": 0.7660449743270874,
-      "learning_rate": 0.0002310138993642926,
-      "loss": 3.2556,
+      "epoch": 6.16711590296496,
+      "grad_norm": 0.7702175378799438,
+      "learning_rate": 0.00023037021046950888,
+      "loss": 3.2516,
       "step": 57200
     },
     {
-      "epoch": 6.161877085351415,
-      "grad_norm": 0.7167730331420898,
-      "learning_rate": 0.0002306906583342312,
-      "loss": 3.2519,
+      "epoch": 6.172506738544475,
+      "grad_norm": 0.6899598240852356,
+      "learning_rate": 0.00023004641122504046,
+      "loss": 3.2521,
       "step": 57250
     },
     {
-      "epoch": 6.167258637391024,
-      "grad_norm": 0.7189817428588867,
-      "learning_rate": 0.00023036741730416977,
-      "loss": 3.2575,
+      "epoch": 6.177897574123989,
+      "grad_norm": 0.7570024132728577,
+      "learning_rate": 0.00022972261198057203,
+      "loss": 3.265,
       "step": 57300
     },
     {
-      "epoch": 6.1726401894306315,
-      "grad_norm": 0.7526877522468567,
-      "learning_rate": 0.00023004417627410837,
-      "loss": 3.264,
+      "epoch": 6.183288409703504,
+      "grad_norm": 0.6939279437065125,
+      "learning_rate": 0.0002293988127361036,
+      "loss": 3.2533,
       "step": 57350
     },
     {
-      "epoch": 6.17802174147024,
-      "grad_norm": 0.7627101540565491,
-      "learning_rate": 0.00022972093524404696,
-      "loss": 3.24,
+      "epoch": 6.188679245283019,
+      "grad_norm": 0.7017226815223694,
+      "learning_rate": 0.00022907501349163516,
+      "loss": 3.2629,
       "step": 57400
     },
     {
-      "epoch": 6.183403293509849,
-      "grad_norm": 0.7467467784881592,
-      "learning_rate": 0.00022939769421398553,
-      "loss": 3.2537,
+      "epoch": 6.1940700808625335,
+      "grad_norm": 0.7507325410842896,
+      "learning_rate": 0.00022875121424716674,
+      "loss": 3.2578,
       "step": 57450
     },
     {
-      "epoch": 6.188784845549456,
-      "grad_norm": 0.7287554144859314,
-      "learning_rate": 0.0002290744531839241,
-      "loss": 3.2784,
+      "epoch": 6.199460916442049,
+      "grad_norm": 0.7436730265617371,
+      "learning_rate": 0.00022842741500269832,
+      "loss": 3.252,
       "step": 57500
     },
     {
-      "epoch": 6.194166397589065,
-      "grad_norm": 0.7858840227127075,
-      "learning_rate": 0.00022875121215386272,
-      "loss": 3.2711,
+      "epoch": 6.204851752021563,
+      "grad_norm": 0.7096840143203735,
+      "learning_rate": 0.00022810361575822987,
+      "loss": 3.2719,
       "step": 57550
     },
     {
-      "epoch": 6.1995479496286725,
-      "grad_norm": 0.7679668664932251,
-      "learning_rate": 0.0002284279711238013,
-      "loss": 3.2466,
+      "epoch": 6.210242587601078,
+      "grad_norm": 0.7345703840255737,
+      "learning_rate": 0.00022777981651376145,
+      "loss": 3.2618,
       "step": 57600
     },
     {
-      "epoch": 6.204929501668281,
-      "grad_norm": 0.7927324175834656,
-      "learning_rate": 0.00022810473009373986,
-      "loss": 3.2551,
+      "epoch": 6.215633423180593,
+      "grad_norm": 0.7381048798561096,
+      "learning_rate": 0.00022745601726929302,
+      "loss": 3.2755,
       "step": 57650
     },
     {
-      "epoch": 6.21031105370789,
-      "grad_norm": 0.7803001403808594,
-      "learning_rate": 0.00022778148906367848,
-      "loss": 3.2639,
+      "epoch": 6.2210242587601075,
+      "grad_norm": 0.7399464845657349,
+      "learning_rate": 0.00022713221802482457,
+      "loss": 3.2659,
       "step": 57700
     },
     {
-      "epoch": 6.215692605747497,
-      "grad_norm": 0.7728279232978821,
-      "learning_rate": 0.00022745824803361705,
-      "loss": 3.2457,
+      "epoch": 6.226415094339623,
+      "grad_norm": 0.7439436316490173,
+      "learning_rate": 0.00022680841878035615,
+      "loss": 3.2576,
       "step": 57750
     },
     {
-      "epoch": 6.221074157787106,
-      "grad_norm": 0.7537714838981628,
-      "learning_rate": 0.00022713500700355564,
-      "loss": 3.2434,
+      "epoch": 6.231805929919138,
+      "grad_norm": 0.7336342334747314,
+      "learning_rate": 0.00022648461953588773,
+      "loss": 3.2533,
       "step": 57800
     },
     {
-      "epoch": 6.226455709826714,
-      "grad_norm": 0.7308183908462524,
-      "learning_rate": 0.0002268117659734942,
-      "loss": 3.2817,
+      "epoch": 6.237196765498652,
+      "grad_norm": 0.7191817760467529,
+      "learning_rate": 0.0002261608202914193,
+      "loss": 3.2674,
       "step": 57850
     },
     {
-      "epoch": 6.231837261866322,
-      "grad_norm": 0.7599002122879028,
-      "learning_rate": 0.0002264885249434328,
-      "loss": 3.2703,
+      "epoch": 6.242587601078167,
+      "grad_norm": 0.7303761839866638,
+      "learning_rate": 0.00022583702104695088,
+      "loss": 3.2754,
       "step": 57900
     },
     {
-      "epoch": 6.237218813905931,
-      "grad_norm": 0.7640085816383362,
-      "learning_rate": 0.0002261652839133714,
-      "loss": 3.2687,
+      "epoch": 6.247978436657682,
+      "grad_norm": 0.7404301762580872,
+      "learning_rate": 0.00022551322180248246,
+      "loss": 3.2668,
       "step": 57950
     },
     {
-      "epoch": 6.242600365945538,
-      "grad_norm": 0.7159485220909119,
-      "learning_rate": 0.00022584850770391118,
-      "loss": 3.2748,
+      "epoch": 6.2533692722371965,
+      "grad_norm": 0.7766687870025635,
+      "learning_rate": 0.00022518942255801399,
+      "loss": 3.2693,
       "step": 58000
     },
     {
-      "epoch": 6.242600365945538,
-      "eval_accuracy": 0.3850757218273112,
-      "eval_loss": 3.3764970302581787,
-      "eval_runtime": 198.0743,
-      "eval_samples_per_second": 90.931,
-      "eval_steps_per_second": 5.685,
+      "epoch": 6.2533692722371965,
+      "eval_accuracy": 0.3849878216410693,
+      "eval_loss": 3.3774213790893555,
+      "eval_runtime": 185.4873,
+      "eval_samples_per_second": 97.101,
+      "eval_steps_per_second": 6.07,
       "step": 58000
     },
     {
-      "epoch": 6.247981917985147,
-      "grad_norm": 0.7499499917030334,
-      "learning_rate": 0.0002255252666738498,
-      "loss": 3.2711,
+      "epoch": 6.258760107816712,
+      "grad_norm": 0.7129665613174438,
+      "learning_rate": 0.00022486562331354556,
+      "loss": 3.255,
       "step": 58050
     },
     {
-      "epoch": 6.253363470024755,
-      "grad_norm": 0.8008130788803101,
-      "learning_rate": 0.00022520202564378837,
-      "loss": 3.281,
+      "epoch": 6.264150943396227,
+      "grad_norm": 0.7495144605636597,
+      "learning_rate": 0.00022454182406907714,
+      "loss": 3.2688,
       "step": 58100
     },
     {
-      "epoch": 6.258745022064363,
-      "grad_norm": 0.7483268976211548,
-      "learning_rate": 0.00022487878461372693,
-      "loss": 3.2487,
+      "epoch": 6.269541778975741,
+      "grad_norm": 0.7381498217582703,
+      "learning_rate": 0.00022421802482460872,
+      "loss": 3.263,
       "step": 58150
     },
     {
-      "epoch": 6.264126574103972,
-      "grad_norm": 0.7136285305023193,
-      "learning_rate": 0.00022455554358366555,
-      "loss": 3.276,
+      "epoch": 6.274932614555256,
+      "grad_norm": 0.733660101890564,
+      "learning_rate": 0.0002238942255801403,
+      "loss": 3.2842,
       "step": 58200
     },
     {
-      "epoch": 6.26950812614358,
-      "grad_norm": 0.7955875992774963,
-      "learning_rate": 0.00022423230255360412,
-      "loss": 3.2697,
+      "epoch": 6.280323450134771,
+      "grad_norm": 0.7204157114028931,
+      "learning_rate": 0.00022357042633567187,
+      "loss": 3.2661,
       "step": 58250
     },
     {
-      "epoch": 6.274889678183188,
-      "grad_norm": 0.7933011651039124,
-      "learning_rate": 0.0002239090615235427,
-      "loss": 3.2698,
+      "epoch": 6.285714285714286,
+      "grad_norm": 0.7055049538612366,
+      "learning_rate": 0.00022324662709120345,
+      "loss": 3.257,
       "step": 58300
     },
     {
-      "epoch": 6.280271230222796,
-      "grad_norm": 0.7357099056243896,
-      "learning_rate": 0.00022358582049348128,
-      "loss": 3.2896,
+      "epoch": 6.291105121293801,
+      "grad_norm": 0.7392539381980896,
+      "learning_rate": 0.00022292282784673503,
+      "loss": 3.2796,
       "step": 58350
     },
     {
-      "epoch": 6.285652782262405,
-      "grad_norm": 0.7623982429504395,
-      "learning_rate": 0.00022326257946341988,
-      "loss": 3.2657,
+      "epoch": 6.296495956873315,
+      "grad_norm": 0.6930847764015198,
+      "learning_rate": 0.00022259902860226655,
+      "loss": 3.2618,
       "step": 58400
     },
     {
-      "epoch": 6.2910343343020125,
-      "grad_norm": 0.750810444355011,
-      "learning_rate": 0.00022293933843335845,
-      "loss": 3.2644,
+      "epoch": 6.30188679245283,
+      "grad_norm": 0.7264565229415894,
+      "learning_rate": 0.00022227522935779813,
+      "loss": 3.2887,
       "step": 58450
     },
     {
-      "epoch": 6.296415886341621,
-      "grad_norm": 0.7263330817222595,
-      "learning_rate": 0.00022261609740329704,
-      "loss": 3.2587,
+      "epoch": 6.307277628032345,
+      "grad_norm": 0.7048168778419495,
+      "learning_rate": 0.0002219514301133297,
+      "loss": 3.2859,
       "step": 58500
     },
     {
-      "epoch": 6.301797438381229,
-      "grad_norm": 0.7506089210510254,
-      "learning_rate": 0.0002222928563732356,
-      "loss": 3.252,
+      "epoch": 6.3126684636118595,
+      "grad_norm": 0.7758840322494507,
+      "learning_rate": 0.00022162763086886129,
+      "loss": 3.2758,
       "step": 58550
     },
     {
-      "epoch": 6.307178990420837,
-      "grad_norm": 0.7367995977401733,
-      "learning_rate": 0.00022196961534317423,
-      "loss": 3.2646,
+      "epoch": 6.318059299191375,
+      "grad_norm": 0.7596189379692078,
+      "learning_rate": 0.00022130383162439286,
+      "loss": 3.2785,
       "step": 58600
     },
     {
-      "epoch": 6.312560542460446,
-      "grad_norm": 0.8068094253540039,
-      "learning_rate": 0.000221652839133714,
-      "loss": 3.2717,
+      "epoch": 6.32345013477089,
+      "grad_norm": 0.7099698185920715,
+      "learning_rate": 0.00022098003237992444,
+      "loss": 3.2781,
       "step": 58650
     },
     {
-      "epoch": 6.3179420945000535,
-      "grad_norm": 0.758985161781311,
-      "learning_rate": 0.0002213295981036526,
-      "loss": 3.2693,
+      "epoch": 6.328840970350404,
+      "grad_norm": 0.7199275493621826,
+      "learning_rate": 0.00022065623313545602,
+      "loss": 3.2759,
       "step": 58700
     },
     {
-      "epoch": 6.323323646539662,
-      "grad_norm": 0.7637356519699097,
-      "learning_rate": 0.0002210063570735912,
-      "loss": 3.2843,
+      "epoch": 6.334231805929919,
+      "grad_norm": 0.7280547022819519,
+      "learning_rate": 0.00022033243389098757,
+      "loss": 3.2655,
       "step": 58750
     },
     {
-      "epoch": 6.328705198579271,
-      "grad_norm": 0.758793294429779,
-      "learning_rate": 0.00022068311604352977,
-      "loss": 3.249,
+      "epoch": 6.339622641509434,
+      "grad_norm": 0.7119089961051941,
+      "learning_rate": 0.00022000863464651915,
+      "loss": 3.2741,
       "step": 58800
     },
     {
-      "epoch": 6.334086750618878,
-      "grad_norm": 0.7414732575416565,
-      "learning_rate": 0.00022035987501346836,
-      "loss": 3.2642,
+      "epoch": 6.345013477088949,
+      "grad_norm": 0.7322922945022583,
+      "learning_rate": 0.0002196848354020507,
+      "loss": 3.286,
       "step": 58850
     },
     {
-      "epoch": 6.339468302658487,
-      "grad_norm": 0.7935400605201721,
-      "learning_rate": 0.00022003663398340693,
-      "loss": 3.2584,
+      "epoch": 6.350404312668464,
+      "grad_norm": 0.7355718016624451,
+      "learning_rate": 0.00021936103615758227,
+      "loss": 3.2721,
       "step": 58900
     },
     {
-      "epoch": 6.344849854698095,
-      "grad_norm": 0.8070433735847473,
-      "learning_rate": 0.00021971339295334552,
-      "loss": 3.2647,
+      "epoch": 6.355795148247978,
+      "grad_norm": 0.7406314015388489,
+      "learning_rate": 0.00021903723691311385,
+      "loss": 3.2692,
       "step": 58950
     },
     {
-      "epoch": 6.350231406737703,
-      "grad_norm": 0.7544232606887817,
-      "learning_rate": 0.00021939015192328412,
-      "loss": 3.2771,
+      "epoch": 6.361185983827493,
+      "grad_norm": 0.7210497856140137,
+      "learning_rate": 0.00021871343766864543,
+      "loss": 3.2868,
       "step": 59000
     },
     {
-      "epoch": 6.350231406737703,
-      "eval_accuracy": 0.38530650056102916,
-      "eval_loss": 3.3725526332855225,
-      "eval_runtime": 194.8804,
-      "eval_samples_per_second": 92.421,
-      "eval_steps_per_second": 5.778,
+      "epoch": 6.361185983827493,
+      "eval_accuracy": 0.3853350762705244,
+      "eval_loss": 3.373208522796631,
+      "eval_runtime": 184.9801,
+      "eval_samples_per_second": 97.367,
+      "eval_steps_per_second": 6.087,
       "step": 59000
     },
     {
-      "epoch": 6.355612958777312,
-      "grad_norm": 0.7711015939712524,
-      "learning_rate": 0.00021906691089322269,
-      "loss": 3.2659,
+      "epoch": 6.366576819407008,
+      "grad_norm": 0.7338991165161133,
+      "learning_rate": 0.00021838963842417698,
+      "loss": 3.2746,
       "step": 59050
     },
     {
-      "epoch": 6.360994510816919,
-      "grad_norm": 0.7567399740219116,
-      "learning_rate": 0.00021874366986316125,
-      "loss": 3.2869,
+      "epoch": 6.3719676549865225,
+      "grad_norm": 0.6929742097854614,
+      "learning_rate": 0.00021806583917970856,
+      "loss": 3.2719,
       "step": 59100
     },
     {
-      "epoch": 6.366376062856528,
-      "grad_norm": 0.8418982028961182,
-      "learning_rate": 0.00021842042883309988,
-      "loss": 3.2786,
+      "epoch": 6.377358490566038,
+      "grad_norm": 0.7513812184333801,
+      "learning_rate": 0.00021774203993524014,
+      "loss": 3.2666,
       "step": 59150
     },
     {
-      "epoch": 6.371757614896136,
-      "grad_norm": 0.7360608577728271,
-      "learning_rate": 0.00021809718780303844,
-      "loss": 3.2816,
+      "epoch": 6.382749326145553,
+      "grad_norm": 0.7775893807411194,
+      "learning_rate": 0.00021741824069077171,
+      "loss": 3.2702,
       "step": 59200
     },
     {
-      "epoch": 6.377139166935744,
-      "grad_norm": 0.7634429931640625,
-      "learning_rate": 0.000217773946772977,
-      "loss": 3.2519,
+      "epoch": 6.388140161725067,
+      "grad_norm": 0.7301859259605408,
+      "learning_rate": 0.00021709444144630326,
+      "loss": 3.2745,
       "step": 59250
     },
     {
-      "epoch": 6.382520718975353,
-      "grad_norm": 0.7280207872390747,
-      "learning_rate": 0.00021745070574291563,
-      "loss": 3.2656,
+      "epoch": 6.393530997304582,
+      "grad_norm": 0.7141174077987671,
+      "learning_rate": 0.00021677064220183484,
+      "loss": 3.2756,
       "step": 59300
     },
     {
-      "epoch": 6.387902271014961,
-      "grad_norm": 0.7721338868141174,
-      "learning_rate": 0.0002171274647128542,
-      "loss": 3.2525,
+      "epoch": 6.398921832884097,
+      "grad_norm": 0.70742267370224,
+      "learning_rate": 0.00021644684295736642,
+      "loss": 3.2724,
       "step": 59350
     },
     {
-      "epoch": 6.393283823054569,
-      "grad_norm": 0.7533543109893799,
-      "learning_rate": 0.0002168042236827928,
-      "loss": 3.262,
+      "epoch": 6.404312668463612,
+      "grad_norm": 0.732514500617981,
+      "learning_rate": 0.00021612304371289797,
+      "loss": 3.2977,
       "step": 59400
     },
     {
-      "epoch": 6.398665375094177,
-      "grad_norm": 0.7705949544906616,
-      "learning_rate": 0.00021648098265273136,
-      "loss": 3.2626,
+      "epoch": 6.409703504043127,
+      "grad_norm": 0.8343349099159241,
+      "learning_rate": 0.00021579924446842955,
+      "loss": 3.2829,
       "step": 59450
     },
     {
-      "epoch": 6.404046927133785,
-      "grad_norm": 0.7239140868186951,
-      "learning_rate": 0.00021615774162266996,
-      "loss": 3.2778,
+      "epoch": 6.415094339622642,
+      "grad_norm": 0.7617896199226379,
+      "learning_rate": 0.00021547544522396113,
+      "loss": 3.2697,
       "step": 59500
     },
     {
-      "epoch": 6.4094284791733935,
-      "grad_norm": 0.802168071269989,
-      "learning_rate": 0.00021583450059260855,
-      "loss": 3.2704,
+      "epoch": 6.420485175202156,
+      "grad_norm": 0.7627173662185669,
+      "learning_rate": 0.0002151516459794927,
+      "loss": 3.2653,
       "step": 59550
     },
     {
-      "epoch": 6.414810031213002,
-      "grad_norm": 0.7422902584075928,
-      "learning_rate": 0.00021551125956254712,
-      "loss": 3.2768,
+      "epoch": 6.425876010781671,
+      "grad_norm": 0.7342140078544617,
+      "learning_rate": 0.00021482784673502428,
+      "loss": 3.2902,
       "step": 59600
     },
     {
-      "epoch": 6.42019158325261,
-      "grad_norm": 0.737062394618988,
-      "learning_rate": 0.00021518801853248569,
-      "loss": 3.2592,
+      "epoch": 6.431266846361186,
+      "grad_norm": 0.7224741578102112,
+      "learning_rate": 0.00021450404749055586,
+      "loss": 3.2801,
       "step": 59650
     },
     {
-      "epoch": 6.425573135292218,
-      "grad_norm": 0.7138918042182922,
-      "learning_rate": 0.0002148647775024243,
-      "loss": 3.2859,
+      "epoch": 6.436657681940701,
+      "grad_norm": 0.7586390376091003,
+      "learning_rate": 0.00021418024824608738,
+      "loss": 3.2881,
       "step": 59700
     },
     {
-      "epoch": 6.430954687331827,
-      "grad_norm": 0.8071754574775696,
-      "learning_rate": 0.00021454153647236288,
-      "loss": 3.2687,
+      "epoch": 6.442048517520216,
+      "grad_norm": 0.7512621283531189,
+      "learning_rate": 0.00021385644900161896,
+      "loss": 3.287,
       "step": 59750
     },
     {
-      "epoch": 6.4363362393714345,
-      "grad_norm": 0.7855433821678162,
-      "learning_rate": 0.00021421829544230144,
-      "loss": 3.2828,
+      "epoch": 6.44743935309973,
+      "grad_norm": 0.7532150745391846,
+      "learning_rate": 0.00021353264975715054,
+      "loss": 3.2775,
       "step": 59800
     },
     {
-      "epoch": 6.441717791411043,
-      "grad_norm": 0.7580162882804871,
-      "learning_rate": 0.00021389505441224007,
-      "loss": 3.275,
+      "epoch": 6.452830188679245,
+      "grad_norm": 0.7296251058578491,
+      "learning_rate": 0.00021320885051268211,
+      "loss": 3.2628,
       "step": 59850
     },
     {
-      "epoch": 6.447099343450651,
-      "grad_norm": 0.8188996911048889,
-      "learning_rate": 0.00021357181338217863,
-      "loss": 3.2723,
+      "epoch": 6.45822102425876,
+      "grad_norm": 0.7432535290718079,
+      "learning_rate": 0.0002128850512682137,
+      "loss": 3.2747,
       "step": 59900
     },
     {
-      "epoch": 6.452480895490259,
-      "grad_norm": 0.7686909437179565,
-      "learning_rate": 0.0002132485723521172,
-      "loss": 3.2703,
+      "epoch": 6.463611859838275,
+      "grad_norm": 0.7279696464538574,
+      "learning_rate": 0.00021256125202374527,
+      "loss": 3.2652,
       "step": 59950
     },
     {
-      "epoch": 6.457862447529868,
-      "grad_norm": 0.8067818880081177,
-      "learning_rate": 0.0002129253313220558,
-      "loss": 3.2855,
+      "epoch": 6.46900269541779,
+      "grad_norm": 0.7320942282676697,
+      "learning_rate": 0.00021223745277927685,
+      "loss": 3.2781,
       "step": 60000
     },
     {
-      "epoch": 6.457862447529868,
-      "eval_accuracy": 0.3853550684018823,
-      "eval_loss": 3.369478464126587,
-      "eval_runtime": 188.6392,
-      "eval_samples_per_second": 95.479,
-      "eval_steps_per_second": 5.969,
+      "epoch": 6.46900269541779,
+      "eval_accuracy": 0.38602882495921986,
+      "eval_loss": 3.3693761825561523,
+      "eval_runtime": 185.2629,
+      "eval_samples_per_second": 97.219,
+      "eval_steps_per_second": 6.078,
       "step": 60000
     },
     {
-      "epoch": 6.4632439995694755,
-      "grad_norm": 0.7619180083274841,
-      "learning_rate": 0.00021260209029199436,
-      "loss": 3.2692,
+      "epoch": 6.474393530997305,
+      "grad_norm": 0.7365020513534546,
+      "learning_rate": 0.00021191365353480843,
+      "loss": 3.2728,
       "step": 60050
     },
     {
-      "epoch": 6.468625551609084,
-      "grad_norm": 0.7411639094352722,
-      "learning_rate": 0.00021227884926193296,
-      "loss": 3.2696,
+      "epoch": 6.479784366576819,
+      "grad_norm": 0.7278674840927124,
+      "learning_rate": 0.00021159633027522932,
+      "loss": 3.2865,
       "step": 60100
     },
     {
-      "epoch": 6.474007103648693,
-      "grad_norm": 0.7599124312400818,
-      "learning_rate": 0.00021195560823187155,
-      "loss": 3.2765,
+      "epoch": 6.485175202156334,
+      "grad_norm": 0.7251003384590149,
+      "learning_rate": 0.0002112725310307609,
+      "loss": 3.2635,
       "step": 60150
     },
     {
-      "epoch": 6.4793886556883,
-      "grad_norm": 0.7548618912696838,
-      "learning_rate": 0.00021163236720181012,
-      "loss": 3.2775,
+      "epoch": 6.490566037735849,
+      "grad_norm": 0.7077476978302002,
+      "learning_rate": 0.00021094873178629248,
+      "loss": 3.2768,
       "step": 60200
     },
     {
-      "epoch": 6.484770207727909,
-      "grad_norm": 0.7517174482345581,
-      "learning_rate": 0.0002113091261717487,
-      "loss": 3.3001,
+      "epoch": 6.495956873315364,
+      "grad_norm": 0.7250546813011169,
+      "learning_rate": 0.00021062493254182406,
+      "loss": 3.2729,
       "step": 60250
     },
     {
-      "epoch": 6.490151759767517,
-      "grad_norm": 0.7505078315734863,
-      "learning_rate": 0.0002109858851416873,
-      "loss": 3.2905,
+      "epoch": 6.501347708894879,
+      "grad_norm": 0.7367532849311829,
+      "learning_rate": 0.00021030113329735563,
+      "loss": 3.2688,
       "step": 60300
     },
     {
-      "epoch": 6.495533311807125,
-      "grad_norm": 0.7407610416412354,
-      "learning_rate": 0.00021066264411162588,
-      "loss": 3.2746,
+      "epoch": 6.506738544474393,
+      "grad_norm": 0.7516636252403259,
+      "learning_rate": 0.0002099773340528872,
+      "loss": 3.2542,
       "step": 60350
     },
     {
-      "epoch": 6.500914863846734,
-      "grad_norm": 0.8135571479797363,
-      "learning_rate": 0.00021033940308156447,
-      "loss": 3.2772,
+      "epoch": 6.512129380053908,
+      "grad_norm": 0.7502709627151489,
+      "learning_rate": 0.0002096535348084188,
+      "loss": 3.2852,
       "step": 60400
     },
     {
-      "epoch": 6.506296415886341,
-      "grad_norm": 0.7895756363868713,
-      "learning_rate": 0.00021001616205150307,
+      "epoch": 6.517520215633423,
+      "grad_norm": 0.7901685833930969,
+      "learning_rate": 0.0002093297355639503,
       "loss": 3.2731,
       "step": 60450
     },
     {
-      "epoch": 6.51167796792595,
-      "grad_norm": 0.7601107954978943,
-      "learning_rate": 0.00020969292102144163,
-      "loss": 3.2905,
+      "epoch": 6.5229110512129385,
+      "grad_norm": 0.7563238739967346,
+      "learning_rate": 0.0002090059363194819,
+      "loss": 3.2696,
       "step": 60500
     },
     {
-      "epoch": 6.517059519965558,
-      "grad_norm": 0.8015565872192383,
-      "learning_rate": 0.00020936967999138023,
-      "loss": 3.276,
+      "epoch": 6.528301886792453,
+      "grad_norm": 0.7389952540397644,
+      "learning_rate": 0.00020868213707501347,
+      "loss": 3.2925,
       "step": 60550
     },
     {
-      "epoch": 6.522441072005166,
-      "grad_norm": 0.8048959970474243,
-      "learning_rate": 0.0002090464389613188,
-      "loss": 3.2836,
+      "epoch": 6.533692722371968,
+      "grad_norm": 0.7402742505073547,
+      "learning_rate": 0.00020835833783054505,
+      "loss": 3.2653,
       "step": 60600
     },
     {
-      "epoch": 6.5278226240447745,
-      "grad_norm": 0.7634525895118713,
-      "learning_rate": 0.0002087231979312574,
-      "loss": 3.2812,
+      "epoch": 6.539083557951482,
+      "grad_norm": 0.7205880880355835,
+      "learning_rate": 0.00020803453858607662,
+      "loss": 3.2817,
       "step": 60650
     },
     {
-      "epoch": 6.533204176084383,
-      "grad_norm": 0.7490655183792114,
-      "learning_rate": 0.00020839995690119598,
-      "loss": 3.2882,
+      "epoch": 6.544474393530997,
+      "grad_norm": 0.7250434160232544,
+      "learning_rate": 0.0002077107393416082,
+      "loss": 3.2812,
       "step": 60700
     },
     {
-      "epoch": 6.538585728123991,
-      "grad_norm": 0.7527089715003967,
-      "learning_rate": 0.00020807671587113455,
+      "epoch": 6.549865229110512,
+      "grad_norm": 0.7092582583427429,
+      "learning_rate": 0.00020738694009713975,
       "loss": 3.283,
       "step": 60750
     },
     {
-      "epoch": 6.543967280163599,
-      "grad_norm": 0.7919941544532776,
-      "learning_rate": 0.00020775347484107312,
-      "loss": 3.2847,
+      "epoch": 6.555256064690027,
+      "grad_norm": 0.7592464685440063,
+      "learning_rate": 0.00020706314085267133,
+      "loss": 3.278,
       "step": 60800
     },
     {
-      "epoch": 6.549348832203208,
-      "grad_norm": 0.7546364665031433,
-      "learning_rate": 0.00020743023381101174,
-      "loss": 3.2682,
+      "epoch": 6.560646900269542,
+      "grad_norm": 0.7507815361022949,
+      "learning_rate": 0.00020673934160820288,
+      "loss": 3.2695,
       "step": 60850
     },
     {
-      "epoch": 6.5547303842428155,
-      "grad_norm": 0.7347162961959839,
-      "learning_rate": 0.0002071069927809503,
-      "loss": 3.2528,
+      "epoch": 6.566037735849057,
+      "grad_norm": 0.7489672899246216,
+      "learning_rate": 0.00020642201834862383,
+      "loss": 3.2636,
       "step": 60900
     },
     {
-      "epoch": 6.560111936282424,
-      "grad_norm": 0.7587774991989136,
-      "learning_rate": 0.00020678375175088888,
-      "loss": 3.2634,
+      "epoch": 6.571428571428571,
+      "grad_norm": 0.7646813988685608,
+      "learning_rate": 0.0002060982191041554,
+      "loss": 3.2822,
       "step": 60950
     },
     {
-      "epoch": 6.565493488322032,
-      "grad_norm": 0.7877184748649597,
-      "learning_rate": 0.0002064605107208275,
-      "loss": 3.274,
+      "epoch": 6.576819407008086,
+      "grad_norm": 0.7060157656669617,
+      "learning_rate": 0.000205774419859687,
+      "loss": 3.271,
       "step": 61000
     },
     {
-      "epoch": 6.565493488322032,
-      "eval_accuracy": 0.38548545186725963,
-      "eval_loss": 3.365142822265625,
-      "eval_runtime": 189.5759,
-      "eval_samples_per_second": 95.007,
-      "eval_steps_per_second": 5.94,
+      "epoch": 6.576819407008086,
+      "eval_accuracy": 0.3863094753684447,
+      "eval_loss": 3.3651201725006104,
+      "eval_runtime": 185.4691,
+      "eval_samples_per_second": 97.11,
+      "eval_steps_per_second": 6.071,
       "step": 61000
     },
     {
-      "epoch": 6.57087504036164,
-      "grad_norm": 0.7893941402435303,
-      "learning_rate": 0.00020613726969076607,
-      "loss": 3.2793,
+      "epoch": 6.5822102425876015,
+      "grad_norm": 0.7125557661056519,
+      "learning_rate": 0.00020545062061521857,
+      "loss": 3.2676,
       "step": 61050
     },
     {
-      "epoch": 6.576256592401249,
-      "grad_norm": 0.7795023322105408,
-      "learning_rate": 0.00020581402866070463,
-      "loss": 3.2695,
+      "epoch": 6.587601078167116,
+      "grad_norm": 0.728756308555603,
+      "learning_rate": 0.00020512682137075012,
+      "loss": 3.275,
       "step": 61100
     },
     {
-      "epoch": 6.5816381444408565,
-      "grad_norm": 0.7646381258964539,
-      "learning_rate": 0.00020549078763064323,
-      "loss": 3.2903,
+      "epoch": 6.592991913746631,
+      "grad_norm": 0.7177598476409912,
+      "learning_rate": 0.0002048030221262817,
+      "loss": 3.2906,
       "step": 61150
     },
     {
-      "epoch": 6.587019696480465,
-      "grad_norm": 0.8036014437675476,
-      "learning_rate": 0.00020516754660058182,
-      "loss": 3.2689,
+      "epoch": 6.598382749326145,
+      "grad_norm": 0.7181693911552429,
+      "learning_rate": 0.00020447922288181324,
+      "loss": 3.2873,
       "step": 61200
     },
     {
-      "epoch": 6.592401248520073,
-      "grad_norm": 0.7395958304405212,
-      "learning_rate": 0.0002048443055705204,
-      "loss": 3.2766,
+      "epoch": 6.60377358490566,
+      "grad_norm": 0.7636637687683105,
+      "learning_rate": 0.00020415542363734482,
+      "loss": 3.2776,
       "step": 61250
     },
     {
-      "epoch": 6.597782800559681,
-      "grad_norm": 0.7397066354751587,
-      "learning_rate": 0.00020452106454045898,
-      "loss": 3.2817,
+      "epoch": 6.609164420485175,
+      "grad_norm": 0.7477648854255676,
+      "learning_rate": 0.0002038316243928764,
+      "loss": 3.2703,
       "step": 61300
     },
     {
-      "epoch": 6.60316435259929,
-      "grad_norm": 0.7858966588973999,
-      "learning_rate": 0.00020419782351039755,
-      "loss": 3.2744,
+      "epoch": 6.6145552560646905,
+      "grad_norm": 0.8419035077095032,
+      "learning_rate": 0.00020350782514840798,
+      "loss": 3.2828,
       "step": 61350
     },
     {
-      "epoch": 6.608545904638898,
-      "grad_norm": 0.9551375508308411,
-      "learning_rate": 0.00020387458248033617,
-      "loss": 3.2923,
+      "epoch": 6.619946091644205,
+      "grad_norm": 0.7486584782600403,
+      "learning_rate": 0.00020318402590393953,
+      "loss": 3.2835,
       "step": 61400
     },
     {
-      "epoch": 6.613927456678506,
-      "grad_norm": 0.7568668723106384,
-      "learning_rate": 0.00020355134145027474,
-      "loss": 3.2774,
+      "epoch": 6.62533692722372,
+      "grad_norm": 0.7403963804244995,
+      "learning_rate": 0.0002028602266594711,
+      "loss": 3.2799,
       "step": 61450
     },
     {
-      "epoch": 6.619309008718115,
-      "grad_norm": 0.8731978535652161,
-      "learning_rate": 0.0002032281004202133,
-      "loss": 3.284,
+      "epoch": 6.630727762803234,
+      "grad_norm": 0.7406607866287231,
+      "learning_rate": 0.00020253642741500268,
+      "loss": 3.2743,
       "step": 61500
     },
     {
-      "epoch": 6.624690560757722,
-      "grad_norm": 0.7871314883232117,
-      "learning_rate": 0.00020290485939015193,
-      "loss": 3.2765,
+      "epoch": 6.636118598382749,
+      "grad_norm": 0.7461050152778625,
+      "learning_rate": 0.00020221262817053426,
+      "loss": 3.2737,
       "step": 61550
     },
     {
-      "epoch": 6.630072112797331,
-      "grad_norm": 0.7354938983917236,
-      "learning_rate": 0.0002025816183600905,
-      "loss": 3.2641,
+      "epoch": 6.6415094339622645,
+      "grad_norm": 0.7818842530250549,
+      "learning_rate": 0.00020188882892606584,
+      "loss": 3.3006,
       "step": 61600
     },
     {
-      "epoch": 6.635453664836939,
-      "grad_norm": 0.7317883968353271,
-      "learning_rate": 0.00020225837733002907,
-      "loss": 3.2747,
+      "epoch": 6.646900269541779,
+      "grad_norm": 0.7182121872901917,
+      "learning_rate": 0.0002015650296815974,
+      "loss": 3.2571,
       "step": 61650
     },
     {
-      "epoch": 6.640835216876547,
-      "grad_norm": 0.7435038089752197,
-      "learning_rate": 0.00020193513629996766,
-      "loss": 3.2945,
+      "epoch": 6.652291105121294,
+      "grad_norm": 0.79462730884552,
+      "learning_rate": 0.00020124123043712897,
+      "loss": 3.2773,
       "step": 61700
     },
     {
-      "epoch": 6.6462167689161555,
-      "grad_norm": 0.7606987953186035,
-      "learning_rate": 0.00020161189526990626,
-      "loss": 3.2753,
+      "epoch": 6.657681940700809,
+      "grad_norm": 0.790791392326355,
+      "learning_rate": 0.00020091743119266052,
+      "loss": 3.2923,
       "step": 61750
     },
     {
-      "epoch": 6.651598320955763,
-      "grad_norm": 0.8060020804405212,
-      "learning_rate": 0.00020128865423984482,
-      "loss": 3.2658,
+      "epoch": 6.663072776280323,
+      "grad_norm": 0.779761016368866,
+      "learning_rate": 0.0002005936319481921,
+      "loss": 3.2776,
       "step": 61800
     },
     {
-      "epoch": 6.656979872995372,
-      "grad_norm": 0.7530675530433655,
-      "learning_rate": 0.00020096541320978342,
-      "loss": 3.2544,
+      "epoch": 6.668463611859838,
+      "grad_norm": 0.7565503120422363,
+      "learning_rate": 0.00020026983270372367,
+      "loss": 3.2889,
       "step": 61850
     },
     {
-      "epoch": 6.66236142503498,
-      "grad_norm": 0.7579043507575989,
-      "learning_rate": 0.00020064217217972199,
-      "loss": 3.2715,
+      "epoch": 6.6738544474393535,
+      "grad_norm": 0.7549183964729309,
+      "learning_rate": 0.00019994603345925525,
+      "loss": 3.2779,
       "step": 61900
     },
     {
-      "epoch": 6.667742977074588,
-      "grad_norm": 0.7606813311576843,
-      "learning_rate": 0.00020031893114966058,
-      "loss": 3.2528,
+      "epoch": 6.679245283018868,
+      "grad_norm": 0.7439727783203125,
+      "learning_rate": 0.00019962223421478683,
+      "loss": 3.2726,
       "step": 61950
     },
     {
-      "epoch": 6.6731245291141965,
-      "grad_norm": 0.8275402784347534,
-      "learning_rate": 0.00019999569011959917,
-      "loss": 3.2694,
+      "epoch": 6.684636118598383,
+      "grad_norm": 0.7546395659446716,
+      "learning_rate": 0.0001992984349703184,
+      "loss": 3.2725,
       "step": 62000
     },
     {
-      "epoch": 6.6731245291141965,
-      "eval_accuracy": 0.3865296061191568,
-      "eval_loss": 3.3617382049560547,
-      "eval_runtime": 198.9452,
-      "eval_samples_per_second": 90.532,
-      "eval_steps_per_second": 5.66,
+      "epoch": 6.684636118598383,
+      "eval_accuracy": 0.38652612922674673,
+      "eval_loss": 3.361262798309326,
+      "eval_runtime": 183.7014,
+      "eval_samples_per_second": 98.045,
+      "eval_steps_per_second": 6.13,
       "step": 62000
     },
     {
-      "epoch": 6.678506081153805,
-      "grad_norm": 0.7591835856437683,
-      "learning_rate": 0.00019967244908953774,
-      "loss": 3.2691,
+      "epoch": 6.690026954177897,
+      "grad_norm": 0.7109931111335754,
+      "learning_rate": 0.00019897463572584993,
+      "loss": 3.2688,
       "step": 62050
     },
     {
-      "epoch": 6.683887633193413,
-      "grad_norm": 0.7736212015151978,
-      "learning_rate": 0.0001993492080594763,
-      "loss": 3.2896,
+      "epoch": 6.695417789757412,
+      "grad_norm": 0.7498511672019958,
+      "learning_rate": 0.0001986508364813815,
+      "loss": 3.2773,
       "step": 62100
     },
     {
-      "epoch": 6.689269185233021,
-      "grad_norm": 0.7579230070114136,
-      "learning_rate": 0.00019902596702941493,
-      "loss": 3.2785,
+      "epoch": 6.7008086253369274,
+      "grad_norm": 0.8148815035820007,
+      "learning_rate": 0.00019832703723691308,
+      "loss": 3.2928,
       "step": 62150
     },
     {
-      "epoch": 6.69465073727263,
-      "grad_norm": 0.7463725209236145,
-      "learning_rate": 0.0001987027259993535,
-      "loss": 3.2967,
+      "epoch": 6.706199460916442,
+      "grad_norm": 0.751923143863678,
+      "learning_rate": 0.00019800323799244466,
+      "loss": 3.3001,
       "step": 62200
     },
     {
-      "epoch": 6.7000322893122375,
-      "grad_norm": 0.7582423090934753,
-      "learning_rate": 0.00019837948496929207,
-      "loss": 3.2791,
+      "epoch": 6.711590296495957,
+      "grad_norm": 0.7420632839202881,
+      "learning_rate": 0.00019767943874797624,
+      "loss": 3.2741,
       "step": 62250
     },
     {
-      "epoch": 6.705413841351846,
-      "grad_norm": 0.7430950999259949,
-      "learning_rate": 0.00019805624393923066,
-      "loss": 3.2677,
+      "epoch": 6.716981132075472,
+      "grad_norm": 0.755337655544281,
+      "learning_rate": 0.00019735563950350782,
+      "loss": 3.2794,
       "step": 62300
     },
     {
-      "epoch": 6.710795393391454,
-      "grad_norm": 0.7181265950202942,
-      "learning_rate": 0.00019773300290916926,
-      "loss": 3.2925,
+      "epoch": 6.722371967654986,
+      "grad_norm": 0.7848864197731018,
+      "learning_rate": 0.0001970318402590394,
+      "loss": 3.2819,
       "step": 62350
     },
     {
-      "epoch": 6.716176945431062,
-      "grad_norm": 0.7656122446060181,
-      "learning_rate": 0.00019740976187910785,
-      "loss": 3.2809,
+      "epoch": 6.727762803234501,
+      "grad_norm": 0.7375281453132629,
+      "learning_rate": 0.00019670804101457097,
+      "loss": 3.2642,
       "step": 62400
     },
     {
-      "epoch": 6.721558497470671,
-      "grad_norm": 0.7912488579750061,
-      "learning_rate": 0.00019708652084904642,
-      "loss": 3.2577,
+      "epoch": 6.7331536388140165,
+      "grad_norm": 0.7862948775291443,
+      "learning_rate": 0.00019638424177010252,
+      "loss": 3.2709,
       "step": 62450
     },
     {
-      "epoch": 6.7269400495102785,
-      "grad_norm": 0.7884469628334045,
-      "learning_rate": 0.00019676327981898499,
-      "loss": 3.2782,
+      "epoch": 6.738544474393531,
+      "grad_norm": 0.7952997088432312,
+      "learning_rate": 0.00019606044252563407,
+      "loss": 3.2813,
       "step": 62500
     },
     {
-      "epoch": 6.732321601549887,
-      "grad_norm": 0.818396806716919,
-      "learning_rate": 0.0001964400387889236,
-      "loss": 3.2791,
+      "epoch": 6.743935309973046,
+      "grad_norm": 0.7619608044624329,
+      "learning_rate": 0.00019573664328116565,
+      "loss": 3.2781,
       "step": 62550
     },
     {
-      "epoch": 6.737703153589496,
-      "grad_norm": 0.8610832691192627,
-      "learning_rate": 0.00019611679775886217,
-      "loss": 3.2645,
+      "epoch": 6.74932614555256,
+      "grad_norm": 0.7676348686218262,
+      "learning_rate": 0.00019541284403669723,
+      "loss": 3.273,
       "step": 62600
     },
     {
-      "epoch": 6.743084705629103,
-      "grad_norm": 0.7635974884033203,
-      "learning_rate": 0.00019579355672880074,
-      "loss": 3.265,
+      "epoch": 6.754716981132075,
+      "grad_norm": 0.792212963104248,
+      "learning_rate": 0.0001950890447922288,
+      "loss": 3.2846,
       "step": 62650
     },
     {
-      "epoch": 6.748466257668712,
-      "grad_norm": 0.7944522500038147,
-      "learning_rate": 0.00019547031569873936,
-      "loss": 3.2626,
+      "epoch": 6.7601078167115904,
+      "grad_norm": 0.8594814538955688,
+      "learning_rate": 0.00019476524554776038,
+      "loss": 3.2743,
       "step": 62700
     },
     {
-      "epoch": 6.75384780970832,
-      "grad_norm": 0.7731636762619019,
-      "learning_rate": 0.00019514707466867793,
-      "loss": 3.3033,
+      "epoch": 6.765498652291106,
+      "grad_norm": 0.706619918346405,
+      "learning_rate": 0.00019444144630329193,
+      "loss": 3.2912,
       "step": 62750
     },
     {
-      "epoch": 6.759229361747928,
-      "grad_norm": 0.7597733736038208,
-      "learning_rate": 0.0001948238336386165,
-      "loss": 3.2774,
+      "epoch": 6.77088948787062,
+      "grad_norm": 0.7490671873092651,
+      "learning_rate": 0.0001941176470588235,
+      "loss": 3.289,
       "step": 62800
     },
     {
-      "epoch": 6.7646109137875365,
-      "grad_norm": 0.7862128019332886,
-      "learning_rate": 0.0001945005926085551,
-      "loss": 3.2861,
+      "epoch": 6.776280323450135,
+      "grad_norm": 0.7888407111167908,
+      "learning_rate": 0.0001937938478143551,
+      "loss": 3.2825,
       "step": 62850
     },
     {
-      "epoch": 6.769992465827144,
-      "grad_norm": 0.7566255331039429,
-      "learning_rate": 0.0001941773515784937,
-      "loss": 3.2723,
+      "epoch": 6.781671159029649,
+      "grad_norm": 0.7556346654891968,
+      "learning_rate": 0.00019347004856988664,
+      "loss": 3.2768,
       "step": 62900
     },
     {
-      "epoch": 6.775374017866753,
-      "grad_norm": 0.7786105871200562,
-      "learning_rate": 0.00019385411054843226,
-      "loss": 3.2988,
+      "epoch": 6.787061994609164,
+      "grad_norm": 0.7548149228096008,
+      "learning_rate": 0.00019314624932541822,
+      "loss": 3.2774,
       "step": 62950
     },
     {
-      "epoch": 6.780755569906361,
-      "grad_norm": 0.7658774256706238,
-      "learning_rate": 0.00019353086951837085,
-      "loss": 3.2922,
+      "epoch": 6.7924528301886795,
+      "grad_norm": 0.7699045538902283,
+      "learning_rate": 0.0001928224500809498,
+      "loss": 3.2818,
       "step": 63000
     },
     {
-      "epoch": 6.780755569906361,
-      "eval_accuracy": 0.386855130171049,
-      "eval_loss": 3.356630563735962,
-      "eval_runtime": 189.3045,
-      "eval_samples_per_second": 95.143,
-      "eval_steps_per_second": 5.948,
+      "epoch": 6.7924528301886795,
+      "eval_accuracy": 0.3871022068379391,
+      "eval_loss": 3.3578391075134277,
+      "eval_runtime": 184.1148,
+      "eval_samples_per_second": 97.825,
+      "eval_steps_per_second": 6.116,
       "step": 63000
     },
     {
-      "epoch": 6.786137121945969,
-      "grad_norm": 0.8002941608428955,
-      "learning_rate": 0.00019321409330891066,
-      "loss": 3.2842,
+      "epoch": 6.797843665768194,
+      "grad_norm": 0.7240121960639954,
+      "learning_rate": 0.00019249865083648137,
+      "loss": 3.2666,
       "step": 63050
     },
     {
-      "epoch": 6.7915186739855775,
-      "grad_norm": 0.7723585963249207,
-      "learning_rate": 0.00019289085227884925,
-      "loss": 3.28,
+      "epoch": 6.803234501347709,
+      "grad_norm": 0.825818657875061,
+      "learning_rate": 0.00019217485159201292,
+      "loss": 3.2924,
       "step": 63100
     },
     {
-      "epoch": 6.796900226025185,
-      "grad_norm": 0.7752054333686829,
-      "learning_rate": 0.00019256761124878782,
-      "loss": 3.2727,
+      "epoch": 6.808625336927224,
+      "grad_norm": 0.7854468822479248,
+      "learning_rate": 0.0001918510523475445,
+      "loss": 3.2781,
       "step": 63150
     },
     {
-      "epoch": 6.802281778064794,
-      "grad_norm": 0.7694298624992371,
-      "learning_rate": 0.00019224437021872644,
-      "loss": 3.2716,
+      "epoch": 6.814016172506738,
+      "grad_norm": 0.745103120803833,
+      "learning_rate": 0.00019152725310307608,
+      "loss": 3.2773,
       "step": 63200
     },
     {
-      "epoch": 6.807663330104402,
-      "grad_norm": 0.7400794625282288,
-      "learning_rate": 0.000191921129188665,
-      "loss": 3.2845,
+      "epoch": 6.819407008086253,
+      "grad_norm": 0.7766667008399963,
+      "learning_rate": 0.00019120345385860766,
+      "loss": 3.2856,
       "step": 63250
     },
     {
-      "epoch": 6.813044882144011,
-      "grad_norm": 0.7604348659515381,
-      "learning_rate": 0.00019159788815860358,
-      "loss": 3.2862,
+      "epoch": 6.824797843665769,
+      "grad_norm": 0.8103490471839905,
+      "learning_rate": 0.00019087965461413923,
+      "loss": 3.2834,
       "step": 63300
     },
     {
-      "epoch": 6.8184264341836185,
-      "grad_norm": 0.7525448203086853,
-      "learning_rate": 0.00019127464712854217,
-      "loss": 3.2706,
+      "epoch": 6.830188679245283,
+      "grad_norm": 0.7124233245849609,
+      "learning_rate": 0.00019055585536967079,
+      "loss": 3.2765,
       "step": 63350
     },
     {
-      "epoch": 6.823807986223227,
-      "grad_norm": 0.7643925547599792,
-      "learning_rate": 0.00019095140609848077,
-      "loss": 3.2678,
+      "epoch": 6.835579514824798,
+      "grad_norm": 0.7679892778396606,
+      "learning_rate": 0.00019023205612520234,
+      "loss": 3.2712,
       "step": 63400
     },
     {
-      "epoch": 6.829189538262835,
-      "grad_norm": 0.7530627250671387,
-      "learning_rate": 0.00019062816506841933,
-      "loss": 3.2508,
+      "epoch": 6.840970350404312,
+      "grad_norm": 0.7390249371528625,
+      "learning_rate": 0.00018990825688073391,
+      "loss": 3.2741,
       "step": 63450
     },
     {
-      "epoch": 6.834571090302443,
-      "grad_norm": 0.7988157272338867,
-      "learning_rate": 0.00019030492403835793,
-      "loss": 3.2714,
+      "epoch": 6.846361185983827,
+      "grad_norm": 0.6974737644195557,
+      "learning_rate": 0.0001895844576362655,
+      "loss": 3.2758,
       "step": 63500
     },
     {
-      "epoch": 6.839952642342052,
-      "grad_norm": 0.8077282905578613,
-      "learning_rate": 0.0001899816830082965,
-      "loss": 3.2773,
+      "epoch": 6.8517520215633425,
+      "grad_norm": 0.7622225284576416,
+      "learning_rate": 0.00018926065839179707,
+      "loss": 3.2937,
       "step": 63550
     },
     {
-      "epoch": 6.8453341943816595,
-      "grad_norm": 0.7729781866073608,
-      "learning_rate": 0.00018965844197823506,
-      "loss": 3.2604,
+      "epoch": 6.857142857142857,
+      "grad_norm": 0.818612813949585,
+      "learning_rate": 0.00018893685914732865,
+      "loss": 3.28,
       "step": 63600
     },
     {
-      "epoch": 6.850715746421268,
-      "grad_norm": 0.7184460163116455,
-      "learning_rate": 0.00018933520094817368,
-      "loss": 3.2719,
+      "epoch": 6.862533692722372,
+      "grad_norm": 0.7427046895027161,
+      "learning_rate": 0.00018861305990286022,
+      "loss": 3.2822,
       "step": 63650
     },
     {
-      "epoch": 6.856097298460876,
-      "grad_norm": 0.7712047100067139,
-      "learning_rate": 0.00018901195991811225,
-      "loss": 3.2841,
+      "epoch": 6.867924528301887,
+      "grad_norm": 0.7735974788665771,
+      "learning_rate": 0.0001882892606583918,
+      "loss": 3.2591,
       "step": 63700
     },
     {
-      "epoch": 6.861478850500484,
-      "grad_norm": 0.8100636005401611,
-      "learning_rate": 0.00018868871888805082,
-      "loss": 3.2932,
+      "epoch": 6.873315363881401,
+      "grad_norm": 0.7264222502708435,
+      "learning_rate": 0.00018796546141392333,
+      "loss": 3.2696,
       "step": 63750
     },
     {
-      "epoch": 6.866860402540093,
-      "grad_norm": 0.7884566187858582,
-      "learning_rate": 0.00018836547785798944,
-      "loss": 3.2776,
+      "epoch": 6.878706199460916,
+      "grad_norm": 0.7692385911941528,
+      "learning_rate": 0.0001876416621694549,
+      "loss": 3.2821,
       "step": 63800
     },
     {
-      "epoch": 6.8722419545797,
-      "grad_norm": 0.7687737345695496,
-      "learning_rate": 0.000188042236827928,
-      "loss": 3.2631,
+      "epoch": 6.884097035040432,
+      "grad_norm": 0.7711987495422363,
+      "learning_rate": 0.00018731786292498648,
+      "loss": 3.2555,
       "step": 63850
     },
     {
-      "epoch": 6.877623506619309,
-      "grad_norm": 0.787807047367096,
-      "learning_rate": 0.00018771899579786658,
-      "loss": 3.2586,
+      "epoch": 6.889487870619946,
+      "grad_norm": 0.7359189987182617,
+      "learning_rate": 0.00018699406368051806,
+      "loss": 3.2698,
       "step": 63900
     },
     {
-      "epoch": 6.8830050586589175,
-      "grad_norm": 0.8006682395935059,
-      "learning_rate": 0.00018739575476780517,
-      "loss": 3.2689,
+      "epoch": 6.894878706199461,
+      "grad_norm": 0.7642732858657837,
+      "learning_rate": 0.00018667026443604964,
+      "loss": 3.2864,
       "step": 63950
     },
     {
-      "epoch": 6.888386610698525,
-      "grad_norm": 0.7582936882972717,
-      "learning_rate": 0.00018707251373774377,
-      "loss": 3.2819,
+      "epoch": 6.900269541778976,
+      "grad_norm": 0.7783806324005127,
+      "learning_rate": 0.00018634646519158121,
+      "loss": 3.2955,
       "step": 64000
     },
     {
-      "epoch": 6.888386610698525,
-      "eval_accuracy": 0.387549748082847,
-      "eval_loss": 3.3533716201782227,
-      "eval_runtime": 184.1372,
-      "eval_samples_per_second": 97.813,
-      "eval_steps_per_second": 6.115,
+      "epoch": 6.900269541778976,
+      "eval_accuracy": 0.3873601487936107,
+      "eval_loss": 3.3531534671783447,
+      "eval_runtime": 183.7258,
+      "eval_samples_per_second": 98.032,
+      "eval_steps_per_second": 6.129,
       "step": 64000
     },
     {
-      "epoch": 6.893768162738134,
-      "grad_norm": 0.7883850932121277,
-      "learning_rate": 0.00018674927270768233,
-      "loss": 3.2751,
+      "epoch": 6.90566037735849,
+      "grad_norm": 0.7426219582557678,
+      "learning_rate": 0.0001860226659471128,
+      "loss": 3.2768,
       "step": 64050
     },
     {
-      "epoch": 6.899149714777742,
-      "grad_norm": 0.7839645743370056,
-      "learning_rate": 0.00018642603167762093,
-      "loss": 3.2778,
+      "epoch": 6.9110512129380055,
+      "grad_norm": 0.7826731204986572,
+      "learning_rate": 0.00018569886670264434,
+      "loss": 3.2768,
       "step": 64100
     },
     {
-      "epoch": 6.90453126681735,
-      "grad_norm": 0.7904356122016907,
-      "learning_rate": 0.0001861027906475595,
-      "loss": 3.2729,
+      "epoch": 6.916442048517521,
+      "grad_norm": 0.6981860399246216,
+      "learning_rate": 0.00018537506745817592,
+      "loss": 3.2768,
       "step": 64150
     },
     {
-      "epoch": 6.9099128188569585,
-      "grad_norm": 0.7667959332466125,
-      "learning_rate": 0.00018577954961749812,
-      "loss": 3.2584,
+      "epoch": 6.921832884097035,
+      "grad_norm": 0.7405015230178833,
+      "learning_rate": 0.00018505126821370747,
+      "loss": 3.3001,
       "step": 64200
     },
     {
-      "epoch": 6.915294370896566,
-      "grad_norm": 0.7473137974739075,
-      "learning_rate": 0.00018545630858743669,
-      "loss": 3.2792,
+      "epoch": 6.92722371967655,
+      "grad_norm": 0.8087053298950195,
+      "learning_rate": 0.00018472746896923905,
+      "loss": 3.2719,
       "step": 64250
     },
     {
-      "epoch": 6.920675922936175,
-      "grad_norm": 0.775063693523407,
-      "learning_rate": 0.00018513306755737525,
-      "loss": 3.2819,
+      "epoch": 6.932614555256064,
+      "grad_norm": 0.7994046807289124,
+      "learning_rate": 0.00018440366972477063,
+      "loss": 3.2774,
       "step": 64300
     },
     {
-      "epoch": 6.926057474975783,
-      "grad_norm": 0.7284723520278931,
-      "learning_rate": 0.00018480982652731387,
-      "loss": 3.2736,
+      "epoch": 6.938005390835579,
+      "grad_norm": 0.7336465716362,
+      "learning_rate": 0.0001840798704803022,
+      "loss": 3.2681,
       "step": 64350
     },
     {
-      "epoch": 6.931439027015391,
-      "grad_norm": 0.779125988483429,
-      "learning_rate": 0.00018448658549725244,
-      "loss": 3.2844,
+      "epoch": 6.943396226415095,
+      "grad_norm": 0.7434164881706238,
+      "learning_rate": 0.00018375607123583378,
+      "loss": 3.2784,
       "step": 64400
     },
     {
-      "epoch": 6.9368205790549995,
-      "grad_norm": 0.8092654347419739,
-      "learning_rate": 0.000184163344467191,
-      "loss": 3.2749,
+      "epoch": 6.948787061994609,
+      "grad_norm": 0.7420080900192261,
+      "learning_rate": 0.00018343227199136533,
+      "loss": 3.2529,
       "step": 64450
     },
     {
-      "epoch": 6.942202131094608,
-      "grad_norm": 0.8407021760940552,
-      "learning_rate": 0.0001838401034371296,
-      "loss": 3.291,
+      "epoch": 6.954177897574124,
+      "grad_norm": 0.7828781008720398,
+      "learning_rate": 0.00018311494873178628,
+      "loss": 3.296,
       "step": 64500
     },
     {
-      "epoch": 6.947583683134216,
-      "grad_norm": 0.8245097398757935,
-      "learning_rate": 0.0001835168624070682,
-      "loss": 3.2682,
+      "epoch": 6.959568733153639,
+      "grad_norm": 0.7523041367530823,
+      "learning_rate": 0.00018279114948731783,
+      "loss": 3.2885,
       "step": 64550
     },
     {
-      "epoch": 6.952965235173824,
-      "grad_norm": 0.7837095260620117,
-      "learning_rate": 0.00018319362137700677,
-      "loss": 3.2662,
+      "epoch": 6.964959568733153,
+      "grad_norm": 0.7568721771240234,
+      "learning_rate": 0.0001824673502428494,
+      "loss": 3.2922,
       "step": 64600
     },
     {
-      "epoch": 6.958346787213433,
-      "grad_norm": 0.7676137685775757,
-      "learning_rate": 0.00018287038034694536,
-      "loss": 3.2644,
+      "epoch": 6.9703504043126685,
+      "grad_norm": 0.8034500479698181,
+      "learning_rate": 0.000182143550998381,
+      "loss": 3.3018,
       "step": 64650
     },
     {
-      "epoch": 6.9637283392530405,
-      "grad_norm": 0.7583410143852234,
-      "learning_rate": 0.00018254713931688393,
-      "loss": 3.2667,
+      "epoch": 6.975741239892184,
+      "grad_norm": 0.7914881706237793,
+      "learning_rate": 0.00018181975175391257,
+      "loss": 3.293,
       "step": 64700
     },
     {
-      "epoch": 6.969109891292649,
-      "grad_norm": 0.7313645482063293,
-      "learning_rate": 0.00018222389828682252,
-      "loss": 3.2846,
+      "epoch": 6.981132075471698,
+      "grad_norm": 0.8011124134063721,
+      "learning_rate": 0.00018149595250944414,
+      "loss": 3.2808,
       "step": 64750
     },
     {
-      "epoch": 6.974491443332257,
-      "grad_norm": 0.7626355290412903,
-      "learning_rate": 0.00018190065725676112,
-      "loss": 3.277,
+      "epoch": 6.986522911051213,
+      "grad_norm": 0.8010050058364868,
+      "learning_rate": 0.0001811721532649757,
+      "loss": 3.259,
       "step": 64800
     },
     {
-      "epoch": 6.979872995371865,
-      "grad_norm": 0.734296977519989,
-      "learning_rate": 0.00018157741622669969,
-      "loss": 3.2821,
+      "epoch": 6.991913746630727,
+      "grad_norm": 0.7234358787536621,
+      "learning_rate": 0.00018084835402050727,
+      "loss": 3.2707,
       "step": 64850
     },
     {
-      "epoch": 6.985254547411474,
-      "grad_norm": 0.8134587407112122,
-      "learning_rate": 0.00018125417519663825,
-      "loss": 3.2753,
+      "epoch": 6.997304582210242,
+      "grad_norm": 0.7819421887397766,
+      "learning_rate": 0.00018052455477603885,
+      "loss": 3.273,
       "step": 64900
     },
     {
-      "epoch": 6.990636099451081,
-      "grad_norm": 0.783110499382019,
-      "learning_rate": 0.00018093093416657687,
-      "loss": 3.2841,
+      "epoch": 7.002695417789758,
+      "grad_norm": 0.7928745150566101,
+      "learning_rate": 0.0001802007555315704,
+      "loss": 3.2246,
       "step": 64950
     },
     {
-      "epoch": 6.99601765149069,
-      "grad_norm": 0.8081582188606262,
-      "learning_rate": 0.00018060769313651544,
-      "loss": 3.2671,
+      "epoch": 7.008086253369272,
+      "grad_norm": 0.8181132674217224,
+      "learning_rate": 0.00017987695628710198,
+      "loss": 3.2045,
       "step": 65000
     },
     {
-      "epoch": 6.99601765149069,
-      "eval_accuracy": 0.3880192372110933,
-      "eval_loss": 3.3486366271972656,
-      "eval_runtime": 183.9327,
-      "eval_samples_per_second": 97.922,
-      "eval_steps_per_second": 6.122,
+      "epoch": 7.008086253369272,
+      "eval_accuracy": 0.38724986611247897,
+      "eval_loss": 3.3570756912231445,
+      "eval_runtime": 183.7885,
+      "eval_samples_per_second": 97.999,
+      "eval_steps_per_second": 6.127,
       "step": 65000
     },
     {
-      "epoch": 7.0013992035302985,
-      "grad_norm": 0.7595446109771729,
-      "learning_rate": 0.000180284452106454,
-      "loss": 3.2612,
+      "epoch": 7.013477088948787,
+      "grad_norm": 0.7654605507850647,
+      "learning_rate": 0.00017955315704263356,
+      "loss": 3.1885,
       "step": 65050
     },
     {
-      "epoch": 7.006780755569906,
-      "grad_norm": 0.7790128588676453,
-      "learning_rate": 0.00017996121107639263,
-      "loss": 3.1886,
+      "epoch": 7.018867924528302,
+      "grad_norm": 0.8949223160743713,
+      "learning_rate": 0.0001792293577981651,
+      "loss": 3.203,
       "step": 65100
     },
     {
-      "epoch": 7.012162307609515,
-      "grad_norm": 0.8460402488708496,
-      "learning_rate": 0.0001796379700463312,
-      "loss": 3.1878,
+      "epoch": 7.024258760107816,
+      "grad_norm": 0.7771431803703308,
+      "learning_rate": 0.00017890555855369668,
+      "loss": 3.2139,
       "step": 65150
     },
     {
-      "epoch": 7.017543859649122,
-      "grad_norm": 0.794796884059906,
-      "learning_rate": 0.0001793147290162698,
-      "loss": 3.1727,
+      "epoch": 7.0296495956873315,
+      "grad_norm": 0.7502090334892273,
+      "learning_rate": 0.00017858175930922826,
+      "loss": 3.1923,
       "step": 65200
     },
     {
-      "epoch": 7.022925411688731,
-      "grad_norm": 0.8340461850166321,
-      "learning_rate": 0.00017899148798620836,
-      "loss": 3.2121,
+      "epoch": 7.035040431266847,
+      "grad_norm": 0.7991393804550171,
+      "learning_rate": 0.00017825796006475984,
+      "loss": 3.187,
       "step": 65250
     },
     {
-      "epoch": 7.0283069637283395,
-      "grad_norm": 0.7978731989860535,
-      "learning_rate": 0.0001786747117767482,
-      "loss": 3.1929,
+      "epoch": 7.040431266846361,
+      "grad_norm": 0.7929309606552124,
+      "learning_rate": 0.00017793416082029142,
+      "loss": 3.2047,
       "step": 65300
     },
     {
-      "epoch": 7.033688515767947,
-      "grad_norm": 0.7534997463226318,
-      "learning_rate": 0.00017835147074668676,
-      "loss": 3.1643,
+      "epoch": 7.045822102425876,
+      "grad_norm": 0.7534916400909424,
+      "learning_rate": 0.000177610361575823,
+      "loss": 3.1833,
       "step": 65350
     },
     {
-      "epoch": 7.039070067807556,
-      "grad_norm": 0.8307495713233948,
-      "learning_rate": 0.00017802822971662533,
-      "loss": 3.1987,
+      "epoch": 7.051212938005391,
+      "grad_norm": 0.7809228897094727,
+      "learning_rate": 0.00017728656233135452,
+      "loss": 3.2045,
       "step": 65400
     },
     {
-      "epoch": 7.044451619847164,
-      "grad_norm": 0.8099019527435303,
-      "learning_rate": 0.00017770498868656395,
-      "loss": 3.1938,
+      "epoch": 7.056603773584905,
+      "grad_norm": 0.7841359972953796,
+      "learning_rate": 0.0001769627630868861,
+      "loss": 3.2029,
       "step": 65450
     },
     {
-      "epoch": 7.049833171886772,
-      "grad_norm": 0.812278151512146,
-      "learning_rate": 0.00017738174765650252,
-      "loss": 3.1768,
+      "epoch": 7.061994609164421,
+      "grad_norm": 0.7379030585289001,
+      "learning_rate": 0.00017663896384241767,
+      "loss": 3.1991,
       "step": 65500
     },
     {
-      "epoch": 7.0552147239263805,
-      "grad_norm": 0.7854213714599609,
-      "learning_rate": 0.0001770585066264411,
-      "loss": 3.1905,
+      "epoch": 7.067385444743936,
+      "grad_norm": 0.7718530297279358,
+      "learning_rate": 0.00017631516459794925,
+      "loss": 3.2111,
       "step": 65550
     },
     {
-      "epoch": 7.060596275965988,
-      "grad_norm": 0.8177743554115295,
-      "learning_rate": 0.00017673526559637968,
-      "loss": 3.2111,
+      "epoch": 7.07277628032345,
+      "grad_norm": 0.7789267897605896,
+      "learning_rate": 0.00017599136535348083,
+      "loss": 3.2116,
       "step": 65600
     },
     {
-      "epoch": 7.065977828005597,
-      "grad_norm": 0.7576644420623779,
-      "learning_rate": 0.00017641202456631828,
-      "loss": 3.2025,
+      "epoch": 7.078167115902965,
+      "grad_norm": 0.7927533984184265,
+      "learning_rate": 0.0001756675661090124,
+      "loss": 3.197,
       "step": 65650
     },
     {
-      "epoch": 7.071359380045205,
-      "grad_norm": 0.8074950575828552,
-      "learning_rate": 0.00017608878353625684,
-      "loss": 3.2064,
+      "epoch": 7.083557951482479,
+      "grad_norm": 0.7786009907722473,
+      "learning_rate": 0.00017534376686454398,
+      "loss": 3.2089,
       "step": 65700
     },
     {
-      "epoch": 7.076740932084813,
-      "grad_norm": 0.7767495512962341,
-      "learning_rate": 0.00017576554250619544,
-      "loss": 3.1986,
+      "epoch": 7.0889487870619945,
+      "grad_norm": 0.7937779426574707,
+      "learning_rate": 0.00017501996762007556,
+      "loss": 3.2009,
       "step": 65750
     },
     {
-      "epoch": 7.0821224841244215,
-      "grad_norm": 0.770216166973114,
-      "learning_rate": 0.000175442301476134,
-      "loss": 3.195,
+      "epoch": 7.09433962264151,
+      "grad_norm": 0.7974300384521484,
+      "learning_rate": 0.00017469616837560709,
+      "loss": 3.2076,
       "step": 65800
     },
     {
-      "epoch": 7.08750403616403,
-      "grad_norm": 0.7485193610191345,
-      "learning_rate": 0.0001751190604460726,
-      "loss": 3.2142,
+      "epoch": 7.099730458221024,
+      "grad_norm": 0.763508677482605,
+      "learning_rate": 0.00017437236913113866,
+      "loss": 3.1875,
       "step": 65850
     },
     {
-      "epoch": 7.092885588203638,
-      "grad_norm": 0.7984924912452698,
-      "learning_rate": 0.0001747958194160112,
-      "loss": 3.2052,
+      "epoch": 7.105121293800539,
+      "grad_norm": 0.7986451983451843,
+      "learning_rate": 0.00017404856988667024,
+      "loss": 3.2016,
       "step": 65900
     },
     {
-      "epoch": 7.098267140243246,
-      "grad_norm": 0.8029812574386597,
-      "learning_rate": 0.00017447257838594976,
-      "loss": 3.2028,
+      "epoch": 7.110512129380054,
+      "grad_norm": 0.7924427390098572,
+      "learning_rate": 0.00017372477064220182,
+      "loss": 3.1884,
       "step": 65950
     },
     {
-      "epoch": 7.103648692282855,
-      "grad_norm": 0.7820143103599548,
-      "learning_rate": 0.00017414933735588838,
-      "loss": 3.212,
+      "epoch": 7.115902964959568,
+      "grad_norm": 0.7830730080604553,
+      "learning_rate": 0.0001734009713977334,
+      "loss": 3.2137,
       "step": 66000
     },
     {
-      "epoch": 7.103648692282855,
-      "eval_accuracy": 0.38781649092243153,
-      "eval_loss": 3.3556320667266846,
-      "eval_runtime": 183.7973,
-      "eval_samples_per_second": 97.994,
-      "eval_steps_per_second": 6.126,
+      "epoch": 7.115902964959568,
+      "eval_accuracy": 0.38791069297616665,
+      "eval_loss": 3.356255292892456,
+      "eval_runtime": 184.1892,
+      "eval_samples_per_second": 97.785,
+      "eval_steps_per_second": 6.113,
       "step": 66000
     },
     {
-      "epoch": 7.109030244322462,
-      "grad_norm": 0.8352454304695129,
-      "learning_rate": 0.00017382609632582695,
-      "loss": 3.1852,
+      "epoch": 7.121293800539084,
+      "grad_norm": 0.7967952489852905,
+      "learning_rate": 0.00017307717215326497,
+      "loss": 3.2032,
       "step": 66050
     },
     {
-      "epoch": 7.114411796362071,
-      "grad_norm": 0.8147093653678894,
-      "learning_rate": 0.00017350285529576552,
-      "loss": 3.1959,
+      "epoch": 7.126684636118599,
+      "grad_norm": 0.7771111726760864,
+      "learning_rate": 0.00017275337290879655,
+      "loss": 3.2054,
       "step": 66100
     },
     {
-      "epoch": 7.119793348401679,
-      "grad_norm": 0.7820026874542236,
-      "learning_rate": 0.00017317961426570411,
-      "loss": 3.2127,
+      "epoch": 7.132075471698113,
+      "grad_norm": 0.7791370153427124,
+      "learning_rate": 0.0001724295736643281,
+      "loss": 3.2137,
       "step": 66150
     },
     {
-      "epoch": 7.125174900441287,
-      "grad_norm": 0.7874245047569275,
-      "learning_rate": 0.0001728563732356427,
-      "loss": 3.2118,
+      "epoch": 7.137466307277628,
+      "grad_norm": 0.7515457272529602,
+      "learning_rate": 0.00017210577441985968,
+      "loss": 3.2013,
       "step": 66200
     },
     {
-      "epoch": 7.130556452480896,
-      "grad_norm": 0.7995700836181641,
-      "learning_rate": 0.00017253313220558128,
-      "loss": 3.2178,
+      "epoch": 7.142857142857143,
+      "grad_norm": 0.7740494012832642,
+      "learning_rate": 0.00017178197517539123,
+      "loss": 3.2199,
       "step": 66250
     },
     {
-      "epoch": 7.135938004520503,
-      "grad_norm": 0.7790077924728394,
-      "learning_rate": 0.00017220989117551987,
-      "loss": 3.2109,
+      "epoch": 7.1482479784366575,
+      "grad_norm": 0.8139404654502869,
+      "learning_rate": 0.0001714581759309228,
+      "loss": 3.2026,
       "step": 66300
     },
     {
-      "epoch": 7.141319556560112,
-      "grad_norm": 0.7603393793106079,
-      "learning_rate": 0.00017188665014545844,
-      "loss": 3.1997,
+      "epoch": 7.153638814016173,
+      "grad_norm": 0.7923606634140015,
+      "learning_rate": 0.00017113437668645439,
+      "loss": 3.2076,
       "step": 66350
     },
     {
-      "epoch": 7.1467011085997205,
-      "grad_norm": 0.7857191562652588,
-      "learning_rate": 0.00017156340911539703,
-      "loss": 3.1844,
+      "epoch": 7.159029649595688,
+      "grad_norm": 0.8198365569114685,
+      "learning_rate": 0.00017081057744198596,
+      "loss": 3.2099,
       "step": 66400
     },
     {
-      "epoch": 7.152082660639328,
-      "grad_norm": 0.8000122308731079,
-      "learning_rate": 0.00017124016808533563,
-      "loss": 3.2166,
+      "epoch": 7.164420485175202,
+      "grad_norm": 0.7356213927268982,
+      "learning_rate": 0.00017048677819751751,
+      "loss": 3.2183,
       "step": 66450
     },
     {
-      "epoch": 7.157464212678937,
-      "grad_norm": 0.7864665389060974,
-      "learning_rate": 0.0001709169270552742,
-      "loss": 3.2022,
+      "epoch": 7.169811320754717,
+      "grad_norm": 0.7708905339241028,
+      "learning_rate": 0.0001701629789530491,
+      "loss": 3.2062,
       "step": 66500
     },
     {
-      "epoch": 7.162845764718545,
-      "grad_norm": 0.837347149848938,
-      "learning_rate": 0.00017059368602521276,
-      "loss": 3.2122,
+      "epoch": 7.175202156334231,
+      "grad_norm": 0.8079206347465515,
+      "learning_rate": 0.00016983917970858067,
+      "loss": 3.2209,
       "step": 66550
     },
     {
-      "epoch": 7.168227316758153,
-      "grad_norm": 0.8074911832809448,
-      "learning_rate": 0.00017027044499515139,
-      "loss": 3.2052,
+      "epoch": 7.180592991913747,
+      "grad_norm": 0.8059882521629333,
+      "learning_rate": 0.00016951538046411225,
+      "loss": 3.2069,
       "step": 66600
     },
     {
-      "epoch": 7.1736088687977615,
-      "grad_norm": 0.791344940662384,
-      "learning_rate": 0.00016994720396508995,
-      "loss": 3.22,
+      "epoch": 7.185983827493262,
+      "grad_norm": 0.7901318073272705,
+      "learning_rate": 0.0001691915812196438,
+      "loss": 3.2223,
       "step": 66650
     },
     {
-      "epoch": 7.178990420837369,
-      "grad_norm": 0.809097409248352,
-      "learning_rate": 0.00016962396293502852,
-      "loss": 3.2121,
+      "epoch": 7.191374663072776,
+      "grad_norm": 0.7910335659980774,
+      "learning_rate": 0.00016886778197517538,
+      "loss": 3.2139,
       "step": 66700
     },
     {
-      "epoch": 7.184371972876978,
-      "grad_norm": 0.7525076866149902,
-      "learning_rate": 0.00016930072190496714,
-      "loss": 3.192,
+      "epoch": 7.196765498652291,
+      "grad_norm": 0.7589784264564514,
+      "learning_rate": 0.00016854398273070693,
+      "loss": 3.2206,
       "step": 66750
     },
     {
-      "epoch": 7.189753524916586,
-      "grad_norm": 0.8340684771537781,
-      "learning_rate": 0.0001689774808749057,
-      "loss": 3.1797,
+      "epoch": 7.202156334231806,
+      "grad_norm": 0.7931374907493591,
+      "learning_rate": 0.0001682201834862385,
+      "loss": 3.1961,
       "step": 66800
     },
     {
-      "epoch": 7.195135076956194,
-      "grad_norm": 0.8415242433547974,
-      "learning_rate": 0.00016865423984484428,
-      "loss": 3.2191,
+      "epoch": 7.2075471698113205,
+      "grad_norm": 0.8265644907951355,
+      "learning_rate": 0.00016789638424177008,
+      "loss": 3.2136,
       "step": 66850
     },
     {
-      "epoch": 7.2005166289958025,
-      "grad_norm": 0.8338217735290527,
-      "learning_rate": 0.00016833099881478287,
-      "loss": 3.2045,
+      "epoch": 7.212938005390836,
+      "grad_norm": 0.820319652557373,
+      "learning_rate": 0.00016757258499730166,
+      "loss": 3.232,
       "step": 66900
     },
     {
-      "epoch": 7.205898181035411,
-      "grad_norm": 0.8091142773628235,
-      "learning_rate": 0.00016800775778472144,
-      "loss": 3.2198,
+      "epoch": 7.218328840970351,
+      "grad_norm": 0.8143536448478699,
+      "learning_rate": 0.00016724878575283324,
+      "loss": 3.2209,
       "step": 66950
     },
     {
-      "epoch": 7.211279733075019,
-      "grad_norm": 0.7719037532806396,
-      "learning_rate": 0.00016768451675466006,
-      "loss": 3.2224,
+      "epoch": 7.223719676549865,
+      "grad_norm": 0.7919588685035706,
+      "learning_rate": 0.00016692498650836481,
+      "loss": 3.2254,
       "step": 67000
     },
     {
-      "epoch": 7.211279733075019,
-      "eval_accuracy": 0.38794513594160385,
-      "eval_loss": 3.3524861335754395,
-      "eval_runtime": 183.9648,
-      "eval_samples_per_second": 97.905,
-      "eval_steps_per_second": 6.121,
+      "epoch": 7.223719676549865,
+      "eval_accuracy": 0.38788407301865213,
+      "eval_loss": 3.3532602787017822,
+      "eval_runtime": 183.7056,
+      "eval_samples_per_second": 98.043,
+      "eval_steps_per_second": 6.129,
       "step": 67000
     },
     {
-      "epoch": 7.216661285114627,
-      "grad_norm": 0.8019943833351135,
-      "learning_rate": 0.00016736127572459863,
-      "loss": 3.2131,
+      "epoch": 7.22911051212938,
+      "grad_norm": 0.784161388874054,
+      "learning_rate": 0.0001666011872638964,
+      "loss": 3.2196,
       "step": 67050
     },
     {
-      "epoch": 7.222042837154235,
-      "grad_norm": 0.8244363069534302,
-      "learning_rate": 0.0001670380346945372,
-      "loss": 3.2046,
+      "epoch": 7.234501347708895,
+      "grad_norm": 0.8608620166778564,
+      "learning_rate": 0.00016627738801942792,
+      "loss": 3.2218,
       "step": 67100
     },
     {
-      "epoch": 7.2274243891938434,
-      "grad_norm": 0.8283898234367371,
-      "learning_rate": 0.00016671479366447582,
-      "loss": 3.2243,
+      "epoch": 7.2398921832884096,
+      "grad_norm": 0.7472286820411682,
+      "learning_rate": 0.0001659535887749595,
+      "loss": 3.2223,
       "step": 67150
     },
     {
-      "epoch": 7.232805941233452,
-      "grad_norm": 0.7998861074447632,
-      "learning_rate": 0.00016639155263441439,
-      "loss": 3.219,
+      "epoch": 7.245283018867925,
+      "grad_norm": 0.783933699131012,
+      "learning_rate": 0.00016562978953049107,
+      "loss": 3.2127,
       "step": 67200
     },
     {
-      "epoch": 7.23818749327306,
-      "grad_norm": 0.848159909248352,
-      "learning_rate": 0.00016606831160435295,
-      "loss": 3.2326,
+      "epoch": 7.250673854447439,
+      "grad_norm": 0.8143625259399414,
+      "learning_rate": 0.00016530599028602265,
+      "loss": 3.2117,
       "step": 67250
     },
     {
-      "epoch": 7.243569045312668,
-      "grad_norm": 0.7724121809005737,
-      "learning_rate": 0.00016574507057429155,
-      "loss": 3.2117,
+      "epoch": 7.256064690026954,
+      "grad_norm": 0.7995963096618652,
+      "learning_rate": 0.00016498219104155423,
+      "loss": 3.2283,
       "step": 67300
     },
     {
-      "epoch": 7.248950597352277,
-      "grad_norm": 0.7755951881408691,
-      "learning_rate": 0.00016542182954423014,
-      "loss": 3.2282,
+      "epoch": 7.261455525606469,
+      "grad_norm": 0.7936796545982361,
+      "learning_rate": 0.0001646583917970858,
+      "loss": 3.2163,
       "step": 67350
     },
     {
-      "epoch": 7.254332149391884,
-      "grad_norm": 0.780353307723999,
-      "learning_rate": 0.0001650985885141687,
-      "loss": 3.2344,
+      "epoch": 7.2668463611859835,
+      "grad_norm": 0.816673755645752,
+      "learning_rate": 0.00016433459255261738,
+      "loss": 3.223,
       "step": 67400
     },
     {
-      "epoch": 7.259713701431493,
-      "grad_norm": 0.7967930436134338,
-      "learning_rate": 0.0001647753474841073,
-      "loss": 3.2135,
+      "epoch": 7.272237196765499,
+      "grad_norm": 0.7831211090087891,
+      "learning_rate": 0.00016401079330814896,
+      "loss": 3.2359,
       "step": 67450
     },
     {
-      "epoch": 7.265095253471101,
-      "grad_norm": 0.7403193116188049,
-      "learning_rate": 0.00016445210645404587,
-      "loss": 3.2216,
+      "epoch": 7.277628032345014,
+      "grad_norm": 0.7967017292976379,
+      "learning_rate": 0.00016368699406368048,
+      "loss": 3.2288,
       "step": 67500
     },
     {
-      "epoch": 7.270476805510709,
-      "grad_norm": 0.8273634314537048,
-      "learning_rate": 0.00016412886542398447,
-      "loss": 3.215,
+      "epoch": 7.283018867924528,
+      "grad_norm": 0.8944386839866638,
+      "learning_rate": 0.00016336319481921206,
+      "loss": 3.2204,
       "step": 67550
     },
     {
-      "epoch": 7.275858357550318,
-      "grad_norm": 0.8118467926979065,
-      "learning_rate": 0.00016380562439392306,
-      "loss": 3.2236,
+      "epoch": 7.288409703504043,
+      "grad_norm": 0.7435490489006042,
+      "learning_rate": 0.00016303939557474364,
+      "loss": 3.2419,
       "step": 67600
     },
     {
-      "epoch": 7.281239909589925,
-      "grad_norm": 0.7936812043190002,
-      "learning_rate": 0.00016348238336386163,
-      "loss": 3.2275,
+      "epoch": 7.293800539083558,
+      "grad_norm": 0.861144483089447,
+      "learning_rate": 0.00016271559633027522,
+      "loss": 3.2236,
       "step": 67650
     },
     {
-      "epoch": 7.286621461629534,
-      "grad_norm": 0.886665940284729,
-      "learning_rate": 0.0001631591423338002,
-      "loss": 3.2133,
+      "epoch": 7.2991913746630726,
+      "grad_norm": 0.8081782460212708,
+      "learning_rate": 0.0001623917970858068,
+      "loss": 3.225,
       "step": 67700
     },
     {
-      "epoch": 7.2920030136691425,
-      "grad_norm": 0.8477235436439514,
-      "learning_rate": 0.00016283590130373882,
-      "loss": 3.2252,
+      "epoch": 7.304582210242588,
+      "grad_norm": 0.7873321175575256,
+      "learning_rate": 0.00016206799784133837,
+      "loss": 3.2273,
       "step": 67750
     },
     {
-      "epoch": 7.29738456570875,
-      "grad_norm": 0.7720082402229309,
-      "learning_rate": 0.00016251266027367739,
-      "loss": 3.2076,
+      "epoch": 7.309973045822103,
+      "grad_norm": 0.7682214379310608,
+      "learning_rate": 0.00016174419859686992,
+      "loss": 3.209,
       "step": 67800
     },
     {
-      "epoch": 7.302766117748359,
-      "grad_norm": 0.7984726428985596,
-      "learning_rate": 0.00016218941924361595,
-      "loss": 3.201,
+      "epoch": 7.315363881401617,
+      "grad_norm": 0.8452110886573792,
+      "learning_rate": 0.0001614203993524015,
+      "loss": 3.2454,
       "step": 67850
     },
     {
-      "epoch": 7.308147669787967,
-      "grad_norm": 0.8324883580207825,
-      "learning_rate": 0.00016186617821355458,
-      "loss": 3.1971,
+      "epoch": 7.320754716981132,
+      "grad_norm": 0.8070641756057739,
+      "learning_rate": 0.00016109660010793308,
+      "loss": 3.2246,
       "step": 67900
     },
     {
-      "epoch": 7.313529221827575,
-      "grad_norm": 0.8051575422286987,
-      "learning_rate": 0.00016154293718349314,
-      "loss": 3.2159,
+      "epoch": 7.3261455525606465,
+      "grad_norm": 0.8329002857208252,
+      "learning_rate": 0.00016077280086346463,
+      "loss": 3.2303,
       "step": 67950
     },
     {
-      "epoch": 7.3189107738671835,
-      "grad_norm": 0.8427183032035828,
-      "learning_rate": 0.00016121969615343174,
-      "loss": 3.2356,
+      "epoch": 7.331536388140162,
+      "grad_norm": 0.8193233013153076,
+      "learning_rate": 0.0001604490016189962,
+      "loss": 3.2077,
       "step": 68000
     },
     {
-      "epoch": 7.3189107738671835,
-      "eval_accuracy": 0.38804335815218816,
-      "eval_loss": 3.3509562015533447,
-      "eval_runtime": 183.863,
-      "eval_samples_per_second": 97.959,
-      "eval_steps_per_second": 6.124,
+      "epoch": 7.331536388140162,
+      "eval_accuracy": 0.3883694254685194,
+      "eval_loss": 3.3502113819122314,
+      "eval_runtime": 183.932,
+      "eval_samples_per_second": 97.922,
+      "eval_steps_per_second": 6.122,
       "step": 68000
     },
     {
-      "epoch": 7.324292325906791,
-      "grad_norm": 0.8195973038673401,
-      "learning_rate": 0.0001608964551233703,
-      "loss": 3.2297,
+      "epoch": 7.336927223719677,
+      "grad_norm": 0.8045969605445862,
+      "learning_rate": 0.00016012520237452778,
+      "loss": 3.2285,
       "step": 68050
     },
     {
-      "epoch": 7.3296738779464,
-      "grad_norm": 0.8639369606971741,
-      "learning_rate": 0.0001605732140933089,
-      "loss": 3.2161,
+      "epoch": 7.342318059299191,
+      "grad_norm": 0.7933641672134399,
+      "learning_rate": 0.00015980140313005933,
+      "loss": 3.2351,
       "step": 68100
     },
     {
-      "epoch": 7.335055429986008,
-      "grad_norm": 0.8031762838363647,
-      "learning_rate": 0.0001602499730632475,
-      "loss": 3.2191,
+      "epoch": 7.347708894878706,
+      "grad_norm": 0.8304075598716736,
+      "learning_rate": 0.0001594776038855909,
+      "loss": 3.2184,
       "step": 68150
     },
     {
-      "epoch": 7.340436982025616,
-      "grad_norm": 0.8548082709312439,
-      "learning_rate": 0.00015992673203318606,
-      "loss": 3.2057,
+      "epoch": 7.353099730458221,
+      "grad_norm": 0.8163561820983887,
+      "learning_rate": 0.0001591538046411225,
+      "loss": 3.2405,
       "step": 68200
     },
     {
-      "epoch": 7.3458185340652244,
-      "grad_norm": 0.8049331307411194,
-      "learning_rate": 0.0001596099558237259,
-      "loss": 3.2228,
+      "epoch": 7.3584905660377355,
+      "grad_norm": 0.8061451315879822,
+      "learning_rate": 0.00015883000539665407,
+      "loss": 3.2421,
       "step": 68250
     },
     {
-      "epoch": 7.351200086104833,
-      "grad_norm": 0.8107957243919373,
-      "learning_rate": 0.00015928671479366446,
-      "loss": 3.2238,
+      "epoch": 7.363881401617251,
+      "grad_norm": 0.7901108264923096,
+      "learning_rate": 0.00015850620615218564,
+      "loss": 3.23,
       "step": 68300
     },
     {
-      "epoch": 7.356581638144441,
-      "grad_norm": 0.8472992181777954,
-      "learning_rate": 0.00015896347376360303,
-      "loss": 3.2053,
+      "epoch": 7.369272237196766,
+      "grad_norm": 0.798541247844696,
+      "learning_rate": 0.0001581824069077172,
+      "loss": 3.2292,
       "step": 68350
     },
     {
-      "epoch": 7.361963190184049,
-      "grad_norm": 0.8222916126251221,
-      "learning_rate": 0.00015864023273354165,
-      "loss": 3.2277,
+      "epoch": 7.37466307277628,
+      "grad_norm": 0.8197664618492126,
+      "learning_rate": 0.00015785860766324877,
+      "loss": 3.2205,
       "step": 68400
     },
     {
-      "epoch": 7.367344742223658,
-      "grad_norm": 0.826626718044281,
-      "learning_rate": 0.00015831699170348022,
-      "loss": 3.2274,
+      "epoch": 7.380053908355795,
+      "grad_norm": 0.8161022067070007,
+      "learning_rate": 0.00015753480841878032,
+      "loss": 3.2056,
       "step": 68450
     },
     {
-      "epoch": 7.372726294263265,
-      "grad_norm": 0.8423181772232056,
-      "learning_rate": 0.0001579937506734188,
-      "loss": 3.2316,
+      "epoch": 7.38544474393531,
+      "grad_norm": 0.7873967885971069,
+      "learning_rate": 0.00015721748515920127,
+      "loss": 3.2098,
       "step": 68500
     },
     {
-      "epoch": 7.378107846302874,
-      "grad_norm": 0.8112326264381409,
-      "learning_rate": 0.00015767050964335738,
-      "loss": 3.2284,
+      "epoch": 7.390835579514825,
+      "grad_norm": 0.7904083132743835,
+      "learning_rate": 0.00015689368591473285,
+      "loss": 3.2214,
       "step": 68550
     },
     {
-      "epoch": 7.383489398342482,
-      "grad_norm": 0.7790609002113342,
-      "learning_rate": 0.00015734726861329595,
-      "loss": 3.2156,
+      "epoch": 7.39622641509434,
+      "grad_norm": 0.7788252234458923,
+      "learning_rate": 0.00015656988667026443,
+      "loss": 3.2373,
       "step": 68600
     },
     {
-      "epoch": 7.38887095038209,
-      "grad_norm": 0.7766805291175842,
-      "learning_rate": 0.00015702402758323454,
-      "loss": 3.2235,
+      "epoch": 7.401617250673855,
+      "grad_norm": 0.8234322667121887,
+      "learning_rate": 0.000156246087425796,
+      "loss": 3.2098,
       "step": 68650
     },
     {
-      "epoch": 7.394252502421699,
-      "grad_norm": 0.8133030533790588,
-      "learning_rate": 0.00015670078655317314,
-      "loss": 3.2166,
+      "epoch": 7.407008086253369,
+      "grad_norm": 0.8141764402389526,
+      "learning_rate": 0.00015592228818132756,
+      "loss": 3.2145,
       "step": 68700
     },
     {
-      "epoch": 7.399634054461306,
-      "grad_norm": 0.8183771967887878,
-      "learning_rate": 0.0001563775455231117,
-      "loss": 3.2174,
+      "epoch": 7.412398921832884,
+      "grad_norm": 0.7871109247207642,
+      "learning_rate": 0.0001555984889368591,
+      "loss": 3.2187,
       "step": 68750
     },
     {
-      "epoch": 7.405015606500915,
-      "grad_norm": 0.8056250810623169,
-      "learning_rate": 0.00015605430449305033,
-      "loss": 3.2086,
+      "epoch": 7.4177897574123985,
+      "grad_norm": 0.8350339531898499,
+      "learning_rate": 0.0001552746896923907,
+      "loss": 3.2296,
       "step": 68800
     },
     {
-      "epoch": 7.4103971585405235,
-      "grad_norm": 0.8371083736419678,
-      "learning_rate": 0.0001557310634629889,
-      "loss": 3.2284,
+      "epoch": 7.423180592991914,
+      "grad_norm": 0.807808518409729,
+      "learning_rate": 0.00015495089044792226,
+      "loss": 3.2123,
       "step": 68850
     },
     {
-      "epoch": 7.415778710580131,
-      "grad_norm": 0.8087985515594482,
-      "learning_rate": 0.00015540782243292746,
-      "loss": 3.2382,
+      "epoch": 7.428571428571429,
+      "grad_norm": 0.8097934126853943,
+      "learning_rate": 0.00015462709120345384,
+      "loss": 3.2271,
       "step": 68900
     },
     {
-      "epoch": 7.42116026261974,
-      "grad_norm": 0.7947363257408142,
-      "learning_rate": 0.00015508458140286606,
-      "loss": 3.2169,
+      "epoch": 7.433962264150943,
+      "grad_norm": 0.8217871189117432,
+      "learning_rate": 0.00015430329195898542,
+      "loss": 3.2196,
       "step": 68950
     },
     {
-      "epoch": 7.426541814659347,
-      "grad_norm": 0.8020465970039368,
-      "learning_rate": 0.00015476134037280465,
-      "loss": 3.2185,
+      "epoch": 7.439353099730458,
+      "grad_norm": 0.8195081949234009,
+      "learning_rate": 0.000153979492714517,
+      "loss": 3.2431,
       "step": 69000
     },
     {
-      "epoch": 7.426541814659347,
-      "eval_accuracy": 0.38862247804423933,
-      "eval_loss": 3.347287654876709,
-      "eval_runtime": 184.0756,
-      "eval_samples_per_second": 97.846,
-      "eval_steps_per_second": 6.117,
+      "epoch": 7.439353099730458,
+      "eval_accuracy": 0.38852316930477687,
+      "eval_loss": 3.3481428623199463,
+      "eval_runtime": 184.114,
+      "eval_samples_per_second": 97.825,
+      "eval_steps_per_second": 6.116,
       "step": 69000
     },
     {
-      "epoch": 7.431923366698956,
-      "grad_norm": 0.8113294243812561,
-      "learning_rate": 0.00015443809934274322,
-      "loss": 3.2167,
+      "epoch": 7.444743935309973,
+      "grad_norm": 0.8148500919342041,
+      "learning_rate": 0.00015365569347004858,
+      "loss": 3.2229,
       "step": 69050
     },
     {
-      "epoch": 7.4373049187385645,
-      "grad_norm": 0.7670662999153137,
-      "learning_rate": 0.00015411485831268182,
-      "loss": 3.2246,
+      "epoch": 7.450134770889488,
+      "grad_norm": 0.8321552872657776,
+      "learning_rate": 0.00015333189422558015,
+      "loss": 3.2144,
       "step": 69100
     },
     {
-      "epoch": 7.442686470778172,
-      "grad_norm": 0.8148250579833984,
-      "learning_rate": 0.00015379161728262038,
-      "loss": 3.2062,
+      "epoch": 7.455525606469003,
+      "grad_norm": 0.842375636100769,
+      "learning_rate": 0.00015300809498111168,
+      "loss": 3.199,
       "step": 69150
     },
     {
-      "epoch": 7.448068022817781,
-      "grad_norm": 0.8673095107078552,
-      "learning_rate": 0.00015346837625255898,
-      "loss": 3.2256,
+      "epoch": 7.460916442048518,
+      "grad_norm": 0.815597414970398,
+      "learning_rate": 0.00015268429573664325,
+      "loss": 3.2414,
       "step": 69200
     },
     {
-      "epoch": 7.453449574857389,
-      "grad_norm": 0.7744132280349731,
-      "learning_rate": 0.00015314513522249757,
-      "loss": 3.2148,
+      "epoch": 7.466307277628032,
+      "grad_norm": 0.7956526875495911,
+      "learning_rate": 0.00015236049649217483,
+      "loss": 3.2261,
       "step": 69250
     },
     {
-      "epoch": 7.458831126896997,
-      "grad_norm": 0.831725001335144,
-      "learning_rate": 0.00015282189419243614,
-      "loss": 3.2241,
+      "epoch": 7.471698113207547,
+      "grad_norm": 0.8070045113563538,
+      "learning_rate": 0.0001520366972477064,
+      "loss": 3.2143,
       "step": 69300
     },
     {
-      "epoch": 7.4642126789366054,
-      "grad_norm": 0.8336020708084106,
-      "learning_rate": 0.0001524986531623747,
-      "loss": 3.2037,
+      "epoch": 7.4770889487870615,
+      "grad_norm": 0.816788375377655,
+      "learning_rate": 0.000151712898003238,
+      "loss": 3.2341,
       "step": 69350
     },
     {
-      "epoch": 7.469594230976213,
-      "grad_norm": 0.8407599925994873,
-      "learning_rate": 0.00015217541213231333,
-      "loss": 3.2314,
+      "epoch": 7.482479784366577,
+      "grad_norm": 0.7771405577659607,
+      "learning_rate": 0.00015138909875876956,
+      "loss": 3.2364,
       "step": 69400
     },
     {
-      "epoch": 7.474975783015822,
-      "grad_norm": 0.9136285185813904,
-      "learning_rate": 0.0001518521711022519,
-      "loss": 3.2255,
+      "epoch": 7.487870619946092,
+      "grad_norm": 0.7785300016403198,
+      "learning_rate": 0.00015106529951430114,
+      "loss": 3.2301,
       "step": 69450
     },
     {
-      "epoch": 7.48035733505543,
-      "grad_norm": 0.8944079875946045,
-      "learning_rate": 0.00015152893007219046,
-      "loss": 3.2225,
+      "epoch": 7.493261455525606,
+      "grad_norm": 0.82381272315979,
+      "learning_rate": 0.0001507415002698327,
+      "loss": 3.2202,
       "step": 69500
     },
     {
-      "epoch": 7.485738887095038,
-      "grad_norm": 0.7992327809333801,
-      "learning_rate": 0.00015120568904212909,
-      "loss": 3.2074,
+      "epoch": 7.498652291105121,
+      "grad_norm": 0.8353618383407593,
+      "learning_rate": 0.00015041770102536424,
+      "loss": 3.2239,
       "step": 69550
     },
     {
-      "epoch": 7.491120439134646,
-      "grad_norm": 0.8009116053581238,
-      "learning_rate": 0.00015088244801206765,
-      "loss": 3.2045,
+      "epoch": 7.504043126684636,
+      "grad_norm": 0.8109288215637207,
+      "learning_rate": 0.00015009390178089582,
+      "loss": 3.2196,
       "step": 69600
     },
     {
-      "epoch": 7.496501991174255,
-      "grad_norm": 0.7832297086715698,
-      "learning_rate": 0.00015055920698200622,
-      "loss": 3.217,
+      "epoch": 7.509433962264151,
+      "grad_norm": 0.8037685751914978,
+      "learning_rate": 0.0001497701025364274,
+      "loss": 3.2237,
       "step": 69650
     },
     {
-      "epoch": 7.501883543213863,
-      "grad_norm": 0.8556872010231018,
-      "learning_rate": 0.00015023596595194482,
-      "loss": 3.2139,
+      "epoch": 7.514824797843666,
+      "grad_norm": 0.8085172176361084,
+      "learning_rate": 0.00014944630329195898,
+      "loss": 3.2343,
       "step": 69700
     },
     {
-      "epoch": 7.507265095253471,
-      "grad_norm": 0.8262065649032593,
-      "learning_rate": 0.0001499127249218834,
-      "loss": 3.2283,
+      "epoch": 7.520215633423181,
+      "grad_norm": 0.8349311947822571,
+      "learning_rate": 0.00014912250404749055,
+      "loss": 3.2244,
       "step": 69750
     },
     {
-      "epoch": 7.51264664729308,
-      "grad_norm": 0.8240439891815186,
-      "learning_rate": 0.00014958948389182198,
-      "loss": 3.2444,
+      "epoch": 7.525606469002695,
+      "grad_norm": 0.8080737590789795,
+      "learning_rate": 0.0001487987048030221,
+      "loss": 3.2404,
       "step": 69800
     },
     {
-      "epoch": 7.518028199332687,
-      "grad_norm": 0.7957839369773865,
-      "learning_rate": 0.00014926624286176057,
-      "loss": 3.2229,
+      "epoch": 7.53099730458221,
+      "grad_norm": 0.7988405227661133,
+      "learning_rate": 0.00014847490555855368,
+      "loss": 3.2161,
       "step": 69850
     },
     {
-      "epoch": 7.523409751372296,
-      "grad_norm": 0.8447942733764648,
-      "learning_rate": 0.00014894300183169914,
-      "loss": 3.2171,
+      "epoch": 7.536388140161725,
+      "grad_norm": 0.7778878211975098,
+      "learning_rate": 0.00014815110631408526,
+      "loss": 3.2236,
       "step": 69900
     },
     {
-      "epoch": 7.528791303411904,
-      "grad_norm": 0.801649808883667,
-      "learning_rate": 0.00014861976080163773,
-      "loss": 3.2341,
+      "epoch": 7.54177897574124,
+      "grad_norm": 0.8272253274917603,
+      "learning_rate": 0.0001478273070696168,
+      "loss": 3.2307,
       "step": 69950
     },
     {
-      "epoch": 7.534172855451512,
-      "grad_norm": 0.8142421841621399,
-      "learning_rate": 0.00014829651977157633,
-      "loss": 3.2095,
+      "epoch": 7.547169811320755,
+      "grad_norm": 0.8347529172897339,
+      "learning_rate": 0.0001475035078251484,
+      "loss": 3.2283,
       "step": 70000
     },
     {
-      "epoch": 7.534172855451512,
-      "eval_accuracy": 0.38903296865440246,
-      "eval_loss": 3.343346357345581,
-      "eval_runtime": 184.3428,
-      "eval_samples_per_second": 97.704,
-      "eval_steps_per_second": 6.108,
+      "epoch": 7.547169811320755,
+      "eval_accuracy": 0.3891295610716695,
+      "eval_loss": 3.342951774597168,
+      "eval_runtime": 183.6962,
+      "eval_samples_per_second": 98.048,
+      "eval_steps_per_second": 6.13,
       "step": 70000
     }
   ],
   "logging_steps": 50,
-  "max_steps": 92910,
+  "max_steps": 92750,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 10,
   "save_steps": 10000,
@@ -10456,7 +10456,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 5.85268545060864e+17,
+  "total_flos": 5.852484255744e+17,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null