diff --git "a/last-checkpoint/trainer_state.json" "b/last-checkpoint/trainer_state.json"
--- "a/last-checkpoint/trainer_state.json"
+++ "b/last-checkpoint/trainer_state.json"
@@ -1,6721 +1,99 @@
 {
-  "best_global_step": 38250,
-  "best_metric": -30.471132278442383,
-  "best_model_checkpoint": "./ar-diffusion-checkpoints-progressive-attention/checkpoint-37500",
-  "epoch": 2.999769248519345,
+  "best_global_step": 500,
+  "best_metric": 1.7362440824508667,
+  "best_model_checkpoint": "./ar-diffusion-checkpoints-progressive-attention/checkpoint-500",
+  "epoch": 0.038458580109222366,
   "eval_steps": 250,
-  "global_step": 39000,
+  "global_step": 500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.003845858010922237,
-      "grad_norm": 13.142335891723633,
-      "learning_rate": 1.84e-05,
-      "loss": 12.0117,
+      "grad_norm": 8.293161392211914,
+      "learning_rate": 1.88e-05,
+      "loss": 8.4888,
       "step": 50
     },
     {
       "epoch": 0.007691716021844474,
-      "grad_norm": 5.153238296508789,
-      "learning_rate": 3.8e-05,
-      "loss": 5.8408,
+      "grad_norm": 2.511484384536743,
+      "learning_rate": 3.88e-05,
+      "loss": 2.9957,
       "step": 100
     },
     {
       "epoch": 0.01153757403276671,
-      "grad_norm": 7.471889019012451,
-      "learning_rate": 5.8e-05,
-      "loss": 4.5173,
+      "grad_norm": 3.1082534790039062,
+      "learning_rate": 5.88e-05,
+      "loss": 2.2168,
       "step": 150
     },
     {
       "epoch": 0.015383432043688947,
-      "grad_norm": 4.756172180175781,
-      "learning_rate": 7.800000000000001e-05,
-      "loss": 3.8946,
+      "grad_norm": 3.4769539833068848,
+      "learning_rate": 7.88e-05,
+      "loss": 1.9786,
       "step": 200
     },
     {
       "epoch": 0.019229290054611183,
-      "grad_norm": 4.0329108238220215,
-      "learning_rate": 9.8e-05,
-      "loss": 3.842,
+      "grad_norm": 1.49607253074646,
+      "learning_rate": 9.88e-05,
+      "loss": 1.9748,
       "step": 250
     },
     {
       "epoch": 0.019229290054611183,
-      "eval_loss": 3.7754223346710205,
-      "eval_runtime": 18.0149,
-      "eval_samples_per_second": 55.51,
-      "eval_steps_per_second": 13.877,
+      "eval_loss": 1.9188385009765625,
+      "eval_runtime": 16.8659,
+      "eval_samples_per_second": 59.291,
+      "eval_steps_per_second": 14.823,
       "step": 250
     },
     {
       "epoch": 0.02307514806553342,
-      "grad_norm": 7.410353660583496,
-      "learning_rate": 0.000118,
-      "loss": 3.5501,
+      "grad_norm": 2.4030940532684326,
+      "learning_rate": 0.0001188,
+      "loss": 1.7869,
       "step": 300
     },
     {
       "epoch": 0.02692100607645566,
-      "grad_norm": 4.085294246673584,
-      "learning_rate": 0.000138,
-      "loss": 3.4015,
+      "grad_norm": 8.926054000854492,
+      "learning_rate": 0.00013879999999999999,
+      "loss": 1.7282,
       "step": 350
     },
     {
       "epoch": 0.030766864087377895,
-      "grad_norm": 5.78131103515625,
-      "learning_rate": 0.00015800000000000002,
-      "loss": 3.4988,
+      "grad_norm": 2.0970349311828613,
+      "learning_rate": 0.0001588,
+      "loss": 1.8283,
       "step": 400
     },
     {
       "epoch": 0.03461272209830013,
-      "grad_norm": 3.2009644508361816,
-      "learning_rate": 0.00017800000000000002,
-      "loss": 3.4145,
+      "grad_norm": 1.1687132120132446,
+      "learning_rate": 0.0001788,
+      "loss": 1.7503,
       "step": 450
     },
     {
       "epoch": 0.038458580109222366,
-      "grad_norm": 4.380056858062744,
-      "learning_rate": 0.00019800000000000002,
-      "loss": 3.3179,
+      "grad_norm": 1.5801700353622437,
+      "learning_rate": 0.0001988,
+      "loss": 1.7329,
       "step": 500
     },
     {
       "epoch": 0.038458580109222366,
-      "eval_loss": 3.3119897842407227,
-      "eval_runtime": 17.9539,
-      "eval_samples_per_second": 55.698,
-      "eval_steps_per_second": 13.925,
+      "eval_loss": 1.7362440824508667,
+      "eval_runtime": 16.8354,
+      "eval_samples_per_second": 59.399,
+      "eval_steps_per_second": 14.85,
       "step": 500
-    },
-    {
-      "epoch": 0.0423044381201446,
-      "grad_norm": 4.549246788024902,
-      "learning_rate": 0.0001997662519803652,
-      "loss": 3.3535,
-      "step": 550
-    },
-    {
-      "epoch": 0.04615029613106684,
-      "grad_norm": 4.421209335327148,
-      "learning_rate": 0.0001995065319585487,
-      "loss": 3.252,
-      "step": 600
-    },
-    {
-      "epoch": 0.04999615414198908,
-      "grad_norm": 3.3526384830474854,
-      "learning_rate": 0.0001992468119367322,
-      "loss": 3.1192,
-      "step": 650
-    },
-    {
-      "epoch": 0.05384201215291132,
-      "grad_norm": 3.0326781272888184,
-      "learning_rate": 0.00019898709191491573,
-      "loss": 2.9977,
-      "step": 700
-    },
-    {
-      "epoch": 0.05768787016383355,
-      "grad_norm": 5.362135410308838,
-      "learning_rate": 0.00019872737189309924,
-      "loss": 3.2504,
-      "step": 750
-    },
-    {
-      "epoch": 0.05768787016383355,
-      "eval_loss": 3.141200542449951,
-      "eval_runtime": 17.9291,
-      "eval_samples_per_second": 55.775,
-      "eval_steps_per_second": 13.944,
-      "step": 750
-    },
-    {
-      "epoch": 0.06153372817475579,
-      "grad_norm": 4.031102180480957,
-      "learning_rate": 0.00019846765187128275,
-      "loss": 3.2034,
-      "step": 800
-    },
-    {
-      "epoch": 0.06537958618567802,
-      "grad_norm": 4.671544551849365,
-      "learning_rate": 0.0001982079318494663,
-      "loss": 3.1486,
-      "step": 850
-    },
-    {
-      "epoch": 0.06922544419660026,
-      "grad_norm": 2.8522539138793945,
-      "learning_rate": 0.0001979482118276498,
-      "loss": 3.1832,
-      "step": 900
-    },
-    {
-      "epoch": 0.0730713022075225,
-      "grad_norm": 1.8937668800354004,
-      "learning_rate": 0.00019768849180583334,
-      "loss": 2.9562,
-      "step": 950
-    },
-    {
-      "epoch": 0.07691716021844473,
-      "grad_norm": 3.348437547683716,
-      "learning_rate": 0.00019742877178401683,
-      "loss": 3.0789,
-      "step": 1000
-    },
-    {
-      "epoch": 0.07691716021844473,
-      "eval_loss": 3.0555787086486816,
-      "eval_runtime": 17.9146,
-      "eval_samples_per_second": 55.82,
-      "eval_steps_per_second": 13.955,
-      "step": 1000
-    },
-    {
-      "epoch": 0.08076301822936698,
-      "grad_norm": 2.3448729515075684,
-      "learning_rate": 0.00019716905176220034,
-      "loss": 2.9338,
-      "step": 1050
-    },
-    {
-      "epoch": 0.0846088762402892,
-      "grad_norm": 2.609321355819702,
-      "learning_rate": 0.00019690933174038388,
-      "loss": 3.0214,
-      "step": 1100
-    },
-    {
-      "epoch": 0.08845473425121145,
-      "grad_norm": 2.4843454360961914,
-      "learning_rate": 0.0001966496117185674,
-      "loss": 2.9918,
-      "step": 1150
-    },
-    {
-      "epoch": 0.09230059226213368,
-      "grad_norm": 4.419650077819824,
-      "learning_rate": 0.0001963898916967509,
-      "loss": 2.9023,
-      "step": 1200
-    },
-    {
-      "epoch": 0.09614645027305592,
-      "grad_norm": 3.8419339656829834,
-      "learning_rate": 0.00019613017167493445,
-      "loss": 3.0145,
-      "step": 1250
-    },
-    {
-      "epoch": 0.09614645027305592,
-      "eval_loss": 2.9297268390655518,
-      "eval_runtime": 17.9152,
-      "eval_samples_per_second": 55.819,
-      "eval_steps_per_second": 13.955,
-      "step": 1250
-    },
-    {
-      "epoch": 0.09999230828397816,
-      "grad_norm": 7.558617115020752,
-      "learning_rate": 0.00019587045165311796,
-      "loss": 2.9151,
-      "step": 1300
-    },
-    {
-      "epoch": 0.10383816629490039,
-      "grad_norm": 2.1020314693450928,
-      "learning_rate": 0.00019561073163130145,
-      "loss": 2.8712,
-      "step": 1350
-    },
-    {
-      "epoch": 0.10768402430582263,
-      "grad_norm": 2.2210774421691895,
-      "learning_rate": 0.000195351011609485,
-      "loss": 2.8326,
-      "step": 1400
-    },
-    {
-      "epoch": 0.11152988231674486,
-      "grad_norm": 6.891756534576416,
-      "learning_rate": 0.0001950912915876685,
-      "loss": 2.8415,
-      "step": 1450
-    },
-    {
-      "epoch": 0.1153757403276671,
-      "grad_norm": 4.5944061279296875,
-      "learning_rate": 0.000194831571565852,
-      "loss": 2.9745,
-      "step": 1500
-    },
-    {
-      "epoch": 0.1153757403276671,
-      "eval_loss": 2.887022018432617,
-      "eval_runtime": 17.9841,
-      "eval_samples_per_second": 55.605,
-      "eval_steps_per_second": 13.901,
-      "step": 1500
-    },
-    {
-      "epoch": 0.11922159833858934,
-      "grad_norm": 3.0902695655822754,
-      "learning_rate": 0.00019457185154403555,
-      "loss": 2.8636,
-      "step": 1550
-    },
-    {
-      "epoch": 0.12306745634951158,
-      "grad_norm": 5.035081386566162,
-      "learning_rate": 0.00019431213152221907,
-      "loss": 2.9387,
-      "step": 1600
-    },
-    {
-      "epoch": 0.12691331436043382,
-      "grad_norm": 2.6864142417907715,
-      "learning_rate": 0.00019405241150040255,
-      "loss": 2.8064,
-      "step": 1650
-    },
-    {
-      "epoch": 0.13075917237135604,
-      "grad_norm": 4.986116409301758,
-      "learning_rate": 0.0001937926914785861,
-      "loss": 2.9818,
-      "step": 1700
-    },
-    {
-      "epoch": 0.13460503038227828,
-      "grad_norm": 2.94878888130188,
-      "learning_rate": 0.0001935329714567696,
-      "loss": 2.8934,
-      "step": 1750
-    },
-    {
-      "epoch": 0.13460503038227828,
-      "eval_loss": 2.7845654487609863,
-      "eval_runtime": 17.7992,
-      "eval_samples_per_second": 56.182,
-      "eval_steps_per_second": 14.046,
-      "step": 1750
-    },
-    {
-      "epoch": 0.13845088839320052,
-      "grad_norm": 1.6211572885513306,
-      "learning_rate": 0.00019327844583538946,
-      "loss": 2.8786,
-      "step": 1800
-    },
-    {
-      "epoch": 0.14229674640412276,
-      "grad_norm": 2.8289225101470947,
-      "learning_rate": 0.000193018725813573,
-      "loss": 2.702,
-      "step": 1850
-    },
-    {
-      "epoch": 0.146142604415045,
-      "grad_norm": 3.591913938522339,
-      "learning_rate": 0.0001927590057917565,
-      "loss": 2.8623,
-      "step": 1900
-    },
-    {
-      "epoch": 0.14998846242596722,
-      "grad_norm": 2.2412045001983643,
-      "learning_rate": 0.00019249928576994002,
-      "loss": 2.7517,
-      "step": 1950
-    },
-    {
-      "epoch": 0.15383432043688947,
-      "grad_norm": 2.419358015060425,
-      "learning_rate": 0.00019223956574812353,
-      "loss": 2.8122,
-      "step": 2000
-    },
-    {
-      "epoch": 0.15383432043688947,
-      "eval_loss": 2.7242305278778076,
-      "eval_runtime": 18.2018,
-      "eval_samples_per_second": 54.94,
-      "eval_steps_per_second": 13.735,
-      "step": 2000
-    },
-    {
-      "epoch": 0.1576801784478117,
-      "grad_norm": 2.873013973236084,
-      "learning_rate": 0.00019197984572630705,
-      "loss": 2.8007,
-      "step": 2050
-    },
-    {
-      "epoch": 0.16152603645873395,
-      "grad_norm": 2.716311454772949,
-      "learning_rate": 0.00019172012570449056,
-      "loss": 2.7435,
-      "step": 2100
-    },
-    {
-      "epoch": 0.16537189446965617,
-      "grad_norm": 2.8333351612091064,
-      "learning_rate": 0.0001914604056826741,
-      "loss": 2.6415,
-      "step": 2150
-    },
-    {
-      "epoch": 0.1692177524805784,
-      "grad_norm": 2.460752248764038,
-      "learning_rate": 0.0001912006856608576,
-      "loss": 2.64,
-      "step": 2200
-    },
-    {
-      "epoch": 0.17306361049150065,
-      "grad_norm": 2.181720733642578,
-      "learning_rate": 0.00019094096563904112,
-      "loss": 2.5928,
-      "step": 2250
-    },
-    {
-      "epoch": 0.17306361049150065,
-      "eval_loss": 2.620633363723755,
-      "eval_runtime": 17.9269,
-      "eval_samples_per_second": 55.782,
-      "eval_steps_per_second": 13.946,
-      "step": 2250
-    },
-    {
-      "epoch": 0.1769094685024229,
-      "grad_norm": 2.5329666137695312,
-      "learning_rate": 0.00019068124561722464,
-      "loss": 2.7231,
-      "step": 2300
-    },
-    {
-      "epoch": 0.18075532651334514,
-      "grad_norm": 3.047774314880371,
-      "learning_rate": 0.00019042152559540815,
-      "loss": 2.6484,
-      "step": 2350
-    },
-    {
-      "epoch": 0.18460118452426735,
-      "grad_norm": 3.1008481979370117,
-      "learning_rate": 0.00019016180557359166,
-      "loss": 2.6458,
-      "step": 2400
-    },
-    {
-      "epoch": 0.1884470425351896,
-      "grad_norm": 2.8875646591186523,
-      "learning_rate": 0.0001899020855517752,
-      "loss": 2.6817,
-      "step": 2450
-    },
-    {
-      "epoch": 0.19229290054611184,
-      "grad_norm": 2.3887267112731934,
-      "learning_rate": 0.00018964236552995872,
-      "loss": 2.6372,
-      "step": 2500
-    },
-    {
-      "epoch": 0.19229290054611184,
-      "eval_loss": 2.6067891120910645,
-      "eval_runtime": 18.0696,
-      "eval_samples_per_second": 55.342,
-      "eval_steps_per_second": 13.835,
-      "step": 2500
-    },
-    {
-      "epoch": 0.19613875855703408,
-      "grad_norm": 2.686868190765381,
-      "learning_rate": 0.00018938264550814223,
-      "loss": 2.498,
-      "step": 2550
-    },
-    {
-      "epoch": 0.19998461656795632,
-      "grad_norm": 2.962398052215576,
-      "learning_rate": 0.00018912292548632574,
-      "loss": 2.6486,
-      "step": 2600
-    },
-    {
-      "epoch": 0.20383047457887854,
-      "grad_norm": 1.946420669555664,
-      "learning_rate": 0.00018886320546450925,
-      "loss": 2.5939,
-      "step": 2650
-    },
-    {
-      "epoch": 0.20767633258980078,
-      "grad_norm": 3.541088104248047,
-      "learning_rate": 0.0001886034854426928,
-      "loss": 2.5497,
-      "step": 2700
-    },
-    {
-      "epoch": 0.21152219060072303,
-      "grad_norm": 2.2219605445861816,
-      "learning_rate": 0.0001883437654208763,
-      "loss": 2.4533,
-      "step": 2750
-    },
-    {
-      "epoch": 0.21152219060072303,
-      "eval_loss": 2.572129011154175,
-      "eval_runtime": 17.9723,
-      "eval_samples_per_second": 55.641,
-      "eval_steps_per_second": 13.91,
-      "step": 2750
-    },
-    {
-      "epoch": 0.21536804861164527,
-      "grad_norm": 3.535885810852051,
-      "learning_rate": 0.00018808404539905982,
-      "loss": 2.4784,
-      "step": 2800
-    },
-    {
-      "epoch": 0.21921390662256748,
-      "grad_norm": 3.0803003311157227,
-      "learning_rate": 0.00018782432537724336,
-      "loss": 2.5209,
-      "step": 2850
-    },
-    {
-      "epoch": 0.22305976463348973,
-      "grad_norm": 2.4540822505950928,
-      "learning_rate": 0.00018756460535542685,
-      "loss": 2.6128,
-      "step": 2900
-    },
-    {
-      "epoch": 0.22690562264441197,
-      "grad_norm": 2.583606243133545,
-      "learning_rate": 0.00018730488533361036,
-      "loss": 2.464,
-      "step": 2950
-    },
-    {
-      "epoch": 0.2307514806553342,
-      "grad_norm": 3.963115930557251,
-      "learning_rate": 0.0001870451653117939,
-      "loss": 2.5803,
-      "step": 3000
-    },
-    {
-      "epoch": 0.2307514806553342,
-      "eval_loss": 2.4989535808563232,
-      "eval_runtime": 17.9286,
-      "eval_samples_per_second": 55.777,
-      "eval_steps_per_second": 13.944,
-      "step": 3000
-    },
-    {
-      "epoch": 0.23459733866625646,
-      "grad_norm": 3.3239638805389404,
-      "learning_rate": 0.0001867854452899774,
-      "loss": 2.5628,
-      "step": 3050
-    },
-    {
-      "epoch": 0.23844319667717867,
-      "grad_norm": 2.3675172328948975,
-      "learning_rate": 0.00018652572526816092,
-      "loss": 2.4826,
-      "step": 3100
-    },
-    {
-      "epoch": 0.2422890546881009,
-      "grad_norm": 2.950798988342285,
-      "learning_rate": 0.00018626600524634446,
-      "loss": 2.518,
-      "step": 3150
-    },
-    {
-      "epoch": 0.24613491269902316,
-      "grad_norm": 2.4416730403900146,
-      "learning_rate": 0.00018600628522452798,
-      "loss": 2.6569,
-      "step": 3200
-    },
-    {
-      "epoch": 0.2499807707099454,
-      "grad_norm": 3.964357376098633,
-      "learning_rate": 0.00018574656520271146,
-      "loss": 2.4808,
-      "step": 3250
-    },
-    {
-      "epoch": 0.2499807707099454,
-      "eval_loss": 2.4578912258148193,
-      "eval_runtime": 17.8773,
-      "eval_samples_per_second": 55.937,
-      "eval_steps_per_second": 13.984,
-      "step": 3250
-    },
-    {
-      "epoch": 0.25382662872086764,
-      "grad_norm": 2.267664670944214,
-      "learning_rate": 0.000185486845180895,
-      "loss": 2.3763,
-      "step": 3300
-    },
-    {
-      "epoch": 0.2576724867317899,
-      "grad_norm": 2.5599284172058105,
-      "learning_rate": 0.00018522712515907852,
-      "loss": 2.4935,
-      "step": 3350
-    },
-    {
-      "epoch": 0.26151834474271207,
-      "grad_norm": 1.8299232721328735,
-      "learning_rate": 0.00018496740513726206,
-      "loss": 2.4674,
-      "step": 3400
-    },
-    {
-      "epoch": 0.2653642027536343,
-      "grad_norm": 1.7302981615066528,
-      "learning_rate": 0.00018470768511544557,
-      "loss": 2.4121,
-      "step": 3450
-    },
-    {
-      "epoch": 0.26921006076455656,
-      "grad_norm": 1.7371883392333984,
-      "learning_rate": 0.00018444796509362908,
-      "loss": 2.3909,
-      "step": 3500
-    },
-    {
-      "epoch": 0.26921006076455656,
-      "eval_loss": 2.427211046218872,
-      "eval_runtime": 17.8942,
-      "eval_samples_per_second": 55.884,
-      "eval_steps_per_second": 13.971,
-      "step": 3500
-    },
-    {
-      "epoch": 0.2730559187754788,
-      "grad_norm": 3.6324281692504883,
-      "learning_rate": 0.0001841882450718126,
-      "loss": 2.4369,
-      "step": 3550
-    },
-    {
-      "epoch": 0.27690177678640104,
-      "grad_norm": 3.5725560188293457,
-      "learning_rate": 0.0001839285250499961,
-      "loss": 2.4912,
-      "step": 3600
-    },
-    {
-      "epoch": 0.2807476347973233,
-      "grad_norm": 2.492825508117676,
-      "learning_rate": 0.00018366880502817962,
-      "loss": 2.4626,
-      "step": 3650
-    },
-    {
-      "epoch": 0.28459349280824553,
-      "grad_norm": 2.1992433071136475,
-      "learning_rate": 0.00018340908500636316,
-      "loss": 2.337,
-      "step": 3700
-    },
-    {
-      "epoch": 0.2884393508191678,
-      "grad_norm": 2.287579298019409,
-      "learning_rate": 0.00018314936498454667,
-      "loss": 2.4293,
-      "step": 3750
-    },
-    {
-      "epoch": 0.2884393508191678,
-      "eval_loss": 2.403160810470581,
-      "eval_runtime": 18.0987,
-      "eval_samples_per_second": 55.252,
-      "eval_steps_per_second": 13.813,
-      "step": 3750
-    },
-    {
-      "epoch": 0.29228520883009,
-      "grad_norm": 2.5130691528320312,
-      "learning_rate": 0.00018288964496273019,
-      "loss": 2.3591,
-      "step": 3800
-    },
-    {
-      "epoch": 0.2961310668410122,
-      "grad_norm": 2.376176357269287,
-      "learning_rate": 0.0001826299249409137,
-      "loss": 2.3887,
-      "step": 3850
-    },
-    {
-      "epoch": 0.29997692485193445,
-      "grad_norm": 2.7165868282318115,
-      "learning_rate": 0.0001823702049190972,
-      "loss": 2.4395,
-      "step": 3900
-    },
-    {
-      "epoch": 0.3038227828628567,
-      "grad_norm": 2.4704644680023193,
-      "learning_rate": 0.00018211048489728072,
-      "loss": 2.3314,
-      "step": 3950
-    },
-    {
-      "epoch": 0.30766864087377893,
-      "grad_norm": 1.2658872604370117,
-      "learning_rate": 0.00018185076487546426,
-      "loss": 2.3133,
-      "step": 4000
-    },
-    {
-      "epoch": 0.30766864087377893,
-      "eval_loss": 2.3614165782928467,
-      "eval_runtime": 17.9421,
-      "eval_samples_per_second": 55.735,
-      "eval_steps_per_second": 13.934,
-      "step": 4000
-    },
-    {
-      "epoch": 0.3115144988847012,
-      "grad_norm": 1.7358697652816772,
-      "learning_rate": 0.00018159104485364778,
-      "loss": 2.4004,
-      "step": 4050
-    },
-    {
-      "epoch": 0.3153603568956234,
-      "grad_norm": 2.5807759761810303,
-      "learning_rate": 0.0001813313248318313,
-      "loss": 2.3131,
-      "step": 4100
-    },
-    {
-      "epoch": 0.31920621490654566,
-      "grad_norm": 2.713397741317749,
-      "learning_rate": 0.00018107160481001483,
-      "loss": 2.3477,
-      "step": 4150
-    },
-    {
-      "epoch": 0.3230520729174679,
-      "grad_norm": 2.12007737159729,
-      "learning_rate": 0.00018081188478819832,
-      "loss": 2.3955,
-      "step": 4200
-    },
-    {
-      "epoch": 0.32689793092839015,
-      "grad_norm": 3.2964043617248535,
-      "learning_rate": 0.00018055216476638186,
-      "loss": 2.388,
-      "step": 4250
-    },
-    {
-      "epoch": 0.32689793092839015,
-      "eval_loss": 2.338472604751587,
-      "eval_runtime": 17.8233,
-      "eval_samples_per_second": 56.106,
-      "eval_steps_per_second": 14.027,
-      "step": 4250
-    },
-    {
-      "epoch": 0.33074378893931233,
-      "grad_norm": 1.9492484331130981,
-      "learning_rate": 0.00018029244474456537,
-      "loss": 2.3843,
-      "step": 4300
-    },
-    {
-      "epoch": 0.3345896469502346,
-      "grad_norm": 1.6252554655075073,
-      "learning_rate": 0.00018003272472274888,
-      "loss": 2.2891,
-      "step": 4350
-    },
-    {
-      "epoch": 0.3384355049611568,
-      "grad_norm": 2.1422605514526367,
-      "learning_rate": 0.00017977300470093242,
-      "loss": 2.3104,
-      "step": 4400
-    },
-    {
-      "epoch": 0.34228136297207906,
-      "grad_norm": 2.899003267288208,
-      "learning_rate": 0.00017951328467911593,
-      "loss": 2.2582,
-      "step": 4450
-    },
-    {
-      "epoch": 0.3461272209830013,
-      "grad_norm": 1.2482810020446777,
-      "learning_rate": 0.00017925356465729945,
-      "loss": 2.2296,
-      "step": 4500
-    },
-    {
-      "epoch": 0.3461272209830013,
-      "eval_loss": 2.2705459594726562,
-      "eval_runtime": 18.0389,
-      "eval_samples_per_second": 55.436,
-      "eval_steps_per_second": 13.859,
-      "step": 4500
-    },
-    {
-      "epoch": 0.34997307899392355,
-      "grad_norm": 2.5873026847839355,
-      "learning_rate": 0.00017899384463548296,
-      "loss": 2.3245,
-      "step": 4550
-    },
-    {
-      "epoch": 0.3538189370048458,
-      "grad_norm": 2.199103355407715,
-      "learning_rate": 0.00017873412461366647,
-      "loss": 2.322,
-      "step": 4600
-    },
-    {
-      "epoch": 0.35766479501576803,
-      "grad_norm": 2.3591790199279785,
-      "learning_rate": 0.00017847440459184999,
-      "loss": 2.3104,
-      "step": 4650
-    },
-    {
-      "epoch": 0.3615106530266903,
-      "grad_norm": 2.2885327339172363,
-      "learning_rate": 0.00017821468457003353,
-      "loss": 2.2148,
-      "step": 4700
-    },
-    {
-      "epoch": 0.36535651103761246,
-      "grad_norm": 1.8642834424972534,
-      "learning_rate": 0.00017795496454821704,
-      "loss": 2.2191,
-      "step": 4750
-    },
-    {
-      "epoch": 0.36535651103761246,
-      "eval_loss": 2.264026403427124,
-      "eval_runtime": 17.8717,
-      "eval_samples_per_second": 55.955,
-      "eval_steps_per_second": 13.989,
-      "step": 4750
-    },
-    {
-      "epoch": 0.3692023690485347,
-      "grad_norm": 1.7243143320083618,
-      "learning_rate": 0.00017769524452640055,
-      "loss": 2.2907,
-      "step": 4800
-    },
-    {
-      "epoch": 0.37304822705945695,
-      "grad_norm": 2.168203592300415,
-      "learning_rate": 0.00017743552450458406,
-      "loss": 2.2912,
-      "step": 4850
-    },
-    {
-      "epoch": 0.3768940850703792,
-      "grad_norm": 1.8623658418655396,
-      "learning_rate": 0.00017717580448276758,
-      "loss": 2.2809,
-      "step": 4900
-    },
-    {
-      "epoch": 0.38073994308130144,
-      "grad_norm": 1.9947059154510498,
-      "learning_rate": 0.0001769160844609511,
-      "loss": 2.2542,
-      "step": 4950
-    },
-    {
-      "epoch": 0.3845858010922237,
-      "grad_norm": 3.0198140144348145,
-      "learning_rate": 0.00017665636443913463,
-      "loss": 2.2352,
-      "step": 5000
-    },
-    {
-      "epoch": 0.3845858010922237,
-      "eval_loss": 2.2291696071624756,
-      "eval_runtime": 17.8435,
-      "eval_samples_per_second": 56.043,
-      "eval_steps_per_second": 14.011,
-      "step": 5000
-    },
-    {
-      "epoch": 0.3884316591031459,
-      "grad_norm": 1.8238661289215088,
-      "learning_rate": 0.00017639664441731814,
-      "loss": 2.3479,
-      "step": 5050
-    },
-    {
-      "epoch": 0.39227751711406816,
-      "grad_norm": 1.2968848943710327,
-      "learning_rate": 0.00017613692439550165,
-      "loss": 2.2648,
-      "step": 5100
-    },
-    {
-      "epoch": 0.3961233751249904,
-      "grad_norm": 3.036280393600464,
-      "learning_rate": 0.00017587720437368517,
-      "loss": 2.1473,
-      "step": 5150
-    },
-    {
-      "epoch": 0.39996923313591265,
-      "grad_norm": 1.7635585069656372,
-      "learning_rate": 0.00017561748435186868,
-      "loss": 2.2847,
-      "step": 5200
-    },
-    {
-      "epoch": 0.40381509114683484,
-      "grad_norm": 1.4038410186767578,
-      "learning_rate": 0.00017535776433005222,
-      "loss": 2.2621,
-      "step": 5250
-    },
-    {
-      "epoch": 0.40381509114683484,
-      "eval_loss": 2.214146614074707,
-      "eval_runtime": 17.7776,
-      "eval_samples_per_second": 56.25,
-      "eval_steps_per_second": 14.063,
-      "step": 5250
-    },
-    {
-      "epoch": 0.4076609491577571,
-      "grad_norm": 1.6948449611663818,
-      "learning_rate": 0.00017509804430823573,
-      "loss": 2.1945,
-      "step": 5300
-    },
-    {
-      "epoch": 0.4115068071686793,
-      "grad_norm": 2.7417380809783936,
-      "learning_rate": 0.00017483832428641925,
-      "loss": 2.2123,
-      "step": 5350
-    },
-    {
-      "epoch": 0.41535266517960157,
-      "grad_norm": 2.041032075881958,
-      "learning_rate": 0.00017457860426460279,
-      "loss": 2.2078,
-      "step": 5400
-    },
-    {
-      "epoch": 0.4191985231905238,
-      "grad_norm": 2.755319356918335,
-      "learning_rate": 0.00017431888424278627,
-      "loss": 2.1637,
-      "step": 5450
-    },
-    {
-      "epoch": 0.42304438120144605,
-      "grad_norm": 4.371281623840332,
-      "learning_rate": 0.00017405916422096978,
-      "loss": 2.3042,
-      "step": 5500
-    },
-    {
-      "epoch": 0.42304438120144605,
-      "eval_loss": 2.1678926944732666,
-      "eval_runtime": 17.7194,
-      "eval_samples_per_second": 56.435,
-      "eval_steps_per_second": 14.109,
-      "step": 5500
-    },
-    {
-      "epoch": 0.4268902392123683,
-      "grad_norm": 2.5908162593841553,
-      "learning_rate": 0.00017379944419915332,
-      "loss": 2.1524,
-      "step": 5550
-    },
-    {
-      "epoch": 0.43073609722329054,
-      "grad_norm": 2.3773181438446045,
-      "learning_rate": 0.00017353972417733684,
-      "loss": 2.1081,
-      "step": 5600
-    },
-    {
-      "epoch": 0.4345819552342128,
-      "grad_norm": 1.7582311630249023,
-      "learning_rate": 0.00017328000415552035,
-      "loss": 2.1667,
-      "step": 5650
-    },
-    {
-      "epoch": 0.43842781324513497,
-      "grad_norm": 2.112551689147949,
-      "learning_rate": 0.0001730202841337039,
-      "loss": 2.3053,
-      "step": 5700
-    },
-    {
-      "epoch": 0.4422736712560572,
-      "grad_norm": 1.6515388488769531,
-      "learning_rate": 0.0001727605641118874,
-      "loss": 2.1822,
-      "step": 5750
-    },
-    {
-      "epoch": 0.4422736712560572,
-      "eval_loss": 2.1589457988739014,
-      "eval_runtime": 18.0719,
-      "eval_samples_per_second": 55.335,
-      "eval_steps_per_second": 13.834,
-      "step": 5750
-    },
-    {
-      "epoch": 0.44611952926697945,
-      "grad_norm": 2.3416104316711426,
-      "learning_rate": 0.0001725008440900709,
-      "loss": 2.2417,
-      "step": 5800
-    },
-    {
-      "epoch": 0.4499653872779017,
-      "grad_norm": 2.031121015548706,
-      "learning_rate": 0.00017224112406825443,
-      "loss": 2.2381,
-      "step": 5850
-    },
-    {
-      "epoch": 0.45381124528882394,
-      "grad_norm": 1.880008339881897,
-      "learning_rate": 0.00017198140404643794,
-      "loss": 2.117,
-      "step": 5900
-    },
-    {
-      "epoch": 0.4576571032997462,
-      "grad_norm": 2.0337576866149902,
-      "learning_rate": 0.00017172168402462148,
-      "loss": 2.2118,
-      "step": 5950
-    },
-    {
-      "epoch": 0.4615029613106684,
-      "grad_norm": 2.1007370948791504,
-      "learning_rate": 0.000171461964002805,
-      "loss": 2.1743,
-      "step": 6000
-    },
-    {
-      "epoch": 0.4615029613106684,
-      "eval_loss": 2.130246877670288,
-      "eval_runtime": 17.7749,
-      "eval_samples_per_second": 56.259,
-      "eval_steps_per_second": 14.065,
-      "step": 6000
-    },
-    {
-      "epoch": 0.46534881932159067,
-      "grad_norm": 2.155015230178833,
-      "learning_rate": 0.0001712022439809885,
-      "loss": 2.1705,
-      "step": 6050
-    },
-    {
-      "epoch": 0.4691946773325129,
-      "grad_norm": 1.6260873079299927,
-      "learning_rate": 0.00017094252395917202,
-      "loss": 2.1626,
-      "step": 6100
-    },
-    {
-      "epoch": 0.4730405353434351,
-      "grad_norm": 1.473543405532837,
-      "learning_rate": 0.00017068280393735553,
-      "loss": 2.1957,
-      "step": 6150
-    },
-    {
-      "epoch": 0.47688639335435734,
-      "grad_norm": 1.728969931602478,
-      "learning_rate": 0.00017042308391553905,
-      "loss": 2.1864,
-      "step": 6200
-    },
-    {
-      "epoch": 0.4807322513652796,
-      "grad_norm": 1.990909218788147,
-      "learning_rate": 0.00017016336389372259,
-      "loss": 2.1385,
-      "step": 6250
-    },
-    {
-      "epoch": 0.4807322513652796,
-      "eval_loss": 2.0987420082092285,
-      "eval_runtime": 17.8525,
-      "eval_samples_per_second": 56.015,
-      "eval_steps_per_second": 14.004,
-      "step": 6250
-    },
-    {
-      "epoch": 0.4845781093762018,
-      "grad_norm": 1.9462237358093262,
-      "learning_rate": 0.0001699036438719061,
-      "loss": 2.1086,
-      "step": 6300
-    },
-    {
-      "epoch": 0.48842396738712407,
-      "grad_norm": 1.9223551750183105,
-      "learning_rate": 0.0001696439238500896,
-      "loss": 2.1302,
-      "step": 6350
-    },
-    {
-      "epoch": 0.4922698253980463,
-      "grad_norm": 2.4909427165985107,
-      "learning_rate": 0.00016938420382827312,
-      "loss": 2.2133,
-      "step": 6400
-    },
-    {
-      "epoch": 0.49611568340896856,
-      "grad_norm": 2.9876811504364014,
-      "learning_rate": 0.00016912448380645664,
-      "loss": 2.1354,
-      "step": 6450
-    },
-    {
-      "epoch": 0.4999615414198908,
-      "grad_norm": 1.96725594997406,
-      "learning_rate": 0.00016886476378464015,
-      "loss": 2.0744,
-      "step": 6500
-    },
-    {
-      "epoch": 0.4999615414198908,
-      "eval_loss": 2.083791494369507,
-      "eval_runtime": 17.9344,
-      "eval_samples_per_second": 55.759,
-      "eval_steps_per_second": 13.94,
-      "step": 6500
-    },
-    {
-      "epoch": 0.503807399430813,
-      "grad_norm": 2.44826602935791,
-      "learning_rate": 0.0001686050437628237,
-      "loss": 2.1554,
-      "step": 6550
-    },
-    {
-      "epoch": 0.5076532574417353,
-      "grad_norm": 3.768937826156616,
-      "learning_rate": 0.0001683453237410072,
-      "loss": 2.1123,
-      "step": 6600
-    },
-    {
-      "epoch": 0.5114991154526575,
-      "grad_norm": 1.6884543895721436,
-      "learning_rate": 0.00016808560371919072,
-      "loss": 2.0963,
-      "step": 6650
-    },
-    {
-      "epoch": 0.5153449734635798,
-      "grad_norm": 3.2465906143188477,
-      "learning_rate": 0.00016782588369737426,
-      "loss": 2.0452,
-      "step": 6700
-    },
-    {
-      "epoch": 0.519190831474502,
-      "grad_norm": 3.1067276000976562,
-      "learning_rate": 0.00016756616367555774,
-      "loss": 2.1565,
-      "step": 6750
-    },
-    {
-      "epoch": 0.519190831474502,
-      "eval_loss": 2.0547163486480713,
-      "eval_runtime": 17.7909,
-      "eval_samples_per_second": 56.208,
-      "eval_steps_per_second": 14.052,
-      "step": 6750
-    },
-    {
-      "epoch": 0.5230366894854241,
-      "grad_norm": 1.615390419960022,
-      "learning_rate": 0.00016730644365374128,
-      "loss": 2.0498,
-      "step": 6800
-    },
-    {
-      "epoch": 0.5268825474963464,
-      "grad_norm": 3.027007818222046,
-      "learning_rate": 0.0001670467236319248,
-      "loss": 2.1043,
-      "step": 6850
-    },
-    {
-      "epoch": 0.5307284055072686,
-      "grad_norm": 2.288503408432007,
-      "learning_rate": 0.0001667870036101083,
-      "loss": 2.019,
-      "step": 6900
-    },
-    {
-      "epoch": 0.5345742635181909,
-      "grad_norm": 2.088984727859497,
-      "learning_rate": 0.00016652728358829185,
-      "loss": 2.1001,
-      "step": 6950
-    },
-    {
-      "epoch": 0.5384201215291131,
-      "grad_norm": 1.8852577209472656,
-      "learning_rate": 0.00016626756356647536,
-      "loss": 2.0464,
-      "step": 7000
-    },
-    {
-      "epoch": 0.5384201215291131,
-      "eval_loss": 2.042011022567749,
-      "eval_runtime": 17.752,
-      "eval_samples_per_second": 56.332,
-      "eval_steps_per_second": 14.083,
-      "step": 7000
-    },
-    {
-      "epoch": 0.5422659795400354,
-      "grad_norm": 2.664232015609741,
-      "learning_rate": 0.00016600784354465885,
-      "loss": 2.068,
-      "step": 7050
-    },
-    {
-      "epoch": 0.5461118375509576,
-      "grad_norm": 1.5149424076080322,
-      "learning_rate": 0.00016574812352284239,
-      "loss": 2.1542,
-      "step": 7100
-    },
-    {
-      "epoch": 0.5499576955618799,
-      "grad_norm": 1.6383947134017944,
-      "learning_rate": 0.0001654884035010259,
-      "loss": 2.1201,
-      "step": 7150
-    },
-    {
-      "epoch": 0.5538035535728021,
-      "grad_norm": 2.146315097808838,
-      "learning_rate": 0.0001652286834792094,
-      "loss": 2.084,
-      "step": 7200
-    },
-    {
-      "epoch": 0.5576494115837243,
-      "grad_norm": 3.742957592010498,
-      "learning_rate": 0.00016496896345739295,
-      "loss": 2.1831,
-      "step": 7250
-    },
-    {
-      "epoch": 0.5576494115837243,
-      "eval_loss": 2.0231454372406006,
-      "eval_runtime": 17.88,
-      "eval_samples_per_second": 55.928,
-      "eval_steps_per_second": 13.982,
-      "step": 7250
-    },
-    {
-      "epoch": 0.5614952695946466,
-      "grad_norm": 2.808338165283203,
-      "learning_rate": 0.00016470924343557646,
-      "loss": 2.067,
-      "step": 7300
-    },
-    {
-      "epoch": 0.5653411276055688,
-      "grad_norm": 1.7112871408462524,
-      "learning_rate": 0.00016444952341375998,
-      "loss": 2.0396,
-      "step": 7350
-    },
-    {
-      "epoch": 0.5691869856164911,
-      "grad_norm": 1.9912537336349487,
-      "learning_rate": 0.0001641898033919435,
-      "loss": 2.0466,
-      "step": 7400
-    },
-    {
-      "epoch": 0.5730328436274132,
-      "grad_norm": 1.753312110900879,
-      "learning_rate": 0.000163930083370127,
-      "loss": 2.1138,
-      "step": 7450
-    },
-    {
-      "epoch": 0.5768787016383355,
-      "grad_norm": 1.4950071573257446,
-      "learning_rate": 0.00016367036334831052,
-      "loss": 2.1113,
-      "step": 7500
-    },
-    {
-      "epoch": 0.5768787016383355,
-      "eval_loss": 2.014023542404175,
-      "eval_runtime": 17.708,
-      "eval_samples_per_second": 56.472,
-      "eval_steps_per_second": 14.118,
-      "step": 7500
-    },
-    {
-      "epoch": 0.5807245596492577,
-      "grad_norm": 1.7512377500534058,
-      "learning_rate": 0.00016341064332649406,
-      "loss": 2.0485,
-      "step": 7550
-    },
-    {
-      "epoch": 0.58457041766018,
-      "grad_norm": 1.7805777788162231,
-      "learning_rate": 0.00016315092330467757,
-      "loss": 1.9163,
-      "step": 7600
-    },
-    {
-      "epoch": 0.5884162756711022,
-      "grad_norm": 2.1236133575439453,
-      "learning_rate": 0.00016289120328286108,
-      "loss": 2.1195,
-      "step": 7650
-    },
-    {
-      "epoch": 0.5922621336820244,
-      "grad_norm": 1.6518317461013794,
-      "learning_rate": 0.0001626314832610446,
-      "loss": 1.9555,
-      "step": 7700
-    },
-    {
-      "epoch": 0.5961079916929467,
-      "grad_norm": 6.697778701782227,
-      "learning_rate": 0.0001623717632392281,
-      "loss": 2.0895,
-      "step": 7750
-    },
-    {
-      "epoch": 0.5961079916929467,
-      "eval_loss": 1.9862221479415894,
-      "eval_runtime": 17.9057,
-      "eval_samples_per_second": 55.848,
-      "eval_steps_per_second": 13.962,
-      "step": 7750
-    },
-    {
-      "epoch": 0.5999538497038689,
-      "grad_norm": 1.7682440280914307,
-      "learning_rate": 0.00016211204321741165,
-      "loss": 2.0741,
-      "step": 7800
-    },
-    {
-      "epoch": 0.6037997077147912,
-      "grad_norm": 1.3762458562850952,
-      "learning_rate": 0.00016185232319559516,
-      "loss": 1.9891,
-      "step": 7850
-    },
-    {
-      "epoch": 0.6076455657257134,
-      "grad_norm": 1.802490234375,
-      "learning_rate": 0.00016159260317377867,
-      "loss": 1.9971,
-      "step": 7900
-    },
-    {
-      "epoch": 0.6114914237366357,
-      "grad_norm": 1.8477146625518799,
-      "learning_rate": 0.0001613328831519622,
-      "loss": 2.0318,
-      "step": 7950
-    },
-    {
-      "epoch": 0.6153372817475579,
-      "grad_norm": 2.1450252532958984,
-      "learning_rate": 0.0001610731631301457,
-      "loss": 2.1171,
-      "step": 8000
-    },
-    {
-      "epoch": 0.6153372817475579,
-      "eval_loss": 1.9715532064437866,
-      "eval_runtime": 17.7264,
-      "eval_samples_per_second": 56.413,
-      "eval_steps_per_second": 14.103,
-      "step": 8000
-    },
-    {
-      "epoch": 0.6191831397584802,
-      "grad_norm": 2.724039077758789,
-      "learning_rate": 0.0001608134431083292,
-      "loss": 1.9564,
-      "step": 8050
-    },
-    {
-      "epoch": 0.6230289977694023,
-      "grad_norm": 2.1357760429382324,
-      "learning_rate": 0.00016055372308651275,
-      "loss": 2.044,
-      "step": 8100
-    },
-    {
-      "epoch": 0.6268748557803245,
-      "grad_norm": 2.747109889984131,
-      "learning_rate": 0.00016029400306469626,
-      "loss": 2.0157,
-      "step": 8150
-    },
-    {
-      "epoch": 0.6307207137912468,
-      "grad_norm": 2.979858160018921,
-      "learning_rate": 0.00016003428304287978,
-      "loss": 2.0219,
-      "step": 8200
-    },
-    {
-      "epoch": 0.634566571802169,
-      "grad_norm": 1.5936884880065918,
-      "learning_rate": 0.00015977456302106332,
-      "loss": 1.9364,
-      "step": 8250
-    },
-    {
-      "epoch": 0.634566571802169,
-      "eval_loss": 1.9464218616485596,
-      "eval_runtime": 17.7894,
-      "eval_samples_per_second": 56.213,
-      "eval_steps_per_second": 14.053,
-      "step": 8250
-    },
-    {
-      "epoch": 0.6384124298130913,
-      "grad_norm": 2.502946376800537,
-      "learning_rate": 0.00015951484299924683,
-      "loss": 1.9746,
-      "step": 8300
-    },
-    {
-      "epoch": 0.6422582878240135,
-      "grad_norm": 1.307023048400879,
-      "learning_rate": 0.00015925512297743034,
-      "loss": 1.9936,
-      "step": 8350
-    },
-    {
-      "epoch": 0.6461041458349358,
-      "grad_norm": 2.037578582763672,
-      "learning_rate": 0.00015899540295561385,
-      "loss": 1.9515,
-      "step": 8400
-    },
-    {
-      "epoch": 0.649950003845858,
-      "grad_norm": 1.4244835376739502,
-      "learning_rate": 0.00015873568293379737,
-      "loss": 1.9988,
-      "step": 8450
-    },
-    {
-      "epoch": 0.6537958618567803,
-      "grad_norm": 2.5362133979797363,
-      "learning_rate": 0.0001584759629119809,
-      "loss": 2.1192,
-      "step": 8500
-    },
-    {
-      "epoch": 0.6537958618567803,
-      "eval_loss": 1.9391300678253174,
-      "eval_runtime": 17.9326,
-      "eval_samples_per_second": 55.764,
-      "eval_steps_per_second": 13.941,
-      "step": 8500
-    },
-    {
-      "epoch": 0.6576417198677025,
-      "grad_norm": 1.2576918601989746,
-      "learning_rate": 0.00015821624289016442,
-      "loss": 1.9734,
-      "step": 8550
-    },
-    {
-      "epoch": 0.6614875778786247,
-      "grad_norm": 2.036285400390625,
-      "learning_rate": 0.00015795652286834793,
-      "loss": 1.9899,
-      "step": 8600
-    },
-    {
-      "epoch": 0.665333435889547,
-      "grad_norm": 1.6453627347946167,
-      "learning_rate": 0.00015769680284653145,
-      "loss": 1.9238,
-      "step": 8650
-    },
-    {
-      "epoch": 0.6691792939004692,
-      "grad_norm": 1.4801263809204102,
-      "learning_rate": 0.00015743708282471496,
-      "loss": 1.9491,
-      "step": 8700
-    },
-    {
-      "epoch": 0.6730251519113915,
-      "grad_norm": 2.176879405975342,
-      "learning_rate": 0.00015717736280289847,
-      "loss": 1.9754,
-      "step": 8750
-    },
-    {
-      "epoch": 0.6730251519113915,
-      "eval_loss": 1.9268388748168945,
-      "eval_runtime": 17.7426,
-      "eval_samples_per_second": 56.362,
-      "eval_steps_per_second": 14.09,
-      "step": 8750
-    },
-    {
-      "epoch": 0.6768710099223136,
-      "grad_norm": 2.347804546356201,
-      "learning_rate": 0.000156917642781082,
-      "loss": 2.0034,
-      "step": 8800
-    },
-    {
-      "epoch": 0.6807168679332359,
-      "grad_norm": 1.8260661363601685,
-      "learning_rate": 0.00015665792275926552,
-      "loss": 1.9851,
-      "step": 8850
-    },
-    {
-      "epoch": 0.6845627259441581,
-      "grad_norm": 2.697566509246826,
-      "learning_rate": 0.00015639820273744904,
-      "loss": 1.9259,
-      "step": 8900
-    },
-    {
-      "epoch": 0.6884085839550804,
-      "grad_norm": 2.5771334171295166,
-      "learning_rate": 0.00015613848271563255,
-      "loss": 1.9216,
-      "step": 8950
-    },
-    {
-      "epoch": 0.6922544419660026,
-      "grad_norm": 1.098548173904419,
-      "learning_rate": 0.00015587876269381606,
-      "loss": 1.8965,
-      "step": 9000
-    },
-    {
-      "epoch": 0.6922544419660026,
-      "eval_loss": 1.900924801826477,
-      "eval_runtime": 17.6818,
-      "eval_samples_per_second": 56.555,
-      "eval_steps_per_second": 14.139,
-      "step": 9000
-    },
-    {
-      "epoch": 0.6961002999769248,
-      "grad_norm": 3.657864809036255,
-      "learning_rate": 0.00015561904267199958,
-      "loss": 1.95,
-      "step": 9050
-    },
-    {
-      "epoch": 0.6999461579878471,
-      "grad_norm": 1.5260729789733887,
-      "learning_rate": 0.00015535932265018312,
-      "loss": 1.9086,
-      "step": 9100
-    },
-    {
-      "epoch": 0.7037920159987693,
-      "grad_norm": 1.523361086845398,
-      "learning_rate": 0.00015509960262836663,
-      "loss": 1.8866,
-      "step": 9150
-    },
-    {
-      "epoch": 0.7076378740096916,
-      "grad_norm": 2.4977364540100098,
-      "learning_rate": 0.00015483988260655017,
-      "loss": 1.9033,
-      "step": 9200
-    },
-    {
-      "epoch": 0.7114837320206138,
-      "grad_norm": 2.8058366775512695,
-      "learning_rate": 0.00015458016258473365,
-      "loss": 1.9326,
-      "step": 9250
-    },
-    {
-      "epoch": 0.7114837320206138,
-      "eval_loss": 1.9121650457382202,
-      "eval_runtime": 17.715,
-      "eval_samples_per_second": 56.449,
-      "eval_steps_per_second": 14.112,
-      "step": 9250
-    },
-    {
-      "epoch": 0.7153295900315361,
-      "grad_norm": 2.2067878246307373,
-      "learning_rate": 0.00015432044256291717,
-      "loss": 1.9433,
-      "step": 9300
-    },
-    {
-      "epoch": 0.7191754480424583,
-      "grad_norm": 2.8370649814605713,
-      "learning_rate": 0.0001540607225411007,
-      "loss": 1.8972,
-      "step": 9350
-    },
-    {
-      "epoch": 0.7230213060533806,
-      "grad_norm": 1.707829236984253,
-      "learning_rate": 0.00015380100251928422,
-      "loss": 1.9328,
-      "step": 9400
-    },
-    {
-      "epoch": 0.7268671640643027,
-      "grad_norm": 2.2784628868103027,
-      "learning_rate": 0.00015354128249746773,
-      "loss": 1.9882,
-      "step": 9450
-    },
-    {
-      "epoch": 0.7307130220752249,
-      "grad_norm": 0.9409534931182861,
-      "learning_rate": 0.00015328156247565127,
-      "loss": 1.865,
-      "step": 9500
-    },
-    {
-      "epoch": 0.7307130220752249,
-      "eval_loss": 1.8746944665908813,
-      "eval_runtime": 17.5929,
-      "eval_samples_per_second": 56.841,
-      "eval_steps_per_second": 14.21,
-      "step": 9500
-    },
-    {
-      "epoch": 0.7345588800861472,
-      "grad_norm": 1.6800929307937622,
-      "learning_rate": 0.00015302184245383479,
-      "loss": 1.8351,
-      "step": 9550
-    },
-    {
-      "epoch": 0.7384047380970694,
-      "grad_norm": 2.518232822418213,
-      "learning_rate": 0.00015276212243201827,
-      "loss": 1.9335,
-      "step": 9600
-    },
-    {
-      "epoch": 0.7422505961079917,
-      "grad_norm": 1.3104406595230103,
-      "learning_rate": 0.0001525024024102018,
-      "loss": 1.9058,
-      "step": 9650
-    },
-    {
-      "epoch": 0.7460964541189139,
-      "grad_norm": 1.8274422883987427,
-      "learning_rate": 0.00015224268238838532,
-      "loss": 1.8246,
-      "step": 9700
-    },
-    {
-      "epoch": 0.7499423121298362,
-      "grad_norm": 1.5810478925704956,
-      "learning_rate": 0.00015198296236656884,
-      "loss": 1.9504,
-      "step": 9750
-    },
-    {
-      "epoch": 0.7499423121298362,
-      "eval_loss": 1.8743510246276855,
-      "eval_runtime": 17.783,
-      "eval_samples_per_second": 56.233,
-      "eval_steps_per_second": 14.058,
-      "step": 9750
-    },
-    {
-      "epoch": 0.7537881701407584,
-      "grad_norm": 1.7910431623458862,
-      "learning_rate": 0.00015172324234475238,
-      "loss": 1.8525,
-      "step": 9800
-    },
-    {
-      "epoch": 0.7576340281516807,
-      "grad_norm": 0.7378084063529968,
-      "learning_rate": 0.0001514635223229359,
-      "loss": 1.8688,
-      "step": 9850
-    },
-    {
-      "epoch": 0.7614798861626029,
-      "grad_norm": 1.357282280921936,
-      "learning_rate": 0.0001512038023011194,
-      "loss": 1.855,
-      "step": 9900
-    },
-    {
-      "epoch": 0.7653257441735252,
-      "grad_norm": 2.7199490070343018,
-      "learning_rate": 0.00015094408227930292,
-      "loss": 1.8655,
-      "step": 9950
-    },
-    {
-      "epoch": 0.7691716021844474,
-      "grad_norm": 1.9351532459259033,
-      "learning_rate": 0.00015068436225748643,
-      "loss": 1.9272,
-      "step": 10000
-    },
-    {
-      "epoch": 0.7691716021844474,
-      "eval_loss": 1.8518301248550415,
-      "eval_runtime": 17.6154,
-      "eval_samples_per_second": 56.769,
-      "eval_steps_per_second": 14.192,
-      "step": 10000
-    },
-    {
-      "epoch": 0.7730174601953695,
-      "grad_norm": 1.4824655055999756,
-      "learning_rate": 0.00015042464223566997,
-      "loss": 1.9359,
-      "step": 10050
-    },
-    {
-      "epoch": 0.7768633182062918,
-      "grad_norm": 1.8489737510681152,
-      "learning_rate": 0.00015016492221385348,
-      "loss": 1.7524,
-      "step": 10100
-    },
-    {
-      "epoch": 0.780709176217214,
-      "grad_norm": 1.8048781156539917,
-      "learning_rate": 0.000149905202192037,
-      "loss": 1.9209,
-      "step": 10150
-    },
-    {
-      "epoch": 0.7845550342281363,
-      "grad_norm": 1.5031684637069702,
-      "learning_rate": 0.0001496454821702205,
-      "loss": 1.8224,
-      "step": 10200
-    },
-    {
-      "epoch": 0.7884008922390585,
-      "grad_norm": 3.326998710632324,
-      "learning_rate": 0.00014938576214840402,
-      "loss": 1.8585,
-      "step": 10250
-    },
-    {
-      "epoch": 0.7884008922390585,
-      "eval_loss": 1.840669870376587,
-      "eval_runtime": 17.7052,
-      "eval_samples_per_second": 56.48,
-      "eval_steps_per_second": 14.12,
-      "step": 10250
-    },
-    {
-      "epoch": 0.7922467502499808,
-      "grad_norm": 1.9124550819396973,
-      "learning_rate": 0.00014912604212658753,
-      "loss": 1.8978,
-      "step": 10300
-    },
-    {
-      "epoch": 0.796092608260903,
-      "grad_norm": 1.2113343477249146,
-      "learning_rate": 0.00014886632210477107,
-      "loss": 1.9814,
-      "step": 10350
-    },
-    {
-      "epoch": 0.7999384662718253,
-      "grad_norm": 3.08606219291687,
-      "learning_rate": 0.00014860660208295459,
-      "loss": 1.7926,
-      "step": 10400
-    },
-    {
-      "epoch": 0.8037843242827475,
-      "grad_norm": 3.1372311115264893,
-      "learning_rate": 0.0001483468820611381,
-      "loss": 1.9039,
-      "step": 10450
-    },
-    {
-      "epoch": 0.8076301822936697,
-      "grad_norm": 2.0522525310516357,
-      "learning_rate": 0.00014808716203932164,
-      "loss": 1.91,
-      "step": 10500
-    },
-    {
-      "epoch": 0.8076301822936697,
-      "eval_loss": 1.8211458921432495,
-      "eval_runtime": 17.6697,
-      "eval_samples_per_second": 56.594,
-      "eval_steps_per_second": 14.149,
-      "step": 10500
-    },
-    {
-      "epoch": 0.811476040304592,
-      "grad_norm": 2.5157995223999023,
-      "learning_rate": 0.00014782744201750512,
-      "loss": 1.9747,
-      "step": 10550
-    },
-    {
-      "epoch": 0.8153218983155142,
-      "grad_norm": 1.821583867073059,
-      "learning_rate": 0.00014756772199568864,
-      "loss": 1.8858,
-      "step": 10600
-    },
-    {
-      "epoch": 0.8191677563264365,
-      "grad_norm": 1.9474884271621704,
-      "learning_rate": 0.00014730800197387218,
-      "loss": 1.7796,
-      "step": 10650
-    },
-    {
-      "epoch": 0.8230136143373586,
-      "grad_norm": 1.889302372932434,
-      "learning_rate": 0.0001470482819520557,
-      "loss": 2.0,
-      "step": 10700
-    },
-    {
-      "epoch": 0.8268594723482809,
-      "grad_norm": 2.596470594406128,
-      "learning_rate": 0.0001467885619302392,
-      "loss": 1.8673,
-      "step": 10750
-    },
-    {
-      "epoch": 0.8268594723482809,
-      "eval_loss": 1.821048617362976,
-      "eval_runtime": 17.7145,
-      "eval_samples_per_second": 56.451,
-      "eval_steps_per_second": 14.113,
-      "step": 10750
-    },
-    {
-      "epoch": 0.8307053303592031,
-      "grad_norm": 1.4734866619110107,
-      "learning_rate": 0.00014652884190842274,
-      "loss": 1.794,
-      "step": 10800
-    },
-    {
-      "epoch": 0.8345511883701254,
-      "grad_norm": 2.4944119453430176,
-      "learning_rate": 0.00014626912188660625,
-      "loss": 1.8974,
-      "step": 10850
-    },
-    {
-      "epoch": 0.8383970463810476,
-      "grad_norm": 2.175258159637451,
-      "learning_rate": 0.00014600940186478977,
-      "loss": 1.9176,
-      "step": 10900
-    },
-    {
-      "epoch": 0.8422429043919698,
-      "grad_norm": 1.6867458820343018,
-      "learning_rate": 0.00014574968184297328,
-      "loss": 1.8364,
-      "step": 10950
-    },
-    {
-      "epoch": 0.8460887624028921,
-      "grad_norm": 1.8447539806365967,
-      "learning_rate": 0.0001454899618211568,
-      "loss": 1.7942,
-      "step": 11000
-    },
-    {
-      "epoch": 0.8460887624028921,
-      "eval_loss": 1.7994624376296997,
-      "eval_runtime": 17.4954,
-      "eval_samples_per_second": 57.158,
-      "eval_steps_per_second": 14.289,
-      "step": 11000
-    },
-    {
-      "epoch": 0.8499346204138143,
-      "grad_norm": 2.359623432159424,
-      "learning_rate": 0.00014523024179934033,
-      "loss": 1.8082,
-      "step": 11050
-    },
-    {
-      "epoch": 0.8537804784247366,
-      "grad_norm": 1.690730333328247,
-      "learning_rate": 0.00014497052177752385,
-      "loss": 1.9029,
-      "step": 11100
-    },
-    {
-      "epoch": 0.8576263364356588,
-      "grad_norm": 1.5211533308029175,
-      "learning_rate": 0.00014471080175570736,
-      "loss": 1.836,
-      "step": 11150
-    },
-    {
-      "epoch": 0.8614721944465811,
-      "grad_norm": 2.294943332672119,
-      "learning_rate": 0.00014445108173389087,
-      "loss": 1.8554,
-      "step": 11200
-    },
-    {
-      "epoch": 0.8653180524575033,
-      "grad_norm": 1.7020373344421387,
-      "learning_rate": 0.00014419136171207438,
-      "loss": 1.8487,
-      "step": 11250
-    },
-    {
-      "epoch": 0.8653180524575033,
-      "eval_loss": 1.8003779649734497,
-      "eval_runtime": 17.5162,
-      "eval_samples_per_second": 57.09,
-      "eval_steps_per_second": 14.272,
-      "step": 11250
-    },
-    {
-      "epoch": 0.8691639104684256,
-      "grad_norm": 1.6584821939468384,
-      "learning_rate": 0.0001439316416902579,
-      "loss": 1.8169,
-      "step": 11300
-    },
-    {
-      "epoch": 0.8730097684793477,
-      "grad_norm": 1.5835644006729126,
-      "learning_rate": 0.00014367192166844144,
-      "loss": 1.8344,
-      "step": 11350
-    },
-    {
-      "epoch": 0.8768556264902699,
-      "grad_norm": 1.37686026096344,
-      "learning_rate": 0.00014341220164662495,
-      "loss": 1.839,
-      "step": 11400
-    },
-    {
-      "epoch": 0.8807014845011922,
-      "grad_norm": 1.2032794952392578,
-      "learning_rate": 0.00014315248162480846,
-      "loss": 1.8392,
-      "step": 11450
-    },
-    {
-      "epoch": 0.8845473425121144,
-      "grad_norm": 2.025676727294922,
-      "learning_rate": 0.00014289276160299198,
-      "loss": 1.7939,
-      "step": 11500
-    },
-    {
-      "epoch": 0.8845473425121144,
-      "eval_loss": 1.7949305772781372,
-      "eval_runtime": 17.7014,
-      "eval_samples_per_second": 56.493,
-      "eval_steps_per_second": 14.123,
-      "step": 11500
-    },
-    {
-      "epoch": 0.8883932005230367,
-      "grad_norm": 2.710430383682251,
-      "learning_rate": 0.0001426330415811755,
-      "loss": 1.8816,
-      "step": 11550
-    },
-    {
-      "epoch": 0.8922390585339589,
-      "grad_norm": 1.4110090732574463,
-      "learning_rate": 0.000142373321559359,
-      "loss": 1.8762,
-      "step": 11600
-    },
-    {
-      "epoch": 0.8960849165448812,
-      "grad_norm": 1.4623394012451172,
-      "learning_rate": 0.00014211360153754254,
-      "loss": 1.8463,
-      "step": 11650
-    },
-    {
-      "epoch": 0.8999307745558034,
-      "grad_norm": 1.2035316228866577,
-      "learning_rate": 0.00014185388151572605,
-      "loss": 1.8143,
-      "step": 11700
-    },
-    {
-      "epoch": 0.9037766325667257,
-      "grad_norm": 3.2838220596313477,
-      "learning_rate": 0.0001415941614939096,
-      "loss": -16.199,
-      "step": 11750
-    },
-    {
-      "epoch": 0.9037766325667257,
-      "eval_loss": -26.813920974731445,
-      "eval_runtime": 17.131,
-      "eval_samples_per_second": 58.374,
-      "eval_steps_per_second": 14.593,
-      "step": 11750
-    },
-    {
-      "epoch": 0.9076224905776479,
-      "grad_norm": 9.025236129760742,
-      "learning_rate": 0.00014133963587252942,
-      "loss": -28.4398,
-      "step": 11800
-    },
-    {
-      "epoch": 0.9114683485885701,
-      "grad_norm": 9.17872428894043,
-      "learning_rate": 0.00014109030465158558,
-      "loss": -29.3786,
-      "step": 11850
-    },
-    {
-      "epoch": 0.9153142065994924,
-      "grad_norm": 13.099103927612305,
-      "learning_rate": 0.00014083058462976912,
-      "loss": -29.5958,
-      "step": 11900
-    },
-    {
-      "epoch": 0.9191600646104146,
-      "grad_norm": 7.575841426849365,
-      "learning_rate": 0.00014057086460795263,
-      "loss": -29.7462,
-      "step": 11950
-    },
-    {
-      "epoch": 0.9230059226213368,
-      "grad_norm": 8.827557563781738,
-      "learning_rate": 0.00014031114458613615,
-      "loss": -29.9022,
-      "step": 12000
-    },
-    {
-      "epoch": 0.9230059226213368,
-      "eval_loss": -29.983943939208984,
-      "eval_runtime": 17.1961,
-      "eval_samples_per_second": 58.153,
-      "eval_steps_per_second": 14.538,
-      "step": 12000
-    },
-    {
-      "epoch": 0.926851780632259,
-      "grad_norm": 9.088982582092285,
-      "learning_rate": 0.0001400514245643197,
-      "loss": -29.9348,
-      "step": 12050
-    },
-    {
-      "epoch": 0.9306976386431813,
-      "grad_norm": 3.507662534713745,
-      "learning_rate": 0.0001397917045425032,
-      "loss": -30.0066,
-      "step": 12100
-    },
-    {
-      "epoch": 0.9345434966541035,
-      "grad_norm": 5.43391227722168,
-      "learning_rate": 0.00013953198452068669,
-      "loss": -30.0515,
-      "step": 12150
-    },
-    {
-      "epoch": 0.9383893546650258,
-      "grad_norm": 6.1119704246521,
-      "learning_rate": 0.00013927226449887023,
-      "loss": -30.0768,
-      "step": 12200
-    },
-    {
-      "epoch": 0.942235212675948,
-      "grad_norm": 4.298177242279053,
-      "learning_rate": 0.00013901254447705374,
-      "loss": -30.0727,
-      "step": 12250
-    },
-    {
-      "epoch": 0.942235212675948,
-      "eval_loss": -30.140077590942383,
-      "eval_runtime": 17.3124,
-      "eval_samples_per_second": 57.762,
-      "eval_steps_per_second": 14.441,
-      "step": 12250
-    },
-    {
-      "epoch": 0.9460810706868702,
-      "grad_norm": 5.769038677215576,
-      "learning_rate": 0.00013875282445523725,
-      "loss": -30.0775,
-      "step": 12300
-    },
-    {
-      "epoch": 0.9499269286977925,
-      "grad_norm": 5.405306816101074,
-      "learning_rate": 0.0001384931044334208,
-      "loss": -30.1151,
-      "step": 12350
-    },
-    {
-      "epoch": 0.9537727867087147,
-      "grad_norm": 7.666799545288086,
-      "learning_rate": 0.0001382333844116043,
-      "loss": -30.1243,
-      "step": 12400
-    },
-    {
-      "epoch": 0.957618644719637,
-      "grad_norm": 8.419403076171875,
-      "learning_rate": 0.00013797366438978782,
-      "loss": -30.1474,
-      "step": 12450
-    },
-    {
-      "epoch": 0.9614645027305592,
-      "grad_norm": 2.181037187576294,
-      "learning_rate": 0.00013771394436797133,
-      "loss": -30.1508,
-      "step": 12500
-    },
-    {
-      "epoch": 0.9614645027305592,
-      "eval_loss": -30.163711547851562,
-      "eval_runtime": 17.0325,
-      "eval_samples_per_second": 58.711,
-      "eval_steps_per_second": 14.678,
-      "step": 12500
-    },
-    {
-      "epoch": 0.9653103607414815,
-      "grad_norm": 8.547541618347168,
-      "learning_rate": 0.00013745422434615484,
-      "loss": -30.1494,
-      "step": 12550
-    },
-    {
-      "epoch": 0.9691562187524037,
-      "grad_norm": 4.740926265716553,
-      "learning_rate": 0.00013719450432433838,
-      "loss": -30.1558,
-      "step": 12600
-    },
-    {
-      "epoch": 0.973002076763326,
-      "grad_norm": 5.869479179382324,
-      "learning_rate": 0.0001369347843025219,
-      "loss": -30.1616,
-      "step": 12650
-    },
-    {
-      "epoch": 0.9768479347742481,
-      "grad_norm": 7.4348673820495605,
-      "learning_rate": 0.0001366750642807054,
-      "loss": -30.1561,
-      "step": 12700
-    },
-    {
-      "epoch": 0.9806937927851703,
-      "grad_norm": 8.506593704223633,
-      "learning_rate": 0.00013641534425888892,
-      "loss": -30.1737,
-      "step": 12750
-    },
-    {
-      "epoch": 0.9806937927851703,
-      "eval_loss": -30.198650360107422,
-      "eval_runtime": 16.9999,
-      "eval_samples_per_second": 58.824,
-      "eval_steps_per_second": 14.706,
-      "step": 12750
-    },
-    {
-      "epoch": 0.9845396507960926,
-      "grad_norm": 5.528530120849609,
-      "learning_rate": 0.00013615562423707243,
-      "loss": -30.1728,
-      "step": 12800
-    },
-    {
-      "epoch": 0.9883855088070148,
-      "grad_norm": 3.630002021789551,
-      "learning_rate": 0.00013589590421525595,
-      "loss": -30.1832,
-      "step": 12850
-    },
-    {
-      "epoch": 0.9922313668179371,
-      "grad_norm": 5.898171424865723,
-      "learning_rate": 0.00013563618419343949,
-      "loss": -30.1849,
-      "step": 12900
-    },
-    {
-      "epoch": 0.9960772248288593,
-      "grad_norm": 5.75803279876709,
-      "learning_rate": 0.000135376464171623,
-      "loss": -30.1971,
-      "step": 12950
-    },
-    {
-      "epoch": 0.9999230828397816,
-      "grad_norm": 8.700019836425781,
-      "learning_rate": 0.0001351167441498065,
-      "loss": -30.2117,
-      "step": 13000
-    },
-    {
-      "epoch": 0.9999230828397816,
-      "eval_loss": -30.213665008544922,
-      "eval_runtime": 17.0537,
-      "eval_samples_per_second": 58.638,
-      "eval_steps_per_second": 14.66,
-      "step": 13000
-    },
-    {
-      "epoch": 1.0037689408507038,
-      "grad_norm": 6.23037052154541,
-      "learning_rate": 0.00013485702412799005,
-      "loss": -30.1957,
-      "step": 13050
-    },
-    {
-      "epoch": 1.007614798861626,
-      "grad_norm": 4.10649299621582,
-      "learning_rate": 0.00013459730410617354,
-      "loss": -30.2093,
-      "step": 13100
-    },
-    {
-      "epoch": 1.0114606568725482,
-      "grad_norm": 6.44740629196167,
-      "learning_rate": 0.00013433758408435705,
-      "loss": -30.1925,
-      "step": 13150
-    },
-    {
-      "epoch": 1.0153065148834706,
-      "grad_norm": 3.919144630432129,
-      "learning_rate": 0.0001340778640625406,
-      "loss": -30.2131,
-      "step": 13200
-    },
-    {
-      "epoch": 1.0191523728943928,
-      "grad_norm": 3.7862091064453125,
-      "learning_rate": 0.0001338181440407241,
-      "loss": -30.2237,
-      "step": 13250
-    },
-    {
-      "epoch": 1.0191523728943928,
-      "eval_loss": -30.221071243286133,
-      "eval_runtime": 17.0391,
-      "eval_samples_per_second": 58.688,
-      "eval_steps_per_second": 14.672,
-      "step": 13250
-    },
-    {
-      "epoch": 1.022998230905315,
-      "grad_norm": 2.9611716270446777,
-      "learning_rate": 0.00013355842401890762,
-      "loss": -30.2087,
-      "step": 13300
-    },
-    {
-      "epoch": 1.0268440889162371,
-      "grad_norm": 3.5864665508270264,
-      "learning_rate": 0.00013329870399709116,
-      "loss": -30.2231,
-      "step": 13350
-    },
-    {
-      "epoch": 1.0306899469271595,
-      "grad_norm": 4.609787940979004,
-      "learning_rate": 0.00013304417837571098,
-      "loss": -30.2137,
-      "step": 13400
-    },
-    {
-      "epoch": 1.0345358049380817,
-      "grad_norm": 13.944046974182129,
-      "learning_rate": 0.0001327844583538945,
-      "loss": -30.2348,
-      "step": 13450
-    },
-    {
-      "epoch": 1.038381662949004,
-      "grad_norm": 3.965855836868286,
-      "learning_rate": 0.00013252473833207803,
-      "loss": -30.2325,
-      "step": 13500
-    },
-    {
-      "epoch": 1.038381662949004,
-      "eval_loss": -30.246578216552734,
-      "eval_runtime": 17.0659,
-      "eval_samples_per_second": 58.596,
-      "eval_steps_per_second": 14.649,
-      "step": 13500
-    },
-    {
-      "epoch": 1.042227520959926,
-      "grad_norm": 11.001680374145508,
-      "learning_rate": 0.00013226501831026155,
-      "loss": -30.2413,
-      "step": 13550
-    },
-    {
-      "epoch": 1.0460733789708483,
-      "grad_norm": 6.924397945404053,
-      "learning_rate": 0.00013200529828844506,
-      "loss": -30.2411,
-      "step": 13600
-    },
-    {
-      "epoch": 1.0499192369817707,
-      "grad_norm": 4.740690231323242,
-      "learning_rate": 0.0001317455782666286,
-      "loss": -30.2412,
-      "step": 13650
-    },
-    {
-      "epoch": 1.0537650949926929,
-      "grad_norm": 6.716350078582764,
-      "learning_rate": 0.0001314858582448121,
-      "loss": -30.243,
-      "step": 13700
-    },
-    {
-      "epoch": 1.057610953003615,
-      "grad_norm": 3.5048656463623047,
-      "learning_rate": 0.0001312261382229956,
-      "loss": -30.2447,
-      "step": 13750
-    },
-    {
-      "epoch": 1.057610953003615,
-      "eval_loss": -30.26629066467285,
-      "eval_runtime": 17.0438,
-      "eval_samples_per_second": 58.672,
-      "eval_steps_per_second": 14.668,
-      "step": 13750
-    },
-    {
-      "epoch": 1.0614568110145373,
-      "grad_norm": 12.012964248657227,
-      "learning_rate": 0.00013096641820117914,
-      "loss": -30.2578,
-      "step": 13800
-    },
-    {
-      "epoch": 1.0653026690254597,
-      "grad_norm": 5.072702407836914,
-      "learning_rate": 0.00013070669817936265,
-      "loss": -30.2674,
-      "step": 13850
-    },
-    {
-      "epoch": 1.0691485270363819,
-      "grad_norm": 6.612403392791748,
-      "learning_rate": 0.00013044697815754616,
-      "loss": -30.281,
-      "step": 13900
-    },
-    {
-      "epoch": 1.072994385047304,
-      "grad_norm": 5.599041938781738,
-      "learning_rate": 0.0001301872581357297,
-      "loss": -30.2823,
-      "step": 13950
-    },
-    {
-      "epoch": 1.0768402430582262,
-      "grad_norm": 3.985121726989746,
-      "learning_rate": 0.00012992753811391322,
-      "loss": -30.2907,
-      "step": 14000
-    },
-    {
-      "epoch": 1.0768402430582262,
-      "eval_loss": -30.32036781311035,
-      "eval_runtime": 16.9552,
-      "eval_samples_per_second": 58.979,
-      "eval_steps_per_second": 14.745,
-      "step": 14000
-    },
-    {
-      "epoch": 1.0806861010691486,
-      "grad_norm": 8.776793479919434,
-      "learning_rate": 0.00012966781809209673,
-      "loss": -30.2985,
-      "step": 14050
-    },
-    {
-      "epoch": 1.0845319590800708,
-      "grad_norm": 6.127963066101074,
-      "learning_rate": 0.00012940809807028024,
-      "loss": -30.3066,
-      "step": 14100
-    },
-    {
-      "epoch": 1.088377817090993,
-      "grad_norm": 7.999585151672363,
-      "learning_rate": 0.00012914837804846375,
-      "loss": -30.2914,
-      "step": 14150
-    },
-    {
-      "epoch": 1.0922236751019152,
-      "grad_norm": 7.283746719360352,
-      "learning_rate": 0.0001288886580266473,
-      "loss": -30.3119,
-      "step": 14200
-    },
-    {
-      "epoch": 1.0960695331128374,
-      "grad_norm": 10.978534698486328,
-      "learning_rate": 0.0001286289380048308,
-      "loss": -30.3062,
-      "step": 14250
-    },
-    {
-      "epoch": 1.0960695331128374,
-      "eval_loss": -30.304922103881836,
-      "eval_runtime": 16.9338,
-      "eval_samples_per_second": 59.054,
-      "eval_steps_per_second": 14.763,
-      "step": 14250
-    },
-    {
-      "epoch": 1.0999153911237598,
-      "grad_norm": 4.0659918785095215,
-      "learning_rate": 0.00012836921798301432,
-      "loss": -30.3197,
-      "step": 14300
-    },
-    {
-      "epoch": 1.103761249134682,
-      "grad_norm": 6.115564346313477,
-      "learning_rate": 0.00012810949796119783,
-      "loss": -30.3324,
-      "step": 14350
-    },
-    {
-      "epoch": 1.1076071071456042,
-      "grad_norm": 7.9282331466674805,
-      "learning_rate": 0.00012784977793938135,
-      "loss": -30.3401,
-      "step": 14400
-    },
-    {
-      "epoch": 1.1114529651565264,
-      "grad_norm": 5.920327663421631,
-      "learning_rate": 0.00012759005791756486,
-      "loss": -30.3405,
-      "step": 14450
-    },
-    {
-      "epoch": 1.1152988231674485,
-      "grad_norm": 3.5994579792022705,
-      "learning_rate": 0.0001273303378957484,
-      "loss": -30.3499,
-      "step": 14500
-    },
-    {
-      "epoch": 1.1152988231674485,
-      "eval_loss": -30.38400650024414,
-      "eval_runtime": 16.9883,
-      "eval_samples_per_second": 58.864,
-      "eval_steps_per_second": 14.716,
-      "step": 14500
-    },
-    {
-      "epoch": 1.119144681178371,
-      "grad_norm": 4.9496750831604,
-      "learning_rate": 0.0001270706178739319,
-      "loss": -30.352,
-      "step": 14550
-    },
-    {
-      "epoch": 1.1229905391892931,
-      "grad_norm": 4.976698875427246,
-      "learning_rate": 0.00012681089785211542,
-      "loss": -30.3553,
-      "step": 14600
-    },
-    {
-      "epoch": 1.1268363972002153,
-      "grad_norm": 3.139697790145874,
-      "learning_rate": 0.00012655117783029896,
-      "loss": -30.357,
-      "step": 14650
-    },
-    {
-      "epoch": 1.1306822552111375,
-      "grad_norm": 5.011180400848389,
-      "learning_rate": 0.00012629145780848245,
-      "loss": -30.3574,
-      "step": 14700
-    },
-    {
-      "epoch": 1.13452811322206,
-      "grad_norm": 8.093709945678711,
-      "learning_rate": 0.00012603173778666596,
-      "loss": -30.3587,
-      "step": 14750
-    },
-    {
-      "epoch": 1.13452811322206,
-      "eval_loss": -30.386066436767578,
-      "eval_runtime": 16.9522,
-      "eval_samples_per_second": 58.989,
-      "eval_steps_per_second": 14.747,
-      "step": 14750
-    },
-    {
-      "epoch": 1.1383739712329821,
-      "grad_norm": 4.983211040496826,
-      "learning_rate": 0.0001257720177648495,
-      "loss": -30.3616,
-      "step": 14800
-    },
-    {
-      "epoch": 1.1422198292439043,
-      "grad_norm": 3.838836431503296,
-      "learning_rate": 0.00012551229774303301,
-      "loss": -30.3639,
-      "step": 14850
-    },
-    {
-      "epoch": 1.1460656872548265,
-      "grad_norm": 53.798492431640625,
-      "learning_rate": 0.00012525257772121653,
-      "loss": -30.3554,
-      "step": 14900
-    },
-    {
-      "epoch": 1.149911545265749,
-      "grad_norm": 2.782118558883667,
-      "learning_rate": 0.00012499285769940007,
-      "loss": -30.3685,
-      "step": 14950
-    },
-    {
-      "epoch": 1.153757403276671,
-      "grad_norm": 4.757026195526123,
-      "learning_rate": 0.00012473313767758355,
-      "loss": -30.3738,
-      "step": 15000
-    },
-    {
-      "epoch": 1.153757403276671,
-      "eval_loss": -30.39963722229004,
-      "eval_runtime": 16.9126,
-      "eval_samples_per_second": 59.128,
-      "eval_steps_per_second": 14.782,
-      "step": 15000
-    },
-    {
-      "epoch": 1.1576032612875933,
-      "grad_norm": 5.459719657897949,
-      "learning_rate": 0.0001244734176557671,
-      "loss": -30.3726,
-      "step": 15050
-    },
-    {
-      "epoch": 1.1614491192985155,
-      "grad_norm": 5.796365737915039,
-      "learning_rate": 0.0001242136976339506,
-      "loss": -30.3743,
-      "step": 15100
-    },
-    {
-      "epoch": 1.1652949773094377,
-      "grad_norm": 8.96560287475586,
-      "learning_rate": 0.00012395397761213412,
-      "loss": -30.3713,
-      "step": 15150
-    },
-    {
-      "epoch": 1.16914083532036,
-      "grad_norm": 3.6820108890533447,
-      "learning_rate": 0.00012369425759031766,
-      "loss": -30.3737,
-      "step": 15200
-    },
-    {
-      "epoch": 1.1729866933312822,
-      "grad_norm": 4.876701831817627,
-      "learning_rate": 0.00012343453756850117,
-      "loss": -30.3783,
-      "step": 15250
-    },
-    {
-      "epoch": 1.1729866933312822,
-      "eval_loss": -30.406328201293945,
-      "eval_runtime": 17.0903,
-      "eval_samples_per_second": 58.513,
-      "eval_steps_per_second": 14.628,
-      "step": 15250
-    },
-    {
-      "epoch": 1.1768325513422044,
-      "grad_norm": 6.255270957946777,
-      "learning_rate": 0.00012317481754668468,
-      "loss": -30.3819,
-      "step": 15300
-    },
-    {
-      "epoch": 1.1806784093531266,
-      "grad_norm": 3.348052978515625,
-      "learning_rate": 0.00012293068072617719,
-      "loss": -30.0431,
-      "step": 15350
-    },
-    {
-      "epoch": 1.1845242673640488,
-      "grad_norm": 6.610827922821045,
-      "learning_rate": 0.0001226709607043607,
-      "loss": -30.3818,
-      "step": 15400
-    },
-    {
-      "epoch": 1.1883701253749712,
-      "grad_norm": 5.610541343688965,
-      "learning_rate": 0.0001224112406825442,
-      "loss": -30.3844,
-      "step": 15450
-    },
-    {
-      "epoch": 1.1922159833858934,
-      "grad_norm": 6.599088191986084,
-      "learning_rate": 0.00012215152066072775,
-      "loss": -30.3857,
-      "step": 15500
-    },
-    {
-      "epoch": 1.1922159833858934,
-      "eval_loss": -30.417774200439453,
-      "eval_runtime": 17.1082,
-      "eval_samples_per_second": 58.451,
-      "eval_steps_per_second": 14.613,
-      "step": 15500
-    },
-    {
-      "epoch": 1.1960618413968156,
-      "grad_norm": 5.156207084655762,
-      "learning_rate": 0.00012189180063891126,
-      "loss": -30.3882,
-      "step": 15550
-    },
-    {
-      "epoch": 1.1999076994077378,
-      "grad_norm": 6.738996982574463,
-      "learning_rate": 0.00012163208061709476,
-      "loss": -30.3916,
-      "step": 15600
-    },
-    {
-      "epoch": 1.2037535574186602,
-      "grad_norm": 3.4522907733917236,
-      "learning_rate": 0.0001213723605952783,
-      "loss": -30.3917,
-      "step": 15650
-    },
-    {
-      "epoch": 1.2075994154295824,
-      "grad_norm": 2.062731981277466,
-      "learning_rate": 0.00012111264057346182,
-      "loss": -30.3951,
-      "step": 15700
-    },
-    {
-      "epoch": 1.2114452734405046,
-      "grad_norm": 6.478760242462158,
-      "learning_rate": 0.00012085292055164532,
-      "loss": -30.3963,
-      "step": 15750
-    },
-    {
-      "epoch": 1.2114452734405046,
-      "eval_loss": -30.418058395385742,
-      "eval_runtime": 16.8984,
-      "eval_samples_per_second": 59.177,
-      "eval_steps_per_second": 14.794,
-      "step": 15750
-    },
-    {
-      "epoch": 1.2152911314514268,
-      "grad_norm": 3.3898251056671143,
-      "learning_rate": 0.00012059320052982886,
-      "loss": -30.3988,
-      "step": 15800
-    },
-    {
-      "epoch": 1.2191369894623492,
-      "grad_norm": 5.295444965362549,
-      "learning_rate": 0.00012033348050801237,
-      "loss": -30.4014,
-      "step": 15850
-    },
-    {
-      "epoch": 1.2229828474732714,
-      "grad_norm": 3.9401540756225586,
-      "learning_rate": 0.0001200737604861959,
-      "loss": -30.4028,
-      "step": 15900
-    },
-    {
-      "epoch": 1.2268287054841935,
-      "grad_norm": 6.728087902069092,
-      "learning_rate": 0.00011981404046437941,
-      "loss": -30.403,
-      "step": 15950
-    },
-    {
-      "epoch": 1.2306745634951157,
-      "grad_norm": 4.500195503234863,
-      "learning_rate": 0.00011955432044256292,
-      "loss": -30.4058,
-      "step": 16000
-    },
-    {
-      "epoch": 1.2306745634951157,
-      "eval_loss": -30.41932487487793,
-      "eval_runtime": 16.9437,
-      "eval_samples_per_second": 59.019,
-      "eval_steps_per_second": 14.755,
-      "step": 16000
-    },
-    {
-      "epoch": 1.234520421506038,
-      "grad_norm": 4.910549163818359,
-      "learning_rate": 0.00011929460042074645,
-      "loss": -30.4013,
-      "step": 16050
-    },
-    {
-      "epoch": 1.2383662795169603,
-      "grad_norm": 3.5906295776367188,
-      "learning_rate": 0.00011903488039892996,
-      "loss": -30.4036,
-      "step": 16100
-    },
-    {
-      "epoch": 1.2422121375278825,
-      "grad_norm": 4.437080383300781,
-      "learning_rate": 0.00011877516037711347,
-      "loss": -30.4115,
-      "step": 16150
-    },
-    {
-      "epoch": 1.2460579955388047,
-      "grad_norm": 6.500924110412598,
-      "learning_rate": 0.000118515440355297,
-      "loss": -30.4133,
-      "step": 16200
-    },
-    {
-      "epoch": 1.2499038535497269,
-      "grad_norm": 5.0980143547058105,
-      "learning_rate": 0.00011825572033348051,
-      "loss": -30.4132,
-      "step": 16250
-    },
-    {
-      "epoch": 1.2499038535497269,
-      "eval_loss": -30.43558120727539,
-      "eval_runtime": 17.0068,
-      "eval_samples_per_second": 58.8,
-      "eval_steps_per_second": 14.7,
-      "step": 16250
-    },
-    {
-      "epoch": 1.253749711560649,
-      "grad_norm": 4.151224613189697,
-      "learning_rate": 0.00011799600031166402,
-      "loss": -30.4139,
-      "step": 16300
-    },
-    {
-      "epoch": 1.2575955695715715,
-      "grad_norm": 5.165994167327881,
-      "learning_rate": 0.00011773628028984755,
-      "loss": -30.4154,
-      "step": 16350
-    },
-    {
-      "epoch": 1.2614414275824937,
-      "grad_norm": 3.3392717838287354,
-      "learning_rate": 0.00011747656026803106,
-      "loss": -30.4177,
-      "step": 16400
-    },
-    {
-      "epoch": 1.2652872855934159,
-      "grad_norm": 5.80817985534668,
-      "learning_rate": 0.00011721684024621458,
-      "loss": -30.4187,
-      "step": 16450
-    },
-    {
-      "epoch": 1.2691331436043383,
-      "grad_norm": 4.31649112701416,
-      "learning_rate": 0.0001169571202243981,
-      "loss": -30.4183,
-      "step": 16500
-    },
-    {
-      "epoch": 1.2691331436043383,
-      "eval_loss": -30.43385124206543,
-      "eval_runtime": 16.9325,
-      "eval_samples_per_second": 59.058,
-      "eval_steps_per_second": 14.765,
-      "step": 16500
-    },
-    {
-      "epoch": 1.2729790016152602,
-      "grad_norm": 2.9843568801879883,
-      "learning_rate": 0.00011669740020258162,
-      "loss": -30.4202,
-      "step": 16550
-    },
-    {
-      "epoch": 1.2768248596261826,
-      "grad_norm": 2.4206202030181885,
-      "learning_rate": 0.00011643768018076516,
-      "loss": -30.4172,
-      "step": 16600
-    },
-    {
-      "epoch": 1.2806707176371048,
-      "grad_norm": 4.983964443206787,
-      "learning_rate": 0.00011617796015894867,
-      "loss": -30.4214,
-      "step": 16650
-    },
-    {
-      "epoch": 1.284516575648027,
-      "grad_norm": 1.8332034349441528,
-      "learning_rate": 0.00011591824013713217,
-      "loss": -30.4232,
-      "step": 16700
-    },
-    {
-      "epoch": 1.2883624336589494,
-      "grad_norm": 5.537328720092773,
-      "learning_rate": 0.00011565852011531571,
-      "loss": -30.4212,
-      "step": 16750
-    },
-    {
-      "epoch": 1.2883624336589494,
-      "eval_loss": -30.440034866333008,
-      "eval_runtime": 16.9934,
-      "eval_samples_per_second": 58.846,
-      "eval_steps_per_second": 14.712,
-      "step": 16750
-    },
-    {
-      "epoch": 1.2922082916698716,
-      "grad_norm": 4.200454235076904,
-      "learning_rate": 0.00011539880009349922,
-      "loss": -30.4222,
-      "step": 16800
-    },
-    {
-      "epoch": 1.2960541496807938,
-      "grad_norm": 3.8615283966064453,
-      "learning_rate": 0.00011513908007168272,
-      "loss": -30.4218,
-      "step": 16850
-    },
-    {
-      "epoch": 1.299900007691716,
-      "grad_norm": 4.854475975036621,
-      "learning_rate": 0.00011487936004986626,
-      "loss": -30.4196,
-      "step": 16900
-    },
-    {
-      "epoch": 1.3037458657026382,
-      "grad_norm": 3.457418441772461,
-      "learning_rate": 0.00011461964002804977,
-      "loss": -30.4238,
-      "step": 16950
-    },
-    {
-      "epoch": 1.3075917237135606,
-      "grad_norm": 3.361154317855835,
-      "learning_rate": 0.00011435992000623327,
-      "loss": -30.4236,
-      "step": 17000
-    },
-    {
-      "epoch": 1.3075917237135606,
-      "eval_loss": -30.437725067138672,
-      "eval_runtime": 16.8779,
-      "eval_samples_per_second": 59.249,
-      "eval_steps_per_second": 14.812,
-      "step": 17000
-    },
-    {
-      "epoch": 1.3114375817244828,
-      "grad_norm": 5.688414096832275,
-      "learning_rate": 0.00011410019998441681,
-      "loss": -30.4228,
-      "step": 17050
-    },
-    {
-      "epoch": 1.315283439735405,
-      "grad_norm": 4.790516376495361,
-      "learning_rate": 0.00011384047996260032,
-      "loss": -30.4245,
-      "step": 17100
-    },
-    {
-      "epoch": 1.3191292977463271,
-      "grad_norm": 4.5773539543151855,
-      "learning_rate": 0.00011358075994078384,
-      "loss": -30.4263,
-      "step": 17150
-    },
-    {
-      "epoch": 1.3229751557572493,
-      "grad_norm": 3.8185110092163086,
-      "learning_rate": 0.00011332103991896736,
-      "loss": -30.4269,
-      "step": 17200
-    },
-    {
-      "epoch": 1.3268210137681717,
-      "grad_norm": 2.325967788696289,
-      "learning_rate": 0.00011306131989715088,
-      "loss": -30.4272,
-      "step": 17250
-    },
-    {
-      "epoch": 1.3268210137681717,
-      "eval_loss": -30.445838928222656,
-      "eval_runtime": 16.9252,
-      "eval_samples_per_second": 59.083,
-      "eval_steps_per_second": 14.771,
-      "step": 17250
-    },
-    {
-      "epoch": 1.330666871779094,
-      "grad_norm": 2.729612112045288,
-      "learning_rate": 0.00011280159987533439,
-      "loss": -30.4271,
-      "step": 17300
-    },
-    {
-      "epoch": 1.3345127297900161,
-      "grad_norm": 3.422671318054199,
-      "learning_rate": 0.00011254187985351792,
-      "loss": -30.4248,
-      "step": 17350
-    },
-    {
-      "epoch": 1.3383585878009385,
-      "grad_norm": 3.345399856567383,
-      "learning_rate": 0.00011228215983170143,
-      "loss": -30.4281,
-      "step": 17400
-    },
-    {
-      "epoch": 1.3422044458118605,
-      "grad_norm": 3.8958306312561035,
-      "learning_rate": 0.00011202243980988496,
-      "loss": -30.4291,
-      "step": 17450
-    },
-    {
-      "epoch": 1.346050303822783,
-      "grad_norm": 6.417865753173828,
-      "learning_rate": 0.00011176271978806847,
-      "loss": -30.4293,
-      "step": 17500
-    },
-    {
-      "epoch": 1.346050303822783,
-      "eval_loss": -30.43779182434082,
-      "eval_runtime": 16.8798,
-      "eval_samples_per_second": 59.242,
-      "eval_steps_per_second": 14.811,
-      "step": 17500
-    },
-    {
-      "epoch": 1.349896161833705,
-      "grad_norm": 3.9683895111083984,
-      "learning_rate": 0.00011150299976625198,
-      "loss": -30.43,
-      "step": 17550
-    },
-    {
-      "epoch": 1.3537420198446273,
-      "grad_norm": 2.845992088317871,
-      "learning_rate": 0.00011124327974443551,
-      "loss": -30.4299,
-      "step": 17600
-    },
-    {
-      "epoch": 1.3575878778555497,
-      "grad_norm": 2.662853240966797,
-      "learning_rate": 0.00011098355972261902,
-      "loss": -30.4296,
-      "step": 17650
-    },
-    {
-      "epoch": 1.3614337358664719,
-      "grad_norm": 4.240821838378906,
-      "learning_rate": 0.00011072383970080253,
-      "loss": -30.4296,
-      "step": 17700
-    },
-    {
-      "epoch": 1.365279593877394,
-      "grad_norm": 3.456580400466919,
-      "learning_rate": 0.00011046411967898607,
-      "loss": -30.4301,
-      "step": 17750
-    },
-    {
-      "epoch": 1.365279593877394,
-      "eval_loss": -30.446062088012695,
-      "eval_runtime": 17.169,
-      "eval_samples_per_second": 58.244,
-      "eval_steps_per_second": 14.561,
-      "step": 17750
-    },
-    {
-      "epoch": 1.3691254518883162,
-      "grad_norm": 2.481637477874756,
-      "learning_rate": 0.00011020439965716957,
-      "loss": -30.43,
-      "step": 17800
-    },
-    {
-      "epoch": 1.3729713098992384,
-      "grad_norm": 5.5576558113098145,
-      "learning_rate": 0.00010994467963535309,
-      "loss": -30.4315,
-      "step": 17850
-    },
-    {
-      "epoch": 1.3768171679101608,
-      "grad_norm": 3.786583662033081,
-      "learning_rate": 0.00010968495961353663,
-      "loss": -30.4279,
-      "step": 17900
-    },
-    {
-      "epoch": 1.380663025921083,
-      "grad_norm": 3.966721773147583,
-      "learning_rate": 0.00010942523959172012,
-      "loss": -30.4325,
-      "step": 17950
-    },
-    {
-      "epoch": 1.3845088839320052,
-      "grad_norm": 5.2827582359313965,
-      "learning_rate": 0.00010916551956990364,
-      "loss": -30.433,
-      "step": 18000
-    },
-    {
-      "epoch": 1.3845088839320052,
-      "eval_loss": -30.445518493652344,
-      "eval_runtime": 16.9753,
-      "eval_samples_per_second": 58.909,
-      "eval_steps_per_second": 14.727,
-      "step": 18000
-    },
-    {
-      "epoch": 1.3883547419429274,
-      "grad_norm": 4.988203048706055,
-      "learning_rate": 0.00010890579954808718,
-      "loss": -30.4311,
-      "step": 18050
-    },
-    {
-      "epoch": 1.3922005999538496,
-      "grad_norm": 4.5886454582214355,
-      "learning_rate": 0.00010864607952627068,
-      "loss": -30.4332,
-      "step": 18100
-    },
-    {
-      "epoch": 1.396046457964772,
-      "grad_norm": 2.1658132076263428,
-      "learning_rate": 0.00010838635950445419,
-      "loss": -30.434,
-      "step": 18150
-    },
-    {
-      "epoch": 1.3998923159756942,
-      "grad_norm": 3.8353946208953857,
-      "learning_rate": 0.00010812663948263773,
-      "loss": -30.4336,
-      "step": 18200
-    },
-    {
-      "epoch": 1.4037381739866164,
-      "grad_norm": 4.345909118652344,
-      "learning_rate": 0.00010786691946082124,
-      "loss": -30.4363,
-      "step": 18250
-    },
-    {
-      "epoch": 1.4037381739866164,
-      "eval_loss": -30.447927474975586,
-      "eval_runtime": 16.9767,
-      "eval_samples_per_second": 58.904,
-      "eval_steps_per_second": 14.726,
-      "step": 18250
-    },
-    {
-      "epoch": 1.4075840319975388,
-      "grad_norm": 2.635801076889038,
-      "learning_rate": 0.00010760719943900477,
-      "loss": -30.4351,
-      "step": 18300
-    },
-    {
-      "epoch": 1.4114298900084608,
-      "grad_norm": 4.603756427764893,
-      "learning_rate": 0.00010734747941718828,
-      "loss": -30.4351,
-      "step": 18350
-    },
-    {
-      "epoch": 1.4152757480193832,
-      "grad_norm": 4.313344955444336,
-      "learning_rate": 0.0001070877593953718,
-      "loss": -30.4361,
-      "step": 18400
-    },
-    {
-      "epoch": 1.4191216060303053,
-      "grad_norm": 4.869143486022949,
-      "learning_rate": 0.00010682803937355532,
-      "loss": -30.4351,
-      "step": 18450
-    },
-    {
-      "epoch": 1.4229674640412275,
-      "grad_norm": 2.338141679763794,
-      "learning_rate": 0.00010656831935173883,
-      "loss": -30.435,
-      "step": 18500
-    },
-    {
-      "epoch": 1.4229674640412275,
-      "eval_loss": -30.445350646972656,
-      "eval_runtime": 16.9508,
-      "eval_samples_per_second": 58.994,
-      "eval_steps_per_second": 14.749,
-      "step": 18500
-    },
-    {
-      "epoch": 1.42681332205215,
-      "grad_norm": 3.5870954990386963,
-      "learning_rate": 0.00010630859932992235,
-      "loss": -30.4369,
-      "step": 18550
-    },
-    {
-      "epoch": 1.4306591800630721,
-      "grad_norm": 2.8752095699310303,
-      "learning_rate": 0.00010604887930810587,
-      "loss": -30.4345,
-      "step": 18600
-    },
-    {
-      "epoch": 1.4345050380739943,
-      "grad_norm": 3.7183678150177,
-      "learning_rate": 0.00010578915928628939,
-      "loss": -30.4362,
-      "step": 18650
-    },
-    {
-      "epoch": 1.4383508960849165,
-      "grad_norm": 3.3645851612091064,
-      "learning_rate": 0.0001055294392644729,
-      "loss": -30.4363,
-      "step": 18700
-    },
-    {
-      "epoch": 1.4421967540958387,
-      "grad_norm": 3.521252393722534,
-      "learning_rate": 0.00010526971924265642,
-      "loss": -30.4371,
-      "step": 18750
-    },
-    {
-      "epoch": 1.4421967540958387,
-      "eval_loss": -30.448963165283203,
-      "eval_runtime": 16.966,
-      "eval_samples_per_second": 58.941,
-      "eval_steps_per_second": 14.735,
-      "step": 18750
-    },
-    {
-      "epoch": 1.446042612106761,
-      "grad_norm": 4.74116325378418,
-      "learning_rate": 0.00010500999922083994,
-      "loss": -30.4378,
-      "step": 18800
-    },
-    {
-      "epoch": 1.4498884701176833,
-      "grad_norm": 3.7258331775665283,
-      "learning_rate": 0.00010475027919902345,
-      "loss": -30.4376,
-      "step": 18850
-    },
-    {
-      "epoch": 1.4537343281286055,
-      "grad_norm": 4.193502426147461,
-      "learning_rate": 0.00010449055917720698,
-      "loss": -30.4384,
-      "step": 18900
-    },
-    {
-      "epoch": 1.4575801861395277,
-      "grad_norm": 2.630093812942505,
-      "learning_rate": 0.00010423083915539049,
-      "loss": -30.4363,
-      "step": 18950
-    },
-    {
-      "epoch": 1.4614260441504499,
-      "grad_norm": 5.140982151031494,
-      "learning_rate": 0.000103971119133574,
-      "loss": -30.433,
-      "step": 19000
-    },
-    {
-      "epoch": 1.4614260441504499,
-      "eval_loss": -30.441884994506836,
-      "eval_runtime": 16.8889,
-      "eval_samples_per_second": 59.21,
-      "eval_steps_per_second": 14.803,
-      "step": 19000
-    },
-    {
-      "epoch": 1.4652719021613723,
-      "grad_norm": 4.889688968658447,
-      "learning_rate": 0.00010371139911175753,
-      "loss": -30.4382,
-      "step": 19050
-    },
-    {
-      "epoch": 1.4691177601722945,
-      "grad_norm": 3.332848072052002,
-      "learning_rate": 0.00010345167908994104,
-      "loss": -30.4385,
-      "step": 19100
-    },
-    {
-      "epoch": 1.4729636181832166,
-      "grad_norm": 3.6298229694366455,
-      "learning_rate": 0.00010319195906812458,
-      "loss": -30.4384,
-      "step": 19150
-    },
-    {
-      "epoch": 1.476809476194139,
-      "grad_norm": 3.3661463260650635,
-      "learning_rate": 0.0001029322390463081,
-      "loss": -30.4383,
-      "step": 19200
-    },
-    {
-      "epoch": 1.480655334205061,
-      "grad_norm": 3.7535996437072754,
-      "learning_rate": 0.0001026725190244916,
-      "loss": -30.4388,
-      "step": 19250
-    },
-    {
-      "epoch": 1.480655334205061,
-      "eval_loss": -30.449628829956055,
-      "eval_runtime": 16.977,
-      "eval_samples_per_second": 58.903,
-      "eval_steps_per_second": 14.726,
-      "step": 19250
-    },
-    {
-      "epoch": 1.4845011922159834,
-      "grad_norm": 2.860837936401367,
-      "learning_rate": 0.00010241279900267513,
-      "loss": -30.4381,
-      "step": 19300
-    },
-    {
-      "epoch": 1.4883470502269056,
-      "grad_norm": 3.559368848800659,
-      "learning_rate": 0.00010215307898085865,
-      "loss": -30.4401,
-      "step": 19350
-    },
-    {
-      "epoch": 1.4921929082378278,
-      "grad_norm": 4.874238967895508,
-      "learning_rate": 0.00010189335895904215,
-      "loss": -30.4403,
-      "step": 19400
-    },
-    {
-      "epoch": 1.4960387662487502,
-      "grad_norm": 3.3331515789031982,
-      "learning_rate": 0.00010163363893722569,
-      "loss": -30.4398,
-      "step": 19450
-    },
-    {
-      "epoch": 1.4998846242596724,
-      "grad_norm": 2.6575047969818115,
-      "learning_rate": 0.0001013739189154092,
-      "loss": -30.4338,
-      "step": 19500
-    },
-    {
-      "epoch": 1.4998846242596724,
-      "eval_loss": -30.449535369873047,
-      "eval_runtime": 16.7967,
-      "eval_samples_per_second": 59.536,
-      "eval_steps_per_second": 14.884,
-      "step": 19500
-    },
-    {
-      "epoch": 1.5037304822705946,
-      "grad_norm": 3.4910714626312256,
-      "learning_rate": 0.0001011141988935927,
-      "loss": -30.4402,
-      "step": 19550
-    },
-    {
-      "epoch": 1.5075763402815168,
-      "grad_norm": 3.4878039360046387,
-      "learning_rate": 0.00010085447887177624,
-      "loss": -30.4406,
-      "step": 19600
-    },
-    {
-      "epoch": 1.511422198292439,
-      "grad_norm": 1.7369427680969238,
-      "learning_rate": 0.00010059475884995975,
-      "loss": -30.441,
-      "step": 19650
-    },
-    {
-      "epoch": 1.5152680563033614,
-      "grad_norm": 2.82442045211792,
-      "learning_rate": 0.00010033503882814326,
-      "loss": -30.4407,
-      "step": 19700
-    },
-    {
-      "epoch": 1.5191139143142836,
-      "grad_norm": 3.5565500259399414,
-      "learning_rate": 0.00010007531880632679,
-      "loss": -30.4322,
-      "step": 19750
-    },
-    {
-      "epoch": 1.5191139143142836,
-      "eval_loss": -30.445045471191406,
-      "eval_runtime": 16.8955,
-      "eval_samples_per_second": 59.187,
-      "eval_steps_per_second": 14.797,
-      "step": 19750
-    },
-    {
-      "epoch": 1.5229597723252057,
-      "grad_norm": 2.4976837635040283,
-      "learning_rate": 9.98155987845103e-05,
-      "loss": -30.4389,
-      "step": 19800
-    },
-    {
-      "epoch": 1.5268056303361282,
-      "grad_norm": 2.888719320297241,
-      "learning_rate": 9.955587876269382e-05,
-      "loss": -30.4332,
-      "step": 19850
-    },
-    {
-      "epoch": 1.5306514883470501,
-      "grad_norm": 4.438738822937012,
-      "learning_rate": 9.929615874087734e-05,
-      "loss": -30.4415,
-      "step": 19900
-    },
-    {
-      "epoch": 1.5344973463579725,
-      "grad_norm": 1.5556750297546387,
-      "learning_rate": 9.903643871906085e-05,
-      "loss": -30.4422,
-      "step": 19950
-    },
-    {
-      "epoch": 1.5383432043688947,
-      "grad_norm": 3.4446284770965576,
-      "learning_rate": 9.877671869724438e-05,
-      "loss": -30.4399,
-      "step": 20000
-    },
-    {
-      "epoch": 1.5383432043688947,
-      "eval_loss": -30.452638626098633,
-      "eval_runtime": 16.7438,
-      "eval_samples_per_second": 59.724,
-      "eval_steps_per_second": 14.931,
-      "step": 20000
-    },
-    {
-      "epoch": 1.542189062379817,
-      "grad_norm": 2.2135956287384033,
-      "learning_rate": 9.85169986754279e-05,
-      "loss": -30.4423,
-      "step": 20050
-    },
-    {
-      "epoch": 1.5460349203907393,
-      "grad_norm": 3.5614891052246094,
-      "learning_rate": 9.825727865361141e-05,
-      "loss": -30.4419,
-      "step": 20100
-    },
-    {
-      "epoch": 1.5498807784016613,
-      "grad_norm": 3.5207202434539795,
-      "learning_rate": 9.799755863179493e-05,
-      "loss": -30.4427,
-      "step": 20150
-    },
-    {
-      "epoch": 1.5537266364125837,
-      "grad_norm": 2.915543794631958,
-      "learning_rate": 9.773783860997845e-05,
-      "loss": -30.4403,
-      "step": 20200
-    },
-    {
-      "epoch": 1.5575724944235059,
-      "grad_norm": 2.3293378353118896,
-      "learning_rate": 9.747811858816197e-05,
-      "loss": -30.4429,
-      "step": 20250
-    },
-    {
-      "epoch": 1.5575724944235059,
-      "eval_loss": -30.45326042175293,
-      "eval_runtime": 16.8328,
-      "eval_samples_per_second": 59.408,
-      "eval_steps_per_second": 14.852,
-      "step": 20250
-    },
-    {
-      "epoch": 1.561418352434428,
-      "grad_norm": 4.711287021636963,
-      "learning_rate": 9.721839856634549e-05,
-      "loss": -30.4434,
-      "step": 20300
-    },
-    {
-      "epoch": 1.5652642104453505,
-      "grad_norm": 1.71726393699646,
-      "learning_rate": 9.6958678544529e-05,
-      "loss": -30.4421,
-      "step": 20350
-    },
-    {
-      "epoch": 1.5691100684562724,
-      "grad_norm": 3.524540424346924,
-      "learning_rate": 9.669895852271252e-05,
-      "loss": -30.4437,
-      "step": 20400
-    },
-    {
-      "epoch": 1.5729559264671948,
-      "grad_norm": 3.0880210399627686,
-      "learning_rate": 9.643923850089604e-05,
-      "loss": -30.4424,
-      "step": 20450
-    },
-    {
-      "epoch": 1.576801784478117,
-      "grad_norm": 4.263181686401367,
-      "learning_rate": 9.617951847907955e-05,
-      "loss": -30.444,
-      "step": 20500
-    },
-    {
-      "epoch": 1.576801784478117,
-      "eval_loss": -30.45319175720215,
-      "eval_runtime": 16.9713,
-      "eval_samples_per_second": 58.923,
-      "eval_steps_per_second": 14.731,
-      "step": 20500
-    },
-    {
-      "epoch": 1.5806476424890392,
-      "grad_norm": 7.002903938293457,
-      "learning_rate": 9.591979845726308e-05,
-      "loss": -30.4429,
-      "step": 20550
-    },
-    {
-      "epoch": 1.5844935004999616,
-      "grad_norm": 3.438530445098877,
-      "learning_rate": 9.56600784354466e-05,
-      "loss": -30.4417,
-      "step": 20600
-    },
-    {
-      "epoch": 1.5883393585108838,
-      "grad_norm": 3.099423408508301,
-      "learning_rate": 9.54003584136301e-05,
-      "loss": -30.444,
-      "step": 20650
-    },
-    {
-      "epoch": 1.592185216521806,
-      "grad_norm": 3.872269868850708,
-      "learning_rate": 9.514063839181363e-05,
-      "loss": -30.4439,
-      "step": 20700
-    },
-    {
-      "epoch": 1.5960310745327284,
-      "grad_norm": 2.9675283432006836,
-      "learning_rate": 9.488091836999716e-05,
-      "loss": -30.4453,
-      "step": 20750
-    },
-    {
-      "epoch": 1.5960310745327284,
-      "eval_loss": -30.452062606811523,
-      "eval_runtime": 16.9677,
-      "eval_samples_per_second": 58.936,
-      "eval_steps_per_second": 14.734,
-      "step": 20750
-    },
-    {
-      "epoch": 1.5998769325436504,
-      "grad_norm": 3.0117275714874268,
-      "learning_rate": 9.462119834818067e-05,
-      "loss": -30.4438,
-      "step": 20800
-    },
-    {
-      "epoch": 1.6037227905545728,
-      "grad_norm": 3.3223464488983154,
-      "learning_rate": 9.436147832636418e-05,
-      "loss": -30.4445,
-      "step": 20850
-    },
-    {
-      "epoch": 1.607568648565495,
-      "grad_norm": 2.780449390411377,
-      "learning_rate": 9.410175830454771e-05,
-      "loss": -30.4441,
-      "step": 20900
-    },
-    {
-      "epoch": 1.6114145065764172,
-      "grad_norm": 2.49845027923584,
-      "learning_rate": 9.384203828273122e-05,
-      "loss": -30.4451,
-      "step": 20950
-    },
-    {
-      "epoch": 1.6152603645873396,
-      "grad_norm": 2.1651992797851562,
-      "learning_rate": 9.358231826091473e-05,
-      "loss": -30.4443,
-      "step": 21000
-    },
-    {
-      "epoch": 1.6152603645873396,
-      "eval_loss": -30.446992874145508,
-      "eval_runtime": 16.808,
-      "eval_samples_per_second": 59.495,
-      "eval_steps_per_second": 14.874,
-      "step": 21000
-    },
-    {
-      "epoch": 1.6191062225982615,
-      "grad_norm": 2.3256795406341553,
-      "learning_rate": 9.332259823909826e-05,
-      "loss": -30.4447,
-      "step": 21050
-    },
-    {
-      "epoch": 1.622952080609184,
-      "grad_norm": 3.201892852783203,
-      "learning_rate": 9.306287821728179e-05,
-      "loss": -30.4439,
-      "step": 21100
-    },
-    {
-      "epoch": 1.6267979386201061,
-      "grad_norm": 2.861118793487549,
-      "learning_rate": 9.280315819546528e-05,
-      "loss": -30.4453,
-      "step": 21150
-    },
-    {
-      "epoch": 1.6306437966310283,
-      "grad_norm": 2.3513760566711426,
-      "learning_rate": 9.254343817364881e-05,
-      "loss": -30.4454,
-      "step": 21200
-    },
-    {
-      "epoch": 1.6344896546419507,
-      "grad_norm": 2.680166482925415,
-      "learning_rate": 9.228371815183234e-05,
-      "loss": -30.4461,
-      "step": 21250
-    },
-    {
-      "epoch": 1.6344896546419507,
-      "eval_loss": -30.453556060791016,
-      "eval_runtime": 16.9036,
-      "eval_samples_per_second": 59.159,
-      "eval_steps_per_second": 14.79,
-      "step": 21250
-    },
-    {
-      "epoch": 1.6383355126528727,
-      "grad_norm": 3.896608352661133,
-      "learning_rate": 9.202399813001584e-05,
-      "loss": -30.4461,
-      "step": 21300
-    },
-    {
-      "epoch": 1.642181370663795,
-      "grad_norm": 3.4098875522613525,
-      "learning_rate": 9.176427810819936e-05,
-      "loss": -30.4464,
-      "step": 21350
-    },
-    {
-      "epoch": 1.6460272286747173,
-      "grad_norm": 2.555978536605835,
-      "learning_rate": 9.150455808638289e-05,
-      "loss": -30.4457,
-      "step": 21400
-    },
-    {
-      "epoch": 1.6498730866856395,
-      "grad_norm": 3.0330915451049805,
-      "learning_rate": 9.12448380645664e-05,
-      "loss": -30.4463,
-      "step": 21450
-    },
-    {
-      "epoch": 1.6537189446965619,
-      "grad_norm": 4.583216190338135,
-      "learning_rate": 9.098511804274992e-05,
-      "loss": -30.445,
-      "step": 21500
-    },
-    {
-      "epoch": 1.6537189446965619,
-      "eval_loss": -30.445945739746094,
-      "eval_runtime": 16.9709,
-      "eval_samples_per_second": 58.925,
-      "eval_steps_per_second": 14.731,
-      "step": 21500
-    },
-    {
-      "epoch": 1.657564802707484,
-      "grad_norm": 2.692993402481079,
-      "learning_rate": 9.072539802093344e-05,
-      "loss": -30.4452,
-      "step": 21550
-    },
-    {
-      "epoch": 1.6614106607184063,
-      "grad_norm": 3.2486062049865723,
-      "learning_rate": 9.046567799911695e-05,
-      "loss": -30.4466,
-      "step": 21600
-    },
-    {
-      "epoch": 1.6652565187293287,
-      "grad_norm": 3.774552345275879,
-      "learning_rate": 9.020595797730047e-05,
-      "loss": -30.4468,
-      "step": 21650
-    },
-    {
-      "epoch": 1.6691023767402506,
-      "grad_norm": 2.923283338546753,
-      "learning_rate": 8.9946237955484e-05,
-      "loss": -30.4472,
-      "step": 21700
-    },
-    {
-      "epoch": 1.672948234751173,
-      "grad_norm": 2.804466724395752,
-      "learning_rate": 8.96865179336675e-05,
-      "loss": -30.4471,
-      "step": 21750
-    },
-    {
-      "epoch": 1.672948234751173,
-      "eval_loss": -30.455671310424805,
-      "eval_runtime": 16.8224,
-      "eval_samples_per_second": 59.445,
-      "eval_steps_per_second": 14.861,
-      "step": 21750
-    },
-    {
-      "epoch": 1.6767940927620952,
-      "grad_norm": 2.091228485107422,
-      "learning_rate": 8.942679791185102e-05,
-      "loss": -30.4476,
-      "step": 21800
-    },
-    {
-      "epoch": 1.6806399507730174,
-      "grad_norm": 3.4288341999053955,
-      "learning_rate": 8.916707789003455e-05,
-      "loss": -30.4466,
-      "step": 21850
-    },
-    {
-      "epoch": 1.6844858087839398,
-      "grad_norm": 2.85774827003479,
-      "learning_rate": 8.890735786821807e-05,
-      "loss": -30.4471,
-      "step": 21900
-    },
-    {
-      "epoch": 1.6883316667948618,
-      "grad_norm": 3.7691524028778076,
-      "learning_rate": 8.864763784640159e-05,
-      "loss": -30.4473,
-      "step": 21950
-    },
-    {
-      "epoch": 1.6921775248057842,
-      "grad_norm": 3.015897035598755,
-      "learning_rate": 8.83879178245851e-05,
-      "loss": -30.4478,
-      "step": 22000
-    },
-    {
-      "epoch": 1.6921775248057842,
-      "eval_loss": -30.455175399780273,
-      "eval_runtime": 16.9353,
-      "eval_samples_per_second": 59.048,
-      "eval_steps_per_second": 14.762,
-      "step": 22000
-    },
-    {
-      "epoch": 1.6960233828167064,
-      "grad_norm": 2.7589683532714844,
-      "learning_rate": 8.812819780276862e-05,
-      "loss": -30.4483,
-      "step": 22050
-    },
-    {
-      "epoch": 1.6998692408276286,
-      "grad_norm": 3.7240102291107178,
-      "learning_rate": 8.786847778095214e-05,
-      "loss": -30.448,
-      "step": 22100
-    },
-    {
-      "epoch": 1.703715098838551,
-      "grad_norm": 3.830146074295044,
-      "learning_rate": 8.760875775913565e-05,
-      "loss": -30.448,
-      "step": 22150
-    },
-    {
-      "epoch": 1.707560956849473,
-      "grad_norm": 2.670171022415161,
-      "learning_rate": 8.734903773731918e-05,
-      "loss": -30.4485,
-      "step": 22200
-    },
-    {
-      "epoch": 1.7114068148603954,
-      "grad_norm": 2.8607327938079834,
-      "learning_rate": 8.708931771550269e-05,
-      "loss": -30.4486,
-      "step": 22250
-    },
-    {
-      "epoch": 1.7114068148603954,
-      "eval_loss": -30.454792022705078,
-      "eval_runtime": 16.9895,
-      "eval_samples_per_second": 58.86,
-      "eval_steps_per_second": 14.715,
-      "step": 22250
-    },
-    {
-      "epoch": 1.7152526728713176,
-      "grad_norm": 4.332964897155762,
-      "learning_rate": 8.682959769368622e-05,
-      "loss": -30.4481,
-      "step": 22300
-    },
-    {
-      "epoch": 1.7190985308822397,
-      "grad_norm": 3.3097095489501953,
-      "learning_rate": 8.656987767186973e-05,
-      "loss": -30.4482,
-      "step": 22350
-    },
-    {
-      "epoch": 1.7229443888931621,
-      "grad_norm": 2.8887455463409424,
-      "learning_rate": 8.631015765005324e-05,
-      "loss": -30.4493,
-      "step": 22400
-    },
-    {
-      "epoch": 1.7267902469040843,
-      "grad_norm": 2.452583074569702,
-      "learning_rate": 8.605043762823677e-05,
-      "loss": -30.4492,
-      "step": 22450
-    },
-    {
-      "epoch": 1.7306361049150065,
-      "grad_norm": 3.4899165630340576,
-      "learning_rate": 8.579071760642028e-05,
-      "loss": -30.4492,
-      "step": 22500
-    },
-    {
-      "epoch": 1.7306361049150065,
-      "eval_loss": -30.454774856567383,
-      "eval_runtime": 16.8487,
-      "eval_samples_per_second": 59.352,
-      "eval_steps_per_second": 14.838,
-      "step": 22500
-    },
-    {
-      "epoch": 1.734481962925929,
-      "grad_norm": 2.8990135192871094,
-      "learning_rate": 8.553099758460381e-05,
-      "loss": -30.448,
-      "step": 22550
-    },
-    {
-      "epoch": 1.738327820936851,
-      "grad_norm": 2.8500287532806396,
-      "learning_rate": 8.527127756278732e-05,
-      "loss": -30.4489,
-      "step": 22600
-    },
-    {
-      "epoch": 1.7421736789477733,
-      "grad_norm": 1.9518436193466187,
-      "learning_rate": 8.501155754097083e-05,
-      "loss": -30.4487,
-      "step": 22650
-    },
-    {
-      "epoch": 1.7460195369586955,
-      "grad_norm": 2.950298309326172,
-      "learning_rate": 8.475183751915436e-05,
-      "loss": -30.4484,
-      "step": 22700
-    },
-    {
-      "epoch": 1.7498653949696177,
-      "grad_norm": 3.065047264099121,
-      "learning_rate": 8.449211749733787e-05,
-      "loss": -30.4494,
-      "step": 22750
-    },
-    {
-      "epoch": 1.7498653949696177,
-      "eval_loss": -30.456340789794922,
-      "eval_runtime": 16.963,
-      "eval_samples_per_second": 58.952,
-      "eval_steps_per_second": 14.738,
-      "step": 22750
-    },
-    {
-      "epoch": 1.75371125298054,
-      "grad_norm": 2.389820098876953,
-      "learning_rate": 8.42323974755214e-05,
-      "loss": -30.4503,
-      "step": 22800
-    },
-    {
-      "epoch": 1.757557110991462,
-      "grad_norm": 4.366456985473633,
-      "learning_rate": 8.397267745370491e-05,
-      "loss": -30.4499,
-      "step": 22850
-    },
-    {
-      "epoch": 1.7614029690023845,
-      "grad_norm": 1.5029981136322021,
-      "learning_rate": 8.371295743188842e-05,
-      "loss": -30.4503,
-      "step": 22900
-    },
-    {
-      "epoch": 1.7652488270133067,
-      "grad_norm": 2.083980083465576,
-      "learning_rate": 8.345323741007195e-05,
-      "loss": -30.4501,
-      "step": 22950
-    },
-    {
-      "epoch": 1.7690946850242288,
-      "grad_norm": 3.1565330028533936,
-      "learning_rate": 8.319351738825546e-05,
-      "loss": -30.4494,
-      "step": 23000
-    },
-    {
-      "epoch": 1.7690946850242288,
-      "eval_loss": -30.45475959777832,
-      "eval_runtime": 16.8819,
-      "eval_samples_per_second": 59.235,
-      "eval_steps_per_second": 14.809,
-      "step": 23000
-    },
-    {
-      "epoch": 1.7729405430351513,
-      "grad_norm": 3.5162878036499023,
-      "learning_rate": 8.293379736643898e-05,
-      "loss": -30.4499,
-      "step": 23050
-    },
-    {
-      "epoch": 1.7767864010460732,
-      "grad_norm": 2.4113125801086426,
-      "learning_rate": 8.26740773446225e-05,
-      "loss": -30.4491,
-      "step": 23100
-    },
-    {
-      "epoch": 1.7806322590569956,
-      "grad_norm": 1.8185194730758667,
-      "learning_rate": 8.241435732280603e-05,
-      "loss": -30.4506,
-      "step": 23150
-    },
-    {
-      "epoch": 1.7844781170679178,
-      "grad_norm": 3.115666627883911,
-      "learning_rate": 8.215463730098953e-05,
-      "loss": -30.4495,
-      "step": 23200
-    },
-    {
-      "epoch": 1.78832397507884,
-      "grad_norm": 2.227865695953369,
-      "learning_rate": 8.189491727917305e-05,
-      "loss": -30.4502,
-      "step": 23250
-    },
-    {
-      "epoch": 1.78832397507884,
-      "eval_loss": -30.456247329711914,
-      "eval_runtime": 16.8967,
-      "eval_samples_per_second": 59.183,
-      "eval_steps_per_second": 14.796,
-      "step": 23250
-    },
-    {
-      "epoch": 1.7921698330897624,
-      "grad_norm": 2.8278579711914062,
-      "learning_rate": 8.163519725735658e-05,
-      "loss": -30.4501,
-      "step": 23300
-    },
-    {
-      "epoch": 1.7960156911006846,
-      "grad_norm": 1.8039246797561646,
-      "learning_rate": 8.13754772355401e-05,
-      "loss": -30.4503,
-      "step": 23350
-    },
-    {
-      "epoch": 1.7998615491116068,
-      "grad_norm": 2.9235339164733887,
-      "learning_rate": 8.11157572137236e-05,
-      "loss": -30.4506,
-      "step": 23400
-    },
-    {
-      "epoch": 1.8037074071225292,
-      "grad_norm": 1.375192642211914,
-      "learning_rate": 8.085603719190713e-05,
-      "loss": -30.4514,
-      "step": 23450
-    },
-    {
-      "epoch": 1.8075532651334512,
-      "grad_norm": 2.6810855865478516,
-      "learning_rate": 8.059631717009065e-05,
-      "loss": -30.451,
-      "step": 23500
-    },
-    {
-      "epoch": 1.8075532651334512,
-      "eval_loss": -30.456506729125977,
-      "eval_runtime": 16.8767,
-      "eval_samples_per_second": 59.253,
-      "eval_steps_per_second": 14.813,
-      "step": 23500
-    },
-    {
-      "epoch": 1.8113991231443736,
-      "grad_norm": 3.5937023162841797,
-      "learning_rate": 8.033659714827416e-05,
-      "loss": -30.4489,
-      "step": 23550
-    },
-    {
-      "epoch": 1.8152449811552958,
-      "grad_norm": 2.0678956508636475,
-      "learning_rate": 8.008207152689401e-05,
-      "loss": -30.4469,
-      "step": 23600
-    },
-    {
-      "epoch": 1.819090839166218,
-      "grad_norm": 3.0057597160339355,
-      "learning_rate": 7.982235150507754e-05,
-      "loss": -30.451,
-      "step": 23650
-    },
-    {
-      "epoch": 1.8229366971771404,
-      "grad_norm": 2.567552089691162,
-      "learning_rate": 7.956263148326105e-05,
-      "loss": -30.4511,
-      "step": 23700
-    },
-    {
-      "epoch": 1.8267825551880623,
-      "grad_norm": 2.3882832527160645,
-      "learning_rate": 7.930291146144456e-05,
-      "loss": -30.4512,
-      "step": 23750
-    },
-    {
-      "epoch": 1.8267825551880623,
-      "eval_loss": -30.457120895385742,
-      "eval_runtime": 16.7786,
-      "eval_samples_per_second": 59.6,
-      "eval_steps_per_second": 14.9,
-      "step": 23750
-    },
-    {
-      "epoch": 1.8306284131989847,
-      "grad_norm": 2.1864283084869385,
-      "learning_rate": 7.904319143962809e-05,
-      "loss": -30.4511,
-      "step": 23800
-    },
-    {
-      "epoch": 1.834474271209907,
-      "grad_norm": 2.5760741233825684,
-      "learning_rate": 7.87834714178116e-05,
-      "loss": -30.4512,
-      "step": 23850
-    },
-    {
-      "epoch": 1.838320129220829,
-      "grad_norm": 2.0894832611083984,
-      "learning_rate": 7.852375139599513e-05,
-      "loss": -30.4514,
-      "step": 23900
-    },
-    {
-      "epoch": 1.8421659872317515,
-      "grad_norm": 2.8342926502227783,
-      "learning_rate": 7.826403137417864e-05,
-      "loss": -30.4509,
-      "step": 23950
-    },
-    {
-      "epoch": 1.8460118452426735,
-      "grad_norm": 2.050407648086548,
-      "learning_rate": 7.800431135236215e-05,
-      "loss": -30.4518,
-      "step": 24000
-    },
-    {
-      "epoch": 1.8460118452426735,
-      "eval_loss": -30.43790626525879,
-      "eval_runtime": 16.6887,
-      "eval_samples_per_second": 59.921,
-      "eval_steps_per_second": 14.98,
-      "step": 24000
-    },
-    {
-      "epoch": 1.8498577032535959,
-      "grad_norm": 3.1422579288482666,
-      "learning_rate": 7.774459133054568e-05,
-      "loss": -30.4522,
-      "step": 24050
-    },
-    {
-      "epoch": 1.853703561264518,
-      "grad_norm": 2.7372827529907227,
-      "learning_rate": 7.748487130872919e-05,
-      "loss": -30.4523,
-      "step": 24100
-    },
-    {
-      "epoch": 1.8575494192754403,
-      "grad_norm": 2.3861794471740723,
-      "learning_rate": 7.72251512869127e-05,
-      "loss": -30.4523,
-      "step": 24150
-    },
-    {
-      "epoch": 1.8613952772863627,
-      "grad_norm": 2.221836805343628,
-      "learning_rate": 7.696543126509623e-05,
-      "loss": -30.453,
-      "step": 24200
-    },
-    {
-      "epoch": 1.8652411352972849,
-      "grad_norm": 3.2006218433380127,
-      "learning_rate": 7.670571124327974e-05,
-      "loss": -30.453,
-      "step": 24250
-    },
-    {
-      "epoch": 1.8652411352972849,
-      "eval_loss": -30.45906639099121,
-      "eval_runtime": 16.6647,
-      "eval_samples_per_second": 60.007,
-      "eval_steps_per_second": 15.002,
-      "step": 24250
-    },
-    {
-      "epoch": 1.869086993308207,
-      "grad_norm": 3.0464978218078613,
-      "learning_rate": 7.644599122146327e-05,
-      "loss": -30.4534,
-      "step": 24300
-    },
-    {
-      "epoch": 1.8729328513191295,
-      "grad_norm": 2.5272178649902344,
-      "learning_rate": 7.618627119964678e-05,
-      "loss": -30.4532,
-      "step": 24350
-    },
-    {
-      "epoch": 1.8767787093300514,
-      "grad_norm": 1.6283899545669556,
-      "learning_rate": 7.592655117783031e-05,
-      "loss": -30.4525,
-      "step": 24400
-    },
-    {
-      "epoch": 1.8806245673409738,
-      "grad_norm": 3.1398937702178955,
-      "learning_rate": 7.566683115601382e-05,
-      "loss": -30.4527,
-      "step": 24450
-    },
-    {
-      "epoch": 1.884470425351896,
-      "grad_norm": 2.703652858734131,
-      "learning_rate": 7.540711113419734e-05,
-      "loss": -30.453,
-      "step": 24500
-    },
-    {
-      "epoch": 1.884470425351896,
-      "eval_loss": -30.457090377807617,
-      "eval_runtime": 16.8086,
-      "eval_samples_per_second": 59.493,
-      "eval_steps_per_second": 14.873,
-      "step": 24500
-    },
-    {
-      "epoch": 1.8883162833628182,
-      "grad_norm": 2.3130578994750977,
-      "learning_rate": 7.514739111238086e-05,
-      "loss": -30.4525,
-      "step": 24550
-    },
-    {
-      "epoch": 1.8921621413737406,
-      "grad_norm": 1.720234990119934,
-      "learning_rate": 7.488767109056438e-05,
-      "loss": -30.4533,
-      "step": 24600
-    },
-    {
-      "epoch": 1.8960079993846626,
-      "grad_norm": 2.150315761566162,
-      "learning_rate": 7.462795106874789e-05,
-      "loss": -30.4528,
-      "step": 24650
-    },
-    {
-      "epoch": 1.899853857395585,
-      "grad_norm": 2.7258620262145996,
-      "learning_rate": 7.436823104693141e-05,
-      "loss": -30.4538,
-      "step": 24700
-    },
-    {
-      "epoch": 1.9036997154065072,
-      "grad_norm": 3.188485860824585,
-      "learning_rate": 7.410851102511494e-05,
-      "loss": -30.4522,
-      "step": 24750
-    },
-    {
-      "epoch": 1.9036997154065072,
-      "eval_loss": -30.41309356689453,
-      "eval_runtime": 16.8987,
-      "eval_samples_per_second": 59.176,
-      "eval_steps_per_second": 14.794,
-      "step": 24750
-    },
-    {
-      "epoch": 1.9075455734174294,
-      "grad_norm": 3.489452600479126,
-      "learning_rate": 7.38591798041711e-05,
-      "loss": -29.9165,
-      "step": 24800
-    },
-    {
-      "epoch": 1.9113914314283518,
-      "grad_norm": 2.326321601867676,
-      "learning_rate": 7.359945978235463e-05,
-      "loss": -30.4499,
-      "step": 24850
-    },
-    {
-      "epoch": 1.9152372894392737,
-      "grad_norm": 3.053919792175293,
-      "learning_rate": 7.333973976053814e-05,
-      "loss": -30.4527,
-      "step": 24900
-    },
-    {
-      "epoch": 1.9190831474501961,
-      "grad_norm": 3.227989435195923,
-      "learning_rate": 7.308001973872166e-05,
-      "loss": -30.4533,
-      "step": 24950
-    },
-    {
-      "epoch": 1.9229290054611183,
-      "grad_norm": 2.026668071746826,
-      "learning_rate": 7.282029971690518e-05,
-      "loss": -30.4532,
-      "step": 25000
-    },
-    {
-      "epoch": 1.9229290054611183,
-      "eval_loss": -30.459980010986328,
-      "eval_runtime": 16.8538,
-      "eval_samples_per_second": 59.334,
-      "eval_steps_per_second": 14.833,
-      "step": 25000
-    },
-    {
-      "epoch": 1.9267748634720405,
-      "grad_norm": 2.0744123458862305,
-      "learning_rate": 7.25605796950887e-05,
-      "loss": -30.4534,
-      "step": 25050
-    },
-    {
-      "epoch": 1.930620721482963,
-      "grad_norm": 1.4959275722503662,
-      "learning_rate": 7.230085967327222e-05,
-      "loss": -30.4542,
-      "step": 25100
-    },
-    {
-      "epoch": 1.9344665794938851,
-      "grad_norm": 2.123647689819336,
-      "learning_rate": 7.204113965145573e-05,
-      "loss": -30.4521,
-      "step": 25150
-    },
-    {
-      "epoch": 1.9383124375048073,
-      "grad_norm": 2.204134702682495,
-      "learning_rate": 7.178141962963926e-05,
-      "loss": -30.4519,
-      "step": 25200
-    },
-    {
-      "epoch": 1.9421582955157297,
-      "grad_norm": 1.7041330337524414,
-      "learning_rate": 7.152169960782277e-05,
-      "loss": -30.4539,
-      "step": 25250
-    },
-    {
-      "epoch": 1.9421582955157297,
-      "eval_loss": -30.45870590209961,
-      "eval_runtime": 16.7732,
-      "eval_samples_per_second": 59.619,
-      "eval_steps_per_second": 14.905,
-      "step": 25250
-    },
-    {
-      "epoch": 1.9460041535266517,
-      "grad_norm": 2.355837345123291,
-      "learning_rate": 7.126197958600629e-05,
-      "loss": -30.4536,
-      "step": 25300
-    },
-    {
-      "epoch": 1.949850011537574,
-      "grad_norm": 2.4595673084259033,
-      "learning_rate": 7.100225956418981e-05,
-      "loss": -30.4529,
-      "step": 25350
-    },
-    {
-      "epoch": 1.9536958695484963,
-      "grad_norm": 3.4583332538604736,
-      "learning_rate": 7.074253954237333e-05,
-      "loss": -30.4544,
-      "step": 25400
-    },
-    {
-      "epoch": 1.9575417275594185,
-      "grad_norm": 2.7398970127105713,
-      "learning_rate": 7.048281952055684e-05,
-      "loss": -30.4548,
-      "step": 25450
-    },
-    {
-      "epoch": 1.9613875855703409,
-      "grad_norm": 2.3612735271453857,
-      "learning_rate": 7.022309949874036e-05,
-      "loss": -30.454,
-      "step": 25500
-    },
-    {
-      "epoch": 1.9613875855703409,
-      "eval_loss": -30.459121704101562,
-      "eval_runtime": 16.7228,
-      "eval_samples_per_second": 59.799,
-      "eval_steps_per_second": 14.95,
-      "step": 25500
-    },
-    {
-      "epoch": 1.9652334435812628,
-      "grad_norm": 2.4880611896514893,
-      "learning_rate": 6.996337947692389e-05,
-      "loss": -30.4548,
-      "step": 25550
-    },
-    {
-      "epoch": 1.9690793015921852,
-      "grad_norm": 1.9865983724594116,
-      "learning_rate": 6.970365945510739e-05,
-      "loss": -30.4522,
-      "step": 25600
-    },
-    {
-      "epoch": 1.9729251596031074,
-      "grad_norm": 2.1370553970336914,
-      "learning_rate": 6.944393943329092e-05,
-      "loss": -30.4546,
-      "step": 25650
-    },
-    {
-      "epoch": 1.9767710176140296,
-      "grad_norm": 2.5807433128356934,
-      "learning_rate": 6.918421941147444e-05,
-      "loss": -30.4552,
-      "step": 25700
-    },
-    {
-      "epoch": 1.980616875624952,
-      "grad_norm": 2.2055130004882812,
-      "learning_rate": 6.892449938965794e-05,
-      "loss": -30.4554,
-      "step": 25750
-    },
-    {
-      "epoch": 1.980616875624952,
-      "eval_loss": -30.4610538482666,
-      "eval_runtime": 16.9285,
-      "eval_samples_per_second": 59.072,
-      "eval_steps_per_second": 14.768,
-      "step": 25750
-    },
-    {
-      "epoch": 1.984462733635874,
-      "grad_norm": 2.3505399227142334,
-      "learning_rate": 6.866477936784147e-05,
-      "loss": -30.4548,
-      "step": 25800
-    },
-    {
-      "epoch": 1.9883085916467964,
-      "grad_norm": 2.190473794937134,
-      "learning_rate": 6.8405059346025e-05,
-      "loss": -30.455,
-      "step": 25850
-    },
-    {
-      "epoch": 1.9921544496577186,
-      "grad_norm": 2.6653695106506348,
-      "learning_rate": 6.814533932420851e-05,
-      "loss": -30.4553,
-      "step": 25900
-    },
-    {
-      "epoch": 1.9960003076686408,
-      "grad_norm": 1.8680397272109985,
-      "learning_rate": 6.788561930239202e-05,
-      "loss": -30.456,
-      "step": 25950
-    },
-    {
-      "epoch": 1.9998461656795632,
-      "grad_norm": 2.7857582569122314,
-      "learning_rate": 6.762589928057555e-05,
-      "loss": -30.4556,
-      "step": 26000
-    },
-    {
-      "epoch": 1.9998461656795632,
-      "eval_loss": -30.46038246154785,
-      "eval_runtime": 16.741,
-      "eval_samples_per_second": 59.734,
-      "eval_steps_per_second": 14.933,
-      "step": 26000
-    },
-    {
-      "epoch": 2.003692023690485,
-      "grad_norm": 1.8869682550430298,
-      "learning_rate": 6.736617925875906e-05,
-      "loss": -30.4556,
-      "step": 26050
-    },
-    {
-      "epoch": 2.0075378817014076,
-      "grad_norm": 3.1017348766326904,
-      "learning_rate": 6.710645923694257e-05,
-      "loss": -30.4554,
-      "step": 26100
-    },
-    {
-      "epoch": 2.01138373971233,
-      "grad_norm": 2.1093878746032715,
-      "learning_rate": 6.68467392151261e-05,
-      "loss": -30.4563,
-      "step": 26150
-    },
-    {
-      "epoch": 2.015229597723252,
-      "grad_norm": 2.2458508014678955,
-      "learning_rate": 6.658701919330963e-05,
-      "loss": -30.4562,
-      "step": 26200
-    },
-    {
-      "epoch": 2.0190754557341744,
-      "grad_norm": 2.072725296020508,
-      "learning_rate": 6.632729917149312e-05,
-      "loss": -30.4556,
-      "step": 26250
-    },
-    {
-      "epoch": 2.0190754557341744,
-      "eval_loss": -30.46011734008789,
-      "eval_runtime": 16.7329,
-      "eval_samples_per_second": 59.762,
-      "eval_steps_per_second": 14.941,
-      "step": 26250
-    },
-    {
-      "epoch": 2.0229213137450963,
-      "grad_norm": 2.431828260421753,
-      "learning_rate": 6.606757914967665e-05,
-      "loss": -30.4549,
-      "step": 26300
-    },
-    {
-      "epoch": 2.0267671717560187,
-      "grad_norm": 3.0434165000915527,
-      "learning_rate": 6.580785912786018e-05,
-      "loss": -30.4562,
-      "step": 26350
-    },
-    {
-      "epoch": 2.030613029766941,
-      "grad_norm": 2.3322978019714355,
-      "learning_rate": 6.554813910604369e-05,
-      "loss": -30.4542,
-      "step": 26400
-    },
-    {
-      "epoch": 2.034458887777863,
-      "grad_norm": 3.0168204307556152,
-      "learning_rate": 6.52884190842272e-05,
-      "loss": -30.4566,
-      "step": 26450
-    },
-    {
-      "epoch": 2.0383047457887855,
-      "grad_norm": 1.6167207956314087,
-      "learning_rate": 6.502869906241073e-05,
-      "loss": -30.4568,
-      "step": 26500
-    },
-    {
-      "epoch": 2.0383047457887855,
-      "eval_loss": -30.460376739501953,
-      "eval_runtime": 16.813,
-      "eval_samples_per_second": 59.478,
-      "eval_steps_per_second": 14.869,
-      "step": 26500
-    },
-    {
-      "epoch": 2.042150603799708,
-      "grad_norm": 2.257894992828369,
-      "learning_rate": 6.476897904059424e-05,
-      "loss": -30.4568,
-      "step": 26550
-    },
-    {
-      "epoch": 2.04599646181063,
-      "grad_norm": 3.5271377563476562,
-      "learning_rate": 6.450925901877776e-05,
-      "loss": -30.4569,
-      "step": 26600
-    },
-    {
-      "epoch": 2.0498423198215523,
-      "grad_norm": 2.4597764015197754,
-      "learning_rate": 6.424953899696128e-05,
-      "loss": -30.4565,
-      "step": 26650
-    },
-    {
-      "epoch": 2.0536881778324743,
-      "grad_norm": 2.209780693054199,
-      "learning_rate": 6.39898189751448e-05,
-      "loss": -30.4567,
-      "step": 26700
-    },
-    {
-      "epoch": 2.0575340358433967,
-      "grad_norm": 2.887272834777832,
-      "learning_rate": 6.373009895332831e-05,
-      "loss": -30.4562,
-      "step": 26750
-    },
-    {
-      "epoch": 2.0575340358433967,
-      "eval_loss": -30.461790084838867,
-      "eval_runtime": 16.8176,
-      "eval_samples_per_second": 59.462,
-      "eval_steps_per_second": 14.865,
-      "step": 26750
-    },
-    {
-      "epoch": 2.061379893854319,
-      "grad_norm": 1.9459458589553833,
-      "learning_rate": 6.347037893151183e-05,
-      "loss": -30.4557,
-      "step": 26800
-    },
-    {
-      "epoch": 2.065225751865241,
-      "grad_norm": 1.4567102193832397,
-      "learning_rate": 6.321065890969535e-05,
-      "loss": -30.4562,
-      "step": 26850
-    },
-    {
-      "epoch": 2.0690716098761635,
-      "grad_norm": 2.8089401721954346,
-      "learning_rate": 6.295093888787887e-05,
-      "loss": -30.4562,
-      "step": 26900
-    },
-    {
-      "epoch": 2.0729174678870854,
-      "grad_norm": 1.6846117973327637,
-      "learning_rate": 6.269121886606239e-05,
-      "loss": -30.4563,
-      "step": 26950
-    },
-    {
-      "epoch": 2.076763325898008,
-      "grad_norm": 2.10868501663208,
-      "learning_rate": 6.243149884424591e-05,
-      "loss": -30.4575,
-      "step": 27000
-    },
-    {
-      "epoch": 2.076763325898008,
-      "eval_loss": -30.46200942993164,
-      "eval_runtime": 16.733,
-      "eval_samples_per_second": 59.762,
-      "eval_steps_per_second": 14.941,
-      "step": 27000
-    },
-    {
-      "epoch": 2.0806091839089302,
-      "grad_norm": 2.005777359008789,
-      "learning_rate": 6.217177882242942e-05,
-      "loss": -30.457,
-      "step": 27050
-    },
-    {
-      "epoch": 2.084455041919852,
-      "grad_norm": 3.0324509143829346,
-      "learning_rate": 6.191205880061294e-05,
-      "loss": -30.4566,
-      "step": 27100
-    },
-    {
-      "epoch": 2.0883008999307746,
-      "grad_norm": 2.6647162437438965,
-      "learning_rate": 6.165233877879646e-05,
-      "loss": -30.4574,
-      "step": 27150
-    },
-    {
-      "epoch": 2.0921467579416966,
-      "grad_norm": 2.078568458557129,
-      "learning_rate": 6.139261875697998e-05,
-      "loss": -30.4579,
-      "step": 27200
-    },
-    {
-      "epoch": 2.095992615952619,
-      "grad_norm": 1.8361953496932983,
-      "learning_rate": 6.11328987351635e-05,
-      "loss": -30.4566,
-      "step": 27250
-    },
-    {
-      "epoch": 2.095992615952619,
-      "eval_loss": -30.461936950683594,
-      "eval_runtime": 16.9333,
-      "eval_samples_per_second": 59.055,
-      "eval_steps_per_second": 14.764,
-      "step": 27250
-    },
-    {
-      "epoch": 2.0998384739635414,
-      "grad_norm": 2.176724433898926,
-      "learning_rate": 6.087317871334701e-05,
-      "loss": -30.4579,
-      "step": 27300
-    },
-    {
-      "epoch": 2.1036843319744634,
-      "grad_norm": 2.137570858001709,
-      "learning_rate": 6.0613458691530536e-05,
-      "loss": -30.4578,
-      "step": 27350
-    },
-    {
-      "epoch": 2.1075301899853858,
-      "grad_norm": 2.3757429122924805,
-      "learning_rate": 6.0353738669714055e-05,
-      "loss": -30.4576,
-      "step": 27400
-    },
-    {
-      "epoch": 2.111376047996308,
-      "grad_norm": 2.0854310989379883,
-      "learning_rate": 6.009401864789757e-05,
-      "loss": -30.4571,
-      "step": 27450
-    },
-    {
-      "epoch": 2.11522190600723,
-      "grad_norm": 2.1995458602905273,
-      "learning_rate": 5.983429862608109e-05,
-      "loss": -30.4575,
-      "step": 27500
-    },
-    {
-      "epoch": 2.11522190600723,
-      "eval_loss": -30.462617874145508,
-      "eval_runtime": 16.8806,
-      "eval_samples_per_second": 59.24,
-      "eval_steps_per_second": 14.81,
-      "step": 27500
-    },
-    {
-      "epoch": 2.1190677640181526,
-      "grad_norm": 1.9954910278320312,
-      "learning_rate": 5.957457860426461e-05,
-      "loss": -30.458,
-      "step": 27550
-    },
-    {
-      "epoch": 2.1229136220290745,
-      "grad_norm": 1.2620997428894043,
-      "learning_rate": 5.931485858244812e-05,
-      "loss": -30.4578,
-      "step": 27600
-    },
-    {
-      "epoch": 2.126759480039997,
-      "grad_norm": 1.8231252431869507,
-      "learning_rate": 5.905513856063164e-05,
-      "loss": -30.4571,
-      "step": 27650
-    },
-    {
-      "epoch": 2.1306053380509193,
-      "grad_norm": 2.8753602504730225,
-      "learning_rate": 5.879541853881516e-05,
-      "loss": -30.4576,
-      "step": 27700
-    },
-    {
-      "epoch": 2.1344511960618413,
-      "grad_norm": 2.1304397583007812,
-      "learning_rate": 5.8535698516998686e-05,
-      "loss": -30.4579,
-      "step": 27750
-    },
-    {
-      "epoch": 2.1344511960618413,
-      "eval_loss": -30.463788986206055,
-      "eval_runtime": 16.8747,
-      "eval_samples_per_second": 59.26,
-      "eval_steps_per_second": 14.815,
-      "step": 27750
-    },
-    {
-      "epoch": 2.1382970540727637,
-      "grad_norm": 2.8373219966888428,
-      "learning_rate": 5.827597849518219e-05,
-      "loss": -30.458,
-      "step": 27800
-    },
-    {
-      "epoch": 2.1421429120836857,
-      "grad_norm": 2.214996576309204,
-      "learning_rate": 5.801625847336571e-05,
-      "loss": -30.4577,
-      "step": 27850
-    },
-    {
-      "epoch": 2.145988770094608,
-      "grad_norm": 2.79176664352417,
-      "learning_rate": 5.775653845154924e-05,
-      "loss": -30.4586,
-      "step": 27900
-    },
-    {
-      "epoch": 2.1498346281055305,
-      "grad_norm": 3.1659934520721436,
-      "learning_rate": 5.7496818429732744e-05,
-      "loss": -30.4584,
-      "step": 27950
-    },
-    {
-      "epoch": 2.1536804861164525,
-      "grad_norm": 3.1562278270721436,
-      "learning_rate": 5.723709840791627e-05,
-      "loss": -30.4578,
-      "step": 28000
-    },
-    {
-      "epoch": 2.1536804861164525,
-      "eval_loss": -30.461944580078125,
-      "eval_runtime": 16.8316,
-      "eval_samples_per_second": 59.412,
-      "eval_steps_per_second": 14.853,
-      "step": 28000
-    },
-    {
-      "epoch": 2.157526344127375,
-      "grad_norm": 1.7737715244293213,
-      "learning_rate": 5.697737838609979e-05,
-      "loss": -30.4579,
-      "step": 28050
-    },
-    {
-      "epoch": 2.1613722021382973,
-      "grad_norm": 2.3100192546844482,
-      "learning_rate": 5.671765836428331e-05,
-      "loss": -30.4582,
-      "step": 28100
-    },
-    {
-      "epoch": 2.1652180601492192,
-      "grad_norm": 2.7968549728393555,
-      "learning_rate": 5.645793834246682e-05,
-      "loss": -30.4588,
-      "step": 28150
-    },
-    {
-      "epoch": 2.1690639181601417,
-      "grad_norm": 1.6727192401885986,
-      "learning_rate": 5.619821832065034e-05,
-      "loss": -30.4585,
-      "step": 28200
-    },
-    {
-      "epoch": 2.1729097761710636,
-      "grad_norm": 1.8535244464874268,
-      "learning_rate": 5.593849829883386e-05,
-      "loss": -30.4589,
-      "step": 28250
-    },
-    {
-      "epoch": 2.1729097761710636,
-      "eval_loss": -30.463205337524414,
-      "eval_runtime": 16.861,
-      "eval_samples_per_second": 59.308,
-      "eval_steps_per_second": 14.827,
-      "step": 28250
-    },
-    {
-      "epoch": 2.176755634181986,
-      "grad_norm": 1.9916008710861206,
-      "learning_rate": 5.5678778277017375e-05,
-      "loss": -30.4585,
-      "step": 28300
-    },
-    {
-      "epoch": 2.1806014921929084,
-      "grad_norm": 1.7134790420532227,
-      "learning_rate": 5.5419058255200894e-05,
-      "loss": -30.4589,
-      "step": 28350
-    },
-    {
-      "epoch": 2.1844473502038304,
-      "grad_norm": 1.7053008079528809,
-      "learning_rate": 5.5159338233384414e-05,
-      "loss": -30.4587,
-      "step": 28400
-    },
-    {
-      "epoch": 2.188293208214753,
-      "grad_norm": 2.964104413986206,
-      "learning_rate": 5.489961821156793e-05,
-      "loss": -30.4586,
-      "step": 28450
-    },
-    {
-      "epoch": 2.192139066225675,
-      "grad_norm": 1.901634931564331,
-      "learning_rate": 5.4639898189751446e-05,
-      "loss": -30.4595,
-      "step": 28500
-    },
-    {
-      "epoch": 2.192139066225675,
-      "eval_loss": -30.463890075683594,
-      "eval_runtime": 16.9187,
-      "eval_samples_per_second": 59.106,
-      "eval_steps_per_second": 14.777,
-      "step": 28500
-    },
-    {
-      "epoch": 2.195984924236597,
-      "grad_norm": 1.420623779296875,
-      "learning_rate": 5.438017816793497e-05,
-      "loss": -30.4589,
-      "step": 28550
-    },
-    {
-      "epoch": 2.1998307822475196,
-      "grad_norm": 2.0934250354766846,
-      "learning_rate": 5.412045814611849e-05,
-      "loss": -30.4585,
-      "step": 28600
-    },
-    {
-      "epoch": 2.2036766402584416,
-      "grad_norm": 1.6350470781326294,
-      "learning_rate": 5.3860738124302e-05,
-      "loss": -30.459,
-      "step": 28650
-    },
-    {
-      "epoch": 2.207522498269364,
-      "grad_norm": 1.5302081108093262,
-      "learning_rate": 5.3601018102485525e-05,
-      "loss": -30.4597,
-      "step": 28700
-    },
-    {
-      "epoch": 2.211368356280286,
-      "grad_norm": 1.9414024353027344,
-      "learning_rate": 5.3341298080669044e-05,
-      "loss": -30.4578,
-      "step": 28750
-    },
-    {
-      "epoch": 2.211368356280286,
-      "eval_loss": -30.459186553955078,
-      "eval_runtime": 16.8332,
-      "eval_samples_per_second": 59.406,
-      "eval_steps_per_second": 14.852,
-      "step": 28750
-    },
-    {
-      "epoch": 2.2152142142912084,
-      "grad_norm": 2.1578445434570312,
-      "learning_rate": 5.308157805885256e-05,
-      "loss": -30.4588,
-      "step": 28800
-    },
-    {
-      "epoch": 2.2190600723021308,
-      "grad_norm": 1.992924690246582,
-      "learning_rate": 5.282185803703608e-05,
-      "loss": -30.4565,
-      "step": 28850
-    },
-    {
-      "epoch": 2.2229059303130527,
-      "grad_norm": 1.7855414152145386,
-      "learning_rate": 5.2562138015219596e-05,
-      "loss": -30.4597,
-      "step": 28900
-    },
-    {
-      "epoch": 2.226751788323975,
-      "grad_norm": 2.323654890060425,
-      "learning_rate": 5.2302417993403116e-05,
-      "loss": -30.4597,
-      "step": 28950
-    },
-    {
-      "epoch": 2.230597646334897,
-      "grad_norm": 2.070629835128784,
-      "learning_rate": 5.204269797158663e-05,
-      "loss": -30.4596,
-      "step": 29000
-    },
-    {
-      "epoch": 2.230597646334897,
-      "eval_loss": -30.4643497467041,
-      "eval_runtime": 16.8617,
-      "eval_samples_per_second": 59.306,
-      "eval_steps_per_second": 14.827,
-      "step": 29000
-    },
-    {
-      "epoch": 2.2344435043458195,
-      "grad_norm": 2.0510427951812744,
-      "learning_rate": 5.178297794977015e-05,
-      "loss": -30.4589,
-      "step": 29050
-    },
-    {
-      "epoch": 2.238289362356742,
-      "grad_norm": 1.5907058715820312,
-      "learning_rate": 5.1523257927953675e-05,
-      "loss": -30.4597,
-      "step": 29100
-    },
-    {
-      "epoch": 2.242135220367664,
-      "grad_norm": 2.039212226867676,
-      "learning_rate": 5.126353790613718e-05,
-      "loss": -30.4604,
-      "step": 29150
-    },
-    {
-      "epoch": 2.2459810783785863,
-      "grad_norm": 1.4781228303909302,
-      "learning_rate": 5.10038178843207e-05,
-      "loss": -30.4605,
-      "step": 29200
-    },
-    {
-      "epoch": 2.2498269363895087,
-      "grad_norm": 1.5673744678497314,
-      "learning_rate": 5.074409786250423e-05,
-      "loss": -30.4595,
-      "step": 29250
-    },
-    {
-      "epoch": 2.2498269363895087,
-      "eval_loss": -30.46396827697754,
-      "eval_runtime": 16.8497,
-      "eval_samples_per_second": 59.348,
-      "eval_steps_per_second": 14.837,
-      "step": 29250
-    },
-    {
-      "epoch": 2.2536727944004307,
-      "grad_norm": 1.5207022428512573,
-      "learning_rate": 5.048437784068775e-05,
-      "loss": -30.4592,
-      "step": 29300
-    },
-    {
-      "epoch": 2.257518652411353,
-      "grad_norm": 1.9671216011047363,
-      "learning_rate": 5.022465781887126e-05,
-      "loss": -30.46,
-      "step": 29350
-    },
-    {
-      "epoch": 2.261364510422275,
-      "grad_norm": 1.6448663473129272,
-      "learning_rate": 4.996493779705478e-05,
-      "loss": -30.4604,
-      "step": 29400
-    },
-    {
-      "epoch": 2.2652103684331975,
-      "grad_norm": 2.366356372833252,
-      "learning_rate": 4.970521777523829e-05,
-      "loss": -30.4599,
-      "step": 29450
-    },
-    {
-      "epoch": 2.26905622644412,
-      "grad_norm": 2.137887716293335,
-      "learning_rate": 4.944549775342182e-05,
-      "loss": -30.4605,
-      "step": 29500
-    },
-    {
-      "epoch": 2.26905622644412,
-      "eval_loss": -30.46525001525879,
-      "eval_runtime": 16.856,
-      "eval_samples_per_second": 59.326,
-      "eval_steps_per_second": 14.832,
-      "step": 29500
-    },
-    {
-      "epoch": 2.272902084455042,
-      "grad_norm": 2.2537872791290283,
-      "learning_rate": 4.918577773160533e-05,
-      "loss": -30.4599,
-      "step": 29550
-    },
-    {
-      "epoch": 2.2767479424659642,
-      "grad_norm": 1.9328151941299438,
-      "learning_rate": 4.8926057709788844e-05,
-      "loss": -30.4602,
-      "step": 29600
-    },
-    {
-      "epoch": 2.280593800476886,
-      "grad_norm": 2.076399326324463,
-      "learning_rate": 4.866633768797237e-05,
-      "loss": -30.4607,
-      "step": 29650
-    },
-    {
-      "epoch": 2.2844396584878086,
-      "grad_norm": 1.552717924118042,
-      "learning_rate": 4.840661766615588e-05,
-      "loss": -30.46,
-      "step": 29700
-    },
-    {
-      "epoch": 2.288285516498731,
-      "grad_norm": 1.7150416374206543,
-      "learning_rate": 4.81468976443394e-05,
-      "loss": -30.4601,
-      "step": 29750
-    },
-    {
-      "epoch": 2.288285516498731,
-      "eval_loss": -30.464597702026367,
-      "eval_runtime": 16.899,
-      "eval_samples_per_second": 59.175,
-      "eval_steps_per_second": 14.794,
-      "step": 29750
-    },
-    {
-      "epoch": 2.292131374509653,
-      "grad_norm": 2.387169122695923,
-      "learning_rate": 4.788717762252292e-05,
-      "loss": -30.4604,
-      "step": 29800
-    },
-    {
-      "epoch": 2.2959772325205754,
-      "grad_norm": 2.8682591915130615,
-      "learning_rate": 4.762745760070644e-05,
-      "loss": -30.4607,
-      "step": 29850
-    },
-    {
-      "epoch": 2.299823090531498,
-      "grad_norm": 2.2639787197113037,
-      "learning_rate": 4.736773757888996e-05,
-      "loss": -30.4605,
-      "step": 29900
-    },
-    {
-      "epoch": 2.3036689485424198,
-      "grad_norm": 1.8516710996627808,
-      "learning_rate": 4.7108017557073475e-05,
-      "loss": -30.4601,
-      "step": 29950
-    },
-    {
-      "epoch": 2.307514806553342,
-      "grad_norm": 2.0204548835754395,
-      "learning_rate": 4.6848297535256994e-05,
-      "loss": -30.4602,
-      "step": 30000
-    },
-    {
-      "epoch": 2.307514806553342,
-      "eval_loss": -30.46541976928711,
-      "eval_runtime": 16.9539,
-      "eval_samples_per_second": 58.983,
-      "eval_steps_per_second": 14.746,
-      "step": 30000
-    },
-    {
-      "epoch": 2.311360664564264,
-      "grad_norm": 1.7279092073440552,
-      "learning_rate": 4.6588577513440514e-05,
-      "loss": -30.4612,
-      "step": 30050
-    },
-    {
-      "epoch": 2.3152065225751866,
-      "grad_norm": 1.8211647272109985,
-      "learning_rate": 4.632885749162403e-05,
-      "loss": -30.4612,
-      "step": 30100
-    },
-    {
-      "epoch": 2.3190523805861085,
-      "grad_norm": 2.0550918579101562,
-      "learning_rate": 4.6069137469807546e-05,
-      "loss": -30.4605,
-      "step": 30150
-    },
-    {
-      "epoch": 2.322898238597031,
-      "grad_norm": 1.8465496301651,
-      "learning_rate": 4.5809417447991066e-05,
-      "loss": -30.461,
-      "step": 30200
-    },
-    {
-      "epoch": 2.3267440966079533,
-      "grad_norm": 2.338982582092285,
-      "learning_rate": 4.5549697426174585e-05,
-      "loss": -30.4609,
-      "step": 30250
-    },
-    {
-      "epoch": 2.3267440966079533,
-      "eval_loss": -30.466022491455078,
-      "eval_runtime": 16.8437,
-      "eval_samples_per_second": 59.37,
-      "eval_steps_per_second": 14.842,
-      "step": 30250
-    },
-    {
-      "epoch": 2.3305899546188753,
-      "grad_norm": 2.0003364086151123,
-      "learning_rate": 4.5289977404358105e-05,
-      "loss": -30.4608,
-      "step": 30300
-    },
-    {
-      "epoch": 2.3344358126297977,
-      "grad_norm": 1.2678554058074951,
-      "learning_rate": 4.5030257382541625e-05,
-      "loss": -30.4611,
-      "step": 30350
-    },
-    {
-      "epoch": 2.33828167064072,
-      "grad_norm": 1.6590012311935425,
-      "learning_rate": 4.477053736072514e-05,
-      "loss": -30.4602,
-      "step": 30400
-    },
-    {
-      "epoch": 2.342127528651642,
-      "grad_norm": 2.2486422061920166,
-      "learning_rate": 4.4510817338908664e-05,
-      "loss": -30.4607,
-      "step": 30450
-    },
-    {
-      "epoch": 2.3459733866625645,
-      "grad_norm": 1.6569013595581055,
-      "learning_rate": 4.425109731709218e-05,
-      "loss": -30.4617,
-      "step": 30500
-    },
-    {
-      "epoch": 2.3459733866625645,
-      "eval_loss": -30.465835571289062,
-      "eval_runtime": 17.0015,
-      "eval_samples_per_second": 58.818,
-      "eval_steps_per_second": 14.705,
-      "step": 30500
-    },
-    {
-      "epoch": 2.3498192446734865,
-      "grad_norm": 1.567749261856079,
-      "learning_rate": 4.3991377295275696e-05,
-      "loss": -30.4615,
-      "step": 30550
-    },
-    {
-      "epoch": 2.353665102684409,
-      "grad_norm": 1.2779616117477417,
-      "learning_rate": 4.3731657273459216e-05,
-      "loss": -30.4614,
-      "step": 30600
-    },
-    {
-      "epoch": 2.3575109606953313,
-      "grad_norm": 1.9518656730651855,
-      "learning_rate": 4.347193725164273e-05,
-      "loss": -30.4614,
-      "step": 30650
-    },
-    {
-      "epoch": 2.3613568187062532,
-      "grad_norm": 1.6477199792861938,
-      "learning_rate": 4.321221722982625e-05,
-      "loss": -30.4617,
-      "step": 30700
-    },
-    {
-      "epoch": 2.3652026767171757,
-      "grad_norm": 1.1383932828903198,
-      "learning_rate": 4.295249720800977e-05,
-      "loss": -30.462,
-      "step": 30750
-    },
-    {
-      "epoch": 2.3652026767171757,
-      "eval_loss": -30.466596603393555,
-      "eval_runtime": 16.9152,
-      "eval_samples_per_second": 59.119,
-      "eval_steps_per_second": 14.78,
-      "step": 30750
-    },
-    {
-      "epoch": 2.3690485347280976,
-      "grad_norm": 1.8575818538665771,
-      "learning_rate": 4.269277718619328e-05,
-      "loss": -30.4619,
-      "step": 30800
-    },
-    {
-      "epoch": 2.37289439273902,
-      "grad_norm": 1.8746001720428467,
-      "learning_rate": 4.243305716437681e-05,
-      "loss": -30.4617,
-      "step": 30850
-    },
-    {
-      "epoch": 2.3767402507499424,
-      "grad_norm": 1.870802402496338,
-      "learning_rate": 4.217333714256032e-05,
-      "loss": -30.4619,
-      "step": 30900
-    },
-    {
-      "epoch": 2.3805861087608644,
-      "grad_norm": 1.6408015489578247,
-      "learning_rate": 4.191361712074384e-05,
-      "loss": -30.462,
-      "step": 30950
-    },
-    {
-      "epoch": 2.384431966771787,
-      "grad_norm": 1.2397960424423218,
-      "learning_rate": 4.165389709892736e-05,
-      "loss": -30.4618,
-      "step": 31000
-    },
-    {
-      "epoch": 2.384431966771787,
-      "eval_loss": -30.46613121032715,
-      "eval_runtime": 16.8356,
-      "eval_samples_per_second": 59.398,
-      "eval_steps_per_second": 14.85,
-      "step": 31000
-    },
-    {
-      "epoch": 2.3882778247827092,
-      "grad_norm": 1.8404053449630737,
-      "learning_rate": 4.139417707711087e-05,
-      "loss": -30.462,
-      "step": 31050
-    },
-    {
-      "epoch": 2.392123682793631,
-      "grad_norm": 2.467785596847534,
-      "learning_rate": 4.11344570552944e-05,
-      "loss": -30.4619,
-      "step": 31100
-    },
-    {
-      "epoch": 2.3959695408045536,
-      "grad_norm": 2.1997389793395996,
-      "learning_rate": 4.087473703347791e-05,
-      "loss": -30.462,
-      "step": 31150
-    },
-    {
-      "epoch": 2.3998153988154756,
-      "grad_norm": 2.1173219680786133,
-      "learning_rate": 4.061501701166143e-05,
-      "loss": -30.4621,
-      "step": 31200
-    },
-    {
-      "epoch": 2.403661256826398,
-      "grad_norm": 1.7325748205184937,
-      "learning_rate": 4.035529698984495e-05,
-      "loss": -30.4621,
-      "step": 31250
-    },
-    {
-      "epoch": 2.403661256826398,
-      "eval_loss": -30.46693229675293,
-      "eval_runtime": 16.8671,
-      "eval_samples_per_second": 59.287,
-      "eval_steps_per_second": 14.822,
-      "step": 31250
-    },
-    {
-      "epoch": 2.4075071148373204,
-      "grad_norm": 1.5806682109832764,
-      "learning_rate": 4.009557696802847e-05,
-      "loss": -30.4619,
-      "step": 31300
-    },
-    {
-      "epoch": 2.4113529728482423,
-      "grad_norm": 2.6714529991149902,
-      "learning_rate": 3.983585694621198e-05,
-      "loss": -30.4617,
-      "step": 31350
-    },
-    {
-      "epoch": 2.4151988308591648,
-      "grad_norm": 1.4386422634124756,
-      "learning_rate": 3.95761369243955e-05,
-      "loss": -30.462,
-      "step": 31400
-    },
-    {
-      "epoch": 2.4190446888700867,
-      "grad_norm": 1.7787779569625854,
-      "learning_rate": 3.931641690257902e-05,
-      "loss": -30.462,
-      "step": 31450
-    },
-    {
-      "epoch": 2.422890546881009,
-      "grad_norm": 1.5804333686828613,
-      "learning_rate": 3.905669688076254e-05,
-      "loss": -30.4619,
-      "step": 31500
-    },
-    {
-      "epoch": 2.422890546881009,
-      "eval_loss": -30.466127395629883,
-      "eval_runtime": 16.9571,
-      "eval_samples_per_second": 58.972,
-      "eval_steps_per_second": 14.743,
-      "step": 31500
-    },
-    {
-      "epoch": 2.4267364048919315,
-      "grad_norm": 2.5385239124298096,
-      "learning_rate": 3.879697685894606e-05,
-      "loss": -30.4623,
-      "step": 31550
-    },
-    {
-      "epoch": 2.4305822629028535,
-      "grad_norm": 2.142542839050293,
-      "learning_rate": 3.8537256837129574e-05,
-      "loss": -30.4624,
-      "step": 31600
-    },
-    {
-      "epoch": 2.434428120913776,
-      "grad_norm": 1.5194621086120605,
-      "learning_rate": 3.8277536815313094e-05,
-      "loss": -30.4625,
-      "step": 31650
-    },
-    {
-      "epoch": 2.4382739789246983,
-      "grad_norm": 1.2033811807632446,
-      "learning_rate": 3.8017816793496614e-05,
-      "loss": -30.4624,
-      "step": 31700
-    },
-    {
-      "epoch": 2.4421198369356203,
-      "grad_norm": 3.1859564781188965,
-      "learning_rate": 3.7758096771680126e-05,
-      "loss": -30.4621,
-      "step": 31750
-    },
-    {
-      "epoch": 2.4421198369356203,
-      "eval_loss": -30.46595573425293,
-      "eval_runtime": 16.8541,
-      "eval_samples_per_second": 59.333,
-      "eval_steps_per_second": 14.833,
-      "step": 31750
-    },
-    {
-      "epoch": 2.4459656949465427,
-      "grad_norm": 2.844355821609497,
-      "learning_rate": 3.749837674986365e-05,
-      "loss": -30.4622,
-      "step": 31800
-    },
-    {
-      "epoch": 2.4498115529574647,
-      "grad_norm": 1.9936915636062622,
-      "learning_rate": 3.7238656728047166e-05,
-      "loss": -30.4626,
-      "step": 31850
-    },
-    {
-      "epoch": 2.453657410968387,
-      "grad_norm": 1.0412544012069702,
-      "learning_rate": 3.6978936706230685e-05,
-      "loss": -30.4626,
-      "step": 31900
-    },
-    {
-      "epoch": 2.457503268979309,
-      "grad_norm": 2.455996513366699,
-      "learning_rate": 3.6719216684414205e-05,
-      "loss": -30.4624,
-      "step": 31950
-    },
-    {
-      "epoch": 2.4613491269902315,
-      "grad_norm": 1.5153650045394897,
-      "learning_rate": 3.645949666259772e-05,
-      "loss": -30.4627,
-      "step": 32000
-    },
-    {
-      "epoch": 2.4613491269902315,
-      "eval_loss": -30.466768264770508,
-      "eval_runtime": 17.0502,
-      "eval_samples_per_second": 58.65,
-      "eval_steps_per_second": 14.663,
-      "step": 32000
-    },
-    {
-      "epoch": 2.465194985001154,
-      "grad_norm": 1.4773210287094116,
-      "learning_rate": 3.6199776640781244e-05,
-      "loss": -30.4631,
-      "step": 32050
-    },
-    {
-      "epoch": 2.469040843012076,
-      "grad_norm": 2.181157112121582,
-      "learning_rate": 3.594005661896476e-05,
-      "loss": -30.4627,
-      "step": 32100
-    },
-    {
-      "epoch": 2.4728867010229982,
-      "grad_norm": 1.5130398273468018,
-      "learning_rate": 3.5680336597148277e-05,
-      "loss": -30.463,
-      "step": 32150
-    },
-    {
-      "epoch": 2.4767325590339206,
-      "grad_norm": 1.2299845218658447,
-      "learning_rate": 3.5420616575331796e-05,
-      "loss": -30.4632,
-      "step": 32200
-    },
-    {
-      "epoch": 2.4805784170448426,
-      "grad_norm": 1.3962304592132568,
-      "learning_rate": 3.516089655351531e-05,
-      "loss": -30.4632,
-      "step": 32250
-    },
-    {
-      "epoch": 2.4805784170448426,
-      "eval_loss": -30.467247009277344,
-      "eval_runtime": 16.7784,
-      "eval_samples_per_second": 59.6,
-      "eval_steps_per_second": 14.9,
-      "step": 32250
-    },
-    {
-      "epoch": 2.484424275055765,
-      "grad_norm": 1.6033949851989746,
-      "learning_rate": 3.490117653169883e-05,
-      "loss": -30.4632,
-      "step": 32300
-    },
-    {
-      "epoch": 2.488270133066687,
-      "grad_norm": 1.4287469387054443,
-      "learning_rate": 3.464145650988235e-05,
-      "loss": -30.4626,
-      "step": 32350
-    },
-    {
-      "epoch": 2.4921159910776094,
-      "grad_norm": 2.5753841400146484,
-      "learning_rate": 3.438173648806587e-05,
-      "loss": -30.4633,
-      "step": 32400
-    },
-    {
-      "epoch": 2.495961849088532,
-      "grad_norm": 1.5068596601486206,
-      "learning_rate": 3.412201646624939e-05,
-      "loss": -30.4631,
-      "step": 32450
-    },
-    {
-      "epoch": 2.4998077070994538,
-      "grad_norm": 1.6092793941497803,
-      "learning_rate": 3.38622964444329e-05,
-      "loss": -30.4633,
-      "step": 32500
-    },
-    {
-      "epoch": 2.4998077070994538,
-      "eval_loss": -30.467578887939453,
-      "eval_runtime": 16.8772,
-      "eval_samples_per_second": 59.252,
-      "eval_steps_per_second": 14.813,
-      "step": 32500
-    },
-    {
-      "epoch": 2.503653565110376,
-      "grad_norm": 0.964199423789978,
-      "learning_rate": 3.360257642261642e-05,
-      "loss": -30.463,
-      "step": 32550
-    },
-    {
-      "epoch": 2.507499423121298,
-      "grad_norm": 2.097522497177124,
-      "learning_rate": 3.334285640079994e-05,
-      "loss": -30.4637,
-      "step": 32600
-    },
-    {
-      "epoch": 2.5113452811322206,
-      "grad_norm": 1.688252568244934,
-      "learning_rate": 3.308313637898346e-05,
-      "loss": -30.4634,
-      "step": 32650
-    },
-    {
-      "epoch": 2.515191139143143,
-      "grad_norm": 1.733709454536438,
-      "learning_rate": 3.282341635716697e-05,
-      "loss": -30.4629,
-      "step": 32700
-    },
-    {
-      "epoch": 2.519036997154065,
-      "grad_norm": 1.897979974746704,
-      "learning_rate": 3.25636963353505e-05,
-      "loss": -30.4629,
-      "step": 32750
-    },
-    {
-      "epoch": 2.519036997154065,
-      "eval_loss": -30.46739387512207,
-      "eval_runtime": 16.8323,
-      "eval_samples_per_second": 59.41,
-      "eval_steps_per_second": 14.852,
-      "step": 32750
-    },
-    {
-      "epoch": 2.5228828551649873,
-      "grad_norm": 0.9445510506629944,
-      "learning_rate": 3.230397631353401e-05,
-      "loss": -30.4633,
-      "step": 32800
-    },
-    {
-      "epoch": 2.5267287131759097,
-      "grad_norm": 1.7918905019760132,
-      "learning_rate": 3.204425629171753e-05,
-      "loss": -30.4636,
-      "step": 32850
-    },
-    {
-      "epoch": 2.5305745711868317,
-      "grad_norm": 2.108637571334839,
-      "learning_rate": 3.178453626990105e-05,
-      "loss": -30.4635,
-      "step": 32900
-    },
-    {
-      "epoch": 2.534420429197754,
-      "grad_norm": 1.0502158403396606,
-      "learning_rate": 3.152481624808456e-05,
-      "loss": -30.4638,
-      "step": 32950
-    },
-    {
-      "epoch": 2.5382662872086765,
-      "grad_norm": 1.5265525579452515,
-      "learning_rate": 3.126509622626809e-05,
-      "loss": -30.4632,
-      "step": 33000
-    },
-    {
-      "epoch": 2.5382662872086765,
-      "eval_loss": -30.46713638305664,
-      "eval_runtime": 16.8812,
-      "eval_samples_per_second": 59.237,
-      "eval_steps_per_second": 14.809,
-      "step": 33000
-    },
-    {
-      "epoch": 2.5421121452195985,
-      "grad_norm": 1.2125709056854248,
-      "learning_rate": 3.10053762044516e-05,
-      "loss": -30.4629,
-      "step": 33050
-    },
-    {
-      "epoch": 2.5459580032305205,
-      "grad_norm": 1.4404875040054321,
-      "learning_rate": 3.0745656182635115e-05,
-      "loss": -30.464,
-      "step": 33100
-    },
-    {
-      "epoch": 2.549803861241443,
-      "grad_norm": 1.889853835105896,
-      "learning_rate": 3.048593616081864e-05,
-      "loss": -30.4633,
-      "step": 33150
-    },
-    {
-      "epoch": 2.5536497192523653,
-      "grad_norm": 1.402807593345642,
-      "learning_rate": 3.0226216139002155e-05,
-      "loss": -30.4639,
-      "step": 33200
-    },
-    {
-      "epoch": 2.5574955772632872,
-      "grad_norm": 1.044944405555725,
-      "learning_rate": 2.9966496117185678e-05,
-      "loss": -30.4639,
-      "step": 33250
-    },
-    {
-      "epoch": 2.5574955772632872,
-      "eval_loss": -30.468294143676758,
-      "eval_runtime": 16.9528,
-      "eval_samples_per_second": 58.987,
-      "eval_steps_per_second": 14.747,
-      "step": 33250
-    },
-    {
-      "epoch": 2.5613414352742097,
-      "grad_norm": 1.4782401323318481,
-      "learning_rate": 2.9706776095369194e-05,
-      "loss": -30.4639,
-      "step": 33300
-    },
-    {
-      "epoch": 2.565187293285132,
-      "grad_norm": 1.0335558652877808,
-      "learning_rate": 2.944705607355271e-05,
-      "loss": -30.4635,
-      "step": 33350
-    },
-    {
-      "epoch": 2.569033151296054,
-      "grad_norm": 1.1205657720565796,
-      "learning_rate": 2.918733605173623e-05,
-      "loss": -30.4643,
-      "step": 33400
-    },
-    {
-      "epoch": 2.5728790093069764,
-      "grad_norm": 1.6154303550720215,
-      "learning_rate": 2.8927616029919746e-05,
-      "loss": -30.4641,
-      "step": 33450
-    },
-    {
-      "epoch": 2.576724867317899,
-      "grad_norm": 1.305467128753662,
-      "learning_rate": 2.866789600810327e-05,
-      "loss": -30.4642,
-      "step": 33500
-    },
-    {
-      "epoch": 2.576724867317899,
-      "eval_loss": -30.468229293823242,
-      "eval_runtime": 16.8335,
-      "eval_samples_per_second": 59.405,
-      "eval_steps_per_second": 14.851,
-      "step": 33500
-    },
-    {
-      "epoch": 2.580570725328821,
-      "grad_norm": 1.4428735971450806,
-      "learning_rate": 2.8408175986286785e-05,
-      "loss": -30.4636,
-      "step": 33550
-    },
-    {
-      "epoch": 2.5844165833397432,
-      "grad_norm": 1.3220546245574951,
-      "learning_rate": 2.8148455964470305e-05,
-      "loss": -30.4638,
-      "step": 33600
-    },
-    {
-      "epoch": 2.588262441350665,
-      "grad_norm": 1.3444154262542725,
-      "learning_rate": 2.788873594265382e-05,
-      "loss": -30.464,
-      "step": 33650
-    },
-    {
-      "epoch": 2.5921082993615876,
-      "grad_norm": 1.2507829666137695,
-      "learning_rate": 2.7629015920837337e-05,
-      "loss": -30.464,
-      "step": 33700
-    },
-    {
-      "epoch": 2.5959541573725096,
-      "grad_norm": 1.4041513204574585,
-      "learning_rate": 2.7369295899020857e-05,
-      "loss": -30.4643,
-      "step": 33750
-    },
-    {
-      "epoch": 2.5959541573725096,
-      "eval_loss": -30.468578338623047,
-      "eval_runtime": 16.9279,
-      "eval_samples_per_second": 59.074,
-      "eval_steps_per_second": 14.769,
-      "step": 33750
-    },
-    {
-      "epoch": 2.599800015383432,
-      "grad_norm": 1.2238858938217163,
-      "learning_rate": 2.7109575877204373e-05,
-      "loss": -30.4641,
-      "step": 33800
-    },
-    {
-      "epoch": 2.6036458733943544,
-      "grad_norm": 0.8141158819198608,
-      "learning_rate": 2.6849855855387896e-05,
-      "loss": -30.4641,
-      "step": 33850
-    },
-    {
-      "epoch": 2.6074917314052763,
-      "grad_norm": 3.0219566822052,
-      "learning_rate": 2.6590135833571412e-05,
-      "loss": -30.4642,
-      "step": 33900
-    },
-    {
-      "epoch": 2.6113375894161988,
-      "grad_norm": 1.9201821088790894,
-      "learning_rate": 2.633041581175493e-05,
-      "loss": -30.4641,
-      "step": 33950
-    },
-    {
-      "epoch": 2.615183447427121,
-      "grad_norm": 1.4936727285385132,
-      "learning_rate": 2.6070695789938448e-05,
-      "loss": -30.4645,
-      "step": 34000
-    },
-    {
-      "epoch": 2.615183447427121,
-      "eval_loss": -30.46862030029297,
-      "eval_runtime": 16.8191,
-      "eval_samples_per_second": 59.456,
-      "eval_steps_per_second": 14.864,
-      "step": 34000
-    },
-    {
-      "epoch": 2.619029305438043,
-      "grad_norm": 0.9636225700378418,
-      "learning_rate": 2.5810975768121964e-05,
-      "loss": -30.4646,
-      "step": 34050
-    },
-    {
-      "epoch": 2.6228751634489655,
-      "grad_norm": 1.7576632499694824,
-      "learning_rate": 2.5551255746305487e-05,
-      "loss": -30.4644,
-      "step": 34100
-    },
-    {
-      "epoch": 2.626721021459888,
-      "grad_norm": 1.2852941751480103,
-      "learning_rate": 2.5291535724489e-05,
-      "loss": -30.4646,
-      "step": 34150
-    },
-    {
-      "epoch": 2.63056687947081,
-      "grad_norm": 1.7288777828216553,
-      "learning_rate": 2.5031815702672516e-05,
-      "loss": -30.4642,
-      "step": 34200
-    },
-    {
-      "epoch": 2.6344127374817323,
-      "grad_norm": 1.7094942331314087,
-      "learning_rate": 2.477209568085604e-05,
-      "loss": -30.4648,
-      "step": 34250
-    },
-    {
-      "epoch": 2.6344127374817323,
-      "eval_loss": -30.468746185302734,
-      "eval_runtime": 16.7565,
-      "eval_samples_per_second": 59.678,
-      "eval_steps_per_second": 14.92,
-      "step": 34250
-    },
-    {
-      "epoch": 2.6382585954926543,
-      "grad_norm": 1.2149583101272583,
-      "learning_rate": 2.451237565903956e-05,
-      "loss": -30.4644,
-      "step": 34300
-    },
-    {
-      "epoch": 2.6421044535035767,
-      "grad_norm": 1.0909748077392578,
-      "learning_rate": 2.4252655637223072e-05,
-      "loss": -30.4647,
-      "step": 34350
-    },
-    {
-      "epoch": 2.6459503115144987,
-      "grad_norm": 1.1797704696655273,
-      "learning_rate": 2.399293561540659e-05,
-      "loss": -30.4645,
-      "step": 34400
-    },
-    {
-      "epoch": 2.649796169525421,
-      "grad_norm": 1.0700589418411255,
-      "learning_rate": 2.373321559359011e-05,
-      "loss": -30.465,
-      "step": 34450
-    },
-    {
-      "epoch": 2.6536420275363435,
-      "grad_norm": 1.4579436779022217,
-      "learning_rate": 2.347349557177363e-05,
-      "loss": -30.4646,
-      "step": 34500
-    },
-    {
-      "epoch": 2.6536420275363435,
-      "eval_loss": -30.46938705444336,
-      "eval_runtime": 16.8439,
-      "eval_samples_per_second": 59.369,
-      "eval_steps_per_second": 14.842,
-      "step": 34500
-    },
-    {
-      "epoch": 2.6574878855472654,
-      "grad_norm": 0.7920024991035461,
-      "learning_rate": 2.3213775549957147e-05,
-      "loss": -30.4646,
-      "step": 34550
-    },
-    {
-      "epoch": 2.661333743558188,
-      "grad_norm": 1.470051646232605,
-      "learning_rate": 2.2954055528140667e-05,
-      "loss": -30.4651,
-      "step": 34600
-    },
-    {
-      "epoch": 2.6651796015691103,
-      "grad_norm": 0.7992119193077087,
-      "learning_rate": 2.2694335506324183e-05,
-      "loss": -30.4645,
-      "step": 34650
-    },
-    {
-      "epoch": 2.6690254595800322,
-      "grad_norm": 0.9150895476341248,
-      "learning_rate": 2.2434615484507702e-05,
-      "loss": -30.4651,
-      "step": 34700
-    },
-    {
-      "epoch": 2.6728713175909546,
-      "grad_norm": 1.7401885986328125,
-      "learning_rate": 2.217489546269122e-05,
-      "loss": -30.4645,
-      "step": 34750
-    },
-    {
-      "epoch": 2.6728713175909546,
-      "eval_loss": -30.468637466430664,
-      "eval_runtime": 16.8394,
-      "eval_samples_per_second": 59.384,
-      "eval_steps_per_second": 14.846,
-      "step": 34750
-    },
-    {
-      "epoch": 2.676717175601877,
-      "grad_norm": 1.014699935913086,
-      "learning_rate": 2.1915175440874738e-05,
-      "loss": -30.465,
-      "step": 34800
-    },
-    {
-      "epoch": 2.680563033612799,
-      "grad_norm": 0.9657973051071167,
-      "learning_rate": 2.1655455419058258e-05,
-      "loss": -30.465,
-      "step": 34850
-    },
-    {
-      "epoch": 2.684408891623721,
-      "grad_norm": 1.1033378839492798,
-      "learning_rate": 2.1395735397241774e-05,
-      "loss": -30.465,
-      "step": 34900
-    },
-    {
-      "epoch": 2.6882547496346434,
-      "grad_norm": 3.0065767765045166,
-      "learning_rate": 2.113601537542529e-05,
-      "loss": -30.4651,
-      "step": 34950
-    },
-    {
-      "epoch": 2.692100607645566,
-      "grad_norm": 1.226468563079834,
-      "learning_rate": 2.087629535360881e-05,
-      "loss": -30.4653,
-      "step": 35000
-    },
-    {
-      "epoch": 2.692100607645566,
-      "eval_loss": -30.46967315673828,
-      "eval_runtime": 16.9427,
-      "eval_samples_per_second": 59.022,
-      "eval_steps_per_second": 14.756,
-      "step": 35000
-    },
-    {
-      "epoch": 2.6959464656564878,
-      "grad_norm": 1.8203473091125488,
-      "learning_rate": 2.061657533179233e-05,
-      "loss": -30.4653,
-      "step": 35050
-    },
-    {
-      "epoch": 2.69979232366741,
-      "grad_norm": 1.8032426834106445,
-      "learning_rate": 2.0356855309975846e-05,
-      "loss": -30.4652,
-      "step": 35100
-    },
-    {
-      "epoch": 2.7036381816783326,
-      "grad_norm": 1.6480307579040527,
-      "learning_rate": 2.0097135288159365e-05,
-      "loss": -30.465,
-      "step": 35150
-    },
-    {
-      "epoch": 2.7074840396892546,
-      "grad_norm": 1.4139653444290161,
-      "learning_rate": 1.983741526634288e-05,
-      "loss": -30.4647,
-      "step": 35200
-    },
-    {
-      "epoch": 2.711329897700177,
-      "grad_norm": 1.9126203060150146,
-      "learning_rate": 1.95776952445264e-05,
-      "loss": -30.4651,
-      "step": 35250
-    },
-    {
-      "epoch": 2.711329897700177,
-      "eval_loss": -30.4685115814209,
-      "eval_runtime": 17.0035,
-      "eval_samples_per_second": 58.812,
-      "eval_steps_per_second": 14.703,
-      "step": 35250
-    },
-    {
-      "epoch": 2.7151757557110994,
-      "grad_norm": 1.5345407724380493,
-      "learning_rate": 1.931797522270992e-05,
-      "loss": -30.465,
-      "step": 35300
-    },
-    {
-      "epoch": 2.7190216137220213,
-      "grad_norm": 0.8343766331672668,
-      "learning_rate": 1.9058255200893437e-05,
-      "loss": -30.4653,
-      "step": 35350
-    },
-    {
-      "epoch": 2.7228674717329437,
-      "grad_norm": 0.7670063972473145,
-      "learning_rate": 1.8798535179076957e-05,
-      "loss": -30.4652,
-      "step": 35400
-    },
-    {
-      "epoch": 2.7267133297438657,
-      "grad_norm": 0.913152813911438,
-      "learning_rate": 1.8538815157260476e-05,
-      "loss": -30.4652,
-      "step": 35450
-    },
-    {
-      "epoch": 2.730559187754788,
-      "grad_norm": 1.1838785409927368,
-      "learning_rate": 1.8279095135443993e-05,
-      "loss": -30.4654,
-      "step": 35500
-    },
-    {
-      "epoch": 2.730559187754788,
-      "eval_loss": -30.469755172729492,
-      "eval_runtime": 16.8431,
-      "eval_samples_per_second": 59.371,
-      "eval_steps_per_second": 14.843,
-      "step": 35500
-    },
-    {
-      "epoch": 2.73440504576571,
-      "grad_norm": 1.0075780153274536,
-      "learning_rate": 1.801937511362751e-05,
-      "loss": -30.4655,
-      "step": 35550
-    },
-    {
-      "epoch": 2.7382509037766325,
-      "grad_norm": 1.856740117073059,
-      "learning_rate": 1.775965509181103e-05,
-      "loss": -30.4653,
-      "step": 35600
-    },
-    {
-      "epoch": 2.742096761787555,
-      "grad_norm": 0.6788634657859802,
-      "learning_rate": 1.7499935069994548e-05,
-      "loss": -30.4655,
-      "step": 35650
-    },
-    {
-      "epoch": 2.745942619798477,
-      "grad_norm": 0.9009542465209961,
-      "learning_rate": 1.7240215048178064e-05,
-      "loss": -30.4656,
-      "step": 35700
-    },
-    {
-      "epoch": 2.7497884778093993,
-      "grad_norm": 1.9731862545013428,
-      "learning_rate": 1.6980495026361584e-05,
-      "loss": -30.4643,
-      "step": 35750
-    },
-    {
-      "epoch": 2.7497884778093993,
-      "eval_loss": -30.46780014038086,
-      "eval_runtime": 16.8878,
-      "eval_samples_per_second": 59.214,
-      "eval_steps_per_second": 14.804,
-      "step": 35750
-    },
-    {
-      "epoch": 2.7536343358203217,
-      "grad_norm": 1.0164039134979248,
-      "learning_rate": 1.67207750045451e-05,
-      "loss": -30.4655,
-      "step": 35800
-    },
-    {
-      "epoch": 2.7574801938312437,
-      "grad_norm": 1.1127376556396484,
-      "learning_rate": 1.646105498272862e-05,
-      "loss": -30.4657,
-      "step": 35850
-    },
-    {
-      "epoch": 2.761326051842166,
-      "grad_norm": 1.2492812871932983,
-      "learning_rate": 1.6201334960912136e-05,
-      "loss": -30.4659,
-      "step": 35900
-    },
-    {
-      "epoch": 2.7651719098530885,
-      "grad_norm": 0.944760799407959,
-      "learning_rate": 1.5941614939095655e-05,
-      "loss": -30.4654,
-      "step": 35950
-    },
-    {
-      "epoch": 2.7690177678640104,
-      "grad_norm": 0.7091095447540283,
-      "learning_rate": 1.5681894917279175e-05,
-      "loss": -30.4654,
-      "step": 36000
-    },
-    {
-      "epoch": 2.7690177678640104,
-      "eval_loss": -30.46934700012207,
-      "eval_runtime": 16.7482,
-      "eval_samples_per_second": 59.708,
-      "eval_steps_per_second": 14.927,
-      "step": 36000
-    },
-    {
-      "epoch": 2.772863625874933,
-      "grad_norm": 2.303079128265381,
-      "learning_rate": 1.5422174895462695e-05,
-      "loss": -30.4658,
-      "step": 36050
-    },
-    {
-      "epoch": 2.776709483885855,
-      "grad_norm": 1.1409249305725098,
-      "learning_rate": 1.516245487364621e-05,
-      "loss": -30.4659,
-      "step": 36100
-    },
-    {
-      "epoch": 2.780555341896777,
-      "grad_norm": 0.9111425280570984,
-      "learning_rate": 1.4902734851829727e-05,
-      "loss": -30.4658,
-      "step": 36150
-    },
-    {
-      "epoch": 2.784401199907699,
-      "grad_norm": 1.3415731191635132,
-      "learning_rate": 1.4643014830013247e-05,
-      "loss": -30.4659,
-      "step": 36200
-    },
-    {
-      "epoch": 2.7882470579186216,
-      "grad_norm": 0.8532393574714661,
-      "learning_rate": 1.4383294808196765e-05,
-      "loss": -30.4658,
-      "step": 36250
-    },
-    {
-      "epoch": 2.7882470579186216,
-      "eval_loss": -30.469867706298828,
-      "eval_runtime": 17.0192,
-      "eval_samples_per_second": 58.757,
-      "eval_steps_per_second": 14.689,
-      "step": 36250
-    },
-    {
-      "epoch": 2.792092915929544,
-      "grad_norm": 1.6142336130142212,
-      "learning_rate": 1.4123574786380284e-05,
-      "loss": -30.4662,
-      "step": 36300
-    },
-    {
-      "epoch": 2.795938773940466,
-      "grad_norm": 1.9979597330093384,
-      "learning_rate": 1.3863854764563799e-05,
-      "loss": -30.4652,
-      "step": 36350
-    },
-    {
-      "epoch": 2.7997846319513884,
-      "grad_norm": 1.4881914854049683,
-      "learning_rate": 1.3604134742747318e-05,
-      "loss": -30.466,
-      "step": 36400
-    },
-    {
-      "epoch": 2.803630489962311,
-      "grad_norm": 2.0614233016967773,
-      "learning_rate": 1.3344414720930836e-05,
-      "loss": -30.4658,
-      "step": 36450
-    },
-    {
-      "epoch": 2.8074763479732328,
-      "grad_norm": 1.0104310512542725,
-      "learning_rate": 1.3084694699114356e-05,
-      "loss": -30.466,
-      "step": 36500
-    },
-    {
-      "epoch": 2.8074763479732328,
-      "eval_loss": -30.47022247314453,
-      "eval_runtime": 16.8782,
-      "eval_samples_per_second": 59.248,
-      "eval_steps_per_second": 14.812,
-      "step": 36500
-    },
-    {
-      "epoch": 2.811322205984155,
-      "grad_norm": 0.8461622595787048,
-      "learning_rate": 1.2824974677297874e-05,
-      "loss": -30.466,
-      "step": 36550
-    },
-    {
-      "epoch": 2.8151680639950776,
-      "grad_norm": 1.3280683755874634,
-      "learning_rate": 1.2565254655481392e-05,
-      "loss": -30.466,
-      "step": 36600
-    },
-    {
-      "epoch": 2.8190139220059995,
-      "grad_norm": 1.5778801441192627,
-      "learning_rate": 1.230553463366491e-05,
-      "loss": -30.4659,
-      "step": 36650
-    },
-    {
-      "epoch": 2.8228597800169215,
-      "grad_norm": 1.6054023504257202,
-      "learning_rate": 1.204581461184843e-05,
-      "loss": -30.4661,
-      "step": 36700
-    },
-    {
-      "epoch": 2.826705638027844,
-      "grad_norm": 1.3517789840698242,
-      "learning_rate": 1.1786094590031946e-05,
-      "loss": -30.4661,
-      "step": 36750
-    },
-    {
-      "epoch": 2.826705638027844,
-      "eval_loss": -30.470460891723633,
-      "eval_runtime": 16.9586,
-      "eval_samples_per_second": 58.967,
-      "eval_steps_per_second": 14.742,
-      "step": 36750
-    },
-    {
-      "epoch": 2.8305514960387663,
-      "grad_norm": 1.1399978399276733,
-      "learning_rate": 1.1526374568215465e-05,
-      "loss": -30.4658,
-      "step": 36800
-    },
-    {
-      "epoch": 2.8343973540496883,
-      "grad_norm": 0.8231783509254456,
-      "learning_rate": 1.1266654546398983e-05,
-      "loss": -30.4663,
-      "step": 36850
-    },
-    {
-      "epoch": 2.8382432120606107,
-      "grad_norm": 0.8370407819747925,
-      "learning_rate": 1.1006934524582501e-05,
-      "loss": -30.4662,
-      "step": 36900
-    },
-    {
-      "epoch": 2.842089070071533,
-      "grad_norm": 0.8367822170257568,
-      "learning_rate": 1.0747214502766019e-05,
-      "loss": -30.4662,
-      "step": 36950
-    },
-    {
-      "epoch": 2.845934928082455,
-      "grad_norm": 0.8438307642936707,
-      "learning_rate": 1.0487494480949537e-05,
-      "loss": -30.4661,
-      "step": 37000
-    },
-    {
-      "epoch": 2.845934928082455,
-      "eval_loss": -30.470365524291992,
-      "eval_runtime": 16.8139,
-      "eval_samples_per_second": 59.475,
-      "eval_steps_per_second": 14.869,
-      "step": 37000
-    },
-    {
-      "epoch": 2.8497807860933775,
-      "grad_norm": 0.8433590531349182,
-      "learning_rate": 1.0227774459133055e-05,
-      "loss": -30.4661,
-      "step": 37050
-    },
-    {
-      "epoch": 2.8536266441043,
-      "grad_norm": 0.6508035063743591,
-      "learning_rate": 9.968054437316573e-06,
-      "loss": -30.4663,
-      "step": 37100
-    },
-    {
-      "epoch": 2.857472502115222,
-      "grad_norm": 1.3801597356796265,
-      "learning_rate": 9.70833441550009e-06,
-      "loss": -30.4662,
-      "step": 37150
-    },
-    {
-      "epoch": 2.8613183601261443,
-      "grad_norm": 0.7447367906570435,
-      "learning_rate": 9.448614393683609e-06,
-      "loss": -30.4663,
-      "step": 37200
-    },
-    {
-      "epoch": 2.8651642181370662,
-      "grad_norm": 1.370768666267395,
-      "learning_rate": 9.188894371867128e-06,
-      "loss": -30.4661,
-      "step": 37250
-    },
-    {
-      "epoch": 2.8651642181370662,
-      "eval_loss": -30.470088958740234,
-      "eval_runtime": 16.7498,
-      "eval_samples_per_second": 59.702,
-      "eval_steps_per_second": 14.926,
-      "step": 37250
-    },
-    {
-      "epoch": 2.8690100761479886,
-      "grad_norm": 1.3880547285079956,
-      "learning_rate": 8.929174350050646e-06,
-      "loss": -30.4665,
-      "step": 37300
-    },
-    {
-      "epoch": 2.8728559341589106,
-      "grad_norm": 1.428084135055542,
-      "learning_rate": 8.669454328234164e-06,
-      "loss": -30.4665,
-      "step": 37350
-    },
-    {
-      "epoch": 2.876701792169833,
-      "grad_norm": 0.4759369492530823,
-      "learning_rate": 8.409734306417682e-06,
-      "loss": -30.4665,
-      "step": 37400
-    },
-    {
-      "epoch": 2.8805476501807554,
-      "grad_norm": 1.0323377847671509,
-      "learning_rate": 8.1500142846012e-06,
-      "loss": -30.4666,
-      "step": 37450
-    },
-    {
-      "epoch": 2.8843935081916774,
-      "grad_norm": 0.9753648638725281,
-      "learning_rate": 7.890294262784718e-06,
-      "loss": -30.4666,
-      "step": 37500
-    },
-    {
-      "epoch": 2.8843935081916774,
-      "eval_loss": -30.470531463623047,
-      "eval_runtime": 16.7944,
-      "eval_samples_per_second": 59.544,
-      "eval_steps_per_second": 14.886,
-      "step": 37500
-    },
-    {
-      "epoch": 2.8882393662026,
-      "grad_norm": 0.6514129042625427,
-      "learning_rate": 7.630574240968237e-06,
-      "loss": -30.4666,
-      "step": 37550
-    },
-    {
-      "epoch": 2.892085224213522,
-      "grad_norm": 1.0104998350143433,
-      "learning_rate": 7.3708542191517545e-06,
-      "loss": -30.4667,
-      "step": 37600
-    },
-    {
-      "epoch": 2.895931082224444,
-      "grad_norm": 0.713027834892273,
-      "learning_rate": 7.111134197335273e-06,
-      "loss": -30.4665,
-      "step": 37650
-    },
-    {
-      "epoch": 2.8997769402353666,
-      "grad_norm": 0.6001731157302856,
-      "learning_rate": 6.851414175518792e-06,
-      "loss": -30.4666,
-      "step": 37700
-    },
-    {
-      "epoch": 2.903622798246289,
-      "grad_norm": 0.6344665288925171,
-      "learning_rate": 6.591694153702309e-06,
-      "loss": -30.4664,
-      "step": 37750
-    },
-    {
-      "epoch": 2.903622798246289,
-      "eval_loss": -30.47093391418457,
-      "eval_runtime": 16.9622,
-      "eval_samples_per_second": 58.955,
-      "eval_steps_per_second": 14.739,
-      "step": 37750
-    },
-    {
-      "epoch": 2.907468656257211,
-      "grad_norm": 0.5279808640480042,
-      "learning_rate": 6.331974131885828e-06,
-      "loss": -30.4667,
-      "step": 37800
-    },
-    {
-      "epoch": 2.9113145142681334,
-      "grad_norm": 0.9036094546318054,
-      "learning_rate": 6.072254110069346e-06,
-      "loss": -30.4667,
-      "step": 37850
-    },
-    {
-      "epoch": 2.9151603722790553,
-      "grad_norm": 0.601686954498291,
-      "learning_rate": 5.812534088252864e-06,
-      "loss": -30.4668,
-      "step": 37900
-    },
-    {
-      "epoch": 2.9190062302899777,
-      "grad_norm": 1.0191963911056519,
-      "learning_rate": 5.552814066436382e-06,
-      "loss": -30.4667,
-      "step": 37950
-    },
-    {
-      "epoch": 2.9228520883008997,
-      "grad_norm": 0.9260501265525818,
-      "learning_rate": 5.2930940446198996e-06,
-      "loss": -30.4667,
-      "step": 38000
-    },
-    {
-      "epoch": 2.9228520883008997,
-      "eval_loss": -30.47071075439453,
-      "eval_runtime": 16.8261,
-      "eval_samples_per_second": 59.431,
-      "eval_steps_per_second": 14.858,
-      "step": 38000
-    },
-    {
-      "epoch": 2.926697946311822,
-      "grad_norm": 0.9110496044158936,
-      "learning_rate": 5.033374022803418e-06,
-      "loss": -30.4668,
-      "step": 38050
-    },
-    {
-      "epoch": 2.9305438043227445,
-      "grad_norm": 0.7347224950790405,
-      "learning_rate": 4.773654000986936e-06,
-      "loss": -30.4668,
-      "step": 38100
-    },
-    {
-      "epoch": 2.9343896623336665,
-      "grad_norm": 0.946358859539032,
-      "learning_rate": 4.513933979170454e-06,
-      "loss": -30.4669,
-      "step": 38150
-    },
-    {
-      "epoch": 2.938235520344589,
-      "grad_norm": 0.5716750621795654,
-      "learning_rate": 4.254213957353972e-06,
-      "loss": -30.4669,
-      "step": 38200
-    },
-    {
-      "epoch": 2.9420813783555113,
-      "grad_norm": 1.8405364751815796,
-      "learning_rate": 3.994493935537491e-06,
-      "loss": -30.4667,
-      "step": 38250
-    },
-    {
-      "epoch": 2.9420813783555113,
-      "eval_loss": -30.471132278442383,
-      "eval_runtime": 16.8489,
-      "eval_samples_per_second": 59.351,
-      "eval_steps_per_second": 14.838,
-      "step": 38250
-    },
-    {
-      "epoch": 2.9459272363664333,
-      "grad_norm": 0.6454173922538757,
-      "learning_rate": 3.734773913721009e-06,
-      "loss": -30.467,
-      "step": 38300
-    },
-    {
-      "epoch": 2.9497730943773557,
-      "grad_norm": 0.7232244610786438,
-      "learning_rate": 3.475053891904527e-06,
-      "loss": -30.4668,
-      "step": 38350
-    },
-    {
-      "epoch": 2.953618952388278,
-      "grad_norm": 1.0730299949645996,
-      "learning_rate": 3.215333870088045e-06,
-      "loss": -30.4668,
-      "step": 38400
-    },
-    {
-      "epoch": 2.9574648103992,
-      "grad_norm": 0.4886613190174103,
-      "learning_rate": 2.9556138482715634e-06,
-      "loss": -30.4669,
-      "step": 38450
-    },
-    {
-      "epoch": 2.961310668410122,
-      "grad_norm": 0.5141581296920776,
-      "learning_rate": 2.6958938264550813e-06,
-      "loss": -30.4669,
-      "step": 38500
-    },
-    {
-      "epoch": 2.961310668410122,
-      "eval_loss": -30.471107482910156,
-      "eval_runtime": 16.867,
-      "eval_samples_per_second": 59.287,
-      "eval_steps_per_second": 14.822,
-      "step": 38500
-    },
-    {
-      "epoch": 2.9651565264210444,
-      "grad_norm": 0.5108075737953186,
-      "learning_rate": 2.4361738046385997e-06,
-      "loss": -30.467,
-      "step": 38550
-    },
-    {
-      "epoch": 2.969002384431967,
-      "grad_norm": 0.608711838722229,
-      "learning_rate": 2.1764537828221176e-06,
-      "loss": -30.467,
-      "step": 38600
-    },
-    {
-      "epoch": 2.972848242442889,
-      "grad_norm": 0.9723970890045166,
-      "learning_rate": 1.916733761005636e-06,
-      "loss": -30.4672,
-      "step": 38650
-    },
-    {
-      "epoch": 2.976694100453811,
-      "grad_norm": 0.4361710846424103,
-      "learning_rate": 1.6570137391891543e-06,
-      "loss": -30.467,
-      "step": 38700
-    },
-    {
-      "epoch": 2.9805399584647336,
-      "grad_norm": 1.1543878316879272,
-      "learning_rate": 1.3972937173726722e-06,
-      "loss": -30.4669,
-      "step": 38750
-    },
-    {
-      "epoch": 2.9805399584647336,
-      "eval_loss": -30.47084617614746,
-      "eval_runtime": 16.7842,
-      "eval_samples_per_second": 59.58,
-      "eval_steps_per_second": 14.895,
-      "step": 38750
-    },
-    {
-      "epoch": 2.9843858164756556,
-      "grad_norm": 0.4299401640892029,
-      "learning_rate": 1.1375736955561905e-06,
-      "loss": -30.467,
-      "step": 38800
-    },
-    {
-      "epoch": 2.988231674486578,
-      "grad_norm": 0.8476674556732178,
-      "learning_rate": 8.778536737397087e-07,
-      "loss": -30.467,
-      "step": 38850
-    },
-    {
-      "epoch": 2.9920775324975004,
-      "grad_norm": 1.1194632053375244,
-      "learning_rate": 6.181336519232268e-07,
-      "loss": -30.4671,
-      "step": 38900
-    },
-    {
-      "epoch": 2.9959233905084224,
-      "grad_norm": 0.7896573543548584,
-      "learning_rate": 3.584136301067449e-07,
-      "loss": -30.4671,
-      "step": 38950
-    },
-    {
-      "epoch": 2.999769248519345,
-      "grad_norm": 0.32893452048301697,
-      "learning_rate": 9.869360829026311e-08,
-      "loss": -30.4671,
-      "step": 39000
-    },
-    {
-      "epoch": 2.999769248519345,
-      "eval_loss": -30.471084594726562,
-      "eval_runtime": 16.7188,
-      "eval_samples_per_second": 59.813,
-      "eval_steps_per_second": 14.953,
-      "step": 39000
     }
   ],
   "logging_steps": 50,