diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,3366 +1,5039 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 50.0,
+  "epoch": 30.0,
   "eval_steps": 500,
-  "global_step": 238700,
+  "global_step": 71610,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.10473397570171764,
-      "grad_norm": 1.1384308338165283,
-      "learning_rate": 4.989526602429829e-05,
-      "loss": 8.9672,
-      "step": 500
+      "epoch": 0.04189359028068706,
+      "grad_norm": 1.1499052047729492,
+      "learning_rate": 4.993017734953219e-05,
+      "loss": 9.973,
+      "step": 100
     },
     {
-      "epoch": 0.20946795140343527,
-      "grad_norm": 0.9985808730125427,
+      "epoch": 0.08378718056137412,
+      "grad_norm": 1.1211209297180176,
+      "learning_rate": 4.986035469906438e-05,
+      "loss": 9.315,
+      "step": 200
+    },
+    {
+      "epoch": 0.12568077084206117,
+      "grad_norm": 1.0281990766525269,
       "learning_rate": 4.979053204859657e-05,
-      "loss": 7.7253,
+      "loss": 8.8239,
+      "step": 300
+    },
+    {
+      "epoch": 0.16757436112274823,
+      "grad_norm": 1.1534616947174072,
+      "learning_rate": 4.972070939812876e-05,
+      "loss": 8.4254,
+      "step": 400
+    },
+    {
+      "epoch": 0.20946795140343527,
+      "grad_norm": 0.9906216859817505,
+      "learning_rate": 4.9650886747660944e-05,
+      "loss": 8.1225,
+      "step": 500
+    },
+    {
+      "epoch": 0.25136154168412234,
+      "grad_norm": 0.9559820294380188,
+      "learning_rate": 4.958106409719313e-05,
+      "loss": 7.8557,
+      "step": 600
+    },
+    {
+      "epoch": 0.2932551319648094,
+      "grad_norm": 0.7887147665023804,
+      "learning_rate": 4.951124144672532e-05,
+      "loss": 7.7282,
+      "step": 700
+    },
+    {
+      "epoch": 0.33514872224549647,
+      "grad_norm": 0.6530473232269287,
+      "learning_rate": 4.944141879625751e-05,
+      "loss": 7.6586,
+      "step": 800
+    },
+    {
+      "epoch": 0.3770423125261835,
+      "grad_norm": 0.6546779870986938,
+      "learning_rate": 4.9371596145789694e-05,
+      "loss": 7.6438,
+      "step": 900
+    },
+    {
+      "epoch": 0.41893590280687054,
+      "grad_norm": 0.5660322308540344,
+      "learning_rate": 4.9301773495321885e-05,
+      "loss": 7.6237,
       "step": 1000
     },
     {
-      "epoch": 0.31420192710515293,
-      "grad_norm": 1.0521348714828491,
-      "learning_rate": 4.968579807289485e-05,
-      "loss": 7.5614,
+      "epoch": 0.4608294930875576,
+      "grad_norm": 0.6094719171524048,
+      "learning_rate": 4.9231950844854076e-05,
+      "loss": 7.5937,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5027230833682447,
+      "grad_norm": 0.6101346015930176,
+      "learning_rate": 4.916212819438626e-05,
+      "loss": 7.5661,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5446166736489317,
+      "grad_norm": 0.5943477749824524,
+      "learning_rate": 4.909230554391845e-05,
+      "loss": 7.6007,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5865102639296188,
+      "grad_norm": 0.6604776382446289,
+      "learning_rate": 4.902248289345064e-05,
+      "loss": 7.568,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6284038542103059,
+      "grad_norm": 0.6151777505874634,
+      "learning_rate": 4.8952660242982826e-05,
+      "loss": 7.5521,
       "step": 1500
     },
     {
-      "epoch": 0.41893590280687054,
-      "grad_norm": 1.0602227449417114,
-      "learning_rate": 4.958106409719313e-05,
-      "loss": 7.5037,
+      "epoch": 0.6702974444909929,
+      "grad_norm": 0.6381381750106812,
+      "learning_rate": 4.888283759251501e-05,
+      "loss": 7.5478,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7121910347716799,
+      "grad_norm": 0.6552081108093262,
+      "learning_rate": 4.88130149420472e-05,
+      "loss": 7.5293,
+      "step": 1700
+    },
+    {
+      "epoch": 0.754084625052367,
+      "grad_norm": 0.6973659992218018,
+      "learning_rate": 4.874319229157939e-05,
+      "loss": 7.4983,
+      "step": 1800
+    },
+    {
+      "epoch": 0.795978215333054,
+      "grad_norm": 0.8130584955215454,
+      "learning_rate": 4.8673369641111576e-05,
+      "loss": 7.5213,
+      "step": 1900
+    },
+    {
+      "epoch": 0.8378718056137411,
+      "grad_norm": 0.8530446887016296,
+      "learning_rate": 4.860354699064377e-05,
+      "loss": 7.5105,
       "step": 2000
     },
     {
-      "epoch": 0.5236698785085881,
-      "grad_norm": 1.6102268695831299,
-      "learning_rate": 4.9476330121491414e-05,
-      "loss": 7.4595,
+      "epoch": 0.8797653958944281,
+      "grad_norm": 0.8059477210044861,
+      "learning_rate": 4.853372434017596e-05,
+      "loss": 7.508,
+      "step": 2100
+    },
+    {
+      "epoch": 0.9216589861751152,
+      "grad_norm": 0.7378331422805786,
+      "learning_rate": 4.846390168970814e-05,
+      "loss": 7.482,
+      "step": 2200
+    },
+    {
+      "epoch": 0.9635525764558023,
+      "grad_norm": 1.1823194026947021,
+      "learning_rate": 4.839407903924033e-05,
+      "loss": 7.4778,
+      "step": 2300
+    },
+    {
+      "epoch": 1.0054461667364893,
+      "grad_norm": 0.8547298908233643,
+      "learning_rate": 4.832425638877252e-05,
+      "loss": 7.4529,
+      "step": 2400
+    },
+    {
+      "epoch": 1.0473397570171763,
+      "grad_norm": 0.8535734415054321,
+      "learning_rate": 4.825443373830471e-05,
+      "loss": 7.4374,
       "step": 2500
     },
     {
-      "epoch": 0.6284038542103059,
-      "grad_norm": 1.335976004600525,
-      "learning_rate": 4.9371596145789694e-05,
-      "loss": 7.4267,
+      "epoch": 1.0892333472978635,
+      "grad_norm": 0.994597852230072,
+      "learning_rate": 4.818461108783689e-05,
+      "loss": 7.4306,
+      "step": 2600
+    },
+    {
+      "epoch": 1.1311269375785504,
+      "grad_norm": 1.2056490182876587,
+      "learning_rate": 4.8114788437369084e-05,
+      "loss": 7.4322,
+      "step": 2700
+    },
+    {
+      "epoch": 1.1730205278592376,
+      "grad_norm": 1.2451157569885254,
+      "learning_rate": 4.8044965786901275e-05,
+      "loss": 7.4205,
+      "step": 2800
+    },
+    {
+      "epoch": 1.2149141181399246,
+      "grad_norm": 0.9964780211448669,
+      "learning_rate": 4.797514313643346e-05,
+      "loss": 7.4031,
+      "step": 2900
+    },
+    {
+      "epoch": 1.2568077084206117,
+      "grad_norm": 0.8989804983139038,
+      "learning_rate": 4.790532048596565e-05,
+      "loss": 7.4043,
       "step": 3000
     },
     {
-      "epoch": 0.7331378299120235,
-      "grad_norm": 1.340728998184204,
-      "learning_rate": 4.926686217008798e-05,
-      "loss": 7.392,
+      "epoch": 1.2987012987012987,
+      "grad_norm": 1.1330469846725464,
+      "learning_rate": 4.783549783549784e-05,
+      "loss": 7.4031,
+      "step": 3100
+    },
+    {
+      "epoch": 1.3405948889819856,
+      "grad_norm": 0.9531299471855164,
+      "learning_rate": 4.7765675185030025e-05,
+      "loss": 7.3866,
+      "step": 3200
+    },
+    {
+      "epoch": 1.3824884792626728,
+      "grad_norm": 1.0342323780059814,
+      "learning_rate": 4.7695852534562216e-05,
+      "loss": 7.3477,
+      "step": 3300
+    },
+    {
+      "epoch": 1.42438206954336,
+      "grad_norm": 1.0523111820220947,
+      "learning_rate": 4.76260298840944e-05,
+      "loss": 7.3726,
+      "step": 3400
+    },
+    {
+      "epoch": 1.466275659824047,
+      "grad_norm": 1.298751711845398,
+      "learning_rate": 4.755620723362659e-05,
+      "loss": 7.3484,
       "step": 3500
     },
     {
-      "epoch": 0.8378718056137411,
-      "grad_norm": 1.4520059823989868,
-      "learning_rate": 4.916212819438626e-05,
-      "loss": 7.3253,
+      "epoch": 1.508169250104734,
+      "grad_norm": 1.0065233707427979,
+      "learning_rate": 4.7486384583158775e-05,
+      "loss": 7.3567,
+      "step": 3600
+    },
+    {
+      "epoch": 1.550062840385421,
+      "grad_norm": 1.2989579439163208,
+      "learning_rate": 4.7416561932690966e-05,
+      "loss": 7.3275,
+      "step": 3700
+    },
+    {
+      "epoch": 1.591956430666108,
+      "grad_norm": 1.0343406200408936,
+      "learning_rate": 4.734673928222316e-05,
+      "loss": 7.3353,
+      "step": 3800
+    },
+    {
+      "epoch": 1.6338500209467952,
+      "grad_norm": 0.9944115281105042,
+      "learning_rate": 4.727691663175534e-05,
+      "loss": 7.3304,
+      "step": 3900
+    },
+    {
+      "epoch": 1.6757436112274822,
+      "grad_norm": 1.102974534034729,
+      "learning_rate": 4.720709398128753e-05,
+      "loss": 7.3128,
       "step": 4000
     },
     {
-      "epoch": 0.9426057813154587,
-      "grad_norm": 1.7685532569885254,
-      "learning_rate": 4.905760368663595e-05,
-      "loss": 7.3204,
-      "step": 4500
+      "epoch": 1.7176372015081691,
+      "grad_norm": 1.112282156944275,
+      "learning_rate": 4.713727133081972e-05,
+      "loss": 7.333,
+      "step": 4100
+    },
+    {
+      "epoch": 1.7595307917888563,
+      "grad_norm": 1.2143328189849854,
+      "learning_rate": 4.7067448680351914e-05,
+      "loss": 7.3342,
+      "step": 4200
+    },
+    {
+      "epoch": 1.8014243820695435,
+      "grad_norm": 1.1656922101974487,
+      "learning_rate": 4.69976260298841e-05,
+      "loss": 7.2995,
+      "step": 4300
+    },
+    {
+      "epoch": 1.8433179723502304,
+      "grad_norm": 1.2085694074630737,
+      "learning_rate": 4.692780337941628e-05,
+      "loss": 7.2939,
+      "step": 4400
+    },
+    {
+      "epoch": 1.8852115626309174,
+      "grad_norm": 1.1366217136383057,
+      "learning_rate": 4.685798072894847e-05,
+      "loss": 7.3079,
+      "step": 4500
+    },
+    {
+      "epoch": 1.9271051529116046,
+      "grad_norm": 1.5368098020553589,
+      "learning_rate": 4.678815807848066e-05,
+      "loss": 7.2682,
+      "step": 4600
+    },
+    {
+      "epoch": 1.9689987431922917,
+      "grad_norm": 1.1548763513565063,
+      "learning_rate": 4.671833542801285e-05,
+      "loss": 7.2708,
+      "step": 4700
+    },
+    {
+      "epoch": 2.0108923334729787,
+      "grad_norm": 1.1510928869247437,
+      "learning_rate": 4.664851277754504e-05,
+      "loss": 7.2794,
+      "step": 4800
+    },
+    {
+      "epoch": 2.0527859237536656,
+      "grad_norm": 1.103461503982544,
+      "learning_rate": 4.6578690127077224e-05,
+      "loss": 7.2612,
+      "step": 4900
+    },
+    {
+      "epoch": 2.0946795140343526,
+      "grad_norm": 1.2767215967178345,
+      "learning_rate": 4.6508867476609414e-05,
+      "loss": 7.2436,
+      "step": 5000
+    },
+    {
+      "epoch": 2.13657310431504,
+      "grad_norm": 1.3062710762023926,
+      "learning_rate": 4.6439044826141605e-05,
+      "loss": 7.2162,
+      "step": 5100
+    },
+    {
+      "epoch": 2.178466694595727,
+      "grad_norm": 1.2461299896240234,
+      "learning_rate": 4.6369222175673796e-05,
+      "loss": 7.2387,
+      "step": 5200
+    },
+    {
+      "epoch": 2.220360284876414,
+      "grad_norm": 1.5427358150482178,
+      "learning_rate": 4.629939952520598e-05,
+      "loss": 7.2004,
+      "step": 5300
+    },
+    {
+      "epoch": 2.262253875157101,
+      "grad_norm": 1.390331506729126,
+      "learning_rate": 4.6229576874738165e-05,
+      "loss": 7.2403,
+      "step": 5400
+    },
+    {
+      "epoch": 2.3041474654377883,
+      "grad_norm": 1.4087032079696655,
+      "learning_rate": 4.6159754224270356e-05,
+      "loss": 7.2157,
+      "step": 5500
+    },
+    {
+      "epoch": 2.346041055718475,
+      "grad_norm": 1.2417359352111816,
+      "learning_rate": 4.608993157380254e-05,
+      "loss": 7.2277,
+      "step": 5600
+    },
+    {
+      "epoch": 2.387934645999162,
+      "grad_norm": 1.4267281293869019,
+      "learning_rate": 4.602010892333473e-05,
+      "loss": 7.1999,
+      "step": 5700
+    },
+    {
+      "epoch": 2.429828236279849,
+      "grad_norm": 1.3897684812545776,
+      "learning_rate": 4.595028627286692e-05,
+      "loss": 7.2155,
+      "step": 5800
+    },
+    {
+      "epoch": 2.471721826560536,
+      "grad_norm": 1.326821208000183,
+      "learning_rate": 4.5880463622399106e-05,
+      "loss": 7.1705,
+      "step": 5900
+    },
+    {
+      "epoch": 2.5136154168412235,
+      "grad_norm": 1.2585749626159668,
+      "learning_rate": 4.58106409719313e-05,
+      "loss": 7.1787,
+      "step": 6000
+    },
+    {
+      "epoch": 2.5555090071219104,
+      "grad_norm": 1.4856244325637817,
+      "learning_rate": 4.574081832146349e-05,
+      "loss": 7.1923,
+      "step": 6100
+    },
+    {
+      "epoch": 2.5974025974025974,
+      "grad_norm": 1.2883421182632446,
+      "learning_rate": 4.567099567099568e-05,
+      "loss": 7.1776,
+      "step": 6200
+    },
+    {
+      "epoch": 2.6392961876832843,
+      "grad_norm": 1.4935518503189087,
+      "learning_rate": 4.560117302052786e-05,
+      "loss": 7.1711,
+      "step": 6300
+    },
+    {
+      "epoch": 2.6811897779639713,
+      "grad_norm": 1.3920152187347412,
+      "learning_rate": 4.553135037006005e-05,
+      "loss": 7.1292,
+      "step": 6400
+    },
+    {
+      "epoch": 2.7230833682446587,
+      "grad_norm": 1.2802495956420898,
+      "learning_rate": 4.546152771959224e-05,
+      "loss": 7.1558,
+      "step": 6500
+    },
+    {
+      "epoch": 2.7649769585253456,
+      "grad_norm": 1.4111789464950562,
+      "learning_rate": 4.5392403295629106e-05,
+      "loss": 7.172,
+      "step": 6600
+    },
+    {
+      "epoch": 2.8068705488060326,
+      "grad_norm": 1.6390964984893799,
+      "learning_rate": 4.53225806451613e-05,
+      "loss": 7.1263,
+      "step": 6700
+    },
+    {
+      "epoch": 2.84876413908672,
+      "grad_norm": 1.4132812023162842,
+      "learning_rate": 4.525275799469348e-05,
+      "loss": 7.1259,
+      "step": 6800
+    },
+    {
+      "epoch": 2.890657729367407,
+      "grad_norm": 1.4943978786468506,
+      "learning_rate": 4.518293534422567e-05,
+      "loss": 7.1252,
+      "step": 6900
+    },
+    {
+      "epoch": 2.932551319648094,
+      "grad_norm": 1.3022414445877075,
+      "learning_rate": 4.5113112693757856e-05,
+      "loss": 7.107,
+      "step": 7000
+    },
+    {
+      "epoch": 2.974444909928781,
+      "grad_norm": 1.4270446300506592,
+      "learning_rate": 4.504329004329004e-05,
+      "loss": 7.1278,
+      "step": 7100
+    },
+    {
+      "epoch": 3.016338500209468,
+      "grad_norm": 1.3672137260437012,
+      "learning_rate": 4.497346739282223e-05,
+      "loss": 7.1282,
+      "step": 7200
+    },
+    {
+      "epoch": 3.058232090490155,
+      "grad_norm": 1.955368995666504,
+      "learning_rate": 4.490364474235442e-05,
+      "loss": 7.1123,
+      "step": 7300
+    },
+    {
+      "epoch": 3.100125680770842,
+      "grad_norm": 1.3990498781204224,
+      "learning_rate": 4.483382209188661e-05,
+      "loss": 7.1117,
+      "step": 7400
+    },
+    {
+      "epoch": 3.142019271051529,
+      "grad_norm": 1.6294671297073364,
+      "learning_rate": 4.47639994414188e-05,
+      "loss": 7.0721,
+      "step": 7500
+    },
+    {
+      "epoch": 3.183912861332216,
+      "grad_norm": 1.3939063549041748,
+      "learning_rate": 4.469417679095099e-05,
+      "loss": 7.0599,
+      "step": 7600
+    },
+    {
+      "epoch": 3.225806451612903,
+      "grad_norm": 1.918155312538147,
+      "learning_rate": 4.462435414048318e-05,
+      "loss": 7.0759,
+      "step": 7700
+    },
+    {
+      "epoch": 3.2677000418935904,
+      "grad_norm": 1.3072093725204468,
+      "learning_rate": 4.455453149001536e-05,
+      "loss": 7.0784,
+      "step": 7800
+    },
+    {
+      "epoch": 3.3095936321742774,
+      "grad_norm": 1.3794573545455933,
+      "learning_rate": 4.4484708839547554e-05,
+      "loss": 7.0726,
+      "step": 7900
+    },
+    {
+      "epoch": 3.3514872224549643,
+      "grad_norm": 1.4223533868789673,
+      "learning_rate": 4.441488618907974e-05,
+      "loss": 7.0513,
+      "step": 8000
+    },
+    {
+      "epoch": 3.3933808127356513,
+      "grad_norm": 1.695520043373108,
+      "learning_rate": 4.434506353861192e-05,
+      "loss": 7.0586,
+      "step": 8100
+    },
+    {
+      "epoch": 3.4352744030163387,
+      "grad_norm": 2.290275812149048,
+      "learning_rate": 4.4275240888144113e-05,
+      "loss": 7.0429,
+      "step": 8200
+    },
+    {
+      "epoch": 3.4771679932970256,
+      "grad_norm": 1.5322943925857544,
+      "learning_rate": 4.4205418237676304e-05,
+      "loss": 7.0196,
+      "step": 8300
+    },
+    {
+      "epoch": 3.5190615835777126,
+      "grad_norm": 1.4767639636993408,
+      "learning_rate": 4.4135595587208495e-05,
+      "loss": 7.0643,
+      "step": 8400
+    },
+    {
+      "epoch": 3.5609551738583995,
+      "grad_norm": 1.4343881607055664,
+      "learning_rate": 4.406577293674068e-05,
+      "loss": 7.0153,
+      "step": 8500
+    },
+    {
+      "epoch": 3.602848764139087,
+      "grad_norm": 1.7641897201538086,
+      "learning_rate": 4.399664851277755e-05,
+      "loss": 7.013,
+      "step": 8600
+    },
+    {
+      "epoch": 3.644742354419774,
+      "grad_norm": 1.9688279628753662,
+      "learning_rate": 4.392682586230973e-05,
+      "loss": 6.9985,
+      "step": 8700
+    },
+    {
+      "epoch": 3.686635944700461,
+      "grad_norm": 1.7434871196746826,
+      "learning_rate": 4.385700321184192e-05,
+      "loss": 7.0142,
+      "step": 8800
+    },
+    {
+      "epoch": 3.728529534981148,
+      "grad_norm": 1.550470232963562,
+      "learning_rate": 4.378718056137411e-05,
+      "loss": 6.9975,
+      "step": 8900
+    },
+    {
+      "epoch": 3.7704231252618348,
+      "grad_norm": 1.759869933128357,
+      "learning_rate": 4.37173579109063e-05,
+      "loss": 6.9821,
+      "step": 9000
+    },
+    {
+      "epoch": 3.812316715542522,
+      "grad_norm": 2.052905797958374,
+      "learning_rate": 4.364753526043849e-05,
+      "loss": 6.9917,
+      "step": 9100
+    },
+    {
+      "epoch": 3.854210305823209,
+      "grad_norm": 1.8859872817993164,
+      "learning_rate": 4.357771260997068e-05,
+      "loss": 6.9934,
+      "step": 9200
+    },
+    {
+      "epoch": 3.896103896103896,
+      "grad_norm": 1.8349354267120361,
+      "learning_rate": 4.3507889959502863e-05,
+      "loss": 6.9861,
+      "step": 9300
+    },
+    {
+      "epoch": 3.937997486384583,
+      "grad_norm": 2.519893169403076,
+      "learning_rate": 4.3438067309035054e-05,
+      "loss": 6.9611,
+      "step": 9400
+    },
+    {
+      "epoch": 3.97989107666527,
+      "grad_norm": 1.506072759628296,
+      "learning_rate": 4.336824465856724e-05,
+      "loss": 6.9688,
+      "step": 9500
+    },
+    {
+      "epoch": 4.021784666945957,
+      "grad_norm": 1.5342004299163818,
+      "learning_rate": 4.329842200809943e-05,
+      "loss": 6.9582,
+      "step": 9600
+    },
+    {
+      "epoch": 4.063678257226645,
+      "grad_norm": 1.6476861238479614,
+      "learning_rate": 4.3228599357631614e-05,
+      "loss": 6.9712,
+      "step": 9700
+    },
+    {
+      "epoch": 4.105571847507331,
+      "grad_norm": 2.112595319747925,
+      "learning_rate": 4.3158776707163805e-05,
+      "loss": 6.9568,
+      "step": 9800
+    },
+    {
+      "epoch": 4.147465437788019,
+      "grad_norm": 2.390101194381714,
+      "learning_rate": 4.3088954056695996e-05,
+      "loss": 6.9501,
+      "step": 9900
+    },
+    {
+      "epoch": 4.189359028068705,
+      "grad_norm": 1.900177240371704,
+      "learning_rate": 4.301913140622818e-05,
+      "loss": 6.935,
+      "step": 10000
+    },
+    {
+      "epoch": 4.231252618349393,
+      "grad_norm": 1.7032443284988403,
+      "learning_rate": 4.294930875576037e-05,
+      "loss": 6.9343,
+      "step": 10100
+    },
+    {
+      "epoch": 4.27314620863008,
+      "grad_norm": 1.8393847942352295,
+      "learning_rate": 4.287948610529256e-05,
+      "loss": 6.8927,
+      "step": 10200
+    },
+    {
+      "epoch": 4.3150397989107665,
+      "grad_norm": 2.046727180480957,
+      "learning_rate": 4.280966345482475e-05,
+      "loss": 6.9156,
+      "step": 10300
+    },
+    {
+      "epoch": 4.356933389191454,
+      "grad_norm": 1.832216501235962,
+      "learning_rate": 4.273984080435694e-05,
+      "loss": 6.89,
+      "step": 10400
+    },
+    {
+      "epoch": 4.39882697947214,
+      "grad_norm": 1.8682448863983154,
+      "learning_rate": 4.267001815388912e-05,
+      "loss": 6.8995,
+      "step": 10500
+    },
+    {
+      "epoch": 4.440720569752828,
+      "grad_norm": 2.0732340812683105,
+      "learning_rate": 4.260089372992599e-05,
+      "loss": 6.9094,
+      "step": 10600
+    },
+    {
+      "epoch": 4.482614160033515,
+      "grad_norm": 1.6016206741333008,
+      "learning_rate": 4.253107107945818e-05,
+      "loss": 6.913,
+      "step": 10700
+    },
+    {
+      "epoch": 4.524507750314202,
+      "grad_norm": 2.063062906265259,
+      "learning_rate": 4.246124842899037e-05,
+      "loss": 6.8943,
+      "step": 10800
+    },
+    {
+      "epoch": 4.566401340594889,
+      "grad_norm": 1.9563026428222656,
+      "learning_rate": 4.2391425778522555e-05,
+      "loss": 6.8986,
+      "step": 10900
+    },
+    {
+      "epoch": 4.6082949308755765,
+      "grad_norm": 1.8872498273849487,
+      "learning_rate": 4.2321603128054746e-05,
+      "loss": 6.8883,
+      "step": 11000
+    },
+    {
+      "epoch": 4.650188521156263,
+      "grad_norm": 2.1376144886016846,
+      "learning_rate": 4.225178047758693e-05,
+      "loss": 6.8716,
+      "step": 11100
+    },
+    {
+      "epoch": 4.69208211143695,
+      "grad_norm": 1.938679575920105,
+      "learning_rate": 4.218195782711912e-05,
+      "loss": 6.8836,
+      "step": 11200
+    },
+    {
+      "epoch": 4.733975701717637,
+      "grad_norm": 1.9372957944869995,
+      "learning_rate": 4.2112135176651305e-05,
+      "loss": 6.8925,
+      "step": 11300
+    },
+    {
+      "epoch": 4.775869291998324,
+      "grad_norm": 2.716827630996704,
+      "learning_rate": 4.2042312526183496e-05,
+      "loss": 6.8284,
+      "step": 11400
+    },
+    {
+      "epoch": 4.817762882279011,
+      "grad_norm": 1.942700743675232,
+      "learning_rate": 4.197248987571568e-05,
+      "loss": 6.8753,
+      "step": 11500
+    },
+    {
+      "epoch": 4.859656472559698,
+      "grad_norm": 2.026385545730591,
+      "learning_rate": 4.190266722524787e-05,
+      "loss": 6.8707,
+      "step": 11600
+    },
+    {
+      "epoch": 4.901550062840386,
+      "grad_norm": 1.7594517469406128,
+      "learning_rate": 4.183284457478006e-05,
+      "loss": 6.8427,
+      "step": 11700
+    },
+    {
+      "epoch": 4.943443653121072,
+      "grad_norm": 1.8161801099777222,
+      "learning_rate": 4.176302192431225e-05,
+      "loss": 6.8519,
+      "step": 11800
+    },
+    {
+      "epoch": 4.9853372434017595,
+      "grad_norm": 2.6034481525421143,
+      "learning_rate": 4.169319927384444e-05,
+      "loss": 6.8448,
+      "step": 11900
+    },
+    {
+      "epoch": 5.027230833682447,
+      "grad_norm": 1.93776535987854,
+      "learning_rate": 4.162337662337663e-05,
+      "loss": 6.8071,
+      "step": 12000
+    },
+    {
+      "epoch": 5.0691244239631335,
+      "grad_norm": 2.0754964351654053,
+      "learning_rate": 4.155355397290881e-05,
+      "loss": 6.8386,
+      "step": 12100
+    },
+    {
+      "epoch": 5.111018014243821,
+      "grad_norm": 2.0640342235565186,
+      "learning_rate": 4.1483731322440996e-05,
+      "loss": 6.8402,
+      "step": 12200
+    },
+    {
+      "epoch": 5.152911604524507,
+      "grad_norm": 1.8218064308166504,
+      "learning_rate": 4.141390867197319e-05,
+      "loss": 6.8153,
+      "step": 12300
+    },
+    {
+      "epoch": 5.194805194805195,
+      "grad_norm": 2.0181634426116943,
+      "learning_rate": 4.134408602150538e-05,
+      "loss": 6.8104,
+      "step": 12400
+    },
+    {
+      "epoch": 5.236698785085882,
+      "grad_norm": 2.5224316120147705,
+      "learning_rate": 4.127426337103757e-05,
+      "loss": 6.8355,
+      "step": 12500
+    },
+    {
+      "epoch": 5.278592375366569,
+      "grad_norm": 3.1008002758026123,
+      "learning_rate": 4.120513894707444e-05,
+      "loss": 6.8384,
+      "step": 12600
+    },
+    {
+      "epoch": 5.320485965647256,
+      "grad_norm": 1.8872394561767578,
+      "learning_rate": 4.113531629660662e-05,
+      "loss": 6.8087,
+      "step": 12700
+    },
+    {
+      "epoch": 5.362379555927943,
+      "grad_norm": 2.109281063079834,
+      "learning_rate": 4.1065493646138805e-05,
+      "loss": 6.8161,
+      "step": 12800
+    },
+    {
+      "epoch": 5.40427314620863,
+      "grad_norm": 1.7881128787994385,
+      "learning_rate": 4.0995670995670996e-05,
+      "loss": 6.8215,
+      "step": 12900
+    },
+    {
+      "epoch": 5.446166736489317,
+      "grad_norm": 2.5179624557495117,
+      "learning_rate": 4.092584834520319e-05,
+      "loss": 6.7883,
+      "step": 13000
+    },
+    {
+      "epoch": 5.488060326770004,
+      "grad_norm": 2.4349751472473145,
+      "learning_rate": 4.085602569473537e-05,
+      "loss": 6.792,
+      "step": 13100
+    },
+    {
+      "epoch": 5.529953917050691,
+      "grad_norm": 2.011018991470337,
+      "learning_rate": 4.078620304426756e-05,
+      "loss": 6.7846,
+      "step": 13200
+    },
+    {
+      "epoch": 5.571847507331379,
+      "grad_norm": 2.519958019256592,
+      "learning_rate": 4.071638039379975e-05,
+      "loss": 6.7887,
+      "step": 13300
+    },
+    {
+      "epoch": 5.613741097612065,
+      "grad_norm": 1.9241886138916016,
+      "learning_rate": 4.064655774333194e-05,
+      "loss": 6.7662,
+      "step": 13400
+    },
+    {
+      "epoch": 5.655634687892753,
+      "grad_norm": 1.8995391130447388,
+      "learning_rate": 4.057673509286413e-05,
+      "loss": 6.7672,
+      "step": 13500
+    },
+    {
+      "epoch": 5.697528278173439,
+      "grad_norm": 2.1511363983154297,
+      "learning_rate": 4.050691244239632e-05,
+      "loss": 6.7867,
+      "step": 13600
+    },
+    {
+      "epoch": 5.7394218684541265,
+      "grad_norm": 1.8995012044906616,
+      "learning_rate": 4.04370897919285e-05,
+      "loss": 6.7563,
+      "step": 13700
+    },
+    {
+      "epoch": 5.781315458734814,
+      "grad_norm": 1.83163321018219,
+      "learning_rate": 4.036726714146069e-05,
+      "loss": 6.7848,
+      "step": 13800
+    },
+    {
+      "epoch": 5.8232090490155,
+      "grad_norm": 2.2616159915924072,
+      "learning_rate": 4.029744449099288e-05,
+      "loss": 6.7896,
+      "step": 13900
+    },
+    {
+      "epoch": 5.865102639296188,
+      "grad_norm": 2.0548572540283203,
+      "learning_rate": 4.0228320067029746e-05,
+      "loss": 6.7633,
+      "step": 14000
+    },
+    {
+      "epoch": 5.906996229576874,
+      "grad_norm": 2.4749302864074707,
+      "learning_rate": 4.015849741656194e-05,
+      "loss": 6.7267,
+      "step": 14100
+    },
+    {
+      "epoch": 5.948889819857562,
+      "grad_norm": 1.906648874282837,
+      "learning_rate": 4.008867476609413e-05,
+      "loss": 6.7645,
+      "step": 14200
+    },
+    {
+      "epoch": 5.990783410138249,
+      "grad_norm": 2.0839619636535645,
+      "learning_rate": 4.001885211562631e-05,
+      "loss": 6.8082,
+      "step": 14300
+    },
+    {
+      "epoch": 6.032677000418936,
+      "grad_norm": 2.1202664375305176,
+      "learning_rate": 3.9949029465158496e-05,
+      "loss": 6.7625,
+      "step": 14400
+    },
+    {
+      "epoch": 6.074570590699623,
+      "grad_norm": 1.988951563835144,
+      "learning_rate": 3.987920681469069e-05,
+      "loss": 6.7413,
+      "step": 14500
+    },
+    {
+      "epoch": 6.11646418098031,
+      "grad_norm": 2.4327659606933594,
+      "learning_rate": 3.980938416422287e-05,
+      "loss": 6.7123,
+      "step": 14600
+    },
+    {
+      "epoch": 6.158357771260997,
+      "grad_norm": 2.07710599899292,
+      "learning_rate": 3.973956151375506e-05,
+      "loss": 6.7316,
+      "step": 14700
+    },
+    {
+      "epoch": 6.200251361541684,
+      "grad_norm": 1.9640876054763794,
+      "learning_rate": 3.966973886328725e-05,
+      "loss": 6.752,
+      "step": 14800
+    },
+    {
+      "epoch": 6.242144951822371,
+      "grad_norm": 2.3012888431549072,
+      "learning_rate": 3.959991621281944e-05,
+      "loss": 6.7188,
+      "step": 14900
+    },
+    {
+      "epoch": 6.284038542103058,
+      "grad_norm": 2.0262773036956787,
+      "learning_rate": 3.953009356235163e-05,
+      "loss": 6.7255,
+      "step": 15000
+    },
+    {
+      "epoch": 6.325932132383746,
+      "grad_norm": 1.8689815998077393,
+      "learning_rate": 3.946027091188382e-05,
+      "loss": 6.7132,
+      "step": 15100
+    },
+    {
+      "epoch": 6.367825722664432,
+      "grad_norm": 2.188612937927246,
+      "learning_rate": 3.939044826141601e-05,
+      "loss": 6.7407,
+      "step": 15200
+    },
+    {
+      "epoch": 6.4097193129451195,
+      "grad_norm": 2.0168368816375732,
+      "learning_rate": 3.9320625610948195e-05,
+      "loss": 6.7132,
+      "step": 15300
+    },
+    {
+      "epoch": 6.451612903225806,
+      "grad_norm": 2.496889352798462,
+      "learning_rate": 3.925080296048038e-05,
+      "loss": 6.7003,
+      "step": 15400
+    },
+    {
+      "epoch": 6.4935064935064934,
+      "grad_norm": 2.1601486206054688,
+      "learning_rate": 3.918098031001257e-05,
+      "loss": 6.7056,
+      "step": 15500
+    },
+    {
+      "epoch": 6.535400083787181,
+      "grad_norm": 2.300112009048462,
+      "learning_rate": 3.9111157659544754e-05,
+      "loss": 6.7314,
+      "step": 15600
+    },
+    {
+      "epoch": 6.577293674067867,
+      "grad_norm": 2.321880578994751,
+      "learning_rate": 3.9041335009076945e-05,
+      "loss": 6.7166,
+      "step": 15700
+    },
+    {
+      "epoch": 6.619187264348555,
+      "grad_norm": 2.029465913772583,
+      "learning_rate": 3.8971512358609136e-05,
+      "loss": 6.6908,
+      "step": 15800
+    },
+    {
+      "epoch": 6.661080854629242,
+      "grad_norm": 2.258577585220337,
+      "learning_rate": 3.890168970814133e-05,
+      "loss": 6.7359,
+      "step": 15900
+    },
+    {
+      "epoch": 6.702974444909929,
+      "grad_norm": 2.3579437732696533,
+      "learning_rate": 3.883186705767351e-05,
+      "loss": 6.7021,
+      "step": 16000
+    },
+    {
+      "epoch": 6.744868035190616,
+      "grad_norm": 2.236828565597534,
+      "learning_rate": 3.87620444072057e-05,
+      "loss": 6.6897,
+      "step": 16100
+    },
+    {
+      "epoch": 6.786761625471303,
+      "grad_norm": 2.6255593299865723,
+      "learning_rate": 3.869222175673789e-05,
+      "loss": 6.6899,
+      "step": 16200
+    },
+    {
+      "epoch": 6.82865521575199,
+      "grad_norm": 2.297067880630493,
+      "learning_rate": 3.862239910627008e-05,
+      "loss": 6.7058,
+      "step": 16300
+    },
+    {
+      "epoch": 6.870548806032677,
+      "grad_norm": 2.440605640411377,
+      "learning_rate": 3.8553274682306945e-05,
+      "loss": 6.6559,
+      "step": 16400
+    },
+    {
+      "epoch": 6.912442396313364,
+      "grad_norm": 2.0427000522613525,
+      "learning_rate": 3.848345203183913e-05,
+      "loss": 6.6799,
+      "step": 16500
+    },
+    {
+      "epoch": 6.954335986594051,
+      "grad_norm": 2.0323081016540527,
+      "learning_rate": 3.841362938137132e-05,
+      "loss": 6.6863,
+      "step": 16600
+    },
+    {
+      "epoch": 6.996229576874738,
+      "grad_norm": 3.407731533050537,
+      "learning_rate": 3.834380673090351e-05,
+      "loss": 6.6767,
+      "step": 16700
+    },
+    {
+      "epoch": 7.038123167155425,
+      "grad_norm": 2.112870931625366,
+      "learning_rate": 3.8273984080435695e-05,
+      "loss": 6.682,
+      "step": 16800
+    },
+    {
+      "epoch": 7.080016757436113,
+      "grad_norm": 2.710810422897339,
+      "learning_rate": 3.8204161429967886e-05,
+      "loss": 6.7046,
+      "step": 16900
+    },
+    {
+      "epoch": 7.121910347716799,
+      "grad_norm": 2.0754942893981934,
+      "learning_rate": 3.813433877950007e-05,
+      "loss": 6.6511,
+      "step": 17000
+    },
+    {
+      "epoch": 7.1638039379974865,
+      "grad_norm": 3.1009552478790283,
+      "learning_rate": 3.8064516129032254e-05,
+      "loss": 6.666,
+      "step": 17100
+    },
+    {
+      "epoch": 7.205697528278174,
+      "grad_norm": 2.1582441329956055,
+      "learning_rate": 3.7994693478564445e-05,
+      "loss": 6.6574,
+      "step": 17200
+    },
+    {
+      "epoch": 7.24759111855886,
+      "grad_norm": 2.680147647857666,
+      "learning_rate": 3.7924870828096636e-05,
+      "loss": 6.6814,
+      "step": 17300
+    },
+    {
+      "epoch": 7.289484708839548,
+      "grad_norm": 2.0264320373535156,
+      "learning_rate": 3.785504817762883e-05,
+      "loss": 6.668,
+      "step": 17400
+    },
+    {
+      "epoch": 7.331378299120234,
+      "grad_norm": 2.032093048095703,
+      "learning_rate": 3.778522552716101e-05,
+      "loss": 6.6603,
+      "step": 17500
+    },
+    {
+      "epoch": 7.373271889400922,
+      "grad_norm": 2.4837894439697266,
+      "learning_rate": 3.77154028766932e-05,
+      "loss": 6.6817,
+      "step": 17600
+    },
+    {
+      "epoch": 7.415165479681609,
+      "grad_norm": 2.70166015625,
+      "learning_rate": 3.764558022622539e-05,
+      "loss": 6.6657,
+      "step": 17700
+    },
+    {
+      "epoch": 7.457059069962296,
+      "grad_norm": 2.3508477210998535,
+      "learning_rate": 3.757575757575758e-05,
+      "loss": 6.6314,
+      "step": 17800
+    },
+    {
+      "epoch": 7.498952660242983,
+      "grad_norm": 2.450437307357788,
+      "learning_rate": 3.750593492528977e-05,
+      "loss": 6.6551,
+      "step": 17900
+    },
+    {
+      "epoch": 7.5408462505236695,
+      "grad_norm": 1.9939864873886108,
+      "learning_rate": 3.743611227482195e-05,
+      "loss": 6.6128,
+      "step": 18000
+    },
+    {
+      "epoch": 7.582739840804357,
+      "grad_norm": 2.470285177230835,
+      "learning_rate": 3.736628962435414e-05,
+      "loss": 6.6126,
+      "step": 18100
+    },
+    {
+      "epoch": 7.624633431085044,
+      "grad_norm": 2.5651469230651855,
+      "learning_rate": 3.729646697388633e-05,
+      "loss": 6.6694,
+      "step": 18200
+    },
+    {
+      "epoch": 7.666527021365731,
+      "grad_norm": 2.361785650253296,
+      "learning_rate": 3.722664432341852e-05,
+      "loss": 6.6349,
+      "step": 18300
+    },
+    {
+      "epoch": 7.708420611646418,
+      "grad_norm": 2.371994972229004,
+      "learning_rate": 3.715682167295071e-05,
+      "loss": 6.6483,
+      "step": 18400
+    },
+    {
+      "epoch": 7.750314201927106,
+      "grad_norm": 2.862107038497925,
+      "learning_rate": 3.708769724898758e-05,
+      "loss": 6.634,
+      "step": 18500
+    },
+    {
+      "epoch": 7.792207792207792,
+      "grad_norm": 2.815486192703247,
+      "learning_rate": 3.701787459851976e-05,
+      "loss": 6.6324,
+      "step": 18600
+    },
+    {
+      "epoch": 7.8341013824884795,
+      "grad_norm": 1.930017352104187,
+      "learning_rate": 3.6948051948051945e-05,
+      "loss": 6.6275,
+      "step": 18700
+    },
+    {
+      "epoch": 7.875994972769166,
+      "grad_norm": 3.1758625507354736,
+      "learning_rate": 3.6878229297584136e-05,
+      "loss": 6.6529,
+      "step": 18800
+    },
+    {
+      "epoch": 7.9178885630498534,
+      "grad_norm": 2.1219429969787598,
+      "learning_rate": 3.680840664711633e-05,
+      "loss": 6.6085,
+      "step": 18900
+    },
+    {
+      "epoch": 7.95978215333054,
+      "grad_norm": 2.1965785026550293,
+      "learning_rate": 3.673858399664851e-05,
+      "loss": 6.6206,
+      "step": 19000
+    },
+    {
+      "epoch": 8.001675743611228,
+      "grad_norm": 2.489473581314087,
+      "learning_rate": 3.66687613461807e-05,
+      "loss": 6.6089,
+      "step": 19100
+    },
+    {
+      "epoch": 8.043569333891915,
+      "grad_norm": 2.3411850929260254,
+      "learning_rate": 3.659893869571289e-05,
+      "loss": 6.6286,
+      "step": 19200
+    },
+    {
+      "epoch": 8.085462924172601,
+      "grad_norm": 2.32071590423584,
+      "learning_rate": 3.6529116045245084e-05,
+      "loss": 6.5984,
+      "step": 19300
+    },
+    {
+      "epoch": 8.12735651445329,
+      "grad_norm": 2.402956247329712,
+      "learning_rate": 3.645929339477727e-05,
+      "loss": 6.5952,
+      "step": 19400
+    },
+    {
+      "epoch": 8.169250104733976,
+      "grad_norm": 2.6951029300689697,
+      "learning_rate": 3.638947074430946e-05,
+      "loss": 6.6106,
+      "step": 19500
+    },
+    {
+      "epoch": 8.211143695014663,
+      "grad_norm": 2.807187080383301,
+      "learning_rate": 3.6319648093841643e-05,
+      "loss": 6.6109,
+      "step": 19600
+    },
+    {
+      "epoch": 8.253037285295349,
+      "grad_norm": 2.798614025115967,
+      "learning_rate": 3.624982544337383e-05,
+      "loss": 6.6052,
+      "step": 19700
+    },
+    {
+      "epoch": 8.294930875576037,
+      "grad_norm": 4.015589237213135,
+      "learning_rate": 3.618000279290602e-05,
+      "loss": 6.5995,
+      "step": 19800
+    },
+    {
+      "epoch": 8.336824465856724,
+      "grad_norm": 2.6923959255218506,
+      "learning_rate": 3.611018014243821e-05,
+      "loss": 6.5855,
+      "step": 19900
+    },
+    {
+      "epoch": 8.37871805613741,
+      "grad_norm": 2.112994909286499,
+      "learning_rate": 3.6040357491970394e-05,
+      "loss": 6.5968,
+      "step": 20000
+    },
+    {
+      "epoch": 8.420611646418099,
+      "grad_norm": 2.8196451663970947,
+      "learning_rate": 3.5970534841502585e-05,
+      "loss": 6.5977,
+      "step": 20100
+    },
+    {
+      "epoch": 8.462505236698785,
+      "grad_norm": 2.2421326637268066,
+      "learning_rate": 3.5900712191034776e-05,
+      "loss": 6.5846,
+      "step": 20200
+    },
+    {
+      "epoch": 8.504398826979472,
+      "grad_norm": 2.634634256362915,
+      "learning_rate": 3.583088954056697e-05,
+      "loss": 6.5955,
+      "step": 20300
+    },
+    {
+      "epoch": 8.54629241726016,
+      "grad_norm": 2.101125955581665,
+      "learning_rate": 3.576106689009915e-05,
+      "loss": 6.6013,
+      "step": 20400
+    },
+    {
+      "epoch": 8.588186007540846,
+      "grad_norm": 2.719330072402954,
+      "learning_rate": 3.569194246613601e-05,
+      "loss": 6.5668,
+      "step": 20500
+    },
+    {
+      "epoch": 8.630079597821533,
+      "grad_norm": 2.283790349960327,
+      "learning_rate": 3.56221198156682e-05,
+      "loss": 6.6107,
+      "step": 20600
+    },
+    {
+      "epoch": 8.671973188102221,
+      "grad_norm": 2.1805171966552734,
+      "learning_rate": 3.5552297165200393e-05,
+      "loss": 6.5875,
+      "step": 20700
+    },
+    {
+      "epoch": 8.713866778382908,
+      "grad_norm": 2.6632487773895264,
+      "learning_rate": 3.5482474514732584e-05,
+      "loss": 6.613,
+      "step": 20800
+    },
+    {
+      "epoch": 8.755760368663594,
+      "grad_norm": 2.3296337127685547,
+      "learning_rate": 3.541265186426477e-05,
+      "loss": 6.5628,
+      "step": 20900
+    },
+    {
+      "epoch": 8.79765395894428,
+      "grad_norm": 2.8429343700408936,
+      "learning_rate": 3.534282921379696e-05,
+      "loss": 6.5823,
+      "step": 21000
+    },
+    {
+      "epoch": 8.839547549224969,
+      "grad_norm": 2.4361233711242676,
+      "learning_rate": 3.527300656332915e-05,
+      "loss": 6.5853,
+      "step": 21100
+    },
+    {
+      "epoch": 8.881441139505656,
+      "grad_norm": 2.5633111000061035,
+      "learning_rate": 3.5203183912861335e-05,
+      "loss": 6.5979,
+      "step": 21200
+    },
+    {
+      "epoch": 8.923334729786342,
+      "grad_norm": 2.350463628768921,
+      "learning_rate": 3.513336126239352e-05,
+      "loss": 6.5744,
+      "step": 21300
+    },
+    {
+      "epoch": 8.96522832006703,
+      "grad_norm": 2.456291675567627,
+      "learning_rate": 3.506353861192571e-05,
+      "loss": 6.57,
+      "step": 21400
+    },
+    {
+      "epoch": 9.007121910347717,
+      "grad_norm": 2.401036262512207,
+      "learning_rate": 3.49937159614579e-05,
+      "loss": 6.5614,
+      "step": 21500
+    },
+    {
+      "epoch": 9.049015500628403,
+      "grad_norm": 2.5537233352661133,
+      "learning_rate": 3.4923893310990085e-05,
+      "loss": 6.5836,
+      "step": 21600
+    },
+    {
+      "epoch": 9.090909090909092,
+      "grad_norm": 2.6386375427246094,
+      "learning_rate": 3.4854070660522276e-05,
+      "loss": 6.6178,
+      "step": 21700
+    },
+    {
+      "epoch": 9.132802681189778,
+      "grad_norm": 2.508533477783203,
+      "learning_rate": 3.478424801005447e-05,
+      "loss": 6.5761,
+      "step": 21800
+    },
+    {
+      "epoch": 9.174696271470465,
+      "grad_norm": 3.1510419845581055,
+      "learning_rate": 3.471442535958665e-05,
+      "loss": 6.558,
+      "step": 21900
+    },
+    {
+      "epoch": 9.216589861751151,
+      "grad_norm": 2.6325526237487793,
+      "learning_rate": 3.464460270911884e-05,
+      "loss": 6.5661,
+      "step": 22000
+    },
+    {
+      "epoch": 9.25848345203184,
+      "grad_norm": 2.9870827198028564,
+      "learning_rate": 3.457478005865103e-05,
+      "loss": 6.5392,
+      "step": 22100
+    },
+    {
+      "epoch": 9.300377042312526,
+      "grad_norm": 2.4924209117889404,
+      "learning_rate": 3.450495740818322e-05,
+      "loss": 6.547,
+      "step": 22200
+    },
+    {
+      "epoch": 9.342270632593213,
+      "grad_norm": 2.3227298259735107,
+      "learning_rate": 3.44351347577154e-05,
+      "loss": 6.5306,
+      "step": 22300
+    },
+    {
+      "epoch": 9.3841642228739,
+      "grad_norm": 2.867182731628418,
+      "learning_rate": 3.436531210724759e-05,
+      "loss": 6.5628,
+      "step": 22400
+    },
+    {
+      "epoch": 9.426057813154587,
+      "grad_norm": 2.2619149684906006,
+      "learning_rate": 3.429548945677978e-05,
+      "loss": 6.5192,
+      "step": 22500
+    },
+    {
+      "epoch": 9.467951403435274,
+      "grad_norm": 2.232321262359619,
+      "learning_rate": 3.422636503281665e-05,
+      "loss": 6.56,
+      "step": 22600
+    },
+    {
+      "epoch": 9.509844993715962,
+      "grad_norm": 2.4485862255096436,
+      "learning_rate": 3.4156542382348835e-05,
+      "loss": 6.557,
+      "step": 22700
+    },
+    {
+      "epoch": 9.551738583996649,
+      "grad_norm": 2.4476943016052246,
+      "learning_rate": 3.4086719731881026e-05,
+      "loss": 6.5314,
+      "step": 22800
+    },
+    {
+      "epoch": 9.593632174277335,
+      "grad_norm": 2.491731643676758,
+      "learning_rate": 3.401689708141321e-05,
+      "loss": 6.4952,
+      "step": 22900
+    },
+    {
+      "epoch": 9.635525764558023,
+      "grad_norm": 2.6474783420562744,
+      "learning_rate": 3.39470744309454e-05,
+      "loss": 6.5499,
+      "step": 23000
+    },
+    {
+      "epoch": 9.67741935483871,
+      "grad_norm": 2.5691514015197754,
+      "learning_rate": 3.3877251780477585e-05,
+      "loss": 6.5417,
+      "step": 23100
+    },
+    {
+      "epoch": 9.719312945119396,
+      "grad_norm": 2.601832151412964,
+      "learning_rate": 3.3807429130009776e-05,
+      "loss": 6.5584,
+      "step": 23200
+    },
+    {
+      "epoch": 9.761206535400083,
+      "grad_norm": 3.481239080429077,
+      "learning_rate": 3.373760647954197e-05,
+      "loss": 6.5403,
+      "step": 23300
+    },
+    {
+      "epoch": 9.803100125680771,
+      "grad_norm": 3.0747485160827637,
+      "learning_rate": 3.366778382907415e-05,
+      "loss": 6.5751,
+      "step": 23400
+    },
+    {
+      "epoch": 9.844993715961458,
+      "grad_norm": 2.2310988903045654,
+      "learning_rate": 3.359796117860634e-05,
+      "loss": 6.5046,
+      "step": 23500
+    },
+    {
+      "epoch": 9.886887306242144,
+      "grad_norm": 2.4555273056030273,
+      "learning_rate": 3.352813852813853e-05,
+      "loss": 6.5544,
+      "step": 23600
+    },
+    {
+      "epoch": 9.928780896522833,
+      "grad_norm": 3.1235666275024414,
+      "learning_rate": 3.345831587767072e-05,
+      "loss": 6.5396,
+      "step": 23700
+    },
+    {
+      "epoch": 9.970674486803519,
+      "grad_norm": 2.2766611576080322,
+      "learning_rate": 3.338849322720291e-05,
+      "loss": 6.5542,
+      "step": 23800
+    },
+    {
+      "epoch": 10.012568077084206,
+      "grad_norm": 3.0408995151519775,
+      "learning_rate": 3.331867057673509e-05,
+      "loss": 6.4978,
+      "step": 23900
+    },
+    {
+      "epoch": 10.054461667364894,
+      "grad_norm": 2.8702831268310547,
+      "learning_rate": 3.3248847926267283e-05,
+      "loss": 6.5264,
+      "step": 24000
+    },
+    {
+      "epoch": 10.09635525764558,
+      "grad_norm": 2.9117937088012695,
+      "learning_rate": 3.317902527579947e-05,
+      "loss": 6.5028,
+      "step": 24100
+    },
+    {
+      "epoch": 10.138248847926267,
+      "grad_norm": 2.925631046295166,
+      "learning_rate": 3.310920262533166e-05,
+      "loss": 6.5143,
+      "step": 24200
+    },
+    {
+      "epoch": 10.180142438206955,
+      "grad_norm": 2.6605536937713623,
+      "learning_rate": 3.303937997486385e-05,
+      "loss": 6.5394,
+      "step": 24300
+    },
+    {
+      "epoch": 10.222036028487642,
+      "grad_norm": 2.31357479095459,
+      "learning_rate": 3.2969557324396034e-05,
+      "loss": 6.5224,
+      "step": 24400
+    },
+    {
+      "epoch": 10.263929618768328,
+      "grad_norm": 2.6544747352600098,
+      "learning_rate": 3.2899734673928225e-05,
+      "loss": 6.5035,
+      "step": 24500
+    },
+    {
+      "epoch": 10.305823209049015,
+      "grad_norm": 2.5945372581481934,
+      "learning_rate": 3.2830610249965085e-05,
+      "loss": 6.4977,
+      "step": 24600
+    },
+    {
+      "epoch": 10.347716799329703,
+      "grad_norm": 3.120873212814331,
+      "learning_rate": 3.2760787599497276e-05,
+      "loss": 6.5399,
+      "step": 24700
+    },
+    {
+      "epoch": 10.38961038961039,
+      "grad_norm": 2.705008029937744,
+      "learning_rate": 3.269096494902947e-05,
+      "loss": 6.4938,
+      "step": 24800
     },
     {
-      "epoch": 1.0473397570171763,
-      "grad_norm": 1.396130084991455,
-      "learning_rate": 4.8952869710934226e-05,
-      "loss": 7.2682,
-      "step": 5000
+      "epoch": 10.431503979891076,
+      "grad_norm": 2.2395503520965576,
+      "learning_rate": 3.262114229856166e-05,
+      "loss": 6.4854,
+      "step": 24900
     },
     {
-      "epoch": 1.1520737327188941,
-      "grad_norm": 1.5962079763412476,
-      "learning_rate": 4.884813573523251e-05,
-      "loss": 7.2441,
-      "step": 5500
+      "epoch": 10.473397570171764,
+      "grad_norm": 2.5891764163970947,
+      "learning_rate": 3.255131964809384e-05,
+      "loss": 6.5107,
+      "step": 25000
     },
     {
-      "epoch": 1.2568077084206117,
-      "grad_norm": 1.6328166723251343,
-      "learning_rate": 4.874340175953079e-05,
-      "loss": 7.2216,
-      "step": 6000
+      "epoch": 10.51529116045245,
+      "grad_norm": 3.115931749343872,
+      "learning_rate": 3.248149699762603e-05,
+      "loss": 6.5389,
+      "step": 25100
     },
     {
-      "epoch": 1.3615416841223293,
-      "grad_norm": 1.8534362316131592,
-      "learning_rate": 4.863887725178048e-05,
-      "loss": 7.1957,
-      "step": 6500
+      "epoch": 10.557184750733137,
+      "grad_norm": 2.264437675476074,
+      "learning_rate": 3.2411674347158224e-05,
+      "loss": 6.51,
+      "step": 25200
     },
     {
-      "epoch": 1.466275659824047,
-      "grad_norm": 1.4871692657470703,
-      "learning_rate": 4.8534143276078766e-05,
-      "loss": 7.1561,
-      "step": 7000
+      "epoch": 10.599078341013826,
+      "grad_norm": 3.449631690979004,
+      "learning_rate": 3.234185169669041e-05,
+      "loss": 6.5161,
+      "step": 25300
     },
     {
-      "epoch": 1.5710096355257646,
-      "grad_norm": 1.8590672016143799,
-      "learning_rate": 4.8429409300377045e-05,
-      "loss": 7.1397,
-      "step": 7500
+      "epoch": 10.640971931294512,
+      "grad_norm": 2.478337526321411,
+      "learning_rate": 3.227202904622259e-05,
+      "loss": 6.5019,
+      "step": 25400
     },
     {
-      "epoch": 1.6757436112274822,
-      "grad_norm": 1.7009446620941162,
-      "learning_rate": 4.8324675324675325e-05,
-      "loss": 7.1149,
-      "step": 8000
+      "epoch": 10.682865521575199,
+      "grad_norm": 3.2756478786468506,
+      "learning_rate": 3.2202206395754784e-05,
+      "loss": 6.4869,
+      "step": 25500
     },
     {
-      "epoch": 1.7804775869291998,
-      "grad_norm": 1.9020010232925415,
-      "learning_rate": 4.822015081692501e-05,
-      "loss": 7.1215,
-      "step": 8500
+      "epoch": 10.724759111855885,
+      "grad_norm": 2.7576985359191895,
+      "learning_rate": 3.213238374528697e-05,
+      "loss": 6.5206,
+      "step": 25600
     },
     {
-      "epoch": 1.8852115626309174,
-      "grad_norm": 2.912442445755005,
-      "learning_rate": 4.811541684122329e-05,
-      "loss": 7.0616,
-      "step": 9000
+      "epoch": 10.766652702136573,
+      "grad_norm": 2.200963020324707,
+      "learning_rate": 3.206256109481916e-05,
+      "loss": 6.48,
+      "step": 25700
     },
     {
-      "epoch": 1.989945538332635,
-      "grad_norm": 2.654263496398926,
-      "learning_rate": 4.801068286552158e-05,
-      "loss": 7.0508,
-      "step": 9500
+      "epoch": 10.80854629241726,
+      "grad_norm": 2.7358744144439697,
+      "learning_rate": 3.199273844435135e-05,
+      "loss": 6.5126,
+      "step": 25800
     },
     {
-      "epoch": 2.0946795140343526,
-      "grad_norm": 2.1642003059387207,
-      "learning_rate": 4.790594888981986e-05,
-      "loss": 7.0251,
-      "step": 10000
+      "epoch": 10.850439882697946,
+      "grad_norm": 2.7179319858551025,
+      "learning_rate": 3.192291579388354e-05,
+      "loss": 6.4699,
+      "step": 25900
     },
     {
-      "epoch": 2.19941348973607,
-      "grad_norm": 1.9420874118804932,
-      "learning_rate": 4.7801424382069544e-05,
-      "loss": 6.9806,
-      "step": 10500
+      "epoch": 10.892333472978635,
+      "grad_norm": 2.811340808868408,
+      "learning_rate": 3.1853093143415725e-05,
+      "loss": 6.5056,
+      "step": 26000
     },
     {
-      "epoch": 2.3041474654377883,
-      "grad_norm": 2.2306201457977295,
-      "learning_rate": 4.769669040636783e-05,
-      "loss": 6.9721,
-      "step": 11000
+      "epoch": 10.934227063259321,
+      "grad_norm": 3.010690450668335,
+      "learning_rate": 3.1783270492947916e-05,
+      "loss": 6.5103,
+      "step": 26100
     },
     {
-      "epoch": 2.4088814411395054,
-      "grad_norm": 2.8180389404296875,
-      "learning_rate": 4.759195643066611e-05,
-      "loss": 6.9582,
-      "step": 11500
+      "epoch": 10.976120653540008,
+      "grad_norm": 3.213487148284912,
+      "learning_rate": 3.171344784248011e-05,
+      "loss": 6.4874,
+      "step": 26200
     },
     {
-      "epoch": 2.5136154168412235,
-      "grad_norm": 2.387949228286743,
-      "learning_rate": 4.748722245496439e-05,
-      "loss": 6.9211,
-      "step": 12000
+      "epoch": 11.018014243820696,
+      "grad_norm": 2.5710039138793945,
+      "learning_rate": 3.164362519201229e-05,
+      "loss": 6.4919,
+      "step": 26300
     },
     {
-      "epoch": 2.618349392542941,
-      "grad_norm": 3.3709394931793213,
-      "learning_rate": 4.7382697947214076e-05,
-      "loss": 6.9183,
-      "step": 12500
+      "epoch": 11.059907834101383,
+      "grad_norm": 2.6933746337890625,
+      "learning_rate": 3.1573802541544475e-05,
+      "loss": 6.5284,
+      "step": 26400
     },
     {
-      "epoch": 2.7230833682446587,
-      "grad_norm": 2.567798376083374,
-      "learning_rate": 4.727796397151236e-05,
-      "loss": 6.8732,
-      "step": 13000
+      "epoch": 11.101801424382069,
+      "grad_norm": 3.775012254714966,
+      "learning_rate": 3.1503979891076666e-05,
+      "loss": 6.4894,
+      "step": 26500
     },
     {
-      "epoch": 2.8278173439463763,
-      "grad_norm": 2.6373414993286133,
-      "learning_rate": 4.717322999581064e-05,
-      "loss": 6.8658,
-      "step": 13500
+      "epoch": 11.143695014662757,
+      "grad_norm": 3.2401301860809326,
+      "learning_rate": 3.1434855467113534e-05,
+      "loss": 6.4721,
+      "step": 26600
     },
     {
-      "epoch": 2.932551319648094,
-      "grad_norm": 2.2950875759124756,
-      "learning_rate": 4.706849602010893e-05,
-      "loss": 6.8436,
-      "step": 14000
+      "epoch": 11.185588604943444,
+      "grad_norm": 2.642794132232666,
+      "learning_rate": 3.1365032816645725e-05,
+      "loss": 6.4797,
+      "step": 26700
     },
     {
-      "epoch": 3.0372852953498115,
-      "grad_norm": 4.0021514892578125,
-      "learning_rate": 4.696397151235861e-05,
-      "loss": 6.8496,
-      "step": 14500
+      "epoch": 11.22748219522413,
+      "grad_norm": 3.191567897796631,
+      "learning_rate": 3.129521016617791e-05,
+      "loss": 6.5022,
+      "step": 26800
     },
     {
-      "epoch": 3.142019271051529,
-      "grad_norm": 3.289193630218506,
-      "learning_rate": 4.685923753665689e-05,
-      "loss": 6.8047,
-      "step": 15000
+      "epoch": 11.269375785504817,
+      "grad_norm": 2.816554307937622,
+      "learning_rate": 3.12253875157101e-05,
+      "loss": 6.4853,
+      "step": 26900
     },
     {
-      "epoch": 3.2467532467532467,
-      "grad_norm": 2.9973654747009277,
-      "learning_rate": 4.6754503560955175e-05,
-      "loss": 6.7651,
-      "step": 15500
+      "epoch": 11.311269375785505,
+      "grad_norm": 2.8666136264801025,
+      "learning_rate": 3.1155564865242284e-05,
+      "loss": 6.4839,
+      "step": 27000
     },
     {
-      "epoch": 3.3514872224549643,
-      "grad_norm": 2.9979376792907715,
-      "learning_rate": 4.664976958525346e-05,
-      "loss": 6.7768,
-      "step": 16000
+      "epoch": 11.353162966066192,
+      "grad_norm": 2.9831254482269287,
+      "learning_rate": 3.1085742214774475e-05,
+      "loss": 6.5082,
+      "step": 27100
     },
     {
-      "epoch": 3.456221198156682,
-      "grad_norm": 3.263784885406494,
-      "learning_rate": 4.654524507750315e-05,
-      "loss": 6.7617,
-      "step": 16500
+      "epoch": 11.395056556346878,
+      "grad_norm": 2.7065083980560303,
+      "learning_rate": 3.101591956430666e-05,
+      "loss": 6.4412,
+      "step": 27200
     },
     {
-      "epoch": 3.5609551738583995,
-      "grad_norm": 3.330116033554077,
-      "learning_rate": 4.644051110180143e-05,
-      "loss": 6.7417,
-      "step": 17000
+      "epoch": 11.436950146627566,
+      "grad_norm": 2.5580694675445557,
+      "learning_rate": 3.094609691383885e-05,
+      "loss": 6.4849,
+      "step": 27300
     },
     {
-      "epoch": 3.665689149560117,
-      "grad_norm": 3.224337339401245,
-      "learning_rate": 4.6335777126099714e-05,
-      "loss": 6.7028,
-      "step": 17500
+      "epoch": 11.478843736908253,
+      "grad_norm": 2.571390390396118,
+      "learning_rate": 3.087627426337104e-05,
+      "loss": 6.4689,
+      "step": 27400
     },
     {
-      "epoch": 3.7704231252618348,
-      "grad_norm": 3.21891450881958,
-      "learning_rate": 4.623104315039799e-05,
-      "loss": 6.7101,
-      "step": 18000
+      "epoch": 11.52073732718894,
+      "grad_norm": 2.835906982421875,
+      "learning_rate": 3.0806451612903225e-05,
+      "loss": 6.4887,
+      "step": 27500
     },
     {
-      "epoch": 3.875157100963553,
-      "grad_norm": 2.3559324741363525,
-      "learning_rate": 4.6126518642647674e-05,
-      "loss": 6.6892,
-      "step": 18500
+      "epoch": 11.562630917469628,
+      "grad_norm": 3.1355161666870117,
+      "learning_rate": 3.0736628962435416e-05,
+      "loss": 6.4568,
+      "step": 27600
     },
     {
-      "epoch": 3.97989107666527,
-      "grad_norm": 3.1527633666992188,
-      "learning_rate": 4.602178466694596e-05,
-      "loss": 6.6851,
-      "step": 19000
+      "epoch": 11.604524507750314,
+      "grad_norm": 3.0155599117279053,
+      "learning_rate": 3.066680631196761e-05,
+      "loss": 6.4607,
+      "step": 27700
     },
     {
-      "epoch": 4.084625052366988,
-      "grad_norm": 2.9760189056396484,
-      "learning_rate": 4.591705069124424e-05,
-      "loss": 6.6704,
-      "step": 19500
+      "epoch": 11.646418098031,
+      "grad_norm": 2.6346957683563232,
+      "learning_rate": 3.059698366149979e-05,
+      "loss": 6.4706,
+      "step": 27800
     },
     {
-      "epoch": 4.189359028068705,
-      "grad_norm": 2.8135318756103516,
-      "learning_rate": 4.5812316715542526e-05,
-      "loss": 6.6621,
-      "step": 20000
+      "epoch": 11.688311688311689,
+      "grad_norm": 2.4353625774383545,
+      "learning_rate": 3.052716101103198e-05,
+      "loss": 6.482,
+      "step": 27900
     },
     {
-      "epoch": 4.294093003770423,
-      "grad_norm": 3.060316324234009,
-      "learning_rate": 4.570779220779221e-05,
-      "loss": 6.6298,
-      "step": 20500
+      "epoch": 11.730205278592376,
+      "grad_norm": 3.29835844039917,
+      "learning_rate": 3.045733836056417e-05,
+      "loss": 6.4625,
+      "step": 28000
     },
     {
-      "epoch": 4.39882697947214,
-      "grad_norm": 2.7130279541015625,
-      "learning_rate": 4.560305823209049e-05,
-      "loss": 6.628,
-      "step": 21000
+      "epoch": 11.772098868873062,
+      "grad_norm": 2.233579158782959,
+      "learning_rate": 3.038751571009636e-05,
+      "loss": 6.4727,
+      "step": 28100
     },
     {
-      "epoch": 4.5035609551738585,
-      "grad_norm": 3.156386613845825,
-      "learning_rate": 4.549853372434017e-05,
-      "loss": 6.6423,
-      "step": 21500
+      "epoch": 11.813992459153749,
+      "grad_norm": 2.5708439350128174,
+      "learning_rate": 3.0317693059628545e-05,
+      "loss": 6.4751,
+      "step": 28200
     },
     {
-      "epoch": 4.6082949308755765,
-      "grad_norm": 3.039471387863159,
-      "learning_rate": 4.539379974863846e-05,
-      "loss": 6.6343,
-      "step": 22000
+      "epoch": 11.855886049434437,
+      "grad_norm": 2.29488205909729,
+      "learning_rate": 3.0247870409160732e-05,
+      "loss": 6.4599,
+      "step": 28300
     },
     {
-      "epoch": 4.713028906577294,
-      "grad_norm": 3.976949453353882,
-      "learning_rate": 4.5289065772936745e-05,
-      "loss": 6.606,
-      "step": 22500
+      "epoch": 11.897779639715123,
+      "grad_norm": 2.858208179473877,
+      "learning_rate": 3.0178047758692923e-05,
+      "loss": 6.469,
+      "step": 28400
     },
     {
-      "epoch": 4.817762882279011,
-      "grad_norm": 3.310382604598999,
-      "learning_rate": 4.5184331797235025e-05,
-      "loss": 6.5956,
-      "step": 23000
+      "epoch": 11.93967322999581,
+      "grad_norm": 2.854923725128174,
+      "learning_rate": 3.0108225108225107e-05,
+      "loss": 6.4995,
+      "step": 28500
     },
     {
-      "epoch": 4.922496857980729,
-      "grad_norm": 3.5924322605133057,
-      "learning_rate": 4.507959782153331e-05,
-      "loss": 6.5965,
-      "step": 23500
+      "epoch": 11.981566820276498,
+      "grad_norm": 2.590484857559204,
+      "learning_rate": 3.003910068426198e-05,
+      "loss": 6.4508,
+      "step": 28600
     },
     {
-      "epoch": 5.027230833682447,
-      "grad_norm": 2.616468667984009,
-      "learning_rate": 4.497486384583159e-05,
-      "loss": 6.5788,
-      "step": 24000
+      "epoch": 12.023460410557185,
+      "grad_norm": 3.3479676246643066,
+      "learning_rate": 2.9969278033794163e-05,
+      "loss": 6.4581,
+      "step": 28700
     },
     {
-      "epoch": 5.131964809384164,
-      "grad_norm": 3.3178062438964844,
-      "learning_rate": 4.487012987012987e-05,
-      "loss": 6.5684,
-      "step": 24500
+      "epoch": 12.065354000837871,
+      "grad_norm": 2.7855923175811768,
+      "learning_rate": 2.9899455383326354e-05,
+      "loss": 6.4744,
+      "step": 28800
     },
     {
-      "epoch": 5.236698785085882,
-      "grad_norm": 3.7108089923858643,
-      "learning_rate": 4.476539589442815e-05,
-      "loss": 6.5756,
-      "step": 25000
+      "epoch": 12.10724759111856,
+      "grad_norm": 3.2668962478637695,
+      "learning_rate": 2.982963273285854e-05,
+      "loss": 6.4731,
+      "step": 28900
     },
     {
-      "epoch": 5.341432760787599,
-      "grad_norm": 3.396498918533325,
-      "learning_rate": 4.466087138667784e-05,
-      "loss": 6.5678,
-      "step": 25500
+      "epoch": 12.149141181399246,
+      "grad_norm": 2.850735664367676,
+      "learning_rate": 2.9759810082390725e-05,
+      "loss": 6.4788,
+      "step": 29000
     },
     {
-      "epoch": 5.446166736489317,
-      "grad_norm": 3.7245748043060303,
-      "learning_rate": 4.4556137410976123e-05,
-      "loss": 6.5578,
-      "step": 26000
+      "epoch": 12.191034771679933,
+      "grad_norm": 2.9676952362060547,
+      "learning_rate": 2.9689987431922916e-05,
+      "loss": 6.4525,
+      "step": 29100
     },
     {
-      "epoch": 5.5509007121910345,
-      "grad_norm": 3.6525135040283203,
-      "learning_rate": 4.44514034352744e-05,
-      "loss": 6.5385,
-      "step": 26500
+      "epoch": 12.23292836196062,
+      "grad_norm": 2.604408025741577,
+      "learning_rate": 2.9620164781455107e-05,
+      "loss": 6.4564,
+      "step": 29200
     },
     {
-      "epoch": 5.655634687892753,
-      "grad_norm": 3.4302523136138916,
-      "learning_rate": 4.434666945957269e-05,
-      "loss": 6.5143,
-      "step": 27000
+      "epoch": 12.274821952241307,
+      "grad_norm": 2.974653482437134,
+      "learning_rate": 2.9550342130987295e-05,
+      "loss": 6.463,
+      "step": 29300
     },
     {
-      "epoch": 5.76036866359447,
-      "grad_norm": 3.762871503829956,
-      "learning_rate": 4.4242144951822376e-05,
-      "loss": 6.52,
-      "step": 27500
+      "epoch": 12.316715542521994,
+      "grad_norm": 3.372664213180542,
+      "learning_rate": 2.9480519480519482e-05,
+      "loss": 6.464,
+      "step": 29400
     },
     {
-      "epoch": 5.865102639296188,
-      "grad_norm": 2.8195388317108154,
-      "learning_rate": 4.4137410976120656e-05,
-      "loss": 6.5213,
-      "step": 28000
+      "epoch": 12.35860913280268,
+      "grad_norm": 2.6891355514526367,
+      "learning_rate": 2.941069683005167e-05,
+      "loss": 6.4674,
+      "step": 29500
     },
     {
-      "epoch": 5.969836614997905,
-      "grad_norm": 3.013187885284424,
-      "learning_rate": 4.4032677000418936e-05,
-      "loss": 6.5052,
-      "step": 28500
+      "epoch": 12.400502723083369,
+      "grad_norm": 2.964113473892212,
+      "learning_rate": 2.934087417958386e-05,
+      "loss": 6.4539,
+      "step": 29600
     },
     {
-      "epoch": 6.074570590699623,
-      "grad_norm": 2.9772274494171143,
-      "learning_rate": 4.392794302471722e-05,
-      "loss": 6.502,
-      "step": 29000
+      "epoch": 12.442396313364055,
+      "grad_norm": 2.7328097820281982,
+      "learning_rate": 2.9271051529116045e-05,
+      "loss": 6.4224,
+      "step": 29700
     },
     {
-      "epoch": 6.17930456640134,
-      "grad_norm": 3.2228713035583496,
-      "learning_rate": 4.38232090490155e-05,
-      "loss": 6.4889,
-      "step": 29500
+      "epoch": 12.484289903644742,
+      "grad_norm": 2.6205203533172607,
+      "learning_rate": 2.9201228878648236e-05,
+      "loss": 6.4266,
+      "step": 29800
     },
     {
-      "epoch": 6.284038542103058,
-      "grad_norm": 3.824286937713623,
-      "learning_rate": 4.371847507331379e-05,
-      "loss": 6.4792,
-      "step": 30000
+      "epoch": 12.52618349392543,
+      "grad_norm": 3.681053400039673,
+      "learning_rate": 2.9131406228180424e-05,
+      "loss": 6.4549,
+      "step": 29900
     },
     {
-      "epoch": 6.388772517804776,
-      "grad_norm": 3.100308656692505,
-      "learning_rate": 4.361374109761207e-05,
-      "loss": 6.4816,
-      "step": 30500
+      "epoch": 12.568077084206116,
+      "grad_norm": 2.9732627868652344,
+      "learning_rate": 2.9061583577712608e-05,
+      "loss": 6.4466,
+      "step": 30000
     },
     {
-      "epoch": 6.4935064935064934,
-      "grad_norm": 3.4449245929718018,
-      "learning_rate": 4.350900712191035e-05,
-      "loss": 6.4786,
-      "step": 31000
+      "epoch": 12.609970674486803,
+      "grad_norm": 3.47816801071167,
+      "learning_rate": 2.89917609272448e-05,
+      "loss": 6.4408,
+      "step": 30100
     },
     {
-      "epoch": 6.5982404692082115,
-      "grad_norm": 3.6803085803985596,
-      "learning_rate": 4.3404482614160034e-05,
-      "loss": 6.4778,
-      "step": 31500
+      "epoch": 12.651864264767491,
+      "grad_norm": 2.70326566696167,
+      "learning_rate": 2.892193827677699e-05,
+      "loss": 6.4444,
+      "step": 30200
     },
     {
-      "epoch": 6.702974444909929,
-      "grad_norm": 3.6413722038269043,
-      "learning_rate": 4.329974863845832e-05,
-      "loss": 6.4782,
-      "step": 32000
+      "epoch": 12.693757855048178,
+      "grad_norm": 2.9219532012939453,
+      "learning_rate": 2.8852115626309177e-05,
+      "loss": 6.4183,
+      "step": 30300
     },
     {
-      "epoch": 6.807708420611647,
-      "grad_norm": 3.482905626296997,
-      "learning_rate": 4.31950146627566e-05,
-      "loss": 6.4719,
-      "step": 32500
+      "epoch": 12.735651445328864,
+      "grad_norm": 2.8546571731567383,
+      "learning_rate": 2.878229297584136e-05,
+      "loss": 6.4399,
+      "step": 30400
     },
     {
-      "epoch": 6.912442396313364,
-      "grad_norm": 3.4605376720428467,
-      "learning_rate": 4.3090280687054887e-05,
-      "loss": 6.4458,
-      "step": 33000
+      "epoch": 12.777545035609553,
+      "grad_norm": 2.95047926902771,
+      "learning_rate": 2.8712470325373552e-05,
+      "loss": 6.4396,
+      "step": 30500
     },
     {
-      "epoch": 7.017176372015082,
-      "grad_norm": 3.814375877380371,
-      "learning_rate": 4.2985965647255974e-05,
-      "loss": 6.4539,
-      "step": 33500
+      "epoch": 12.819438625890239,
+      "grad_norm": 3.397934675216675,
+      "learning_rate": 2.8643345901410416e-05,
+      "loss": 6.438,
+      "step": 30600
     },
     {
-      "epoch": 7.121910347716799,
-      "grad_norm": 4.238844871520996,
-      "learning_rate": 4.288123167155425e-05,
-      "loss": 6.4522,
-      "step": 34000
+      "epoch": 12.861332216170926,
+      "grad_norm": 2.625852346420288,
+      "learning_rate": 2.8573523250942607e-05,
+      "loss": 6.4363,
+      "step": 30700
     },
     {
-      "epoch": 7.226644323418517,
-      "grad_norm": 2.8327670097351074,
-      "learning_rate": 4.277649769585253e-05,
-      "loss": 6.4383,
-      "step": 34500
+      "epoch": 12.903225806451612,
+      "grad_norm": 2.5299527645111084,
+      "learning_rate": 2.85037006004748e-05,
+      "loss": 6.3952,
+      "step": 30800
     },
     {
-      "epoch": 7.331378299120234,
-      "grad_norm": 3.1451475620269775,
-      "learning_rate": 4.267176372015082e-05,
-      "loss": 6.4451,
-      "step": 35000
+      "epoch": 12.9451193967323,
+      "grad_norm": 2.6445415019989014,
+      "learning_rate": 2.8433877950006983e-05,
+      "loss": 6.4559,
+      "step": 30900
     },
     {
-      "epoch": 7.436112274821952,
-      "grad_norm": 3.6858575344085693,
-      "learning_rate": 4.2567239212400506e-05,
-      "loss": 6.4313,
-      "step": 35500
+      "epoch": 12.987012987012987,
+      "grad_norm": 2.9675769805908203,
+      "learning_rate": 2.836405529953917e-05,
+      "loss": 6.4447,
+      "step": 31000
     },
     {
-      "epoch": 7.5408462505236695,
-      "grad_norm": 4.258295059204102,
-      "learning_rate": 4.2462505236698786e-05,
-      "loss": 6.4143,
-      "step": 36000
+      "epoch": 13.028906577293673,
+      "grad_norm": 2.607391119003296,
+      "learning_rate": 2.829423264907136e-05,
+      "loss": 6.446,
+      "step": 31100
     },
     {
-      "epoch": 7.645580226225388,
-      "grad_norm": 4.3574676513671875,
-      "learning_rate": 4.235777126099707e-05,
-      "loss": 6.4061,
-      "step": 36500
+      "epoch": 13.070800167574362,
+      "grad_norm": 3.196765661239624,
+      "learning_rate": 2.8224409998603545e-05,
+      "loss": 6.4336,
+      "step": 31200
     },
     {
-      "epoch": 7.750314201927106,
-      "grad_norm": 3.8001816272735596,
-      "learning_rate": 4.225303728529535e-05,
-      "loss": 6.3994,
-      "step": 37000
+      "epoch": 13.112693757855048,
+      "grad_norm": 5.778535842895508,
+      "learning_rate": 2.8154587348135736e-05,
+      "loss": 6.4339,
+      "step": 31300
     },
     {
-      "epoch": 7.855048177628823,
-      "grad_norm": 3.487893581390381,
-      "learning_rate": 4.214851277754504e-05,
-      "loss": 6.4192,
-      "step": 37500
+      "epoch": 13.154587348135735,
+      "grad_norm": 3.0479419231414795,
+      "learning_rate": 2.8084764697667927e-05,
+      "loss": 6.4147,
+      "step": 31400
     },
     {
-      "epoch": 7.95978215333054,
-      "grad_norm": 3.9729723930358887,
-      "learning_rate": 4.204377880184332e-05,
-      "loss": 6.407,
-      "step": 38000
+      "epoch": 13.196480938416423,
+      "grad_norm": 2.6787302494049072,
+      "learning_rate": 2.8014942047200115e-05,
+      "loss": 6.4312,
+      "step": 31500
     },
     {
-      "epoch": 8.064516129032258,
-      "grad_norm": 3.4465062618255615,
-      "learning_rate": 4.1939044826141604e-05,
-      "loss": 6.3867,
-      "step": 38500
+      "epoch": 13.23837452869711,
+      "grad_norm": 2.7929670810699463,
+      "learning_rate": 2.79451193967323e-05,
+      "loss": 6.4224,
+      "step": 31600
     },
     {
-      "epoch": 8.169250104733976,
-      "grad_norm": 3.706404685974121,
-      "learning_rate": 4.1834310850439884e-05,
-      "loss": 6.3877,
-      "step": 39000
+      "epoch": 13.280268118977796,
+      "grad_norm": 2.722101926803589,
+      "learning_rate": 2.787529674626449e-05,
+      "loss": 6.4247,
+      "step": 31700
     },
     {
-      "epoch": 8.273984080435694,
-      "grad_norm": 3.8204259872436523,
-      "learning_rate": 4.172957687473817e-05,
-      "loss": 6.3921,
-      "step": 39500
+      "epoch": 13.322161709258484,
+      "grad_norm": 3.295348644256592,
+      "learning_rate": 2.780547409579668e-05,
+      "loss": 6.4435,
+      "step": 31800
     },
     {
-      "epoch": 8.37871805613741,
-      "grad_norm": 3.4868948459625244,
-      "learning_rate": 4.162505236698786e-05,
-      "loss": 6.3729,
-      "step": 40000
+      "epoch": 13.36405529953917,
+      "grad_norm": 2.5780696868896484,
+      "learning_rate": 2.7735651445328865e-05,
+      "loss": 6.406,
+      "step": 31900
     },
     {
-      "epoch": 8.483452031839128,
-      "grad_norm": 3.5007081031799316,
-      "learning_rate": 4.152031839128614e-05,
-      "loss": 6.3886,
-      "step": 40500
+      "epoch": 13.405948889819857,
+      "grad_norm": 2.955299139022827,
+      "learning_rate": 2.7665828794861053e-05,
+      "loss": 6.4633,
+      "step": 32000
     },
     {
-      "epoch": 8.588186007540846,
-      "grad_norm": 2.937894582748413,
-      "learning_rate": 4.1415584415584417e-05,
-      "loss": 6.3814,
-      "step": 41000
+      "epoch": 13.447842480100544,
+      "grad_norm": 3.8027708530426025,
+      "learning_rate": 2.7596006144393244e-05,
+      "loss": 6.4445,
+      "step": 32100
     },
     {
-      "epoch": 8.692919983242565,
-      "grad_norm": 3.529237985610962,
-      "learning_rate": 4.1310850439882696e-05,
-      "loss": 6.3722,
-      "step": 41500
+      "epoch": 13.489736070381232,
+      "grad_norm": 2.6895995140075684,
+      "learning_rate": 2.7526183493925428e-05,
+      "loss": 6.4015,
+      "step": 32200
     },
     {
-      "epoch": 8.79765395894428,
-      "grad_norm": 3.883575677871704,
-      "learning_rate": 4.120611646418098e-05,
-      "loss": 6.3655,
-      "step": 42000
+      "epoch": 13.531629660661919,
+      "grad_norm": 2.6936516761779785,
+      "learning_rate": 2.745636084345762e-05,
+      "loss": 6.4211,
+      "step": 32300
     },
     {
-      "epoch": 8.902387934645999,
-      "grad_norm": 4.439103603363037,
-      "learning_rate": 4.110159195643067e-05,
-      "loss": 6.3673,
-      "step": 42500
+      "epoch": 13.573523250942605,
+      "grad_norm": 2.948420763015747,
+      "learning_rate": 2.738653819298981e-05,
+      "loss": 6.4042,
+      "step": 32400
     },
     {
-      "epoch": 9.007121910347717,
-      "grad_norm": 4.103298664093018,
-      "learning_rate": 4.099685798072895e-05,
-      "loss": 6.3659,
-      "step": 43000
+      "epoch": 13.615416841223293,
+      "grad_norm": 2.763885974884033,
+      "learning_rate": 2.7316715542521997e-05,
+      "loss": 6.393,
+      "step": 32500
     },
     {
-      "epoch": 9.111855886049435,
-      "grad_norm": 3.491204023361206,
-      "learning_rate": 4.0892124005027235e-05,
-      "loss": 6.3744,
-      "step": 43500
+      "epoch": 13.65731043150398,
+      "grad_norm": 3.1601672172546387,
+      "learning_rate": 2.724759111855886e-05,
+      "loss": 6.4398,
+      "step": 32600
     },
     {
-      "epoch": 9.216589861751151,
-      "grad_norm": 3.441976547241211,
-      "learning_rate": 4.0787390029325515e-05,
-      "loss": 6.3573,
-      "step": 44000
+      "epoch": 13.699204021784666,
+      "grad_norm": 2.4161715507507324,
+      "learning_rate": 2.7177768468091052e-05,
+      "loss": 6.401,
+      "step": 32700
     },
     {
-      "epoch": 9.32132383745287,
-      "grad_norm": 3.58134126663208,
-      "learning_rate": 4.0682656053623795e-05,
-      "loss": 6.3407,
-      "step": 44500
+      "epoch": 13.741097612065355,
+      "grad_norm": 3.0796055793762207,
+      "learning_rate": 2.7107945817623236e-05,
+      "loss": 6.4265,
+      "step": 32800
     },
     {
-      "epoch": 9.426057813154587,
-      "grad_norm": 3.274592638015747,
-      "learning_rate": 4.057813154587348e-05,
-      "loss": 6.3373,
-      "step": 45000
+      "epoch": 13.782991202346041,
+      "grad_norm": 3.6223697662353516,
+      "learning_rate": 2.7038123167155427e-05,
+      "loss": 6.4075,
+      "step": 32900
     },
     {
-      "epoch": 9.530791788856305,
-      "grad_norm": 4.296390533447266,
-      "learning_rate": 4.047339757017177e-05,
-      "loss": 6.3499,
-      "step": 45500
+      "epoch": 13.824884792626728,
+      "grad_norm": 2.6991615295410156,
+      "learning_rate": 2.696830051668762e-05,
+      "loss": 6.3912,
+      "step": 33000
     },
     {
-      "epoch": 9.635525764558023,
-      "grad_norm": 3.5000336170196533,
-      "learning_rate": 4.036866359447005e-05,
-      "loss": 6.3199,
-      "step": 46000
+      "epoch": 13.866778382907416,
+      "grad_norm": 3.1701860427856445,
+      "learning_rate": 2.6898477866219803e-05,
+      "loss": 6.4173,
+      "step": 33100
     },
     {
-      "epoch": 9.74025974025974,
-      "grad_norm": 3.4947054386138916,
-      "learning_rate": 4.0263929618768334e-05,
-      "loss": 6.3474,
-      "step": 46500
+      "epoch": 13.908671973188103,
+      "grad_norm": 2.915432929992676,
+      "learning_rate": 2.682865521575199e-05,
+      "loss": 6.4179,
+      "step": 33200
     },
     {
-      "epoch": 9.844993715961458,
-      "grad_norm": 3.3658857345581055,
-      "learning_rate": 4.0159195643066614e-05,
-      "loss": 6.3296,
-      "step": 47000
+      "epoch": 13.950565563468789,
+      "grad_norm": 3.155080795288086,
+      "learning_rate": 2.675883256528418e-05,
+      "loss": 6.3895,
+      "step": 33300
     },
     {
-      "epoch": 9.949727691663176,
-      "grad_norm": 2.9811642169952393,
-      "learning_rate": 4.0054671135316294e-05,
-      "loss": 6.345,
-      "step": 47500
+      "epoch": 13.992459153749476,
+      "grad_norm": 3.3861114978790283,
+      "learning_rate": 2.6689009914816365e-05,
+      "loss": 6.4279,
+      "step": 33400
     },
     {
-      "epoch": 10.054461667364894,
-      "grad_norm": 4.165875434875488,
-      "learning_rate": 3.994993715961458e-05,
-      "loss": 6.3209,
-      "step": 48000
+      "epoch": 14.034352744030164,
+      "grad_norm": 3.301805019378662,
+      "learning_rate": 2.6619187264348556e-05,
+      "loss": 6.4072,
+      "step": 33500
     },
     {
-      "epoch": 10.15919564306661,
-      "grad_norm": 3.6118202209472656,
-      "learning_rate": 3.9845203183912866e-05,
-      "loss": 6.3175,
-      "step": 48500
+      "epoch": 14.07624633431085,
+      "grad_norm": 3.305147171020508,
+      "learning_rate": 2.6549364613880744e-05,
+      "loss": 6.3949,
+      "step": 33600
     },
     {
-      "epoch": 10.263929618768328,
-      "grad_norm": 3.930669069290161,
-      "learning_rate": 3.9740469208211146e-05,
-      "loss": 6.3231,
-      "step": 49000
+      "epoch": 14.118139924591537,
+      "grad_norm": 2.7602477073669434,
+      "learning_rate": 2.6479541963412935e-05,
+      "loss": 6.4048,
+      "step": 33700
     },
     {
-      "epoch": 10.368663594470046,
-      "grad_norm": 3.1688554286956787,
-      "learning_rate": 3.963594470046083e-05,
-      "loss": 6.309,
-      "step": 49500
+      "epoch": 14.160033514872225,
+      "grad_norm": 2.5257952213287354,
+      "learning_rate": 2.640971931294512e-05,
+      "loss": 6.4033,
+      "step": 33800
     },
     {
-      "epoch": 10.473397570171764,
-      "grad_norm": 3.6746394634246826,
-      "learning_rate": 3.953121072475911e-05,
-      "loss": 6.3077,
-      "step": 50000
+      "epoch": 14.201927105152912,
+      "grad_norm": 2.4649853706359863,
+      "learning_rate": 2.633989666247731e-05,
+      "loss": 6.374,
+      "step": 33900
     },
     {
-      "epoch": 10.57813154587348,
-      "grad_norm": 3.5134785175323486,
-      "learning_rate": 3.942647674905739e-05,
-      "loss": 6.3299,
-      "step": 50500
+      "epoch": 14.243820695433598,
+      "grad_norm": 2.7136335372924805,
+      "learning_rate": 2.6270074012009497e-05,
+      "loss": 6.3993,
+      "step": 34000
     },
     {
-      "epoch": 10.682865521575199,
-      "grad_norm": 3.2903287410736084,
-      "learning_rate": 3.932174277335568e-05,
-      "loss": 6.3178,
-      "step": 51000
+      "epoch": 14.285714285714286,
+      "grad_norm": 2.801712989807129,
+      "learning_rate": 2.6200251361541685e-05,
+      "loss": 6.4059,
+      "step": 34100
     },
     {
-      "epoch": 10.787599497276917,
-      "grad_norm": 3.5344769954681396,
-      "learning_rate": 3.921700879765396e-05,
-      "loss": 6.3139,
-      "step": 51500
+      "epoch": 14.327607875994973,
+      "grad_norm": 2.7054030895233154,
+      "learning_rate": 2.6130428711073873e-05,
+      "loss": 6.431,
+      "step": 34200
     },
     {
-      "epoch": 10.892333472978635,
-      "grad_norm": 3.5710573196411133,
-      "learning_rate": 3.9112274821952245e-05,
-      "loss": 6.306,
-      "step": 52000
+      "epoch": 14.36950146627566,
+      "grad_norm": 2.653932809829712,
+      "learning_rate": 2.6060606060606063e-05,
+      "loss": 6.4035,
+      "step": 34300
     },
     {
-      "epoch": 10.997067448680351,
-      "grad_norm": 3.4193336963653564,
-      "learning_rate": 3.9007540846250524e-05,
-      "loss": 6.3129,
-      "step": 52500
+      "epoch": 14.411395056556348,
+      "grad_norm": 2.5450570583343506,
+      "learning_rate": 2.5990783410138248e-05,
+      "loss": 6.417,
+      "step": 34400
     },
     {
-      "epoch": 11.101801424382069,
-      "grad_norm": 3.683143377304077,
-      "learning_rate": 3.890301633850021e-05,
-      "loss": 6.3084,
-      "step": 53000
+      "epoch": 14.453288646837034,
+      "grad_norm": 2.9578003883361816,
+      "learning_rate": 2.592096075967044e-05,
+      "loss": 6.4087,
+      "step": 34500
     },
     {
-      "epoch": 11.206535400083787,
-      "grad_norm": 3.214221239089966,
-      "learning_rate": 3.879828236279849e-05,
-      "loss": 6.302,
-      "step": 53500
+      "epoch": 14.49518223711772,
+      "grad_norm": 2.9408493041992188,
+      "learning_rate": 2.5851836335707303e-05,
+      "loss": 6.3936,
+      "step": 34600
     },
     {
-      "epoch": 11.311269375785505,
-      "grad_norm": 3.5691747665405273,
-      "learning_rate": 3.869354838709678e-05,
-      "loss": 6.3164,
-      "step": 54000
+      "epoch": 14.537075827398407,
+      "grad_norm": 2.756441116333008,
+      "learning_rate": 2.5782013685239494e-05,
+      "loss": 6.404,
+      "step": 34700
     },
     {
-      "epoch": 11.416003351487223,
-      "grad_norm": 3.2734036445617676,
-      "learning_rate": 3.858881441139506e-05,
-      "loss": 6.2889,
-      "step": 54500
+      "epoch": 14.578969417679096,
+      "grad_norm": 3.685004711151123,
+      "learning_rate": 2.571219103477168e-05,
+      "loss": 6.3932,
+      "step": 34800
     },
     {
-      "epoch": 11.52073732718894,
-      "grad_norm": 4.049854278564453,
-      "learning_rate": 3.848428990364474e-05,
-      "loss": 6.2957,
-      "step": 55000
+      "epoch": 14.620863007959782,
+      "grad_norm": 2.670825719833374,
+      "learning_rate": 2.5642368384303872e-05,
+      "loss": 6.3839,
+      "step": 34900
     },
     {
-      "epoch": 11.625471302890658,
-      "grad_norm": 3.837921380996704,
-      "learning_rate": 3.837955592794303e-05,
-      "loss": 6.2806,
-      "step": 55500
+      "epoch": 14.662756598240469,
+      "grad_norm": 3.0986082553863525,
+      "learning_rate": 2.5572545733836056e-05,
+      "loss": 6.3782,
+      "step": 35000
     },
     {
-      "epoch": 11.730205278592376,
-      "grad_norm": 3.4606828689575195,
-      "learning_rate": 3.827482195224131e-05,
-      "loss": 6.2896,
-      "step": 56000
+      "epoch": 14.704650188521157,
+      "grad_norm": 3.003432273864746,
+      "learning_rate": 2.5502723083368247e-05,
+      "loss": 6.3775,
+      "step": 35100
     },
     {
-      "epoch": 11.834939254294094,
-      "grad_norm": 4.859198093414307,
-      "learning_rate": 3.8170087976539596e-05,
-      "loss": 6.273,
-      "step": 56500
+      "epoch": 14.746543778801843,
+      "grad_norm": 2.752516269683838,
+      "learning_rate": 2.5432900432900435e-05,
+      "loss": 6.3731,
+      "step": 35200
     },
     {
-      "epoch": 11.93967322999581,
-      "grad_norm": 4.689023494720459,
-      "learning_rate": 3.806535400083787e-05,
-      "loss": 6.2776,
-      "step": 57000
+      "epoch": 14.78843736908253,
+      "grad_norm": 2.7697649002075195,
+      "learning_rate": 2.536307778243262e-05,
+      "loss": 6.3701,
+      "step": 35300
     },
     {
-      "epoch": 12.044407205697528,
-      "grad_norm": 4.234752178192139,
-      "learning_rate": 3.7960829493087555e-05,
-      "loss": 6.282,
-      "step": 57500
+      "epoch": 14.830330959363218,
+      "grad_norm": 3.0245521068573,
+      "learning_rate": 2.529325513196481e-05,
+      "loss": 6.3916,
+      "step": 35400
     },
     {
-      "epoch": 12.149141181399246,
-      "grad_norm": 3.950773239135742,
-      "learning_rate": 3.785609551738584e-05,
-      "loss": 6.283,
-      "step": 58000
+      "epoch": 14.872224549643905,
+      "grad_norm": 3.1849350929260254,
+      "learning_rate": 2.5223432481497e-05,
+      "loss": 6.3993,
+      "step": 35500
     },
     {
-      "epoch": 12.253875157100964,
-      "grad_norm": 4.1780548095703125,
-      "learning_rate": 3.775136154168412e-05,
-      "loss": 6.2635,
-      "step": 58500
+      "epoch": 14.914118139924591,
+      "grad_norm": 3.6655123233795166,
+      "learning_rate": 2.5153609831029185e-05,
+      "loss": 6.3791,
+      "step": 35600
     },
     {
-      "epoch": 12.35860913280268,
-      "grad_norm": 3.2049672603607178,
-      "learning_rate": 3.764662756598241e-05,
-      "loss": 6.2858,
-      "step": 59000
+      "epoch": 14.95601173020528,
+      "grad_norm": 3.2252790927886963,
+      "learning_rate": 2.5083787180561376e-05,
+      "loss": 6.3865,
+      "step": 35700
     },
     {
-      "epoch": 12.463343108504398,
-      "grad_norm": 3.863649606704712,
-      "learning_rate": 3.7541893590280694e-05,
-      "loss": 6.2609,
-      "step": 59500
+      "epoch": 14.997905320485966,
+      "grad_norm": 2.8366169929504395,
+      "learning_rate": 2.5013964530093564e-05,
+      "loss": 6.3897,
+      "step": 35800
     },
     {
-      "epoch": 12.568077084206116,
-      "grad_norm": 3.881343364715576,
-      "learning_rate": 3.743715961457897e-05,
-      "loss": 6.2695,
-      "step": 60000
+      "epoch": 15.039798910766653,
+      "grad_norm": 2.757725715637207,
+      "learning_rate": 2.494414187962575e-05,
+      "loss": 6.376,
+      "step": 35900
     },
     {
-      "epoch": 12.672811059907835,
-      "grad_norm": 3.522132635116577,
-      "learning_rate": 3.7332635106828654e-05,
-      "loss": 6.2523,
-      "step": 60500
+      "epoch": 15.081692501047339,
+      "grad_norm": 3.1640422344207764,
+      "learning_rate": 2.4874319229157942e-05,
+      "loss": 6.3796,
+      "step": 36000
     },
     {
-      "epoch": 12.777545035609553,
-      "grad_norm": 4.043595790863037,
-      "learning_rate": 3.722790113112694e-05,
-      "loss": 6.2546,
-      "step": 61000
+      "epoch": 15.123586091328027,
+      "grad_norm": 2.849719285964966,
+      "learning_rate": 2.480449657869013e-05,
+      "loss": 6.3765,
+      "step": 36100
     },
     {
-      "epoch": 12.882279011311269,
-      "grad_norm": 3.4860141277313232,
-      "learning_rate": 3.712316715542522e-05,
-      "loss": 6.2468,
-      "step": 61500
+      "epoch": 15.165479681608714,
+      "grad_norm": 2.7223923206329346,
+      "learning_rate": 2.4734673928222314e-05,
+      "loss": 6.3953,
+      "step": 36200
     },
     {
-      "epoch": 12.987012987012987,
-      "grad_norm": 3.9201574325561523,
-      "learning_rate": 3.7018433179723506e-05,
-      "loss": 6.2615,
-      "step": 62000
+      "epoch": 15.2073732718894,
+      "grad_norm": 3.173750162124634,
+      "learning_rate": 2.4664851277754505e-05,
+      "loss": 6.3724,
+      "step": 36300
     },
     {
-      "epoch": 13.091746962714705,
-      "grad_norm": 3.5582118034362793,
-      "learning_rate": 3.6913699204021786e-05,
-      "loss": 6.2529,
-      "step": 62500
+      "epoch": 15.249266862170089,
+      "grad_norm": 3.054779529571533,
+      "learning_rate": 2.4595028627286692e-05,
+      "loss": 6.3764,
+      "step": 36400
     },
     {
-      "epoch": 13.196480938416423,
-      "grad_norm": 3.1254093647003174,
-      "learning_rate": 3.680917469627147e-05,
-      "loss": 6.2418,
-      "step": 63000
+      "epoch": 15.291160452450775,
+      "grad_norm": 3.277862071990967,
+      "learning_rate": 2.4525205976818883e-05,
+      "loss": 6.3583,
+      "step": 36500
     },
     {
-      "epoch": 13.30121491411814,
-      "grad_norm": 4.058616638183594,
-      "learning_rate": 3.670444072056975e-05,
-      "loss": 6.243,
-      "step": 63500
+      "epoch": 15.333054042731462,
+      "grad_norm": 2.9208297729492188,
+      "learning_rate": 2.4456081552855748e-05,
+      "loss": 6.3878,
+      "step": 36600
     },
     {
-      "epoch": 13.405948889819857,
-      "grad_norm": 3.5146963596343994,
-      "learning_rate": 3.659970674486803e-05,
-      "loss": 6.2595,
-      "step": 64000
+      "epoch": 15.37494763301215,
+      "grad_norm": 2.5356411933898926,
+      "learning_rate": 2.4386258902387935e-05,
+      "loss": 6.3705,
+      "step": 36700
     },
     {
-      "epoch": 13.510682865521575,
-      "grad_norm": 3.804818630218506,
-      "learning_rate": 3.649497276916632e-05,
-      "loss": 6.2438,
-      "step": 64500
+      "epoch": 15.416841223292836,
+      "grad_norm": 2.8953468799591064,
+      "learning_rate": 2.4316436251920126e-05,
+      "loss": 6.3947,
+      "step": 36800
     },
     {
-      "epoch": 13.615416841223293,
-      "grad_norm": 3.591214179992676,
-      "learning_rate": 3.6390238793464605e-05,
-      "loss": 6.2266,
-      "step": 65000
+      "epoch": 15.458734813573523,
+      "grad_norm": 2.9166266918182373,
+      "learning_rate": 2.424661360145231e-05,
+      "loss": 6.3809,
+      "step": 36900
     },
     {
-      "epoch": 13.72015081692501,
-      "grad_norm": 4.973635196685791,
-      "learning_rate": 3.6285504817762885e-05,
-      "loss": 6.2501,
-      "step": 65500
+      "epoch": 15.50062840385421,
+      "grad_norm": 3.4554710388183594,
+      "learning_rate": 2.41767909509845e-05,
+      "loss": 6.3746,
+      "step": 37000
     },
     {
-      "epoch": 13.824884792626728,
-      "grad_norm": 4.189575672149658,
-      "learning_rate": 3.618098031001257e-05,
-      "loss": 6.2341,
-      "step": 66000
+      "epoch": 15.542521994134898,
+      "grad_norm": 3.7208077907562256,
+      "learning_rate": 2.410696830051669e-05,
+      "loss": 6.3758,
+      "step": 37100
     },
     {
-      "epoch": 13.929618768328446,
-      "grad_norm": 4.408186912536621,
-      "learning_rate": 3.607624633431085e-05,
-      "loss": 6.227,
-      "step": 66500
+      "epoch": 15.584415584415584,
+      "grad_norm": 3.3161842823028564,
+      "learning_rate": 2.4037145650048876e-05,
+      "loss": 6.3744,
+      "step": 37200
     },
     {
-      "epoch": 14.034352744030164,
-      "grad_norm": 4.066199779510498,
-      "learning_rate": 3.597151235860913e-05,
-      "loss": 6.2316,
-      "step": 67000
+      "epoch": 15.62630917469627,
+      "grad_norm": 2.4062047004699707,
+      "learning_rate": 2.3967322999581064e-05,
+      "loss": 6.381,
+      "step": 37300
     },
     {
-      "epoch": 14.139086719731882,
-      "grad_norm": 3.8263683319091797,
-      "learning_rate": 3.586677838290742e-05,
-      "loss": 6.2322,
-      "step": 67500
+      "epoch": 15.668202764976959,
+      "grad_norm": 3.1894476413726807,
+      "learning_rate": 2.389750034911325e-05,
+      "loss": 6.3895,
+      "step": 37400
     },
     {
-      "epoch": 14.243820695433598,
-      "grad_norm": 4.787020206451416,
-      "learning_rate": 3.57620444072057e-05,
-      "loss": 6.2099,
-      "step": 68000
+      "epoch": 15.710096355257646,
+      "grad_norm": 2.9203104972839355,
+      "learning_rate": 2.3827677698645442e-05,
+      "loss": 6.363,
+      "step": 37500
     },
     {
-      "epoch": 14.348554671135316,
-      "grad_norm": 3.5196545124053955,
-      "learning_rate": 3.565751989945538e-05,
-      "loss": 6.2425,
-      "step": 68500
+      "epoch": 15.751989945538332,
+      "grad_norm": 3.000694513320923,
+      "learning_rate": 2.375785504817763e-05,
+      "loss": 6.3837,
+      "step": 37600
     },
     {
-      "epoch": 14.453288646837034,
-      "grad_norm": 4.1746439933776855,
-      "learning_rate": 3.555278592375367e-05,
-      "loss": 6.2373,
-      "step": 69000
+      "epoch": 15.79388353581902,
+      "grad_norm": 2.838684558868408,
+      "learning_rate": 2.368803239770982e-05,
+      "loss": 6.3859,
+      "step": 37700
     },
     {
-      "epoch": 14.558022622538752,
-      "grad_norm": 4.07820463180542,
-      "learning_rate": 3.544805194805195e-05,
-      "loss": 6.2099,
-      "step": 69500
+      "epoch": 15.835777126099707,
+      "grad_norm": 2.648862361907959,
+      "learning_rate": 2.3618209747242005e-05,
+      "loss": 6.3411,
+      "step": 37800
     },
     {
-      "epoch": 14.662756598240469,
-      "grad_norm": 3.400038242340088,
-      "learning_rate": 3.534331797235023e-05,
-      "loss": 6.216,
-      "step": 70000
+      "epoch": 15.877670716380393,
+      "grad_norm": 3.5438232421875,
+      "learning_rate": 2.3548387096774193e-05,
+      "loss": 6.3627,
+      "step": 37900
     },
     {
-      "epoch": 14.767490573942187,
-      "grad_norm": 4.578042030334473,
-      "learning_rate": 3.5238793464599916e-05,
-      "loss": 6.2091,
-      "step": 70500
+      "epoch": 15.91956430666108,
+      "grad_norm": 2.8182501792907715,
+      "learning_rate": 2.3478564446306384e-05,
+      "loss": 6.3731,
+      "step": 38000
     },
     {
-      "epoch": 14.872224549643905,
-      "grad_norm": 3.6254208087921143,
-      "learning_rate": 3.51340594888982e-05,
-      "loss": 6.2258,
-      "step": 71000
+      "epoch": 15.961457896941768,
+      "grad_norm": 3.3253772258758545,
+      "learning_rate": 2.340874179583857e-05,
+      "loss": 6.396,
+      "step": 38100
     },
     {
-      "epoch": 14.976958525345623,
-      "grad_norm": 3.496166467666626,
-      "learning_rate": 3.502932551319648e-05,
-      "loss": 6.2138,
-      "step": 71500
+      "epoch": 16.003351487222456,
+      "grad_norm": 3.668926954269409,
+      "learning_rate": 2.3338919145370762e-05,
+      "loss": 6.3437,
+      "step": 38200
     },
     {
-      "epoch": 15.081692501047339,
-      "grad_norm": 3.5367865562438965,
-      "learning_rate": 3.492459153749477e-05,
-      "loss": 6.213,
-      "step": 72000
+      "epoch": 16.045245077503143,
+      "grad_norm": 3.028989315032959,
+      "learning_rate": 2.3269096494902946e-05,
+      "loss": 6.3837,
+      "step": 38300
     },
     {
-      "epoch": 15.186426476749057,
-      "grad_norm": 3.4754440784454346,
-      "learning_rate": 3.481985756179305e-05,
-      "loss": 6.2153,
-      "step": 72500
+      "epoch": 16.08713866778383,
+      "grad_norm": 3.220702648162842,
+      "learning_rate": 2.3199273844435134e-05,
+      "loss": 6.3609,
+      "step": 38400
     },
     {
-      "epoch": 15.291160452450775,
-      "grad_norm": 4.432271957397461,
-      "learning_rate": 3.4715333054042735e-05,
-      "loss": 6.2007,
-      "step": 73000
+      "epoch": 16.129032258064516,
+      "grad_norm": 3.1788036823272705,
+      "learning_rate": 2.3129451193967325e-05,
+      "loss": 6.3723,
+      "step": 38500
     },
     {
-      "epoch": 15.395894428152493,
-      "grad_norm": 3.8427770137786865,
-      "learning_rate": 3.4610599078341014e-05,
-      "loss": 6.2071,
-      "step": 73500
+      "epoch": 16.170925848345203,
+      "grad_norm": 3.351151466369629,
+      "learning_rate": 2.306032677000419e-05,
+      "loss": 6.3731,
+      "step": 38600
     },
     {
-      "epoch": 15.50062840385421,
-      "grad_norm": 3.9617857933044434,
-      "learning_rate": 3.4505865102639294e-05,
-      "loss": 6.2142,
-      "step": 74000
+      "epoch": 16.21281943862589,
+      "grad_norm": 2.933992862701416,
+      "learning_rate": 2.299050411953638e-05,
+      "loss": 6.3654,
+      "step": 38700
     },
     {
-      "epoch": 15.605362379555928,
-      "grad_norm": 3.769693613052368,
-      "learning_rate": 3.440113112693758e-05,
-      "loss": 6.2065,
-      "step": 74500
+      "epoch": 16.25471302890658,
+      "grad_norm": 4.2123589515686035,
+      "learning_rate": 2.2920681469068568e-05,
+      "loss": 6.3364,
+      "step": 38800
     },
     {
-      "epoch": 15.710096355257646,
-      "grad_norm": 3.825507402420044,
-      "learning_rate": 3.429639715123587e-05,
-      "loss": 6.2072,
-      "step": 75000
+      "epoch": 16.296606619187266,
+      "grad_norm": 2.9287397861480713,
+      "learning_rate": 2.2850858818600755e-05,
+      "loss": 6.3643,
+      "step": 38900
     },
     {
-      "epoch": 15.814830330959364,
-      "grad_norm": 3.982872724533081,
-      "learning_rate": 3.4191872643485554e-05,
-      "loss": 6.2003,
-      "step": 75500
+      "epoch": 16.338500209467952,
+      "grad_norm": 2.6518173217773438,
+      "learning_rate": 2.2781036168132943e-05,
+      "loss": 6.3538,
+      "step": 39000
     },
     {
-      "epoch": 15.91956430666108,
-      "grad_norm": 3.9958648681640625,
-      "learning_rate": 3.408713866778383e-05,
-      "loss": 6.1913,
-      "step": 76000
+      "epoch": 16.38039379974864,
+      "grad_norm": 3.490497589111328,
+      "learning_rate": 2.271121351766513e-05,
+      "loss": 6.365,
+      "step": 39100
     },
     {
-      "epoch": 16.0242982823628,
-      "grad_norm": 3.947957754135132,
-      "learning_rate": 3.398240469208211e-05,
-      "loss": 6.2019,
-      "step": 76500
+      "epoch": 16.422287390029325,
+      "grad_norm": 3.090874195098877,
+      "learning_rate": 2.264139086719732e-05,
+      "loss": 6.3513,
+      "step": 39200
     },
     {
-      "epoch": 16.129032258064516,
-      "grad_norm": 3.8135411739349365,
-      "learning_rate": 3.387767071638039e-05,
-      "loss": 6.1944,
-      "step": 77000
+      "epoch": 16.46418098031001,
+      "grad_norm": 2.793083429336548,
+      "learning_rate": 2.257156821672951e-05,
+      "loss": 6.3815,
+      "step": 39300
     },
     {
-      "epoch": 16.233766233766232,
-      "grad_norm": 3.940861701965332,
-      "learning_rate": 3.377293674067868e-05,
-      "loss": 6.1893,
-      "step": 77500
+      "epoch": 16.506074570590698,
+      "grad_norm": 2.656334638595581,
+      "learning_rate": 2.2501745566261696e-05,
+      "loss": 6.3677,
+      "step": 39400
     },
     {
-      "epoch": 16.338500209467952,
-      "grad_norm": 5.24894905090332,
-      "learning_rate": 3.3668412232928366e-05,
-      "loss": 6.1984,
-      "step": 78000
+      "epoch": 16.547968160871388,
+      "grad_norm": 2.950857162475586,
+      "learning_rate": 2.2431922915793884e-05,
+      "loss": 6.3601,
+      "step": 39500
     },
     {
-      "epoch": 16.44323418516967,
-      "grad_norm": 4.470870494842529,
-      "learning_rate": 3.3563678257226645e-05,
-      "loss": 6.1958,
-      "step": 78500
+      "epoch": 16.589861751152075,
+      "grad_norm": 2.948397636413574,
+      "learning_rate": 2.236210026532607e-05,
+      "loss": 6.3633,
+      "step": 39600
     },
     {
-      "epoch": 16.547968160871388,
-      "grad_norm": 3.699892282485962,
-      "learning_rate": 3.345894428152493e-05,
-      "loss": 6.1952,
-      "step": 79000
+      "epoch": 16.63175534143276,
+      "grad_norm": 3.759934902191162,
+      "learning_rate": 2.2292277614858262e-05,
+      "loss": 6.3664,
+      "step": 39700
     },
     {
-      "epoch": 16.652702136573104,
-      "grad_norm": 4.136711120605469,
-      "learning_rate": 3.335421030582321e-05,
-      "loss": 6.1896,
-      "step": 79500
+      "epoch": 16.673648931713448,
+      "grad_norm": 2.6607794761657715,
+      "learning_rate": 2.222245496439045e-05,
+      "loss": 6.3659,
+      "step": 39800
     },
     {
-      "epoch": 16.75743611227482,
-      "grad_norm": 4.904257297515869,
-      "learning_rate": 3.324968579807289e-05,
-      "loss": 6.1715,
-      "step": 80000
+      "epoch": 16.715542521994134,
+      "grad_norm": 3.2569267749786377,
+      "learning_rate": 2.2152632313922638e-05,
+      "loss": 6.3477,
+      "step": 39900
     },
     {
-      "epoch": 16.86217008797654,
-      "grad_norm": 4.219280242919922,
-      "learning_rate": 3.314495182237118e-05,
-      "loss": 6.1829,
-      "step": 80500
+      "epoch": 16.75743611227482,
+      "grad_norm": 3.1701977252960205,
+      "learning_rate": 2.2082809663454825e-05,
+      "loss": 6.3466,
+      "step": 40000
     },
     {
-      "epoch": 16.966904063678257,
-      "grad_norm": 4.426414489746094,
-      "learning_rate": 3.3040217846669464e-05,
-      "loss": 6.1782,
-      "step": 81000
+      "epoch": 16.79932970255551,
+      "grad_norm": 2.8855369091033936,
+      "learning_rate": 2.2012987012987013e-05,
+      "loss": 6.3774,
+      "step": 40100
     },
     {
-      "epoch": 17.071638039379973,
-      "grad_norm": 4.792020797729492,
-      "learning_rate": 3.2935483870967744e-05,
-      "loss": 6.1675,
-      "step": 81500
+      "epoch": 16.841223292836197,
+      "grad_norm": 2.8468215465545654,
+      "learning_rate": 2.1943164362519204e-05,
+      "loss": 6.3388,
+      "step": 40200
     },
     {
-      "epoch": 17.176372015081693,
-      "grad_norm": 3.9796903133392334,
-      "learning_rate": 3.283095936321743e-05,
-      "loss": 6.179,
-      "step": 82000
+      "epoch": 16.883116883116884,
+      "grad_norm": 3.3314404487609863,
+      "learning_rate": 2.187334171205139e-05,
+      "loss": 6.3658,
+      "step": 40300
     },
     {
-      "epoch": 17.28110599078341,
-      "grad_norm": 4.554388046264648,
-      "learning_rate": 3.272622538751571e-05,
-      "loss": 6.1756,
-      "step": 82500
+      "epoch": 16.92501047339757,
+      "grad_norm": 3.023106336593628,
+      "learning_rate": 2.180351906158358e-05,
+      "loss": 6.3443,
+      "step": 40400
     },
     {
-      "epoch": 17.38583996648513,
-      "grad_norm": 4.024316787719727,
-      "learning_rate": 3.262149141181399e-05,
-      "loss": 6.177,
-      "step": 83000
+      "epoch": 16.966904063678257,
+      "grad_norm": 3.2845230102539062,
+      "learning_rate": 2.1733696411115766e-05,
+      "loss": 6.3785,
+      "step": 40500
     },
     {
-      "epoch": 17.490573942186845,
-      "grad_norm": 4.059772968292236,
-      "learning_rate": 3.2516757436112276e-05,
-      "loss": 6.1818,
-      "step": 83500
+      "epoch": 17.008797653958943,
+      "grad_norm": 2.805790424346924,
+      "learning_rate": 2.166457198715263e-05,
+      "loss": 6.3792,
+      "step": 40600
     },
     {
-      "epoch": 17.59530791788856,
-      "grad_norm": 4.296391487121582,
-      "learning_rate": 3.241223292836196e-05,
-      "loss": 6.1866,
-      "step": 84000
+      "epoch": 17.05069124423963,
+      "grad_norm": 2.893737554550171,
+      "learning_rate": 2.159474933668482e-05,
+      "loss": 6.3138,
+      "step": 40700
     },
     {
-      "epoch": 17.70004189359028,
-      "grad_norm": 4.008220672607422,
-      "learning_rate": 3.230749895266024e-05,
-      "loss": 6.172,
-      "step": 84500
+      "epoch": 17.09258483452032,
+      "grad_norm": 3.238863945007324,
+      "learning_rate": 2.1525624912721686e-05,
+      "loss": 6.3686,
+      "step": 40800
     },
     {
-      "epoch": 17.804775869291998,
-      "grad_norm": 4.639082908630371,
-      "learning_rate": 3.220276497695853e-05,
-      "loss": 6.169,
-      "step": 85000
+      "epoch": 17.134478424801006,
+      "grad_norm": 3.403582811355591,
+      "learning_rate": 2.1455802262253877e-05,
+      "loss": 6.3312,
+      "step": 40900
     },
     {
-      "epoch": 17.909509844993718,
-      "grad_norm": 4.635848522186279,
-      "learning_rate": 3.209803100125681e-05,
-      "loss": 6.1721,
-      "step": 85500
+      "epoch": 17.176372015081693,
+      "grad_norm": 2.963287353515625,
+      "learning_rate": 2.1385979611786064e-05,
+      "loss": 6.3515,
+      "step": 41000
     },
     {
-      "epoch": 18.014243820695434,
-      "grad_norm": 4.662270545959473,
-      "learning_rate": 3.199329702555509e-05,
-      "loss": 6.1575,
-      "step": 86000
+      "epoch": 17.21826560536238,
+      "grad_norm": 3.867340087890625,
+      "learning_rate": 2.1316156961318255e-05,
+      "loss": 6.3566,
+      "step": 41100
     },
     {
-      "epoch": 18.11897779639715,
-      "grad_norm": 4.280701637268066,
-      "learning_rate": 3.1888772517804775e-05,
-      "loss": 6.1482,
-      "step": 86500
+      "epoch": 17.260159195643066,
+      "grad_norm": 2.841190814971924,
+      "learning_rate": 2.124633431085044e-05,
+      "loss": 6.3308,
+      "step": 41200
     },
     {
-      "epoch": 18.22371177209887,
-      "grad_norm": 3.8602380752563477,
-      "learning_rate": 3.178403854210306e-05,
-      "loss": 6.1565,
-      "step": 87000
+      "epoch": 17.302052785923753,
+      "grad_norm": 2.872523307800293,
+      "learning_rate": 2.1176511660382627e-05,
+      "loss": 6.3433,
+      "step": 41300
     },
     {
-      "epoch": 18.328445747800586,
-      "grad_norm": 4.634263515472412,
-      "learning_rate": 3.167930456640134e-05,
-      "loss": 6.1587,
-      "step": 87500
+      "epoch": 17.34394637620444,
+      "grad_norm": 3.156465530395508,
+      "learning_rate": 2.1106689009914818e-05,
+      "loss": 6.3779,
+      "step": 41400
     },
     {
-      "epoch": 18.433179723502302,
-      "grad_norm": 4.115392208099365,
-      "learning_rate": 3.157457059069963e-05,
-      "loss": 6.1436,
-      "step": 88000
+      "epoch": 17.38583996648513,
+      "grad_norm": 3.5904667377471924,
+      "learning_rate": 2.1036866359447005e-05,
+      "loss": 6.3402,
+      "step": 41500
     },
     {
-      "epoch": 18.537913699204022,
-      "grad_norm": 3.6665916442871094,
-      "learning_rate": 3.146983661499791e-05,
-      "loss": 6.1442,
-      "step": 88500
+      "epoch": 17.427733556765816,
+      "grad_norm": 3.5753939151763916,
+      "learning_rate": 2.0967043708979196e-05,
+      "loss": 6.3572,
+      "step": 41600
     },
     {
-      "epoch": 18.64264767490574,
-      "grad_norm": 4.444345951080322,
-      "learning_rate": 3.1365312107247594e-05,
-      "loss": 6.1571,
-      "step": 89000
+      "epoch": 17.469627147046502,
+      "grad_norm": 3.129514217376709,
+      "learning_rate": 2.089722105851138e-05,
+      "loss": 6.3302,
+      "step": 41700
     },
     {
-      "epoch": 18.74738165060746,
-      "grad_norm": 3.792792558670044,
-      "learning_rate": 3.1260578131545873e-05,
-      "loss": 6.1535,
-      "step": 89500
+      "epoch": 17.51152073732719,
+      "grad_norm": 2.988732099533081,
+      "learning_rate": 2.0827398408043568e-05,
+      "loss": 6.3807,
+      "step": 41800
     },
     {
-      "epoch": 18.852115626309175,
-      "grad_norm": 3.904019832611084,
-      "learning_rate": 3.115584415584415e-05,
-      "loss": 6.1501,
-      "step": 90000
+      "epoch": 17.553414327607875,
+      "grad_norm": 2.857875108718872,
+      "learning_rate": 2.075757575757576e-05,
+      "loss": 6.3519,
+      "step": 41900
     },
     {
-      "epoch": 18.95684960201089,
-      "grad_norm": 4.531284332275391,
-      "learning_rate": 3.105111018014244e-05,
-      "loss": 6.1592,
-      "step": 90500
+      "epoch": 17.59530791788856,
+      "grad_norm": 4.023842811584473,
+      "learning_rate": 2.0687753107107947e-05,
+      "loss": 6.3467,
+      "step": 42000
     },
     {
-      "epoch": 19.06158357771261,
-      "grad_norm": 3.7976317405700684,
-      "learning_rate": 3.0946376204440726e-05,
-      "loss": 6.1474,
-      "step": 91000
+      "epoch": 17.63720150816925,
+      "grad_norm": 3.049686908721924,
+      "learning_rate": 2.0617930456640137e-05,
+      "loss": 6.3306,
+      "step": 42100
     },
     {
-      "epoch": 19.166317553414327,
-      "grad_norm": 3.8021469116210938,
-      "learning_rate": 3.084185169669041e-05,
-      "loss": 6.1408,
-      "step": 91500
+      "epoch": 17.679095098449938,
+      "grad_norm": 3.3211073875427246,
+      "learning_rate": 2.054810780617232e-05,
+      "loss": 6.3611,
+      "step": 42200
     },
     {
-      "epoch": 19.271051529116047,
-      "grad_norm": 4.194758892059326,
-      "learning_rate": 3.073711772098869e-05,
-      "loss": 6.1476,
-      "step": 92000
+      "epoch": 17.720988688730625,
+      "grad_norm": 3.064138174057007,
+      "learning_rate": 2.047828515570451e-05,
+      "loss": 6.3217,
+      "step": 42300
     },
     {
-      "epoch": 19.375785504817763,
-      "grad_norm": 4.084668159484863,
-      "learning_rate": 3.063238374528697e-05,
-      "loss": 6.1443,
-      "step": 92500
+      "epoch": 17.76288227901131,
+      "grad_norm": 2.7812724113464355,
+      "learning_rate": 2.04084625052367e-05,
+      "loss": 6.3131,
+      "step": 42400
     },
     {
-      "epoch": 19.48051948051948,
-      "grad_norm": 4.383222579956055,
-      "learning_rate": 3.052764976958525e-05,
-      "loss": 6.1422,
-      "step": 93000
+      "epoch": 17.804775869291998,
+      "grad_norm": 2.5516164302825928,
+      "learning_rate": 2.0338639854768888e-05,
+      "loss": 6.3428,
+      "step": 42500
     },
     {
-      "epoch": 19.5852534562212,
-      "grad_norm": 4.250995635986328,
-      "learning_rate": 3.042312526183494e-05,
-      "loss": 6.1375,
-      "step": 93500
+      "epoch": 17.846669459572684,
+      "grad_norm": 2.9599711894989014,
+      "learning_rate": 2.026881720430108e-05,
+      "loss": 6.3545,
+      "step": 42600
     },
     {
-      "epoch": 19.689987431922916,
-      "grad_norm": 4.78529691696167,
-      "learning_rate": 3.0318391286133225e-05,
-      "loss": 6.1368,
-      "step": 94000
+      "epoch": 17.88856304985337,
+      "grad_norm": 2.8674137592315674,
+      "learning_rate": 2.0198994553833263e-05,
+      "loss": 6.3302,
+      "step": 42700
     },
     {
-      "epoch": 19.794721407624632,
-      "grad_norm": 3.4997754096984863,
-      "learning_rate": 3.0213657310431504e-05,
-      "loss": 6.1432,
-      "step": 94500
+      "epoch": 17.93045664013406,
+      "grad_norm": 3.3227078914642334,
+      "learning_rate": 2.012917190336545e-05,
+      "loss": 6.3278,
+      "step": 42800
     },
     {
-      "epoch": 19.89945538332635,
-      "grad_norm": 4.723648548126221,
-      "learning_rate": 3.0108923334729787e-05,
-      "loss": 6.1263,
-      "step": 95000
+      "epoch": 17.972350230414747,
+      "grad_norm": 3.080399751663208,
+      "learning_rate": 2.005934925289764e-05,
+      "loss": 6.3206,
+      "step": 42900
     },
     {
-      "epoch": 20.004189359028068,
-      "grad_norm": 3.930859088897705,
-      "learning_rate": 3.0004398826979474e-05,
-      "loss": 6.1387,
-      "step": 95500
+      "epoch": 18.014243820695434,
+      "grad_norm": 4.004719257354736,
+      "learning_rate": 1.998952660242983e-05,
+      "loss": 6.3407,
+      "step": 43000
     },
     {
-      "epoch": 20.108923334729788,
-      "grad_norm": 4.286599159240723,
-      "learning_rate": 2.9899664851277754e-05,
-      "loss": 6.1187,
-      "step": 96000
+      "epoch": 18.05613741097612,
+      "grad_norm": 2.8186423778533936,
+      "learning_rate": 1.991970395196202e-05,
+      "loss": 6.3136,
+      "step": 43100
     },
     {
-      "epoch": 20.213657310431504,
-      "grad_norm": 3.8475680351257324,
-      "learning_rate": 2.9794930875576037e-05,
-      "loss": 6.1312,
-      "step": 96500
+      "epoch": 18.098031001256807,
+      "grad_norm": 2.81748104095459,
+      "learning_rate": 1.9849881301494204e-05,
+      "loss": 6.3353,
+      "step": 43200
     },
     {
-      "epoch": 20.31839128613322,
-      "grad_norm": 4.844906806945801,
-      "learning_rate": 2.9690196899874323e-05,
-      "loss": 6.1457,
-      "step": 97000
+      "epoch": 18.139924591537493,
+      "grad_norm": 2.9991416931152344,
+      "learning_rate": 1.9780058651026395e-05,
+      "loss": 6.3194,
+      "step": 43300
     },
     {
-      "epoch": 20.42312526183494,
-      "grad_norm": 4.691315174102783,
-      "learning_rate": 2.958567239212401e-05,
-      "loss": 6.1351,
-      "step": 97500
+      "epoch": 18.181818181818183,
+      "grad_norm": 3.4876794815063477,
+      "learning_rate": 1.9710236000558583e-05,
+      "loss": 6.3293,
+      "step": 43400
     },
     {
-      "epoch": 20.527859237536656,
-      "grad_norm": 6.15250825881958,
-      "learning_rate": 2.9480938416422286e-05,
-      "loss": 6.1275,
-      "step": 98000
+      "epoch": 18.22371177209887,
+      "grad_norm": 3.0756711959838867,
+      "learning_rate": 1.964041335009077e-05,
+      "loss": 6.341,
+      "step": 43500
     },
     {
-      "epoch": 20.632593213238373,
-      "grad_norm": 3.8872599601745605,
-      "learning_rate": 2.9376204440720573e-05,
-      "loss": 6.1275,
-      "step": 98500
+      "epoch": 18.265605362379556,
+      "grad_norm": 3.171670436859131,
+      "learning_rate": 1.9570590699622958e-05,
+      "loss": 6.3075,
+      "step": 43600
     },
     {
-      "epoch": 20.737327188940093,
-      "grad_norm": 4.541051864624023,
-      "learning_rate": 2.9271470465018852e-05,
-      "loss": 6.1472,
-      "step": 99000
+      "epoch": 18.307498952660243,
+      "grad_norm": 3.3317439556121826,
+      "learning_rate": 1.9500768049155145e-05,
+      "loss": 6.3436,
+      "step": 43700
     },
     {
-      "epoch": 20.84206116464181,
-      "grad_norm": 4.556408405303955,
-      "learning_rate": 2.9166736489317135e-05,
-      "loss": 6.1369,
-      "step": 99500
+      "epoch": 18.34939254294093,
+      "grad_norm": 2.924349308013916,
+      "learning_rate": 1.9430945398687336e-05,
+      "loss": 6.3217,
+      "step": 43800
     },
     {
-      "epoch": 20.94679514034353,
-      "grad_norm": 4.5567498207092285,
-      "learning_rate": 2.9062211981566822e-05,
-      "loss": 6.1148,
-      "step": 100000
+      "epoch": 18.391286133221616,
+      "grad_norm": 3.247955560684204,
+      "learning_rate": 1.9361122748219524e-05,
+      "loss": 6.3324,
+      "step": 43900
     },
     {
-      "epoch": 21.051529116045245,
-      "grad_norm": 4.647518634796143,
-      "learning_rate": 2.8957478005865102e-05,
-      "loss": 6.1281,
-      "step": 100500
+      "epoch": 18.433179723502302,
+      "grad_norm": 3.340263843536377,
+      "learning_rate": 1.929130009775171e-05,
+      "loss": 6.2993,
+      "step": 44000
     },
     {
-      "epoch": 21.15626309174696,
-      "grad_norm": 4.372421741485596,
-      "learning_rate": 2.8852744030163388e-05,
-      "loss": 6.1226,
-      "step": 101000
+      "epoch": 18.475073313782993,
+      "grad_norm": 2.973019599914551,
+      "learning_rate": 1.92214774472839e-05,
+      "loss": 6.3292,
+      "step": 44100
     },
     {
-      "epoch": 21.26099706744868,
-      "grad_norm": 4.270533084869385,
-      "learning_rate": 2.874801005446167e-05,
-      "loss": 6.123,
-      "step": 101500
+      "epoch": 18.51696690406368,
+      "grad_norm": 3.5055582523345947,
+      "learning_rate": 1.9151654796816086e-05,
+      "loss": 6.3175,
+      "step": 44200
     },
     {
-      "epoch": 21.365731043150397,
-      "grad_norm": 3.5596370697021484,
-      "learning_rate": 2.8643485546711358e-05,
-      "loss": 6.1402,
-      "step": 102000
+      "epoch": 18.558860494344366,
+      "grad_norm": 2.9543776512145996,
+      "learning_rate": 1.9081832146348277e-05,
+      "loss": 6.3206,
+      "step": 44300
     },
     {
-      "epoch": 21.470465018852117,
-      "grad_norm": 5.230384826660156,
-      "learning_rate": 2.8538751571009638e-05,
-      "loss": 6.1199,
-      "step": 102500
+      "epoch": 18.600754084625052,
+      "grad_norm": 2.790940284729004,
+      "learning_rate": 1.9012009495880465e-05,
+      "loss": 6.3383,
+      "step": 44400
     },
     {
-      "epoch": 21.575198994553833,
-      "grad_norm": 3.9881417751312256,
-      "learning_rate": 2.843401759530792e-05,
-      "loss": 6.1244,
-      "step": 103000
+      "epoch": 18.64264767490574,
+      "grad_norm": 3.419908285140991,
+      "learning_rate": 1.8942186845412653e-05,
+      "loss": 6.3329,
+      "step": 44500
     },
     {
-      "epoch": 21.67993297025555,
-      "grad_norm": 4.617568016052246,
-      "learning_rate": 2.83292836196062e-05,
-      "loss": 6.1154,
-      "step": 103500
+      "epoch": 18.684541265186425,
+      "grad_norm": 3.3396215438842773,
+      "learning_rate": 1.887236419494484e-05,
+      "loss": 6.312,
+      "step": 44600
     },
     {
-      "epoch": 21.78466694595727,
-      "grad_norm": 4.641009330749512,
-      "learning_rate": 2.8224759111855887e-05,
-      "loss": 6.113,
-      "step": 104000
+      "epoch": 18.726434855467115,
+      "grad_norm": 2.6713643074035645,
+      "learning_rate": 1.8802541544477028e-05,
+      "loss": 6.315,
+      "step": 44700
     },
     {
-      "epoch": 21.889400921658986,
-      "grad_norm": 4.005772113800049,
-      "learning_rate": 2.812002513615417e-05,
-      "loss": 6.1163,
-      "step": 104500
+      "epoch": 18.7683284457478,
+      "grad_norm": 3.2764880657196045,
+      "learning_rate": 1.8733417120513895e-05,
+      "loss": 6.3311,
+      "step": 44800
     },
     {
-      "epoch": 21.994134897360702,
-      "grad_norm": 4.2611799240112305,
-      "learning_rate": 2.801529116045245e-05,
-      "loss": 6.1023,
-      "step": 105000
+      "epoch": 18.810222036028488,
+      "grad_norm": 3.602581739425659,
+      "learning_rate": 1.8663594470046083e-05,
+      "loss": 6.327,
+      "step": 44900
     },
     {
-      "epoch": 22.098868873062422,
-      "grad_norm": 4.568357467651367,
-      "learning_rate": 2.7910557184750736e-05,
-      "loss": 6.1169,
-      "step": 105500
+      "epoch": 18.852115626309175,
+      "grad_norm": 3.052971124649048,
+      "learning_rate": 1.8593771819578274e-05,
+      "loss": 6.2911,
+      "step": 45000
     },
     {
-      "epoch": 22.203602848764138,
-      "grad_norm": 4.323103427886963,
-      "learning_rate": 2.780603267700042e-05,
-      "loss": 6.1226,
-      "step": 106000
+      "epoch": 18.89400921658986,
+      "grad_norm": 3.0912699699401855,
+      "learning_rate": 1.852394916911046e-05,
+      "loss": 6.3057,
+      "step": 45100
     },
     {
-      "epoch": 22.308336824465858,
-      "grad_norm": 4.507444381713867,
-      "learning_rate": 2.77012987012987e-05,
-      "loss": 6.1052,
-      "step": 106500
+      "epoch": 18.935902806870548,
+      "grad_norm": 2.631545305252075,
+      "learning_rate": 1.845412651864265e-05,
+      "loss": 6.3381,
+      "step": 45200
     },
     {
-      "epoch": 22.413070800167574,
-      "grad_norm": 4.301244735717773,
-      "learning_rate": 2.7596564725596985e-05,
-      "loss": 6.0944,
-      "step": 107000
+      "epoch": 18.977796397151234,
+      "grad_norm": 3.8213324546813965,
+      "learning_rate": 1.8384303868174836e-05,
+      "loss": 6.3123,
+      "step": 45300
     },
     {
-      "epoch": 22.51780477586929,
-      "grad_norm": 4.984853267669678,
-      "learning_rate": 2.749183074989527e-05,
-      "loss": 6.1125,
-      "step": 107500
+      "epoch": 19.019689987431924,
+      "grad_norm": 3.3717353343963623,
+      "learning_rate": 1.8314481217707024e-05,
+      "loss": 6.3194,
+      "step": 45400
     },
     {
-      "epoch": 22.62253875157101,
-      "grad_norm": 4.682931423187256,
-      "learning_rate": 2.7387096774193548e-05,
-      "loss": 6.1158,
-      "step": 108000
+      "epoch": 19.06158357771261,
+      "grad_norm": 2.831409215927124,
+      "learning_rate": 1.8244658567239215e-05,
+      "loss": 6.3383,
+      "step": 45500
     },
     {
-      "epoch": 22.727272727272727,
-      "grad_norm": 4.494015693664551,
-      "learning_rate": 2.7282572266443235e-05,
-      "loss": 6.1035,
-      "step": 108500
+      "epoch": 19.103477167993297,
+      "grad_norm": 2.915093183517456,
+      "learning_rate": 1.8174835916771403e-05,
+      "loss": 6.3208,
+      "step": 45600
     },
     {
-      "epoch": 22.832006702974446,
-      "grad_norm": 3.880779981613159,
-      "learning_rate": 2.717783829074152e-05,
-      "loss": 6.1084,
-      "step": 109000
+      "epoch": 19.145370758273984,
+      "grad_norm": 3.1236917972564697,
+      "learning_rate": 1.810501326630359e-05,
+      "loss": 6.3089,
+      "step": 45700
     },
     {
-      "epoch": 22.936740678676163,
-      "grad_norm": 4.154653072357178,
-      "learning_rate": 2.7073104315039798e-05,
-      "loss": 6.0945,
-      "step": 109500
+      "epoch": 19.18726434855467,
+      "grad_norm": 3.2876298427581787,
+      "learning_rate": 1.8035190615835778e-05,
+      "loss": 6.2975,
+      "step": 45800
     },
     {
-      "epoch": 23.04147465437788,
-      "grad_norm": 5.443271160125732,
-      "learning_rate": 2.6968370339338084e-05,
-      "loss": 6.1016,
-      "step": 110000
+      "epoch": 19.229157938835357,
+      "grad_norm": 2.6437103748321533,
+      "learning_rate": 1.7965367965367965e-05,
+      "loss": 6.3341,
+      "step": 45900
     },
     {
-      "epoch": 23.1462086300796,
-      "grad_norm": 4.298133373260498,
-      "learning_rate": 2.686384583158777e-05,
-      "loss": 6.103,
-      "step": 110500
+      "epoch": 19.271051529116047,
+      "grad_norm": 2.9252028465270996,
+      "learning_rate": 1.7895545314900156e-05,
+      "loss": 6.3404,
+      "step": 46000
     },
     {
-      "epoch": 23.250942605781315,
-      "grad_norm": 4.379884243011475,
-      "learning_rate": 2.675911185588605e-05,
-      "loss": 6.0814,
-      "step": 111000
+      "epoch": 19.312945119396733,
+      "grad_norm": 3.4250340461730957,
+      "learning_rate": 1.7825722664432344e-05,
+      "loss": 6.3072,
+      "step": 46100
     },
     {
-      "epoch": 23.35567658148303,
-      "grad_norm": 6.175398349761963,
-      "learning_rate": 2.6654377880184333e-05,
-      "loss": 6.1088,
-      "step": 111500
+      "epoch": 19.35483870967742,
+      "grad_norm": 3.1287946701049805,
+      "learning_rate": 1.775590001396453e-05,
+      "loss": 6.3022,
+      "step": 46200
     },
     {
-      "epoch": 23.46041055718475,
-      "grad_norm": 4.121715068817139,
-      "learning_rate": 2.6549643904482613e-05,
-      "loss": 6.0966,
-      "step": 112000
+      "epoch": 19.396732299958106,
+      "grad_norm": 3.4577419757843018,
+      "learning_rate": 1.76867755900014e-05,
+      "loss": 6.2938,
+      "step": 46300
     },
     {
-      "epoch": 23.565144532886467,
-      "grad_norm": 5.040287494659424,
-      "learning_rate": 2.6444909928780896e-05,
-      "loss": 6.098,
-      "step": 112500
+      "epoch": 19.438625890238793,
+      "grad_norm": 3.7131240367889404,
+      "learning_rate": 1.7616952939533586e-05,
+      "loss": 6.3088,
+      "step": 46400
     },
     {
-      "epoch": 23.669878508588187,
-      "grad_norm": 4.766879081726074,
-      "learning_rate": 2.6340175953079182e-05,
-      "loss": 6.0957,
-      "step": 113000
+      "epoch": 19.48051948051948,
+      "grad_norm": 3.6799802780151367,
+      "learning_rate": 1.7547130289065774e-05,
+      "loss": 6.3326,
+      "step": 46500
     },
     {
-      "epoch": 23.774612484289904,
-      "grad_norm": 5.87930965423584,
-      "learning_rate": 2.623565144532887e-05,
-      "loss": 6.1089,
-      "step": 113500
+      "epoch": 19.522413070800166,
+      "grad_norm": 2.834351062774658,
+      "learning_rate": 1.747730763859796e-05,
+      "loss": 6.2952,
+      "step": 46600
     },
     {
-      "epoch": 23.87934645999162,
-      "grad_norm": 5.318653583526611,
-      "learning_rate": 2.613091746962715e-05,
-      "loss": 6.0761,
-      "step": 114000
+      "epoch": 19.564306661080856,
+      "grad_norm": 3.0629451274871826,
+      "learning_rate": 1.7407484988130152e-05,
+      "loss": 6.3185,
+      "step": 46700
     },
     {
-      "epoch": 23.98408043569334,
-      "grad_norm": 4.465319633483887,
-      "learning_rate": 2.6026183493925432e-05,
-      "loss": 6.0826,
-      "step": 114500
+      "epoch": 19.606200251361543,
+      "grad_norm": 3.4801712036132812,
+      "learning_rate": 1.733766233766234e-05,
+      "loss": 6.3003,
+      "step": 46800
     },
     {
-      "epoch": 24.088814411395056,
-      "grad_norm": 4.640571594238281,
-      "learning_rate": 2.592144951822371e-05,
-      "loss": 6.0805,
-      "step": 115000
+      "epoch": 19.64809384164223,
+      "grad_norm": 2.8250389099121094,
+      "learning_rate": 1.7267839687194524e-05,
+      "loss": 6.3033,
+      "step": 46900
     },
     {
-      "epoch": 24.193548387096776,
-      "grad_norm": 4.252554416656494,
-      "learning_rate": 2.5816925010473398e-05,
-      "loss": 6.0701,
-      "step": 115500
+      "epoch": 19.689987431922916,
+      "grad_norm": 3.5964672565460205,
+      "learning_rate": 1.7198017036726715e-05,
+      "loss": 6.293,
+      "step": 47000
     },
     {
-      "epoch": 24.298282362798492,
-      "grad_norm": 4.704644203186035,
-      "learning_rate": 2.571219103477168e-05,
-      "loss": 6.0936,
-      "step": 116000
+      "epoch": 19.731881022203602,
+      "grad_norm": 2.7947146892547607,
+      "learning_rate": 1.7128194386258903e-05,
+      "loss": 6.2884,
+      "step": 47100
     },
     {
-      "epoch": 24.40301633850021,
-      "grad_norm": 4.601324558258057,
-      "learning_rate": 2.560745705906996e-05,
-      "loss": 6.0758,
-      "step": 116500
+      "epoch": 19.77377461248429,
+      "grad_norm": 3.0473551750183105,
+      "learning_rate": 1.7058371735791094e-05,
+      "loss": 6.312,
+      "step": 47200
     },
     {
-      "epoch": 24.507750314201928,
-      "grad_norm": 4.380444526672363,
-      "learning_rate": 2.5502723083368247e-05,
-      "loss": 6.0871,
-      "step": 117000
+      "epoch": 19.81566820276498,
+      "grad_norm": 3.1810736656188965,
+      "learning_rate": 1.698854908532328e-05,
+      "loss": 6.3102,
+      "step": 47300
     },
     {
-      "epoch": 24.612484289903644,
-      "grad_norm": 4.119806289672852,
-      "learning_rate": 2.5397989107666527e-05,
-      "loss": 6.102,
-      "step": 117500
+      "epoch": 19.857561793045665,
+      "grad_norm": 3.0046746730804443,
+      "learning_rate": 1.6918726434855465e-05,
+      "loss": 6.3115,
+      "step": 47400
     },
     {
-      "epoch": 24.71721826560536,
-      "grad_norm": 3.9712698459625244,
-      "learning_rate": 2.5293464599916217e-05,
-      "loss": 6.0999,
-      "step": 118000
+      "epoch": 19.89945538332635,
+      "grad_norm": 2.6985220909118652,
+      "learning_rate": 1.6848903784387656e-05,
+      "loss": 6.3132,
+      "step": 47500
     },
     {
-      "epoch": 24.82195224130708,
-      "grad_norm": 5.146612167358398,
-      "learning_rate": 2.5188730624214497e-05,
-      "loss": 6.0947,
-      "step": 118500
+      "epoch": 19.941348973607038,
+      "grad_norm": 2.958906650543213,
+      "learning_rate": 1.6779081133919844e-05,
+      "loss": 6.3024,
+      "step": 47600
     },
     {
-      "epoch": 24.926686217008797,
-      "grad_norm": 4.406741142272949,
-      "learning_rate": 2.508399664851278e-05,
-      "loss": 6.0829,
-      "step": 119000
+      "epoch": 19.983242563887725,
+      "grad_norm": 3.5484089851379395,
+      "learning_rate": 1.6709258483452035e-05,
+      "loss": 6.2989,
+      "step": 47700
     },
     {
-      "epoch": 25.031420192710517,
-      "grad_norm": 5.4739766120910645,
-      "learning_rate": 2.497926267281106e-05,
-      "loss": 6.0598,
-      "step": 119500
+      "epoch": 20.02513615416841,
+      "grad_norm": 4.328272342681885,
+      "learning_rate": 1.6639435832984222e-05,
+      "loss": 6.3162,
+      "step": 47800
     },
     {
-      "epoch": 25.136154168412233,
-      "grad_norm": 4.6231794357299805,
-      "learning_rate": 2.4874738165060746e-05,
-      "loss": 6.072,
-      "step": 120000
+      "epoch": 20.067029744449098,
+      "grad_norm": 3.0396926403045654,
+      "learning_rate": 1.6569613182516407e-05,
+      "loss": 6.3004,
+      "step": 47900
     },
     {
-      "epoch": 25.24088814411395,
-      "grad_norm": 4.47750186920166,
-      "learning_rate": 2.477000418935903e-05,
-      "loss": 6.0664,
-      "step": 120500
+      "epoch": 20.108923334729788,
+      "grad_norm": 3.328972339630127,
+      "learning_rate": 1.6499790532048598e-05,
+      "loss": 6.2855,
+      "step": 48000
     },
     {
-      "epoch": 25.34562211981567,
-      "grad_norm": 5.023014068603516,
-      "learning_rate": 2.4665270213657312e-05,
-      "loss": 6.0894,
-      "step": 121000
+      "epoch": 20.150816925010474,
+      "grad_norm": 3.301114320755005,
+      "learning_rate": 1.6429967881580785e-05,
+      "loss": 6.2874,
+      "step": 48100
     },
     {
-      "epoch": 25.450356095517385,
-      "grad_norm": 5.687644004821777,
-      "learning_rate": 2.4560536237955595e-05,
-      "loss": 6.0936,
-      "step": 121500
+      "epoch": 20.19271051529116,
+      "grad_norm": 3.297041177749634,
+      "learning_rate": 1.6360145231112976e-05,
+      "loss": 6.3089,
+      "step": 48200
     },
     {
-      "epoch": 25.555090071219105,
-      "grad_norm": 4.534958362579346,
-      "learning_rate": 2.4455802262253878e-05,
-      "loss": 6.0794,
-      "step": 122000
+      "epoch": 20.234604105571847,
+      "grad_norm": 2.9122605323791504,
+      "learning_rate": 1.6290322580645164e-05,
+      "loss": 6.3157,
+      "step": 48300
     },
     {
-      "epoch": 25.65982404692082,
-      "grad_norm": 5.563751697540283,
-      "learning_rate": 2.435127775450356e-05,
-      "loss": 6.081,
-      "step": 122500
+      "epoch": 20.276497695852534,
+      "grad_norm": 2.8182084560394287,
+      "learning_rate": 1.6220499930177348e-05,
+      "loss": 6.3118,
+      "step": 48400
     },
     {
-      "epoch": 25.764558022622538,
-      "grad_norm": 4.613626956939697,
-      "learning_rate": 2.4246543778801845e-05,
-      "loss": 6.0982,
-      "step": 123000
+      "epoch": 20.31839128613322,
+      "grad_norm": 3.8560192584991455,
+      "learning_rate": 1.615067727970954e-05,
+      "loss": 6.2858,
+      "step": 48500
     },
     {
-      "epoch": 25.869291998324258,
-      "grad_norm": 4.645818710327148,
-      "learning_rate": 2.4141809803100128e-05,
-      "loss": 6.0665,
-      "step": 123500
+      "epoch": 20.36028487641391,
+      "grad_norm": 2.457240581512451,
+      "learning_rate": 1.6080854629241726e-05,
+      "loss": 6.3077,
+      "step": 48600
     },
     {
-      "epoch": 25.974025974025974,
-      "grad_norm": 4.9156928062438965,
-      "learning_rate": 2.4037075827398407e-05,
-      "loss": 6.0674,
-      "step": 124000
+      "epoch": 20.402178466694597,
+      "grad_norm": 3.5376362800598145,
+      "learning_rate": 1.6011031978773917e-05,
+      "loss": 6.2892,
+      "step": 48700
     },
     {
-      "epoch": 26.07875994972769,
-      "grad_norm": 4.7342305183410645,
-      "learning_rate": 2.3932551319648094e-05,
-      "loss": 6.0741,
-      "step": 124500
+      "epoch": 20.444072056975283,
+      "grad_norm": 3.3489222526550293,
+      "learning_rate": 1.59412093283061e-05,
+      "loss": 6.2973,
+      "step": 48800
     },
     {
-      "epoch": 26.18349392542941,
-      "grad_norm": 4.607081413269043,
-      "learning_rate": 2.3827817343946377e-05,
-      "loss": 6.0626,
-      "step": 125000
+      "epoch": 20.48596564725597,
+      "grad_norm": 3.600166082382202,
+      "learning_rate": 1.587138667783829e-05,
+      "loss": 6.31,
+      "step": 48900
     },
     {
-      "epoch": 26.288227901131126,
-      "grad_norm": 4.820442199707031,
-      "learning_rate": 2.372308336824466e-05,
-      "loss": 6.0936,
-      "step": 125500
+      "epoch": 20.527859237536656,
+      "grad_norm": 3.255598783493042,
+      "learning_rate": 1.580156402737048e-05,
+      "loss": 6.2389,
+      "step": 49000
     },
     {
-      "epoch": 26.392961876832846,
-      "grad_norm": 4.549975395202637,
-      "learning_rate": 2.3618349392542943e-05,
-      "loss": 6.0804,
-      "step": 126000
+      "epoch": 20.569752827817343,
+      "grad_norm": 3.166994094848633,
+      "learning_rate": 1.5731741376902668e-05,
+      "loss": 6.303,
+      "step": 49100
     },
     {
-      "epoch": 26.497695852534562,
-      "grad_norm": 4.722150802612305,
-      "learning_rate": 2.351382488479263e-05,
-      "loss": 6.0555,
-      "step": 126500
+      "epoch": 20.61164641809803,
+      "grad_norm": 3.615269184112549,
+      "learning_rate": 1.566191872643486e-05,
+      "loss": 6.281,
+      "step": 49200
     },
     {
-      "epoch": 26.60242982823628,
-      "grad_norm": 4.2948408126831055,
-      "learning_rate": 2.340909090909091e-05,
-      "loss": 6.0626,
-      "step": 127000
+      "epoch": 20.65354000837872,
+      "grad_norm": 3.1495063304901123,
+      "learning_rate": 1.5592096075967043e-05,
+      "loss": 6.2666,
+      "step": 49300
     },
     {
-      "epoch": 26.707163803938,
-      "grad_norm": 4.246878623962402,
-      "learning_rate": 2.3304356933389193e-05,
-      "loss": 6.0577,
-      "step": 127500
+      "epoch": 20.695433598659406,
+      "grad_norm": 2.9170730113983154,
+      "learning_rate": 1.552227342549923e-05,
+      "loss": 6.2738,
+      "step": 49400
     },
     {
-      "epoch": 26.811897779639715,
-      "grad_norm": 4.165809154510498,
-      "learning_rate": 2.3199622957687476e-05,
-      "loss": 6.0608,
-      "step": 128000
+      "epoch": 20.737327188940093,
+      "grad_norm": 3.0922224521636963,
+      "learning_rate": 1.545245077503142e-05,
+      "loss": 6.2805,
+      "step": 49500
     },
     {
-      "epoch": 26.916631755341434,
-      "grad_norm": 4.3159894943237305,
-      "learning_rate": 2.309488898198576e-05,
-      "loss": 6.0806,
-      "step": 128500
+      "epoch": 20.77922077922078,
+      "grad_norm": 3.088012933731079,
+      "learning_rate": 1.538262812456361e-05,
+      "loss": 6.2906,
+      "step": 49600
     },
     {
-      "epoch": 27.02136573104315,
-      "grad_norm": 4.15300989151001,
-      "learning_rate": 2.2990364474235442e-05,
-      "loss": 6.0654,
-      "step": 129000
+      "epoch": 20.821114369501466,
+      "grad_norm": 2.939486503601074,
+      "learning_rate": 1.53128054740958e-05,
+      "loss": 6.2636,
+      "step": 49700
     },
     {
-      "epoch": 27.126099706744867,
-      "grad_norm": 4.730154991149902,
-      "learning_rate": 2.2885630498533725e-05,
-      "loss": 6.0567,
-      "step": 129500
+      "epoch": 20.863007959782152,
+      "grad_norm": 3.597949743270874,
+      "learning_rate": 1.5242982823627986e-05,
+      "loss": 6.2745,
+      "step": 49800
     },
     {
-      "epoch": 27.230833682446587,
-      "grad_norm": 4.300974369049072,
-      "learning_rate": 2.2780896522832008e-05,
-      "loss": 6.0449,
-      "step": 130000
+      "epoch": 20.90490155006284,
+      "grad_norm": 3.4760777950286865,
+      "learning_rate": 1.5173160173160175e-05,
+      "loss": 6.2702,
+      "step": 49900
     },
     {
-      "epoch": 27.335567658148303,
-      "grad_norm": 4.148283958435059,
-      "learning_rate": 2.2676162547130288e-05,
-      "loss": 6.0491,
-      "step": 130500
+      "epoch": 20.94679514034353,
+      "grad_norm": 3.04856014251709,
+      "learning_rate": 1.5103337522692362e-05,
+      "loss": 6.2841,
+      "step": 50000
     },
     {
-      "epoch": 27.44030163385002,
-      "grad_norm": 4.9924421310424805,
-      "learning_rate": 2.2571638039379974e-05,
-      "loss": 6.0491,
-      "step": 131000
+      "epoch": 20.988688730624215,
+      "grad_norm": 2.849895477294922,
+      "learning_rate": 1.503351487222455e-05,
+      "loss": 6.2814,
+      "step": 50100
     },
     {
-      "epoch": 27.54503560955174,
-      "grad_norm": 5.713706016540527,
-      "learning_rate": 2.246690406367826e-05,
-      "loss": 6.0396,
-      "step": 131500
+      "epoch": 21.0305823209049,
+      "grad_norm": 3.1246280670166016,
+      "learning_rate": 1.496369222175674e-05,
+      "loss": 6.2754,
+      "step": 50200
     },
     {
-      "epoch": 27.649769585253456,
-      "grad_norm": 5.007369518280029,
-      "learning_rate": 2.236217008797654e-05,
-      "loss": 6.0378,
-      "step": 132000
+      "epoch": 21.072475911185588,
+      "grad_norm": 3.303846836090088,
+      "learning_rate": 1.4894567797793605e-05,
+      "loss": 6.2661,
+      "step": 50300
     },
     {
-      "epoch": 27.754503560955175,
-      "grad_norm": 4.500640392303467,
-      "learning_rate": 2.2257436112274823e-05,
-      "loss": 6.0313,
-      "step": 132500
+      "epoch": 21.114369501466275,
+      "grad_norm": 3.5818755626678467,
+      "learning_rate": 1.4824745147325794e-05,
+      "loss": 6.2804,
+      "step": 50400
     },
     {
-      "epoch": 27.85923753665689,
-      "grad_norm": 4.709275722503662,
-      "learning_rate": 2.2152702136573107e-05,
-      "loss": 6.0278,
-      "step": 133000
+      "epoch": 21.15626309174696,
+      "grad_norm": 3.0695786476135254,
+      "learning_rate": 1.4754922496857982e-05,
+      "loss": 6.284,
+      "step": 50500
     },
     {
-      "epoch": 27.963971512358608,
-      "grad_norm": 4.891386032104492,
-      "learning_rate": 2.204817762882279e-05,
-      "loss": 6.0267,
-      "step": 133500
+      "epoch": 21.19815668202765,
+      "grad_norm": 3.6067614555358887,
+      "learning_rate": 1.4685099846390168e-05,
+      "loss": 6.2863,
+      "step": 50600
     },
     {
-      "epoch": 28.068705488060328,
-      "grad_norm": 4.82666540145874,
-      "learning_rate": 2.1943443653121073e-05,
-      "loss": 5.986,
-      "step": 134000
+      "epoch": 21.240050272308338,
+      "grad_norm": 3.2230417728424072,
+      "learning_rate": 1.4615277195922359e-05,
+      "loss": 6.287,
+      "step": 50700
     },
     {
-      "epoch": 28.173439463762044,
-      "grad_norm": 4.489607810974121,
-      "learning_rate": 2.1838709677419356e-05,
-      "loss": 6.0209,
-      "step": 134500
+      "epoch": 21.281943862589024,
+      "grad_norm": 3.059466600418091,
+      "learning_rate": 1.4545454545454545e-05,
+      "loss": 6.2442,
+      "step": 50800
     },
     {
-      "epoch": 28.278173439463764,
-      "grad_norm": 4.719301700592041,
-      "learning_rate": 2.173397570171764e-05,
-      "loss": 5.9998,
-      "step": 135000
+      "epoch": 21.32383745286971,
+      "grad_norm": 3.7770040035247803,
+      "learning_rate": 1.4475631894986736e-05,
+      "loss": 6.2612,
+      "step": 50900
     },
     {
-      "epoch": 28.38290741516548,
-      "grad_norm": 5.639565467834473,
-      "learning_rate": 2.1629451193967322e-05,
-      "loss": 5.9881,
-      "step": 135500
+      "epoch": 21.365731043150397,
+      "grad_norm": 3.3269879817962646,
+      "learning_rate": 1.4405809244518923e-05,
+      "loss": 6.2985,
+      "step": 51000
     },
     {
-      "epoch": 28.487641390867196,
-      "grad_norm": 4.745512008666992,
-      "learning_rate": 2.1524717218265605e-05,
-      "loss": 6.0002,
-      "step": 136000
+      "epoch": 21.407624633431084,
+      "grad_norm": 2.649940252304077,
+      "learning_rate": 1.4335986594051109e-05,
+      "loss": 6.2343,
+      "step": 51100
     },
     {
-      "epoch": 28.592375366568916,
-      "grad_norm": 5.661725997924805,
-      "learning_rate": 2.141998324256389e-05,
-      "loss": 5.9967,
-      "step": 136500
+      "epoch": 21.44951822371177,
+      "grad_norm": 3.4042983055114746,
+      "learning_rate": 1.42661639435833e-05,
+      "loss": 6.2701,
+      "step": 51200
     },
     {
-      "epoch": 28.697109342270632,
-      "grad_norm": 5.3391194343566895,
-      "learning_rate": 2.131524926686217e-05,
-      "loss": 6.0024,
-      "step": 137000
+      "epoch": 21.49141181399246,
+      "grad_norm": 3.1958000659942627,
+      "learning_rate": 1.4196341293115486e-05,
+      "loss": 6.2866,
+      "step": 51300
     },
     {
-      "epoch": 28.80184331797235,
-      "grad_norm": 5.1614089012146,
-      "learning_rate": 2.1210724759111858e-05,
-      "loss": 5.9992,
-      "step": 137500
+      "epoch": 21.533305404273147,
+      "grad_norm": 3.6010313034057617,
+      "learning_rate": 1.4126518642647677e-05,
+      "loss": 6.2683,
+      "step": 51400
     },
     {
-      "epoch": 28.90657729367407,
-      "grad_norm": 5.429248332977295,
-      "learning_rate": 2.110599078341014e-05,
-      "loss": 5.9929,
-      "step": 138000
+      "epoch": 21.575198994553833,
+      "grad_norm": 3.429414749145508,
+      "learning_rate": 1.4056695992179864e-05,
+      "loss": 6.2408,
+      "step": 51500
     },
     {
-      "epoch": 29.011311269375785,
-      "grad_norm": 5.1270012855529785,
-      "learning_rate": 2.100125680770842e-05,
-      "loss": 5.9883,
-      "step": 138500
+      "epoch": 21.61709258483452,
+      "grad_norm": 3.069561004638672,
+      "learning_rate": 1.3986873341712054e-05,
+      "loss": 6.2641,
+      "step": 51600
     },
     {
-      "epoch": 29.116045245077505,
-      "grad_norm": 5.027891159057617,
-      "learning_rate": 2.0896522832006704e-05,
-      "loss": 5.963,
-      "step": 139000
+      "epoch": 21.658986175115206,
+      "grad_norm": 3.575247287750244,
+      "learning_rate": 1.3917050691244241e-05,
+      "loss": 6.2722,
+      "step": 51700
     },
     {
-      "epoch": 29.22077922077922,
-      "grad_norm": 5.712099552154541,
-      "learning_rate": 2.079199832425639e-05,
-      "loss": 5.9811,
-      "step": 139500
+      "epoch": 21.700879765395893,
+      "grad_norm": 3.033505439758301,
+      "learning_rate": 1.3847228040776427e-05,
+      "loss": 6.2424,
+      "step": 51800
     },
     {
-      "epoch": 29.325513196480937,
-      "grad_norm": 4.954220294952393,
-      "learning_rate": 2.0687264348554674e-05,
-      "loss": 5.9808,
-      "step": 140000
+      "epoch": 21.742773355676583,
+      "grad_norm": 3.287740707397461,
+      "learning_rate": 1.3777405390308618e-05,
+      "loss": 6.2516,
+      "step": 51900
     },
     {
-      "epoch": 29.430247172182657,
-      "grad_norm": 5.713419437408447,
-      "learning_rate": 2.0582530372852953e-05,
-      "loss": 5.9887,
-      "step": 140500
+      "epoch": 21.78466694595727,
+      "grad_norm": 3.0363028049468994,
+      "learning_rate": 1.3707582739840804e-05,
+      "loss": 6.2641,
+      "step": 52000
     },
     {
-      "epoch": 29.534981147884373,
-      "grad_norm": 4.683711528778076,
-      "learning_rate": 2.0477796397151236e-05,
-      "loss": 5.9612,
-      "step": 141000
+      "epoch": 21.826560536237956,
+      "grad_norm": 3.1549689769744873,
+      "learning_rate": 1.3637760089372995e-05,
+      "loss": 6.2335,
+      "step": 52100
     },
     {
-      "epoch": 29.63971512358609,
-      "grad_norm": 5.164538383483887,
-      "learning_rate": 2.0373271889400923e-05,
-      "loss": 5.993,
-      "step": 141500
+      "epoch": 21.868454126518643,
+      "grad_norm": 3.8512282371520996,
+      "learning_rate": 1.3567937438905182e-05,
+      "loss": 6.2729,
+      "step": 52200
     },
     {
-      "epoch": 29.74444909928781,
-      "grad_norm": 5.386078357696533,
-      "learning_rate": 2.0268537913699203e-05,
-      "loss": 5.9735,
-      "step": 142000
+      "epoch": 21.91034771679933,
+      "grad_norm": 4.0751824378967285,
+      "learning_rate": 1.3498813014942047e-05,
+      "loss": 6.2397,
+      "step": 52300
     },
     {
-      "epoch": 29.849183074989526,
-      "grad_norm": 4.4406418800354,
-      "learning_rate": 2.016380393799749e-05,
-      "loss": 5.9672,
-      "step": 142500
+      "epoch": 21.952241307080016,
+      "grad_norm": 3.375235080718994,
+      "learning_rate": 1.3428990364474236e-05,
+      "loss": 6.2316,
+      "step": 52400
     },
     {
-      "epoch": 29.953917050691246,
-      "grad_norm": 5.029815673828125,
-      "learning_rate": 2.0059069962295772e-05,
-      "loss": 5.961,
-      "step": 143000
+      "epoch": 21.994134897360702,
+      "grad_norm": 3.093156337738037,
+      "learning_rate": 1.3359167714006423e-05,
+      "loss": 6.2468,
+      "step": 52500
     },
     {
-      "epoch": 30.058651026392962,
-      "grad_norm": 4.666591167449951,
-      "learning_rate": 1.9954335986594052e-05,
-      "loss": 5.9505,
-      "step": 143500
+      "epoch": 22.036028487641392,
+      "grad_norm": 3.729182243347168,
+      "learning_rate": 1.3289345063538614e-05,
+      "loss": 6.2366,
+      "step": 52600
     },
     {
-      "epoch": 30.163385002094678,
-      "grad_norm": 6.975547790527344,
-      "learning_rate": 1.9849602010892335e-05,
-      "loss": 5.956,
-      "step": 144000
+      "epoch": 22.07792207792208,
+      "grad_norm": 3.4075732231140137,
+      "learning_rate": 1.32195224130708e-05,
+      "loss": 6.2693,
+      "step": 52700
     },
     {
-      "epoch": 30.268118977796398,
-      "grad_norm": 4.687684535980225,
-      "learning_rate": 1.974507750314202e-05,
-      "loss": 5.9475,
-      "step": 144500
+      "epoch": 22.119815668202765,
+      "grad_norm": 2.9553005695343018,
+      "learning_rate": 1.3149699762602988e-05,
+      "loss": 6.2592,
+      "step": 52800
     },
     {
-      "epoch": 30.372852953498114,
-      "grad_norm": 5.594231605529785,
-      "learning_rate": 1.96403435274403e-05,
-      "loss": 5.9496,
-      "step": 145000
+      "epoch": 22.16170925848345,
+      "grad_norm": 3.094538688659668,
+      "learning_rate": 1.3079877112135177e-05,
+      "loss": 6.26,
+      "step": 52900
     },
     {
-      "epoch": 30.477586929199834,
-      "grad_norm": 4.879722595214844,
-      "learning_rate": 1.9535609551738584e-05,
-      "loss": 5.9577,
-      "step": 145500
+      "epoch": 22.203602848764138,
+      "grad_norm": 3.907914161682129,
+      "learning_rate": 1.3010054461667365e-05,
+      "loss": 6.2711,
+      "step": 53000
     },
     {
-      "epoch": 30.58232090490155,
-      "grad_norm": 5.470447540283203,
-      "learning_rate": 1.9430875576036867e-05,
-      "loss": 5.9672,
-      "step": 146000
+      "epoch": 22.245496439044825,
+      "grad_norm": 3.7182159423828125,
+      "learning_rate": 1.2940231811199554e-05,
+      "loss": 6.2713,
+      "step": 53100
     },
     {
-      "epoch": 30.687054880603267,
-      "grad_norm": 5.818385124206543,
-      "learning_rate": 1.932614160033515e-05,
-      "loss": 5.9501,
-      "step": 146500
+      "epoch": 22.287390029325515,
+      "grad_norm": 2.8652303218841553,
+      "learning_rate": 1.2870409160731741e-05,
+      "loss": 6.2325,
+      "step": 53200
     },
     {
-      "epoch": 30.791788856304986,
-      "grad_norm": 5.907487392425537,
-      "learning_rate": 1.9221617092584834e-05,
-      "loss": 5.9458,
-      "step": 147000
+      "epoch": 22.3292836196062,
+      "grad_norm": 3.190359592437744,
+      "learning_rate": 1.2800586510263929e-05,
+      "loss": 6.2563,
+      "step": 53300
     },
     {
-      "epoch": 30.896522832006703,
-      "grad_norm": 4.739224433898926,
-      "learning_rate": 1.911688311688312e-05,
-      "loss": 5.935,
-      "step": 147500
+      "epoch": 22.371177209886888,
+      "grad_norm": 3.372394561767578,
+      "learning_rate": 1.2730763859796118e-05,
+      "loss": 6.2489,
+      "step": 53400
     },
     {
-      "epoch": 31.00125680770842,
-      "grad_norm": 4.57131814956665,
-      "learning_rate": 1.90121491411814e-05,
-      "loss": 5.945,
-      "step": 148000
+      "epoch": 22.413070800167574,
+      "grad_norm": 3.340397596359253,
+      "learning_rate": 1.2660941209328306e-05,
+      "loss": 6.2147,
+      "step": 53500
     },
     {
-      "epoch": 31.10599078341014,
-      "grad_norm": 5.128586769104004,
-      "learning_rate": 1.8907415165479683e-05,
-      "loss": 5.9494,
-      "step": 148500
+      "epoch": 22.45496439044826,
+      "grad_norm": 3.1127400398254395,
+      "learning_rate": 1.2591118558860495e-05,
+      "loss": 6.2588,
+      "step": 53600
     },
     {
-      "epoch": 31.210724759111855,
-      "grad_norm": 4.871676921844482,
-      "learning_rate": 1.880289065772937e-05,
-      "loss": 5.9415,
-      "step": 149000
+      "epoch": 22.496857980728947,
+      "grad_norm": 4.315746307373047,
+      "learning_rate": 1.2521295908392683e-05,
+      "loss": 6.2641,
+      "step": 53700
     },
     {
-      "epoch": 31.315458734813575,
-      "grad_norm": 5.380068778991699,
-      "learning_rate": 1.8698156682027652e-05,
-      "loss": 5.939,
-      "step": 149500
+      "epoch": 22.538751571009634,
+      "grad_norm": 3.204827070236206,
+      "learning_rate": 1.2451473257924872e-05,
+      "loss": 6.2506,
+      "step": 53800
     },
     {
-      "epoch": 31.42019271051529,
-      "grad_norm": 5.430812835693359,
-      "learning_rate": 1.8593422706325932e-05,
-      "loss": 5.9276,
-      "step": 150000
+      "epoch": 22.580645161290324,
+      "grad_norm": 3.653074026107788,
+      "learning_rate": 1.238165060745706e-05,
+      "loss": 6.2512,
+      "step": 53900
     },
     {
-      "epoch": 31.524926686217007,
-      "grad_norm": 4.7710442543029785,
-      "learning_rate": 1.8488688730624215e-05,
-      "loss": 5.9413,
-      "step": 150500
+      "epoch": 22.62253875157101,
+      "grad_norm": 3.8693697452545166,
+      "learning_rate": 1.2311827956989249e-05,
+      "loss": 6.2515,
+      "step": 54000
     },
     {
-      "epoch": 31.629660661918727,
-      "grad_norm": 5.183919906616211,
-      "learning_rate": 1.8383954754922498e-05,
-      "loss": 5.9257,
-      "step": 151000
+      "epoch": 22.664432341851697,
+      "grad_norm": 3.9418985843658447,
+      "learning_rate": 1.2242005306521436e-05,
+      "loss": 6.2522,
+      "step": 54100
     },
     {
-      "epoch": 31.734394637620444,
-      "grad_norm": 4.851598739624023,
-      "learning_rate": 1.827943024717218e-05,
-      "loss": 5.9251,
-      "step": 151500
+      "epoch": 22.706325932132383,
+      "grad_norm": 3.328951358795166,
+      "learning_rate": 1.2172182656053624e-05,
+      "loss": 6.2244,
+      "step": 54200
     },
     {
-      "epoch": 31.839128613322163,
-      "grad_norm": 4.835882663726807,
-      "learning_rate": 1.8174696271470464e-05,
-      "loss": 5.92,
-      "step": 152000
+      "epoch": 22.74821952241307,
+      "grad_norm": 3.251552104949951,
+      "learning_rate": 1.210305823209049e-05,
+      "loss": 6.2413,
+      "step": 54300
     },
     {
-      "epoch": 31.94386258902388,
-      "grad_norm": 5.428823947906494,
-      "learning_rate": 1.8069962295768748e-05,
-      "loss": 5.9146,
-      "step": 152500
+      "epoch": 22.790113112693756,
+      "grad_norm": 3.0756313800811768,
+      "learning_rate": 1.2033235581622679e-05,
+      "loss": 6.2343,
+      "step": 54400
     },
     {
-      "epoch": 32.0485965647256,
-      "grad_norm": 6.11329984664917,
-      "learning_rate": 1.796522832006703e-05,
-      "loss": 5.9179,
-      "step": 153000
+      "epoch": 22.832006702974446,
+      "grad_norm": 3.174830913543701,
+      "learning_rate": 1.1963412931154867e-05,
+      "loss": 6.2445,
+      "step": 54500
     },
     {
-      "epoch": 32.15333054042731,
-      "grad_norm": 4.836859226226807,
-      "learning_rate": 1.7860703812316717e-05,
-      "loss": 5.9189,
-      "step": 153500
+      "epoch": 22.873900293255133,
+      "grad_norm": 2.831454038619995,
+      "learning_rate": 1.1893590280687056e-05,
+      "loss": 6.2457,
+      "step": 54600
     },
     {
-      "epoch": 32.25806451612903,
-      "grad_norm": 4.598475456237793,
-      "learning_rate": 1.7755969836615e-05,
-      "loss": 5.9207,
-      "step": 154000
+      "epoch": 22.91579388353582,
+      "grad_norm": 3.3783247470855713,
+      "learning_rate": 1.1823767630219245e-05,
+      "loss": 6.2202,
+      "step": 54700
     },
     {
-      "epoch": 32.36279849183075,
-      "grad_norm": 4.638394832611084,
-      "learning_rate": 1.765123586091328e-05,
-      "loss": 5.915,
-      "step": 154500
+      "epoch": 22.957687473816506,
+      "grad_norm": 3.4505226612091064,
+      "learning_rate": 1.1753944979751431e-05,
+      "loss": 6.2329,
+      "step": 54800
     },
     {
-      "epoch": 32.467532467532465,
-      "grad_norm": 5.637279987335205,
-      "learning_rate": 1.7546501885211563e-05,
-      "loss": 5.9228,
-      "step": 155000
+      "epoch": 22.999581064097192,
+      "grad_norm": 4.203530311584473,
+      "learning_rate": 1.168412232928362e-05,
+      "loss": 6.2464,
+      "step": 54900
     },
     {
-      "epoch": 32.572266443234184,
-      "grad_norm": 4.516068458557129,
-      "learning_rate": 1.7441767909509846e-05,
-      "loss": 5.9322,
-      "step": 155500
+      "epoch": 23.04147465437788,
+      "grad_norm": 3.295198678970337,
+      "learning_rate": 1.1614299678815808e-05,
+      "loss": 6.2163,
+      "step": 55000
     },
     {
-      "epoch": 32.677000418935904,
-      "grad_norm": 4.652084827423096,
-      "learning_rate": 1.7337243401759533e-05,
-      "loss": 5.9395,
-      "step": 156000
+      "epoch": 23.083368244658566,
+      "grad_norm": 3.6795082092285156,
+      "learning_rate": 1.1544477028347997e-05,
+      "loss": 6.2108,
+      "step": 55100
     },
     {
-      "epoch": 32.78173439463762,
-      "grad_norm": 5.667607307434082,
-      "learning_rate": 1.7232509426057812e-05,
-      "loss": 5.9172,
-      "step": 156500
+      "epoch": 23.125261834939256,
+      "grad_norm": 3.7577404975891113,
+      "learning_rate": 1.1474654377880186e-05,
+      "loss": 6.2406,
+      "step": 55200
     },
     {
-      "epoch": 32.88646837033934,
-      "grad_norm": 4.980391025543213,
-      "learning_rate": 1.7127775450356095e-05,
-      "loss": 5.934,
-      "step": 157000
+      "epoch": 23.167155425219942,
+      "grad_norm": 4.524641036987305,
+      "learning_rate": 1.1404831727412372e-05,
+      "loss": 6.2449,
+      "step": 55300
     },
     {
-      "epoch": 32.99120234604106,
-      "grad_norm": 4.39646053314209,
-      "learning_rate": 1.702304147465438e-05,
-      "loss": 5.914,
-      "step": 157500
+      "epoch": 23.20904901550063,
+      "grad_norm": 3.3049490451812744,
+      "learning_rate": 1.1335009076944561e-05,
+      "loss": 6.202,
+      "step": 55400
     },
     {
-      "epoch": 33.095936321742776,
-      "grad_norm": 5.263533115386963,
-      "learning_rate": 1.6918516966904062e-05,
-      "loss": 5.9094,
-      "step": 158000
+      "epoch": 23.250942605781315,
+      "grad_norm": 3.6244115829467773,
+      "learning_rate": 1.1265186426476749e-05,
+      "loss": 6.2214,
+      "step": 55500
     },
     {
-      "epoch": 33.20067029744449,
-      "grad_norm": 4.661126136779785,
-      "learning_rate": 1.6813782991202348e-05,
-      "loss": 5.9189,
-      "step": 158500
+      "epoch": 23.292836196062,
+      "grad_norm": 3.1158556938171387,
+      "learning_rate": 1.1195363776008938e-05,
+      "loss": 6.2247,
+      "step": 55600
     },
     {
-      "epoch": 33.30540427314621,
-      "grad_norm": 4.935306549072266,
-      "learning_rate": 1.670904901550063e-05,
-      "loss": 5.8944,
-      "step": 159000
+      "epoch": 23.334729786342688,
+      "grad_norm": 3.208771228790283,
+      "learning_rate": 1.1125541125541126e-05,
+      "loss": 6.2416,
+      "step": 55700
     },
     {
-      "epoch": 33.41013824884793,
-      "grad_norm": 5.82065486907959,
-      "learning_rate": 1.660431503979891e-05,
-      "loss": 5.892,
-      "step": 159500
+      "epoch": 23.376623376623378,
+      "grad_norm": 4.181106090545654,
+      "learning_rate": 1.1055718475073313e-05,
+      "loss": 6.2343,
+      "step": 55800
     },
     {
-      "epoch": 33.51487222454964,
-      "grad_norm": 4.6220927238464355,
-      "learning_rate": 1.6499581064097194e-05,
-      "loss": 5.9022,
-      "step": 160000
+      "epoch": 23.418516966904065,
+      "grad_norm": 2.8972866535186768,
+      "learning_rate": 1.0985895824605503e-05,
+      "loss": 6.2186,
+      "step": 55900
     },
     {
-      "epoch": 33.61960620025136,
-      "grad_norm": 5.109046936035156,
-      "learning_rate": 1.639505655634688e-05,
-      "loss": 5.8933,
-      "step": 160500
+      "epoch": 23.46041055718475,
+      "grad_norm": 3.1691384315490723,
+      "learning_rate": 1.091607317413769e-05,
+      "loss": 6.2328,
+      "step": 56000
     },
     {
-      "epoch": 33.72434017595308,
-      "grad_norm": 5.230437278747559,
-      "learning_rate": 1.6290322580645164e-05,
-      "loss": 5.9107,
-      "step": 161000
+      "epoch": 23.502304147465438,
+      "grad_norm": 3.214346408843994,
+      "learning_rate": 1.084625052366988e-05,
+      "loss": 6.2356,
+      "step": 56100
     },
     {
-      "epoch": 33.829074151654794,
-      "grad_norm": 6.466080188751221,
-      "learning_rate": 1.6185588604943443e-05,
-      "loss": 5.9077,
-      "step": 161500
+      "epoch": 23.544197737746124,
+      "grad_norm": 3.0547690391540527,
+      "learning_rate": 1.0776427873202067e-05,
+      "loss": 6.2245,
+      "step": 56200
     },
     {
-      "epoch": 33.933808127356514,
-      "grad_norm": 4.655428409576416,
-      "learning_rate": 1.6080854629241726e-05,
-      "loss": 5.9186,
-      "step": 162000
+      "epoch": 23.58609132802681,
+      "grad_norm": 3.6090760231018066,
+      "learning_rate": 1.0707303449238935e-05,
+      "loss": 6.2634,
+      "step": 56300
     },
     {
-      "epoch": 34.038542103058234,
-      "grad_norm": 6.136354923248291,
-      "learning_rate": 1.597612065354001e-05,
-      "loss": 5.8815,
-      "step": 162500
+      "epoch": 23.627984918307497,
+      "grad_norm": 3.210068702697754,
+      "learning_rate": 1.0637480798771122e-05,
+      "loss": 6.2126,
+      "step": 56400
     },
     {
-      "epoch": 34.143276078759946,
-      "grad_norm": 5.668376445770264,
-      "learning_rate": 1.587138667783829e-05,
-      "loss": 5.8872,
-      "step": 163000
+      "epoch": 23.669878508588187,
+      "grad_norm": 3.872507095336914,
+      "learning_rate": 1.056765814830331e-05,
+      "loss": 6.2286,
+      "step": 56500
     },
     {
-      "epoch": 34.248010054461666,
-      "grad_norm": 5.579314708709717,
-      "learning_rate": 1.5766862170087976e-05,
-      "loss": 5.8966,
-      "step": 163500
+      "epoch": 23.711772098868874,
+      "grad_norm": 4.503695011138916,
+      "learning_rate": 1.0497835497835499e-05,
+      "loss": 6.2156,
+      "step": 56600
     },
     {
-      "epoch": 34.352744030163386,
-      "grad_norm": 5.474893569946289,
-      "learning_rate": 1.5662128194386262e-05,
-      "loss": 5.8865,
-      "step": 164000
+      "epoch": 23.75366568914956,
+      "grad_norm": 3.963315486907959,
+      "learning_rate": 1.0428012847367686e-05,
+      "loss": 6.2247,
+      "step": 56700
     },
     {
-      "epoch": 34.457478005865106,
-      "grad_norm": 4.853377342224121,
-      "learning_rate": 1.5557394218684542e-05,
-      "loss": 5.8718,
-      "step": 164500
+      "epoch": 23.795559279430247,
+      "grad_norm": 3.4394917488098145,
+      "learning_rate": 1.0358190196899876e-05,
+      "loss": 6.234,
+      "step": 56800
     },
     {
-      "epoch": 34.56221198156682,
-      "grad_norm": 4.8684563636779785,
-      "learning_rate": 1.5452660242982825e-05,
-      "loss": 5.8896,
-      "step": 165000
+      "epoch": 23.837452869710933,
+      "grad_norm": 3.403167724609375,
+      "learning_rate": 1.0288367546432063e-05,
+      "loss": 6.2045,
+      "step": 56900
     },
     {
-      "epoch": 34.66694595726854,
-      "grad_norm": 4.851233959197998,
-      "learning_rate": 1.5347926267281108e-05,
-      "loss": 5.8908,
-      "step": 165500
+      "epoch": 23.87934645999162,
+      "grad_norm": 2.8274378776550293,
+      "learning_rate": 1.0218544895964251e-05,
+      "loss": 6.2121,
+      "step": 57000
     },
     {
-      "epoch": 34.77167993297026,
-      "grad_norm": 6.647628307342529,
-      "learning_rate": 1.524319229157939e-05,
-      "loss": 5.8972,
-      "step": 166000
+      "epoch": 23.92124005027231,
+      "grad_norm": 3.277188301086426,
+      "learning_rate": 1.0148722245496438e-05,
+      "loss": 6.222,
+      "step": 57100
     },
     {
-      "epoch": 34.87641390867197,
-      "grad_norm": 5.826745986938477,
-      "learning_rate": 1.5138667783829074e-05,
-      "loss": 5.8852,
-      "step": 166500
+      "epoch": 23.963133640552996,
+      "grad_norm": 3.0735063552856445,
+      "learning_rate": 1.0078899595028628e-05,
+      "loss": 6.2257,
+      "step": 57200
     },
     {
-      "epoch": 34.98114788437369,
-      "grad_norm": 5.109675407409668,
-      "learning_rate": 1.5033933808127357e-05,
-      "loss": 5.8971,
-      "step": 167000
+      "epoch": 24.005027230833683,
+      "grad_norm": 3.6680026054382324,
+      "learning_rate": 1.0009076944560817e-05,
+      "loss": 6.2131,
+      "step": 57300
     },
     {
-      "epoch": 35.08588186007541,
-      "grad_norm": 5.743017196655273,
-      "learning_rate": 1.4929199832425639e-05,
-      "loss": 5.8892,
-      "step": 167500
+      "epoch": 24.04692082111437,
+      "grad_norm": 3.134713888168335,
+      "learning_rate": 9.939254294093005e-06,
+      "loss": 6.2241,
+      "step": 57400
     },
     {
-      "epoch": 35.19061583577712,
-      "grad_norm": 4.862270355224609,
-      "learning_rate": 1.482446585672392e-05,
-      "loss": 5.8976,
-      "step": 168000
+      "epoch": 24.088814411395056,
+      "grad_norm": 2.9466712474823,
+      "learning_rate": 9.869431643625192e-06,
+      "loss": 6.2158,
+      "step": 57500
     },
     {
-      "epoch": 35.29534981147884,
-      "grad_norm": 5.532686233520508,
-      "learning_rate": 1.4719941348973607e-05,
-      "loss": 5.8716,
-      "step": 168500
+      "epoch": 24.130708001675742,
+      "grad_norm": 3.468949794769287,
+      "learning_rate": 9.79960899315738e-06,
+      "loss": 6.1793,
+      "step": 57600
     },
     {
-      "epoch": 35.40008378718056,
-      "grad_norm": 5.498019695281982,
-      "learning_rate": 1.4615207373271891e-05,
-      "loss": 5.8768,
-      "step": 169000
+      "epoch": 24.17260159195643,
+      "grad_norm": 3.5487060546875,
+      "learning_rate": 9.729786342689569e-06,
+      "loss": 6.2218,
+      "step": 57700
     },
     {
-      "epoch": 35.504817762882276,
-      "grad_norm": 4.7324042320251465,
-      "learning_rate": 1.4510473397570173e-05,
-      "loss": 5.8739,
-      "step": 169500
+      "epoch": 24.21449518223712,
+      "grad_norm": 4.345893383026123,
+      "learning_rate": 9.659963692221758e-06,
+      "loss": 6.2023,
+      "step": 57800
     },
     {
-      "epoch": 35.609551738583995,
-      "grad_norm": 4.973413944244385,
-      "learning_rate": 1.4405739421868456e-05,
-      "loss": 5.8801,
-      "step": 170000
+      "epoch": 24.256388772517806,
+      "grad_norm": 2.9016401767730713,
+      "learning_rate": 9.590141041753946e-06,
+      "loss": 6.23,
+      "step": 57900
     },
     {
-      "epoch": 35.714285714285715,
-      "grad_norm": 4.977658271789551,
-      "learning_rate": 1.430121491411814e-05,
-      "loss": 5.8644,
-      "step": 170500
+      "epoch": 24.298282362798492,
+      "grad_norm": 4.17023229598999,
+      "learning_rate": 9.520318391286135e-06,
+      "loss": 6.2114,
+      "step": 58000
     },
     {
-      "epoch": 35.819019689987435,
-      "grad_norm": 5.551715850830078,
-      "learning_rate": 1.4196480938416424e-05,
-      "loss": 5.8789,
-      "step": 171000
+      "epoch": 24.34017595307918,
+      "grad_norm": 3.322115421295166,
+      "learning_rate": 9.45049574081832e-06,
+      "loss": 6.204,
+      "step": 58100
     },
     {
-      "epoch": 35.92375366568915,
-      "grad_norm": 5.135740756988525,
-      "learning_rate": 1.4091746962714705e-05,
-      "loss": 5.8862,
-      "step": 171500
+      "epoch": 24.382069543359865,
+      "grad_norm": 3.709805488586426,
+      "learning_rate": 9.38067309035051e-06,
+      "loss": 6.2087,
+      "step": 58200
     },
     {
-      "epoch": 36.02848764139087,
-      "grad_norm": 5.068655967712402,
-      "learning_rate": 1.3987012987012987e-05,
-      "loss": 5.879,
-      "step": 172000
+      "epoch": 24.42396313364055,
+      "grad_norm": 3.225588798522949,
+      "learning_rate": 9.311548666387376e-06,
+      "loss": 6.2436,
+      "step": 58300
     },
     {
-      "epoch": 36.13322161709259,
-      "grad_norm": 5.393857479095459,
-      "learning_rate": 1.388227901131127e-05,
-      "loss": 5.8544,
-      "step": 172500
+      "epoch": 24.46585672392124,
+      "grad_norm": 3.1229472160339355,
+      "learning_rate": 9.241726015919565e-06,
+      "loss": 6.2253,
+      "step": 58400
     },
     {
-      "epoch": 36.2379555927943,
-      "grad_norm": 5.854538440704346,
-      "learning_rate": 1.3777545035609551e-05,
-      "loss": 5.8824,
-      "step": 173000
+      "epoch": 24.507750314201928,
+      "grad_norm": 3.4445230960845947,
+      "learning_rate": 9.171903365451753e-06,
+      "loss": 6.2254,
+      "step": 58500
     },
     {
-      "epoch": 36.34268956849602,
-      "grad_norm": 5.566401481628418,
-      "learning_rate": 1.3673020527859238e-05,
-      "loss": 5.8587,
-      "step": 173500
+      "epoch": 24.549643904482615,
+      "grad_norm": 4.2796807289123535,
+      "learning_rate": 9.102080714983942e-06,
+      "loss": 6.221,
+      "step": 58600
     },
     {
-      "epoch": 36.44742354419774,
-      "grad_norm": 6.091250896453857,
-      "learning_rate": 1.3568286552157519e-05,
-      "loss": 5.8624,
-      "step": 174000
+      "epoch": 24.5915374947633,
+      "grad_norm": 3.2323966026306152,
+      "learning_rate": 9.03225806451613e-06,
+      "loss": 6.228,
+      "step": 58700
     },
     {
-      "epoch": 36.55215751989945,
-      "grad_norm": 4.826417922973633,
-      "learning_rate": 1.3463552576455804e-05,
-      "loss": 5.879,
-      "step": 174500
+      "epoch": 24.633431085043988,
+      "grad_norm": 4.064596652984619,
+      "learning_rate": 8.962435414048317e-06,
+      "loss": 6.2363,
+      "step": 58800
     },
     {
-      "epoch": 36.65689149560117,
-      "grad_norm": 5.28770637512207,
-      "learning_rate": 1.3358818600754087e-05,
-      "loss": 5.8632,
-      "step": 175000
+      "epoch": 24.675324675324674,
+      "grad_norm": 3.068544864654541,
+      "learning_rate": 8.893310990085183e-06,
+      "loss": 6.2508,
+      "step": 58900
     },
     {
-      "epoch": 36.76162547130289,
-      "grad_norm": 5.072086811065674,
-      "learning_rate": 1.3254084625052368e-05,
-      "loss": 5.8698,
-      "step": 175500
+      "epoch": 24.71721826560536,
+      "grad_norm": 2.6201155185699463,
+      "learning_rate": 8.823488339617372e-06,
+      "loss": 6.2193,
+      "step": 59000
     },
     {
-      "epoch": 36.866359447004605,
-      "grad_norm": 6.194067001342773,
-      "learning_rate": 1.3149560117302053e-05,
-      "loss": 5.8701,
-      "step": 176000
+      "epoch": 24.75911185588605,
+      "grad_norm": 4.960629463195801,
+      "learning_rate": 8.753665689149562e-06,
+      "loss": 6.1999,
+      "step": 59100
     },
     {
-      "epoch": 36.971093422706325,
-      "grad_norm": 5.250491142272949,
-      "learning_rate": 1.3044826141600336e-05,
-      "loss": 5.855,
-      "step": 176500
+      "epoch": 24.801005446166737,
+      "grad_norm": 3.191586971282959,
+      "learning_rate": 8.683843038681749e-06,
+      "loss": 6.2203,
+      "step": 59200
     },
     {
-      "epoch": 37.075827398408045,
-      "grad_norm": 4.9726080894470215,
-      "learning_rate": 1.2940092165898618e-05,
-      "loss": 5.8613,
-      "step": 177000
+      "epoch": 24.842899036447424,
+      "grad_norm": 3.224745512008667,
+      "learning_rate": 8.614020388213937e-06,
+      "loss": 6.212,
+      "step": 59300
     },
     {
-      "epoch": 37.180561374109764,
-      "grad_norm": 5.526548385620117,
-      "learning_rate": 1.28353581901969e-05,
-      "loss": 5.8519,
-      "step": 177500
+      "epoch": 24.88479262672811,
+      "grad_norm": 3.450741767883301,
+      "learning_rate": 8.544197737746124e-06,
+      "loss": 6.2386,
+      "step": 59400
     },
     {
-      "epoch": 37.28529534981148,
-      "grad_norm": 5.5989861488342285,
-      "learning_rate": 1.2730624214495182e-05,
-      "loss": 5.8642,
-      "step": 178000
+      "epoch": 24.926686217008797,
+      "grad_norm": 4.297729969024658,
+      "learning_rate": 8.474375087278313e-06,
+      "loss": 6.2088,
+      "step": 59500
     },
     {
-      "epoch": 37.3900293255132,
-      "grad_norm": 5.138686180114746,
-      "learning_rate": 1.2625890238793465e-05,
-      "loss": 5.852,
-      "step": 178500
+      "epoch": 24.968579807289483,
+      "grad_norm": 3.376110553741455,
+      "learning_rate": 8.404552436810503e-06,
+      "loss": 6.2176,
+      "step": 59600
     },
     {
-      "epoch": 37.49476330121492,
-      "grad_norm": 5.0514326095581055,
-      "learning_rate": 1.252115626309175e-05,
-      "loss": 5.8484,
-      "step": 179000
+      "epoch": 25.010473397570173,
+      "grad_norm": 3.0211358070373535,
+      "learning_rate": 8.33472978634269e-06,
+      "loss": 6.1906,
+      "step": 59700
     },
     {
-      "epoch": 37.59949727691663,
-      "grad_norm": 4.9300360679626465,
-      "learning_rate": 1.241642228739003e-05,
-      "loss": 5.849,
-      "step": 179500
+      "epoch": 25.05236698785086,
+      "grad_norm": 2.8490803241729736,
+      "learning_rate": 8.264907135874878e-06,
+      "loss": 6.2,
+      "step": 59800
     },
     {
-      "epoch": 37.70423125261835,
-      "grad_norm": 5.487224102020264,
-      "learning_rate": 1.2311897779639716e-05,
-      "loss": 5.8562,
-      "step": 180000
+      "epoch": 25.094260578131546,
+      "grad_norm": 3.0233705043792725,
+      "learning_rate": 8.195084485407065e-06,
+      "loss": 6.1886,
+      "step": 59900
     },
     {
-      "epoch": 37.80896522832007,
-      "grad_norm": 5.826539516448975,
-      "learning_rate": 1.2207163803937999e-05,
-      "loss": 5.8665,
-      "step": 180500
+      "epoch": 25.136154168412233,
+      "grad_norm": 3.7582995891571045,
+      "learning_rate": 8.125261834939255e-06,
+      "loss": 6.2064,
+      "step": 60000
     },
     {
-      "epoch": 37.91369920402178,
-      "grad_norm": 5.733819961547852,
-      "learning_rate": 1.2102639296187684e-05,
-      "loss": 5.8569,
-      "step": 181000
+      "epoch": 25.17804775869292,
+      "grad_norm": 3.128079891204834,
+      "learning_rate": 8.055439184471442e-06,
+      "loss": 6.2264,
+      "step": 60100
     },
     {
-      "epoch": 38.0184331797235,
-      "grad_norm": 4.917960166931152,
-      "learning_rate": 1.1997905320485967e-05,
-      "loss": 5.8395,
-      "step": 181500
+      "epoch": 25.219941348973606,
+      "grad_norm": 3.1808972358703613,
+      "learning_rate": 7.985616534003632e-06,
+      "loss": 6.2149,
+      "step": 60200
     },
     {
-      "epoch": 38.12316715542522,
-      "grad_norm": 5.337119102478027,
-      "learning_rate": 1.1893171344784248e-05,
-      "loss": 5.854,
-      "step": 182000
+      "epoch": 25.261834939254292,
+      "grad_norm": 3.2326996326446533,
+      "learning_rate": 7.91579388353582e-06,
+      "loss": 6.2142,
+      "step": 60300
     },
     {
-      "epoch": 38.227901131126934,
-      "grad_norm": 5.299139022827148,
-      "learning_rate": 1.178843736908253e-05,
-      "loss": 5.8433,
-      "step": 182500
+      "epoch": 25.303728529534983,
+      "grad_norm": 3.267465114593506,
+      "learning_rate": 7.845971233068007e-06,
+      "loss": 6.2439,
+      "step": 60400
     },
     {
-      "epoch": 38.332635106828654,
-      "grad_norm": 5.900153160095215,
-      "learning_rate": 1.1683703393380813e-05,
-      "loss": 5.8535,
-      "step": 183000
+      "epoch": 25.34562211981567,
+      "grad_norm": 3.691075563430786,
+      "learning_rate": 7.776148582600196e-06,
+      "loss": 6.2178,
+      "step": 60500
     },
     {
-      "epoch": 38.437369082530374,
-      "grad_norm": 6.776584625244141,
-      "learning_rate": 1.15791788856305e-05,
-      "loss": 5.8454,
-      "step": 183500
+      "epoch": 25.387515710096356,
+      "grad_norm": 3.290562152862549,
+      "learning_rate": 7.706325932132383e-06,
+      "loss": 6.2165,
+      "step": 60600
     },
     {
-      "epoch": 38.542103058232094,
-      "grad_norm": 6.258368015289307,
-      "learning_rate": 1.1474444909928783e-05,
-      "loss": 5.8354,
-      "step": 184000
+      "epoch": 25.429409300377042,
+      "grad_norm": 4.553886413574219,
+      "learning_rate": 7.636503281664573e-06,
+      "loss": 6.2165,
+      "step": 60700
     },
     {
-      "epoch": 38.64683703393381,
-      "grad_norm": 5.288670539855957,
-      "learning_rate": 1.1369710934227064e-05,
-      "loss": 5.8458,
-      "step": 184500
+      "epoch": 25.47130289065773,
+      "grad_norm": 4.013444423675537,
+      "learning_rate": 7.566680631196761e-06,
+      "loss": 6.2122,
+      "step": 60800
     },
     {
-      "epoch": 38.751571009635526,
-      "grad_norm": 5.596650123596191,
-      "learning_rate": 1.1264976958525345e-05,
-      "loss": 5.8387,
-      "step": 185000
+      "epoch": 25.513196480938415,
+      "grad_norm": 4.044810771942139,
+      "learning_rate": 7.496857980728948e-06,
+      "loss": 6.2533,
+      "step": 60900
     },
     {
-      "epoch": 38.856304985337246,
-      "grad_norm": 5.121638774871826,
-      "learning_rate": 1.1160452450775032e-05,
-      "loss": 5.8268,
-      "step": 185500
+      "epoch": 25.555090071219105,
+      "grad_norm": 3.788613796234131,
+      "learning_rate": 7.427035330261137e-06,
+      "loss": 6.2039,
+      "step": 61000
     },
     {
-      "epoch": 38.96103896103896,
-      "grad_norm": 4.5758442878723145,
-      "learning_rate": 1.1055718475073313e-05,
-      "loss": 5.83,
-      "step": 186000
+      "epoch": 25.59698366149979,
+      "grad_norm": 3.317281484603882,
+      "learning_rate": 7.3572126797933255e-06,
+      "loss": 6.2228,
+      "step": 61100
     },
     {
-      "epoch": 39.06577293674068,
-      "grad_norm": 5.161282539367676,
-      "learning_rate": 1.0950984499371596e-05,
-      "loss": 5.8544,
-      "step": 186500
+      "epoch": 25.638877251780478,
+      "grad_norm": 3.4238085746765137,
+      "learning_rate": 7.287390029325514e-06,
+      "loss": 6.1979,
+      "step": 61200
     },
     {
-      "epoch": 39.1705069124424,
-      "grad_norm": 4.628884315490723,
-      "learning_rate": 1.084625052366988e-05,
-      "loss": 5.8474,
-      "step": 187000
+      "epoch": 25.680770842061165,
+      "grad_norm": 3.1558725833892822,
+      "learning_rate": 7.217567378857702e-06,
+      "loss": 6.2044,
+      "step": 61300
     },
     {
-      "epoch": 39.27524088814411,
-      "grad_norm": 5.854598045349121,
-      "learning_rate": 1.074151654796816e-05,
-      "loss": 5.8501,
-      "step": 187500
+      "epoch": 25.72266443234185,
+      "grad_norm": 2.939328670501709,
+      "learning_rate": 7.147744728389889e-06,
+      "loss": 6.2312,
+      "step": 61400
     },
     {
-      "epoch": 39.37997486384583,
-      "grad_norm": 5.315525054931641,
-      "learning_rate": 1.0636782572266444e-05,
-      "loss": 5.8265,
-      "step": 188000
+      "epoch": 25.764558022622538,
+      "grad_norm": 4.0037455558776855,
+      "learning_rate": 7.0779220779220775e-06,
+      "loss": 6.228,
+      "step": 61500
     },
     {
-      "epoch": 39.48470883954755,
-      "grad_norm": 6.078185081481934,
-      "learning_rate": 1.0532048596564727e-05,
-      "loss": 5.8419,
-      "step": 188500
+      "epoch": 25.806451612903224,
+      "grad_norm": 4.4582343101501465,
+      "learning_rate": 7.008099427454267e-06,
+      "loss": 6.2132,
+      "step": 61600
     },
     {
-      "epoch": 39.589442815249264,
-      "grad_norm": 5.223086357116699,
-      "learning_rate": 1.0427524088814412e-05,
-      "loss": 5.8197,
-      "step": 189000
+      "epoch": 25.848345203183914,
+      "grad_norm": 3.006201982498169,
+      "learning_rate": 6.938276776986455e-06,
+      "loss": 6.2242,
+      "step": 61700
     },
     {
-      "epoch": 39.69417679095098,
-      "grad_norm": 5.235757827758789,
-      "learning_rate": 1.0322790113112695e-05,
-      "loss": 5.8245,
-      "step": 189500
+      "epoch": 25.8902387934646,
+      "grad_norm": 3.6898059844970703,
+      "learning_rate": 6.8684541265186436e-06,
+      "loss": 6.2134,
+      "step": 61800
     },
     {
-      "epoch": 39.7989107666527,
-      "grad_norm": 5.124643325805664,
-      "learning_rate": 1.0218056137410976e-05,
-      "loss": 5.839,
-      "step": 190000
+      "epoch": 25.932132383745287,
+      "grad_norm": 3.3489785194396973,
+      "learning_rate": 6.798631476050832e-06,
+      "loss": 6.2042,
+      "step": 61900
     },
     {
-      "epoch": 39.90364474235442,
-      "grad_norm": 5.613321304321289,
-      "learning_rate": 1.011332216170926e-05,
-      "loss": 5.844,
-      "step": 190500
+      "epoch": 25.974025974025974,
+      "grad_norm": 3.2489922046661377,
+      "learning_rate": 6.729507052087698e-06,
+      "loss": 6.2212,
+      "step": 62000
     },
     {
-      "epoch": 40.008378718056136,
-      "grad_norm": 5.873430252075195,
-      "learning_rate": 1.0008588186007542e-05,
-      "loss": 5.837,
-      "step": 191000
+      "epoch": 26.01591956430666,
+      "grad_norm": 4.022356033325195,
+      "learning_rate": 6.659684401619885e-06,
+      "loss": 6.2423,
+      "step": 62100
     },
     {
-      "epoch": 40.113112693757856,
-      "grad_norm": 5.089309215545654,
-      "learning_rate": 9.903854210305824e-06,
-      "loss": 5.8384,
-      "step": 191500
+      "epoch": 26.057813154587347,
+      "grad_norm": 4.803937911987305,
+      "learning_rate": 6.589861751152074e-06,
+      "loss": 6.2319,
+      "step": 62200
     },
     {
-      "epoch": 40.217846669459576,
-      "grad_norm": 7.3569817543029785,
-      "learning_rate": 9.79932970255551e-06,
-      "loss": 5.8232,
-      "step": 192000
+      "epoch": 26.099706744868037,
+      "grad_norm": 3.7283337116241455,
+      "learning_rate": 6.520039100684262e-06,
+      "loss": 6.1924,
+      "step": 62300
     },
     {
-      "epoch": 40.32258064516129,
-      "grad_norm": 6.024489402770996,
-      "learning_rate": 9.694595726853792e-06,
-      "loss": 5.8292,
-      "step": 192500
+      "epoch": 26.141600335148723,
+      "grad_norm": 3.817946672439575,
+      "learning_rate": 6.450216450216451e-06,
+      "loss": 6.2039,
+      "step": 62400
     },
     {
-      "epoch": 40.42731462086301,
-      "grad_norm": 5.7150983810424805,
-      "learning_rate": 9.589861751152073e-06,
-      "loss": 5.8614,
-      "step": 193000
+      "epoch": 26.18349392542941,
+      "grad_norm": 3.4621963500976562,
+      "learning_rate": 6.380393799748639e-06,
+      "loss": 6.214,
+      "step": 62500
     },
     {
-      "epoch": 40.53204859656473,
-      "grad_norm": 4.717107772827148,
-      "learning_rate": 9.485127775450356e-06,
-      "loss": 5.8092,
-      "step": 193500
+      "epoch": 26.225387515710096,
+      "grad_norm": 4.458475112915039,
+      "learning_rate": 6.310571149280827e-06,
+      "loss": 6.2327,
+      "step": 62600
     },
     {
-      "epoch": 40.63678257226644,
-      "grad_norm": 4.9722490310668945,
-      "learning_rate": 9.380603267700043e-06,
-      "loss": 5.8231,
-      "step": 194000
+      "epoch": 26.267281105990783,
+      "grad_norm": 3.1324493885040283,
+      "learning_rate": 6.240748498813015e-06,
+      "loss": 6.2518,
+      "step": 62700
     },
     {
-      "epoch": 40.74151654796816,
-      "grad_norm": 5.593094825744629,
-      "learning_rate": 9.275869291998326e-06,
-      "loss": 5.8339,
-      "step": 194500
+      "epoch": 26.30917469627147,
+      "grad_norm": 3.410626173019409,
+      "learning_rate": 6.1709258483452034e-06,
+      "loss": 6.2054,
+      "step": 62800
     },
     {
-      "epoch": 40.84625052366988,
-      "grad_norm": 5.731310844421387,
-      "learning_rate": 9.171135316296607e-06,
-      "loss": 5.8381,
-      "step": 195000
+      "epoch": 26.351068286552156,
+      "grad_norm": 3.221602201461792,
+      "learning_rate": 6.101103197877392e-06,
+      "loss": 6.2297,
+      "step": 62900
     },
     {
-      "epoch": 40.95098449937159,
-      "grad_norm": 5.072065353393555,
-      "learning_rate": 9.066401340594889e-06,
-      "loss": 5.8367,
-      "step": 195500
+      "epoch": 26.392961876832846,
+      "grad_norm": 3.1413893699645996,
+      "learning_rate": 6.031280547409579e-06,
+      "loss": 6.2134,
+      "step": 63000
     },
     {
-      "epoch": 41.05571847507331,
-      "grad_norm": 5.219040870666504,
-      "learning_rate": 8.961667364893172e-06,
-      "loss": 5.8234,
-      "step": 196000
+      "epoch": 26.434855467113533,
+      "grad_norm": 3.3834433555603027,
+      "learning_rate": 5.961457896941768e-06,
+      "loss": 6.167,
+      "step": 63100
     },
     {
-      "epoch": 41.16045245077503,
-      "grad_norm": 5.844238758087158,
-      "learning_rate": 8.856933389191455e-06,
-      "loss": 5.8347,
-      "step": 196500
+      "epoch": 26.47674905739422,
+      "grad_norm": 3.016921281814575,
+      "learning_rate": 5.891635246473957e-06,
+      "loss": 6.2181,
+      "step": 63200
     },
     {
-      "epoch": 41.26518642647675,
-      "grad_norm": 6.088447093963623,
-      "learning_rate": 8.752199413489736e-06,
-      "loss": 5.8178,
-      "step": 197000
+      "epoch": 26.518642647674906,
+      "grad_norm": 3.4190244674682617,
+      "learning_rate": 5.821812596006145e-06,
+      "loss": 6.172,
+      "step": 63300
     },
     {
-      "epoch": 41.369920402178465,
-      "grad_norm": 5.14108943939209,
-      "learning_rate": 8.647465437788019e-06,
-      "loss": 5.8248,
-      "step": 197500
+      "epoch": 26.560536237955592,
+      "grad_norm": 3.519742488861084,
+      "learning_rate": 5.751989945538333e-06,
+      "loss": 6.2144,
+      "step": 63400
     },
     {
-      "epoch": 41.474654377880185,
-      "grad_norm": 5.424249172210693,
-      "learning_rate": 8.542940930037704e-06,
-      "loss": 5.8113,
-      "step": 198000
+      "epoch": 26.60242982823628,
+      "grad_norm": 3.083923101425171,
+      "learning_rate": 5.682167295070521e-06,
+      "loss": 6.204,
+      "step": 63500
     },
     {
-      "epoch": 41.579388353581905,
-      "grad_norm": 4.888121604919434,
-      "learning_rate": 8.43841642228739e-06,
-      "loss": 5.8111,
-      "step": 198500
+      "epoch": 26.64432341851697,
+      "grad_norm": 3.8977878093719482,
+      "learning_rate": 5.612344644602709e-06,
+      "loss": 6.1759,
+      "step": 63600
     },
     {
-      "epoch": 41.68412232928362,
-      "grad_norm": 4.9909515380859375,
-      "learning_rate": 8.333682446585672e-06,
-      "loss": 5.8276,
-      "step": 199000
+      "epoch": 26.686217008797655,
+      "grad_norm": 3.5598249435424805,
+      "learning_rate": 5.5425219941348974e-06,
+      "loss": 6.2233,
+      "step": 63700
     },
     {
-      "epoch": 41.78885630498534,
-      "grad_norm": 5.032175540924072,
-      "learning_rate": 8.228948470883955e-06,
-      "loss": 5.8332,
-      "step": 199500
+      "epoch": 26.72811059907834,
+      "grad_norm": 3.6333513259887695,
+      "learning_rate": 5.472699343667086e-06,
+      "loss": 6.2133,
+      "step": 63800
     },
     {
-      "epoch": 41.89359028068706,
-      "grad_norm": 5.116880416870117,
-      "learning_rate": 8.124214495182238e-06,
-      "loss": 5.8233,
-      "step": 200000
+      "epoch": 26.770004189359028,
+      "grad_norm": 3.2468085289001465,
+      "learning_rate": 5.402876693199274e-06,
+      "loss": 6.2081,
+      "step": 63900
     },
     {
-      "epoch": 41.99832425638877,
-      "grad_norm": 5.235647678375244,
-      "learning_rate": 8.019689987431923e-06,
-      "loss": 5.8297,
-      "step": 200500
+      "epoch": 26.811897779639715,
+      "grad_norm": 3.6896772384643555,
+      "learning_rate": 5.333054042731463e-06,
+      "loss": 6.1935,
+      "step": 64000
     },
     {
-      "epoch": 42.10305823209049,
-      "grad_norm": 5.445380210876465,
-      "learning_rate": 7.914956011730206e-06,
-      "loss": 5.8159,
-      "step": 201000
+      "epoch": 26.8537913699204,
+      "grad_norm": 3.263144016265869,
+      "learning_rate": 5.26323139226365e-06,
+      "loss": 6.2127,
+      "step": 64100
     },
     {
-      "epoch": 42.20779220779221,
-      "grad_norm": 4.979036331176758,
-      "learning_rate": 7.810222036028488e-06,
-      "loss": 5.809,
-      "step": 201500
+      "epoch": 26.895684960201088,
+      "grad_norm": 3.2848362922668457,
+      "learning_rate": 5.193408741795839e-06,
+      "loss": 6.2074,
+      "step": 64200
     },
     {
-      "epoch": 42.31252618349392,
-      "grad_norm": 5.359362602233887,
-      "learning_rate": 7.70548806032677e-06,
-      "loss": 5.8346,
-      "step": 202000
+      "epoch": 26.937578550481778,
+      "grad_norm": 3.675541639328003,
+      "learning_rate": 5.123586091328027e-06,
+      "loss": 6.2015,
+      "step": 64300
     },
     {
-      "epoch": 42.41726015919564,
-      "grad_norm": 5.264519214630127,
-      "learning_rate": 7.600754084625053e-06,
-      "loss": 5.8089,
-      "step": 202500
+      "epoch": 26.979472140762464,
+      "grad_norm": 3.413780689239502,
+      "learning_rate": 5.0537634408602155e-06,
+      "loss": 6.2218,
+      "step": 64400
     },
     {
-      "epoch": 42.52199413489736,
-      "grad_norm": 5.985982894897461,
-      "learning_rate": 7.496020108923335e-06,
-      "loss": 5.8192,
-      "step": 203000
+      "epoch": 27.02136573104315,
+      "grad_norm": 4.108157634735107,
+      "learning_rate": 4.983940790392404e-06,
+      "loss": 6.212,
+      "step": 64500
     },
     {
-      "epoch": 42.626728110599075,
-      "grad_norm": 5.505626201629639,
-      "learning_rate": 7.391286133221617e-06,
-      "loss": 5.8095,
-      "step": 203500
+      "epoch": 27.063259321323837,
+      "grad_norm": 3.7690155506134033,
+      "learning_rate": 4.9141181399245915e-06,
+      "loss": 6.22,
+      "step": 64600
     },
     {
-      "epoch": 42.731462086300795,
-      "grad_norm": 5.069738388061523,
-      "learning_rate": 7.286552157519899e-06,
-      "loss": 5.8186,
-      "step": 204000
+      "epoch": 27.105152911604524,
+      "grad_norm": 3.379786491394043,
+      "learning_rate": 4.84429548945678e-06,
+      "loss": 6.2334,
+      "step": 64700
     },
     {
-      "epoch": 42.836196062002514,
-      "grad_norm": 6.004745960235596,
-      "learning_rate": 7.182027649769586e-06,
-      "loss": 5.8136,
-      "step": 204500
+      "epoch": 27.14704650188521,
+      "grad_norm": 3.5175390243530273,
+      "learning_rate": 4.774472838988968e-06,
+      "loss": 6.1932,
+      "step": 64800
     },
     {
-      "epoch": 42.940930037704234,
-      "grad_norm": 6.299502372741699,
-      "learning_rate": 7.077293674067868e-06,
-      "loss": 5.8213,
-      "step": 205000
+      "epoch": 27.1889400921659,
+      "grad_norm": 2.8454129695892334,
+      "learning_rate": 4.704650188521157e-06,
+      "loss": 6.2087,
+      "step": 64900
     },
     {
-      "epoch": 43.04566401340595,
-      "grad_norm": 6.302718162536621,
-      "learning_rate": 6.97255969836615e-06,
-      "loss": 5.8075,
-      "step": 205500
+      "epoch": 27.230833682446587,
+      "grad_norm": 3.4630961418151855,
+      "learning_rate": 4.634827538053345e-06,
+      "loss": 6.2142,
+      "step": 65000
     },
     {
-      "epoch": 43.15039798910767,
-      "grad_norm": 5.921250343322754,
-      "learning_rate": 6.867825722664433e-06,
-      "loss": 5.8276,
-      "step": 206000
+      "epoch": 27.272727272727273,
+      "grad_norm": 3.339860677719116,
+      "learning_rate": 4.565004887585533e-06,
+      "loss": 6.1772,
+      "step": 65100
     },
     {
-      "epoch": 43.25513196480939,
-      "grad_norm": 5.123110771179199,
-      "learning_rate": 6.763091746962715e-06,
-      "loss": 5.7965,
-      "step": 206500
+      "epoch": 27.31462086300796,
+      "grad_norm": 3.0743260383605957,
+      "learning_rate": 4.495182237117721e-06,
+      "loss": 6.2044,
+      "step": 65200
     },
     {
-      "epoch": 43.3598659405111,
-      "grad_norm": 5.187294006347656,
-      "learning_rate": 6.658357771260998e-06,
-      "loss": 5.8137,
-      "step": 207000
+      "epoch": 27.356514453288646,
+      "grad_norm": 3.2576496601104736,
+      "learning_rate": 4.4253595866499095e-06,
+      "loss": 6.2191,
+      "step": 65300
     },
     {
-      "epoch": 43.46459991621282,
-      "grad_norm": 5.407510757446289,
-      "learning_rate": 6.55362379555928e-06,
-      "loss": 5.8305,
-      "step": 207500
+      "epoch": 27.398408043569333,
+      "grad_norm": 3.326819896697998,
+      "learning_rate": 4.355536936182097e-06,
+      "loss": 6.1762,
+      "step": 65400
     },
     {
-      "epoch": 43.56933389191454,
-      "grad_norm": 5.892600059509277,
-      "learning_rate": 6.449099287808966e-06,
-      "loss": 5.8167,
-      "step": 208000
+      "epoch": 27.44030163385002,
+      "grad_norm": 3.4447667598724365,
+      "learning_rate": 4.285714285714286e-06,
+      "loss": 6.1823,
+      "step": 65500
     },
     {
-      "epoch": 43.67406786761625,
-      "grad_norm": 5.39382266998291,
-      "learning_rate": 6.344365312107248e-06,
-      "loss": 5.8185,
-      "step": 208500
+      "epoch": 27.48219522413071,
+      "grad_norm": 3.4771687984466553,
+      "learning_rate": 4.215891635246475e-06,
+      "loss": 6.228,
+      "step": 65600
     },
     {
-      "epoch": 43.77880184331797,
-      "grad_norm": 5.608034133911133,
-      "learning_rate": 6.23963133640553e-06,
-      "loss": 5.8051,
-      "step": 209000
+      "epoch": 27.524088814411396,
+      "grad_norm": 3.3457424640655518,
+      "learning_rate": 4.146068984778662e-06,
+      "loss": 6.1651,
+      "step": 65700
     },
     {
-      "epoch": 43.88353581901969,
-      "grad_norm": 6.069722652435303,
-      "learning_rate": 6.1348973607038125e-06,
-      "loss": 5.8101,
-      "step": 209500
+      "epoch": 27.565982404692082,
+      "grad_norm": 3.006155490875244,
+      "learning_rate": 4.076246334310851e-06,
+      "loss": 6.2026,
+      "step": 65800
     },
     {
-      "epoch": 43.988269794721404,
-      "grad_norm": 5.938599109649658,
-      "learning_rate": 6.0301633850020955e-06,
-      "loss": 5.807,
-      "step": 210000
+      "epoch": 27.60787599497277,
+      "grad_norm": 4.228708744049072,
+      "learning_rate": 4.006423683843038e-06,
+      "loss": 6.1923,
+      "step": 65900
     },
     {
-      "epoch": 44.093003770423124,
-      "grad_norm": 5.808456897735596,
-      "learning_rate": 5.925429409300378e-06,
-      "loss": 5.8246,
-      "step": 210500
+      "epoch": 27.649769585253456,
+      "grad_norm": 3.4744226932525635,
+      "learning_rate": 3.937299259879905e-06,
+      "loss": 6.1891,
+      "step": 66000
     },
     {
-      "epoch": 44.197737746124844,
-      "grad_norm": 5.229996681213379,
-      "learning_rate": 5.820904901550063e-06,
-      "loss": 5.8029,
-      "step": 211000
+      "epoch": 27.691663175534142,
+      "grad_norm": 3.8300633430480957,
+      "learning_rate": 3.867476609412093e-06,
+      "loss": 6.2237,
+      "step": 66100
     },
     {
-      "epoch": 44.302471721826564,
-      "grad_norm": 5.295706748962402,
-      "learning_rate": 5.716170925848346e-06,
-      "loss": 5.8094,
-      "step": 211500
+      "epoch": 27.733556765814832,
+      "grad_norm": 2.9689528942108154,
+      "learning_rate": 3.7976539589442818e-06,
+      "loss": 6.217,
+      "step": 66200
     },
     {
-      "epoch": 44.407205697528276,
-      "grad_norm": 5.649194240570068,
-      "learning_rate": 5.611436950146628e-06,
-      "loss": 5.811,
-      "step": 212000
+      "epoch": 27.77545035609552,
+      "grad_norm": 3.1309947967529297,
+      "learning_rate": 3.7278313084764698e-06,
+      "loss": 6.2061,
+      "step": 66300
     },
     {
-      "epoch": 44.511939673229996,
-      "grad_norm": 6.5928521156311035,
-      "learning_rate": 5.50670297444491e-06,
-      "loss": 5.7974,
-      "step": 212500
+      "epoch": 27.817343946376205,
+      "grad_norm": 3.4571166038513184,
+      "learning_rate": 3.658008658008658e-06,
+      "loss": 6.1863,
+      "step": 66400
     },
     {
-      "epoch": 44.616673648931716,
-      "grad_norm": 6.246605396270752,
-      "learning_rate": 5.401968998743192e-06,
-      "loss": 5.8011,
-      "step": 213000
+      "epoch": 27.85923753665689,
+      "grad_norm": 3.354229211807251,
+      "learning_rate": 3.5881860075408466e-06,
+      "loss": 6.1996,
+      "step": 66500
     },
     {
-      "epoch": 44.72140762463343,
-      "grad_norm": 5.312093734741211,
-      "learning_rate": 5.2972350230414745e-06,
-      "loss": 5.7819,
-      "step": 213500
+      "epoch": 27.901131126937578,
+      "grad_norm": 3.745568037033081,
+      "learning_rate": 3.5183633570730346e-06,
+      "loss": 6.1839,
+      "step": 66600
     },
     {
-      "epoch": 44.82614160033515,
-      "grad_norm": 5.348554611206055,
-      "learning_rate": 5.19271051529116e-06,
-      "loss": 5.8027,
-      "step": 214000
+      "epoch": 27.943024717218265,
+      "grad_norm": 3.356715440750122,
+      "learning_rate": 3.448540706605223e-06,
+      "loss": 6.2048,
+      "step": 66700
     },
     {
-      "epoch": 44.93087557603687,
-      "grad_norm": 5.95352029800415,
-      "learning_rate": 5.087976539589443e-06,
-      "loss": 5.8046,
-      "step": 214500
+      "epoch": 27.98491830749895,
+      "grad_norm": 2.964492082595825,
+      "learning_rate": 3.378718056137411e-06,
+      "loss": 6.228,
+      "step": 66800
     },
     {
-      "epoch": 45.03560955173858,
-      "grad_norm": 5.978014945983887,
-      "learning_rate": 4.983242563887726e-06,
-      "loss": 5.8021,
-      "step": 215000
+      "epoch": 28.02681189777964,
+      "grad_norm": 3.336606502532959,
+      "learning_rate": 3.3088954056695994e-06,
+      "loss": 6.1953,
+      "step": 66900
     },
     {
-      "epoch": 45.1403435274403,
-      "grad_norm": 5.595849990844727,
-      "learning_rate": 4.878508588186008e-06,
-      "loss": 5.7996,
-      "step": 215500
+      "epoch": 28.068705488060328,
+      "grad_norm": 3.264971971511841,
+      "learning_rate": 3.239072755201788e-06,
+      "loss": 6.1783,
+      "step": 67000
     },
     {
-      "epoch": 45.24507750314202,
-      "grad_norm": 5.570345401763916,
-      "learning_rate": 4.77377461248429e-06,
-      "loss": 5.7973,
-      "step": 216000
+      "epoch": 28.110599078341014,
+      "grad_norm": 3.4968082904815674,
+      "learning_rate": 3.169250104733976e-06,
+      "loss": 6.2117,
+      "step": 67100
     },
     {
-      "epoch": 45.34981147884373,
-      "grad_norm": 5.320748805999756,
-      "learning_rate": 4.669040636782573e-06,
-      "loss": 5.7886,
-      "step": 216500
+      "epoch": 28.1524926686217,
+      "grad_norm": 3.4082252979278564,
+      "learning_rate": 3.099427454266164e-06,
+      "loss": 6.2278,
+      "step": 67200
     },
     {
-      "epoch": 45.45454545454545,
-      "grad_norm": 4.676185607910156,
-      "learning_rate": 4.564516129032258e-06,
-      "loss": 5.7874,
-      "step": 217000
+      "epoch": 28.194386258902387,
+      "grad_norm": 3.52056884765625,
+      "learning_rate": 3.029604803798352e-06,
+      "loss": 6.2037,
+      "step": 67300
     },
     {
-      "epoch": 45.55927943024717,
-      "grad_norm": 5.7768473625183105,
-      "learning_rate": 4.45978215333054e-06,
-      "loss": 5.7875,
-      "step": 217500
+      "epoch": 28.236279849183074,
+      "grad_norm": 3.6062779426574707,
+      "learning_rate": 2.9597821533305406e-06,
+      "loss": 6.1952,
+      "step": 67400
     },
     {
-      "epoch": 45.66401340594889,
-      "grad_norm": 5.668895244598389,
-      "learning_rate": 4.355048177628823e-06,
-      "loss": 5.8121,
-      "step": 218000
+      "epoch": 28.278173439463764,
+      "grad_norm": 3.158705472946167,
+      "learning_rate": 2.889959502862729e-06,
+      "loss": 6.2067,
+      "step": 67500
     },
     {
-      "epoch": 45.768747381650606,
-      "grad_norm": 5.033557891845703,
-      "learning_rate": 4.2503142019271055e-06,
-      "loss": 5.8137,
-      "step": 218500
+      "epoch": 28.32006702974445,
+      "grad_norm": 3.6732075214385986,
+      "learning_rate": 2.820136852394917e-06,
+      "loss": 6.1752,
+      "step": 67600
     },
     {
-      "epoch": 45.873481357352325,
-      "grad_norm": 6.15772819519043,
-      "learning_rate": 4.145580226225388e-06,
-      "loss": 5.8137,
-      "step": 219000
+      "epoch": 28.361960620025137,
+      "grad_norm": 2.842560291290283,
+      "learning_rate": 2.7503142019271054e-06,
+      "loss": 6.1823,
+      "step": 67700
     },
     {
-      "epoch": 45.978215333054045,
-      "grad_norm": 6.617910861968994,
-      "learning_rate": 4.04084625052367e-06,
-      "loss": 5.8079,
-      "step": 219500
+      "epoch": 28.403854210305823,
+      "grad_norm": 3.412233591079712,
+      "learning_rate": 2.6804915514592934e-06,
+      "loss": 6.1997,
+      "step": 67800
     },
     {
-      "epoch": 46.08294930875576,
-      "grad_norm": 5.210205554962158,
-      "learning_rate": 3.936112274821952e-06,
-      "loss": 5.8189,
-      "step": 220000
+      "epoch": 28.44574780058651,
+      "grad_norm": 2.8313143253326416,
+      "learning_rate": 2.610668900991482e-06,
+      "loss": 6.187,
+      "step": 67900
     },
     {
-      "epoch": 46.18768328445748,
-      "grad_norm": 5.630945205688477,
-      "learning_rate": 3.831587767071639e-06,
-      "loss": 5.795,
-      "step": 220500
+      "epoch": 28.487641390867196,
+      "grad_norm": 3.122307300567627,
+      "learning_rate": 2.541544477028348e-06,
+      "loss": 6.1977,
+      "step": 68000
     },
     {
-      "epoch": 46.2924172601592,
-      "grad_norm": 5.8690032958984375,
-      "learning_rate": 3.7268537913699205e-06,
-      "loss": 5.8019,
-      "step": 221000
+      "epoch": 28.529534981147883,
+      "grad_norm": 3.4732697010040283,
+      "learning_rate": 2.4717218265605365e-06,
+      "loss": 6.1913,
+      "step": 68100
     },
     {
-      "epoch": 46.39715123586091,
-      "grad_norm": 5.787112712860107,
-      "learning_rate": 3.6221198156682027e-06,
-      "loss": 5.7901,
-      "step": 221500
+      "epoch": 28.571428571428573,
+      "grad_norm": 3.3936917781829834,
+      "learning_rate": 2.4018991760927245e-06,
+      "loss": 6.2216,
+      "step": 68200
     },
     {
-      "epoch": 46.50188521156263,
-      "grad_norm": 5.568469524383545,
-      "learning_rate": 3.5173858399664853e-06,
-      "loss": 5.803,
-      "step": 222000
+      "epoch": 28.61332216170926,
+      "grad_norm": 3.2980170249938965,
+      "learning_rate": 2.332076525624913e-06,
+      "loss": 6.1989,
+      "step": 68300
     },
     {
-      "epoch": 46.60661918726435,
-      "grad_norm": 5.671326637268066,
-      "learning_rate": 3.4126518642647675e-06,
-      "loss": 5.7931,
-      "step": 222500
+      "epoch": 28.655215751989946,
+      "grad_norm": 4.099823951721191,
+      "learning_rate": 2.2622538751571013e-06,
+      "loss": 6.1755,
+      "step": 68400
     },
     {
-      "epoch": 46.71135316296606,
-      "grad_norm": 5.4085307121276855,
-      "learning_rate": 3.30791788856305e-06,
-      "loss": 5.7903,
-      "step": 223000
+      "epoch": 28.697109342270632,
+      "grad_norm": 3.7930960655212402,
+      "learning_rate": 2.1924312246892893e-06,
+      "loss": 6.2155,
+      "step": 68500
     },
     {
-      "epoch": 46.81608713866778,
-      "grad_norm": 5.7440571784973145,
-      "learning_rate": 3.203393380812736e-06,
-      "loss": 5.7897,
-      "step": 223500
+      "epoch": 28.73900293255132,
+      "grad_norm": 3.620065212249756,
+      "learning_rate": 2.1226085742214777e-06,
+      "loss": 6.2041,
+      "step": 68600
     },
     {
-      "epoch": 46.9208211143695,
-      "grad_norm": 5.144542217254639,
-      "learning_rate": 3.098659405111018e-06,
-      "loss": 5.8145,
-      "step": 224000
+      "epoch": 28.780896522832006,
+      "grad_norm": 3.3451735973358154,
+      "learning_rate": 2.0527859237536657e-06,
+      "loss": 6.2095,
+      "step": 68700
     },
     {
-      "epoch": 47.02555509007122,
-      "grad_norm": 5.842213153839111,
-      "learning_rate": 2.9939254294093008e-06,
-      "loss": 5.8046,
-      "step": 224500
+      "epoch": 28.822790113112696,
+      "grad_norm": 4.007857799530029,
+      "learning_rate": 1.982963273285854e-06,
+      "loss": 6.2283,
+      "step": 68800
     },
     {
-      "epoch": 47.130289065772935,
-      "grad_norm": 6.161410331726074,
-      "learning_rate": 2.8891914537075826e-06,
-      "loss": 5.7965,
-      "step": 225000
+      "epoch": 28.864683703393382,
+      "grad_norm": 4.236888885498047,
+      "learning_rate": 1.9131406228180425e-06,
+      "loss": 6.1799,
+      "step": 68900
     },
     {
-      "epoch": 47.235023041474655,
-      "grad_norm": 6.173724174499512,
-      "learning_rate": 2.784457478005865e-06,
-      "loss": 5.7854,
-      "step": 225500
+      "epoch": 28.90657729367407,
+      "grad_norm": 3.222273111343384,
+      "learning_rate": 1.8433179723502305e-06,
+      "loss": 6.2008,
+      "step": 69000
     },
     {
-      "epoch": 47.339757017176375,
-      "grad_norm": 5.132796287536621,
-      "learning_rate": 2.679932970255551e-06,
-      "loss": 5.791,
-      "step": 226000
+      "epoch": 28.948470883954755,
+      "grad_norm": 3.8649580478668213,
+      "learning_rate": 1.7734953218824187e-06,
+      "loss": 6.194,
+      "step": 69100
     },
     {
-      "epoch": 47.44449099287809,
-      "grad_norm": 6.053417205810547,
-      "learning_rate": 2.5751989945538332e-06,
-      "loss": 5.7935,
-      "step": 226500
+      "epoch": 28.99036447423544,
+      "grad_norm": 3.9630191326141357,
+      "learning_rate": 1.7036726714146071e-06,
+      "loss": 6.2083,
+      "step": 69200
     },
     {
-      "epoch": 47.54922496857981,
-      "grad_norm": 5.08466911315918,
-      "learning_rate": 2.470465018852116e-06,
-      "loss": 5.8006,
-      "step": 227000
+      "epoch": 29.032258064516128,
+      "grad_norm": 3.9617035388946533,
+      "learning_rate": 1.6338500209467953e-06,
+      "loss": 6.2197,
+      "step": 69300
     },
     {
-      "epoch": 47.65395894428153,
-      "grad_norm": 6.060305595397949,
-      "learning_rate": 2.365731043150398e-06,
-      "loss": 5.7839,
-      "step": 227500
+      "epoch": 29.074151654796815,
+      "grad_norm": 3.4647514820098877,
+      "learning_rate": 1.5640273704789835e-06,
+      "loss": 6.1919,
+      "step": 69400
     },
     {
-      "epoch": 47.75869291998324,
-      "grad_norm": 5.808520317077637,
-      "learning_rate": 2.2609970674486806e-06,
-      "loss": 5.7933,
-      "step": 228000
+      "epoch": 29.116045245077505,
+      "grad_norm": 3.7548468112945557,
+      "learning_rate": 1.4942047200111717e-06,
+      "loss": 6.171,
+      "step": 69500
     },
     {
-      "epoch": 47.86342689568496,
-      "grad_norm": 5.413971424102783,
-      "learning_rate": 2.156472559698366e-06,
-      "loss": 5.7985,
-      "step": 228500
+      "epoch": 29.15793883535819,
+      "grad_norm": 3.4267735481262207,
+      "learning_rate": 1.42438206954336e-06,
+      "loss": 6.2031,
+      "step": 69600
     },
     {
-      "epoch": 47.96816087138668,
-      "grad_norm": 6.4786529541015625,
-      "learning_rate": 2.0517385839966487e-06,
-      "loss": 5.8059,
-      "step": 229000
+      "epoch": 29.199832425638878,
+      "grad_norm": 3.166888952255249,
+      "learning_rate": 1.3545594190755483e-06,
+      "loss": 6.1937,
+      "step": 69700
     },
     {
-      "epoch": 48.07289484708839,
-      "grad_norm": 5.606147289276123,
-      "learning_rate": 1.947004608294931e-06,
-      "loss": 5.7934,
-      "step": 229500
+      "epoch": 29.241726015919564,
+      "grad_norm": 2.9794344902038574,
+      "learning_rate": 1.2847367686077363e-06,
+      "loss": 6.2068,
+      "step": 69800
     },
     {
-      "epoch": 48.17762882279011,
-      "grad_norm": 5.827240467071533,
-      "learning_rate": 1.8422706325932133e-06,
-      "loss": 5.8007,
-      "step": 230000
+      "epoch": 29.28361960620025,
+      "grad_norm": 3.056293249130249,
+      "learning_rate": 1.2149141181399247e-06,
+      "loss": 6.1652,
+      "step": 69900
     },
     {
-      "epoch": 48.28236279849183,
-      "grad_norm": 5.678854465484619,
-      "learning_rate": 1.7375366568914957e-06,
-      "loss": 5.8038,
-      "step": 230500
+      "epoch": 29.325513196480937,
+      "grad_norm": 3.851149320602417,
+      "learning_rate": 1.145789694176791e-06,
+      "loss": 6.2129,
+      "step": 70000
     },
     {
-      "epoch": 48.38709677419355,
-      "grad_norm": 5.42138671875,
-      "learning_rate": 1.632802681189778e-06,
-      "loss": 5.7976,
-      "step": 231000
+      "epoch": 29.367406786761624,
+      "grad_norm": 3.670929193496704,
+      "learning_rate": 1.0759670437089792e-06,
+      "loss": 6.2068,
+      "step": 70100
     },
     {
-      "epoch": 48.491830749895264,
-      "grad_norm": 6.03090238571167,
-      "learning_rate": 1.5280687054880603e-06,
-      "loss": 5.7874,
-      "step": 231500
+      "epoch": 29.409300377042314,
+      "grad_norm": 3.3581252098083496,
+      "learning_rate": 1.0061443932411674e-06,
+      "loss": 6.2208,
+      "step": 70200
     },
     {
-      "epoch": 48.596564725596984,
-      "grad_norm": 4.660221099853516,
-      "learning_rate": 1.4233347297863427e-06,
-      "loss": 5.7742,
-      "step": 232000
+      "epoch": 29.451193967323,
+      "grad_norm": 3.7551257610321045,
+      "learning_rate": 9.363217427733557e-07,
+      "loss": 6.179,
+      "step": 70300
     },
     {
-      "epoch": 48.701298701298704,
-      "grad_norm": 6.032063961029053,
-      "learning_rate": 1.3188102220360285e-06,
-      "loss": 5.7866,
-      "step": 232500
+      "epoch": 29.493087557603687,
+      "grad_norm": 2.9767682552337646,
+      "learning_rate": 8.664990923055439e-07,
+      "loss": 6.2157,
+      "step": 70400
     },
     {
-      "epoch": 48.80603267700042,
-      "grad_norm": 6.296925067901611,
-      "learning_rate": 1.2142857142857144e-06,
-      "loss": 5.7922,
-      "step": 233000
+      "epoch": 29.534981147884373,
+      "grad_norm": 3.3218774795532227,
+      "learning_rate": 7.966764418377322e-07,
+      "loss": 6.1773,
+      "step": 70500
     },
     {
-      "epoch": 48.91076665270214,
-      "grad_norm": 5.2991437911987305,
-      "learning_rate": 1.1095517385839968e-06,
-      "loss": 5.7873,
-      "step": 233500
+      "epoch": 29.57687473816506,
+      "grad_norm": 4.360437870025635,
+      "learning_rate": 7.268537913699204e-07,
+      "loss": 6.1985,
+      "step": 70600
     },
     {
-      "epoch": 49.015500628403856,
-      "grad_norm": 6.130777835845947,
-      "learning_rate": 1.0048177628822792e-06,
-      "loss": 5.8074,
-      "step": 234000
+      "epoch": 29.618768328445746,
+      "grad_norm": 3.544264078140259,
+      "learning_rate": 6.570311409021087e-07,
+      "loss": 6.1973,
+      "step": 70700
     },
     {
-      "epoch": 49.12023460410557,
-      "grad_norm": 6.305094242095947,
-      "learning_rate": 9.000837871805614e-07,
-      "loss": 5.7847,
-      "step": 234500
+      "epoch": 29.660661918726436,
+      "grad_norm": 3.7416069507598877,
+      "learning_rate": 5.872084904342969e-07,
+      "loss": 6.2031,
+      "step": 70800
     },
     {
-      "epoch": 49.22496857980729,
-      "grad_norm": 6.156949996948242,
-      "learning_rate": 7.953498114788438e-07,
-      "loss": 5.7741,
-      "step": 235000
+      "epoch": 29.702555509007123,
+      "grad_norm": 3.0346035957336426,
+      "learning_rate": 5.173858399664851e-07,
+      "loss": 6.2123,
+      "step": 70900
     },
     {
-      "epoch": 49.32970255550901,
-      "grad_norm": 6.475966930389404,
-      "learning_rate": 6.908253037285296e-07,
-      "loss": 5.8083,
-      "step": 235500
+      "epoch": 29.74444909928781,
+      "grad_norm": 3.2308425903320312,
+      "learning_rate": 4.4756318949867344e-07,
+      "loss": 6.2106,
+      "step": 71000
     },
     {
-      "epoch": 49.43443653121072,
-      "grad_norm": 5.687895774841309,
-      "learning_rate": 5.86091328026812e-07,
-      "loss": 5.7976,
-      "step": 236000
+      "epoch": 29.786342689568496,
+      "grad_norm": 3.0109570026397705,
+      "learning_rate": 3.7774053903086163e-07,
+      "loss": 6.213,
+      "step": 71100
     },
     {
-      "epoch": 49.53917050691244,
-      "grad_norm": 5.615401744842529,
-      "learning_rate": 4.813573523250943e-07,
-      "loss": 5.7907,
-      "step": 236500
+      "epoch": 29.828236279849182,
+      "grad_norm": 3.733609199523926,
+      "learning_rate": 3.079178885630499e-07,
+      "loss": 6.1984,
+      "step": 71200
     },
     {
-      "epoch": 49.64390448261416,
-      "grad_norm": 6.177160263061523,
-      "learning_rate": 3.7662337662337666e-07,
-      "loss": 5.792,
-      "step": 237000
+      "epoch": 29.87012987012987,
+      "grad_norm": 3.5430541038513184,
+      "learning_rate": 2.3809523809523814e-07,
+      "loss": 6.183,
+      "step": 71300
     },
     {
-      "epoch": 49.74863845831588,
-      "grad_norm": 5.604287624359131,
-      "learning_rate": 2.71889400921659e-07,
-      "loss": 5.7837,
-      "step": 237500
+      "epoch": 29.912023460410555,
+      "grad_norm": 3.1964950561523438,
+      "learning_rate": 1.6827258762742634e-07,
+      "loss": 6.1817,
+      "step": 71400
     },
     {
-      "epoch": 49.853372434017594,
-      "grad_norm": 4.826539039611816,
-      "learning_rate": 1.673648931713448e-07,
-      "loss": 5.7808,
-      "step": 238000
+      "epoch": 29.953917050691246,
+      "grad_norm": 3.6197755336761475,
+      "learning_rate": 9.844993715961458e-08,
+      "loss": 6.1907,
+      "step": 71500
     },
     {
-      "epoch": 49.95810640971931,
-      "grad_norm": 5.6649250984191895,
-      "learning_rate": 6.263091746962715e-08,
-      "loss": 5.7888,
-      "step": 238500
+      "epoch": 29.995810640971932,
+      "grad_norm": 3.035473346710205,
+      "learning_rate": 2.862728669180282e-08,
+      "loss": 6.1677,
+      "step": 71600
     },
     {
-      "epoch": 50.0,
-      "step": 238700,
-      "total_flos": 5161725447936000.0,
-      "train_loss": 6.140905278960581,
-      "train_runtime": 7883.0646,
-      "train_samples_per_second": 484.444,
-      "train_steps_per_second": 30.28
+      "epoch": 30.0,
+      "step": 71610,
+      "total_flos": 781486986700800.0,
+      "train_loss": 6.528281962779993,
+      "train_runtime": 3948.1737,
+      "train_samples_per_second": 580.354,
+      "train_steps_per_second": 18.137
     }
   ],
-  "logging_steps": 500,
-  "max_steps": 238700,
+  "logging_steps": 100,
+  "max_steps": 71610,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 50,
+  "num_train_epochs": 30,
   "save_steps": 5000,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -3374,8 +5047,8 @@
       "attributes": {}
     }
   },
-  "total_flos": 5161725447936000.0,
-  "train_batch_size": 16,
+  "total_flos": 781486986700800.0,
+  "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null
 }