diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6480 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.22494348142214404,
+  "eval_steps": 100000000,
+  "global_step": 27600,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 8.15012613848348e-06,
+      "grad_norm": 3.4710192680358887,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 11.2605,
+      "step": 1
+    },
+    {
+      "epoch": 0.0002445037841545044,
+      "grad_norm": 5.464046955108643,
+      "learning_rate": 3e-06,
+      "loss": 10.9977,
+      "step": 30
+    },
+    {
+      "epoch": 0.0004890075683090088,
+      "grad_norm": 1.1376756429672241,
+      "learning_rate": 6e-06,
+      "loss": 9.4082,
+      "step": 60
+    },
+    {
+      "epoch": 0.0007335113524635132,
+      "grad_norm": 1.2878344058990479,
+      "learning_rate": 9e-06,
+      "loss": 8.549,
+      "step": 90
+    },
+    {
+      "epoch": 0.0009780151366180175,
+      "grad_norm": 1.2800588607788086,
+      "learning_rate": 1.2e-05,
+      "loss": 8.0361,
+      "step": 120
+    },
+    {
+      "epoch": 0.001222518920772522,
+      "grad_norm": 1.4315314292907715,
+      "learning_rate": 1.5e-05,
+      "loss": 7.653,
+      "step": 150
+    },
+    {
+      "epoch": 0.0014670227049270264,
+      "grad_norm": 1.381317377090454,
+      "learning_rate": 1.8e-05,
+      "loss": 7.4179,
+      "step": 180
+    },
+    {
+      "epoch": 0.0017115264890815308,
+      "grad_norm": 1.7989460229873657,
+      "learning_rate": 2.1e-05,
+      "loss": 7.2012,
+      "step": 210
+    },
+    {
+      "epoch": 0.001956030273236035,
+      "grad_norm": 1.3431414365768433,
+      "learning_rate": 2.4e-05,
+      "loss": 7.0383,
+      "step": 240
+    },
+    {
+      "epoch": 0.0022005340573905395,
+      "grad_norm": 1.028826117515564,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 6.866,
+      "step": 270
+    },
+    {
+      "epoch": 0.002445037841545044,
+      "grad_norm": 1.201025128364563,
+      "learning_rate": 3e-05,
+      "loss": 6.717,
+      "step": 300
+    },
+    {
+      "epoch": 0.0026895416256995483,
+      "grad_norm": 1.1023098230361938,
+      "learning_rate": 3.3e-05,
+      "loss": 6.5535,
+      "step": 330
+    },
+    {
+      "epoch": 0.0029340454098540527,
+      "grad_norm": 1.2839674949645996,
+      "learning_rate": 3.6e-05,
+      "loss": 6.4022,
+      "step": 360
+    },
+    {
+      "epoch": 0.003178549194008557,
+      "grad_norm": 2.267265796661377,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 6.2858,
+      "step": 390
+    },
+    {
+      "epoch": 0.0034230529781630616,
+      "grad_norm": 1.0635628700256348,
+      "learning_rate": 4.2e-05,
+      "loss": 6.1681,
+      "step": 420
+    },
+    {
+      "epoch": 0.0036675567623175656,
+      "grad_norm": 1.263838768005371,
+      "learning_rate": 4.5e-05,
+      "loss": 6.0728,
+      "step": 450
+    },
+    {
+      "epoch": 0.00391206054647207,
+      "grad_norm": 1.4611454010009766,
+      "learning_rate": 4.8e-05,
+      "loss": 5.972,
+      "step": 480
+    },
+    {
+      "epoch": 0.0041565643306265745,
+      "grad_norm": 1.0120561122894287,
+      "learning_rate": 4.999999990869806e-05,
+      "loss": 5.8619,
+      "step": 510
+    },
+    {
+      "epoch": 0.004401068114781079,
+      "grad_norm": 1.1349974870681763,
+      "learning_rate": 4.999999853916893e-05,
+      "loss": 5.785,
+      "step": 540
+    },
+    {
+      "epoch": 0.004645571898935583,
+      "grad_norm": 1.0840613842010498,
+      "learning_rate": 4.9999995526204936e-05,
+      "loss": 5.7071,
+      "step": 570
+    },
+    {
+      "epoch": 0.004890075683090088,
+      "grad_norm": 1.258074402809143,
+      "learning_rate": 4.999999086980628e-05,
+      "loss": 5.6199,
+      "step": 600
+    },
+    {
+      "epoch": 0.005134579467244592,
+      "grad_norm": 1.284726858139038,
+      "learning_rate": 4.999998456997326e-05,
+      "loss": 5.5465,
+      "step": 630
+    },
+    {
+      "epoch": 0.005379083251399097,
+      "grad_norm": 1.2079874277114868,
+      "learning_rate": 4.999997662670628e-05,
+      "loss": 5.4816,
+      "step": 660
+    },
+    {
+      "epoch": 0.005623587035553601,
+      "grad_norm": 1.3364052772521973,
+      "learning_rate": 4.999996704000589e-05,
+      "loss": 5.4079,
+      "step": 690
+    },
+    {
+      "epoch": 0.0058680908197081055,
+      "grad_norm": 0.9860705137252808,
+      "learning_rate": 4.99999558098727e-05,
+      "loss": 5.3598,
+      "step": 720
+    },
+    {
+      "epoch": 0.00611259460386261,
+      "grad_norm": 1.2071930170059204,
+      "learning_rate": 4.9999942936307445e-05,
+      "loss": 5.2884,
+      "step": 750
+    },
+    {
+      "epoch": 0.006357098388017114,
+      "grad_norm": 0.8959563970565796,
+      "learning_rate": 4.9999928419310994e-05,
+      "loss": 5.2391,
+      "step": 780
+    },
+    {
+      "epoch": 0.006601602172171619,
+      "grad_norm": 1.2356096506118774,
+      "learning_rate": 4.999991225888427e-05,
+      "loss": 5.1879,
+      "step": 810
+    },
+    {
+      "epoch": 0.006846105956326123,
+      "grad_norm": 0.9705113172531128,
+      "learning_rate": 4.999989445502837e-05,
+      "loss": 5.1424,
+      "step": 840
+    },
+    {
+      "epoch": 0.007090609740480627,
+      "grad_norm": 0.9504437446594238,
+      "learning_rate": 4.9999875007744436e-05,
+      "loss": 5.0966,
+      "step": 870
+    },
+    {
+      "epoch": 0.007335113524635131,
+      "grad_norm": 0.9488673806190491,
+      "learning_rate": 4.9999853917033756e-05,
+      "loss": 5.0424,
+      "step": 900
+    },
+    {
+      "epoch": 0.007579617308789636,
+      "grad_norm": 0.959373950958252,
+      "learning_rate": 4.999983118289773e-05,
+      "loss": 5.0387,
+      "step": 930
+    },
+    {
+      "epoch": 0.00782412109294414,
+      "grad_norm": 0.8465414643287659,
+      "learning_rate": 4.999980680533782e-05,
+      "loss": 4.9769,
+      "step": 960
+    },
+    {
+      "epoch": 0.008068624877098645,
+      "grad_norm": 0.8328993916511536,
+      "learning_rate": 4.999978078435567e-05,
+      "loss": 4.9335,
+      "step": 990
+    },
+    {
+      "epoch": 0.008313128661253149,
+      "grad_norm": 0.8107655644416809,
+      "learning_rate": 4.999975311995295e-05,
+      "loss": 4.9214,
+      "step": 1020
+    },
+    {
+      "epoch": 0.008557632445407654,
+      "grad_norm": 0.8149654865264893,
+      "learning_rate": 4.99997238121315e-05,
+      "loss": 4.8651,
+      "step": 1050
+    },
+    {
+      "epoch": 0.008802136229562158,
+      "grad_norm": 0.8837414979934692,
+      "learning_rate": 4.999969286089325e-05,
+      "loss": 4.8327,
+      "step": 1080
+    },
+    {
+      "epoch": 0.009046640013716663,
+      "grad_norm": 1.1360137462615967,
+      "learning_rate": 4.9999660266240235e-05,
+      "loss": 4.7906,
+      "step": 1110
+    },
+    {
+      "epoch": 0.009291143797871167,
+      "grad_norm": 0.7992026209831238,
+      "learning_rate": 4.9999626028174585e-05,
+      "loss": 4.7612,
+      "step": 1140
+    },
+    {
+      "epoch": 0.00953564758202567,
+      "grad_norm": 0.8481825590133667,
+      "learning_rate": 4.999959014669856e-05,
+      "loss": 4.7106,
+      "step": 1170
+    },
+    {
+      "epoch": 0.009780151366180176,
+      "grad_norm": 0.8183879256248474,
+      "learning_rate": 4.9999552621814513e-05,
+      "loss": 4.6993,
+      "step": 1200
+    },
+    {
+      "epoch": 0.010024655150334679,
+      "grad_norm": 0.8460689187049866,
+      "learning_rate": 4.9999513453524917e-05,
+      "loss": 4.6664,
+      "step": 1230
+    },
+    {
+      "epoch": 0.010269158934489184,
+      "grad_norm": 0.8723706007003784,
+      "learning_rate": 4.9999472641832336e-05,
+      "loss": 4.6371,
+      "step": 1260
+    },
+    {
+      "epoch": 0.010513662718643688,
+      "grad_norm": 0.7682787179946899,
+      "learning_rate": 4.999943018673946e-05,
+      "loss": 4.6184,
+      "step": 1290
+    },
+    {
+      "epoch": 0.010758166502798193,
+      "grad_norm": 0.9049955606460571,
+      "learning_rate": 4.999938608824909e-05,
+      "loss": 4.5968,
+      "step": 1320
+    },
+    {
+      "epoch": 0.011002670286952697,
+      "grad_norm": 0.7860899567604065,
+      "learning_rate": 4.999934034636411e-05,
+      "loss": 4.5266,
+      "step": 1350
+    },
+    {
+      "epoch": 0.011247174071107202,
+      "grad_norm": 0.7918768525123596,
+      "learning_rate": 4.999929296108753e-05,
+      "loss": 4.5069,
+      "step": 1380
+    },
+    {
+      "epoch": 0.011491677855261706,
+      "grad_norm": 0.6713089942932129,
+      "learning_rate": 4.9999243932422466e-05,
+      "loss": 4.4663,
+      "step": 1410
+    },
+    {
+      "epoch": 0.011736181639416211,
+      "grad_norm": 0.756101131439209,
+      "learning_rate": 4.999919326037215e-05,
+      "loss": 4.4782,
+      "step": 1440
+    },
+    {
+      "epoch": 0.011980685423570715,
+      "grad_norm": 0.7067996263504028,
+      "learning_rate": 4.99991409449399e-05,
+      "loss": 4.4388,
+      "step": 1470
+    },
+    {
+      "epoch": 0.01222518920772522,
+      "grad_norm": 0.7132194638252258,
+      "learning_rate": 4.999908698612916e-05,
+      "loss": 4.4135,
+      "step": 1500
+    },
+    {
+      "epoch": 0.012469692991879723,
+      "grad_norm": 0.7282501459121704,
+      "learning_rate": 4.9999031383943486e-05,
+      "loss": 4.4057,
+      "step": 1530
+    },
+    {
+      "epoch": 0.012714196776034229,
+      "grad_norm": 0.6851722598075867,
+      "learning_rate": 4.999897413838651e-05,
+      "loss": 4.3612,
+      "step": 1560
+    },
+    {
+      "epoch": 0.012958700560188732,
+      "grad_norm": 0.7953110337257385,
+      "learning_rate": 4.999891524946202e-05,
+      "loss": 4.3268,
+      "step": 1590
+    },
+    {
+      "epoch": 0.013203204344343238,
+      "grad_norm": 0.726719319820404,
+      "learning_rate": 4.999885471717387e-05,
+      "loss": 4.314,
+      "step": 1620
+    },
+    {
+      "epoch": 0.013447708128497741,
+      "grad_norm": 0.7354792356491089,
+      "learning_rate": 4.999879254152605e-05,
+      "loss": 4.2807,
+      "step": 1650
+    },
+    {
+      "epoch": 0.013692211912652246,
+      "grad_norm": 0.7072017788887024,
+      "learning_rate": 4.999872872252265e-05,
+      "loss": 4.282,
+      "step": 1680
+    },
+    {
+      "epoch": 0.01393671569680675,
+      "grad_norm": 0.690112292766571,
+      "learning_rate": 4.999866326016785e-05,
+      "loss": 4.2686,
+      "step": 1710
+    },
+    {
+      "epoch": 0.014181219480961254,
+      "grad_norm": 0.697634756565094,
+      "learning_rate": 4.999859615446596e-05,
+      "loss": 4.2662,
+      "step": 1740
+    },
+    {
+      "epoch": 0.014425723265115759,
+      "grad_norm": 0.6999862790107727,
+      "learning_rate": 4.99985274054214e-05,
+      "loss": 4.2207,
+      "step": 1770
+    },
+    {
+      "epoch": 0.014670227049270262,
+      "grad_norm": 0.6920334100723267,
+      "learning_rate": 4.999845701303868e-05,
+      "loss": 4.2163,
+      "step": 1800
+    },
+    {
+      "epoch": 0.014914730833424768,
+      "grad_norm": 0.6884493827819824,
+      "learning_rate": 4.999838497732243e-05,
+      "loss": 4.209,
+      "step": 1830
+    },
+    {
+      "epoch": 0.015159234617579271,
+      "grad_norm": 0.776447057723999,
+      "learning_rate": 4.999831129827739e-05,
+      "loss": 4.1856,
+      "step": 1860
+    },
+    {
+      "epoch": 0.015403738401733777,
+      "grad_norm": 0.7332949042320251,
+      "learning_rate": 4.9998235975908394e-05,
+      "loss": 4.156,
+      "step": 1890
+    },
+    {
+      "epoch": 0.01564824218588828,
+      "grad_norm": 0.6691558361053467,
+      "learning_rate": 4.99981590102204e-05,
+      "loss": 4.1339,
+      "step": 1920
+    },
+    {
+      "epoch": 0.015892745970042785,
+      "grad_norm": 0.6789201498031616,
+      "learning_rate": 4.9998080401218464e-05,
+      "loss": 4.1344,
+      "step": 1950
+    },
+    {
+      "epoch": 0.01613724975419729,
+      "grad_norm": 0.6654175519943237,
+      "learning_rate": 4.999800014890777e-05,
+      "loss": 4.1295,
+      "step": 1980
+    },
+    {
+      "epoch": 0.016381753538351793,
+      "grad_norm": 0.6632519364356995,
+      "learning_rate": 4.9997918253293555e-05,
+      "loss": 4.1036,
+      "step": 2010
+    },
+    {
+      "epoch": 0.016626257322506298,
+      "grad_norm": 0.6618740558624268,
+      "learning_rate": 4.999783471438124e-05,
+      "loss": 4.0857,
+      "step": 2040
+    },
+    {
+      "epoch": 0.016870761106660803,
+      "grad_norm": 0.6384952068328857,
+      "learning_rate": 4.999774953217631e-05,
+      "loss": 4.0755,
+      "step": 2070
+    },
+    {
+      "epoch": 0.01711526489081531,
+      "grad_norm": 0.6808215975761414,
+      "learning_rate": 4.9997662706684345e-05,
+      "loss": 4.0568,
+      "step": 2100
+    },
+    {
+      "epoch": 0.01735976867496981,
+      "grad_norm": 0.6514068841934204,
+      "learning_rate": 4.999757423791107e-05,
+      "loss": 4.0474,
+      "step": 2130
+    },
+    {
+      "epoch": 0.017604272459124316,
+      "grad_norm": 0.6691142320632935,
+      "learning_rate": 4.9997484125862306e-05,
+      "loss": 4.0467,
+      "step": 2160
+    },
+    {
+      "epoch": 0.01784877624327882,
+      "grad_norm": 0.6240984201431274,
+      "learning_rate": 4.999739237054395e-05,
+      "loss": 4.0259,
+      "step": 2190
+    },
+    {
+      "epoch": 0.018093280027433326,
+      "grad_norm": 0.6286123991012573,
+      "learning_rate": 4.9997298971962065e-05,
+      "loss": 4.0098,
+      "step": 2220
+    },
+    {
+      "epoch": 0.018337783811587828,
+      "grad_norm": 0.6232910752296448,
+      "learning_rate": 4.999720393012277e-05,
+      "loss": 4.0203,
+      "step": 2250
+    },
+    {
+      "epoch": 0.018582287595742333,
+      "grad_norm": 0.6792376041412354,
+      "learning_rate": 4.999710724503233e-05,
+      "loss": 3.9909,
+      "step": 2280
+    },
+    {
+      "epoch": 0.01882679137989684,
+      "grad_norm": 0.6824718117713928,
+      "learning_rate": 4.9997008916697075e-05,
+      "loss": 3.9922,
+      "step": 2310
+    },
+    {
+      "epoch": 0.01907129516405134,
+      "grad_norm": 0.6520410776138306,
+      "learning_rate": 4.999690894512349e-05,
+      "loss": 3.9533,
+      "step": 2340
+    },
+    {
+      "epoch": 0.019315798948205846,
+      "grad_norm": 0.6467030048370361,
+      "learning_rate": 4.999680733031814e-05,
+      "loss": 3.9464,
+      "step": 2370
+    },
+    {
+      "epoch": 0.01956030273236035,
+      "grad_norm": 0.6356080770492554,
+      "learning_rate": 4.9996704072287716e-05,
+      "loss": 3.9288,
+      "step": 2400
+    },
+    {
+      "epoch": 0.019804806516514856,
+      "grad_norm": 0.6498362421989441,
+      "learning_rate": 4.9996599171038984e-05,
+      "loss": 3.9345,
+      "step": 2430
+    },
+    {
+      "epoch": 0.020049310300669358,
+      "grad_norm": 0.6107562184333801,
+      "learning_rate": 4.999649262657886e-05,
+      "loss": 3.9368,
+      "step": 2460
+    },
+    {
+      "epoch": 0.020293814084823864,
+      "grad_norm": 0.6478250622749329,
+      "learning_rate": 4.999638443891434e-05,
+      "loss": 3.9273,
+      "step": 2490
+    },
+    {
+      "epoch": 0.02053831786897837,
+      "grad_norm": 0.631263256072998,
+      "learning_rate": 4.999627460805253e-05,
+      "loss": 3.9136,
+      "step": 2520
+    },
+    {
+      "epoch": 0.020782821653132874,
+      "grad_norm": 0.6120920777320862,
+      "learning_rate": 4.999616313400066e-05,
+      "loss": 3.8834,
+      "step": 2550
+    },
+    {
+      "epoch": 0.021027325437287376,
+      "grad_norm": 0.6151197552680969,
+      "learning_rate": 4.999605001676605e-05,
+      "loss": 3.8994,
+      "step": 2580
+    },
+    {
+      "epoch": 0.02127182922144188,
+      "grad_norm": 0.623715877532959,
+      "learning_rate": 4.9995935256356144e-05,
+      "loss": 3.8929,
+      "step": 2610
+    },
+    {
+      "epoch": 0.021516333005596387,
+      "grad_norm": 0.6688769459724426,
+      "learning_rate": 4.9995818852778476e-05,
+      "loss": 3.8499,
+      "step": 2640
+    },
+    {
+      "epoch": 0.021760836789750892,
+      "grad_norm": 0.6272155046463013,
+      "learning_rate": 4.999570080604071e-05,
+      "loss": 3.8861,
+      "step": 2670
+    },
+    {
+      "epoch": 0.022005340573905394,
+      "grad_norm": 0.597653329372406,
+      "learning_rate": 4.99955811161506e-05,
+      "loss": 3.8674,
+      "step": 2700
+    },
+    {
+      "epoch": 0.0222498443580599,
+      "grad_norm": 0.5608483552932739,
+      "learning_rate": 4.9995459783116004e-05,
+      "loss": 3.8493,
+      "step": 2730
+    },
+    {
+      "epoch": 0.022494348142214404,
+      "grad_norm": 0.5991063117980957,
+      "learning_rate": 4.999533680694493e-05,
+      "loss": 3.8454,
+      "step": 2760
+    },
+    {
+      "epoch": 0.02273885192636891,
+      "grad_norm": 0.5738102793693542,
+      "learning_rate": 4.9995212187645416e-05,
+      "loss": 3.8395,
+      "step": 2790
+    },
+    {
+      "epoch": 0.02298335571052341,
+      "grad_norm": 0.6145568490028381,
+      "learning_rate": 4.9995085925225693e-05,
+      "loss": 3.8313,
+      "step": 2820
+    },
+    {
+      "epoch": 0.023227859494677917,
+      "grad_norm": 0.6019515991210938,
+      "learning_rate": 4.999495801969404e-05,
+      "loss": 3.8277,
+      "step": 2850
+    },
+    {
+      "epoch": 0.023472363278832422,
+      "grad_norm": 0.6177758574485779,
+      "learning_rate": 4.9994828471058876e-05,
+      "loss": 3.8083,
+      "step": 2880
+    },
+    {
+      "epoch": 0.023716867062986924,
+      "grad_norm": 0.5616528987884521,
+      "learning_rate": 4.9994697279328714e-05,
+      "loss": 3.8114,
+      "step": 2910
+    },
+    {
+      "epoch": 0.02396137084714143,
+      "grad_norm": 0.6194447875022888,
+      "learning_rate": 4.9994564444512176e-05,
+      "loss": 3.8183,
+      "step": 2940
+    },
+    {
+      "epoch": 0.024205874631295934,
+      "grad_norm": 0.578895092010498,
+      "learning_rate": 4.9994429966618e-05,
+      "loss": 3.7871,
+      "step": 2970
+    },
+    {
+      "epoch": 0.02445037841545044,
+      "grad_norm": 0.6014060378074646,
+      "learning_rate": 4.999429384565502e-05,
+      "loss": 3.7711,
+      "step": 3000
+    },
+    {
+      "epoch": 0.02469488219960494,
+      "grad_norm": 0.5589067339897156,
+      "learning_rate": 4.999415608163217e-05,
+      "loss": 3.7533,
+      "step": 3030
+    },
+    {
+      "epoch": 0.024939385983759447,
+      "grad_norm": 0.5722873210906982,
+      "learning_rate": 4.999401667455854e-05,
+      "loss": 3.7585,
+      "step": 3060
+    },
+    {
+      "epoch": 0.025183889767913952,
+      "grad_norm": 0.5734965801239014,
+      "learning_rate": 4.9993875624443274e-05,
+      "loss": 3.77,
+      "step": 3090
+    },
+    {
+      "epoch": 0.025428393552068457,
+      "grad_norm": 0.5524207353591919,
+      "learning_rate": 4.9993732931295646e-05,
+      "loss": 3.718,
+      "step": 3120
+    },
+    {
+      "epoch": 0.02567289733622296,
+      "grad_norm": 0.5776082277297974,
+      "learning_rate": 4.999358859512503e-05,
+      "loss": 3.7573,
+      "step": 3150
+    },
+    {
+      "epoch": 0.025917401120377465,
+      "grad_norm": 0.5799595713615417,
+      "learning_rate": 4.9993442615940936e-05,
+      "loss": 3.7552,
+      "step": 3180
+    },
+    {
+      "epoch": 0.02616190490453197,
+      "grad_norm": 0.5820346474647522,
+      "learning_rate": 4.999329499375292e-05,
+      "loss": 3.7394,
+      "step": 3210
+    },
+    {
+      "epoch": 0.026406408688686475,
+      "grad_norm": 0.5450282096862793,
+      "learning_rate": 4.999314572857074e-05,
+      "loss": 3.7393,
+      "step": 3240
+    },
+    {
+      "epoch": 0.026650912472840977,
+      "grad_norm": 0.6213249564170837,
+      "learning_rate": 4.9992994820404174e-05,
+      "loss": 3.7191,
+      "step": 3270
+    },
+    {
+      "epoch": 0.026895416256995482,
+      "grad_norm": 0.5940688848495483,
+      "learning_rate": 4.999284226926314e-05,
+      "loss": 3.719,
+      "step": 3300
+    },
+    {
+      "epoch": 0.027139920041149988,
+      "grad_norm": 0.5779993534088135,
+      "learning_rate": 4.999268807515768e-05,
+      "loss": 3.7092,
+      "step": 3330
+    },
+    {
+      "epoch": 0.027384423825304493,
+      "grad_norm": 0.5899255871772766,
+      "learning_rate": 4.999253223809792e-05,
+      "loss": 3.6939,
+      "step": 3360
+    },
+    {
+      "epoch": 0.027628927609458995,
+      "grad_norm": 0.6330375671386719,
+      "learning_rate": 4.999237475809411e-05,
+      "loss": 3.7102,
+      "step": 3390
+    },
+    {
+      "epoch": 0.0278734313936135,
+      "grad_norm": 0.5771914720535278,
+      "learning_rate": 4.99922156351566e-05,
+      "loss": 3.7196,
+      "step": 3420
+    },
+    {
+      "epoch": 0.028117935177768005,
+      "grad_norm": 0.5769143104553223,
+      "learning_rate": 4.999205486929586e-05,
+      "loss": 3.7127,
+      "step": 3450
+    },
+    {
+      "epoch": 0.028362438961922507,
+      "grad_norm": 0.5581954121589661,
+      "learning_rate": 4.999189246052245e-05,
+      "loss": 3.689,
+      "step": 3480
+    },
+    {
+      "epoch": 0.028606942746077012,
+      "grad_norm": 0.6041043400764465,
+      "learning_rate": 4.999172840884704e-05,
+      "loss": 3.6831,
+      "step": 3510
+    },
+    {
+      "epoch": 0.028851446530231518,
+      "grad_norm": 0.5458335280418396,
+      "learning_rate": 4.999156271428043e-05,
+      "loss": 3.6694,
+      "step": 3540
+    },
+    {
+      "epoch": 0.029095950314386023,
+      "grad_norm": 0.5470607280731201,
+      "learning_rate": 4.9991395376833496e-05,
+      "loss": 3.6702,
+      "step": 3570
+    },
+    {
+      "epoch": 0.029340454098540525,
+      "grad_norm": 0.5878787040710449,
+      "learning_rate": 4.999122639651725e-05,
+      "loss": 3.6492,
+      "step": 3600
+    },
+    {
+      "epoch": 0.02958495788269503,
+      "grad_norm": 0.5691691637039185,
+      "learning_rate": 4.9991055773342795e-05,
+      "loss": 3.6812,
+      "step": 3630
+    },
+    {
+      "epoch": 0.029829461666849535,
+      "grad_norm": 0.5548356771469116,
+      "learning_rate": 4.9990883507321354e-05,
+      "loss": 3.645,
+      "step": 3660
+    },
+    {
+      "epoch": 0.03007396545100404,
+      "grad_norm": 0.5686156153678894,
+      "learning_rate": 4.999070959846424e-05,
+      "loss": 3.6505,
+      "step": 3690
+    },
+    {
+      "epoch": 0.030318469235158543,
+      "grad_norm": 0.5596534013748169,
+      "learning_rate": 4.999053404678289e-05,
+      "loss": 3.6532,
+      "step": 3720
+    },
+    {
+      "epoch": 0.030562973019313048,
+      "grad_norm": 0.54421067237854,
+      "learning_rate": 4.999035685228884e-05,
+      "loss": 3.6238,
+      "step": 3750
+    },
+    {
+      "epoch": 0.030807476803467553,
+      "grad_norm": 0.56732177734375,
+      "learning_rate": 4.999017801499375e-05,
+      "loss": 3.6449,
+      "step": 3780
+    },
+    {
+      "epoch": 0.03105198058762206,
+      "grad_norm": 0.5608410239219666,
+      "learning_rate": 4.998999753490937e-05,
+      "loss": 3.6344,
+      "step": 3810
+    },
+    {
+      "epoch": 0.03129648437177656,
+      "grad_norm": 0.5919491648674011,
+      "learning_rate": 4.998981541204757e-05,
+      "loss": 3.6213,
+      "step": 3840
+    },
+    {
+      "epoch": 0.031540988155931066,
+      "grad_norm": 0.5795233249664307,
+      "learning_rate": 4.998963164642031e-05,
+      "loss": 3.6239,
+      "step": 3870
+    },
+    {
+      "epoch": 0.03178549194008557,
+      "grad_norm": 0.5669205784797668,
+      "learning_rate": 4.9989446238039676e-05,
+      "loss": 3.5831,
+      "step": 3900
+    },
+    {
+      "epoch": 0.032029995724240076,
+      "grad_norm": 0.5817368626594543,
+      "learning_rate": 4.998925918691786e-05,
+      "loss": 3.6074,
+      "step": 3930
+    },
+    {
+      "epoch": 0.03227449950839458,
+      "grad_norm": 0.5662333369255066,
+      "learning_rate": 4.998907049306715e-05,
+      "loss": 3.5874,
+      "step": 3960
+    },
+    {
+      "epoch": 0.03251900329254909,
+      "grad_norm": 0.5641735792160034,
+      "learning_rate": 4.998888015649996e-05,
+      "loss": 3.6056,
+      "step": 3990
+    },
+    {
+      "epoch": 0.032763507076703585,
+      "grad_norm": 0.524918794631958,
+      "learning_rate": 4.99886881772288e-05,
+      "loss": 3.5963,
+      "step": 4020
+    },
+    {
+      "epoch": 0.03300801086085809,
+      "grad_norm": 0.5625722408294678,
+      "learning_rate": 4.998849455526628e-05,
+      "loss": 3.5917,
+      "step": 4050
+    },
+    {
+      "epoch": 0.033252514645012596,
+      "grad_norm": 0.5612478852272034,
+      "learning_rate": 4.998829929062515e-05,
+      "loss": 3.5792,
+      "step": 4080
+    },
+    {
+      "epoch": 0.0334970184291671,
+      "grad_norm": 0.5586293935775757,
+      "learning_rate": 4.998810238331822e-05,
+      "loss": 3.5708,
+      "step": 4110
+    },
+    {
+      "epoch": 0.033741522213321606,
+      "grad_norm": 0.53324955701828,
+      "learning_rate": 4.998790383335845e-05,
+      "loss": 3.5686,
+      "step": 4140
+    },
+    {
+      "epoch": 0.03398602599747611,
+      "grad_norm": 0.5210742950439453,
+      "learning_rate": 4.9987703640758894e-05,
+      "loss": 3.575,
+      "step": 4170
+    },
+    {
+      "epoch": 0.03423052978163062,
+      "grad_norm": 0.5591189861297607,
+      "learning_rate": 4.99875018055327e-05,
+      "loss": 3.5717,
+      "step": 4200
+    },
+    {
+      "epoch": 0.034475033565785115,
+      "grad_norm": 0.5435970425605774,
+      "learning_rate": 4.998729832769315e-05,
+      "loss": 3.5638,
+      "step": 4230
+    },
+    {
+      "epoch": 0.03471953734993962,
+      "grad_norm": 0.5489551424980164,
+      "learning_rate": 4.998709320725361e-05,
+      "loss": 3.574,
+      "step": 4260
+    },
+    {
+      "epoch": 0.034964041134094126,
+      "grad_norm": 0.5275290012359619,
+      "learning_rate": 4.998688644422756e-05,
+      "loss": 3.5695,
+      "step": 4290
+    },
+    {
+      "epoch": 0.03520854491824863,
+      "grad_norm": 0.583881139755249,
+      "learning_rate": 4.998667803862861e-05,
+      "loss": 3.5703,
+      "step": 4320
+    },
+    {
+      "epoch": 0.03545304870240314,
+      "grad_norm": 0.5317121744155884,
+      "learning_rate": 4.9986467990470445e-05,
+      "loss": 3.5668,
+      "step": 4350
+    },
+    {
+      "epoch": 0.03569755248655764,
+      "grad_norm": 0.5981696248054504,
+      "learning_rate": 4.998625629976688e-05,
+      "loss": 3.5411,
+      "step": 4380
+    },
+    {
+      "epoch": 0.03594205627071215,
+      "grad_norm": 0.55133455991745,
+      "learning_rate": 4.998604296653182e-05,
+      "loss": 3.521,
+      "step": 4410
+    },
+    {
+      "epoch": 0.03618656005486665,
+      "grad_norm": 0.5481094717979431,
+      "learning_rate": 4.99858279907793e-05,
+      "loss": 3.5421,
+      "step": 4440
+    },
+    {
+      "epoch": 0.03643106383902115,
+      "grad_norm": 0.5218031406402588,
+      "learning_rate": 4.998561137252346e-05,
+      "loss": 3.5305,
+      "step": 4470
+    },
+    {
+      "epoch": 0.036675567623175656,
+      "grad_norm": 0.5458360910415649,
+      "learning_rate": 4.9985393111778525e-05,
+      "loss": 3.5332,
+      "step": 4500
+    },
+    {
+      "epoch": 0.03692007140733016,
+      "grad_norm": 0.5501233339309692,
+      "learning_rate": 4.998517320855884e-05,
+      "loss": 3.5485,
+      "step": 4530
+    },
+    {
+      "epoch": 0.03716457519148467,
+      "grad_norm": 0.5359978675842285,
+      "learning_rate": 4.998495166287887e-05,
+      "loss": 3.534,
+      "step": 4560
+    },
+    {
+      "epoch": 0.03740907897563917,
+      "grad_norm": 0.5447133183479309,
+      "learning_rate": 4.998472847475318e-05,
+      "loss": 3.5176,
+      "step": 4590
+    },
+    {
+      "epoch": 0.03765358275979368,
+      "grad_norm": 0.5201069712638855,
+      "learning_rate": 4.998450364419643e-05,
+      "loss": 3.518,
+      "step": 4620
+    },
+    {
+      "epoch": 0.03789808654394818,
+      "grad_norm": 0.5251840353012085,
+      "learning_rate": 4.998427717122342e-05,
+      "loss": 3.5021,
+      "step": 4650
+    },
+    {
+      "epoch": 0.03814259032810268,
+      "grad_norm": 0.5204648375511169,
+      "learning_rate": 4.9984049055849024e-05,
+      "loss": 3.505,
+      "step": 4680
+    },
+    {
+      "epoch": 0.038387094112257186,
+      "grad_norm": 0.5137141942977905,
+      "learning_rate": 4.9983819298088234e-05,
+      "loss": 3.4997,
+      "step": 4710
+    },
+    {
+      "epoch": 0.03863159789641169,
+      "grad_norm": 0.5232805609703064,
+      "learning_rate": 4.9983587897956166e-05,
+      "loss": 3.5049,
+      "step": 4740
+    },
+    {
+      "epoch": 0.0388761016805662,
+      "grad_norm": 0.5415229201316833,
+      "learning_rate": 4.998335485546802e-05,
+      "loss": 3.5123,
+      "step": 4770
+    },
+    {
+      "epoch": 0.0391206054647207,
+      "grad_norm": 0.5097187161445618,
+      "learning_rate": 4.998312017063912e-05,
+      "loss": 3.4839,
+      "step": 4800
+    },
+    {
+      "epoch": 0.03936510924887521,
+      "grad_norm": 0.5637670755386353,
+      "learning_rate": 4.9982883843484895e-05,
+      "loss": 3.5084,
+      "step": 4830
+    },
+    {
+      "epoch": 0.03960961303302971,
+      "grad_norm": 0.5436129570007324,
+      "learning_rate": 4.998264587402088e-05,
+      "loss": 3.5184,
+      "step": 4860
+    },
+    {
+      "epoch": 0.03985411681718422,
+      "grad_norm": 0.5133230090141296,
+      "learning_rate": 4.9982406262262726e-05,
+      "loss": 3.48,
+      "step": 4890
+    },
+    {
+      "epoch": 0.040098620601338716,
+      "grad_norm": 0.5678510069847107,
+      "learning_rate": 4.9982165008226175e-05,
+      "loss": 3.4899,
+      "step": 4920
+    },
+    {
+      "epoch": 0.04034312438549322,
+      "grad_norm": 0.5121597647666931,
+      "learning_rate": 4.998192211192708e-05,
+      "loss": 3.4942,
+      "step": 4950
+    },
+    {
+      "epoch": 0.04058762816964773,
+      "grad_norm": 0.537929117679596,
+      "learning_rate": 4.9981677573381423e-05,
+      "loss": 3.4721,
+      "step": 4980
+    },
+    {
+      "epoch": 0.04083213195380223,
+      "grad_norm": 0.537289023399353,
+      "learning_rate": 4.9981431392605274e-05,
+      "loss": 3.4919,
+      "step": 5010
+    },
+    {
+      "epoch": 0.04107663573795674,
+      "grad_norm": 0.5594364404678345,
+      "learning_rate": 4.998118356961481e-05,
+      "loss": 3.4665,
+      "step": 5040
+    },
+    {
+      "epoch": 0.04132113952211124,
+      "grad_norm": 0.5009840726852417,
+      "learning_rate": 4.998093410442632e-05,
+      "loss": 3.4524,
+      "step": 5070
+    },
+    {
+      "epoch": 0.04156564330626575,
+      "grad_norm": 0.53801429271698,
+      "learning_rate": 4.998068299705623e-05,
+      "loss": 3.4629,
+      "step": 5100
+    },
+    {
+      "epoch": 0.041810147090420254,
+      "grad_norm": 0.5134426355361938,
+      "learning_rate": 4.9980430247521014e-05,
+      "loss": 3.4504,
+      "step": 5130
+    },
+    {
+      "epoch": 0.04205465087457475,
+      "grad_norm": 0.5095092058181763,
+      "learning_rate": 4.998017585583731e-05,
+      "loss": 3.4684,
+      "step": 5160
+    },
+    {
+      "epoch": 0.04229915465872926,
+      "grad_norm": 0.5233584642410278,
+      "learning_rate": 4.9979919822021824e-05,
+      "loss": 3.4726,
+      "step": 5190
+    },
+    {
+      "epoch": 0.04254365844288376,
+      "grad_norm": 0.5375773906707764,
+      "learning_rate": 4.99796621460914e-05,
+      "loss": 3.4502,
+      "step": 5220
+    },
+    {
+      "epoch": 0.04278816222703827,
+      "grad_norm": 0.5292348265647888,
+      "learning_rate": 4.9979402828062963e-05,
+      "loss": 3.4494,
+      "step": 5250
+    },
+    {
+      "epoch": 0.04303266601119277,
+      "grad_norm": 0.5170340538024902,
+      "learning_rate": 4.997914186795358e-05,
+      "loss": 3.4671,
+      "step": 5280
+    },
+    {
+      "epoch": 0.04327716979534728,
+      "grad_norm": 0.5027357339859009,
+      "learning_rate": 4.9978879265780385e-05,
+      "loss": 3.4424,
+      "step": 5310
+    },
+    {
+      "epoch": 0.043521673579501784,
+      "grad_norm": 0.4898991584777832,
+      "learning_rate": 4.997861502156066e-05,
+      "loss": 3.4515,
+      "step": 5340
+    },
+    {
+      "epoch": 0.04376617736365628,
+      "grad_norm": 0.5362656712532043,
+      "learning_rate": 4.997834913531176e-05,
+      "loss": 3.4209,
+      "step": 5370
+    },
+    {
+      "epoch": 0.04401068114781079,
+      "grad_norm": 0.5150277614593506,
+      "learning_rate": 4.9978081607051176e-05,
+      "loss": 3.4575,
+      "step": 5400
+    },
+    {
+      "epoch": 0.04425518493196529,
+      "grad_norm": 0.5086013674736023,
+      "learning_rate": 4.997781243679648e-05,
+      "loss": 3.4377,
+      "step": 5430
+    },
+    {
+      "epoch": 0.0444996887161198,
+      "grad_norm": 0.538194477558136,
+      "learning_rate": 4.9977541624565374e-05,
+      "loss": 3.4282,
+      "step": 5460
+    },
+    {
+      "epoch": 0.0447441925002743,
+      "grad_norm": 0.5045614838600159,
+      "learning_rate": 4.9977269170375665e-05,
+      "loss": 3.4109,
+      "step": 5490
+    },
+    {
+      "epoch": 0.04498869628442881,
+      "grad_norm": 0.5368480086326599,
+      "learning_rate": 4.997699507424526e-05,
+      "loss": 3.4341,
+      "step": 5520
+    },
+    {
+      "epoch": 0.045233200068583314,
+      "grad_norm": 0.6185830235481262,
+      "learning_rate": 4.997671933619218e-05,
+      "loss": 3.4203,
+      "step": 5550
+    },
+    {
+      "epoch": 0.04547770385273782,
+      "grad_norm": 0.4984918534755707,
+      "learning_rate": 4.9976441956234546e-05,
+      "loss": 3.4309,
+      "step": 5580
+    },
+    {
+      "epoch": 0.04572220763689232,
+      "grad_norm": 0.5066754221916199,
+      "learning_rate": 4.99761629343906e-05,
+      "loss": 3.3986,
+      "step": 5610
+    },
+    {
+      "epoch": 0.04596671142104682,
+      "grad_norm": 0.5132448673248291,
+      "learning_rate": 4.9975882270678676e-05,
+      "loss": 3.4126,
+      "step": 5640
+    },
+    {
+      "epoch": 0.04621121520520133,
+      "grad_norm": 0.5501627922058105,
+      "learning_rate": 4.997559996511723e-05,
+      "loss": 3.4057,
+      "step": 5670
+    },
+    {
+      "epoch": 0.04645571898935583,
+      "grad_norm": 0.4679185152053833,
+      "learning_rate": 4.997531601772481e-05,
+      "loss": 3.428,
+      "step": 5700
+    },
+    {
+      "epoch": 0.04670022277351034,
+      "grad_norm": 0.5020308494567871,
+      "learning_rate": 4.9975030428520084e-05,
+      "loss": 3.3977,
+      "step": 5730
+    },
+    {
+      "epoch": 0.046944726557664844,
+      "grad_norm": 0.5141638517379761,
+      "learning_rate": 4.997474319752184e-05,
+      "loss": 3.4044,
+      "step": 5760
+    },
+    {
+      "epoch": 0.04718923034181935,
+      "grad_norm": 0.5127893090248108,
+      "learning_rate": 4.997445432474895e-05,
+      "loss": 3.4071,
+      "step": 5790
+    },
+    {
+      "epoch": 0.04743373412597385,
+      "grad_norm": 0.5528402924537659,
+      "learning_rate": 4.9974163810220406e-05,
+      "loss": 3.3891,
+      "step": 5820
+    },
+    {
+      "epoch": 0.04767823791012835,
+      "grad_norm": 0.5092859864234924,
+      "learning_rate": 4.99738716539553e-05,
+      "loss": 3.4118,
+      "step": 5850
+    },
+    {
+      "epoch": 0.04792274169428286,
+      "grad_norm": 0.49626457691192627,
+      "learning_rate": 4.997357785597284e-05,
+      "loss": 3.3834,
+      "step": 5880
+    },
+    {
+      "epoch": 0.048167245478437364,
+      "grad_norm": 0.5132192969322205,
+      "learning_rate": 4.997328241629234e-05,
+      "loss": 3.3944,
+      "step": 5910
+    },
+    {
+      "epoch": 0.04841174926259187,
+      "grad_norm": 0.5104256868362427,
+      "learning_rate": 4.997298533493323e-05,
+      "loss": 3.4001,
+      "step": 5940
+    },
+    {
+      "epoch": 0.048656253046746374,
+      "grad_norm": 0.507205069065094,
+      "learning_rate": 4.997268661191503e-05,
+      "loss": 3.3867,
+      "step": 5970
+    },
+    {
+      "epoch": 0.04890075683090088,
+      "grad_norm": 0.6091153025627136,
+      "learning_rate": 4.9972386247257385e-05,
+      "loss": 3.3984,
+      "step": 6000
+    },
+    {
+      "epoch": 0.049145260615055385,
+      "grad_norm": 0.4803309142589569,
+      "learning_rate": 4.9972084240980025e-05,
+      "loss": 3.3703,
+      "step": 6030
+    },
+    {
+      "epoch": 0.04938976439920988,
+      "grad_norm": 0.515164852142334,
+      "learning_rate": 4.997178059310281e-05,
+      "loss": 3.3733,
+      "step": 6060
+    },
+    {
+      "epoch": 0.04963426818336439,
+      "grad_norm": 0.515275776386261,
+      "learning_rate": 4.997147530364571e-05,
+      "loss": 3.3802,
+      "step": 6090
+    },
+    {
+      "epoch": 0.049878771967518894,
+      "grad_norm": 0.5258405208587646,
+      "learning_rate": 4.9971168372628793e-05,
+      "loss": 3.3705,
+      "step": 6120
+    },
+    {
+      "epoch": 0.0501232757516734,
+      "grad_norm": 0.49037066102027893,
+      "learning_rate": 4.997085980007222e-05,
+      "loss": 3.37,
+      "step": 6150
+    },
+    {
+      "epoch": 0.050367779535827904,
+      "grad_norm": 0.47182220220565796,
+      "learning_rate": 4.99705495859963e-05,
+      "loss": 3.3705,
+      "step": 6180
+    },
+    {
+      "epoch": 0.05061228331998241,
+      "grad_norm": 0.4967211186885834,
+      "learning_rate": 4.99702377304214e-05,
+      "loss": 3.3743,
+      "step": 6210
+    },
+    {
+      "epoch": 0.050856787104136915,
+      "grad_norm": 0.7469291090965271,
+      "learning_rate": 4.9969924233368036e-05,
+      "loss": 3.3732,
+      "step": 6240
+    },
+    {
+      "epoch": 0.05110129088829142,
+      "grad_norm": 0.5237293839454651,
+      "learning_rate": 4.996960909485681e-05,
+      "loss": 3.3646,
+      "step": 6270
+    },
+    {
+      "epoch": 0.05134579467244592,
+      "grad_norm": 0.48839274048805237,
+      "learning_rate": 4.9969292314908446e-05,
+      "loss": 3.3665,
+      "step": 6300
+    },
+    {
+      "epoch": 0.051590298456600424,
+      "grad_norm": 0.48733407258987427,
+      "learning_rate": 4.996897389354376e-05,
+      "loss": 3.3646,
+      "step": 6330
+    },
+    {
+      "epoch": 0.05183480224075493,
+      "grad_norm": 0.4938340187072754,
+      "learning_rate": 4.9968653830783684e-05,
+      "loss": 3.3542,
+      "step": 6360
+    },
+    {
+      "epoch": 0.052079306024909434,
+      "grad_norm": 0.5001193284988403,
+      "learning_rate": 4.996833212664927e-05,
+      "loss": 3.3583,
+      "step": 6390
+    },
+    {
+      "epoch": 0.05232380980906394,
+      "grad_norm": 0.4772029221057892,
+      "learning_rate": 4.996800878116166e-05,
+      "loss": 3.336,
+      "step": 6420
+    },
+    {
+      "epoch": 0.052568313593218445,
+      "grad_norm": 0.49077659845352173,
+      "learning_rate": 4.99676837943421e-05,
+      "loss": 3.3681,
+      "step": 6450
+    },
+    {
+      "epoch": 0.05281281737737295,
+      "grad_norm": 0.5102148056030273,
+      "learning_rate": 4.996735716621196e-05,
+      "loss": 3.3437,
+      "step": 6480
+    },
+    {
+      "epoch": 0.05305732116152745,
+      "grad_norm": 0.5012289881706238,
+      "learning_rate": 4.996702889679272e-05,
+      "loss": 3.3536,
+      "step": 6510
+    },
+    {
+      "epoch": 0.053301824945681954,
+      "grad_norm": 0.4870162308216095,
+      "learning_rate": 4.996669898610595e-05,
+      "loss": 3.3513,
+      "step": 6540
+    },
+    {
+      "epoch": 0.05354632872983646,
+      "grad_norm": 0.48452699184417725,
+      "learning_rate": 4.996636743417334e-05,
+      "loss": 3.3415,
+      "step": 6570
+    },
+    {
+      "epoch": 0.053790832513990965,
+      "grad_norm": 0.4973999559879303,
+      "learning_rate": 4.996603424101669e-05,
+      "loss": 3.3304,
+      "step": 6600
+    },
+    {
+      "epoch": 0.05403533629814547,
+      "grad_norm": 0.4833717942237854,
+      "learning_rate": 4.996569940665789e-05,
+      "loss": 3.3424,
+      "step": 6630
+    },
+    {
+      "epoch": 0.054279840082299975,
+      "grad_norm": 0.5074206590652466,
+      "learning_rate": 4.996536293111896e-05,
+      "loss": 3.3302,
+      "step": 6660
+    },
+    {
+      "epoch": 0.05452434386645448,
+      "grad_norm": 0.4932290315628052,
+      "learning_rate": 4.996502481442202e-05,
+      "loss": 3.3388,
+      "step": 6690
+    },
+    {
+      "epoch": 0.054768847650608986,
+      "grad_norm": 0.48740679025650024,
+      "learning_rate": 4.9964685056589314e-05,
+      "loss": 3.3182,
+      "step": 6720
+    },
+    {
+      "epoch": 0.055013351434763484,
+      "grad_norm": 0.48752760887145996,
+      "learning_rate": 4.996434365764314e-05,
+      "loss": 3.3065,
+      "step": 6750
+    },
+    {
+      "epoch": 0.05525785521891799,
+      "grad_norm": 0.50692218542099,
+      "learning_rate": 4.996400061760597e-05,
+      "loss": 3.3379,
+      "step": 6780
+    },
+    {
+      "epoch": 0.055502359003072495,
+      "grad_norm": 0.479159414768219,
+      "learning_rate": 4.996365593650033e-05,
+      "loss": 3.3317,
+      "step": 6810
+    },
+    {
+      "epoch": 0.055746862787227,
+      "grad_norm": 0.498662531375885,
+      "learning_rate": 4.99633096143489e-05,
+      "loss": 3.3306,
+      "step": 6840
+    },
+    {
+      "epoch": 0.055991366571381505,
+      "grad_norm": 1.4371449947357178,
+      "learning_rate": 4.9962961651174436e-05,
+      "loss": 3.3334,
+      "step": 6870
+    },
+    {
+      "epoch": 0.05623587035553601,
+      "grad_norm": 0.49862873554229736,
+      "learning_rate": 4.9962612046999827e-05,
+      "loss": 3.3142,
+      "step": 6900
+    },
+    {
+      "epoch": 0.056480374139690516,
+      "grad_norm": 0.4759610593318939,
+      "learning_rate": 4.996226080184803e-05,
+      "loss": 3.3238,
+      "step": 6930
+    },
+    {
+      "epoch": 0.056724877923845014,
+      "grad_norm": 0.4844242334365845,
+      "learning_rate": 4.996190791574215e-05,
+      "loss": 3.3197,
+      "step": 6960
+    },
+    {
+      "epoch": 0.05696938170799952,
+      "grad_norm": 0.46844130754470825,
+      "learning_rate": 4.996155338870538e-05,
+      "loss": 3.2949,
+      "step": 6990
+    },
+    {
+      "epoch": 0.057213885492154025,
+      "grad_norm": 0.4850478768348694,
+      "learning_rate": 4.9961197220761035e-05,
+      "loss": 3.3143,
+      "step": 7020
+    },
+    {
+      "epoch": 0.05745838927630853,
+      "grad_norm": 0.4838846027851105,
+      "learning_rate": 4.996083941193252e-05,
+      "loss": 3.3015,
+      "step": 7050
+    },
+    {
+      "epoch": 0.057702893060463036,
+      "grad_norm": 0.49992483854293823,
+      "learning_rate": 4.9960479962243367e-05,
+      "loss": 3.3099,
+      "step": 7080
+    },
+    {
+      "epoch": 0.05794739684461754,
+      "grad_norm": 0.49964553117752075,
+      "learning_rate": 4.996011887171719e-05,
+      "loss": 3.3046,
+      "step": 7110
+    },
+    {
+      "epoch": 0.058191900628772046,
+      "grad_norm": 0.4723115563392639,
+      "learning_rate": 4.995975614037773e-05,
+      "loss": 3.3009,
+      "step": 7140
+    },
+    {
+      "epoch": 0.05843640441292655,
+      "grad_norm": 0.48575958609580994,
+      "learning_rate": 4.995939176824883e-05,
+      "loss": 3.3018,
+      "step": 7170
+    },
+    {
+      "epoch": 0.05868090819708105,
+      "grad_norm": 0.5264491438865662,
+      "learning_rate": 4.995902575535446e-05,
+      "loss": 3.2877,
+      "step": 7200
+    },
+    {
+      "epoch": 0.058925411981235555,
+      "grad_norm": 0.4813016355037689,
+      "learning_rate": 4.995865810171866e-05,
+      "loss": 3.2933,
+      "step": 7230
+    },
+    {
+      "epoch": 0.05916991576539006,
+      "grad_norm": 0.47151580452919006,
+      "learning_rate": 4.995828880736561e-05,
+      "loss": 3.3143,
+      "step": 7260
+    },
+    {
+      "epoch": 0.059414419549544566,
+      "grad_norm": 0.4812193512916565,
+      "learning_rate": 4.995791787231958e-05,
+      "loss": 3.2914,
+      "step": 7290
+    },
+    {
+      "epoch": 0.05965892333369907,
+      "grad_norm": 0.4926256239414215,
+      "learning_rate": 4.9957545296604965e-05,
+      "loss": 3.2756,
+      "step": 7320
+    },
+    {
+      "epoch": 0.059903427117853576,
+      "grad_norm": 0.47530651092529297,
+      "learning_rate": 4.9957171080246245e-05,
+      "loss": 3.3075,
+      "step": 7350
+    },
+    {
+      "epoch": 0.06014793090200808,
+      "grad_norm": 0.4838476777076721,
+      "learning_rate": 4.995679522326803e-05,
+      "loss": 3.292,
+      "step": 7380
+    },
+    {
+      "epoch": 0.06039243468616259,
+      "grad_norm": 0.48683232069015503,
+      "learning_rate": 4.995641772569502e-05,
+      "loss": 3.269,
+      "step": 7410
+    },
+    {
+      "epoch": 0.060636938470317085,
+      "grad_norm": 0.4818269908428192,
+      "learning_rate": 4.995603858755203e-05,
+      "loss": 3.2798,
+      "step": 7440
+    },
+    {
+      "epoch": 0.06088144225447159,
+      "grad_norm": 0.46415388584136963,
+      "learning_rate": 4.9955657808863985e-05,
+      "loss": 3.2768,
+      "step": 7470
+    },
+    {
+      "epoch": 0.061125946038626096,
+      "grad_norm": 0.48808780312538147,
+      "learning_rate": 4.995527538965593e-05,
+      "loss": 3.2797,
+      "step": 7500
+    },
+    {
+      "epoch": 0.0613704498227806,
+      "grad_norm": 0.4957239627838135,
+      "learning_rate": 4.995489132995298e-05,
+      "loss": 3.2912,
+      "step": 7530
+    },
+    {
+      "epoch": 0.061614953606935106,
+      "grad_norm": 0.4858773350715637,
+      "learning_rate": 4.99545056297804e-05,
+      "loss": 3.269,
+      "step": 7560
+    },
+    {
+      "epoch": 0.06185945739108961,
+      "grad_norm": 0.46054506301879883,
+      "learning_rate": 4.995411828916354e-05,
+      "loss": 3.2663,
+      "step": 7590
+    },
+    {
+      "epoch": 0.06210396117524412,
+      "grad_norm": 0.4704018831253052,
+      "learning_rate": 4.9953729308127874e-05,
+      "loss": 3.263,
+      "step": 7620
+    },
+    {
+      "epoch": 0.062348464959398615,
+      "grad_norm": 0.491974413394928,
+      "learning_rate": 4.995333868669895e-05,
+      "loss": 3.2709,
+      "step": 7650
+    },
+    {
+      "epoch": 0.06259296874355312,
+      "grad_norm": 0.46538054943084717,
+      "learning_rate": 4.995294642490246e-05,
+      "loss": 3.2818,
+      "step": 7680
+    },
+    {
+      "epoch": 0.06283747252770763,
+      "grad_norm": 0.49248039722442627,
+      "learning_rate": 4.995255252276418e-05,
+      "loss": 3.2581,
+      "step": 7710
+    },
+    {
+      "epoch": 0.06308197631186213,
+      "grad_norm": 0.45257478952407837,
+      "learning_rate": 4.9952156980310016e-05,
+      "loss": 3.2691,
+      "step": 7740
+    },
+    {
+      "epoch": 0.06332648009601663,
+      "grad_norm": 0.479942262172699,
+      "learning_rate": 4.9951759797565965e-05,
+      "loss": 3.276,
+      "step": 7770
+    },
+    {
+      "epoch": 0.06357098388017114,
+      "grad_norm": 0.48799383640289307,
+      "learning_rate": 4.995136097455815e-05,
+      "loss": 3.2668,
+      "step": 7800
+    },
+    {
+      "epoch": 0.06381548766432564,
+      "grad_norm": 0.46520474553108215,
+      "learning_rate": 4.995096051131276e-05,
+      "loss": 3.2509,
+      "step": 7830
+    },
+    {
+      "epoch": 0.06405999144848015,
+      "grad_norm": 0.4901852309703827,
+      "learning_rate": 4.995055840785614e-05,
+      "loss": 3.2383,
+      "step": 7860
+    },
+    {
+      "epoch": 0.06430449523263465,
+      "grad_norm": 0.46401068568229675,
+      "learning_rate": 4.995015466421473e-05,
+      "loss": 3.2626,
+      "step": 7890
+    },
+    {
+      "epoch": 0.06454899901678916,
+      "grad_norm": 0.46814706921577454,
+      "learning_rate": 4.9949749280415056e-05,
+      "loss": 3.2488,
+      "step": 7920
+    },
+    {
+      "epoch": 0.06479350280094366,
+      "grad_norm": 0.48936349153518677,
+      "learning_rate": 4.9949342256483766e-05,
+      "loss": 3.2368,
+      "step": 7950
+    },
+    {
+      "epoch": 0.06503800658509817,
+      "grad_norm": 0.4871247708797455,
+      "learning_rate": 4.9948933592447636e-05,
+      "loss": 3.2603,
+      "step": 7980
+    },
+    {
+      "epoch": 0.06528251036925267,
+      "grad_norm": 0.4689597487449646,
+      "learning_rate": 4.9948523288333506e-05,
+      "loss": 3.2437,
+      "step": 8010
+    },
+    {
+      "epoch": 0.06552701415340717,
+      "grad_norm": 0.45705971121788025,
+      "learning_rate": 4.994811134416836e-05,
+      "loss": 3.2415,
+      "step": 8040
+    },
+    {
+      "epoch": 0.06577151793756168,
+      "grad_norm": 0.46996039152145386,
+      "learning_rate": 4.994769775997927e-05,
+      "loss": 3.2545,
+      "step": 8070
+    },
+    {
+      "epoch": 0.06601602172171618,
+      "grad_norm": 0.4835875928401947,
+      "learning_rate": 4.994728253579345e-05,
+      "loss": 3.2587,
+      "step": 8100
+    },
+    {
+      "epoch": 0.0662605255058707,
+      "grad_norm": 0.4678030014038086,
+      "learning_rate": 4.9946865671638166e-05,
+      "loss": 3.2368,
+      "step": 8130
+    },
+    {
+      "epoch": 0.06650502929002519,
+      "grad_norm": 0.5105451345443726,
+      "learning_rate": 4.9946447167540835e-05,
+      "loss": 3.259,
+      "step": 8160
+    },
+    {
+      "epoch": 0.0667495330741797,
+      "grad_norm": 0.471935898065567,
+      "learning_rate": 4.994602702352896e-05,
+      "loss": 3.2631,
+      "step": 8190
+    },
+    {
+      "epoch": 0.0669940368583342,
+      "grad_norm": 0.48006853461265564,
+      "learning_rate": 4.994560523963018e-05,
+      "loss": 3.2313,
+      "step": 8220
+    },
+    {
+      "epoch": 0.0672385406424887,
+      "grad_norm": 0.46124544739723206,
+      "learning_rate": 4.9945181815872196e-05,
+      "loss": 3.215,
+      "step": 8250
+    },
+    {
+      "epoch": 0.06748304442664321,
+      "grad_norm": 0.481571763753891,
+      "learning_rate": 4.9944756752282855e-05,
+      "loss": 3.2357,
+      "step": 8280
+    },
+    {
+      "epoch": 0.06772754821079771,
+      "grad_norm": 0.47422316670417786,
+      "learning_rate": 4.994433004889011e-05,
+      "loss": 3.2279,
+      "step": 8310
+    },
+    {
+      "epoch": 0.06797205199495222,
+      "grad_norm": 0.48801445960998535,
+      "learning_rate": 4.994390170572199e-05,
+      "loss": 3.2369,
+      "step": 8340
+    },
+    {
+      "epoch": 0.06821655577910672,
+      "grad_norm": 0.46795913577079773,
+      "learning_rate": 4.994347172280667e-05,
+      "loss": 3.2187,
+      "step": 8370
+    },
+    {
+      "epoch": 0.06846105956326123,
+      "grad_norm": 0.4645237326622009,
+      "learning_rate": 4.994304010017241e-05,
+      "loss": 3.2183,
+      "step": 8400
+    },
+    {
+      "epoch": 0.06870556334741573,
+      "grad_norm": 0.45294952392578125,
+      "learning_rate": 4.994260683784758e-05,
+      "loss": 3.2449,
+      "step": 8430
+    },
+    {
+      "epoch": 0.06895006713157023,
+      "grad_norm": 0.47424617409706116,
+      "learning_rate": 4.9942171935860674e-05,
+      "loss": 3.2249,
+      "step": 8460
+    },
+    {
+      "epoch": 0.06919457091572474,
+      "grad_norm": 0.4657289683818817,
+      "learning_rate": 4.994173539424026e-05,
+      "loss": 3.2313,
+      "step": 8490
+    },
+    {
+      "epoch": 0.06943907469987924,
+      "grad_norm": 0.4828115701675415,
+      "learning_rate": 4.994129721301506e-05,
+      "loss": 3.236,
+      "step": 8520
+    },
+    {
+      "epoch": 0.06968357848403375,
+      "grad_norm": 0.4601866602897644,
+      "learning_rate": 4.994085739221386e-05,
+      "loss": 3.2459,
+      "step": 8550
+    },
+    {
+      "epoch": 0.06992808226818825,
+      "grad_norm": 0.45241278409957886,
+      "learning_rate": 4.994041593186558e-05,
+      "loss": 3.2077,
+      "step": 8580
+    },
+    {
+      "epoch": 0.07017258605234276,
+      "grad_norm": 0.47955596446990967,
+      "learning_rate": 4.993997283199924e-05,
+      "loss": 3.2378,
+      "step": 8610
+    },
+    {
+      "epoch": 0.07041708983649726,
+      "grad_norm": 0.4656619727611542,
+      "learning_rate": 4.993952809264397e-05,
+      "loss": 3.2277,
+      "step": 8640
+    },
+    {
+      "epoch": 0.07066159362065176,
+      "grad_norm": 0.45982396602630615,
+      "learning_rate": 4.9939081713829006e-05,
+      "loss": 3.2192,
+      "step": 8670
+    },
+    {
+      "epoch": 0.07090609740480627,
+      "grad_norm": 0.4598824381828308,
+      "learning_rate": 4.993863369558369e-05,
+      "loss": 3.2286,
+      "step": 8700
+    },
+    {
+      "epoch": 0.07115060118896077,
+      "grad_norm": 0.4744945168495178,
+      "learning_rate": 4.9938184037937466e-05,
+      "loss": 3.2201,
+      "step": 8730
+    },
+    {
+      "epoch": 0.07139510497311528,
+      "grad_norm": 0.46979376673698425,
+      "learning_rate": 4.993773274091991e-05,
+      "loss": 3.1986,
+      "step": 8760
+    },
+    {
+      "epoch": 0.07163960875726978,
+      "grad_norm": 0.46936362981796265,
+      "learning_rate": 4.993727980456067e-05,
+      "loss": 3.2047,
+      "step": 8790
+    },
+    {
+      "epoch": 0.0718841125414243,
+      "grad_norm": 0.4529063105583191,
+      "learning_rate": 4.993682522888954e-05,
+      "loss": 3.221,
+      "step": 8820
+    },
+    {
+      "epoch": 0.07212861632557879,
+      "grad_norm": 0.46270012855529785,
+      "learning_rate": 4.993636901393639e-05,
+      "loss": 3.2054,
+      "step": 8850
+    },
+    {
+      "epoch": 0.0723731201097333,
+      "grad_norm": 0.6890011429786682,
+      "learning_rate": 4.993591115973121e-05,
+      "loss": 3.2124,
+      "step": 8880
+    },
+    {
+      "epoch": 0.0726176238938878,
+      "grad_norm": 0.48755085468292236,
+      "learning_rate": 4.9935451666304105e-05,
+      "loss": 3.2177,
+      "step": 8910
+    },
+    {
+      "epoch": 0.0728621276780423,
+      "grad_norm": 0.4768828749656677,
+      "learning_rate": 4.993499053368528e-05,
+      "loss": 3.2075,
+      "step": 8940
+    },
+    {
+      "epoch": 0.07310663146219681,
+      "grad_norm": 0.5368286371231079,
+      "learning_rate": 4.993452776190504e-05,
+      "loss": 3.1981,
+      "step": 8970
+    },
+    {
+      "epoch": 0.07335113524635131,
+      "grad_norm": 0.45793619751930237,
+      "learning_rate": 4.993406335099382e-05,
+      "loss": 3.2143,
+      "step": 9000
+    },
+    {
+      "epoch": 0.07359563903050582,
+      "grad_norm": 0.459844172000885,
+      "learning_rate": 4.993359730098214e-05,
+      "loss": 3.2107,
+      "step": 9030
+    },
+    {
+      "epoch": 0.07384014281466032,
+      "grad_norm": 0.4570547640323639,
+      "learning_rate": 4.993312961190064e-05,
+      "loss": 3.1962,
+      "step": 9060
+    },
+    {
+      "epoch": 0.07408464659881484,
+      "grad_norm": 0.4817084074020386,
+      "learning_rate": 4.993266028378006e-05,
+      "loss": 3.1895,
+      "step": 9090
+    },
+    {
+      "epoch": 0.07432915038296933,
+      "grad_norm": 0.46276503801345825,
+      "learning_rate": 4.993218931665126e-05,
+      "loss": 3.1916,
+      "step": 9120
+    },
+    {
+      "epoch": 0.07457365416712383,
+      "grad_norm": 0.4731460511684418,
+      "learning_rate": 4.993171671054519e-05,
+      "loss": 3.2119,
+      "step": 9150
+    },
+    {
+      "epoch": 0.07481815795127834,
+      "grad_norm": 0.4732898473739624,
+      "learning_rate": 4.993124246549293e-05,
+      "loss": 3.1993,
+      "step": 9180
+    },
+    {
+      "epoch": 0.07506266173543284,
+      "grad_norm": 0.4521377682685852,
+      "learning_rate": 4.9930766581525645e-05,
+      "loss": 3.1948,
+      "step": 9210
+    },
+    {
+      "epoch": 0.07530716551958735,
+      "grad_norm": 0.466330885887146,
+      "learning_rate": 4.993028905867463e-05,
+      "loss": 3.1929,
+      "step": 9240
+    },
+    {
+      "epoch": 0.07555166930374185,
+      "grad_norm": 0.4678347110748291,
+      "learning_rate": 4.992980989697126e-05,
+      "loss": 3.1962,
+      "step": 9270
+    },
+    {
+      "epoch": 0.07579617308789637,
+      "grad_norm": 0.4650745391845703,
+      "learning_rate": 4.992932909644705e-05,
+      "loss": 3.19,
+      "step": 9300
+    },
+    {
+      "epoch": 0.07604067687205086,
+      "grad_norm": 0.45810264348983765,
+      "learning_rate": 4.9928846657133596e-05,
+      "loss": 3.1937,
+      "step": 9330
+    },
+    {
+      "epoch": 0.07628518065620536,
+      "grad_norm": 0.4602174162864685,
+      "learning_rate": 4.992836257906262e-05,
+      "loss": 3.2078,
+      "step": 9360
+    },
+    {
+      "epoch": 0.07652968444035987,
+      "grad_norm": 0.44673460721969604,
+      "learning_rate": 4.9927876862265935e-05,
+      "loss": 3.1726,
+      "step": 9390
+    },
+    {
+      "epoch": 0.07677418822451437,
+      "grad_norm": 0.4698408246040344,
+      "learning_rate": 4.992738950677548e-05,
+      "loss": 3.1658,
+      "step": 9420
+    },
+    {
+      "epoch": 0.07701869200866888,
+      "grad_norm": 0.47652050852775574,
+      "learning_rate": 4.992690051262329e-05,
+      "loss": 3.1642,
+      "step": 9450
+    },
+    {
+      "epoch": 0.07726319579282338,
+      "grad_norm": 0.4623103737831116,
+      "learning_rate": 4.99264098798415e-05,
+      "loss": 3.1788,
+      "step": 9480
+    },
+    {
+      "epoch": 0.0775076995769779,
+      "grad_norm": 0.4699987471103668,
+      "learning_rate": 4.992591760846238e-05,
+      "loss": 3.1679,
+      "step": 9510
+    },
+    {
+      "epoch": 0.0777522033611324,
+      "grad_norm": 0.466349720954895,
+      "learning_rate": 4.9925423698518274e-05,
+      "loss": 3.1674,
+      "step": 9540
+    },
+    {
+      "epoch": 0.0779967071452869,
+      "grad_norm": 0.4621599316596985,
+      "learning_rate": 4.992492815004166e-05,
+      "loss": 3.193,
+      "step": 9570
+    },
+    {
+      "epoch": 0.0782412109294414,
+      "grad_norm": 0.4567157030105591,
+      "learning_rate": 4.992443096306512e-05,
+      "loss": 3.1681,
+      "step": 9600
+    },
+    {
+      "epoch": 0.0784857147135959,
+      "grad_norm": 0.45757701992988586,
+      "learning_rate": 4.992393213762132e-05,
+      "loss": 3.2002,
+      "step": 9630
+    },
+    {
+      "epoch": 0.07873021849775041,
+      "grad_norm": 0.45566242933273315,
+      "learning_rate": 4.992343167374307e-05,
+      "loss": 3.1651,
+      "step": 9660
+    },
+    {
+      "epoch": 0.07897472228190491,
+      "grad_norm": 0.4485650360584259,
+      "learning_rate": 4.992292957146326e-05,
+      "loss": 3.1613,
+      "step": 9690
+    },
+    {
+      "epoch": 0.07921922606605943,
+      "grad_norm": 0.4732878804206848,
+      "learning_rate": 4.992242583081489e-05,
+      "loss": 3.1602,
+      "step": 9720
+    },
+    {
+      "epoch": 0.07946372985021392,
+      "grad_norm": 0.4469018280506134,
+      "learning_rate": 4.992192045183109e-05,
+      "loss": 3.1487,
+      "step": 9750
+    },
+    {
+      "epoch": 0.07970823363436844,
+      "grad_norm": 0.45860594511032104,
+      "learning_rate": 4.9921413434545075e-05,
+      "loss": 3.1516,
+      "step": 9780
+    },
+    {
+      "epoch": 0.07995273741852293,
+      "grad_norm": 0.45328783988952637,
+      "learning_rate": 4.992090477899018e-05,
+      "loss": 3.1481,
+      "step": 9810
+    },
+    {
+      "epoch": 0.08019724120267743,
+      "grad_norm": 0.48180997371673584,
+      "learning_rate": 4.992039448519982e-05,
+      "loss": 3.1459,
+      "step": 9840
+    },
+    {
+      "epoch": 0.08044174498683195,
+      "grad_norm": 0.45059671998023987,
+      "learning_rate": 4.9919882553207566e-05,
+      "loss": 3.1432,
+      "step": 9870
+    },
+    {
+      "epoch": 0.08068624877098644,
+      "grad_norm": 0.4592690169811249,
+      "learning_rate": 4.9919368983047066e-05,
+      "loss": 3.1532,
+      "step": 9900
+    },
+    {
+      "epoch": 0.08093075255514096,
+      "grad_norm": 0.4452771842479706,
+      "learning_rate": 4.9918853774752074e-05,
+      "loss": 3.171,
+      "step": 9930
+    },
+    {
+      "epoch": 0.08117525633929545,
+      "grad_norm": 0.45305460691452026,
+      "learning_rate": 4.991833692835646e-05,
+      "loss": 3.1422,
+      "step": 9960
+    },
+    {
+      "epoch": 0.08141976012344997,
+      "grad_norm": 0.47526153922080994,
+      "learning_rate": 4.9917818443894203e-05,
+      "loss": 3.1565,
+      "step": 9990
+    },
+    {
+      "epoch": 0.08166426390760446,
+      "grad_norm": 0.4707624614238739,
+      "learning_rate": 4.991729832139939e-05,
+      "loss": 3.1481,
+      "step": 10020
+    },
+    {
+      "epoch": 0.08190876769175896,
+      "grad_norm": 0.44442108273506165,
+      "learning_rate": 4.991677656090621e-05,
+      "loss": 3.1455,
+      "step": 10050
+    },
+    {
+      "epoch": 0.08215327147591348,
+      "grad_norm": 0.4681081473827362,
+      "learning_rate": 4.991625316244896e-05,
+      "loss": 3.1703,
+      "step": 10080
+    },
+    {
+      "epoch": 0.08239777526006797,
+      "grad_norm": 0.4762340486049652,
+      "learning_rate": 4.991572812606205e-05,
+      "loss": 3.1636,
+      "step": 10110
+    },
+    {
+      "epoch": 0.08264227904422249,
+      "grad_norm": 0.4579741954803467,
+      "learning_rate": 4.991520145177998e-05,
+      "loss": 3.1546,
+      "step": 10140
+    },
+    {
+      "epoch": 0.08288678282837698,
+      "grad_norm": 0.45810720324516296,
+      "learning_rate": 4.991467313963739e-05,
+      "loss": 3.1456,
+      "step": 10170
+    },
+    {
+      "epoch": 0.0831312866125315,
+      "grad_norm": 0.4368288516998291,
+      "learning_rate": 4.991414318966901e-05,
+      "loss": 3.1393,
+      "step": 10200
+    },
+    {
+      "epoch": 0.083375790396686,
+      "grad_norm": 0.4572373330593109,
+      "learning_rate": 4.991361160190966e-05,
+      "loss": 3.1258,
+      "step": 10230
+    },
+    {
+      "epoch": 0.08362029418084051,
+      "grad_norm": 0.488158255815506,
+      "learning_rate": 4.9913078376394304e-05,
+      "loss": 3.1563,
+      "step": 10260
+    },
+    {
+      "epoch": 0.083864797964995,
+      "grad_norm": 0.4564719498157501,
+      "learning_rate": 4.991254351315799e-05,
+      "loss": 3.1344,
+      "step": 10290
+    },
+    {
+      "epoch": 0.0841093017491495,
+      "grad_norm": 0.4469270408153534,
+      "learning_rate": 4.991200701223587e-05,
+      "loss": 3.1493,
+      "step": 10320
+    },
+    {
+      "epoch": 0.08435380553330402,
+      "grad_norm": 0.46140623092651367,
+      "learning_rate": 4.991146887366323e-05,
+      "loss": 3.1517,
+      "step": 10350
+    },
+    {
+      "epoch": 0.08459830931745851,
+      "grad_norm": 0.4541148841381073,
+      "learning_rate": 4.991092909747542e-05,
+      "loss": 3.1442,
+      "step": 10380
+    },
+    {
+      "epoch": 0.08484281310161303,
+      "grad_norm": 0.4556724727153778,
+      "learning_rate": 4.9910387683707946e-05,
+      "loss": 3.1429,
+      "step": 10410
+    },
+    {
+      "epoch": 0.08508731688576752,
+      "grad_norm": 0.4350408613681793,
+      "learning_rate": 4.9909844632396386e-05,
+      "loss": 3.1356,
+      "step": 10440
+    },
+    {
+      "epoch": 0.08533182066992204,
+      "grad_norm": 0.45419394969940186,
+      "learning_rate": 4.9909299943576445e-05,
+      "loss": 3.1321,
+      "step": 10470
+    },
+    {
+      "epoch": 0.08557632445407654,
+      "grad_norm": 0.4699922204017639,
+      "learning_rate": 4.990875361728393e-05,
+      "loss": 3.1406,
+      "step": 10500
+    },
+    {
+      "epoch": 0.08582082823823103,
+      "grad_norm": 0.45109736919403076,
+      "learning_rate": 4.990820565355475e-05,
+      "loss": 3.1369,
+      "step": 10530
+    },
+    {
+      "epoch": 0.08606533202238555,
+      "grad_norm": 0.45368725061416626,
+      "learning_rate": 4.990765605242493e-05,
+      "loss": 3.1299,
+      "step": 10560
+    },
+    {
+      "epoch": 0.08630983580654004,
+      "grad_norm": 0.4491060674190521,
+      "learning_rate": 4.990710481393061e-05,
+      "loss": 3.1139,
+      "step": 10590
+    },
+    {
+      "epoch": 0.08655433959069456,
+      "grad_norm": 0.47071024775505066,
+      "learning_rate": 4.9906551938108003e-05,
+      "loss": 3.159,
+      "step": 10620
+    },
+    {
+      "epoch": 0.08679884337484906,
+      "grad_norm": 0.4578341841697693,
+      "learning_rate": 4.990599742499347e-05,
+      "loss": 3.1449,
+      "step": 10650
+    },
+    {
+      "epoch": 0.08704334715900357,
+      "grad_norm": 0.4422541856765747,
+      "learning_rate": 4.990544127462346e-05,
+      "loss": 3.125,
+      "step": 10680
+    },
+    {
+      "epoch": 0.08728785094315807,
+      "grad_norm": 0.44820329546928406,
+      "learning_rate": 4.9904883487034537e-05,
+      "loss": 3.1328,
+      "step": 10710
+    },
+    {
+      "epoch": 0.08753235472731256,
+      "grad_norm": 0.4683224856853485,
+      "learning_rate": 4.990432406226336e-05,
+      "loss": 3.1296,
+      "step": 10740
+    },
+    {
+      "epoch": 0.08777685851146708,
+      "grad_norm": 0.4463740885257721,
+      "learning_rate": 4.990376300034671e-05,
+      "loss": 3.1177,
+      "step": 10770
+    },
+    {
+      "epoch": 0.08802136229562157,
+      "grad_norm": 0.4537316560745239,
+      "learning_rate": 4.990320030132147e-05,
+      "loss": 3.1269,
+      "step": 10800
+    },
+    {
+      "epoch": 0.08826586607977609,
+      "grad_norm": 0.472412109375,
+      "learning_rate": 4.990263596522462e-05,
+      "loss": 3.1378,
+      "step": 10830
+    },
+    {
+      "epoch": 0.08851036986393059,
+      "grad_norm": 0.4640551805496216,
+      "learning_rate": 4.9902069992093275e-05,
+      "loss": 3.1381,
+      "step": 10860
+    },
+    {
+      "epoch": 0.0887548736480851,
+      "grad_norm": 0.4387233853340149,
+      "learning_rate": 4.990150238196463e-05,
+      "loss": 3.1299,
+      "step": 10890
+    },
+    {
+      "epoch": 0.0889993774322396,
+      "grad_norm": 0.4460201561450958,
+      "learning_rate": 4.9900933134876e-05,
+      "loss": 3.127,
+      "step": 10920
+    },
+    {
+      "epoch": 0.0892438812163941,
+      "grad_norm": 0.4335034191608429,
+      "learning_rate": 4.990036225086481e-05,
+      "loss": 3.1032,
+      "step": 10950
+    },
+    {
+      "epoch": 0.0894883850005486,
+      "grad_norm": 0.44630929827690125,
+      "learning_rate": 4.9899789729968585e-05,
+      "loss": 3.119,
+      "step": 10980
+    },
+    {
+      "epoch": 0.0897328887847031,
+      "grad_norm": 0.4360102117061615,
+      "learning_rate": 4.989921557222496e-05,
+      "loss": 3.1123,
+      "step": 11010
+    },
+    {
+      "epoch": 0.08997739256885762,
+      "grad_norm": 0.45332589745521545,
+      "learning_rate": 4.989863977767167e-05,
+      "loss": 3.123,
+      "step": 11040
+    },
+    {
+      "epoch": 0.09022189635301212,
+      "grad_norm": 0.4539777338504791,
+      "learning_rate": 4.989806234634659e-05,
+      "loss": 3.1144,
+      "step": 11070
+    },
+    {
+      "epoch": 0.09046640013716663,
+      "grad_norm": 0.5205842852592468,
+      "learning_rate": 4.989748327828765e-05,
+      "loss": 3.1199,
+      "step": 11100
+    },
+    {
+      "epoch": 0.09071090392132113,
+      "grad_norm": 0.44270429015159607,
+      "learning_rate": 4.989690257353294e-05,
+      "loss": 3.0783,
+      "step": 11130
+    },
+    {
+      "epoch": 0.09095540770547564,
+      "grad_norm": 0.4286257326602936,
+      "learning_rate": 4.989632023212062e-05,
+      "loss": 3.1162,
+      "step": 11160
+    },
+    {
+      "epoch": 0.09119991148963014,
+      "grad_norm": 0.49412211775779724,
+      "learning_rate": 4.9895736254088975e-05,
+      "loss": 3.1142,
+      "step": 11190
+    },
+    {
+      "epoch": 0.09144441527378463,
+      "grad_norm": 0.4755348563194275,
+      "learning_rate": 4.989515063947641e-05,
+      "loss": 3.1173,
+      "step": 11220
+    },
+    {
+      "epoch": 0.09168891905793915,
+      "grad_norm": 0.48216554522514343,
+      "learning_rate": 4.9894563388321395e-05,
+      "loss": 3.1027,
+      "step": 11250
+    },
+    {
+      "epoch": 0.09193342284209365,
+      "grad_norm": 0.4571470320224762,
+      "learning_rate": 4.989397450066254e-05,
+      "loss": 3.1191,
+      "step": 11280
+    },
+    {
+      "epoch": 0.09217792662624816,
+      "grad_norm": 0.4681689441204071,
+      "learning_rate": 4.989338397653858e-05,
+      "loss": 3.097,
+      "step": 11310
+    },
+    {
+      "epoch": 0.09242243041040266,
+      "grad_norm": 0.4475450813770294,
+      "learning_rate": 4.98927918159883e-05,
+      "loss": 3.118,
+      "step": 11340
+    },
+    {
+      "epoch": 0.09266693419455717,
+      "grad_norm": 0.4555697739124298,
+      "learning_rate": 4.989219801905066e-05,
+      "loss": 3.1078,
+      "step": 11370
+    },
+    {
+      "epoch": 0.09291143797871167,
+      "grad_norm": 0.45538437366485596,
+      "learning_rate": 4.989160258576469e-05,
+      "loss": 3.1106,
+      "step": 11400
+    },
+    {
+      "epoch": 0.09315594176286617,
+      "grad_norm": 0.48153620958328247,
+      "learning_rate": 4.98910055161695e-05,
+      "loss": 3.0844,
+      "step": 11430
+    },
+    {
+      "epoch": 0.09340044554702068,
+      "grad_norm": 0.42441487312316895,
+      "learning_rate": 4.989040681030437e-05,
+      "loss": 3.1159,
+      "step": 11460
+    },
+    {
+      "epoch": 0.09364494933117518,
+      "grad_norm": 0.4604051411151886,
+      "learning_rate": 4.988980646820865e-05,
+      "loss": 3.1035,
+      "step": 11490
+    },
+    {
+      "epoch": 0.09388945311532969,
+      "grad_norm": 0.43700629472732544,
+      "learning_rate": 4.9889204489921804e-05,
+      "loss": 3.0811,
+      "step": 11520
+    },
+    {
+      "epoch": 0.09413395689948419,
+      "grad_norm": 0.4468238949775696,
+      "learning_rate": 4.9888600875483404e-05,
+      "loss": 3.1104,
+      "step": 11550
+    },
+    {
+      "epoch": 0.0943784606836387,
+      "grad_norm": 0.44943588972091675,
+      "learning_rate": 4.9887995624933137e-05,
+      "loss": 3.0917,
+      "step": 11580
+    },
+    {
+      "epoch": 0.0946229644677932,
+      "grad_norm": 0.4467644691467285,
+      "learning_rate": 4.988738873831078e-05,
+      "loss": 3.087,
+      "step": 11610
+    },
+    {
+      "epoch": 0.0948674682519477,
+      "grad_norm": 0.44013822078704834,
+      "learning_rate": 4.988678021565623e-05,
+      "loss": 3.0868,
+      "step": 11640
+    },
+    {
+      "epoch": 0.09511197203610221,
+      "grad_norm": 0.4420963525772095,
+      "learning_rate": 4.988617005700949e-05,
+      "loss": 3.102,
+      "step": 11670
+    },
+    {
+      "epoch": 0.0953564758202567,
+      "grad_norm": 0.45339009165763855,
+      "learning_rate": 4.988555826241068e-05,
+      "loss": 3.1001,
+      "step": 11700
+    },
+    {
+      "epoch": 0.09560097960441122,
+      "grad_norm": 0.4557483196258545,
+      "learning_rate": 4.988494483190001e-05,
+      "loss": 3.0894,
+      "step": 11730
+    },
+    {
+      "epoch": 0.09584548338856572,
+      "grad_norm": 0.43180930614471436,
+      "learning_rate": 4.98843297655178e-05,
+      "loss": 3.1116,
+      "step": 11760
+    },
+    {
+      "epoch": 0.09608998717272023,
+      "grad_norm": 0.43689119815826416,
+      "learning_rate": 4.988371306330449e-05,
+      "loss": 3.0829,
+      "step": 11790
+    },
+    {
+      "epoch": 0.09633449095687473,
+      "grad_norm": 0.45857593417167664,
+      "learning_rate": 4.9883094725300625e-05,
+      "loss": 3.0991,
+      "step": 11820
+    },
+    {
+      "epoch": 0.09657899474102924,
+      "grad_norm": 0.46954238414764404,
+      "learning_rate": 4.9882474751546846e-05,
+      "loss": 3.0792,
+      "step": 11850
+    },
+    {
+      "epoch": 0.09682349852518374,
+      "grad_norm": 0.4400712847709656,
+      "learning_rate": 4.98818531420839e-05,
+      "loss": 3.0839,
+      "step": 11880
+    },
+    {
+      "epoch": 0.09706800230933824,
+      "grad_norm": 0.45475077629089355,
+      "learning_rate": 4.988122989695267e-05,
+      "loss": 3.1036,
+      "step": 11910
+    },
+    {
+      "epoch": 0.09731250609349275,
+      "grad_norm": 0.451103150844574,
+      "learning_rate": 4.9880605016194116e-05,
+      "loss": 3.0762,
+      "step": 11940
+    },
+    {
+      "epoch": 0.09755700987764725,
+      "grad_norm": 0.44475677609443665,
+      "learning_rate": 4.9879978499849304e-05,
+      "loss": 3.082,
+      "step": 11970
+    },
+    {
+      "epoch": 0.09780151366180176,
+      "grad_norm": 0.4441121518611908,
+      "learning_rate": 4.987935034795944e-05,
+      "loss": 3.0826,
+      "step": 12000
+    },
+    {
+      "epoch": 0.09804601744595626,
+      "grad_norm": 0.4929714798927307,
+      "learning_rate": 4.9878720560565803e-05,
+      "loss": 3.0762,
+      "step": 12030
+    },
+    {
+      "epoch": 0.09829052123011077,
+      "grad_norm": 0.4454306364059448,
+      "learning_rate": 4.9878089137709806e-05,
+      "loss": 3.0864,
+      "step": 12060
+    },
+    {
+      "epoch": 0.09853502501426527,
+      "grad_norm": 0.45685863494873047,
+      "learning_rate": 4.987745607943295e-05,
+      "loss": 3.1007,
+      "step": 12090
+    },
+    {
+      "epoch": 0.09877952879841977,
+      "grad_norm": 0.43002381920814514,
+      "learning_rate": 4.9876821385776854e-05,
+      "loss": 3.0768,
+      "step": 12120
+    },
+    {
+      "epoch": 0.09902403258257428,
+      "grad_norm": 0.4407506585121155,
+      "learning_rate": 4.9876185056783226e-05,
+      "loss": 3.0818,
+      "step": 12150
+    },
+    {
+      "epoch": 0.09926853636672878,
+      "grad_norm": 0.4490843117237091,
+      "learning_rate": 4.987554709249391e-05,
+      "loss": 3.0837,
+      "step": 12180
+    },
+    {
+      "epoch": 0.09951304015088329,
+      "grad_norm": 0.45774146914482117,
+      "learning_rate": 4.987490749295085e-05,
+      "loss": 3.0757,
+      "step": 12210
+    },
+    {
+      "epoch": 0.09975754393503779,
+      "grad_norm": 0.45156994462013245,
+      "learning_rate": 4.9874266258196084e-05,
+      "loss": 3.0738,
+      "step": 12240
+    },
+    {
+      "epoch": 0.1000020477191923,
+      "grad_norm": 0.42710986733436584,
+      "learning_rate": 4.9873623388271764e-05,
+      "loss": 3.0838,
+      "step": 12270
+    },
+    {
+      "epoch": 0.1002465515033468,
+      "grad_norm": 0.4447808563709259,
+      "learning_rate": 4.987297888322015e-05,
+      "loss": 3.0669,
+      "step": 12300
+    },
+    {
+      "epoch": 0.1004910552875013,
+      "grad_norm": 0.4497504234313965,
+      "learning_rate": 4.9872332743083615e-05,
+      "loss": 3.0806,
+      "step": 12330
+    },
+    {
+      "epoch": 0.10073555907165581,
+      "grad_norm": 0.44097018241882324,
+      "learning_rate": 4.987168496790463e-05,
+      "loss": 3.0687,
+      "step": 12360
+    },
+    {
+      "epoch": 0.10098006285581031,
+      "grad_norm": 0.4529583752155304,
+      "learning_rate": 4.9871035557725774e-05,
+      "loss": 3.0707,
+      "step": 12390
+    },
+    {
+      "epoch": 0.10122456663996482,
+      "grad_norm": 0.4495998024940491,
+      "learning_rate": 4.987038451258975e-05,
+      "loss": 3.0772,
+      "step": 12420
+    },
+    {
+      "epoch": 0.10146907042411932,
+      "grad_norm": 0.46496671438217163,
+      "learning_rate": 4.9869731832539346e-05,
+      "loss": 3.0572,
+      "step": 12450
+    },
+    {
+      "epoch": 0.10171357420827383,
+      "grad_norm": 0.43100932240486145,
+      "learning_rate": 4.9869077517617474e-05,
+      "loss": 3.0545,
+      "step": 12480
+    },
+    {
+      "epoch": 0.10195807799242833,
+      "grad_norm": 0.4375341534614563,
+      "learning_rate": 4.986842156786714e-05,
+      "loss": 3.0411,
+      "step": 12510
+    },
+    {
+      "epoch": 0.10220258177658284,
+      "grad_norm": 0.463009774684906,
+      "learning_rate": 4.9867763983331474e-05,
+      "loss": 3.0525,
+      "step": 12540
+    },
+    {
+      "epoch": 0.10244708556073734,
+      "grad_norm": 0.44937971234321594,
+      "learning_rate": 4.9867104764053697e-05,
+      "loss": 3.0572,
+      "step": 12570
+    },
+    {
+      "epoch": 0.10269158934489184,
+      "grad_norm": 0.43725940585136414,
+      "learning_rate": 4.986644391007714e-05,
+      "loss": 3.0464,
+      "step": 12600
+    },
+    {
+      "epoch": 0.10293609312904635,
+      "grad_norm": 0.4457106292247772,
+      "learning_rate": 4.986578142144525e-05,
+      "loss": 3.0667,
+      "step": 12630
+    },
+    {
+      "epoch": 0.10318059691320085,
+      "grad_norm": 0.43846726417541504,
+      "learning_rate": 4.986511729820158e-05,
+      "loss": 3.064,
+      "step": 12660
+    },
+    {
+      "epoch": 0.10342510069735536,
+      "grad_norm": 0.4581950306892395,
+      "learning_rate": 4.986445154038979e-05,
+      "loss": 3.0567,
+      "step": 12690
+    },
+    {
+      "epoch": 0.10366960448150986,
+      "grad_norm": 0.45388519763946533,
+      "learning_rate": 4.986378414805364e-05,
+      "loss": 3.0679,
+      "step": 12720
+    },
+    {
+      "epoch": 0.10391410826566437,
+      "grad_norm": 0.43947508931159973,
+      "learning_rate": 4.9863115121237006e-05,
+      "loss": 3.0475,
+      "step": 12750
+    },
+    {
+      "epoch": 0.10415861204981887,
+      "grad_norm": 0.4460260570049286,
+      "learning_rate": 4.986244445998386e-05,
+      "loss": 3.0657,
+      "step": 12780
+    },
+    {
+      "epoch": 0.10440311583397337,
+      "grad_norm": 0.44423437118530273,
+      "learning_rate": 4.9861772164338304e-05,
+      "loss": 3.052,
+      "step": 12810
+    },
+    {
+      "epoch": 0.10464761961812788,
+      "grad_norm": 0.4410192668437958,
+      "learning_rate": 4.986109823434452e-05,
+      "loss": 3.0528,
+      "step": 12840
+    },
+    {
+      "epoch": 0.10489212340228238,
+      "grad_norm": 0.46037909388542175,
+      "learning_rate": 4.986042267004681e-05,
+      "loss": 3.0629,
+      "step": 12870
+    },
+    {
+      "epoch": 0.10513662718643689,
+      "grad_norm": 0.4454433023929596,
+      "learning_rate": 4.98597454714896e-05,
+      "loss": 3.0608,
+      "step": 12900
+    },
+    {
+      "epoch": 0.10538113097059139,
+      "grad_norm": 0.47135457396507263,
+      "learning_rate": 4.985906663871739e-05,
+      "loss": 3.066,
+      "step": 12930
+    },
+    {
+      "epoch": 0.1056256347547459,
+      "grad_norm": 0.4219493865966797,
+      "learning_rate": 4.9858386171774804e-05,
+      "loss": 3.0631,
+      "step": 12960
+    },
+    {
+      "epoch": 0.1058701385389004,
+      "grad_norm": 0.4326363205909729,
+      "learning_rate": 4.985770407070659e-05,
+      "loss": 3.0365,
+      "step": 12990
+    },
+    {
+      "epoch": 0.1061146423230549,
+      "grad_norm": 0.44705840945243835,
+      "learning_rate": 4.985702033555757e-05,
+      "loss": 3.0607,
+      "step": 13020
+    },
+    {
+      "epoch": 0.10635914610720941,
+      "grad_norm": 0.44811519980430603,
+      "learning_rate": 4.9856334966372705e-05,
+      "loss": 3.05,
+      "step": 13050
+    },
+    {
+      "epoch": 0.10660364989136391,
+      "grad_norm": 0.4571828544139862,
+      "learning_rate": 4.985564796319705e-05,
+      "loss": 3.0619,
+      "step": 13080
+    },
+    {
+      "epoch": 0.10684815367551842,
+      "grad_norm": 0.4468649923801422,
+      "learning_rate": 4.985495932607576e-05,
+      "loss": 3.0704,
+      "step": 13110
+    },
+    {
+      "epoch": 0.10709265745967292,
+      "grad_norm": 0.4374684989452362,
+      "learning_rate": 4.9854269055054096e-05,
+      "loss": 3.0614,
+      "step": 13140
+    },
+    {
+      "epoch": 0.10733716124382743,
+      "grad_norm": 0.4390091598033905,
+      "learning_rate": 4.985357715017744e-05,
+      "loss": 3.0511,
+      "step": 13170
+    },
+    {
+      "epoch": 0.10758166502798193,
+      "grad_norm": 0.4404134154319763,
+      "learning_rate": 4.985288361149129e-05,
+      "loss": 3.0577,
+      "step": 13200
+    },
+    {
+      "epoch": 0.10782616881213643,
+      "grad_norm": 0.4556957185268402,
+      "learning_rate": 4.985218843904122e-05,
+      "loss": 3.0632,
+      "step": 13230
+    },
+    {
+      "epoch": 0.10807067259629094,
+      "grad_norm": 0.4744901657104492,
+      "learning_rate": 4.985149163287294e-05,
+      "loss": 3.0525,
+      "step": 13260
+    },
+    {
+      "epoch": 0.10831517638044544,
+      "grad_norm": 0.42146775126457214,
+      "learning_rate": 4.985079319303225e-05,
+      "loss": 3.0456,
+      "step": 13290
+    },
+    {
+      "epoch": 0.10855968016459995,
+      "grad_norm": 0.4721812605857849,
+      "learning_rate": 4.985009311956507e-05,
+      "loss": 3.0412,
+      "step": 13320
+    },
+    {
+      "epoch": 0.10880418394875445,
+      "grad_norm": 0.44272100925445557,
+      "learning_rate": 4.984939141251741e-05,
+      "loss": 3.0591,
+      "step": 13350
+    },
+    {
+      "epoch": 0.10904868773290896,
+      "grad_norm": 0.4616432785987854,
+      "learning_rate": 4.9848688071935415e-05,
+      "loss": 3.0384,
+      "step": 13380
+    },
+    {
+      "epoch": 0.10929319151706346,
+      "grad_norm": 0.4424038231372833,
+      "learning_rate": 4.98479830978653e-05,
+      "loss": 3.0442,
+      "step": 13410
+    },
+    {
+      "epoch": 0.10953769530121797,
+      "grad_norm": 0.4387684464454651,
+      "learning_rate": 4.9847276490353425e-05,
+      "loss": 3.0406,
+      "step": 13440
+    },
+    {
+      "epoch": 0.10978219908537247,
+      "grad_norm": 0.44404199719429016,
+      "learning_rate": 4.984656824944623e-05,
+      "loss": 3.0146,
+      "step": 13470
+    },
+    {
+      "epoch": 0.11002670286952697,
+      "grad_norm": 0.4519595503807068,
+      "learning_rate": 4.984585837519028e-05,
+      "loss": 3.0314,
+      "step": 13500
+    },
+    {
+      "epoch": 0.11027120665368148,
+      "grad_norm": 0.42673036456108093,
+      "learning_rate": 4.984514686763224e-05,
+      "loss": 3.0193,
+      "step": 13530
+    },
+    {
+      "epoch": 0.11051571043783598,
+      "grad_norm": 0.44936153292655945,
+      "learning_rate": 4.984443372681887e-05,
+      "loss": 3.0275,
+      "step": 13560
+    },
+    {
+      "epoch": 0.11076021422199049,
+      "grad_norm": 0.4443502128124237,
+      "learning_rate": 4.9843718952797074e-05,
+      "loss": 3.0316,
+      "step": 13590
+    },
+    {
+      "epoch": 0.11100471800614499,
+      "grad_norm": 0.46171683073043823,
+      "learning_rate": 4.984300254561382e-05,
+      "loss": 3.0148,
+      "step": 13620
+    },
+    {
+      "epoch": 0.1112492217902995,
+      "grad_norm": 0.4363279342651367,
+      "learning_rate": 4.9842284505316206e-05,
+      "loss": 3.0381,
+      "step": 13650
+    },
+    {
+      "epoch": 0.111493725574454,
+      "grad_norm": 0.4243530035018921,
+      "learning_rate": 4.984156483195144e-05,
+      "loss": 3.0072,
+      "step": 13680
+    },
+    {
+      "epoch": 0.1117382293586085,
+      "grad_norm": 0.4298417270183563,
+      "learning_rate": 4.984084352556683e-05,
+      "loss": 3.026,
+      "step": 13710
+    },
+    {
+      "epoch": 0.11198273314276301,
+      "grad_norm": 0.437334805727005,
+      "learning_rate": 4.984012058620979e-05,
+      "loss": 3.0263,
+      "step": 13740
+    },
+    {
+      "epoch": 0.11222723692691751,
+      "grad_norm": 0.44620245695114136,
+      "learning_rate": 4.983939601392784e-05,
+      "loss": 3.0338,
+      "step": 13770
+    },
+    {
+      "epoch": 0.11247174071107202,
+      "grad_norm": 0.4280051290988922,
+      "learning_rate": 4.9838669808768616e-05,
+      "loss": 3.037,
+      "step": 13800
+    },
+    {
+      "epoch": 0.11271624449522652,
+      "grad_norm": 0.4300813376903534,
+      "learning_rate": 4.9837941970779865e-05,
+      "loss": 3.0322,
+      "step": 13830
+    },
+    {
+      "epoch": 0.11296074827938103,
+      "grad_norm": 0.4635563790798187,
+      "learning_rate": 4.983721250000942e-05,
+      "loss": 3.0247,
+      "step": 13860
+    },
+    {
+      "epoch": 0.11320525206353553,
+      "grad_norm": 0.4525061845779419,
+      "learning_rate": 4.983648139650524e-05,
+      "loss": 3.0177,
+      "step": 13890
+    },
+    {
+      "epoch": 0.11344975584769003,
+      "grad_norm": 0.4455706477165222,
+      "learning_rate": 4.983574866031538e-05,
+      "loss": 3.019,
+      "step": 13920
+    },
+    {
+      "epoch": 0.11369425963184454,
+      "grad_norm": 0.43036848306655884,
+      "learning_rate": 4.983501429148802e-05,
+      "loss": 3.016,
+      "step": 13950
+    },
+    {
+      "epoch": 0.11393876341599904,
+      "grad_norm": 0.44014501571655273,
+      "learning_rate": 4.983427829007144e-05,
+      "loss": 3.0157,
+      "step": 13980
+    },
+    {
+      "epoch": 0.11418326720015355,
+      "grad_norm": 0.43280887603759766,
+      "learning_rate": 4.9833540656114004e-05,
+      "loss": 3.0267,
+      "step": 14010
+    },
+    {
+      "epoch": 0.11442777098430805,
+      "grad_norm": 0.4454876482486725,
+      "learning_rate": 4.983280138966421e-05,
+      "loss": 3.0367,
+      "step": 14040
+    },
+    {
+      "epoch": 0.11467227476846256,
+      "grad_norm": 0.42322927713394165,
+      "learning_rate": 4.983206049077065e-05,
+      "loss": 3.0045,
+      "step": 14070
+    },
+    {
+      "epoch": 0.11491677855261706,
+      "grad_norm": 0.44999608397483826,
+      "learning_rate": 4.9831317959482044e-05,
+      "loss": 3.0393,
+      "step": 14100
+    },
+    {
+      "epoch": 0.11516128233677157,
+      "grad_norm": 0.44285348057746887,
+      "learning_rate": 4.983057379584719e-05,
+      "loss": 3.0189,
+      "step": 14130
+    },
+    {
+      "epoch": 0.11540578612092607,
+      "grad_norm": 0.42630505561828613,
+      "learning_rate": 4.982982799991501e-05,
+      "loss": 3.0244,
+      "step": 14160
+    },
+    {
+      "epoch": 0.11565028990508057,
+      "grad_norm": 0.44115307927131653,
+      "learning_rate": 4.982908057173454e-05,
+      "loss": 3.0091,
+      "step": 14190
+    },
+    {
+      "epoch": 0.11589479368923508,
+      "grad_norm": 0.43445199728012085,
+      "learning_rate": 4.98283315113549e-05,
+      "loss": 3.0119,
+      "step": 14220
+    },
+    {
+      "epoch": 0.11613929747338958,
+      "grad_norm": 0.44717735052108765,
+      "learning_rate": 4.982758081882534e-05,
+      "loss": 3.0177,
+      "step": 14250
+    },
+    {
+      "epoch": 0.11638380125754409,
+      "grad_norm": 0.43636298179626465,
+      "learning_rate": 4.982682849419521e-05,
+      "loss": 3.0186,
+      "step": 14280
+    },
+    {
+      "epoch": 0.11662830504169859,
+      "grad_norm": 0.4618465304374695,
+      "learning_rate": 4.982607453751396e-05,
+      "loss": 3.0074,
+      "step": 14310
+    },
+    {
+      "epoch": 0.1168728088258531,
+      "grad_norm": 0.4471168518066406,
+      "learning_rate": 4.982531894883116e-05,
+      "loss": 3.0304,
+      "step": 14340
+    },
+    {
+      "epoch": 0.1171173126100076,
+      "grad_norm": 0.45274320244789124,
+      "learning_rate": 4.9824561728196465e-05,
+      "loss": 3.0252,
+      "step": 14370
+    },
+    {
+      "epoch": 0.1173618163941621,
+      "grad_norm": 0.45815110206604004,
+      "learning_rate": 4.9823802875659674e-05,
+      "loss": 3.0253,
+      "step": 14400
+    },
+    {
+      "epoch": 0.11760632017831661,
+      "grad_norm": 0.43060410022735596,
+      "learning_rate": 4.9823042391270657e-05,
+      "loss": 2.9925,
+      "step": 14430
+    },
+    {
+      "epoch": 0.11785082396247111,
+      "grad_norm": 0.45071136951446533,
+      "learning_rate": 4.9822280275079416e-05,
+      "loss": 3.0117,
+      "step": 14460
+    },
+    {
+      "epoch": 0.11809532774662562,
+      "grad_norm": 0.45227503776550293,
+      "learning_rate": 4.9821516527136034e-05,
+      "loss": 3.0199,
+      "step": 14490
+    },
+    {
+      "epoch": 0.11833983153078012,
+      "grad_norm": 0.4271472990512848,
+      "learning_rate": 4.982075114749074e-05,
+      "loss": 3.013,
+      "step": 14520
+    },
+    {
+      "epoch": 0.11858433531493463,
+      "grad_norm": 0.4607602655887604,
+      "learning_rate": 4.9819984136193835e-05,
+      "loss": 3.0219,
+      "step": 14550
+    },
+    {
+      "epoch": 0.11882883909908913,
+      "grad_norm": 0.448803573846817,
+      "learning_rate": 4.9819215493295746e-05,
+      "loss": 3.0206,
+      "step": 14580
+    },
+    {
+      "epoch": 0.11907334288324363,
+      "grad_norm": 0.4245416820049286,
+      "learning_rate": 4.9818445218846995e-05,
+      "loss": 2.9972,
+      "step": 14610
+    },
+    {
+      "epoch": 0.11931784666739814,
+      "grad_norm": 0.4541032910346985,
+      "learning_rate": 4.9817673312898215e-05,
+      "loss": 3.0341,
+      "step": 14640
+    },
+    {
+      "epoch": 0.11956235045155264,
+      "grad_norm": 0.44764143228530884,
+      "learning_rate": 4.981689977550017e-05,
+      "loss": 2.9945,
+      "step": 14670
+    },
+    {
+      "epoch": 0.11980685423570715,
+      "grad_norm": 0.45025044679641724,
+      "learning_rate": 4.9816124606703684e-05,
+      "loss": 3.0092,
+      "step": 14700
+    },
+    {
+      "epoch": 0.12005135801986165,
+      "grad_norm": 0.43760672211647034,
+      "learning_rate": 4.9815347806559734e-05,
+      "loss": 2.9982,
+      "step": 14730
+    },
+    {
+      "epoch": 0.12029586180401616,
+      "grad_norm": 0.47016751766204834,
+      "learning_rate": 4.981456937511937e-05,
+      "loss": 2.9968,
+      "step": 14760
+    },
+    {
+      "epoch": 0.12054036558817066,
+      "grad_norm": 0.42403316497802734,
+      "learning_rate": 4.981378931243377e-05,
+      "loss": 2.9937,
+      "step": 14790
+    },
+    {
+      "epoch": 0.12078486937232517,
+      "grad_norm": 0.44598811864852905,
+      "learning_rate": 4.981300761855422e-05,
+      "loss": 3.0144,
+      "step": 14820
+    },
+    {
+      "epoch": 0.12102937315647967,
+      "grad_norm": 0.42306816577911377,
+      "learning_rate": 4.981222429353211e-05,
+      "loss": 3.017,
+      "step": 14850
+    },
+    {
+      "epoch": 0.12127387694063417,
+      "grad_norm": 0.45428961515426636,
+      "learning_rate": 4.981143933741891e-05,
+      "loss": 3.009,
+      "step": 14880
+    },
+    {
+      "epoch": 0.12151838072478868,
+      "grad_norm": 0.4344576299190521,
+      "learning_rate": 4.981065275026624e-05,
+      "loss": 3.0311,
+      "step": 14910
+    },
+    {
+      "epoch": 0.12176288450894318,
+      "grad_norm": 0.4361552298069,
+      "learning_rate": 4.9809864532125797e-05,
+      "loss": 2.9942,
+      "step": 14940
+    },
+    {
+      "epoch": 0.1220073882930977,
+      "grad_norm": 0.45136308670043945,
+      "learning_rate": 4.9809074683049415e-05,
+      "loss": 2.9886,
+      "step": 14970
+    },
+    {
+      "epoch": 0.12225189207725219,
+      "grad_norm": 0.4495120942592621,
+      "learning_rate": 4.9808283203089e-05,
+      "loss": 2.9799,
+      "step": 15000
+    },
+    {
+      "epoch": 0.1224963958614067,
+      "grad_norm": 0.45297014713287354,
+      "learning_rate": 4.9807490092296586e-05,
+      "loss": 3.0067,
+      "step": 15030
+    },
+    {
+      "epoch": 0.1227408996455612,
+      "grad_norm": 0.4508165419101715,
+      "learning_rate": 4.980669535072431e-05,
+      "loss": 3.0063,
+      "step": 15060
+    },
+    {
+      "epoch": 0.1229854034297157,
+      "grad_norm": 0.453021764755249,
+      "learning_rate": 4.980589897842441e-05,
+      "loss": 2.9853,
+      "step": 15090
+    },
+    {
+      "epoch": 0.12322990721387021,
+      "grad_norm": 0.45669054985046387,
+      "learning_rate": 4.9805100975449256e-05,
+      "loss": 2.9788,
+      "step": 15120
+    },
+    {
+      "epoch": 0.12347441099802471,
+      "grad_norm": 0.45175668597221375,
+      "learning_rate": 4.98043013418513e-05,
+      "loss": 2.9949,
+      "step": 15150
+    },
+    {
+      "epoch": 0.12371891478217922,
+      "grad_norm": 0.4426526129245758,
+      "learning_rate": 4.980350007768309e-05,
+      "loss": 2.9853,
+      "step": 15180
+    },
+    {
+      "epoch": 0.12396341856633372,
+      "grad_norm": 0.4432315528392792,
+      "learning_rate": 4.9802697182997316e-05,
+      "loss": 3.0084,
+      "step": 15210
+    },
+    {
+      "epoch": 0.12420792235048823,
+      "grad_norm": 0.4267187714576721,
+      "learning_rate": 4.980189265784676e-05,
+      "loss": 2.9764,
+      "step": 15240
+    },
+    {
+      "epoch": 0.12445242613464273,
+      "grad_norm": 0.45367953181266785,
+      "learning_rate": 4.98010865022843e-05,
+      "loss": 2.997,
+      "step": 15270
+    },
+    {
+      "epoch": 0.12469692991879723,
+      "grad_norm": 0.4352944791316986,
+      "learning_rate": 4.980027871636293e-05,
+      "loss": 2.989,
+      "step": 15300
+    },
+    {
+      "epoch": 0.12494143370295174,
+      "grad_norm": 0.43728742003440857,
+      "learning_rate": 4.979946930013576e-05,
+      "loss": 2.9732,
+      "step": 15330
+    },
+    {
+      "epoch": 0.12518593748710624,
+      "grad_norm": 0.43509089946746826,
+      "learning_rate": 4.9798658253655996e-05,
+      "loss": 2.9815,
+      "step": 15360
+    },
+    {
+      "epoch": 0.12543044127126074,
+      "grad_norm": 0.4618440866470337,
+      "learning_rate": 4.9797845576976945e-05,
+      "loss": 2.9945,
+      "step": 15390
+    },
+    {
+      "epoch": 0.12567494505541527,
+      "grad_norm": 0.44366157054901123,
+      "learning_rate": 4.979703127015205e-05,
+      "loss": 3.0167,
+      "step": 15420
+    },
+    {
+      "epoch": 0.12591944883956976,
+      "grad_norm": 0.43511441349983215,
+      "learning_rate": 4.979621533323482e-05,
+      "loss": 2.977,
+      "step": 15450
+    },
+    {
+      "epoch": 0.12616395262372426,
+      "grad_norm": 0.4174363911151886,
+      "learning_rate": 4.97953977662789e-05,
+      "loss": 2.9767,
+      "step": 15480
+    },
+    {
+      "epoch": 0.12640845640787876,
+      "grad_norm": 0.4421631395816803,
+      "learning_rate": 4.9794578569338046e-05,
+      "loss": 3.0044,
+      "step": 15510
+    },
+    {
+      "epoch": 0.12665296019203326,
+      "grad_norm": 0.4244731664657593,
+      "learning_rate": 4.979375774246609e-05,
+      "loss": 2.9865,
+      "step": 15540
+    },
+    {
+      "epoch": 0.12689746397618779,
+      "grad_norm": 0.4276571273803711,
+      "learning_rate": 4.979293528571701e-05,
+      "loss": 2.9608,
+      "step": 15570
+    },
+    {
+      "epoch": 0.12714196776034228,
+      "grad_norm": 0.44067880511283875,
+      "learning_rate": 4.979211119914486e-05,
+      "loss": 2.9793,
+      "step": 15600
+    },
+    {
+      "epoch": 0.12738647154449678,
+      "grad_norm": 0.4460121691226959,
+      "learning_rate": 4.9791285482803824e-05,
+      "loss": 2.9795,
+      "step": 15630
+    },
+    {
+      "epoch": 0.12763097532865128,
+      "grad_norm": 0.44630923867225647,
+      "learning_rate": 4.979045813674817e-05,
+      "loss": 2.9655,
+      "step": 15660
+    },
+    {
+      "epoch": 0.1278754791128058,
+      "grad_norm": 0.4406234323978424,
+      "learning_rate": 4.9789629161032294e-05,
+      "loss": 2.9752,
+      "step": 15690
+    },
+    {
+      "epoch": 0.1281199828969603,
+      "grad_norm": 0.45534536242485046,
+      "learning_rate": 4.978879855571068e-05,
+      "loss": 2.9874,
+      "step": 15720
+    },
+    {
+      "epoch": 0.1283644866811148,
+      "grad_norm": 0.4341179132461548,
+      "learning_rate": 4.9787966320837946e-05,
+      "loss": 2.9983,
+      "step": 15750
+    },
+    {
+      "epoch": 0.1286089904652693,
+      "grad_norm": 0.4406190514564514,
+      "learning_rate": 4.97871324564688e-05,
+      "loss": 2.9798,
+      "step": 15780
+    },
+    {
+      "epoch": 0.1288534942494238,
+      "grad_norm": 0.4403591752052307,
+      "learning_rate": 4.978629696265804e-05,
+      "loss": 2.9673,
+      "step": 15810
+    },
+    {
+      "epoch": 0.12909799803357833,
+      "grad_norm": 0.43992435932159424,
+      "learning_rate": 4.9785459839460595e-05,
+      "loss": 2.9673,
+      "step": 15840
+    },
+    {
+      "epoch": 0.12934250181773282,
+      "grad_norm": 0.4353017508983612,
+      "learning_rate": 4.97846210869315e-05,
+      "loss": 2.9677,
+      "step": 15870
+    },
+    {
+      "epoch": 0.12958700560188732,
+      "grad_norm": 0.4335249066352844,
+      "learning_rate": 4.978378070512591e-05,
+      "loss": 2.9625,
+      "step": 15900
+    },
+    {
+      "epoch": 0.12983150938604182,
+      "grad_norm": 0.4469107389450073,
+      "learning_rate": 4.978293869409903e-05,
+      "loss": 2.9773,
+      "step": 15930
+    },
+    {
+      "epoch": 0.13007601317019635,
+      "grad_norm": 0.4289211630821228,
+      "learning_rate": 4.978209505390625e-05,
+      "loss": 2.9819,
+      "step": 15960
+    },
+    {
+      "epoch": 0.13032051695435085,
+      "grad_norm": 0.45461341738700867,
+      "learning_rate": 4.978124978460301e-05,
+      "loss": 2.983,
+      "step": 15990
+    },
+    {
+      "epoch": 0.13056502073850534,
+      "grad_norm": 0.4294207990169525,
+      "learning_rate": 4.978040288624488e-05,
+      "loss": 2.9757,
+      "step": 16020
+    },
+    {
+      "epoch": 0.13080952452265984,
+      "grad_norm": 0.4259834289550781,
+      "learning_rate": 4.977955435888753e-05,
+      "loss": 2.97,
+      "step": 16050
+    },
+    {
+      "epoch": 0.13105402830681434,
+      "grad_norm": 0.4382823407649994,
+      "learning_rate": 4.9778704202586736e-05,
+      "loss": 2.9689,
+      "step": 16080
+    },
+    {
+      "epoch": 0.13129853209096887,
+      "grad_norm": 0.4408438503742218,
+      "learning_rate": 4.9777852417398394e-05,
+      "loss": 2.9689,
+      "step": 16110
+    },
+    {
+      "epoch": 0.13154303587512337,
+      "grad_norm": 0.440287321805954,
+      "learning_rate": 4.977699900337849e-05,
+      "loss": 2.9807,
+      "step": 16140
+    },
+    {
+      "epoch": 0.13178753965927786,
+      "grad_norm": 0.4474526643753052,
+      "learning_rate": 4.977614396058313e-05,
+      "loss": 2.9752,
+      "step": 16170
+    },
+    {
+      "epoch": 0.13203204344343236,
+      "grad_norm": 0.4456516206264496,
+      "learning_rate": 4.977528728906853e-05,
+      "loss": 2.9669,
+      "step": 16200
+    },
+    {
+      "epoch": 0.13227654722758686,
+      "grad_norm": 0.44889894127845764,
+      "learning_rate": 4.977442898889099e-05,
+      "loss": 2.9576,
+      "step": 16230
+    },
+    {
+      "epoch": 0.1325210510117414,
+      "grad_norm": 0.42843857407569885,
+      "learning_rate": 4.977356906010694e-05,
+      "loss": 2.9739,
+      "step": 16260
+    },
+    {
+      "epoch": 0.13276555479589588,
+      "grad_norm": 0.43735024333000183,
+      "learning_rate": 4.977270750277291e-05,
+      "loss": 2.9683,
+      "step": 16290
+    },
+    {
+      "epoch": 0.13301005858005038,
+      "grad_norm": 0.4472540318965912,
+      "learning_rate": 4.9771844316945536e-05,
+      "loss": 2.956,
+      "step": 16320
+    },
+    {
+      "epoch": 0.13325456236420488,
+      "grad_norm": 0.45051297545433044,
+      "learning_rate": 4.9770979502681566e-05,
+      "loss": 2.9828,
+      "step": 16350
+    },
+    {
+      "epoch": 0.1334990661483594,
+      "grad_norm": 0.4349469542503357,
+      "learning_rate": 4.977011306003784e-05,
+      "loss": 2.9661,
+      "step": 16380
+    },
+    {
+      "epoch": 0.1337435699325139,
+      "grad_norm": 0.4395361840724945,
+      "learning_rate": 4.976924498907133e-05,
+      "loss": 2.9624,
+      "step": 16410
+    },
+    {
+      "epoch": 0.1339880737166684,
+      "grad_norm": 0.4418069124221802,
+      "learning_rate": 4.9768375289839083e-05,
+      "loss": 2.9623,
+      "step": 16440
+    },
+    {
+      "epoch": 0.1342325775008229,
+      "grad_norm": 0.41756564378738403,
+      "learning_rate": 4.976750396239828e-05,
+      "loss": 2.949,
+      "step": 16470
+    },
+    {
+      "epoch": 0.1344770812849774,
+      "grad_norm": 0.4466126263141632,
+      "learning_rate": 4.9766631006806205e-05,
+      "loss": 2.9694,
+      "step": 16500
+    },
+    {
+      "epoch": 0.13472158506913193,
+      "grad_norm": 0.4336778223514557,
+      "learning_rate": 4.9765756423120235e-05,
+      "loss": 2.9652,
+      "step": 16530
+    },
+    {
+      "epoch": 0.13496608885328643,
+      "grad_norm": 0.42424583435058594,
+      "learning_rate": 4.976488021139787e-05,
+      "loss": 2.9487,
+      "step": 16560
+    },
+    {
+      "epoch": 0.13521059263744092,
+      "grad_norm": 0.4211365878582001,
+      "learning_rate": 4.9764002371696704e-05,
+      "loss": 2.9573,
+      "step": 16590
+    },
+    {
+      "epoch": 0.13545509642159542,
+      "grad_norm": 0.42899149656295776,
+      "learning_rate": 4.976312290407445e-05,
+      "loss": 2.968,
+      "step": 16620
+    },
+    {
+      "epoch": 0.13569960020574992,
+      "grad_norm": 0.4524397850036621,
+      "learning_rate": 4.9762241808588915e-05,
+      "loss": 2.9687,
+      "step": 16650
+    },
+    {
+      "epoch": 0.13594410398990445,
+      "grad_norm": 0.4466136693954468,
+      "learning_rate": 4.976135908529802e-05,
+      "loss": 2.9614,
+      "step": 16680
+    },
+    {
+      "epoch": 0.13618860777405895,
+      "grad_norm": 0.4260255694389343,
+      "learning_rate": 4.97604747342598e-05,
+      "loss": 2.9626,
+      "step": 16710
+    },
+    {
+      "epoch": 0.13643311155821344,
+      "grad_norm": 0.432644248008728,
+      "learning_rate": 4.9759588755532394e-05,
+      "loss": 2.9642,
+      "step": 16740
+    },
+    {
+      "epoch": 0.13667761534236794,
+      "grad_norm": 0.4311199188232422,
+      "learning_rate": 4.9758701149174026e-05,
+      "loss": 2.946,
+      "step": 16770
+    },
+    {
+      "epoch": 0.13692211912652247,
+      "grad_norm": 0.44559580087661743,
+      "learning_rate": 4.975781191524306e-05,
+      "loss": 2.9699,
+      "step": 16800
+    },
+    {
+      "epoch": 0.13716662291067697,
+      "grad_norm": 0.4649328291416168,
+      "learning_rate": 4.975692105379794e-05,
+      "loss": 2.9657,
+      "step": 16830
+    },
+    {
+      "epoch": 0.13741112669483146,
+      "grad_norm": 0.43768975138664246,
+      "learning_rate": 4.9756028564897236e-05,
+      "loss": 2.9407,
+      "step": 16860
+    },
+    {
+      "epoch": 0.13765563047898596,
+      "grad_norm": 0.42234501242637634,
+      "learning_rate": 4.975513444859963e-05,
+      "loss": 2.9506,
+      "step": 16890
+    },
+    {
+      "epoch": 0.13790013426314046,
+      "grad_norm": 0.4412541389465332,
+      "learning_rate": 4.9754238704963875e-05,
+      "loss": 2.9661,
+      "step": 16920
+    },
+    {
+      "epoch": 0.138144638047295,
+      "grad_norm": 0.43382880091667175,
+      "learning_rate": 4.975334133404887e-05,
+      "loss": 2.9706,
+      "step": 16950
+    },
+    {
+      "epoch": 0.13838914183144949,
+      "grad_norm": 0.4374088943004608,
+      "learning_rate": 4.97524423359136e-05,
+      "loss": 2.9393,
+      "step": 16980
+    },
+    {
+      "epoch": 0.13863364561560398,
+      "grad_norm": 0.4129140377044678,
+      "learning_rate": 4.9751541710617165e-05,
+      "loss": 2.9577,
+      "step": 17010
+    },
+    {
+      "epoch": 0.13887814939975848,
+      "grad_norm": 0.43458837270736694,
+      "learning_rate": 4.9750639458218774e-05,
+      "loss": 2.9739,
+      "step": 17040
+    },
+    {
+      "epoch": 0.139122653183913,
+      "grad_norm": 0.4235387444496155,
+      "learning_rate": 4.974973557877773e-05,
+      "loss": 2.9592,
+      "step": 17070
+    },
+    {
+      "epoch": 0.1393671569680675,
+      "grad_norm": 0.4440494775772095,
+      "learning_rate": 4.9748830072353455e-05,
+      "loss": 2.9502,
+      "step": 17100
+    },
+    {
+      "epoch": 0.139611660752222,
+      "grad_norm": 0.4453585147857666,
+      "learning_rate": 4.9747922939005484e-05,
+      "loss": 2.9565,
+      "step": 17130
+    },
+    {
+      "epoch": 0.1398561645363765,
+      "grad_norm": 0.4328203797340393,
+      "learning_rate": 4.974701417879344e-05,
+      "loss": 2.9465,
+      "step": 17160
+    },
+    {
+      "epoch": 0.140100668320531,
+      "grad_norm": 0.4327595829963684,
+      "learning_rate": 4.974610379177706e-05,
+      "loss": 2.9355,
+      "step": 17190
+    },
+    {
+      "epoch": 0.14034517210468553,
+      "grad_norm": 0.4403304159641266,
+      "learning_rate": 4.9745191778016196e-05,
+      "loss": 2.9428,
+      "step": 17220
+    },
+    {
+      "epoch": 0.14058967588884003,
+      "grad_norm": 0.430889368057251,
+      "learning_rate": 4.97442781375708e-05,
+      "loss": 2.9428,
+      "step": 17250
+    },
+    {
+      "epoch": 0.14083417967299453,
+      "grad_norm": 0.4505791664123535,
+      "learning_rate": 4.974336287050093e-05,
+      "loss": 2.9362,
+      "step": 17280
+    },
+    {
+      "epoch": 0.14107868345714902,
+      "grad_norm": 0.43023189902305603,
+      "learning_rate": 4.974244597686676e-05,
+      "loss": 2.9485,
+      "step": 17310
+    },
+    {
+      "epoch": 0.14132318724130352,
+      "grad_norm": 0.41717565059661865,
+      "learning_rate": 4.974152745672855e-05,
+      "loss": 2.9298,
+      "step": 17340
+    },
+    {
+      "epoch": 0.14156769102545805,
+      "grad_norm": 0.4348154366016388,
+      "learning_rate": 4.9740607310146706e-05,
+      "loss": 2.9531,
+      "step": 17370
+    },
+    {
+      "epoch": 0.14181219480961255,
+      "grad_norm": 0.43419626355171204,
+      "learning_rate": 4.973968553718169e-05,
+      "loss": 2.9381,
+      "step": 17400
+    },
+    {
+      "epoch": 0.14205669859376704,
+      "grad_norm": 0.43489935994148254,
+      "learning_rate": 4.9738762137894106e-05,
+      "loss": 2.9605,
+      "step": 17430
+    },
+    {
+      "epoch": 0.14230120237792154,
+      "grad_norm": 0.4242507219314575,
+      "learning_rate": 4.973783711234467e-05,
+      "loss": 2.9653,
+      "step": 17460
+    },
+    {
+      "epoch": 0.14254570616207607,
+      "grad_norm": 0.46746551990509033,
+      "learning_rate": 4.9736910460594164e-05,
+      "loss": 2.9525,
+      "step": 17490
+    },
+    {
+      "epoch": 0.14279020994623057,
+      "grad_norm": 0.4230245053768158,
+      "learning_rate": 4.973598218270352e-05,
+      "loss": 2.9399,
+      "step": 17520
+    },
+    {
+      "epoch": 0.14303471373038507,
+      "grad_norm": 0.451238751411438,
+      "learning_rate": 4.973505227873377e-05,
+      "loss": 2.9428,
+      "step": 17550
+    },
+    {
+      "epoch": 0.14327921751453956,
+      "grad_norm": 0.43941864371299744,
+      "learning_rate": 4.9734120748746026e-05,
+      "loss": 2.9424,
+      "step": 17580
+    },
+    {
+      "epoch": 0.14352372129869406,
+      "grad_norm": 0.43805423378944397,
+      "learning_rate": 4.973318759280153e-05,
+      "loss": 2.9324,
+      "step": 17610
+    },
+    {
+      "epoch": 0.1437682250828486,
+      "grad_norm": 0.46820029616355896,
+      "learning_rate": 4.973225281096162e-05,
+      "loss": 2.9415,
+      "step": 17640
+    },
+    {
+      "epoch": 0.1440127288670031,
+      "grad_norm": 0.44885364174842834,
+      "learning_rate": 4.973131640328776e-05,
+      "loss": 2.9438,
+      "step": 17670
+    },
+    {
+      "epoch": 0.14425723265115759,
+      "grad_norm": 0.43716761469841003,
+      "learning_rate": 4.97303783698415e-05,
+      "loss": 2.929,
+      "step": 17700
+    },
+    {
+      "epoch": 0.14450173643531208,
+      "grad_norm": 0.4408782124519348,
+      "learning_rate": 4.97294387106845e-05,
+      "loss": 2.9418,
+      "step": 17730
+    },
+    {
+      "epoch": 0.1447462402194666,
+      "grad_norm": 0.4411155581474304,
+      "learning_rate": 4.972849742587853e-05,
+      "loss": 2.9492,
+      "step": 17760
+    },
+    {
+      "epoch": 0.1449907440036211,
+      "grad_norm": 0.44424518942832947,
+      "learning_rate": 4.972755451548548e-05,
+      "loss": 2.9455,
+      "step": 17790
+    },
+    {
+      "epoch": 0.1452352477877756,
+      "grad_norm": 0.4497133195400238,
+      "learning_rate": 4.9726609979567316e-05,
+      "loss": 2.9273,
+      "step": 17820
+    },
+    {
+      "epoch": 0.1454797515719301,
+      "grad_norm": 0.4441046416759491,
+      "learning_rate": 4.972566381818614e-05,
+      "loss": 2.9529,
+      "step": 17850
+    },
+    {
+      "epoch": 0.1457242553560846,
+      "grad_norm": 0.4226887822151184,
+      "learning_rate": 4.972471603140415e-05,
+      "loss": 2.9358,
+      "step": 17880
+    },
+    {
+      "epoch": 0.14596875914023913,
+      "grad_norm": 0.4297832250595093,
+      "learning_rate": 4.972376661928365e-05,
+      "loss": 2.9362,
+      "step": 17910
+    },
+    {
+      "epoch": 0.14621326292439363,
+      "grad_norm": 0.4289223551750183,
+      "learning_rate": 4.972281558188706e-05,
+      "loss": 2.9203,
+      "step": 17940
+    },
+    {
+      "epoch": 0.14645776670854813,
+      "grad_norm": 0.45321324467658997,
+      "learning_rate": 4.972186291927689e-05,
+      "loss": 2.941,
+      "step": 17970
+    },
+    {
+      "epoch": 0.14670227049270262,
+      "grad_norm": 0.4406587481498718,
+      "learning_rate": 4.9720908631515757e-05,
+      "loss": 2.9364,
+      "step": 18000
+    },
+    {
+      "epoch": 0.14694677427685712,
+      "grad_norm": 0.44196027517318726,
+      "learning_rate": 4.9719952718666406e-05,
+      "loss": 2.93,
+      "step": 18030
+    },
+    {
+      "epoch": 0.14719127806101165,
+      "grad_norm": 0.41876834630966187,
+      "learning_rate": 4.971899518079167e-05,
+      "loss": 2.9286,
+      "step": 18060
+    },
+    {
+      "epoch": 0.14743578184516615,
+      "grad_norm": 0.42476391792297363,
+      "learning_rate": 4.971803601795451e-05,
+      "loss": 2.9239,
+      "step": 18090
+    },
+    {
+      "epoch": 0.14768028562932065,
+      "grad_norm": 0.43276798725128174,
+      "learning_rate": 4.9717075230217955e-05,
+      "loss": 2.9469,
+      "step": 18120
+    },
+    {
+      "epoch": 0.14792478941347514,
+      "grad_norm": 0.42786705493927,
+      "learning_rate": 4.9716112817645176e-05,
+      "loss": 2.9442,
+      "step": 18150
+    },
+    {
+      "epoch": 0.14816929319762967,
+      "grad_norm": 0.4399658739566803,
+      "learning_rate": 4.971514878029945e-05,
+      "loss": 2.9334,
+      "step": 18180
+    },
+    {
+      "epoch": 0.14841379698178417,
+      "grad_norm": 0.4298873841762543,
+      "learning_rate": 4.971418311824413e-05,
+      "loss": 2.9176,
+      "step": 18210
+    },
+    {
+      "epoch": 0.14865830076593867,
+      "grad_norm": 0.43509942293167114,
+      "learning_rate": 4.971321583154271e-05,
+      "loss": 2.9345,
+      "step": 18240
+    },
+    {
+      "epoch": 0.14890280455009317,
+      "grad_norm": 0.42680835723876953,
+      "learning_rate": 4.9712246920258784e-05,
+      "loss": 2.9274,
+      "step": 18270
+    },
+    {
+      "epoch": 0.14914730833424766,
+      "grad_norm": 0.4409879446029663,
+      "learning_rate": 4.9711276384456024e-05,
+      "loss": 2.9231,
+      "step": 18300
+    },
+    {
+      "epoch": 0.1493918121184022,
+      "grad_norm": 0.4345747232437134,
+      "learning_rate": 4.9710304224198245e-05,
+      "loss": 2.9267,
+      "step": 18330
+    },
+    {
+      "epoch": 0.1496363159025567,
+      "grad_norm": 0.4443499743938446,
+      "learning_rate": 4.970933043954935e-05,
+      "loss": 2.9307,
+      "step": 18360
+    },
+    {
+      "epoch": 0.1498808196867112,
+      "grad_norm": 0.4309675991535187,
+      "learning_rate": 4.970835503057335e-05,
+      "loss": 2.9174,
+      "step": 18390
+    },
+    {
+      "epoch": 0.15012532347086568,
+      "grad_norm": 0.4251437187194824,
+      "learning_rate": 4.970737799733438e-05,
+      "loss": 2.9375,
+      "step": 18420
+    },
+    {
+      "epoch": 0.1503698272550202,
+      "grad_norm": 0.4523163139820099,
+      "learning_rate": 4.970639933989665e-05,
+      "loss": 2.9424,
+      "step": 18450
+    },
+    {
+      "epoch": 0.1506143310391747,
+      "grad_norm": 0.43151840567588806,
+      "learning_rate": 4.97054190583245e-05,
+      "loss": 2.9284,
+      "step": 18480
+    },
+    {
+      "epoch": 0.1508588348233292,
+      "grad_norm": 0.43805721402168274,
+      "learning_rate": 4.970443715268238e-05,
+      "loss": 2.9266,
+      "step": 18510
+    },
+    {
+      "epoch": 0.1511033386074837,
+      "grad_norm": 0.4656009078025818,
+      "learning_rate": 4.970345362303482e-05,
+      "loss": 2.9274,
+      "step": 18540
+    },
+    {
+      "epoch": 0.1513478423916382,
+      "grad_norm": 0.4599588215351105,
+      "learning_rate": 4.9702468469446496e-05,
+      "loss": 2.9297,
+      "step": 18570
+    },
+    {
+      "epoch": 0.15159234617579273,
+      "grad_norm": 0.44205912947654724,
+      "learning_rate": 4.9701481691982157e-05,
+      "loss": 2.9084,
+      "step": 18600
+    },
+    {
+      "epoch": 0.15183684995994723,
+      "grad_norm": 0.43253156542778015,
+      "learning_rate": 4.9700493290706665e-05,
+      "loss": 2.9237,
+      "step": 18630
+    },
+    {
+      "epoch": 0.15208135374410173,
+      "grad_norm": 0.42774561047554016,
+      "learning_rate": 4.969950326568501e-05,
+      "loss": 2.9248,
+      "step": 18660
+    },
+    {
+      "epoch": 0.15232585752825623,
+      "grad_norm": 0.4493556618690491,
+      "learning_rate": 4.969851161698227e-05,
+      "loss": 2.9248,
+      "step": 18690
+    },
+    {
+      "epoch": 0.15257036131241072,
+      "grad_norm": 0.4323256015777588,
+      "learning_rate": 4.969751834466363e-05,
+      "loss": 2.923,
+      "step": 18720
+    },
+    {
+      "epoch": 0.15281486509656525,
+      "grad_norm": 0.42869383096694946,
+      "learning_rate": 4.969652344879438e-05,
+      "loss": 2.9237,
+      "step": 18750
+    },
+    {
+      "epoch": 0.15305936888071975,
+      "grad_norm": 0.44113072752952576,
+      "learning_rate": 4.9695526929439925e-05,
+      "loss": 2.9323,
+      "step": 18780
+    },
+    {
+      "epoch": 0.15330387266487425,
+      "grad_norm": 0.4490866959095001,
+      "learning_rate": 4.969452878666578e-05,
+      "loss": 2.9142,
+      "step": 18810
+    },
+    {
+      "epoch": 0.15354837644902875,
+      "grad_norm": 0.41400545835494995,
+      "learning_rate": 4.969352902053756e-05,
+      "loss": 2.9306,
+      "step": 18840
+    },
+    {
+      "epoch": 0.15379288023318327,
+      "grad_norm": 0.42999714612960815,
+      "learning_rate": 4.9692527631120976e-05,
+      "loss": 2.922,
+      "step": 18870
+    },
+    {
+      "epoch": 0.15403738401733777,
+      "grad_norm": 0.4393715262413025,
+      "learning_rate": 4.969152461848187e-05,
+      "loss": 2.9129,
+      "step": 18900
+    },
+    {
+      "epoch": 0.15428188780149227,
+      "grad_norm": 0.4195369780063629,
+      "learning_rate": 4.969051998268617e-05,
+      "loss": 2.9257,
+      "step": 18930
+    },
+    {
+      "epoch": 0.15452639158564677,
+      "grad_norm": 0.44224682450294495,
+      "learning_rate": 4.968951372379992e-05,
+      "loss": 2.9088,
+      "step": 18960
+    },
+    {
+      "epoch": 0.15477089536980126,
+      "grad_norm": 0.42597490549087524,
+      "learning_rate": 4.9688505841889266e-05,
+      "loss": 2.9001,
+      "step": 18990
+    },
+    {
+      "epoch": 0.1550153991539558,
+      "grad_norm": 0.4543912708759308,
+      "learning_rate": 4.968749633702047e-05,
+      "loss": 2.932,
+      "step": 19020
+    },
+    {
+      "epoch": 0.1552599029381103,
+      "grad_norm": 0.41840681433677673,
+      "learning_rate": 4.968648520925988e-05,
+      "loss": 2.927,
+      "step": 19050
+    },
+    {
+      "epoch": 0.1555044067222648,
+      "grad_norm": 0.4398231506347656,
+      "learning_rate": 4.9685472458673986e-05,
+      "loss": 2.9166,
+      "step": 19080
+    },
+    {
+      "epoch": 0.15574891050641929,
+      "grad_norm": 0.4245362877845764,
+      "learning_rate": 4.968445808532935e-05,
+      "loss": 2.9119,
+      "step": 19110
+    },
+    {
+      "epoch": 0.1559934142905738,
+      "grad_norm": 0.4447348415851593,
+      "learning_rate": 4.9683442089292656e-05,
+      "loss": 2.9158,
+      "step": 19140
+    },
+    {
+      "epoch": 0.1562379180747283,
+      "grad_norm": 0.45135971903800964,
+      "learning_rate": 4.9682424470630685e-05,
+      "loss": 2.9048,
+      "step": 19170
+    },
+    {
+      "epoch": 0.1564824218588828,
+      "grad_norm": 0.42726004123687744,
+      "learning_rate": 4.968140522941035e-05,
+      "loss": 2.9085,
+      "step": 19200
+    },
+    {
+      "epoch": 0.1567269256430373,
+      "grad_norm": 0.42716965079307556,
+      "learning_rate": 4.9680384365698644e-05,
+      "loss": 2.9229,
+      "step": 19230
+    },
+    {
+      "epoch": 0.1569714294271918,
+      "grad_norm": 0.4451904594898224,
+      "learning_rate": 4.967936187956268e-05,
+      "loss": 2.921,
+      "step": 19260
+    },
+    {
+      "epoch": 0.15721593321134633,
+      "grad_norm": 0.4485413134098053,
+      "learning_rate": 4.967833777106966e-05,
+      "loss": 2.9194,
+      "step": 19290
+    },
+    {
+      "epoch": 0.15746043699550083,
+      "grad_norm": 0.44135797023773193,
+      "learning_rate": 4.9677312040286926e-05,
+      "loss": 2.885,
+      "step": 19320
+    },
+    {
+      "epoch": 0.15770494077965533,
+      "grad_norm": 0.4556313753128052,
+      "learning_rate": 4.967628468728189e-05,
+      "loss": 2.923,
+      "step": 19350
+    },
+    {
+      "epoch": 0.15794944456380983,
+      "grad_norm": 0.4418553411960602,
+      "learning_rate": 4.9675255712122105e-05,
+      "loss": 2.9064,
+      "step": 19380
+    },
+    {
+      "epoch": 0.15819394834796432,
+      "grad_norm": 0.443284273147583,
+      "learning_rate": 4.967422511487519e-05,
+      "loss": 2.8951,
+      "step": 19410
+    },
+    {
+      "epoch": 0.15843845213211885,
+      "grad_norm": 0.4480453431606293,
+      "learning_rate": 4.967319289560892e-05,
+      "loss": 2.9151,
+      "step": 19440
+    },
+    {
+      "epoch": 0.15868295591627335,
+      "grad_norm": 0.4382655620574951,
+      "learning_rate": 4.967215905439113e-05,
+      "loss": 2.9263,
+      "step": 19470
+    },
+    {
+      "epoch": 0.15892745970042785,
+      "grad_norm": 0.44708484411239624,
+      "learning_rate": 4.967112359128979e-05,
+      "loss": 2.9088,
+      "step": 19500
+    },
+    {
+      "epoch": 0.15917196348458235,
+      "grad_norm": 0.4404008984565735,
+      "learning_rate": 4.967008650637297e-05,
+      "loss": 2.8915,
+      "step": 19530
+    },
+    {
+      "epoch": 0.15941646726873687,
+      "grad_norm": 0.4447616636753082,
+      "learning_rate": 4.9669047799708834e-05,
+      "loss": 2.9082,
+      "step": 19560
+    },
+    {
+      "epoch": 0.15966097105289137,
+      "grad_norm": 0.41954031586647034,
+      "learning_rate": 4.966800747136568e-05,
+      "loss": 2.8891,
+      "step": 19590
+    },
+    {
+      "epoch": 0.15990547483704587,
+      "grad_norm": 0.4296923279762268,
+      "learning_rate": 4.966696552141189e-05,
+      "loss": 2.9053,
+      "step": 19620
+    },
+    {
+      "epoch": 0.16014997862120037,
+      "grad_norm": 0.4200703203678131,
+      "learning_rate": 4.966592194991596e-05,
+      "loss": 2.9006,
+      "step": 19650
+    },
+    {
+      "epoch": 0.16039448240535487,
+      "grad_norm": 0.42826735973358154,
+      "learning_rate": 4.966487675694649e-05,
+      "loss": 2.908,
+      "step": 19680
+    },
+    {
+      "epoch": 0.1606389861895094,
+      "grad_norm": 0.43031007051467896,
+      "learning_rate": 4.966382994257218e-05,
+      "loss": 2.9067,
+      "step": 19710
+    },
+    {
+      "epoch": 0.1608834899736639,
+      "grad_norm": 0.41722506284713745,
+      "learning_rate": 4.966278150686186e-05,
+      "loss": 2.882,
+      "step": 19740
+    },
+    {
+      "epoch": 0.1611279937578184,
+      "grad_norm": 0.4380910098552704,
+      "learning_rate": 4.9661731449884444e-05,
+      "loss": 2.8822,
+      "step": 19770
+    },
+    {
+      "epoch": 0.1613724975419729,
+      "grad_norm": 0.42654240131378174,
+      "learning_rate": 4.966067977170896e-05,
+      "loss": 2.9046,
+      "step": 19800
+    },
+    {
+      "epoch": 0.1616170013261274,
+      "grad_norm": 0.4350351393222809,
+      "learning_rate": 4.9659626472404545e-05,
+      "loss": 2.8921,
+      "step": 19830
+    },
+    {
+      "epoch": 0.1618615051102819,
+      "grad_norm": 0.45913204550743103,
+      "learning_rate": 4.965857155204044e-05,
+      "loss": 2.9024,
+      "step": 19860
+    },
+    {
+      "epoch": 0.1621060088944364,
+      "grad_norm": 0.448477566242218,
+      "learning_rate": 4.965751501068598e-05,
+      "loss": 2.9039,
+      "step": 19890
+    },
+    {
+      "epoch": 0.1623505126785909,
+      "grad_norm": 0.42940184473991394,
+      "learning_rate": 4.965645684841065e-05,
+      "loss": 2.9179,
+      "step": 19920
+    },
+    {
+      "epoch": 0.1625950164627454,
+      "grad_norm": 0.4369218647480011,
+      "learning_rate": 4.965539706528398e-05,
+      "loss": 2.9041,
+      "step": 19950
+    },
+    {
+      "epoch": 0.16283952024689993,
+      "grad_norm": 0.43815574049949646,
+      "learning_rate": 4.965433566137565e-05,
+      "loss": 2.8966,
+      "step": 19980
+    },
+    {
+      "epoch": 0.16308402403105443,
+      "grad_norm": 0.42448291182518005,
+      "learning_rate": 4.965327263675543e-05,
+      "loss": 2.8959,
+      "step": 20010
+    },
+    {
+      "epoch": 0.16332852781520893,
+      "grad_norm": 0.4157077968120575,
+      "learning_rate": 4.965220799149321e-05,
+      "loss": 2.8876,
+      "step": 20040
+    },
+    {
+      "epoch": 0.16357303159936343,
+      "grad_norm": 0.4500816762447357,
+      "learning_rate": 4.9651141725658955e-05,
+      "loss": 2.9075,
+      "step": 20070
+    },
+    {
+      "epoch": 0.16381753538351793,
+      "grad_norm": 0.4356308579444885,
+      "learning_rate": 4.9650073839322786e-05,
+      "loss": 2.9147,
+      "step": 20100
+    },
+    {
+      "epoch": 0.16406203916767245,
+      "grad_norm": 0.4416338801383972,
+      "learning_rate": 4.964900433255489e-05,
+      "loss": 2.9017,
+      "step": 20130
+    },
+    {
+      "epoch": 0.16430654295182695,
+      "grad_norm": 0.4255659580230713,
+      "learning_rate": 4.9647933205425574e-05,
+      "loss": 2.9,
+      "step": 20160
+    },
+    {
+      "epoch": 0.16455104673598145,
+      "grad_norm": 0.4527672827243805,
+      "learning_rate": 4.9646860458005254e-05,
+      "loss": 2.9076,
+      "step": 20190
+    },
+    {
+      "epoch": 0.16479555052013595,
+      "grad_norm": 0.45242664217948914,
+      "learning_rate": 4.9645786090364436e-05,
+      "loss": 2.9177,
+      "step": 20220
+    },
+    {
+      "epoch": 0.16504005430429047,
+      "grad_norm": 0.43450960516929626,
+      "learning_rate": 4.964471010257377e-05,
+      "loss": 2.8809,
+      "step": 20250
+    },
+    {
+      "epoch": 0.16528455808844497,
+      "grad_norm": 0.44094353914260864,
+      "learning_rate": 4.964363249470396e-05,
+      "loss": 2.8996,
+      "step": 20280
+    },
+    {
+      "epoch": 0.16552906187259947,
+      "grad_norm": 0.4236229658126831,
+      "learning_rate": 4.964255326682588e-05,
+      "loss": 2.9014,
+      "step": 20310
+    },
+    {
+      "epoch": 0.16577356565675397,
+      "grad_norm": 0.43658915162086487,
+      "learning_rate": 4.964147241901045e-05,
+      "loss": 2.8835,
+      "step": 20340
+    },
+    {
+      "epoch": 0.16601806944090847,
+      "grad_norm": 0.4210447371006012,
+      "learning_rate": 4.964038995132872e-05,
+      "loss": 2.9015,
+      "step": 20370
+    },
+    {
+      "epoch": 0.166262573225063,
+      "grad_norm": 0.4408273994922638,
+      "learning_rate": 4.963930586385186e-05,
+      "loss": 2.8933,
+      "step": 20400
+    },
+    {
+      "epoch": 0.1665070770092175,
+      "grad_norm": 0.4378444254398346,
+      "learning_rate": 4.963822015665114e-05,
+      "loss": 2.8775,
+      "step": 20430
+    },
+    {
+      "epoch": 0.166751580793372,
+      "grad_norm": 0.4379393458366394,
+      "learning_rate": 4.963713282979792e-05,
+      "loss": 2.8828,
+      "step": 20460
+    },
+    {
+      "epoch": 0.1669960845775265,
+      "grad_norm": 0.4111761748790741,
+      "learning_rate": 4.963604388336368e-05,
+      "loss": 2.9,
+      "step": 20490
+    },
+    {
+      "epoch": 0.16724058836168101,
+      "grad_norm": 0.4153888523578644,
+      "learning_rate": 4.9634953317420006e-05,
+      "loss": 2.8834,
+      "step": 20520
+    },
+    {
+      "epoch": 0.1674850921458355,
+      "grad_norm": 0.44652920961380005,
+      "learning_rate": 4.963386113203859e-05,
+      "loss": 2.8591,
+      "step": 20550
+    },
+    {
+      "epoch": 0.16772959592999,
+      "grad_norm": 0.4109705090522766,
+      "learning_rate": 4.963276732729124e-05,
+      "loss": 2.8794,
+      "step": 20580
+    },
+    {
+      "epoch": 0.1679740997141445,
+      "grad_norm": 0.417043000459671,
+      "learning_rate": 4.9631671903249835e-05,
+      "loss": 2.8822,
+      "step": 20610
+    },
+    {
+      "epoch": 0.168218603498299,
+      "grad_norm": 0.4661671221256256,
+      "learning_rate": 4.963057485998641e-05,
+      "loss": 2.8884,
+      "step": 20640
+    },
+    {
+      "epoch": 0.16846310728245353,
+      "grad_norm": 0.4327886700630188,
+      "learning_rate": 4.962947619757307e-05,
+      "loss": 2.8988,
+      "step": 20670
+    },
+    {
+      "epoch": 0.16870761106660803,
+      "grad_norm": 0.4549929201602936,
+      "learning_rate": 4.962837591608203e-05,
+      "loss": 2.8748,
+      "step": 20700
+    },
+    {
+      "epoch": 0.16895211485076253,
+      "grad_norm": 0.4588701128959656,
+      "learning_rate": 4.962727401558563e-05,
+      "loss": 2.8811,
+      "step": 20730
+    },
+    {
+      "epoch": 0.16919661863491703,
+      "grad_norm": 0.4168790578842163,
+      "learning_rate": 4.962617049615631e-05,
+      "loss": 2.8648,
+      "step": 20760
+    },
+    {
+      "epoch": 0.16944112241907153,
+      "grad_norm": 0.440508633852005,
+      "learning_rate": 4.96250653578666e-05,
+      "loss": 2.8815,
+      "step": 20790
+    },
+    {
+      "epoch": 0.16968562620322605,
+      "grad_norm": 0.41507402062416077,
+      "learning_rate": 4.962395860078917e-05,
+      "loss": 2.8689,
+      "step": 20820
+    },
+    {
+      "epoch": 0.16993012998738055,
+      "grad_norm": 0.4498741924762726,
+      "learning_rate": 4.9622850224996744e-05,
+      "loss": 2.8794,
+      "step": 20850
+    },
+    {
+      "epoch": 0.17017463377153505,
+      "grad_norm": 0.42679455876350403,
+      "learning_rate": 4.962174023056221e-05,
+      "loss": 2.8815,
+      "step": 20880
+    },
+    {
+      "epoch": 0.17041913755568955,
+      "grad_norm": 0.4224587380886078,
+      "learning_rate": 4.962062861755853e-05,
+      "loss": 2.8964,
+      "step": 20910
+    },
+    {
+      "epoch": 0.17066364133984407,
+      "grad_norm": 0.4392225444316864,
+      "learning_rate": 4.9619515386058775e-05,
+      "loss": 2.8821,
+      "step": 20940
+    },
+    {
+      "epoch": 0.17090814512399857,
+      "grad_norm": 0.43538761138916016,
+      "learning_rate": 4.9618400536136124e-05,
+      "loss": 2.8867,
+      "step": 20970
+    },
+    {
+      "epoch": 0.17115264890815307,
+      "grad_norm": 0.4309751093387604,
+      "learning_rate": 4.9617284067863866e-05,
+      "loss": 2.8721,
+      "step": 21000
+    },
+    {
+      "epoch": 0.17139715269230757,
+      "grad_norm": 0.43987441062927246,
+      "learning_rate": 4.9616165981315397e-05,
+      "loss": 2.8818,
+      "step": 21030
+    },
+    {
+      "epoch": 0.17164165647646207,
+      "grad_norm": 0.449082612991333,
+      "learning_rate": 4.961504627656422e-05,
+      "loss": 2.8882,
+      "step": 21060
+    },
+    {
+      "epoch": 0.1718861602606166,
+      "grad_norm": 0.4370419979095459,
+      "learning_rate": 4.961392495368393e-05,
+      "loss": 2.8902,
+      "step": 21090
+    },
+    {
+      "epoch": 0.1721306640447711,
+      "grad_norm": 0.44723600149154663,
+      "learning_rate": 4.961280201274826e-05,
+      "loss": 2.9021,
+      "step": 21120
+    },
+    {
+      "epoch": 0.1723751678289256,
+      "grad_norm": 0.45456555485725403,
+      "learning_rate": 4.9611677453831005e-05,
+      "loss": 2.8602,
+      "step": 21150
+    },
+    {
+      "epoch": 0.1726196716130801,
+      "grad_norm": 0.4356628954410553,
+      "learning_rate": 4.9610551277006115e-05,
+      "loss": 2.876,
+      "step": 21180
+    },
+    {
+      "epoch": 0.1728641753972346,
+      "grad_norm": 0.4403671324253082,
+      "learning_rate": 4.96094234823476e-05,
+      "loss": 2.8809,
+      "step": 21210
+    },
+    {
+      "epoch": 0.1731086791813891,
+      "grad_norm": 0.4314226806163788,
+      "learning_rate": 4.9608294069929605e-05,
+      "loss": 2.8596,
+      "step": 21240
+    },
+    {
+      "epoch": 0.1733531829655436,
+      "grad_norm": 0.4421268701553345,
+      "learning_rate": 4.960716303982638e-05,
+      "loss": 2.8562,
+      "step": 21270
+    },
+    {
+      "epoch": 0.1735976867496981,
+      "grad_norm": 0.4297054409980774,
+      "learning_rate": 4.960603039211227e-05,
+      "loss": 2.8632,
+      "step": 21300
+    },
+    {
+      "epoch": 0.1738421905338526,
+      "grad_norm": 0.42435142397880554,
+      "learning_rate": 4.960489612686174e-05,
+      "loss": 2.8827,
+      "step": 21330
+    },
+    {
+      "epoch": 0.17408669431800713,
+      "grad_norm": 0.42332854866981506,
+      "learning_rate": 4.960376024414935e-05,
+      "loss": 2.8725,
+      "step": 21360
+    },
+    {
+      "epoch": 0.17433119810216163,
+      "grad_norm": 0.43352165818214417,
+      "learning_rate": 4.9602622744049765e-05,
+      "loss": 2.8714,
+      "step": 21390
+    },
+    {
+      "epoch": 0.17457570188631613,
+      "grad_norm": 0.44356927275657654,
+      "learning_rate": 4.960148362663777e-05,
+      "loss": 2.8753,
+      "step": 21420
+    },
+    {
+      "epoch": 0.17482020567047063,
+      "grad_norm": 0.43293410539627075,
+      "learning_rate": 4.960034289198824e-05,
+      "loss": 2.8716,
+      "step": 21450
+    },
+    {
+      "epoch": 0.17506470945462513,
+      "grad_norm": 0.4402593970298767,
+      "learning_rate": 4.959920054017616e-05,
+      "loss": 2.8837,
+      "step": 21480
+    },
+    {
+      "epoch": 0.17530921323877965,
+      "grad_norm": 0.4285391867160797,
+      "learning_rate": 4.959805657127664e-05,
+      "loss": 2.8776,
+      "step": 21510
+    },
+    {
+      "epoch": 0.17555371702293415,
+      "grad_norm": 0.4400288164615631,
+      "learning_rate": 4.959691098536487e-05,
+      "loss": 2.8794,
+      "step": 21540
+    },
+    {
+      "epoch": 0.17579822080708865,
+      "grad_norm": 0.4306489825248718,
+      "learning_rate": 4.959576378251617e-05,
+      "loss": 2.8744,
+      "step": 21570
+    },
+    {
+      "epoch": 0.17604272459124315,
+      "grad_norm": 0.4402148723602295,
+      "learning_rate": 4.959461496280594e-05,
+      "loss": 2.8749,
+      "step": 21600
+    },
+    {
+      "epoch": 0.17628722837539768,
+      "grad_norm": 0.44051840901374817,
+      "learning_rate": 4.95934645263097e-05,
+      "loss": 2.8714,
+      "step": 21630
+    },
+    {
+      "epoch": 0.17653173215955217,
+      "grad_norm": 0.4164459705352783,
+      "learning_rate": 4.9592312473103086e-05,
+      "loss": 2.8757,
+      "step": 21660
+    },
+    {
+      "epoch": 0.17677623594370667,
+      "grad_norm": 0.4289766550064087,
+      "learning_rate": 4.9591158803261834e-05,
+      "loss": 2.8675,
+      "step": 21690
+    },
+    {
+      "epoch": 0.17702073972786117,
+      "grad_norm": 0.42621496319770813,
+      "learning_rate": 4.9590003516861774e-05,
+      "loss": 2.852,
+      "step": 21720
+    },
+    {
+      "epoch": 0.17726524351201567,
+      "grad_norm": 0.43924206495285034,
+      "learning_rate": 4.958884661397885e-05,
+      "loss": 2.867,
+      "step": 21750
+    },
+    {
+      "epoch": 0.1775097472961702,
+      "grad_norm": 0.4479053020477295,
+      "learning_rate": 4.958768809468912e-05,
+      "loss": 2.8768,
+      "step": 21780
+    },
+    {
+      "epoch": 0.1777542510803247,
+      "grad_norm": 0.4436418414115906,
+      "learning_rate": 4.958652795906874e-05,
+      "loss": 2.8795,
+      "step": 21810
+    },
+    {
+      "epoch": 0.1779987548644792,
+      "grad_norm": 0.4646727740764618,
+      "learning_rate": 4.9585366207193984e-05,
+      "loss": 2.8738,
+      "step": 21840
+    },
+    {
+      "epoch": 0.1782432586486337,
+      "grad_norm": 0.44066762924194336,
+      "learning_rate": 4.95842028391412e-05,
+      "loss": 2.8975,
+      "step": 21870
+    },
+    {
+      "epoch": 0.1784877624327882,
+      "grad_norm": 0.4433143138885498,
+      "learning_rate": 4.958303785498689e-05,
+      "loss": 2.8845,
+      "step": 21900
+    },
+    {
+      "epoch": 0.17873226621694271,
+      "grad_norm": 0.45124372839927673,
+      "learning_rate": 4.958187125480761e-05,
+      "loss": 2.8767,
+      "step": 21930
+    },
+    {
+      "epoch": 0.1789767700010972,
+      "grad_norm": 0.44482797384262085,
+      "learning_rate": 4.958070303868008e-05,
+      "loss": 2.8577,
+      "step": 21960
+    },
+    {
+      "epoch": 0.1792212737852517,
+      "grad_norm": 0.4298214018344879,
+      "learning_rate": 4.9579533206681066e-05,
+      "loss": 2.8613,
+      "step": 21990
+    },
+    {
+      "epoch": 0.1794657775694062,
+      "grad_norm": 0.4243757426738739,
+      "learning_rate": 4.957836175888749e-05,
+      "loss": 2.8533,
+      "step": 22020
+    },
+    {
+      "epoch": 0.17971028135356074,
+      "grad_norm": 0.43930861353874207,
+      "learning_rate": 4.957718869537635e-05,
+      "loss": 2.8845,
+      "step": 22050
+    },
+    {
+      "epoch": 0.17995478513771523,
+      "grad_norm": 0.42320388555526733,
+      "learning_rate": 4.957601401622477e-05,
+      "loss": 2.8757,
+      "step": 22080
+    },
+    {
+      "epoch": 0.18019928892186973,
+      "grad_norm": 0.4299392104148865,
+      "learning_rate": 4.957483772150996e-05,
+      "loss": 2.8794,
+      "step": 22110
+    },
+    {
+      "epoch": 0.18044379270602423,
+      "grad_norm": 0.42219462990760803,
+      "learning_rate": 4.957365981130925e-05,
+      "loss": 2.8451,
+      "step": 22140
+    },
+    {
+      "epoch": 0.18068829649017873,
+      "grad_norm": 0.4573104977607727,
+      "learning_rate": 4.957248028570007e-05,
+      "loss": 2.8643,
+      "step": 22170
+    },
+    {
+      "epoch": 0.18093280027433326,
+      "grad_norm": 0.41791173815727234,
+      "learning_rate": 4.9571299144759965e-05,
+      "loss": 2.872,
+      "step": 22200
+    },
+    {
+      "epoch": 0.18117730405848775,
+      "grad_norm": 0.42580413818359375,
+      "learning_rate": 4.957011638856658e-05,
+      "loss": 2.8639,
+      "step": 22230
+    },
+    {
+      "epoch": 0.18142180784264225,
+      "grad_norm": 0.4491201937198639,
+      "learning_rate": 4.956893201719766e-05,
+      "loss": 2.8571,
+      "step": 22260
+    },
+    {
+      "epoch": 0.18166631162679675,
+      "grad_norm": 0.4404660761356354,
+      "learning_rate": 4.956774603073107e-05,
+      "loss": 2.8626,
+      "step": 22290
+    },
+    {
+      "epoch": 0.18191081541095128,
+      "grad_norm": 0.4323899447917938,
+      "learning_rate": 4.956655842924477e-05,
+      "loss": 2.877,
+      "step": 22320
+    },
+    {
+      "epoch": 0.18215531919510577,
+      "grad_norm": 0.42996013164520264,
+      "learning_rate": 4.9565369212816825e-05,
+      "loss": 2.858,
+      "step": 22350
+    },
+    {
+      "epoch": 0.18239982297926027,
+      "grad_norm": 0.42361024022102356,
+      "learning_rate": 4.9564178381525414e-05,
+      "loss": 2.8497,
+      "step": 22380
+    },
+    {
+      "epoch": 0.18264432676341477,
+      "grad_norm": 0.42209547758102417,
+      "learning_rate": 4.956298593544883e-05,
+      "loss": 2.8664,
+      "step": 22410
+    },
+    {
+      "epoch": 0.18288883054756927,
+      "grad_norm": 0.42259085178375244,
+      "learning_rate": 4.9561791874665445e-05,
+      "loss": 2.8416,
+      "step": 22440
+    },
+    {
+      "epoch": 0.1831333343317238,
+      "grad_norm": 0.4607374668121338,
+      "learning_rate": 4.956059619925376e-05,
+      "loss": 2.8708,
+      "step": 22470
+    },
+    {
+      "epoch": 0.1833778381158783,
+      "grad_norm": 0.43727099895477295,
+      "learning_rate": 4.9559398909292376e-05,
+      "loss": 2.8641,
+      "step": 22500
+    },
+    {
+      "epoch": 0.1836223419000328,
+      "grad_norm": 0.42699259519577026,
+      "learning_rate": 4.955820000486e-05,
+      "loss": 2.8532,
+      "step": 22530
+    },
+    {
+      "epoch": 0.1838668456841873,
+      "grad_norm": 0.4515063464641571,
+      "learning_rate": 4.9556999486035444e-05,
+      "loss": 2.8692,
+      "step": 22560
+    },
+    {
+      "epoch": 0.1841113494683418,
+      "grad_norm": 0.4204477369785309,
+      "learning_rate": 4.9555797352897625e-05,
+      "loss": 2.8586,
+      "step": 22590
+    },
+    {
+      "epoch": 0.18435585325249632,
+      "grad_norm": 0.43957340717315674,
+      "learning_rate": 4.955459360552558e-05,
+      "loss": 2.8553,
+      "step": 22620
+    },
+    {
+      "epoch": 0.18460035703665081,
+      "grad_norm": 0.4496155381202698,
+      "learning_rate": 4.9553388243998424e-05,
+      "loss": 2.8722,
+      "step": 22650
+    },
+    {
+      "epoch": 0.1848448608208053,
+      "grad_norm": 0.4415959119796753,
+      "learning_rate": 4.9552181268395404e-05,
+      "loss": 2.8521,
+      "step": 22680
+    },
+    {
+      "epoch": 0.1850893646049598,
+      "grad_norm": 0.43609628081321716,
+      "learning_rate": 4.955097267879586e-05,
+      "loss": 2.8404,
+      "step": 22710
+    },
+    {
+      "epoch": 0.18533386838911434,
+      "grad_norm": 0.43372586369514465,
+      "learning_rate": 4.954976247527924e-05,
+      "loss": 2.8526,
+      "step": 22740
+    },
+    {
+      "epoch": 0.18557837217326884,
+      "grad_norm": 0.4543731212615967,
+      "learning_rate": 4.9548550657925095e-05,
+      "loss": 2.8468,
+      "step": 22770
+    },
+    {
+      "epoch": 0.18582287595742333,
+      "grad_norm": 0.4426426589488983,
+      "learning_rate": 4.954733722681311e-05,
+      "loss": 2.8554,
+      "step": 22800
+    },
+    {
+      "epoch": 0.18606737974157783,
+      "grad_norm": 0.44420433044433594,
+      "learning_rate": 4.954612218202303e-05,
+      "loss": 2.8546,
+      "step": 22830
+    },
+    {
+      "epoch": 0.18631188352573233,
+      "grad_norm": 0.41634631156921387,
+      "learning_rate": 4.954490552363473e-05,
+      "loss": 2.8629,
+      "step": 22860
+    },
+    {
+      "epoch": 0.18655638730988686,
+      "grad_norm": 0.44227173924446106,
+      "learning_rate": 4.95436872517282e-05,
+      "loss": 2.8612,
+      "step": 22890
+    },
+    {
+      "epoch": 0.18680089109404135,
+      "grad_norm": 0.4149799048900604,
+      "learning_rate": 4.954246736638352e-05,
+      "loss": 2.8627,
+      "step": 22920
+    },
+    {
+      "epoch": 0.18704539487819585,
+      "grad_norm": 0.43874844908714294,
+      "learning_rate": 4.954124586768088e-05,
+      "loss": 2.8516,
+      "step": 22950
+    },
+    {
+      "epoch": 0.18728989866235035,
+      "grad_norm": 0.4350124001502991,
+      "learning_rate": 4.9540022755700587e-05,
+      "loss": 2.8583,
+      "step": 22980
+    },
+    {
+      "epoch": 0.18753440244650488,
+      "grad_norm": 0.43529069423675537,
+      "learning_rate": 4.953879803052304e-05,
+      "loss": 2.8631,
+      "step": 23010
+    },
+    {
+      "epoch": 0.18777890623065938,
+      "grad_norm": 0.43385937809944153,
+      "learning_rate": 4.9537571692228744e-05,
+      "loss": 2.8493,
+      "step": 23040
+    },
+    {
+      "epoch": 0.18802341001481387,
+      "grad_norm": 0.4345625936985016,
+      "learning_rate": 4.953634374089832e-05,
+      "loss": 2.8463,
+      "step": 23070
+    },
+    {
+      "epoch": 0.18826791379896837,
+      "grad_norm": 0.44867751002311707,
+      "learning_rate": 4.953511417661249e-05,
+      "loss": 2.8507,
+      "step": 23100
+    },
+    {
+      "epoch": 0.18851241758312287,
+      "grad_norm": 0.4386790692806244,
+      "learning_rate": 4.953388299945209e-05,
+      "loss": 2.847,
+      "step": 23130
+    },
+    {
+      "epoch": 0.1887569213672774,
+      "grad_norm": 0.4632754325866699,
+      "learning_rate": 4.9532650209498045e-05,
+      "loss": 2.8569,
+      "step": 23160
+    },
+    {
+      "epoch": 0.1890014251514319,
+      "grad_norm": 0.4407399296760559,
+      "learning_rate": 4.9531415806831394e-05,
+      "loss": 2.8541,
+      "step": 23190
+    },
+    {
+      "epoch": 0.1892459289355864,
+      "grad_norm": 0.45401519536972046,
+      "learning_rate": 4.953017979153329e-05,
+      "loss": 2.8702,
+      "step": 23220
+    },
+    {
+      "epoch": 0.1894904327197409,
+      "grad_norm": 0.42486798763275146,
+      "learning_rate": 4.952894216368498e-05,
+      "loss": 2.8701,
+      "step": 23250
+    },
+    {
+      "epoch": 0.1897349365038954,
+      "grad_norm": 0.4279034733772278,
+      "learning_rate": 4.952770292336782e-05,
+      "loss": 2.8429,
+      "step": 23280
+    },
+    {
+      "epoch": 0.18997944028804992,
+      "grad_norm": 0.4184633791446686,
+      "learning_rate": 4.9526462070663294e-05,
+      "loss": 2.8358,
+      "step": 23310
+    },
+    {
+      "epoch": 0.19022394407220442,
+      "grad_norm": 0.4264056086540222,
+      "learning_rate": 4.952521960565295e-05,
+      "loss": 2.8448,
+      "step": 23340
+    },
+    {
+      "epoch": 0.1904684478563589,
+      "grad_norm": 0.43503764271736145,
+      "learning_rate": 4.952397552841846e-05,
+      "loss": 2.851,
+      "step": 23370
+    },
+    {
+      "epoch": 0.1907129516405134,
+      "grad_norm": 0.4345065653324127,
+      "learning_rate": 4.952272983904164e-05,
+      "loss": 2.8468,
+      "step": 23400
+    },
+    {
+      "epoch": 0.19095745542466794,
+      "grad_norm": 0.4394157826900482,
+      "learning_rate": 4.9521482537604345e-05,
+      "loss": 2.837,
+      "step": 23430
+    },
+    {
+      "epoch": 0.19120195920882244,
+      "grad_norm": 0.4336976408958435,
+      "learning_rate": 4.9520233624188585e-05,
+      "loss": 2.8516,
+      "step": 23460
+    },
+    {
+      "epoch": 0.19144646299297693,
+      "grad_norm": 0.42082545161247253,
+      "learning_rate": 4.951898309887646e-05,
+      "loss": 2.8616,
+      "step": 23490
+    },
+    {
+      "epoch": 0.19169096677713143,
+      "grad_norm": 0.44079628586769104,
+      "learning_rate": 4.951773096175016e-05,
+      "loss": 2.8371,
+      "step": 23520
+    },
+    {
+      "epoch": 0.19193547056128593,
+      "grad_norm": 0.4336305558681488,
+      "learning_rate": 4.9516477212892024e-05,
+      "loss": 2.8398,
+      "step": 23550
+    },
+    {
+      "epoch": 0.19217997434544046,
+      "grad_norm": 0.42380383610725403,
+      "learning_rate": 4.9515221852384455e-05,
+      "loss": 2.8444,
+      "step": 23580
+    },
+    {
+      "epoch": 0.19242447812959496,
+      "grad_norm": 0.43711090087890625,
+      "learning_rate": 4.9513964880309974e-05,
+      "loss": 2.8399,
+      "step": 23610
+    },
+    {
+      "epoch": 0.19266898191374945,
+      "grad_norm": 0.44011425971984863,
+      "learning_rate": 4.951270629675121e-05,
+      "loss": 2.8585,
+      "step": 23640
+    },
+    {
+      "epoch": 0.19291348569790395,
+      "grad_norm": 0.4214531183242798,
+      "learning_rate": 4.951144610179092e-05,
+      "loss": 2.8444,
+      "step": 23670
+    },
+    {
+      "epoch": 0.19315798948205848,
+      "grad_norm": 0.43613433837890625,
+      "learning_rate": 4.9510184295511915e-05,
+      "loss": 2.85,
+      "step": 23700
+    },
+    {
+      "epoch": 0.19340249326621298,
+      "grad_norm": 0.43342262506484985,
+      "learning_rate": 4.950892087799716e-05,
+      "loss": 2.8515,
+      "step": 23730
+    },
+    {
+      "epoch": 0.19364699705036748,
+      "grad_norm": 0.46385619044303894,
+      "learning_rate": 4.950765584932971e-05,
+      "loss": 2.8542,
+      "step": 23760
+    },
+    {
+      "epoch": 0.19389150083452197,
+      "grad_norm": 0.43448546528816223,
+      "learning_rate": 4.950638920959273e-05,
+      "loss": 2.8407,
+      "step": 23790
+    },
+    {
+      "epoch": 0.19413600461867647,
+      "grad_norm": 0.43169257044792175,
+      "learning_rate": 4.950512095886947e-05,
+      "loss": 2.8203,
+      "step": 23820
+    },
+    {
+      "epoch": 0.194380508402831,
+      "grad_norm": 0.4383741617202759,
+      "learning_rate": 4.950385109724331e-05,
+      "loss": 2.8434,
+      "step": 23850
+    },
+    {
+      "epoch": 0.1946250121869855,
+      "grad_norm": 0.44045546650886536,
+      "learning_rate": 4.950257962479773e-05,
+      "loss": 2.8484,
+      "step": 23880
+    },
+    {
+      "epoch": 0.19486951597114,
+      "grad_norm": 0.4332634210586548,
+      "learning_rate": 4.95013065416163e-05,
+      "loss": 2.8428,
+      "step": 23910
+    },
+    {
+      "epoch": 0.1951140197552945,
+      "grad_norm": 0.4307137429714203,
+      "learning_rate": 4.950003184778272e-05,
+      "loss": 2.8264,
+      "step": 23940
+    },
+    {
+      "epoch": 0.195358523539449,
+      "grad_norm": 0.4232560694217682,
+      "learning_rate": 4.9498755543380795e-05,
+      "loss": 2.8307,
+      "step": 23970
+    },
+    {
+      "epoch": 0.19560302732360352,
+      "grad_norm": 0.42863690853118896,
+      "learning_rate": 4.949747762849441e-05,
+      "loss": 2.8349,
+      "step": 24000
+    },
+    {
+      "epoch": 0.19584753110775802,
+      "grad_norm": 0.4395289719104767,
+      "learning_rate": 4.949619810320757e-05,
+      "loss": 2.838,
+      "step": 24030
+    },
+    {
+      "epoch": 0.19609203489191251,
+      "grad_norm": 0.4364834725856781,
+      "learning_rate": 4.9494916967604404e-05,
+      "loss": 2.8399,
+      "step": 24060
+    },
+    {
+      "epoch": 0.196336538676067,
+      "grad_norm": 0.4557901918888092,
+      "learning_rate": 4.949363422176911e-05,
+      "loss": 2.8317,
+      "step": 24090
+    },
+    {
+      "epoch": 0.19658104246022154,
+      "grad_norm": 0.4531985819339752,
+      "learning_rate": 4.9492349865786034e-05,
+      "loss": 2.8307,
+      "step": 24120
+    },
+    {
+      "epoch": 0.19682554624437604,
+      "grad_norm": 0.4378800392150879,
+      "learning_rate": 4.9491063899739596e-05,
+      "loss": 2.8351,
+      "step": 24150
+    },
+    {
+      "epoch": 0.19707005002853054,
+      "grad_norm": 0.4349973201751709,
+      "learning_rate": 4.948977632371432e-05,
+      "loss": 2.832,
+      "step": 24180
+    },
+    {
+      "epoch": 0.19731455381268503,
+      "grad_norm": 0.4363759756088257,
+      "learning_rate": 4.948848713779487e-05,
+      "loss": 2.8357,
+      "step": 24210
+    },
+    {
+      "epoch": 0.19755905759683953,
+      "grad_norm": 0.42884907126426697,
+      "learning_rate": 4.948719634206598e-05,
+      "loss": 2.8272,
+      "step": 24240
+    },
+    {
+      "epoch": 0.19780356138099406,
+      "grad_norm": 0.4273601174354553,
+      "learning_rate": 4.9485903936612504e-05,
+      "loss": 2.8675,
+      "step": 24270
+    },
+    {
+      "epoch": 0.19804806516514856,
+      "grad_norm": 0.4261528253555298,
+      "learning_rate": 4.948460992151941e-05,
+      "loss": 2.8254,
+      "step": 24300
+    },
+    {
+      "epoch": 0.19829256894930306,
+      "grad_norm": 0.41921886801719666,
+      "learning_rate": 4.948331429687175e-05,
+      "loss": 2.8469,
+      "step": 24330
+    },
+    {
+      "epoch": 0.19853707273345755,
+      "grad_norm": 0.436755508184433,
+      "learning_rate": 4.9482017062754704e-05,
+      "loss": 2.8461,
+      "step": 24360
+    },
+    {
+      "epoch": 0.19878157651761208,
+      "grad_norm": 0.4177142381668091,
+      "learning_rate": 4.9480718219253555e-05,
+      "loss": 2.8252,
+      "step": 24390
+    },
+    {
+      "epoch": 0.19902608030176658,
+      "grad_norm": 0.4457404911518097,
+      "learning_rate": 4.947941776645366e-05,
+      "loss": 2.8434,
+      "step": 24420
+    },
+    {
+      "epoch": 0.19927058408592108,
+      "grad_norm": 0.44040533900260925,
+      "learning_rate": 4.9478115704440545e-05,
+      "loss": 2.8336,
+      "step": 24450
+    },
+    {
+      "epoch": 0.19951508787007557,
+      "grad_norm": 0.4262928068637848,
+      "learning_rate": 4.947681203329977e-05,
+      "loss": 2.8407,
+      "step": 24480
+    },
+    {
+      "epoch": 0.19975959165423007,
+      "grad_norm": 0.4242199957370758,
+      "learning_rate": 4.947550675311706e-05,
+      "loss": 2.8235,
+      "step": 24510
+    },
+    {
+      "epoch": 0.2000040954383846,
+      "grad_norm": 0.4429243803024292,
+      "learning_rate": 4.94741998639782e-05,
+      "loss": 2.8327,
+      "step": 24540
+    },
+    {
+      "epoch": 0.2002485992225391,
+      "grad_norm": 0.42853012681007385,
+      "learning_rate": 4.947289136596912e-05,
+      "loss": 2.844,
+      "step": 24570
+    },
+    {
+      "epoch": 0.2004931030066936,
+      "grad_norm": 0.4213114082813263,
+      "learning_rate": 4.9471581259175816e-05,
+      "loss": 2.8381,
+      "step": 24600
+    },
+    {
+      "epoch": 0.2007376067908481,
+      "grad_norm": 0.42277562618255615,
+      "learning_rate": 4.9470269543684436e-05,
+      "loss": 2.8443,
+      "step": 24630
+    },
+    {
+      "epoch": 0.2009821105750026,
+      "grad_norm": 0.42599666118621826,
+      "learning_rate": 4.9468956219581186e-05,
+      "loss": 2.8377,
+      "step": 24660
+    },
+    {
+      "epoch": 0.20122661435915712,
+      "grad_norm": 0.4356340169906616,
+      "learning_rate": 4.946764128695242e-05,
+      "loss": 2.8322,
+      "step": 24690
+    },
+    {
+      "epoch": 0.20147111814331162,
+      "grad_norm": 0.42555105686187744,
+      "learning_rate": 4.9466324745884565e-05,
+      "loss": 2.8418,
+      "step": 24720
+    },
+    {
+      "epoch": 0.20171562192746612,
+      "grad_norm": 0.4337792992591858,
+      "learning_rate": 4.946500659646417e-05,
+      "loss": 2.8426,
+      "step": 24750
+    },
+    {
+      "epoch": 0.20196012571162061,
+      "grad_norm": 0.4453223943710327,
+      "learning_rate": 4.9463686838777887e-05,
+      "loss": 2.8329,
+      "step": 24780
+    },
+    {
+      "epoch": 0.20220462949577514,
+      "grad_norm": 0.45427435636520386,
+      "learning_rate": 4.946236547291248e-05,
+      "loss": 2.825,
+      "step": 24810
+    },
+    {
+      "epoch": 0.20244913327992964,
+      "grad_norm": 0.42173105478286743,
+      "learning_rate": 4.94610424989548e-05,
+      "loss": 2.8412,
+      "step": 24840
+    },
+    {
+      "epoch": 0.20269363706408414,
+      "grad_norm": 0.4376053810119629,
+      "learning_rate": 4.9459717916991834e-05,
+      "loss": 2.8071,
+      "step": 24870
+    },
+    {
+      "epoch": 0.20293814084823864,
+      "grad_norm": 0.4222453832626343,
+      "learning_rate": 4.945839172711063e-05,
+      "loss": 2.8369,
+      "step": 24900
+    },
+    {
+      "epoch": 0.20318264463239313,
+      "grad_norm": 0.4197344481945038,
+      "learning_rate": 4.9457063929398394e-05,
+      "loss": 2.8411,
+      "step": 24930
+    },
+    {
+      "epoch": 0.20342714841654766,
+      "grad_norm": 0.41742411255836487,
+      "learning_rate": 4.94557345239424e-05,
+      "loss": 2.8138,
+      "step": 24960
+    },
+    {
+      "epoch": 0.20367165220070216,
+      "grad_norm": 0.4322751462459564,
+      "learning_rate": 4.945440351083003e-05,
+      "loss": 2.832,
+      "step": 24990
+    },
+    {
+      "epoch": 0.20391615598485666,
+      "grad_norm": 0.42838114500045776,
+      "learning_rate": 4.94530708901488e-05,
+      "loss": 2.8517,
+      "step": 25020
+    },
+    {
+      "epoch": 0.20416065976901115,
+      "grad_norm": 0.44520124793052673,
+      "learning_rate": 4.9451736661986306e-05,
+      "loss": 2.8259,
+      "step": 25050
+    },
+    {
+      "epoch": 0.20440516355316568,
+      "grad_norm": 0.4274967610836029,
+      "learning_rate": 4.945040082643026e-05,
+      "loss": 2.8189,
+      "step": 25080
+    },
+    {
+      "epoch": 0.20464966733732018,
+      "grad_norm": 0.5348169207572937,
+      "learning_rate": 4.944906338356846e-05,
+      "loss": 2.8208,
+      "step": 25110
+    },
+    {
+      "epoch": 0.20489417112147468,
+      "grad_norm": 0.43682798743247986,
+      "learning_rate": 4.944772433348884e-05,
+      "loss": 2.8243,
+      "step": 25140
+    },
+    {
+      "epoch": 0.20513867490562918,
+      "grad_norm": 0.44586774706840515,
+      "learning_rate": 4.9446383676279435e-05,
+      "loss": 2.8349,
+      "step": 25170
+    },
+    {
+      "epoch": 0.20538317868978367,
+      "grad_norm": 0.42513328790664673,
+      "learning_rate": 4.944504141202836e-05,
+      "loss": 2.8211,
+      "step": 25200
+    },
+    {
+      "epoch": 0.2056276824739382,
+      "grad_norm": 0.41674286127090454,
+      "learning_rate": 4.944369754082385e-05,
+      "loss": 2.815,
+      "step": 25230
+    },
+    {
+      "epoch": 0.2058721862580927,
+      "grad_norm": 0.4303615391254425,
+      "learning_rate": 4.944235206275426e-05,
+      "loss": 2.8233,
+      "step": 25260
+    },
+    {
+      "epoch": 0.2061166900422472,
+      "grad_norm": 0.44005194306373596,
+      "learning_rate": 4.9441004977908036e-05,
+      "loss": 2.8205,
+      "step": 25290
+    },
+    {
+      "epoch": 0.2063611938264017,
+      "grad_norm": 0.43432343006134033,
+      "learning_rate": 4.943965628637373e-05,
+      "loss": 2.8139,
+      "step": 25320
+    },
+    {
+      "epoch": 0.2066056976105562,
+      "grad_norm": 0.42410290241241455,
+      "learning_rate": 4.943830598823999e-05,
+      "loss": 2.8063,
+      "step": 25350
+    },
+    {
+      "epoch": 0.20685020139471072,
+      "grad_norm": 0.4405938386917114,
+      "learning_rate": 4.94369540835956e-05,
+      "loss": 2.8065,
+      "step": 25380
+    },
+    {
+      "epoch": 0.20709470517886522,
+      "grad_norm": 0.42279052734375,
+      "learning_rate": 4.9435600572529415e-05,
+      "loss": 2.8279,
+      "step": 25410
+    },
+    {
+      "epoch": 0.20733920896301972,
+      "grad_norm": 0.42218053340911865,
+      "learning_rate": 4.9434245455130425e-05,
+      "loss": 2.8041,
+      "step": 25440
+    },
+    {
+      "epoch": 0.20758371274717421,
+      "grad_norm": 0.43804386258125305,
+      "learning_rate": 4.9432888731487705e-05,
+      "loss": 2.8311,
+      "step": 25470
+    },
+    {
+      "epoch": 0.20782821653132874,
+      "grad_norm": 0.4266342222690582,
+      "learning_rate": 4.943153040169044e-05,
+      "loss": 2.8288,
+      "step": 25500
+    },
+    {
+      "epoch": 0.20807272031548324,
+      "grad_norm": 0.4279581904411316,
+      "learning_rate": 4.943017046582793e-05,
+      "loss": 2.8138,
+      "step": 25530
+    },
+    {
+      "epoch": 0.20831722409963774,
+      "grad_norm": 0.42624422907829285,
+      "learning_rate": 4.942880892398957e-05,
+      "loss": 2.815,
+      "step": 25560
+    },
+    {
+      "epoch": 0.20856172788379224,
+      "grad_norm": 0.4279497265815735,
+      "learning_rate": 4.942744577626485e-05,
+      "loss": 2.8178,
+      "step": 25590
+    },
+    {
+      "epoch": 0.20880623166794673,
+      "grad_norm": 0.4416673481464386,
+      "learning_rate": 4.942608102274341e-05,
+      "loss": 2.8055,
+      "step": 25620
+    },
+    {
+      "epoch": 0.20905073545210126,
+      "grad_norm": 0.4616566300392151,
+      "learning_rate": 4.942471466351494e-05,
+      "loss": 2.8229,
+      "step": 25650
+    },
+    {
+      "epoch": 0.20929523923625576,
+      "grad_norm": 0.4455949366092682,
+      "learning_rate": 4.942334669866927e-05,
+      "loss": 2.8176,
+      "step": 25680
+    },
+    {
+      "epoch": 0.20953974302041026,
+      "grad_norm": 0.4414938986301422,
+      "learning_rate": 4.9421977128296326e-05,
+      "loss": 2.8182,
+      "step": 25710
+    },
+    {
+      "epoch": 0.20978424680456476,
+      "grad_norm": 0.4289626181125641,
+      "learning_rate": 4.942060595248614e-05,
+      "loss": 2.7938,
+      "step": 25740
+    },
+    {
+      "epoch": 0.21002875058871925,
+      "grad_norm": 0.450878381729126,
+      "learning_rate": 4.941923317132885e-05,
+      "loss": 2.7949,
+      "step": 25770
+    },
+    {
+      "epoch": 0.21027325437287378,
+      "grad_norm": 0.44763535261154175,
+      "learning_rate": 4.941785878491471e-05,
+      "loss": 2.7994,
+      "step": 25800
+    },
+    {
+      "epoch": 0.21051775815702828,
+      "grad_norm": 0.4248661696910858,
+      "learning_rate": 4.941648279333404e-05,
+      "loss": 2.8302,
+      "step": 25830
+    },
+    {
+      "epoch": 0.21076226194118278,
+      "grad_norm": 0.435785174369812,
+      "learning_rate": 4.941510519667732e-05,
+      "loss": 2.8257,
+      "step": 25860
+    },
+    {
+      "epoch": 0.21100676572533728,
+      "grad_norm": 0.43406128883361816,
+      "learning_rate": 4.94137259950351e-05,
+      "loss": 2.8225,
+      "step": 25890
+    },
+    {
+      "epoch": 0.2112512695094918,
+      "grad_norm": 0.45253458619117737,
+      "learning_rate": 4.9412345188498046e-05,
+      "loss": 2.8392,
+      "step": 25920
+    },
+    {
+      "epoch": 0.2114957732936463,
+      "grad_norm": 0.417129784822464,
+      "learning_rate": 4.9410962777156925e-05,
+      "loss": 2.7987,
+      "step": 25950
+    },
+    {
+      "epoch": 0.2117402770778008,
+      "grad_norm": 0.4531177580356598,
+      "learning_rate": 4.940957876110263e-05,
+      "loss": 2.8175,
+      "step": 25980
+    },
+    {
+      "epoch": 0.2119847808619553,
+      "grad_norm": 0.41148507595062256,
+      "learning_rate": 4.940819314042611e-05,
+      "loss": 2.8274,
+      "step": 26010
+    },
+    {
+      "epoch": 0.2122292846461098,
+      "grad_norm": 0.4292708933353424,
+      "learning_rate": 4.9406805915218486e-05,
+      "loss": 2.8122,
+      "step": 26040
+    },
+    {
+      "epoch": 0.21247378843026432,
+      "grad_norm": 0.4388817548751831,
+      "learning_rate": 4.940541708557093e-05,
+      "loss": 2.8086,
+      "step": 26070
+    },
+    {
+      "epoch": 0.21271829221441882,
+      "grad_norm": 0.4264972507953644,
+      "learning_rate": 4.940402665157475e-05,
+      "loss": 2.7905,
+      "step": 26100
+    },
+    {
+      "epoch": 0.21296279599857332,
+      "grad_norm": 0.4268186092376709,
+      "learning_rate": 4.9402634613321344e-05,
+      "loss": 2.819,
+      "step": 26130
+    },
+    {
+      "epoch": 0.21320729978272782,
+      "grad_norm": 0.4356623589992523,
+      "learning_rate": 4.940124097090222e-05,
+      "loss": 2.7987,
+      "step": 26160
+    },
+    {
+      "epoch": 0.21345180356688234,
+      "grad_norm": 0.43201175332069397,
+      "learning_rate": 4.9399845724409e-05,
+      "loss": 2.8197,
+      "step": 26190
+    },
+    {
+      "epoch": 0.21369630735103684,
+      "grad_norm": 0.4420289099216461,
+      "learning_rate": 4.939844887393339e-05,
+      "loss": 2.8036,
+      "step": 26220
+    },
+    {
+      "epoch": 0.21394081113519134,
+      "grad_norm": 0.4327016770839691,
+      "learning_rate": 4.939705041956724e-05,
+      "loss": 2.8299,
+      "step": 26250
+    },
+    {
+      "epoch": 0.21418531491934584,
+      "grad_norm": 0.4344867467880249,
+      "learning_rate": 4.939565036140245e-05,
+      "loss": 2.8083,
+      "step": 26280
+    },
+    {
+      "epoch": 0.21442981870350034,
+      "grad_norm": 0.4580075144767761,
+      "learning_rate": 4.939424869953108e-05,
+      "loss": 2.8023,
+      "step": 26310
+    },
+    {
+      "epoch": 0.21467432248765486,
+      "grad_norm": 0.4417720139026642,
+      "learning_rate": 4.939284543404525e-05,
+      "loss": 2.8054,
+      "step": 26340
+    },
+    {
+      "epoch": 0.21491882627180936,
+      "grad_norm": 0.42408108711242676,
+      "learning_rate": 4.939144056503724e-05,
+      "loss": 2.8288,
+      "step": 26370
+    },
+    {
+      "epoch": 0.21516333005596386,
+      "grad_norm": 0.44218704104423523,
+      "learning_rate": 4.939003409259937e-05,
+      "loss": 2.7957,
+      "step": 26400
+    },
+    {
+      "epoch": 0.21540783384011836,
+      "grad_norm": 0.4381102919578552,
+      "learning_rate": 4.9388626016824115e-05,
+      "loss": 2.8045,
+      "step": 26430
+    },
+    {
+      "epoch": 0.21565233762427286,
+      "grad_norm": 0.42545047402381897,
+      "learning_rate": 4.9387216337804035e-05,
+      "loss": 2.7998,
+      "step": 26460
+    },
+    {
+      "epoch": 0.21589684140842738,
+      "grad_norm": 0.41947245597839355,
+      "learning_rate": 4.938580505563179e-05,
+      "loss": 2.8053,
+      "step": 26490
+    },
+    {
+      "epoch": 0.21614134519258188,
+      "grad_norm": 0.42647966742515564,
+      "learning_rate": 4.9384392170400165e-05,
+      "loss": 2.8039,
+      "step": 26520
+    },
+    {
+      "epoch": 0.21638584897673638,
+      "grad_norm": 0.41264015436172485,
+      "learning_rate": 4.938297768220203e-05,
+      "loss": 2.7871,
+      "step": 26550
+    },
+    {
+      "epoch": 0.21663035276089088,
+      "grad_norm": 0.43038272857666016,
+      "learning_rate": 4.938156159113038e-05,
+      "loss": 2.8032,
+      "step": 26580
+    },
+    {
+      "epoch": 0.2168748565450454,
+      "grad_norm": 0.4356917142868042,
+      "learning_rate": 4.9380143897278294e-05,
+      "loss": 2.7942,
+      "step": 26610
+    },
+    {
+      "epoch": 0.2171193603291999,
+      "grad_norm": 0.4203088879585266,
+      "learning_rate": 4.937872460073898e-05,
+      "loss": 2.801,
+      "step": 26640
+    },
+    {
+      "epoch": 0.2173638641133544,
+      "grad_norm": 0.41007059812545776,
+      "learning_rate": 4.937730370160574e-05,
+      "loss": 2.804,
+      "step": 26670
+    },
+    {
+      "epoch": 0.2176083678975089,
+      "grad_norm": 0.41945499181747437,
+      "learning_rate": 4.937588119997196e-05,
+      "loss": 2.7834,
+      "step": 26700
+    },
+    {
+      "epoch": 0.2178528716816634,
+      "grad_norm": 0.4370434880256653,
+      "learning_rate": 4.937445709593117e-05,
+      "loss": 2.802,
+      "step": 26730
+    },
+    {
+      "epoch": 0.21809737546581792,
+      "grad_norm": 0.428362101316452,
+      "learning_rate": 4.937303138957698e-05,
+      "loss": 2.7942,
+      "step": 26760
+    },
+    {
+      "epoch": 0.21834187924997242,
+      "grad_norm": 0.42694905400276184,
+      "learning_rate": 4.937160408100312e-05,
+      "loss": 2.8051,
+      "step": 26790
+    },
+    {
+      "epoch": 0.21858638303412692,
+      "grad_norm": 0.4423834979534149,
+      "learning_rate": 4.937017517030341e-05,
+      "loss": 2.8012,
+      "step": 26820
+    },
+    {
+      "epoch": 0.21883088681828142,
+      "grad_norm": 0.48314982652664185,
+      "learning_rate": 4.9368744657571776e-05,
+      "loss": 2.7941,
+      "step": 26850
+    },
+    {
+      "epoch": 0.21907539060243594,
+      "grad_norm": 0.4422082006931305,
+      "learning_rate": 4.936731254290227e-05,
+      "loss": 2.8067,
+      "step": 26880
+    },
+    {
+      "epoch": 0.21931989438659044,
+      "grad_norm": 0.43247443437576294,
+      "learning_rate": 4.936587882638903e-05,
+      "loss": 2.7909,
+      "step": 26910
+    },
+    {
+      "epoch": 0.21956439817074494,
+      "grad_norm": 0.42312097549438477,
+      "learning_rate": 4.93644435081263e-05,
+      "loss": 2.8253,
+      "step": 26940
+    },
+    {
+      "epoch": 0.21980890195489944,
+      "grad_norm": 0.42186856269836426,
+      "learning_rate": 4.936300658820844e-05,
+      "loss": 2.7949,
+      "step": 26970
+    },
+    {
+      "epoch": 0.22005340573905394,
+      "grad_norm": 0.4168436527252197,
+      "learning_rate": 4.936156806672991e-05,
+      "loss": 2.8036,
+      "step": 27000
+    },
+    {
+      "epoch": 0.22029790952320846,
+      "grad_norm": 0.42488592863082886,
+      "learning_rate": 4.9360127943785275e-05,
+      "loss": 2.7866,
+      "step": 27030
+    },
+    {
+      "epoch": 0.22054241330736296,
+      "grad_norm": 0.41458776593208313,
+      "learning_rate": 4.93586862194692e-05,
+      "loss": 2.7945,
+      "step": 27060
+    },
+    {
+      "epoch": 0.22078691709151746,
+      "grad_norm": 0.43253812193870544,
+      "learning_rate": 4.935724289387647e-05,
+      "loss": 2.8058,
+      "step": 27090
+    },
+    {
+      "epoch": 0.22103142087567196,
+      "grad_norm": 0.42101171612739563,
+      "learning_rate": 4.9355797967101946e-05,
+      "loss": 2.7928,
+      "step": 27120
+    },
+    {
+      "epoch": 0.22127592465982646,
+      "grad_norm": 0.4362278878688812,
+      "learning_rate": 4.935435143924063e-05,
+      "loss": 2.7996,
+      "step": 27150
+    },
+    {
+      "epoch": 0.22152042844398098,
+      "grad_norm": 0.4580351412296295,
+      "learning_rate": 4.935290331038761e-05,
+      "loss": 2.8006,
+      "step": 27180
+    },
+    {
+      "epoch": 0.22176493222813548,
+      "grad_norm": 0.4496763348579407,
+      "learning_rate": 4.935145358063808e-05,
+      "loss": 2.7921,
+      "step": 27210
+    },
+    {
+      "epoch": 0.22200943601228998,
+      "grad_norm": 0.43551400303840637,
+      "learning_rate": 4.9350002250087344e-05,
+      "loss": 2.8082,
+      "step": 27240
+    },
+    {
+      "epoch": 0.22225393979644448,
+      "grad_norm": 0.4334968626499176,
+      "learning_rate": 4.934854931883081e-05,
+      "loss": 2.7888,
+      "step": 27270
+    },
+    {
+      "epoch": 0.222498443580599,
+      "grad_norm": 0.4326109290122986,
+      "learning_rate": 4.934709478696398e-05,
+      "loss": 2.7832,
+      "step": 27300
+    },
+    {
+      "epoch": 0.2227429473647535,
+      "grad_norm": 0.4907919466495514,
+      "learning_rate": 4.934563865458248e-05,
+      "loss": 2.7945,
+      "step": 27330
+    },
+    {
+      "epoch": 0.222987451148908,
+      "grad_norm": 0.45372188091278076,
+      "learning_rate": 4.934418092178203e-05,
+      "loss": 2.8092,
+      "step": 27360
+    },
+    {
+      "epoch": 0.2232319549330625,
+      "grad_norm": 0.4431178867816925,
+      "learning_rate": 4.934272158865847e-05,
+      "loss": 2.8066,
+      "step": 27390
+    },
+    {
+      "epoch": 0.223476458717217,
+      "grad_norm": 0.4186871349811554,
+      "learning_rate": 4.934126065530771e-05,
+      "loss": 2.8006,
+      "step": 27420
+    },
+    {
+      "epoch": 0.22372096250137152,
+      "grad_norm": 0.428605854511261,
+      "learning_rate": 4.93397981218258e-05,
+      "loss": 2.7793,
+      "step": 27450
+    },
+    {
+      "epoch": 0.22396546628552602,
+      "grad_norm": 0.44454094767570496,
+      "learning_rate": 4.9338333988308887e-05,
+      "loss": 2.7975,
+      "step": 27480
+    },
+    {
+      "epoch": 0.22420997006968052,
+      "grad_norm": 0.4257412254810333,
+      "learning_rate": 4.933686825485321e-05,
+      "loss": 2.7831,
+      "step": 27510
+    },
+    {
+      "epoch": 0.22445447385383502,
+      "grad_norm": 0.4227311909198761,
+      "learning_rate": 4.933540092155513e-05,
+      "loss": 2.8197,
+      "step": 27540
+    },
+    {
+      "epoch": 0.22469897763798954,
+      "grad_norm": 0.4631737470626831,
+      "learning_rate": 4.933393198851111e-05,
+      "loss": 2.8038,
+      "step": 27570
+    },
+    {
+      "epoch": 0.22494348142214404,
+      "grad_norm": 0.4268395006656647,
+      "learning_rate": 4.933246145581769e-05,
+      "loss": 2.7793,
+      "step": 27600
+    }
+  ],
+  "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": 2.640147320615731e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}