diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5920 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.20538317868978367,
+  "eval_steps": 100000000,
+  "global_step": 25200,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 8.15012613848348e-06,
+      "grad_norm": 43.29102325439453,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 81.0533,
+      "step": 1
+    },
+    {
+      "epoch": 0.0002445037841545044,
+      "grad_norm": 42.03493881225586,
+      "learning_rate": 3e-06,
+      "loss": 70.3325,
+      "step": 30
+    },
+    {
+      "epoch": 0.0004890075683090088,
+      "grad_norm": 10.497812271118164,
+      "learning_rate": 6e-06,
+      "loss": 23.8844,
+      "step": 60
+    },
+    {
+      "epoch": 0.0007335113524635132,
+      "grad_norm": 6.692718029022217,
+      "learning_rate": 9e-06,
+      "loss": 11.7198,
+      "step": 90
+    },
+    {
+      "epoch": 0.0009780151366180175,
+      "grad_norm": 9.11446475982666,
+      "learning_rate": 1.2e-05,
+      "loss": 10.3199,
+      "step": 120
+    },
+    {
+      "epoch": 0.001222518920772522,
+      "grad_norm": 16.92243766784668,
+      "learning_rate": 1.5e-05,
+      "loss": 9.5338,
+      "step": 150
+    },
+    {
+      "epoch": 0.0014670227049270264,
+      "grad_norm": 12.929216384887695,
+      "learning_rate": 1.8e-05,
+      "loss": 8.9114,
+      "step": 180
+    },
+    {
+      "epoch": 0.0017115264890815308,
+      "grad_norm": 8.830116271972656,
+      "learning_rate": 2.1e-05,
+      "loss": 8.3867,
+      "step": 210
+    },
+    {
+      "epoch": 0.001956030273236035,
+      "grad_norm": 7.348124980926514,
+      "learning_rate": 2.4e-05,
+      "loss": 8.0113,
+      "step": 240
+    },
+    {
+      "epoch": 0.0022005340573905395,
+      "grad_norm": 12.751787185668945,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 7.6093,
+      "step": 270
+    },
+    {
+      "epoch": 0.002445037841545044,
+      "grad_norm": 15.10444164276123,
+      "learning_rate": 3e-05,
+      "loss": 7.3484,
+      "step": 300
+    },
+    {
+      "epoch": 0.0026895416256995483,
+      "grad_norm": 6.653383731842041,
+      "learning_rate": 3.3e-05,
+      "loss": 7.2105,
+      "step": 330
+    },
+    {
+      "epoch": 0.0029340454098540527,
+      "grad_norm": 6.986039161682129,
+      "learning_rate": 3.6e-05,
+      "loss": 7.0228,
+      "step": 360
+    },
+    {
+      "epoch": 0.003178549194008557,
+      "grad_norm": 6.230088710784912,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 6.8948,
+      "step": 390
+    },
+    {
+      "epoch": 0.0034230529781630616,
+      "grad_norm": 8.170981407165527,
+      "learning_rate": 4.2e-05,
+      "loss": 6.6965,
+      "step": 420
+    },
+    {
+      "epoch": 0.0036675567623175656,
+      "grad_norm": 5.6345930099487305,
+      "learning_rate": 4.5e-05,
+      "loss": 6.5988,
+      "step": 450
+    },
+    {
+      "epoch": 0.00391206054647207,
+      "grad_norm": 5.156513214111328,
+      "learning_rate": 4.8e-05,
+      "loss": 6.4795,
+      "step": 480
+    },
+    {
+      "epoch": 0.0041565643306265745,
+      "grad_norm": 5.4964189529418945,
+      "learning_rate": 4.999999990869806e-05,
+      "loss": 6.333,
+      "step": 510
+    },
+    {
+      "epoch": 0.004401068114781079,
+      "grad_norm": 3.67378306388855,
+      "learning_rate": 4.999999853916893e-05,
+      "loss": 6.2208,
+      "step": 540
+    },
+    {
+      "epoch": 0.004645571898935583,
+      "grad_norm": 8.507222175598145,
+      "learning_rate": 4.9999995526204936e-05,
+      "loss": 6.1097,
+      "step": 570
+    },
+    {
+      "epoch": 0.004890075683090088,
+      "grad_norm": 3.756618022918701,
+      "learning_rate": 4.999999086980628e-05,
+      "loss": 5.9886,
+      "step": 600
+    },
+    {
+      "epoch": 0.005134579467244592,
+      "grad_norm": 3.8149781227111816,
+      "learning_rate": 4.999998456997326e-05,
+      "loss": 5.8779,
+      "step": 630
+    },
+    {
+      "epoch": 0.005379083251399097,
+      "grad_norm": 3.840543270111084,
+      "learning_rate": 4.999997662670628e-05,
+      "loss": 5.805,
+      "step": 660
+    },
+    {
+      "epoch": 0.005623587035553601,
+      "grad_norm": 3.4208931922912598,
+      "learning_rate": 4.999996704000589e-05,
+      "loss": 5.6992,
+      "step": 690
+    },
+    {
+      "epoch": 0.0058680908197081055,
+      "grad_norm": 3.2975683212280273,
+      "learning_rate": 4.99999558098727e-05,
+      "loss": 5.6531,
+      "step": 720
+    },
+    {
+      "epoch": 0.00611259460386261,
+      "grad_norm": 4.05631160736084,
+      "learning_rate": 4.9999942936307445e-05,
+      "loss": 5.554,
+      "step": 750
+    },
+    {
+      "epoch": 0.006357098388017114,
+      "grad_norm": 3.1539864540100098,
+      "learning_rate": 4.9999928419310994e-05,
+      "loss": 5.4931,
+      "step": 780
+    },
+    {
+      "epoch": 0.006601602172171619,
+      "grad_norm": 4.811732292175293,
+      "learning_rate": 4.999991225888427e-05,
+      "loss": 5.4204,
+      "step": 810
+    },
+    {
+      "epoch": 0.006846105956326123,
+      "grad_norm": 2.9593210220336914,
+      "learning_rate": 4.999989445502837e-05,
+      "loss": 5.3687,
+      "step": 840
+    },
+    {
+      "epoch": 0.007090609740480627,
+      "grad_norm": 3.942239284515381,
+      "learning_rate": 4.9999875007744436e-05,
+      "loss": 5.3238,
+      "step": 870
+    },
+    {
+      "epoch": 0.007335113524635131,
+      "grad_norm": 2.29752254486084,
+      "learning_rate": 4.9999853917033756e-05,
+      "loss": 5.2423,
+      "step": 900
+    },
+    {
+      "epoch": 0.007579617308789636,
+      "grad_norm": 2.243770122528076,
+      "learning_rate": 4.999983118289773e-05,
+      "loss": 5.2384,
+      "step": 930
+    },
+    {
+      "epoch": 0.00782412109294414,
+      "grad_norm": 2.5572686195373535,
+      "learning_rate": 4.999980680533782e-05,
+      "loss": 5.1761,
+      "step": 960
+    },
+    {
+      "epoch": 0.008068624877098645,
+      "grad_norm": 2.4739913940429688,
+      "learning_rate": 4.999978078435567e-05,
+      "loss": 5.1215,
+      "step": 990
+    },
+    {
+      "epoch": 0.008313128661253149,
+      "grad_norm": 1.6927807331085205,
+      "learning_rate": 4.999975311995295e-05,
+      "loss": 5.1264,
+      "step": 1020
+    },
+    {
+      "epoch": 0.008557632445407654,
+      "grad_norm": 2.3166885375976562,
+      "learning_rate": 4.99997238121315e-05,
+      "loss": 5.0469,
+      "step": 1050
+    },
+    {
+      "epoch": 0.008802136229562158,
+      "grad_norm": 1.69430410861969,
+      "learning_rate": 4.999969286089325e-05,
+      "loss": 5.005,
+      "step": 1080
+    },
+    {
+      "epoch": 0.009046640013716663,
+      "grad_norm": 2.0859012603759766,
+      "learning_rate": 4.9999660266240235e-05,
+      "loss": 4.9693,
+      "step": 1110
+    },
+    {
+      "epoch": 0.009291143797871167,
+      "grad_norm": 1.7291619777679443,
+      "learning_rate": 4.9999626028174585e-05,
+      "loss": 4.9352,
+      "step": 1140
+    },
+    {
+      "epoch": 0.00953564758202567,
+      "grad_norm": 1.5966330766677856,
+      "learning_rate": 4.999959014669856e-05,
+      "loss": 4.87,
+      "step": 1170
+    },
+    {
+      "epoch": 0.009780151366180176,
+      "grad_norm": 1.9011143445968628,
+      "learning_rate": 4.9999552621814513e-05,
+      "loss": 4.8772,
+      "step": 1200
+    },
+    {
+      "epoch": 0.010024655150334679,
+      "grad_norm": 1.701004981994629,
+      "learning_rate": 4.9999513453524917e-05,
+      "loss": 4.8484,
+      "step": 1230
+    },
+    {
+      "epoch": 0.010269158934489184,
+      "grad_norm": 2.171808958053589,
+      "learning_rate": 4.9999472641832336e-05,
+      "loss": 4.8233,
+      "step": 1260
+    },
+    {
+      "epoch": 0.010513662718643688,
+      "grad_norm": 1.4896094799041748,
+      "learning_rate": 4.999943018673946e-05,
+      "loss": 4.7921,
+      "step": 1290
+    },
+    {
+      "epoch": 0.010758166502798193,
+      "grad_norm": 2.112971544265747,
+      "learning_rate": 4.999938608824909e-05,
+      "loss": 4.7908,
+      "step": 1320
+    },
+    {
+      "epoch": 0.011002670286952697,
+      "grad_norm": 1.7113062143325806,
+      "learning_rate": 4.999934034636411e-05,
+      "loss": 4.7186,
+      "step": 1350
+    },
+    {
+      "epoch": 0.011247174071107202,
+      "grad_norm": 2.099663019180298,
+      "learning_rate": 4.999929296108753e-05,
+      "loss": 4.7112,
+      "step": 1380
+    },
+    {
+      "epoch": 0.011491677855261706,
+      "grad_norm": 1.40205979347229,
+      "learning_rate": 4.9999243932422466e-05,
+      "loss": 4.6701,
+      "step": 1410
+    },
+    {
+      "epoch": 0.011736181639416211,
+      "grad_norm": 1.53098726272583,
+      "learning_rate": 4.999919326037215e-05,
+      "loss": 4.6786,
+      "step": 1440
+    },
+    {
+      "epoch": 0.011980685423570715,
+      "grad_norm": 1.7461609840393066,
+      "learning_rate": 4.99991409449399e-05,
+      "loss": 4.6614,
+      "step": 1470
+    },
+    {
+      "epoch": 0.01222518920772522,
+      "grad_norm": 1.6045819520950317,
+      "learning_rate": 4.999908698612916e-05,
+      "loss": 4.6233,
+      "step": 1500
+    },
+    {
+      "epoch": 0.012469692991879723,
+      "grad_norm": 1.351077675819397,
+      "learning_rate": 4.9999031383943486e-05,
+      "loss": 4.626,
+      "step": 1530
+    },
+    {
+      "epoch": 0.012714196776034229,
+      "grad_norm": 1.4363055229187012,
+      "learning_rate": 4.999897413838651e-05,
+      "loss": 4.5781,
+      "step": 1560
+    },
+    {
+      "epoch": 0.012958700560188732,
+      "grad_norm": 1.1321336030960083,
+      "learning_rate": 4.999891524946202e-05,
+      "loss": 4.5492,
+      "step": 1590
+    },
+    {
+      "epoch": 0.013203204344343238,
+      "grad_norm": 1.2755215167999268,
+      "learning_rate": 4.999885471717387e-05,
+      "loss": 4.5444,
+      "step": 1620
+    },
+    {
+      "epoch": 0.013447708128497741,
+      "grad_norm": 1.471504807472229,
+      "learning_rate": 4.999879254152605e-05,
+      "loss": 4.5247,
+      "step": 1650
+    },
+    {
+      "epoch": 0.013692211912652246,
+      "grad_norm": 1.3664900064468384,
+      "learning_rate": 4.999872872252265e-05,
+      "loss": 4.5158,
+      "step": 1680
+    },
+    {
+      "epoch": 0.01393671569680675,
+      "grad_norm": 1.360611915588379,
+      "learning_rate": 4.999866326016785e-05,
+      "loss": 4.498,
+      "step": 1710
+    },
+    {
+      "epoch": 0.014181219480961254,
+      "grad_norm": 1.43533456325531,
+      "learning_rate": 4.999859615446596e-05,
+      "loss": 4.5035,
+      "step": 1740
+    },
+    {
+      "epoch": 0.014425723265115759,
+      "grad_norm": 1.4639947414398193,
+      "learning_rate": 4.99985274054214e-05,
+      "loss": 4.4533,
+      "step": 1770
+    },
+    {
+      "epoch": 0.014670227049270262,
+      "grad_norm": 1.438585638999939,
+      "learning_rate": 4.999845701303868e-05,
+      "loss": 4.4484,
+      "step": 1800
+    },
+    {
+      "epoch": 0.014914730833424768,
+      "grad_norm": 1.1868066787719727,
+      "learning_rate": 4.999838497732243e-05,
+      "loss": 4.4516,
+      "step": 1830
+    },
+    {
+      "epoch": 0.015159234617579271,
+      "grad_norm": 1.1804672479629517,
+      "learning_rate": 4.999831129827739e-05,
+      "loss": 4.4378,
+      "step": 1860
+    },
+    {
+      "epoch": 0.015403738401733777,
+      "grad_norm": 1.127652645111084,
+      "learning_rate": 4.9998235975908394e-05,
+      "loss": 4.4,
+      "step": 1890
+    },
+    {
+      "epoch": 0.01564824218588828,
+      "grad_norm": 1.2617137432098389,
+      "learning_rate": 4.99981590102204e-05,
+      "loss": 4.3802,
+      "step": 1920
+    },
+    {
+      "epoch": 0.015892745970042785,
+      "grad_norm": 1.4249097108840942,
+      "learning_rate": 4.9998080401218464e-05,
+      "loss": 4.3689,
+      "step": 1950
+    },
+    {
+      "epoch": 0.01613724975419729,
+      "grad_norm": 1.0717036724090576,
+      "learning_rate": 4.999800014890777e-05,
+      "loss": 4.3645,
+      "step": 1980
+    },
+    {
+      "epoch": 0.016381753538351793,
+      "grad_norm": 1.414404273033142,
+      "learning_rate": 4.9997918253293555e-05,
+      "loss": 4.343,
+      "step": 2010
+    },
+    {
+      "epoch": 0.016626257322506298,
+      "grad_norm": 1.1941649913787842,
+      "learning_rate": 4.999783471438124e-05,
+      "loss": 4.3271,
+      "step": 2040
+    },
+    {
+      "epoch": 0.016870761106660803,
+      "grad_norm": 1.0737369060516357,
+      "learning_rate": 4.999774953217631e-05,
+      "loss": 4.3223,
+      "step": 2070
+    },
+    {
+      "epoch": 0.01711526489081531,
+      "grad_norm": 1.2367453575134277,
+      "learning_rate": 4.9997662706684345e-05,
+      "loss": 4.3106,
+      "step": 2100
+    },
+    {
+      "epoch": 0.01735976867496981,
+      "grad_norm": 1.0919674634933472,
+      "learning_rate": 4.999757423791107e-05,
+      "loss": 4.2916,
+      "step": 2130
+    },
+    {
+      "epoch": 0.017604272459124316,
+      "grad_norm": 1.2348527908325195,
+      "learning_rate": 4.9997484125862306e-05,
+      "loss": 4.2935,
+      "step": 2160
+    },
+    {
+      "epoch": 0.01784877624327882,
+      "grad_norm": 1.173093318939209,
+      "learning_rate": 4.999739237054395e-05,
+      "loss": 4.2761,
+      "step": 2190
+    },
+    {
+      "epoch": 0.018093280027433326,
+      "grad_norm": 1.1982988119125366,
+      "learning_rate": 4.9997298971962065e-05,
+      "loss": 4.2619,
+      "step": 2220
+    },
+    {
+      "epoch": 0.018337783811587828,
+      "grad_norm": 1.4401408433914185,
+      "learning_rate": 4.999720393012277e-05,
+      "loss": 4.2672,
+      "step": 2250
+    },
+    {
+      "epoch": 0.018582287595742333,
+      "grad_norm": 1.0311753749847412,
+      "learning_rate": 4.999710724503233e-05,
+      "loss": 4.2272,
+      "step": 2280
+    },
+    {
+      "epoch": 0.01882679137989684,
+      "grad_norm": 1.4242304563522339,
+      "learning_rate": 4.9997008916697075e-05,
+      "loss": 4.2415,
+      "step": 2310
+    },
+    {
+      "epoch": 0.01907129516405134,
+      "grad_norm": 1.039548397064209,
+      "learning_rate": 4.999690894512349e-05,
+      "loss": 4.2086,
+      "step": 2340
+    },
+    {
+      "epoch": 0.019315798948205846,
+      "grad_norm": 1.1213486194610596,
+      "learning_rate": 4.999680733031814e-05,
+      "loss": 4.2043,
+      "step": 2370
+    },
+    {
+      "epoch": 0.01956030273236035,
+      "grad_norm": 1.0591011047363281,
+      "learning_rate": 4.9996704072287716e-05,
+      "loss": 4.1894,
+      "step": 2400
+    },
+    {
+      "epoch": 0.019804806516514856,
+      "grad_norm": 1.0503829717636108,
+      "learning_rate": 4.9996599171038984e-05,
+      "loss": 4.184,
+      "step": 2430
+    },
+    {
+      "epoch": 0.020049310300669358,
+      "grad_norm": 1.068516492843628,
+      "learning_rate": 4.999649262657886e-05,
+      "loss": 4.1918,
+      "step": 2460
+    },
+    {
+      "epoch": 0.020293814084823864,
+      "grad_norm": 0.9620047211647034,
+      "learning_rate": 4.999638443891434e-05,
+      "loss": 4.182,
+      "step": 2490
+    },
+    {
+      "epoch": 0.02053831786897837,
+      "grad_norm": 0.9728343486785889,
+      "learning_rate": 4.999627460805253e-05,
+      "loss": 4.1746,
+      "step": 2520
+    },
+    {
+      "epoch": 0.020782821653132874,
+      "grad_norm": 0.990840494632721,
+      "learning_rate": 4.999616313400066e-05,
+      "loss": 4.1339,
+      "step": 2550
+    },
+    {
+      "epoch": 0.021027325437287376,
+      "grad_norm": 1.0678601264953613,
+      "learning_rate": 4.999605001676605e-05,
+      "loss": 4.1548,
+      "step": 2580
+    },
+    {
+      "epoch": 0.02127182922144188,
+      "grad_norm": 1.1837236881256104,
+      "learning_rate": 4.9995935256356144e-05,
+      "loss": 4.1508,
+      "step": 2610
+    },
+    {
+      "epoch": 0.021516333005596387,
+      "grad_norm": 1.0381749868392944,
+      "learning_rate": 4.9995818852778476e-05,
+      "loss": 4.0977,
+      "step": 2640
+    },
+    {
+      "epoch": 0.021760836789750892,
+      "grad_norm": 0.9649094343185425,
+      "learning_rate": 4.999570080604071e-05,
+      "loss": 4.1405,
+      "step": 2670
+    },
+    {
+      "epoch": 0.022005340573905394,
+      "grad_norm": 1.1858196258544922,
+      "learning_rate": 4.99955811161506e-05,
+      "loss": 4.1242,
+      "step": 2700
+    },
+    {
+      "epoch": 0.0222498443580599,
+      "grad_norm": 1.035134196281433,
+      "learning_rate": 4.9995459783116004e-05,
+      "loss": 4.1042,
+      "step": 2730
+    },
+    {
+      "epoch": 0.022494348142214404,
+      "grad_norm": 1.0874470472335815,
+      "learning_rate": 4.999533680694493e-05,
+      "loss": 4.0951,
+      "step": 2760
+    },
+    {
+      "epoch": 0.02273885192636891,
+      "grad_norm": 1.5923190116882324,
+      "learning_rate": 4.9995212187645416e-05,
+      "loss": 4.1031,
+      "step": 2790
+    },
+    {
+      "epoch": 0.02298335571052341,
+      "grad_norm": 0.935354471206665,
+      "learning_rate": 4.9995085925225693e-05,
+      "loss": 4.0804,
+      "step": 2820
+    },
+    {
+      "epoch": 0.023227859494677917,
+      "grad_norm": 1.6059459447860718,
+      "learning_rate": 4.999495801969404e-05,
+      "loss": 4.0703,
+      "step": 2850
+    },
+    {
+      "epoch": 0.023472363278832422,
+      "grad_norm": 1.1116045713424683,
+      "learning_rate": 4.9994828471058876e-05,
+      "loss": 4.0579,
+      "step": 2880
+    },
+    {
+      "epoch": 0.023716867062986924,
+      "grad_norm": 1.0068392753601074,
+      "learning_rate": 4.9994697279328714e-05,
+      "loss": 4.0635,
+      "step": 2910
+    },
+    {
+      "epoch": 0.02396137084714143,
+      "grad_norm": 0.9715657234191895,
+      "learning_rate": 4.9994564444512176e-05,
+      "loss": 4.068,
+      "step": 2940
+    },
+    {
+      "epoch": 0.024205874631295934,
+      "grad_norm": 1.1550418138504028,
+      "learning_rate": 4.9994429966618e-05,
+      "loss": 4.0416,
+      "step": 2970
+    },
+    {
+      "epoch": 0.02445037841545044,
+      "grad_norm": 1.1463241577148438,
+      "learning_rate": 4.999429384565502e-05,
+      "loss": 4.0224,
+      "step": 3000
+    },
+    {
+      "epoch": 0.02469488219960494,
+      "grad_norm": 0.9614786505699158,
+      "learning_rate": 4.999415608163217e-05,
+      "loss": 4.0092,
+      "step": 3030
+    },
+    {
+      "epoch": 0.024939385983759447,
+      "grad_norm": 1.0928614139556885,
+      "learning_rate": 4.999401667455854e-05,
+      "loss": 4.0083,
+      "step": 3060
+    },
+    {
+      "epoch": 0.025183889767913952,
+      "grad_norm": 1.0884851217269897,
+      "learning_rate": 4.9993875624443274e-05,
+      "loss": 4.0251,
+      "step": 3090
+    },
+    {
+      "epoch": 0.025428393552068457,
+      "grad_norm": 1.0158635377883911,
+      "learning_rate": 4.9993732931295646e-05,
+      "loss": 3.9629,
+      "step": 3120
+    },
+    {
+      "epoch": 0.02567289733622296,
+      "grad_norm": 1.0032926797866821,
+      "learning_rate": 4.999358859512503e-05,
+      "loss": 4.0015,
+      "step": 3150
+    },
+    {
+      "epoch": 0.025917401120377465,
+      "grad_norm": 0.9303460121154785,
+      "learning_rate": 4.9993442615940936e-05,
+      "loss": 4.0066,
+      "step": 3180
+    },
+    {
+      "epoch": 0.02616190490453197,
+      "grad_norm": 1.0750280618667603,
+      "learning_rate": 4.999329499375292e-05,
+      "loss": 3.9937,
+      "step": 3210
+    },
+    {
+      "epoch": 0.026406408688686475,
+      "grad_norm": 1.2123857736587524,
+      "learning_rate": 4.999314572857074e-05,
+      "loss": 3.9891,
+      "step": 3240
+    },
+    {
+      "epoch": 0.026650912472840977,
+      "grad_norm": 0.988030731678009,
+      "learning_rate": 4.9992994820404174e-05,
+      "loss": 3.9779,
+      "step": 3270
+    },
+    {
+      "epoch": 0.026895416256995482,
+      "grad_norm": 1.0025092363357544,
+      "learning_rate": 4.999284226926314e-05,
+      "loss": 3.9684,
+      "step": 3300
+    },
+    {
+      "epoch": 0.027139920041149988,
+      "grad_norm": 0.9497402310371399,
+      "learning_rate": 4.999268807515768e-05,
+      "loss": 3.9542,
+      "step": 3330
+    },
+    {
+      "epoch": 0.027384423825304493,
+      "grad_norm": 0.9898300766944885,
+      "learning_rate": 4.999253223809792e-05,
+      "loss": 3.9462,
+      "step": 3360
+    },
+    {
+      "epoch": 0.027628927609458995,
+      "grad_norm": 0.8951054215431213,
+      "learning_rate": 4.999237475809411e-05,
+      "loss": 3.9535,
+      "step": 3390
+    },
+    {
+      "epoch": 0.0278734313936135,
+      "grad_norm": 1.1410372257232666,
+      "learning_rate": 4.99922156351566e-05,
+      "loss": 3.9701,
+      "step": 3420
+    },
+    {
+      "epoch": 0.028117935177768005,
+      "grad_norm": 0.9642576575279236,
+      "learning_rate": 4.999205486929586e-05,
+      "loss": 3.9532,
+      "step": 3450
+    },
+    {
+      "epoch": 0.028362438961922507,
+      "grad_norm": 0.9304039478302002,
+      "learning_rate": 4.999189246052245e-05,
+      "loss": 3.9335,
+      "step": 3480
+    },
+    {
+      "epoch": 0.028606942746077012,
+      "grad_norm": 0.9684301614761353,
+      "learning_rate": 4.999172840884704e-05,
+      "loss": 3.9289,
+      "step": 3510
+    },
+    {
+      "epoch": 0.028851446530231518,
+      "grad_norm": 0.9566449522972107,
+      "learning_rate": 4.999156271428043e-05,
+      "loss": 3.92,
+      "step": 3540
+    },
+    {
+      "epoch": 0.029095950314386023,
+      "grad_norm": 0.9796332120895386,
+      "learning_rate": 4.9991395376833496e-05,
+      "loss": 3.9173,
+      "step": 3570
+    },
+    {
+      "epoch": 0.029340454098540525,
+      "grad_norm": 0.939052164554596,
+      "learning_rate": 4.999122639651725e-05,
+      "loss": 3.8938,
+      "step": 3600
+    },
+    {
+      "epoch": 0.02958495788269503,
+      "grad_norm": 0.8773409724235535,
+      "learning_rate": 4.9991055773342795e-05,
+      "loss": 3.9225,
+      "step": 3630
+    },
+    {
+      "epoch": 0.029829461666849535,
+      "grad_norm": 0.9443736672401428,
+      "learning_rate": 4.9990883507321354e-05,
+      "loss": 3.8889,
+      "step": 3660
+    },
+    {
+      "epoch": 0.03007396545100404,
+      "grad_norm": 0.9129781723022461,
+      "learning_rate": 4.999070959846424e-05,
+      "loss": 3.894,
+      "step": 3690
+    },
+    {
+      "epoch": 0.030318469235158543,
+      "grad_norm": 0.988441526889801,
+      "learning_rate": 4.999053404678289e-05,
+      "loss": 3.8968,
+      "step": 3720
+    },
+    {
+      "epoch": 0.030562973019313048,
+      "grad_norm": 1.0134508609771729,
+      "learning_rate": 4.999035685228884e-05,
+      "loss": 3.873,
+      "step": 3750
+    },
+    {
+      "epoch": 0.030807476803467553,
+      "grad_norm": 0.9853084087371826,
+      "learning_rate": 4.999017801499375e-05,
+      "loss": 3.8921,
+      "step": 3780
+    },
+    {
+      "epoch": 0.03105198058762206,
+      "grad_norm": 1.0263316631317139,
+      "learning_rate": 4.998999753490937e-05,
+      "loss": 3.8691,
+      "step": 3810
+    },
+    {
+      "epoch": 0.03129648437177656,
+      "grad_norm": 1.2496927976608276,
+      "learning_rate": 4.998981541204757e-05,
+      "loss": 3.8631,
+      "step": 3840
+    },
+    {
+      "epoch": 0.031540988155931066,
+      "grad_norm": 1.2647733688354492,
+      "learning_rate": 4.998963164642031e-05,
+      "loss": 3.8676,
+      "step": 3870
+    },
+    {
+      "epoch": 0.03178549194008557,
+      "grad_norm": 0.8532811403274536,
+      "learning_rate": 4.9989446238039676e-05,
+      "loss": 3.8196,
+      "step": 3900
+    },
+    {
+      "epoch": 0.032029995724240076,
+      "grad_norm": 0.9394932985305786,
+      "learning_rate": 4.998925918691786e-05,
+      "loss": 3.8464,
+      "step": 3930
+    },
+    {
+      "epoch": 0.03227449950839458,
+      "grad_norm": 0.9216808676719666,
+      "learning_rate": 4.998907049306715e-05,
+      "loss": 3.8294,
+      "step": 3960
+    },
+    {
+      "epoch": 0.03251900329254909,
+      "grad_norm": 0.9735797047615051,
+      "learning_rate": 4.998888015649996e-05,
+      "loss": 3.8412,
+      "step": 3990
+    },
+    {
+      "epoch": 0.032763507076703585,
+      "grad_norm": 0.9803655743598938,
+      "learning_rate": 4.99886881772288e-05,
+      "loss": 3.834,
+      "step": 4020
+    },
+    {
+      "epoch": 0.03300801086085809,
+      "grad_norm": 0.8677938580513,
+      "learning_rate": 4.998849455526628e-05,
+      "loss": 3.8325,
+      "step": 4050
+    },
+    {
+      "epoch": 0.033252514645012596,
+      "grad_norm": 1.0530720949172974,
+      "learning_rate": 4.998829929062515e-05,
+      "loss": 3.8125,
+      "step": 4080
+    },
+    {
+      "epoch": 0.0334970184291671,
+      "grad_norm": 0.9517835378646851,
+      "learning_rate": 4.998810238331822e-05,
+      "loss": 3.8079,
+      "step": 4110
+    },
+    {
+      "epoch": 0.033741522213321606,
+      "grad_norm": 0.9056320190429688,
+      "learning_rate": 4.998790383335845e-05,
+      "loss": 3.8023,
+      "step": 4140
+    },
+    {
+      "epoch": 0.03398602599747611,
+      "grad_norm": 1.042631983757019,
+      "learning_rate": 4.9987703640758894e-05,
+      "loss": 3.8089,
+      "step": 4170
+    },
+    {
+      "epoch": 0.03423052978163062,
+      "grad_norm": 1.003548502922058,
+      "learning_rate": 4.99875018055327e-05,
+      "loss": 3.8088,
+      "step": 4200
+    },
+    {
+      "epoch": 0.034475033565785115,
+      "grad_norm": 0.9754064083099365,
+      "learning_rate": 4.998729832769315e-05,
+      "loss": 3.7969,
+      "step": 4230
+    },
+    {
+      "epoch": 0.03471953734993962,
+      "grad_norm": 0.9784180521965027,
+      "learning_rate": 4.998709320725361e-05,
+      "loss": 3.8066,
+      "step": 4260
+    },
+    {
+      "epoch": 0.034964041134094126,
+      "grad_norm": 0.9522731900215149,
+      "learning_rate": 4.998688644422756e-05,
+      "loss": 3.8062,
+      "step": 4290
+    },
+    {
+      "epoch": 0.03520854491824863,
+      "grad_norm": 1.0362435579299927,
+      "learning_rate": 4.998667803862861e-05,
+      "loss": 3.8007,
+      "step": 4320
+    },
+    {
+      "epoch": 0.03545304870240314,
+      "grad_norm": 0.899608850479126,
+      "learning_rate": 4.9986467990470445e-05,
+      "loss": 3.8013,
+      "step": 4350
+    },
+    {
+      "epoch": 0.03569755248655764,
+      "grad_norm": 0.9725902676582336,
+      "learning_rate": 4.998625629976688e-05,
+      "loss": 3.7688,
+      "step": 4380
+    },
+    {
+      "epoch": 0.03594205627071215,
+      "grad_norm": 0.9904030561447144,
+      "learning_rate": 4.998604296653182e-05,
+      "loss": 3.7503,
+      "step": 4410
+    },
+    {
+      "epoch": 0.03618656005486665,
+      "grad_norm": 0.9177742004394531,
+      "learning_rate": 4.99858279907793e-05,
+      "loss": 3.7664,
+      "step": 4440
+    },
+    {
+      "epoch": 0.03643106383902115,
+      "grad_norm": 0.964356005191803,
+      "learning_rate": 4.998561137252346e-05,
+      "loss": 3.7574,
+      "step": 4470
+    },
+    {
+      "epoch": 0.036675567623175656,
+      "grad_norm": 1.0006535053253174,
+      "learning_rate": 4.9985393111778525e-05,
+      "loss": 3.7616,
+      "step": 4500
+    },
+    {
+      "epoch": 0.03692007140733016,
+      "grad_norm": 0.9334198236465454,
+      "learning_rate": 4.998517320855884e-05,
+      "loss": 3.776,
+      "step": 4530
+    },
+    {
+      "epoch": 0.03716457519148467,
+      "grad_norm": 0.9192038774490356,
+      "learning_rate": 4.998495166287887e-05,
+      "loss": 3.7552,
+      "step": 4560
+    },
+    {
+      "epoch": 0.03740907897563917,
+      "grad_norm": 0.9278284311294556,
+      "learning_rate": 4.998472847475318e-05,
+      "loss": 3.7508,
+      "step": 4590
+    },
+    {
+      "epoch": 0.03765358275979368,
+      "grad_norm": 0.870951771736145,
+      "learning_rate": 4.998450364419643e-05,
+      "loss": 3.7419,
+      "step": 4620
+    },
+    {
+      "epoch": 0.03789808654394818,
+      "grad_norm": 0.8986937403678894,
+      "learning_rate": 4.998427717122342e-05,
+      "loss": 3.7285,
+      "step": 4650
+    },
+    {
+      "epoch": 0.03814259032810268,
+      "grad_norm": 0.8922073841094971,
+      "learning_rate": 4.9984049055849024e-05,
+      "loss": 3.7252,
+      "step": 4680
+    },
+    {
+      "epoch": 0.038387094112257186,
+      "grad_norm": 0.8716844916343689,
+      "learning_rate": 4.9983819298088234e-05,
+      "loss": 3.7243,
+      "step": 4710
+    },
+    {
+      "epoch": 0.03863159789641169,
+      "grad_norm": 0.9074581861495972,
+      "learning_rate": 4.9983587897956166e-05,
+      "loss": 3.7254,
+      "step": 4740
+    },
+    {
+      "epoch": 0.0388761016805662,
+      "grad_norm": 0.8903953433036804,
+      "learning_rate": 4.998335485546802e-05,
+      "loss": 3.7363,
+      "step": 4770
+    },
+    {
+      "epoch": 0.0391206054647207,
+      "grad_norm": 1.0858160257339478,
+      "learning_rate": 4.998312017063912e-05,
+      "loss": 3.708,
+      "step": 4800
+    },
+    {
+      "epoch": 0.03936510924887521,
+      "grad_norm": 1.1008403301239014,
+      "learning_rate": 4.9982883843484895e-05,
+      "loss": 3.727,
+      "step": 4830
+    },
+    {
+      "epoch": 0.03960961303302971,
+      "grad_norm": 0.9895632863044739,
+      "learning_rate": 4.998264587402088e-05,
+      "loss": 3.7383,
+      "step": 4860
+    },
+    {
+      "epoch": 0.03985411681718422,
+      "grad_norm": 1.0174016952514648,
+      "learning_rate": 4.9982406262262726e-05,
+      "loss": 3.6967,
+      "step": 4890
+    },
+    {
+      "epoch": 0.040098620601338716,
+      "grad_norm": 0.9164692163467407,
+      "learning_rate": 4.9982165008226175e-05,
+      "loss": 3.7038,
+      "step": 4920
+    },
+    {
+      "epoch": 0.04034312438549322,
+      "grad_norm": 0.8442469835281372,
+      "learning_rate": 4.998192211192708e-05,
+      "loss": 3.7045,
+      "step": 4950
+    },
+    {
+      "epoch": 0.04058762816964773,
+      "grad_norm": 0.9134944677352905,
+      "learning_rate": 4.9981677573381423e-05,
+      "loss": 3.6909,
+      "step": 4980
+    },
+    {
+      "epoch": 0.04083213195380223,
+      "grad_norm": 0.9397178292274475,
+      "learning_rate": 4.9981431392605274e-05,
+      "loss": 3.7087,
+      "step": 5010
+    },
+    {
+      "epoch": 0.04107663573795674,
+      "grad_norm": 1.1442530155181885,
+      "learning_rate": 4.998118356961481e-05,
+      "loss": 3.6804,
+      "step": 5040
+    },
+    {
+      "epoch": 0.04132113952211124,
+      "grad_norm": 0.8651236891746521,
+      "learning_rate": 4.998093410442632e-05,
+      "loss": 3.668,
+      "step": 5070
+    },
+    {
+      "epoch": 0.04156564330626575,
+      "grad_norm": 1.0404279232025146,
+      "learning_rate": 4.998068299705623e-05,
+      "loss": 3.673,
+      "step": 5100
+    },
+    {
+      "epoch": 0.041810147090420254,
+      "grad_norm": 0.9545896053314209,
+      "learning_rate": 4.9980430247521014e-05,
+      "loss": 3.6635,
+      "step": 5130
+    },
+    {
+      "epoch": 0.04205465087457475,
+      "grad_norm": 0.9344103932380676,
+      "learning_rate": 4.998017585583731e-05,
+      "loss": 3.6834,
+      "step": 5160
+    },
+    {
+      "epoch": 0.04229915465872926,
+      "grad_norm": 1.0414049625396729,
+      "learning_rate": 4.9979919822021824e-05,
+      "loss": 3.6868,
+      "step": 5190
+    },
+    {
+      "epoch": 0.04254365844288376,
+      "grad_norm": 0.9391636252403259,
+      "learning_rate": 4.99796621460914e-05,
+      "loss": 3.6671,
+      "step": 5220
+    },
+    {
+      "epoch": 0.04278816222703827,
+      "grad_norm": 0.8789961934089661,
+      "learning_rate": 4.9979402828062963e-05,
+      "loss": 3.6575,
+      "step": 5250
+    },
+    {
+      "epoch": 0.04303266601119277,
+      "grad_norm": 0.8953958749771118,
+      "learning_rate": 4.997914186795358e-05,
+      "loss": 3.6741,
+      "step": 5280
+    },
+    {
+      "epoch": 0.04327716979534728,
+      "grad_norm": 0.8470034003257751,
+      "learning_rate": 4.9978879265780385e-05,
+      "loss": 3.6517,
+      "step": 5310
+    },
+    {
+      "epoch": 0.043521673579501784,
+      "grad_norm": 0.8863801956176758,
+      "learning_rate": 4.997861502156066e-05,
+      "loss": 3.6667,
+      "step": 5340
+    },
+    {
+      "epoch": 0.04376617736365628,
+      "grad_norm": 0.9392006993293762,
+      "learning_rate": 4.997834913531176e-05,
+      "loss": 3.6278,
+      "step": 5370
+    },
+    {
+      "epoch": 0.04401068114781079,
+      "grad_norm": 0.8935637474060059,
+      "learning_rate": 4.9978081607051176e-05,
+      "loss": 3.6656,
+      "step": 5400
+    },
+    {
+      "epoch": 0.04425518493196529,
+      "grad_norm": 0.9051403403282166,
+      "learning_rate": 4.997781243679648e-05,
+      "loss": 3.6439,
+      "step": 5430
+    },
+    {
+      "epoch": 0.0444996887161198,
+      "grad_norm": 0.9147741794586182,
+      "learning_rate": 4.9977541624565374e-05,
+      "loss": 3.6362,
+      "step": 5460
+    },
+    {
+      "epoch": 0.0447441925002743,
+      "grad_norm": 0.8926962614059448,
+      "learning_rate": 4.9977269170375665e-05,
+      "loss": 3.6179,
+      "step": 5490
+    },
+    {
+      "epoch": 0.04498869628442881,
+      "grad_norm": 0.9078648686408997,
+      "learning_rate": 4.997699507424526e-05,
+      "loss": 3.6323,
+      "step": 5520
+    },
+    {
+      "epoch": 0.045233200068583314,
+      "grad_norm": 0.8844118118286133,
+      "learning_rate": 4.997671933619218e-05,
+      "loss": 3.6219,
+      "step": 5550
+    },
+    {
+      "epoch": 0.04547770385273782,
+      "grad_norm": 0.9098038077354431,
+      "learning_rate": 4.9976441956234546e-05,
+      "loss": 3.6341,
+      "step": 5580
+    },
+    {
+      "epoch": 0.04572220763689232,
+      "grad_norm": 0.8704332709312439,
+      "learning_rate": 4.99761629343906e-05,
+      "loss": 3.6038,
+      "step": 5610
+    },
+    {
+      "epoch": 0.04596671142104682,
+      "grad_norm": 0.9353320002555847,
+      "learning_rate": 4.9975882270678676e-05,
+      "loss": 3.6171,
+      "step": 5640
+    },
+    {
+      "epoch": 0.04621121520520133,
+      "grad_norm": 1.0025008916854858,
+      "learning_rate": 4.997559996511723e-05,
+      "loss": 3.6084,
+      "step": 5670
+    },
+    {
+      "epoch": 0.04645571898935583,
+      "grad_norm": 0.8339236974716187,
+      "learning_rate": 4.997531601772481e-05,
+      "loss": 3.6271,
+      "step": 5700
+    },
+    {
+      "epoch": 0.04670022277351034,
+      "grad_norm": 0.8991410136222839,
+      "learning_rate": 4.9975030428520084e-05,
+      "loss": 3.5897,
+      "step": 5730
+    },
+    {
+      "epoch": 0.046944726557664844,
+      "grad_norm": 0.8854517936706543,
+      "learning_rate": 4.997474319752184e-05,
+      "loss": 3.603,
+      "step": 5760
+    },
+    {
+      "epoch": 0.04718923034181935,
+      "grad_norm": 0.9445385932922363,
+      "learning_rate": 4.997445432474895e-05,
+      "loss": 3.6084,
+      "step": 5790
+    },
+    {
+      "epoch": 0.04743373412597385,
+      "grad_norm": 0.871096134185791,
+      "learning_rate": 4.9974163810220406e-05,
+      "loss": 3.5892,
+      "step": 5820
+    },
+    {
+      "epoch": 0.04767823791012835,
+      "grad_norm": 0.8828051686286926,
+      "learning_rate": 4.99738716539553e-05,
+      "loss": 3.6075,
+      "step": 5850
+    },
+    {
+      "epoch": 0.04792274169428286,
+      "grad_norm": 0.8362672328948975,
+      "learning_rate": 4.997357785597284e-05,
+      "loss": 3.5772,
+      "step": 5880
+    },
+    {
+      "epoch": 0.048167245478437364,
+      "grad_norm": 0.9075796008110046,
+      "learning_rate": 4.997328241629234e-05,
+      "loss": 3.5881,
+      "step": 5910
+    },
+    {
+      "epoch": 0.04841174926259187,
+      "grad_norm": 0.9475303888320923,
+      "learning_rate": 4.997298533493323e-05,
+      "loss": 3.5963,
+      "step": 5940
+    },
+    {
+      "epoch": 0.048656253046746374,
+      "grad_norm": 0.807908296585083,
+      "learning_rate": 4.997268661191503e-05,
+      "loss": 3.5776,
+      "step": 5970
+    },
+    {
+      "epoch": 0.04890075683090088,
+      "grad_norm": 0.9100547432899475,
+      "learning_rate": 4.9972386247257385e-05,
+      "loss": 3.5924,
+      "step": 6000
+    },
+    {
+      "epoch": 0.049145260615055385,
+      "grad_norm": 0.79295814037323,
+      "learning_rate": 4.9972084240980025e-05,
+      "loss": 3.5603,
+      "step": 6030
+    },
+    {
+      "epoch": 0.04938976439920988,
+      "grad_norm": 0.8548495173454285,
+      "learning_rate": 4.997178059310281e-05,
+      "loss": 3.5624,
+      "step": 6060
+    },
+    {
+      "epoch": 0.04963426818336439,
+      "grad_norm": 0.8862835168838501,
+      "learning_rate": 4.997147530364571e-05,
+      "loss": 3.5738,
+      "step": 6090
+    },
+    {
+      "epoch": 0.049878771967518894,
+      "grad_norm": 1.0974136590957642,
+      "learning_rate": 4.9971168372628793e-05,
+      "loss": 3.5549,
+      "step": 6120
+    },
+    {
+      "epoch": 0.0501232757516734,
+      "grad_norm": 0.9224910140037537,
+      "learning_rate": 4.997085980007222e-05,
+      "loss": 3.5632,
+      "step": 6150
+    },
+    {
+      "epoch": 0.050367779535827904,
+      "grad_norm": 0.8416465520858765,
+      "learning_rate": 4.99705495859963e-05,
+      "loss": 3.5605,
+      "step": 6180
+    },
+    {
+      "epoch": 0.05061228331998241,
+      "grad_norm": 0.9750398993492126,
+      "learning_rate": 4.99702377304214e-05,
+      "loss": 3.5652,
+      "step": 6210
+    },
+    {
+      "epoch": 0.050856787104136915,
+      "grad_norm": 1.0585671663284302,
+      "learning_rate": 4.9969924233368036e-05,
+      "loss": 3.5605,
+      "step": 6240
+    },
+    {
+      "epoch": 0.05110129088829142,
+      "grad_norm": 1.0180178880691528,
+      "learning_rate": 4.996960909485681e-05,
+      "loss": 3.5543,
+      "step": 6270
+    },
+    {
+      "epoch": 0.05134579467244592,
+      "grad_norm": 0.8081057667732239,
+      "learning_rate": 4.9969292314908446e-05,
+      "loss": 3.5551,
+      "step": 6300
+    },
+    {
+      "epoch": 0.051590298456600424,
+      "grad_norm": 0.8196061253547668,
+      "learning_rate": 4.996897389354376e-05,
+      "loss": 3.5566,
+      "step": 6330
+    },
+    {
+      "epoch": 0.05183480224075493,
+      "grad_norm": 0.9229000806808472,
+      "learning_rate": 4.9968653830783684e-05,
+      "loss": 3.5439,
+      "step": 6360
+    },
+    {
+      "epoch": 0.052079306024909434,
+      "grad_norm": 0.8422220945358276,
+      "learning_rate": 4.996833212664927e-05,
+      "loss": 3.5413,
+      "step": 6390
+    },
+    {
+      "epoch": 0.05232380980906394,
+      "grad_norm": 0.9290680885314941,
+      "learning_rate": 4.996800878116166e-05,
+      "loss": 3.5245,
+      "step": 6420
+    },
+    {
+      "epoch": 0.052568313593218445,
+      "grad_norm": 0.7990279793739319,
+      "learning_rate": 4.99676837943421e-05,
+      "loss": 3.5519,
+      "step": 6450
+    },
+    {
+      "epoch": 0.05281281737737295,
+      "grad_norm": 0.8968884348869324,
+      "learning_rate": 4.996735716621196e-05,
+      "loss": 3.5306,
+      "step": 6480
+    },
+    {
+      "epoch": 0.05305732116152745,
+      "grad_norm": 0.8280816674232483,
+      "learning_rate": 4.996702889679272e-05,
+      "loss": 3.5407,
+      "step": 6510
+    },
+    {
+      "epoch": 0.053301824945681954,
+      "grad_norm": 0.9637565612792969,
+      "learning_rate": 4.996669898610595e-05,
+      "loss": 3.5361,
+      "step": 6540
+    },
+    {
+      "epoch": 0.05354632872983646,
+      "grad_norm": 0.8355570435523987,
+      "learning_rate": 4.996636743417334e-05,
+      "loss": 3.5215,
+      "step": 6570
+    },
+    {
+      "epoch": 0.053790832513990965,
+      "grad_norm": 0.7987068295478821,
+      "learning_rate": 4.996603424101669e-05,
+      "loss": 3.5135,
+      "step": 6600
+    },
+    {
+      "epoch": 0.05403533629814547,
+      "grad_norm": 0.8417894840240479,
+      "learning_rate": 4.996569940665789e-05,
+      "loss": 3.5254,
+      "step": 6630
+    },
+    {
+      "epoch": 0.054279840082299975,
+      "grad_norm": 0.8814358115196228,
+      "learning_rate": 4.996536293111896e-05,
+      "loss": 3.5091,
+      "step": 6660
+    },
+    {
+      "epoch": 0.05452434386645448,
+      "grad_norm": 0.8625611662864685,
+      "learning_rate": 4.996502481442202e-05,
+      "loss": 3.5207,
+      "step": 6690
+    },
+    {
+      "epoch": 0.054768847650608986,
+      "grad_norm": 0.8265979886054993,
+      "learning_rate": 4.9964685056589314e-05,
+      "loss": 3.5011,
+      "step": 6720
+    },
+    {
+      "epoch": 0.055013351434763484,
+      "grad_norm": 0.8092629313468933,
+      "learning_rate": 4.996434365764314e-05,
+      "loss": 3.4897,
+      "step": 6750
+    },
+    {
+      "epoch": 0.05525785521891799,
+      "grad_norm": 0.917871356010437,
+      "learning_rate": 4.996400061760597e-05,
+      "loss": 3.5188,
+      "step": 6780
+    },
+    {
+      "epoch": 0.055502359003072495,
+      "grad_norm": 0.8477138876914978,
+      "learning_rate": 4.996365593650033e-05,
+      "loss": 3.5142,
+      "step": 6810
+    },
+    {
+      "epoch": 0.055746862787227,
+      "grad_norm": 0.7970532178878784,
+      "learning_rate": 4.99633096143489e-05,
+      "loss": 3.5112,
+      "step": 6840
+    },
+    {
+      "epoch": 0.055991366571381505,
+      "grad_norm": 0.8973976969718933,
+      "learning_rate": 4.9962961651174436e-05,
+      "loss": 3.5156,
+      "step": 6870
+    },
+    {
+      "epoch": 0.05623587035553601,
+      "grad_norm": 0.962066113948822,
+      "learning_rate": 4.9962612046999827e-05,
+      "loss": 3.4848,
+      "step": 6900
+    },
+    {
+      "epoch": 0.056480374139690516,
+      "grad_norm": 0.9020524621009827,
+      "learning_rate": 4.996226080184803e-05,
+      "loss": 3.503,
+      "step": 6930
+    },
+    {
+      "epoch": 0.056724877923845014,
+      "grad_norm": 0.8064045906066895,
+      "learning_rate": 4.996190791574215e-05,
+      "loss": 3.4982,
+      "step": 6960
+    },
+    {
+      "epoch": 0.05696938170799952,
+      "grad_norm": 0.8507282733917236,
+      "learning_rate": 4.996155338870538e-05,
+      "loss": 3.4713,
+      "step": 6990
+    },
+    {
+      "epoch": 0.057213885492154025,
+      "grad_norm": 0.8574703931808472,
+      "learning_rate": 4.9961197220761035e-05,
+      "loss": 3.4924,
+      "step": 7020
+    },
+    {
+      "epoch": 0.05745838927630853,
+      "grad_norm": 0.8214017152786255,
+      "learning_rate": 4.996083941193252e-05,
+      "loss": 3.4773,
+      "step": 7050
+    },
+    {
+      "epoch": 0.057702893060463036,
+      "grad_norm": 0.8267714977264404,
+      "learning_rate": 4.9960479962243367e-05,
+      "loss": 3.4836,
+      "step": 7080
+    },
+    {
+      "epoch": 0.05794739684461754,
+      "grad_norm": 0.8415769934654236,
+      "learning_rate": 4.996011887171719e-05,
+      "loss": 3.4871,
+      "step": 7110
+    },
+    {
+      "epoch": 0.058191900628772046,
+      "grad_norm": 0.9093459248542786,
+      "learning_rate": 4.995975614037773e-05,
+      "loss": 3.4777,
+      "step": 7140
+    },
+    {
+      "epoch": 0.05843640441292655,
+      "grad_norm": 0.865168571472168,
+      "learning_rate": 4.995939176824883e-05,
+      "loss": 3.4798,
+      "step": 7170
+    },
+    {
+      "epoch": 0.05868090819708105,
+      "grad_norm": 0.8812352418899536,
+      "learning_rate": 4.995902575535446e-05,
+      "loss": 3.4634,
+      "step": 7200
+    },
+    {
+      "epoch": 0.058925411981235555,
+      "grad_norm": 0.8318623900413513,
+      "learning_rate": 4.995865810171866e-05,
+      "loss": 3.4691,
+      "step": 7230
+    },
+    {
+      "epoch": 0.05916991576539006,
+      "grad_norm": 0.8475897312164307,
+      "learning_rate": 4.995828880736561e-05,
+      "loss": 3.4894,
+      "step": 7260
+    },
+    {
+      "epoch": 0.059414419549544566,
+      "grad_norm": 0.8317278027534485,
+      "learning_rate": 4.995791787231958e-05,
+      "loss": 3.4597,
+      "step": 7290
+    },
+    {
+      "epoch": 0.05965892333369907,
+      "grad_norm": 0.8189125657081604,
+      "learning_rate": 4.9957545296604965e-05,
+      "loss": 3.4495,
+      "step": 7320
+    },
+    {
+      "epoch": 0.059903427117853576,
+      "grad_norm": 0.8047211766242981,
+      "learning_rate": 4.9957171080246245e-05,
+      "loss": 3.4865,
+      "step": 7350
+    },
+    {
+      "epoch": 0.06014793090200808,
+      "grad_norm": 0.8353013396263123,
+      "learning_rate": 4.995679522326803e-05,
+      "loss": 3.4671,
+      "step": 7380
+    },
+    {
+      "epoch": 0.06039243468616259,
+      "grad_norm": 0.7747114300727844,
+      "learning_rate": 4.995641772569502e-05,
+      "loss": 3.4423,
+      "step": 7410
+    },
+    {
+      "epoch": 0.060636938470317085,
+      "grad_norm": 0.792785108089447,
+      "learning_rate": 4.995603858755203e-05,
+      "loss": 3.448,
+      "step": 7440
+    },
+    {
+      "epoch": 0.06088144225447159,
+      "grad_norm": 0.9009000658988953,
+      "learning_rate": 4.9955657808863985e-05,
+      "loss": 3.4471,
+      "step": 7470
+    },
+    {
+      "epoch": 0.061125946038626096,
+      "grad_norm": 0.8719607591629028,
+      "learning_rate": 4.995527538965593e-05,
+      "loss": 3.4512,
+      "step": 7500
+    },
+    {
+      "epoch": 0.0613704498227806,
+      "grad_norm": 0.840308666229248,
+      "learning_rate": 4.995489132995298e-05,
+      "loss": 3.4576,
+      "step": 7530
+    },
+    {
+      "epoch": 0.061614953606935106,
+      "grad_norm": 0.8216726779937744,
+      "learning_rate": 4.99545056297804e-05,
+      "loss": 3.4369,
+      "step": 7560
+    },
+    {
+      "epoch": 0.06185945739108961,
+      "grad_norm": 0.7841954827308655,
+      "learning_rate": 4.995411828916354e-05,
+      "loss": 3.4338,
+      "step": 7590
+    },
+    {
+      "epoch": 0.06210396117524412,
+      "grad_norm": 0.7832449674606323,
+      "learning_rate": 4.9953729308127874e-05,
+      "loss": 3.431,
+      "step": 7620
+    },
+    {
+      "epoch": 0.062348464959398615,
+      "grad_norm": 0.821207582950592,
+      "learning_rate": 4.995333868669895e-05,
+      "loss": 3.44,
+      "step": 7650
+    },
+    {
+      "epoch": 0.06259296874355312,
+      "grad_norm": 0.7705381512641907,
+      "learning_rate": 4.995294642490246e-05,
+      "loss": 3.4479,
+      "step": 7680
+    },
+    {
+      "epoch": 0.06283747252770763,
+      "grad_norm": 0.7850989699363708,
+      "learning_rate": 4.995255252276418e-05,
+      "loss": 3.4276,
+      "step": 7710
+    },
+    {
+      "epoch": 0.06308197631186213,
+      "grad_norm": 0.7798038721084595,
+      "learning_rate": 4.9952156980310016e-05,
+      "loss": 3.4376,
+      "step": 7740
+    },
+    {
+      "epoch": 0.06332648009601663,
+      "grad_norm": 0.8357362151145935,
+      "learning_rate": 4.9951759797565965e-05,
+      "loss": 3.4408,
+      "step": 7770
+    },
+    {
+      "epoch": 0.06357098388017114,
+      "grad_norm": 0.8627969622612,
+      "learning_rate": 4.995136097455815e-05,
+      "loss": 3.4328,
+      "step": 7800
+    },
+    {
+      "epoch": 0.06381548766432564,
+      "grad_norm": 0.7546987533569336,
+      "learning_rate": 4.995096051131276e-05,
+      "loss": 3.4168,
+      "step": 7830
+    },
+    {
+      "epoch": 0.06405999144848015,
+      "grad_norm": 0.8405272960662842,
+      "learning_rate": 4.995055840785614e-05,
+      "loss": 3.402,
+      "step": 7860
+    },
+    {
+      "epoch": 0.06430449523263465,
+      "grad_norm": 0.8333033919334412,
+      "learning_rate": 4.995015466421473e-05,
+      "loss": 3.4333,
+      "step": 7890
+    },
+    {
+      "epoch": 0.06454899901678916,
+      "grad_norm": 0.787031352519989,
+      "learning_rate": 4.9949749280415056e-05,
+      "loss": 3.4065,
+      "step": 7920
+    },
+    {
+      "epoch": 0.06479350280094366,
+      "grad_norm": 0.7898181676864624,
+      "learning_rate": 4.9949342256483766e-05,
+      "loss": 3.4023,
+      "step": 7950
+    },
+    {
+      "epoch": 0.06503800658509817,
+      "grad_norm": 0.7997327446937561,
+      "learning_rate": 4.9948933592447636e-05,
+      "loss": 3.4268,
+      "step": 7980
+    },
+    {
+      "epoch": 0.06528251036925267,
+      "grad_norm": 0.8030378818511963,
+      "learning_rate": 4.9948523288333506e-05,
+      "loss": 3.4106,
+      "step": 8010
+    },
+    {
+      "epoch": 0.06552701415340717,
+      "grad_norm": 0.8793920278549194,
+      "learning_rate": 4.994811134416836e-05,
+      "loss": 3.4075,
+      "step": 8040
+    },
+    {
+      "epoch": 0.06577151793756168,
+      "grad_norm": 0.810756504535675,
+      "learning_rate": 4.994769775997927e-05,
+      "loss": 3.4177,
+      "step": 8070
+    },
+    {
+      "epoch": 0.06601602172171618,
+      "grad_norm": 0.8193936347961426,
+      "learning_rate": 4.994728253579345e-05,
+      "loss": 3.4195,
+      "step": 8100
+    },
+    {
+      "epoch": 0.0662605255058707,
+      "grad_norm": 0.8704981803894043,
+      "learning_rate": 4.9946865671638166e-05,
+      "loss": 3.3981,
+      "step": 8130
+    },
+    {
+      "epoch": 0.06650502929002519,
+      "grad_norm": 0.9237441420555115,
+      "learning_rate": 4.9946447167540835e-05,
+      "loss": 3.4177,
+      "step": 8160
+    },
+    {
+      "epoch": 0.0667495330741797,
+      "grad_norm": 0.7941862344741821,
+      "learning_rate": 4.994602702352896e-05,
+      "loss": 3.4252,
+      "step": 8190
+    },
+    {
+      "epoch": 0.0669940368583342,
+      "grad_norm": 0.8230422735214233,
+      "learning_rate": 4.994560523963018e-05,
+      "loss": 3.3915,
+      "step": 8220
+    },
+    {
+      "epoch": 0.0672385406424887,
+      "grad_norm": 0.7928892970085144,
+      "learning_rate": 4.9945181815872196e-05,
+      "loss": 3.3738,
+      "step": 8250
+    },
+    {
+      "epoch": 0.06748304442664321,
+      "grad_norm": 0.798755407333374,
+      "learning_rate": 4.9944756752282855e-05,
+      "loss": 3.3991,
+      "step": 8280
+    },
+    {
+      "epoch": 0.06772754821079771,
+      "grad_norm": 0.744189977645874,
+      "learning_rate": 4.994433004889011e-05,
+      "loss": 3.3874,
+      "step": 8310
+    },
+    {
+      "epoch": 0.06797205199495222,
+      "grad_norm": 0.7570254802703857,
+      "learning_rate": 4.994390170572199e-05,
+      "loss": 3.3991,
+      "step": 8340
+    },
+    {
+      "epoch": 0.06821655577910672,
+      "grad_norm": 0.7630516886711121,
+      "learning_rate": 4.994347172280667e-05,
+      "loss": 3.38,
+      "step": 8370
+    },
+    {
+      "epoch": 0.06846105956326123,
+      "grad_norm": 0.8707906603813171,
+      "learning_rate": 4.994304010017241e-05,
+      "loss": 3.3794,
+      "step": 8400
+    },
+    {
+      "epoch": 0.06870556334741573,
+      "grad_norm": 0.7646734714508057,
+      "learning_rate": 4.994260683784758e-05,
+      "loss": 3.4037,
+      "step": 8430
+    },
+    {
+      "epoch": 0.06895006713157023,
+      "grad_norm": 0.8288954496383667,
+      "learning_rate": 4.9942171935860674e-05,
+      "loss": 3.3847,
+      "step": 8460
+    },
+    {
+      "epoch": 0.06919457091572474,
+      "grad_norm": 0.7819532752037048,
+      "learning_rate": 4.994173539424026e-05,
+      "loss": 3.3854,
+      "step": 8490
+    },
+    {
+      "epoch": 0.06943907469987924,
+      "grad_norm": 0.7958394289016724,
+      "learning_rate": 4.994129721301506e-05,
+      "loss": 3.3937,
+      "step": 8520
+    },
+    {
+      "epoch": 0.06968357848403375,
+      "grad_norm": 0.8431640267372131,
+      "learning_rate": 4.994085739221386e-05,
+      "loss": 3.4072,
+      "step": 8550
+    },
+    {
+      "epoch": 0.06992808226818825,
+      "grad_norm": 0.7959804534912109,
+      "learning_rate": 4.994041593186558e-05,
+      "loss": 3.3662,
+      "step": 8580
+    },
+    {
+      "epoch": 0.07017258605234276,
+      "grad_norm": 0.7684959769248962,
+      "learning_rate": 4.993997283199924e-05,
+      "loss": 3.395,
+      "step": 8610
+    },
+    {
+      "epoch": 0.07041708983649726,
+      "grad_norm": 0.7869526147842407,
+      "learning_rate": 4.993952809264397e-05,
+      "loss": 3.3903,
+      "step": 8640
+    },
+    {
+      "epoch": 0.07066159362065176,
+      "grad_norm": 0.8075766563415527,
+      "learning_rate": 4.9939081713829006e-05,
+      "loss": 3.3732,
+      "step": 8670
+    },
+    {
+      "epoch": 0.07090609740480627,
+      "grad_norm": 0.7754388451576233,
+      "learning_rate": 4.993863369558369e-05,
+      "loss": 3.3836,
+      "step": 8700
+    },
+    {
+      "epoch": 0.07115060118896077,
+      "grad_norm": 0.7476597428321838,
+      "learning_rate": 4.9938184037937466e-05,
+      "loss": 3.3767,
+      "step": 8730
+    },
+    {
+      "epoch": 0.07139510497311528,
+      "grad_norm": 0.7789965271949768,
+      "learning_rate": 4.993773274091991e-05,
+      "loss": 3.3518,
+      "step": 8760
+    },
+    {
+      "epoch": 0.07163960875726978,
+      "grad_norm": 0.8229824900627136,
+      "learning_rate": 4.993727980456067e-05,
+      "loss": 3.3629,
+      "step": 8790
+    },
+    {
+      "epoch": 0.0718841125414243,
+      "grad_norm": 0.762668251991272,
+      "learning_rate": 4.993682522888954e-05,
+      "loss": 3.375,
+      "step": 8820
+    },
+    {
+      "epoch": 0.07212861632557879,
+      "grad_norm": 0.8019202947616577,
+      "learning_rate": 4.993636901393639e-05,
+      "loss": 3.3635,
+      "step": 8850
+    },
+    {
+      "epoch": 0.0723731201097333,
+      "grad_norm": 0.9177828431129456,
+      "learning_rate": 4.993591115973121e-05,
+      "loss": 3.37,
+      "step": 8880
+    },
+    {
+      "epoch": 0.0726176238938878,
+      "grad_norm": 0.8312643766403198,
+      "learning_rate": 4.9935451666304105e-05,
+      "loss": 3.3763,
+      "step": 8910
+    },
+    {
+      "epoch": 0.0728621276780423,
+      "grad_norm": 0.828384518623352,
+      "learning_rate": 4.993499053368528e-05,
+      "loss": 3.368,
+      "step": 8940
+    },
+    {
+      "epoch": 0.07310663146219681,
+      "grad_norm": 1.0234721899032593,
+      "learning_rate": 4.993452776190504e-05,
+      "loss": 3.3534,
+      "step": 8970
+    },
+    {
+      "epoch": 0.07335113524635131,
+      "grad_norm": 0.8042975068092346,
+      "learning_rate": 4.993406335099382e-05,
+      "loss": 3.3677,
+      "step": 9000
+    },
+    {
+      "epoch": 0.07359563903050582,
+      "grad_norm": 0.753710150718689,
+      "learning_rate": 4.993359730098214e-05,
+      "loss": 3.3652,
+      "step": 9030
+    },
+    {
+      "epoch": 0.07384014281466032,
+      "grad_norm": 0.8520479798316956,
+      "learning_rate": 4.993312961190064e-05,
+      "loss": 3.3511,
+      "step": 9060
+    },
+    {
+      "epoch": 0.07408464659881484,
+      "grad_norm": 0.7559372186660767,
+      "learning_rate": 4.993266028378006e-05,
+      "loss": 3.3438,
+      "step": 9090
+    },
+    {
+      "epoch": 0.07432915038296933,
+      "grad_norm": 0.7419788837432861,
+      "learning_rate": 4.993218931665126e-05,
+      "loss": 3.3453,
+      "step": 9120
+    },
+    {
+      "epoch": 0.07457365416712383,
+      "grad_norm": 0.8156657218933105,
+      "learning_rate": 4.993171671054519e-05,
+      "loss": 3.3654,
+      "step": 9150
+    },
+    {
+      "epoch": 0.07481815795127834,
+      "grad_norm": 0.8189958930015564,
+      "learning_rate": 4.993124246549293e-05,
+      "loss": 3.3534,
+      "step": 9180
+    },
+    {
+      "epoch": 0.07506266173543284,
+      "grad_norm": 0.7935733795166016,
+      "learning_rate": 4.9930766581525645e-05,
+      "loss": 3.3472,
+      "step": 9210
+    },
+    {
+      "epoch": 0.07530716551958735,
+      "grad_norm": 0.823512852191925,
+      "learning_rate": 4.993028905867463e-05,
+      "loss": 3.3476,
+      "step": 9240
+    },
+    {
+      "epoch": 0.07555166930374185,
+      "grad_norm": 0.8196141719818115,
+      "learning_rate": 4.992980989697126e-05,
+      "loss": 3.3479,
+      "step": 9270
+    },
+    {
+      "epoch": 0.07579617308789637,
+      "grad_norm": 0.7773461937904358,
+      "learning_rate": 4.992932909644705e-05,
+      "loss": 3.344,
+      "step": 9300
+    },
+    {
+      "epoch": 0.07604067687205086,
+      "grad_norm": 0.7703831791877747,
+      "learning_rate": 4.9928846657133596e-05,
+      "loss": 3.3495,
+      "step": 9330
+    },
+    {
+      "epoch": 0.07628518065620536,
+      "grad_norm": 0.7672068476676941,
+      "learning_rate": 4.992836257906262e-05,
+      "loss": 3.3607,
+      "step": 9360
+    },
+    {
+      "epoch": 0.07652968444035987,
+      "grad_norm": 0.7359826564788818,
+      "learning_rate": 4.9927876862265935e-05,
+      "loss": 3.3213,
+      "step": 9390
+    },
+    {
+      "epoch": 0.07677418822451437,
+      "grad_norm": 0.8856120705604553,
+      "learning_rate": 4.992738950677548e-05,
+      "loss": 3.3132,
+      "step": 9420
+    },
+    {
+      "epoch": 0.07701869200866888,
+      "grad_norm": 0.7555311918258667,
+      "learning_rate": 4.992690051262329e-05,
+      "loss": 3.3171,
+      "step": 9450
+    },
+    {
+      "epoch": 0.07726319579282338,
+      "grad_norm": 0.8101349472999573,
+      "learning_rate": 4.99264098798415e-05,
+      "loss": 3.3356,
+      "step": 9480
+    },
+    {
+      "epoch": 0.0775076995769779,
+      "grad_norm": 0.7647647857666016,
+      "learning_rate": 4.992591760846238e-05,
+      "loss": 3.318,
+      "step": 9510
+    },
+    {
+      "epoch": 0.0777522033611324,
+      "grad_norm": 0.7218281030654907,
+      "learning_rate": 4.9925423698518274e-05,
+      "loss": 3.3136,
+      "step": 9540
+    },
+    {
+      "epoch": 0.0779967071452869,
+      "grad_norm": 0.8410701155662537,
+      "learning_rate": 4.992492815004166e-05,
+      "loss": 3.3436,
+      "step": 9570
+    },
+    {
+      "epoch": 0.0782412109294414,
+      "grad_norm": 0.7226998209953308,
+      "learning_rate": 4.992443096306512e-05,
+      "loss": 3.3171,
+      "step": 9600
+    },
+    {
+      "epoch": 0.0784857147135959,
+      "grad_norm": 0.7702776193618774,
+      "learning_rate": 4.992393213762132e-05,
+      "loss": 3.3512,
+      "step": 9630
+    },
+    {
+      "epoch": 0.07873021849775041,
+      "grad_norm": 0.816752016544342,
+      "learning_rate": 4.992343167374307e-05,
+      "loss": 3.3189,
+      "step": 9660
+    },
+    {
+      "epoch": 0.07897472228190491,
+      "grad_norm": 0.854523241519928,
+      "learning_rate": 4.992292957146326e-05,
+      "loss": 3.315,
+      "step": 9690
+    },
+    {
+      "epoch": 0.07921922606605943,
+      "grad_norm": 0.7951149940490723,
+      "learning_rate": 4.992242583081489e-05,
+      "loss": 3.314,
+      "step": 9720
+    },
+    {
+      "epoch": 0.07946372985021392,
+      "grad_norm": 0.7649703621864319,
+      "learning_rate": 4.992192045183109e-05,
+      "loss": 3.2931,
+      "step": 9750
+    },
+    {
+      "epoch": 0.07970823363436844,
+      "grad_norm": 0.8076454401016235,
+      "learning_rate": 4.9921413434545075e-05,
+      "loss": 3.2975,
+      "step": 9780
+    },
+    {
+      "epoch": 0.07995273741852293,
+      "grad_norm": 0.7618278861045837,
+      "learning_rate": 4.992090477899018e-05,
+      "loss": 3.2909,
+      "step": 9810
+    },
+    {
+      "epoch": 0.08019724120267743,
+      "grad_norm": 0.748193085193634,
+      "learning_rate": 4.992039448519982e-05,
+      "loss": 3.2894,
+      "step": 9840
+    },
+    {
+      "epoch": 0.08044174498683195,
+      "grad_norm": 0.7488676309585571,
+      "learning_rate": 4.9919882553207566e-05,
+      "loss": 3.2905,
+      "step": 9870
+    },
+    {
+      "epoch": 0.08068624877098644,
+      "grad_norm": 0.7544564604759216,
+      "learning_rate": 4.9919368983047066e-05,
+      "loss": 3.2983,
+      "step": 9900
+    },
+    {
+      "epoch": 0.08093075255514096,
+      "grad_norm": 0.7570173144340515,
+      "learning_rate": 4.9918853774752074e-05,
+      "loss": 3.3151,
+      "step": 9930
+    },
+    {
+      "epoch": 0.08117525633929545,
+      "grad_norm": 0.7475308775901794,
+      "learning_rate": 4.991833692835646e-05,
+      "loss": 3.2878,
+      "step": 9960
+    },
+    {
+      "epoch": 0.08141976012344997,
+      "grad_norm": 0.858024001121521,
+      "learning_rate": 4.9917818443894203e-05,
+      "loss": 3.298,
+      "step": 9990
+    },
+    {
+      "epoch": 0.08166426390760446,
+      "grad_norm": 0.9019065499305725,
+      "learning_rate": 4.991729832139939e-05,
+      "loss": 3.2961,
+      "step": 10020
+    },
+    {
+      "epoch": 0.08190876769175896,
+      "grad_norm": 0.757036030292511,
+      "learning_rate": 4.991677656090621e-05,
+      "loss": 3.2939,
+      "step": 10050
+    },
+    {
+      "epoch": 0.08215327147591348,
+      "grad_norm": 0.7498080730438232,
+      "learning_rate": 4.991625316244896e-05,
+      "loss": 3.3187,
+      "step": 10080
+    },
+    {
+      "epoch": 0.08239777526006797,
+      "grad_norm": 0.7562547922134399,
+      "learning_rate": 4.991572812606205e-05,
+      "loss": 3.3113,
+      "step": 10110
+    },
+    {
+      "epoch": 0.08264227904422249,
+      "grad_norm": 0.7378952503204346,
+      "learning_rate": 4.991520145177998e-05,
+      "loss": 3.3021,
+      "step": 10140
+    },
+    {
+      "epoch": 0.08288678282837698,
+      "grad_norm": 0.7791491150856018,
+      "learning_rate": 4.991467313963739e-05,
+      "loss": 3.2856,
+      "step": 10170
+    },
+    {
+      "epoch": 0.0831312866125315,
+      "grad_norm": 0.7295765280723572,
+      "learning_rate": 4.991414318966901e-05,
+      "loss": 3.2815,
+      "step": 10200
+    },
+    {
+      "epoch": 0.083375790396686,
+      "grad_norm": 0.7607465982437134,
+      "learning_rate": 4.991361160190966e-05,
+      "loss": 3.2709,
+      "step": 10230
+    },
+    {
+      "epoch": 0.08362029418084051,
+      "grad_norm": 0.7715272903442383,
+      "learning_rate": 4.9913078376394304e-05,
+      "loss": 3.3011,
+      "step": 10260
+    },
+    {
+      "epoch": 0.083864797964995,
+      "grad_norm": 0.7366231679916382,
+      "learning_rate": 4.991254351315799e-05,
+      "loss": 3.2798,
+      "step": 10290
+    },
+    {
+      "epoch": 0.0841093017491495,
+      "grad_norm": 0.7520325183868408,
+      "learning_rate": 4.991200701223587e-05,
+      "loss": 3.2925,
+      "step": 10320
+    },
+    {
+      "epoch": 0.08435380553330402,
+      "grad_norm": 0.7658284902572632,
+      "learning_rate": 4.991146887366323e-05,
+      "loss": 3.2954,
+      "step": 10350
+    },
+    {
+      "epoch": 0.08459830931745851,
+      "grad_norm": 0.7013542056083679,
+      "learning_rate": 4.991092909747542e-05,
+      "loss": 3.2897,
+      "step": 10380
+    },
+    {
+      "epoch": 0.08484281310161303,
+      "grad_norm": 0.8035088181495667,
+      "learning_rate": 4.9910387683707946e-05,
+      "loss": 3.2876,
+      "step": 10410
+    },
+    {
+      "epoch": 0.08508731688576752,
+      "grad_norm": 0.7619040608406067,
+      "learning_rate": 4.9909844632396386e-05,
+      "loss": 3.278,
+      "step": 10440
+    },
+    {
+      "epoch": 0.08533182066992204,
+      "grad_norm": 0.7567503452301025,
+      "learning_rate": 4.9909299943576445e-05,
+      "loss": 3.2774,
+      "step": 10470
+    },
+    {
+      "epoch": 0.08557632445407654,
+      "grad_norm": 0.7545759677886963,
+      "learning_rate": 4.990875361728393e-05,
+      "loss": 3.2824,
+      "step": 10500
+    },
+    {
+      "epoch": 0.08582082823823103,
+      "grad_norm": 0.7955304980278015,
+      "learning_rate": 4.990820565355475e-05,
+      "loss": 3.2809,
+      "step": 10530
+    },
+    {
+      "epoch": 0.08606533202238555,
+      "grad_norm": 0.8105009198188782,
+      "learning_rate": 4.990765605242493e-05,
+      "loss": 3.2745,
+      "step": 10560
+    },
+    {
+      "epoch": 0.08630983580654004,
+      "grad_norm": 0.758289098739624,
+      "learning_rate": 4.990710481393061e-05,
+      "loss": 3.2549,
+      "step": 10590
+    },
+    {
+      "epoch": 0.08655433959069456,
+      "grad_norm": 0.7958084344863892,
+      "learning_rate": 4.9906551938108003e-05,
+      "loss": 3.3004,
+      "step": 10620
+    },
+    {
+      "epoch": 0.08679884337484906,
+      "grad_norm": 0.7208577394485474,
+      "learning_rate": 4.990599742499347e-05,
+      "loss": 3.291,
+      "step": 10650
+    },
+    {
+      "epoch": 0.08704334715900357,
+      "grad_norm": 0.7224754691123962,
+      "learning_rate": 4.990544127462346e-05,
+      "loss": 3.2676,
+      "step": 10680
+    },
+    {
+      "epoch": 0.08728785094315807,
+      "grad_norm": 0.7175268530845642,
+      "learning_rate": 4.9904883487034537e-05,
+      "loss": 3.2753,
+      "step": 10710
+    },
+    {
+      "epoch": 0.08753235472731256,
+      "grad_norm": 0.75492262840271,
+      "learning_rate": 4.990432406226336e-05,
+      "loss": 3.2707,
+      "step": 10740
+    },
+    {
+      "epoch": 0.08777685851146708,
+      "grad_norm": 0.7345853447914124,
+      "learning_rate": 4.990376300034671e-05,
+      "loss": 3.2593,
+      "step": 10770
+    },
+    {
+      "epoch": 0.08802136229562157,
+      "grad_norm": 0.719987154006958,
+      "learning_rate": 4.990320030132147e-05,
+      "loss": 3.2689,
+      "step": 10800
+    },
+    {
+      "epoch": 0.08826586607977609,
+      "grad_norm": 0.7458385229110718,
+      "learning_rate": 4.990263596522462e-05,
+      "loss": 3.2822,
+      "step": 10830
+    },
+    {
+      "epoch": 0.08851036986393059,
+      "grad_norm": 0.7266592383384705,
+      "learning_rate": 4.9902069992093275e-05,
+      "loss": 3.2818,
+      "step": 10860
+    },
+    {
+      "epoch": 0.0887548736480851,
+      "grad_norm": 0.6844702363014221,
+      "learning_rate": 4.990150238196463e-05,
+      "loss": 3.2711,
+      "step": 10890
+    },
+    {
+      "epoch": 0.0889993774322396,
+      "grad_norm": 0.7185301184654236,
+      "learning_rate": 4.9900933134876e-05,
+      "loss": 3.2682,
+      "step": 10920
+    },
+    {
+      "epoch": 0.0892438812163941,
+      "grad_norm": 0.7460073232650757,
+      "learning_rate": 4.990036225086481e-05,
+      "loss": 3.2419,
+      "step": 10950
+    },
+    {
+      "epoch": 0.0894883850005486,
+      "grad_norm": 0.7382895946502686,
+      "learning_rate": 4.9899789729968585e-05,
+      "loss": 3.2612,
+      "step": 10980
+    },
+    {
+      "epoch": 0.0897328887847031,
+      "grad_norm": 0.7204514741897583,
+      "learning_rate": 4.989921557222496e-05,
+      "loss": 3.2525,
+      "step": 11010
+    },
+    {
+      "epoch": 0.08997739256885762,
+      "grad_norm": 0.7512257695198059,
+      "learning_rate": 4.989863977767167e-05,
+      "loss": 3.2656,
+      "step": 11040
+    },
+    {
+      "epoch": 0.09022189635301212,
+      "grad_norm": 0.7579988837242126,
+      "learning_rate": 4.989806234634659e-05,
+      "loss": 3.2545,
+      "step": 11070
+    },
+    {
+      "epoch": 0.09046640013716663,
+      "grad_norm": 0.7989354729652405,
+      "learning_rate": 4.989748327828765e-05,
+      "loss": 3.2588,
+      "step": 11100
+    },
+    {
+      "epoch": 0.09071090392132113,
+      "grad_norm": 0.7362430095672607,
+      "learning_rate": 4.989690257353294e-05,
+      "loss": 3.2205,
+      "step": 11130
+    },
+    {
+      "epoch": 0.09095540770547564,
+      "grad_norm": 0.6724596619606018,
+      "learning_rate": 4.989632023212062e-05,
+      "loss": 3.252,
+      "step": 11160
+    },
+    {
+      "epoch": 0.09119991148963014,
+      "grad_norm": 0.804040789604187,
+      "learning_rate": 4.9895736254088975e-05,
+      "loss": 3.2554,
+      "step": 11190
+    },
+    {
+      "epoch": 0.09144441527378463,
+      "grad_norm": 0.7371083498001099,
+      "learning_rate": 4.989515063947641e-05,
+      "loss": 3.2551,
+      "step": 11220
+    },
+    {
+      "epoch": 0.09168891905793915,
+      "grad_norm": 0.8144158124923706,
+      "learning_rate": 4.9894563388321395e-05,
+      "loss": 3.2396,
+      "step": 11250
+    },
+    {
+      "epoch": 0.09193342284209365,
+      "grad_norm": 0.7379621863365173,
+      "learning_rate": 4.989397450066254e-05,
+      "loss": 3.2588,
+      "step": 11280
+    },
+    {
+      "epoch": 0.09217792662624816,
+      "grad_norm": 0.7908419370651245,
+      "learning_rate": 4.989338397653858e-05,
+      "loss": 3.2316,
+      "step": 11310
+    },
+    {
+      "epoch": 0.09242243041040266,
+      "grad_norm": 0.763767421245575,
+      "learning_rate": 4.98927918159883e-05,
+      "loss": 3.2585,
+      "step": 11340
+    },
+    {
+      "epoch": 0.09266693419455717,
+      "grad_norm": 0.7726889252662659,
+      "learning_rate": 4.989219801905066e-05,
+      "loss": 3.2481,
+      "step": 11370
+    },
+    {
+      "epoch": 0.09291143797871167,
+      "grad_norm": 0.7199206352233887,
+      "learning_rate": 4.989160258576469e-05,
+      "loss": 3.2468,
+      "step": 11400
+    },
+    {
+      "epoch": 0.09315594176286617,
+      "grad_norm": 0.7201483249664307,
+      "learning_rate": 4.98910055161695e-05,
+      "loss": 3.2226,
+      "step": 11430
+    },
+    {
+      "epoch": 0.09340044554702068,
+      "grad_norm": 0.724740743637085,
+      "learning_rate": 4.989040681030437e-05,
+      "loss": 3.2546,
+      "step": 11460
+    },
+    {
+      "epoch": 0.09364494933117518,
+      "grad_norm": 0.7120088338851929,
+      "learning_rate": 4.988980646820865e-05,
+      "loss": 3.2397,
+      "step": 11490
+    },
+    {
+      "epoch": 0.09388945311532969,
+      "grad_norm": 0.7296812534332275,
+      "learning_rate": 4.9889204489921804e-05,
+      "loss": 3.2201,
+      "step": 11520
+    },
+    {
+      "epoch": 0.09413395689948419,
+      "grad_norm": 0.7710677981376648,
+      "learning_rate": 4.9888600875483404e-05,
+      "loss": 3.2475,
+      "step": 11550
+    },
+    {
+      "epoch": 0.0943784606836387,
+      "grad_norm": 0.7161293029785156,
+      "learning_rate": 4.9887995624933137e-05,
+      "loss": 3.2289,
+      "step": 11580
+    },
+    {
+      "epoch": 0.0946229644677932,
+      "grad_norm": 0.6972730755805969,
+      "learning_rate": 4.988738873831078e-05,
+      "loss": 3.2247,
+      "step": 11610
+    },
+    {
+      "epoch": 0.0948674682519477,
+      "grad_norm": 0.7108049988746643,
+      "learning_rate": 4.988678021565623e-05,
+      "loss": 3.2227,
+      "step": 11640
+    },
+    {
+      "epoch": 0.09511197203610221,
+      "grad_norm": 0.729718029499054,
+      "learning_rate": 4.988617005700949e-05,
+      "loss": 3.2396,
+      "step": 11670
+    },
+    {
+      "epoch": 0.0953564758202567,
+      "grad_norm": 0.6947882771492004,
+      "learning_rate": 4.988555826241068e-05,
+      "loss": 3.238,
+      "step": 11700
+    },
+    {
+      "epoch": 0.09560097960441122,
+      "grad_norm": 0.7448226809501648,
+      "learning_rate": 4.988494483190001e-05,
+      "loss": 3.2274,
+      "step": 11730
+    },
+    {
+      "epoch": 0.09584548338856572,
+      "grad_norm": 0.6681617498397827,
+      "learning_rate": 4.98843297655178e-05,
+      "loss": 3.2504,
+      "step": 11760
+    },
+    {
+      "epoch": 0.09608998717272023,
+      "grad_norm": 0.7603419423103333,
+      "learning_rate": 4.988371306330449e-05,
+      "loss": 3.2206,
+      "step": 11790
+    },
+    {
+      "epoch": 0.09633449095687473,
+      "grad_norm": 0.7187109589576721,
+      "learning_rate": 4.9883094725300625e-05,
+      "loss": 3.2313,
+      "step": 11820
+    },
+    {
+      "epoch": 0.09657899474102924,
+      "grad_norm": 0.7375375628471375,
+      "learning_rate": 4.9882474751546846e-05,
+      "loss": 3.214,
+      "step": 11850
+    },
+    {
+      "epoch": 0.09682349852518374,
+      "grad_norm": 0.7366687059402466,
+      "learning_rate": 4.98818531420839e-05,
+      "loss": 3.2191,
+      "step": 11880
+    },
+    {
+      "epoch": 0.09706800230933824,
+      "grad_norm": 0.690647304058075,
+      "learning_rate": 4.988122989695267e-05,
+      "loss": 3.2378,
+      "step": 11910
+    },
+    {
+      "epoch": 0.09731250609349275,
+      "grad_norm": 0.7315085530281067,
+      "learning_rate": 4.9880605016194116e-05,
+      "loss": 3.2121,
+      "step": 11940
+    },
+    {
+      "epoch": 0.09755700987764725,
+      "grad_norm": 0.7334670424461365,
+      "learning_rate": 4.9879978499849304e-05,
+      "loss": 3.2194,
+      "step": 11970
+    },
+    {
+      "epoch": 0.09780151366180176,
+      "grad_norm": 0.7092124223709106,
+      "learning_rate": 4.987935034795944e-05,
+      "loss": 3.2162,
+      "step": 12000
+    },
+    {
+      "epoch": 0.09804601744595626,
+      "grad_norm": 0.7557071447372437,
+      "learning_rate": 4.9878720560565803e-05,
+      "loss": 3.21,
+      "step": 12030
+    },
+    {
+      "epoch": 0.09829052123011077,
+      "grad_norm": 0.7767185568809509,
+      "learning_rate": 4.9878089137709806e-05,
+      "loss": 3.2221,
+      "step": 12060
+    },
+    {
+      "epoch": 0.09853502501426527,
+      "grad_norm": 0.7004647850990295,
+      "learning_rate": 4.987745607943295e-05,
+      "loss": 3.2333,
+      "step": 12090
+    },
+    {
+      "epoch": 0.09877952879841977,
+      "grad_norm": 0.6746474504470825,
+      "learning_rate": 4.9876821385776854e-05,
+      "loss": 3.2185,
+      "step": 12120
+    },
+    {
+      "epoch": 0.09902403258257428,
+      "grad_norm": 0.7061980962753296,
+      "learning_rate": 4.9876185056783226e-05,
+      "loss": 3.2159,
+      "step": 12150
+    },
+    {
+      "epoch": 0.09926853636672878,
+      "grad_norm": 0.7409046292304993,
+      "learning_rate": 4.987554709249391e-05,
+      "loss": 3.2214,
+      "step": 12180
+    },
+    {
+      "epoch": 0.09951304015088329,
+      "grad_norm": 0.6951879858970642,
+      "learning_rate": 4.987490749295085e-05,
+      "loss": 3.2124,
+      "step": 12210
+    },
+    {
+      "epoch": 0.09975754393503779,
+      "grad_norm": 0.7459700107574463,
+      "learning_rate": 4.9874266258196084e-05,
+      "loss": 3.2074,
+      "step": 12240
+    },
+    {
+      "epoch": 0.1000020477191923,
+      "grad_norm": 0.654021680355072,
+      "learning_rate": 4.9873623388271764e-05,
+      "loss": 3.2199,
+      "step": 12270
+    },
+    {
+      "epoch": 0.1002465515033468,
+      "grad_norm": 0.7051617503166199,
+      "learning_rate": 4.987297888322015e-05,
+      "loss": 3.1998,
+      "step": 12300
+    },
+    {
+      "epoch": 0.1004910552875013,
+      "grad_norm": 0.7212511301040649,
+      "learning_rate": 4.9872332743083615e-05,
+      "loss": 3.2144,
+      "step": 12330
+    },
+    {
+      "epoch": 0.10073555907165581,
+      "grad_norm": 0.6988245844841003,
+      "learning_rate": 4.987168496790463e-05,
+      "loss": 3.2036,
+      "step": 12360
+    },
+    {
+      "epoch": 0.10098006285581031,
+      "grad_norm": 0.7479727864265442,
+      "learning_rate": 4.9871035557725774e-05,
+      "loss": 3.2035,
+      "step": 12390
+    },
+    {
+      "epoch": 0.10122456663996482,
+      "grad_norm": 0.6703733801841736,
+      "learning_rate": 4.987038451258975e-05,
+      "loss": 3.2119,
+      "step": 12420
+    },
+    {
+      "epoch": 0.10146907042411932,
+      "grad_norm": 0.7142158150672913,
+      "learning_rate": 4.9869731832539346e-05,
+      "loss": 3.1866,
+      "step": 12450
+    },
+    {
+      "epoch": 0.10171357420827383,
+      "grad_norm": 0.6590874195098877,
+      "learning_rate": 4.9869077517617474e-05,
+      "loss": 3.1854,
+      "step": 12480
+    },
+    {
+      "epoch": 0.10195807799242833,
+      "grad_norm": 0.7379269599914551,
+      "learning_rate": 4.986842156786714e-05,
+      "loss": 3.1755,
+      "step": 12510
+    },
+    {
+      "epoch": 0.10220258177658284,
+      "grad_norm": 0.7689487934112549,
+      "learning_rate": 4.9867763983331474e-05,
+      "loss": 3.1838,
+      "step": 12540
+    },
+    {
+      "epoch": 0.10244708556073734,
+      "grad_norm": 0.6958329677581787,
+      "learning_rate": 4.9867104764053697e-05,
+      "loss": 3.191,
+      "step": 12570
+    },
+    {
+      "epoch": 0.10269158934489184,
+      "grad_norm": 0.7052045464515686,
+      "learning_rate": 4.986644391007714e-05,
+      "loss": 3.1773,
+      "step": 12600
+    },
+    {
+      "epoch": 0.10293609312904635,
+      "grad_norm": 0.6961473226547241,
+      "learning_rate": 4.986578142144525e-05,
+      "loss": 3.1982,
+      "step": 12630
+    },
+    {
+      "epoch": 0.10318059691320085,
+      "grad_norm": 0.6776019930839539,
+      "learning_rate": 4.986511729820158e-05,
+      "loss": 3.1932,
+      "step": 12660
+    },
+    {
+      "epoch": 0.10342510069735536,
+      "grad_norm": 0.7005933523178101,
+      "learning_rate": 4.986445154038979e-05,
+      "loss": 3.1892,
+      "step": 12690
+    },
+    {
+      "epoch": 0.10366960448150986,
+      "grad_norm": 0.6863512992858887,
+      "learning_rate": 4.986378414805364e-05,
+      "loss": 3.199,
+      "step": 12720
+    },
+    {
+      "epoch": 0.10391410826566437,
+      "grad_norm": 0.7182413339614868,
+      "learning_rate": 4.9863115121237006e-05,
+      "loss": 3.178,
+      "step": 12750
+    },
+    {
+      "epoch": 0.10415861204981887,
+      "grad_norm": 0.7226789593696594,
+      "learning_rate": 4.986244445998386e-05,
+      "loss": 3.197,
+      "step": 12780
+    },
+    {
+      "epoch": 0.10440311583397337,
+      "grad_norm": 0.7314379215240479,
+      "learning_rate": 4.9861772164338304e-05,
+      "loss": 3.184,
+      "step": 12810
+    },
+    {
+      "epoch": 0.10464761961812788,
+      "grad_norm": 0.7286015152931213,
+      "learning_rate": 4.986109823434452e-05,
+      "loss": 3.1846,
+      "step": 12840
+    },
+    {
+      "epoch": 0.10489212340228238,
+      "grad_norm": 0.6875571608543396,
+      "learning_rate": 4.986042267004681e-05,
+      "loss": 3.1965,
+      "step": 12870
+    },
+    {
+      "epoch": 0.10513662718643689,
+      "grad_norm": 0.7378973364830017,
+      "learning_rate": 4.98597454714896e-05,
+      "loss": 3.1946,
+      "step": 12900
+    },
+    {
+      "epoch": 0.10538113097059139,
+      "grad_norm": 0.7709534764289856,
+      "learning_rate": 4.985906663871739e-05,
+      "loss": 3.195,
+      "step": 12930
+    },
+    {
+      "epoch": 0.1056256347547459,
+      "grad_norm": 0.6385577917098999,
+      "learning_rate": 4.9858386171774804e-05,
+      "loss": 3.1958,
+      "step": 12960
+    },
+    {
+      "epoch": 0.1058701385389004,
+      "grad_norm": 0.7177209258079529,
+      "learning_rate": 4.985770407070659e-05,
+      "loss": 3.1669,
+      "step": 12990
+    },
+    {
+      "epoch": 0.1061146423230549,
+      "grad_norm": 0.6712928414344788,
+      "learning_rate": 4.985702033555757e-05,
+      "loss": 3.193,
+      "step": 13020
+    },
+    {
+      "epoch": 0.10635914610720941,
+      "grad_norm": 0.670972466468811,
+      "learning_rate": 4.9856334966372705e-05,
+      "loss": 3.1788,
+      "step": 13050
+    },
+    {
+      "epoch": 0.10660364989136391,
+      "grad_norm": 0.7206739187240601,
+      "learning_rate": 4.985564796319705e-05,
+      "loss": 3.1928,
+      "step": 13080
+    },
+    {
+      "epoch": 0.10684815367551842,
+      "grad_norm": 0.7591316103935242,
+      "learning_rate": 4.985495932607576e-05,
+      "loss": 3.196,
+      "step": 13110
+    },
+    {
+      "epoch": 0.10709265745967292,
+      "grad_norm": 0.6864051222801208,
+      "learning_rate": 4.9854269055054096e-05,
+      "loss": 3.1929,
+      "step": 13140
+    },
+    {
+      "epoch": 0.10733716124382743,
+      "grad_norm": 0.6852788329124451,
+      "learning_rate": 4.985357715017744e-05,
+      "loss": 3.1824,
+      "step": 13170
+    },
+    {
+      "epoch": 0.10758166502798193,
+      "grad_norm": 0.6889312863349915,
+      "learning_rate": 4.985288361149129e-05,
+      "loss": 3.1918,
+      "step": 13200
+    },
+    {
+      "epoch": 0.10782616881213643,
+      "grad_norm": 0.724328875541687,
+      "learning_rate": 4.985218843904122e-05,
+      "loss": 3.1944,
+      "step": 13230
+    },
+    {
+      "epoch": 0.10807067259629094,
+      "grad_norm": 0.7326011061668396,
+      "learning_rate": 4.985149163287294e-05,
+      "loss": 3.1815,
+      "step": 13260
+    },
+    {
+      "epoch": 0.10831517638044544,
+      "grad_norm": 0.7652591466903687,
+      "learning_rate": 4.985079319303225e-05,
+      "loss": 3.1747,
+      "step": 13290
+    },
+    {
+      "epoch": 0.10855968016459995,
+      "grad_norm": 0.7179089188575745,
+      "learning_rate": 4.985009311956507e-05,
+      "loss": 3.1729,
+      "step": 13320
+    },
+    {
+      "epoch": 0.10880418394875445,
+      "grad_norm": 0.6816122531890869,
+      "learning_rate": 4.984939141251741e-05,
+      "loss": 3.1877,
+      "step": 13350
+    },
+    {
+      "epoch": 0.10904868773290896,
+      "grad_norm": 0.660858690738678,
+      "learning_rate": 4.9848688071935415e-05,
+      "loss": 3.1658,
+      "step": 13380
+    },
+    {
+      "epoch": 0.10929319151706346,
+      "grad_norm": 0.6714891195297241,
+      "learning_rate": 4.98479830978653e-05,
+      "loss": 3.1753,
+      "step": 13410
+    },
+    {
+      "epoch": 0.10953769530121797,
+      "grad_norm": 0.6714405417442322,
+      "learning_rate": 4.9847276490353425e-05,
+      "loss": 3.167,
+      "step": 13440
+    },
+    {
+      "epoch": 0.10978219908537247,
+      "grad_norm": 0.669769823551178,
+      "learning_rate": 4.984656824944623e-05,
+      "loss": 3.1433,
+      "step": 13470
+    },
+    {
+      "epoch": 0.11002670286952697,
+      "grad_norm": 0.7423032522201538,
+      "learning_rate": 4.984585837519028e-05,
+      "loss": 3.1587,
+      "step": 13500
+    },
+    {
+      "epoch": 0.11027120665368148,
+      "grad_norm": 0.6712002754211426,
+      "learning_rate": 4.984514686763224e-05,
+      "loss": 3.1472,
+      "step": 13530
+    },
+    {
+      "epoch": 0.11051571043783598,
+      "grad_norm": 0.7179352045059204,
+      "learning_rate": 4.984443372681887e-05,
+      "loss": 3.1537,
+      "step": 13560
+    },
+    {
+      "epoch": 0.11076021422199049,
+      "grad_norm": 0.7070680260658264,
+      "learning_rate": 4.9843718952797074e-05,
+      "loss": 3.1565,
+      "step": 13590
+    },
+    {
+      "epoch": 0.11100471800614499,
+      "grad_norm": 0.7316999435424805,
+      "learning_rate": 4.984300254561382e-05,
+      "loss": 3.1408,
+      "step": 13620
+    },
+    {
+      "epoch": 0.1112492217902995,
+      "grad_norm": 0.7282266616821289,
+      "learning_rate": 4.9842284505316206e-05,
+      "loss": 3.1646,
+      "step": 13650
+    },
+    {
+      "epoch": 0.111493725574454,
+      "grad_norm": 0.6478064060211182,
+      "learning_rate": 4.984156483195144e-05,
+      "loss": 3.134,
+      "step": 13680
+    },
+    {
+      "epoch": 0.1117382293586085,
+      "grad_norm": 0.6633407473564148,
+      "learning_rate": 4.984084352556683e-05,
+      "loss": 3.1501,
+      "step": 13710
+    },
+    {
+      "epoch": 0.11198273314276301,
+      "grad_norm": 0.6864104866981506,
+      "learning_rate": 4.984012058620979e-05,
+      "loss": 3.154,
+      "step": 13740
+    },
+    {
+      "epoch": 0.11222723692691751,
+      "grad_norm": 0.7092240452766418,
+      "learning_rate": 4.983939601392784e-05,
+      "loss": 3.1625,
+      "step": 13770
+    },
+    {
+      "epoch": 0.11247174071107202,
+      "grad_norm": 0.6620602607727051,
+      "learning_rate": 4.9838669808768616e-05,
+      "loss": 3.1681,
+      "step": 13800
+    },
+    {
+      "epoch": 0.11271624449522652,
+      "grad_norm": 0.6808717250823975,
+      "learning_rate": 4.9837941970779865e-05,
+      "loss": 3.1612,
+      "step": 13830
+    },
+    {
+      "epoch": 0.11296074827938103,
+      "grad_norm": 0.7544457316398621,
+      "learning_rate": 4.983721250000942e-05,
+      "loss": 3.1518,
+      "step": 13860
+    },
+    {
+      "epoch": 0.11320525206353553,
+      "grad_norm": 0.6600370407104492,
+      "learning_rate": 4.983648139650524e-05,
+      "loss": 3.1432,
+      "step": 13890
+    },
+    {
+      "epoch": 0.11344975584769003,
+      "grad_norm": 0.6779924035072327,
+      "learning_rate": 4.983574866031538e-05,
+      "loss": 3.1451,
+      "step": 13920
+    },
+    {
+      "epoch": 0.11369425963184454,
+      "grad_norm": 0.7038785815238953,
+      "learning_rate": 4.983501429148802e-05,
+      "loss": 3.1389,
+      "step": 13950
+    },
+    {
+      "epoch": 0.11393876341599904,
+      "grad_norm": 0.7131289839744568,
+      "learning_rate": 4.983427829007144e-05,
+      "loss": 3.1389,
+      "step": 13980
+    },
+    {
+      "epoch": 0.11418326720015355,
+      "grad_norm": 0.6724546551704407,
+      "learning_rate": 4.9833540656114004e-05,
+      "loss": 3.1516,
+      "step": 14010
+    },
+    {
+      "epoch": 0.11442777098430805,
+      "grad_norm": 0.6874014139175415,
+      "learning_rate": 4.983280138966421e-05,
+      "loss": 3.1616,
+      "step": 14040
+    },
+    {
+      "epoch": 0.11467227476846256,
+      "grad_norm": 0.722695529460907,
+      "learning_rate": 4.983206049077065e-05,
+      "loss": 3.1323,
+      "step": 14070
+    },
+    {
+      "epoch": 0.11491677855261706,
+      "grad_norm": 0.7108673453330994,
+      "learning_rate": 4.9831317959482044e-05,
+      "loss": 3.1676,
+      "step": 14100
+    },
+    {
+      "epoch": 0.11516128233677157,
+      "grad_norm": 0.673359215259552,
+      "learning_rate": 4.983057379584719e-05,
+      "loss": 3.1468,
+      "step": 14130
+    },
+    {
+      "epoch": 0.11540578612092607,
+      "grad_norm": 0.7127307653427124,
+      "learning_rate": 4.982982799991501e-05,
+      "loss": 3.1485,
+      "step": 14160
+    },
+    {
+      "epoch": 0.11565028990508057,
+      "grad_norm": 0.698799729347229,
+      "learning_rate": 4.982908057173454e-05,
+      "loss": 3.1342,
+      "step": 14190
+    },
+    {
+      "epoch": 0.11589479368923508,
+      "grad_norm": 0.6260579228401184,
+      "learning_rate": 4.98283315113549e-05,
+      "loss": 3.1381,
+      "step": 14220
+    },
+    {
+      "epoch": 0.11613929747338958,
+      "grad_norm": 0.6754068732261658,
+      "learning_rate": 4.982758081882534e-05,
+      "loss": 3.1409,
+      "step": 14250
+    },
+    {
+      "epoch": 0.11638380125754409,
+      "grad_norm": 0.6846606135368347,
+      "learning_rate": 4.982682849419521e-05,
+      "loss": 3.1422,
+      "step": 14280
+    },
+    {
+      "epoch": 0.11662830504169859,
+      "grad_norm": 0.6726362109184265,
+      "learning_rate": 4.982607453751396e-05,
+      "loss": 3.1339,
+      "step": 14310
+    },
+    {
+      "epoch": 0.1168728088258531,
+      "grad_norm": 0.6798800230026245,
+      "learning_rate": 4.982531894883116e-05,
+      "loss": 3.1571,
+      "step": 14340
+    },
+    {
+      "epoch": 0.1171173126100076,
+      "grad_norm": 0.7045117616653442,
+      "learning_rate": 4.9824561728196465e-05,
+      "loss": 3.1529,
+      "step": 14370
+    },
+    {
+      "epoch": 0.1173618163941621,
+      "grad_norm": 0.7069810032844543,
+      "learning_rate": 4.9823802875659674e-05,
+      "loss": 3.1498,
+      "step": 14400
+    },
+    {
+      "epoch": 0.11760632017831661,
+      "grad_norm": 0.6553217768669128,
+      "learning_rate": 4.9823042391270657e-05,
+      "loss": 3.1188,
+      "step": 14430
+    },
+    {
+      "epoch": 0.11785082396247111,
+      "grad_norm": 0.6937916874885559,
+      "learning_rate": 4.9822280275079416e-05,
+      "loss": 3.1363,
+      "step": 14460
+    },
+    {
+      "epoch": 0.11809532774662562,
+      "grad_norm": 0.6913928389549255,
+      "learning_rate": 4.9821516527136034e-05,
+      "loss": 3.1434,
+      "step": 14490
+    },
+    {
+      "epoch": 0.11833983153078012,
+      "grad_norm": 0.6402164101600647,
+      "learning_rate": 4.982075114749074e-05,
+      "loss": 3.1379,
+      "step": 14520
+    },
+    {
+      "epoch": 0.11858433531493463,
+      "grad_norm": 0.7302058935165405,
+      "learning_rate": 4.9819984136193835e-05,
+      "loss": 3.1501,
+      "step": 14550
+    },
+    {
+      "epoch": 0.11882883909908913,
+      "grad_norm": 0.6861427426338196,
+      "learning_rate": 4.9819215493295746e-05,
+      "loss": 3.1443,
+      "step": 14580
+    },
+    {
+      "epoch": 0.11907334288324363,
+      "grad_norm": 0.6565712094306946,
+      "learning_rate": 4.9818445218846995e-05,
+      "loss": 3.0966,
+      "step": 14610
+    },
+    {
+      "epoch": 0.11931784666739814,
+      "grad_norm": 0.6484554409980774,
+      "learning_rate": 4.9817673312898215e-05,
+      "loss": 3.1578,
+      "step": 14640
+    },
+    {
+      "epoch": 0.11956235045155264,
+      "grad_norm": 0.6579607129096985,
+      "learning_rate": 4.981689977550017e-05,
+      "loss": 3.1177,
+      "step": 14670
+    },
+    {
+      "epoch": 0.11980685423570715,
+      "grad_norm": 0.6820259690284729,
+      "learning_rate": 4.9816124606703684e-05,
+      "loss": 3.1332,
+      "step": 14700
+    },
+    {
+      "epoch": 0.12005135801986165,
+      "grad_norm": 0.6516405940055847,
+      "learning_rate": 4.9815347806559734e-05,
+      "loss": 3.1205,
+      "step": 14730
+    },
+    {
+      "epoch": 0.12029586180401616,
+      "grad_norm": 0.681978166103363,
+      "learning_rate": 4.981456937511937e-05,
+      "loss": 3.1184,
+      "step": 14760
+    },
+    {
+      "epoch": 0.12054036558817066,
+      "grad_norm": 0.6539319753646851,
+      "learning_rate": 4.981378931243377e-05,
+      "loss": 3.1198,
+      "step": 14790
+    },
+    {
+      "epoch": 0.12078486937232517,
+      "grad_norm": 0.6952561140060425,
+      "learning_rate": 4.981300761855422e-05,
+      "loss": 3.1392,
+      "step": 14820
+    },
+    {
+      "epoch": 0.12102937315647967,
+      "grad_norm": 0.6566502451896667,
+      "learning_rate": 4.981222429353211e-05,
+      "loss": 3.141,
+      "step": 14850
+    },
+    {
+      "epoch": 0.12127387694063417,
+      "grad_norm": 0.7198945879936218,
+      "learning_rate": 4.981143933741891e-05,
+      "loss": 3.1337,
+      "step": 14880
+    },
+    {
+      "epoch": 0.12151838072478868,
+      "grad_norm": 0.6870773434638977,
+      "learning_rate": 4.981065275026624e-05,
+      "loss": 3.1529,
+      "step": 14910
+    },
+    {
+      "epoch": 0.12176288450894318,
+      "grad_norm": 0.6778687238693237,
+      "learning_rate": 4.9809864532125797e-05,
+      "loss": 3.1149,
+      "step": 14940
+    },
+    {
+      "epoch": 0.1220073882930977,
+      "grad_norm": 0.6705347895622253,
+      "learning_rate": 4.9809074683049415e-05,
+      "loss": 3.1061,
+      "step": 14970
+    },
+    {
+      "epoch": 0.12225189207725219,
+      "grad_norm": 0.662770688533783,
+      "learning_rate": 4.9808283203089e-05,
+      "loss": 3.1039,
+      "step": 15000
+    },
+    {
+      "epoch": 0.1224963958614067,
+      "grad_norm": 0.680558443069458,
+      "learning_rate": 4.9807490092296586e-05,
+      "loss": 3.1306,
+      "step": 15030
+    },
+    {
+      "epoch": 0.1227408996455612,
+      "grad_norm": 0.692523717880249,
+      "learning_rate": 4.980669535072431e-05,
+      "loss": 3.1291,
+      "step": 15060
+    },
+    {
+      "epoch": 0.1229854034297157,
+      "grad_norm": 0.6542308926582336,
+      "learning_rate": 4.980589897842441e-05,
+      "loss": 3.1058,
+      "step": 15090
+    },
+    {
+      "epoch": 0.12322990721387021,
+      "grad_norm": 0.6776884198188782,
+      "learning_rate": 4.9805100975449256e-05,
+      "loss": 3.0983,
+      "step": 15120
+    },
+    {
+      "epoch": 0.12347441099802471,
+      "grad_norm": 0.6549819111824036,
+      "learning_rate": 4.98043013418513e-05,
+      "loss": 3.1171,
+      "step": 15150
+    },
+    {
+      "epoch": 0.12371891478217922,
+      "grad_norm": 0.6724957823753357,
+      "learning_rate": 4.980350007768309e-05,
+      "loss": 3.107,
+      "step": 15180
+    },
+    {
+      "epoch": 0.12396341856633372,
+      "grad_norm": 0.7151206731796265,
+      "learning_rate": 4.9802697182997316e-05,
+      "loss": 3.1308,
+      "step": 15210
+    },
+    {
+      "epoch": 0.12420792235048823,
+      "grad_norm": 0.6585493683815002,
+      "learning_rate": 4.980189265784676e-05,
+      "loss": 3.0973,
+      "step": 15240
+    },
+    {
+      "epoch": 0.12445242613464273,
+      "grad_norm": 0.6829107403755188,
+      "learning_rate": 4.98010865022843e-05,
+      "loss": 3.117,
+      "step": 15270
+    },
+    {
+      "epoch": 0.12469692991879723,
+      "grad_norm": 0.6554434895515442,
+      "learning_rate": 4.980027871636293e-05,
+      "loss": 3.1089,
+      "step": 15300
+    },
+    {
+      "epoch": 0.12494143370295174,
+      "grad_norm": 0.6667256355285645,
+      "learning_rate": 4.979946930013576e-05,
+      "loss": 3.0926,
+      "step": 15330
+    },
+    {
+      "epoch": 0.12518593748710624,
+      "grad_norm": 0.6656805872917175,
+      "learning_rate": 4.9798658253655996e-05,
+      "loss": 3.103,
+      "step": 15360
+    },
+    {
+      "epoch": 0.12543044127126074,
+      "grad_norm": 0.7125841379165649,
+      "learning_rate": 4.9797845576976945e-05,
+      "loss": 3.1153,
+      "step": 15390
+    },
+    {
+      "epoch": 0.12567494505541527,
+      "grad_norm": 0.6778884530067444,
+      "learning_rate": 4.979703127015205e-05,
+      "loss": 3.1402,
+      "step": 15420
+    },
+    {
+      "epoch": 0.12591944883956976,
+      "grad_norm": 0.6417499780654907,
+      "learning_rate": 4.979621533323482e-05,
+      "loss": 3.1006,
+      "step": 15450
+    },
+    {
+      "epoch": 0.12616395262372426,
+      "grad_norm": 0.5959713459014893,
+      "learning_rate": 4.97953977662789e-05,
+      "loss": 3.0978,
+      "step": 15480
+    },
+    {
+      "epoch": 0.12640845640787876,
+      "grad_norm": 0.6567413806915283,
+      "learning_rate": 4.9794578569338046e-05,
+      "loss": 3.1093,
+      "step": 15510
+    },
+    {
+      "epoch": 0.12665296019203326,
+      "grad_norm": 0.630220353603363,
+      "learning_rate": 4.979375774246609e-05,
+      "loss": 3.106,
+      "step": 15540
+    },
+    {
+      "epoch": 0.12689746397618779,
+      "grad_norm": 0.6512444615364075,
+      "learning_rate": 4.979293528571701e-05,
+      "loss": 3.0794,
+      "step": 15570
+    },
+    {
+      "epoch": 0.12714196776034228,
+      "grad_norm": 0.69533771276474,
+      "learning_rate": 4.979211119914486e-05,
+      "loss": 3.0998,
+      "step": 15600
+    },
+    {
+      "epoch": 0.12738647154449678,
+      "grad_norm": 0.6545675992965698,
+      "learning_rate": 4.9791285482803824e-05,
+      "loss": 3.0992,
+      "step": 15630
+    },
+    {
+      "epoch": 0.12763097532865128,
+      "grad_norm": 0.6943343877792358,
+      "learning_rate": 4.979045813674817e-05,
+      "loss": 3.0894,
+      "step": 15660
+    },
+    {
+      "epoch": 0.1278754791128058,
+      "grad_norm": 0.6437342166900635,
+      "learning_rate": 4.9789629161032294e-05,
+      "loss": 3.093,
+      "step": 15690
+    },
+    {
+      "epoch": 0.1281199828969603,
+      "grad_norm": 0.6412293910980225,
+      "learning_rate": 4.978879855571068e-05,
+      "loss": 3.1068,
+      "step": 15720
+    },
+    {
+      "epoch": 0.1283644866811148,
+      "grad_norm": 0.6677573323249817,
+      "learning_rate": 4.9787966320837946e-05,
+      "loss": 3.122,
+      "step": 15750
+    },
+    {
+      "epoch": 0.1286089904652693,
+      "grad_norm": 0.6717284321784973,
+      "learning_rate": 4.97871324564688e-05,
+      "loss": 3.1016,
+      "step": 15780
+    },
+    {
+      "epoch": 0.1288534942494238,
+      "grad_norm": 0.674525260925293,
+      "learning_rate": 4.978629696265804e-05,
+      "loss": 3.0843,
+      "step": 15810
+    },
+    {
+      "epoch": 0.12909799803357833,
+      "grad_norm": 0.6576509475708008,
+      "learning_rate": 4.9785459839460595e-05,
+      "loss": 3.0866,
+      "step": 15840
+    },
+    {
+      "epoch": 0.12934250181773282,
+      "grad_norm": 0.6330334544181824,
+      "learning_rate": 4.97846210869315e-05,
+      "loss": 3.0852,
+      "step": 15870
+    },
+    {
+      "epoch": 0.12958700560188732,
+      "grad_norm": 0.6420540809631348,
+      "learning_rate": 4.978378070512591e-05,
+      "loss": 3.0794,
+      "step": 15900
+    },
+    {
+      "epoch": 0.12983150938604182,
+      "grad_norm": 0.6615996956825256,
+      "learning_rate": 4.978293869409903e-05,
+      "loss": 3.0974,
+      "step": 15930
+    },
+    {
+      "epoch": 0.13007601317019635,
+      "grad_norm": 0.6461077332496643,
+      "learning_rate": 4.978209505390625e-05,
+      "loss": 3.1028,
+      "step": 15960
+    },
+    {
+      "epoch": 0.13032051695435085,
+      "grad_norm": 0.761465311050415,
+      "learning_rate": 4.978124978460301e-05,
+      "loss": 3.1031,
+      "step": 15990
+    },
+    {
+      "epoch": 0.13056502073850534,
+      "grad_norm": 0.7302458882331848,
+      "learning_rate": 4.978040288624488e-05,
+      "loss": 3.0962,
+      "step": 16020
+    },
+    {
+      "epoch": 0.13080952452265984,
+      "grad_norm": 0.6438547372817993,
+      "learning_rate": 4.977955435888753e-05,
+      "loss": 3.0893,
+      "step": 16050
+    },
+    {
+      "epoch": 0.13105402830681434,
+      "grad_norm": 0.6334322690963745,
+      "learning_rate": 4.9778704202586736e-05,
+      "loss": 3.088,
+      "step": 16080
+    },
+    {
+      "epoch": 0.13129853209096887,
+      "grad_norm": 0.6831603646278381,
+      "learning_rate": 4.9777852417398394e-05,
+      "loss": 3.0889,
+      "step": 16110
+    },
+    {
+      "epoch": 0.13154303587512337,
+      "grad_norm": 0.6886951327323914,
+      "learning_rate": 4.977699900337849e-05,
+      "loss": 3.1016,
+      "step": 16140
+    },
+    {
+      "epoch": 0.13178753965927786,
+      "grad_norm": 0.6405287384986877,
+      "learning_rate": 4.977614396058313e-05,
+      "loss": 3.0914,
+      "step": 16170
+    },
+    {
+      "epoch": 0.13203204344343236,
+      "grad_norm": 0.671773374080658,
+      "learning_rate": 4.977528728906853e-05,
+      "loss": 3.0843,
+      "step": 16200
+    },
+    {
+      "epoch": 0.13227654722758686,
+      "grad_norm": 0.6488827466964722,
+      "learning_rate": 4.977442898889099e-05,
+      "loss": 3.0759,
+      "step": 16230
+    },
+    {
+      "epoch": 0.1325210510117414,
+      "grad_norm": 0.6638045310974121,
+      "learning_rate": 4.977356906010694e-05,
+      "loss": 3.0921,
+      "step": 16260
+    },
+    {
+      "epoch": 0.13276555479589588,
+      "grad_norm": 0.6270018219947815,
+      "learning_rate": 4.977270750277291e-05,
+      "loss": 3.0881,
+      "step": 16290
+    },
+    {
+      "epoch": 0.13301005858005038,
+      "grad_norm": 0.6884603500366211,
+      "learning_rate": 4.9771844316945536e-05,
+      "loss": 3.076,
+      "step": 16320
+    },
+    {
+      "epoch": 0.13325456236420488,
+      "grad_norm": 0.6531707644462585,
+      "learning_rate": 4.9770979502681566e-05,
+      "loss": 3.1,
+      "step": 16350
+    },
+    {
+      "epoch": 0.1334990661483594,
+      "grad_norm": 0.6543397903442383,
+      "learning_rate": 4.977011306003784e-05,
+      "loss": 3.0849,
+      "step": 16380
+    },
+    {
+      "epoch": 0.1337435699325139,
+      "grad_norm": 0.6806967258453369,
+      "learning_rate": 4.976924498907133e-05,
+      "loss": 3.0794,
+      "step": 16410
+    },
+    {
+      "epoch": 0.1339880737166684,
+      "grad_norm": 0.6400375962257385,
+      "learning_rate": 4.9768375289839083e-05,
+      "loss": 3.0783,
+      "step": 16440
+    },
+    {
+      "epoch": 0.1342325775008229,
+      "grad_norm": 0.6495615243911743,
+      "learning_rate": 4.976750396239828e-05,
+      "loss": 3.0674,
+      "step": 16470
+    },
+    {
+      "epoch": 0.1344770812849774,
+      "grad_norm": 0.6477648019790649,
+      "learning_rate": 4.9766631006806205e-05,
+      "loss": 3.0868,
+      "step": 16500
+    },
+    {
+      "epoch": 0.13472158506913193,
+      "grad_norm": 0.6622707843780518,
+      "learning_rate": 4.9765756423120235e-05,
+      "loss": 3.086,
+      "step": 16530
+    },
+    {
+      "epoch": 0.13496608885328643,
+      "grad_norm": 0.6572865843772888,
+      "learning_rate": 4.976488021139787e-05,
+      "loss": 3.0662,
+      "step": 16560
+    },
+    {
+      "epoch": 0.13521059263744092,
+      "grad_norm": 0.6188306212425232,
+      "learning_rate": 4.9764002371696704e-05,
+      "loss": 3.0723,
+      "step": 16590
+    },
+    {
+      "epoch": 0.13545509642159542,
+      "grad_norm": 0.6763888597488403,
+      "learning_rate": 4.976312290407445e-05,
+      "loss": 3.0863,
+      "step": 16620
+    },
+    {
+      "epoch": 0.13569960020574992,
+      "grad_norm": 0.6841058135032654,
+      "learning_rate": 4.9762241808588915e-05,
+      "loss": 3.0839,
+      "step": 16650
+    },
+    {
+      "epoch": 0.13594410398990445,
+      "grad_norm": 0.6681898832321167,
+      "learning_rate": 4.976135908529802e-05,
+      "loss": 3.0775,
+      "step": 16680
+    },
+    {
+      "epoch": 0.13618860777405895,
+      "grad_norm": 0.6338205337524414,
+      "learning_rate": 4.97604747342598e-05,
+      "loss": 3.0795,
+      "step": 16710
+    },
+    {
+      "epoch": 0.13643311155821344,
+      "grad_norm": 0.6305208206176758,
+      "learning_rate": 4.9759588755532394e-05,
+      "loss": 3.0809,
+      "step": 16740
+    },
+    {
+      "epoch": 0.13667761534236794,
+      "grad_norm": 0.6307396292686462,
+      "learning_rate": 4.9758701149174026e-05,
+      "loss": 3.0635,
+      "step": 16770
+    },
+    {
+      "epoch": 0.13692211912652247,
+      "grad_norm": 0.6475804448127747,
+      "learning_rate": 4.975781191524306e-05,
+      "loss": 3.0886,
+      "step": 16800
+    },
+    {
+      "epoch": 0.13716662291067697,
+      "grad_norm": 0.6961448788642883,
+      "learning_rate": 4.975692105379794e-05,
+      "loss": 3.0827,
+      "step": 16830
+    },
+    {
+      "epoch": 0.13741112669483146,
+      "grad_norm": 0.6734959483146667,
+      "learning_rate": 4.9756028564897236e-05,
+      "loss": 3.0541,
+      "step": 16860
+    },
+    {
+      "epoch": 0.13765563047898596,
+      "grad_norm": 0.6229287981987,
+      "learning_rate": 4.975513444859963e-05,
+      "loss": 3.0651,
+      "step": 16890
+    },
+    {
+      "epoch": 0.13790013426314046,
+      "grad_norm": 0.6343866586685181,
+      "learning_rate": 4.9754238704963875e-05,
+      "loss": 3.0831,
+      "step": 16920
+    },
+    {
+      "epoch": 0.138144638047295,
+      "grad_norm": 0.6195032596588135,
+      "learning_rate": 4.975334133404887e-05,
+      "loss": 3.0846,
+      "step": 16950
+    },
+    {
+      "epoch": 0.13838914183144949,
+      "grad_norm": 0.6339040398597717,
+      "learning_rate": 4.97524423359136e-05,
+      "loss": 3.0565,
+      "step": 16980
+    },
+    {
+      "epoch": 0.13863364561560398,
+      "grad_norm": 0.6364467144012451,
+      "learning_rate": 4.9751541710617165e-05,
+      "loss": 3.0752,
+      "step": 17010
+    },
+    {
+      "epoch": 0.13887814939975848,
+      "grad_norm": 0.715610921382904,
+      "learning_rate": 4.9750639458218774e-05,
+      "loss": 3.0903,
+      "step": 17040
+    },
+    {
+      "epoch": 0.139122653183913,
+      "grad_norm": 0.6401960253715515,
+      "learning_rate": 4.974973557877773e-05,
+      "loss": 3.0751,
+      "step": 17070
+    },
+    {
+      "epoch": 0.1393671569680675,
+      "grad_norm": 0.660082995891571,
+      "learning_rate": 4.9748830072353455e-05,
+      "loss": 3.0684,
+      "step": 17100
+    },
+    {
+      "epoch": 0.139611660752222,
+      "grad_norm": 0.6341212391853333,
+      "learning_rate": 4.9747922939005484e-05,
+      "loss": 3.0755,
+      "step": 17130
+    },
+    {
+      "epoch": 0.1398561645363765,
+      "grad_norm": 0.7254164814949036,
+      "learning_rate": 4.974701417879344e-05,
+      "loss": 3.0668,
+      "step": 17160
+    },
+    {
+      "epoch": 0.140100668320531,
+      "grad_norm": 0.649255633354187,
+      "learning_rate": 4.974610379177706e-05,
+      "loss": 3.0527,
+      "step": 17190
+    },
+    {
+      "epoch": 0.14034517210468553,
+      "grad_norm": 0.6330759525299072,
+      "learning_rate": 4.9745191778016196e-05,
+      "loss": 3.0587,
+      "step": 17220
+    },
+    {
+      "epoch": 0.14058967588884003,
+      "grad_norm": 0.6298040747642517,
+      "learning_rate": 4.97442781375708e-05,
+      "loss": 3.0576,
+      "step": 17250
+    },
+    {
+      "epoch": 0.14083417967299453,
+      "grad_norm": 0.6595584750175476,
+      "learning_rate": 4.974336287050093e-05,
+      "loss": 3.0518,
+      "step": 17280
+    },
+    {
+      "epoch": 0.14107868345714902,
+      "grad_norm": 0.634013831615448,
+      "learning_rate": 4.974244597686676e-05,
+      "loss": 3.0628,
+      "step": 17310
+    },
+    {
+      "epoch": 0.14132318724130352,
+      "grad_norm": 0.6461661458015442,
+      "learning_rate": 4.974152745672855e-05,
+      "loss": 3.0455,
+      "step": 17340
+    },
+    {
+      "epoch": 0.14156769102545805,
+      "grad_norm": 0.635123074054718,
+      "learning_rate": 4.9740607310146706e-05,
+      "loss": 3.0677,
+      "step": 17370
+    },
+    {
+      "epoch": 0.14181219480961255,
+      "grad_norm": 0.6322553753852844,
+      "learning_rate": 4.973968553718169e-05,
+      "loss": 3.0534,
+      "step": 17400
+    },
+    {
+      "epoch": 0.14205669859376704,
+      "grad_norm": 0.6354007720947266,
+      "learning_rate": 4.9738762137894106e-05,
+      "loss": 3.0785,
+      "step": 17430
+    },
+    {
+      "epoch": 0.14230120237792154,
+      "grad_norm": 0.6522131562232971,
+      "learning_rate": 4.973783711234467e-05,
+      "loss": 3.0831,
+      "step": 17460
+    },
+    {
+      "epoch": 0.14254570616207607,
+      "grad_norm": 0.691369354724884,
+      "learning_rate": 4.9736910460594164e-05,
+      "loss": 3.0671,
+      "step": 17490
+    },
+    {
+      "epoch": 0.14279020994623057,
+      "grad_norm": 0.6239174604415894,
+      "learning_rate": 4.973598218270352e-05,
+      "loss": 3.0541,
+      "step": 17520
+    },
+    {
+      "epoch": 0.14303471373038507,
+      "grad_norm": 0.649414598941803,
+      "learning_rate": 4.973505227873377e-05,
+      "loss": 3.0581,
+      "step": 17550
+    },
+    {
+      "epoch": 0.14327921751453956,
+      "grad_norm": 0.6302400827407837,
+      "learning_rate": 4.9734120748746026e-05,
+      "loss": 3.056,
+      "step": 17580
+    },
+    {
+      "epoch": 0.14352372129869406,
+      "grad_norm": 0.6369176506996155,
+      "learning_rate": 4.973318759280153e-05,
+      "loss": 3.0468,
+      "step": 17610
+    },
+    {
+      "epoch": 0.1437682250828486,
+      "grad_norm": 0.6598733067512512,
+      "learning_rate": 4.973225281096162e-05,
+      "loss": 3.055,
+      "step": 17640
+    },
+    {
+      "epoch": 0.1440127288670031,
+      "grad_norm": 0.6348416805267334,
+      "learning_rate": 4.973131640328776e-05,
+      "loss": 3.0594,
+      "step": 17670
+    },
+    {
+      "epoch": 0.14425723265115759,
+      "grad_norm": 0.646902322769165,
+      "learning_rate": 4.97303783698415e-05,
+      "loss": 3.042,
+      "step": 17700
+    },
+    {
+      "epoch": 0.14450173643531208,
+      "grad_norm": 0.6352052688598633,
+      "learning_rate": 4.97294387106845e-05,
+      "loss": 3.0574,
+      "step": 17730
+    },
+    {
+      "epoch": 0.1447462402194666,
+      "grad_norm": 0.6383568048477173,
+      "learning_rate": 4.972849742587853e-05,
+      "loss": 3.0635,
+      "step": 17760
+    },
+    {
+      "epoch": 0.1449907440036211,
+      "grad_norm": 0.6384525895118713,
+      "learning_rate": 4.972755451548548e-05,
+      "loss": 3.0595,
+      "step": 17790
+    },
+    {
+      "epoch": 0.1452352477877756,
+      "grad_norm": 0.635008692741394,
+      "learning_rate": 4.9726609979567316e-05,
+      "loss": 3.0414,
+      "step": 17820
+    },
+    {
+      "epoch": 0.1454797515719301,
+      "grad_norm": 0.6321554183959961,
+      "learning_rate": 4.972566381818614e-05,
+      "loss": 3.0696,
+      "step": 17850
+    },
+    {
+      "epoch": 0.1457242553560846,
+      "grad_norm": 0.6157737970352173,
+      "learning_rate": 4.972471603140415e-05,
+      "loss": 3.0504,
+      "step": 17880
+    },
+    {
+      "epoch": 0.14596875914023913,
+      "grad_norm": 0.6098494529724121,
+      "learning_rate": 4.972376661928365e-05,
+      "loss": 3.0512,
+      "step": 17910
+    },
+    {
+      "epoch": 0.14621326292439363,
+      "grad_norm": 0.6314985156059265,
+      "learning_rate": 4.972281558188706e-05,
+      "loss": 3.0349,
+      "step": 17940
+    },
+    {
+      "epoch": 0.14645776670854813,
+      "grad_norm": 0.628455638885498,
+      "learning_rate": 4.972186291927689e-05,
+      "loss": 3.0559,
+      "step": 17970
+    },
+    {
+      "epoch": 0.14670227049270262,
+      "grad_norm": 0.6198962926864624,
+      "learning_rate": 4.9720908631515757e-05,
+      "loss": 3.0521,
+      "step": 18000
+    },
+    {
+      "epoch": 0.14694677427685712,
+      "grad_norm": 0.6339931488037109,
+      "learning_rate": 4.9719952718666406e-05,
+      "loss": 3.0446,
+      "step": 18030
+    },
+    {
+      "epoch": 0.14719127806101165,
+      "grad_norm": 0.5936692357063293,
+      "learning_rate": 4.971899518079167e-05,
+      "loss": 3.0442,
+      "step": 18060
+    },
+    {
+      "epoch": 0.14743578184516615,
+      "grad_norm": 0.6095844507217407,
+      "learning_rate": 4.971803601795451e-05,
+      "loss": 3.0377,
+      "step": 18090
+    },
+    {
+      "epoch": 0.14768028562932065,
+      "grad_norm": 0.6403419375419617,
+      "learning_rate": 4.9717075230217955e-05,
+      "loss": 3.0631,
+      "step": 18120
+    },
+    {
+      "epoch": 0.14792478941347514,
+      "grad_norm": 0.6064668893814087,
+      "learning_rate": 4.9716112817645176e-05,
+      "loss": 3.0596,
+      "step": 18150
+    },
+    {
+      "epoch": 0.14816929319762967,
+      "grad_norm": 0.6483209729194641,
+      "learning_rate": 4.971514878029945e-05,
+      "loss": 3.0478,
+      "step": 18180
+    },
+    {
+      "epoch": 0.14841379698178417,
+      "grad_norm": 0.6281334757804871,
+      "learning_rate": 4.971418311824413e-05,
+      "loss": 3.0314,
+      "step": 18210
+    },
+    {
+      "epoch": 0.14865830076593867,
+      "grad_norm": 0.6215760707855225,
+      "learning_rate": 4.971321583154271e-05,
+      "loss": 3.048,
+      "step": 18240
+    },
+    {
+      "epoch": 0.14890280455009317,
+      "grad_norm": 0.6094553470611572,
+      "learning_rate": 4.9712246920258784e-05,
+      "loss": 3.0391,
+      "step": 18270
+    },
+    {
+      "epoch": 0.14914730833424766,
+      "grad_norm": 0.6209439635276794,
+      "learning_rate": 4.9711276384456024e-05,
+      "loss": 3.0353,
+      "step": 18300
+    },
+    {
+      "epoch": 0.1493918121184022,
+      "grad_norm": 0.6350612044334412,
+      "learning_rate": 4.9710304224198245e-05,
+      "loss": 3.0395,
+      "step": 18330
+    },
+    {
+      "epoch": 0.1496363159025567,
+      "grad_norm": 0.6057417392730713,
+      "learning_rate": 4.970933043954935e-05,
+      "loss": 3.0448,
+      "step": 18360
+    },
+    {
+      "epoch": 0.1498808196867112,
+      "grad_norm": 0.6194915175437927,
+      "learning_rate": 4.970835503057335e-05,
+      "loss": 3.0291,
+      "step": 18390
+    },
+    {
+      "epoch": 0.15012532347086568,
+      "grad_norm": 0.6081920862197876,
+      "learning_rate": 4.970737799733438e-05,
+      "loss": 3.0509,
+      "step": 18420
+    },
+    {
+      "epoch": 0.1503698272550202,
+      "grad_norm": 0.6237574815750122,
+      "learning_rate": 4.970639933989665e-05,
+      "loss": 3.0534,
+      "step": 18450
+    },
+    {
+      "epoch": 0.1506143310391747,
+      "grad_norm": 0.6470500230789185,
+      "learning_rate": 4.97054190583245e-05,
+      "loss": 3.0419,
+      "step": 18480
+    },
+    {
+      "epoch": 0.1508588348233292,
+      "grad_norm": 0.6340089440345764,
+      "learning_rate": 4.970443715268238e-05,
+      "loss": 3.0389,
+      "step": 18510
+    },
+    {
+      "epoch": 0.1511033386074837,
+      "grad_norm": 0.6185374855995178,
+      "learning_rate": 4.970345362303482e-05,
+      "loss": 3.0408,
+      "step": 18540
+    },
+    {
+      "epoch": 0.1513478423916382,
+      "grad_norm": 0.6287412643432617,
+      "learning_rate": 4.9702468469446496e-05,
+      "loss": 3.0409,
+      "step": 18570
+    },
+    {
+      "epoch": 0.15159234617579273,
+      "grad_norm": 0.651208221912384,
+      "learning_rate": 4.9701481691982157e-05,
+      "loss": 3.024,
+      "step": 18600
+    },
+    {
+      "epoch": 0.15183684995994723,
+      "grad_norm": 0.6494688987731934,
+      "learning_rate": 4.9700493290706665e-05,
+      "loss": 3.0333,
+      "step": 18630
+    },
+    {
+      "epoch": 0.15208135374410173,
+      "grad_norm": 0.6297596096992493,
+      "learning_rate": 4.969950326568501e-05,
+      "loss": 3.0369,
+      "step": 18660
+    },
+    {
+      "epoch": 0.15232585752825623,
+      "grad_norm": 0.6509076356887817,
+      "learning_rate": 4.969851161698227e-05,
+      "loss": 3.0342,
+      "step": 18690
+    },
+    {
+      "epoch": 0.15257036131241072,
+      "grad_norm": 0.6396616101264954,
+      "learning_rate": 4.969751834466363e-05,
+      "loss": 3.0346,
+      "step": 18720
+    },
+    {
+      "epoch": 0.15281486509656525,
+      "grad_norm": 0.603057324886322,
+      "learning_rate": 4.969652344879438e-05,
+      "loss": 3.0336,
+      "step": 18750
+    },
+    {
+      "epoch": 0.15305936888071975,
+      "grad_norm": 0.6266219019889832,
+      "learning_rate": 4.9695526929439925e-05,
+      "loss": 3.0433,
+      "step": 18780
+    },
+    {
+      "epoch": 0.15330387266487425,
+      "grad_norm": 0.6538950800895691,
+      "learning_rate": 4.969452878666578e-05,
+      "loss": 3.0271,
+      "step": 18810
+    },
+    {
+      "epoch": 0.15354837644902875,
+      "grad_norm": 0.6142851710319519,
+      "learning_rate": 4.969352902053756e-05,
+      "loss": 3.0436,
+      "step": 18840
+    },
+    {
+      "epoch": 0.15379288023318327,
+      "grad_norm": 0.6135523319244385,
+      "learning_rate": 4.9692527631120976e-05,
+      "loss": 3.0306,
+      "step": 18870
+    },
+    {
+      "epoch": 0.15403738401733777,
+      "grad_norm": 0.6318328976631165,
+      "learning_rate": 4.969152461848187e-05,
+      "loss": 3.0241,
+      "step": 18900
+    },
+    {
+      "epoch": 0.15428188780149227,
+      "grad_norm": 0.6185058355331421,
+      "learning_rate": 4.969051998268617e-05,
+      "loss": 3.0375,
+      "step": 18930
+    },
+    {
+      "epoch": 0.15452639158564677,
+      "grad_norm": 0.6477802991867065,
+      "learning_rate": 4.968951372379992e-05,
+      "loss": 3.0224,
+      "step": 18960
+    },
+    {
+      "epoch": 0.15477089536980126,
+      "grad_norm": 0.6331843733787537,
+      "learning_rate": 4.9688505841889266e-05,
+      "loss": 3.0114,
+      "step": 18990
+    },
+    {
+      "epoch": 0.1550153991539558,
+      "grad_norm": 0.6481664180755615,
+      "learning_rate": 4.968749633702047e-05,
+      "loss": 3.0443,
+      "step": 19020
+    },
+    {
+      "epoch": 0.1552599029381103,
+      "grad_norm": 0.6068332195281982,
+      "learning_rate": 4.968648520925988e-05,
+      "loss": 3.0362,
+      "step": 19050
+    },
+    {
+      "epoch": 0.1555044067222648,
+      "grad_norm": 0.6112623810768127,
+      "learning_rate": 4.9685472458673986e-05,
+      "loss": 3.0289,
+      "step": 19080
+    },
+    {
+      "epoch": 0.15574891050641929,
+      "grad_norm": 0.5991814136505127,
+      "learning_rate": 4.968445808532935e-05,
+      "loss": 3.0233,
+      "step": 19110
+    },
+    {
+      "epoch": 0.1559934142905738,
+      "grad_norm": 0.64973384141922,
+      "learning_rate": 4.9683442089292656e-05,
+      "loss": 3.0264,
+      "step": 19140
+    },
+    {
+      "epoch": 0.1562379180747283,
+      "grad_norm": 0.6454912424087524,
+      "learning_rate": 4.9682424470630685e-05,
+      "loss": 3.0168,
+      "step": 19170
+    },
+    {
+      "epoch": 0.1564824218588828,
+      "grad_norm": 0.6001042723655701,
+      "learning_rate": 4.968140522941035e-05,
+      "loss": 3.0211,
+      "step": 19200
+    },
+    {
+      "epoch": 0.1567269256430373,
+      "grad_norm": 0.6036773324012756,
+      "learning_rate": 4.9680384365698644e-05,
+      "loss": 3.0354,
+      "step": 19230
+    },
+    {
+      "epoch": 0.1569714294271918,
+      "grad_norm": 0.64256352186203,
+      "learning_rate": 4.967936187956268e-05,
+      "loss": 3.0359,
+      "step": 19260
+    },
+    {
+      "epoch": 0.15721593321134633,
+      "grad_norm": 0.6243559718132019,
+      "learning_rate": 4.967833777106966e-05,
+      "loss": 3.0307,
+      "step": 19290
+    },
+    {
+      "epoch": 0.15746043699550083,
+      "grad_norm": 0.6249656081199646,
+      "learning_rate": 4.9677312040286926e-05,
+      "loss": 2.9937,
+      "step": 19320
+    },
+    {
+      "epoch": 0.15770494077965533,
+      "grad_norm": 0.654472827911377,
+      "learning_rate": 4.967628468728189e-05,
+      "loss": 3.0364,
+      "step": 19350
+    },
+    {
+      "epoch": 0.15794944456380983,
+      "grad_norm": 0.6340649127960205,
+      "learning_rate": 4.9675255712122105e-05,
+      "loss": 3.0216,
+      "step": 19380
+    },
+    {
+      "epoch": 0.15819394834796432,
+      "grad_norm": 0.6119604110717773,
+      "learning_rate": 4.967422511487519e-05,
+      "loss": 3.0084,
+      "step": 19410
+    },
+    {
+      "epoch": 0.15843845213211885,
+      "grad_norm": 0.6218209266662598,
+      "learning_rate": 4.967319289560892e-05,
+      "loss": 3.0254,
+      "step": 19440
+    },
+    {
+      "epoch": 0.15868295591627335,
+      "grad_norm": 0.6509430408477783,
+      "learning_rate": 4.967215905439113e-05,
+      "loss": 3.0377,
+      "step": 19470
+    },
+    {
+      "epoch": 0.15892745970042785,
+      "grad_norm": 0.62874436378479,
+      "learning_rate": 4.967112359128979e-05,
+      "loss": 3.0205,
+      "step": 19500
+    },
+    {
+      "epoch": 0.15917196348458235,
+      "grad_norm": 0.6077570915222168,
+      "learning_rate": 4.967008650637297e-05,
+      "loss": 3.0021,
+      "step": 19530
+    },
+    {
+      "epoch": 0.15941646726873687,
+      "grad_norm": 0.6242923736572266,
+      "learning_rate": 4.9669047799708834e-05,
+      "loss": 3.0195,
+      "step": 19560
+    },
+    {
+      "epoch": 0.15966097105289137,
+      "grad_norm": 0.58917236328125,
+      "learning_rate": 4.966800747136568e-05,
+      "loss": 3.0002,
+      "step": 19590
+    },
+    {
+      "epoch": 0.15990547483704587,
+      "grad_norm": 0.611606776714325,
+      "learning_rate": 4.966696552141189e-05,
+      "loss": 3.015,
+      "step": 19620
+    },
+    {
+      "epoch": 0.16014997862120037,
+      "grad_norm": 0.5987648367881775,
+      "learning_rate": 4.966592194991596e-05,
+      "loss": 3.0078,
+      "step": 19650
+    },
+    {
+      "epoch": 0.16039448240535487,
+      "grad_norm": 0.6070423126220703,
+      "learning_rate": 4.966487675694649e-05,
+      "loss": 3.0171,
+      "step": 19680
+    },
+    {
+      "epoch": 0.1606389861895094,
+      "grad_norm": 0.6137816905975342,
+      "learning_rate": 4.966382994257218e-05,
+      "loss": 3.0158,
+      "step": 19710
+    },
+    {
+      "epoch": 0.1608834899736639,
+      "grad_norm": 0.6102014183998108,
+      "learning_rate": 4.966278150686186e-05,
+      "loss": 2.9924,
+      "step": 19740
+    },
+    {
+      "epoch": 0.1611279937578184,
+      "grad_norm": 0.641494870185852,
+      "learning_rate": 4.9661731449884444e-05,
+      "loss": 2.9951,
+      "step": 19770
+    },
+    {
+      "epoch": 0.1613724975419729,
+      "grad_norm": 0.6112068295478821,
+      "learning_rate": 4.966067977170896e-05,
+      "loss": 3.0151,
+      "step": 19800
+    },
+    {
+      "epoch": 0.1616170013261274,
+      "grad_norm": 0.6172027587890625,
+      "learning_rate": 4.9659626472404545e-05,
+      "loss": 3.0018,
+      "step": 19830
+    },
+    {
+      "epoch": 0.1618615051102819,
+      "grad_norm": 0.6285238862037659,
+      "learning_rate": 4.965857155204044e-05,
+      "loss": 3.0134,
+      "step": 19860
+    },
+    {
+      "epoch": 0.1621060088944364,
+      "grad_norm": 0.6203986406326294,
+      "learning_rate": 4.965751501068598e-05,
+      "loss": 3.0134,
+      "step": 19890
+    },
+    {
+      "epoch": 0.1623505126785909,
+      "grad_norm": 0.6211456656455994,
+      "learning_rate": 4.965645684841065e-05,
+      "loss": 3.0258,
+      "step": 19920
+    },
+    {
+      "epoch": 0.1625950164627454,
+      "grad_norm": 0.5944324135780334,
+      "learning_rate": 4.965539706528398e-05,
+      "loss": 3.0115,
+      "step": 19950
+    },
+    {
+      "epoch": 0.16283952024689993,
+      "grad_norm": 0.640683650970459,
+      "learning_rate": 4.965433566137565e-05,
+      "loss": 3.0065,
+      "step": 19980
+    },
+    {
+      "epoch": 0.16308402403105443,
+      "grad_norm": 0.5876209139823914,
+      "learning_rate": 4.965327263675543e-05,
+      "loss": 3.0062,
+      "step": 20010
+    },
+    {
+      "epoch": 0.16332852781520893,
+      "grad_norm": 0.5797463655471802,
+      "learning_rate": 4.965220799149321e-05,
+      "loss": 2.9948,
+      "step": 20040
+    },
+    {
+      "epoch": 0.16357303159936343,
+      "grad_norm": 0.6228756308555603,
+      "learning_rate": 4.9651141725658955e-05,
+      "loss": 3.0144,
+      "step": 20070
+    },
+    {
+      "epoch": 0.16381753538351793,
+      "grad_norm": 0.6091254949569702,
+      "learning_rate": 4.9650073839322786e-05,
+      "loss": 3.0243,
+      "step": 20100
+    },
+    {
+      "epoch": 0.16406203916767245,
+      "grad_norm": 0.600600004196167,
+      "learning_rate": 4.964900433255489e-05,
+      "loss": 3.012,
+      "step": 20130
+    },
+    {
+      "epoch": 0.16430654295182695,
+      "grad_norm": 0.6189032196998596,
+      "learning_rate": 4.9647933205425574e-05,
+      "loss": 3.0066,
+      "step": 20160
+    },
+    {
+      "epoch": 0.16455104673598145,
+      "grad_norm": 0.6211121678352356,
+      "learning_rate": 4.9646860458005254e-05,
+      "loss": 3.0187,
+      "step": 20190
+    },
+    {
+      "epoch": 0.16479555052013595,
+      "grad_norm": 0.639061450958252,
+      "learning_rate": 4.9645786090364436e-05,
+      "loss": 3.0281,
+      "step": 20220
+    },
+    {
+      "epoch": 0.16504005430429047,
+      "grad_norm": 0.5996602177619934,
+      "learning_rate": 4.964471010257377e-05,
+      "loss": 2.9862,
+      "step": 20250
+    },
+    {
+      "epoch": 0.16528455808844497,
+      "grad_norm": 0.6031808257102966,
+      "learning_rate": 4.964363249470396e-05,
+      "loss": 3.0069,
+      "step": 20280
+    },
+    {
+      "epoch": 0.16552906187259947,
+      "grad_norm": 0.6389682292938232,
+      "learning_rate": 4.964255326682588e-05,
+      "loss": 3.0099,
+      "step": 20310
+    },
+    {
+      "epoch": 0.16577356565675397,
+      "grad_norm": 0.6115438342094421,
+      "learning_rate": 4.964147241901045e-05,
+      "loss": 2.9926,
+      "step": 20340
+    },
+    {
+      "epoch": 0.16601806944090847,
+      "grad_norm": 0.5930183529853821,
+      "learning_rate": 4.964038995132872e-05,
+      "loss": 3.0108,
+      "step": 20370
+    },
+    {
+      "epoch": 0.166262573225063,
+      "grad_norm": 0.6384121775627136,
+      "learning_rate": 4.963930586385186e-05,
+      "loss": 3.002,
+      "step": 20400
+    },
+    {
+      "epoch": 0.1665070770092175,
+      "grad_norm": 0.6042916178703308,
+      "learning_rate": 4.963822015665114e-05,
+      "loss": 2.9867,
+      "step": 20430
+    },
+    {
+      "epoch": 0.166751580793372,
+      "grad_norm": 0.6278625130653381,
+      "learning_rate": 4.963713282979792e-05,
+      "loss": 2.9896,
+      "step": 20460
+    },
+    {
+      "epoch": 0.1669960845775265,
+      "grad_norm": 0.5804896354675293,
+      "learning_rate": 4.963604388336368e-05,
+      "loss": 3.0091,
+      "step": 20490
+    },
+    {
+      "epoch": 0.16724058836168101,
+      "grad_norm": 0.6057926416397095,
+      "learning_rate": 4.9634953317420006e-05,
+      "loss": 2.9925,
+      "step": 20520
+    },
+    {
+      "epoch": 0.1674850921458355,
+      "grad_norm": 0.6216057538986206,
+      "learning_rate": 4.963386113203859e-05,
+      "loss": 2.9662,
+      "step": 20550
+    },
+    {
+      "epoch": 0.16772959592999,
+      "grad_norm": 0.5599972605705261,
+      "learning_rate": 4.963276732729124e-05,
+      "loss": 2.9848,
+      "step": 20580
+    },
+    {
+      "epoch": 0.1679740997141445,
+      "grad_norm": 0.5946940183639526,
+      "learning_rate": 4.9631671903249835e-05,
+      "loss": 2.99,
+      "step": 20610
+    },
+    {
+      "epoch": 0.168218603498299,
+      "grad_norm": 0.663993239402771,
+      "learning_rate": 4.963057485998641e-05,
+      "loss": 2.9959,
+      "step": 20640
+    },
+    {
+      "epoch": 0.16846310728245353,
+      "grad_norm": 0.579962432384491,
+      "learning_rate": 4.962947619757307e-05,
+      "loss": 3.0081,
+      "step": 20670
+    },
+    {
+      "epoch": 0.16870761106660803,
+      "grad_norm": 0.6171576380729675,
+      "learning_rate": 4.962837591608203e-05,
+      "loss": 2.9801,
+      "step": 20700
+    },
+    {
+      "epoch": 0.16895211485076253,
+      "grad_norm": 0.6330801844596863,
+      "learning_rate": 4.962727401558563e-05,
+      "loss": 2.986,
+      "step": 20730
+    },
+    {
+      "epoch": 0.16919661863491703,
+      "grad_norm": 0.5701897740364075,
+      "learning_rate": 4.962617049615631e-05,
+      "loss": 2.9687,
+      "step": 20760
+    },
+    {
+      "epoch": 0.16944112241907153,
+      "grad_norm": 0.6304188966751099,
+      "learning_rate": 4.96250653578666e-05,
+      "loss": 2.9874,
+      "step": 20790
+    },
+    {
+      "epoch": 0.16968562620322605,
+      "grad_norm": 0.594829261302948,
+      "learning_rate": 4.962395860078917e-05,
+      "loss": 2.9743,
+      "step": 20820
+    },
+    {
+      "epoch": 0.16993012998738055,
+      "grad_norm": 0.6388981938362122,
+      "learning_rate": 4.9622850224996744e-05,
+      "loss": 2.9882,
+      "step": 20850
+    },
+    {
+      "epoch": 0.17017463377153505,
+      "grad_norm": 0.5935282111167908,
+      "learning_rate": 4.962174023056221e-05,
+      "loss": 2.9889,
+      "step": 20880
+    },
+    {
+      "epoch": 0.17041913755568955,
+      "grad_norm": 0.6028230786323547,
+      "learning_rate": 4.962062861755853e-05,
+      "loss": 3.0032,
+      "step": 20910
+    },
+    {
+      "epoch": 0.17066364133984407,
+      "grad_norm": 0.6068926453590393,
+      "learning_rate": 4.9619515386058775e-05,
+      "loss": 2.9902,
+      "step": 20940
+    },
+    {
+      "epoch": 0.17090814512399857,
+      "grad_norm": 0.5927624106407166,
+      "learning_rate": 4.9618400536136124e-05,
+      "loss": 2.993,
+      "step": 20970
+    },
+    {
+      "epoch": 0.17115264890815307,
+      "grad_norm": 0.5808185338973999,
+      "learning_rate": 4.9617284067863866e-05,
+      "loss": 2.9785,
+      "step": 21000
+    },
+    {
+      "epoch": 0.17139715269230757,
+      "grad_norm": 0.6751015186309814,
+      "learning_rate": 4.9616165981315397e-05,
+      "loss": 2.987,
+      "step": 21030
+    },
+    {
+      "epoch": 0.17164165647646207,
+      "grad_norm": 0.6041668057441711,
+      "learning_rate": 4.961504627656422e-05,
+      "loss": 2.9953,
+      "step": 21060
+    },
+    {
+      "epoch": 0.1718861602606166,
+      "grad_norm": 0.6176959276199341,
+      "learning_rate": 4.961392495368393e-05,
+      "loss": 2.9975,
+      "step": 21090
+    },
+    {
+      "epoch": 0.1721306640447711,
+      "grad_norm": 0.6609038710594177,
+      "learning_rate": 4.961280201274826e-05,
+      "loss": 3.0082,
+      "step": 21120
+    },
+    {
+      "epoch": 0.1723751678289256,
+      "grad_norm": 0.6228735446929932,
+      "learning_rate": 4.9611677453831005e-05,
+      "loss": 2.9649,
+      "step": 21150
+    },
+    {
+      "epoch": 0.1726196716130801,
+      "grad_norm": 0.5892258882522583,
+      "learning_rate": 4.9610551277006115e-05,
+      "loss": 2.9809,
+      "step": 21180
+    },
+    {
+      "epoch": 0.1728641753972346,
+      "grad_norm": 0.6028584837913513,
+      "learning_rate": 4.96094234823476e-05,
+      "loss": 2.989,
+      "step": 21210
+    },
+    {
+      "epoch": 0.1731086791813891,
+      "grad_norm": 0.583513081073761,
+      "learning_rate": 4.9608294069929605e-05,
+      "loss": 2.9662,
+      "step": 21240
+    },
+    {
+      "epoch": 0.1733531829655436,
+      "grad_norm": 0.6293556094169617,
+      "learning_rate": 4.960716303982638e-05,
+      "loss": 2.9646,
+      "step": 21270
+    },
+    {
+      "epoch": 0.1735976867496981,
+      "grad_norm": 0.5999639630317688,
+      "learning_rate": 4.960603039211227e-05,
+      "loss": 2.9678,
+      "step": 21300
+    },
+    {
+      "epoch": 0.1738421905338526,
+      "grad_norm": 0.5838433504104614,
+      "learning_rate": 4.960489612686174e-05,
+      "loss": 2.9879,
+      "step": 21330
+    },
+    {
+      "epoch": 0.17408669431800713,
+      "grad_norm": 0.5837694406509399,
+      "learning_rate": 4.960376024414935e-05,
+      "loss": 2.9804,
+      "step": 21360
+    },
+    {
+      "epoch": 0.17433119810216163,
+      "grad_norm": 0.5874453186988831,
+      "learning_rate": 4.9602622744049765e-05,
+      "loss": 2.9765,
+      "step": 21390
+    },
+    {
+      "epoch": 0.17457570188631613,
+      "grad_norm": 0.6144499778747559,
+      "learning_rate": 4.960148362663777e-05,
+      "loss": 2.9787,
+      "step": 21420
+    },
+    {
+      "epoch": 0.17482020567047063,
+      "grad_norm": 0.6012053489685059,
+      "learning_rate": 4.960034289198824e-05,
+      "loss": 2.9816,
+      "step": 21450
+    },
+    {
+      "epoch": 0.17506470945462513,
+      "grad_norm": 0.6054373383522034,
+      "learning_rate": 4.959920054017616e-05,
+      "loss": 2.9876,
+      "step": 21480
+    },
+    {
+      "epoch": 0.17530921323877965,
+      "grad_norm": 0.5829117894172668,
+      "learning_rate": 4.959805657127664e-05,
+      "loss": 2.9821,
+      "step": 21510
+    },
+    {
+      "epoch": 0.17555371702293415,
+      "grad_norm": 0.5930308699607849,
+      "learning_rate": 4.959691098536487e-05,
+      "loss": 2.9861,
+      "step": 21540
+    },
+    {
+      "epoch": 0.17579822080708865,
+      "grad_norm": 0.5863514542579651,
+      "learning_rate": 4.959576378251617e-05,
+      "loss": 2.9819,
+      "step": 21570
+    },
+    {
+      "epoch": 0.17604272459124315,
+      "grad_norm": 0.6006020307540894,
+      "learning_rate": 4.959461496280594e-05,
+      "loss": 2.9804,
+      "step": 21600
+    },
+    {
+      "epoch": 0.17628722837539768,
+      "grad_norm": 0.5972082018852234,
+      "learning_rate": 4.95934645263097e-05,
+      "loss": 2.9772,
+      "step": 21630
+    },
+    {
+      "epoch": 0.17653173215955217,
+      "grad_norm": 0.5825967192649841,
+      "learning_rate": 4.9592312473103086e-05,
+      "loss": 2.9835,
+      "step": 21660
+    },
+    {
+      "epoch": 0.17677623594370667,
+      "grad_norm": 0.5810343623161316,
+      "learning_rate": 4.9591158803261834e-05,
+      "loss": 2.9717,
+      "step": 21690
+    },
+    {
+      "epoch": 0.17702073972786117,
+      "grad_norm": 0.5918793678283691,
+      "learning_rate": 4.9590003516861774e-05,
+      "loss": 2.9554,
+      "step": 21720
+    },
+    {
+      "epoch": 0.17726524351201567,
+      "grad_norm": 0.5869923233985901,
+      "learning_rate": 4.958884661397885e-05,
+      "loss": 2.971,
+      "step": 21750
+    },
+    {
+      "epoch": 0.1775097472961702,
+      "grad_norm": 0.6326466202735901,
+      "learning_rate": 4.958768809468912e-05,
+      "loss": 2.9822,
+      "step": 21780
+    },
+    {
+      "epoch": 0.1777542510803247,
+      "grad_norm": 0.6132392883300781,
+      "learning_rate": 4.958652795906874e-05,
+      "loss": 2.9869,
+      "step": 21810
+    },
+    {
+      "epoch": 0.1779987548644792,
+      "grad_norm": 0.6292781829833984,
+      "learning_rate": 4.9585366207193984e-05,
+      "loss": 2.979,
+      "step": 21840
+    },
+    {
+      "epoch": 0.1782432586486337,
+      "grad_norm": 0.603531002998352,
+      "learning_rate": 4.95842028391412e-05,
+      "loss": 3.0038,
+      "step": 21870
+    },
+    {
+      "epoch": 0.1784877624327882,
+      "grad_norm": 0.6049748659133911,
+      "learning_rate": 4.958303785498689e-05,
+      "loss": 2.9919,
+      "step": 21900
+    },
+    {
+      "epoch": 0.17873226621694271,
+      "grad_norm": 0.6037519574165344,
+      "learning_rate": 4.958187125480761e-05,
+      "loss": 2.98,
+      "step": 21930
+    },
+    {
+      "epoch": 0.1789767700010972,
+      "grad_norm": 0.5910034775733948,
+      "learning_rate": 4.958070303868008e-05,
+      "loss": 2.9605,
+      "step": 21960
+    },
+    {
+      "epoch": 0.1792212737852517,
+      "grad_norm": 0.5778069496154785,
+      "learning_rate": 4.9579533206681066e-05,
+      "loss": 2.9668,
+      "step": 21990
+    },
+    {
+      "epoch": 0.1794657775694062,
+      "grad_norm": 0.5993695855140686,
+      "learning_rate": 4.957836175888749e-05,
+      "loss": 2.9568,
+      "step": 22020
+    },
+    {
+      "epoch": 0.17971028135356074,
+      "grad_norm": 0.6744295358657837,
+      "learning_rate": 4.957718869537635e-05,
+      "loss": 2.9903,
+      "step": 22050
+    },
+    {
+      "epoch": 0.17995478513771523,
+      "grad_norm": 0.6041865944862366,
+      "learning_rate": 4.957601401622477e-05,
+      "loss": 2.9791,
+      "step": 22080
+    },
+    {
+      "epoch": 0.18019928892186973,
+      "grad_norm": 0.6148335933685303,
+      "learning_rate": 4.957483772150996e-05,
+      "loss": 2.9856,
+      "step": 22110
+    },
+    {
+      "epoch": 0.18044379270602423,
+      "grad_norm": 0.6091467142105103,
+      "learning_rate": 4.957365981130925e-05,
+      "loss": 2.9485,
+      "step": 22140
+    },
+    {
+      "epoch": 0.18068829649017873,
+      "grad_norm": 0.6290934681892395,
+      "learning_rate": 4.957248028570007e-05,
+      "loss": 2.9712,
+      "step": 22170
+    },
+    {
+      "epoch": 0.18093280027433326,
+      "grad_norm": 0.570002019405365,
+      "learning_rate": 4.9571299144759965e-05,
+      "loss": 2.9787,
+      "step": 22200
+    },
+    {
+      "epoch": 0.18117730405848775,
+      "grad_norm": 0.5980287790298462,
+      "learning_rate": 4.957011638856658e-05,
+      "loss": 2.9696,
+      "step": 22230
+    },
+    {
+      "epoch": 0.18142180784264225,
+      "grad_norm": 0.6118261814117432,
+      "learning_rate": 4.956893201719766e-05,
+      "loss": 2.9635,
+      "step": 22260
+    },
+    {
+      "epoch": 0.18166631162679675,
+      "grad_norm": 0.578364372253418,
+      "learning_rate": 4.956774603073107e-05,
+      "loss": 2.9691,
+      "step": 22290
+    },
+    {
+      "epoch": 0.18191081541095128,
+      "grad_norm": 0.6028851270675659,
+      "learning_rate": 4.956655842924477e-05,
+      "loss": 2.9829,
+      "step": 22320
+    },
+    {
+      "epoch": 0.18215531919510577,
+      "grad_norm": 0.5682802200317383,
+      "learning_rate": 4.9565369212816825e-05,
+      "loss": 2.9643,
+      "step": 22350
+    },
+    {
+      "epoch": 0.18239982297926027,
+      "grad_norm": 0.611182689666748,
+      "learning_rate": 4.9564178381525414e-05,
+      "loss": 2.956,
+      "step": 22380
+    },
+    {
+      "epoch": 0.18264432676341477,
+      "grad_norm": 0.5692561864852905,
+      "learning_rate": 4.956298593544883e-05,
+      "loss": 2.9766,
+      "step": 22410
+    },
+    {
+      "epoch": 0.18288883054756927,
+      "grad_norm": 0.5628278255462646,
+      "learning_rate": 4.9561791874665445e-05,
+      "loss": 2.9449,
+      "step": 22440
+    },
+    {
+      "epoch": 0.1831333343317238,
+      "grad_norm": 0.6359434723854065,
+      "learning_rate": 4.956059619925376e-05,
+      "loss": 2.9749,
+      "step": 22470
+    },
+    {
+      "epoch": 0.1833778381158783,
+      "grad_norm": 0.5779123306274414,
+      "learning_rate": 4.9559398909292376e-05,
+      "loss": 2.9668,
+      "step": 22500
+    },
+    {
+      "epoch": 0.1836223419000328,
+      "grad_norm": 0.5938737988471985,
+      "learning_rate": 4.955820000486e-05,
+      "loss": 2.9592,
+      "step": 22530
+    },
+    {
+      "epoch": 0.1838668456841873,
+      "grad_norm": 0.6009770035743713,
+      "learning_rate": 4.9556999486035444e-05,
+      "loss": 2.975,
+      "step": 22560
+    },
+    {
+      "epoch": 0.1841113494683418,
+      "grad_norm": 0.549861490726471,
+      "learning_rate": 4.9555797352897625e-05,
+      "loss": 2.9631,
+      "step": 22590
+    },
+    {
+      "epoch": 0.18435585325249632,
+      "grad_norm": 0.613752543926239,
+      "learning_rate": 4.955459360552558e-05,
+      "loss": 2.959,
+      "step": 22620
+    },
+    {
+      "epoch": 0.18460035703665081,
+      "grad_norm": 0.5959582924842834,
+      "learning_rate": 4.9553388243998424e-05,
+      "loss": 2.9752,
+      "step": 22650
+    },
+    {
+      "epoch": 0.1848448608208053,
+      "grad_norm": 0.5848221182823181,
+      "learning_rate": 4.9552181268395404e-05,
+      "loss": 2.9545,
+      "step": 22680
+    },
+    {
+      "epoch": 0.1850893646049598,
+      "grad_norm": 0.5932062864303589,
+      "learning_rate": 4.955097267879586e-05,
+      "loss": 2.9452,
+      "step": 22710
+    },
+    {
+      "epoch": 0.18533386838911434,
+      "grad_norm": 0.5847275853157043,
+      "learning_rate": 4.954976247527924e-05,
+      "loss": 2.9562,
+      "step": 22740
+    },
+    {
+      "epoch": 0.18557837217326884,
+      "grad_norm": 0.6085349917411804,
+      "learning_rate": 4.9548550657925095e-05,
+      "loss": 2.9498,
+      "step": 22770
+    },
+    {
+      "epoch": 0.18582287595742333,
+      "grad_norm": 0.5871360301971436,
+      "learning_rate": 4.954733722681311e-05,
+      "loss": 2.9585,
+      "step": 22800
+    },
+    {
+      "epoch": 0.18606737974157783,
+      "grad_norm": 0.5915745496749878,
+      "learning_rate": 4.954612218202303e-05,
+      "loss": 2.9549,
+      "step": 22830
+    },
+    {
+      "epoch": 0.18631188352573233,
+      "grad_norm": 0.5842469930648804,
+      "learning_rate": 4.954490552363473e-05,
+      "loss": 2.9667,
+      "step": 22860
+    },
+    {
+      "epoch": 0.18655638730988686,
+      "grad_norm": 0.5912492871284485,
+      "learning_rate": 4.95436872517282e-05,
+      "loss": 2.9672,
+      "step": 22890
+    },
+    {
+      "epoch": 0.18680089109404135,
+      "grad_norm": 0.558610200881958,
+      "learning_rate": 4.954246736638352e-05,
+      "loss": 2.9644,
+      "step": 22920
+    },
+    {
+      "epoch": 0.18704539487819585,
+      "grad_norm": 0.6107169985771179,
+      "learning_rate": 4.954124586768088e-05,
+      "loss": 2.955,
+      "step": 22950
+    },
+    {
+      "epoch": 0.18728989866235035,
+      "grad_norm": 0.6018122434616089,
+      "learning_rate": 4.9540022755700587e-05,
+      "loss": 2.963,
+      "step": 22980
+    },
+    {
+      "epoch": 0.18753440244650488,
+      "grad_norm": 0.6130113005638123,
+      "learning_rate": 4.953879803052304e-05,
+      "loss": 2.9665,
+      "step": 23010
+    },
+    {
+      "epoch": 0.18777890623065938,
+      "grad_norm": 0.6094121336936951,
+      "learning_rate": 4.9537571692228744e-05,
+      "loss": 2.9505,
+      "step": 23040
+    },
+    {
+      "epoch": 0.18802341001481387,
+      "grad_norm": 0.5776805877685547,
+      "learning_rate": 4.953634374089832e-05,
+      "loss": 2.9508,
+      "step": 23070
+    },
+    {
+      "epoch": 0.18826791379896837,
+      "grad_norm": 0.6227549314498901,
+      "learning_rate": 4.953511417661249e-05,
+      "loss": 2.9521,
+      "step": 23100
+    },
+    {
+      "epoch": 0.18851241758312287,
+      "grad_norm": 0.5928400754928589,
+      "learning_rate": 4.953388299945209e-05,
+      "loss": 2.9481,
+      "step": 23130
+    },
+    {
+      "epoch": 0.1887569213672774,
+      "grad_norm": 0.6355761289596558,
+      "learning_rate": 4.9532650209498045e-05,
+      "loss": 2.9619,
+      "step": 23160
+    },
+    {
+      "epoch": 0.1890014251514319,
+      "grad_norm": 0.5992570519447327,
+      "learning_rate": 4.9531415806831394e-05,
+      "loss": 2.9551,
+      "step": 23190
+    },
+    {
+      "epoch": 0.1892459289355864,
+      "grad_norm": 0.611517071723938,
+      "learning_rate": 4.953017979153329e-05,
+      "loss": 2.9727,
+      "step": 23220
+    },
+    {
+      "epoch": 0.1894904327197409,
+      "grad_norm": 0.5785388350486755,
+      "learning_rate": 4.952894216368498e-05,
+      "loss": 2.9751,
+      "step": 23250
+    },
+    {
+      "epoch": 0.1897349365038954,
+      "grad_norm": 0.5729740858078003,
+      "learning_rate": 4.952770292336782e-05,
+      "loss": 2.9428,
+      "step": 23280
+    },
+    {
+      "epoch": 0.18997944028804992,
+      "grad_norm": 0.5732846856117249,
+      "learning_rate": 4.9526462070663294e-05,
+      "loss": 2.9361,
+      "step": 23310
+    },
+    {
+      "epoch": 0.19022394407220442,
+      "grad_norm": 0.5825323462486267,
+      "learning_rate": 4.952521960565295e-05,
+      "loss": 2.9467,
+      "step": 23340
+    },
+    {
+      "epoch": 0.1904684478563589,
+      "grad_norm": 0.5768432021141052,
+      "learning_rate": 4.952397552841846e-05,
+      "loss": 2.9534,
+      "step": 23370
+    },
+    {
+      "epoch": 0.1907129516405134,
+      "grad_norm": 0.5852957367897034,
+      "learning_rate": 4.952272983904164e-05,
+      "loss": 2.9504,
+      "step": 23400
+    },
+    {
+      "epoch": 0.19095745542466794,
+      "grad_norm": 0.575799822807312,
+      "learning_rate": 4.9521482537604345e-05,
+      "loss": 2.9356,
+      "step": 23430
+    },
+    {
+      "epoch": 0.19120195920882244,
+      "grad_norm": 0.5817368626594543,
+      "learning_rate": 4.9520233624188585e-05,
+      "loss": 2.9553,
+      "step": 23460
+    },
+    {
+      "epoch": 0.19144646299297693,
+      "grad_norm": 0.5669567584991455,
+      "learning_rate": 4.951898309887646e-05,
+      "loss": 2.9629,
+      "step": 23490
+    },
+    {
+      "epoch": 0.19169096677713143,
+      "grad_norm": 0.6092284321784973,
+      "learning_rate": 4.951773096175016e-05,
+      "loss": 2.9396,
+      "step": 23520
+    },
+    {
+      "epoch": 0.19193547056128593,
+      "grad_norm": 0.5807421803474426,
+      "learning_rate": 4.9516477212892024e-05,
+      "loss": 2.9425,
+      "step": 23550
+    },
+    {
+      "epoch": 0.19217997434544046,
+      "grad_norm": 0.5865337252616882,
+      "learning_rate": 4.9515221852384455e-05,
+      "loss": 2.9476,
+      "step": 23580
+    },
+    {
+      "epoch": 0.19242447812959496,
+      "grad_norm": 0.583296000957489,
+      "learning_rate": 4.9513964880309974e-05,
+      "loss": 2.9411,
+      "step": 23610
+    },
+    {
+      "epoch": 0.19266898191374945,
+      "grad_norm": 0.5733707547187805,
+      "learning_rate": 4.951270629675121e-05,
+      "loss": 2.9603,
+      "step": 23640
+    },
+    {
+      "epoch": 0.19291348569790395,
+      "grad_norm": 0.5758892297744751,
+      "learning_rate": 4.951144610179092e-05,
+      "loss": 2.9481,
+      "step": 23670
+    },
+    {
+      "epoch": 0.19315798948205848,
+      "grad_norm": 0.5843274593353271,
+      "learning_rate": 4.9510184295511915e-05,
+      "loss": 2.9508,
+      "step": 23700
+    },
+    {
+      "epoch": 0.19340249326621298,
+      "grad_norm": 0.5948002338409424,
+      "learning_rate": 4.950892087799716e-05,
+      "loss": 2.9555,
+      "step": 23730
+    },
+    {
+      "epoch": 0.19364699705036748,
+      "grad_norm": 0.6155719757080078,
+      "learning_rate": 4.950765584932971e-05,
+      "loss": 2.957,
+      "step": 23760
+    },
+    {
+      "epoch": 0.19389150083452197,
+      "grad_norm": 0.5775892734527588,
+      "learning_rate": 4.950638920959273e-05,
+      "loss": 2.9414,
+      "step": 23790
+    },
+    {
+      "epoch": 0.19413600461867647,
+      "grad_norm": 0.5580487847328186,
+      "learning_rate": 4.950512095886947e-05,
+      "loss": 2.9214,
+      "step": 23820
+    },
+    {
+      "epoch": 0.194380508402831,
+      "grad_norm": 0.5733426213264465,
+      "learning_rate": 4.950385109724331e-05,
+      "loss": 2.9431,
+      "step": 23850
+    },
+    {
+      "epoch": 0.1946250121869855,
+      "grad_norm": 0.5649263262748718,
+      "learning_rate": 4.950257962479773e-05,
+      "loss": 2.949,
+      "step": 23880
+    },
+    {
+      "epoch": 0.19486951597114,
+      "grad_norm": 0.5632926225662231,
+      "learning_rate": 4.95013065416163e-05,
+      "loss": 2.9438,
+      "step": 23910
+    },
+    {
+      "epoch": 0.1951140197552945,
+      "grad_norm": 0.5577961206436157,
+      "learning_rate": 4.950003184778272e-05,
+      "loss": 2.9253,
+      "step": 23940
+    },
+    {
+      "epoch": 0.195358523539449,
+      "grad_norm": 0.5746459364891052,
+      "learning_rate": 4.9498755543380795e-05,
+      "loss": 2.9316,
+      "step": 23970
+    },
+    {
+      "epoch": 0.19560302732360352,
+      "grad_norm": 0.5570378303527832,
+      "learning_rate": 4.949747762849441e-05,
+      "loss": 2.9356,
+      "step": 24000
+    },
+    {
+      "epoch": 0.19584753110775802,
+      "grad_norm": 0.5748524069786072,
+      "learning_rate": 4.949619810320757e-05,
+      "loss": 2.9405,
+      "step": 24030
+    },
+    {
+      "epoch": 0.19609203489191251,
+      "grad_norm": 0.5601252913475037,
+      "learning_rate": 4.9494916967604404e-05,
+      "loss": 2.9411,
+      "step": 24060
+    },
+    {
+      "epoch": 0.196336538676067,
+      "grad_norm": 0.6282344460487366,
+      "learning_rate": 4.949363422176911e-05,
+      "loss": 2.9335,
+      "step": 24090
+    },
+    {
+      "epoch": 0.19658104246022154,
+      "grad_norm": 0.5971565842628479,
+      "learning_rate": 4.9492349865786034e-05,
+      "loss": 2.9311,
+      "step": 24120
+    },
+    {
+      "epoch": 0.19682554624437604,
+      "grad_norm": 0.5764152407646179,
+      "learning_rate": 4.9491063899739596e-05,
+      "loss": 2.9343,
+      "step": 24150
+    },
+    {
+      "epoch": 0.19707005002853054,
+      "grad_norm": 0.5799585580825806,
+      "learning_rate": 4.948977632371432e-05,
+      "loss": 2.9357,
+      "step": 24180
+    },
+    {
+      "epoch": 0.19731455381268503,
+      "grad_norm": 0.6082672476768494,
+      "learning_rate": 4.948848713779487e-05,
+      "loss": 2.9373,
+      "step": 24210
+    },
+    {
+      "epoch": 0.19755905759683953,
+      "grad_norm": 0.5643133521080017,
+      "learning_rate": 4.948719634206598e-05,
+      "loss": 2.9291,
+      "step": 24240
+    },
+    {
+      "epoch": 0.19780356138099406,
+      "grad_norm": 0.5771824717521667,
+      "learning_rate": 4.9485903936612504e-05,
+      "loss": 2.969,
+      "step": 24270
+    },
+    {
+      "epoch": 0.19804806516514856,
+      "grad_norm": 0.5739302039146423,
+      "learning_rate": 4.948460992151941e-05,
+      "loss": 2.9263,
+      "step": 24300
+    },
+    {
+      "epoch": 0.19829256894930306,
+      "grad_norm": 0.5707888007164001,
+      "learning_rate": 4.948331429687175e-05,
+      "loss": 2.948,
+      "step": 24330
+    },
+    {
+      "epoch": 0.19853707273345755,
+      "grad_norm": 0.5753827095031738,
+      "learning_rate": 4.9482017062754704e-05,
+      "loss": 2.9476,
+      "step": 24360
+    },
+    {
+      "epoch": 0.19878157651761208,
+      "grad_norm": 0.5887372493743896,
+      "learning_rate": 4.9480718219253555e-05,
+      "loss": 2.9257,
+      "step": 24390
+    },
+    {
+      "epoch": 0.19902608030176658,
+      "grad_norm": 0.5863837599754333,
+      "learning_rate": 4.947941776645366e-05,
+      "loss": 2.9433,
+      "step": 24420
+    },
+    {
+      "epoch": 0.19927058408592108,
+      "grad_norm": 0.5440607070922852,
+      "learning_rate": 4.9478115704440545e-05,
+      "loss": 2.9332,
+      "step": 24450
+    },
+    {
+      "epoch": 0.19951508787007557,
+      "grad_norm": 0.5498874187469482,
+      "learning_rate": 4.947681203329977e-05,
+      "loss": 2.9388,
+      "step": 24480
+    },
+    {
+      "epoch": 0.19975959165423007,
+      "grad_norm": 0.5828962326049805,
+      "learning_rate": 4.947550675311706e-05,
+      "loss": 2.9198,
+      "step": 24510
+    },
+    {
+      "epoch": 0.2000040954383846,
+      "grad_norm": 0.5751544833183289,
+      "learning_rate": 4.94741998639782e-05,
+      "loss": 2.9326,
+      "step": 24540
+    },
+    {
+      "epoch": 0.2002485992225391,
+      "grad_norm": 0.5590956807136536,
+      "learning_rate": 4.947289136596912e-05,
+      "loss": 2.9443,
+      "step": 24570
+    },
+    {
+      "epoch": 0.2004931030066936,
+      "grad_norm": 0.575299084186554,
+      "learning_rate": 4.9471581259175816e-05,
+      "loss": 2.9392,
+      "step": 24600
+    },
+    {
+      "epoch": 0.2007376067908481,
+      "grad_norm": 0.5502616167068481,
+      "learning_rate": 4.9470269543684436e-05,
+      "loss": 2.9482,
+      "step": 24630
+    },
+    {
+      "epoch": 0.2009821105750026,
+      "grad_norm": 0.5680727958679199,
+      "learning_rate": 4.9468956219581186e-05,
+      "loss": 2.9367,
+      "step": 24660
+    },
+    {
+      "epoch": 0.20122661435915712,
+      "grad_norm": 0.6132157444953918,
+      "learning_rate": 4.946764128695242e-05,
+      "loss": 2.9319,
+      "step": 24690
+    },
+    {
+      "epoch": 0.20147111814331162,
+      "grad_norm": 0.5580316185951233,
+      "learning_rate": 4.9466324745884565e-05,
+      "loss": 2.9407,
+      "step": 24720
+    },
+    {
+      "epoch": 0.20171562192746612,
+      "grad_norm": 0.5621688961982727,
+      "learning_rate": 4.946500659646417e-05,
+      "loss": 2.9404,
+      "step": 24750
+    },
+    {
+      "epoch": 0.20196012571162061,
+      "grad_norm": 0.5874739289283752,
+      "learning_rate": 4.9463686838777887e-05,
+      "loss": 2.9338,
+      "step": 24780
+    },
+    {
+      "epoch": 0.20220462949577514,
+      "grad_norm": 0.575962483882904,
+      "learning_rate": 4.946236547291248e-05,
+      "loss": 2.9223,
+      "step": 24810
+    },
+    {
+      "epoch": 0.20244913327992964,
+      "grad_norm": 0.5589853525161743,
+      "learning_rate": 4.94610424989548e-05,
+      "loss": 2.9414,
+      "step": 24840
+    },
+    {
+      "epoch": 0.20269363706408414,
+      "grad_norm": 0.6121220588684082,
+      "learning_rate": 4.9459717916991834e-05,
+      "loss": 2.9061,
+      "step": 24870
+    },
+    {
+      "epoch": 0.20293814084823864,
+      "grad_norm": 0.5621205568313599,
+      "learning_rate": 4.945839172711063e-05,
+      "loss": 2.9375,
+      "step": 24900
+    },
+    {
+      "epoch": 0.20318264463239313,
+      "grad_norm": 0.5536648631095886,
+      "learning_rate": 4.9457063929398394e-05,
+      "loss": 2.9395,
+      "step": 24930
+    },
+    {
+      "epoch": 0.20342714841654766,
+      "grad_norm": 0.5474145412445068,
+      "learning_rate": 4.94557345239424e-05,
+      "loss": 2.9126,
+      "step": 24960
+    },
+    {
+      "epoch": 0.20367165220070216,
+      "grad_norm": 0.5837785601615906,
+      "learning_rate": 4.945440351083003e-05,
+      "loss": 2.9347,
+      "step": 24990
+    },
+    {
+      "epoch": 0.20391615598485666,
+      "grad_norm": 0.5928856134414673,
+      "learning_rate": 4.94530708901488e-05,
+      "loss": 2.9504,
+      "step": 25020
+    },
+    {
+      "epoch": 0.20416065976901115,
+      "grad_norm": 0.5786676406860352,
+      "learning_rate": 4.9451736661986306e-05,
+      "loss": 2.9265,
+      "step": 25050
+    },
+    {
+      "epoch": 0.20440516355316568,
+      "grad_norm": 0.5716553330421448,
+      "learning_rate": 4.945040082643026e-05,
+      "loss": 2.9159,
+      "step": 25080
+    },
+    {
+      "epoch": 0.20464966733732018,
+      "grad_norm": 0.676673412322998,
+      "learning_rate": 4.944906338356846e-05,
+      "loss": 2.9183,
+      "step": 25110
+    },
+    {
+      "epoch": 0.20489417112147468,
+      "grad_norm": 0.556983232498169,
+      "learning_rate": 4.944772433348884e-05,
+      "loss": 2.9248,
+      "step": 25140
+    },
+    {
+      "epoch": 0.20513867490562918,
+      "grad_norm": 0.5894055962562561,
+      "learning_rate": 4.9446383676279435e-05,
+      "loss": 2.9333,
+      "step": 25170
+    },
+    {
+      "epoch": 0.20538317868978367,
+      "grad_norm": 0.5735295414924622,
+      "learning_rate": 4.944504141202836e-05,
+      "loss": 2.9187,
+      "step": 25200
+    }
+  ],
+  "logging_steps": 30,
+  "max_steps": 368091,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.452144150020096e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}