diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5466 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 0,
+  "global_step": 388311,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0038628831014315846,
+      "grad_norm": 0.028026340529322624,
+      "learning_rate": 2.5701335714040445e-06,
+      "loss": 0.0035,
+      "step": 500
+    },
+    {
+      "epoch": 0.007725766202863169,
+      "grad_norm": 0.028945209458470345,
+      "learning_rate": 5.145417711087456e-06,
+      "loss": 0.0024,
+      "step": 1000
+    },
+    {
+      "epoch": 0.011588649304294754,
+      "grad_norm": 0.023790603503584862,
+      "learning_rate": 7.720701850770868e-06,
+      "loss": 0.0022,
+      "step": 1500
+    },
+    {
+      "epoch": 0.015451532405726338,
+      "grad_norm": 0.012105235829949379,
+      "learning_rate": 1.0295985990454279e-05,
+      "loss": 0.002,
+      "step": 2000
+    },
+    {
+      "epoch": 0.01931441550715792,
+      "grad_norm": 0.007807546760886908,
+      "learning_rate": 1.2871270130137691e-05,
+      "loss": 0.0019,
+      "step": 2500
+    },
+    {
+      "epoch": 0.023177298608589508,
+      "grad_norm": 0.004866514354944229,
+      "learning_rate": 1.54465542698211e-05,
+      "loss": 0.0019,
+      "step": 3000
+    },
+    {
+      "epoch": 0.02704018171002109,
+      "grad_norm": 0.004499210510402918,
+      "learning_rate": 1.8021838409504514e-05,
+      "loss": 0.0018,
+      "step": 3500
+    },
+    {
+      "epoch": 0.030903064811452677,
+      "grad_norm": 0.0029713744297623634,
+      "learning_rate": 2.0597122549187926e-05,
+      "loss": 0.0018,
+      "step": 4000
+    },
+    {
+      "epoch": 0.03476594791288426,
+      "grad_norm": 0.002635620068758726,
+      "learning_rate": 2.3172406688871335e-05,
+      "loss": 0.0017,
+      "step": 4500
+    },
+    {
+      "epoch": 0.03862883101431584,
+      "grad_norm": 0.0026769228279590607,
+      "learning_rate": 2.5747690828554747e-05,
+      "loss": 0.0017,
+      "step": 5000
+    },
+    {
+      "epoch": 0.04249171411574743,
+      "grad_norm": 0.004493319429457188,
+      "learning_rate": 2.8322974968238163e-05,
+      "loss": 0.0016,
+      "step": 5500
+    },
+    {
+      "epoch": 0.046354597217179015,
+      "grad_norm": 0.002984994789585471,
+      "learning_rate": 3.089825910792157e-05,
+      "loss": 0.0016,
+      "step": 6000
+    },
+    {
+      "epoch": 0.0502174803186106,
+      "grad_norm": 0.002657888922840357,
+      "learning_rate": 3.3473543247604984e-05,
+      "loss": 0.0016,
+      "step": 6500
+    },
+    {
+      "epoch": 0.05408036342004218,
+      "grad_norm": 0.0023709412198513746,
+      "learning_rate": 3.604882738728839e-05,
+      "loss": 0.0016,
+      "step": 7000
+    },
+    {
+      "epoch": 0.057943246521473764,
+      "grad_norm": 0.00325986510142684,
+      "learning_rate": 3.862411152697181e-05,
+      "loss": 0.0015,
+      "step": 7500
+    },
+    {
+      "epoch": 0.061806129622905354,
+      "grad_norm": 0.0019731384236365557,
+      "learning_rate": 4.119939566665521e-05,
+      "loss": 0.0015,
+      "step": 8000
+    },
+    {
+      "epoch": 0.06566901272433694,
+      "grad_norm": 0.002514706226065755,
+      "learning_rate": 4.377467980633863e-05,
+      "loss": 0.0015,
+      "step": 8500
+    },
+    {
+      "epoch": 0.06953189582576852,
+      "grad_norm": 0.005514668766409159,
+      "learning_rate": 4.634996394602204e-05,
+      "loss": 0.0014,
+      "step": 9000
+    },
+    {
+      "epoch": 0.0733947789272001,
+      "grad_norm": 0.002325449837371707,
+      "learning_rate": 4.892524808570545e-05,
+      "loss": 0.0014,
+      "step": 9500
+    },
+    {
+      "epoch": 0.07725766202863169,
+      "grad_norm": 0.00256099714897573,
+      "learning_rate": 5.150053222538886e-05,
+      "loss": 0.0014,
+      "step": 10000
+    },
+    {
+      "epoch": 0.08112054513006327,
+      "grad_norm": 0.002316329861059785,
+      "learning_rate": 5.4075816365072276e-05,
+      "loss": 0.0013,
+      "step": 10500
+    },
+    {
+      "epoch": 0.08498342823149486,
+      "grad_norm": 0.002595042111352086,
+      "learning_rate": 5.6651100504755685e-05,
+      "loss": 0.0013,
+      "step": 11000
+    },
+    {
+      "epoch": 0.08884631133292645,
+      "grad_norm": 0.002004101872444153,
+      "learning_rate": 5.92263846444391e-05,
+      "loss": 0.0013,
+      "step": 11500
+    },
+    {
+      "epoch": 0.09270919443435803,
+      "grad_norm": 0.001948538818396628,
+      "learning_rate": 6.180166878412251e-05,
+      "loss": 0.0012,
+      "step": 12000
+    },
+    {
+      "epoch": 0.09657207753578961,
+      "grad_norm": 0.002029005205258727,
+      "learning_rate": 6.437695292380593e-05,
+      "loss": 0.0012,
+      "step": 12500
+    },
+    {
+      "epoch": 0.1004349606372212,
+      "grad_norm": 0.001954406499862671,
+      "learning_rate": 6.695223706348933e-05,
+      "loss": 0.0012,
+      "step": 13000
+    },
+    {
+      "epoch": 0.10429784373865278,
+      "grad_norm": 0.001649603364057839,
+      "learning_rate": 6.952752120317274e-05,
+      "loss": 0.0012,
+      "step": 13500
+    },
+    {
+      "epoch": 0.10816072684008436,
+      "grad_norm": 0.002146226353943348,
+      "learning_rate": 7.210280534285616e-05,
+      "loss": 0.0011,
+      "step": 14000
+    },
+    {
+      "epoch": 0.11202360994151594,
+      "grad_norm": 0.002416988369077444,
+      "learning_rate": 7.467808948253957e-05,
+      "loss": 0.0011,
+      "step": 14500
+    },
+    {
+      "epoch": 0.11588649304294753,
+      "grad_norm": 0.0018118915613740683,
+      "learning_rate": 7.725337362222298e-05,
+      "loss": 0.0011,
+      "step": 15000
+    },
+    {
+      "epoch": 0.11974937614437912,
+      "grad_norm": 0.0018452458316460252,
+      "learning_rate": 7.982865776190639e-05,
+      "loss": 0.0011,
+      "step": 15500
+    },
+    {
+      "epoch": 0.12361225924581071,
+      "grad_norm": 0.0015237265033647418,
+      "learning_rate": 8.240394190158981e-05,
+      "loss": 0.0011,
+      "step": 16000
+    },
+    {
+      "epoch": 0.12747514234724228,
+      "grad_norm": 0.0019381915917620063,
+      "learning_rate": 8.497922604127322e-05,
+      "loss": 0.001,
+      "step": 16500
+    },
+    {
+      "epoch": 0.13133802544867387,
+      "grad_norm": 0.0018706010887399316,
+      "learning_rate": 8.755451018095661e-05,
+      "loss": 0.001,
+      "step": 17000
+    },
+    {
+      "epoch": 0.13520090855010547,
+      "grad_norm": 0.0016168917063623667,
+      "learning_rate": 9.012979432064003e-05,
+      "loss": 0.001,
+      "step": 17500
+    },
+    {
+      "epoch": 0.13906379165153704,
+      "grad_norm": 0.0015931539237499237,
+      "learning_rate": 9.270507846032344e-05,
+      "loss": 0.001,
+      "step": 18000
+    },
+    {
+      "epoch": 0.14292667475296864,
+      "grad_norm": 0.0014986397000029683,
+      "learning_rate": 9.528036260000686e-05,
+      "loss": 0.001,
+      "step": 18500
+    },
+    {
+      "epoch": 0.1467895578544002,
+      "grad_norm": 0.0014668807853013277,
+      "learning_rate": 9.785564673969028e-05,
+      "loss": 0.0009,
+      "step": 19000
+    },
+    {
+      "epoch": 0.1506524409558318,
+      "grad_norm": 0.0013408615486696362,
+      "learning_rate": 0.00010043093087937369,
+      "loss": 0.0009,
+      "step": 19500
+    },
+    {
+      "epoch": 0.15451532405726337,
+      "grad_norm": 0.0015196386957541108,
+      "learning_rate": 0.00010300621501905708,
+      "loss": 0.0009,
+      "step": 20000
+    },
+    {
+      "epoch": 0.15837820715869497,
+      "grad_norm": 0.0013054045848548412,
+      "learning_rate": 0.0001055814991587405,
+      "loss": 0.0009,
+      "step": 20500
+    },
+    {
+      "epoch": 0.16224109026012654,
+      "grad_norm": 0.001247019157744944,
+      "learning_rate": 0.00010815678329842391,
+      "loss": 0.0009,
+      "step": 21000
+    },
+    {
+      "epoch": 0.16610397336155813,
+      "grad_norm": 0.001298928284086287,
+      "learning_rate": 0.00011073206743810733,
+      "loss": 0.0008,
+      "step": 21500
+    },
+    {
+      "epoch": 0.16996685646298973,
+      "grad_norm": 0.0011672589462250471,
+      "learning_rate": 0.00011330735157779074,
+      "loss": 0.0008,
+      "step": 22000
+    },
+    {
+      "epoch": 0.1738297395644213,
+      "grad_norm": 0.0012577781453728676,
+      "learning_rate": 0.00011588263571747416,
+      "loss": 0.0008,
+      "step": 22500
+    },
+    {
+      "epoch": 0.1776926226658529,
+      "grad_norm": 0.0011562498984858394,
+      "learning_rate": 0.00011845791985715756,
+      "loss": 0.0008,
+      "step": 23000
+    },
+    {
+      "epoch": 0.18155550576728446,
+      "grad_norm": 0.0011603948660194874,
+      "learning_rate": 0.00012103320399684098,
+      "loss": 0.0008,
+      "step": 23500
+    },
+    {
+      "epoch": 0.18541838886871606,
+      "grad_norm": 0.0011685957433655858,
+      "learning_rate": 0.00012360848813652438,
+      "loss": 0.0008,
+      "step": 24000
+    },
+    {
+      "epoch": 0.18928127197014763,
+      "grad_norm": 0.0012253538006916642,
+      "learning_rate": 0.0001261837722762078,
+      "loss": 0.0008,
+      "step": 24500
+    },
+    {
+      "epoch": 0.19314415507157923,
+      "grad_norm": 0.0009547853842377663,
+      "learning_rate": 0.0001287590564158912,
+      "loss": 0.0008,
+      "step": 25000
+    },
+    {
+      "epoch": 0.1970070381730108,
+      "grad_norm": 0.0009738618973642588,
+      "learning_rate": 0.0001313343405555746,
+      "loss": 0.0007,
+      "step": 25500
+    },
+    {
+      "epoch": 0.2008699212744424,
+      "grad_norm": 0.0009221744840033352,
+      "learning_rate": 0.00013390962469525801,
+      "loss": 0.0007,
+      "step": 26000
+    },
+    {
+      "epoch": 0.204732804375874,
+      "grad_norm": 0.0010526985861361027,
+      "learning_rate": 0.00013648490883494144,
+      "loss": 0.0007,
+      "step": 26500
+    },
+    {
+      "epoch": 0.20859568747730556,
+      "grad_norm": 0.0010718920966610312,
+      "learning_rate": 0.00013906019297462485,
+      "loss": 0.0007,
+      "step": 27000
+    },
+    {
+      "epoch": 0.21245857057873715,
+      "grad_norm": 0.0010513877496123314,
+      "learning_rate": 0.00014163547711430828,
+      "loss": 0.0007,
+      "step": 27500
+    },
+    {
+      "epoch": 0.21632145368016872,
+      "grad_norm": 0.0010410523973405361,
+      "learning_rate": 0.00014421076125399168,
+      "loss": 0.0007,
+      "step": 28000
+    },
+    {
+      "epoch": 0.22018433678160032,
+      "grad_norm": 0.0010957131162285805,
+      "learning_rate": 0.00014678604539367508,
+      "loss": 0.0007,
+      "step": 28500
+    },
+    {
+      "epoch": 0.2240472198830319,
+      "grad_norm": 0.0011075297370553017,
+      "learning_rate": 0.0001493613295333585,
+      "loss": 0.0007,
+      "step": 29000
+    },
+    {
+      "epoch": 0.2279101029844635,
+      "grad_norm": 0.0009366936865262687,
+      "learning_rate": 0.00015193661367304188,
+      "loss": 0.0007,
+      "step": 29500
+    },
+    {
+      "epoch": 0.23177298608589506,
+      "grad_norm": 0.0009237458580173552,
+      "learning_rate": 0.00015451189781272534,
+      "loss": 0.0007,
+      "step": 30000
+    },
+    {
+      "epoch": 0.23563586918732665,
+      "grad_norm": 0.0009658120106905699,
+      "learning_rate": 0.00015708718195240872,
+      "loss": 0.0007,
+      "step": 30500
+    },
+    {
+      "epoch": 0.23949875228875825,
+      "grad_norm": 0.0009536503930576146,
+      "learning_rate": 0.00015966246609209214,
+      "loss": 0.0007,
+      "step": 31000
+    },
+    {
+      "epoch": 0.24336163539018982,
+      "grad_norm": 0.0009778173407539725,
+      "learning_rate": 0.00016223775023177555,
+      "loss": 0.0006,
+      "step": 31500
+    },
+    {
+      "epoch": 0.24722451849162141,
+      "grad_norm": 0.0010041093919426203,
+      "learning_rate": 0.00016481303437145898,
+      "loss": 0.0006,
+      "step": 32000
+    },
+    {
+      "epoch": 0.251087401593053,
+      "grad_norm": 0.0008569998317398131,
+      "learning_rate": 0.00016738831851114238,
+      "loss": 0.0006,
+      "step": 32500
+    },
+    {
+      "epoch": 0.25495028469448455,
+      "grad_norm": 0.0009386276942677796,
+      "learning_rate": 0.00016996360265082578,
+      "loss": 0.0006,
+      "step": 33000
+    },
+    {
+      "epoch": 0.2588131677959162,
+      "grad_norm": 0.0008849843870848417,
+      "learning_rate": 0.0001725388867905092,
+      "loss": 0.0006,
+      "step": 33500
+    },
+    {
+      "epoch": 0.26267605089734775,
+      "grad_norm": 0.0008240257739089429,
+      "learning_rate": 0.0001751141709301926,
+      "loss": 0.0006,
+      "step": 34000
+    },
+    {
+      "epoch": 0.2665389339987793,
+      "grad_norm": 0.0009251784067600965,
+      "learning_rate": 0.00017768945506987604,
+      "loss": 0.0006,
+      "step": 34500
+    },
+    {
+      "epoch": 0.27040181710021094,
+      "grad_norm": 0.0009474306716583669,
+      "learning_rate": 0.00018026473920955944,
+      "loss": 0.0006,
+      "step": 35000
+    },
+    {
+      "epoch": 0.2742647002016425,
+      "grad_norm": 0.0008043038542382419,
+      "learning_rate": 0.00018284002334924285,
+      "loss": 0.0006,
+      "step": 35500
+    },
+    {
+      "epoch": 0.2781275833030741,
+      "grad_norm": 0.0007399991736747324,
+      "learning_rate": 0.00018541530748892627,
+      "loss": 0.0006,
+      "step": 36000
+    },
+    {
+      "epoch": 0.28199046640450565,
+      "grad_norm": 0.0008425785345025361,
+      "learning_rate": 0.00018799059162860968,
+      "loss": 0.0006,
+      "step": 36500
+    },
+    {
+      "epoch": 0.28585334950593727,
+      "grad_norm": 0.0008299360633827746,
+      "learning_rate": 0.0001905658757682931,
+      "loss": 0.0006,
+      "step": 37000
+    },
+    {
+      "epoch": 0.28971623260736884,
+      "grad_norm": 0.0007926743710413575,
+      "learning_rate": 0.0001931411599079765,
+      "loss": 0.0006,
+      "step": 37500
+    },
+    {
+      "epoch": 0.2935791157088004,
+      "grad_norm": 0.0008646908099763095,
+      "learning_rate": 0.00019571644404765988,
+      "loss": 0.0006,
+      "step": 38000
+    },
+    {
+      "epoch": 0.297441998810232,
+      "grad_norm": 0.000800468900706619,
+      "learning_rate": 0.0001982917281873433,
+      "loss": 0.0006,
+      "step": 38500
+    },
+    {
+      "epoch": 0.3013048819116636,
+      "grad_norm": 0.0007611193577758968,
+      "learning_rate": 0.00020086701232702671,
+      "loss": 0.0006,
+      "step": 39000
+    },
+    {
+      "epoch": 0.30516776501309517,
+      "grad_norm": 0.0007989577716216445,
+      "learning_rate": 0.00020344229646671014,
+      "loss": 0.0006,
+      "step": 39500
+    },
+    {
+      "epoch": 0.30903064811452674,
+      "grad_norm": 0.0007914176676422358,
+      "learning_rate": 0.00020601758060639355,
+      "loss": 0.0006,
+      "step": 40000
+    },
+    {
+      "epoch": 0.31289353121595836,
+      "grad_norm": 0.0007808151422068477,
+      "learning_rate": 0.00020859286474607698,
+      "loss": 0.0006,
+      "step": 40500
+    },
+    {
+      "epoch": 0.31675641431738993,
+      "grad_norm": 0.0007411599508486688,
+      "learning_rate": 0.00021116814888576038,
+      "loss": 0.0006,
+      "step": 41000
+    },
+    {
+      "epoch": 0.3206192974188215,
+      "grad_norm": 0.0007389839738607407,
+      "learning_rate": 0.00021374343302544378,
+      "loss": 0.0005,
+      "step": 41500
+    },
+    {
+      "epoch": 0.32448218052025307,
+      "grad_norm": 0.0008718458120711148,
+      "learning_rate": 0.0002163187171651272,
+      "loss": 0.0005,
+      "step": 42000
+    },
+    {
+      "epoch": 0.3283450636216847,
+      "grad_norm": 0.0007400435279123485,
+      "learning_rate": 0.0002188940013048106,
+      "loss": 0.0005,
+      "step": 42500
+    },
+    {
+      "epoch": 0.33220794672311627,
+      "grad_norm": 0.0006650601280853152,
+      "learning_rate": 0.00022146928544449404,
+      "loss": 0.0005,
+      "step": 43000
+    },
+    {
+      "epoch": 0.33607082982454783,
+      "grad_norm": 0.0006877753767184913,
+      "learning_rate": 0.00022404456958417744,
+      "loss": 0.0005,
+      "step": 43500
+    },
+    {
+      "epoch": 0.33993371292597946,
+      "grad_norm": 0.000792111677583307,
+      "learning_rate": 0.00022661985372386084,
+      "loss": 0.0005,
+      "step": 44000
+    },
+    {
+      "epoch": 0.343796596027411,
+      "grad_norm": 0.0006559982430189848,
+      "learning_rate": 0.00022919513786354427,
+      "loss": 0.0005,
+      "step": 44500
+    },
+    {
+      "epoch": 0.3476594791288426,
+      "grad_norm": 0.000791021331679076,
+      "learning_rate": 0.00023177042200322768,
+      "loss": 0.0005,
+      "step": 45000
+    },
+    {
+      "epoch": 0.35152236223027417,
+      "grad_norm": 0.000694429618306458,
+      "learning_rate": 0.0002343457061429111,
+      "loss": 0.0005,
+      "step": 45500
+    },
+    {
+      "epoch": 0.3553852453317058,
+      "grad_norm": 0.0006511561223305762,
+      "learning_rate": 0.00023692099028259448,
+      "loss": 0.0005,
+      "step": 46000
+    },
+    {
+      "epoch": 0.35924812843313736,
+      "grad_norm": 0.0007060636417008936,
+      "learning_rate": 0.00023949627442227788,
+      "loss": 0.0005,
+      "step": 46500
+    },
+    {
+      "epoch": 0.36311101153456893,
+      "grad_norm": 0.0006298599764704704,
+      "learning_rate": 0.0002420715585619613,
+      "loss": 0.0005,
+      "step": 47000
+    },
+    {
+      "epoch": 0.36697389463600055,
+      "grad_norm": 0.0006566342781297863,
+      "learning_rate": 0.00024464684270164474,
+      "loss": 0.0005,
+      "step": 47500
+    },
+    {
+      "epoch": 0.3708367777374321,
+      "grad_norm": 0.0006428054184652865,
+      "learning_rate": 0.00024722212684132814,
+      "loss": 0.0005,
+      "step": 48000
+    },
+    {
+      "epoch": 0.3746996608388637,
+      "grad_norm": 0.0006790259503759444,
+      "learning_rate": 0.00024979741098101155,
+      "loss": 0.0005,
+      "step": 48500
+    },
+    {
+      "epoch": 0.37856254394029526,
+      "grad_norm": 0.0006684795371256769,
+      "learning_rate": 0.00025237269512069495,
+      "loss": 0.0005,
+      "step": 49000
+    },
+    {
+      "epoch": 0.3824254270417269,
+      "grad_norm": 0.0006840928690508008,
+      "learning_rate": 0.00025494797926037835,
+      "loss": 0.0005,
+      "step": 49500
+    },
+    {
+      "epoch": 0.38628831014315845,
+      "grad_norm": 0.0005870233871974051,
+      "learning_rate": 0.00025752326340006175,
+      "loss": 0.0005,
+      "step": 50000
+    },
+    {
+      "epoch": 0.39015119324459,
+      "grad_norm": 0.0006383236614055932,
+      "learning_rate": 0.0002600985475397452,
+      "loss": 0.0005,
+      "step": 50500
+    },
+    {
+      "epoch": 0.3940140763460216,
+      "grad_norm": 0.0006128019886091352,
+      "learning_rate": 0.0002626738316794286,
+      "loss": 0.0005,
+      "step": 51000
+    },
+    {
+      "epoch": 0.3978769594474532,
+      "grad_norm": 0.0006441852310672402,
+      "learning_rate": 0.000265249115819112,
+      "loss": 0.0005,
+      "step": 51500
+    },
+    {
+      "epoch": 0.4017398425488848,
+      "grad_norm": 0.0005380984512157738,
+      "learning_rate": 0.0002678243999587954,
+      "loss": 0.0005,
+      "step": 52000
+    },
+    {
+      "epoch": 0.40560272565031635,
+      "grad_norm": 0.0006370097398757935,
+      "learning_rate": 0.0002703996840984788,
+      "loss": 0.0005,
+      "step": 52500
+    },
+    {
+      "epoch": 0.409465608751748,
+      "grad_norm": 0.0006471597589552402,
+      "learning_rate": 0.0002729749682381623,
+      "loss": 0.0005,
+      "step": 53000
+    },
+    {
+      "epoch": 0.41332849185317955,
+      "grad_norm": 0.0005368549609556794,
+      "learning_rate": 0.0002755502523778457,
+      "loss": 0.0005,
+      "step": 53500
+    },
+    {
+      "epoch": 0.4171913749546111,
+      "grad_norm": 0.0005874933558516204,
+      "learning_rate": 0.0002781255365175291,
+      "loss": 0.0005,
+      "step": 54000
+    },
+    {
+      "epoch": 0.4210542580560427,
+      "grad_norm": 0.0005409115110523999,
+      "learning_rate": 0.0002807008206572125,
+      "loss": 0.0005,
+      "step": 54500
+    },
+    {
+      "epoch": 0.4249171411574743,
+      "grad_norm": 0.0005910882027819753,
+      "learning_rate": 0.0002832761047968959,
+      "loss": 0.0005,
+      "step": 55000
+    },
+    {
+      "epoch": 0.4287800242589059,
+      "grad_norm": 0.0006001652218401432,
+      "learning_rate": 0.00028585138893657934,
+      "loss": 0.0005,
+      "step": 55500
+    },
+    {
+      "epoch": 0.43264290736033745,
+      "grad_norm": 0.000514556304551661,
+      "learning_rate": 0.00028842667307626274,
+      "loss": 0.0005,
+      "step": 56000
+    },
+    {
+      "epoch": 0.43650579046176907,
+      "grad_norm": 0.0005608577048406005,
+      "learning_rate": 0.00029100195721594614,
+      "loss": 0.0005,
+      "step": 56500
+    },
+    {
+      "epoch": 0.44036867356320064,
+      "grad_norm": 0.0007016558083705604,
+      "learning_rate": 0.00029357724135562954,
+      "loss": 0.0005,
+      "step": 57000
+    },
+    {
+      "epoch": 0.4442315566646322,
+      "grad_norm": 0.0005601047305390239,
+      "learning_rate": 0.00029615252549531295,
+      "loss": 0.0005,
+      "step": 57500
+    },
+    {
+      "epoch": 0.4480944397660638,
+      "grad_norm": 0.0004940334474667907,
+      "learning_rate": 0.00029872780963499635,
+      "loss": 0.0005,
+      "step": 58000
+    },
+    {
+      "epoch": 0.4519573228674954,
+      "grad_norm": 0.000568122137337923,
+      "learning_rate": 0.00029977004320400405,
+      "loss": 0.0005,
+      "step": 58500
+    },
+    {
+      "epoch": 0.455820205968927,
+      "grad_norm": 0.0005688059609383345,
+      "learning_rate": 0.0002993155831328659,
+      "loss": 0.0005,
+      "step": 59000
+    },
+    {
+      "epoch": 0.45968308907035854,
+      "grad_norm": 0.0005606883205473423,
+      "learning_rate": 0.0002988611230617278,
+      "loss": 0.0005,
+      "step": 59500
+    },
+    {
+      "epoch": 0.4635459721717901,
+      "grad_norm": 0.0005766854155808687,
+      "learning_rate": 0.0002984066629905896,
+      "loss": 0.0005,
+      "step": 60000
+    },
+    {
+      "epoch": 0.46740885527322173,
+      "grad_norm": 0.000506491691339761,
+      "learning_rate": 0.00029795220291945146,
+      "loss": 0.0005,
+      "step": 60500
+    },
+    {
+      "epoch": 0.4712717383746533,
+      "grad_norm": 0.0005349202547222376,
+      "learning_rate": 0.0002974977428483133,
+      "loss": 0.0005,
+      "step": 61000
+    },
+    {
+      "epoch": 0.4751346214760849,
+      "grad_norm": 0.0005307575920596719,
+      "learning_rate": 0.00029704328277717514,
+      "loss": 0.0005,
+      "step": 61500
+    },
+    {
+      "epoch": 0.4789975045775165,
+      "grad_norm": 0.0005095390370115638,
+      "learning_rate": 0.00029658882270603706,
+      "loss": 0.0005,
+      "step": 62000
+    },
+    {
+      "epoch": 0.48286038767894807,
+      "grad_norm": 0.0005189366056583822,
+      "learning_rate": 0.00029613436263489887,
+      "loss": 0.0005,
+      "step": 62500
+    },
+    {
+      "epoch": 0.48672327078037964,
+      "grad_norm": 0.0005579465650953352,
+      "learning_rate": 0.0002956799025637607,
+      "loss": 0.0005,
+      "step": 63000
+    },
+    {
+      "epoch": 0.4905861538818112,
+      "grad_norm": 0.000490622769575566,
+      "learning_rate": 0.0002952254424926226,
+      "loss": 0.0005,
+      "step": 63500
+    },
+    {
+      "epoch": 0.49444903698324283,
+      "grad_norm": 0.00047076333430595696,
+      "learning_rate": 0.0002947709824214844,
+      "loss": 0.0005,
+      "step": 64000
+    },
+    {
+      "epoch": 0.4983119200846744,
+      "grad_norm": 0.0005202812608331442,
+      "learning_rate": 0.0002943165223503463,
+      "loss": 0.0005,
+      "step": 64500
+    },
+    {
+      "epoch": 0.502174803186106,
+      "grad_norm": 0.0005745698581449687,
+      "learning_rate": 0.00029386206227920814,
+      "loss": 0.0005,
+      "step": 65000
+    },
+    {
+      "epoch": 0.5060376862875375,
+      "grad_norm": 0.0005540290731005371,
+      "learning_rate": 0.00029340760220806996,
+      "loss": 0.0004,
+      "step": 65500
+    },
+    {
+      "epoch": 0.5099005693889691,
+      "grad_norm": 0.0005399156943894923,
+      "learning_rate": 0.0002929531421369318,
+      "loss": 0.0004,
+      "step": 66000
+    },
+    {
+      "epoch": 0.5137634524904008,
+      "grad_norm": 0.0005016261129640043,
+      "learning_rate": 0.0002924986820657937,
+      "loss": 0.0004,
+      "step": 66500
+    },
+    {
+      "epoch": 0.5176263355918324,
+      "grad_norm": 0.0004883984220214188,
+      "learning_rate": 0.00029204422199465555,
+      "loss": 0.0004,
+      "step": 67000
+    },
+    {
+      "epoch": 0.5214892186932639,
+      "grad_norm": 0.0004978585639037192,
+      "learning_rate": 0.00029158976192351736,
+      "loss": 0.0004,
+      "step": 67500
+    },
+    {
+      "epoch": 0.5253521017946955,
+      "grad_norm": 0.00048788345884531736,
+      "learning_rate": 0.00029113530185237923,
+      "loss": 0.0004,
+      "step": 68000
+    },
+    {
+      "epoch": 0.5292149848961271,
+      "grad_norm": 0.0005122956354171038,
+      "learning_rate": 0.0002906808417812411,
+      "loss": 0.0004,
+      "step": 68500
+    },
+    {
+      "epoch": 0.5330778679975586,
+      "grad_norm": 0.0005075953667983413,
+      "learning_rate": 0.0002902263817101029,
+      "loss": 0.0004,
+      "step": 69000
+    },
+    {
+      "epoch": 0.5369407510989902,
+      "grad_norm": 0.010060743428766727,
+      "learning_rate": 0.00028977192163896477,
+      "loss": 0.0004,
+      "step": 69500
+    },
+    {
+      "epoch": 0.5408036342004219,
+      "grad_norm": 0.00045171918463893235,
+      "learning_rate": 0.00028931746156782664,
+      "loss": 0.0004,
+      "step": 70000
+    },
+    {
+      "epoch": 0.5446665173018534,
+      "grad_norm": 0.00047910166904330254,
+      "learning_rate": 0.00028886300149668845,
+      "loss": 0.0004,
+      "step": 70500
+    },
+    {
+      "epoch": 0.548529400403285,
+      "grad_norm": 0.0004093217430636287,
+      "learning_rate": 0.0002884085414255503,
+      "loss": 0.0004,
+      "step": 71000
+    },
+    {
+      "epoch": 0.5523922835047166,
+      "grad_norm": 0.0005109374760650098,
+      "learning_rate": 0.0002879540813544122,
+      "loss": 0.0004,
+      "step": 71500
+    },
+    {
+      "epoch": 0.5562551666061482,
+      "grad_norm": 0.00047159241512417793,
+      "learning_rate": 0.00028749962128327405,
+      "loss": 0.0004,
+      "step": 72000
+    },
+    {
+      "epoch": 0.5601180497075797,
+      "grad_norm": 0.00048621243331581354,
+      "learning_rate": 0.00028704516121213586,
+      "loss": 0.0004,
+      "step": 72500
+    },
+    {
+      "epoch": 0.5639809328090113,
+      "grad_norm": 0.0005125854513607919,
+      "learning_rate": 0.0002865907011409977,
+      "loss": 0.0004,
+      "step": 73000
+    },
+    {
+      "epoch": 0.567843815910443,
+      "grad_norm": 0.00048265952500514686,
+      "learning_rate": 0.0002861362410698596,
+      "loss": 0.0004,
+      "step": 73500
+    },
+    {
+      "epoch": 0.5717066990118745,
+      "grad_norm": 0.0004471069551073015,
+      "learning_rate": 0.0002856817809987214,
+      "loss": 0.0004,
+      "step": 74000
+    },
+    {
+      "epoch": 0.5755695821133061,
+      "grad_norm": 0.0005123110022395849,
+      "learning_rate": 0.00028522732092758327,
+      "loss": 0.0004,
+      "step": 74500
+    },
+    {
+      "epoch": 0.5794324652147377,
+      "grad_norm": 0.00046480647870339453,
+      "learning_rate": 0.00028477286085644513,
+      "loss": 0.0004,
+      "step": 75000
+    },
+    {
+      "epoch": 0.5832953483161692,
+      "grad_norm": 0.000416387541918084,
+      "learning_rate": 0.00028431840078530694,
+      "loss": 0.0004,
+      "step": 75500
+    },
+    {
+      "epoch": 0.5871582314176008,
+      "grad_norm": 0.00046883479808457196,
+      "learning_rate": 0.0002838639407141688,
+      "loss": 0.0004,
+      "step": 76000
+    },
+    {
+      "epoch": 0.5910211145190324,
+      "grad_norm": 0.00046609266428276896,
+      "learning_rate": 0.0002834094806430307,
+      "loss": 0.0004,
+      "step": 76500
+    },
+    {
+      "epoch": 0.594883997620464,
+      "grad_norm": 0.0004421696939971298,
+      "learning_rate": 0.00028295502057189254,
+      "loss": 0.0004,
+      "step": 77000
+    },
+    {
+      "epoch": 0.5987468807218956,
+      "grad_norm": 0.0005302332574501634,
+      "learning_rate": 0.0002825005605007544,
+      "loss": 0.0004,
+      "step": 77500
+    },
+    {
+      "epoch": 0.6026097638233272,
+      "grad_norm": 0.0004425778752192855,
+      "learning_rate": 0.0002820461004296162,
+      "loss": 0.0004,
+      "step": 78000
+    },
+    {
+      "epoch": 0.6064726469247588,
+      "grad_norm": 0.00043949965038336813,
+      "learning_rate": 0.0002815916403584781,
+      "loss": 0.0004,
+      "step": 78500
+    },
+    {
+      "epoch": 0.6103355300261903,
+      "grad_norm": 0.00048047167365439236,
+      "learning_rate": 0.00028113718028733995,
+      "loss": 0.0004,
+      "step": 79000
+    },
+    {
+      "epoch": 0.6141984131276219,
+      "grad_norm": 0.00046403607120737433,
+      "learning_rate": 0.0002806827202162018,
+      "loss": 0.0004,
+      "step": 79500
+    },
+    {
+      "epoch": 0.6180612962290535,
+      "grad_norm": 0.0004179133102297783,
+      "learning_rate": 0.00028022826014506363,
+      "loss": 0.0004,
+      "step": 80000
+    },
+    {
+      "epoch": 0.621924179330485,
+      "grad_norm": 0.0005025283899158239,
+      "learning_rate": 0.0002797738000739255,
+      "loss": 0.0004,
+      "step": 80500
+    },
+    {
+      "epoch": 0.6257870624319167,
+      "grad_norm": 0.00045130823855288327,
+      "learning_rate": 0.00027931934000278736,
+      "loss": 0.0004,
+      "step": 81000
+    },
+    {
+      "epoch": 0.6296499455333483,
+      "grad_norm": 0.00047835326404310763,
+      "learning_rate": 0.00027886487993164917,
+      "loss": 0.0004,
+      "step": 81500
+    },
+    {
+      "epoch": 0.6335128286347799,
+      "grad_norm": 0.0004951538285240531,
+      "learning_rate": 0.00027841041986051104,
+      "loss": 0.0004,
+      "step": 82000
+    },
+    {
+      "epoch": 0.6373757117362114,
+      "grad_norm": 0.0004999985103495419,
+      "learning_rate": 0.0002779559597893729,
+      "loss": 0.0004,
+      "step": 82500
+    },
+    {
+      "epoch": 0.641238594837643,
+      "grad_norm": 0.00045166266500018537,
+      "learning_rate": 0.0002775014997182347,
+      "loss": 0.0004,
+      "step": 83000
+    },
+    {
+      "epoch": 0.6451014779390746,
+      "grad_norm": 0.0004108847351744771,
+      "learning_rate": 0.0002770470396470966,
+      "loss": 0.0004,
+      "step": 83500
+    },
+    {
+      "epoch": 0.6489643610405061,
+      "grad_norm": 0.0004798057198058814,
+      "learning_rate": 0.00027659257957595844,
+      "loss": 0.0004,
+      "step": 84000
+    },
+    {
+      "epoch": 0.6528272441419378,
+      "grad_norm": 0.0004300622676964849,
+      "learning_rate": 0.0002761381195048203,
+      "loss": 0.0004,
+      "step": 84500
+    },
+    {
+      "epoch": 0.6566901272433694,
+      "grad_norm": 0.00042947594192810357,
+      "learning_rate": 0.0002756836594336821,
+      "loss": 0.0004,
+      "step": 85000
+    },
+    {
+      "epoch": 0.660553010344801,
+      "grad_norm": 0.0004585100687108934,
+      "learning_rate": 0.000275229199362544,
+      "loss": 0.0004,
+      "step": 85500
+    },
+    {
+      "epoch": 0.6644158934462325,
+      "grad_norm": 0.00048575850087217987,
+      "learning_rate": 0.00027477473929140585,
+      "loss": 0.0004,
+      "step": 86000
+    },
+    {
+      "epoch": 0.6682787765476641,
+      "grad_norm": 0.0004099896177649498,
+      "learning_rate": 0.00027432027922026766,
+      "loss": 0.0004,
+      "step": 86500
+    },
+    {
+      "epoch": 0.6721416596490957,
+      "grad_norm": 0.000426844839239493,
+      "learning_rate": 0.00027386581914912953,
+      "loss": 0.0004,
+      "step": 87000
+    },
+    {
+      "epoch": 0.6760045427505272,
+      "grad_norm": 0.000434650486567989,
+      "learning_rate": 0.0002734113590779914,
+      "loss": 0.0004,
+      "step": 87500
+    },
+    {
+      "epoch": 0.6798674258519589,
+      "grad_norm": 0.0004488716658670455,
+      "learning_rate": 0.0002729568990068532,
+      "loss": 0.0004,
+      "step": 88000
+    },
+    {
+      "epoch": 0.6837303089533905,
+      "grad_norm": 0.0004696206597145647,
+      "learning_rate": 0.0002725024389357151,
+      "loss": 0.0004,
+      "step": 88500
+    },
+    {
+      "epoch": 0.687593192054822,
+      "grad_norm": 0.00042553144157864153,
+      "learning_rate": 0.00027204797886457694,
+      "loss": 0.0004,
+      "step": 89000
+    },
+    {
+      "epoch": 0.6914560751562536,
+      "grad_norm": 0.0004377859877422452,
+      "learning_rate": 0.00027159351879343875,
+      "loss": 0.0004,
+      "step": 89500
+    },
+    {
+      "epoch": 0.6953189582576852,
+      "grad_norm": 0.00040723910205997527,
+      "learning_rate": 0.0002711390587223006,
+      "loss": 0.0004,
+      "step": 90000
+    },
+    {
+      "epoch": 0.6991818413591168,
+      "grad_norm": 0.0004500757495407015,
+      "learning_rate": 0.0002706845986511625,
+      "loss": 0.0004,
+      "step": 90500
+    },
+    {
+      "epoch": 0.7030447244605483,
+      "grad_norm": 0.00040941766928881407,
+      "learning_rate": 0.00027023013858002435,
+      "loss": 0.0004,
+      "step": 91000
+    },
+    {
+      "epoch": 0.70690760756198,
+      "grad_norm": 0.00045324547681957483,
+      "learning_rate": 0.00026977567850888616,
+      "loss": 0.0004,
+      "step": 91500
+    },
+    {
+      "epoch": 0.7107704906634116,
+      "grad_norm": 0.00047752921818755567,
+      "learning_rate": 0.000269321218437748,
+      "loss": 0.0004,
+      "step": 92000
+    },
+    {
+      "epoch": 0.7146333737648431,
+      "grad_norm": 0.0004383829946164042,
+      "learning_rate": 0.0002688667583666099,
+      "loss": 0.0004,
+      "step": 92500
+    },
+    {
+      "epoch": 0.7184962568662747,
+      "grad_norm": 0.00045177084393799305,
+      "learning_rate": 0.0002684122982954717,
+      "loss": 0.0004,
+      "step": 93000
+    },
+    {
+      "epoch": 0.7223591399677063,
+      "grad_norm": 0.0004552272439468652,
+      "learning_rate": 0.0002679578382243336,
+      "loss": 0.0004,
+      "step": 93500
+    },
+    {
+      "epoch": 0.7262220230691379,
+      "grad_norm": 0.00043098500464111567,
+      "learning_rate": 0.00026750337815319543,
+      "loss": 0.0004,
+      "step": 94000
+    },
+    {
+      "epoch": 0.7300849061705694,
+      "grad_norm": 0.00043315321090631187,
+      "learning_rate": 0.0002670489180820573,
+      "loss": 0.0004,
+      "step": 94500
+    },
+    {
+      "epoch": 0.7339477892720011,
+      "grad_norm": 0.00043436733540147543,
+      "learning_rate": 0.00026659445801091916,
+      "loss": 0.0004,
+      "step": 95000
+    },
+    {
+      "epoch": 0.7378106723734327,
+      "grad_norm": 0.00040910765528678894,
+      "learning_rate": 0.000266139997939781,
+      "loss": 0.0004,
+      "step": 95500
+    },
+    {
+      "epoch": 0.7416735554748642,
+      "grad_norm": 0.0004274219973012805,
+      "learning_rate": 0.00026568553786864284,
+      "loss": 0.0004,
+      "step": 96000
+    },
+    {
+      "epoch": 0.7455364385762958,
+      "grad_norm": 0.0004410296387504786,
+      "learning_rate": 0.0002652310777975047,
+      "loss": 0.0004,
+      "step": 96500
+    },
+    {
+      "epoch": 0.7493993216777274,
+      "grad_norm": 0.0004376559518277645,
+      "learning_rate": 0.0002647766177263665,
+      "loss": 0.0004,
+      "step": 97000
+    },
+    {
+      "epoch": 0.753262204779159,
+      "grad_norm": 0.00047215758240781724,
+      "learning_rate": 0.0002643221576552284,
+      "loss": 0.0004,
+      "step": 97500
+    },
+    {
+      "epoch": 0.7571250878805905,
+      "grad_norm": 0.00042207614751532674,
+      "learning_rate": 0.00026386769758409025,
+      "loss": 0.0004,
+      "step": 98000
+    },
+    {
+      "epoch": 0.7609879709820221,
+      "grad_norm": 0.00042739673517644405,
+      "learning_rate": 0.0002634132375129521,
+      "loss": 0.0004,
+      "step": 98500
+    },
+    {
+      "epoch": 0.7648508540834538,
+      "grad_norm": 0.00041113453335128725,
+      "learning_rate": 0.00026295877744181393,
+      "loss": 0.0004,
+      "step": 99000
+    },
+    {
+      "epoch": 0.7687137371848853,
+      "grad_norm": 0.00043188530253246427,
+      "learning_rate": 0.0002625043173706758,
+      "loss": 0.0004,
+      "step": 99500
+    },
+    {
+      "epoch": 0.7725766202863169,
+      "grad_norm": 0.00045034836512058973,
+      "learning_rate": 0.00026204985729953766,
+      "loss": 0.0004,
+      "step": 100000
+    },
+    {
+      "epoch": 0.7764395033877485,
+      "grad_norm": 0.0004060341161675751,
+      "learning_rate": 0.00026159539722839947,
+      "loss": 0.0004,
+      "step": 100500
+    },
+    {
+      "epoch": 0.78030238648918,
+      "grad_norm": 0.00040883294423110783,
+      "learning_rate": 0.00026114093715726134,
+      "loss": 0.0004,
+      "step": 101000
+    },
+    {
+      "epoch": 0.7841652695906116,
+      "grad_norm": 0.00044574079220183194,
+      "learning_rate": 0.0002606864770861232,
+      "loss": 0.0004,
+      "step": 101500
+    },
+    {
+      "epoch": 0.7880281526920432,
+      "grad_norm": 0.0004099115321878344,
+      "learning_rate": 0.000260232017014985,
+      "loss": 0.0004,
+      "step": 102000
+    },
+    {
+      "epoch": 0.7918910357934749,
+      "grad_norm": 0.0004170772444922477,
+      "learning_rate": 0.0002597775569438469,
+      "loss": 0.0004,
+      "step": 102500
+    },
+    {
+      "epoch": 0.7957539188949064,
+      "grad_norm": 0.0004352552932687104,
+      "learning_rate": 0.00025932309687270875,
+      "loss": 0.0004,
+      "step": 103000
+    },
+    {
+      "epoch": 0.799616801996338,
+      "grad_norm": 0.00040374931995756924,
+      "learning_rate": 0.0002588686368015706,
+      "loss": 0.0004,
+      "step": 103500
+    },
+    {
+      "epoch": 0.8034796850977696,
+      "grad_norm": 0.00044298075954429805,
+      "learning_rate": 0.0002584141767304324,
+      "loss": 0.0004,
+      "step": 104000
+    },
+    {
+      "epoch": 0.8073425681992011,
+      "grad_norm": 0.00042256712913513184,
+      "learning_rate": 0.0002579597166592943,
+      "loss": 0.0004,
+      "step": 104500
+    },
+    {
+      "epoch": 0.8112054513006327,
+      "grad_norm": 0.00036991469096392393,
+      "learning_rate": 0.00025750525658815615,
+      "loss": 0.0004,
+      "step": 105000
+    },
+    {
+      "epoch": 0.8150683344020643,
+      "grad_norm": 0.0004396368167363107,
+      "learning_rate": 0.00025705079651701797,
+      "loss": 0.0004,
+      "step": 105500
+    },
+    {
+      "epoch": 0.818931217503496,
+      "grad_norm": 0.0004093723255209625,
+      "learning_rate": 0.00025659633644587983,
+      "loss": 0.0004,
+      "step": 106000
+    },
+    {
+      "epoch": 0.8227941006049275,
+      "grad_norm": 0.000397693132981658,
+      "learning_rate": 0.0002561418763747417,
+      "loss": 0.0004,
+      "step": 106500
+    },
+    {
+      "epoch": 0.8266569837063591,
+      "grad_norm": 0.0004059737839270383,
+      "learning_rate": 0.0002556874163036035,
+      "loss": 0.0004,
+      "step": 107000
+    },
+    {
+      "epoch": 0.8305198668077907,
+      "grad_norm": 0.0004193147469777614,
+      "learning_rate": 0.00025523295623246543,
+      "loss": 0.0004,
+      "step": 107500
+    },
+    {
+      "epoch": 0.8343827499092222,
+      "grad_norm": 0.00043243260006420314,
+      "learning_rate": 0.00025477849616132724,
+      "loss": 0.0004,
+      "step": 108000
+    },
+    {
+      "epoch": 0.8382456330106538,
+      "grad_norm": 0.0004471674910746515,
+      "learning_rate": 0.0002543240360901891,
+      "loss": 0.0004,
+      "step": 108500
+    },
+    {
+      "epoch": 0.8421085161120854,
+      "grad_norm": 0.00039739784551784396,
+      "learning_rate": 0.00025386957601905097,
+      "loss": 0.0004,
+      "step": 109000
+    },
+    {
+      "epoch": 0.845971399213517,
+      "grad_norm": 0.00039233858115039766,
+      "learning_rate": 0.0002534151159479128,
+      "loss": 0.0004,
+      "step": 109500
+    },
+    {
+      "epoch": 0.8498342823149486,
+      "grad_norm": 0.00042606148053891957,
+      "learning_rate": 0.00025296065587677465,
+      "loss": 0.0004,
+      "step": 110000
+    },
+    {
+      "epoch": 0.8536971654163802,
+      "grad_norm": 0.0004140585951972753,
+      "learning_rate": 0.0002525061958056365,
+      "loss": 0.0004,
+      "step": 110500
+    },
+    {
+      "epoch": 0.8575600485178118,
+      "grad_norm": 0.0004395502619445324,
+      "learning_rate": 0.0002520517357344984,
+      "loss": 0.0004,
+      "step": 111000
+    },
+    {
+      "epoch": 0.8614229316192433,
+      "grad_norm": 0.0004029794654343277,
+      "learning_rate": 0.0002515972756633602,
+      "loss": 0.0004,
+      "step": 111500
+    },
+    {
+      "epoch": 0.8652858147206749,
+      "grad_norm": 0.00038435307214967906,
+      "learning_rate": 0.00025114281559222206,
+      "loss": 0.0004,
+      "step": 112000
+    },
+    {
+      "epoch": 0.8691486978221065,
+      "grad_norm": 0.00039446449954994023,
+      "learning_rate": 0.0002506883555210839,
+      "loss": 0.0004,
+      "step": 112500
+    },
+    {
+      "epoch": 0.8730115809235381,
+      "grad_norm": 0.0004027863615192473,
+      "learning_rate": 0.00025023389544994573,
+      "loss": 0.0004,
+      "step": 113000
+    },
+    {
+      "epoch": 0.8768744640249697,
+      "grad_norm": 0.00038630064227618277,
+      "learning_rate": 0.0002497794353788076,
+      "loss": 0.0004,
+      "step": 113500
+    },
+    {
+      "epoch": 0.8807373471264013,
+      "grad_norm": 0.0003898652794305235,
+      "learning_rate": 0.00024932497530766947,
+      "loss": 0.0004,
+      "step": 114000
+    },
+    {
+      "epoch": 0.8846002302278329,
+      "grad_norm": 0.0003884608449880034,
+      "learning_rate": 0.0002488705152365313,
+      "loss": 0.0004,
+      "step": 114500
+    },
+    {
+      "epoch": 0.8884631133292644,
+      "grad_norm": 0.0004448189283721149,
+      "learning_rate": 0.00024841605516539314,
+      "loss": 0.0004,
+      "step": 115000
+    },
+    {
+      "epoch": 0.892325996430696,
+      "grad_norm": 0.0004219826078042388,
+      "learning_rate": 0.000247961595094255,
+      "loss": 0.0004,
+      "step": 115500
+    },
+    {
+      "epoch": 0.8961888795321276,
+      "grad_norm": 0.0004262237052898854,
+      "learning_rate": 0.0002475071350231168,
+      "loss": 0.0004,
+      "step": 116000
+    },
+    {
+      "epoch": 0.9000517626335592,
+      "grad_norm": 0.0004264891322236508,
+      "learning_rate": 0.0002470526749519787,
+      "loss": 0.0004,
+      "step": 116500
+    },
+    {
+      "epoch": 0.9039146457349908,
+      "grad_norm": 0.0004334848199505359,
+      "learning_rate": 0.00024659821488084055,
+      "loss": 0.0004,
+      "step": 117000
+    },
+    {
+      "epoch": 0.9077775288364224,
+      "grad_norm": 0.0003926197241526097,
+      "learning_rate": 0.0002461437548097024,
+      "loss": 0.0004,
+      "step": 117500
+    },
+    {
+      "epoch": 0.911640411937854,
+      "grad_norm": 0.0004120821540709585,
+      "learning_rate": 0.00024568929473856423,
+      "loss": 0.0004,
+      "step": 118000
+    },
+    {
+      "epoch": 0.9155032950392855,
+      "grad_norm": 0.00037737927050329745,
+      "learning_rate": 0.0002452348346674261,
+      "loss": 0.0004,
+      "step": 118500
+    },
+    {
+      "epoch": 0.9193661781407171,
+      "grad_norm": 0.0003999023465439677,
+      "learning_rate": 0.00024478037459628796,
+      "loss": 0.0004,
+      "step": 119000
+    },
+    {
+      "epoch": 0.9232290612421487,
+      "grad_norm": 0.00039843874401412904,
+      "learning_rate": 0.00024432591452514977,
+      "loss": 0.0004,
+      "step": 119500
+    },
+    {
+      "epoch": 0.9270919443435802,
+      "grad_norm": 0.0004592763143591583,
+      "learning_rate": 0.00024387145445401166,
+      "loss": 0.0004,
+      "step": 120000
+    },
+    {
+      "epoch": 0.9309548274450119,
+      "grad_norm": 0.00041658771806396544,
+      "learning_rate": 0.0002434169943828735,
+      "loss": 0.0004,
+      "step": 120500
+    },
+    {
+      "epoch": 0.9348177105464435,
+      "grad_norm": 0.000446771620772779,
+      "learning_rate": 0.00024296253431173534,
+      "loss": 0.0004,
+      "step": 121000
+    },
+    {
+      "epoch": 0.938680593647875,
+      "grad_norm": 0.00039099738933146,
+      "learning_rate": 0.0002425080742405972,
+      "loss": 0.0004,
+      "step": 121500
+    },
+    {
+      "epoch": 0.9425434767493066,
+      "grad_norm": 0.0004380798782221973,
+      "learning_rate": 0.00024205361416945905,
+      "loss": 0.0004,
+      "step": 122000
+    },
+    {
+      "epoch": 0.9464063598507382,
+      "grad_norm": 0.0004325969784986228,
+      "learning_rate": 0.0002415991540983209,
+      "loss": 0.0004,
+      "step": 122500
+    },
+    {
+      "epoch": 0.9502692429521697,
+      "grad_norm": 0.00042953985393978655,
+      "learning_rate": 0.00024114469402718275,
+      "loss": 0.0004,
+      "step": 123000
+    },
+    {
+      "epoch": 0.9541321260536013,
+      "grad_norm": 0.00042357659549452364,
+      "learning_rate": 0.0002406902339560446,
+      "loss": 0.0004,
+      "step": 123500
+    },
+    {
+      "epoch": 0.957995009155033,
+      "grad_norm": 0.00041984530980698764,
+      "learning_rate": 0.00024023577388490645,
+      "loss": 0.0004,
+      "step": 124000
+    },
+    {
+      "epoch": 0.9618578922564646,
+      "grad_norm": 0.0003806228924077004,
+      "learning_rate": 0.0002397813138137683,
+      "loss": 0.0004,
+      "step": 124500
+    },
+    {
+      "epoch": 0.9657207753578961,
+      "grad_norm": 0.00043450563680380583,
+      "learning_rate": 0.00023932685374263016,
+      "loss": 0.0004,
+      "step": 125000
+    },
+    {
+      "epoch": 0.9695836584593277,
+      "grad_norm": 0.00040280865505337715,
+      "learning_rate": 0.000238872393671492,
+      "loss": 0.0004,
+      "step": 125500
+    },
+    {
+      "epoch": 0.9734465415607593,
+      "grad_norm": 0.00039368733996525407,
+      "learning_rate": 0.00023841793360035384,
+      "loss": 0.0004,
+      "step": 126000
+    },
+    {
+      "epoch": 0.9773094246621908,
+      "grad_norm": 0.0004507504927460104,
+      "learning_rate": 0.0002379634735292157,
+      "loss": 0.0004,
+      "step": 126500
+    },
+    {
+      "epoch": 0.9811723077636224,
+      "grad_norm": 0.00042234413558617234,
+      "learning_rate": 0.00023750901345807754,
+      "loss": 0.0004,
+      "step": 127000
+    },
+    {
+      "epoch": 0.9850351908650541,
+      "grad_norm": 0.00039364254917018116,
+      "learning_rate": 0.0002370545533869394,
+      "loss": 0.0004,
+      "step": 127500
+    },
+    {
+      "epoch": 0.9888980739664857,
+      "grad_norm": 0.00035987445153295994,
+      "learning_rate": 0.00023660009331580124,
+      "loss": 0.0004,
+      "step": 128000
+    },
+    {
+      "epoch": 0.9927609570679172,
+      "grad_norm": 0.00039482262218371034,
+      "learning_rate": 0.00023614563324466308,
+      "loss": 0.0004,
+      "step": 128500
+    },
+    {
+      "epoch": 0.9966238401693488,
+      "grad_norm": 0.0004762003954965621,
+      "learning_rate": 0.00023569117317352495,
+      "loss": 0.0004,
+      "step": 129000
+    },
+    {
+      "epoch": 1.0004867232707804,
+      "grad_norm": 0.0004270481294952333,
+      "learning_rate": 0.0002352367131023868,
+      "loss": 0.0004,
+      "step": 129500
+    },
+    {
+      "epoch": 1.004349606372212,
+      "grad_norm": 0.00039340462535619736,
+      "learning_rate": 0.00023478225303124868,
+      "loss": 0.0004,
+      "step": 130000
+    },
+    {
+      "epoch": 1.0082124894736435,
+      "grad_norm": 0.0004116026684641838,
+      "learning_rate": 0.00023432779296011052,
+      "loss": 0.0004,
+      "step": 130500
+    },
+    {
+      "epoch": 1.012075372575075,
+      "grad_norm": 0.0003860292781610042,
+      "learning_rate": 0.00023387333288897233,
+      "loss": 0.0004,
+      "step": 131000
+    },
+    {
+      "epoch": 1.0159382556765066,
+      "grad_norm": 0.00040404967148788273,
+      "learning_rate": 0.00023341887281783422,
+      "loss": 0.0004,
+      "step": 131500
+    },
+    {
+      "epoch": 1.0198011387779382,
+      "grad_norm": 0.0004312015080358833,
+      "learning_rate": 0.00023296441274669606,
+      "loss": 0.0004,
+      "step": 132000
+    },
+    {
+      "epoch": 1.0236640218793698,
+      "grad_norm": 0.0004256974207237363,
+      "learning_rate": 0.00023250995267555793,
+      "loss": 0.0004,
+      "step": 132500
+    },
+    {
+      "epoch": 1.0275269049808016,
+      "grad_norm": 0.0003821359423454851,
+      "learning_rate": 0.00023205549260441977,
+      "loss": 0.0004,
+      "step": 133000
+    },
+    {
+      "epoch": 1.0313897880822331,
+      "grad_norm": 0.000378616590751335,
+      "learning_rate": 0.0002316010325332816,
+      "loss": 0.0004,
+      "step": 133500
+    },
+    {
+      "epoch": 1.0352526711836647,
+      "grad_norm": 0.0003908917715307325,
+      "learning_rate": 0.00023114657246214347,
+      "loss": 0.0004,
+      "step": 134000
+    },
+    {
+      "epoch": 1.0391155542850963,
+      "grad_norm": 0.00039952111546881497,
+      "learning_rate": 0.0002306921123910053,
+      "loss": 0.0004,
+      "step": 134500
+    },
+    {
+      "epoch": 1.0429784373865278,
+      "grad_norm": 0.0004296097031328827,
+      "learning_rate": 0.00023023765231986717,
+      "loss": 0.0004,
+      "step": 135000
+    },
+    {
+      "epoch": 1.0468413204879594,
+      "grad_norm": 0.00045530457282438874,
+      "learning_rate": 0.000229783192248729,
+      "loss": 0.0004,
+      "step": 135500
+    },
+    {
+      "epoch": 1.050704203589391,
+      "grad_norm": 0.0004504481330513954,
+      "learning_rate": 0.00022932873217759085,
+      "loss": 0.0004,
+      "step": 136000
+    },
+    {
+      "epoch": 1.0545670866908226,
+      "grad_norm": 0.0003698394284583628,
+      "learning_rate": 0.00022887427210645272,
+      "loss": 0.0004,
+      "step": 136500
+    },
+    {
+      "epoch": 1.0584299697922541,
+      "grad_norm": 0.00040104315849021077,
+      "learning_rate": 0.00022841981203531456,
+      "loss": 0.0004,
+      "step": 137000
+    },
+    {
+      "epoch": 1.0622928528936857,
+      "grad_norm": 0.00039140283479355276,
+      "learning_rate": 0.00022796535196417642,
+      "loss": 0.0004,
+      "step": 137500
+    },
+    {
+      "epoch": 1.0661557359951173,
+      "grad_norm": 0.0004166768048889935,
+      "learning_rate": 0.00022751089189303826,
+      "loss": 0.0004,
+      "step": 138000
+    },
+    {
+      "epoch": 1.0700186190965488,
+      "grad_norm": 0.00044842908391728997,
+      "learning_rate": 0.0002270564318219001,
+      "loss": 0.0004,
+      "step": 138500
+    },
+    {
+      "epoch": 1.0738815021979804,
+      "grad_norm": 0.00040768241160549223,
+      "learning_rate": 0.00022660197175076196,
+      "loss": 0.0004,
+      "step": 139000
+    },
+    {
+      "epoch": 1.077744385299412,
+      "grad_norm": 0.0004266213800292462,
+      "learning_rate": 0.0002261475116796238,
+      "loss": 0.0004,
+      "step": 139500
+    },
+    {
+      "epoch": 1.0816072684008438,
+      "grad_norm": 0.0003989775141235441,
+      "learning_rate": 0.00022569305160848564,
+      "loss": 0.0004,
+      "step": 140000
+    },
+    {
+      "epoch": 1.0854701515022753,
+      "grad_norm": 0.0003877159033436328,
+      "learning_rate": 0.0002252385915373475,
+      "loss": 0.0004,
+      "step": 140500
+    },
+    {
+      "epoch": 1.089333034603707,
+      "grad_norm": 0.000394685281207785,
+      "learning_rate": 0.00022478413146620935,
+      "loss": 0.0004,
+      "step": 141000
+    },
+    {
+      "epoch": 1.0931959177051385,
+      "grad_norm": 0.00042225164361298084,
+      "learning_rate": 0.0002243296713950712,
+      "loss": 0.0004,
+      "step": 141500
+    },
+    {
+      "epoch": 1.09705880080657,
+      "grad_norm": 0.00043836236000061035,
+      "learning_rate": 0.00022387521132393305,
+      "loss": 0.0004,
+      "step": 142000
+    },
+    {
+      "epoch": 1.1009216839080016,
+      "grad_norm": 0.00039904468576423824,
+      "learning_rate": 0.0002234207512527949,
+      "loss": 0.0004,
+      "step": 142500
+    },
+    {
+      "epoch": 1.1047845670094332,
+      "grad_norm": 0.0003950648242607713,
+      "learning_rate": 0.00022296629118165675,
+      "loss": 0.0004,
+      "step": 143000
+    },
+    {
+      "epoch": 1.1086474501108647,
+      "grad_norm": 0.0003653796447906643,
+      "learning_rate": 0.0002225118311105186,
+      "loss": 0.0004,
+      "step": 143500
+    },
+    {
+      "epoch": 1.1125103332122963,
+      "grad_norm": 0.0004174882487859577,
+      "learning_rate": 0.00022205737103938046,
+      "loss": 0.0004,
+      "step": 144000
+    },
+    {
+      "epoch": 1.1163732163137279,
+      "grad_norm": 0.00038426730316132307,
+      "learning_rate": 0.0002216029109682423,
+      "loss": 0.0004,
+      "step": 144500
+    },
+    {
+      "epoch": 1.1202360994151594,
+      "grad_norm": 0.0004381178005132824,
+      "learning_rate": 0.00022114845089710414,
+      "loss": 0.0004,
+      "step": 145000
+    },
+    {
+      "epoch": 1.124098982516591,
+      "grad_norm": 0.00036438272218219936,
+      "learning_rate": 0.00022069399082596603,
+      "loss": 0.0004,
+      "step": 145500
+    },
+    {
+      "epoch": 1.1279618656180226,
+      "grad_norm": 0.00037406469346024096,
+      "learning_rate": 0.00022023953075482784,
+      "loss": 0.0004,
+      "step": 146000
+    },
+    {
+      "epoch": 1.1318247487194544,
+      "grad_norm": 0.0004003301728516817,
+      "learning_rate": 0.00021978507068368973,
+      "loss": 0.0004,
+      "step": 146500
+    },
+    {
+      "epoch": 1.1356876318208857,
+      "grad_norm": 0.00044306329800747335,
+      "learning_rate": 0.00021933061061255157,
+      "loss": 0.0004,
+      "step": 147000
+    },
+    {
+      "epoch": 1.1395505149223175,
+      "grad_norm": 0.00036800315137952566,
+      "learning_rate": 0.0002188761505414134,
+      "loss": 0.0004,
+      "step": 147500
+    },
+    {
+      "epoch": 1.143413398023749,
+      "grad_norm": 0.0003712923207785934,
+      "learning_rate": 0.00021842169047027528,
+      "loss": 0.0004,
+      "step": 148000
+    },
+    {
+      "epoch": 1.1472762811251807,
+      "grad_norm": 0.000403615616960451,
+      "learning_rate": 0.00021796723039913711,
+      "loss": 0.0004,
+      "step": 148500
+    },
+    {
+      "epoch": 1.1511391642266122,
+      "grad_norm": 0.00039503254811279476,
+      "learning_rate": 0.00021751277032799898,
+      "loss": 0.0004,
+      "step": 149000
+    },
+    {
+      "epoch": 1.1550020473280438,
+      "grad_norm": 0.00040640676161274314,
+      "learning_rate": 0.00021705831025686082,
+      "loss": 0.0004,
+      "step": 149500
+    },
+    {
+      "epoch": 1.1588649304294754,
+      "grad_norm": 0.00039924593875184655,
+      "learning_rate": 0.00021660385018572266,
+      "loss": 0.0004,
+      "step": 150000
+    },
+    {
+      "epoch": 1.162727813530907,
+      "grad_norm": 0.00041021950892172754,
+      "learning_rate": 0.00021614939011458452,
+      "loss": 0.0004,
+      "step": 150500
+    },
+    {
+      "epoch": 1.1665906966323385,
+      "grad_norm": 0.0003925120981875807,
+      "learning_rate": 0.00021569493004344636,
+      "loss": 0.0004,
+      "step": 151000
+    },
+    {
+      "epoch": 1.17045357973377,
+      "grad_norm": 0.0003923355252481997,
+      "learning_rate": 0.00021524046997230823,
+      "loss": 0.0004,
+      "step": 151500
+    },
+    {
+      "epoch": 1.1743164628352016,
+      "grad_norm": 0.00036843877751380205,
+      "learning_rate": 0.00021478600990117007,
+      "loss": 0.0004,
+      "step": 152000
+    },
+    {
+      "epoch": 1.1781793459366332,
+      "grad_norm": 0.0003888097999151796,
+      "learning_rate": 0.0002143315498300319,
+      "loss": 0.0004,
+      "step": 152500
+    },
+    {
+      "epoch": 1.1820422290380648,
+      "grad_norm": 0.00041298934957012534,
+      "learning_rate": 0.00021387708975889377,
+      "loss": 0.0004,
+      "step": 153000
+    },
+    {
+      "epoch": 1.1859051121394963,
+      "grad_norm": 0.0004346439673099667,
+      "learning_rate": 0.0002134226296877556,
+      "loss": 0.0004,
+      "step": 153500
+    },
+    {
+      "epoch": 1.1897679952409281,
+      "grad_norm": 0.0004072737356182188,
+      "learning_rate": 0.00021296816961661748,
+      "loss": 0.0004,
+      "step": 154000
+    },
+    {
+      "epoch": 1.1936308783423595,
+      "grad_norm": 0.00042254169238731265,
+      "learning_rate": 0.00021251370954547931,
+      "loss": 0.0004,
+      "step": 154500
+    },
+    {
+      "epoch": 1.1974937614437913,
+      "grad_norm": 0.00044913325109519064,
+      "learning_rate": 0.00021205924947434115,
+      "loss": 0.0004,
+      "step": 155000
+    },
+    {
+      "epoch": 1.2013566445452228,
+      "grad_norm": 0.0004437014285940677,
+      "learning_rate": 0.00021160478940320302,
+      "loss": 0.0003,
+      "step": 155500
+    },
+    {
+      "epoch": 1.2052195276466544,
+      "grad_norm": 0.00040918594459071755,
+      "learning_rate": 0.00021115032933206486,
+      "loss": 0.0003,
+      "step": 156000
+    },
+    {
+      "epoch": 1.209082410748086,
+      "grad_norm": 0.0004345249617472291,
+      "learning_rate": 0.00021069586926092672,
+      "loss": 0.0004,
+      "step": 156500
+    },
+    {
+      "epoch": 1.2129452938495175,
+      "grad_norm": 0.00039289359119720757,
+      "learning_rate": 0.00021024140918978856,
+      "loss": 0.0003,
+      "step": 157000
+    },
+    {
+      "epoch": 1.2168081769509491,
+      "grad_norm": 0.0003551707195583731,
+      "learning_rate": 0.0002097869491186504,
+      "loss": 0.0004,
+      "step": 157500
+    },
+    {
+      "epoch": 1.2206710600523807,
+      "grad_norm": 0.0004176785587333143,
+      "learning_rate": 0.00020933248904751227,
+      "loss": 0.0003,
+      "step": 158000
+    },
+    {
+      "epoch": 1.2245339431538123,
+      "grad_norm": 0.00039580900920554996,
+      "learning_rate": 0.0002088780289763741,
+      "loss": 0.0003,
+      "step": 158500
+    },
+    {
+      "epoch": 1.2283968262552438,
+      "grad_norm": 0.0004138547810725868,
+      "learning_rate": 0.00020842356890523597,
+      "loss": 0.0003,
+      "step": 159000
+    },
+    {
+      "epoch": 1.2322597093566754,
+      "grad_norm": 0.0004359263402875513,
+      "learning_rate": 0.0002079691088340978,
+      "loss": 0.0003,
+      "step": 159500
+    },
+    {
+      "epoch": 1.236122592458107,
+      "grad_norm": 0.00038180273259058595,
+      "learning_rate": 0.00020751464876295965,
+      "loss": 0.0003,
+      "step": 160000
+    },
+    {
+      "epoch": 1.2399854755595385,
+      "grad_norm": 0.0003900369047187269,
+      "learning_rate": 0.00020706018869182154,
+      "loss": 0.0003,
+      "step": 160500
+    },
+    {
+      "epoch": 1.24384835866097,
+      "grad_norm": 0.0004283432208467275,
+      "learning_rate": 0.00020660572862068335,
+      "loss": 0.0003,
+      "step": 161000
+    },
+    {
+      "epoch": 1.247711241762402,
+      "grad_norm": 0.0003838810371235013,
+      "learning_rate": 0.00020615126854954524,
+      "loss": 0.0003,
+      "step": 161500
+    },
+    {
+      "epoch": 1.2515741248638332,
+      "grad_norm": 0.00044562091352418065,
+      "learning_rate": 0.00020569680847840708,
+      "loss": 0.0003,
+      "step": 162000
+    },
+    {
+      "epoch": 1.255437007965265,
+      "grad_norm": 0.00038624508306384087,
+      "learning_rate": 0.00020524234840726892,
+      "loss": 0.0003,
+      "step": 162500
+    },
+    {
+      "epoch": 1.2592998910666966,
+      "grad_norm": 0.00038577744271606207,
+      "learning_rate": 0.0002047878883361308,
+      "loss": 0.0003,
+      "step": 163000
+    },
+    {
+      "epoch": 1.2631627741681282,
+      "grad_norm": 0.0003903675533365458,
+      "learning_rate": 0.00020433342826499263,
+      "loss": 0.0003,
+      "step": 163500
+    },
+    {
+      "epoch": 1.2670256572695597,
+      "grad_norm": 0.0003998366591986269,
+      "learning_rate": 0.0002038789681938545,
+      "loss": 0.0003,
+      "step": 164000
+    },
+    {
+      "epoch": 1.2708885403709913,
+      "grad_norm": 0.0003925804339814931,
+      "learning_rate": 0.00020342450812271633,
+      "loss": 0.0003,
+      "step": 164500
+    },
+    {
+      "epoch": 1.2747514234724229,
+      "grad_norm": 0.00036960531724616885,
+      "learning_rate": 0.00020297004805157817,
+      "loss": 0.0003,
+      "step": 165000
+    },
+    {
+      "epoch": 1.2786143065738544,
+      "grad_norm": 0.0004027937538921833,
+      "learning_rate": 0.00020251558798044003,
+      "loss": 0.0003,
+      "step": 165500
+    },
+    {
+      "epoch": 1.282477189675286,
+      "grad_norm": 0.0004089568683411926,
+      "learning_rate": 0.00020206112790930187,
+      "loss": 0.0003,
+      "step": 166000
+    },
+    {
+      "epoch": 1.2863400727767176,
+      "grad_norm": 0.0003960439353249967,
+      "learning_rate": 0.00020160666783816374,
+      "loss": 0.0003,
+      "step": 166500
+    },
+    {
+      "epoch": 1.2902029558781491,
+      "grad_norm": 0.00039318378549069166,
+      "learning_rate": 0.00020115220776702558,
+      "loss": 0.0003,
+      "step": 167000
+    },
+    {
+      "epoch": 1.2940658389795807,
+      "grad_norm": 0.0003687525459099561,
+      "learning_rate": 0.00020069774769588742,
+      "loss": 0.0003,
+      "step": 167500
+    },
+    {
+      "epoch": 1.2979287220810125,
+      "grad_norm": 0.0003783398715313524,
+      "learning_rate": 0.00020024328762474928,
+      "loss": 0.0003,
+      "step": 168000
+    },
+    {
+      "epoch": 1.3017916051824439,
+      "grad_norm": 0.0004212854546494782,
+      "learning_rate": 0.00019978882755361112,
+      "loss": 0.0003,
+      "step": 168500
+    },
+    {
+      "epoch": 1.3056544882838756,
+      "grad_norm": 0.0004270244389772415,
+      "learning_rate": 0.00019933436748247296,
+      "loss": 0.0003,
+      "step": 169000
+    },
+    {
+      "epoch": 1.3095173713853072,
+      "grad_norm": 0.00039981535519473255,
+      "learning_rate": 0.00019887990741133482,
+      "loss": 0.0003,
+      "step": 169500
+    },
+    {
+      "epoch": 1.3133802544867388,
+      "grad_norm": 0.00041499731014482677,
+      "learning_rate": 0.00019842544734019666,
+      "loss": 0.0003,
+      "step": 170000
+    },
+    {
+      "epoch": 1.3172431375881704,
+      "grad_norm": 0.0004115419287700206,
+      "learning_rate": 0.00019797098726905853,
+      "loss": 0.0003,
+      "step": 170500
+    },
+    {
+      "epoch": 1.321106020689602,
+      "grad_norm": 0.00044279222493059933,
+      "learning_rate": 0.00019751652719792037,
+      "loss": 0.0003,
+      "step": 171000
+    },
+    {
+      "epoch": 1.3249689037910335,
+      "grad_norm": 0.0004417746386025101,
+      "learning_rate": 0.0001970620671267822,
+      "loss": 0.0003,
+      "step": 171500
+    },
+    {
+      "epoch": 1.328831786892465,
+      "grad_norm": 0.0003883994068019092,
+      "learning_rate": 0.00019660760705564407,
+      "loss": 0.0003,
+      "step": 172000
+    },
+    {
+      "epoch": 1.3326946699938966,
+      "grad_norm": 0.0003858699055854231,
+      "learning_rate": 0.0001961531469845059,
+      "loss": 0.0003,
+      "step": 172500
+    },
+    {
+      "epoch": 1.3365575530953282,
+      "grad_norm": 0.0004506374825723469,
+      "learning_rate": 0.00019569868691336778,
+      "loss": 0.0003,
+      "step": 173000
+    },
+    {
+      "epoch": 1.3404204361967598,
+      "grad_norm": 0.0003691207093652338,
+      "learning_rate": 0.00019524422684222961,
+      "loss": 0.0003,
+      "step": 173500
+    },
+    {
+      "epoch": 1.3442833192981913,
+      "grad_norm": 0.00040023517794907093,
+      "learning_rate": 0.00019478976677109145,
+      "loss": 0.0003,
+      "step": 174000
+    },
+    {
+      "epoch": 1.348146202399623,
+      "grad_norm": 0.00040831902879290283,
+      "learning_rate": 0.00019433530669995332,
+      "loss": 0.0003,
+      "step": 174500
+    },
+    {
+      "epoch": 1.3520090855010545,
+      "grad_norm": 0.00039210630347952247,
+      "learning_rate": 0.00019388084662881516,
+      "loss": 0.0003,
+      "step": 175000
+    },
+    {
+      "epoch": 1.3558719686024863,
+      "grad_norm": 0.0004167640581727028,
+      "learning_rate": 0.00019342638655767705,
+      "loss": 0.0003,
+      "step": 175500
+    },
+    {
+      "epoch": 1.3597348517039176,
+      "grad_norm": 0.0004036327882204205,
+      "learning_rate": 0.00019297192648653886,
+      "loss": 0.0003,
+      "step": 176000
+    },
+    {
+      "epoch": 1.3635977348053494,
+      "grad_norm": 0.0004186241130810231,
+      "learning_rate": 0.0001925174664154007,
+      "loss": 0.0003,
+      "step": 176500
+    },
+    {
+      "epoch": 1.367460617906781,
+      "grad_norm": 0.00042677807505242527,
+      "learning_rate": 0.0001920630063442626,
+      "loss": 0.0003,
+      "step": 177000
+    },
+    {
+      "epoch": 1.3713235010082125,
+      "grad_norm": 0.0003888640203513205,
+      "learning_rate": 0.00019160854627312443,
+      "loss": 0.0003,
+      "step": 177500
+    },
+    {
+      "epoch": 1.375186384109644,
+      "grad_norm": 0.00038852522266097367,
+      "learning_rate": 0.0001911540862019863,
+      "loss": 0.0003,
+      "step": 178000
+    },
+    {
+      "epoch": 1.3790492672110757,
+      "grad_norm": 0.00048570215585641563,
+      "learning_rate": 0.00019069962613084814,
+      "loss": 0.0003,
+      "step": 178500
+    },
+    {
+      "epoch": 1.3829121503125072,
+      "grad_norm": 0.0004395216528791934,
+      "learning_rate": 0.00019024516605970997,
+      "loss": 0.0003,
+      "step": 179000
+    },
+    {
+      "epoch": 1.3867750334139388,
+      "grad_norm": 0.00039375273627229035,
+      "learning_rate": 0.00018979070598857184,
+      "loss": 0.0003,
+      "step": 179500
+    },
+    {
+      "epoch": 1.3906379165153704,
+      "grad_norm": 0.00038475391920655966,
+      "learning_rate": 0.00018933624591743368,
+      "loss": 0.0003,
+      "step": 180000
+    },
+    {
+      "epoch": 1.394500799616802,
+      "grad_norm": 0.0003974135033786297,
+      "learning_rate": 0.00018888178584629554,
+      "loss": 0.0003,
+      "step": 180500
+    },
+    {
+      "epoch": 1.3983636827182335,
+      "grad_norm": 0.00041810667607933283,
+      "learning_rate": 0.00018842732577515738,
+      "loss": 0.0003,
+      "step": 181000
+    },
+    {
+      "epoch": 1.402226565819665,
+      "grad_norm": 0.00039118548738770187,
+      "learning_rate": 0.00018797286570401922,
+      "loss": 0.0003,
+      "step": 181500
+    },
+    {
+      "epoch": 1.4060894489210969,
+      "grad_norm": 0.0004240362031850964,
+      "learning_rate": 0.0001875184056328811,
+      "loss": 0.0003,
+      "step": 182000
+    },
+    {
+      "epoch": 1.4099523320225282,
+      "grad_norm": 0.00044009560951963067,
+      "learning_rate": 0.00018706394556174293,
+      "loss": 0.0003,
+      "step": 182500
+    },
+    {
+      "epoch": 1.41381521512396,
+      "grad_norm": 0.00038786401273682714,
+      "learning_rate": 0.0001866094854906048,
+      "loss": 0.0003,
+      "step": 183000
+    },
+    {
+      "epoch": 1.4176780982253914,
+      "grad_norm": 0.0003855082031805068,
+      "learning_rate": 0.00018615502541946663,
+      "loss": 0.0003,
+      "step": 183500
+    },
+    {
+      "epoch": 1.4215409813268232,
+      "grad_norm": 0.00035129828029312193,
+      "learning_rate": 0.00018570056534832847,
+      "loss": 0.0003,
+      "step": 184000
+    },
+    {
+      "epoch": 1.4254038644282547,
+      "grad_norm": 0.0004144856648053974,
+      "learning_rate": 0.00018524610527719033,
+      "loss": 0.0003,
+      "step": 184500
+    },
+    {
+      "epoch": 1.4292667475296863,
+      "grad_norm": 0.00038916736957617104,
+      "learning_rate": 0.00018479164520605217,
+      "loss": 0.0003,
+      "step": 185000
+    },
+    {
+      "epoch": 1.4331296306311179,
+      "grad_norm": 0.00037746617454104125,
+      "learning_rate": 0.00018433718513491404,
+      "loss": 0.0003,
+      "step": 185500
+    },
+    {
+      "epoch": 1.4369925137325494,
+      "grad_norm": 0.0004193016211502254,
+      "learning_rate": 0.00018388272506377588,
+      "loss": 0.0003,
+      "step": 186000
+    },
+    {
+      "epoch": 1.440855396833981,
+      "grad_norm": 0.0003769499890040606,
+      "learning_rate": 0.00018342826499263772,
+      "loss": 0.0003,
+      "step": 186500
+    },
+    {
+      "epoch": 1.4447182799354126,
+      "grad_norm": 0.00040736031951382756,
+      "learning_rate": 0.00018297380492149958,
+      "loss": 0.0003,
+      "step": 187000
+    },
+    {
+      "epoch": 1.4485811630368441,
+      "grad_norm": 0.00042587725329212844,
+      "learning_rate": 0.00018251934485036142,
+      "loss": 0.0003,
+      "step": 187500
+    },
+    {
+      "epoch": 1.4524440461382757,
+      "grad_norm": 0.0004074377939105034,
+      "learning_rate": 0.00018206488477922329,
+      "loss": 0.0003,
+      "step": 188000
+    },
+    {
+      "epoch": 1.4563069292397073,
+      "grad_norm": 0.000427786901127547,
+      "learning_rate": 0.00018161042470808512,
+      "loss": 0.0003,
+      "step": 188500
+    },
+    {
+      "epoch": 1.4601698123411389,
+      "grad_norm": 0.0004227866302244365,
+      "learning_rate": 0.00018115596463694696,
+      "loss": 0.0003,
+      "step": 189000
+    },
+    {
+      "epoch": 1.4640326954425706,
+      "grad_norm": 0.00043102685594931245,
+      "learning_rate": 0.00018070150456580883,
+      "loss": 0.0003,
+      "step": 189500
+    },
+    {
+      "epoch": 1.467895578544002,
+      "grad_norm": 0.0003675154584925622,
+      "learning_rate": 0.00018024704449467067,
+      "loss": 0.0003,
+      "step": 190000
+    },
+    {
+      "epoch": 1.4717584616454338,
+      "grad_norm": 0.0004437029710970819,
+      "learning_rate": 0.00017979258442353256,
+      "loss": 0.0003,
+      "step": 190500
+    },
+    {
+      "epoch": 1.4756213447468653,
+      "grad_norm": 0.00044143290142528713,
+      "learning_rate": 0.0001793381243523944,
+      "loss": 0.0003,
+      "step": 191000
+    },
+    {
+      "epoch": 1.479484227848297,
+      "grad_norm": 0.00038855167804285884,
+      "learning_rate": 0.0001788836642812562,
+      "loss": 0.0003,
+      "step": 191500
+    },
+    {
+      "epoch": 1.4833471109497285,
+      "grad_norm": 0.00041832137503661215,
+      "learning_rate": 0.0001784292042101181,
+      "loss": 0.0003,
+      "step": 192000
+    },
+    {
+      "epoch": 1.48720999405116,
+      "grad_norm": 0.00042216904694214463,
+      "learning_rate": 0.00017797474413897994,
+      "loss": 0.0003,
+      "step": 192500
+    },
+    {
+      "epoch": 1.4910728771525916,
+      "grad_norm": 0.0003619820927269757,
+      "learning_rate": 0.0001775202840678418,
+      "loss": 0.0003,
+      "step": 193000
+    },
+    {
+      "epoch": 1.4949357602540232,
+      "grad_norm": 0.0004148418374825269,
+      "learning_rate": 0.00017706582399670365,
+      "loss": 0.0003,
+      "step": 193500
+    },
+    {
+      "epoch": 1.4987986433554548,
+      "grad_norm": 0.0003889719373546541,
+      "learning_rate": 0.00017661136392556548,
+      "loss": 0.0003,
+      "step": 194000
+    },
+    {
+      "epoch": 1.5026615264568863,
+      "grad_norm": 0.00035413564182817936,
+      "learning_rate": 0.00017615690385442735,
+      "loss": 0.0003,
+      "step": 194500
+    },
+    {
+      "epoch": 1.506524409558318,
+      "grad_norm": 0.00041614958900026977,
+      "learning_rate": 0.0001757024437832892,
+      "loss": 0.0003,
+      "step": 195000
+    },
+    {
+      "epoch": 1.5103872926597495,
+      "grad_norm": 0.00038349401438608766,
+      "learning_rate": 0.00017524798371215103,
+      "loss": 0.0003,
+      "step": 195500
+    },
+    {
+      "epoch": 1.5142501757611813,
+      "grad_norm": 0.0004736898990813643,
+      "learning_rate": 0.0001747935236410129,
+      "loss": 0.0003,
+      "step": 196000
+    },
+    {
+      "epoch": 1.5181130588626126,
+      "grad_norm": 0.0003606528916861862,
+      "learning_rate": 0.00017433906356987473,
+      "loss": 0.0003,
+      "step": 196500
+    },
+    {
+      "epoch": 1.5219759419640444,
+      "grad_norm": 0.00039227615343406796,
+      "learning_rate": 0.0001738846034987366,
+      "loss": 0.0003,
+      "step": 197000
+    },
+    {
+      "epoch": 1.5258388250654757,
+      "grad_norm": 0.00038570541073568165,
+      "learning_rate": 0.00017343014342759844,
+      "loss": 0.0003,
+      "step": 197500
+    },
+    {
+      "epoch": 1.5297017081669075,
+      "grad_norm": 0.0003896007256116718,
+      "learning_rate": 0.00017297568335646028,
+      "loss": 0.0003,
+      "step": 198000
+    },
+    {
+      "epoch": 1.5335645912683389,
+      "grad_norm": 0.00037037860602140427,
+      "learning_rate": 0.00017252122328532214,
+      "loss": 0.0003,
+      "step": 198500
+    },
+    {
+      "epoch": 1.5374274743697707,
+      "grad_norm": 0.0004178315575700253,
+      "learning_rate": 0.00017206676321418398,
+      "loss": 0.0003,
+      "step": 199000
+    },
+    {
+      "epoch": 1.5412903574712022,
+      "grad_norm": 0.0004494404711294919,
+      "learning_rate": 0.00017161230314304585,
+      "loss": 0.0003,
+      "step": 199500
+    },
+    {
+      "epoch": 1.5451532405726338,
+      "grad_norm": 0.00043085639481432736,
+      "learning_rate": 0.00017115784307190768,
+      "loss": 0.0003,
+      "step": 200000
+    },
+    {
+      "epoch": 1.5490161236740654,
+      "grad_norm": 0.0003994444850832224,
+      "learning_rate": 0.00017070338300076952,
+      "loss": 0.0003,
+      "step": 200500
+    },
+    {
+      "epoch": 1.552879006775497,
+      "grad_norm": 0.00038502513780258596,
+      "learning_rate": 0.0001702489229296314,
+      "loss": 0.0003,
+      "step": 201000
+    },
+    {
+      "epoch": 1.5567418898769285,
+      "grad_norm": 0.00040339346742257476,
+      "learning_rate": 0.00016979446285849323,
+      "loss": 0.0003,
+      "step": 201500
+    },
+    {
+      "epoch": 1.56060477297836,
+      "grad_norm": 0.0004164897254668176,
+      "learning_rate": 0.0001693400027873551,
+      "loss": 0.0003,
+      "step": 202000
+    },
+    {
+      "epoch": 1.5644676560797919,
+      "grad_norm": 0.0004254815576132387,
+      "learning_rate": 0.00016888554271621693,
+      "loss": 0.0003,
+      "step": 202500
+    },
+    {
+      "epoch": 1.5683305391812232,
+      "grad_norm": 0.00037603103555738926,
+      "learning_rate": 0.00016843108264507877,
+      "loss": 0.0003,
+      "step": 203000
+    },
+    {
+      "epoch": 1.572193422282655,
+      "grad_norm": 0.0004439698241185397,
+      "learning_rate": 0.00016797662257394064,
+      "loss": 0.0003,
+      "step": 203500
+    },
+    {
+      "epoch": 1.5760563053840864,
+      "grad_norm": 0.0004298971325624734,
+      "learning_rate": 0.00016752216250280247,
+      "loss": 0.0003,
+      "step": 204000
+    },
+    {
+      "epoch": 1.5799191884855182,
+      "grad_norm": 0.00040610809810459614,
+      "learning_rate": 0.00016706770243166434,
+      "loss": 0.0003,
+      "step": 204500
+    },
+    {
+      "epoch": 1.5837820715869495,
+      "grad_norm": 0.00037626715493388474,
+      "learning_rate": 0.00016661324236052618,
+      "loss": 0.0003,
+      "step": 205000
+    },
+    {
+      "epoch": 1.5876449546883813,
+      "grad_norm": 0.0004322736931499094,
+      "learning_rate": 0.00016615878228938802,
+      "loss": 0.0003,
+      "step": 205500
+    },
+    {
+      "epoch": 1.5915078377898126,
+      "grad_norm": 0.0003791552153415978,
+      "learning_rate": 0.0001657043222182499,
+      "loss": 0.0003,
+      "step": 206000
+    },
+    {
+      "epoch": 1.5953707208912444,
+      "grad_norm": 0.0004476773028727621,
+      "learning_rate": 0.00016524986214711172,
+      "loss": 0.0003,
+      "step": 206500
+    },
+    {
+      "epoch": 1.599233603992676,
+      "grad_norm": 0.0004227587196510285,
+      "learning_rate": 0.00016479540207597361,
+      "loss": 0.0003,
+      "step": 207000
+    },
+    {
+      "epoch": 1.6030964870941076,
+      "grad_norm": 0.0004329047806095332,
+      "learning_rate": 0.00016434094200483545,
+      "loss": 0.0003,
+      "step": 207500
+    },
+    {
+      "epoch": 1.6069593701955391,
+      "grad_norm": 0.0004244181327521801,
+      "learning_rate": 0.0001638864819336973,
+      "loss": 0.0003,
+      "step": 208000
+    },
+    {
+      "epoch": 1.6108222532969707,
+      "grad_norm": 0.0004290763463359326,
+      "learning_rate": 0.00016343202186255916,
+      "loss": 0.0003,
+      "step": 208500
+    },
+    {
+      "epoch": 1.6146851363984023,
+      "grad_norm": 0.00040158795309253037,
+      "learning_rate": 0.000162977561791421,
+      "loss": 0.0003,
+      "step": 209000
+    },
+    {
+      "epoch": 1.6185480194998338,
+      "grad_norm": 0.00042520073475316167,
+      "learning_rate": 0.00016252310172028286,
+      "loss": 0.0003,
+      "step": 209500
+    },
+    {
+      "epoch": 1.6224109026012656,
+      "grad_norm": 0.0004901690408587456,
+      "learning_rate": 0.0001620686416491447,
+      "loss": 0.0003,
+      "step": 210000
+    },
+    {
+      "epoch": 1.626273785702697,
+      "grad_norm": 0.00038485441473312676,
+      "learning_rate": 0.00016161418157800654,
+      "loss": 0.0003,
+      "step": 210500
+    },
+    {
+      "epoch": 1.6301366688041288,
+      "grad_norm": 0.00040810354403220117,
+      "learning_rate": 0.0001611597215068684,
+      "loss": 0.0003,
+      "step": 211000
+    },
+    {
+      "epoch": 1.6339995519055601,
+      "grad_norm": 0.00039863213896751404,
+      "learning_rate": 0.00016070526143573024,
+      "loss": 0.0003,
+      "step": 211500
+    },
+    {
+      "epoch": 1.637862435006992,
+      "grad_norm": 0.0004517484048847109,
+      "learning_rate": 0.0001602508013645921,
+      "loss": 0.0003,
+      "step": 212000
+    },
+    {
+      "epoch": 1.6417253181084233,
+      "grad_norm": 0.0003811582282651216,
+      "learning_rate": 0.00015979634129345395,
+      "loss": 0.0003,
+      "step": 212500
+    },
+    {
+      "epoch": 1.645588201209855,
+      "grad_norm": 0.000378860771888867,
+      "learning_rate": 0.00015934188122231579,
+      "loss": 0.0003,
+      "step": 213000
+    },
+    {
+      "epoch": 1.6494510843112866,
+      "grad_norm": 0.00045831358875148,
+      "learning_rate": 0.00015888742115117765,
+      "loss": 0.0003,
+      "step": 213500
+    },
+    {
+      "epoch": 1.6533139674127182,
+      "grad_norm": 0.0004245250893291086,
+      "learning_rate": 0.0001584329610800395,
+      "loss": 0.0003,
+      "step": 214000
+    },
+    {
+      "epoch": 1.6571768505141498,
+      "grad_norm": 0.0004042689106427133,
+      "learning_rate": 0.00015797850100890136,
+      "loss": 0.0003,
+      "step": 214500
+    },
+    {
+      "epoch": 1.6610397336155813,
+      "grad_norm": 0.0003839880519080907,
+      "learning_rate": 0.0001575240409377632,
+      "loss": 0.0003,
+      "step": 215000
+    },
+    {
+      "epoch": 1.664902616717013,
+      "grad_norm": 0.0004094207542948425,
+      "learning_rate": 0.00015706958086662503,
+      "loss": 0.0003,
+      "step": 215500
+    },
+    {
+      "epoch": 1.6687654998184445,
+      "grad_norm": 0.0004042787477374077,
+      "learning_rate": 0.0001566151207954869,
+      "loss": 0.0003,
+      "step": 216000
+    },
+    {
+      "epoch": 1.672628382919876,
+      "grad_norm": 0.0003925784258171916,
+      "learning_rate": 0.00015616066072434874,
+      "loss": 0.0003,
+      "step": 216500
+    },
+    {
+      "epoch": 1.6764912660213076,
+      "grad_norm": 0.00037218185025267303,
+      "learning_rate": 0.0001557062006532106,
+      "loss": 0.0003,
+      "step": 217000
+    },
+    {
+      "epoch": 1.6803541491227394,
+      "grad_norm": 0.0004271686193533242,
+      "learning_rate": 0.00015525174058207244,
+      "loss": 0.0003,
+      "step": 217500
+    },
+    {
+      "epoch": 1.6842170322241707,
+      "grad_norm": 0.0004018540494143963,
+      "learning_rate": 0.00015479728051093428,
+      "loss": 0.0003,
+      "step": 218000
+    },
+    {
+      "epoch": 1.6880799153256025,
+      "grad_norm": 0.00041026700637303293,
+      "learning_rate": 0.00015434282043979615,
+      "loss": 0.0003,
+      "step": 218500
+    },
+    {
+      "epoch": 1.6919427984270339,
+      "grad_norm": 0.00038516061613336205,
+      "learning_rate": 0.00015388836036865798,
+      "loss": 0.0003,
+      "step": 219000
+    },
+    {
+      "epoch": 1.6958056815284657,
+      "grad_norm": 0.00036750882281921804,
+      "learning_rate": 0.00015343390029751985,
+      "loss": 0.0003,
+      "step": 219500
+    },
+    {
+      "epoch": 1.699668564629897,
+      "grad_norm": 0.00046027119969949126,
+      "learning_rate": 0.0001529794402263817,
+      "loss": 0.0003,
+      "step": 220000
+    },
+    {
+      "epoch": 1.7035314477313288,
+      "grad_norm": 0.00037043061456643045,
+      "learning_rate": 0.00015252498015524353,
+      "loss": 0.0003,
+      "step": 220500
+    },
+    {
+      "epoch": 1.7073943308327604,
+      "grad_norm": 0.0004286552139092237,
+      "learning_rate": 0.00015207052008410542,
+      "loss": 0.0003,
+      "step": 221000
+    },
+    {
+      "epoch": 1.711257213934192,
+      "grad_norm": 0.0004185969883110374,
+      "learning_rate": 0.00015161606001296723,
+      "loss": 0.0003,
+      "step": 221500
+    },
+    {
+      "epoch": 1.7151200970356235,
+      "grad_norm": 0.0003913992550224066,
+      "learning_rate": 0.00015116159994182907,
+      "loss": 0.0003,
+      "step": 222000
+    },
+    {
+      "epoch": 1.718982980137055,
+      "grad_norm": 0.00038768764352425933,
+      "learning_rate": 0.00015070713987069096,
+      "loss": 0.0003,
+      "step": 222500
+    },
+    {
+      "epoch": 1.7228458632384867,
+      "grad_norm": 0.00038810214027762413,
+      "learning_rate": 0.0001502526797995528,
+      "loss": 0.0003,
+      "step": 223000
+    },
+    {
+      "epoch": 1.7267087463399182,
+      "grad_norm": 0.00042798841604962945,
+      "learning_rate": 0.00014979821972841464,
+      "loss": 0.0003,
+      "step": 223500
+    },
+    {
+      "epoch": 1.73057162944135,
+      "grad_norm": 0.00038101564859971404,
+      "learning_rate": 0.0001493437596572765,
+      "loss": 0.0003,
+      "step": 224000
+    },
+    {
+      "epoch": 1.7344345125427814,
+      "grad_norm": 0.0003932290128432214,
+      "learning_rate": 0.00014888929958613834,
+      "loss": 0.0003,
+      "step": 224500
+    },
+    {
+      "epoch": 1.7382973956442132,
+      "grad_norm": 0.00039348064456135035,
+      "learning_rate": 0.0001484348395150002,
+      "loss": 0.0003,
+      "step": 225000
+    },
+    {
+      "epoch": 1.7421602787456445,
+      "grad_norm": 0.00038930537994019687,
+      "learning_rate": 0.00014798037944386205,
+      "loss": 0.0003,
+      "step": 225500
+    },
+    {
+      "epoch": 1.7460231618470763,
+      "grad_norm": 0.00040366893517784774,
+      "learning_rate": 0.0001475259193727239,
+      "loss": 0.0003,
+      "step": 226000
+    },
+    {
+      "epoch": 1.7498860449485076,
+      "grad_norm": 0.0004505430406425148,
+      "learning_rate": 0.00014707145930158575,
+      "loss": 0.0003,
+      "step": 226500
+    },
+    {
+      "epoch": 1.7537489280499394,
+      "grad_norm": 0.0003945432254113257,
+      "learning_rate": 0.0001466169992304476,
+      "loss": 0.0003,
+      "step": 227000
+    },
+    {
+      "epoch": 1.7576118111513708,
+      "grad_norm": 0.0004522661038208753,
+      "learning_rate": 0.00014616253915930946,
+      "loss": 0.0003,
+      "step": 227500
+    },
+    {
+      "epoch": 1.7614746942528026,
+      "grad_norm": 0.0003905027697328478,
+      "learning_rate": 0.0001457080790881713,
+      "loss": 0.0003,
+      "step": 228000
+    },
+    {
+      "epoch": 1.7653375773542341,
+      "grad_norm": 0.00039171066600829363,
+      "learning_rate": 0.00014525361901703313,
+      "loss": 0.0003,
+      "step": 228500
+    },
+    {
+      "epoch": 1.7692004604556657,
+      "grad_norm": 0.00041425134986639023,
+      "learning_rate": 0.000144799158945895,
+      "loss": 0.0003,
+      "step": 229000
+    },
+    {
+      "epoch": 1.7730633435570973,
+      "grad_norm": 0.00039298326009884477,
+      "learning_rate": 0.00014434469887475687,
+      "loss": 0.0003,
+      "step": 229500
+    },
+    {
+      "epoch": 1.7769262266585288,
+      "grad_norm": 0.00038489836151711643,
+      "learning_rate": 0.0001438902388036187,
+      "loss": 0.0003,
+      "step": 230000
+    },
+    {
+      "epoch": 1.7807891097599604,
+      "grad_norm": 0.0004210007027722895,
+      "learning_rate": 0.00014343577873248054,
+      "loss": 0.0003,
+      "step": 230500
+    },
+    {
+      "epoch": 1.784651992861392,
+      "grad_norm": 0.00042244375799782574,
+      "learning_rate": 0.0001429813186613424,
+      "loss": 0.0003,
+      "step": 231000
+    },
+    {
+      "epoch": 1.7885148759628238,
+      "grad_norm": 0.00044437378528527915,
+      "learning_rate": 0.00014252685859020425,
+      "loss": 0.0003,
+      "step": 231500
+    },
+    {
+      "epoch": 1.7923777590642551,
+      "grad_norm": 0.0003892597451340407,
+      "learning_rate": 0.0001420723985190661,
+      "loss": 0.0003,
+      "step": 232000
+    },
+    {
+      "epoch": 1.796240642165687,
+      "grad_norm": 0.00042284041410312057,
+      "learning_rate": 0.00014161793844792795,
+      "loss": 0.0003,
+      "step": 232500
+    },
+    {
+      "epoch": 1.8001035252671183,
+      "grad_norm": 0.00037430404336191714,
+      "learning_rate": 0.0001411634783767898,
+      "loss": 0.0003,
+      "step": 233000
+    },
+    {
+      "epoch": 1.80396640836855,
+      "grad_norm": 0.0004308871866669506,
+      "learning_rate": 0.00014070901830565166,
+      "loss": 0.0003,
+      "step": 233500
+    },
+    {
+      "epoch": 1.8078292914699814,
+      "grad_norm": 0.0004083754902239889,
+      "learning_rate": 0.0001402545582345135,
+      "loss": 0.0003,
+      "step": 234000
+    },
+    {
+      "epoch": 1.8116921745714132,
+      "grad_norm": 0.00041078985668718815,
+      "learning_rate": 0.00013980009816337536,
+      "loss": 0.0003,
+      "step": 234500
+    },
+    {
+      "epoch": 1.8155550576728448,
+      "grad_norm": 0.0004413351707626134,
+      "learning_rate": 0.0001393456380922372,
+      "loss": 0.0003,
+      "step": 235000
+    },
+    {
+      "epoch": 1.8194179407742763,
+      "grad_norm": 0.00039201107574626803,
+      "learning_rate": 0.00013889117802109904,
+      "loss": 0.0003,
+      "step": 235500
+    },
+    {
+      "epoch": 1.823280823875708,
+      "grad_norm": 0.00039923322037793696,
+      "learning_rate": 0.0001384367179499609,
+      "loss": 0.0003,
+      "step": 236000
+    },
+    {
+      "epoch": 1.8271437069771395,
+      "grad_norm": 0.0004075372125953436,
+      "learning_rate": 0.00013798225787882274,
+      "loss": 0.0003,
+      "step": 236500
+    },
+    {
+      "epoch": 1.831006590078571,
+      "grad_norm": 0.00040116155287250876,
+      "learning_rate": 0.0001375277978076846,
+      "loss": 0.0003,
+      "step": 237000
+    },
+    {
+      "epoch": 1.8348694731800026,
+      "grad_norm": 0.00041852614958770573,
+      "learning_rate": 0.00013707333773654647,
+      "loss": 0.0003,
+      "step": 237500
+    },
+    {
+      "epoch": 1.8387323562814342,
+      "grad_norm": 0.0004257333348505199,
+      "learning_rate": 0.0001366188776654083,
+      "loss": 0.0003,
+      "step": 238000
+    },
+    {
+      "epoch": 1.8425952393828657,
+      "grad_norm": 0.0003763121203519404,
+      "learning_rate": 0.00013616441759427015,
+      "loss": 0.0003,
+      "step": 238500
+    },
+    {
+      "epoch": 1.8464581224842975,
+      "grad_norm": 0.00041360125760547817,
+      "learning_rate": 0.00013570995752313202,
+      "loss": 0.0003,
+      "step": 239000
+    },
+    {
+      "epoch": 1.8503210055857289,
+      "grad_norm": 0.00042873475467786193,
+      "learning_rate": 0.00013525549745199385,
+      "loss": 0.0003,
+      "step": 239500
+    },
+    {
+      "epoch": 1.8541838886871607,
+      "grad_norm": 0.00039483854197897017,
+      "learning_rate": 0.00013480103738085572,
+      "loss": 0.0003,
+      "step": 240000
+    },
+    {
+      "epoch": 1.858046771788592,
+      "grad_norm": 0.00039821091922931373,
+      "learning_rate": 0.00013434657730971756,
+      "loss": 0.0003,
+      "step": 240500
+    },
+    {
+      "epoch": 1.8619096548900238,
+      "grad_norm": 0.0003865666803903878,
+      "learning_rate": 0.0001338921172385794,
+      "loss": 0.0003,
+      "step": 241000
+    },
+    {
+      "epoch": 1.8657725379914551,
+      "grad_norm": 0.00039214183925651014,
+      "learning_rate": 0.00013343765716744126,
+      "loss": 0.0003,
+      "step": 241500
+    },
+    {
+      "epoch": 1.869635421092887,
+      "grad_norm": 0.0004279193817637861,
+      "learning_rate": 0.0001329831970963031,
+      "loss": 0.0003,
+      "step": 242000
+    },
+    {
+      "epoch": 1.8734983041943185,
+      "grad_norm": 0.00042890734039247036,
+      "learning_rate": 0.00013252873702516494,
+      "loss": 0.0003,
+      "step": 242500
+    },
+    {
+      "epoch": 1.87736118729575,
+      "grad_norm": 0.00040093736606650054,
+      "learning_rate": 0.0001320742769540268,
+      "loss": 0.0003,
+      "step": 243000
+    },
+    {
+      "epoch": 1.8812240703971816,
+      "grad_norm": 0.00038263245369307697,
+      "learning_rate": 0.00013161981688288864,
+      "loss": 0.0003,
+      "step": 243500
+    },
+    {
+      "epoch": 1.8850869534986132,
+      "grad_norm": 0.00043640268268063664,
+      "learning_rate": 0.0001311653568117505,
+      "loss": 0.0003,
+      "step": 244000
+    },
+    {
+      "epoch": 1.8889498366000448,
+      "grad_norm": 0.0004298704443499446,
+      "learning_rate": 0.00013071089674061238,
+      "loss": 0.0003,
+      "step": 244500
+    },
+    {
+      "epoch": 1.8928127197014764,
+      "grad_norm": 0.0004053697339259088,
+      "learning_rate": 0.0001302564366694742,
+      "loss": 0.0003,
+      "step": 245000
+    },
+    {
+      "epoch": 1.8966756028029081,
+      "grad_norm": 0.0004045435052830726,
+      "learning_rate": 0.00012980197659833605,
+      "loss": 0.0003,
+      "step": 245500
+    },
+    {
+      "epoch": 1.9005384859043395,
+      "grad_norm": 0.00037249817978590727,
+      "learning_rate": 0.00012934751652719792,
+      "loss": 0.0003,
+      "step": 246000
+    },
+    {
+      "epoch": 1.9044013690057713,
+      "grad_norm": 0.0004331369127612561,
+      "learning_rate": 0.00012889305645605976,
+      "loss": 0.0003,
+      "step": 246500
+    },
+    {
+      "epoch": 1.9082642521072026,
+      "grad_norm": 0.00039425501017831266,
+      "learning_rate": 0.00012843859638492162,
+      "loss": 0.0003,
+      "step": 247000
+    },
+    {
+      "epoch": 1.9121271352086344,
+      "grad_norm": 0.00040850628283806145,
+      "learning_rate": 0.00012798413631378346,
+      "loss": 0.0003,
+      "step": 247500
+    },
+    {
+      "epoch": 1.9159900183100658,
+      "grad_norm": 0.0004171692708041519,
+      "learning_rate": 0.0001275296762426453,
+      "loss": 0.0003,
+      "step": 248000
+    },
+    {
+      "epoch": 1.9198529014114976,
+      "grad_norm": 0.00042486467282287776,
+      "learning_rate": 0.00012707521617150717,
+      "loss": 0.0003,
+      "step": 248500
+    },
+    {
+      "epoch": 1.923715784512929,
+      "grad_norm": 0.0003857310221064836,
+      "learning_rate": 0.000126620756100369,
+      "loss": 0.0003,
+      "step": 249000
+    },
+    {
+      "epoch": 1.9275786676143607,
+      "grad_norm": 0.00043836142867803574,
+      "learning_rate": 0.00012616629602923087,
+      "loss": 0.0003,
+      "step": 249500
+    },
+    {
+      "epoch": 1.9314415507157923,
+      "grad_norm": 0.00039056732202880085,
+      "learning_rate": 0.0001257118359580927,
+      "loss": 0.0003,
+      "step": 250000
+    },
+    {
+      "epoch": 1.9353044338172238,
+      "grad_norm": 0.0004863255890086293,
+      "learning_rate": 0.00012525737588695455,
+      "loss": 0.0003,
+      "step": 250500
+    },
+    {
+      "epoch": 1.9391673169186554,
+      "grad_norm": 0.00042994500836357474,
+      "learning_rate": 0.00012480291581581641,
+      "loss": 0.0003,
+      "step": 251000
+    },
+    {
+      "epoch": 1.943030200020087,
+      "grad_norm": 0.0004507972626015544,
+      "learning_rate": 0.00012434845574467828,
+      "loss": 0.0003,
+      "step": 251500
+    },
+    {
+      "epoch": 1.9468930831215185,
+      "grad_norm": 0.000748885446228087,
+      "learning_rate": 0.00012389399567354012,
+      "loss": 0.0003,
+      "step": 252000
+    },
+    {
+      "epoch": 1.95075596622295,
+      "grad_norm": 0.0003968368982896209,
+      "learning_rate": 0.00012343953560240196,
+      "loss": 0.0003,
+      "step": 252500
+    },
+    {
+      "epoch": 1.954618849324382,
+      "grad_norm": 0.0004240422276780009,
+      "learning_rate": 0.00012298507553126382,
+      "loss": 0.0003,
+      "step": 253000
+    },
+    {
+      "epoch": 1.9584817324258132,
+      "grad_norm": 0.00044167175656184554,
+      "learning_rate": 0.00012253061546012566,
+      "loss": 0.0003,
+      "step": 253500
+    },
+    {
+      "epoch": 1.962344615527245,
+      "grad_norm": 0.00036484008887782693,
+      "learning_rate": 0.00012207615538898753,
+      "loss": 0.0003,
+      "step": 254000
+    },
+    {
+      "epoch": 1.9662074986286764,
+      "grad_norm": 0.0004044172237627208,
+      "learning_rate": 0.00012162169531784937,
+      "loss": 0.0003,
+      "step": 254500
+    },
+    {
+      "epoch": 1.9700703817301082,
+      "grad_norm": 0.0004093232564628124,
+      "learning_rate": 0.0001211672352467112,
+      "loss": 0.0003,
+      "step": 255000
+    },
+    {
+      "epoch": 1.9739332648315395,
+      "grad_norm": 0.0003724129928741604,
+      "learning_rate": 0.00012071277517557306,
+      "loss": 0.0003,
+      "step": 255500
+    },
+    {
+      "epoch": 1.9777961479329713,
+      "grad_norm": 0.0003869113279506564,
+      "learning_rate": 0.00012025831510443492,
+      "loss": 0.0003,
+      "step": 256000
+    },
+    {
+      "epoch": 1.9816590310344029,
+      "grad_norm": 0.00036866377922706306,
+      "learning_rate": 0.00011980385503329677,
+      "loss": 0.0003,
+      "step": 256500
+    },
+    {
+      "epoch": 1.9855219141358345,
+      "grad_norm": 0.00042447738815099,
+      "learning_rate": 0.00011934939496215861,
+      "loss": 0.0003,
+      "step": 257000
+    },
+    {
+      "epoch": 1.989384797237266,
+      "grad_norm": 0.0003843741142190993,
+      "learning_rate": 0.00011889493489102046,
+      "loss": 0.0003,
+      "step": 257500
+    },
+    {
+      "epoch": 1.9932476803386976,
+      "grad_norm": 0.0004299108113627881,
+      "learning_rate": 0.00011844047481988232,
+      "loss": 0.0003,
+      "step": 258000
+    },
+    {
+      "epoch": 1.9971105634401292,
+      "grad_norm": 0.00039060041308403015,
+      "learning_rate": 0.00011798601474874417,
+      "loss": 0.0003,
+      "step": 258500
+    },
+    {
+      "epoch": 2.0009734465415607,
+      "grad_norm": 0.0004127940337639302,
+      "learning_rate": 0.00011753155467760602,
+      "loss": 0.0003,
+      "step": 259000
+    },
+    {
+      "epoch": 2.0048363296429925,
+      "grad_norm": 0.0003710352466441691,
+      "learning_rate": 0.00011707709460646786,
+      "loss": 0.0003,
+      "step": 259500
+    },
+    {
+      "epoch": 2.008699212744424,
+      "grad_norm": 0.00043169083073735237,
+      "learning_rate": 0.00011662263453532971,
+      "loss": 0.0003,
+      "step": 260000
+    },
+    {
+      "epoch": 2.0125620958458557,
+      "grad_norm": 0.00042857075459323823,
+      "learning_rate": 0.00011616817446419156,
+      "loss": 0.0003,
+      "step": 260500
+    },
+    {
+      "epoch": 2.016424978947287,
+      "grad_norm": 0.00046496940194629133,
+      "learning_rate": 0.00011571371439305342,
+      "loss": 0.0003,
+      "step": 261000
+    },
+    {
+      "epoch": 2.020287862048719,
+      "grad_norm": 0.00039056496461853385,
+      "learning_rate": 0.00011525925432191527,
+      "loss": 0.0003,
+      "step": 261500
+    },
+    {
+      "epoch": 2.02415074515015,
+      "grad_norm": 0.000406107195885852,
+      "learning_rate": 0.00011480479425077711,
+      "loss": 0.0003,
+      "step": 262000
+    },
+    {
+      "epoch": 2.028013628251582,
+      "grad_norm": 0.00039516916149295866,
+      "learning_rate": 0.00011435033417963896,
+      "loss": 0.0003,
+      "step": 262500
+    },
+    {
+      "epoch": 2.0318765113530133,
+      "grad_norm": 0.0003748384478967637,
+      "learning_rate": 0.00011389587410850081,
+      "loss": 0.0003,
+      "step": 263000
+    },
+    {
+      "epoch": 2.035739394454445,
+      "grad_norm": 0.0003946254728361964,
+      "learning_rate": 0.00011344141403736268,
+      "loss": 0.0003,
+      "step": 263500
+    },
+    {
+      "epoch": 2.0396022775558764,
+      "grad_norm": 0.00039266227395273745,
+      "learning_rate": 0.00011298695396622453,
+      "loss": 0.0003,
+      "step": 264000
+    },
+    {
+      "epoch": 2.043465160657308,
+      "grad_norm": 0.0003905179037246853,
+      "learning_rate": 0.00011253249389508637,
+      "loss": 0.0003,
+      "step": 264500
+    },
+    {
+      "epoch": 2.0473280437587396,
+      "grad_norm": 0.00040395421092398465,
+      "learning_rate": 0.00011207803382394822,
+      "loss": 0.0003,
+      "step": 265000
+    },
+    {
+      "epoch": 2.0511909268601713,
+      "grad_norm": 0.00040508355596102774,
+      "learning_rate": 0.00011162357375281007,
+      "loss": 0.0003,
+      "step": 265500
+    },
+    {
+      "epoch": 2.055053809961603,
+      "grad_norm": 0.0004007483657915145,
+      "learning_rate": 0.00011116911368167192,
+      "loss": 0.0003,
+      "step": 266000
+    },
+    {
+      "epoch": 2.0589166930630345,
+      "grad_norm": 0.00040279951645061374,
+      "learning_rate": 0.00011071465361053378,
+      "loss": 0.0003,
+      "step": 266500
+    },
+    {
+      "epoch": 2.0627795761644663,
+      "grad_norm": 0.0004587202565744519,
+      "learning_rate": 0.00011026019353939561,
+      "loss": 0.0003,
+      "step": 267000
+    },
+    {
+      "epoch": 2.0666424592658976,
+      "grad_norm": 0.0003896713024005294,
+      "learning_rate": 0.00010980573346825747,
+      "loss": 0.0003,
+      "step": 267500
+    },
+    {
+      "epoch": 2.0705053423673294,
+      "grad_norm": 0.00039689254481345415,
+      "learning_rate": 0.00010935127339711932,
+      "loss": 0.0003,
+      "step": 268000
+    },
+    {
+      "epoch": 2.0743682254687608,
+      "grad_norm": 0.00041530997259542346,
+      "learning_rate": 0.00010889681332598117,
+      "loss": 0.0003,
+      "step": 268500
+    },
+    {
+      "epoch": 2.0782311085701926,
+      "grad_norm": 0.00047515874030068517,
+      "learning_rate": 0.00010844235325484301,
+      "loss": 0.0003,
+      "step": 269000
+    },
+    {
+      "epoch": 2.082093991671624,
+      "grad_norm": 0.0003935998538509011,
+      "learning_rate": 0.00010798789318370486,
+      "loss": 0.0003,
+      "step": 269500
+    },
+    {
+      "epoch": 2.0859568747730557,
+      "grad_norm": 0.00048738697660155594,
+      "learning_rate": 0.00010753343311256671,
+      "loss": 0.0003,
+      "step": 270000
+    },
+    {
+      "epoch": 2.089819757874487,
+      "grad_norm": 0.00038931367453187704,
+      "learning_rate": 0.00010707897304142857,
+      "loss": 0.0003,
+      "step": 270500
+    },
+    {
+      "epoch": 2.093682640975919,
+      "grad_norm": 0.0004456916358321905,
+      "learning_rate": 0.00010662451297029043,
+      "loss": 0.0003,
+      "step": 271000
+    },
+    {
+      "epoch": 2.09754552407735,
+      "grad_norm": 0.00042469296022318304,
+      "learning_rate": 0.00010617005289915226,
+      "loss": 0.0003,
+      "step": 271500
+    },
+    {
+      "epoch": 2.101408407178782,
+      "grad_norm": 0.0004127691499888897,
+      "learning_rate": 0.00010571559282801412,
+      "loss": 0.0003,
+      "step": 272000
+    },
+    {
+      "epoch": 2.1052712902802133,
+      "grad_norm": 0.0004185699508525431,
+      "learning_rate": 0.00010526113275687598,
+      "loss": 0.0003,
+      "step": 272500
+    },
+    {
+      "epoch": 2.109134173381645,
+      "grad_norm": 0.0004032867436762899,
+      "learning_rate": 0.00010480667268573783,
+      "loss": 0.0003,
+      "step": 273000
+    },
+    {
+      "epoch": 2.112997056483077,
+      "grad_norm": 0.00039662199560552835,
+      "learning_rate": 0.00010435221261459968,
+      "loss": 0.0003,
+      "step": 273500
+    },
+    {
+      "epoch": 2.1168599395845082,
+      "grad_norm": 0.0004110955342184752,
+      "learning_rate": 0.00010389775254346152,
+      "loss": 0.0003,
+      "step": 274000
+    },
+    {
+      "epoch": 2.12072282268594,
+      "grad_norm": 0.0004184690478723496,
+      "learning_rate": 0.00010344329247232337,
+      "loss": 0.0003,
+      "step": 274500
+    },
+    {
+      "epoch": 2.1245857057873714,
+      "grad_norm": 0.0004132980538997799,
+      "learning_rate": 0.00010298883240118522,
+      "loss": 0.0003,
+      "step": 275000
+    },
+    {
+      "epoch": 2.128448588888803,
+      "grad_norm": 0.0004049827402923256,
+      "learning_rate": 0.00010253437233004707,
+      "loss": 0.0003,
+      "step": 275500
+    },
+    {
+      "epoch": 2.1323114719902345,
+      "grad_norm": 0.00039436816587112844,
+      "learning_rate": 0.00010207991225890893,
+      "loss": 0.0003,
+      "step": 276000
+    },
+    {
+      "epoch": 2.1361743550916663,
+      "grad_norm": 0.00044596754014492035,
+      "learning_rate": 0.00010162545218777077,
+      "loss": 0.0003,
+      "step": 276500
+    },
+    {
+      "epoch": 2.1400372381930977,
+      "grad_norm": 0.0004116700729355216,
+      "learning_rate": 0.00010117099211663262,
+      "loss": 0.0003,
+      "step": 277000
+    },
+    {
+      "epoch": 2.1439001212945294,
+      "grad_norm": 0.00036984015605412424,
+      "learning_rate": 0.00010071653204549447,
+      "loss": 0.0003,
+      "step": 277500
+    },
+    {
+      "epoch": 2.147763004395961,
+      "grad_norm": 0.0004131913010496646,
+      "learning_rate": 0.00010026207197435632,
+      "loss": 0.0003,
+      "step": 278000
+    },
+    {
+      "epoch": 2.1516258874973926,
+      "grad_norm": 0.00039576427661813796,
+      "learning_rate": 9.980761190321819e-05,
+      "loss": 0.0003,
+      "step": 278500
+    },
+    {
+      "epoch": 2.155488770598824,
+      "grad_norm": 0.0003931752871721983,
+      "learning_rate": 9.935315183208001e-05,
+      "loss": 0.0003,
+      "step": 279000
+    },
+    {
+      "epoch": 2.1593516537002557,
+      "grad_norm": 0.0004005175724159926,
+      "learning_rate": 9.889869176094188e-05,
+      "loss": 0.0003,
+      "step": 279500
+    },
+    {
+      "epoch": 2.1632145368016875,
+      "grad_norm": 0.00038955482887104154,
+      "learning_rate": 9.844423168980373e-05,
+      "loss": 0.0003,
+      "step": 280000
+    },
+    {
+      "epoch": 2.167077419903119,
+      "grad_norm": 0.0004296032420825213,
+      "learning_rate": 9.798977161866558e-05,
+      "loss": 0.0003,
+      "step": 280500
+    },
+    {
+      "epoch": 2.1709403030045507,
+      "grad_norm": 0.0003988723619841039,
+      "learning_rate": 9.753531154752743e-05,
+      "loss": 0.0003,
+      "step": 281000
+    },
+    {
+      "epoch": 2.174803186105982,
+      "grad_norm": 0.0004284825408831239,
+      "learning_rate": 9.708085147638927e-05,
+      "loss": 0.0003,
+      "step": 281500
+    },
+    {
+      "epoch": 2.178666069207414,
+      "grad_norm": 0.00043784480658359826,
+      "learning_rate": 9.662639140525113e-05,
+      "loss": 0.0003,
+      "step": 282000
+    },
+    {
+      "epoch": 2.182528952308845,
+      "grad_norm": 0.0004204624274279922,
+      "learning_rate": 9.617193133411298e-05,
+      "loss": 0.0003,
+      "step": 282500
+    },
+    {
+      "epoch": 2.186391835410277,
+      "grad_norm": 0.00045462080743163824,
+      "learning_rate": 9.571747126297483e-05,
+      "loss": 0.0003,
+      "step": 283000
+    },
+    {
+      "epoch": 2.1902547185117083,
+      "grad_norm": 0.000426099868491292,
+      "learning_rate": 9.526301119183667e-05,
+      "loss": 0.0003,
+      "step": 283500
+    },
+    {
+      "epoch": 2.19411760161314,
+      "grad_norm": 0.0004363481712061912,
+      "learning_rate": 9.480855112069852e-05,
+      "loss": 0.0003,
+      "step": 284000
+    },
+    {
+      "epoch": 2.1979804847145714,
+      "grad_norm": 0.00041816759039647877,
+      "learning_rate": 9.435409104956037e-05,
+      "loss": 0.0003,
+      "step": 284500
+    },
+    {
+      "epoch": 2.201843367816003,
+      "grad_norm": 0.00039576852577738464,
+      "learning_rate": 9.389963097842222e-05,
+      "loss": 0.0003,
+      "step": 285000
+    },
+    {
+      "epoch": 2.2057062509174346,
+      "grad_norm": 0.0003871625813189894,
+      "learning_rate": 9.344517090728409e-05,
+      "loss": 0.0003,
+      "step": 285500
+    },
+    {
+      "epoch": 2.2095691340188663,
+      "grad_norm": 0.0004058276826981455,
+      "learning_rate": 9.299071083614592e-05,
+      "loss": 0.0003,
+      "step": 286000
+    },
+    {
+      "epoch": 2.213432017120298,
+      "grad_norm": 0.0003888069768436253,
+      "learning_rate": 9.253625076500778e-05,
+      "loss": 0.0003,
+      "step": 286500
+    },
+    {
+      "epoch": 2.2172949002217295,
+      "grad_norm": 0.0003826101601589471,
+      "learning_rate": 9.208179069386963e-05,
+      "loss": 0.0003,
+      "step": 287000
+    },
+    {
+      "epoch": 2.2211577833231613,
+      "grad_norm": 0.0004133163602091372,
+      "learning_rate": 9.162733062273149e-05,
+      "loss": 0.0003,
+      "step": 287500
+    },
+    {
+      "epoch": 2.2250206664245926,
+      "grad_norm": 0.0003799792903009802,
+      "learning_rate": 9.117287055159334e-05,
+      "loss": 0.0003,
+      "step": 288000
+    },
+    {
+      "epoch": 2.2288835495260244,
+      "grad_norm": 0.0004137216310482472,
+      "learning_rate": 9.071841048045518e-05,
+      "loss": 0.0003,
+      "step": 288500
+    },
+    {
+      "epoch": 2.2327464326274558,
+      "grad_norm": 0.0004602668632287532,
+      "learning_rate": 9.026395040931703e-05,
+      "loss": 0.0003,
+      "step": 289000
+    },
+    {
+      "epoch": 2.2366093157288875,
+      "grad_norm": 0.0004129618755541742,
+      "learning_rate": 8.980949033817888e-05,
+      "loss": 0.0003,
+      "step": 289500
+    },
+    {
+      "epoch": 2.240472198830319,
+      "grad_norm": 0.0004122887330595404,
+      "learning_rate": 8.935503026704073e-05,
+      "loss": 0.0003,
+      "step": 290000
+    },
+    {
+      "epoch": 2.2443350819317507,
+      "grad_norm": 0.00042714428855106235,
+      "learning_rate": 8.890057019590258e-05,
+      "loss": 0.0003,
+      "step": 290500
+    },
+    {
+      "epoch": 2.248197965033182,
+      "grad_norm": 0.0004340863088145852,
+      "learning_rate": 8.844611012476442e-05,
+      "loss": 0.0003,
+      "step": 291000
+    },
+    {
+      "epoch": 2.252060848134614,
+      "grad_norm": 0.00038818069151602685,
+      "learning_rate": 8.799165005362628e-05,
+      "loss": 0.0003,
+      "step": 291500
+    },
+    {
+      "epoch": 2.255923731236045,
+      "grad_norm": 0.00039604620542377234,
+      "learning_rate": 8.753718998248813e-05,
+      "loss": 0.0003,
+      "step": 292000
+    },
+    {
+      "epoch": 2.259786614337477,
+      "grad_norm": 0.00037946528755128384,
+      "learning_rate": 8.708272991134998e-05,
+      "loss": 0.0003,
+      "step": 292500
+    },
+    {
+      "epoch": 2.2636494974389088,
+      "grad_norm": 0.00045615085400640965,
+      "learning_rate": 8.662826984021185e-05,
+      "loss": 0.0003,
+      "step": 293000
+    },
+    {
+      "epoch": 2.26751238054034,
+      "grad_norm": 0.0004230223421473056,
+      "learning_rate": 8.617380976907367e-05,
+      "loss": 0.0003,
+      "step": 293500
+    },
+    {
+      "epoch": 2.2713752636417714,
+      "grad_norm": 0.0004022732318844646,
+      "learning_rate": 8.571934969793554e-05,
+      "loss": 0.0003,
+      "step": 294000
+    },
+    {
+      "epoch": 2.2752381467432032,
+      "grad_norm": 0.0004257276304997504,
+      "learning_rate": 8.526488962679739e-05,
+      "loss": 0.0003,
+      "step": 294500
+    },
+    {
+      "epoch": 2.279101029844635,
+      "grad_norm": 0.00045862089609727263,
+      "learning_rate": 8.481042955565924e-05,
+      "loss": 0.0003,
+      "step": 295000
+    },
+    {
+      "epoch": 2.2829639129460664,
+      "grad_norm": 0.0004398869932629168,
+      "learning_rate": 8.435596948452108e-05,
+      "loss": 0.0003,
+      "step": 295500
+    },
+    {
+      "epoch": 2.286826796047498,
+      "grad_norm": 0.00044305986375547945,
+      "learning_rate": 8.390150941338293e-05,
+      "loss": 0.0003,
+      "step": 296000
+    },
+    {
+      "epoch": 2.2906896791489295,
+      "grad_norm": 0.0004642159619834274,
+      "learning_rate": 8.344704934224478e-05,
+      "loss": 0.0003,
+      "step": 296500
+    },
+    {
+      "epoch": 2.2945525622503613,
+      "grad_norm": 0.00043294980423524976,
+      "learning_rate": 8.299258927110664e-05,
+      "loss": 0.0003,
+      "step": 297000
+    },
+    {
+      "epoch": 2.2984154453517927,
+      "grad_norm": 0.0004789868835359812,
+      "learning_rate": 8.253812919996849e-05,
+      "loss": 0.0003,
+      "step": 297500
+    },
+    {
+      "epoch": 2.3022783284532244,
+      "grad_norm": 0.00042497721733525395,
+      "learning_rate": 8.208366912883033e-05,
+      "loss": 0.0003,
+      "step": 298000
+    },
+    {
+      "epoch": 2.306141211554656,
+      "grad_norm": 0.00044182001147419214,
+      "learning_rate": 8.162920905769218e-05,
+      "loss": 0.0003,
+      "step": 298500
+    },
+    {
+      "epoch": 2.3100040946560876,
+      "grad_norm": 0.00042325007962062955,
+      "learning_rate": 8.117474898655403e-05,
+      "loss": 0.0003,
+      "step": 299000
+    },
+    {
+      "epoch": 2.313866977757519,
+      "grad_norm": 0.0004093611496500671,
+      "learning_rate": 8.072028891541588e-05,
+      "loss": 0.0003,
+      "step": 299500
+    },
+    {
+      "epoch": 2.3177298608589507,
+      "grad_norm": 0.00038856887840665877,
+      "learning_rate": 8.026582884427774e-05,
+      "loss": 0.0003,
+      "step": 300000
+    },
+    {
+      "epoch": 2.321592743960382,
+      "grad_norm": 0.00041343894554302096,
+      "learning_rate": 7.981136877313957e-05,
+      "loss": 0.0003,
+      "step": 300500
+    },
+    {
+      "epoch": 2.325455627061814,
+      "grad_norm": 0.0004005722003057599,
+      "learning_rate": 7.935690870200143e-05,
+      "loss": 0.0003,
+      "step": 301000
+    },
+    {
+      "epoch": 2.3293185101632456,
+      "grad_norm": 0.00039640217437408864,
+      "learning_rate": 7.890244863086329e-05,
+      "loss": 0.0003,
+      "step": 301500
+    },
+    {
+      "epoch": 2.333181393264677,
+      "grad_norm": 0.00041457550833001733,
+      "learning_rate": 7.844798855972514e-05,
+      "loss": 0.0003,
+      "step": 302000
+    },
+    {
+      "epoch": 2.337044276366109,
+      "grad_norm": 0.0004301823501009494,
+      "learning_rate": 7.7993528488587e-05,
+      "loss": 0.0003,
+      "step": 302500
+    },
+    {
+      "epoch": 2.34090715946754,
+      "grad_norm": 0.0004285312897991389,
+      "learning_rate": 7.753906841744883e-05,
+      "loss": 0.0003,
+      "step": 303000
+    },
+    {
+      "epoch": 2.344770042568972,
+      "grad_norm": 0.0004082127125002444,
+      "learning_rate": 7.708460834631069e-05,
+      "loss": 0.0003,
+      "step": 303500
+    },
+    {
+      "epoch": 2.3486329256704033,
+      "grad_norm": 0.0004119858785998076,
+      "learning_rate": 7.663014827517254e-05,
+      "loss": 0.0003,
+      "step": 304000
+    },
+    {
+      "epoch": 2.352495808771835,
+      "grad_norm": 0.0004263624723535031,
+      "learning_rate": 7.617568820403439e-05,
+      "loss": 0.0003,
+      "step": 304500
+    },
+    {
+      "epoch": 2.3563586918732664,
+      "grad_norm": 0.00039540533907711506,
+      "learning_rate": 7.572122813289624e-05,
+      "loss": 0.0003,
+      "step": 305000
+    },
+    {
+      "epoch": 2.360221574974698,
+      "grad_norm": 0.0004299231804907322,
+      "learning_rate": 7.526676806175808e-05,
+      "loss": 0.0003,
+      "step": 305500
+    },
+    {
+      "epoch": 2.3640844580761295,
+      "grad_norm": 0.0004845628864131868,
+      "learning_rate": 7.481230799061993e-05,
+      "loss": 0.0003,
+      "step": 306000
+    },
+    {
+      "epoch": 2.3679473411775613,
+      "grad_norm": 0.00041427643736824393,
+      "learning_rate": 7.435784791948179e-05,
+      "loss": 0.0003,
+      "step": 306500
+    },
+    {
+      "epoch": 2.3718102242789927,
+      "grad_norm": 0.0004160687094554305,
+      "learning_rate": 7.390338784834364e-05,
+      "loss": 0.0003,
+      "step": 307000
+    },
+    {
+      "epoch": 2.3756731073804245,
+      "grad_norm": 0.00044749100925400853,
+      "learning_rate": 7.344892777720549e-05,
+      "loss": 0.0003,
+      "step": 307500
+    },
+    {
+      "epoch": 2.3795359904818563,
+      "grad_norm": 0.0004799690213985741,
+      "learning_rate": 7.299446770606734e-05,
+      "loss": 0.0003,
+      "step": 308000
+    },
+    {
+      "epoch": 2.3833988735832876,
+      "grad_norm": 0.0004245197051204741,
+      "learning_rate": 7.254000763492918e-05,
+      "loss": 0.0003,
+      "step": 308500
+    },
+    {
+      "epoch": 2.387261756684719,
+      "grad_norm": 0.00034600042272359133,
+      "learning_rate": 7.208554756379105e-05,
+      "loss": 0.0003,
+      "step": 309000
+    },
+    {
+      "epoch": 2.3911246397861508,
+      "grad_norm": 0.0004245495656505227,
+      "learning_rate": 7.163108749265289e-05,
+      "loss": 0.0003,
+      "step": 309500
+    },
+    {
+      "epoch": 2.3949875228875825,
+      "grad_norm": 0.0004331233212724328,
+      "learning_rate": 7.117662742151474e-05,
+      "loss": 0.0003,
+      "step": 310000
+    },
+    {
+      "epoch": 2.398850405989014,
+      "grad_norm": 0.00041013164445757866,
+      "learning_rate": 7.072216735037659e-05,
+      "loss": 0.0003,
+      "step": 310500
+    },
+    {
+      "epoch": 2.4027132890904457,
+      "grad_norm": 0.00046506704529747367,
+      "learning_rate": 7.026770727923844e-05,
+      "loss": 0.0003,
+      "step": 311000
+    },
+    {
+      "epoch": 2.406576172191877,
+      "grad_norm": 0.0004329345829319209,
+      "learning_rate": 6.98132472081003e-05,
+      "loss": 0.0003,
+      "step": 311500
+    },
+    {
+      "epoch": 2.410439055293309,
+      "grad_norm": 0.00037213502218946815,
+      "learning_rate": 6.935878713696213e-05,
+      "loss": 0.0003,
+      "step": 312000
+    },
+    {
+      "epoch": 2.41430193839474,
+      "grad_norm": 0.0004223994619678706,
+      "learning_rate": 6.890432706582398e-05,
+      "loss": 0.0003,
+      "step": 312500
+    },
+    {
+      "epoch": 2.418164821496172,
+      "grad_norm": 0.00041198599501512945,
+      "learning_rate": 6.844986699468584e-05,
+      "loss": 0.0003,
+      "step": 313000
+    },
+    {
+      "epoch": 2.4220277045976033,
+      "grad_norm": 0.00044434002484194934,
+      "learning_rate": 6.799540692354769e-05,
+      "loss": 0.0003,
+      "step": 313500
+    },
+    {
+      "epoch": 2.425890587699035,
+      "grad_norm": 0.0003673033497761935,
+      "learning_rate": 6.754094685240954e-05,
+      "loss": 0.0003,
+      "step": 314000
+    },
+    {
+      "epoch": 2.429753470800467,
+      "grad_norm": 0.0004261393041815609,
+      "learning_rate": 6.70864867812714e-05,
+      "loss": 0.0003,
+      "step": 314500
+    },
+    {
+      "epoch": 2.4336163539018982,
+      "grad_norm": 0.0004654694930650294,
+      "learning_rate": 6.663202671013325e-05,
+      "loss": 0.0003,
+      "step": 315000
+    },
+    {
+      "epoch": 2.4374792370033296,
+      "grad_norm": 0.00038143552956171334,
+      "learning_rate": 6.617756663899508e-05,
+      "loss": 0.0003,
+      "step": 315500
+    },
+    {
+      "epoch": 2.4413421201047614,
+      "grad_norm": 0.00042159578879363835,
+      "learning_rate": 6.572310656785694e-05,
+      "loss": 0.0003,
+      "step": 316000
+    },
+    {
+      "epoch": 2.445205003206193,
+      "grad_norm": 0.0004413559508975595,
+      "learning_rate": 6.52686464967188e-05,
+      "loss": 0.0003,
+      "step": 316500
+    },
+    {
+      "epoch": 2.4490678863076245,
+      "grad_norm": 0.0005215974524617195,
+      "learning_rate": 6.481418642558064e-05,
+      "loss": 0.0003,
+      "step": 317000
+    },
+    {
+      "epoch": 2.4529307694090563,
+      "grad_norm": 0.000413423782447353,
+      "learning_rate": 6.435972635444249e-05,
+      "loss": 0.0003,
+      "step": 317500
+    },
+    {
+      "epoch": 2.4567936525104876,
+      "grad_norm": 0.00038620055420324206,
+      "learning_rate": 6.390526628330435e-05,
+      "loss": 0.0003,
+      "step": 318000
+    },
+    {
+      "epoch": 2.4606565356119194,
+      "grad_norm": 0.00041074163164012134,
+      "learning_rate": 6.34508062121662e-05,
+      "loss": 0.0003,
+      "step": 318500
+    },
+    {
+      "epoch": 2.464519418713351,
+      "grad_norm": 0.0004115503397770226,
+      "learning_rate": 6.299634614102805e-05,
+      "loss": 0.0003,
+      "step": 319000
+    },
+    {
+      "epoch": 2.4683823018147826,
+      "grad_norm": 0.00046097987797111273,
+      "learning_rate": 6.254188606988989e-05,
+      "loss": 0.0003,
+      "step": 319500
+    },
+    {
+      "epoch": 2.472245184916214,
+      "grad_norm": 0.0004020912747364491,
+      "learning_rate": 6.208742599875175e-05,
+      "loss": 0.0003,
+      "step": 320000
+    },
+    {
+      "epoch": 2.4761080680176457,
+      "grad_norm": 0.000455143628641963,
+      "learning_rate": 6.163296592761359e-05,
+      "loss": 0.0003,
+      "step": 320500
+    },
+    {
+      "epoch": 2.479970951119077,
+      "grad_norm": 0.00041087414138019085,
+      "learning_rate": 6.117850585647544e-05,
+      "loss": 0.0003,
+      "step": 321000
+    },
+    {
+      "epoch": 2.483833834220509,
+      "grad_norm": 0.00041171652264893055,
+      "learning_rate": 6.072404578533729e-05,
+      "loss": 0.0003,
+      "step": 321500
+    },
+    {
+      "epoch": 2.48769671732194,
+      "grad_norm": 0.0004624698485713452,
+      "learning_rate": 6.026958571419914e-05,
+      "loss": 0.0003,
+      "step": 322000
+    },
+    {
+      "epoch": 2.491559600423372,
+      "grad_norm": 0.0004319006693549454,
+      "learning_rate": 5.9815125643061e-05,
+      "loss": 0.0003,
+      "step": 322500
+    },
+    {
+      "epoch": 2.495422483524804,
+      "grad_norm": 0.00045846853754483163,
+      "learning_rate": 5.9360665571922846e-05,
+      "loss": 0.0003,
+      "step": 323000
+    },
+    {
+      "epoch": 2.499285366626235,
+      "grad_norm": 0.00042378256330266595,
+      "learning_rate": 5.89062055007847e-05,
+      "loss": 0.0003,
+      "step": 323500
+    },
+    {
+      "epoch": 2.5031482497276665,
+      "grad_norm": 0.00043979313340969384,
+      "learning_rate": 5.8451745429646544e-05,
+      "loss": 0.0003,
+      "step": 324000
+    },
+    {
+      "epoch": 2.5070111328290983,
+      "grad_norm": 0.00041039736242964864,
+      "learning_rate": 5.7997285358508396e-05,
+      "loss": 0.0003,
+      "step": 324500
+    },
+    {
+      "epoch": 2.51087401593053,
+      "grad_norm": 0.00043092764099128544,
+      "learning_rate": 5.754282528737025e-05,
+      "loss": 0.0003,
+      "step": 325000
+    },
+    {
+      "epoch": 2.5147368990319614,
+      "grad_norm": 0.0004449010593816638,
+      "learning_rate": 5.7088365216232093e-05,
+      "loss": 0.0003,
+      "step": 325500
+    },
+    {
+      "epoch": 2.518599782133393,
+      "grad_norm": 0.00043555087177082896,
+      "learning_rate": 5.663390514509395e-05,
+      "loss": 0.0003,
+      "step": 326000
+    },
+    {
+      "epoch": 2.5224626652348245,
+      "grad_norm": 0.00046992747229523957,
+      "learning_rate": 5.61794450739558e-05,
+      "loss": 0.0003,
+      "step": 326500
+    },
+    {
+      "epoch": 2.5263255483362563,
+      "grad_norm": 0.0004627726739272475,
+      "learning_rate": 5.572498500281765e-05,
+      "loss": 0.0003,
+      "step": 327000
+    },
+    {
+      "epoch": 2.5301884314376877,
+      "grad_norm": 0.0004344623303040862,
+      "learning_rate": 5.5270524931679495e-05,
+      "loss": 0.0003,
+      "step": 327500
+    },
+    {
+      "epoch": 2.5340513145391195,
+      "grad_norm": 0.0004928177804686129,
+      "learning_rate": 5.481606486054135e-05,
+      "loss": 0.0003,
+      "step": 328000
+    },
+    {
+      "epoch": 2.537914197640551,
+      "grad_norm": 0.0004244371084496379,
+      "learning_rate": 5.43616047894032e-05,
+      "loss": 0.0003,
+      "step": 328500
+    },
+    {
+      "epoch": 2.5417770807419826,
+      "grad_norm": 0.0004467975231818855,
+      "learning_rate": 5.3907144718265045e-05,
+      "loss": 0.0003,
+      "step": 329000
+    },
+    {
+      "epoch": 2.5456399638434144,
+      "grad_norm": 0.00041408653487451375,
+      "learning_rate": 5.34526846471269e-05,
+      "loss": 0.0003,
+      "step": 329500
+    },
+    {
+      "epoch": 2.5495028469448457,
+      "grad_norm": 0.00043784506851807237,
+      "learning_rate": 5.299822457598874e-05,
+      "loss": 0.0003,
+      "step": 330000
+    },
+    {
+      "epoch": 2.553365730046277,
+      "grad_norm": 0.00046141675557009876,
+      "learning_rate": 5.25437645048506e-05,
+      "loss": 0.0003,
+      "step": 330500
+    },
+    {
+      "epoch": 2.557228613147709,
+      "grad_norm": 0.0004683432634919882,
+      "learning_rate": 5.2089304433712454e-05,
+      "loss": 0.0003,
+      "step": 331000
+    },
+    {
+      "epoch": 2.5610914962491407,
+      "grad_norm": 0.0003889546496793628,
+      "learning_rate": 5.16348443625743e-05,
+      "loss": 0.0003,
+      "step": 331500
+    },
+    {
+      "epoch": 2.564954379350572,
+      "grad_norm": 0.0004467256658244878,
+      "learning_rate": 5.118038429143615e-05,
+      "loss": 0.0003,
+      "step": 332000
+    },
+    {
+      "epoch": 2.568817262452004,
+      "grad_norm": 0.00038880863576196134,
+      "learning_rate": 5.0725924220297997e-05,
+      "loss": 0.0003,
+      "step": 332500
+    },
+    {
+      "epoch": 2.572680145553435,
+      "grad_norm": 0.00043496862053871155,
+      "learning_rate": 5.027146414915985e-05,
+      "loss": 0.0003,
+      "step": 333000
+    },
+    {
+      "epoch": 2.576543028654867,
+      "grad_norm": 0.0003715400234796107,
+      "learning_rate": 4.9817004078021694e-05,
+      "loss": 0.0003,
+      "step": 333500
+    },
+    {
+      "epoch": 2.5804059117562983,
+      "grad_norm": 0.000468456739326939,
+      "learning_rate": 4.936254400688355e-05,
+      "loss": 0.0003,
+      "step": 334000
+    },
+    {
+      "epoch": 2.58426879485773,
+      "grad_norm": 0.00040985396481119096,
+      "learning_rate": 4.8908083935745405e-05,
+      "loss": 0.0003,
+      "step": 334500
+    },
+    {
+      "epoch": 2.5881316779591614,
+      "grad_norm": 0.0004460928903426975,
+      "learning_rate": 4.845362386460725e-05,
+      "loss": 0.0003,
+      "step": 335000
+    },
+    {
+      "epoch": 2.5919945610605932,
+      "grad_norm": 0.00039911369094625115,
+      "learning_rate": 4.79991637934691e-05,
+      "loss": 0.0003,
+      "step": 335500
+    },
+    {
+      "epoch": 2.595857444162025,
+      "grad_norm": 0.0004065028333570808,
+      "learning_rate": 4.754470372233095e-05,
+      "loss": 0.0003,
+      "step": 336000
+    },
+    {
+      "epoch": 2.5997203272634564,
+      "grad_norm": 0.0004132339672651142,
+      "learning_rate": 4.70902436511928e-05,
+      "loss": 0.0003,
+      "step": 336500
+    },
+    {
+      "epoch": 2.6035832103648877,
+      "grad_norm": 0.00044415410957299173,
+      "learning_rate": 4.663578358005466e-05,
+      "loss": 0.0003,
+      "step": 337000
+    },
+    {
+      "epoch": 2.6074460934663195,
+      "grad_norm": 0.00043628757703118026,
+      "learning_rate": 4.6181323508916505e-05,
+      "loss": 0.0003,
+      "step": 337500
+    },
+    {
+      "epoch": 2.6113089765677513,
+      "grad_norm": 0.0004943295498378575,
+      "learning_rate": 4.572686343777836e-05,
+      "loss": 0.0003,
+      "step": 338000
+    },
+    {
+      "epoch": 2.6151718596691826,
+      "grad_norm": 0.0004152540641371161,
+      "learning_rate": 4.52724033666402e-05,
+      "loss": 0.0003,
+      "step": 338500
+    },
+    {
+      "epoch": 2.6190347427706144,
+      "grad_norm": 0.0004294981772545725,
+      "learning_rate": 4.4817943295502054e-05,
+      "loss": 0.0003,
+      "step": 339000
+    },
+    {
+      "epoch": 2.622897625872046,
+      "grad_norm": 0.00039649149402976036,
+      "learning_rate": 4.43634832243639e-05,
+      "loss": 0.0003,
+      "step": 339500
+    },
+    {
+      "epoch": 2.6267605089734776,
+      "grad_norm": 0.00043368435581214726,
+      "learning_rate": 4.390902315322575e-05,
+      "loss": 0.0003,
+      "step": 340000
+    },
+    {
+      "epoch": 2.630623392074909,
+      "grad_norm": 0.0004430446424521506,
+      "learning_rate": 4.3454563082087604e-05,
+      "loss": 0.0003,
+      "step": 340500
+    },
+    {
+      "epoch": 2.6344862751763407,
+      "grad_norm": 0.00047145935241132975,
+      "learning_rate": 4.300010301094945e-05,
+      "loss": 0.0003,
+      "step": 341000
+    },
+    {
+      "epoch": 2.638349158277772,
+      "grad_norm": 0.0004192702181171626,
+      "learning_rate": 4.254564293981131e-05,
+      "loss": 0.0003,
+      "step": 341500
+    },
+    {
+      "epoch": 2.642212041379204,
+      "grad_norm": 0.00046445822226814926,
+      "learning_rate": 4.2091182868673154e-05,
+      "loss": 0.0003,
+      "step": 342000
+    },
+    {
+      "epoch": 2.6460749244806356,
+      "grad_norm": 0.000462083553429693,
+      "learning_rate": 4.1636722797535006e-05,
+      "loss": 0.0003,
+      "step": 342500
+    },
+    {
+      "epoch": 2.649937807582067,
+      "grad_norm": 0.00038474082248285413,
+      "learning_rate": 4.118226272639686e-05,
+      "loss": 0.0003,
+      "step": 343000
+    },
+    {
+      "epoch": 2.6538006906834983,
+      "grad_norm": 0.00039861336699686944,
+      "learning_rate": 4.07278026552587e-05,
+      "loss": 0.0003,
+      "step": 343500
+    },
+    {
+      "epoch": 2.65766357378493,
+      "grad_norm": 0.0004491112194955349,
+      "learning_rate": 4.0273342584120555e-05,
+      "loss": 0.0003,
+      "step": 344000
+    },
+    {
+      "epoch": 2.661526456886362,
+      "grad_norm": 0.00046189481508918107,
+      "learning_rate": 3.98188825129824e-05,
+      "loss": 0.0003,
+      "step": 344500
+    },
+    {
+      "epoch": 2.6653893399877933,
+      "grad_norm": 0.00038176134694367647,
+      "learning_rate": 3.936442244184426e-05,
+      "loss": 0.0003,
+      "step": 345000
+    },
+    {
+      "epoch": 2.6692522230892246,
+      "grad_norm": 0.00037889971281401813,
+      "learning_rate": 3.890996237070611e-05,
+      "loss": 0.0003,
+      "step": 345500
+    },
+    {
+      "epoch": 2.6731151061906564,
+      "grad_norm": 0.0004248527984600514,
+      "learning_rate": 3.845550229956796e-05,
+      "loss": 0.0003,
+      "step": 346000
+    },
+    {
+      "epoch": 2.676977989292088,
+      "grad_norm": 0.00037932791747152805,
+      "learning_rate": 3.800104222842981e-05,
+      "loss": 0.0003,
+      "step": 346500
+    },
+    {
+      "epoch": 2.6808408723935195,
+      "grad_norm": 0.0003727327275555581,
+      "learning_rate": 3.7546582157291655e-05,
+      "loss": 0.0003,
+      "step": 347000
+    },
+    {
+      "epoch": 2.6847037554949513,
+      "grad_norm": 0.00040110255940817297,
+      "learning_rate": 3.709212208615351e-05,
+      "loss": 0.0003,
+      "step": 347500
+    },
+    {
+      "epoch": 2.6885666385963827,
+      "grad_norm": 0.00042014423524960876,
+      "learning_rate": 3.663766201501536e-05,
+      "loss": 0.0003,
+      "step": 348000
+    },
+    {
+      "epoch": 2.6924295216978145,
+      "grad_norm": 0.00038637884426862,
+      "learning_rate": 3.6183201943877204e-05,
+      "loss": 0.0003,
+      "step": 348500
+    },
+    {
+      "epoch": 2.696292404799246,
+      "grad_norm": 0.0004331624659243971,
+      "learning_rate": 3.572874187273906e-05,
+      "loss": 0.0003,
+      "step": 349000
+    },
+    {
+      "epoch": 2.7001552879006776,
+      "grad_norm": 0.00040115389856509864,
+      "learning_rate": 3.527428180160091e-05,
+      "loss": 0.0003,
+      "step": 349500
+    },
+    {
+      "epoch": 2.704018171002109,
+      "grad_norm": 0.0004400424659252167,
+      "learning_rate": 3.481982173046276e-05,
+      "loss": 0.0003,
+      "step": 350000
+    },
+    {
+      "epoch": 2.7078810541035407,
+      "grad_norm": 0.000458619964774698,
+      "learning_rate": 3.436536165932461e-05,
+      "loss": 0.0003,
+      "step": 350500
+    },
+    {
+      "epoch": 2.7117439372049725,
+      "grad_norm": 0.00042235825094394386,
+      "learning_rate": 3.391090158818646e-05,
+      "loss": 0.0003,
+      "step": 351000
+    },
+    {
+      "epoch": 2.715606820306404,
+      "grad_norm": 0.00045935032540000975,
+      "learning_rate": 3.345644151704831e-05,
+      "loss": 0.0003,
+      "step": 351500
+    },
+    {
+      "epoch": 2.7194697034078352,
+      "grad_norm": 0.0004515291075222194,
+      "learning_rate": 3.3001981445910156e-05,
+      "loss": 0.0003,
+      "step": 352000
+    },
+    {
+      "epoch": 2.723332586509267,
+      "grad_norm": 0.00042773320456035435,
+      "learning_rate": 3.254752137477201e-05,
+      "loss": 0.0003,
+      "step": 352500
+    },
+    {
+      "epoch": 2.727195469610699,
+      "grad_norm": 0.00045925399172119796,
+      "learning_rate": 3.209306130363386e-05,
+      "loss": 0.0003,
+      "step": 353000
+    },
+    {
+      "epoch": 2.73105835271213,
+      "grad_norm": 0.0004555301566142589,
+      "learning_rate": 3.163860123249571e-05,
+      "loss": 0.0003,
+      "step": 353500
+    },
+    {
+      "epoch": 2.734921235813562,
+      "grad_norm": 0.000439466122770682,
+      "learning_rate": 3.118414116135756e-05,
+      "loss": 0.0003,
+      "step": 354000
+    },
+    {
+      "epoch": 2.7387841189149933,
+      "grad_norm": 0.00042543449671939015,
+      "learning_rate": 3.072968109021941e-05,
+      "loss": 0.0003,
+      "step": 354500
+    },
+    {
+      "epoch": 2.742647002016425,
+      "grad_norm": 0.00039303614175878465,
+      "learning_rate": 3.027522101908126e-05,
+      "loss": 0.0003,
+      "step": 355000
+    },
+    {
+      "epoch": 2.7465098851178564,
+      "grad_norm": 0.0003858358832076192,
+      "learning_rate": 2.9820760947943108e-05,
+      "loss": 0.0003,
+      "step": 355500
+    },
+    {
+      "epoch": 2.750372768219288,
+      "grad_norm": 0.0004583084082696587,
+      "learning_rate": 2.9366300876804963e-05,
+      "loss": 0.0003,
+      "step": 356000
+    },
+    {
+      "epoch": 2.7542356513207196,
+      "grad_norm": 0.00042337161721661687,
+      "learning_rate": 2.8911840805666812e-05,
+      "loss": 0.0003,
+      "step": 356500
+    },
+    {
+      "epoch": 2.7580985344221514,
+      "grad_norm": 0.0004135474155191332,
+      "learning_rate": 2.845738073452866e-05,
+      "loss": 0.0003,
+      "step": 357000
+    },
+    {
+      "epoch": 2.761961417523583,
+      "grad_norm": 0.0004276975814718753,
+      "learning_rate": 2.8002920663390513e-05,
+      "loss": 0.0003,
+      "step": 357500
+    },
+    {
+      "epoch": 2.7658243006250145,
+      "grad_norm": 0.00042497014510445297,
+      "learning_rate": 2.754846059225236e-05,
+      "loss": 0.0003,
+      "step": 358000
+    },
+    {
+      "epoch": 2.769687183726446,
+      "grad_norm": 0.0004969782894477248,
+      "learning_rate": 2.709400052111421e-05,
+      "loss": 0.0003,
+      "step": 358500
+    },
+    {
+      "epoch": 2.7735500668278776,
+      "grad_norm": 0.0004229978076182306,
+      "learning_rate": 2.6639540449976066e-05,
+      "loss": 0.0003,
+      "step": 359000
+    },
+    {
+      "epoch": 2.7774129499293094,
+      "grad_norm": 0.0004182982083875686,
+      "learning_rate": 2.6185080378837915e-05,
+      "loss": 0.0003,
+      "step": 359500
+    },
+    {
+      "epoch": 2.7812758330307408,
+      "grad_norm": 0.00044540813541971147,
+      "learning_rate": 2.5730620307699763e-05,
+      "loss": 0.0003,
+      "step": 360000
+    },
+    {
+      "epoch": 2.7851387161321726,
+      "grad_norm": 0.0004250209603924304,
+      "learning_rate": 2.5276160236561612e-05,
+      "loss": 0.0003,
+      "step": 360500
+    },
+    {
+      "epoch": 2.789001599233604,
+      "grad_norm": 0.00039502166328020394,
+      "learning_rate": 2.482170016542346e-05,
+      "loss": 0.0003,
+      "step": 361000
+    },
+    {
+      "epoch": 2.7928644823350357,
+      "grad_norm": 0.0004499274946283549,
+      "learning_rate": 2.4367240094285313e-05,
+      "loss": 0.0003,
+      "step": 361500
+    },
+    {
+      "epoch": 2.796727365436467,
+      "grad_norm": 0.0003957873850595206,
+      "learning_rate": 2.3912780023147165e-05,
+      "loss": 0.0003,
+      "step": 362000
+    },
+    {
+      "epoch": 2.800590248537899,
+      "grad_norm": 0.000493398925755173,
+      "learning_rate": 2.3458319952009014e-05,
+      "loss": 0.0003,
+      "step": 362500
+    },
+    {
+      "epoch": 2.80445313163933,
+      "grad_norm": 0.00042534625390544534,
+      "learning_rate": 2.3003859880870866e-05,
+      "loss": 0.0003,
+      "step": 363000
+    },
+    {
+      "epoch": 2.808316014740762,
+      "grad_norm": 0.0004305491747800261,
+      "learning_rate": 2.2549399809732715e-05,
+      "loss": 0.0003,
+      "step": 363500
+    },
+    {
+      "epoch": 2.8121788978421938,
+      "grad_norm": 0.0004288312920834869,
+      "learning_rate": 2.2094939738594564e-05,
+      "loss": 0.0003,
+      "step": 364000
+    },
+    {
+      "epoch": 2.816041780943625,
+      "grad_norm": 0.00046938960440456867,
+      "learning_rate": 2.1640479667456416e-05,
+      "loss": 0.0003,
+      "step": 364500
+    },
+    {
+      "epoch": 2.8199046640450565,
+      "grad_norm": 0.00041804375359788537,
+      "learning_rate": 2.1186019596318268e-05,
+      "loss": 0.0003,
+      "step": 365000
+    },
+    {
+      "epoch": 2.8237675471464883,
+      "grad_norm": 0.00041626792517490685,
+      "learning_rate": 2.0731559525180117e-05,
+      "loss": 0.0003,
+      "step": 365500
+    },
+    {
+      "epoch": 2.82763043024792,
+      "grad_norm": 0.0004569618031382561,
+      "learning_rate": 2.0277099454041966e-05,
+      "loss": 0.0003,
+      "step": 366000
+    },
+    {
+      "epoch": 2.8314933133493514,
+      "grad_norm": 0.00040743814315646887,
+      "learning_rate": 1.9822639382903814e-05,
+      "loss": 0.0003,
+      "step": 366500
+    },
+    {
+      "epoch": 2.8353561964507827,
+      "grad_norm": 0.0005269908579066396,
+      "learning_rate": 1.9368179311765666e-05,
+      "loss": 0.0003,
+      "step": 367000
+    },
+    {
+      "epoch": 2.8392190795522145,
+      "grad_norm": 0.0003937318397220224,
+      "learning_rate": 1.891371924062752e-05,
+      "loss": 0.0003,
+      "step": 367500
+    },
+    {
+      "epoch": 2.8430819626536463,
+      "grad_norm": 0.0005019407253712416,
+      "learning_rate": 1.8459259169489367e-05,
+      "loss": 0.0003,
+      "step": 368000
+    },
+    {
+      "epoch": 2.8469448457550777,
+      "grad_norm": 0.00042626584763638675,
+      "learning_rate": 1.8004799098351216e-05,
+      "loss": 0.0003,
+      "step": 368500
+    },
+    {
+      "epoch": 2.8508077288565095,
+      "grad_norm": 0.0004206789599265903,
+      "learning_rate": 1.755033902721307e-05,
+      "loss": 0.0003,
+      "step": 369000
+    },
+    {
+      "epoch": 2.854670611957941,
+      "grad_norm": 0.0004782930773217231,
+      "learning_rate": 1.7095878956074917e-05,
+      "loss": 0.0003,
+      "step": 369500
+    },
+    {
+      "epoch": 2.8585334950593726,
+      "grad_norm": 0.00043975317385047674,
+      "learning_rate": 1.664141888493677e-05,
+      "loss": 0.0003,
+      "step": 370000
+    },
+    {
+      "epoch": 2.862396378160804,
+      "grad_norm": 0.00044531343155540526,
+      "learning_rate": 1.6186958813798618e-05,
+      "loss": 0.0003,
+      "step": 370500
+    },
+    {
+      "epoch": 2.8662592612622357,
+      "grad_norm": 0.0004300449218135327,
+      "learning_rate": 1.5732498742660467e-05,
+      "loss": 0.0003,
+      "step": 371000
+    },
+    {
+      "epoch": 2.870122144363667,
+      "grad_norm": 0.0003669953439384699,
+      "learning_rate": 1.527803867152232e-05,
+      "loss": 0.0003,
+      "step": 371500
+    },
+    {
+      "epoch": 2.873985027465099,
+      "grad_norm": 0.00039420899702236056,
+      "learning_rate": 1.482357860038417e-05,
+      "loss": 0.0003,
+      "step": 372000
+    },
+    {
+      "epoch": 2.8778479105665307,
+      "grad_norm": 0.00044955965131521225,
+      "learning_rate": 1.436911852924602e-05,
+      "loss": 0.0003,
+      "step": 372500
+    },
+    {
+      "epoch": 2.881710793667962,
+      "grad_norm": 0.0004604061250574887,
+      "learning_rate": 1.391465845810787e-05,
+      "loss": 0.0003,
+      "step": 373000
+    },
+    {
+      "epoch": 2.8855736767693934,
+      "grad_norm": 0.00042817433131858706,
+      "learning_rate": 1.3460198386969719e-05,
+      "loss": 0.0003,
+      "step": 373500
+    },
+    {
+      "epoch": 2.889436559870825,
+      "grad_norm": 0.0004006045637652278,
+      "learning_rate": 1.3005738315831571e-05,
+      "loss": 0.0003,
+      "step": 374000
+    },
+    {
+      "epoch": 2.893299442972257,
+      "grad_norm": 0.00040681843529455364,
+      "learning_rate": 1.255127824469342e-05,
+      "loss": 0.0003,
+      "step": 374500
+    },
+    {
+      "epoch": 2.8971623260736883,
+      "grad_norm": 0.0004340211162343621,
+      "learning_rate": 1.209681817355527e-05,
+      "loss": 0.0003,
+      "step": 375000
+    },
+    {
+      "epoch": 2.90102520917512,
+      "grad_norm": 0.0004770898667629808,
+      "learning_rate": 1.1642358102417121e-05,
+      "loss": 0.0003,
+      "step": 375500
+    },
+    {
+      "epoch": 2.9048880922765514,
+      "grad_norm": 0.0004441466589923948,
+      "learning_rate": 1.1187898031278971e-05,
+      "loss": 0.0003,
+      "step": 376000
+    },
+    {
+      "epoch": 2.908750975377983,
+      "grad_norm": 0.00045734934974461794,
+      "learning_rate": 1.073343796014082e-05,
+      "loss": 0.0003,
+      "step": 376500
+    },
+    {
+      "epoch": 2.9126138584794146,
+      "grad_norm": 0.0004386783402878791,
+      "learning_rate": 1.0278977889002672e-05,
+      "loss": 0.0003,
+      "step": 377000
+    },
+    {
+      "epoch": 2.9164767415808464,
+      "grad_norm": 0.0003852332301903516,
+      "learning_rate": 9.824517817864521e-06,
+      "loss": 0.0003,
+      "step": 377500
+    },
+    {
+      "epoch": 2.9203396246822777,
+      "grad_norm": 0.0004367204965092242,
+      "learning_rate": 9.370057746726372e-06,
+      "loss": 0.0003,
+      "step": 378000
+    },
+    {
+      "epoch": 2.9242025077837095,
+      "grad_norm": 0.0004893042496405542,
+      "learning_rate": 8.915597675588222e-06,
+      "loss": 0.0003,
+      "step": 378500
+    },
+    {
+      "epoch": 2.9280653908851413,
+      "grad_norm": 0.0004190478939563036,
+      "learning_rate": 8.461137604450072e-06,
+      "loss": 0.0003,
+      "step": 379000
+    },
+    {
+      "epoch": 2.9319282739865726,
+      "grad_norm": 0.0004581400426104665,
+      "learning_rate": 8.006677533311923e-06,
+      "loss": 0.0003,
+      "step": 379500
+    },
+    {
+      "epoch": 2.935791157088004,
+      "grad_norm": 0.00042551904334686697,
+      "learning_rate": 7.5522174621737725e-06,
+      "loss": 0.0003,
+      "step": 380000
+    },
+    {
+      "epoch": 2.9396540401894358,
+      "grad_norm": 0.0004512323357630521,
+      "learning_rate": 7.097757391035623e-06,
+      "loss": 0.0003,
+      "step": 380500
+    },
+    {
+      "epoch": 2.9435169232908676,
+      "grad_norm": 0.0004587092262227088,
+      "learning_rate": 6.6432973198974734e-06,
+      "loss": 0.0003,
+      "step": 381000
+    },
+    {
+      "epoch": 2.947379806392299,
+      "grad_norm": 0.00046226088306866586,
+      "learning_rate": 6.188837248759323e-06,
+      "loss": 0.0003,
+      "step": 381500
+    },
+    {
+      "epoch": 2.9512426894937307,
+      "grad_norm": 0.0004404925275593996,
+      "learning_rate": 5.7343771776211735e-06,
+      "loss": 0.0003,
+      "step": 382000
+    },
+    {
+      "epoch": 2.955105572595162,
+      "grad_norm": 0.00045682353083975613,
+      "learning_rate": 5.279917106483024e-06,
+      "loss": 0.0003,
+      "step": 382500
+    },
+    {
+      "epoch": 2.958968455696594,
+      "grad_norm": 0.0004541674570646137,
+      "learning_rate": 4.825457035344874e-06,
+      "loss": 0.0003,
+      "step": 383000
+    },
+    {
+      "epoch": 2.962831338798025,
+      "grad_norm": 0.00042987189954146743,
+      "learning_rate": 4.370996964206724e-06,
+      "loss": 0.0003,
+      "step": 383500
+    },
+    {
+      "epoch": 2.966694221899457,
+      "grad_norm": 0.00045570009388029575,
+      "learning_rate": 3.9165368930685745e-06,
+      "loss": 0.0003,
+      "step": 384000
+    },
+    {
+      "epoch": 2.9705571050008883,
+      "grad_norm": 0.0004511605075094849,
+      "learning_rate": 3.462076821930425e-06,
+      "loss": 0.0003,
+      "step": 384500
+    },
+    {
+      "epoch": 2.97441998810232,
+      "grad_norm": 0.000448725710157305,
+      "learning_rate": 3.007616750792275e-06,
+      "loss": 0.0003,
+      "step": 385000
+    },
+    {
+      "epoch": 2.978282871203752,
+      "grad_norm": 0.0004488198319450021,
+      "learning_rate": 2.553156679654125e-06,
+      "loss": 0.0003,
+      "step": 385500
+    },
+    {
+      "epoch": 2.9821457543051832,
+      "grad_norm": 0.000420234544435516,
+      "learning_rate": 2.0986966085159755e-06,
+      "loss": 0.0003,
+      "step": 386000
+    },
+    {
+      "epoch": 2.9860086374066146,
+      "grad_norm": 0.00040884234476834536,
+      "learning_rate": 1.6442365373778258e-06,
+      "loss": 0.0003,
+      "step": 386500
+    },
+    {
+      "epoch": 2.9898715205080464,
+      "grad_norm": 0.0005011099274270236,
+      "learning_rate": 1.189776466239676e-06,
+      "loss": 0.0003,
+      "step": 387000
+    },
+    {
+      "epoch": 2.993734403609478,
+      "grad_norm": 0.00048375382903032005,
+      "learning_rate": 7.353163951015264e-07,
+      "loss": 0.0003,
+      "step": 387500
+    },
+    {
+      "epoch": 2.9975972867109095,
+      "grad_norm": 0.00044487029663287103,
+      "learning_rate": 2.8085632396337657e-07,
+      "loss": 0.0003,
+      "step": 388000
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 388311,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 129437,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 64,
+  "trial_name": null,
+  "trial_params": null
+}