diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,19643 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 1000,
+  "global_step": 27688,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.000361167292689974,
+      "grad_norm": 44.511689821462525,
+      "learning_rate": 3.2490974729241873e-08,
+      "loss": 1.0199,
+      "step": 10
+    },
+    {
+      "epoch": 0.000722334585379948,
+      "grad_norm": 63.11283329459405,
+      "learning_rate": 6.859205776173286e-08,
+      "loss": 1.015,
+      "step": 20
+    },
+    {
+      "epoch": 0.001083501878069922,
+      "grad_norm": 50.092974513563895,
+      "learning_rate": 1.0469314079422382e-07,
+      "loss": 1.0055,
+      "step": 30
+    },
+    {
+      "epoch": 0.001444669170759896,
+      "grad_norm": 46.521355817624354,
+      "learning_rate": 1.407942238267148e-07,
+      "loss": 0.9607,
+      "step": 40
+    },
+    {
+      "epoch": 0.00180583646344987,
+      "grad_norm": 40.7924605496559,
+      "learning_rate": 1.7689530685920579e-07,
+      "loss": 0.8974,
+      "step": 50
+    },
+    {
+      "epoch": 0.002167003756139844,
+      "grad_norm": 36.08859546007994,
+      "learning_rate": 2.1299638989169677e-07,
+      "loss": 0.7474,
+      "step": 60
+    },
+    {
+      "epoch": 0.002528171048829818,
+      "grad_norm": 25.21625175303931,
+      "learning_rate": 2.490974729241877e-07,
+      "loss": 0.6421,
+      "step": 70
+    },
+    {
+      "epoch": 0.002889338341519792,
+      "grad_norm": 21.55773343540464,
+      "learning_rate": 2.851985559566787e-07,
+      "loss": 0.4558,
+      "step": 80
+    },
+    {
+      "epoch": 0.0032505056342097658,
+      "grad_norm": 15.331142402093759,
+      "learning_rate": 3.212996389891697e-07,
+      "loss": 0.3809,
+      "step": 90
+    },
+    {
+      "epoch": 0.00361167292689974,
+      "grad_norm": 8.866126675891149,
+      "learning_rate": 3.574007220216607e-07,
+      "loss": 0.3099,
+      "step": 100
+    },
+    {
+      "epoch": 0.003972840219589714,
+      "grad_norm": 10.827932188798304,
+      "learning_rate": 3.935018050541516e-07,
+      "loss": 0.276,
+      "step": 110
+    },
+    {
+      "epoch": 0.004334007512279688,
+      "grad_norm": 7.413047936592497,
+      "learning_rate": 4.296028880866426e-07,
+      "loss": 0.2693,
+      "step": 120
+    },
+    {
+      "epoch": 0.004695174804969662,
+      "grad_norm": 6.293675260731944,
+      "learning_rate": 4.6570397111913365e-07,
+      "loss": 0.2547,
+      "step": 130
+    },
+    {
+      "epoch": 0.005056342097659636,
+      "grad_norm": 5.534540103488713,
+      "learning_rate": 5.018050541516246e-07,
+      "loss": 0.249,
+      "step": 140
+    },
+    {
+      "epoch": 0.00541750939034961,
+      "grad_norm": 4.510805370269695,
+      "learning_rate": 5.379061371841155e-07,
+      "loss": 0.2393,
+      "step": 150
+    },
+    {
+      "epoch": 0.005778676683039584,
+      "grad_norm": 5.742831395432012,
+      "learning_rate": 5.740072202166065e-07,
+      "loss": 0.2329,
+      "step": 160
+    },
+    {
+      "epoch": 0.006139843975729558,
+      "grad_norm": 5.970864473220756,
+      "learning_rate": 6.101083032490975e-07,
+      "loss": 0.2273,
+      "step": 170
+    },
+    {
+      "epoch": 0.0065010112684195315,
+      "grad_norm": 4.378568717415189,
+      "learning_rate": 6.462093862815884e-07,
+      "loss": 0.2275,
+      "step": 180
+    },
+    {
+      "epoch": 0.006862178561109506,
+      "grad_norm": 7.135518696599739,
+      "learning_rate": 6.823104693140795e-07,
+      "loss": 0.2097,
+      "step": 190
+    },
+    {
+      "epoch": 0.00722334585379948,
+      "grad_norm": 4.529986883296516,
+      "learning_rate": 7.184115523465705e-07,
+      "loss": 0.2049,
+      "step": 200
+    },
+    {
+      "epoch": 0.007584513146489454,
+      "grad_norm": 7.533456886066814,
+      "learning_rate": 7.545126353790614e-07,
+      "loss": 0.2009,
+      "step": 210
+    },
+    {
+      "epoch": 0.007945680439179427,
+      "grad_norm": 4.010874822985454,
+      "learning_rate": 7.906137184115524e-07,
+      "loss": 0.1979,
+      "step": 220
+    },
+    {
+      "epoch": 0.008306847731869403,
+      "grad_norm": 4.307069216740719,
+      "learning_rate": 8.267148014440433e-07,
+      "loss": 0.1881,
+      "step": 230
+    },
+    {
+      "epoch": 0.008668015024559377,
+      "grad_norm": 4.496212995694617,
+      "learning_rate": 8.628158844765343e-07,
+      "loss": 0.1835,
+      "step": 240
+    },
+    {
+      "epoch": 0.00902918231724935,
+      "grad_norm": 4.362755829318934,
+      "learning_rate": 8.989169675090252e-07,
+      "loss": 0.1804,
+      "step": 250
+    },
+    {
+      "epoch": 0.009390349609939324,
+      "grad_norm": 4.3392286093970736,
+      "learning_rate": 9.350180505415163e-07,
+      "loss": 0.1754,
+      "step": 260
+    },
+    {
+      "epoch": 0.009751516902629298,
+      "grad_norm": 4.947227513055756,
+      "learning_rate": 9.71119133574007e-07,
+      "loss": 0.1696,
+      "step": 270
+    },
+    {
+      "epoch": 0.010112684195319271,
+      "grad_norm": 4.302209764509094,
+      "learning_rate": 1.0072202166064982e-06,
+      "loss": 0.1644,
+      "step": 280
+    },
+    {
+      "epoch": 0.010473851488009245,
+      "grad_norm": 5.224064928650982,
+      "learning_rate": 1.0433212996389893e-06,
+      "loss": 0.1583,
+      "step": 290
+    },
+    {
+      "epoch": 0.01083501878069922,
+      "grad_norm": 5.123057604746107,
+      "learning_rate": 1.0794223826714802e-06,
+      "loss": 0.1568,
+      "step": 300
+    },
+    {
+      "epoch": 0.011196186073389194,
+      "grad_norm": 9.362488595106852,
+      "learning_rate": 1.1155234657039712e-06,
+      "loss": 0.1451,
+      "step": 310
+    },
+    {
+      "epoch": 0.011557353366079168,
+      "grad_norm": 3.534571639131067,
+      "learning_rate": 1.1516245487364621e-06,
+      "loss": 0.1461,
+      "step": 320
+    },
+    {
+      "epoch": 0.011918520658769142,
+      "grad_norm": 5.376353939907483,
+      "learning_rate": 1.187725631768953e-06,
+      "loss": 0.1402,
+      "step": 330
+    },
+    {
+      "epoch": 0.012279687951459116,
+      "grad_norm": 3.5604782392308043,
+      "learning_rate": 1.2238267148014441e-06,
+      "loss": 0.1419,
+      "step": 340
+    },
+    {
+      "epoch": 0.01264085524414909,
+      "grad_norm": 3.4583118929152317,
+      "learning_rate": 1.259927797833935e-06,
+      "loss": 0.1329,
+      "step": 350
+    },
+    {
+      "epoch": 0.013002022536839063,
+      "grad_norm": 3.488206183338124,
+      "learning_rate": 1.296028880866426e-06,
+      "loss": 0.1314,
+      "step": 360
+    },
+    {
+      "epoch": 0.013363189829529038,
+      "grad_norm": 5.654974532326477,
+      "learning_rate": 1.332129963898917e-06,
+      "loss": 0.1288,
+      "step": 370
+    },
+    {
+      "epoch": 0.013724357122219012,
+      "grad_norm": 3.8124297181543088,
+      "learning_rate": 1.368231046931408e-06,
+      "loss": 0.1249,
+      "step": 380
+    },
+    {
+      "epoch": 0.014085524414908986,
+      "grad_norm": 5.85995467310869,
+      "learning_rate": 1.404332129963899e-06,
+      "loss": 0.1219,
+      "step": 390
+    },
+    {
+      "epoch": 0.01444669170759896,
+      "grad_norm": 3.3127581958825862,
+      "learning_rate": 1.4404332129963898e-06,
+      "loss": 0.1185,
+      "step": 400
+    },
+    {
+      "epoch": 0.014807859000288933,
+      "grad_norm": 4.179744416992595,
+      "learning_rate": 1.476534296028881e-06,
+      "loss": 0.115,
+      "step": 410
+    },
+    {
+      "epoch": 0.015169026292978907,
+      "grad_norm": 3.120459253546715,
+      "learning_rate": 1.5126353790613718e-06,
+      "loss": 0.1132,
+      "step": 420
+    },
+    {
+      "epoch": 0.015530193585668883,
+      "grad_norm": 4.2096986538576315,
+      "learning_rate": 1.548736462093863e-06,
+      "loss": 0.1147,
+      "step": 430
+    },
+    {
+      "epoch": 0.015891360878358855,
+      "grad_norm": 4.8798744139511,
+      "learning_rate": 1.584837545126354e-06,
+      "loss": 0.1074,
+      "step": 440
+    },
+    {
+      "epoch": 0.01625252817104883,
+      "grad_norm": 4.958199878035793,
+      "learning_rate": 1.6209386281588449e-06,
+      "loss": 0.1084,
+      "step": 450
+    },
+    {
+      "epoch": 0.016613695463738806,
+      "grad_norm": 3.664975582757708,
+      "learning_rate": 1.6570397111913358e-06,
+      "loss": 0.1032,
+      "step": 460
+    },
+    {
+      "epoch": 0.016974862756428778,
+      "grad_norm": 4.846403581032119,
+      "learning_rate": 1.6931407942238267e-06,
+      "loss": 0.1034,
+      "step": 470
+    },
+    {
+      "epoch": 0.017336030049118753,
+      "grad_norm": 3.7583761664026714,
+      "learning_rate": 1.7292418772563178e-06,
+      "loss": 0.0992,
+      "step": 480
+    },
+    {
+      "epoch": 0.017697197341808725,
+      "grad_norm": 3.0213421700340892,
+      "learning_rate": 1.7653429602888088e-06,
+      "loss": 0.0986,
+      "step": 490
+    },
+    {
+      "epoch": 0.0180583646344987,
+      "grad_norm": 2.844434950375668,
+      "learning_rate": 1.8014440433212995e-06,
+      "loss": 0.0964,
+      "step": 500
+    },
+    {
+      "epoch": 0.018419531927188672,
+      "grad_norm": 3.1235860706910734,
+      "learning_rate": 1.8375451263537906e-06,
+      "loss": 0.0985,
+      "step": 510
+    },
+    {
+      "epoch": 0.018780699219878648,
+      "grad_norm": 3.5676891636215697,
+      "learning_rate": 1.8736462093862817e-06,
+      "loss": 0.092,
+      "step": 520
+    },
+    {
+      "epoch": 0.019141866512568623,
+      "grad_norm": 4.962164843230351,
+      "learning_rate": 1.909747292418773e-06,
+      "loss": 0.0894,
+      "step": 530
+    },
+    {
+      "epoch": 0.019503033805258595,
+      "grad_norm": 3.54948960921166,
+      "learning_rate": 1.945848375451264e-06,
+      "loss": 0.0878,
+      "step": 540
+    },
+    {
+      "epoch": 0.01986420109794857,
+      "grad_norm": 5.0876394550805975,
+      "learning_rate": 1.9819494584837546e-06,
+      "loss": 0.0895,
+      "step": 550
+    },
+    {
+      "epoch": 0.020225368390638543,
+      "grad_norm": 4.469519280174299,
+      "learning_rate": 2.0180505415162457e-06,
+      "loss": 0.0847,
+      "step": 560
+    },
+    {
+      "epoch": 0.02058653568332852,
+      "grad_norm": 5.535587540313216,
+      "learning_rate": 2.0541516245487363e-06,
+      "loss": 0.0904,
+      "step": 570
+    },
+    {
+      "epoch": 0.02094770297601849,
+      "grad_norm": 5.472278706118331,
+      "learning_rate": 2.0902527075812274e-06,
+      "loss": 0.0836,
+      "step": 580
+    },
+    {
+      "epoch": 0.021308870268708466,
+      "grad_norm": 5.37037548976207,
+      "learning_rate": 2.1263537906137185e-06,
+      "loss": 0.0829,
+      "step": 590
+    },
+    {
+      "epoch": 0.02167003756139844,
+      "grad_norm": 4.680625344711115,
+      "learning_rate": 2.162454873646209e-06,
+      "loss": 0.0851,
+      "step": 600
+    },
+    {
+      "epoch": 0.022031204854088413,
+      "grad_norm": 10.604777155650206,
+      "learning_rate": 2.1985559566787003e-06,
+      "loss": 0.0791,
+      "step": 610
+    },
+    {
+      "epoch": 0.02239237214677839,
+      "grad_norm": 3.2201306453403564,
+      "learning_rate": 2.2346570397111914e-06,
+      "loss": 0.0793,
+      "step": 620
+    },
+    {
+      "epoch": 0.02275353943946836,
+      "grad_norm": 2.9473569856286637,
+      "learning_rate": 2.2707581227436825e-06,
+      "loss": 0.0792,
+      "step": 630
+    },
+    {
+      "epoch": 0.023114706732158336,
+      "grad_norm": 3.3120823694071424,
+      "learning_rate": 2.3068592057761736e-06,
+      "loss": 0.078,
+      "step": 640
+    },
+    {
+      "epoch": 0.023475874024848308,
+      "grad_norm": 4.23187141175949,
+      "learning_rate": 2.3429602888086643e-06,
+      "loss": 0.077,
+      "step": 650
+    },
+    {
+      "epoch": 0.023837041317538284,
+      "grad_norm": 3.03657532236381,
+      "learning_rate": 2.3790613718411554e-06,
+      "loss": 0.0755,
+      "step": 660
+    },
+    {
+      "epoch": 0.02419820861022826,
+      "grad_norm": 3.467494811487888,
+      "learning_rate": 2.4151624548736464e-06,
+      "loss": 0.0747,
+      "step": 670
+    },
+    {
+      "epoch": 0.02455937590291823,
+      "grad_norm": 2.7848848689495997,
+      "learning_rate": 2.451263537906137e-06,
+      "loss": 0.0697,
+      "step": 680
+    },
+    {
+      "epoch": 0.024920543195608207,
+      "grad_norm": 3.318196400189554,
+      "learning_rate": 2.4873646209386282e-06,
+      "loss": 0.0724,
+      "step": 690
+    },
+    {
+      "epoch": 0.02528171048829818,
+      "grad_norm": 3.916929496783019,
+      "learning_rate": 2.523465703971119e-06,
+      "loss": 0.0702,
+      "step": 700
+    },
+    {
+      "epoch": 0.025642877780988154,
+      "grad_norm": 3.332282508420442,
+      "learning_rate": 2.55956678700361e-06,
+      "loss": 0.0725,
+      "step": 710
+    },
+    {
+      "epoch": 0.026004045073678126,
+      "grad_norm": 5.80855395813645,
+      "learning_rate": 2.595667870036101e-06,
+      "loss": 0.0707,
+      "step": 720
+    },
+    {
+      "epoch": 0.0263652123663681,
+      "grad_norm": 6.14610435325239,
+      "learning_rate": 2.631768953068592e-06,
+      "loss": 0.0679,
+      "step": 730
+    },
+    {
+      "epoch": 0.026726379659058077,
+      "grad_norm": 8.005685584394707,
+      "learning_rate": 2.6678700361010833e-06,
+      "loss": 0.069,
+      "step": 740
+    },
+    {
+      "epoch": 0.02708754695174805,
+      "grad_norm": 6.012594609413831,
+      "learning_rate": 2.703971119133574e-06,
+      "loss": 0.0655,
+      "step": 750
+    },
+    {
+      "epoch": 0.027448714244438024,
+      "grad_norm": 4.601017601119632,
+      "learning_rate": 2.740072202166065e-06,
+      "loss": 0.0687,
+      "step": 760
+    },
+    {
+      "epoch": 0.027809881537127996,
+      "grad_norm": 7.9239090471861795,
+      "learning_rate": 2.776173285198556e-06,
+      "loss": 0.0737,
+      "step": 770
+    },
+    {
+      "epoch": 0.028171048829817972,
+      "grad_norm": 5.507024289456398,
+      "learning_rate": 2.8122743682310472e-06,
+      "loss": 0.0791,
+      "step": 780
+    },
+    {
+      "epoch": 0.028532216122507947,
+      "grad_norm": 4.720177385115065,
+      "learning_rate": 2.848375451263538e-06,
+      "loss": 0.0758,
+      "step": 790
+    },
+    {
+      "epoch": 0.02889338341519792,
+      "grad_norm": 3.297894545538613,
+      "learning_rate": 2.884476534296029e-06,
+      "loss": 0.066,
+      "step": 800
+    },
+    {
+      "epoch": 0.029254550707887895,
+      "grad_norm": 6.200039437688624,
+      "learning_rate": 2.92057761732852e-06,
+      "loss": 0.067,
+      "step": 810
+    },
+    {
+      "epoch": 0.029615718000577867,
+      "grad_norm": 3.487346070844962,
+      "learning_rate": 2.9566787003610108e-06,
+      "loss": 0.0639,
+      "step": 820
+    },
+    {
+      "epoch": 0.029976885293267842,
+      "grad_norm": 2.57436955105714,
+      "learning_rate": 2.992779783393502e-06,
+      "loss": 0.0617,
+      "step": 830
+    },
+    {
+      "epoch": 0.030338052585957814,
+      "grad_norm": 2.923705957430496,
+      "learning_rate": 2.9999993432108755e-06,
+      "loss": 0.0634,
+      "step": 840
+    },
+    {
+      "epoch": 0.03069921987864779,
+      "grad_norm": 2.6326278395261955,
+      "learning_rate": 2.999996675006044e-06,
+      "loss": 0.0605,
+      "step": 850
+    },
+    {
+      "epoch": 0.031060387171337765,
+      "grad_norm": 4.536855915896529,
+      "learning_rate": 2.9999919543398325e-06,
+      "loss": 0.0574,
+      "step": 860
+    },
+    {
+      "epoch": 0.03142155446402774,
+      "grad_norm": 2.444449225726957,
+      "learning_rate": 2.9999851812187014e-06,
+      "loss": 0.0574,
+      "step": 870
+    },
+    {
+      "epoch": 0.03178272175671771,
+      "grad_norm": 2.57087389586591,
+      "learning_rate": 2.999976355651917e-06,
+      "loss": 0.0604,
+      "step": 880
+    },
+    {
+      "epoch": 0.03214388904940769,
+      "grad_norm": 10.571157168733338,
+      "learning_rate": 2.999965477651557e-06,
+      "loss": 0.0592,
+      "step": 890
+    },
+    {
+      "epoch": 0.03250505634209766,
+      "grad_norm": 3.433956868218409,
+      "learning_rate": 2.999952547232504e-06,
+      "loss": 0.0568,
+      "step": 900
+    },
+    {
+      "epoch": 0.03286622363478763,
+      "grad_norm": 3.704131029913721,
+      "learning_rate": 2.999937564412453e-06,
+      "loss": 0.058,
+      "step": 910
+    },
+    {
+      "epoch": 0.03322739092747761,
+      "grad_norm": 2.696350593798532,
+      "learning_rate": 2.999920529211904e-06,
+      "loss": 0.0577,
+      "step": 920
+    },
+    {
+      "epoch": 0.03358855822016758,
+      "grad_norm": 2.604201242464702,
+      "learning_rate": 2.999901441654166e-06,
+      "loss": 0.0543,
+      "step": 930
+    },
+    {
+      "epoch": 0.033949725512857555,
+      "grad_norm": 3.6411147175342164,
+      "learning_rate": 2.9998803017653572e-06,
+      "loss": 0.0564,
+      "step": 940
+    },
+    {
+      "epoch": 0.03431089280554753,
+      "grad_norm": 7.818200873904804,
+      "learning_rate": 2.999857109574404e-06,
+      "loss": 0.0553,
+      "step": 950
+    },
+    {
+      "epoch": 0.034672060098237506,
+      "grad_norm": 3.69420185090383,
+      "learning_rate": 2.99983186511304e-06,
+      "loss": 0.0533,
+      "step": 960
+    },
+    {
+      "epoch": 0.03503322739092748,
+      "grad_norm": 2.5370668016569144,
+      "learning_rate": 2.9998045684158073e-06,
+      "loss": 0.0547,
+      "step": 970
+    },
+    {
+      "epoch": 0.03539439468361745,
+      "grad_norm": 2.311965768201871,
+      "learning_rate": 2.9997752195200567e-06,
+      "loss": 0.0553,
+      "step": 980
+    },
+    {
+      "epoch": 0.03575556197630743,
+      "grad_norm": 3.5169154125571738,
+      "learning_rate": 2.9997438184659467e-06,
+      "loss": 0.0535,
+      "step": 990
+    },
+    {
+      "epoch": 0.0361167292689974,
+      "grad_norm": 2.3245698833622237,
+      "learning_rate": 2.9997103652964425e-06,
+      "loss": 0.0539,
+      "step": 1000
+    },
+    {
+      "epoch": 0.0361167292689974,
+      "eval_loss": 0.14118143916130066,
+      "eval_runtime": 38.7748,
+      "eval_samples_per_second": 105.636,
+      "eval_steps_per_second": 0.206,
+      "step": 1000
+    },
+    {
+      "epoch": 0.03647789656168737,
+      "grad_norm": 2.275034347833786,
+      "learning_rate": 2.9996748600573205e-06,
+      "loss": 0.0525,
+      "step": 1010
+    },
+    {
+      "epoch": 0.036839063854377345,
+      "grad_norm": 2.7945503402831706,
+      "learning_rate": 2.999637302797161e-06,
+      "loss": 0.0484,
+      "step": 1020
+    },
+    {
+      "epoch": 0.037200231147067324,
+      "grad_norm": 2.7887122022618085,
+      "learning_rate": 2.999597693567355e-06,
+      "loss": 0.0522,
+      "step": 1030
+    },
+    {
+      "epoch": 0.037561398439757296,
+      "grad_norm": 4.080998761106243,
+      "learning_rate": 2.9995560324220996e-06,
+      "loss": 0.0513,
+      "step": 1040
+    },
+    {
+      "epoch": 0.03792256573244727,
+      "grad_norm": 3.07147088349794,
+      "learning_rate": 2.999512319418401e-06,
+      "loss": 0.0515,
+      "step": 1050
+    },
+    {
+      "epoch": 0.03828373302513725,
+      "grad_norm": 2.211424487202645,
+      "learning_rate": 2.999466554616072e-06,
+      "loss": 0.0497,
+      "step": 1060
+    },
+    {
+      "epoch": 0.03864490031782722,
+      "grad_norm": 2.8028348454915344,
+      "learning_rate": 2.9994187380777322e-06,
+      "loss": 0.051,
+      "step": 1070
+    },
+    {
+      "epoch": 0.03900606761051719,
+      "grad_norm": 2.1713908105814683,
+      "learning_rate": 2.99936886986881e-06,
+      "loss": 0.0485,
+      "step": 1080
+    },
+    {
+      "epoch": 0.03936723490320716,
+      "grad_norm": 3.5430166816016384,
+      "learning_rate": 2.999316950057541e-06,
+      "loss": 0.0485,
+      "step": 1090
+    },
+    {
+      "epoch": 0.03972840219589714,
+      "grad_norm": 2.244187451963215,
+      "learning_rate": 2.9992629787149674e-06,
+      "loss": 0.0469,
+      "step": 1100
+    },
+    {
+      "epoch": 0.040089569488587114,
+      "grad_norm": 2.3016357125689653,
+      "learning_rate": 2.9992069559149382e-06,
+      "loss": 0.0483,
+      "step": 1110
+    },
+    {
+      "epoch": 0.040450736781277086,
+      "grad_norm": 2.3332804042079833,
+      "learning_rate": 2.9991488817341107e-06,
+      "loss": 0.0474,
+      "step": 1120
+    },
+    {
+      "epoch": 0.040811904073967065,
+      "grad_norm": 2.269634121177138,
+      "learning_rate": 2.9990887562519483e-06,
+      "loss": 0.0471,
+      "step": 1130
+    },
+    {
+      "epoch": 0.04117307136665704,
+      "grad_norm": 3.4255764659801686,
+      "learning_rate": 2.999026579550721e-06,
+      "loss": 0.0475,
+      "step": 1140
+    },
+    {
+      "epoch": 0.04153423865934701,
+      "grad_norm": 3.4248271220795132,
+      "learning_rate": 2.9989623517155064e-06,
+      "loss": 0.0462,
+      "step": 1150
+    },
+    {
+      "epoch": 0.04189540595203698,
+      "grad_norm": 4.8811710660721355,
+      "learning_rate": 2.9988960728341868e-06,
+      "loss": 0.0452,
+      "step": 1160
+    },
+    {
+      "epoch": 0.04225657324472696,
+      "grad_norm": 2.2724522861243934,
+      "learning_rate": 2.998827742997454e-06,
+      "loss": 0.0456,
+      "step": 1170
+    },
+    {
+      "epoch": 0.04261774053741693,
+      "grad_norm": 3.8425070528642276,
+      "learning_rate": 2.9987573622988037e-06,
+      "loss": 0.0461,
+      "step": 1180
+    },
+    {
+      "epoch": 0.042978907830106904,
+      "grad_norm": 2.893236736938225,
+      "learning_rate": 2.9986849308345387e-06,
+      "loss": 0.045,
+      "step": 1190
+    },
+    {
+      "epoch": 0.04334007512279688,
+      "grad_norm": 2.2037419476139704,
+      "learning_rate": 2.9986104487037672e-06,
+      "loss": 0.046,
+      "step": 1200
+    },
+    {
+      "epoch": 0.043701242415486855,
+      "grad_norm": 2.704715861668526,
+      "learning_rate": 2.9985339160084046e-06,
+      "loss": 0.0452,
+      "step": 1210
+    },
+    {
+      "epoch": 0.044062409708176827,
+      "grad_norm": 2.2842391334101815,
+      "learning_rate": 2.998455332853171e-06,
+      "loss": 0.0429,
+      "step": 1220
+    },
+    {
+      "epoch": 0.0444235770008668,
+      "grad_norm": 2.3759783466757227,
+      "learning_rate": 2.9983746993455926e-06,
+      "loss": 0.0441,
+      "step": 1230
+    },
+    {
+      "epoch": 0.04478474429355678,
+      "grad_norm": 4.11286972759774,
+      "learning_rate": 2.9982920155960017e-06,
+      "loss": 0.0438,
+      "step": 1240
+    },
+    {
+      "epoch": 0.04514591158624675,
+      "grad_norm": 2.389568650481896,
+      "learning_rate": 2.998207281717534e-06,
+      "loss": 0.0431,
+      "step": 1250
+    },
+    {
+      "epoch": 0.04550707887893672,
+      "grad_norm": 2.4839424873106055,
+      "learning_rate": 2.998120497826133e-06,
+      "loss": 0.043,
+      "step": 1260
+    },
+    {
+      "epoch": 0.0458682461716267,
+      "grad_norm": 3.998985247532826,
+      "learning_rate": 2.9980316640405454e-06,
+      "loss": 0.0456,
+      "step": 1270
+    },
+    {
+      "epoch": 0.04622941346431667,
+      "grad_norm": 4.0504151987012404,
+      "learning_rate": 2.997940780482324e-06,
+      "loss": 0.0417,
+      "step": 1280
+    },
+    {
+      "epoch": 0.046590580757006644,
+      "grad_norm": 2.154531970849283,
+      "learning_rate": 2.9978478472758248e-06,
+      "loss": 0.0428,
+      "step": 1290
+    },
+    {
+      "epoch": 0.046951748049696616,
+      "grad_norm": 2.1029830201057615,
+      "learning_rate": 2.99775286454821e-06,
+      "loss": 0.043,
+      "step": 1300
+    },
+    {
+      "epoch": 0.047312915342386595,
+      "grad_norm": 4.636005214244119,
+      "learning_rate": 2.9976558324294453e-06,
+      "loss": 0.0409,
+      "step": 1310
+    },
+    {
+      "epoch": 0.04767408263507657,
+      "grad_norm": 2.3192560454039124,
+      "learning_rate": 2.9975567510523007e-06,
+      "loss": 0.0424,
+      "step": 1320
+    },
+    {
+      "epoch": 0.04803524992776654,
+      "grad_norm": 3.365909304241527,
+      "learning_rate": 2.997455620552351e-06,
+      "loss": 0.0402,
+      "step": 1330
+    },
+    {
+      "epoch": 0.04839641722045652,
+      "grad_norm": 1.9971640207765038,
+      "learning_rate": 2.9973524410679727e-06,
+      "loss": 0.042,
+      "step": 1340
+    },
+    {
+      "epoch": 0.04875758451314649,
+      "grad_norm": 16.495246123031166,
+      "learning_rate": 2.9972472127403488e-06,
+      "loss": 0.0406,
+      "step": 1350
+    },
+    {
+      "epoch": 0.04911875180583646,
+      "grad_norm": 2.2051174511826943,
+      "learning_rate": 2.9971399357134632e-06,
+      "loss": 0.041,
+      "step": 1360
+    },
+    {
+      "epoch": 0.049479919098526434,
+      "grad_norm": 2.9204981239827887,
+      "learning_rate": 2.9970306101341054e-06,
+      "loss": 0.0418,
+      "step": 1370
+    },
+    {
+      "epoch": 0.04984108639121641,
+      "grad_norm": 2.1921209294857698,
+      "learning_rate": 2.9969192361518653e-06,
+      "loss": 0.0391,
+      "step": 1380
+    },
+    {
+      "epoch": 0.050202253683906385,
+      "grad_norm": 1.7612561179362265,
+      "learning_rate": 2.9968058139191384e-06,
+      "loss": 0.0395,
+      "step": 1390
+    },
+    {
+      "epoch": 0.05056342097659636,
+      "grad_norm": 3.072867318911964,
+      "learning_rate": 2.9966903435911213e-06,
+      "loss": 0.0395,
+      "step": 1400
+    },
+    {
+      "epoch": 0.050924588269286336,
+      "grad_norm": 3.782356297131538,
+      "learning_rate": 2.9965728253258127e-06,
+      "loss": 0.0376,
+      "step": 1410
+    },
+    {
+      "epoch": 0.05128575556197631,
+      "grad_norm": 3.893556138587198,
+      "learning_rate": 2.996453259284014e-06,
+      "loss": 0.0382,
+      "step": 1420
+    },
+    {
+      "epoch": 0.05164692285466628,
+      "grad_norm": 1.8508286629956445,
+      "learning_rate": 2.9963316456293293e-06,
+      "loss": 0.0381,
+      "step": 1430
+    },
+    {
+      "epoch": 0.05200809014735625,
+      "grad_norm": 1.6496900361058333,
+      "learning_rate": 2.996207984528164e-06,
+      "loss": 0.0378,
+      "step": 1440
+    },
+    {
+      "epoch": 0.05236925744004623,
+      "grad_norm": 2.142318287885142,
+      "learning_rate": 2.9960822761497246e-06,
+      "loss": 0.0393,
+      "step": 1450
+    },
+    {
+      "epoch": 0.0527304247327362,
+      "grad_norm": 1.8299909287540943,
+      "learning_rate": 2.9959545206660186e-06,
+      "loss": 0.0372,
+      "step": 1460
+    },
+    {
+      "epoch": 0.053091592025426175,
+      "grad_norm": 1.5034325758655063,
+      "learning_rate": 2.995824718251856e-06,
+      "loss": 0.0377,
+      "step": 1470
+    },
+    {
+      "epoch": 0.053452759318116154,
+      "grad_norm": 2.614056239339914,
+      "learning_rate": 2.9956928690848467e-06,
+      "loss": 0.04,
+      "step": 1480
+    },
+    {
+      "epoch": 0.053813926610806126,
+      "grad_norm": 3.009970342239538,
+      "learning_rate": 2.9955589733454013e-06,
+      "loss": 0.0384,
+      "step": 1490
+    },
+    {
+      "epoch": 0.0541750939034961,
+      "grad_norm": 2.7584337748577927,
+      "learning_rate": 2.9954230312167303e-06,
+      "loss": 0.0367,
+      "step": 1500
+    },
+    {
+      "epoch": 0.05453626119618607,
+      "grad_norm": 2.0480230261674603,
+      "learning_rate": 2.995285042884845e-06,
+      "loss": 0.0387,
+      "step": 1510
+    },
+    {
+      "epoch": 0.05489742848887605,
+      "grad_norm": 2.6432487555409776,
+      "learning_rate": 2.995145008538557e-06,
+      "loss": 0.0357,
+      "step": 1520
+    },
+    {
+      "epoch": 0.05525859578156602,
+      "grad_norm": 2.0654779056302384,
+      "learning_rate": 2.9950029283694755e-06,
+      "loss": 0.0377,
+      "step": 1530
+    },
+    {
+      "epoch": 0.05561976307425599,
+      "grad_norm": 1.5026108243439562,
+      "learning_rate": 2.9948588025720117e-06,
+      "loss": 0.0375,
+      "step": 1540
+    },
+    {
+      "epoch": 0.05598093036694597,
+      "grad_norm": 1.7922665184670086,
+      "learning_rate": 2.994712631343373e-06,
+      "loss": 0.0393,
+      "step": 1550
+    },
+    {
+      "epoch": 0.056342097659635944,
+      "grad_norm": 1.8260157666473342,
+      "learning_rate": 2.994564414883568e-06,
+      "loss": 0.0366,
+      "step": 1560
+    },
+    {
+      "epoch": 0.056703264952325916,
+      "grad_norm": 1.9852197088096093,
+      "learning_rate": 2.994414153395403e-06,
+      "loss": 0.0364,
+      "step": 1570
+    },
+    {
+      "epoch": 0.057064432245015895,
+      "grad_norm": 2.3328697394427818,
+      "learning_rate": 2.9942618470844814e-06,
+      "loss": 0.0343,
+      "step": 1580
+    },
+    {
+      "epoch": 0.05742559953770587,
+      "grad_norm": 1.4260001952146015,
+      "learning_rate": 2.9941074961592065e-06,
+      "loss": 0.0356,
+      "step": 1590
+    },
+    {
+      "epoch": 0.05778676683039584,
+      "grad_norm": 2.2159261913608996,
+      "learning_rate": 2.993951100830778e-06,
+      "loss": 0.0354,
+      "step": 1600
+    },
+    {
+      "epoch": 0.05814793412308581,
+      "grad_norm": 1.5660141161713501,
+      "learning_rate": 2.9937926613131934e-06,
+      "loss": 0.0348,
+      "step": 1610
+    },
+    {
+      "epoch": 0.05850910141577579,
+      "grad_norm": 2.0434678368844743,
+      "learning_rate": 2.993632177823247e-06,
+      "loss": 0.0358,
+      "step": 1620
+    },
+    {
+      "epoch": 0.05887026870846576,
+      "grad_norm": 1.8037819407293025,
+      "learning_rate": 2.9934696505805297e-06,
+      "loss": 0.0349,
+      "step": 1630
+    },
+    {
+      "epoch": 0.059231436001155734,
+      "grad_norm": 1.4390947903338005,
+      "learning_rate": 2.99330507980743e-06,
+      "loss": 0.0346,
+      "step": 1640
+    },
+    {
+      "epoch": 0.05959260329384571,
+      "grad_norm": 2.217174079747424,
+      "learning_rate": 2.993138465729132e-06,
+      "loss": 0.035,
+      "step": 1650
+    },
+    {
+      "epoch": 0.059953770586535685,
+      "grad_norm": 2.135695550889573,
+      "learning_rate": 2.9929698085736147e-06,
+      "loss": 0.0366,
+      "step": 1660
+    },
+    {
+      "epoch": 0.06031493787922566,
+      "grad_norm": 2.744830219350928,
+      "learning_rate": 2.9927991085716533e-06,
+      "loss": 0.0353,
+      "step": 1670
+    },
+    {
+      "epoch": 0.06067610517191563,
+      "grad_norm": 1.9775441229070698,
+      "learning_rate": 2.992626365956819e-06,
+      "loss": 0.0345,
+      "step": 1680
+    },
+    {
+      "epoch": 0.06103727246460561,
+      "grad_norm": 9.124528104080278,
+      "learning_rate": 2.9924515809654776e-06,
+      "loss": 0.0351,
+      "step": 1690
+    },
+    {
+      "epoch": 0.06139843975729558,
+      "grad_norm": 3.5441594688042968,
+      "learning_rate": 2.992274753836789e-06,
+      "loss": 0.0324,
+      "step": 1700
+    },
+    {
+      "epoch": 0.06175960704998555,
+      "grad_norm": 1.5571670067399437,
+      "learning_rate": 2.9920958848127074e-06,
+      "loss": 0.0354,
+      "step": 1710
+    },
+    {
+      "epoch": 0.06212077434267553,
+      "grad_norm": 1.6849743442068315,
+      "learning_rate": 2.9919149741379812e-06,
+      "loss": 0.0347,
+      "step": 1720
+    },
+    {
+      "epoch": 0.0624819416353655,
+      "grad_norm": 2.2217600942055102,
+      "learning_rate": 2.991732022060153e-06,
+      "loss": 0.0342,
+      "step": 1730
+    },
+    {
+      "epoch": 0.06284310892805547,
+      "grad_norm": 1.5986148822758528,
+      "learning_rate": 2.991547028829557e-06,
+      "loss": 0.0337,
+      "step": 1740
+    },
+    {
+      "epoch": 0.06320427622074545,
+      "grad_norm": 1.8949462290838979,
+      "learning_rate": 2.991359994699322e-06,
+      "loss": 0.0342,
+      "step": 1750
+    },
+    {
+      "epoch": 0.06356544351343542,
+      "grad_norm": 1.4388928229927256,
+      "learning_rate": 2.9911709199253694e-06,
+      "loss": 0.0345,
+      "step": 1760
+    },
+    {
+      "epoch": 0.0639266108061254,
+      "grad_norm": 2.12566368582026,
+      "learning_rate": 2.9909798047664118e-06,
+      "loss": 0.0338,
+      "step": 1770
+    },
+    {
+      "epoch": 0.06428777809881538,
+      "grad_norm": 1.7834682248432385,
+      "learning_rate": 2.990786649483953e-06,
+      "loss": 0.0332,
+      "step": 1780
+    },
+    {
+      "epoch": 0.06464894539150534,
+      "grad_norm": 1.3607425776754558,
+      "learning_rate": 2.990591454342291e-06,
+      "loss": 0.032,
+      "step": 1790
+    },
+    {
+      "epoch": 0.06501011268419532,
+      "grad_norm": 1.726520142227843,
+      "learning_rate": 2.990394219608513e-06,
+      "loss": 0.0332,
+      "step": 1800
+    },
+    {
+      "epoch": 0.0653712799768853,
+      "grad_norm": 1.4720293028121976,
+      "learning_rate": 2.9901949455524967e-06,
+      "loss": 0.0341,
+      "step": 1810
+    },
+    {
+      "epoch": 0.06573244726957526,
+      "grad_norm": 1.7394556004230972,
+      "learning_rate": 2.989993632446911e-06,
+      "loss": 0.0319,
+      "step": 1820
+    },
+    {
+      "epoch": 0.06609361456226524,
+      "grad_norm": 2.0702298368460403,
+      "learning_rate": 2.9897902805672152e-06,
+      "loss": 0.0349,
+      "step": 1830
+    },
+    {
+      "epoch": 0.06645478185495522,
+      "grad_norm": 2.246080367853108,
+      "learning_rate": 2.989584890191658e-06,
+      "loss": 0.0317,
+      "step": 1840
+    },
+    {
+      "epoch": 0.06681594914764519,
+      "grad_norm": 2.8804172742658274,
+      "learning_rate": 2.9893774616012773e-06,
+      "loss": 0.0315,
+      "step": 1850
+    },
+    {
+      "epoch": 0.06717711644033517,
+      "grad_norm": 2.179650755545797,
+      "learning_rate": 2.989167995079899e-06,
+      "loss": 0.0316,
+      "step": 1860
+    },
+    {
+      "epoch": 0.06753828373302513,
+      "grad_norm": 1.8124034118565888,
+      "learning_rate": 2.988956490914139e-06,
+      "loss": 0.0321,
+      "step": 1870
+    },
+    {
+      "epoch": 0.06789945102571511,
+      "grad_norm": 1.4863247114854377,
+      "learning_rate": 2.9887429493934003e-06,
+      "loss": 0.0325,
+      "step": 1880
+    },
+    {
+      "epoch": 0.06826061831840509,
+      "grad_norm": 2.697009246253888,
+      "learning_rate": 2.9885273708098747e-06,
+      "loss": 0.0346,
+      "step": 1890
+    },
+    {
+      "epoch": 0.06862178561109505,
+      "grad_norm": 2.0705367030610304,
+      "learning_rate": 2.98830975545854e-06,
+      "loss": 0.0321,
+      "step": 1900
+    },
+    {
+      "epoch": 0.06898295290378503,
+      "grad_norm": 1.8724177371558055,
+      "learning_rate": 2.9880901036371623e-06,
+      "loss": 0.0333,
+      "step": 1910
+    },
+    {
+      "epoch": 0.06934412019647501,
+      "grad_norm": 1.7130216915936365,
+      "learning_rate": 2.987868415646292e-06,
+      "loss": 0.0309,
+      "step": 1920
+    },
+    {
+      "epoch": 0.06970528748916498,
+      "grad_norm": 1.7772708376791948,
+      "learning_rate": 2.9876446917892695e-06,
+      "loss": 0.0315,
+      "step": 1930
+    },
+    {
+      "epoch": 0.07006645478185496,
+      "grad_norm": 1.6074665224849827,
+      "learning_rate": 2.987418932372217e-06,
+      "loss": 0.0322,
+      "step": 1940
+    },
+    {
+      "epoch": 0.07042762207454493,
+      "grad_norm": 1.4154770553531868,
+      "learning_rate": 2.987191137704043e-06,
+      "loss": 0.0312,
+      "step": 1950
+    },
+    {
+      "epoch": 0.0707887893672349,
+      "grad_norm": 1.456821915900298,
+      "learning_rate": 2.9869613080964425e-06,
+      "loss": 0.0316,
+      "step": 1960
+    },
+    {
+      "epoch": 0.07114995665992488,
+      "grad_norm": 1.5857516199911676,
+      "learning_rate": 2.986729443863893e-06,
+      "loss": 0.0301,
+      "step": 1970
+    },
+    {
+      "epoch": 0.07151112395261486,
+      "grad_norm": 1.5373732162487916,
+      "learning_rate": 2.986495545323658e-06,
+      "loss": 0.0317,
+      "step": 1980
+    },
+    {
+      "epoch": 0.07187229124530482,
+      "grad_norm": 3.041721365377869,
+      "learning_rate": 2.986259612795782e-06,
+      "loss": 0.0303,
+      "step": 1990
+    },
+    {
+      "epoch": 0.0722334585379948,
+      "grad_norm": 1.5667474537449642,
+      "learning_rate": 2.986021646603095e-06,
+      "loss": 0.0313,
+      "step": 2000
+    },
+    {
+      "epoch": 0.0722334585379948,
+      "eval_loss": 0.17110520601272583,
+      "eval_runtime": 37.9026,
+      "eval_samples_per_second": 108.066,
+      "eval_steps_per_second": 0.211,
+      "step": 2000
+    },
+    {
+      "epoch": 0.07259462583068477,
+      "grad_norm": 1.7752266494164892,
+      "learning_rate": 2.9857816470712074e-06,
+      "loss": 0.032,
+      "step": 2010
+    },
+    {
+      "epoch": 0.07295579312337475,
+      "grad_norm": 2.9756420458628225,
+      "learning_rate": 2.9855396145285146e-06,
+      "loss": 0.0313,
+      "step": 2020
+    },
+    {
+      "epoch": 0.07331696041606472,
+      "grad_norm": 1.2861222785183546,
+      "learning_rate": 2.9852955493061917e-06,
+      "loss": 0.0309,
+      "step": 2030
+    },
+    {
+      "epoch": 0.07367812770875469,
+      "grad_norm": 1.670601539212067,
+      "learning_rate": 2.9850494517381953e-06,
+      "loss": 0.0326,
+      "step": 2040
+    },
+    {
+      "epoch": 0.07403929500144467,
+      "grad_norm": 3.0325079092660645,
+      "learning_rate": 2.9848013221612638e-06,
+      "loss": 0.0291,
+      "step": 2050
+    },
+    {
+      "epoch": 0.07440046229413465,
+      "grad_norm": 1.7385676497610276,
+      "learning_rate": 2.984551160914915e-06,
+      "loss": 0.0304,
+      "step": 2060
+    },
+    {
+      "epoch": 0.07476162958682461,
+      "grad_norm": 1.7123931333897295,
+      "learning_rate": 2.984298968341448e-06,
+      "loss": 0.0309,
+      "step": 2070
+    },
+    {
+      "epoch": 0.07512279687951459,
+      "grad_norm": 2.2463806080309694,
+      "learning_rate": 2.98404474478594e-06,
+      "loss": 0.032,
+      "step": 2080
+    },
+    {
+      "epoch": 0.07548396417220457,
+      "grad_norm": 2.1525240532607297,
+      "learning_rate": 2.9837884905962478e-06,
+      "loss": 0.0312,
+      "step": 2090
+    },
+    {
+      "epoch": 0.07584513146489454,
+      "grad_norm": 1.3538340809196117,
+      "learning_rate": 2.983530206123007e-06,
+      "loss": 0.0318,
+      "step": 2100
+    },
+    {
+      "epoch": 0.07620629875758451,
+      "grad_norm": 1.3602852877585043,
+      "learning_rate": 2.983269891719631e-06,
+      "loss": 0.0314,
+      "step": 2110
+    },
+    {
+      "epoch": 0.0765674660502745,
+      "grad_norm": 1.456261243201536,
+      "learning_rate": 2.9830075477423105e-06,
+      "loss": 0.0314,
+      "step": 2120
+    },
+    {
+      "epoch": 0.07692863334296446,
+      "grad_norm": 1.1011740359098003,
+      "learning_rate": 2.9827431745500136e-06,
+      "loss": 0.0294,
+      "step": 2130
+    },
+    {
+      "epoch": 0.07728980063565444,
+      "grad_norm": 1.5188476462216016,
+      "learning_rate": 2.9824767725044855e-06,
+      "loss": 0.0311,
+      "step": 2140
+    },
+    {
+      "epoch": 0.0776509679283444,
+      "grad_norm": 2.1117026048797496,
+      "learning_rate": 2.982208341970246e-06,
+      "loss": 0.0319,
+      "step": 2150
+    },
+    {
+      "epoch": 0.07801213522103438,
+      "grad_norm": 1.1548615366758432,
+      "learning_rate": 2.981937883314592e-06,
+      "loss": 0.0304,
+      "step": 2160
+    },
+    {
+      "epoch": 0.07837330251372436,
+      "grad_norm": 1.9621687586252696,
+      "learning_rate": 2.981665396907595e-06,
+      "loss": 0.0283,
+      "step": 2170
+    },
+    {
+      "epoch": 0.07873446980641433,
+      "grad_norm": 1.7656427213380914,
+      "learning_rate": 2.9813908831221007e-06,
+      "loss": 0.0292,
+      "step": 2180
+    },
+    {
+      "epoch": 0.0790956370991043,
+      "grad_norm": 1.357845299375958,
+      "learning_rate": 2.9811143423337296e-06,
+      "loss": 0.0291,
+      "step": 2190
+    },
+    {
+      "epoch": 0.07945680439179428,
+      "grad_norm": 1.9521671401881249,
+      "learning_rate": 2.980835774920875e-06,
+      "loss": 0.0305,
+      "step": 2200
+    },
+    {
+      "epoch": 0.07981797168448425,
+      "grad_norm": 1.5441066170116196,
+      "learning_rate": 2.9805551812647044e-06,
+      "loss": 0.0303,
+      "step": 2210
+    },
+    {
+      "epoch": 0.08017913897717423,
+      "grad_norm": 1.4585903363920794,
+      "learning_rate": 2.980272561749156e-06,
+      "loss": 0.0296,
+      "step": 2220
+    },
+    {
+      "epoch": 0.0805403062698642,
+      "grad_norm": 1.7713146434669313,
+      "learning_rate": 2.9799879167609423e-06,
+      "loss": 0.0293,
+      "step": 2230
+    },
+    {
+      "epoch": 0.08090147356255417,
+      "grad_norm": 1.3107079017886312,
+      "learning_rate": 2.979701246689545e-06,
+      "loss": 0.0304,
+      "step": 2240
+    },
+    {
+      "epoch": 0.08126264085524415,
+      "grad_norm": 1.2536988905395532,
+      "learning_rate": 2.9794125519272183e-06,
+      "loss": 0.0289,
+      "step": 2250
+    },
+    {
+      "epoch": 0.08162380814793413,
+      "grad_norm": 1.3500615866960406,
+      "learning_rate": 2.979121832868986e-06,
+      "loss": 0.0296,
+      "step": 2260
+    },
+    {
+      "epoch": 0.0819849754406241,
+      "grad_norm": 2.124625421233279,
+      "learning_rate": 2.9788290899126426e-06,
+      "loss": 0.0288,
+      "step": 2270
+    },
+    {
+      "epoch": 0.08234614273331407,
+      "grad_norm": 1.7374171746229727,
+      "learning_rate": 2.978534323458751e-06,
+      "loss": 0.0286,
+      "step": 2280
+    },
+    {
+      "epoch": 0.08270731002600404,
+      "grad_norm": 1.483711353818059,
+      "learning_rate": 2.9782375339106432e-06,
+      "loss": 0.0293,
+      "step": 2290
+    },
+    {
+      "epoch": 0.08306847731869402,
+      "grad_norm": 1.4661179141232732,
+      "learning_rate": 2.97793872167442e-06,
+      "loss": 0.0282,
+      "step": 2300
+    },
+    {
+      "epoch": 0.083429644611384,
+      "grad_norm": 1.0685992271375813,
+      "learning_rate": 2.977637887158949e-06,
+      "loss": 0.0269,
+      "step": 2310
+    },
+    {
+      "epoch": 0.08379081190407396,
+      "grad_norm": 1.4166350792931108,
+      "learning_rate": 2.9773350307758653e-06,
+      "loss": 0.0289,
+      "step": 2320
+    },
+    {
+      "epoch": 0.08415197919676394,
+      "grad_norm": 1.1709675417715821,
+      "learning_rate": 2.97703015293957e-06,
+      "loss": 0.0291,
+      "step": 2330
+    },
+    {
+      "epoch": 0.08451314648945392,
+      "grad_norm": 1.5170128486029664,
+      "learning_rate": 2.9767232540672316e-06,
+      "loss": 0.0282,
+      "step": 2340
+    },
+    {
+      "epoch": 0.08487431378214388,
+      "grad_norm": 1.146666847124037,
+      "learning_rate": 2.976414334578783e-06,
+      "loss": 0.0292,
+      "step": 2350
+    },
+    {
+      "epoch": 0.08523548107483386,
+      "grad_norm": 1.55365896069812,
+      "learning_rate": 2.976103394896921e-06,
+      "loss": 0.0278,
+      "step": 2360
+    },
+    {
+      "epoch": 0.08559664836752384,
+      "grad_norm": 1.795971150763761,
+      "learning_rate": 2.9757904354471085e-06,
+      "loss": 0.0272,
+      "step": 2370
+    },
+    {
+      "epoch": 0.08595781566021381,
+      "grad_norm": 1.818444205839746,
+      "learning_rate": 2.975475456657571e-06,
+      "loss": 0.0297,
+      "step": 2380
+    },
+    {
+      "epoch": 0.08631898295290379,
+      "grad_norm": 1.0588246851496752,
+      "learning_rate": 2.9751584589592976e-06,
+      "loss": 0.027,
+      "step": 2390
+    },
+    {
+      "epoch": 0.08668015024559377,
+      "grad_norm": 1.1079913040036007,
+      "learning_rate": 2.974839442786039e-06,
+      "loss": 0.0273,
+      "step": 2400
+    },
+    {
+      "epoch": 0.08704131753828373,
+      "grad_norm": 2.590278769693884,
+      "learning_rate": 2.974518408574309e-06,
+      "loss": 0.0268,
+      "step": 2410
+    },
+    {
+      "epoch": 0.08740248483097371,
+      "grad_norm": 1.5898006212627334,
+      "learning_rate": 2.9741953567633824e-06,
+      "loss": 0.0293,
+      "step": 2420
+    },
+    {
+      "epoch": 0.08776365212366369,
+      "grad_norm": 1.2719550740701915,
+      "learning_rate": 2.9738702877952936e-06,
+      "loss": 0.0281,
+      "step": 2430
+    },
+    {
+      "epoch": 0.08812481941635365,
+      "grad_norm": 1.198345974248335,
+      "learning_rate": 2.9735432021148384e-06,
+      "loss": 0.0284,
+      "step": 2440
+    },
+    {
+      "epoch": 0.08848598670904363,
+      "grad_norm": 1.2308990441362293,
+      "learning_rate": 2.9732141001695714e-06,
+      "loss": 0.0277,
+      "step": 2450
+    },
+    {
+      "epoch": 0.0888471540017336,
+      "grad_norm": 1.1534593975304672,
+      "learning_rate": 2.9728829824098067e-06,
+      "loss": 0.0279,
+      "step": 2460
+    },
+    {
+      "epoch": 0.08920832129442358,
+      "grad_norm": 1.1948268496517236,
+      "learning_rate": 2.9725498492886153e-06,
+      "loss": 0.0268,
+      "step": 2470
+    },
+    {
+      "epoch": 0.08956948858711355,
+      "grad_norm": 1.2011378953178011,
+      "learning_rate": 2.972214701261828e-06,
+      "loss": 0.027,
+      "step": 2480
+    },
+    {
+      "epoch": 0.08993065587980352,
+      "grad_norm": 1.2324335223571459,
+      "learning_rate": 2.9718775387880307e-06,
+      "loss": 0.0277,
+      "step": 2490
+    },
+    {
+      "epoch": 0.0902918231724935,
+      "grad_norm": 1.3574951820515897,
+      "learning_rate": 2.9715383623285665e-06,
+      "loss": 0.0278,
+      "step": 2500
+    },
+    {
+      "epoch": 0.09065299046518348,
+      "grad_norm": 1.4702085737447135,
+      "learning_rate": 2.9711971723475346e-06,
+      "loss": 0.0285,
+      "step": 2510
+    },
+    {
+      "epoch": 0.09101415775787344,
+      "grad_norm": 1.3815043632593236,
+      "learning_rate": 2.9708539693117877e-06,
+      "loss": 0.0287,
+      "step": 2520
+    },
+    {
+      "epoch": 0.09137532505056342,
+      "grad_norm": 1.600383219774193,
+      "learning_rate": 2.9705087536909354e-06,
+      "loss": 0.0282,
+      "step": 2530
+    },
+    {
+      "epoch": 0.0917364923432534,
+      "grad_norm": 1.115657931341337,
+      "learning_rate": 2.970161525957339e-06,
+      "loss": 0.0266,
+      "step": 2540
+    },
+    {
+      "epoch": 0.09209765963594337,
+      "grad_norm": 1.0918117282070394,
+      "learning_rate": 2.9698122865861144e-06,
+      "loss": 0.0286,
+      "step": 2550
+    },
+    {
+      "epoch": 0.09245882692863334,
+      "grad_norm": 1.2391164973337567,
+      "learning_rate": 2.96946103605513e-06,
+      "loss": 0.0272,
+      "step": 2560
+    },
+    {
+      "epoch": 0.09281999422132332,
+      "grad_norm": 1.1293468553986536,
+      "learning_rate": 2.9691077748450044e-06,
+      "loss": 0.0267,
+      "step": 2570
+    },
+    {
+      "epoch": 0.09318116151401329,
+      "grad_norm": 1.390932009459183,
+      "learning_rate": 2.96875250343911e-06,
+      "loss": 0.0268,
+      "step": 2580
+    },
+    {
+      "epoch": 0.09354232880670327,
+      "grad_norm": 1.300315762070993,
+      "learning_rate": 2.968395222323567e-06,
+      "loss": 0.027,
+      "step": 2590
+    },
+    {
+      "epoch": 0.09390349609939323,
+      "grad_norm": 1.08152141569384,
+      "learning_rate": 2.968035931987248e-06,
+      "loss": 0.0272,
+      "step": 2600
+    },
+    {
+      "epoch": 0.09426466339208321,
+      "grad_norm": 1.1627728171553298,
+      "learning_rate": 2.9676746329217737e-06,
+      "loss": 0.027,
+      "step": 2610
+    },
+    {
+      "epoch": 0.09462583068477319,
+      "grad_norm": 1.7607561386804633,
+      "learning_rate": 2.9673113256215132e-06,
+      "loss": 0.0249,
+      "step": 2620
+    },
+    {
+      "epoch": 0.09498699797746316,
+      "grad_norm": 2.4613129761406305,
+      "learning_rate": 2.966946010583584e-06,
+      "loss": 0.0293,
+      "step": 2630
+    },
+    {
+      "epoch": 0.09534816527015313,
+      "grad_norm": 2.2508692293680843,
+      "learning_rate": 2.96657868830785e-06,
+      "loss": 0.0268,
+      "step": 2640
+    },
+    {
+      "epoch": 0.09570933256284311,
+      "grad_norm": 1.4356693884680782,
+      "learning_rate": 2.9662093592969226e-06,
+      "loss": 0.0269,
+      "step": 2650
+    },
+    {
+      "epoch": 0.09607049985553308,
+      "grad_norm": 1.468597854739315,
+      "learning_rate": 2.9658380240561586e-06,
+      "loss": 0.0267,
+      "step": 2660
+    },
+    {
+      "epoch": 0.09643166714822306,
+      "grad_norm": 1.2125997315621115,
+      "learning_rate": 2.96546468309366e-06,
+      "loss": 0.0268,
+      "step": 2670
+    },
+    {
+      "epoch": 0.09679283444091304,
+      "grad_norm": 1.7059066242194365,
+      "learning_rate": 2.9650893369202726e-06,
+      "loss": 0.0278,
+      "step": 2680
+    },
+    {
+      "epoch": 0.097154001733603,
+      "grad_norm": 1.2631228772547884,
+      "learning_rate": 2.9647119860495877e-06,
+      "loss": 0.0279,
+      "step": 2690
+    },
+    {
+      "epoch": 0.09751516902629298,
+      "grad_norm": 1.0222966758643148,
+      "learning_rate": 2.964332630997937e-06,
+      "loss": 0.0265,
+      "step": 2700
+    },
+    {
+      "epoch": 0.09787633631898296,
+      "grad_norm": 1.2027953510282923,
+      "learning_rate": 2.9639512722843974e-06,
+      "loss": 0.0261,
+      "step": 2710
+    },
+    {
+      "epoch": 0.09823750361167292,
+      "grad_norm": 2.2896362012402673,
+      "learning_rate": 2.9635679104307854e-06,
+      "loss": 0.0255,
+      "step": 2720
+    },
+    {
+      "epoch": 0.0985986709043629,
+      "grad_norm": 3.6987256104659596,
+      "learning_rate": 2.9631825459616593e-06,
+      "loss": 0.0264,
+      "step": 2730
+    },
+    {
+      "epoch": 0.09895983819705287,
+      "grad_norm": 1.1372477868488595,
+      "learning_rate": 2.9627951794043174e-06,
+      "loss": 0.0252,
+      "step": 2740
+    },
+    {
+      "epoch": 0.09932100548974285,
+      "grad_norm": 1.2524505352787907,
+      "learning_rate": 2.9624058112887974e-06,
+      "loss": 0.0267,
+      "step": 2750
+    },
+    {
+      "epoch": 0.09968217278243283,
+      "grad_norm": 1.5458537339814469,
+      "learning_rate": 2.9620144421478757e-06,
+      "loss": 0.0277,
+      "step": 2760
+    },
+    {
+      "epoch": 0.10004334007512279,
+      "grad_norm": 0.9685572920431231,
+      "learning_rate": 2.9616210725170674e-06,
+      "loss": 0.0255,
+      "step": 2770
+    },
+    {
+      "epoch": 0.10040450736781277,
+      "grad_norm": 1.3340612419055051,
+      "learning_rate": 2.9612257029346238e-06,
+      "loss": 0.0256,
+      "step": 2780
+    },
+    {
+      "epoch": 0.10076567466050275,
+      "grad_norm": 1.2157576974226674,
+      "learning_rate": 2.960828333941533e-06,
+      "loss": 0.0269,
+      "step": 2790
+    },
+    {
+      "epoch": 0.10112684195319271,
+      "grad_norm": 1.608284889802177,
+      "learning_rate": 2.9604289660815196e-06,
+      "loss": 0.0251,
+      "step": 2800
+    },
+    {
+      "epoch": 0.1014880092458827,
+      "grad_norm": 1.2368396199936416,
+      "learning_rate": 2.960027599901044e-06,
+      "loss": 0.0261,
+      "step": 2810
+    },
+    {
+      "epoch": 0.10184917653857267,
+      "grad_norm": 1.1665385231852943,
+      "learning_rate": 2.9596242359492977e-06,
+      "loss": 0.0253,
+      "step": 2820
+    },
+    {
+      "epoch": 0.10221034383126264,
+      "grad_norm": 1.1647186290579725,
+      "learning_rate": 2.9592188747782094e-06,
+      "loss": 0.0251,
+      "step": 2830
+    },
+    {
+      "epoch": 0.10257151112395262,
+      "grad_norm": 1.1691076164147158,
+      "learning_rate": 2.9588115169424383e-06,
+      "loss": 0.025,
+      "step": 2840
+    },
+    {
+      "epoch": 0.1029326784166426,
+      "grad_norm": 2.740498684644438,
+      "learning_rate": 2.958402162999377e-06,
+      "loss": 0.025,
+      "step": 2850
+    },
+    {
+      "epoch": 0.10329384570933256,
+      "grad_norm": 1.3997384148578422,
+      "learning_rate": 2.957990813509149e-06,
+      "loss": 0.0271,
+      "step": 2860
+    },
+    {
+      "epoch": 0.10365501300202254,
+      "grad_norm": 2.081911624846036,
+      "learning_rate": 2.9575774690346066e-06,
+      "loss": 0.0256,
+      "step": 2870
+    },
+    {
+      "epoch": 0.1040161802947125,
+      "grad_norm": 1.155059916149935,
+      "learning_rate": 2.957162130141336e-06,
+      "loss": 0.0252,
+      "step": 2880
+    },
+    {
+      "epoch": 0.10437734758740248,
+      "grad_norm": 1.50789020771039,
+      "learning_rate": 2.9567447973976474e-06,
+      "loss": 0.026,
+      "step": 2890
+    },
+    {
+      "epoch": 0.10473851488009246,
+      "grad_norm": 1.096624520751321,
+      "learning_rate": 2.956325471374583e-06,
+      "loss": 0.0248,
+      "step": 2900
+    },
+    {
+      "epoch": 0.10509968217278243,
+      "grad_norm": 1.5388202225904897,
+      "learning_rate": 2.955904152645911e-06,
+      "loss": 0.026,
+      "step": 2910
+    },
+    {
+      "epoch": 0.1054608494654724,
+      "grad_norm": 1.3929214053191044,
+      "learning_rate": 2.9554808417881262e-06,
+      "loss": 0.0254,
+      "step": 2920
+    },
+    {
+      "epoch": 0.10582201675816238,
+      "grad_norm": 1.2101183492293703,
+      "learning_rate": 2.9550555393804483e-06,
+      "loss": 0.025,
+      "step": 2930
+    },
+    {
+      "epoch": 0.10618318405085235,
+      "grad_norm": 1.1324372937351337,
+      "learning_rate": 2.954628246004825e-06,
+      "loss": 0.0257,
+      "step": 2940
+    },
+    {
+      "epoch": 0.10654435134354233,
+      "grad_norm": 1.1723225252194924,
+      "learning_rate": 2.9541989622459244e-06,
+      "loss": 0.0268,
+      "step": 2950
+    },
+    {
+      "epoch": 0.10690551863623231,
+      "grad_norm": 1.121118272190178,
+      "learning_rate": 2.9537676886911413e-06,
+      "loss": 0.024,
+      "step": 2960
+    },
+    {
+      "epoch": 0.10726668592892227,
+      "grad_norm": 0.9671994789081405,
+      "learning_rate": 2.9533344259305915e-06,
+      "loss": 0.0263,
+      "step": 2970
+    },
+    {
+      "epoch": 0.10762785322161225,
+      "grad_norm": 1.0036426808611114,
+      "learning_rate": 2.9528991745571134e-06,
+      "loss": 0.0251,
+      "step": 2980
+    },
+    {
+      "epoch": 0.10798902051430223,
+      "grad_norm": 1.4554253136562012,
+      "learning_rate": 2.952461935166265e-06,
+      "loss": 0.0249,
+      "step": 2990
+    },
+    {
+      "epoch": 0.1083501878069922,
+      "grad_norm": 1.0906119953275397,
+      "learning_rate": 2.9520227083563273e-06,
+      "loss": 0.0259,
+      "step": 3000
+    },
+    {
+      "epoch": 0.1083501878069922,
+      "eval_loss": 0.16691112518310547,
+      "eval_runtime": 38.9846,
+      "eval_samples_per_second": 105.067,
+      "eval_steps_per_second": 0.205,
+      "step": 3000
+    },
+    {
+      "epoch": 0.10871135509968217,
+      "grad_norm": 1.2489725463980832,
+      "learning_rate": 2.9515814947282974e-06,
+      "loss": 0.026,
+      "step": 3010
+    },
+    {
+      "epoch": 0.10907252239237214,
+      "grad_norm": 1.0612786957693097,
+      "learning_rate": 2.9511382948858933e-06,
+      "loss": 0.0258,
+      "step": 3020
+    },
+    {
+      "epoch": 0.10943368968506212,
+      "grad_norm": 1.4330317683303864,
+      "learning_rate": 2.95069310943555e-06,
+      "loss": 0.0248,
+      "step": 3030
+    },
+    {
+      "epoch": 0.1097948569777521,
+      "grad_norm": 1.1441534762493286,
+      "learning_rate": 2.95024593898642e-06,
+      "loss": 0.0243,
+      "step": 3040
+    },
+    {
+      "epoch": 0.11015602427044206,
+      "grad_norm": 1.3473539880379115,
+      "learning_rate": 2.9497967841503703e-06,
+      "loss": 0.0222,
+      "step": 3050
+    },
+    {
+      "epoch": 0.11051719156313204,
+      "grad_norm": 1.0154564520130183,
+      "learning_rate": 2.9493456455419848e-06,
+      "loss": 0.0261,
+      "step": 3060
+    },
+    {
+      "epoch": 0.11087835885582202,
+      "grad_norm": 1.281689539513682,
+      "learning_rate": 2.9488925237785613e-06,
+      "loss": 0.0249,
+      "step": 3070
+    },
+    {
+      "epoch": 0.11123952614851199,
+      "grad_norm": 1.1483257658607477,
+      "learning_rate": 2.948437419480111e-06,
+      "loss": 0.024,
+      "step": 3080
+    },
+    {
+      "epoch": 0.11160069344120196,
+      "grad_norm": 1.2353600840898928,
+      "learning_rate": 2.9479803332693583e-06,
+      "loss": 0.0238,
+      "step": 3090
+    },
+    {
+      "epoch": 0.11196186073389194,
+      "grad_norm": 2.356853655078678,
+      "learning_rate": 2.9475212657717386e-06,
+      "loss": 0.0242,
+      "step": 3100
+    },
+    {
+      "epoch": 0.11232302802658191,
+      "grad_norm": 1.5559978484022756,
+      "learning_rate": 2.9470602176153988e-06,
+      "loss": 0.0257,
+      "step": 3110
+    },
+    {
+      "epoch": 0.11268419531927189,
+      "grad_norm": 0.9927314908980774,
+      "learning_rate": 2.9465971894311967e-06,
+      "loss": 0.0246,
+      "step": 3120
+    },
+    {
+      "epoch": 0.11304536261196187,
+      "grad_norm": 0.9902341363006447,
+      "learning_rate": 2.946132181852698e-06,
+      "loss": 0.0238,
+      "step": 3130
+    },
+    {
+      "epoch": 0.11340652990465183,
+      "grad_norm": 1.1540415264456618,
+      "learning_rate": 2.9456651955161777e-06,
+      "loss": 0.0226,
+      "step": 3140
+    },
+    {
+      "epoch": 0.11376769719734181,
+      "grad_norm": 1.121846062432073,
+      "learning_rate": 2.945196231060618e-06,
+      "loss": 0.0246,
+      "step": 3150
+    },
+    {
+      "epoch": 0.11412886449003179,
+      "grad_norm": 0.8805464236352655,
+      "learning_rate": 2.944725289127708e-06,
+      "loss": 0.0233,
+      "step": 3160
+    },
+    {
+      "epoch": 0.11449003178272175,
+      "grad_norm": 0.8374471230422528,
+      "learning_rate": 2.9442523703618433e-06,
+      "loss": 0.0252,
+      "step": 3170
+    },
+    {
+      "epoch": 0.11485119907541173,
+      "grad_norm": 1.1477807929035493,
+      "learning_rate": 2.943777475410123e-06,
+      "loss": 0.0234,
+      "step": 3180
+    },
+    {
+      "epoch": 0.1152123663681017,
+      "grad_norm": 1.0341394484605884,
+      "learning_rate": 2.9433006049223506e-06,
+      "loss": 0.0234,
+      "step": 3190
+    },
+    {
+      "epoch": 0.11557353366079168,
+      "grad_norm": 1.0792842110990093,
+      "learning_rate": 2.9428217595510333e-06,
+      "loss": 0.0245,
+      "step": 3200
+    },
+    {
+      "epoch": 0.11593470095348166,
+      "grad_norm": 1.3107374510313599,
+      "learning_rate": 2.9423409399513808e-06,
+      "loss": 0.0252,
+      "step": 3210
+    },
+    {
+      "epoch": 0.11629586824617162,
+      "grad_norm": 1.0024381999846992,
+      "learning_rate": 2.941858146781303e-06,
+      "loss": 0.0233,
+      "step": 3220
+    },
+    {
+      "epoch": 0.1166570355388616,
+      "grad_norm": 1.0894409503989062,
+      "learning_rate": 2.941373380701411e-06,
+      "loss": 0.025,
+      "step": 3230
+    },
+    {
+      "epoch": 0.11701820283155158,
+      "grad_norm": 1.2533442390035525,
+      "learning_rate": 2.9408866423750163e-06,
+      "loss": 0.0241,
+      "step": 3240
+    },
+    {
+      "epoch": 0.11737937012424154,
+      "grad_norm": 1.0001851141318647,
+      "learning_rate": 2.940397932468127e-06,
+      "loss": 0.0239,
+      "step": 3250
+    },
+    {
+      "epoch": 0.11774053741693152,
+      "grad_norm": 0.9522643669682826,
+      "learning_rate": 2.9399072516494504e-06,
+      "loss": 0.0238,
+      "step": 3260
+    },
+    {
+      "epoch": 0.1181017047096215,
+      "grad_norm": 1.0729448450117833,
+      "learning_rate": 2.9394146005903902e-06,
+      "loss": 0.0228,
+      "step": 3270
+    },
+    {
+      "epoch": 0.11846287200231147,
+      "grad_norm": 1.0895083668150138,
+      "learning_rate": 2.9389199799650463e-06,
+      "loss": 0.0252,
+      "step": 3280
+    },
+    {
+      "epoch": 0.11882403929500145,
+      "grad_norm": 1.0152855618243894,
+      "learning_rate": 2.9384233904502133e-06,
+      "loss": 0.0244,
+      "step": 3290
+    },
+    {
+      "epoch": 0.11918520658769143,
+      "grad_norm": 1.0640907086273215,
+      "learning_rate": 2.9379248327253797e-06,
+      "loss": 0.0227,
+      "step": 3300
+    },
+    {
+      "epoch": 0.11954637388038139,
+      "grad_norm": 1.1753102566216322,
+      "learning_rate": 2.9374243074727276e-06,
+      "loss": 0.0239,
+      "step": 3310
+    },
+    {
+      "epoch": 0.11990754117307137,
+      "grad_norm": 0.7327412933269778,
+      "learning_rate": 2.936921815377131e-06,
+      "loss": 0.0237,
+      "step": 3320
+    },
+    {
+      "epoch": 0.12026870846576133,
+      "grad_norm": 0.9993457603508268,
+      "learning_rate": 2.936417357126155e-06,
+      "loss": 0.0234,
+      "step": 3330
+    },
+    {
+      "epoch": 0.12062987575845131,
+      "grad_norm": 1.015863824891173,
+      "learning_rate": 2.9359109334100553e-06,
+      "loss": 0.0235,
+      "step": 3340
+    },
+    {
+      "epoch": 0.12099104305114129,
+      "grad_norm": 1.016675672283532,
+      "learning_rate": 2.935402544921777e-06,
+      "loss": 0.023,
+      "step": 3350
+    },
+    {
+      "epoch": 0.12135221034383126,
+      "grad_norm": 1.404004265575021,
+      "learning_rate": 2.9348921923569533e-06,
+      "loss": 0.0249,
+      "step": 3360
+    },
+    {
+      "epoch": 0.12171337763652124,
+      "grad_norm": 1.0643985530349154,
+      "learning_rate": 2.9343798764139044e-06,
+      "loss": 0.0242,
+      "step": 3370
+    },
+    {
+      "epoch": 0.12207454492921122,
+      "grad_norm": 0.8355683306557175,
+      "learning_rate": 2.9338655977936385e-06,
+      "loss": 0.0229,
+      "step": 3380
+    },
+    {
+      "epoch": 0.12243571222190118,
+      "grad_norm": 0.9754645280675069,
+      "learning_rate": 2.9333493571998487e-06,
+      "loss": 0.0247,
+      "step": 3390
+    },
+    {
+      "epoch": 0.12279687951459116,
+      "grad_norm": 0.9704305269290203,
+      "learning_rate": 2.9328311553389116e-06,
+      "loss": 0.024,
+      "step": 3400
+    },
+    {
+      "epoch": 0.12315804680728114,
+      "grad_norm": 0.9388102495469629,
+      "learning_rate": 2.9323109929198887e-06,
+      "loss": 0.0235,
+      "step": 3410
+    },
+    {
+      "epoch": 0.1235192140999711,
+      "grad_norm": 0.9811007074393793,
+      "learning_rate": 2.9317888706545245e-06,
+      "loss": 0.0222,
+      "step": 3420
+    },
+    {
+      "epoch": 0.12388038139266108,
+      "grad_norm": 1.0526223836192772,
+      "learning_rate": 2.931264789257243e-06,
+      "loss": 0.0246,
+      "step": 3430
+    },
+    {
+      "epoch": 0.12424154868535106,
+      "grad_norm": 0.9730388397955962,
+      "learning_rate": 2.9307387494451514e-06,
+      "loss": 0.0233,
+      "step": 3440
+    },
+    {
+      "epoch": 0.12460271597804103,
+      "grad_norm": 1.130144884156202,
+      "learning_rate": 2.930210751938035e-06,
+      "loss": 0.023,
+      "step": 3450
+    },
+    {
+      "epoch": 0.124963883270731,
+      "grad_norm": 1.039552843784386,
+      "learning_rate": 2.929680797458359e-06,
+      "loss": 0.0227,
+      "step": 3460
+    },
+    {
+      "epoch": 0.12532505056342097,
+      "grad_norm": 1.333084493194798,
+      "learning_rate": 2.929148886731265e-06,
+      "loss": 0.022,
+      "step": 3470
+    },
+    {
+      "epoch": 0.12568621785611095,
+      "grad_norm": 1.013123663778419,
+      "learning_rate": 2.928615020484573e-06,
+      "loss": 0.0234,
+      "step": 3480
+    },
+    {
+      "epoch": 0.12604738514880093,
+      "grad_norm": 0.9718306574080104,
+      "learning_rate": 2.928079199448777e-06,
+      "loss": 0.022,
+      "step": 3490
+    },
+    {
+      "epoch": 0.1264085524414909,
+      "grad_norm": 1.065953820852412,
+      "learning_rate": 2.927541424357047e-06,
+      "loss": 0.0232,
+      "step": 3500
+    },
+    {
+      "epoch": 0.12676971973418089,
+      "grad_norm": 1.0613788591926656,
+      "learning_rate": 2.9270016959452265e-06,
+      "loss": 0.0247,
+      "step": 3510
+    },
+    {
+      "epoch": 0.12713088702687084,
+      "grad_norm": 0.9544275164490684,
+      "learning_rate": 2.926460014951831e-06,
+      "loss": 0.0225,
+      "step": 3520
+    },
+    {
+      "epoch": 0.12749205431956082,
+      "grad_norm": 0.9121036558853413,
+      "learning_rate": 2.9259163821180494e-06,
+      "loss": 0.0226,
+      "step": 3530
+    },
+    {
+      "epoch": 0.1278532216122508,
+      "grad_norm": 1.1275616971380547,
+      "learning_rate": 2.9253707981877395e-06,
+      "loss": 0.0244,
+      "step": 3540
+    },
+    {
+      "epoch": 0.12821438890494077,
+      "grad_norm": 0.9385282390552477,
+      "learning_rate": 2.9248232639074295e-06,
+      "loss": 0.0243,
+      "step": 3550
+    },
+    {
+      "epoch": 0.12857555619763075,
+      "grad_norm": 0.9022355521142872,
+      "learning_rate": 2.9242737800263168e-06,
+      "loss": 0.0238,
+      "step": 3560
+    },
+    {
+      "epoch": 0.1289367234903207,
+      "grad_norm": 1.0275089916732845,
+      "learning_rate": 2.9237223472962658e-06,
+      "loss": 0.0222,
+      "step": 3570
+    },
+    {
+      "epoch": 0.12929789078301068,
+      "grad_norm": 1.2848704799678499,
+      "learning_rate": 2.923168966471808e-06,
+      "loss": 0.0219,
+      "step": 3580
+    },
+    {
+      "epoch": 0.12965905807570066,
+      "grad_norm": 0.9018880879920652,
+      "learning_rate": 2.9226136383101397e-06,
+      "loss": 0.0238,
+      "step": 3590
+    },
+    {
+      "epoch": 0.13002022536839064,
+      "grad_norm": 1.3161552021169505,
+      "learning_rate": 2.922056363571123e-06,
+      "loss": 0.0223,
+      "step": 3600
+    },
+    {
+      "epoch": 0.13038139266108062,
+      "grad_norm": 1.0894394800776752,
+      "learning_rate": 2.9214971430172827e-06,
+      "loss": 0.023,
+      "step": 3610
+    },
+    {
+      "epoch": 0.1307425599537706,
+      "grad_norm": 0.976360477800859,
+      "learning_rate": 2.9209359774138055e-06,
+      "loss": 0.0228,
+      "step": 3620
+    },
+    {
+      "epoch": 0.13110372724646055,
+      "grad_norm": 1.8273995780934837,
+      "learning_rate": 2.9203728675285413e-06,
+      "loss": 0.0234,
+      "step": 3630
+    },
+    {
+      "epoch": 0.13146489453915053,
+      "grad_norm": 1.5530259159301834,
+      "learning_rate": 2.9198078141319992e-06,
+      "loss": 0.0235,
+      "step": 3640
+    },
+    {
+      "epoch": 0.1318260618318405,
+      "grad_norm": 0.851401104214509,
+      "learning_rate": 2.9192408179973467e-06,
+      "loss": 0.0246,
+      "step": 3650
+    },
+    {
+      "epoch": 0.1321872291245305,
+      "grad_norm": 1.263889259573535,
+      "learning_rate": 2.9186718799004115e-06,
+      "loss": 0.023,
+      "step": 3660
+    },
+    {
+      "epoch": 0.13254839641722047,
+      "grad_norm": 0.9922268306662819,
+      "learning_rate": 2.9181010006196777e-06,
+      "loss": 0.0227,
+      "step": 3670
+    },
+    {
+      "epoch": 0.13290956370991044,
+      "grad_norm": 1.0665444138887967,
+      "learning_rate": 2.9175281809362854e-06,
+      "loss": 0.023,
+      "step": 3680
+    },
+    {
+      "epoch": 0.1332707310026004,
+      "grad_norm": 1.1599708522246917,
+      "learning_rate": 2.9169534216340297e-06,
+      "loss": 0.0228,
+      "step": 3690
+    },
+    {
+      "epoch": 0.13363189829529037,
+      "grad_norm": 0.9977348839879102,
+      "learning_rate": 2.9163767234993595e-06,
+      "loss": 0.0215,
+      "step": 3700
+    },
+    {
+      "epoch": 0.13399306558798035,
+      "grad_norm": 1.0628513882332797,
+      "learning_rate": 2.915798087321378e-06,
+      "loss": 0.0238,
+      "step": 3710
+    },
+    {
+      "epoch": 0.13435423288067033,
+      "grad_norm": 1.5602654318316866,
+      "learning_rate": 2.9152175138918384e-06,
+      "loss": 0.0222,
+      "step": 3720
+    },
+    {
+      "epoch": 0.1347154001733603,
+      "grad_norm": 1.0281802543319842,
+      "learning_rate": 2.9146350040051452e-06,
+      "loss": 0.0236,
+      "step": 3730
+    },
+    {
+      "epoch": 0.13507656746605026,
+      "grad_norm": 0.99385396262854,
+      "learning_rate": 2.9140505584583534e-06,
+      "loss": 0.0222,
+      "step": 3740
+    },
+    {
+      "epoch": 0.13543773475874024,
+      "grad_norm": 1.0378552408419297,
+      "learning_rate": 2.9134641780511663e-06,
+      "loss": 0.0206,
+      "step": 3750
+    },
+    {
+      "epoch": 0.13579890205143022,
+      "grad_norm": 1.257975217893925,
+      "learning_rate": 2.912875863585934e-06,
+      "loss": 0.0245,
+      "step": 3760
+    },
+    {
+      "epoch": 0.1361600693441202,
+      "grad_norm": 1.0908447444776546,
+      "learning_rate": 2.912285615867654e-06,
+      "loss": 0.0217,
+      "step": 3770
+    },
+    {
+      "epoch": 0.13652123663681018,
+      "grad_norm": 1.783565680282964,
+      "learning_rate": 2.911693435703968e-06,
+      "loss": 0.0225,
+      "step": 3780
+    },
+    {
+      "epoch": 0.13688240392950016,
+      "grad_norm": 1.088676763256214,
+      "learning_rate": 2.911099323905162e-06,
+      "loss": 0.0224,
+      "step": 3790
+    },
+    {
+      "epoch": 0.1372435712221901,
+      "grad_norm": 1.3965132128713493,
+      "learning_rate": 2.9105032812841663e-06,
+      "loss": 0.0209,
+      "step": 3800
+    },
+    {
+      "epoch": 0.1376047385148801,
+      "grad_norm": 1.3558190123098697,
+      "learning_rate": 2.909905308656552e-06,
+      "loss": 0.0217,
+      "step": 3810
+    },
+    {
+      "epoch": 0.13796590580757007,
+      "grad_norm": 1.1019441274300341,
+      "learning_rate": 2.9093054068405313e-06,
+      "loss": 0.023,
+      "step": 3820
+    },
+    {
+      "epoch": 0.13832707310026005,
+      "grad_norm": 0.7922404378732472,
+      "learning_rate": 2.9087035766569567e-06,
+      "loss": 0.0206,
+      "step": 3830
+    },
+    {
+      "epoch": 0.13868824039295002,
+      "grad_norm": 1.06380228017603,
+      "learning_rate": 2.9080998189293185e-06,
+      "loss": 0.0221,
+      "step": 3840
+    },
+    {
+      "epoch": 0.13904940768563998,
+      "grad_norm": 0.8458839028960506,
+      "learning_rate": 2.907494134483744e-06,
+      "loss": 0.0226,
+      "step": 3850
+    },
+    {
+      "epoch": 0.13941057497832995,
+      "grad_norm": 0.9477301109906983,
+      "learning_rate": 2.9068865241489993e-06,
+      "loss": 0.0217,
+      "step": 3860
+    },
+    {
+      "epoch": 0.13977174227101993,
+      "grad_norm": 0.7264402121492679,
+      "learning_rate": 2.906276988756483e-06,
+      "loss": 0.0201,
+      "step": 3870
+    },
+    {
+      "epoch": 0.1401329095637099,
+      "grad_norm": 1.0131897708854927,
+      "learning_rate": 2.905665529140229e-06,
+      "loss": 0.0211,
+      "step": 3880
+    },
+    {
+      "epoch": 0.1404940768563999,
+      "grad_norm": 1.104902284005428,
+      "learning_rate": 2.9050521461369036e-06,
+      "loss": 0.022,
+      "step": 3890
+    },
+    {
+      "epoch": 0.14085524414908987,
+      "grad_norm": 0.931204089684204,
+      "learning_rate": 2.9044368405858066e-06,
+      "loss": 0.0227,
+      "step": 3900
+    },
+    {
+      "epoch": 0.14121641144177982,
+      "grad_norm": 0.8564549929177447,
+      "learning_rate": 2.9038196133288656e-06,
+      "loss": 0.0217,
+      "step": 3910
+    },
+    {
+      "epoch": 0.1415775787344698,
+      "grad_norm": 1.1976742773180935,
+      "learning_rate": 2.90320046521064e-06,
+      "loss": 0.0224,
+      "step": 3920
+    },
+    {
+      "epoch": 0.14193874602715978,
+      "grad_norm": 1.1934921862119712,
+      "learning_rate": 2.902579397078317e-06,
+      "loss": 0.0214,
+      "step": 3930
+    },
+    {
+      "epoch": 0.14229991331984976,
+      "grad_norm": 0.9955232952738895,
+      "learning_rate": 2.90195640978171e-06,
+      "loss": 0.0223,
+      "step": 3940
+    },
+    {
+      "epoch": 0.14266108061253974,
+      "grad_norm": 2.0546199614166523,
+      "learning_rate": 2.9013315041732597e-06,
+      "loss": 0.0221,
+      "step": 3950
+    },
+    {
+      "epoch": 0.14302224790522972,
+      "grad_norm": 0.8489804832466752,
+      "learning_rate": 2.900704681108031e-06,
+      "loss": 0.0219,
+      "step": 3960
+    },
+    {
+      "epoch": 0.14338341519791967,
+      "grad_norm": 1.0846002317778163,
+      "learning_rate": 2.9000759414437117e-06,
+      "loss": 0.0211,
+      "step": 3970
+    },
+    {
+      "epoch": 0.14374458249060965,
+      "grad_norm": 0.9129940122703538,
+      "learning_rate": 2.899445286040614e-06,
+      "loss": 0.0209,
+      "step": 3980
+    },
+    {
+      "epoch": 0.14410574978329962,
+      "grad_norm": 1.0105505723949408,
+      "learning_rate": 2.89881271576167e-06,
+      "loss": 0.0225,
+      "step": 3990
+    },
+    {
+      "epoch": 0.1444669170759896,
+      "grad_norm": 1.9100785371125943,
+      "learning_rate": 2.8981782314724326e-06,
+      "loss": 0.0216,
+      "step": 4000
+    },
+    {
+      "epoch": 0.1444669170759896,
+      "eval_loss": 0.18387216329574585,
+      "eval_runtime": 39.2998,
+      "eval_samples_per_second": 104.224,
+      "eval_steps_per_second": 0.204,
+      "step": 4000
+    },
+    {
+      "epoch": 0.14482808436867958,
+      "grad_norm": 0.9869730006832528,
+      "learning_rate": 2.897541834041073e-06,
+      "loss": 0.0234,
+      "step": 4010
+    },
+    {
+      "epoch": 0.14518925166136953,
+      "grad_norm": 1.114879778698819,
+      "learning_rate": 2.8969035243383814e-06,
+      "loss": 0.0226,
+      "step": 4020
+    },
+    {
+      "epoch": 0.1455504189540595,
+      "grad_norm": 0.790240572654315,
+      "learning_rate": 2.8962633032377622e-06,
+      "loss": 0.0215,
+      "step": 4030
+    },
+    {
+      "epoch": 0.1459115862467495,
+      "grad_norm": 1.059291917418531,
+      "learning_rate": 2.895621171615238e-06,
+      "loss": 0.0221,
+      "step": 4040
+    },
+    {
+      "epoch": 0.14627275353943947,
+      "grad_norm": 0.7894973985577327,
+      "learning_rate": 2.8949771303494435e-06,
+      "loss": 0.021,
+      "step": 4050
+    },
+    {
+      "epoch": 0.14663392083212945,
+      "grad_norm": 1.4364280047021944,
+      "learning_rate": 2.894331180321628e-06,
+      "loss": 0.0209,
+      "step": 4060
+    },
+    {
+      "epoch": 0.14699508812481943,
+      "grad_norm": 0.7814754148583977,
+      "learning_rate": 2.8936833224156515e-06,
+      "loss": 0.0212,
+      "step": 4070
+    },
+    {
+      "epoch": 0.14735625541750938,
+      "grad_norm": 0.9121036462073369,
+      "learning_rate": 2.893033557517985e-06,
+      "loss": 0.0208,
+      "step": 4080
+    },
+    {
+      "epoch": 0.14771742271019936,
+      "grad_norm": 0.8927341161936723,
+      "learning_rate": 2.892381886517708e-06,
+      "loss": 0.0223,
+      "step": 4090
+    },
+    {
+      "epoch": 0.14807859000288934,
+      "grad_norm": 1.0429587781244647,
+      "learning_rate": 2.8917283103065092e-06,
+      "loss": 0.0216,
+      "step": 4100
+    },
+    {
+      "epoch": 0.14843975729557932,
+      "grad_norm": 1.0774219276556827,
+      "learning_rate": 2.8910728297786848e-06,
+      "loss": 0.021,
+      "step": 4110
+    },
+    {
+      "epoch": 0.1488009245882693,
+      "grad_norm": 0.9659178483768529,
+      "learning_rate": 2.890415445831134e-06,
+      "loss": 0.0223,
+      "step": 4120
+    },
+    {
+      "epoch": 0.14916209188095925,
+      "grad_norm": 0.9347235308319273,
+      "learning_rate": 2.8897561593633637e-06,
+      "loss": 0.0211,
+      "step": 4130
+    },
+    {
+      "epoch": 0.14952325917364923,
+      "grad_norm": 0.8596396659911929,
+      "learning_rate": 2.8890949712774817e-06,
+      "loss": 0.0211,
+      "step": 4140
+    },
+    {
+      "epoch": 0.1498844264663392,
+      "grad_norm": 0.9242889599233269,
+      "learning_rate": 2.8884318824781996e-06,
+      "loss": 0.0231,
+      "step": 4150
+    },
+    {
+      "epoch": 0.15024559375902918,
+      "grad_norm": 0.8693882489059123,
+      "learning_rate": 2.8877668938728283e-06,
+      "loss": 0.021,
+      "step": 4160
+    },
+    {
+      "epoch": 0.15060676105171916,
+      "grad_norm": 0.7204477792939438,
+      "learning_rate": 2.887100006371278e-06,
+      "loss": 0.0217,
+      "step": 4170
+    },
+    {
+      "epoch": 0.15096792834440914,
+      "grad_norm": 0.7320279519525891,
+      "learning_rate": 2.8864312208860597e-06,
+      "loss": 0.0208,
+      "step": 4180
+    },
+    {
+      "epoch": 0.1513290956370991,
+      "grad_norm": 0.7249304604257303,
+      "learning_rate": 2.8857605383322777e-06,
+      "loss": 0.0204,
+      "step": 4190
+    },
+    {
+      "epoch": 0.15169026292978907,
+      "grad_norm": 0.601105593970423,
+      "learning_rate": 2.885087959627636e-06,
+      "loss": 0.0217,
+      "step": 4200
+    },
+    {
+      "epoch": 0.15205143022247905,
+      "grad_norm": 0.6281951118488608,
+      "learning_rate": 2.8844134856924295e-06,
+      "loss": 0.0216,
+      "step": 4210
+    },
+    {
+      "epoch": 0.15241259751516903,
+      "grad_norm": 0.6792882428041858,
+      "learning_rate": 2.8837371174495495e-06,
+      "loss": 0.0207,
+      "step": 4220
+    },
+    {
+      "epoch": 0.152773764807859,
+      "grad_norm": 1.0816130136525495,
+      "learning_rate": 2.8830588558244772e-06,
+      "loss": 0.0213,
+      "step": 4230
+    },
+    {
+      "epoch": 0.153134932100549,
+      "grad_norm": 0.611103970453974,
+      "learning_rate": 2.8823787017452847e-06,
+      "loss": 0.0205,
+      "step": 4240
+    },
+    {
+      "epoch": 0.15349609939323894,
+      "grad_norm": 0.8540799825087138,
+      "learning_rate": 2.881696656142634e-06,
+      "loss": 0.0215,
+      "step": 4250
+    },
+    {
+      "epoch": 0.15385726668592892,
+      "grad_norm": 0.8594632578985726,
+      "learning_rate": 2.8810127199497765e-06,
+      "loss": 0.0212,
+      "step": 4260
+    },
+    {
+      "epoch": 0.1542184339786189,
+      "grad_norm": 0.7351125650829493,
+      "learning_rate": 2.8803268941025487e-06,
+      "loss": 0.0219,
+      "step": 4270
+    },
+    {
+      "epoch": 0.15457960127130888,
+      "grad_norm": 0.7088694082932911,
+      "learning_rate": 2.8796391795393726e-06,
+      "loss": 0.0209,
+      "step": 4280
+    },
+    {
+      "epoch": 0.15494076856399885,
+      "grad_norm": 0.7695480450567029,
+      "learning_rate": 2.878949577201257e-06,
+      "loss": 0.0194,
+      "step": 4290
+    },
+    {
+      "epoch": 0.1553019358566888,
+      "grad_norm": 0.9649068517324223,
+      "learning_rate": 2.8782580880317913e-06,
+      "loss": 0.0205,
+      "step": 4300
+    },
+    {
+      "epoch": 0.15566310314937878,
+      "grad_norm": 0.7873968816796555,
+      "learning_rate": 2.877564712977147e-06,
+      "loss": 0.0215,
+      "step": 4310
+    },
+    {
+      "epoch": 0.15602427044206876,
+      "grad_norm": 0.7250699065225449,
+      "learning_rate": 2.8768694529860765e-06,
+      "loss": 0.0209,
+      "step": 4320
+    },
+    {
+      "epoch": 0.15638543773475874,
+      "grad_norm": 1.2747124903296472,
+      "learning_rate": 2.8761723090099127e-06,
+      "loss": 0.022,
+      "step": 4330
+    },
+    {
+      "epoch": 0.15674660502744872,
+      "grad_norm": 0.7588337257240909,
+      "learning_rate": 2.8754732820025637e-06,
+      "loss": 0.02,
+      "step": 4340
+    },
+    {
+      "epoch": 0.1571077723201387,
+      "grad_norm": 0.7331727785886997,
+      "learning_rate": 2.8747723729205166e-06,
+      "loss": 0.0202,
+      "step": 4350
+    },
+    {
+      "epoch": 0.15746893961282865,
+      "grad_norm": 1.095389615165598,
+      "learning_rate": 2.874069582722832e-06,
+      "loss": 0.0201,
+      "step": 4360
+    },
+    {
+      "epoch": 0.15783010690551863,
+      "grad_norm": 0.7207320583846528,
+      "learning_rate": 2.873364912371145e-06,
+      "loss": 0.0205,
+      "step": 4370
+    },
+    {
+      "epoch": 0.1581912741982086,
+      "grad_norm": 0.6797729791810813,
+      "learning_rate": 2.8726583628296645e-06,
+      "loss": 0.0219,
+      "step": 4380
+    },
+    {
+      "epoch": 0.1585524414908986,
+      "grad_norm": 1.0357944357158078,
+      "learning_rate": 2.8719499350651694e-06,
+      "loss": 0.0209,
+      "step": 4390
+    },
+    {
+      "epoch": 0.15891360878358857,
+      "grad_norm": 0.8358934303561882,
+      "learning_rate": 2.8712396300470083e-06,
+      "loss": 0.0222,
+      "step": 4400
+    },
+    {
+      "epoch": 0.15927477607627855,
+      "grad_norm": 0.8190604182237659,
+      "learning_rate": 2.8705274487471003e-06,
+      "loss": 0.0217,
+      "step": 4410
+    },
+    {
+      "epoch": 0.1596359433689685,
+      "grad_norm": 0.7396361215467311,
+      "learning_rate": 2.8698133921399293e-06,
+      "loss": 0.0198,
+      "step": 4420
+    },
+    {
+      "epoch": 0.15999711066165848,
+      "grad_norm": 0.6833440077415392,
+      "learning_rate": 2.8690974612025474e-06,
+      "loss": 0.0203,
+      "step": 4430
+    },
+    {
+      "epoch": 0.16035827795434845,
+      "grad_norm": 0.8061871756223801,
+      "learning_rate": 2.86837965691457e-06,
+      "loss": 0.0209,
+      "step": 4440
+    },
+    {
+      "epoch": 0.16071944524703843,
+      "grad_norm": 0.6226601513902809,
+      "learning_rate": 2.8676599802581767e-06,
+      "loss": 0.0203,
+      "step": 4450
+    },
+    {
+      "epoch": 0.1610806125397284,
+      "grad_norm": 0.9995808222821183,
+      "learning_rate": 2.8669384322181085e-06,
+      "loss": 0.0197,
+      "step": 4460
+    },
+    {
+      "epoch": 0.16144177983241836,
+      "grad_norm": 1.2523041721731554,
+      "learning_rate": 2.866215013781668e-06,
+      "loss": 0.0209,
+      "step": 4470
+    },
+    {
+      "epoch": 0.16180294712510834,
+      "grad_norm": 1.0680367553445114,
+      "learning_rate": 2.865489725938715e-06,
+      "loss": 0.0206,
+      "step": 4480
+    },
+    {
+      "epoch": 0.16216411441779832,
+      "grad_norm": 0.6533896769086377,
+      "learning_rate": 2.8647625696816694e-06,
+      "loss": 0.0199,
+      "step": 4490
+    },
+    {
+      "epoch": 0.1625252817104883,
+      "grad_norm": 0.7534496111543191,
+      "learning_rate": 2.8640335460055065e-06,
+      "loss": 0.021,
+      "step": 4500
+    },
+    {
+      "epoch": 0.16288644900317828,
+      "grad_norm": 0.89529699938759,
+      "learning_rate": 2.8633026559077573e-06,
+      "loss": 0.0188,
+      "step": 4510
+    },
+    {
+      "epoch": 0.16324761629586826,
+      "grad_norm": 0.767519137201243,
+      "learning_rate": 2.862569900388507e-06,
+      "loss": 0.0196,
+      "step": 4520
+    },
+    {
+      "epoch": 0.1636087835885582,
+      "grad_norm": 0.7761044949469634,
+      "learning_rate": 2.861835280450392e-06,
+      "loss": 0.0206,
+      "step": 4530
+    },
+    {
+      "epoch": 0.1639699508812482,
+      "grad_norm": 0.8306852315492617,
+      "learning_rate": 2.8610987970986008e-06,
+      "loss": 0.0215,
+      "step": 4540
+    },
+    {
+      "epoch": 0.16433111817393817,
+      "grad_norm": 0.7459491908527864,
+      "learning_rate": 2.860360451340871e-06,
+      "loss": 0.02,
+      "step": 4550
+    },
+    {
+      "epoch": 0.16469228546662815,
+      "grad_norm": 0.6953915543999247,
+      "learning_rate": 2.8596202441874898e-06,
+      "loss": 0.0192,
+      "step": 4560
+    },
+    {
+      "epoch": 0.16505345275931813,
+      "grad_norm": 0.7532199001003234,
+      "learning_rate": 2.8588781766512894e-06,
+      "loss": 0.0205,
+      "step": 4570
+    },
+    {
+      "epoch": 0.16541462005200808,
+      "grad_norm": 0.9756393548467054,
+      "learning_rate": 2.85813424974765e-06,
+      "loss": 0.0219,
+      "step": 4580
+    },
+    {
+      "epoch": 0.16577578734469806,
+      "grad_norm": 0.758951753015585,
+      "learning_rate": 2.8573884644944936e-06,
+      "loss": 0.0206,
+      "step": 4590
+    },
+    {
+      "epoch": 0.16613695463738803,
+      "grad_norm": 0.8801615516402983,
+      "learning_rate": 2.856640821912287e-06,
+      "loss": 0.0208,
+      "step": 4600
+    },
+    {
+      "epoch": 0.166498121930078,
+      "grad_norm": 0.9360330936295788,
+      "learning_rate": 2.8558913230240367e-06,
+      "loss": 0.0197,
+      "step": 4610
+    },
+    {
+      "epoch": 0.166859289222768,
+      "grad_norm": 0.8616305073651689,
+      "learning_rate": 2.85513996885529e-06,
+      "loss": 0.0205,
+      "step": 4620
+    },
+    {
+      "epoch": 0.16722045651545797,
+      "grad_norm": 0.9173484208788528,
+      "learning_rate": 2.854386760434133e-06,
+      "loss": 0.0204,
+      "step": 4630
+    },
+    {
+      "epoch": 0.16758162380814792,
+      "grad_norm": 0.7079453703545808,
+      "learning_rate": 2.8536316987911894e-06,
+      "loss": 0.0198,
+      "step": 4640
+    },
+    {
+      "epoch": 0.1679427911008379,
+      "grad_norm": 0.9817401067769341,
+      "learning_rate": 2.8528747849596174e-06,
+      "loss": 0.02,
+      "step": 4650
+    },
+    {
+      "epoch": 0.16830395839352788,
+      "grad_norm": 0.8092878140988818,
+      "learning_rate": 2.8521160199751102e-06,
+      "loss": 0.0198,
+      "step": 4660
+    },
+    {
+      "epoch": 0.16866512568621786,
+      "grad_norm": 0.8492910833268026,
+      "learning_rate": 2.851355404875894e-06,
+      "loss": 0.0201,
+      "step": 4670
+    },
+    {
+      "epoch": 0.16902629297890784,
+      "grad_norm": 0.8167494392522142,
+      "learning_rate": 2.8505929407027266e-06,
+      "loss": 0.0201,
+      "step": 4680
+    },
+    {
+      "epoch": 0.16938746027159782,
+      "grad_norm": 0.9055425224379892,
+      "learning_rate": 2.849828628498896e-06,
+      "loss": 0.0206,
+      "step": 4690
+    },
+    {
+      "epoch": 0.16974862756428777,
+      "grad_norm": 1.244236969013448,
+      "learning_rate": 2.8490624693102183e-06,
+      "loss": 0.0197,
+      "step": 4700
+    },
+    {
+      "epoch": 0.17010979485697775,
+      "grad_norm": 1.4732047619082218,
+      "learning_rate": 2.848294464185038e-06,
+      "loss": 0.0202,
+      "step": 4710
+    },
+    {
+      "epoch": 0.17047096214966773,
+      "grad_norm": 0.8746076221015463,
+      "learning_rate": 2.8475246141742236e-06,
+      "loss": 0.0197,
+      "step": 4720
+    },
+    {
+      "epoch": 0.1708321294423577,
+      "grad_norm": 0.777435066790494,
+      "learning_rate": 2.8467529203311703e-06,
+      "loss": 0.0205,
+      "step": 4730
+    },
+    {
+      "epoch": 0.17119329673504768,
+      "grad_norm": 0.8870768693904497,
+      "learning_rate": 2.8459793837117938e-06,
+      "loss": 0.0202,
+      "step": 4740
+    },
+    {
+      "epoch": 0.17155446402773764,
+      "grad_norm": 0.7655122529065079,
+      "learning_rate": 2.8452040053745333e-06,
+      "loss": 0.0207,
+      "step": 4750
+    },
+    {
+      "epoch": 0.17191563132042761,
+      "grad_norm": 0.7907915334563285,
+      "learning_rate": 2.844426786380347e-06,
+      "loss": 0.019,
+      "step": 4760
+    },
+    {
+      "epoch": 0.1722767986131176,
+      "grad_norm": 0.7582677909902071,
+      "learning_rate": 2.8436477277927112e-06,
+      "loss": 0.0194,
+      "step": 4770
+    },
+    {
+      "epoch": 0.17263796590580757,
+      "grad_norm": 1.0971125773347237,
+      "learning_rate": 2.8428668306776214e-06,
+      "loss": 0.0219,
+      "step": 4780
+    },
+    {
+      "epoch": 0.17299913319849755,
+      "grad_norm": 1.1774175782592375,
+      "learning_rate": 2.842084096103586e-06,
+      "loss": 0.0199,
+      "step": 4790
+    },
+    {
+      "epoch": 0.17336030049118753,
+      "grad_norm": 1.209679616398235,
+      "learning_rate": 2.8412995251416303e-06,
+      "loss": 0.0186,
+      "step": 4800
+    },
+    {
+      "epoch": 0.17372146778387748,
+      "grad_norm": 0.9086087608535283,
+      "learning_rate": 2.8405131188652896e-06,
+      "loss": 0.0196,
+      "step": 4810
+    },
+    {
+      "epoch": 0.17408263507656746,
+      "grad_norm": 0.8905421351552915,
+      "learning_rate": 2.8397248783506135e-06,
+      "loss": 0.0191,
+      "step": 4820
+    },
+    {
+      "epoch": 0.17444380236925744,
+      "grad_norm": 0.7649499199466442,
+      "learning_rate": 2.838934804676159e-06,
+      "loss": 0.0199,
+      "step": 4830
+    },
+    {
+      "epoch": 0.17480496966194742,
+      "grad_norm": 0.8774974335734167,
+      "learning_rate": 2.8381428989229927e-06,
+      "loss": 0.0194,
+      "step": 4840
+    },
+    {
+      "epoch": 0.1751661369546374,
+      "grad_norm": 1.0499361364633673,
+      "learning_rate": 2.837349162174687e-06,
+      "loss": 0.0194,
+      "step": 4850
+    },
+    {
+      "epoch": 0.17552730424732738,
+      "grad_norm": 0.75617108933989,
+      "learning_rate": 2.8365535955173216e-06,
+      "loss": 0.0197,
+      "step": 4860
+    },
+    {
+      "epoch": 0.17588847154001733,
+      "grad_norm": 0.8648660089310318,
+      "learning_rate": 2.835756200039477e-06,
+      "loss": 0.0205,
+      "step": 4870
+    },
+    {
+      "epoch": 0.1762496388327073,
+      "grad_norm": 0.8394800782647598,
+      "learning_rate": 2.8349569768322397e-06,
+      "loss": 0.0187,
+      "step": 4880
+    },
+    {
+      "epoch": 0.17661080612539728,
+      "grad_norm": 1.6550491739355024,
+      "learning_rate": 2.8341559269891938e-06,
+      "loss": 0.0186,
+      "step": 4890
+    },
+    {
+      "epoch": 0.17697197341808726,
+      "grad_norm": 0.6480426163956331,
+      "learning_rate": 2.8333530516064255e-06,
+      "loss": 0.0181,
+      "step": 4900
+    },
+    {
+      "epoch": 0.17733314071077724,
+      "grad_norm": 0.8453176631414692,
+      "learning_rate": 2.8325483517825167e-06,
+      "loss": 0.0197,
+      "step": 4910
+    },
+    {
+      "epoch": 0.1776943080034672,
+      "grad_norm": 0.7744174620317114,
+      "learning_rate": 2.8317418286185475e-06,
+      "loss": 0.0195,
+      "step": 4920
+    },
+    {
+      "epoch": 0.17805547529615717,
+      "grad_norm": 0.6976236002124092,
+      "learning_rate": 2.8309334832180914e-06,
+      "loss": 0.0193,
+      "step": 4930
+    },
+    {
+      "epoch": 0.17841664258884715,
+      "grad_norm": 0.621874975410085,
+      "learning_rate": 2.830123316687217e-06,
+      "loss": 0.0202,
+      "step": 4940
+    },
+    {
+      "epoch": 0.17877780988153713,
+      "grad_norm": 0.7123990615664253,
+      "learning_rate": 2.829311330134483e-06,
+      "loss": 0.02,
+      "step": 4950
+    },
+    {
+      "epoch": 0.1791389771742271,
+      "grad_norm": 1.0146461219620062,
+      "learning_rate": 2.828497524670939e-06,
+      "loss": 0.0197,
+      "step": 4960
+    },
+    {
+      "epoch": 0.1795001444669171,
+      "grad_norm": 0.894733794554196,
+      "learning_rate": 2.827681901410125e-06,
+      "loss": 0.0176,
+      "step": 4970
+    },
+    {
+      "epoch": 0.17986131175960704,
+      "grad_norm": 1.264477198652994,
+      "learning_rate": 2.826864461468065e-06,
+      "loss": 0.0204,
+      "step": 4980
+    },
+    {
+      "epoch": 0.18022247905229702,
+      "grad_norm": 0.863026877610938,
+      "learning_rate": 2.8260452059632727e-06,
+      "loss": 0.0179,
+      "step": 4990
+    },
+    {
+      "epoch": 0.180583646344987,
+      "grad_norm": 0.8550089336488327,
+      "learning_rate": 2.8252241360167424e-06,
+      "loss": 0.0197,
+      "step": 5000
+    },
+    {
+      "epoch": 0.180583646344987,
+      "eval_loss": 0.17660273611545563,
+      "eval_runtime": 38.4477,
+      "eval_samples_per_second": 106.534,
+      "eval_steps_per_second": 0.208,
+      "step": 5000
+    },
+    {
+      "epoch": 0.18094481363767698,
+      "grad_norm": 0.7294922627020959,
+      "learning_rate": 2.8244012527519542e-06,
+      "loss": 0.0202,
+      "step": 5010
+    },
+    {
+      "epoch": 0.18130598093036696,
+      "grad_norm": 1.0930839151855722,
+      "learning_rate": 2.823576557294867e-06,
+      "loss": 0.0187,
+      "step": 5020
+    },
+    {
+      "epoch": 0.1816671482230569,
+      "grad_norm": 6.455763033627483,
+      "learning_rate": 2.822750050773921e-06,
+      "loss": 0.019,
+      "step": 5030
+    },
+    {
+      "epoch": 0.18202831551574689,
+      "grad_norm": 0.8335289258147038,
+      "learning_rate": 2.8219217343200335e-06,
+      "loss": 0.0203,
+      "step": 5040
+    },
+    {
+      "epoch": 0.18238948280843686,
+      "grad_norm": 1.0561981248941772,
+      "learning_rate": 2.821091609066599e-06,
+      "loss": 0.0201,
+      "step": 5050
+    },
+    {
+      "epoch": 0.18275065010112684,
+      "grad_norm": 1.0953321789433221,
+      "learning_rate": 2.8202596761494867e-06,
+      "loss": 0.0191,
+      "step": 5060
+    },
+    {
+      "epoch": 0.18311181739381682,
+      "grad_norm": 1.1162712062864288,
+      "learning_rate": 2.8194259367070384e-06,
+      "loss": 0.0195,
+      "step": 5070
+    },
+    {
+      "epoch": 0.1834729846865068,
+      "grad_norm": 0.9050485375280377,
+      "learning_rate": 2.8185903918800698e-06,
+      "loss": 0.0186,
+      "step": 5080
+    },
+    {
+      "epoch": 0.18383415197919675,
+      "grad_norm": 1.7037354646596115,
+      "learning_rate": 2.817753042811865e-06,
+      "loss": 0.0208,
+      "step": 5090
+    },
+    {
+      "epoch": 0.18419531927188673,
+      "grad_norm": 0.9692644965847659,
+      "learning_rate": 2.8169138906481783e-06,
+      "loss": 0.0202,
+      "step": 5100
+    },
+    {
+      "epoch": 0.1845564865645767,
+      "grad_norm": 0.8809823722445786,
+      "learning_rate": 2.81607293653723e-06,
+      "loss": 0.0189,
+      "step": 5110
+    },
+    {
+      "epoch": 0.1849176538572667,
+      "grad_norm": 1.1813545742508698,
+      "learning_rate": 2.8152301816297066e-06,
+      "loss": 0.0212,
+      "step": 5120
+    },
+    {
+      "epoch": 0.18527882114995667,
+      "grad_norm": 0.6748828518048658,
+      "learning_rate": 2.814385627078758e-06,
+      "loss": 0.0182,
+      "step": 5130
+    },
+    {
+      "epoch": 0.18563998844264665,
+      "grad_norm": 1.1909457843989304,
+      "learning_rate": 2.813539274039999e-06,
+      "loss": 0.02,
+      "step": 5140
+    },
+    {
+      "epoch": 0.1860011557353366,
+      "grad_norm": 0.9105731500240968,
+      "learning_rate": 2.8126911236715013e-06,
+      "loss": 0.019,
+      "step": 5150
+    },
+    {
+      "epoch": 0.18636232302802658,
+      "grad_norm": 0.6172458693909525,
+      "learning_rate": 2.811841177133799e-06,
+      "loss": 0.0201,
+      "step": 5160
+    },
+    {
+      "epoch": 0.18672349032071656,
+      "grad_norm": 0.8691395533579604,
+      "learning_rate": 2.8109894355898837e-06,
+      "loss": 0.0195,
+      "step": 5170
+    },
+    {
+      "epoch": 0.18708465761340654,
+      "grad_norm": 0.6195254255898529,
+      "learning_rate": 2.810135900205201e-06,
+      "loss": 0.0193,
+      "step": 5180
+    },
+    {
+      "epoch": 0.18744582490609651,
+      "grad_norm": 0.7542924282088063,
+      "learning_rate": 2.8092805721476537e-06,
+      "loss": 0.0194,
+      "step": 5190
+    },
+    {
+      "epoch": 0.18780699219878647,
+      "grad_norm": 0.9410990265406827,
+      "learning_rate": 2.8084234525875953e-06,
+      "loss": 0.0203,
+      "step": 5200
+    },
+    {
+      "epoch": 0.18816815949147644,
+      "grad_norm": 0.5680959496426883,
+      "learning_rate": 2.8075645426978314e-06,
+      "loss": 0.0201,
+      "step": 5210
+    },
+    {
+      "epoch": 0.18852932678416642,
+      "grad_norm": 0.6158252691886984,
+      "learning_rate": 2.806703843653619e-06,
+      "loss": 0.0188,
+      "step": 5220
+    },
+    {
+      "epoch": 0.1888904940768564,
+      "grad_norm": 0.5159655563926893,
+      "learning_rate": 2.8058413566326607e-06,
+      "loss": 0.0189,
+      "step": 5230
+    },
+    {
+      "epoch": 0.18925166136954638,
+      "grad_norm": 1.0419110999448908,
+      "learning_rate": 2.8049770828151065e-06,
+      "loss": 0.0189,
+      "step": 5240
+    },
+    {
+      "epoch": 0.18961282866223636,
+      "grad_norm": 0.7786724288499506,
+      "learning_rate": 2.804111023383552e-06,
+      "loss": 0.0184,
+      "step": 5250
+    },
+    {
+      "epoch": 0.1899739959549263,
+      "grad_norm": 0.6473442613392099,
+      "learning_rate": 2.803243179523035e-06,
+      "loss": 0.0184,
+      "step": 5260
+    },
+    {
+      "epoch": 0.1903351632476163,
+      "grad_norm": 0.686598912973445,
+      "learning_rate": 2.8023735524210358e-06,
+      "loss": 0.0198,
+      "step": 5270
+    },
+    {
+      "epoch": 0.19069633054030627,
+      "grad_norm": 0.6636509231772526,
+      "learning_rate": 2.801502143267474e-06,
+      "loss": 0.0187,
+      "step": 5280
+    },
+    {
+      "epoch": 0.19105749783299625,
+      "grad_norm": 0.6711463626314156,
+      "learning_rate": 2.800628953254709e-06,
+      "loss": 0.0197,
+      "step": 5290
+    },
+    {
+      "epoch": 0.19141866512568623,
+      "grad_norm": 0.555119752983314,
+      "learning_rate": 2.799753983577535e-06,
+      "loss": 0.0184,
+      "step": 5300
+    },
+    {
+      "epoch": 0.19177983241837618,
+      "grad_norm": 0.7067090096673176,
+      "learning_rate": 2.7988772354331825e-06,
+      "loss": 0.0185,
+      "step": 5310
+    },
+    {
+      "epoch": 0.19214099971106616,
+      "grad_norm": 0.5353112285649535,
+      "learning_rate": 2.797998710021315e-06,
+      "loss": 0.018,
+      "step": 5320
+    },
+    {
+      "epoch": 0.19250216700375614,
+      "grad_norm": 2.1942426498574488,
+      "learning_rate": 2.797118408544029e-06,
+      "loss": 0.0187,
+      "step": 5330
+    },
+    {
+      "epoch": 0.19286333429644612,
+      "grad_norm": 0.9410646036963118,
+      "learning_rate": 2.79623633220585e-06,
+      "loss": 0.0185,
+      "step": 5340
+    },
+    {
+      "epoch": 0.1932245015891361,
+      "grad_norm": 0.7486040043797707,
+      "learning_rate": 2.795352482213732e-06,
+      "loss": 0.0213,
+      "step": 5350
+    },
+    {
+      "epoch": 0.19358566888182607,
+      "grad_norm": 0.6300583252201749,
+      "learning_rate": 2.7944668597770564e-06,
+      "loss": 0.0181,
+      "step": 5360
+    },
+    {
+      "epoch": 0.19394683617451602,
+      "grad_norm": 0.8903447037481014,
+      "learning_rate": 2.7935794661076304e-06,
+      "loss": 0.0193,
+      "step": 5370
+    },
+    {
+      "epoch": 0.194308003467206,
+      "grad_norm": 0.5918597153484177,
+      "learning_rate": 2.792690302419684e-06,
+      "loss": 0.0189,
+      "step": 5380
+    },
+    {
+      "epoch": 0.19466917075989598,
+      "grad_norm": 0.6394796741795215,
+      "learning_rate": 2.7917993699298688e-06,
+      "loss": 0.02,
+      "step": 5390
+    },
+    {
+      "epoch": 0.19503033805258596,
+      "grad_norm": 0.61125406774695,
+      "learning_rate": 2.7909066698572577e-06,
+      "loss": 0.0187,
+      "step": 5400
+    },
+    {
+      "epoch": 0.19539150534527594,
+      "grad_norm": 0.7254436317763739,
+      "learning_rate": 2.790012203423341e-06,
+      "loss": 0.0191,
+      "step": 5410
+    },
+    {
+      "epoch": 0.19575267263796592,
+      "grad_norm": 0.8339512525323466,
+      "learning_rate": 2.7891159718520276e-06,
+      "loss": 0.0188,
+      "step": 5420
+    },
+    {
+      "epoch": 0.19611383993065587,
+      "grad_norm": 0.9032885777971003,
+      "learning_rate": 2.7882179763696403e-06,
+      "loss": 0.0183,
+      "step": 5430
+    },
+    {
+      "epoch": 0.19647500722334585,
+      "grad_norm": 0.6740117068482366,
+      "learning_rate": 2.7873182182049157e-06,
+      "loss": 0.0176,
+      "step": 5440
+    },
+    {
+      "epoch": 0.19683617451603583,
+      "grad_norm": 0.6016525429183767,
+      "learning_rate": 2.7864166985890016e-06,
+      "loss": 0.0179,
+      "step": 5450
+    },
+    {
+      "epoch": 0.1971973418087258,
+      "grad_norm": 0.8934662131531277,
+      "learning_rate": 2.785513418755459e-06,
+      "loss": 0.0197,
+      "step": 5460
+    },
+    {
+      "epoch": 0.19755850910141579,
+      "grad_norm": 0.6347525354638006,
+      "learning_rate": 2.7846083799402536e-06,
+      "loss": 0.0195,
+      "step": 5470
+    },
+    {
+      "epoch": 0.19791967639410574,
+      "grad_norm": 0.6088789948320684,
+      "learning_rate": 2.7837015833817595e-06,
+      "loss": 0.0187,
+      "step": 5480
+    },
+    {
+      "epoch": 0.19828084368679572,
+      "grad_norm": 1.1093929235607585,
+      "learning_rate": 2.7827930303207576e-06,
+      "loss": 0.0196,
+      "step": 5490
+    },
+    {
+      "epoch": 0.1986420109794857,
+      "grad_norm": 0.5926615276341989,
+      "learning_rate": 2.7818827220004292e-06,
+      "loss": 0.0178,
+      "step": 5500
+    },
+    {
+      "epoch": 0.19900317827217567,
+      "grad_norm": 0.7370739454148899,
+      "learning_rate": 2.7809706596663597e-06,
+      "loss": 0.0194,
+      "step": 5510
+    },
+    {
+      "epoch": 0.19936434556486565,
+      "grad_norm": 0.6238066609647006,
+      "learning_rate": 2.780056844566533e-06,
+      "loss": 0.0189,
+      "step": 5520
+    },
+    {
+      "epoch": 0.19972551285755563,
+      "grad_norm": 0.770778783317666,
+      "learning_rate": 2.7791412779513323e-06,
+      "loss": 0.0185,
+      "step": 5530
+    },
+    {
+      "epoch": 0.20008668015024558,
+      "grad_norm": 0.7476969561022,
+      "learning_rate": 2.778223961073537e-06,
+      "loss": 0.0175,
+      "step": 5540
+    },
+    {
+      "epoch": 0.20044784744293556,
+      "grad_norm": 0.8749823425109092,
+      "learning_rate": 2.7773048951883227e-06,
+      "loss": 0.0175,
+      "step": 5550
+    },
+    {
+      "epoch": 0.20080901473562554,
+      "grad_norm": 0.8646423516256542,
+      "learning_rate": 2.776384081553255e-06,
+      "loss": 0.018,
+      "step": 5560
+    },
+    {
+      "epoch": 0.20117018202831552,
+      "grad_norm": 0.6929291324329181,
+      "learning_rate": 2.7754615214282946e-06,
+      "loss": 0.0184,
+      "step": 5570
+    },
+    {
+      "epoch": 0.2015313493210055,
+      "grad_norm": 0.7022893501744458,
+      "learning_rate": 2.7745372160757893e-06,
+      "loss": 0.0181,
+      "step": 5580
+    },
+    {
+      "epoch": 0.20189251661369548,
+      "grad_norm": 0.6229218450435445,
+      "learning_rate": 2.773611166760477e-06,
+      "loss": 0.0187,
+      "step": 5590
+    },
+    {
+      "epoch": 0.20225368390638543,
+      "grad_norm": 0.7802654302660674,
+      "learning_rate": 2.7726833747494795e-06,
+      "loss": 0.0193,
+      "step": 5600
+    },
+    {
+      "epoch": 0.2026148511990754,
+      "grad_norm": 0.7675532407438775,
+      "learning_rate": 2.7717538413123053e-06,
+      "loss": 0.0192,
+      "step": 5610
+    },
+    {
+      "epoch": 0.2029760184917654,
+      "grad_norm": 0.6036273847380762,
+      "learning_rate": 2.7708225677208445e-06,
+      "loss": 0.0171,
+      "step": 5620
+    },
+    {
+      "epoch": 0.20333718578445537,
+      "grad_norm": 0.723251556347302,
+      "learning_rate": 2.769889555249369e-06,
+      "loss": 0.019,
+      "step": 5630
+    },
+    {
+      "epoch": 0.20369835307714534,
+      "grad_norm": 0.7695225385698774,
+      "learning_rate": 2.76895480517453e-06,
+      "loss": 0.0188,
+      "step": 5640
+    },
+    {
+      "epoch": 0.2040595203698353,
+      "grad_norm": 3.0069684919614486,
+      "learning_rate": 2.768018318775355e-06,
+      "loss": 0.0181,
+      "step": 5650
+    },
+    {
+      "epoch": 0.20442068766252527,
+      "grad_norm": 0.6874729254972032,
+      "learning_rate": 2.7670800973332484e-06,
+      "loss": 0.0194,
+      "step": 5660
+    },
+    {
+      "epoch": 0.20478185495521525,
+      "grad_norm": 0.8612218559138693,
+      "learning_rate": 2.766140142131989e-06,
+      "loss": 0.0172,
+      "step": 5670
+    },
+    {
+      "epoch": 0.20514302224790523,
+      "grad_norm": 1.0646874363513779,
+      "learning_rate": 2.7651984544577276e-06,
+      "loss": 0.0179,
+      "step": 5680
+    },
+    {
+      "epoch": 0.2055041895405952,
+      "grad_norm": 0.9814904506462871,
+      "learning_rate": 2.7642550355989857e-06,
+      "loss": 0.0178,
+      "step": 5690
+    },
+    {
+      "epoch": 0.2058653568332852,
+      "grad_norm": 0.5947362080041848,
+      "learning_rate": 2.763309886846652e-06,
+      "loss": 0.0178,
+      "step": 5700
+    },
+    {
+      "epoch": 0.20622652412597514,
+      "grad_norm": 1.2056635624677656,
+      "learning_rate": 2.7623630094939854e-06,
+      "loss": 0.0182,
+      "step": 5710
+    },
+    {
+      "epoch": 0.20658769141866512,
+      "grad_norm": 0.8345992277922449,
+      "learning_rate": 2.7614144048366065e-06,
+      "loss": 0.0188,
+      "step": 5720
+    },
+    {
+      "epoch": 0.2069488587113551,
+      "grad_norm": 0.6623222955228804,
+      "learning_rate": 2.760464074172502e-06,
+      "loss": 0.0187,
+      "step": 5730
+    },
+    {
+      "epoch": 0.20731002600404508,
+      "grad_norm": 0.840747042592569,
+      "learning_rate": 2.75951201880202e-06,
+      "loss": 0.0188,
+      "step": 5740
+    },
+    {
+      "epoch": 0.20767119329673506,
+      "grad_norm": 0.6473168735322594,
+      "learning_rate": 2.7585582400278672e-06,
+      "loss": 0.0189,
+      "step": 5750
+    },
+    {
+      "epoch": 0.208032360589425,
+      "grad_norm": 0.616900414438496,
+      "learning_rate": 2.757602739155109e-06,
+      "loss": 0.0182,
+      "step": 5760
+    },
+    {
+      "epoch": 0.208393527882115,
+      "grad_norm": 0.8094131376850638,
+      "learning_rate": 2.756645517491168e-06,
+      "loss": 0.0168,
+      "step": 5770
+    },
+    {
+      "epoch": 0.20875469517480497,
+      "grad_norm": 1.8104791558219195,
+      "learning_rate": 2.75568657634582e-06,
+      "loss": 0.0183,
+      "step": 5780
+    },
+    {
+      "epoch": 0.20911586246749495,
+      "grad_norm": 0.6025135011126365,
+      "learning_rate": 2.754725917031195e-06,
+      "loss": 0.0174,
+      "step": 5790
+    },
+    {
+      "epoch": 0.20947702976018492,
+      "grad_norm": 1.6449593012645616,
+      "learning_rate": 2.7537635408617735e-06,
+      "loss": 0.0187,
+      "step": 5800
+    },
+    {
+      "epoch": 0.2098381970528749,
+      "grad_norm": 0.9089870138240269,
+      "learning_rate": 2.752799449154385e-06,
+      "loss": 0.0187,
+      "step": 5810
+    },
+    {
+      "epoch": 0.21019936434556485,
+      "grad_norm": 0.6351236171796764,
+      "learning_rate": 2.7518336432282053e-06,
+      "loss": 0.0172,
+      "step": 5820
+    },
+    {
+      "epoch": 0.21056053163825483,
+      "grad_norm": 0.7787544513099935,
+      "learning_rate": 2.7508661244047586e-06,
+      "loss": 0.019,
+      "step": 5830
+    },
+    {
+      "epoch": 0.2109216989309448,
+      "grad_norm": 0.5993378539850092,
+      "learning_rate": 2.74989689400791e-06,
+      "loss": 0.017,
+      "step": 5840
+    },
+    {
+      "epoch": 0.2112828662236348,
+      "grad_norm": 1.1775716661060012,
+      "learning_rate": 2.7489259533638676e-06,
+      "loss": 0.0179,
+      "step": 5850
+    },
+    {
+      "epoch": 0.21164403351632477,
+      "grad_norm": 0.6895934868327835,
+      "learning_rate": 2.7479533038011813e-06,
+      "loss": 0.018,
+      "step": 5860
+    },
+    {
+      "epoch": 0.21200520080901475,
+      "grad_norm": 0.9613453966365332,
+      "learning_rate": 2.746978946650736e-06,
+      "loss": 0.0182,
+      "step": 5870
+    },
+    {
+      "epoch": 0.2123663681017047,
+      "grad_norm": 0.6822733963768682,
+      "learning_rate": 2.746002883245756e-06,
+      "loss": 0.0185,
+      "step": 5880
+    },
+    {
+      "epoch": 0.21272753539439468,
+      "grad_norm": 0.9473267944681202,
+      "learning_rate": 2.745025114921799e-06,
+      "loss": 0.0184,
+      "step": 5890
+    },
+    {
+      "epoch": 0.21308870268708466,
+      "grad_norm": 0.7151319010505307,
+      "learning_rate": 2.7440456430167556e-06,
+      "loss": 0.0181,
+      "step": 5900
+    },
+    {
+      "epoch": 0.21344986997977464,
+      "grad_norm": 1.1696347286490056,
+      "learning_rate": 2.743064468870848e-06,
+      "loss": 0.0177,
+      "step": 5910
+    },
+    {
+      "epoch": 0.21381103727246462,
+      "grad_norm": 0.5780712710247988,
+      "learning_rate": 2.742081593826627e-06,
+      "loss": 0.0176,
+      "step": 5920
+    },
+    {
+      "epoch": 0.21417220456515457,
+      "grad_norm": 0.675553151593643,
+      "learning_rate": 2.7410970192289714e-06,
+      "loss": 0.0188,
+      "step": 5930
+    },
+    {
+      "epoch": 0.21453337185784455,
+      "grad_norm": 0.9838060591720673,
+      "learning_rate": 2.740110746425084e-06,
+      "loss": 0.0175,
+      "step": 5940
+    },
+    {
+      "epoch": 0.21489453915053452,
+      "grad_norm": 0.7160107122036781,
+      "learning_rate": 2.7391227767644928e-06,
+      "loss": 0.0184,
+      "step": 5950
+    },
+    {
+      "epoch": 0.2152557064432245,
+      "grad_norm": 0.8875699304564796,
+      "learning_rate": 2.7381331115990476e-06,
+      "loss": 0.0184,
+      "step": 5960
+    },
+    {
+      "epoch": 0.21561687373591448,
+      "grad_norm": 0.661227580295918,
+      "learning_rate": 2.7371417522829173e-06,
+      "loss": 0.0174,
+      "step": 5970
+    },
+    {
+      "epoch": 0.21597804102860446,
+      "grad_norm": 0.6468109944446252,
+      "learning_rate": 2.7361487001725897e-06,
+      "loss": 0.0179,
+      "step": 5980
+    },
+    {
+      "epoch": 0.2163392083212944,
+      "grad_norm": 0.611033301616413,
+      "learning_rate": 2.7351539566268683e-06,
+      "loss": 0.0181,
+      "step": 5990
+    },
+    {
+      "epoch": 0.2167003756139844,
+      "grad_norm": 0.653601144147023,
+      "learning_rate": 2.7341575230068717e-06,
+      "loss": 0.0173,
+      "step": 6000
+    },
+    {
+      "epoch": 0.2167003756139844,
+      "eval_loss": 0.18290069699287415,
+      "eval_runtime": 38.8918,
+      "eval_samples_per_second": 105.318,
+      "eval_steps_per_second": 0.206,
+      "step": 6000
+    },
+    {
+      "epoch": 0.21706154290667437,
+      "grad_norm": 0.6729286518408925,
+      "learning_rate": 2.73315940067603e-06,
+      "loss": 0.0175,
+      "step": 6010
+    },
+    {
+      "epoch": 0.21742271019936435,
+      "grad_norm": 0.7313684419505742,
+      "learning_rate": 2.7321595910000846e-06,
+      "loss": 0.0177,
+      "step": 6020
+    },
+    {
+      "epoch": 0.21778387749205433,
+      "grad_norm": 0.5804246507304548,
+      "learning_rate": 2.731158095347087e-06,
+      "loss": 0.0174,
+      "step": 6030
+    },
+    {
+      "epoch": 0.21814504478474428,
+      "grad_norm": 0.5346407950267446,
+      "learning_rate": 2.730154915087392e-06,
+      "loss": 0.0172,
+      "step": 6040
+    },
+    {
+      "epoch": 0.21850621207743426,
+      "grad_norm": 0.6826347470205802,
+      "learning_rate": 2.7291500515936636e-06,
+      "loss": 0.0182,
+      "step": 6050
+    },
+    {
+      "epoch": 0.21886737937012424,
+      "grad_norm": 0.6785430328510087,
+      "learning_rate": 2.7281435062408668e-06,
+      "loss": 0.0172,
+      "step": 6060
+    },
+    {
+      "epoch": 0.21922854666281422,
+      "grad_norm": 0.7563289982559749,
+      "learning_rate": 2.727135280406268e-06,
+      "loss": 0.0186,
+      "step": 6070
+    },
+    {
+      "epoch": 0.2195897139555042,
+      "grad_norm": 0.7247016197114028,
+      "learning_rate": 2.726125375469434e-06,
+      "loss": 0.0173,
+      "step": 6080
+    },
+    {
+      "epoch": 0.21995088124819417,
+      "grad_norm": 0.7305899364417616,
+      "learning_rate": 2.7251137928122275e-06,
+      "loss": 0.0178,
+      "step": 6090
+    },
+    {
+      "epoch": 0.22031204854088413,
+      "grad_norm": 0.5357188587863804,
+      "learning_rate": 2.724100533818808e-06,
+      "loss": 0.0185,
+      "step": 6100
+    },
+    {
+      "epoch": 0.2206732158335741,
+      "grad_norm": 0.5902535054917214,
+      "learning_rate": 2.72308559987563e-06,
+      "loss": 0.0178,
+      "step": 6110
+    },
+    {
+      "epoch": 0.22103438312626408,
+      "grad_norm": 0.6210572440558841,
+      "learning_rate": 2.7220689923714357e-06,
+      "loss": 0.0172,
+      "step": 6120
+    },
+    {
+      "epoch": 0.22139555041895406,
+      "grad_norm": 0.8406653297837968,
+      "learning_rate": 2.7210507126972623e-06,
+      "loss": 0.0181,
+      "step": 6130
+    },
+    {
+      "epoch": 0.22175671771164404,
+      "grad_norm": 1.0146877104816188,
+      "learning_rate": 2.720030762246431e-06,
+      "loss": 0.0178,
+      "step": 6140
+    },
+    {
+      "epoch": 0.22211788500433402,
+      "grad_norm": 0.7429069368896167,
+      "learning_rate": 2.7190091424145524e-06,
+      "loss": 0.0183,
+      "step": 6150
+    },
+    {
+      "epoch": 0.22247905229702397,
+      "grad_norm": 0.6619739962603685,
+      "learning_rate": 2.7179858545995187e-06,
+      "loss": 0.0165,
+      "step": 6160
+    },
+    {
+      "epoch": 0.22284021958971395,
+      "grad_norm": 0.7069603565948127,
+      "learning_rate": 2.716960900201506e-06,
+      "loss": 0.0167,
+      "step": 6170
+    },
+    {
+      "epoch": 0.22320138688240393,
+      "grad_norm": 0.8092566979224886,
+      "learning_rate": 2.71593428062297e-06,
+      "loss": 0.018,
+      "step": 6180
+    },
+    {
+      "epoch": 0.2235625541750939,
+      "grad_norm": 0.5815674794430107,
+      "learning_rate": 2.714905997268645e-06,
+      "loss": 0.0184,
+      "step": 6190
+    },
+    {
+      "epoch": 0.2239237214677839,
+      "grad_norm": 0.707793506393144,
+      "learning_rate": 2.7138760515455426e-06,
+      "loss": 0.0185,
+      "step": 6200
+    },
+    {
+      "epoch": 0.22428488876047384,
+      "grad_norm": 0.596756474938895,
+      "learning_rate": 2.7128444448629482e-06,
+      "loss": 0.0178,
+      "step": 6210
+    },
+    {
+      "epoch": 0.22464605605316382,
+      "grad_norm": 0.6141001364253705,
+      "learning_rate": 2.7118111786324203e-06,
+      "loss": 0.0171,
+      "step": 6220
+    },
+    {
+      "epoch": 0.2250072233458538,
+      "grad_norm": 0.6136452271503211,
+      "learning_rate": 2.7107762542677873e-06,
+      "loss": 0.018,
+      "step": 6230
+    },
+    {
+      "epoch": 0.22536839063854378,
+      "grad_norm": 1.0134284788188042,
+      "learning_rate": 2.7097396731851477e-06,
+      "loss": 0.0191,
+      "step": 6240
+    },
+    {
+      "epoch": 0.22572955793123375,
+      "grad_norm": 0.6510475555187035,
+      "learning_rate": 2.708701436802867e-06,
+      "loss": 0.0172,
+      "step": 6250
+    },
+    {
+      "epoch": 0.22609072522392373,
+      "grad_norm": 0.7065340677014725,
+      "learning_rate": 2.7076615465415746e-06,
+      "loss": 0.0172,
+      "step": 6260
+    },
+    {
+      "epoch": 0.22645189251661368,
+      "grad_norm": 0.7706344325277558,
+      "learning_rate": 2.7066200038241634e-06,
+      "loss": 0.0169,
+      "step": 6270
+    },
+    {
+      "epoch": 0.22681305980930366,
+      "grad_norm": 0.9194451067184941,
+      "learning_rate": 2.705576810075787e-06,
+      "loss": 0.0165,
+      "step": 6280
+    },
+    {
+      "epoch": 0.22717422710199364,
+      "grad_norm": 0.48766907212827326,
+      "learning_rate": 2.7045319667238585e-06,
+      "loss": 0.0177,
+      "step": 6290
+    },
+    {
+      "epoch": 0.22753539439468362,
+      "grad_norm": 0.6659911603165722,
+      "learning_rate": 2.703485475198049e-06,
+      "loss": 0.0168,
+      "step": 6300
+    },
+    {
+      "epoch": 0.2278965616873736,
+      "grad_norm": 0.73179751025868,
+      "learning_rate": 2.7024373369302832e-06,
+      "loss": 0.0171,
+      "step": 6310
+    },
+    {
+      "epoch": 0.22825772898006358,
+      "grad_norm": 0.743912262562094,
+      "learning_rate": 2.70138755335474e-06,
+      "loss": 0.0165,
+      "step": 6320
+    },
+    {
+      "epoch": 0.22861889627275353,
+      "grad_norm": 0.5856741515861786,
+      "learning_rate": 2.7003361259078498e-06,
+      "loss": 0.0177,
+      "step": 6330
+    },
+    {
+      "epoch": 0.2289800635654435,
+      "grad_norm": 0.5460708860860752,
+      "learning_rate": 2.6992830560282914e-06,
+      "loss": 0.017,
+      "step": 6340
+    },
+    {
+      "epoch": 0.2293412308581335,
+      "grad_norm": 0.4901871443064363,
+      "learning_rate": 2.6982283451569917e-06,
+      "loss": 0.0169,
+      "step": 6350
+    },
+    {
+      "epoch": 0.22970239815082347,
+      "grad_norm": 0.993962458655905,
+      "learning_rate": 2.697171994737123e-06,
+      "loss": 0.0166,
+      "step": 6360
+    },
+    {
+      "epoch": 0.23006356544351345,
+      "grad_norm": 0.7603966077225356,
+      "learning_rate": 2.6961140062141e-06,
+      "loss": 0.0174,
+      "step": 6370
+    },
+    {
+      "epoch": 0.2304247327362034,
+      "grad_norm": 1.1383561078952167,
+      "learning_rate": 2.6950543810355803e-06,
+      "loss": 0.0177,
+      "step": 6380
+    },
+    {
+      "epoch": 0.23078590002889338,
+      "grad_norm": 0.6147656614113562,
+      "learning_rate": 2.6939931206514602e-06,
+      "loss": 0.0171,
+      "step": 6390
+    },
+    {
+      "epoch": 0.23114706732158335,
+      "grad_norm": 0.5989492525801406,
+      "learning_rate": 2.692930226513874e-06,
+      "loss": 0.0186,
+      "step": 6400
+    },
+    {
+      "epoch": 0.23150823461427333,
+      "grad_norm": 1.0975929720384332,
+      "learning_rate": 2.6918657000771902e-06,
+      "loss": 0.0164,
+      "step": 6410
+    },
+    {
+      "epoch": 0.2318694019069633,
+      "grad_norm": 0.5660934037294238,
+      "learning_rate": 2.690799542798012e-06,
+      "loss": 0.0177,
+      "step": 6420
+    },
+    {
+      "epoch": 0.2322305691996533,
+      "grad_norm": 2.8907103980020437,
+      "learning_rate": 2.6897317561351736e-06,
+      "loss": 0.018,
+      "step": 6430
+    },
+    {
+      "epoch": 0.23259173649234324,
+      "grad_norm": 0.6950084539920196,
+      "learning_rate": 2.6886623415497395e-06,
+      "loss": 0.0177,
+      "step": 6440
+    },
+    {
+      "epoch": 0.23295290378503322,
+      "grad_norm": 0.7919925581152503,
+      "learning_rate": 2.6875913005050003e-06,
+      "loss": 0.0168,
+      "step": 6450
+    },
+    {
+      "epoch": 0.2333140710777232,
+      "grad_norm": 0.7219950753094981,
+      "learning_rate": 2.686518634466473e-06,
+      "loss": 0.0171,
+      "step": 6460
+    },
+    {
+      "epoch": 0.23367523837041318,
+      "grad_norm": 1.2604660993408334,
+      "learning_rate": 2.685444344901898e-06,
+      "loss": 0.0175,
+      "step": 6470
+    },
+    {
+      "epoch": 0.23403640566310316,
+      "grad_norm": 0.7939799678992704,
+      "learning_rate": 2.6843684332812368e-06,
+      "loss": 0.0178,
+      "step": 6480
+    },
+    {
+      "epoch": 0.2343975729557931,
+      "grad_norm": 1.228093127942303,
+      "learning_rate": 2.683290901076671e-06,
+      "loss": 0.0169,
+      "step": 6490
+    },
+    {
+      "epoch": 0.2347587402484831,
+      "grad_norm": 0.6537856611992066,
+      "learning_rate": 2.6822117497626e-06,
+      "loss": 0.0167,
+      "step": 6500
+    },
+    {
+      "epoch": 0.23511990754117307,
+      "grad_norm": 0.5285368743332429,
+      "learning_rate": 2.6811309808156366e-06,
+      "loss": 0.0169,
+      "step": 6510
+    },
+    {
+      "epoch": 0.23548107483386305,
+      "grad_norm": 1.275794427007645,
+      "learning_rate": 2.68004859571461e-06,
+      "loss": 0.0159,
+      "step": 6520
+    },
+    {
+      "epoch": 0.23584224212655303,
+      "grad_norm": 1.0670147076862766,
+      "learning_rate": 2.6789645959405573e-06,
+      "loss": 0.0157,
+      "step": 6530
+    },
+    {
+      "epoch": 0.236203409419243,
+      "grad_norm": 0.8751427639113412,
+      "learning_rate": 2.6778789829767285e-06,
+      "loss": 0.0176,
+      "step": 6540
+    },
+    {
+      "epoch": 0.23656457671193296,
+      "grad_norm": 0.7564966867865751,
+      "learning_rate": 2.676791758308579e-06,
+      "loss": 0.0174,
+      "step": 6550
+    },
+    {
+      "epoch": 0.23692574400462293,
+      "grad_norm": 0.691611049647592,
+      "learning_rate": 2.6757029234237695e-06,
+      "loss": 0.0167,
+      "step": 6560
+    },
+    {
+      "epoch": 0.2372869112973129,
+      "grad_norm": 0.8486126835831662,
+      "learning_rate": 2.674612479812165e-06,
+      "loss": 0.0156,
+      "step": 6570
+    },
+    {
+      "epoch": 0.2376480785900029,
+      "grad_norm": 0.6698365805871095,
+      "learning_rate": 2.67352042896583e-06,
+      "loss": 0.0163,
+      "step": 6580
+    },
+    {
+      "epoch": 0.23800924588269287,
+      "grad_norm": 0.568519372676868,
+      "learning_rate": 2.67242677237903e-06,
+      "loss": 0.0165,
+      "step": 6590
+    },
+    {
+      "epoch": 0.23837041317538285,
+      "grad_norm": 0.7740352575802951,
+      "learning_rate": 2.671331511548227e-06,
+      "loss": 0.0167,
+      "step": 6600
+    },
+    {
+      "epoch": 0.2387315804680728,
+      "grad_norm": 0.655902358813912,
+      "learning_rate": 2.6702346479720777e-06,
+      "loss": 0.0161,
+      "step": 6610
+    },
+    {
+      "epoch": 0.23909274776076278,
+      "grad_norm": 0.6024478413501874,
+      "learning_rate": 2.6691361831514326e-06,
+      "loss": 0.0159,
+      "step": 6620
+    },
+    {
+      "epoch": 0.23945391505345276,
+      "grad_norm": 0.8538479526203305,
+      "learning_rate": 2.668036118589332e-06,
+      "loss": 0.0177,
+      "step": 6630
+    },
+    {
+      "epoch": 0.23981508234614274,
+      "grad_norm": 0.7638940221229296,
+      "learning_rate": 2.666934455791007e-06,
+      "loss": 0.016,
+      "step": 6640
+    },
+    {
+      "epoch": 0.24017624963883272,
+      "grad_norm": 0.70939384875763,
+      "learning_rate": 2.6658311962638737e-06,
+      "loss": 0.0161,
+      "step": 6650
+    },
+    {
+      "epoch": 0.24053741693152267,
+      "grad_norm": 0.5079506605852535,
+      "learning_rate": 2.664726341517534e-06,
+      "loss": 0.0175,
+      "step": 6660
+    },
+    {
+      "epoch": 0.24089858422421265,
+      "grad_norm": 0.7281859197096429,
+      "learning_rate": 2.6636198930637725e-06,
+      "loss": 0.0164,
+      "step": 6670
+    },
+    {
+      "epoch": 0.24125975151690263,
+      "grad_norm": 1.0804508366870482,
+      "learning_rate": 2.662511852416554e-06,
+      "loss": 0.0171,
+      "step": 6680
+    },
+    {
+      "epoch": 0.2416209188095926,
+      "grad_norm": 0.7107643197021816,
+      "learning_rate": 2.661402221092023e-06,
+      "loss": 0.0164,
+      "step": 6690
+    },
+    {
+      "epoch": 0.24198208610228258,
+      "grad_norm": 0.5997547642089863,
+      "learning_rate": 2.6602910006084993e-06,
+      "loss": 0.0167,
+      "step": 6700
+    },
+    {
+      "epoch": 0.24234325339497256,
+      "grad_norm": 0.7509234530469292,
+      "learning_rate": 2.6591781924864773e-06,
+      "loss": 0.0167,
+      "step": 6710
+    },
+    {
+      "epoch": 0.24270442068766251,
+      "grad_norm": 1.185906345195998,
+      "learning_rate": 2.6580637982486255e-06,
+      "loss": 0.0156,
+      "step": 6720
+    },
+    {
+      "epoch": 0.2430655879803525,
+      "grad_norm": 0.6318196409959989,
+      "learning_rate": 2.6569478194197796e-06,
+      "loss": 0.0171,
+      "step": 6730
+    },
+    {
+      "epoch": 0.24342675527304247,
+      "grad_norm": 0.8018865518685864,
+      "learning_rate": 2.6558302575269473e-06,
+      "loss": 0.0164,
+      "step": 6740
+    },
+    {
+      "epoch": 0.24378792256573245,
+      "grad_norm": 0.8510908665022394,
+      "learning_rate": 2.654711114099298e-06,
+      "loss": 0.0173,
+      "step": 6750
+    },
+    {
+      "epoch": 0.24414908985842243,
+      "grad_norm": 0.5948819872512621,
+      "learning_rate": 2.6535903906681705e-06,
+      "loss": 0.0152,
+      "step": 6760
+    },
+    {
+      "epoch": 0.2445102571511124,
+      "grad_norm": 0.8052567179450079,
+      "learning_rate": 2.6524680887670598e-06,
+      "loss": 0.0174,
+      "step": 6770
+    },
+    {
+      "epoch": 0.24487142444380236,
+      "grad_norm": 0.7694136750104158,
+      "learning_rate": 2.6513442099316245e-06,
+      "loss": 0.0173,
+      "step": 6780
+    },
+    {
+      "epoch": 0.24523259173649234,
+      "grad_norm": 0.874974675750873,
+      "learning_rate": 2.6502187556996806e-06,
+      "loss": 0.0165,
+      "step": 6790
+    },
+    {
+      "epoch": 0.24559375902918232,
+      "grad_norm": 0.6675419290140153,
+      "learning_rate": 2.6490917276111986e-06,
+      "loss": 0.0165,
+      "step": 6800
+    },
+    {
+      "epoch": 0.2459549263218723,
+      "grad_norm": 1.2980616080258962,
+      "learning_rate": 2.647963127208303e-06,
+      "loss": 0.0171,
+      "step": 6810
+    },
+    {
+      "epoch": 0.24631609361456228,
+      "grad_norm": 0.6123417152150482,
+      "learning_rate": 2.64683295603527e-06,
+      "loss": 0.0163,
+      "step": 6820
+    },
+    {
+      "epoch": 0.24667726090725223,
+      "grad_norm": 0.676946535055087,
+      "learning_rate": 2.6457012156385237e-06,
+      "loss": 0.0175,
+      "step": 6830
+    },
+    {
+      "epoch": 0.2470384281999422,
+      "grad_norm": 0.8023339588288672,
+      "learning_rate": 2.6445679075666377e-06,
+      "loss": 0.0165,
+      "step": 6840
+    },
+    {
+      "epoch": 0.24739959549263218,
+      "grad_norm": 0.6605712969817626,
+      "learning_rate": 2.643433033370329e-06,
+      "loss": 0.0169,
+      "step": 6850
+    },
+    {
+      "epoch": 0.24776076278532216,
+      "grad_norm": 0.558282600200642,
+      "learning_rate": 2.6422965946024583e-06,
+      "loss": 0.0175,
+      "step": 6860
+    },
+    {
+      "epoch": 0.24812193007801214,
+      "grad_norm": 0.679351734455463,
+      "learning_rate": 2.6411585928180262e-06,
+      "loss": 0.017,
+      "step": 6870
+    },
+    {
+      "epoch": 0.24848309737070212,
+      "grad_norm": 0.6207810396545327,
+      "learning_rate": 2.6400190295741736e-06,
+      "loss": 0.0166,
+      "step": 6880
+    },
+    {
+      "epoch": 0.24884426466339207,
+      "grad_norm": 0.7321761291013332,
+      "learning_rate": 2.638877906430176e-06,
+      "loss": 0.0159,
+      "step": 6890
+    },
+    {
+      "epoch": 0.24920543195608205,
+      "grad_norm": 0.575863477224899,
+      "learning_rate": 2.637735224947445e-06,
+      "loss": 0.0168,
+      "step": 6900
+    },
+    {
+      "epoch": 0.24956659924877203,
+      "grad_norm": 0.6707778981832054,
+      "learning_rate": 2.636590986689523e-06,
+      "loss": 0.0161,
+      "step": 6910
+    },
+    {
+      "epoch": 0.249927766541462,
+      "grad_norm": 0.644525037107554,
+      "learning_rate": 2.635445193222085e-06,
+      "loss": 0.0164,
+      "step": 6920
+    },
+    {
+      "epoch": 0.25028893383415196,
+      "grad_norm": 0.635806005700264,
+      "learning_rate": 2.63429784611293e-06,
+      "loss": 0.0166,
+      "step": 6930
+    },
+    {
+      "epoch": 0.25065010112684194,
+      "grad_norm": 0.5062149953690753,
+      "learning_rate": 2.6331489469319875e-06,
+      "loss": 0.0155,
+      "step": 6940
+    },
+    {
+      "epoch": 0.2510112684195319,
+      "grad_norm": 0.5943483964550672,
+      "learning_rate": 2.631998497251307e-06,
+      "loss": 0.0174,
+      "step": 6950
+    },
+    {
+      "epoch": 0.2513724357122219,
+      "grad_norm": 0.5790747104664716,
+      "learning_rate": 2.630846498645062e-06,
+      "loss": 0.0162,
+      "step": 6960
+    },
+    {
+      "epoch": 0.2517336030049119,
+      "grad_norm": 0.5109709214912981,
+      "learning_rate": 2.6296929526895435e-06,
+      "loss": 0.0167,
+      "step": 6970
+    },
+    {
+      "epoch": 0.25209477029760186,
+      "grad_norm": 0.5563034999570683,
+      "learning_rate": 2.628537860963161e-06,
+      "loss": 0.0164,
+      "step": 6980
+    },
+    {
+      "epoch": 0.25245593759029183,
+      "grad_norm": 0.6384514435369784,
+      "learning_rate": 2.6273812250464394e-06,
+      "loss": 0.0154,
+      "step": 6990
+    },
+    {
+      "epoch": 0.2528171048829818,
+      "grad_norm": 0.5796060823966428,
+      "learning_rate": 2.626223046522015e-06,
+      "loss": 0.0161,
+      "step": 7000
+    },
+    {
+      "epoch": 0.2528171048829818,
+      "eval_loss": 0.1835169792175293,
+      "eval_runtime": 42.5445,
+      "eval_samples_per_second": 96.276,
+      "eval_steps_per_second": 0.188,
+      "step": 7000
+    },
+    {
+      "epoch": 0.2531782721756718,
+      "grad_norm": 0.6360531559297832,
+      "learning_rate": 2.625063326974636e-06,
+      "loss": 0.0157,
+      "step": 7010
+    },
+    {
+      "epoch": 0.25353943946836177,
+      "grad_norm": 0.7851435973840507,
+      "learning_rate": 2.62390206799116e-06,
+      "loss": 0.0156,
+      "step": 7020
+    },
+    {
+      "epoch": 0.2539006067610517,
+      "grad_norm": 0.5877791206613282,
+      "learning_rate": 2.622739271160548e-06,
+      "loss": 0.0154,
+      "step": 7030
+    },
+    {
+      "epoch": 0.2542617740537417,
+      "grad_norm": 0.6492383368751423,
+      "learning_rate": 2.621574938073869e-06,
+      "loss": 0.0153,
+      "step": 7040
+    },
+    {
+      "epoch": 0.25462294134643165,
+      "grad_norm": 0.9596025894693375,
+      "learning_rate": 2.6204090703242917e-06,
+      "loss": 0.0153,
+      "step": 7050
+    },
+    {
+      "epoch": 0.25498410863912163,
+      "grad_norm": 0.5067788993590288,
+      "learning_rate": 2.619241669507085e-06,
+      "loss": 0.016,
+      "step": 7060
+    },
+    {
+      "epoch": 0.2553452759318116,
+      "grad_norm": 0.5177299337080915,
+      "learning_rate": 2.6180727372196167e-06,
+      "loss": 0.0177,
+      "step": 7070
+    },
+    {
+      "epoch": 0.2557064432245016,
+      "grad_norm": 0.5507875568857453,
+      "learning_rate": 2.6169022750613482e-06,
+      "loss": 0.0144,
+      "step": 7080
+    },
+    {
+      "epoch": 0.25606761051719157,
+      "grad_norm": 0.5583323292160006,
+      "learning_rate": 2.615730284633836e-06,
+      "loss": 0.0155,
+      "step": 7090
+    },
+    {
+      "epoch": 0.25642877780988155,
+      "grad_norm": 0.4867248109605541,
+      "learning_rate": 2.6145567675407267e-06,
+      "loss": 0.0158,
+      "step": 7100
+    },
+    {
+      "epoch": 0.2567899451025715,
+      "grad_norm": 0.5425235066541316,
+      "learning_rate": 2.6133817253877568e-06,
+      "loss": 0.0167,
+      "step": 7110
+    },
+    {
+      "epoch": 0.2571511123952615,
+      "grad_norm": 0.4852630624037428,
+      "learning_rate": 2.6122051597827486e-06,
+      "loss": 0.0158,
+      "step": 7120
+    },
+    {
+      "epoch": 0.2575122796879515,
+      "grad_norm": 0.4465149859591923,
+      "learning_rate": 2.611027072335609e-06,
+      "loss": 0.0163,
+      "step": 7130
+    },
+    {
+      "epoch": 0.2578734469806414,
+      "grad_norm": 0.5140062373237979,
+      "learning_rate": 2.6098474646583273e-06,
+      "loss": 0.0152,
+      "step": 7140
+    },
+    {
+      "epoch": 0.2582346142733314,
+      "grad_norm": 0.5453233242239931,
+      "learning_rate": 2.6086663383649744e-06,
+      "loss": 0.0157,
+      "step": 7150
+    },
+    {
+      "epoch": 0.25859578156602137,
+      "grad_norm": 0.5225117479703978,
+      "learning_rate": 2.607483695071697e-06,
+      "loss": 0.0155,
+      "step": 7160
+    },
+    {
+      "epoch": 0.25895694885871134,
+      "grad_norm": 1.3210585850276513,
+      "learning_rate": 2.606299536396719e-06,
+      "loss": 0.0159,
+      "step": 7170
+    },
+    {
+      "epoch": 0.2593181161514013,
+      "grad_norm": 0.4794750067789923,
+      "learning_rate": 2.6051138639603366e-06,
+      "loss": 0.0161,
+      "step": 7180
+    },
+    {
+      "epoch": 0.2596792834440913,
+      "grad_norm": 0.6446354514715279,
+      "learning_rate": 2.603926679384919e-06,
+      "loss": 0.0152,
+      "step": 7190
+    },
+    {
+      "epoch": 0.2600404507367813,
+      "grad_norm": 0.7331314516206872,
+      "learning_rate": 2.602737984294903e-06,
+      "loss": 0.0163,
+      "step": 7200
+    },
+    {
+      "epoch": 0.26040161802947126,
+      "grad_norm": 0.6405054569251143,
+      "learning_rate": 2.6015477803167917e-06,
+      "loss": 0.0159,
+      "step": 7210
+    },
+    {
+      "epoch": 0.26076278532216124,
+      "grad_norm": 0.5215953193892171,
+      "learning_rate": 2.600356069079155e-06,
+      "loss": 0.0152,
+      "step": 7220
+    },
+    {
+      "epoch": 0.2611239526148512,
+      "grad_norm": 0.543951426260401,
+      "learning_rate": 2.599162852212624e-06,
+      "loss": 0.0166,
+      "step": 7230
+    },
+    {
+      "epoch": 0.2614851199075412,
+      "grad_norm": 0.5560505097976055,
+      "learning_rate": 2.5979681313498903e-06,
+      "loss": 0.0161,
+      "step": 7240
+    },
+    {
+      "epoch": 0.2618462872002311,
+      "grad_norm": 0.6126501232727051,
+      "learning_rate": 2.596771908125702e-06,
+      "loss": 0.0158,
+      "step": 7250
+    },
+    {
+      "epoch": 0.2622074544929211,
+      "grad_norm": 0.5861512438165949,
+      "learning_rate": 2.595574184176865e-06,
+      "loss": 0.0159,
+      "step": 7260
+    },
+    {
+      "epoch": 0.2625686217856111,
+      "grad_norm": 0.5678646492719516,
+      "learning_rate": 2.594374961142237e-06,
+      "loss": 0.0162,
+      "step": 7270
+    },
+    {
+      "epoch": 0.26292978907830106,
+      "grad_norm": 0.7085700649427306,
+      "learning_rate": 2.5931742406627285e-06,
+      "loss": 0.0171,
+      "step": 7280
+    },
+    {
+      "epoch": 0.26329095637099104,
+      "grad_norm": 0.7690826553497311,
+      "learning_rate": 2.5919720243812966e-06,
+      "loss": 0.016,
+      "step": 7290
+    },
+    {
+      "epoch": 0.263652123663681,
+      "grad_norm": 0.5640769589830286,
+      "learning_rate": 2.590768313942949e-06,
+      "loss": 0.0168,
+      "step": 7300
+    },
+    {
+      "epoch": 0.264013290956371,
+      "grad_norm": 0.5936409681828224,
+      "learning_rate": 2.589563110994733e-06,
+      "loss": 0.0151,
+      "step": 7310
+    },
+    {
+      "epoch": 0.264374458249061,
+      "grad_norm": 0.7624007339286821,
+      "learning_rate": 2.588356417185742e-06,
+      "loss": 0.0157,
+      "step": 7320
+    },
+    {
+      "epoch": 0.26473562554175095,
+      "grad_norm": 0.6806255534447179,
+      "learning_rate": 2.5871482341671073e-06,
+      "loss": 0.0163,
+      "step": 7330
+    },
+    {
+      "epoch": 0.26509679283444093,
+      "grad_norm": 0.5879464571426036,
+      "learning_rate": 2.5859385635919993e-06,
+      "loss": 0.0156,
+      "step": 7340
+    },
+    {
+      "epoch": 0.2654579601271309,
+      "grad_norm": 0.48606713566516346,
+      "learning_rate": 2.5847274071156225e-06,
+      "loss": 0.015,
+      "step": 7350
+    },
+    {
+      "epoch": 0.2658191274198209,
+      "grad_norm": 0.8034240169639064,
+      "learning_rate": 2.5835147663952153e-06,
+      "loss": 0.0161,
+      "step": 7360
+    },
+    {
+      "epoch": 0.2661802947125108,
+      "grad_norm": 0.7293206043184376,
+      "learning_rate": 2.5823006430900465e-06,
+      "loss": 0.016,
+      "step": 7370
+    },
+    {
+      "epoch": 0.2665414620052008,
+      "grad_norm": 0.7983112295743728,
+      "learning_rate": 2.581085038861415e-06,
+      "loss": 0.0158,
+      "step": 7380
+    },
+    {
+      "epoch": 0.26690262929789077,
+      "grad_norm": 0.6346402851821236,
+      "learning_rate": 2.579867955372644e-06,
+      "loss": 0.0164,
+      "step": 7390
+    },
+    {
+      "epoch": 0.26726379659058075,
+      "grad_norm": 0.64731614149756,
+      "learning_rate": 2.578649394289082e-06,
+      "loss": 0.0153,
+      "step": 7400
+    },
+    {
+      "epoch": 0.26762496388327073,
+      "grad_norm": 0.5097560434805138,
+      "learning_rate": 2.5774293572780993e-06,
+      "loss": 0.0154,
+      "step": 7410
+    },
+    {
+      "epoch": 0.2679861311759607,
+      "grad_norm": 0.547469623807854,
+      "learning_rate": 2.576207846009085e-06,
+      "loss": 0.0156,
+      "step": 7420
+    },
+    {
+      "epoch": 0.2683472984686507,
+      "grad_norm": 0.7276444497685198,
+      "learning_rate": 2.5749848621534466e-06,
+      "loss": 0.0163,
+      "step": 7430
+    },
+    {
+      "epoch": 0.26870846576134066,
+      "grad_norm": 1.4866988937683931,
+      "learning_rate": 2.5737604073846053e-06,
+      "loss": 0.016,
+      "step": 7440
+    },
+    {
+      "epoch": 0.26906963305403064,
+      "grad_norm": 0.6596414606706901,
+      "learning_rate": 2.572534483377996e-06,
+      "loss": 0.016,
+      "step": 7450
+    },
+    {
+      "epoch": 0.2694308003467206,
+      "grad_norm": 0.4746426910742506,
+      "learning_rate": 2.571307091811063e-06,
+      "loss": 0.0162,
+      "step": 7460
+    },
+    {
+      "epoch": 0.2697919676394106,
+      "grad_norm": 0.5775905084177051,
+      "learning_rate": 2.57007823436326e-06,
+      "loss": 0.0163,
+      "step": 7470
+    },
+    {
+      "epoch": 0.2701531349321005,
+      "grad_norm": 0.8190202750593905,
+      "learning_rate": 2.568847912716045e-06,
+      "loss": 0.0154,
+      "step": 7480
+    },
+    {
+      "epoch": 0.2705143022247905,
+      "grad_norm": 0.4886726304082907,
+      "learning_rate": 2.5676161285528803e-06,
+      "loss": 0.0157,
+      "step": 7490
+    },
+    {
+      "epoch": 0.2708754695174805,
+      "grad_norm": 0.5701787278840702,
+      "learning_rate": 2.5663828835592297e-06,
+      "loss": 0.015,
+      "step": 7500
+    },
+    {
+      "epoch": 0.27123663681017046,
+      "grad_norm": 0.4660995912536972,
+      "learning_rate": 2.565148179422555e-06,
+      "loss": 0.0161,
+      "step": 7510
+    },
+    {
+      "epoch": 0.27159780410286044,
+      "grad_norm": 1.1811483629101052,
+      "learning_rate": 2.563912017832315e-06,
+      "loss": 0.0159,
+      "step": 7520
+    },
+    {
+      "epoch": 0.2719589713955504,
+      "grad_norm": 0.5647814474849634,
+      "learning_rate": 2.562674400479963e-06,
+      "loss": 0.0161,
+      "step": 7530
+    },
+    {
+      "epoch": 0.2723201386882404,
+      "grad_norm": 0.7931059059235102,
+      "learning_rate": 2.5614353290589436e-06,
+      "loss": 0.0153,
+      "step": 7540
+    },
+    {
+      "epoch": 0.2726813059809304,
+      "grad_norm": 0.6459120841023008,
+      "learning_rate": 2.560194805264692e-06,
+      "loss": 0.0165,
+      "step": 7550
+    },
+    {
+      "epoch": 0.27304247327362036,
+      "grad_norm": 0.5598073250234038,
+      "learning_rate": 2.558952830794629e-06,
+      "loss": 0.0162,
+      "step": 7560
+    },
+    {
+      "epoch": 0.27340364056631034,
+      "grad_norm": 0.644561560273708,
+      "learning_rate": 2.5577094073481627e-06,
+      "loss": 0.0162,
+      "step": 7570
+    },
+    {
+      "epoch": 0.2737648078590003,
+      "grad_norm": 0.5346299465766857,
+      "learning_rate": 2.5564645366266817e-06,
+      "loss": 0.0163,
+      "step": 7580
+    },
+    {
+      "epoch": 0.27412597515169024,
+      "grad_norm": 0.6429388854065141,
+      "learning_rate": 2.555218220333556e-06,
+      "loss": 0.0162,
+      "step": 7590
+    },
+    {
+      "epoch": 0.2744871424443802,
+      "grad_norm": 0.7473126984986002,
+      "learning_rate": 2.5539704601741343e-06,
+      "loss": 0.0142,
+      "step": 7600
+    },
+    {
+      "epoch": 0.2748483097370702,
+      "grad_norm": 0.5600846827352068,
+      "learning_rate": 2.5527212578557384e-06,
+      "loss": 0.0151,
+      "step": 7610
+    },
+    {
+      "epoch": 0.2752094770297602,
+      "grad_norm": 0.5706860049658498,
+      "learning_rate": 2.551470615087667e-06,
+      "loss": 0.016,
+      "step": 7620
+    },
+    {
+      "epoch": 0.27557064432245015,
+      "grad_norm": 0.5982059555167951,
+      "learning_rate": 2.550218533581186e-06,
+      "loss": 0.016,
+      "step": 7630
+    },
+    {
+      "epoch": 0.27593181161514013,
+      "grad_norm": 0.6862902934859217,
+      "learning_rate": 2.548965015049533e-06,
+      "loss": 0.0149,
+      "step": 7640
+    },
+    {
+      "epoch": 0.2762929789078301,
+      "grad_norm": 0.5751546919427202,
+      "learning_rate": 2.5477100612079107e-06,
+      "loss": 0.0162,
+      "step": 7650
+    },
+    {
+      "epoch": 0.2766541462005201,
+      "grad_norm": 0.4747969735754964,
+      "learning_rate": 2.5464536737734856e-06,
+      "loss": 0.015,
+      "step": 7660
+    },
+    {
+      "epoch": 0.27701531349321007,
+      "grad_norm": 0.7107651033610941,
+      "learning_rate": 2.5451958544653855e-06,
+      "loss": 0.0155,
+      "step": 7670
+    },
+    {
+      "epoch": 0.27737648078590005,
+      "grad_norm": 0.6310706124299429,
+      "learning_rate": 2.5439366050046985e-06,
+      "loss": 0.0146,
+      "step": 7680
+    },
+    {
+      "epoch": 0.27773764807859,
+      "grad_norm": 0.5414754898900336,
+      "learning_rate": 2.5426759271144687e-06,
+      "loss": 0.015,
+      "step": 7690
+    },
+    {
+      "epoch": 0.27809881537127995,
+      "grad_norm": 0.511025383843564,
+      "learning_rate": 2.5414138225196952e-06,
+      "loss": 0.0155,
+      "step": 7700
+    },
+    {
+      "epoch": 0.27845998266396993,
+      "grad_norm": 0.5633260671079873,
+      "learning_rate": 2.5401502929473296e-06,
+      "loss": 0.0145,
+      "step": 7710
+    },
+    {
+      "epoch": 0.2788211499566599,
+      "grad_norm": 0.6036775849068622,
+      "learning_rate": 2.538885340126272e-06,
+      "loss": 0.015,
+      "step": 7720
+    },
+    {
+      "epoch": 0.2791823172493499,
+      "grad_norm": 0.4978635985774278,
+      "learning_rate": 2.537618965787371e-06,
+      "loss": 0.0162,
+      "step": 7730
+    },
+    {
+      "epoch": 0.27954348454203987,
+      "grad_norm": 0.8130021509575047,
+      "learning_rate": 2.53635117166342e-06,
+      "loss": 0.015,
+      "step": 7740
+    },
+    {
+      "epoch": 0.27990465183472985,
+      "grad_norm": 0.4553723719384611,
+      "learning_rate": 2.5350819594891556e-06,
+      "loss": 0.0139,
+      "step": 7750
+    },
+    {
+      "epoch": 0.2802658191274198,
+      "grad_norm": 0.5895647387190823,
+      "learning_rate": 2.533811331001254e-06,
+      "loss": 0.015,
+      "step": 7760
+    },
+    {
+      "epoch": 0.2806269864201098,
+      "grad_norm": 0.4680510602759471,
+      "learning_rate": 2.5325392879383298e-06,
+      "loss": 0.0157,
+      "step": 7770
+    },
+    {
+      "epoch": 0.2809881537127998,
+      "grad_norm": 0.5003532181130818,
+      "learning_rate": 2.5312658320409324e-06,
+      "loss": 0.0155,
+      "step": 7780
+    },
+    {
+      "epoch": 0.28134932100548976,
+      "grad_norm": 0.45920655174112834,
+      "learning_rate": 2.529990965051546e-06,
+      "loss": 0.0147,
+      "step": 7790
+    },
+    {
+      "epoch": 0.28171048829817974,
+      "grad_norm": 0.38279157080632203,
+      "learning_rate": 2.528714688714584e-06,
+      "loss": 0.0161,
+      "step": 7800
+    },
+    {
+      "epoch": 0.28207165559086966,
+      "grad_norm": 0.5025649766870682,
+      "learning_rate": 2.5274370047763884e-06,
+      "loss": 0.0154,
+      "step": 7810
+    },
+    {
+      "epoch": 0.28243282288355964,
+      "grad_norm": 0.43130060328403746,
+      "learning_rate": 2.5261579149852294e-06,
+      "loss": 0.0149,
+      "step": 7820
+    },
+    {
+      "epoch": 0.2827939901762496,
+      "grad_norm": 0.4459636508519494,
+      "learning_rate": 2.5248774210912978e-06,
+      "loss": 0.0158,
+      "step": 7830
+    },
+    {
+      "epoch": 0.2831551574689396,
+      "grad_norm": 0.4462206084335187,
+      "learning_rate": 2.523595524846707e-06,
+      "loss": 0.0155,
+      "step": 7840
+    },
+    {
+      "epoch": 0.2835163247616296,
+      "grad_norm": 0.46820643476392676,
+      "learning_rate": 2.5223122280054895e-06,
+      "loss": 0.0154,
+      "step": 7850
+    },
+    {
+      "epoch": 0.28387749205431956,
+      "grad_norm": 0.5307119711651603,
+      "learning_rate": 2.5210275323235945e-06,
+      "loss": 0.0148,
+      "step": 7860
+    },
+    {
+      "epoch": 0.28423865934700954,
+      "grad_norm": 0.5535656653898645,
+      "learning_rate": 2.5197414395588836e-06,
+      "loss": 0.0156,
+      "step": 7870
+    },
+    {
+      "epoch": 0.2845998266396995,
+      "grad_norm": 0.5266705164808446,
+      "learning_rate": 2.518453951471132e-06,
+      "loss": 0.0144,
+      "step": 7880
+    },
+    {
+      "epoch": 0.2849609939323895,
+      "grad_norm": 0.4143176782626514,
+      "learning_rate": 2.517165069822023e-06,
+      "loss": 0.0153,
+      "step": 7890
+    },
+    {
+      "epoch": 0.2853221612250795,
+      "grad_norm": 0.5370035353268835,
+      "learning_rate": 2.5158747963751483e-06,
+      "loss": 0.0155,
+      "step": 7900
+    },
+    {
+      "epoch": 0.28568332851776945,
+      "grad_norm": 0.5174674712299218,
+      "learning_rate": 2.5145831328960006e-06,
+      "loss": 0.0153,
+      "step": 7910
+    },
+    {
+      "epoch": 0.28604449581045943,
+      "grad_norm": 0.698858080708865,
+      "learning_rate": 2.513290081151978e-06,
+      "loss": 0.0151,
+      "step": 7920
+    },
+    {
+      "epoch": 0.28640566310314935,
+      "grad_norm": 0.5929900933475408,
+      "learning_rate": 2.511995642912377e-06,
+      "loss": 0.0164,
+      "step": 7930
+    },
+    {
+      "epoch": 0.28676683039583933,
+      "grad_norm": 0.543961182849162,
+      "learning_rate": 2.510699819948391e-06,
+      "loss": 0.0156,
+      "step": 7940
+    },
+    {
+      "epoch": 0.2871279976885293,
+      "grad_norm": 0.7566717148590388,
+      "learning_rate": 2.509402614033108e-06,
+      "loss": 0.0145,
+      "step": 7950
+    },
+    {
+      "epoch": 0.2874891649812193,
+      "grad_norm": 0.47816219452845743,
+      "learning_rate": 2.508104026941509e-06,
+      "loss": 0.0153,
+      "step": 7960
+    },
+    {
+      "epoch": 0.28785033227390927,
+      "grad_norm": 0.9957838956184404,
+      "learning_rate": 2.506804060450464e-06,
+      "loss": 0.015,
+      "step": 7970
+    },
+    {
+      "epoch": 0.28821149956659925,
+      "grad_norm": 0.9558100561971354,
+      "learning_rate": 2.5055027163387324e-06,
+      "loss": 0.0153,
+      "step": 7980
+    },
+    {
+      "epoch": 0.28857266685928923,
+      "grad_norm": 0.5785330474692574,
+      "learning_rate": 2.5041999963869552e-06,
+      "loss": 0.0151,
+      "step": 7990
+    },
+    {
+      "epoch": 0.2889338341519792,
+      "grad_norm": 0.7234871906155617,
+      "learning_rate": 2.5028959023776593e-06,
+      "loss": 0.0161,
+      "step": 8000
+    },
+    {
+      "epoch": 0.2889338341519792,
+      "eval_loss": 0.19060981273651123,
+      "eval_runtime": 40.0575,
+      "eval_samples_per_second": 102.253,
+      "eval_steps_per_second": 0.2,
+      "step": 8000
+    },
+    {
+      "epoch": 0.2892950014446692,
+      "grad_norm": 2.6773462998157407,
+      "learning_rate": 2.5015904360952496e-06,
+      "loss": 0.0157,
+      "step": 8010
+    },
+    {
+      "epoch": 0.28965616873735917,
+      "grad_norm": 0.5289218124639341,
+      "learning_rate": 2.50028359932601e-06,
+      "loss": 0.0149,
+      "step": 8020
+    },
+    {
+      "epoch": 0.29001733603004914,
+      "grad_norm": 0.6194266269140976,
+      "learning_rate": 2.498975393858099e-06,
+      "loss": 0.0159,
+      "step": 8030
+    },
+    {
+      "epoch": 0.29037850332273907,
+      "grad_norm": 0.6348305335951918,
+      "learning_rate": 2.4976658214815482e-06,
+      "loss": 0.0161,
+      "step": 8040
+    },
+    {
+      "epoch": 0.29073967061542905,
+      "grad_norm": 0.5067449833924159,
+      "learning_rate": 2.496354883988259e-06,
+      "loss": 0.0149,
+      "step": 8050
+    },
+    {
+      "epoch": 0.291100837908119,
+      "grad_norm": 0.4959419494905272,
+      "learning_rate": 2.4950425831720017e-06,
+      "loss": 0.0149,
+      "step": 8060
+    },
+    {
+      "epoch": 0.291462005200809,
+      "grad_norm": 0.5189977022859331,
+      "learning_rate": 2.4937289208284107e-06,
+      "loss": 0.015,
+      "step": 8070
+    },
+    {
+      "epoch": 0.291823172493499,
+      "grad_norm": 0.6604502136022056,
+      "learning_rate": 2.4924138987549845e-06,
+      "loss": 0.014,
+      "step": 8080
+    },
+    {
+      "epoch": 0.29218433978618896,
+      "grad_norm": 0.4755043766686694,
+      "learning_rate": 2.4910975187510813e-06,
+      "loss": 0.0153,
+      "step": 8090
+    },
+    {
+      "epoch": 0.29254550707887894,
+      "grad_norm": 0.3625950120946837,
+      "learning_rate": 2.4897797826179186e-06,
+      "loss": 0.0154,
+      "step": 8100
+    },
+    {
+      "epoch": 0.2929066743715689,
+      "grad_norm": 0.6776840017755626,
+      "learning_rate": 2.4884606921585685e-06,
+      "loss": 0.0149,
+      "step": 8110
+    },
+    {
+      "epoch": 0.2932678416642589,
+      "grad_norm": 0.4600457583456973,
+      "learning_rate": 2.4871402491779566e-06,
+      "loss": 0.0153,
+      "step": 8120
+    },
+    {
+      "epoch": 0.2936290089569489,
+      "grad_norm": 0.5056373055215263,
+      "learning_rate": 2.485818455482858e-06,
+      "loss": 0.0156,
+      "step": 8130
+    },
+    {
+      "epoch": 0.29399017624963886,
+      "grad_norm": 0.6188605655545764,
+      "learning_rate": 2.4844953128818986e-06,
+      "loss": 0.0157,
+      "step": 8140
+    },
+    {
+      "epoch": 0.2943513435423288,
+      "grad_norm": 0.45729626067468,
+      "learning_rate": 2.483170823185548e-06,
+      "loss": 0.016,
+      "step": 8150
+    },
+    {
+      "epoch": 0.29471251083501876,
+      "grad_norm": 0.5571185221937847,
+      "learning_rate": 2.4818449882061185e-06,
+      "loss": 0.014,
+      "step": 8160
+    },
+    {
+      "epoch": 0.29507367812770874,
+      "grad_norm": 0.5153585552228814,
+      "learning_rate": 2.4805178097577653e-06,
+      "loss": 0.0155,
+      "step": 8170
+    },
+    {
+      "epoch": 0.2954348454203987,
+      "grad_norm": 0.48696589199654416,
+      "learning_rate": 2.4791892896564805e-06,
+      "loss": 0.0147,
+      "step": 8180
+    },
+    {
+      "epoch": 0.2957960127130887,
+      "grad_norm": 0.49164250869449044,
+      "learning_rate": 2.477859429720093e-06,
+      "loss": 0.0149,
+      "step": 8190
+    },
+    {
+      "epoch": 0.2961571800057787,
+      "grad_norm": 1.0428947770961319,
+      "learning_rate": 2.4765282317682622e-06,
+      "loss": 0.0154,
+      "step": 8200
+    },
+    {
+      "epoch": 0.29651834729846865,
+      "grad_norm": 1.2431359508766013,
+      "learning_rate": 2.4751956976224824e-06,
+      "loss": 0.0158,
+      "step": 8210
+    },
+    {
+      "epoch": 0.29687951459115863,
+      "grad_norm": 0.40363369966726786,
+      "learning_rate": 2.473861829106073e-06,
+      "loss": 0.015,
+      "step": 8220
+    },
+    {
+      "epoch": 0.2972406818838486,
+      "grad_norm": 0.5485252358475459,
+      "learning_rate": 2.4725266280441813e-06,
+      "loss": 0.0159,
+      "step": 8230
+    },
+    {
+      "epoch": 0.2976018491765386,
+      "grad_norm": 0.4840508787177018,
+      "learning_rate": 2.4711900962637765e-06,
+      "loss": 0.0146,
+      "step": 8240
+    },
+    {
+      "epoch": 0.29796301646922857,
+      "grad_norm": 0.6711641297143095,
+      "learning_rate": 2.4698522355936497e-06,
+      "loss": 0.0144,
+      "step": 8250
+    },
+    {
+      "epoch": 0.2983241837619185,
+      "grad_norm": 0.7970530741905051,
+      "learning_rate": 2.4685130478644096e-06,
+      "loss": 0.0139,
+      "step": 8260
+    },
+    {
+      "epoch": 0.29868535105460847,
+      "grad_norm": 0.6249155478091548,
+      "learning_rate": 2.467172534908481e-06,
+      "loss": 0.0144,
+      "step": 8270
+    },
+    {
+      "epoch": 0.29904651834729845,
+      "grad_norm": 0.9870211501986811,
+      "learning_rate": 2.4658306985601017e-06,
+      "loss": 0.0142,
+      "step": 8280
+    },
+    {
+      "epoch": 0.29940768563998843,
+      "grad_norm": 0.7717332364198884,
+      "learning_rate": 2.4644875406553214e-06,
+      "loss": 0.0155,
+      "step": 8290
+    },
+    {
+      "epoch": 0.2997688529326784,
+      "grad_norm": 0.4655854872756121,
+      "learning_rate": 2.463143063031997e-06,
+      "loss": 0.0153,
+      "step": 8300
+    },
+    {
+      "epoch": 0.3001300202253684,
+      "grad_norm": 0.5398783093838468,
+      "learning_rate": 2.4617972675297905e-06,
+      "loss": 0.0139,
+      "step": 8310
+    },
+    {
+      "epoch": 0.30049118751805837,
+      "grad_norm": 0.4402722060533199,
+      "learning_rate": 2.4604501559901698e-06,
+      "loss": 0.0151,
+      "step": 8320
+    },
+    {
+      "epoch": 0.30085235481074835,
+      "grad_norm": 1.2336785932370649,
+      "learning_rate": 2.459101730256401e-06,
+      "loss": 0.0156,
+      "step": 8330
+    },
+    {
+      "epoch": 0.3012135221034383,
+      "grad_norm": 0.4507557242955374,
+      "learning_rate": 2.4577519921735498e-06,
+      "loss": 0.0157,
+      "step": 8340
+    },
+    {
+      "epoch": 0.3015746893961283,
+      "grad_norm": 0.5006992818810972,
+      "learning_rate": 2.4564009435884766e-06,
+      "loss": 0.014,
+      "step": 8350
+    },
+    {
+      "epoch": 0.3019358566888183,
+      "grad_norm": 0.5656610755119181,
+      "learning_rate": 2.455048586349836e-06,
+      "loss": 0.0145,
+      "step": 8360
+    },
+    {
+      "epoch": 0.30229702398150826,
+      "grad_norm": 0.39795237528101685,
+      "learning_rate": 2.4536949223080726e-06,
+      "loss": 0.0151,
+      "step": 8370
+    },
+    {
+      "epoch": 0.3026581912741982,
+      "grad_norm": 0.5441566138830236,
+      "learning_rate": 2.452339953315419e-06,
+      "loss": 0.0152,
+      "step": 8380
+    },
+    {
+      "epoch": 0.30301935856688816,
+      "grad_norm": 0.6547713323736791,
+      "learning_rate": 2.4509836812258943e-06,
+      "loss": 0.0142,
+      "step": 8390
+    },
+    {
+      "epoch": 0.30338052585957814,
+      "grad_norm": 0.43102787129323133,
+      "learning_rate": 2.4496261078952996e-06,
+      "loss": 0.0154,
+      "step": 8400
+    },
+    {
+      "epoch": 0.3037416931522681,
+      "grad_norm": 0.8241821684013333,
+      "learning_rate": 2.448267235181217e-06,
+      "loss": 0.0149,
+      "step": 8410
+    },
+    {
+      "epoch": 0.3041028604449581,
+      "grad_norm": 0.42296164831580063,
+      "learning_rate": 2.4469070649430056e-06,
+      "loss": 0.0148,
+      "step": 8420
+    },
+    {
+      "epoch": 0.3044640277376481,
+      "grad_norm": 0.47456691726928474,
+      "learning_rate": 2.4455455990418024e-06,
+      "loss": 0.0139,
+      "step": 8430
+    },
+    {
+      "epoch": 0.30482519503033806,
+      "grad_norm": 0.4995074170067591,
+      "learning_rate": 2.4441828393405144e-06,
+      "loss": 0.0142,
+      "step": 8440
+    },
+    {
+      "epoch": 0.30518636232302804,
+      "grad_norm": 0.4747186273173266,
+      "learning_rate": 2.442818787703821e-06,
+      "loss": 0.0149,
+      "step": 8450
+    },
+    {
+      "epoch": 0.305547529615718,
+      "grad_norm": 0.7087961635715649,
+      "learning_rate": 2.441453445998168e-06,
+      "loss": 0.0141,
+      "step": 8460
+    },
+    {
+      "epoch": 0.305908696908408,
+      "grad_norm": 0.4123772806809425,
+      "learning_rate": 2.4400868160917673e-06,
+      "loss": 0.0148,
+      "step": 8470
+    },
+    {
+      "epoch": 0.306269864201098,
+      "grad_norm": 0.5740396281891088,
+      "learning_rate": 2.438718899854593e-06,
+      "loss": 0.0151,
+      "step": 8480
+    },
+    {
+      "epoch": 0.3066310314937879,
+      "grad_norm": 0.5490893746367949,
+      "learning_rate": 2.437349699158379e-06,
+      "loss": 0.0154,
+      "step": 8490
+    },
+    {
+      "epoch": 0.3069921987864779,
+      "grad_norm": 0.5133083345648322,
+      "learning_rate": 2.4359792158766174e-06,
+      "loss": 0.0144,
+      "step": 8500
+    },
+    {
+      "epoch": 0.30735336607916786,
+      "grad_norm": 0.43594515627118724,
+      "learning_rate": 2.4346074518845552e-06,
+      "loss": 0.0153,
+      "step": 8510
+    },
+    {
+      "epoch": 0.30771453337185783,
+      "grad_norm": 0.6645329942668388,
+      "learning_rate": 2.433234409059192e-06,
+      "loss": 0.0142,
+      "step": 8520
+    },
+    {
+      "epoch": 0.3080757006645478,
+      "grad_norm": 0.5069324244505184,
+      "learning_rate": 2.431860089279276e-06,
+      "loss": 0.0153,
+      "step": 8530
+    },
+    {
+      "epoch": 0.3084368679572378,
+      "grad_norm": 0.41872535858494814,
+      "learning_rate": 2.430484494425304e-06,
+      "loss": 0.0145,
+      "step": 8540
+    },
+    {
+      "epoch": 0.30879803524992777,
+      "grad_norm": 0.7050968833407338,
+      "learning_rate": 2.429107626379517e-06,
+      "loss": 0.0152,
+      "step": 8550
+    },
+    {
+      "epoch": 0.30915920254261775,
+      "grad_norm": 0.4458180003404862,
+      "learning_rate": 2.427729487025898e-06,
+      "loss": 0.014,
+      "step": 8560
+    },
+    {
+      "epoch": 0.30952036983530773,
+      "grad_norm": 0.4350611116374495,
+      "learning_rate": 2.4263500782501705e-06,
+      "loss": 0.0152,
+      "step": 8570
+    },
+    {
+      "epoch": 0.3098815371279977,
+      "grad_norm": 0.39768289957890485,
+      "learning_rate": 2.4249694019397936e-06,
+      "loss": 0.0142,
+      "step": 8580
+    },
+    {
+      "epoch": 0.3102427044206877,
+      "grad_norm": 0.5067606751169844,
+      "learning_rate": 2.4235874599839612e-06,
+      "loss": 0.0147,
+      "step": 8590
+    },
+    {
+      "epoch": 0.3106038717133776,
+      "grad_norm": 0.39908044026981887,
+      "learning_rate": 2.4222042542735987e-06,
+      "loss": 0.0142,
+      "step": 8600
+    },
+    {
+      "epoch": 0.3109650390060676,
+      "grad_norm": 0.4170468769103302,
+      "learning_rate": 2.4208197867013625e-06,
+      "loss": 0.014,
+      "step": 8610
+    },
+    {
+      "epoch": 0.31132620629875757,
+      "grad_norm": 0.6607055101037482,
+      "learning_rate": 2.419434059161633e-06,
+      "loss": 0.0146,
+      "step": 8620
+    },
+    {
+      "epoch": 0.31168737359144755,
+      "grad_norm": 0.48772131368568983,
+      "learning_rate": 2.4180470735505167e-06,
+      "loss": 0.0144,
+      "step": 8630
+    },
+    {
+      "epoch": 0.3120485408841375,
+      "grad_norm": 0.37854651953374596,
+      "learning_rate": 2.4166588317658405e-06,
+      "loss": 0.0138,
+      "step": 8640
+    },
+    {
+      "epoch": 0.3124097081768275,
+      "grad_norm": 0.36595427292429555,
+      "learning_rate": 2.4152693357071508e-06,
+      "loss": 0.0138,
+      "step": 8650
+    },
+    {
+      "epoch": 0.3127708754695175,
+      "grad_norm": 0.5320282596243177,
+      "learning_rate": 2.413878587275709e-06,
+      "loss": 0.0138,
+      "step": 8660
+    },
+    {
+      "epoch": 0.31313204276220746,
+      "grad_norm": 0.37400201467961863,
+      "learning_rate": 2.412486588374491e-06,
+      "loss": 0.0143,
+      "step": 8670
+    },
+    {
+      "epoch": 0.31349321005489744,
+      "grad_norm": 0.4949290990843776,
+      "learning_rate": 2.411093340908184e-06,
+      "loss": 0.014,
+      "step": 8680
+    },
+    {
+      "epoch": 0.3138543773475874,
+      "grad_norm": 0.42788168160789064,
+      "learning_rate": 2.4096988467831843e-06,
+      "loss": 0.0144,
+      "step": 8690
+    },
+    {
+      "epoch": 0.3142155446402774,
+      "grad_norm": 0.4082316442113492,
+      "learning_rate": 2.4083031079075914e-06,
+      "loss": 0.0153,
+      "step": 8700
+    },
+    {
+      "epoch": 0.3145767119329673,
+      "grad_norm": 0.4381586806495369,
+      "learning_rate": 2.40690612619121e-06,
+      "loss": 0.0143,
+      "step": 8710
+    },
+    {
+      "epoch": 0.3149378792256573,
+      "grad_norm": 0.5890504303953396,
+      "learning_rate": 2.4055079035455456e-06,
+      "loss": 0.0142,
+      "step": 8720
+    },
+    {
+      "epoch": 0.3152990465183473,
+      "grad_norm": 0.38400116485931596,
+      "learning_rate": 2.4041084418838005e-06,
+      "loss": 0.0153,
+      "step": 8730
+    },
+    {
+      "epoch": 0.31566021381103726,
+      "grad_norm": 0.31565310244682315,
+      "learning_rate": 2.402707743120873e-06,
+      "loss": 0.0131,
+      "step": 8740
+    },
+    {
+      "epoch": 0.31602138110372724,
+      "grad_norm": 0.42451570122474475,
+      "learning_rate": 2.4013058091733546e-06,
+      "loss": 0.013,
+      "step": 8750
+    },
+    {
+      "epoch": 0.3163825483964172,
+      "grad_norm": 0.4341315053239454,
+      "learning_rate": 2.399902641959527e-06,
+      "loss": 0.0139,
+      "step": 8760
+    },
+    {
+      "epoch": 0.3167437156891072,
+      "grad_norm": 0.42199025190235323,
+      "learning_rate": 2.3984982433993566e-06,
+      "loss": 0.0143,
+      "step": 8770
+    },
+    {
+      "epoch": 0.3171048829817972,
+      "grad_norm": 0.3198427781155048,
+      "learning_rate": 2.397092615414499e-06,
+      "loss": 0.0142,
+      "step": 8780
+    },
+    {
+      "epoch": 0.31746605027448715,
+      "grad_norm": 0.3347634870031812,
+      "learning_rate": 2.3956857599282887e-06,
+      "loss": 0.0144,
+      "step": 8790
+    },
+    {
+      "epoch": 0.31782721756717713,
+      "grad_norm": 0.5406308100489219,
+      "learning_rate": 2.394277678865741e-06,
+      "loss": 0.0155,
+      "step": 8800
+    },
+    {
+      "epoch": 0.3181883848598671,
+      "grad_norm": 0.41291157598110384,
+      "learning_rate": 2.392868374153549e-06,
+      "loss": 0.0144,
+      "step": 8810
+    },
+    {
+      "epoch": 0.3185495521525571,
+      "grad_norm": 0.47241927550799645,
+      "learning_rate": 2.391457847720079e-06,
+      "loss": 0.014,
+      "step": 8820
+    },
+    {
+      "epoch": 0.318910719445247,
+      "grad_norm": 0.6708740378557301,
+      "learning_rate": 2.3900461014953693e-06,
+      "loss": 0.0136,
+      "step": 8830
+    },
+    {
+      "epoch": 0.319271886737937,
+      "grad_norm": 0.4867882357624585,
+      "learning_rate": 2.388633137411127e-06,
+      "loss": 0.0136,
+      "step": 8840
+    },
+    {
+      "epoch": 0.319633054030627,
+      "grad_norm": 0.3500353662520406,
+      "learning_rate": 2.387218957400726e-06,
+      "loss": 0.0141,
+      "step": 8850
+    },
+    {
+      "epoch": 0.31999422132331695,
+      "grad_norm": 0.5765561926582653,
+      "learning_rate": 2.3858035633992037e-06,
+      "loss": 0.0143,
+      "step": 8860
+    },
+    {
+      "epoch": 0.32035538861600693,
+      "grad_norm": 0.5407756684393482,
+      "learning_rate": 2.3843869573432584e-06,
+      "loss": 0.0153,
+      "step": 8870
+    },
+    {
+      "epoch": 0.3207165559086969,
+      "grad_norm": 0.39256560633094506,
+      "learning_rate": 2.3829691411712485e-06,
+      "loss": 0.0128,
+      "step": 8880
+    },
+    {
+      "epoch": 0.3210777232013869,
+      "grad_norm": 0.46466732730408933,
+      "learning_rate": 2.381550116823185e-06,
+      "loss": 0.0146,
+      "step": 8890
+    },
+    {
+      "epoch": 0.32143889049407687,
+      "grad_norm": 0.3540944735155153,
+      "learning_rate": 2.3801298862407357e-06,
+      "loss": 0.0141,
+      "step": 8900
+    },
+    {
+      "epoch": 0.32180005778676685,
+      "grad_norm": 1.2618192316578538,
+      "learning_rate": 2.3787084513672168e-06,
+      "loss": 0.0143,
+      "step": 8910
+    },
+    {
+      "epoch": 0.3221612250794568,
+      "grad_norm": 0.4363329437951931,
+      "learning_rate": 2.377285814147592e-06,
+      "loss": 0.0138,
+      "step": 8920
+    },
+    {
+      "epoch": 0.3225223923721468,
+      "grad_norm": 0.39197381660133873,
+      "learning_rate": 2.375861976528472e-06,
+      "loss": 0.0145,
+      "step": 8930
+    },
+    {
+      "epoch": 0.32288355966483673,
+      "grad_norm": 0.46484970676346976,
+      "learning_rate": 2.3744369404581074e-06,
+      "loss": 0.0141,
+      "step": 8940
+    },
+    {
+      "epoch": 0.3232447269575267,
+      "grad_norm": 0.5440409468208708,
+      "learning_rate": 2.373010707886392e-06,
+      "loss": 0.0146,
+      "step": 8950
+    },
+    {
+      "epoch": 0.3236058942502167,
+      "grad_norm": 0.41966281918675735,
+      "learning_rate": 2.3715832807648536e-06,
+      "loss": 0.0143,
+      "step": 8960
+    },
+    {
+      "epoch": 0.32396706154290666,
+      "grad_norm": 0.9009489780208302,
+      "learning_rate": 2.3701546610466566e-06,
+      "loss": 0.0148,
+      "step": 8970
+    },
+    {
+      "epoch": 0.32432822883559664,
+      "grad_norm": 0.43732518188508773,
+      "learning_rate": 2.368724850686596e-06,
+      "loss": 0.0138,
+      "step": 8980
+    },
+    {
+      "epoch": 0.3246893961282866,
+      "grad_norm": 0.35831032704078486,
+      "learning_rate": 2.3672938516410972e-06,
+      "loss": 0.0136,
+      "step": 8990
+    },
+    {
+      "epoch": 0.3250505634209766,
+      "grad_norm": 0.39612933920202037,
+      "learning_rate": 2.365861665868211e-06,
+      "loss": 0.0144,
+      "step": 9000
+    },
+    {
+      "epoch": 0.3250505634209766,
+      "eval_loss": 0.2061193287372589,
+      "eval_runtime": 37.6953,
+      "eval_samples_per_second": 108.661,
+      "eval_steps_per_second": 0.212,
+      "step": 9000
+    },
+    {
+      "epoch": 0.3254117307136666,
+      "grad_norm": 0.46217937228583406,
+      "learning_rate": 2.3644282953276125e-06,
+      "loss": 0.0142,
+      "step": 9010
+    },
+    {
+      "epoch": 0.32577289800635656,
+      "grad_norm": 0.3900010432833047,
+      "learning_rate": 2.362993741980597e-06,
+      "loss": 0.0141,
+      "step": 9020
+    },
+    {
+      "epoch": 0.32613406529904654,
+      "grad_norm": 0.34183489954843993,
+      "learning_rate": 2.3615580077900804e-06,
+      "loss": 0.0136,
+      "step": 9030
+    },
+    {
+      "epoch": 0.3264952325917365,
+      "grad_norm": 0.6483846119076843,
+      "learning_rate": 2.3601210947205927e-06,
+      "loss": 0.0145,
+      "step": 9040
+    },
+    {
+      "epoch": 0.32685639988442644,
+      "grad_norm": 0.3992927082197166,
+      "learning_rate": 2.3586830047382766e-06,
+      "loss": 0.0147,
+      "step": 9050
+    },
+    {
+      "epoch": 0.3272175671771164,
+      "grad_norm": 0.40349605760058627,
+      "learning_rate": 2.3572437398108863e-06,
+      "loss": 0.0139,
+      "step": 9060
+    },
+    {
+      "epoch": 0.3275787344698064,
+      "grad_norm": 0.36206085908034324,
+      "learning_rate": 2.3558033019077834e-06,
+      "loss": 0.0142,
+      "step": 9070
+    },
+    {
+      "epoch": 0.3279399017624964,
+      "grad_norm": 0.4789878935056123,
+      "learning_rate": 2.354361692999935e-06,
+      "loss": 0.0133,
+      "step": 9080
+    },
+    {
+      "epoch": 0.32830106905518636,
+      "grad_norm": 0.34108997529820523,
+      "learning_rate": 2.3529189150599086e-06,
+      "loss": 0.0135,
+      "step": 9090
+    },
+    {
+      "epoch": 0.32866223634787634,
+      "grad_norm": 0.45414304216072177,
+      "learning_rate": 2.351474970061873e-06,
+      "loss": 0.0145,
+      "step": 9100
+    },
+    {
+      "epoch": 0.3290234036405663,
+      "grad_norm": 0.4082836721244792,
+      "learning_rate": 2.3500298599815942e-06,
+      "loss": 0.0144,
+      "step": 9110
+    },
+    {
+      "epoch": 0.3293845709332563,
+      "grad_norm": 0.43303333029231333,
+      "learning_rate": 2.348583586796431e-06,
+      "loss": 0.014,
+      "step": 9120
+    },
+    {
+      "epoch": 0.32974573822594627,
+      "grad_norm": 0.4650103442998709,
+      "learning_rate": 2.347136152485334e-06,
+      "loss": 0.014,
+      "step": 9130
+    },
+    {
+      "epoch": 0.33010690551863625,
+      "grad_norm": 0.5290778571338972,
+      "learning_rate": 2.3456875590288437e-06,
+      "loss": 0.0139,
+      "step": 9140
+    },
+    {
+      "epoch": 0.33046807281132623,
+      "grad_norm": 0.4291796668299553,
+      "learning_rate": 2.344237808409086e-06,
+      "loss": 0.0147,
+      "step": 9150
+    },
+    {
+      "epoch": 0.33082924010401615,
+      "grad_norm": 0.35973318893747463,
+      "learning_rate": 2.3427869026097692e-06,
+      "loss": 0.0124,
+      "step": 9160
+    },
+    {
+      "epoch": 0.33119040739670613,
+      "grad_norm": 0.3831597349172342,
+      "learning_rate": 2.3413348436161837e-06,
+      "loss": 0.0144,
+      "step": 9170
+    },
+    {
+      "epoch": 0.3315515746893961,
+      "grad_norm": 0.5857504701519285,
+      "learning_rate": 2.339881633415197e-06,
+      "loss": 0.0147,
+      "step": 9180
+    },
+    {
+      "epoch": 0.3319127419820861,
+      "grad_norm": 0.5182134265398111,
+      "learning_rate": 2.3384272739952528e-06,
+      "loss": 0.0143,
+      "step": 9190
+    },
+    {
+      "epoch": 0.33227390927477607,
+      "grad_norm": 0.34156948971088835,
+      "learning_rate": 2.3369717673463653e-06,
+      "loss": 0.0144,
+      "step": 9200
+    },
+    {
+      "epoch": 0.33263507656746605,
+      "grad_norm": 0.41278373333603285,
+      "learning_rate": 2.335515115460121e-06,
+      "loss": 0.0148,
+      "step": 9210
+    },
+    {
+      "epoch": 0.332996243860156,
+      "grad_norm": 0.44712274774882915,
+      "learning_rate": 2.3340573203296707e-06,
+      "loss": 0.0127,
+      "step": 9220
+    },
+    {
+      "epoch": 0.333357411152846,
+      "grad_norm": 0.5595188386564522,
+      "learning_rate": 2.332598383949733e-06,
+      "loss": 0.0128,
+      "step": 9230
+    },
+    {
+      "epoch": 0.333718578445536,
+      "grad_norm": 0.49650966392404156,
+      "learning_rate": 2.3311383083165833e-06,
+      "loss": 0.0147,
+      "step": 9240
+    },
+    {
+      "epoch": 0.33407974573822596,
+      "grad_norm": 0.4164735315554722,
+      "learning_rate": 2.3296770954280607e-06,
+      "loss": 0.0142,
+      "step": 9250
+    },
+    {
+      "epoch": 0.33444091303091594,
+      "grad_norm": 0.3950733431149686,
+      "learning_rate": 2.3282147472835576e-06,
+      "loss": 0.0141,
+      "step": 9260
+    },
+    {
+      "epoch": 0.3348020803236059,
+      "grad_norm": 0.35670307328637924,
+      "learning_rate": 2.3267512658840216e-06,
+      "loss": 0.014,
+      "step": 9270
+    },
+    {
+      "epoch": 0.33516324761629585,
+      "grad_norm": 0.4058130148793825,
+      "learning_rate": 2.325286653231948e-06,
+      "loss": 0.0138,
+      "step": 9280
+    },
+    {
+      "epoch": 0.3355244149089858,
+      "grad_norm": 0.46538857794799554,
+      "learning_rate": 2.3238209113313824e-06,
+      "loss": 0.0143,
+      "step": 9290
+    },
+    {
+      "epoch": 0.3358855822016758,
+      "grad_norm": 0.45139303694697414,
+      "learning_rate": 2.3223540421879155e-06,
+      "loss": 0.0136,
+      "step": 9300
+    },
+    {
+      "epoch": 0.3362467494943658,
+      "grad_norm": 0.5183883484902561,
+      "learning_rate": 2.32088604780868e-06,
+      "loss": 0.0144,
+      "step": 9310
+    },
+    {
+      "epoch": 0.33660791678705576,
+      "grad_norm": 0.5006024364387778,
+      "learning_rate": 2.3194169302023473e-06,
+      "loss": 0.0139,
+      "step": 9320
+    },
+    {
+      "epoch": 0.33696908407974574,
+      "grad_norm": 0.3860604946238145,
+      "learning_rate": 2.3179466913791276e-06,
+      "loss": 0.0149,
+      "step": 9330
+    },
+    {
+      "epoch": 0.3373302513724357,
+      "grad_norm": 0.34074267593375324,
+      "learning_rate": 2.3164753333507643e-06,
+      "loss": 0.0133,
+      "step": 9340
+    },
+    {
+      "epoch": 0.3376914186651257,
+      "grad_norm": 0.39881953200834713,
+      "learning_rate": 2.3150028581305314e-06,
+      "loss": 0.0134,
+      "step": 9350
+    },
+    {
+      "epoch": 0.3380525859578157,
+      "grad_norm": 0.3828960969380417,
+      "learning_rate": 2.3135292677332327e-06,
+      "loss": 0.0146,
+      "step": 9360
+    },
+    {
+      "epoch": 0.33841375325050566,
+      "grad_norm": 0.4110224064875053,
+      "learning_rate": 2.312054564175198e-06,
+      "loss": 0.0138,
+      "step": 9370
+    },
+    {
+      "epoch": 0.33877492054319563,
+      "grad_norm": 0.5095744083611479,
+      "learning_rate": 2.3105787494742794e-06,
+      "loss": 0.015,
+      "step": 9380
+    },
+    {
+      "epoch": 0.33913608783588556,
+      "grad_norm": 0.4083710793303829,
+      "learning_rate": 2.3091018256498507e-06,
+      "loss": 0.013,
+      "step": 9390
+    },
+    {
+      "epoch": 0.33949725512857554,
+      "grad_norm": 0.4388819552886327,
+      "learning_rate": 2.307623794722801e-06,
+      "loss": 0.0136,
+      "step": 9400
+    },
+    {
+      "epoch": 0.3398584224212655,
+      "grad_norm": 0.412158253719299,
+      "learning_rate": 2.306144658715536e-06,
+      "loss": 0.0135,
+      "step": 9410
+    },
+    {
+      "epoch": 0.3402195897139555,
+      "grad_norm": 0.9338696310350311,
+      "learning_rate": 2.3046644196519743e-06,
+      "loss": 0.0142,
+      "step": 9420
+    },
+    {
+      "epoch": 0.3405807570066455,
+      "grad_norm": 0.40817749331900227,
+      "learning_rate": 2.3031830795575425e-06,
+      "loss": 0.0135,
+      "step": 9430
+    },
+    {
+      "epoch": 0.34094192429933545,
+      "grad_norm": 0.7447455370631857,
+      "learning_rate": 2.301700640459173e-06,
+      "loss": 0.0137,
+      "step": 9440
+    },
+    {
+      "epoch": 0.34130309159202543,
+      "grad_norm": 0.9175613052617605,
+      "learning_rate": 2.3002171043853032e-06,
+      "loss": 0.0134,
+      "step": 9450
+    },
+    {
+      "epoch": 0.3416642588847154,
+      "grad_norm": 9.57300071044564,
+      "learning_rate": 2.298732473365872e-06,
+      "loss": 0.0141,
+      "step": 9460
+    },
+    {
+      "epoch": 0.3420254261774054,
+      "grad_norm": 0.4039151871224139,
+      "learning_rate": 2.2972467494323157e-06,
+      "loss": 0.0128,
+      "step": 9470
+    },
+    {
+      "epoch": 0.34238659347009537,
+      "grad_norm": 0.4003629564158645,
+      "learning_rate": 2.2957599346175656e-06,
+      "loss": 0.0145,
+      "step": 9480
+    },
+    {
+      "epoch": 0.34274776076278535,
+      "grad_norm": 0.37276311471391504,
+      "learning_rate": 2.2942720309560463e-06,
+      "loss": 0.015,
+      "step": 9490
+    },
+    {
+      "epoch": 0.34310892805547527,
+      "grad_norm": 0.5473270139077006,
+      "learning_rate": 2.292783040483672e-06,
+      "loss": 0.0128,
+      "step": 9500
+    },
+    {
+      "epoch": 0.34347009534816525,
+      "grad_norm": 0.39626396946547576,
+      "learning_rate": 2.2912929652378454e-06,
+      "loss": 0.0137,
+      "step": 9510
+    },
+    {
+      "epoch": 0.34383126264085523,
+      "grad_norm": 0.3558736146403505,
+      "learning_rate": 2.2898018072574505e-06,
+      "loss": 0.013,
+      "step": 9520
+    },
+    {
+      "epoch": 0.3441924299335452,
+      "grad_norm": 0.4601341303557844,
+      "learning_rate": 2.2883095685828554e-06,
+      "loss": 0.0141,
+      "step": 9530
+    },
+    {
+      "epoch": 0.3445535972262352,
+      "grad_norm": 0.5503260258220605,
+      "learning_rate": 2.2868162512559065e-06,
+      "loss": 0.0136,
+      "step": 9540
+    },
+    {
+      "epoch": 0.34491476451892517,
+      "grad_norm": 0.421658123122978,
+      "learning_rate": 2.285321857319926e-06,
+      "loss": 0.0138,
+      "step": 9550
+    },
+    {
+      "epoch": 0.34527593181161514,
+      "grad_norm": 0.4331898735371994,
+      "learning_rate": 2.2838263888197077e-06,
+      "loss": 0.0139,
+      "step": 9560
+    },
+    {
+      "epoch": 0.3456370991043051,
+      "grad_norm": 0.4821890061431956,
+      "learning_rate": 2.2823298478015177e-06,
+      "loss": 0.0134,
+      "step": 9570
+    },
+    {
+      "epoch": 0.3459982663969951,
+      "grad_norm": 0.40120785300719386,
+      "learning_rate": 2.280832236313089e-06,
+      "loss": 0.014,
+      "step": 9580
+    },
+    {
+      "epoch": 0.3463594336896851,
+      "grad_norm": 0.4418433846970275,
+      "learning_rate": 2.2793335564036198e-06,
+      "loss": 0.0137,
+      "step": 9590
+    },
+    {
+      "epoch": 0.34672060098237506,
+      "grad_norm": 0.5644118359311556,
+      "learning_rate": 2.277833810123769e-06,
+      "loss": 0.0135,
+      "step": 9600
+    },
+    {
+      "epoch": 0.347081768275065,
+      "grad_norm": 0.4341976046906779,
+      "learning_rate": 2.276332999525656e-06,
+      "loss": 0.0137,
+      "step": 9610
+    },
+    {
+      "epoch": 0.34744293556775496,
+      "grad_norm": 0.44963473213249944,
+      "learning_rate": 2.2748311266628552e-06,
+      "loss": 0.0127,
+      "step": 9620
+    },
+    {
+      "epoch": 0.34780410286044494,
+      "grad_norm": 0.35633473343400285,
+      "learning_rate": 2.273328193590395e-06,
+      "loss": 0.0129,
+      "step": 9630
+    },
+    {
+      "epoch": 0.3481652701531349,
+      "grad_norm": 0.45907079494025504,
+      "learning_rate": 2.271824202364756e-06,
+      "loss": 0.0138,
+      "step": 9640
+    },
+    {
+      "epoch": 0.3485264374458249,
+      "grad_norm": 0.40909791124414047,
+      "learning_rate": 2.2703191550438643e-06,
+      "loss": 0.0134,
+      "step": 9650
+    },
+    {
+      "epoch": 0.3488876047385149,
+      "grad_norm": 0.4698520915896749,
+      "learning_rate": 2.2688130536870924e-06,
+      "loss": 0.0133,
+      "step": 9660
+    },
+    {
+      "epoch": 0.34924877203120486,
+      "grad_norm": 0.3640452106507318,
+      "learning_rate": 2.2673059003552546e-06,
+      "loss": 0.0133,
+      "step": 9670
+    },
+    {
+      "epoch": 0.34960993932389484,
+      "grad_norm": 0.5406642727008902,
+      "learning_rate": 2.2657976971106043e-06,
+      "loss": 0.0133,
+      "step": 9680
+    },
+    {
+      "epoch": 0.3499711066165848,
+      "grad_norm": 0.4862543060805936,
+      "learning_rate": 2.264288446016834e-06,
+      "loss": 0.0138,
+      "step": 9690
+    },
+    {
+      "epoch": 0.3503322739092748,
+      "grad_norm": 0.4856206958458161,
+      "learning_rate": 2.262778149139066e-06,
+      "loss": 0.0142,
+      "step": 9700
+    },
+    {
+      "epoch": 0.3506934412019648,
+      "grad_norm": 0.4486721994918359,
+      "learning_rate": 2.2612668085438563e-06,
+      "loss": 0.0139,
+      "step": 9710
+    },
+    {
+      "epoch": 0.35105460849465475,
+      "grad_norm": 0.3891629978326187,
+      "learning_rate": 2.259754426299188e-06,
+      "loss": 0.0136,
+      "step": 9720
+    },
+    {
+      "epoch": 0.3514157757873447,
+      "grad_norm": 0.389247075936951,
+      "learning_rate": 2.2582410044744705e-06,
+      "loss": 0.015,
+      "step": 9730
+    },
+    {
+      "epoch": 0.35177694308003465,
+      "grad_norm": 0.519866240716715,
+      "learning_rate": 2.256726545140534e-06,
+      "loss": 0.0126,
+      "step": 9740
+    },
+    {
+      "epoch": 0.35213811037272463,
+      "grad_norm": 0.35525086453936605,
+      "learning_rate": 2.25521105036963e-06,
+      "loss": 0.0132,
+      "step": 9750
+    },
+    {
+      "epoch": 0.3524992776654146,
+      "grad_norm": 0.32511138323613137,
+      "learning_rate": 2.2536945222354256e-06,
+      "loss": 0.013,
+      "step": 9760
+    },
+    {
+      "epoch": 0.3528604449581046,
+      "grad_norm": 0.33087813541154015,
+      "learning_rate": 2.2521769628130033e-06,
+      "loss": 0.0125,
+      "step": 9770
+    },
+    {
+      "epoch": 0.35322161225079457,
+      "grad_norm": 0.3578703471072089,
+      "learning_rate": 2.2506583741788547e-06,
+      "loss": 0.0134,
+      "step": 9780
+    },
+    {
+      "epoch": 0.35358277954348455,
+      "grad_norm": 0.4635704569906295,
+      "learning_rate": 2.249138758410881e-06,
+      "loss": 0.0147,
+      "step": 9790
+    },
+    {
+      "epoch": 0.35394394683617453,
+      "grad_norm": 0.6464050849528767,
+      "learning_rate": 2.2476181175883887e-06,
+      "loss": 0.0143,
+      "step": 9800
+    },
+    {
+      "epoch": 0.3543051141288645,
+      "grad_norm": 0.46316500016831935,
+      "learning_rate": 2.2460964537920865e-06,
+      "loss": 0.0127,
+      "step": 9810
+    },
+    {
+      "epoch": 0.3546662814215545,
+      "grad_norm": 0.3469916960696187,
+      "learning_rate": 2.244573769104084e-06,
+      "loss": 0.0133,
+      "step": 9820
+    },
+    {
+      "epoch": 0.35502744871424446,
+      "grad_norm": 0.40858435433184037,
+      "learning_rate": 2.2430500656078857e-06,
+      "loss": 0.0136,
+      "step": 9830
+    },
+    {
+      "epoch": 0.3553886160069344,
+      "grad_norm": 0.33588304391344637,
+      "learning_rate": 2.2415253453883925e-06,
+      "loss": 0.0141,
+      "step": 9840
+    },
+    {
+      "epoch": 0.35574978329962437,
+      "grad_norm": 0.4021006655731499,
+      "learning_rate": 2.2399996105318937e-06,
+      "loss": 0.0129,
+      "step": 9850
+    },
+    {
+      "epoch": 0.35611095059231435,
+      "grad_norm": 0.3191054827354406,
+      "learning_rate": 2.2384728631260704e-06,
+      "loss": 0.013,
+      "step": 9860
+    },
+    {
+      "epoch": 0.3564721178850043,
+      "grad_norm": 0.3599066846330424,
+      "learning_rate": 2.236945105259986e-06,
+      "loss": 0.0138,
+      "step": 9870
+    },
+    {
+      "epoch": 0.3568332851776943,
+      "grad_norm": 0.35382381614991093,
+      "learning_rate": 2.2354163390240874e-06,
+      "loss": 0.0136,
+      "step": 9880
+    },
+    {
+      "epoch": 0.3571944524703843,
+      "grad_norm": 0.3186416968397515,
+      "learning_rate": 2.2338865665102028e-06,
+      "loss": 0.0135,
+      "step": 9890
+    },
+    {
+      "epoch": 0.35755561976307426,
+      "grad_norm": 0.36297065584197014,
+      "learning_rate": 2.2323557898115364e-06,
+      "loss": 0.0134,
+      "step": 9900
+    },
+    {
+      "epoch": 0.35791678705576424,
+      "grad_norm": 0.2897710845772418,
+      "learning_rate": 2.230824011022665e-06,
+      "loss": 0.013,
+      "step": 9910
+    },
+    {
+      "epoch": 0.3582779543484542,
+      "grad_norm": 0.38168384652008097,
+      "learning_rate": 2.2292912322395383e-06,
+      "loss": 0.0126,
+      "step": 9920
+    },
+    {
+      "epoch": 0.3586391216411442,
+      "grad_norm": 0.3447637869378139,
+      "learning_rate": 2.227757455559474e-06,
+      "loss": 0.0134,
+      "step": 9930
+    },
+    {
+      "epoch": 0.3590002889338342,
+      "grad_norm": 0.386497111503044,
+      "learning_rate": 2.226222683081155e-06,
+      "loss": 0.0133,
+      "step": 9940
+    },
+    {
+      "epoch": 0.3593614562265241,
+      "grad_norm": 0.37983124242306243,
+      "learning_rate": 2.2246869169046266e-06,
+      "loss": 0.0133,
+      "step": 9950
+    },
+    {
+      "epoch": 0.3597226235192141,
+      "grad_norm": 0.3319587880623651,
+      "learning_rate": 2.2231501591312934e-06,
+      "loss": 0.013,
+      "step": 9960
+    },
+    {
+      "epoch": 0.36008379081190406,
+      "grad_norm": 0.4002019555843001,
+      "learning_rate": 2.2216124118639184e-06,
+      "loss": 0.0132,
+      "step": 9970
+    },
+    {
+      "epoch": 0.36044495810459404,
+      "grad_norm": 0.5535479578543645,
+      "learning_rate": 2.2200736772066174e-06,
+      "loss": 0.0134,
+      "step": 9980
+    },
+    {
+      "epoch": 0.360806125397284,
+      "grad_norm": 0.578968777197135,
+      "learning_rate": 2.218533957264857e-06,
+      "loss": 0.0138,
+      "step": 9990
+    },
+    {
+      "epoch": 0.361167292689974,
+      "grad_norm": 0.41525900414217104,
+      "learning_rate": 2.216993254145452e-06,
+      "loss": 0.0138,
+      "step": 10000
+    },
+    {
+      "epoch": 0.361167292689974,
+      "eval_loss": 0.19624865055084229,
+      "eval_runtime": 38.6863,
+      "eval_samples_per_second": 105.877,
+      "eval_steps_per_second": 0.207,
+      "step": 10000
+    },
+    {
+      "epoch": 0.361528459982664,
+      "grad_norm": 0.3734667898573586,
+      "learning_rate": 2.215451569956564e-06,
+      "loss": 0.0134,
+      "step": 10010
+    },
+    {
+      "epoch": 0.36188962727535395,
+      "grad_norm": 0.3054959374290739,
+      "learning_rate": 2.2139089068076946e-06,
+      "loss": 0.0133,
+      "step": 10020
+    },
+    {
+      "epoch": 0.36225079456804393,
+      "grad_norm": 0.4168907537792672,
+      "learning_rate": 2.2123652668096866e-06,
+      "loss": 0.0122,
+      "step": 10030
+    },
+    {
+      "epoch": 0.3626119618607339,
+      "grad_norm": 0.33663439164340536,
+      "learning_rate": 2.210820652074719e-06,
+      "loss": 0.0128,
+      "step": 10040
+    },
+    {
+      "epoch": 0.3629731291534239,
+      "grad_norm": 0.41833957339515854,
+      "learning_rate": 2.209275064716305e-06,
+      "loss": 0.0144,
+      "step": 10050
+    },
+    {
+      "epoch": 0.3633342964461138,
+      "grad_norm": 0.2850983136233275,
+      "learning_rate": 2.2077285068492877e-06,
+      "loss": 0.0131,
+      "step": 10060
+    },
+    {
+      "epoch": 0.3636954637388038,
+      "grad_norm": 0.3498244215565677,
+      "learning_rate": 2.206180980589839e-06,
+      "loss": 0.0134,
+      "step": 10070
+    },
+    {
+      "epoch": 0.36405663103149377,
+      "grad_norm": 0.4427287746993334,
+      "learning_rate": 2.204632488055455e-06,
+      "loss": 0.0128,
+      "step": 10080
+    },
+    {
+      "epoch": 0.36441779832418375,
+      "grad_norm": 0.291749684596577,
+      "learning_rate": 2.2030830313649553e-06,
+      "loss": 0.0135,
+      "step": 10090
+    },
+    {
+      "epoch": 0.36477896561687373,
+      "grad_norm": 0.3077792429211624,
+      "learning_rate": 2.2015326126384766e-06,
+      "loss": 0.0131,
+      "step": 10100
+    },
+    {
+      "epoch": 0.3651401329095637,
+      "grad_norm": 0.3771205670687442,
+      "learning_rate": 2.1999812339974746e-06,
+      "loss": 0.0124,
+      "step": 10110
+    },
+    {
+      "epoch": 0.3655013002022537,
+      "grad_norm": 0.3658429538352142,
+      "learning_rate": 2.1984288975647165e-06,
+      "loss": 0.0139,
+      "step": 10120
+    },
+    {
+      "epoch": 0.36586246749494367,
+      "grad_norm": 0.39308995461938995,
+      "learning_rate": 2.1968756054642812e-06,
+      "loss": 0.013,
+      "step": 10130
+    },
+    {
+      "epoch": 0.36622363478763365,
+      "grad_norm": 0.4079413751002693,
+      "learning_rate": 2.1953213598215542e-06,
+      "loss": 0.0137,
+      "step": 10140
+    },
+    {
+      "epoch": 0.3665848020803236,
+      "grad_norm": 0.42152312930794744,
+      "learning_rate": 2.1937661627632266e-06,
+      "loss": 0.0137,
+      "step": 10150
+    },
+    {
+      "epoch": 0.3669459693730136,
+      "grad_norm": 0.37328680781621787,
+      "learning_rate": 2.1922100164172915e-06,
+      "loss": 0.0123,
+      "step": 10160
+    },
+    {
+      "epoch": 0.3673071366657035,
+      "grad_norm": 0.34621134217025434,
+      "learning_rate": 2.19065292291304e-06,
+      "loss": 0.0133,
+      "step": 10170
+    },
+    {
+      "epoch": 0.3676683039583935,
+      "grad_norm": 0.3385174395888218,
+      "learning_rate": 2.1890948843810594e-06,
+      "loss": 0.0125,
+      "step": 10180
+    },
+    {
+      "epoch": 0.3680294712510835,
+      "grad_norm": 0.36958485576923195,
+      "learning_rate": 2.1875359029532308e-06,
+      "loss": 0.0127,
+      "step": 10190
+    },
+    {
+      "epoch": 0.36839063854377346,
+      "grad_norm": 0.3268149942773609,
+      "learning_rate": 2.185975980762725e-06,
+      "loss": 0.0131,
+      "step": 10200
+    },
+    {
+      "epoch": 0.36875180583646344,
+      "grad_norm": 0.3910456357198124,
+      "learning_rate": 2.1844151199439997e-06,
+      "loss": 0.0128,
+      "step": 10210
+    },
+    {
+      "epoch": 0.3691129731291534,
+      "grad_norm": 0.39609096040424197,
+      "learning_rate": 2.1828533226327977e-06,
+      "loss": 0.0125,
+      "step": 10220
+    },
+    {
+      "epoch": 0.3694741404218434,
+      "grad_norm": 0.4656757541822704,
+      "learning_rate": 2.181290590966143e-06,
+      "loss": 0.0128,
+      "step": 10230
+    },
+    {
+      "epoch": 0.3698353077145334,
+      "grad_norm": 0.38150517699431713,
+      "learning_rate": 2.1797269270823376e-06,
+      "loss": 0.0123,
+      "step": 10240
+    },
+    {
+      "epoch": 0.37019647500722336,
+      "grad_norm": 0.37033689110930207,
+      "learning_rate": 2.17816233312096e-06,
+      "loss": 0.0128,
+      "step": 10250
+    },
+    {
+      "epoch": 0.37055764229991334,
+      "grad_norm": 0.4073602887866467,
+      "learning_rate": 2.17659681122286e-06,
+      "loss": 0.0129,
+      "step": 10260
+    },
+    {
+      "epoch": 0.3709188095926033,
+      "grad_norm": 0.3543862730948473,
+      "learning_rate": 2.175030363530158e-06,
+      "loss": 0.0131,
+      "step": 10270
+    },
+    {
+      "epoch": 0.3712799768852933,
+      "grad_norm": 0.4091496466233385,
+      "learning_rate": 2.1734629921862406e-06,
+      "loss": 0.0127,
+      "step": 10280
+    },
+    {
+      "epoch": 0.3716411441779832,
+      "grad_norm": 0.3630132901181736,
+      "learning_rate": 2.1718946993357603e-06,
+      "loss": 0.0127,
+      "step": 10290
+    },
+    {
+      "epoch": 0.3720023114706732,
+      "grad_norm": 0.3199914089367192,
+      "learning_rate": 2.170325487124627e-06,
+      "loss": 0.0123,
+      "step": 10300
+    },
+    {
+      "epoch": 0.3723634787633632,
+      "grad_norm": 0.38158618979590025,
+      "learning_rate": 2.168755357700011e-06,
+      "loss": 0.0132,
+      "step": 10310
+    },
+    {
+      "epoch": 0.37272464605605315,
+      "grad_norm": 0.3238847760860415,
+      "learning_rate": 2.1671843132103377e-06,
+      "loss": 0.0127,
+      "step": 10320
+    },
+    {
+      "epoch": 0.37308581334874313,
+      "grad_norm": 0.328983716670202,
+      "learning_rate": 2.1656123558052832e-06,
+      "loss": 0.0133,
+      "step": 10330
+    },
+    {
+      "epoch": 0.3734469806414331,
+      "grad_norm": 0.3512477248311749,
+      "learning_rate": 2.1640394876357738e-06,
+      "loss": 0.0137,
+      "step": 10340
+    },
+    {
+      "epoch": 0.3738081479341231,
+      "grad_norm": 0.37161459143324377,
+      "learning_rate": 2.162465710853982e-06,
+      "loss": 0.0128,
+      "step": 10350
+    },
+    {
+      "epoch": 0.37416931522681307,
+      "grad_norm": 0.4115791233507006,
+      "learning_rate": 2.1608910276133227e-06,
+      "loss": 0.0133,
+      "step": 10360
+    },
+    {
+      "epoch": 0.37453048251950305,
+      "grad_norm": 0.3258014925565154,
+      "learning_rate": 2.1593154400684525e-06,
+      "loss": 0.0129,
+      "step": 10370
+    },
+    {
+      "epoch": 0.37489164981219303,
+      "grad_norm": 0.311440728057203,
+      "learning_rate": 2.1577389503752637e-06,
+      "loss": 0.0134,
+      "step": 10380
+    },
+    {
+      "epoch": 0.375252817104883,
+      "grad_norm": 0.3502254459277228,
+      "learning_rate": 2.156161560690884e-06,
+      "loss": 0.0125,
+      "step": 10390
+    },
+    {
+      "epoch": 0.37561398439757293,
+      "grad_norm": 0.39418286115132567,
+      "learning_rate": 2.1545832731736734e-06,
+      "loss": 0.0132,
+      "step": 10400
+    },
+    {
+      "epoch": 0.3759751516902629,
+      "grad_norm": 0.3340843154396975,
+      "learning_rate": 2.1530040899832184e-06,
+      "loss": 0.0133,
+      "step": 10410
+    },
+    {
+      "epoch": 0.3763363189829529,
+      "grad_norm": 0.4922901828104574,
+      "learning_rate": 2.1514240132803324e-06,
+      "loss": 0.0121,
+      "step": 10420
+    },
+    {
+      "epoch": 0.37669748627564287,
+      "grad_norm": 0.3844403451225593,
+      "learning_rate": 2.1498430452270518e-06,
+      "loss": 0.0126,
+      "step": 10430
+    },
+    {
+      "epoch": 0.37705865356833285,
+      "grad_norm": 0.3838899808772713,
+      "learning_rate": 2.148261187986631e-06,
+      "loss": 0.0123,
+      "step": 10440
+    },
+    {
+      "epoch": 0.3774198208610228,
+      "grad_norm": 0.33560542921236386,
+      "learning_rate": 2.1466784437235423e-06,
+      "loss": 0.0132,
+      "step": 10450
+    },
+    {
+      "epoch": 0.3777809881537128,
+      "grad_norm": 0.33133612390210315,
+      "learning_rate": 2.145094814603472e-06,
+      "loss": 0.0122,
+      "step": 10460
+    },
+    {
+      "epoch": 0.3781421554464028,
+      "grad_norm": 0.3437152609745163,
+      "learning_rate": 2.1435103027933164e-06,
+      "loss": 0.0123,
+      "step": 10470
+    },
+    {
+      "epoch": 0.37850332273909276,
+      "grad_norm": 0.3191297900323259,
+      "learning_rate": 2.14192491046118e-06,
+      "loss": 0.013,
+      "step": 10480
+    },
+    {
+      "epoch": 0.37886449003178274,
+      "grad_norm": 0.38527441739732116,
+      "learning_rate": 2.1403386397763707e-06,
+      "loss": 0.0132,
+      "step": 10490
+    },
+    {
+      "epoch": 0.3792256573244727,
+      "grad_norm": 0.3512972098526821,
+      "learning_rate": 2.1387514929094012e-06,
+      "loss": 0.013,
+      "step": 10500
+    },
+    {
+      "epoch": 0.37958682461716264,
+      "grad_norm": 0.3607719791095933,
+      "learning_rate": 2.1371634720319807e-06,
+      "loss": 0.0116,
+      "step": 10510
+    },
+    {
+      "epoch": 0.3799479919098526,
+      "grad_norm": 0.3643692961345355,
+      "learning_rate": 2.1355745793170154e-06,
+      "loss": 0.0131,
+      "step": 10520
+    },
+    {
+      "epoch": 0.3803091592025426,
+      "grad_norm": 0.3404177240769873,
+      "learning_rate": 2.1339848169386034e-06,
+      "loss": 0.0126,
+      "step": 10530
+    },
+    {
+      "epoch": 0.3806703264952326,
+      "grad_norm": 0.41779897455787424,
+      "learning_rate": 2.1323941870720337e-06,
+      "loss": 0.0129,
+      "step": 10540
+    },
+    {
+      "epoch": 0.38103149378792256,
+      "grad_norm": 0.36340889710408436,
+      "learning_rate": 2.130802691893782e-06,
+      "loss": 0.0122,
+      "step": 10550
+    },
+    {
+      "epoch": 0.38139266108061254,
+      "grad_norm": 0.35847888864983796,
+      "learning_rate": 2.129210333581509e-06,
+      "loss": 0.0127,
+      "step": 10560
+    },
+    {
+      "epoch": 0.3817538283733025,
+      "grad_norm": 0.38203468644390004,
+      "learning_rate": 2.127617114314054e-06,
+      "loss": 0.0122,
+      "step": 10570
+    },
+    {
+      "epoch": 0.3821149956659925,
+      "grad_norm": 0.37148740434883076,
+      "learning_rate": 2.126023036271437e-06,
+      "loss": 0.0125,
+      "step": 10580
+    },
+    {
+      "epoch": 0.3824761629586825,
+      "grad_norm": 0.31670465142127,
+      "learning_rate": 2.124428101634851e-06,
+      "loss": 0.013,
+      "step": 10590
+    },
+    {
+      "epoch": 0.38283733025137245,
+      "grad_norm": 0.4419112649912604,
+      "learning_rate": 2.1228323125866624e-06,
+      "loss": 0.0117,
+      "step": 10600
+    },
+    {
+      "epoch": 0.38319849754406243,
+      "grad_norm": 0.332385396187242,
+      "learning_rate": 2.1212356713104063e-06,
+      "loss": 0.0124,
+      "step": 10610
+    },
+    {
+      "epoch": 0.38355966483675236,
+      "grad_norm": 0.41821938352530613,
+      "learning_rate": 2.1196381799907843e-06,
+      "loss": 0.0123,
+      "step": 10620
+    },
+    {
+      "epoch": 0.38392083212944234,
+      "grad_norm": 0.28848385850838176,
+      "learning_rate": 2.1180398408136603e-06,
+      "loss": 0.0129,
+      "step": 10630
+    },
+    {
+      "epoch": 0.3842819994221323,
+      "grad_norm": 0.3225625488655691,
+      "learning_rate": 2.1164406559660592e-06,
+      "loss": 0.0132,
+      "step": 10640
+    },
+    {
+      "epoch": 0.3846431667148223,
+      "grad_norm": 0.4194566781861903,
+      "learning_rate": 2.1148406276361617e-06,
+      "loss": 0.0123,
+      "step": 10650
+    },
+    {
+      "epoch": 0.38500433400751227,
+      "grad_norm": 0.36233099513128114,
+      "learning_rate": 2.1132397580133043e-06,
+      "loss": 0.0132,
+      "step": 10660
+    },
+    {
+      "epoch": 0.38536550130020225,
+      "grad_norm": 0.3264366720084577,
+      "learning_rate": 2.1116380492879735e-06,
+      "loss": 0.0128,
+      "step": 10670
+    },
+    {
+      "epoch": 0.38572666859289223,
+      "grad_norm": 0.37458527542445613,
+      "learning_rate": 2.110035503651805e-06,
+      "loss": 0.0116,
+      "step": 10680
+    },
+    {
+      "epoch": 0.3860878358855822,
+      "grad_norm": 0.4992592401900711,
+      "learning_rate": 2.108432123297578e-06,
+      "loss": 0.0134,
+      "step": 10690
+    },
+    {
+      "epoch": 0.3864490031782722,
+      "grad_norm": 0.4057951293771338,
+      "learning_rate": 2.106827910419216e-06,
+      "loss": 0.0124,
+      "step": 10700
+    },
+    {
+      "epoch": 0.38681017047096217,
+      "grad_norm": 0.4330523501161968,
+      "learning_rate": 2.1052228672117805e-06,
+      "loss": 0.0129,
+      "step": 10710
+    },
+    {
+      "epoch": 0.38717133776365215,
+      "grad_norm": 0.4365877798904756,
+      "learning_rate": 2.1036169958714677e-06,
+      "loss": 0.013,
+      "step": 10720
+    },
+    {
+      "epoch": 0.3875325050563421,
+      "grad_norm": 0.4331356209364764,
+      "learning_rate": 2.10201029859561e-06,
+      "loss": 0.0116,
+      "step": 10730
+    },
+    {
+      "epoch": 0.38789367234903205,
+      "grad_norm": 0.3430700699346294,
+      "learning_rate": 2.1004027775826672e-06,
+      "loss": 0.0121,
+      "step": 10740
+    },
+    {
+      "epoch": 0.388254839641722,
+      "grad_norm": 0.34287497936419575,
+      "learning_rate": 2.0987944350322283e-06,
+      "loss": 0.0122,
+      "step": 10750
+    },
+    {
+      "epoch": 0.388616006934412,
+      "grad_norm": 0.32828529651589816,
+      "learning_rate": 2.0971852731450047e-06,
+      "loss": 0.0115,
+      "step": 10760
+    },
+    {
+      "epoch": 0.388977174227102,
+      "grad_norm": 0.33466149855711635,
+      "learning_rate": 2.09557529412283e-06,
+      "loss": 0.0129,
+      "step": 10770
+    },
+    {
+      "epoch": 0.38933834151979196,
+      "grad_norm": 0.32059498088746174,
+      "learning_rate": 2.093964500168656e-06,
+      "loss": 0.0126,
+      "step": 10780
+    },
+    {
+      "epoch": 0.38969950881248194,
+      "grad_norm": 0.5000479447800851,
+      "learning_rate": 2.0923528934865486e-06,
+      "loss": 0.0116,
+      "step": 10790
+    },
+    {
+      "epoch": 0.3900606761051719,
+      "grad_norm": 0.4128821205380068,
+      "learning_rate": 2.090740476281687e-06,
+      "loss": 0.0129,
+      "step": 10800
+    },
+    {
+      "epoch": 0.3904218433978619,
+      "grad_norm": 0.3035725592079718,
+      "learning_rate": 2.089127250760358e-06,
+      "loss": 0.0131,
+      "step": 10810
+    },
+    {
+      "epoch": 0.3907830106905519,
+      "grad_norm": 0.31194960825434015,
+      "learning_rate": 2.0875132191299557e-06,
+      "loss": 0.0129,
+      "step": 10820
+    },
+    {
+      "epoch": 0.39114417798324186,
+      "grad_norm": 0.46580012502584905,
+      "learning_rate": 2.0858983835989764e-06,
+      "loss": 0.013,
+      "step": 10830
+    },
+    {
+      "epoch": 0.39150534527593184,
+      "grad_norm": 0.31596039473408966,
+      "learning_rate": 2.0842827463770173e-06,
+      "loss": 0.013,
+      "step": 10840
+    },
+    {
+      "epoch": 0.39186651256862176,
+      "grad_norm": 0.3364153691653146,
+      "learning_rate": 2.082666309674771e-06,
+      "loss": 0.0128,
+      "step": 10850
+    },
+    {
+      "epoch": 0.39222767986131174,
+      "grad_norm": 0.3426904906553759,
+      "learning_rate": 2.0810490757040257e-06,
+      "loss": 0.0128,
+      "step": 10860
+    },
+    {
+      "epoch": 0.3925888471540017,
+      "grad_norm": 0.4184087986415348,
+      "learning_rate": 2.0794310466776604e-06,
+      "loss": 0.0127,
+      "step": 10870
+    },
+    {
+      "epoch": 0.3929500144466917,
+      "grad_norm": 0.36260269938499007,
+      "learning_rate": 2.07781222480964e-06,
+      "loss": 0.0126,
+      "step": 10880
+    },
+    {
+      "epoch": 0.3933111817393817,
+      "grad_norm": 0.3564255533848866,
+      "learning_rate": 2.0761926123150166e-06,
+      "loss": 0.0116,
+      "step": 10890
+    },
+    {
+      "epoch": 0.39367234903207166,
+      "grad_norm": 0.2863696087887251,
+      "learning_rate": 2.0745722114099225e-06,
+      "loss": 0.0133,
+      "step": 10900
+    },
+    {
+      "epoch": 0.39403351632476163,
+      "grad_norm": 0.3313270651622131,
+      "learning_rate": 2.0729510243115706e-06,
+      "loss": 0.0122,
+      "step": 10910
+    },
+    {
+      "epoch": 0.3943946836174516,
+      "grad_norm": 0.3326728830839401,
+      "learning_rate": 2.071329053238247e-06,
+      "loss": 0.0131,
+      "step": 10920
+    },
+    {
+      "epoch": 0.3947558509101416,
+      "grad_norm": 0.31334415491418144,
+      "learning_rate": 2.069706300409313e-06,
+      "loss": 0.0119,
+      "step": 10930
+    },
+    {
+      "epoch": 0.39511701820283157,
+      "grad_norm": 0.3311837517570839,
+      "learning_rate": 2.0680827680451976e-06,
+      "loss": 0.012,
+      "step": 10940
+    },
+    {
+      "epoch": 0.39547818549552155,
+      "grad_norm": 0.2900172024015571,
+      "learning_rate": 2.0664584583673987e-06,
+      "loss": 0.0131,
+      "step": 10950
+    },
+    {
+      "epoch": 0.3958393527882115,
+      "grad_norm": 0.3354539114483783,
+      "learning_rate": 2.0648333735984757e-06,
+      "loss": 0.0131,
+      "step": 10960
+    },
+    {
+      "epoch": 0.39620052008090145,
+      "grad_norm": 0.48439681944011087,
+      "learning_rate": 2.063207515962049e-06,
+      "loss": 0.0122,
+      "step": 10970
+    },
+    {
+      "epoch": 0.39656168737359143,
+      "grad_norm": 0.32497757100248054,
+      "learning_rate": 2.061580887682797e-06,
+      "loss": 0.0124,
+      "step": 10980
+    },
+    {
+      "epoch": 0.3969228546662814,
+      "grad_norm": 0.32691172367490323,
+      "learning_rate": 2.0599534909864535e-06,
+      "loss": 0.0115,
+      "step": 10990
+    },
+    {
+      "epoch": 0.3972840219589714,
+      "grad_norm": 0.2967189754143977,
+      "learning_rate": 2.0583253280998015e-06,
+      "loss": 0.0123,
+      "step": 11000
+    },
+    {
+      "epoch": 0.3972840219589714,
+      "eval_loss": 0.20797944068908691,
+      "eval_runtime": 38.9436,
+      "eval_samples_per_second": 105.178,
+      "eval_steps_per_second": 0.205,
+      "step": 11000
+    },
+    {
+      "epoch": 0.39764518925166137,
+      "grad_norm": 0.3364005798785596,
+      "learning_rate": 2.056696401250674e-06,
+      "loss": 0.0116,
+      "step": 11010
+    },
+    {
+      "epoch": 0.39800635654435135,
+      "grad_norm": 0.39157032885468157,
+      "learning_rate": 2.0550667126679486e-06,
+      "loss": 0.0113,
+      "step": 11020
+    },
+    {
+      "epoch": 0.3983675238370413,
+      "grad_norm": 0.3495268736411879,
+      "learning_rate": 2.0534362645815452e-06,
+      "loss": 0.0128,
+      "step": 11030
+    },
+    {
+      "epoch": 0.3987286911297313,
+      "grad_norm": 0.3078531434117483,
+      "learning_rate": 2.0518050592224247e-06,
+      "loss": 0.0126,
+      "step": 11040
+    },
+    {
+      "epoch": 0.3990898584224213,
+      "grad_norm": 0.3533381536593918,
+      "learning_rate": 2.0501730988225816e-06,
+      "loss": 0.012,
+      "step": 11050
+    },
+    {
+      "epoch": 0.39945102571511126,
+      "grad_norm": 0.8668164491199899,
+      "learning_rate": 2.0485403856150453e-06,
+      "loss": 0.0118,
+      "step": 11060
+    },
+    {
+      "epoch": 0.3998121930078012,
+      "grad_norm": 0.37254687050714674,
+      "learning_rate": 2.0469069218338733e-06,
+      "loss": 0.0132,
+      "step": 11070
+    },
+    {
+      "epoch": 0.40017336030049117,
+      "grad_norm": 0.2983444824493059,
+      "learning_rate": 2.0452727097141527e-06,
+      "loss": 0.012,
+      "step": 11080
+    },
+    {
+      "epoch": 0.40053452759318114,
+      "grad_norm": 0.3565965352186869,
+      "learning_rate": 2.0436377514919928e-06,
+      "loss": 0.0117,
+      "step": 11090
+    },
+    {
+      "epoch": 0.4008956948858711,
+      "grad_norm": 0.4138155546260618,
+      "learning_rate": 2.0420020494045245e-06,
+      "loss": 0.0123,
+      "step": 11100
+    },
+    {
+      "epoch": 0.4012568621785611,
+      "grad_norm": 0.39339953415600243,
+      "learning_rate": 2.040365605689896e-06,
+      "loss": 0.0124,
+      "step": 11110
+    },
+    {
+      "epoch": 0.4016180294712511,
+      "grad_norm": 0.29524294955392366,
+      "learning_rate": 2.0387284225872713e-06,
+      "loss": 0.0122,
+      "step": 11120
+    },
+    {
+      "epoch": 0.40197919676394106,
+      "grad_norm": 0.3272504098703144,
+      "learning_rate": 2.0370905023368244e-06,
+      "loss": 0.012,
+      "step": 11130
+    },
+    {
+      "epoch": 0.40234036405663104,
+      "grad_norm": 0.3067238006788116,
+      "learning_rate": 2.0354518471797397e-06,
+      "loss": 0.0129,
+      "step": 11140
+    },
+    {
+      "epoch": 0.402701531349321,
+      "grad_norm": 0.3117850603925964,
+      "learning_rate": 2.0338124593582067e-06,
+      "loss": 0.0127,
+      "step": 11150
+    },
+    {
+      "epoch": 0.403062698642011,
+      "grad_norm": 0.361427287614536,
+      "learning_rate": 2.0321723411154164e-06,
+      "loss": 0.0121,
+      "step": 11160
+    },
+    {
+      "epoch": 0.403423865934701,
+      "grad_norm": 0.4084135736557363,
+      "learning_rate": 2.0305314946955606e-06,
+      "loss": 0.0123,
+      "step": 11170
+    },
+    {
+      "epoch": 0.40378503322739095,
+      "grad_norm": 0.6061967653401092,
+      "learning_rate": 2.028889922343826e-06,
+      "loss": 0.013,
+      "step": 11180
+    },
+    {
+      "epoch": 0.4041462005200809,
+      "grad_norm": 0.37625316553971455,
+      "learning_rate": 2.027247626306395e-06,
+      "loss": 0.0121,
+      "step": 11190
+    },
+    {
+      "epoch": 0.40450736781277086,
+      "grad_norm": 0.4297111330123469,
+      "learning_rate": 2.0256046088304374e-06,
+      "loss": 0.0119,
+      "step": 11200
+    },
+    {
+      "epoch": 0.40486853510546084,
+      "grad_norm": 0.3629952993241635,
+      "learning_rate": 2.0239608721641117e-06,
+      "loss": 0.0125,
+      "step": 11210
+    },
+    {
+      "epoch": 0.4052297023981508,
+      "grad_norm": 0.29360936214984923,
+      "learning_rate": 2.022316418556561e-06,
+      "loss": 0.0122,
+      "step": 11220
+    },
+    {
+      "epoch": 0.4055908696908408,
+      "grad_norm": 0.3468276178232923,
+      "learning_rate": 2.0206712502579076e-06,
+      "loss": 0.0115,
+      "step": 11230
+    },
+    {
+      "epoch": 0.4059520369835308,
+      "grad_norm": 0.42525765102761137,
+      "learning_rate": 2.0190253695192533e-06,
+      "loss": 0.0115,
+      "step": 11240
+    },
+    {
+      "epoch": 0.40631320427622075,
+      "grad_norm": 0.4788501047591058,
+      "learning_rate": 2.0173787785926743e-06,
+      "loss": 0.012,
+      "step": 11250
+    },
+    {
+      "epoch": 0.40667437156891073,
+      "grad_norm": 0.38693557599813194,
+      "learning_rate": 2.015731479731218e-06,
+      "loss": 0.0123,
+      "step": 11260
+    },
+    {
+      "epoch": 0.4070355388616007,
+      "grad_norm": 0.41592791644671784,
+      "learning_rate": 2.0140834751889013e-06,
+      "loss": 0.0121,
+      "step": 11270
+    },
+    {
+      "epoch": 0.4073967061542907,
+      "grad_norm": 0.3638989714688471,
+      "learning_rate": 2.0124347672207063e-06,
+      "loss": 0.0118,
+      "step": 11280
+    },
+    {
+      "epoch": 0.40775787344698067,
+      "grad_norm": 0.35713421014071955,
+      "learning_rate": 2.010785358082577e-06,
+      "loss": 0.0116,
+      "step": 11290
+    },
+    {
+      "epoch": 0.4081190407396706,
+      "grad_norm": 0.33585841313189246,
+      "learning_rate": 2.0091352500314187e-06,
+      "loss": 0.0124,
+      "step": 11300
+    },
+    {
+      "epoch": 0.40848020803236057,
+      "grad_norm": 0.34022361048060573,
+      "learning_rate": 2.007484445325091e-06,
+      "loss": 0.0122,
+      "step": 11310
+    },
+    {
+      "epoch": 0.40884137532505055,
+      "grad_norm": 0.4195634891561011,
+      "learning_rate": 2.0058329462224073e-06,
+      "loss": 0.0128,
+      "step": 11320
+    },
+    {
+      "epoch": 0.40920254261774053,
+      "grad_norm": 0.4071188012139823,
+      "learning_rate": 2.0041807549831317e-06,
+      "loss": 0.0126,
+      "step": 11330
+    },
+    {
+      "epoch": 0.4095637099104305,
+      "grad_norm": 0.4818004592585391,
+      "learning_rate": 2.002527873867975e-06,
+      "loss": 0.0114,
+      "step": 11340
+    },
+    {
+      "epoch": 0.4099248772031205,
+      "grad_norm": 0.34467032066458037,
+      "learning_rate": 2.0008743051385916e-06,
+      "loss": 0.0117,
+      "step": 11350
+    },
+    {
+      "epoch": 0.41028604449581046,
+      "grad_norm": 0.3429337871932726,
+      "learning_rate": 1.9992200510575778e-06,
+      "loss": 0.0125,
+      "step": 11360
+    },
+    {
+      "epoch": 0.41064721178850044,
+      "grad_norm": 0.3164924170125475,
+      "learning_rate": 1.997565113888466e-06,
+      "loss": 0.0108,
+      "step": 11370
+    },
+    {
+      "epoch": 0.4110083790811904,
+      "grad_norm": 0.30468867427958196,
+      "learning_rate": 1.995909495895726e-06,
+      "loss": 0.0121,
+      "step": 11380
+    },
+    {
+      "epoch": 0.4113695463738804,
+      "grad_norm": 0.4941950687467099,
+      "learning_rate": 1.994253199344756e-06,
+      "loss": 0.0114,
+      "step": 11390
+    },
+    {
+      "epoch": 0.4117307136665704,
+      "grad_norm": 0.35128116490846495,
+      "learning_rate": 1.9925962265018842e-06,
+      "loss": 0.0124,
+      "step": 11400
+    },
+    {
+      "epoch": 0.4120918809592603,
+      "grad_norm": 0.31649063255083587,
+      "learning_rate": 1.990938579634365e-06,
+      "loss": 0.0117,
+      "step": 11410
+    },
+    {
+      "epoch": 0.4124530482519503,
+      "grad_norm": 0.42655041503110547,
+      "learning_rate": 1.989280261010374e-06,
+      "loss": 0.0122,
+      "step": 11420
+    },
+    {
+      "epoch": 0.41281421554464026,
+      "grad_norm": 0.4983463858702786,
+      "learning_rate": 1.9876212728990055e-06,
+      "loss": 0.0125,
+      "step": 11430
+    },
+    {
+      "epoch": 0.41317538283733024,
+      "grad_norm": 0.36511245239889467,
+      "learning_rate": 1.9859616175702714e-06,
+      "loss": 0.0119,
+      "step": 11440
+    },
+    {
+      "epoch": 0.4135365501300202,
+      "grad_norm": 0.3057282933319697,
+      "learning_rate": 1.984301297295095e-06,
+      "loss": 0.0118,
+      "step": 11450
+    },
+    {
+      "epoch": 0.4138977174227102,
+      "grad_norm": 0.29230240569733373,
+      "learning_rate": 1.982640314345311e-06,
+      "loss": 0.0128,
+      "step": 11460
+    },
+    {
+      "epoch": 0.4142588847154002,
+      "grad_norm": 0.3229792744323045,
+      "learning_rate": 1.980978670993659e-06,
+      "loss": 0.0121,
+      "step": 11470
+    },
+    {
+      "epoch": 0.41462005200809016,
+      "grad_norm": 0.3892049581403922,
+      "learning_rate": 1.9793163695137837e-06,
+      "loss": 0.0118,
+      "step": 11480
+    },
+    {
+      "epoch": 0.41498121930078014,
+      "grad_norm": 0.392658999836149,
+      "learning_rate": 1.97765341218023e-06,
+      "loss": 0.0121,
+      "step": 11490
+    },
+    {
+      "epoch": 0.4153423865934701,
+      "grad_norm": 0.4069182934381257,
+      "learning_rate": 1.975989801268441e-06,
+      "loss": 0.0118,
+      "step": 11500
+    },
+    {
+      "epoch": 0.4157035538861601,
+      "grad_norm": 0.3562602260630977,
+      "learning_rate": 1.974325539054751e-06,
+      "loss": 0.0116,
+      "step": 11510
+    },
+    {
+      "epoch": 0.41606472117885,
+      "grad_norm": 0.524790856960923,
+      "learning_rate": 1.9726606278163888e-06,
+      "loss": 0.012,
+      "step": 11520
+    },
+    {
+      "epoch": 0.41642588847154,
+      "grad_norm": 0.37094067659928986,
+      "learning_rate": 1.9709950698314705e-06,
+      "loss": 0.0125,
+      "step": 11530
+    },
+    {
+      "epoch": 0.41678705576423,
+      "grad_norm": 0.34741937032057796,
+      "learning_rate": 1.9693288673789966e-06,
+      "loss": 0.0116,
+      "step": 11540
+    },
+    {
+      "epoch": 0.41714822305691995,
+      "grad_norm": 0.8308302728716334,
+      "learning_rate": 1.967662022738849e-06,
+      "loss": 0.0124,
+      "step": 11550
+    },
+    {
+      "epoch": 0.41750939034960993,
+      "grad_norm": 0.35536385862604264,
+      "learning_rate": 1.9659945381917896e-06,
+      "loss": 0.0119,
+      "step": 11560
+    },
+    {
+      "epoch": 0.4178705576422999,
+      "grad_norm": 0.3544208536825439,
+      "learning_rate": 1.964326416019455e-06,
+      "loss": 0.0122,
+      "step": 11570
+    },
+    {
+      "epoch": 0.4182317249349899,
+      "grad_norm": 0.37548437190234557,
+      "learning_rate": 1.9626576585043545e-06,
+      "loss": 0.0119,
+      "step": 11580
+    },
+    {
+      "epoch": 0.41859289222767987,
+      "grad_norm": 0.37043713067010153,
+      "learning_rate": 1.9609882679298667e-06,
+      "loss": 0.0115,
+      "step": 11590
+    },
+    {
+      "epoch": 0.41895405952036985,
+      "grad_norm": 0.32104151759825084,
+      "learning_rate": 1.959318246580236e-06,
+      "loss": 0.0127,
+      "step": 11600
+    },
+    {
+      "epoch": 0.4193152268130598,
+      "grad_norm": 0.36133293014641027,
+      "learning_rate": 1.9576475967405715e-06,
+      "loss": 0.0123,
+      "step": 11610
+    },
+    {
+      "epoch": 0.4196763941057498,
+      "grad_norm": 0.35613831084756753,
+      "learning_rate": 1.95597632069684e-06,
+      "loss": 0.0119,
+      "step": 11620
+    },
+    {
+      "epoch": 0.4200375613984398,
+      "grad_norm": 0.4204694592581412,
+      "learning_rate": 1.954304420735867e-06,
+      "loss": 0.0121,
+      "step": 11630
+    },
+    {
+      "epoch": 0.4203987286911297,
+      "grad_norm": 0.7478018654976005,
+      "learning_rate": 1.9526318991453305e-06,
+      "loss": 0.0126,
+      "step": 11640
+    },
+    {
+      "epoch": 0.4207598959838197,
+      "grad_norm": 0.38735359879582887,
+      "learning_rate": 1.950958758213759e-06,
+      "loss": 0.0115,
+      "step": 11650
+    },
+    {
+      "epoch": 0.42112106327650967,
+      "grad_norm": 0.3384807472036846,
+      "learning_rate": 1.9492850002305305e-06,
+      "loss": 0.0118,
+      "step": 11660
+    },
+    {
+      "epoch": 0.42148223056919965,
+      "grad_norm": 0.429775914338008,
+      "learning_rate": 1.947610627485865e-06,
+      "loss": 0.0124,
+      "step": 11670
+    },
+    {
+      "epoch": 0.4218433978618896,
+      "grad_norm": 0.35043068385644766,
+      "learning_rate": 1.945935642270824e-06,
+      "loss": 0.0122,
+      "step": 11680
+    },
+    {
+      "epoch": 0.4222045651545796,
+      "grad_norm": 0.34708942583359537,
+      "learning_rate": 1.9442600468773083e-06,
+      "loss": 0.0123,
+      "step": 11690
+    },
+    {
+      "epoch": 0.4225657324472696,
+      "grad_norm": 0.7605963896548767,
+      "learning_rate": 1.9425838435980533e-06,
+      "loss": 0.0115,
+      "step": 11700
+    },
+    {
+      "epoch": 0.42292689973995956,
+      "grad_norm": 0.4123895746769268,
+      "learning_rate": 1.9409070347266245e-06,
+      "loss": 0.0118,
+      "step": 11710
+    },
+    {
+      "epoch": 0.42328806703264954,
+      "grad_norm": 0.36315904950205796,
+      "learning_rate": 1.939229622557418e-06,
+      "loss": 0.0119,
+      "step": 11720
+    },
+    {
+      "epoch": 0.4236492343253395,
+      "grad_norm": 0.37660376795029304,
+      "learning_rate": 1.937551609385654e-06,
+      "loss": 0.012,
+      "step": 11730
+    },
+    {
+      "epoch": 0.4240104016180295,
+      "grad_norm": 0.3224402487070123,
+      "learning_rate": 1.9358729975073776e-06,
+      "loss": 0.012,
+      "step": 11740
+    },
+    {
+      "epoch": 0.4243715689107194,
+      "grad_norm": 0.32794533599540177,
+      "learning_rate": 1.934193789219449e-06,
+      "loss": 0.0111,
+      "step": 11750
+    },
+    {
+      "epoch": 0.4247327362034094,
+      "grad_norm": 0.3735638633154375,
+      "learning_rate": 1.9325139868195474e-06,
+      "loss": 0.0121,
+      "step": 11760
+    },
+    {
+      "epoch": 0.4250939034960994,
+      "grad_norm": 0.35102494839179366,
+      "learning_rate": 1.9308335926061653e-06,
+      "loss": 0.0125,
+      "step": 11770
+    },
+    {
+      "epoch": 0.42545507078878936,
+      "grad_norm": 0.34007990670852645,
+      "learning_rate": 1.929152608878603e-06,
+      "loss": 0.011,
+      "step": 11780
+    },
+    {
+      "epoch": 0.42581623808147934,
+      "grad_norm": 0.3806628161726908,
+      "learning_rate": 1.9274710379369684e-06,
+      "loss": 0.0113,
+      "step": 11790
+    },
+    {
+      "epoch": 0.4261774053741693,
+      "grad_norm": 0.3190653215474123,
+      "learning_rate": 1.9257888820821736e-06,
+      "loss": 0.0119,
+      "step": 11800
+    },
+    {
+      "epoch": 0.4265385726668593,
+      "grad_norm": 0.2941889028960763,
+      "learning_rate": 1.924106143615929e-06,
+      "loss": 0.0116,
+      "step": 11810
+    },
+    {
+      "epoch": 0.4268997399595493,
+      "grad_norm": 0.3546919849289418,
+      "learning_rate": 1.922422824840745e-06,
+      "loss": 0.0119,
+      "step": 11820
+    },
+    {
+      "epoch": 0.42726090725223925,
+      "grad_norm": 0.3847849568894746,
+      "learning_rate": 1.9207389280599236e-06,
+      "loss": 0.0117,
+      "step": 11830
+    },
+    {
+      "epoch": 0.42762207454492923,
+      "grad_norm": 0.350529479398848,
+      "learning_rate": 1.919054455577559e-06,
+      "loss": 0.0119,
+      "step": 11840
+    },
+    {
+      "epoch": 0.4279832418376192,
+      "grad_norm": 0.3338978808263323,
+      "learning_rate": 1.9173694096985328e-06,
+      "loss": 0.0129,
+      "step": 11850
+    },
+    {
+      "epoch": 0.42834440913030913,
+      "grad_norm": 0.4287384235359753,
+      "learning_rate": 1.915683792728511e-06,
+      "loss": 0.0113,
+      "step": 11860
+    },
+    {
+      "epoch": 0.4287055764229991,
+      "grad_norm": 0.33599707051383426,
+      "learning_rate": 1.913997606973941e-06,
+      "loss": 0.0116,
+      "step": 11870
+    },
+    {
+      "epoch": 0.4290667437156891,
+      "grad_norm": 0.3387264617024684,
+      "learning_rate": 1.9123108547420492e-06,
+      "loss": 0.0114,
+      "step": 11880
+    },
+    {
+      "epoch": 0.42942791100837907,
+      "grad_norm": 0.434346912732661,
+      "learning_rate": 1.9106235383408368e-06,
+      "loss": 0.0113,
+      "step": 11890
+    },
+    {
+      "epoch": 0.42978907830106905,
+      "grad_norm": 0.4033928438591671,
+      "learning_rate": 1.9089356600790754e-06,
+      "loss": 0.0112,
+      "step": 11900
+    },
+    {
+      "epoch": 0.43015024559375903,
+      "grad_norm": 0.3013825043319818,
+      "learning_rate": 1.9072472222663077e-06,
+      "loss": 0.0118,
+      "step": 11910
+    },
+    {
+      "epoch": 0.430511412886449,
+      "grad_norm": 0.7468020378869318,
+      "learning_rate": 1.9055582272128414e-06,
+      "loss": 0.0122,
+      "step": 11920
+    },
+    {
+      "epoch": 0.430872580179139,
+      "grad_norm": 0.29792941777803117,
+      "learning_rate": 1.903868677229746e-06,
+      "loss": 0.0115,
+      "step": 11930
+    },
+    {
+      "epoch": 0.43123374747182897,
+      "grad_norm": 0.3605488086785044,
+      "learning_rate": 1.9021785746288506e-06,
+      "loss": 0.0116,
+      "step": 11940
+    },
+    {
+      "epoch": 0.43159491476451894,
+      "grad_norm": 0.36234067686827715,
+      "learning_rate": 1.9004879217227406e-06,
+      "loss": 0.0111,
+      "step": 11950
+    },
+    {
+      "epoch": 0.4319560820572089,
+      "grad_norm": 0.32995310218548174,
+      "learning_rate": 1.8987967208247542e-06,
+      "loss": 0.0116,
+      "step": 11960
+    },
+    {
+      "epoch": 0.43231724934989885,
+      "grad_norm": 0.3896404439587071,
+      "learning_rate": 1.8971049742489803e-06,
+      "loss": 0.0118,
+      "step": 11970
+    },
+    {
+      "epoch": 0.4326784166425888,
+      "grad_norm": 0.5804754620138586,
+      "learning_rate": 1.8954126843102527e-06,
+      "loss": 0.0122,
+      "step": 11980
+    },
+    {
+      "epoch": 0.4330395839352788,
+      "grad_norm": 0.44503451824770235,
+      "learning_rate": 1.8937198533241497e-06,
+      "loss": 0.0116,
+      "step": 11990
+    },
+    {
+      "epoch": 0.4334007512279688,
+      "grad_norm": 0.2624042871435185,
+      "learning_rate": 1.8920264836069907e-06,
+      "loss": 0.0114,
+      "step": 12000
+    },
+    {
+      "epoch": 0.4334007512279688,
+      "eval_loss": 0.21704500913619995,
+      "eval_runtime": 39.3004,
+      "eval_samples_per_second": 104.223,
+      "eval_steps_per_second": 0.204,
+      "step": 12000
+    },
+    {
+      "epoch": 0.43376191852065876,
+      "grad_norm": 0.35680036368847906,
+      "learning_rate": 1.890332577475831e-06,
+      "loss": 0.0117,
+      "step": 12010
+    },
+    {
+      "epoch": 0.43412308581334874,
+      "grad_norm": 0.3651951006995348,
+      "learning_rate": 1.8886381372484606e-06,
+      "loss": 0.0126,
+      "step": 12020
+    },
+    {
+      "epoch": 0.4344842531060387,
+      "grad_norm": 0.3243360231428399,
+      "learning_rate": 1.8869431652433994e-06,
+      "loss": 0.0127,
+      "step": 12030
+    },
+    {
+      "epoch": 0.4348454203987287,
+      "grad_norm": 0.33533246059687033,
+      "learning_rate": 1.8852476637798962e-06,
+      "loss": 0.0126,
+      "step": 12040
+    },
+    {
+      "epoch": 0.4352065876914187,
+      "grad_norm": 0.34956639245773113,
+      "learning_rate": 1.8835516351779238e-06,
+      "loss": 0.0114,
+      "step": 12050
+    },
+    {
+      "epoch": 0.43556775498410866,
+      "grad_norm": 0.31427546153919295,
+      "learning_rate": 1.881855081758176e-06,
+      "loss": 0.0118,
+      "step": 12060
+    },
+    {
+      "epoch": 0.43592892227679864,
+      "grad_norm": 0.31329645661031535,
+      "learning_rate": 1.8801580058420645e-06,
+      "loss": 0.012,
+      "step": 12070
+    },
+    {
+      "epoch": 0.43629008956948856,
+      "grad_norm": 0.2703014515023635,
+      "learning_rate": 1.8784604097517163e-06,
+      "loss": 0.0117,
+      "step": 12080
+    },
+    {
+      "epoch": 0.43665125686217854,
+      "grad_norm": 0.3368304037369543,
+      "learning_rate": 1.8767622958099707e-06,
+      "loss": 0.0108,
+      "step": 12090
+    },
+    {
+      "epoch": 0.4370124241548685,
+      "grad_norm": 0.27191638801597995,
+      "learning_rate": 1.8750636663403747e-06,
+      "loss": 0.0113,
+      "step": 12100
+    },
+    {
+      "epoch": 0.4373735914475585,
+      "grad_norm": 0.48026484284340076,
+      "learning_rate": 1.8733645236671808e-06,
+      "loss": 0.0122,
+      "step": 12110
+    },
+    {
+      "epoch": 0.4377347587402485,
+      "grad_norm": 0.620474792502596,
+      "learning_rate": 1.8716648701153446e-06,
+      "loss": 0.0114,
+      "step": 12120
+    },
+    {
+      "epoch": 0.43809592603293845,
+      "grad_norm": 0.3070455648505682,
+      "learning_rate": 1.8699647080105195e-06,
+      "loss": 0.012,
+      "step": 12130
+    },
+    {
+      "epoch": 0.43845709332562843,
+      "grad_norm": 0.2925720737749786,
+      "learning_rate": 1.8682640396790554e-06,
+      "loss": 0.0117,
+      "step": 12140
+    },
+    {
+      "epoch": 0.4388182606183184,
+      "grad_norm": 0.34254297205149026,
+      "learning_rate": 1.8665628674479959e-06,
+      "loss": 0.0116,
+      "step": 12150
+    },
+    {
+      "epoch": 0.4391794279110084,
+      "grad_norm": 0.26301651432660533,
+      "learning_rate": 1.8648611936450714e-06,
+      "loss": 0.0121,
+      "step": 12160
+    },
+    {
+      "epoch": 0.43954059520369837,
+      "grad_norm": 0.3448661072130684,
+      "learning_rate": 1.8631590205987011e-06,
+      "loss": 0.0113,
+      "step": 12170
+    },
+    {
+      "epoch": 0.43990176249638835,
+      "grad_norm": 0.39134333095711255,
+      "learning_rate": 1.8614563506379864e-06,
+      "loss": 0.0117,
+      "step": 12180
+    },
+    {
+      "epoch": 0.44026292978907833,
+      "grad_norm": 0.30706126049017507,
+      "learning_rate": 1.8597531860927084e-06,
+      "loss": 0.011,
+      "step": 12190
+    },
+    {
+      "epoch": 0.44062409708176825,
+      "grad_norm": 0.48409384365262775,
+      "learning_rate": 1.8580495292933258e-06,
+      "loss": 0.0115,
+      "step": 12200
+    },
+    {
+      "epoch": 0.44098526437445823,
+      "grad_norm": 0.291804821984284,
+      "learning_rate": 1.8563453825709703e-06,
+      "loss": 0.0108,
+      "step": 12210
+    },
+    {
+      "epoch": 0.4413464316671482,
+      "grad_norm": 0.47595296976543194,
+      "learning_rate": 1.8546407482574432e-06,
+      "loss": 0.0123,
+      "step": 12220
+    },
+    {
+      "epoch": 0.4417075989598382,
+      "grad_norm": 0.33628790199858916,
+      "learning_rate": 1.852935628685214e-06,
+      "loss": 0.0116,
+      "step": 12230
+    },
+    {
+      "epoch": 0.44206876625252817,
+      "grad_norm": 0.33838039531064845,
+      "learning_rate": 1.8512300261874169e-06,
+      "loss": 0.0118,
+      "step": 12240
+    },
+    {
+      "epoch": 0.44242993354521815,
+      "grad_norm": 0.33837166583351896,
+      "learning_rate": 1.8495239430978441e-06,
+      "loss": 0.0121,
+      "step": 12250
+    },
+    {
+      "epoch": 0.4427911008379081,
+      "grad_norm": 0.28850939815761484,
+      "learning_rate": 1.8478173817509487e-06,
+      "loss": 0.0114,
+      "step": 12260
+    },
+    {
+      "epoch": 0.4431522681305981,
+      "grad_norm": 0.3381917530768685,
+      "learning_rate": 1.846110344481836e-06,
+      "loss": 0.0115,
+      "step": 12270
+    },
+    {
+      "epoch": 0.4435134354232881,
+      "grad_norm": 0.3498211208387087,
+      "learning_rate": 1.844402833626264e-06,
+      "loss": 0.0119,
+      "step": 12280
+    },
+    {
+      "epoch": 0.44387460271597806,
+      "grad_norm": 0.3654831594619243,
+      "learning_rate": 1.8426948515206369e-06,
+      "loss": 0.0116,
+      "step": 12290
+    },
+    {
+      "epoch": 0.44423577000866804,
+      "grad_norm": 0.31124484715170625,
+      "learning_rate": 1.8409864005020051e-06,
+      "loss": 0.0118,
+      "step": 12300
+    },
+    {
+      "epoch": 0.44459693730135796,
+      "grad_norm": 0.311478157852358,
+      "learning_rate": 1.8392774829080606e-06,
+      "loss": 0.0121,
+      "step": 12310
+    },
+    {
+      "epoch": 0.44495810459404794,
+      "grad_norm": 0.3146483603465052,
+      "learning_rate": 1.8375681010771334e-06,
+      "loss": 0.0119,
+      "step": 12320
+    },
+    {
+      "epoch": 0.4453192718867379,
+      "grad_norm": 0.3109452051509915,
+      "learning_rate": 1.8358582573481885e-06,
+      "loss": 0.0113,
+      "step": 12330
+    },
+    {
+      "epoch": 0.4456804391794279,
+      "grad_norm": 0.2924045176164658,
+      "learning_rate": 1.8341479540608236e-06,
+      "loss": 0.0116,
+      "step": 12340
+    },
+    {
+      "epoch": 0.4460416064721179,
+      "grad_norm": 0.34507636875238323,
+      "learning_rate": 1.8324371935552643e-06,
+      "loss": 0.0111,
+      "step": 12350
+    },
+    {
+      "epoch": 0.44640277376480786,
+      "grad_norm": 0.38115192167997664,
+      "learning_rate": 1.8307259781723635e-06,
+      "loss": 0.0112,
+      "step": 12360
+    },
+    {
+      "epoch": 0.44676394105749784,
+      "grad_norm": 0.3791123739912063,
+      "learning_rate": 1.8290143102535943e-06,
+      "loss": 0.0115,
+      "step": 12370
+    },
+    {
+      "epoch": 0.4471251083501878,
+      "grad_norm": 0.33579925057815285,
+      "learning_rate": 1.82730219214105e-06,
+      "loss": 0.0111,
+      "step": 12380
+    },
+    {
+      "epoch": 0.4474862756428778,
+      "grad_norm": 0.34754181276689544,
+      "learning_rate": 1.825589626177441e-06,
+      "loss": 0.0112,
+      "step": 12390
+    },
+    {
+      "epoch": 0.4478474429355678,
+      "grad_norm": 0.3555263055896049,
+      "learning_rate": 1.8238766147060887e-06,
+      "loss": 0.0111,
+      "step": 12400
+    },
+    {
+      "epoch": 0.44820861022825775,
+      "grad_norm": 0.40697159711181113,
+      "learning_rate": 1.822163160070925e-06,
+      "loss": 0.0113,
+      "step": 12410
+    },
+    {
+      "epoch": 0.4485697775209477,
+      "grad_norm": 0.3539338521747886,
+      "learning_rate": 1.8204492646164879e-06,
+      "loss": 0.0112,
+      "step": 12420
+    },
+    {
+      "epoch": 0.44893094481363766,
+      "grad_norm": 0.5610696088874352,
+      "learning_rate": 1.8187349306879194e-06,
+      "loss": 0.0114,
+      "step": 12430
+    },
+    {
+      "epoch": 0.44929211210632763,
+      "grad_norm": 0.3560722970705039,
+      "learning_rate": 1.8170201606309608e-06,
+      "loss": 0.0111,
+      "step": 12440
+    },
+    {
+      "epoch": 0.4496532793990176,
+      "grad_norm": 0.4300667081172578,
+      "learning_rate": 1.8153049567919491e-06,
+      "loss": 0.0121,
+      "step": 12450
+    },
+    {
+      "epoch": 0.4500144466917076,
+      "grad_norm": 0.4710474787014308,
+      "learning_rate": 1.813589321517817e-06,
+      "loss": 0.0112,
+      "step": 12460
+    },
+    {
+      "epoch": 0.45037561398439757,
+      "grad_norm": 0.3116816544660811,
+      "learning_rate": 1.8118732571560864e-06,
+      "loss": 0.0115,
+      "step": 12470
+    },
+    {
+      "epoch": 0.45073678127708755,
+      "grad_norm": 0.3241365720716808,
+      "learning_rate": 1.8101567660548656e-06,
+      "loss": 0.0121,
+      "step": 12480
+    },
+    {
+      "epoch": 0.45109794856977753,
+      "grad_norm": 0.4905190586658958,
+      "learning_rate": 1.8084398505628484e-06,
+      "loss": 0.0113,
+      "step": 12490
+    },
+    {
+      "epoch": 0.4514591158624675,
+      "grad_norm": 0.33805804427731545,
+      "learning_rate": 1.806722513029308e-06,
+      "loss": 0.0107,
+      "step": 12500
+    },
+    {
+      "epoch": 0.4518202831551575,
+      "grad_norm": 0.30655143550357883,
+      "learning_rate": 1.8050047558040959e-06,
+      "loss": 0.0113,
+      "step": 12510
+    },
+    {
+      "epoch": 0.45218145044784747,
+      "grad_norm": 0.9069473925025127,
+      "learning_rate": 1.8032865812376373e-06,
+      "loss": 0.0112,
+      "step": 12520
+    },
+    {
+      "epoch": 0.4525426177405374,
+      "grad_norm": 0.5334300562030186,
+      "learning_rate": 1.8015679916809288e-06,
+      "loss": 0.0115,
+      "step": 12530
+    },
+    {
+      "epoch": 0.45290378503322737,
+      "grad_norm": 0.4162738295594657,
+      "learning_rate": 1.7998489894855352e-06,
+      "loss": 0.0109,
+      "step": 12540
+    },
+    {
+      "epoch": 0.45326495232591735,
+      "grad_norm": 0.4120251924982592,
+      "learning_rate": 1.7981295770035849e-06,
+      "loss": 0.0113,
+      "step": 12550
+    },
+    {
+      "epoch": 0.4536261196186073,
+      "grad_norm": 0.4319458974449684,
+      "learning_rate": 1.7964097565877683e-06,
+      "loss": 0.0123,
+      "step": 12560
+    },
+    {
+      "epoch": 0.4539872869112973,
+      "grad_norm": 0.30004515402137905,
+      "learning_rate": 1.7946895305913343e-06,
+      "loss": 0.0117,
+      "step": 12570
+    },
+    {
+      "epoch": 0.4543484542039873,
+      "grad_norm": 0.2537568952883938,
+      "learning_rate": 1.7929689013680858e-06,
+      "loss": 0.011,
+      "step": 12580
+    },
+    {
+      "epoch": 0.45470962149667726,
+      "grad_norm": 0.3252070310309314,
+      "learning_rate": 1.7912478712723785e-06,
+      "loss": 0.011,
+      "step": 12590
+    },
+    {
+      "epoch": 0.45507078878936724,
+      "grad_norm": 0.3823311592041792,
+      "learning_rate": 1.7895264426591165e-06,
+      "loss": 0.0116,
+      "step": 12600
+    },
+    {
+      "epoch": 0.4554319560820572,
+      "grad_norm": 0.3336845156312889,
+      "learning_rate": 1.7878046178837478e-06,
+      "loss": 0.0107,
+      "step": 12610
+    },
+    {
+      "epoch": 0.4557931233747472,
+      "grad_norm": 0.39376427766543676,
+      "learning_rate": 1.7860823993022643e-06,
+      "loss": 0.0116,
+      "step": 12620
+    },
+    {
+      "epoch": 0.4561542906674372,
+      "grad_norm": 0.3440487843296649,
+      "learning_rate": 1.784359789271196e-06,
+      "loss": 0.011,
+      "step": 12630
+    },
+    {
+      "epoch": 0.45651545796012716,
+      "grad_norm": 0.30315125422772293,
+      "learning_rate": 1.782636790147608e-06,
+      "loss": 0.0116,
+      "step": 12640
+    },
+    {
+      "epoch": 0.4568766252528171,
+      "grad_norm": 0.3270198905042072,
+      "learning_rate": 1.7809134042890984e-06,
+      "loss": 0.0111,
+      "step": 12650
+    },
+    {
+      "epoch": 0.45723779254550706,
+      "grad_norm": 0.30126511629093433,
+      "learning_rate": 1.7791896340537948e-06,
+      "loss": 0.0121,
+      "step": 12660
+    },
+    {
+      "epoch": 0.45759895983819704,
+      "grad_norm": 0.321971857852143,
+      "learning_rate": 1.7774654818003506e-06,
+      "loss": 0.0118,
+      "step": 12670
+    },
+    {
+      "epoch": 0.457960127130887,
+      "grad_norm": 0.27768620496596935,
+      "learning_rate": 1.7757409498879402e-06,
+      "loss": 0.0107,
+      "step": 12680
+    },
+    {
+      "epoch": 0.458321294423577,
+      "grad_norm": 0.2700572021956183,
+      "learning_rate": 1.7740160406762603e-06,
+      "loss": 0.0119,
+      "step": 12690
+    },
+    {
+      "epoch": 0.458682461716267,
+      "grad_norm": 0.2969652692196239,
+      "learning_rate": 1.7722907565255222e-06,
+      "loss": 0.0122,
+      "step": 12700
+    },
+    {
+      "epoch": 0.45904362900895695,
+      "grad_norm": 0.5196374195223593,
+      "learning_rate": 1.7705650997964504e-06,
+      "loss": 0.0114,
+      "step": 12710
+    },
+    {
+      "epoch": 0.45940479630164693,
+      "grad_norm": 0.5099280160817607,
+      "learning_rate": 1.768839072850279e-06,
+      "loss": 0.0109,
+      "step": 12720
+    },
+    {
+      "epoch": 0.4597659635943369,
+      "grad_norm": 0.3780096525595169,
+      "learning_rate": 1.767112678048749e-06,
+      "loss": 0.0111,
+      "step": 12730
+    },
+    {
+      "epoch": 0.4601271308870269,
+      "grad_norm": 0.31634840638495326,
+      "learning_rate": 1.7653859177541053e-06,
+      "loss": 0.0108,
+      "step": 12740
+    },
+    {
+      "epoch": 0.46048829817971687,
+      "grad_norm": 0.3964360080866963,
+      "learning_rate": 1.7636587943290926e-06,
+      "loss": 0.011,
+      "step": 12750
+    },
+    {
+      "epoch": 0.4608494654724068,
+      "grad_norm": 0.3129912322594426,
+      "learning_rate": 1.7619313101369513e-06,
+      "loss": 0.0117,
+      "step": 12760
+    },
+    {
+      "epoch": 0.4612106327650968,
+      "grad_norm": 0.2819749292559441,
+      "learning_rate": 1.7602034675414168e-06,
+      "loss": 0.0108,
+      "step": 12770
+    },
+    {
+      "epoch": 0.46157180005778675,
+      "grad_norm": 0.29540342785110285,
+      "learning_rate": 1.7584752689067143e-06,
+      "loss": 0.0108,
+      "step": 12780
+    },
+    {
+      "epoch": 0.46193296735047673,
+      "grad_norm": 0.43220759058556324,
+      "learning_rate": 1.7567467165975569e-06,
+      "loss": 0.0112,
+      "step": 12790
+    },
+    {
+      "epoch": 0.4622941346431667,
+      "grad_norm": 0.29758276512324433,
+      "learning_rate": 1.75501781297914e-06,
+      "loss": 0.0107,
+      "step": 12800
+    },
+    {
+      "epoch": 0.4626553019358567,
+      "grad_norm": 0.29727378434619856,
+      "learning_rate": 1.7532885604171417e-06,
+      "loss": 0.0115,
+      "step": 12810
+    },
+    {
+      "epoch": 0.46301646922854667,
+      "grad_norm": 0.2927335192819003,
+      "learning_rate": 1.7515589612777165e-06,
+      "loss": 0.0108,
+      "step": 12820
+    },
+    {
+      "epoch": 0.46337763652123665,
+      "grad_norm": 0.31913824414275965,
+      "learning_rate": 1.7498290179274933e-06,
+      "loss": 0.011,
+      "step": 12830
+    },
+    {
+      "epoch": 0.4637388038139266,
+      "grad_norm": 0.3753385709554931,
+      "learning_rate": 1.7480987327335715e-06,
+      "loss": 0.0119,
+      "step": 12840
+    },
+    {
+      "epoch": 0.4640999711066166,
+      "grad_norm": 0.3020198057782531,
+      "learning_rate": 1.7463681080635193e-06,
+      "loss": 0.0117,
+      "step": 12850
+    },
+    {
+      "epoch": 0.4644611383993066,
+      "grad_norm": 0.3315660491944055,
+      "learning_rate": 1.7446371462853684e-06,
+      "loss": 0.0112,
+      "step": 12860
+    },
+    {
+      "epoch": 0.4648223056919965,
+      "grad_norm": 0.32611957000258385,
+      "learning_rate": 1.7429058497676122e-06,
+      "loss": 0.0127,
+      "step": 12870
+    },
+    {
+      "epoch": 0.4651834729846865,
+      "grad_norm": 0.29663979329907536,
+      "learning_rate": 1.7411742208792024e-06,
+      "loss": 0.0117,
+      "step": 12880
+    },
+    {
+      "epoch": 0.46554464027737646,
+      "grad_norm": 0.2956765509597888,
+      "learning_rate": 1.7394422619895449e-06,
+      "loss": 0.0113,
+      "step": 12890
+    },
+    {
+      "epoch": 0.46590580757006644,
+      "grad_norm": 0.3064430979790666,
+      "learning_rate": 1.7377099754684977e-06,
+      "loss": 0.0112,
+      "step": 12900
+    },
+    {
+      "epoch": 0.4662669748627564,
+      "grad_norm": 0.33717725982205415,
+      "learning_rate": 1.7359773636863668e-06,
+      "loss": 0.0116,
+      "step": 12910
+    },
+    {
+      "epoch": 0.4666281421554464,
+      "grad_norm": 0.3117434663116077,
+      "learning_rate": 1.7342444290139032e-06,
+      "loss": 0.0109,
+      "step": 12920
+    },
+    {
+      "epoch": 0.4669893094481364,
+      "grad_norm": 0.3967284062190275,
+      "learning_rate": 1.7325111738222995e-06,
+      "loss": 0.0106,
+      "step": 12930
+    },
+    {
+      "epoch": 0.46735047674082636,
+      "grad_norm": 0.2813600485875997,
+      "learning_rate": 1.730777600483188e-06,
+      "loss": 0.011,
+      "step": 12940
+    },
+    {
+      "epoch": 0.46771164403351634,
+      "grad_norm": 0.2714383941112416,
+      "learning_rate": 1.7290437113686348e-06,
+      "loss": 0.0117,
+      "step": 12950
+    },
+    {
+      "epoch": 0.4680728113262063,
+      "grad_norm": 0.31183592868696586,
+      "learning_rate": 1.7273095088511389e-06,
+      "loss": 0.0114,
+      "step": 12960
+    },
+    {
+      "epoch": 0.4684339786188963,
+      "grad_norm": 0.265028233584014,
+      "learning_rate": 1.7255749953036284e-06,
+      "loss": 0.011,
+      "step": 12970
+    },
+    {
+      "epoch": 0.4687951459115862,
+      "grad_norm": 0.2861193377799375,
+      "learning_rate": 1.7238401730994561e-06,
+      "loss": 0.011,
+      "step": 12980
+    },
+    {
+      "epoch": 0.4691563132042762,
+      "grad_norm": 0.31302697687670955,
+      "learning_rate": 1.7221050446123981e-06,
+      "loss": 0.011,
+      "step": 12990
+    },
+    {
+      "epoch": 0.4695174804969662,
+      "grad_norm": 0.2997792865761688,
+      "learning_rate": 1.7203696122166488e-06,
+      "loss": 0.0121,
+      "step": 13000
+    },
+    {
+      "epoch": 0.4695174804969662,
+      "eval_loss": 0.21648573875427246,
+      "eval_runtime": 41.344,
+      "eval_samples_per_second": 99.071,
+      "eval_steps_per_second": 0.193,
+      "step": 13000
+    },
+    {
+      "epoch": 0.46987864778965616,
+      "grad_norm": 0.42707875196858397,
+      "learning_rate": 1.7186338782868185e-06,
+      "loss": 0.0111,
+      "step": 13010
+    },
+    {
+      "epoch": 0.47023981508234614,
+      "grad_norm": 0.32773208435988943,
+      "learning_rate": 1.7168978451979314e-06,
+      "loss": 0.0118,
+      "step": 13020
+    },
+    {
+      "epoch": 0.4706009823750361,
+      "grad_norm": 0.29995901480437526,
+      "learning_rate": 1.7151615153254192e-06,
+      "loss": 0.0103,
+      "step": 13030
+    },
+    {
+      "epoch": 0.4709621496677261,
+      "grad_norm": 0.34167438598016503,
+      "learning_rate": 1.7134248910451206e-06,
+      "loss": 0.0116,
+      "step": 13040
+    },
+    {
+      "epoch": 0.47132331696041607,
+      "grad_norm": 0.30651208884429143,
+      "learning_rate": 1.711687974733277e-06,
+      "loss": 0.0111,
+      "step": 13050
+    },
+    {
+      "epoch": 0.47168448425310605,
+      "grad_norm": 0.30614089032708397,
+      "learning_rate": 1.70995076876653e-06,
+      "loss": 0.0107,
+      "step": 13060
+    },
+    {
+      "epoch": 0.47204565154579603,
+      "grad_norm": 0.32369365787035165,
+      "learning_rate": 1.7082132755219163e-06,
+      "loss": 0.0108,
+      "step": 13070
+    },
+    {
+      "epoch": 0.472406818838486,
+      "grad_norm": 0.292287945495707,
+      "learning_rate": 1.7064754973768662e-06,
+      "loss": 0.011,
+      "step": 13080
+    },
+    {
+      "epoch": 0.472767986131176,
+      "grad_norm": 0.27037859504214157,
+      "learning_rate": 1.7047374367092008e-06,
+      "loss": 0.0106,
+      "step": 13090
+    },
+    {
+      "epoch": 0.4731291534238659,
+      "grad_norm": 0.3118953254992916,
+      "learning_rate": 1.7029990958971265e-06,
+      "loss": 0.0109,
+      "step": 13100
+    },
+    {
+      "epoch": 0.4734903207165559,
+      "grad_norm": 0.3403195827949929,
+      "learning_rate": 1.7012604773192334e-06,
+      "loss": 0.0116,
+      "step": 13110
+    },
+    {
+      "epoch": 0.47385148800924587,
+      "grad_norm": 0.31349241623133917,
+      "learning_rate": 1.699521583354492e-06,
+      "loss": 0.0108,
+      "step": 13120
+    },
+    {
+      "epoch": 0.47421265530193585,
+      "grad_norm": 0.31837895704776925,
+      "learning_rate": 1.697782416382249e-06,
+      "loss": 0.0107,
+      "step": 13130
+    },
+    {
+      "epoch": 0.4745738225946258,
+      "grad_norm": 0.2916119049663039,
+      "learning_rate": 1.696042978782226e-06,
+      "loss": 0.01,
+      "step": 13140
+    },
+    {
+      "epoch": 0.4749349898873158,
+      "grad_norm": 0.26726469171956624,
+      "learning_rate": 1.6943032729345126e-06,
+      "loss": 0.0114,
+      "step": 13150
+    },
+    {
+      "epoch": 0.4752961571800058,
+      "grad_norm": 0.32661046828588025,
+      "learning_rate": 1.6925633012195675e-06,
+      "loss": 0.0117,
+      "step": 13160
+    },
+    {
+      "epoch": 0.47565732447269576,
+      "grad_norm": 0.32227027397651176,
+      "learning_rate": 1.6908230660182127e-06,
+      "loss": 0.0107,
+      "step": 13170
+    },
+    {
+      "epoch": 0.47601849176538574,
+      "grad_norm": 0.32163601411648823,
+      "learning_rate": 1.6890825697116305e-06,
+      "loss": 0.0115,
+      "step": 13180
+    },
+    {
+      "epoch": 0.4763796590580757,
+      "grad_norm": 0.30471985141735247,
+      "learning_rate": 1.6873418146813599e-06,
+      "loss": 0.0108,
+      "step": 13190
+    },
+    {
+      "epoch": 0.4767408263507657,
+      "grad_norm": 0.3433952892415814,
+      "learning_rate": 1.6856008033092946e-06,
+      "loss": 0.0114,
+      "step": 13200
+    },
+    {
+      "epoch": 0.4771019936434556,
+      "grad_norm": 0.33871254598857686,
+      "learning_rate": 1.683859537977679e-06,
+      "loss": 0.0113,
+      "step": 13210
+    },
+    {
+      "epoch": 0.4774631609361456,
+      "grad_norm": 0.3521974001456338,
+      "learning_rate": 1.6821180210691056e-06,
+      "loss": 0.0113,
+      "step": 13220
+    },
+    {
+      "epoch": 0.4778243282288356,
+      "grad_norm": 0.2937199288898133,
+      "learning_rate": 1.6803762549665098e-06,
+      "loss": 0.0111,
+      "step": 13230
+    },
+    {
+      "epoch": 0.47818549552152556,
+      "grad_norm": 0.3172745042953924,
+      "learning_rate": 1.6786342420531687e-06,
+      "loss": 0.011,
+      "step": 13240
+    },
+    {
+      "epoch": 0.47854666281421554,
+      "grad_norm": 0.3376413397845222,
+      "learning_rate": 1.6768919847126973e-06,
+      "loss": 0.0103,
+      "step": 13250
+    },
+    {
+      "epoch": 0.4789078301069055,
+      "grad_norm": 0.3226353555967203,
+      "learning_rate": 1.6751494853290449e-06,
+      "loss": 0.0108,
+      "step": 13260
+    },
+    {
+      "epoch": 0.4792689973995955,
+      "grad_norm": 0.2452305414858768,
+      "learning_rate": 1.6734067462864914e-06,
+      "loss": 0.0099,
+      "step": 13270
+    },
+    {
+      "epoch": 0.4796301646922855,
+      "grad_norm": 0.29505884994327575,
+      "learning_rate": 1.6716637699696454e-06,
+      "loss": 0.0103,
+      "step": 13280
+    },
+    {
+      "epoch": 0.47999133198497546,
+      "grad_norm": 0.290050854551887,
+      "learning_rate": 1.6699205587634402e-06,
+      "loss": 0.0104,
+      "step": 13290
+    },
+    {
+      "epoch": 0.48035249927766543,
+      "grad_norm": 0.3234533375190963,
+      "learning_rate": 1.6681771150531302e-06,
+      "loss": 0.0102,
+      "step": 13300
+    },
+    {
+      "epoch": 0.4807136665703554,
+      "grad_norm": 0.299304548745651,
+      "learning_rate": 1.6664334412242875e-06,
+      "loss": 0.0108,
+      "step": 13310
+    },
+    {
+      "epoch": 0.48107483386304534,
+      "grad_norm": 0.30911825257887177,
+      "learning_rate": 1.6646895396627995e-06,
+      "loss": 0.0103,
+      "step": 13320
+    },
+    {
+      "epoch": 0.4814360011557353,
+      "grad_norm": 0.27877759796659024,
+      "learning_rate": 1.6629454127548656e-06,
+      "loss": 0.011,
+      "step": 13330
+    },
+    {
+      "epoch": 0.4817971684484253,
+      "grad_norm": 0.30587820022755197,
+      "learning_rate": 1.6612010628869938e-06,
+      "loss": 0.011,
+      "step": 13340
+    },
+    {
+      "epoch": 0.4821583357411153,
+      "grad_norm": 0.25304950063657494,
+      "learning_rate": 1.659456492445995e-06,
+      "loss": 0.0106,
+      "step": 13350
+    },
+    {
+      "epoch": 0.48251950303380525,
+      "grad_norm": 0.27164161943443565,
+      "learning_rate": 1.657711703818984e-06,
+      "loss": 0.0104,
+      "step": 13360
+    },
+    {
+      "epoch": 0.48288067032649523,
+      "grad_norm": 0.3565067846156723,
+      "learning_rate": 1.6559666993933743e-06,
+      "loss": 0.0103,
+      "step": 13370
+    },
+    {
+      "epoch": 0.4832418376191852,
+      "grad_norm": 0.3680614131818241,
+      "learning_rate": 1.6542214815568736e-06,
+      "loss": 0.0112,
+      "step": 13380
+    },
+    {
+      "epoch": 0.4836030049118752,
+      "grad_norm": 0.3339060873286559,
+      "learning_rate": 1.6524760526974817e-06,
+      "loss": 0.0107,
+      "step": 13390
+    },
+    {
+      "epoch": 0.48396417220456517,
+      "grad_norm": 0.3405452735294899,
+      "learning_rate": 1.6507304152034878e-06,
+      "loss": 0.0108,
+      "step": 13400
+    },
+    {
+      "epoch": 0.48432533949725515,
+      "grad_norm": 0.3456196859850254,
+      "learning_rate": 1.648984571463466e-06,
+      "loss": 0.0112,
+      "step": 13410
+    },
+    {
+      "epoch": 0.4846865067899451,
+      "grad_norm": 0.2825515095470057,
+      "learning_rate": 1.6472385238662733e-06,
+      "loss": 0.0112,
+      "step": 13420
+    },
+    {
+      "epoch": 0.48504767408263505,
+      "grad_norm": 0.2770701620421487,
+      "learning_rate": 1.6454922748010444e-06,
+      "loss": 0.0109,
+      "step": 13430
+    },
+    {
+      "epoch": 0.48540884137532503,
+      "grad_norm": 0.2905348491804557,
+      "learning_rate": 1.6437458266571909e-06,
+      "loss": 0.0107,
+      "step": 13440
+    },
+    {
+      "epoch": 0.485770008668015,
+      "grad_norm": 0.2814298645642102,
+      "learning_rate": 1.6419991818243962e-06,
+      "loss": 0.01,
+      "step": 13450
+    },
+    {
+      "epoch": 0.486131175960705,
+      "grad_norm": 0.2715244378664662,
+      "learning_rate": 1.6402523426926135e-06,
+      "loss": 0.0107,
+      "step": 13460
+    },
+    {
+      "epoch": 0.48649234325339497,
+      "grad_norm": 0.2764259088354881,
+      "learning_rate": 1.6385053116520603e-06,
+      "loss": 0.0112,
+      "step": 13470
+    },
+    {
+      "epoch": 0.48685351054608494,
+      "grad_norm": 0.33897891077591397,
+      "learning_rate": 1.6367580910932188e-06,
+      "loss": 0.0113,
+      "step": 13480
+    },
+    {
+      "epoch": 0.4872146778387749,
+      "grad_norm": 0.30728440353992476,
+      "learning_rate": 1.6350106834068284e-06,
+      "loss": 0.0099,
+      "step": 13490
+    },
+    {
+      "epoch": 0.4875758451314649,
+      "grad_norm": 0.3249733318522845,
+      "learning_rate": 1.6332630909838866e-06,
+      "loss": 0.0109,
+      "step": 13500
+    },
+    {
+      "epoch": 0.4879370124241549,
+      "grad_norm": 0.3254076714323649,
+      "learning_rate": 1.6315153162156423e-06,
+      "loss": 0.0108,
+      "step": 13510
+    },
+    {
+      "epoch": 0.48829817971684486,
+      "grad_norm": 0.2884248356433701,
+      "learning_rate": 1.6297673614935938e-06,
+      "loss": 0.011,
+      "step": 13520
+    },
+    {
+      "epoch": 0.48865934700953484,
+      "grad_norm": 0.3049821538225533,
+      "learning_rate": 1.628019229209487e-06,
+      "loss": 0.0103,
+      "step": 13530
+    },
+    {
+      "epoch": 0.4890205143022248,
+      "grad_norm": 0.3230294913685499,
+      "learning_rate": 1.626270921755309e-06,
+      "loss": 0.011,
+      "step": 13540
+    },
+    {
+      "epoch": 0.48938168159491474,
+      "grad_norm": 0.30467040704857784,
+      "learning_rate": 1.624522441523288e-06,
+      "loss": 0.01,
+      "step": 13550
+    },
+    {
+      "epoch": 0.4897428488876047,
+      "grad_norm": 0.4226948134198232,
+      "learning_rate": 1.6227737909058875e-06,
+      "loss": 0.0108,
+      "step": 13560
+    },
+    {
+      "epoch": 0.4901040161802947,
+      "grad_norm": 0.27819849976817734,
+      "learning_rate": 1.6210249722958057e-06,
+      "loss": 0.0109,
+      "step": 13570
+    },
+    {
+      "epoch": 0.4904651834729847,
+      "grad_norm": 0.35013522947437486,
+      "learning_rate": 1.6192759880859684e-06,
+      "loss": 0.0109,
+      "step": 13580
+    },
+    {
+      "epoch": 0.49082635076567466,
+      "grad_norm": 0.29096546771260506,
+      "learning_rate": 1.6175268406695308e-06,
+      "loss": 0.0101,
+      "step": 13590
+    },
+    {
+      "epoch": 0.49118751805836464,
+      "grad_norm": 0.3060844230006945,
+      "learning_rate": 1.6157775324398685e-06,
+      "loss": 0.0101,
+      "step": 13600
+    },
+    {
+      "epoch": 0.4915486853510546,
+      "grad_norm": 1.5206548943575253,
+      "learning_rate": 1.614028065790579e-06,
+      "loss": 0.0109,
+      "step": 13610
+    },
+    {
+      "epoch": 0.4919098526437446,
+      "grad_norm": 0.35332467332537143,
+      "learning_rate": 1.6122784431154765e-06,
+      "loss": 0.011,
+      "step": 13620
+    },
+    {
+      "epoch": 0.4922710199364346,
+      "grad_norm": 0.32859618335826996,
+      "learning_rate": 1.6105286668085877e-06,
+      "loss": 0.0108,
+      "step": 13630
+    },
+    {
+      "epoch": 0.49263218722912455,
+      "grad_norm": 0.24992636067454016,
+      "learning_rate": 1.6087787392641506e-06,
+      "loss": 0.0108,
+      "step": 13640
+    },
+    {
+      "epoch": 0.49299335452181453,
+      "grad_norm": 0.3085431072610162,
+      "learning_rate": 1.6070286628766095e-06,
+      "loss": 0.0105,
+      "step": 13650
+    },
+    {
+      "epoch": 0.49335452181450445,
+      "grad_norm": 0.38919494859345366,
+      "learning_rate": 1.6052784400406119e-06,
+      "loss": 0.0103,
+      "step": 13660
+    },
+    {
+      "epoch": 0.49371568910719443,
+      "grad_norm": 0.3358059747948857,
+      "learning_rate": 1.6035280731510068e-06,
+      "loss": 0.0106,
+      "step": 13670
+    },
+    {
+      "epoch": 0.4940768563998844,
+      "grad_norm": 0.3232009630836373,
+      "learning_rate": 1.6017775646028401e-06,
+      "loss": 0.011,
+      "step": 13680
+    },
+    {
+      "epoch": 0.4944380236925744,
+      "grad_norm": 0.29919418061120784,
+      "learning_rate": 1.600026916791351e-06,
+      "loss": 0.0098,
+      "step": 13690
+    },
+    {
+      "epoch": 0.49479919098526437,
+      "grad_norm": 0.28156840466022504,
+      "learning_rate": 1.5982761321119691e-06,
+      "loss": 0.0106,
+      "step": 13700
+    },
+    {
+      "epoch": 0.49516035827795435,
+      "grad_norm": 0.37608698186595535,
+      "learning_rate": 1.5965252129603122e-06,
+      "loss": 0.011,
+      "step": 13710
+    },
+    {
+      "epoch": 0.49552152557064433,
+      "grad_norm": 0.3214222642394809,
+      "learning_rate": 1.5947741617321814e-06,
+      "loss": 0.0105,
+      "step": 13720
+    },
+    {
+      "epoch": 0.4958826928633343,
+      "grad_norm": 0.2633955740274459,
+      "learning_rate": 1.5930229808235589e-06,
+      "loss": 0.0106,
+      "step": 13730
+    },
+    {
+      "epoch": 0.4962438601560243,
+      "grad_norm": 0.2985365061264536,
+      "learning_rate": 1.5912716726306034e-06,
+      "loss": 0.0105,
+      "step": 13740
+    },
+    {
+      "epoch": 0.49660502744871426,
+      "grad_norm": 0.25363355525117975,
+      "learning_rate": 1.589520239549649e-06,
+      "loss": 0.0098,
+      "step": 13750
+    },
+    {
+      "epoch": 0.49696619474140424,
+      "grad_norm": 0.28008680980321937,
+      "learning_rate": 1.5877686839772004e-06,
+      "loss": 0.0104,
+      "step": 13760
+    },
+    {
+      "epoch": 0.49732736203409417,
+      "grad_norm": 0.35914451313959717,
+      "learning_rate": 1.5860170083099298e-06,
+      "loss": 0.0108,
+      "step": 13770
+    },
+    {
+      "epoch": 0.49768852932678415,
+      "grad_norm": 0.2545712614274765,
+      "learning_rate": 1.5842652149446726e-06,
+      "loss": 0.01,
+      "step": 13780
+    },
+    {
+      "epoch": 0.4980496966194741,
+      "grad_norm": 0.3913807118896325,
+      "learning_rate": 1.582513306278427e-06,
+      "loss": 0.0104,
+      "step": 13790
+    },
+    {
+      "epoch": 0.4984108639121641,
+      "grad_norm": 0.2819313638237155,
+      "learning_rate": 1.5807612847083486e-06,
+      "loss": 0.0106,
+      "step": 13800
+    },
+    {
+      "epoch": 0.4987720312048541,
+      "grad_norm": 0.260664714666097,
+      "learning_rate": 1.5790091526317466e-06,
+      "loss": 0.0104,
+      "step": 13810
+    },
+    {
+      "epoch": 0.49913319849754406,
+      "grad_norm": 0.3823240947179682,
+      "learning_rate": 1.5772569124460817e-06,
+      "loss": 0.0107,
+      "step": 13820
+    },
+    {
+      "epoch": 0.49949436579023404,
+      "grad_norm": 0.33383430553476595,
+      "learning_rate": 1.5755045665489634e-06,
+      "loss": 0.0108,
+      "step": 13830
+    },
+    {
+      "epoch": 0.499855533082924,
+      "grad_norm": 0.2765746620935416,
+      "learning_rate": 1.5737521173381444e-06,
+      "loss": 0.0107,
+      "step": 13840
+    },
+    {
+      "epoch": 0.5002167003756139,
+      "grad_norm": 0.3106588971781434,
+      "learning_rate": 1.5719995672115208e-06,
+      "loss": 0.0102,
+      "step": 13850
+    },
+    {
+      "epoch": 0.5005778676683039,
+      "grad_norm": 0.43777147083672896,
+      "learning_rate": 1.570246918567124e-06,
+      "loss": 0.0108,
+      "step": 13860
+    },
+    {
+      "epoch": 0.5009390349609939,
+      "grad_norm": 0.30639685845037784,
+      "learning_rate": 1.5684941738031228e-06,
+      "loss": 0.0106,
+      "step": 13870
+    },
+    {
+      "epoch": 0.5013002022536839,
+      "grad_norm": 0.25916272071943547,
+      "learning_rate": 1.5667413353178162e-06,
+      "loss": 0.0107,
+      "step": 13880
+    },
+    {
+      "epoch": 0.5016613695463739,
+      "grad_norm": 0.2598367946718612,
+      "learning_rate": 1.5649884055096324e-06,
+      "loss": 0.0105,
+      "step": 13890
+    },
+    {
+      "epoch": 0.5020225368390638,
+      "grad_norm": 0.3257167499065903,
+      "learning_rate": 1.5632353867771224e-06,
+      "loss": 0.0097,
+      "step": 13900
+    },
+    {
+      "epoch": 0.5023837041317538,
+      "grad_norm": 0.3341498450446017,
+      "learning_rate": 1.5614822815189612e-06,
+      "loss": 0.0104,
+      "step": 13910
+    },
+    {
+      "epoch": 0.5027448714244438,
+      "grad_norm": 0.2716418115879241,
+      "learning_rate": 1.5597290921339417e-06,
+      "loss": 0.0108,
+      "step": 13920
+    },
+    {
+      "epoch": 0.5031060387171338,
+      "grad_norm": 0.2562260798937972,
+      "learning_rate": 1.557975821020971e-06,
+      "loss": 0.0109,
+      "step": 13930
+    },
+    {
+      "epoch": 0.5034672060098238,
+      "grad_norm": 0.2675083915438243,
+      "learning_rate": 1.5562224705790686e-06,
+      "loss": 0.01,
+      "step": 13940
+    },
+    {
+      "epoch": 0.5038283733025137,
+      "grad_norm": 0.2848371819915795,
+      "learning_rate": 1.5544690432073627e-06,
+      "loss": 0.0105,
+      "step": 13950
+    },
+    {
+      "epoch": 0.5041895405952037,
+      "grad_norm": 0.30299960535255815,
+      "learning_rate": 1.5527155413050864e-06,
+      "loss": 0.0103,
+      "step": 13960
+    },
+    {
+      "epoch": 0.5045507078878937,
+      "grad_norm": 0.2673930553167057,
+      "learning_rate": 1.5509619672715751e-06,
+      "loss": 0.0102,
+      "step": 13970
+    },
+    {
+      "epoch": 0.5049118751805837,
+      "grad_norm": 0.3594407646696452,
+      "learning_rate": 1.5492083235062628e-06,
+      "loss": 0.0098,
+      "step": 13980
+    },
+    {
+      "epoch": 0.5052730424732736,
+      "grad_norm": 0.2795945333473362,
+      "learning_rate": 1.5474546124086791e-06,
+      "loss": 0.0102,
+      "step": 13990
+    },
+    {
+      "epoch": 0.5056342097659636,
+      "grad_norm": 0.37917403674934363,
+      "learning_rate": 1.5457008363784456e-06,
+      "loss": 0.0111,
+      "step": 14000
+    },
+    {
+      "epoch": 0.5056342097659636,
+      "eval_loss": 0.22100868821144104,
+      "eval_runtime": 39.1055,
+      "eval_samples_per_second": 104.742,
+      "eval_steps_per_second": 0.205,
+      "step": 14000
+    },
+    {
+      "epoch": 0.5059953770586536,
+      "grad_norm": 0.32921355095807997,
+      "learning_rate": 1.5439469978152716e-06,
+      "loss": 0.0095,
+      "step": 14010
+    },
+    {
+      "epoch": 0.5063565443513436,
+      "grad_norm": 0.267687251662146,
+      "learning_rate": 1.5421930991189544e-06,
+      "loss": 0.0102,
+      "step": 14020
+    },
+    {
+      "epoch": 0.5067177116440336,
+      "grad_norm": 0.29358778350688636,
+      "learning_rate": 1.5404391426893713e-06,
+      "loss": 0.0099,
+      "step": 14030
+    },
+    {
+      "epoch": 0.5070788789367235,
+      "grad_norm": 0.32020996839099647,
+      "learning_rate": 1.53868513092648e-06,
+      "loss": 0.0106,
+      "step": 14040
+    },
+    {
+      "epoch": 0.5074400462294134,
+      "grad_norm": 0.29738990777062035,
+      "learning_rate": 1.5369310662303132e-06,
+      "loss": 0.01,
+      "step": 14050
+    },
+    {
+      "epoch": 0.5078012135221034,
+      "grad_norm": 0.23409034882222773,
+      "learning_rate": 1.535176951000976e-06,
+      "loss": 0.0106,
+      "step": 14060
+    },
+    {
+      "epoch": 0.5081623808147934,
+      "grad_norm": 0.23954429550296077,
+      "learning_rate": 1.5334227876386433e-06,
+      "loss": 0.0102,
+      "step": 14070
+    },
+    {
+      "epoch": 0.5085235481074833,
+      "grad_norm": 0.28141652425230024,
+      "learning_rate": 1.531668578543556e-06,
+      "loss": 0.0103,
+      "step": 14080
+    },
+    {
+      "epoch": 0.5088847154001733,
+      "grad_norm": 0.24828709508354818,
+      "learning_rate": 1.5299143261160156e-06,
+      "loss": 0.011,
+      "step": 14090
+    },
+    {
+      "epoch": 0.5092458826928633,
+      "grad_norm": 0.2899718548992346,
+      "learning_rate": 1.5281600327563853e-06,
+      "loss": 0.0107,
+      "step": 14100
+    },
+    {
+      "epoch": 0.5096070499855533,
+      "grad_norm": 0.33647779981641485,
+      "learning_rate": 1.526405700865083e-06,
+      "loss": 0.0103,
+      "step": 14110
+    },
+    {
+      "epoch": 0.5099682172782433,
+      "grad_norm": 0.29019722996427644,
+      "learning_rate": 1.5246513328425804e-06,
+      "loss": 0.0101,
+      "step": 14120
+    },
+    {
+      "epoch": 0.5103293845709332,
+      "grad_norm": 0.318393229061606,
+      "learning_rate": 1.522896931089397e-06,
+      "loss": 0.0103,
+      "step": 14130
+    },
+    {
+      "epoch": 0.5106905518636232,
+      "grad_norm": 0.35309232587876205,
+      "learning_rate": 1.5211424980060992e-06,
+      "loss": 0.01,
+      "step": 14140
+    },
+    {
+      "epoch": 0.5110517191563132,
+      "grad_norm": 0.29962561160352535,
+      "learning_rate": 1.5193880359932972e-06,
+      "loss": 0.0097,
+      "step": 14150
+    },
+    {
+      "epoch": 0.5114128864490032,
+      "grad_norm": 0.37249233123144315,
+      "learning_rate": 1.5176335474516398e-06,
+      "loss": 0.0101,
+      "step": 14160
+    },
+    {
+      "epoch": 0.5117740537416932,
+      "grad_norm": 0.2903989879850118,
+      "learning_rate": 1.5158790347818118e-06,
+      "loss": 0.0099,
+      "step": 14170
+    },
+    {
+      "epoch": 0.5121352210343831,
+      "grad_norm": 0.3088066901718896,
+      "learning_rate": 1.514124500384532e-06,
+      "loss": 0.0103,
+      "step": 14180
+    },
+    {
+      "epoch": 0.5124963883270731,
+      "grad_norm": 0.24842034153796524,
+      "learning_rate": 1.5123699466605482e-06,
+      "loss": 0.0103,
+      "step": 14190
+    },
+    {
+      "epoch": 0.5128575556197631,
+      "grad_norm": 0.3068063725395222,
+      "learning_rate": 1.5106153760106356e-06,
+      "loss": 0.0107,
+      "step": 14200
+    },
+    {
+      "epoch": 0.5132187229124531,
+      "grad_norm": 0.31446894314109364,
+      "learning_rate": 1.5088607908355907e-06,
+      "loss": 0.0101,
+      "step": 14210
+    },
+    {
+      "epoch": 0.513579890205143,
+      "grad_norm": 0.8363054692442118,
+      "learning_rate": 1.5071061935362318e-06,
+      "loss": 0.0105,
+      "step": 14220
+    },
+    {
+      "epoch": 0.513941057497833,
+      "grad_norm": 0.3145227216620044,
+      "learning_rate": 1.5053515865133928e-06,
+      "loss": 0.0105,
+      "step": 14230
+    },
+    {
+      "epoch": 0.514302224790523,
+      "grad_norm": 0.2578260158550241,
+      "learning_rate": 1.5035969721679212e-06,
+      "loss": 0.0103,
+      "step": 14240
+    },
+    {
+      "epoch": 0.514663392083213,
+      "grad_norm": 0.2761107569371966,
+      "learning_rate": 1.5018423529006743e-06,
+      "loss": 0.0098,
+      "step": 14250
+    },
+    {
+      "epoch": 0.515024559375903,
+      "grad_norm": 0.32247905192514087,
+      "learning_rate": 1.5000877311125164e-06,
+      "loss": 0.0103,
+      "step": 14260
+    },
+    {
+      "epoch": 0.515385726668593,
+      "grad_norm": 0.24655172170012937,
+      "learning_rate": 1.4983331092043152e-06,
+      "loss": 0.01,
+      "step": 14270
+    },
+    {
+      "epoch": 0.5157468939612828,
+      "grad_norm": 0.2509238376592253,
+      "learning_rate": 1.496578489576938e-06,
+      "loss": 0.01,
+      "step": 14280
+    },
+    {
+      "epoch": 0.5161080612539728,
+      "grad_norm": 0.26198848860303103,
+      "learning_rate": 1.49482387463125e-06,
+      "loss": 0.0098,
+      "step": 14290
+    },
+    {
+      "epoch": 0.5164692285466628,
+      "grad_norm": 0.37310101632546216,
+      "learning_rate": 1.4930692667681096e-06,
+      "loss": 0.0103,
+      "step": 14300
+    },
+    {
+      "epoch": 0.5168303958393528,
+      "grad_norm": 0.3292157004526728,
+      "learning_rate": 1.491314668388365e-06,
+      "loss": 0.0097,
+      "step": 14310
+    },
+    {
+      "epoch": 0.5171915631320427,
+      "grad_norm": 0.37140956602853614,
+      "learning_rate": 1.4895600818928515e-06,
+      "loss": 0.0104,
+      "step": 14320
+    },
+    {
+      "epoch": 0.5175527304247327,
+      "grad_norm": 0.25852311715569865,
+      "learning_rate": 1.4878055096823892e-06,
+      "loss": 0.0099,
+      "step": 14330
+    },
+    {
+      "epoch": 0.5179138977174227,
+      "grad_norm": 0.3757869792417407,
+      "learning_rate": 1.4860509541577773e-06,
+      "loss": 0.0097,
+      "step": 14340
+    },
+    {
+      "epoch": 0.5182750650101127,
+      "grad_norm": 0.2807728763963555,
+      "learning_rate": 1.4842964177197933e-06,
+      "loss": 0.0097,
+      "step": 14350
+    },
+    {
+      "epoch": 0.5186362323028026,
+      "grad_norm": 0.2849024526542563,
+      "learning_rate": 1.4825419027691872e-06,
+      "loss": 0.0101,
+      "step": 14360
+    },
+    {
+      "epoch": 0.5189973995954926,
+      "grad_norm": 0.28167848185783206,
+      "learning_rate": 1.4807874117066815e-06,
+      "loss": 0.0106,
+      "step": 14370
+    },
+    {
+      "epoch": 0.5193585668881826,
+      "grad_norm": 0.2813414290427883,
+      "learning_rate": 1.479032946932964e-06,
+      "loss": 0.0103,
+      "step": 14380
+    },
+    {
+      "epoch": 0.5197197341808726,
+      "grad_norm": 0.24612637682691624,
+      "learning_rate": 1.477278510848688e-06,
+      "loss": 0.0098,
+      "step": 14390
+    },
+    {
+      "epoch": 0.5200809014735626,
+      "grad_norm": 0.3277951445020074,
+      "learning_rate": 1.4755241058544667e-06,
+      "loss": 0.0099,
+      "step": 14400
+    },
+    {
+      "epoch": 0.5204420687662525,
+      "grad_norm": 0.25249313064282414,
+      "learning_rate": 1.4737697343508714e-06,
+      "loss": 0.0099,
+      "step": 14410
+    },
+    {
+      "epoch": 0.5208032360589425,
+      "grad_norm": 0.2650465862873837,
+      "learning_rate": 1.472015398738427e-06,
+      "loss": 0.0104,
+      "step": 14420
+    },
+    {
+      "epoch": 0.5211644033516325,
+      "grad_norm": 0.2755409480532333,
+      "learning_rate": 1.4702611014176099e-06,
+      "loss": 0.0107,
+      "step": 14430
+    },
+    {
+      "epoch": 0.5215255706443225,
+      "grad_norm": 0.2682805458384964,
+      "learning_rate": 1.4685068447888433e-06,
+      "loss": 0.0097,
+      "step": 14440
+    },
+    {
+      "epoch": 0.5218867379370125,
+      "grad_norm": 0.2831372810986366,
+      "learning_rate": 1.4667526312524962e-06,
+      "loss": 0.0101,
+      "step": 14450
+    },
+    {
+      "epoch": 0.5222479052297024,
+      "grad_norm": 0.30612266372899305,
+      "learning_rate": 1.4649984632088762e-06,
+      "loss": 0.0102,
+      "step": 14460
+    },
+    {
+      "epoch": 0.5226090725223924,
+      "grad_norm": 0.3124564650002226,
+      "learning_rate": 1.4632443430582306e-06,
+      "loss": 0.01,
+      "step": 14470
+    },
+    {
+      "epoch": 0.5229702398150824,
+      "grad_norm": 0.30767756856194173,
+      "learning_rate": 1.4614902732007408e-06,
+      "loss": 0.0098,
+      "step": 14480
+    },
+    {
+      "epoch": 0.5233314071077724,
+      "grad_norm": 0.2827883997201114,
+      "learning_rate": 1.459736256036519e-06,
+      "loss": 0.01,
+      "step": 14490
+    },
+    {
+      "epoch": 0.5236925744004622,
+      "grad_norm": 0.34924570193811233,
+      "learning_rate": 1.4579822939656058e-06,
+      "loss": 0.0105,
+      "step": 14500
+    },
+    {
+      "epoch": 0.5240537416931522,
+      "grad_norm": 0.2763985208894854,
+      "learning_rate": 1.4562283893879657e-06,
+      "loss": 0.01,
+      "step": 14510
+    },
+    {
+      "epoch": 0.5244149089858422,
+      "grad_norm": 0.27642482689746106,
+      "learning_rate": 1.4544745447034856e-06,
+      "loss": 0.0102,
+      "step": 14520
+    },
+    {
+      "epoch": 0.5247760762785322,
+      "grad_norm": 0.2620067609291681,
+      "learning_rate": 1.4527207623119699e-06,
+      "loss": 0.0101,
+      "step": 14530
+    },
+    {
+      "epoch": 0.5251372435712222,
+      "grad_norm": 0.2722046866126832,
+      "learning_rate": 1.4509670446131367e-06,
+      "loss": 0.0104,
+      "step": 14540
+    },
+    {
+      "epoch": 0.5254984108639121,
+      "grad_norm": 0.30688934101365084,
+      "learning_rate": 1.449213394006617e-06,
+      "loss": 0.0095,
+      "step": 14550
+    },
+    {
+      "epoch": 0.5258595781566021,
+      "grad_norm": 0.27430449531440454,
+      "learning_rate": 1.4474598128919494e-06,
+      "loss": 0.0101,
+      "step": 14560
+    },
+    {
+      "epoch": 0.5262207454492921,
+      "grad_norm": 0.3511332863893882,
+      "learning_rate": 1.445706303668578e-06,
+      "loss": 0.0107,
+      "step": 14570
+    },
+    {
+      "epoch": 0.5265819127419821,
+      "grad_norm": 0.3650598041749963,
+      "learning_rate": 1.4439528687358478e-06,
+      "loss": 0.0101,
+      "step": 14580
+    },
+    {
+      "epoch": 0.526943080034672,
+      "grad_norm": 0.3131148334579366,
+      "learning_rate": 1.4421995104930025e-06,
+      "loss": 0.0101,
+      "step": 14590
+    },
+    {
+      "epoch": 0.527304247327362,
+      "grad_norm": 0.36861361458034614,
+      "learning_rate": 1.4404462313391807e-06,
+      "loss": 0.01,
+      "step": 14600
+    },
+    {
+      "epoch": 0.527665414620052,
+      "grad_norm": 0.37590043680853025,
+      "learning_rate": 1.4386930336734138e-06,
+      "loss": 0.0102,
+      "step": 14610
+    },
+    {
+      "epoch": 0.528026581912742,
+      "grad_norm": 0.2917213026608246,
+      "learning_rate": 1.4369399198946193e-06,
+      "loss": 0.0106,
+      "step": 14620
+    },
+    {
+      "epoch": 0.528387749205432,
+      "grad_norm": 0.274178981071542,
+      "learning_rate": 1.4351868924016018e-06,
+      "loss": 0.0096,
+      "step": 14630
+    },
+    {
+      "epoch": 0.528748916498122,
+      "grad_norm": 0.3906661557219609,
+      "learning_rate": 1.433433953593048e-06,
+      "loss": 0.0101,
+      "step": 14640
+    },
+    {
+      "epoch": 0.5291100837908119,
+      "grad_norm": 0.2996891514368005,
+      "learning_rate": 1.4316811058675224e-06,
+      "loss": 0.0101,
+      "step": 14650
+    },
+    {
+      "epoch": 0.5294712510835019,
+      "grad_norm": 0.24820207707881337,
+      "learning_rate": 1.429928351623465e-06,
+      "loss": 0.0096,
+      "step": 14660
+    },
+    {
+      "epoch": 0.5298324183761919,
+      "grad_norm": 0.3249716164802091,
+      "learning_rate": 1.4281756932591884e-06,
+      "loss": 0.0102,
+      "step": 14670
+    },
+    {
+      "epoch": 0.5301935856688819,
+      "grad_norm": 0.34261709260208856,
+      "learning_rate": 1.4264231331728733e-06,
+      "loss": 0.0099,
+      "step": 14680
+    },
+    {
+      "epoch": 0.5305547529615718,
+      "grad_norm": 0.2998680212076015,
+      "learning_rate": 1.4246706737625667e-06,
+      "loss": 0.0094,
+      "step": 14690
+    },
+    {
+      "epoch": 0.5309159202542618,
+      "grad_norm": 0.24678811495395428,
+      "learning_rate": 1.4229183174261768e-06,
+      "loss": 0.0093,
+      "step": 14700
+    },
+    {
+      "epoch": 0.5312770875469518,
+      "grad_norm": 0.3918429823574824,
+      "learning_rate": 1.4211660665614719e-06,
+      "loss": 0.0107,
+      "step": 14710
+    },
+    {
+      "epoch": 0.5316382548396418,
+      "grad_norm": 0.32413050127108006,
+      "learning_rate": 1.4194139235660752e-06,
+      "loss": 0.0092,
+      "step": 14720
+    },
+    {
+      "epoch": 0.5319994221323316,
+      "grad_norm": 0.329692426525019,
+      "learning_rate": 1.4176618908374626e-06,
+      "loss": 0.01,
+      "step": 14730
+    },
+    {
+      "epoch": 0.5323605894250216,
+      "grad_norm": 0.3208968014633852,
+      "learning_rate": 1.4159099707729593e-06,
+      "loss": 0.0096,
+      "step": 14740
+    },
+    {
+      "epoch": 0.5327217567177116,
+      "grad_norm": 0.28420081892096777,
+      "learning_rate": 1.4141581657697358e-06,
+      "loss": 0.0104,
+      "step": 14750
+    },
+    {
+      "epoch": 0.5330829240104016,
+      "grad_norm": 0.3154855216552726,
+      "learning_rate": 1.412406478224806e-06,
+      "loss": 0.0103,
+      "step": 14760
+    },
+    {
+      "epoch": 0.5334440913030916,
+      "grad_norm": 0.3168993876985505,
+      "learning_rate": 1.4106549105350227e-06,
+      "loss": 0.0098,
+      "step": 14770
+    },
+    {
+      "epoch": 0.5338052585957815,
+      "grad_norm": 0.2895675546820685,
+      "learning_rate": 1.4089034650970733e-06,
+      "loss": 0.0096,
+      "step": 14780
+    },
+    {
+      "epoch": 0.5341664258884715,
+      "grad_norm": 0.29140366547021257,
+      "learning_rate": 1.4071521443074803e-06,
+      "loss": 0.0105,
+      "step": 14790
+    },
+    {
+      "epoch": 0.5345275931811615,
+      "grad_norm": 0.3657076830116289,
+      "learning_rate": 1.4054009505625942e-06,
+      "loss": 0.0103,
+      "step": 14800
+    },
+    {
+      "epoch": 0.5348887604738515,
+      "grad_norm": 0.33198361812322424,
+      "learning_rate": 1.4036498862585917e-06,
+      "loss": 0.0099,
+      "step": 14810
+    },
+    {
+      "epoch": 0.5352499277665415,
+      "grad_norm": 0.5562512024228276,
+      "learning_rate": 1.4018989537914733e-06,
+      "loss": 0.0094,
+      "step": 14820
+    },
+    {
+      "epoch": 0.5356110950592314,
+      "grad_norm": 0.3407861829591437,
+      "learning_rate": 1.4001481555570582e-06,
+      "loss": 0.0103,
+      "step": 14830
+    },
+    {
+      "epoch": 0.5359722623519214,
+      "grad_norm": 0.2675362811612675,
+      "learning_rate": 1.3983974939509827e-06,
+      "loss": 0.0094,
+      "step": 14840
+    },
+    {
+      "epoch": 0.5363334296446114,
+      "grad_norm": 0.257972193951641,
+      "learning_rate": 1.3966469713686947e-06,
+      "loss": 0.0098,
+      "step": 14850
+    },
+    {
+      "epoch": 0.5366945969373014,
+      "grad_norm": 0.3247153894709782,
+      "learning_rate": 1.3948965902054534e-06,
+      "loss": 0.0099,
+      "step": 14860
+    },
+    {
+      "epoch": 0.5370557642299914,
+      "grad_norm": 0.286452349983971,
+      "learning_rate": 1.3931463528563237e-06,
+      "loss": 0.0103,
+      "step": 14870
+    },
+    {
+      "epoch": 0.5374169315226813,
+      "grad_norm": 0.3060499956864613,
+      "learning_rate": 1.3913962617161742e-06,
+      "loss": 0.0099,
+      "step": 14880
+    },
+    {
+      "epoch": 0.5377780988153713,
+      "grad_norm": 0.3686821486426744,
+      "learning_rate": 1.389646319179673e-06,
+      "loss": 0.009,
+      "step": 14890
+    },
+    {
+      "epoch": 0.5381392661080613,
+      "grad_norm": 0.368357839538451,
+      "learning_rate": 1.3878965276412853e-06,
+      "loss": 0.0096,
+      "step": 14900
+    },
+    {
+      "epoch": 0.5385004334007513,
+      "grad_norm": 0.37338028208342816,
+      "learning_rate": 1.3861468894952694e-06,
+      "loss": 0.0101,
+      "step": 14910
+    },
+    {
+      "epoch": 0.5388616006934412,
+      "grad_norm": 0.3721193043590535,
+      "learning_rate": 1.384397407135674e-06,
+      "loss": 0.01,
+      "step": 14920
+    },
+    {
+      "epoch": 0.5392227679861312,
+      "grad_norm": 0.2969059654844757,
+      "learning_rate": 1.3826480829563337e-06,
+      "loss": 0.0092,
+      "step": 14930
+    },
+    {
+      "epoch": 0.5395839352788212,
+      "grad_norm": 0.39773634701324795,
+      "learning_rate": 1.3808989193508675e-06,
+      "loss": 0.0095,
+      "step": 14940
+    },
+    {
+      "epoch": 0.5399451025715111,
+      "grad_norm": 0.3161665643249287,
+      "learning_rate": 1.3791499187126748e-06,
+      "loss": 0.0101,
+      "step": 14950
+    },
+    {
+      "epoch": 0.540306269864201,
+      "grad_norm": 0.44342290473221385,
+      "learning_rate": 1.3774010834349317e-06,
+      "loss": 0.0092,
+      "step": 14960
+    },
+    {
+      "epoch": 0.540667437156891,
+      "grad_norm": 0.29627663576435403,
+      "learning_rate": 1.3756524159105883e-06,
+      "loss": 0.0094,
+      "step": 14970
+    },
+    {
+      "epoch": 0.541028604449581,
+      "grad_norm": 0.3263240029643017,
+      "learning_rate": 1.3739039185323643e-06,
+      "loss": 0.0097,
+      "step": 14980
+    },
+    {
+      "epoch": 0.541389771742271,
+      "grad_norm": 0.32983295853013334,
+      "learning_rate": 1.3721555936927477e-06,
+      "loss": 0.0099,
+      "step": 14990
+    },
+    {
+      "epoch": 0.541750939034961,
+      "grad_norm": 0.3062725486856836,
+      "learning_rate": 1.3704074437839904e-06,
+      "loss": 0.0095,
+      "step": 15000
+    },
+    {
+      "epoch": 0.541750939034961,
+      "eval_loss": 0.2289300113916397,
+      "eval_runtime": 38.8055,
+      "eval_samples_per_second": 105.552,
+      "eval_steps_per_second": 0.206,
+      "step": 15000
+    },
+    {
+      "epoch": 0.5421121063276509,
+      "grad_norm": 0.3600702993336002,
+      "learning_rate": 1.3686594711981036e-06,
+      "loss": 0.0107,
+      "step": 15010
+    },
+    {
+      "epoch": 0.5424732736203409,
+      "grad_norm": 0.3742463533073402,
+      "learning_rate": 1.3669116783268573e-06,
+      "loss": 0.01,
+      "step": 15020
+    },
+    {
+      "epoch": 0.5428344409130309,
+      "grad_norm": 0.33594779442493394,
+      "learning_rate": 1.365164067561775e-06,
+      "loss": 0.01,
+      "step": 15030
+    },
+    {
+      "epoch": 0.5431956082057209,
+      "grad_norm": 0.27151711863131855,
+      "learning_rate": 1.3634166412941306e-06,
+      "loss": 0.0098,
+      "step": 15040
+    },
+    {
+      "epoch": 0.5435567754984109,
+      "grad_norm": 0.41902150110711217,
+      "learning_rate": 1.3616694019149464e-06,
+      "loss": 0.0095,
+      "step": 15050
+    },
+    {
+      "epoch": 0.5439179427911008,
+      "grad_norm": 0.4365908277376003,
+      "learning_rate": 1.3599223518149886e-06,
+      "loss": 0.0097,
+      "step": 15060
+    },
+    {
+      "epoch": 0.5442791100837908,
+      "grad_norm": 0.7265745787050304,
+      "learning_rate": 1.3581754933847642e-06,
+      "loss": 0.0098,
+      "step": 15070
+    },
+    {
+      "epoch": 0.5446402773764808,
+      "grad_norm": 0.34929042669540683,
+      "learning_rate": 1.356428829014519e-06,
+      "loss": 0.0097,
+      "step": 15080
+    },
+    {
+      "epoch": 0.5450014446691708,
+      "grad_norm": 0.3518994609073389,
+      "learning_rate": 1.354682361094231e-06,
+      "loss": 0.0106,
+      "step": 15090
+    },
+    {
+      "epoch": 0.5453626119618608,
+      "grad_norm": 0.3053798910811591,
+      "learning_rate": 1.3529360920136115e-06,
+      "loss": 0.0092,
+      "step": 15100
+    },
+    {
+      "epoch": 0.5457237792545507,
+      "grad_norm": 0.44094767020307124,
+      "learning_rate": 1.3511900241620991e-06,
+      "loss": 0.0092,
+      "step": 15110
+    },
+    {
+      "epoch": 0.5460849465472407,
+      "grad_norm": 0.2883771273779214,
+      "learning_rate": 1.3494441599288571e-06,
+      "loss": 0.0099,
+      "step": 15120
+    },
+    {
+      "epoch": 0.5464461138399307,
+      "grad_norm": 0.40052552248259,
+      "learning_rate": 1.3476985017027699e-06,
+      "loss": 0.0102,
+      "step": 15130
+    },
+    {
+      "epoch": 0.5468072811326207,
+      "grad_norm": 0.4672416496599479,
+      "learning_rate": 1.3459530518724399e-06,
+      "loss": 0.0096,
+      "step": 15140
+    },
+    {
+      "epoch": 0.5471684484253106,
+      "grad_norm": 0.3588555618491316,
+      "learning_rate": 1.3442078128261853e-06,
+      "loss": 0.0098,
+      "step": 15150
+    },
+    {
+      "epoch": 0.5475296157180006,
+      "grad_norm": 0.2994460295153179,
+      "learning_rate": 1.3424627869520351e-06,
+      "loss": 0.0096,
+      "step": 15160
+    },
+    {
+      "epoch": 0.5478907830106905,
+      "grad_norm": 0.359421358673954,
+      "learning_rate": 1.340717976637726e-06,
+      "loss": 0.0096,
+      "step": 15170
+    },
+    {
+      "epoch": 0.5482519503033805,
+      "grad_norm": 0.30823892459733665,
+      "learning_rate": 1.3389733842707012e-06,
+      "loss": 0.0095,
+      "step": 15180
+    },
+    {
+      "epoch": 0.5486131175960705,
+      "grad_norm": 0.2949328827872679,
+      "learning_rate": 1.3372290122381042e-06,
+      "loss": 0.0089,
+      "step": 15190
+    },
+    {
+      "epoch": 0.5489742848887604,
+      "grad_norm": 0.41197244667643934,
+      "learning_rate": 1.3354848629267787e-06,
+      "loss": 0.0098,
+      "step": 15200
+    },
+    {
+      "epoch": 0.5493354521814504,
+      "grad_norm": 0.3393284170512199,
+      "learning_rate": 1.333740938723262e-06,
+      "loss": 0.0096,
+      "step": 15210
+    },
+    {
+      "epoch": 0.5496966194741404,
+      "grad_norm": 0.29723383795925046,
+      "learning_rate": 1.331997242013785e-06,
+      "loss": 0.0099,
+      "step": 15220
+    },
+    {
+      "epoch": 0.5500577867668304,
+      "grad_norm": 0.3006329589272621,
+      "learning_rate": 1.3302537751842654e-06,
+      "loss": 0.0092,
+      "step": 15230
+    },
+    {
+      "epoch": 0.5504189540595203,
+      "grad_norm": 0.3989787099484312,
+      "learning_rate": 1.3285105406203085e-06,
+      "loss": 0.0099,
+      "step": 15240
+    },
+    {
+      "epoch": 0.5507801213522103,
+      "grad_norm": 0.35691137382770866,
+      "learning_rate": 1.3267675407071994e-06,
+      "loss": 0.0103,
+      "step": 15250
+    },
+    {
+      "epoch": 0.5511412886449003,
+      "grad_norm": 0.48955119596964847,
+      "learning_rate": 1.3250247778299038e-06,
+      "loss": 0.0096,
+      "step": 15260
+    },
+    {
+      "epoch": 0.5515024559375903,
+      "grad_norm": 0.3487378760989221,
+      "learning_rate": 1.323282254373063e-06,
+      "loss": 0.0096,
+      "step": 15270
+    },
+    {
+      "epoch": 0.5518636232302803,
+      "grad_norm": 0.3164752351362649,
+      "learning_rate": 1.3215399727209903e-06,
+      "loss": 0.0101,
+      "step": 15280
+    },
+    {
+      "epoch": 0.5522247905229702,
+      "grad_norm": 0.30874124280657733,
+      "learning_rate": 1.319797935257668e-06,
+      "loss": 0.0093,
+      "step": 15290
+    },
+    {
+      "epoch": 0.5525859578156602,
+      "grad_norm": 0.3212719296663023,
+      "learning_rate": 1.3180561443667444e-06,
+      "loss": 0.0097,
+      "step": 15300
+    },
+    {
+      "epoch": 0.5529471251083502,
+      "grad_norm": 0.43805648325816515,
+      "learning_rate": 1.3163146024315316e-06,
+      "loss": 0.0096,
+      "step": 15310
+    },
+    {
+      "epoch": 0.5533082924010402,
+      "grad_norm": 0.3713534474182569,
+      "learning_rate": 1.314573311834998e-06,
+      "loss": 0.01,
+      "step": 15320
+    },
+    {
+      "epoch": 0.5536694596937302,
+      "grad_norm": 0.3136856717903221,
+      "learning_rate": 1.3128322749597708e-06,
+      "loss": 0.0102,
+      "step": 15330
+    },
+    {
+      "epoch": 0.5540306269864201,
+      "grad_norm": 0.2662161180486923,
+      "learning_rate": 1.31109149418813e-06,
+      "loss": 0.0097,
+      "step": 15340
+    },
+    {
+      "epoch": 0.5543917942791101,
+      "grad_norm": 0.3210731667340902,
+      "learning_rate": 1.3093509719020033e-06,
+      "loss": 0.0094,
+      "step": 15350
+    },
+    {
+      "epoch": 0.5547529615718001,
+      "grad_norm": 0.4796097945986582,
+      "learning_rate": 1.3076107104829664e-06,
+      "loss": 0.0097,
+      "step": 15360
+    },
+    {
+      "epoch": 0.5551141288644901,
+      "grad_norm": 0.3665332274977178,
+      "learning_rate": 1.3058707123122374e-06,
+      "loss": 0.0096,
+      "step": 15370
+    },
+    {
+      "epoch": 0.55547529615718,
+      "grad_norm": 0.3350284087902722,
+      "learning_rate": 1.3041309797706744e-06,
+      "loss": 0.0097,
+      "step": 15380
+    },
+    {
+      "epoch": 0.55583646344987,
+      "grad_norm": 0.23483595272873672,
+      "learning_rate": 1.302391515238772e-06,
+      "loss": 0.0092,
+      "step": 15390
+    },
+    {
+      "epoch": 0.5561976307425599,
+      "grad_norm": 0.32586934687921676,
+      "learning_rate": 1.3006523210966575e-06,
+      "loss": 0.0097,
+      "step": 15400
+    },
+    {
+      "epoch": 0.5565587980352499,
+      "grad_norm": 0.2777961588166937,
+      "learning_rate": 1.2989133997240893e-06,
+      "loss": 0.0099,
+      "step": 15410
+    },
+    {
+      "epoch": 0.5569199653279399,
+      "grad_norm": 0.3272739464312555,
+      "learning_rate": 1.2971747535004515e-06,
+      "loss": 0.0095,
+      "step": 15420
+    },
+    {
+      "epoch": 0.5572811326206298,
+      "grad_norm": 0.2893571731615698,
+      "learning_rate": 1.2954363848047522e-06,
+      "loss": 0.0093,
+      "step": 15430
+    },
+    {
+      "epoch": 0.5576422999133198,
+      "grad_norm": 0.3088079129004796,
+      "learning_rate": 1.29369829601562e-06,
+      "loss": 0.0088,
+      "step": 15440
+    },
+    {
+      "epoch": 0.5580034672060098,
+      "grad_norm": 0.2744541912811685,
+      "learning_rate": 1.2919604895113006e-06,
+      "loss": 0.0093,
+      "step": 15450
+    },
+    {
+      "epoch": 0.5583646344986998,
+      "grad_norm": 0.3704455498129189,
+      "learning_rate": 1.2902229676696523e-06,
+      "loss": 0.0096,
+      "step": 15460
+    },
+    {
+      "epoch": 0.5587258017913898,
+      "grad_norm": 0.48673311811367453,
+      "learning_rate": 1.2884857328681454e-06,
+      "loss": 0.0101,
+      "step": 15470
+    },
+    {
+      "epoch": 0.5590869690840797,
+      "grad_norm": 0.3696486936462307,
+      "learning_rate": 1.286748787483856e-06,
+      "loss": 0.0104,
+      "step": 15480
+    },
+    {
+      "epoch": 0.5594481363767697,
+      "grad_norm": 0.32977810859338114,
+      "learning_rate": 1.2850121338934654e-06,
+      "loss": 0.0094,
+      "step": 15490
+    },
+    {
+      "epoch": 0.5598093036694597,
+      "grad_norm": 0.3502237125793796,
+      "learning_rate": 1.2832757744732552e-06,
+      "loss": 0.0091,
+      "step": 15500
+    },
+    {
+      "epoch": 0.5601704709621497,
+      "grad_norm": 0.3997924091538801,
+      "learning_rate": 1.281539711599104e-06,
+      "loss": 0.0093,
+      "step": 15510
+    },
+    {
+      "epoch": 0.5605316382548396,
+      "grad_norm": 0.4953438455519611,
+      "learning_rate": 1.2798039476464854e-06,
+      "loss": 0.0097,
+      "step": 15520
+    },
+    {
+      "epoch": 0.5608928055475296,
+      "grad_norm": 0.26367444455145383,
+      "learning_rate": 1.2780684849904637e-06,
+      "loss": 0.0093,
+      "step": 15530
+    },
+    {
+      "epoch": 0.5612539728402196,
+      "grad_norm": 0.3477714756781318,
+      "learning_rate": 1.2763333260056907e-06,
+      "loss": 0.0098,
+      "step": 15540
+    },
+    {
+      "epoch": 0.5616151401329096,
+      "grad_norm": 0.3107732134134742,
+      "learning_rate": 1.2745984730664037e-06,
+      "loss": 0.0101,
+      "step": 15550
+    },
+    {
+      "epoch": 0.5619763074255996,
+      "grad_norm": 0.2904466905768486,
+      "learning_rate": 1.2728639285464188e-06,
+      "loss": 0.009,
+      "step": 15560
+    },
+    {
+      "epoch": 0.5623374747182895,
+      "grad_norm": 0.32314488463833857,
+      "learning_rate": 1.2711296948191325e-06,
+      "loss": 0.0095,
+      "step": 15570
+    },
+    {
+      "epoch": 0.5626986420109795,
+      "grad_norm": 0.3460260030207964,
+      "learning_rate": 1.2693957742575156e-06,
+      "loss": 0.0096,
+      "step": 15580
+    },
+    {
+      "epoch": 0.5630598093036695,
+      "grad_norm": 0.30143580019416877,
+      "learning_rate": 1.2676621692341094e-06,
+      "loss": 0.0091,
+      "step": 15590
+    },
+    {
+      "epoch": 0.5634209765963595,
+      "grad_norm": 0.3519178439034617,
+      "learning_rate": 1.2659288821210245e-06,
+      "loss": 0.009,
+      "step": 15600
+    },
+    {
+      "epoch": 0.5637821438890495,
+      "grad_norm": 0.4281840518355516,
+      "learning_rate": 1.2641959152899358e-06,
+      "loss": 0.0103,
+      "step": 15610
+    },
+    {
+      "epoch": 0.5641433111817393,
+      "grad_norm": 0.31578904077233977,
+      "learning_rate": 1.2624632711120803e-06,
+      "loss": 0.0095,
+      "step": 15620
+    },
+    {
+      "epoch": 0.5645044784744293,
+      "grad_norm": 0.3140182806651274,
+      "learning_rate": 1.260730951958254e-06,
+      "loss": 0.0091,
+      "step": 15630
+    },
+    {
+      "epoch": 0.5648656457671193,
+      "grad_norm": 0.43543545058120653,
+      "learning_rate": 1.2589989601988069e-06,
+      "loss": 0.0096,
+      "step": 15640
+    },
+    {
+      "epoch": 0.5652268130598093,
+      "grad_norm": 0.28044592531159834,
+      "learning_rate": 1.2572672982036413e-06,
+      "loss": 0.0094,
+      "step": 15650
+    },
+    {
+      "epoch": 0.5655879803524992,
+      "grad_norm": 0.3535870990678013,
+      "learning_rate": 1.2555359683422095e-06,
+      "loss": 0.0097,
+      "step": 15660
+    },
+    {
+      "epoch": 0.5659491476451892,
+      "grad_norm": 0.3669292005856019,
+      "learning_rate": 1.2538049729835082e-06,
+      "loss": 0.0099,
+      "step": 15670
+    },
+    {
+      "epoch": 0.5663103149378792,
+      "grad_norm": 0.2836912654528039,
+      "learning_rate": 1.252074314496077e-06,
+      "loss": 0.0096,
+      "step": 15680
+    },
+    {
+      "epoch": 0.5666714822305692,
+      "grad_norm": 0.35879885985865634,
+      "learning_rate": 1.2503439952479938e-06,
+      "loss": 0.0102,
+      "step": 15690
+    },
+    {
+      "epoch": 0.5670326495232592,
+      "grad_norm": 0.5689295182325866,
+      "learning_rate": 1.2486140176068736e-06,
+      "loss": 0.009,
+      "step": 15700
+    },
+    {
+      "epoch": 0.5673938168159491,
+      "grad_norm": 0.33716404879296585,
+      "learning_rate": 1.246884383939863e-06,
+      "loss": 0.0098,
+      "step": 15710
+    },
+    {
+      "epoch": 0.5677549841086391,
+      "grad_norm": 0.3751454506727387,
+      "learning_rate": 1.2451550966136372e-06,
+      "loss": 0.0096,
+      "step": 15720
+    },
+    {
+      "epoch": 0.5681161514013291,
+      "grad_norm": 0.400426858451966,
+      "learning_rate": 1.2434261579943988e-06,
+      "loss": 0.0093,
+      "step": 15730
+    },
+    {
+      "epoch": 0.5684773186940191,
+      "grad_norm": 0.3275587326916594,
+      "learning_rate": 1.2416975704478735e-06,
+      "loss": 0.0088,
+      "step": 15740
+    },
+    {
+      "epoch": 0.568838485986709,
+      "grad_norm": 0.47433720622855824,
+      "learning_rate": 1.2399693363393054e-06,
+      "loss": 0.0105,
+      "step": 15750
+    },
+    {
+      "epoch": 0.569199653279399,
+      "grad_norm": 0.42058532951864114,
+      "learning_rate": 1.2382414580334558e-06,
+      "loss": 0.0104,
+      "step": 15760
+    },
+    {
+      "epoch": 0.569560820572089,
+      "grad_norm": 0.40133266684625796,
+      "learning_rate": 1.236513937894599e-06,
+      "loss": 0.0093,
+      "step": 15770
+    },
+    {
+      "epoch": 0.569921987864779,
+      "grad_norm": 0.37543495569049085,
+      "learning_rate": 1.2347867782865192e-06,
+      "loss": 0.0091,
+      "step": 15780
+    },
+    {
+      "epoch": 0.570283155157469,
+      "grad_norm": 0.30375405568611086,
+      "learning_rate": 1.233059981572507e-06,
+      "loss": 0.0097,
+      "step": 15790
+    },
+    {
+      "epoch": 0.570644322450159,
+      "grad_norm": 0.5734991336519728,
+      "learning_rate": 1.2313335501153563e-06,
+      "loss": 0.0094,
+      "step": 15800
+    },
+    {
+      "epoch": 0.5710054897428489,
+      "grad_norm": 0.37476371099074474,
+      "learning_rate": 1.2296074862773625e-06,
+      "loss": 0.0097,
+      "step": 15810
+    },
+    {
+      "epoch": 0.5713666570355389,
+      "grad_norm": 0.403571387584609,
+      "learning_rate": 1.2278817924203165e-06,
+      "loss": 0.0096,
+      "step": 15820
+    },
+    {
+      "epoch": 0.5717278243282289,
+      "grad_norm": 0.46417309236580767,
+      "learning_rate": 1.2261564709055037e-06,
+      "loss": 0.0089,
+      "step": 15830
+    },
+    {
+      "epoch": 0.5720889916209189,
+      "grad_norm": 0.3324165784650547,
+      "learning_rate": 1.2244315240937e-06,
+      "loss": 0.0092,
+      "step": 15840
+    },
+    {
+      "epoch": 0.5724501589136087,
+      "grad_norm": 0.2693252095563595,
+      "learning_rate": 1.222706954345168e-06,
+      "loss": 0.0095,
+      "step": 15850
+    },
+    {
+      "epoch": 0.5728113262062987,
+      "grad_norm": 0.3248612204396875,
+      "learning_rate": 1.220982764019656e-06,
+      "loss": 0.0097,
+      "step": 15860
+    },
+    {
+      "epoch": 0.5731724934989887,
+      "grad_norm": 0.2835247441201622,
+      "learning_rate": 1.2192589554763902e-06,
+      "loss": 0.0091,
+      "step": 15870
+    },
+    {
+      "epoch": 0.5735336607916787,
+      "grad_norm": 0.42976070386544446,
+      "learning_rate": 1.2175355310740769e-06,
+      "loss": 0.0088,
+      "step": 15880
+    },
+    {
+      "epoch": 0.5738948280843686,
+      "grad_norm": 0.39558525028512775,
+      "learning_rate": 1.215812493170896e-06,
+      "loss": 0.0094,
+      "step": 15890
+    },
+    {
+      "epoch": 0.5742559953770586,
+      "grad_norm": 0.5382503034014005,
+      "learning_rate": 1.2140898441244988e-06,
+      "loss": 0.0096,
+      "step": 15900
+    },
+    {
+      "epoch": 0.5746171626697486,
+      "grad_norm": 0.3862476255721364,
+      "learning_rate": 1.212367586292004e-06,
+      "loss": 0.0091,
+      "step": 15910
+    },
+    {
+      "epoch": 0.5749783299624386,
+      "grad_norm": 0.31470362727478224,
+      "learning_rate": 1.2106457220299953e-06,
+      "loss": 0.0095,
+      "step": 15920
+    },
+    {
+      "epoch": 0.5753394972551286,
+      "grad_norm": 0.35477310880491303,
+      "learning_rate": 1.2089242536945183e-06,
+      "loss": 0.0088,
+      "step": 15930
+    },
+    {
+      "epoch": 0.5757006645478185,
+      "grad_norm": 0.38111340361682566,
+      "learning_rate": 1.2072031836410763e-06,
+      "loss": 0.0088,
+      "step": 15940
+    },
+    {
+      "epoch": 0.5760618318405085,
+      "grad_norm": 0.3866266599817926,
+      "learning_rate": 1.2054825142246268e-06,
+      "loss": 0.0093,
+      "step": 15950
+    },
+    {
+      "epoch": 0.5764229991331985,
+      "grad_norm": 0.5346575456927066,
+      "learning_rate": 1.2037622477995809e-06,
+      "loss": 0.0097,
+      "step": 15960
+    },
+    {
+      "epoch": 0.5767841664258885,
+      "grad_norm": 0.32384509382504373,
+      "learning_rate": 1.2020423867197973e-06,
+      "loss": 0.0089,
+      "step": 15970
+    },
+    {
+      "epoch": 0.5771453337185785,
+      "grad_norm": 0.31794102530590257,
+      "learning_rate": 1.2003229333385799e-06,
+      "loss": 0.0097,
+      "step": 15980
+    },
+    {
+      "epoch": 0.5775065010112684,
+      "grad_norm": 0.3168747850126742,
+      "learning_rate": 1.1986038900086753e-06,
+      "loss": 0.0093,
+      "step": 15990
+    },
+    {
+      "epoch": 0.5778676683039584,
+      "grad_norm": 0.37087355717653736,
+      "learning_rate": 1.1968852590822686e-06,
+      "loss": 0.0088,
+      "step": 16000
+    },
+    {
+      "epoch": 0.5778676683039584,
+      "eval_loss": 0.23030215501785278,
+      "eval_runtime": 38.4769,
+      "eval_samples_per_second": 106.454,
+      "eval_steps_per_second": 0.208,
+      "step": 16000
+    },
+    {
+      "epoch": 0.5782288355966484,
+      "grad_norm": 0.3690861298963859,
+      "learning_rate": 1.1951670429109809e-06,
+      "loss": 0.0096,
+      "step": 16010
+    },
+    {
+      "epoch": 0.5785900028893384,
+      "grad_norm": 0.3290269600654781,
+      "learning_rate": 1.1934492438458663e-06,
+      "loss": 0.0089,
+      "step": 16020
+    },
+    {
+      "epoch": 0.5789511701820284,
+      "grad_norm": 0.32722924003884446,
+      "learning_rate": 1.191731864237406e-06,
+      "loss": 0.0093,
+      "step": 16030
+    },
+    {
+      "epoch": 0.5793123374747183,
+      "grad_norm": 0.4532733396810526,
+      "learning_rate": 1.190014906435509e-06,
+      "loss": 0.0098,
+      "step": 16040
+    },
+    {
+      "epoch": 0.5796735047674083,
+      "grad_norm": 1.353380877251689,
+      "learning_rate": 1.1882983727895081e-06,
+      "loss": 0.0088,
+      "step": 16050
+    },
+    {
+      "epoch": 0.5800346720600983,
+      "grad_norm": 0.27593573911084396,
+      "learning_rate": 1.1865822656481536e-06,
+      "loss": 0.0088,
+      "step": 16060
+    },
+    {
+      "epoch": 0.5803958393527882,
+      "grad_norm": 0.48176555720008307,
+      "learning_rate": 1.1848665873596132e-06,
+      "loss": 0.0093,
+      "step": 16070
+    },
+    {
+      "epoch": 0.5807570066454781,
+      "grad_norm": 0.24592332204424794,
+      "learning_rate": 1.1831513402714682e-06,
+      "loss": 0.0094,
+      "step": 16080
+    },
+    {
+      "epoch": 0.5811181739381681,
+      "grad_norm": 0.29389072690770995,
+      "learning_rate": 1.1814365267307088e-06,
+      "loss": 0.0089,
+      "step": 16090
+    },
+    {
+      "epoch": 0.5814793412308581,
+      "grad_norm": 0.3148886614504685,
+      "learning_rate": 1.1797221490837331e-06,
+      "loss": 0.0092,
+      "step": 16100
+    },
+    {
+      "epoch": 0.5818405085235481,
+      "grad_norm": 0.28611631889124334,
+      "learning_rate": 1.1780082096763418e-06,
+      "loss": 0.0088,
+      "step": 16110
+    },
+    {
+      "epoch": 0.582201675816238,
+      "grad_norm": 0.28654340493187835,
+      "learning_rate": 1.1762947108537367e-06,
+      "loss": 0.0095,
+      "step": 16120
+    },
+    {
+      "epoch": 0.582562843108928,
+      "grad_norm": 0.4664716523326948,
+      "learning_rate": 1.1745816549605166e-06,
+      "loss": 0.0094,
+      "step": 16130
+    },
+    {
+      "epoch": 0.582924010401618,
+      "grad_norm": 0.32660503063746,
+      "learning_rate": 1.1728690443406732e-06,
+      "loss": 0.0093,
+      "step": 16140
+    },
+    {
+      "epoch": 0.583285177694308,
+      "grad_norm": 0.3794462428737043,
+      "learning_rate": 1.1711568813375908e-06,
+      "loss": 0.0092,
+      "step": 16150
+    },
+    {
+      "epoch": 0.583646344986998,
+      "grad_norm": 0.36147107355917885,
+      "learning_rate": 1.1694451682940394e-06,
+      "loss": 0.0094,
+      "step": 16160
+    },
+    {
+      "epoch": 0.584007512279688,
+      "grad_norm": 0.45832322034935274,
+      "learning_rate": 1.167733907552175e-06,
+      "loss": 0.0093,
+      "step": 16170
+    },
+    {
+      "epoch": 0.5843686795723779,
+      "grad_norm": 0.2954423626840176,
+      "learning_rate": 1.1660231014535332e-06,
+      "loss": 0.0091,
+      "step": 16180
+    },
+    {
+      "epoch": 0.5847298468650679,
+      "grad_norm": 0.30647173820869633,
+      "learning_rate": 1.1643127523390282e-06,
+      "loss": 0.0089,
+      "step": 16190
+    },
+    {
+      "epoch": 0.5850910141577579,
+      "grad_norm": 0.2685484938129936,
+      "learning_rate": 1.162602862548949e-06,
+      "loss": 0.009,
+      "step": 16200
+    },
+    {
+      "epoch": 0.5854521814504479,
+      "grad_norm": 0.3308173525645922,
+      "learning_rate": 1.160893434422956e-06,
+      "loss": 0.0093,
+      "step": 16210
+    },
+    {
+      "epoch": 0.5858133487431378,
+      "grad_norm": 0.42044940363379935,
+      "learning_rate": 1.1591844703000778e-06,
+      "loss": 0.0088,
+      "step": 16220
+    },
+    {
+      "epoch": 0.5861745160358278,
+      "grad_norm": 0.36603575710150915,
+      "learning_rate": 1.157475972518708e-06,
+      "loss": 0.0091,
+      "step": 16230
+    },
+    {
+      "epoch": 0.5865356833285178,
+      "grad_norm": 0.32467603342117685,
+      "learning_rate": 1.1557679434166027e-06,
+      "loss": 0.009,
+      "step": 16240
+    },
+    {
+      "epoch": 0.5868968506212078,
+      "grad_norm": 0.4589889968750417,
+      "learning_rate": 1.1540603853308763e-06,
+      "loss": 0.0093,
+      "step": 16250
+    },
+    {
+      "epoch": 0.5872580179138978,
+      "grad_norm": 0.28655836063513945,
+      "learning_rate": 1.152353300597998e-06,
+      "loss": 0.0092,
+      "step": 16260
+    },
+    {
+      "epoch": 0.5876191852065877,
+      "grad_norm": 0.43103797293297325,
+      "learning_rate": 1.1506466915537903e-06,
+      "loss": 0.0091,
+      "step": 16270
+    },
+    {
+      "epoch": 0.5879803524992777,
+      "grad_norm": 0.39578487648812594,
+      "learning_rate": 1.1489405605334244e-06,
+      "loss": 0.0096,
+      "step": 16280
+    },
+    {
+      "epoch": 0.5883415197919677,
+      "grad_norm": 0.32977405181099373,
+      "learning_rate": 1.1472349098714177e-06,
+      "loss": 0.0093,
+      "step": 16290
+    },
+    {
+      "epoch": 0.5887026870846576,
+      "grad_norm": 0.3580691347597796,
+      "learning_rate": 1.1455297419016298e-06,
+      "loss": 0.0092,
+      "step": 16300
+    },
+    {
+      "epoch": 0.5890638543773475,
+      "grad_norm": 0.34783234101182237,
+      "learning_rate": 1.1438250589572603e-06,
+      "loss": 0.0097,
+      "step": 16310
+    },
+    {
+      "epoch": 0.5894250216700375,
+      "grad_norm": 0.40392883943105645,
+      "learning_rate": 1.1421208633708452e-06,
+      "loss": 0.0095,
+      "step": 16320
+    },
+    {
+      "epoch": 0.5897861889627275,
+      "grad_norm": 0.39855004655499765,
+      "learning_rate": 1.1404171574742537e-06,
+      "loss": 0.0092,
+      "step": 16330
+    },
+    {
+      "epoch": 0.5901473562554175,
+      "grad_norm": 0.2549516983021182,
+      "learning_rate": 1.1387139435986838e-06,
+      "loss": 0.0088,
+      "step": 16340
+    },
+    {
+      "epoch": 0.5905085235481075,
+      "grad_norm": 0.30130437751923567,
+      "learning_rate": 1.1370112240746614e-06,
+      "loss": 0.0095,
+      "step": 16350
+    },
+    {
+      "epoch": 0.5908696908407974,
+      "grad_norm": 0.3385354949103522,
+      "learning_rate": 1.1353090012320359e-06,
+      "loss": 0.0087,
+      "step": 16360
+    },
+    {
+      "epoch": 0.5912308581334874,
+      "grad_norm": 1.0145553388501027,
+      "learning_rate": 1.1336072773999769e-06,
+      "loss": 0.0095,
+      "step": 16370
+    },
+    {
+      "epoch": 0.5915920254261774,
+      "grad_norm": 0.3554938281204335,
+      "learning_rate": 1.131906054906971e-06,
+      "loss": 0.0088,
+      "step": 16380
+    },
+    {
+      "epoch": 0.5919531927188674,
+      "grad_norm": 0.2583106716957815,
+      "learning_rate": 1.130205336080819e-06,
+      "loss": 0.009,
+      "step": 16390
+    },
+    {
+      "epoch": 0.5923143600115574,
+      "grad_norm": 0.34841022876931005,
+      "learning_rate": 1.1285051232486329e-06,
+      "loss": 0.0092,
+      "step": 16400
+    },
+    {
+      "epoch": 0.5926755273042473,
+      "grad_norm": 0.24693108711267844,
+      "learning_rate": 1.1268054187368324e-06,
+      "loss": 0.0087,
+      "step": 16410
+    },
+    {
+      "epoch": 0.5930366945969373,
+      "grad_norm": 0.31382559916655567,
+      "learning_rate": 1.1251062248711399e-06,
+      "loss": 0.0089,
+      "step": 16420
+    },
+    {
+      "epoch": 0.5933978618896273,
+      "grad_norm": 0.2691156355957071,
+      "learning_rate": 1.123407543976581e-06,
+      "loss": 0.009,
+      "step": 16430
+    },
+    {
+      "epoch": 0.5937590291823173,
+      "grad_norm": 0.3732695425397372,
+      "learning_rate": 1.1217093783774789e-06,
+      "loss": 0.0092,
+      "step": 16440
+    },
+    {
+      "epoch": 0.5941201964750072,
+      "grad_norm": 0.31478284736916656,
+      "learning_rate": 1.1200117303974516e-06,
+      "loss": 0.0096,
+      "step": 16450
+    },
+    {
+      "epoch": 0.5944813637676972,
+      "grad_norm": 0.4252189600756867,
+      "learning_rate": 1.1183146023594088e-06,
+      "loss": 0.0095,
+      "step": 16460
+    },
+    {
+      "epoch": 0.5948425310603872,
+      "grad_norm": 0.8004508128798672,
+      "learning_rate": 1.116617996585549e-06,
+      "loss": 0.009,
+      "step": 16470
+    },
+    {
+      "epoch": 0.5952036983530772,
+      "grad_norm": 0.30528366837270504,
+      "learning_rate": 1.1149219153973557e-06,
+      "loss": 0.0092,
+      "step": 16480
+    },
+    {
+      "epoch": 0.5955648656457672,
+      "grad_norm": 0.37902126733926383,
+      "learning_rate": 1.1132263611155952e-06,
+      "loss": 0.0087,
+      "step": 16490
+    },
+    {
+      "epoch": 0.5959260329384571,
+      "grad_norm": 0.3265213054192932,
+      "learning_rate": 1.111531336060312e-06,
+      "loss": 0.0091,
+      "step": 16500
+    },
+    {
+      "epoch": 0.5962872002311471,
+      "grad_norm": 0.36265412878598896,
+      "learning_rate": 1.1098368425508268e-06,
+      "loss": 0.0093,
+      "step": 16510
+    },
+    {
+      "epoch": 0.596648367523837,
+      "grad_norm": 0.4682693177479052,
+      "learning_rate": 1.1081428829057333e-06,
+      "loss": 0.0091,
+      "step": 16520
+    },
+    {
+      "epoch": 0.597009534816527,
+      "grad_norm": 0.2425192546973366,
+      "learning_rate": 1.1064494594428943e-06,
+      "loss": 0.0089,
+      "step": 16530
+    },
+    {
+      "epoch": 0.5973707021092169,
+      "grad_norm": 0.3218274773431512,
+      "learning_rate": 1.1047565744794395e-06,
+      "loss": 0.0084,
+      "step": 16540
+    },
+    {
+      "epoch": 0.5977318694019069,
+      "grad_norm": 0.4179688818031614,
+      "learning_rate": 1.1030642303317613e-06,
+      "loss": 0.0096,
+      "step": 16550
+    },
+    {
+      "epoch": 0.5980930366945969,
+      "grad_norm": 0.39221770246977483,
+      "learning_rate": 1.101372429315512e-06,
+      "loss": 0.0089,
+      "step": 16560
+    },
+    {
+      "epoch": 0.5984542039872869,
+      "grad_norm": 0.2841093776032645,
+      "learning_rate": 1.0996811737456006e-06,
+      "loss": 0.0084,
+      "step": 16570
+    },
+    {
+      "epoch": 0.5988153712799769,
+      "grad_norm": 0.30523692673957425,
+      "learning_rate": 1.0979904659361903e-06,
+      "loss": 0.0089,
+      "step": 16580
+    },
+    {
+      "epoch": 0.5991765385726668,
+      "grad_norm": 0.3172392182566113,
+      "learning_rate": 1.0963003082006946e-06,
+      "loss": 0.0094,
+      "step": 16590
+    },
+    {
+      "epoch": 0.5995377058653568,
+      "grad_norm": 0.26353776882911595,
+      "learning_rate": 1.0946107028517741e-06,
+      "loss": 0.0089,
+      "step": 16600
+    },
+    {
+      "epoch": 0.5998988731580468,
+      "grad_norm": 0.910257846157668,
+      "learning_rate": 1.0929216522013342e-06,
+      "loss": 0.0094,
+      "step": 16610
+    },
+    {
+      "epoch": 0.6002600404507368,
+      "grad_norm": 0.30559523089616797,
+      "learning_rate": 1.0912331585605203e-06,
+      "loss": 0.0089,
+      "step": 16620
+    },
+    {
+      "epoch": 0.6006212077434268,
+      "grad_norm": 0.2932828151767832,
+      "learning_rate": 1.0895452242397165e-06,
+      "loss": 0.0086,
+      "step": 16630
+    },
+    {
+      "epoch": 0.6009823750361167,
+      "grad_norm": 0.30930145880089954,
+      "learning_rate": 1.0878578515485418e-06,
+      "loss": 0.0091,
+      "step": 16640
+    },
+    {
+      "epoch": 0.6013435423288067,
+      "grad_norm": 0.32741943852694333,
+      "learning_rate": 1.0861710427958451e-06,
+      "loss": 0.0092,
+      "step": 16650
+    },
+    {
+      "epoch": 0.6017047096214967,
+      "grad_norm": 0.31278445183857584,
+      "learning_rate": 1.0844848002897048e-06,
+      "loss": 0.0092,
+      "step": 16660
+    },
+    {
+      "epoch": 0.6020658769141867,
+      "grad_norm": 0.3852443769900257,
+      "learning_rate": 1.0827991263374251e-06,
+      "loss": 0.0086,
+      "step": 16670
+    },
+    {
+      "epoch": 0.6024270442068766,
+      "grad_norm": 0.34942316494312875,
+      "learning_rate": 1.0811140232455314e-06,
+      "loss": 0.0085,
+      "step": 16680
+    },
+    {
+      "epoch": 0.6027882114995666,
+      "grad_norm": 0.27823008305268354,
+      "learning_rate": 1.0794294933197678e-06,
+      "loss": 0.0091,
+      "step": 16690
+    },
+    {
+      "epoch": 0.6031493787922566,
+      "grad_norm": 0.39097372340094194,
+      "learning_rate": 1.0777455388650947e-06,
+      "loss": 0.0089,
+      "step": 16700
+    },
+    {
+      "epoch": 0.6035105460849466,
+      "grad_norm": 0.33190276070738395,
+      "learning_rate": 1.076062162185685e-06,
+      "loss": 0.0093,
+      "step": 16710
+    },
+    {
+      "epoch": 0.6038717133776366,
+      "grad_norm": 0.33535015889896375,
+      "learning_rate": 1.0743793655849216e-06,
+      "loss": 0.0088,
+      "step": 16720
+    },
+    {
+      "epoch": 0.6042328806703265,
+      "grad_norm": 0.2611211174826967,
+      "learning_rate": 1.0726971513653912e-06,
+      "loss": 0.0087,
+      "step": 16730
+    },
+    {
+      "epoch": 0.6045940479630165,
+      "grad_norm": 0.29596354794161445,
+      "learning_rate": 1.0710155218288864e-06,
+      "loss": 0.0088,
+      "step": 16740
+    },
+    {
+      "epoch": 0.6049552152557064,
+      "grad_norm": 0.4230450654072328,
+      "learning_rate": 1.0693344792763986e-06,
+      "loss": 0.0085,
+      "step": 16750
+    },
+    {
+      "epoch": 0.6053163825483964,
+      "grad_norm": 0.26025230971948793,
+      "learning_rate": 1.0676540260081163e-06,
+      "loss": 0.0085,
+      "step": 16760
+    },
+    {
+      "epoch": 0.6056775498410863,
+      "grad_norm": 0.3024122996193209,
+      "learning_rate": 1.0659741643234217e-06,
+      "loss": 0.0091,
+      "step": 16770
+    },
+    {
+      "epoch": 0.6060387171337763,
+      "grad_norm": 0.32327018308746086,
+      "learning_rate": 1.0642948965208868e-06,
+      "loss": 0.0088,
+      "step": 16780
+    },
+    {
+      "epoch": 0.6063998844264663,
+      "grad_norm": 0.430604362789455,
+      "learning_rate": 1.062616224898272e-06,
+      "loss": 0.0089,
+      "step": 16790
+    },
+    {
+      "epoch": 0.6067610517191563,
+      "grad_norm": 0.27614254519564824,
+      "learning_rate": 1.0609381517525222e-06,
+      "loss": 0.009,
+      "step": 16800
+    },
+    {
+      "epoch": 0.6071222190118463,
+      "grad_norm": 0.33496908304997836,
+      "learning_rate": 1.0592606793797614e-06,
+      "loss": 0.0091,
+      "step": 16810
+    },
+    {
+      "epoch": 0.6074833863045362,
+      "grad_norm": 0.32923361894017816,
+      "learning_rate": 1.0575838100752936e-06,
+      "loss": 0.0091,
+      "step": 16820
+    },
+    {
+      "epoch": 0.6078445535972262,
+      "grad_norm": 0.4015734413087467,
+      "learning_rate": 1.0559075461335963e-06,
+      "loss": 0.0084,
+      "step": 16830
+    },
+    {
+      "epoch": 0.6082057208899162,
+      "grad_norm": 0.4978629025926671,
+      "learning_rate": 1.05423188984832e-06,
+      "loss": 0.0089,
+      "step": 16840
+    },
+    {
+      "epoch": 0.6085668881826062,
+      "grad_norm": 0.32885212314668083,
+      "learning_rate": 1.0525568435122823e-06,
+      "loss": 0.0091,
+      "step": 16850
+    },
+    {
+      "epoch": 0.6089280554752962,
+      "grad_norm": 0.32400333142207144,
+      "learning_rate": 1.0508824094174674e-06,
+      "loss": 0.0089,
+      "step": 16860
+    },
+    {
+      "epoch": 0.6092892227679861,
+      "grad_norm": 0.2686377408307983,
+      "learning_rate": 1.049208589855021e-06,
+      "loss": 0.009,
+      "step": 16870
+    },
+    {
+      "epoch": 0.6096503900606761,
+      "grad_norm": 0.37885909381556426,
+      "learning_rate": 1.0475353871152486e-06,
+      "loss": 0.0091,
+      "step": 16880
+    },
+    {
+      "epoch": 0.6100115573533661,
+      "grad_norm": 0.29018815030093564,
+      "learning_rate": 1.045862803487611e-06,
+      "loss": 0.0088,
+      "step": 16890
+    },
+    {
+      "epoch": 0.6103727246460561,
+      "grad_norm": 0.3592579909106019,
+      "learning_rate": 1.0441908412607216e-06,
+      "loss": 0.0093,
+      "step": 16900
+    },
+    {
+      "epoch": 0.610733891938746,
+      "grad_norm": 0.2980004068156313,
+      "learning_rate": 1.0425195027223446e-06,
+      "loss": 0.0093,
+      "step": 16910
+    },
+    {
+      "epoch": 0.611095059231436,
+      "grad_norm": 0.36928851158683434,
+      "learning_rate": 1.0408487901593902e-06,
+      "loss": 0.0083,
+      "step": 16920
+    },
+    {
+      "epoch": 0.611456226524126,
+      "grad_norm": 0.28070335504936716,
+      "learning_rate": 1.0391787058579125e-06,
+      "loss": 0.0082,
+      "step": 16930
+    },
+    {
+      "epoch": 0.611817393816816,
+      "grad_norm": 0.32967086908935517,
+      "learning_rate": 1.037509252103105e-06,
+      "loss": 0.0081,
+      "step": 16940
+    },
+    {
+      "epoch": 0.612178561109506,
+      "grad_norm": 0.28282826842238784,
+      "learning_rate": 1.0358404311792994e-06,
+      "loss": 0.0096,
+      "step": 16950
+    },
+    {
+      "epoch": 0.612539728402196,
+      "grad_norm": 0.31495098122075693,
+      "learning_rate": 1.0341722453699615e-06,
+      "loss": 0.0089,
+      "step": 16960
+    },
+    {
+      "epoch": 0.6129008956948858,
+      "grad_norm": 0.2939733843080595,
+      "learning_rate": 1.0325046969576863e-06,
+      "loss": 0.0088,
+      "step": 16970
+    },
+    {
+      "epoch": 0.6132620629875758,
+      "grad_norm": 0.3867973031713573,
+      "learning_rate": 1.030837788224199e-06,
+      "loss": 0.0086,
+      "step": 16980
+    },
+    {
+      "epoch": 0.6136232302802658,
+      "grad_norm": 0.3345675908949467,
+      "learning_rate": 1.0291715214503483e-06,
+      "loss": 0.0094,
+      "step": 16990
+    },
+    {
+      "epoch": 0.6139843975729558,
+      "grad_norm": 0.3043985118974842,
+      "learning_rate": 1.0275058989161048e-06,
+      "loss": 0.0091,
+      "step": 17000
+    },
+    {
+      "epoch": 0.6139843975729558,
+      "eval_loss": 0.23344512283802032,
+      "eval_runtime": 38.956,
+      "eval_samples_per_second": 105.144,
+      "eval_steps_per_second": 0.205,
+      "step": 17000
+    },
+    {
+      "epoch": 0.6143455648656457,
+      "grad_norm": 0.26606502613702865,
+      "learning_rate": 1.0258409229005572e-06,
+      "loss": 0.009,
+      "step": 17010
+    },
+    {
+      "epoch": 0.6147067321583357,
+      "grad_norm": 0.25594184884708054,
+      "learning_rate": 1.0241765956819102e-06,
+      "loss": 0.0091,
+      "step": 17020
+    },
+    {
+      "epoch": 0.6150678994510257,
+      "grad_norm": 0.3062594193671945,
+      "learning_rate": 1.022512919537481e-06,
+      "loss": 0.0092,
+      "step": 17030
+    },
+    {
+      "epoch": 0.6154290667437157,
+      "grad_norm": 0.3136024967281056,
+      "learning_rate": 1.020849896743694e-06,
+      "loss": 0.0087,
+      "step": 17040
+    },
+    {
+      "epoch": 0.6157902340364056,
+      "grad_norm": 0.41375964072005594,
+      "learning_rate": 1.0191875295760818e-06,
+      "loss": 0.009,
+      "step": 17050
+    },
+    {
+      "epoch": 0.6161514013290956,
+      "grad_norm": 0.5162822164783735,
+      "learning_rate": 1.0175258203092785e-06,
+      "loss": 0.0085,
+      "step": 17060
+    },
+    {
+      "epoch": 0.6165125686217856,
+      "grad_norm": 0.33105740840483405,
+      "learning_rate": 1.0158647712170193e-06,
+      "loss": 0.0092,
+      "step": 17070
+    },
+    {
+      "epoch": 0.6168737359144756,
+      "grad_norm": 0.3062310203104997,
+      "learning_rate": 1.0142043845721345e-06,
+      "loss": 0.0083,
+      "step": 17080
+    },
+    {
+      "epoch": 0.6172349032071656,
+      "grad_norm": 0.32516876749425067,
+      "learning_rate": 1.0125446626465496e-06,
+      "loss": 0.0087,
+      "step": 17090
+    },
+    {
+      "epoch": 0.6175960704998555,
+      "grad_norm": 0.26379606012226087,
+      "learning_rate": 1.0108856077112789e-06,
+      "loss": 0.0088,
+      "step": 17100
+    },
+    {
+      "epoch": 0.6179572377925455,
+      "grad_norm": 0.2850792881815175,
+      "learning_rate": 1.009227222036426e-06,
+      "loss": 0.0084,
+      "step": 17110
+    },
+    {
+      "epoch": 0.6183184050852355,
+      "grad_norm": 0.2504182303143163,
+      "learning_rate": 1.0075695078911766e-06,
+      "loss": 0.0087,
+      "step": 17120
+    },
+    {
+      "epoch": 0.6186795723779255,
+      "grad_norm": 0.3167750798822574,
+      "learning_rate": 1.0059124675437987e-06,
+      "loss": 0.0085,
+      "step": 17130
+    },
+    {
+      "epoch": 0.6190407396706155,
+      "grad_norm": 0.2878930015243878,
+      "learning_rate": 1.0042561032616385e-06,
+      "loss": 0.0087,
+      "step": 17140
+    },
+    {
+      "epoch": 0.6194019069633054,
+      "grad_norm": 0.24823504725468162,
+      "learning_rate": 1.0026004173111165e-06,
+      "loss": 0.0086,
+      "step": 17150
+    },
+    {
+      "epoch": 0.6197630742559954,
+      "grad_norm": 0.34358524560642256,
+      "learning_rate": 1.0009454119577259e-06,
+      "loss": 0.009,
+      "step": 17160
+    },
+    {
+      "epoch": 0.6201242415486854,
+      "grad_norm": 0.25025756788981085,
+      "learning_rate": 9.992910894660274e-07,
+      "loss": 0.0083,
+      "step": 17170
+    },
+    {
+      "epoch": 0.6204854088413754,
+      "grad_norm": 0.3000712218714396,
+      "learning_rate": 9.976374520996489e-07,
+      "loss": 0.009,
+      "step": 17180
+    },
+    {
+      "epoch": 0.6208465761340654,
+      "grad_norm": 0.24933451197801118,
+      "learning_rate": 9.959845021212796e-07,
+      "loss": 0.0081,
+      "step": 17190
+    },
+    {
+      "epoch": 0.6212077434267552,
+      "grad_norm": 0.2274431471522462,
+      "learning_rate": 9.943322417926684e-07,
+      "loss": 0.0086,
+      "step": 17200
+    },
+    {
+      "epoch": 0.6215689107194452,
+      "grad_norm": 0.40606341300056714,
+      "learning_rate": 9.926806733746201e-07,
+      "loss": 0.0091,
+      "step": 17210
+    },
+    {
+      "epoch": 0.6219300780121352,
+      "grad_norm": 0.3077454892095793,
+      "learning_rate": 9.910297991269945e-07,
+      "loss": 0.0088,
+      "step": 17220
+    },
+    {
+      "epoch": 0.6222912453048252,
+      "grad_norm": 0.41504356977804463,
+      "learning_rate": 9.893796213086998e-07,
+      "loss": 0.0087,
+      "step": 17230
+    },
+    {
+      "epoch": 0.6226524125975151,
+      "grad_norm": 0.25695033717420346,
+      "learning_rate": 9.877301421776914e-07,
+      "loss": 0.0086,
+      "step": 17240
+    },
+    {
+      "epoch": 0.6230135798902051,
+      "grad_norm": 0.32852266904495053,
+      "learning_rate": 9.860813639909693e-07,
+      "loss": 0.0089,
+      "step": 17250
+    },
+    {
+      "epoch": 0.6233747471828951,
+      "grad_norm": 0.29996901873243287,
+      "learning_rate": 9.844332890045748e-07,
+      "loss": 0.0083,
+      "step": 17260
+    },
+    {
+      "epoch": 0.6237359144755851,
+      "grad_norm": 0.2506589536509868,
+      "learning_rate": 9.827859194735863e-07,
+      "loss": 0.0088,
+      "step": 17270
+    },
+    {
+      "epoch": 0.624097081768275,
+      "grad_norm": 0.2262146424807176,
+      "learning_rate": 9.81139257652116e-07,
+      "loss": 0.0087,
+      "step": 17280
+    },
+    {
+      "epoch": 0.624458249060965,
+      "grad_norm": 0.3599714769596589,
+      "learning_rate": 9.794933057933097e-07,
+      "loss": 0.0088,
+      "step": 17290
+    },
+    {
+      "epoch": 0.624819416353655,
+      "grad_norm": 0.3041735297836197,
+      "learning_rate": 9.778480661493407e-07,
+      "loss": 0.0085,
+      "step": 17300
+    },
+    {
+      "epoch": 0.625180583646345,
+      "grad_norm": 0.40974108688297517,
+      "learning_rate": 9.762035409714077e-07,
+      "loss": 0.0082,
+      "step": 17310
+    },
+    {
+      "epoch": 0.625541750939035,
+      "grad_norm": 0.2879662709142353,
+      "learning_rate": 9.74559732509732e-07,
+      "loss": 0.0085,
+      "step": 17320
+    },
+    {
+      "epoch": 0.625902918231725,
+      "grad_norm": 0.3618577444764018,
+      "learning_rate": 9.729166430135542e-07,
+      "loss": 0.0085,
+      "step": 17330
+    },
+    {
+      "epoch": 0.6262640855244149,
+      "grad_norm": 0.32560007976073674,
+      "learning_rate": 9.71274274731131e-07,
+      "loss": 0.009,
+      "step": 17340
+    },
+    {
+      "epoch": 0.6266252528171049,
+      "grad_norm": 0.2876799999654926,
+      "learning_rate": 9.696326299097334e-07,
+      "loss": 0.0087,
+      "step": 17350
+    },
+    {
+      "epoch": 0.6269864201097949,
+      "grad_norm": 0.2686090630517681,
+      "learning_rate": 9.679917107956393e-07,
+      "loss": 0.0082,
+      "step": 17360
+    },
+    {
+      "epoch": 0.6273475874024849,
+      "grad_norm": 0.2897355092184923,
+      "learning_rate": 9.663515196341375e-07,
+      "loss": 0.0088,
+      "step": 17370
+    },
+    {
+      "epoch": 0.6277087546951748,
+      "grad_norm": 0.3165403597595294,
+      "learning_rate": 9.647120586695183e-07,
+      "loss": 0.0088,
+      "step": 17380
+    },
+    {
+      "epoch": 0.6280699219878648,
+      "grad_norm": 0.38024897761067367,
+      "learning_rate": 9.630733301450736e-07,
+      "loss": 0.009,
+      "step": 17390
+    },
+    {
+      "epoch": 0.6284310892805548,
+      "grad_norm": 0.3175842121088538,
+      "learning_rate": 9.614353363030933e-07,
+      "loss": 0.0085,
+      "step": 17400
+    },
+    {
+      "epoch": 0.6287922565732448,
+      "grad_norm": 0.25425617624844227,
+      "learning_rate": 9.597980793848617e-07,
+      "loss": 0.0089,
+      "step": 17410
+    },
+    {
+      "epoch": 0.6291534238659346,
+      "grad_norm": 0.30436084946974706,
+      "learning_rate": 9.581615616306544e-07,
+      "loss": 0.0095,
+      "step": 17420
+    },
+    {
+      "epoch": 0.6295145911586246,
+      "grad_norm": 0.23855673758995746,
+      "learning_rate": 9.565257852797372e-07,
+      "loss": 0.0085,
+      "step": 17430
+    },
+    {
+      "epoch": 0.6298757584513146,
+      "grad_norm": 0.28412769395080667,
+      "learning_rate": 9.548907525703588e-07,
+      "loss": 0.0092,
+      "step": 17440
+    },
+    {
+      "epoch": 0.6302369257440046,
+      "grad_norm": 0.2569548468277973,
+      "learning_rate": 9.532564657397522e-07,
+      "loss": 0.0085,
+      "step": 17450
+    },
+    {
+      "epoch": 0.6305980930366946,
+      "grad_norm": 0.3236236807318955,
+      "learning_rate": 9.516229270241298e-07,
+      "loss": 0.0084,
+      "step": 17460
+    },
+    {
+      "epoch": 0.6309592603293845,
+      "grad_norm": 0.3167359431787842,
+      "learning_rate": 9.499901386586797e-07,
+      "loss": 0.0081,
+      "step": 17470
+    },
+    {
+      "epoch": 0.6313204276220745,
+      "grad_norm": 0.2779093249702506,
+      "learning_rate": 9.483581028775639e-07,
+      "loss": 0.0086,
+      "step": 17480
+    },
+    {
+      "epoch": 0.6316815949147645,
+      "grad_norm": 0.25568449343053,
+      "learning_rate": 9.467268219139139e-07,
+      "loss": 0.008,
+      "step": 17490
+    },
+    {
+      "epoch": 0.6320427622074545,
+      "grad_norm": 0.3452572129308364,
+      "learning_rate": 9.450962979998297e-07,
+      "loss": 0.0089,
+      "step": 17500
+    },
+    {
+      "epoch": 0.6324039295001445,
+      "grad_norm": 0.3036583763448961,
+      "learning_rate": 9.43466533366373e-07,
+      "loss": 0.0093,
+      "step": 17510
+    },
+    {
+      "epoch": 0.6327650967928344,
+      "grad_norm": 0.292196774985953,
+      "learning_rate": 9.41837530243569e-07,
+      "loss": 0.0082,
+      "step": 17520
+    },
+    {
+      "epoch": 0.6331262640855244,
+      "grad_norm": 0.28834664392223147,
+      "learning_rate": 9.402092908603994e-07,
+      "loss": 0.0087,
+      "step": 17530
+    },
+    {
+      "epoch": 0.6334874313782144,
+      "grad_norm": 0.24302523025058015,
+      "learning_rate": 9.385818174448021e-07,
+      "loss": 0.0087,
+      "step": 17540
+    },
+    {
+      "epoch": 0.6338485986709044,
+      "grad_norm": 0.24625785696003613,
+      "learning_rate": 9.369551122236658e-07,
+      "loss": 0.0083,
+      "step": 17550
+    },
+    {
+      "epoch": 0.6342097659635944,
+      "grad_norm": 0.2685286516325574,
+      "learning_rate": 9.353291774228287e-07,
+      "loss": 0.0088,
+      "step": 17560
+    },
+    {
+      "epoch": 0.6345709332562843,
+      "grad_norm": 0.22294366359920795,
+      "learning_rate": 9.337040152670746e-07,
+      "loss": 0.0092,
+      "step": 17570
+    },
+    {
+      "epoch": 0.6349321005489743,
+      "grad_norm": 0.2876445939042442,
+      "learning_rate": 9.320796279801306e-07,
+      "loss": 0.0085,
+      "step": 17580
+    },
+    {
+      "epoch": 0.6352932678416643,
+      "grad_norm": 0.2794846277968038,
+      "learning_rate": 9.304560177846622e-07,
+      "loss": 0.0088,
+      "step": 17590
+    },
+    {
+      "epoch": 0.6356544351343543,
+      "grad_norm": 0.23563908110897838,
+      "learning_rate": 9.288331869022729e-07,
+      "loss": 0.0082,
+      "step": 17600
+    },
+    {
+      "epoch": 0.6360156024270442,
+      "grad_norm": 0.2807031802743011,
+      "learning_rate": 9.272111375534997e-07,
+      "loss": 0.0087,
+      "step": 17610
+    },
+    {
+      "epoch": 0.6363767697197342,
+      "grad_norm": 0.26721328381210047,
+      "learning_rate": 9.255898719578094e-07,
+      "loss": 0.0086,
+      "step": 17620
+    },
+    {
+      "epoch": 0.6367379370124242,
+      "grad_norm": 0.33111037283804257,
+      "learning_rate": 9.239693923335975e-07,
+      "loss": 0.0086,
+      "step": 17630
+    },
+    {
+      "epoch": 0.6370991043051142,
+      "grad_norm": 0.2565278951478535,
+      "learning_rate": 9.223497008981829e-07,
+      "loss": 0.0086,
+      "step": 17640
+    },
+    {
+      "epoch": 0.637460271597804,
+      "grad_norm": 0.44218599071435,
+      "learning_rate": 9.207307998678074e-07,
+      "loss": 0.0074,
+      "step": 17650
+    },
+    {
+      "epoch": 0.637821438890494,
+      "grad_norm": 0.26786314350740065,
+      "learning_rate": 9.191126914576299e-07,
+      "loss": 0.0084,
+      "step": 17660
+    },
+    {
+      "epoch": 0.638182606183184,
+      "grad_norm": 0.21843570820388072,
+      "learning_rate": 9.174953778817252e-07,
+      "loss": 0.0081,
+      "step": 17670
+    },
+    {
+      "epoch": 0.638543773475874,
+      "grad_norm": 0.7070641950750955,
+      "learning_rate": 9.158788613530806e-07,
+      "loss": 0.0089,
+      "step": 17680
+    },
+    {
+      "epoch": 0.638904940768564,
+      "grad_norm": 0.2819269366021958,
+      "learning_rate": 9.142631440835932e-07,
+      "loss": 0.0084,
+      "step": 17690
+    },
+    {
+      "epoch": 0.639266108061254,
+      "grad_norm": 0.2966047620092914,
+      "learning_rate": 9.126482282840659e-07,
+      "loss": 0.0079,
+      "step": 17700
+    },
+    {
+      "epoch": 0.6396272753539439,
+      "grad_norm": 0.335259272937138,
+      "learning_rate": 9.110341161642053e-07,
+      "loss": 0.0084,
+      "step": 17710
+    },
+    {
+      "epoch": 0.6399884426466339,
+      "grad_norm": 0.29226599135643655,
+      "learning_rate": 9.09420809932618e-07,
+      "loss": 0.0089,
+      "step": 17720
+    },
+    {
+      "epoch": 0.6403496099393239,
+      "grad_norm": 0.22932977881032743,
+      "learning_rate": 9.078083117968081e-07,
+      "loss": 0.008,
+      "step": 17730
+    },
+    {
+      "epoch": 0.6407107772320139,
+      "grad_norm": 0.2533488511549755,
+      "learning_rate": 9.061966239631748e-07,
+      "loss": 0.0088,
+      "step": 17740
+    },
+    {
+      "epoch": 0.6410719445247038,
+      "grad_norm": 0.2205617368180843,
+      "learning_rate": 9.045857486370059e-07,
+      "loss": 0.008,
+      "step": 17750
+    },
+    {
+      "epoch": 0.6414331118173938,
+      "grad_norm": 0.35202767611716224,
+      "learning_rate": 9.029756880224801e-07,
+      "loss": 0.0089,
+      "step": 17760
+    },
+    {
+      "epoch": 0.6417942791100838,
+      "grad_norm": 0.2581885724686155,
+      "learning_rate": 9.013664443226606e-07,
+      "loss": 0.0081,
+      "step": 17770
+    },
+    {
+      "epoch": 0.6421554464027738,
+      "grad_norm": 0.27874098338329467,
+      "learning_rate": 8.997580197394924e-07,
+      "loss": 0.0084,
+      "step": 17780
+    },
+    {
+      "epoch": 0.6425166136954638,
+      "grad_norm": 0.34176501306500773,
+      "learning_rate": 8.981504164738001e-07,
+      "loss": 0.0081,
+      "step": 17790
+    },
+    {
+      "epoch": 0.6428777809881537,
+      "grad_norm": 0.2670894031851876,
+      "learning_rate": 8.965436367252843e-07,
+      "loss": 0.008,
+      "step": 17800
+    },
+    {
+      "epoch": 0.6432389482808437,
+      "grad_norm": 0.21652063243332767,
+      "learning_rate": 8.949376826925186e-07,
+      "loss": 0.008,
+      "step": 17810
+    },
+    {
+      "epoch": 0.6436001155735337,
+      "grad_norm": 0.30680883617336197,
+      "learning_rate": 8.933325565729477e-07,
+      "loss": 0.0082,
+      "step": 17820
+    },
+    {
+      "epoch": 0.6439612828662237,
+      "grad_norm": 0.26318957793449366,
+      "learning_rate": 8.917282605628816e-07,
+      "loss": 0.0087,
+      "step": 17830
+    },
+    {
+      "epoch": 0.6443224501589137,
+      "grad_norm": 0.23268680087122356,
+      "learning_rate": 8.901247968574963e-07,
+      "loss": 0.0092,
+      "step": 17840
+    },
+    {
+      "epoch": 0.6446836174516036,
+      "grad_norm": 0.24499943638480118,
+      "learning_rate": 8.885221676508277e-07,
+      "loss": 0.0083,
+      "step": 17850
+    },
+    {
+      "epoch": 0.6450447847442936,
+      "grad_norm": 0.26099313005927827,
+      "learning_rate": 8.869203751357708e-07,
+      "loss": 0.0084,
+      "step": 17860
+    },
+    {
+      "epoch": 0.6454059520369835,
+      "grad_norm": 0.30744275228720824,
+      "learning_rate": 8.853194215040751e-07,
+      "loss": 0.0084,
+      "step": 17870
+    },
+    {
+      "epoch": 0.6457671193296735,
+      "grad_norm": 0.2620936307156402,
+      "learning_rate": 8.837193089463427e-07,
+      "loss": 0.0085,
+      "step": 17880
+    },
+    {
+      "epoch": 0.6461282866223634,
+      "grad_norm": 0.279735596467161,
+      "learning_rate": 8.821200396520246e-07,
+      "loss": 0.0081,
+      "step": 17890
+    },
+    {
+      "epoch": 0.6464894539150534,
+      "grad_norm": 0.26400637789458314,
+      "learning_rate": 8.805216158094177e-07,
+      "loss": 0.0081,
+      "step": 17900
+    },
+    {
+      "epoch": 0.6468506212077434,
+      "grad_norm": 0.2517238545969542,
+      "learning_rate": 8.789240396056625e-07,
+      "loss": 0.0087,
+      "step": 17910
+    },
+    {
+      "epoch": 0.6472117885004334,
+      "grad_norm": 0.40156144190851284,
+      "learning_rate": 8.77327313226739e-07,
+      "loss": 0.0087,
+      "step": 17920
+    },
+    {
+      "epoch": 0.6475729557931234,
+      "grad_norm": 0.23693438991843596,
+      "learning_rate": 8.757314388574656e-07,
+      "loss": 0.0086,
+      "step": 17930
+    },
+    {
+      "epoch": 0.6479341230858133,
+      "grad_norm": 0.25372847944441707,
+      "learning_rate": 8.741364186814932e-07,
+      "loss": 0.0077,
+      "step": 17940
+    },
+    {
+      "epoch": 0.6482952903785033,
+      "grad_norm": 0.29769243563514797,
+      "learning_rate": 8.725422548813061e-07,
+      "loss": 0.0083,
+      "step": 17950
+    },
+    {
+      "epoch": 0.6486564576711933,
+      "grad_norm": 0.4280760376305524,
+      "learning_rate": 8.709489496382144e-07,
+      "loss": 0.0087,
+      "step": 17960
+    },
+    {
+      "epoch": 0.6490176249638833,
+      "grad_norm": 0.33098016035152633,
+      "learning_rate": 8.693565051323563e-07,
+      "loss": 0.0093,
+      "step": 17970
+    },
+    {
+      "epoch": 0.6493787922565732,
+      "grad_norm": 0.2513322796792131,
+      "learning_rate": 8.677649235426883e-07,
+      "loss": 0.0084,
+      "step": 17980
+    },
+    {
+      "epoch": 0.6497399595492632,
+      "grad_norm": 0.2582110646925614,
+      "learning_rate": 8.661742070469898e-07,
+      "loss": 0.008,
+      "step": 17990
+    },
+    {
+      "epoch": 0.6501011268419532,
+      "grad_norm": 0.35124837891566857,
+      "learning_rate": 8.64584357821854e-07,
+      "loss": 0.008,
+      "step": 18000
+    },
+    {
+      "epoch": 0.6501011268419532,
+      "eval_loss": 0.24680060148239136,
+      "eval_runtime": 38.5955,
+      "eval_samples_per_second": 106.126,
+      "eval_steps_per_second": 0.207,
+      "step": 18000
+    },
+    {
+      "epoch": 0.6504622941346432,
+      "grad_norm": 0.30385559011118507,
+      "learning_rate": 8.6299537804269e-07,
+      "loss": 0.0085,
+      "step": 18010
+    },
+    {
+      "epoch": 0.6508234614273332,
+      "grad_norm": 0.386482620403904,
+      "learning_rate": 8.614072698837141e-07,
+      "loss": 0.0081,
+      "step": 18020
+    },
+    {
+      "epoch": 0.6511846287200231,
+      "grad_norm": 0.29235737204176093,
+      "learning_rate": 8.59820035517953e-07,
+      "loss": 0.0083,
+      "step": 18030
+    },
+    {
+      "epoch": 0.6515457960127131,
+      "grad_norm": 0.329010767922862,
+      "learning_rate": 8.582336771172349e-07,
+      "loss": 0.0084,
+      "step": 18040
+    },
+    {
+      "epoch": 0.6519069633054031,
+      "grad_norm": 0.48177650333075484,
+      "learning_rate": 8.566481968521932e-07,
+      "loss": 0.0081,
+      "step": 18050
+    },
+    {
+      "epoch": 0.6522681305980931,
+      "grad_norm": 0.28859394258321736,
+      "learning_rate": 8.550635968922542e-07,
+      "loss": 0.0086,
+      "step": 18060
+    },
+    {
+      "epoch": 0.6526292978907831,
+      "grad_norm": 0.2453673297609317,
+      "learning_rate": 8.534798794056453e-07,
+      "loss": 0.0076,
+      "step": 18070
+    },
+    {
+      "epoch": 0.652990465183473,
+      "grad_norm": 0.38972474338660523,
+      "learning_rate": 8.518970465593824e-07,
+      "loss": 0.0081,
+      "step": 18080
+    },
+    {
+      "epoch": 0.653351632476163,
+      "grad_norm": 0.2642121027509173,
+      "learning_rate": 8.503151005192739e-07,
+      "loss": 0.0083,
+      "step": 18090
+    },
+    {
+      "epoch": 0.6537127997688529,
+      "grad_norm": 0.3169476908822654,
+      "learning_rate": 8.487340434499119e-07,
+      "loss": 0.0084,
+      "step": 18100
+    },
+    {
+      "epoch": 0.6540739670615429,
+      "grad_norm": 0.24728512890682072,
+      "learning_rate": 8.471538775146751e-07,
+      "loss": 0.0081,
+      "step": 18110
+    },
+    {
+      "epoch": 0.6544351343542328,
+      "grad_norm": 0.31236087432189535,
+      "learning_rate": 8.4557460487572e-07,
+      "loss": 0.0078,
+      "step": 18120
+    },
+    {
+      "epoch": 0.6547963016469228,
+      "grad_norm": 0.325234470370389,
+      "learning_rate": 8.43996227693983e-07,
+      "loss": 0.0087,
+      "step": 18130
+    },
+    {
+      "epoch": 0.6551574689396128,
+      "grad_norm": 0.2583051911302325,
+      "learning_rate": 8.424187481291743e-07,
+      "loss": 0.0086,
+      "step": 18140
+    },
+    {
+      "epoch": 0.6555186362323028,
+      "grad_norm": 0.2932580314445645,
+      "learning_rate": 8.408421683397751e-07,
+      "loss": 0.0084,
+      "step": 18150
+    },
+    {
+      "epoch": 0.6558798035249928,
+      "grad_norm": 0.24687399275305613,
+      "learning_rate": 8.392664904830374e-07,
+      "loss": 0.0087,
+      "step": 18160
+    },
+    {
+      "epoch": 0.6562409708176827,
+      "grad_norm": 0.3462019089503469,
+      "learning_rate": 8.376917167149768e-07,
+      "loss": 0.0086,
+      "step": 18170
+    },
+    {
+      "epoch": 0.6566021381103727,
+      "grad_norm": 0.2596668716993591,
+      "learning_rate": 8.361178491903742e-07,
+      "loss": 0.008,
+      "step": 18180
+    },
+    {
+      "epoch": 0.6569633054030627,
+      "grad_norm": 0.27228009606333625,
+      "learning_rate": 8.34544890062768e-07,
+      "loss": 0.0081,
+      "step": 18190
+    },
+    {
+      "epoch": 0.6573244726957527,
+      "grad_norm": 0.33391573573930167,
+      "learning_rate": 8.329728414844562e-07,
+      "loss": 0.0084,
+      "step": 18200
+    },
+    {
+      "epoch": 0.6576856399884426,
+      "grad_norm": 0.36667406565232147,
+      "learning_rate": 8.314017056064888e-07,
+      "loss": 0.0084,
+      "step": 18210
+    },
+    {
+      "epoch": 0.6580468072811326,
+      "grad_norm": 0.2969311459770572,
+      "learning_rate": 8.298314845786678e-07,
+      "loss": 0.0081,
+      "step": 18220
+    },
+    {
+      "epoch": 0.6584079745738226,
+      "grad_norm": 0.3153271768470532,
+      "learning_rate": 8.282621805495427e-07,
+      "loss": 0.0087,
+      "step": 18230
+    },
+    {
+      "epoch": 0.6587691418665126,
+      "grad_norm": 0.24953400540686899,
+      "learning_rate": 8.266937956664101e-07,
+      "loss": 0.0079,
+      "step": 18240
+    },
+    {
+      "epoch": 0.6591303091592026,
+      "grad_norm": 0.29621145693914774,
+      "learning_rate": 8.251263320753063e-07,
+      "loss": 0.0087,
+      "step": 18250
+    },
+    {
+      "epoch": 0.6594914764518925,
+      "grad_norm": 1.347861401786446,
+      "learning_rate": 8.235597919210096e-07,
+      "loss": 0.0085,
+      "step": 18260
+    },
+    {
+      "epoch": 0.6598526437445825,
+      "grad_norm": 0.3419230242392899,
+      "learning_rate": 8.219941773470327e-07,
+      "loss": 0.0084,
+      "step": 18270
+    },
+    {
+      "epoch": 0.6602138110372725,
+      "grad_norm": 0.27093581635870856,
+      "learning_rate": 8.204294904956235e-07,
+      "loss": 0.0082,
+      "step": 18280
+    },
+    {
+      "epoch": 0.6605749783299625,
+      "grad_norm": 0.2926777516180426,
+      "learning_rate": 8.188657335077591e-07,
+      "loss": 0.0082,
+      "step": 18290
+    },
+    {
+      "epoch": 0.6609361456226525,
+      "grad_norm": 0.31832849948767483,
+      "learning_rate": 8.17302908523145e-07,
+      "loss": 0.0083,
+      "step": 18300
+    },
+    {
+      "epoch": 0.6612973129153424,
+      "grad_norm": 0.2752400818537562,
+      "learning_rate": 8.157410176802101e-07,
+      "loss": 0.0085,
+      "step": 18310
+    },
+    {
+      "epoch": 0.6616584802080323,
+      "grad_norm": 0.2756133029672652,
+      "learning_rate": 8.141800631161082e-07,
+      "loss": 0.0082,
+      "step": 18320
+    },
+    {
+      "epoch": 0.6620196475007223,
+      "grad_norm": 0.3467847274111164,
+      "learning_rate": 8.126200469667082e-07,
+      "loss": 0.0081,
+      "step": 18330
+    },
+    {
+      "epoch": 0.6623808147934123,
+      "grad_norm": 0.24222770685946765,
+      "learning_rate": 8.11060971366598e-07,
+      "loss": 0.008,
+      "step": 18340
+    },
+    {
+      "epoch": 0.6627419820861022,
+      "grad_norm": 0.266813610784373,
+      "learning_rate": 8.095028384490763e-07,
+      "loss": 0.0086,
+      "step": 18350
+    },
+    {
+      "epoch": 0.6631031493787922,
+      "grad_norm": 0.2540664739640425,
+      "learning_rate": 8.079456503461547e-07,
+      "loss": 0.0075,
+      "step": 18360
+    },
+    {
+      "epoch": 0.6634643166714822,
+      "grad_norm": 0.31070825808571606,
+      "learning_rate": 8.063894091885478e-07,
+      "loss": 0.008,
+      "step": 18370
+    },
+    {
+      "epoch": 0.6638254839641722,
+      "grad_norm": 0.2695984302357673,
+      "learning_rate": 8.048341171056786e-07,
+      "loss": 0.0081,
+      "step": 18380
+    },
+    {
+      "epoch": 0.6641866512568622,
+      "grad_norm": 1.303281046937535,
+      "learning_rate": 8.032797762256688e-07,
+      "loss": 0.0087,
+      "step": 18390
+    },
+    {
+      "epoch": 0.6645478185495521,
+      "grad_norm": 0.2811814431989198,
+      "learning_rate": 8.017263886753405e-07,
+      "loss": 0.0086,
+      "step": 18400
+    },
+    {
+      "epoch": 0.6649089858422421,
+      "grad_norm": 0.34515472415621534,
+      "learning_rate": 8.001739565802093e-07,
+      "loss": 0.008,
+      "step": 18410
+    },
+    {
+      "epoch": 0.6652701531349321,
+      "grad_norm": 0.26760755635823463,
+      "learning_rate": 7.986224820644856e-07,
+      "loss": 0.0086,
+      "step": 18420
+    },
+    {
+      "epoch": 0.6656313204276221,
+      "grad_norm": 0.2753366622252768,
+      "learning_rate": 7.970719672510674e-07,
+      "loss": 0.0085,
+      "step": 18430
+    },
+    {
+      "epoch": 0.665992487720312,
+      "grad_norm": 0.3334479820878204,
+      "learning_rate": 7.955224142615418e-07,
+      "loss": 0.0084,
+      "step": 18440
+    },
+    {
+      "epoch": 0.666353655013002,
+      "grad_norm": 0.26688777498095045,
+      "learning_rate": 7.939738252161778e-07,
+      "loss": 0.0077,
+      "step": 18450
+    },
+    {
+      "epoch": 0.666714822305692,
+      "grad_norm": 0.32677120554888267,
+      "learning_rate": 7.924262022339265e-07,
+      "loss": 0.0078,
+      "step": 18460
+    },
+    {
+      "epoch": 0.667075989598382,
+      "grad_norm": 0.31488246235619316,
+      "learning_rate": 7.90879547432416e-07,
+      "loss": 0.0083,
+      "step": 18470
+    },
+    {
+      "epoch": 0.667437156891072,
+      "grad_norm": 0.2526819373851116,
+      "learning_rate": 7.89333862927952e-07,
+      "loss": 0.0085,
+      "step": 18480
+    },
+    {
+      "epoch": 0.667798324183762,
+      "grad_norm": 0.30700668368096085,
+      "learning_rate": 7.877891508355095e-07,
+      "loss": 0.0085,
+      "step": 18490
+    },
+    {
+      "epoch": 0.6681594914764519,
+      "grad_norm": 0.39559256037199897,
+      "learning_rate": 7.862454132687359e-07,
+      "loss": 0.0083,
+      "step": 18500
+    },
+    {
+      "epoch": 0.6685206587691419,
+      "grad_norm": 0.28035355387158584,
+      "learning_rate": 7.847026523399423e-07,
+      "loss": 0.0084,
+      "step": 18510
+    },
+    {
+      "epoch": 0.6688818260618319,
+      "grad_norm": 0.3428774982161356,
+      "learning_rate": 7.831608701601064e-07,
+      "loss": 0.0082,
+      "step": 18520
+    },
+    {
+      "epoch": 0.6692429933545219,
+      "grad_norm": 0.3631394564923445,
+      "learning_rate": 7.816200688388641e-07,
+      "loss": 0.0079,
+      "step": 18530
+    },
+    {
+      "epoch": 0.6696041606472118,
+      "grad_norm": 0.3364599176389308,
+      "learning_rate": 7.8008025048451e-07,
+      "loss": 0.0083,
+      "step": 18540
+    },
+    {
+      "epoch": 0.6699653279399017,
+      "grad_norm": 0.3421451459912487,
+      "learning_rate": 7.785414172039946e-07,
+      "loss": 0.0076,
+      "step": 18550
+    },
+    {
+      "epoch": 0.6703264952325917,
+      "grad_norm": 0.2317193614726613,
+      "learning_rate": 7.770035711029194e-07,
+      "loss": 0.008,
+      "step": 18560
+    },
+    {
+      "epoch": 0.6706876625252817,
+      "grad_norm": 0.3108475767909017,
+      "learning_rate": 7.754667142855359e-07,
+      "loss": 0.0085,
+      "step": 18570
+    },
+    {
+      "epoch": 0.6710488298179716,
+      "grad_norm": 0.25162757414484,
+      "learning_rate": 7.739308488547411e-07,
+      "loss": 0.0079,
+      "step": 18580
+    },
+    {
+      "epoch": 0.6714099971106616,
+      "grad_norm": 0.24705834543508395,
+      "learning_rate": 7.723959769120767e-07,
+      "loss": 0.0078,
+      "step": 18590
+    },
+    {
+      "epoch": 0.6717711644033516,
+      "grad_norm": 0.24211764885671117,
+      "learning_rate": 7.708621005577239e-07,
+      "loss": 0.0083,
+      "step": 18600
+    },
+    {
+      "epoch": 0.6721323316960416,
+      "grad_norm": 0.3328801343686351,
+      "learning_rate": 7.693292218905019e-07,
+      "loss": 0.0082,
+      "step": 18610
+    },
+    {
+      "epoch": 0.6724934989887316,
+      "grad_norm": 0.22479997677481064,
+      "learning_rate": 7.677973430078645e-07,
+      "loss": 0.0075,
+      "step": 18620
+    },
+    {
+      "epoch": 0.6728546662814215,
+      "grad_norm": 0.29296788337685165,
+      "learning_rate": 7.662664660058987e-07,
+      "loss": 0.0083,
+      "step": 18630
+    },
+    {
+      "epoch": 0.6732158335741115,
+      "grad_norm": 0.3361181188292277,
+      "learning_rate": 7.647365929793186e-07,
+      "loss": 0.0087,
+      "step": 18640
+    },
+    {
+      "epoch": 0.6735770008668015,
+      "grad_norm": 0.29523280915654554,
+      "learning_rate": 7.632077260214671e-07,
+      "loss": 0.0082,
+      "step": 18650
+    },
+    {
+      "epoch": 0.6739381681594915,
+      "grad_norm": 0.30167272895714775,
+      "learning_rate": 7.61679867224308e-07,
+      "loss": 0.0081,
+      "step": 18660
+    },
+    {
+      "epoch": 0.6742993354521815,
+      "grad_norm": 0.3278553920787227,
+      "learning_rate": 7.601530186784276e-07,
+      "loss": 0.0077,
+      "step": 18670
+    },
+    {
+      "epoch": 0.6746605027448714,
+      "grad_norm": 0.3231246724496041,
+      "learning_rate": 7.586271824730286e-07,
+      "loss": 0.0079,
+      "step": 18680
+    },
+    {
+      "epoch": 0.6750216700375614,
+      "grad_norm": 0.2950542806764038,
+      "learning_rate": 7.571023606959286e-07,
+      "loss": 0.0081,
+      "step": 18690
+    },
+    {
+      "epoch": 0.6753828373302514,
+      "grad_norm": 0.26755034505971664,
+      "learning_rate": 7.555785554335575e-07,
+      "loss": 0.0089,
+      "step": 18700
+    },
+    {
+      "epoch": 0.6757440046229414,
+      "grad_norm": 0.33808446091565764,
+      "learning_rate": 7.540557687709545e-07,
+      "loss": 0.008,
+      "step": 18710
+    },
+    {
+      "epoch": 0.6761051719156314,
+      "grad_norm": 0.3200210573066207,
+      "learning_rate": 7.525340027917644e-07,
+      "loss": 0.008,
+      "step": 18720
+    },
+    {
+      "epoch": 0.6764663392083213,
+      "grad_norm": 0.5839165391328461,
+      "learning_rate": 7.510132595782363e-07,
+      "loss": 0.0076,
+      "step": 18730
+    },
+    {
+      "epoch": 0.6768275065010113,
+      "grad_norm": 0.29304601873587005,
+      "learning_rate": 7.494935412112183e-07,
+      "loss": 0.0084,
+      "step": 18740
+    },
+    {
+      "epoch": 0.6771886737937013,
+      "grad_norm": 0.23656019957221028,
+      "learning_rate": 7.479748497701592e-07,
+      "loss": 0.0079,
+      "step": 18750
+    },
+    {
+      "epoch": 0.6775498410863913,
+      "grad_norm": 0.3449033557094587,
+      "learning_rate": 7.464571873330979e-07,
+      "loss": 0.0075,
+      "step": 18760
+    },
+    {
+      "epoch": 0.6779110083790811,
+      "grad_norm": 0.3663154076764176,
+      "learning_rate": 7.449405559766701e-07,
+      "loss": 0.0079,
+      "step": 18770
+    },
+    {
+      "epoch": 0.6782721756717711,
+      "grad_norm": 0.2877317270754981,
+      "learning_rate": 7.434249577760974e-07,
+      "loss": 0.0082,
+      "step": 18780
+    },
+    {
+      "epoch": 0.6786333429644611,
+      "grad_norm": 0.3058342790281808,
+      "learning_rate": 7.419103948051899e-07,
+      "loss": 0.008,
+      "step": 18790
+    },
+    {
+      "epoch": 0.6789945102571511,
+      "grad_norm": 0.3026072862047797,
+      "learning_rate": 7.40396869136339e-07,
+      "loss": 0.0072,
+      "step": 18800
+    },
+    {
+      "epoch": 0.679355677549841,
+      "grad_norm": 0.27210954628038325,
+      "learning_rate": 7.388843828405193e-07,
+      "loss": 0.0081,
+      "step": 18810
+    },
+    {
+      "epoch": 0.679716844842531,
+      "grad_norm": 0.2709850219918483,
+      "learning_rate": 7.373729379872806e-07,
+      "loss": 0.0086,
+      "step": 18820
+    },
+    {
+      "epoch": 0.680078012135221,
+      "grad_norm": 0.3177795608554471,
+      "learning_rate": 7.358625366447502e-07,
+      "loss": 0.0079,
+      "step": 18830
+    },
+    {
+      "epoch": 0.680439179427911,
+      "grad_norm": 0.24099882737371411,
+      "learning_rate": 7.343531808796256e-07,
+      "loss": 0.008,
+      "step": 18840
+    },
+    {
+      "epoch": 0.680800346720601,
+      "grad_norm": 0.2736433494827601,
+      "learning_rate": 7.328448727571738e-07,
+      "loss": 0.008,
+      "step": 18850
+    },
+    {
+      "epoch": 0.681161514013291,
+      "grad_norm": 0.4059274443508309,
+      "learning_rate": 7.313376143412298e-07,
+      "loss": 0.0084,
+      "step": 18860
+    },
+    {
+      "epoch": 0.6815226813059809,
+      "grad_norm": 0.39155708096794617,
+      "learning_rate": 7.298314076941905e-07,
+      "loss": 0.0081,
+      "step": 18870
+    },
+    {
+      "epoch": 0.6818838485986709,
+      "grad_norm": 0.33036574017145376,
+      "learning_rate": 7.283262548770154e-07,
+      "loss": 0.0085,
+      "step": 18880
+    },
+    {
+      "epoch": 0.6822450158913609,
+      "grad_norm": 0.34440478168583916,
+      "learning_rate": 7.268221579492204e-07,
+      "loss": 0.0079,
+      "step": 18890
+    },
+    {
+      "epoch": 0.6826061831840509,
+      "grad_norm": 0.2639805391917245,
+      "learning_rate": 7.25319118968877e-07,
+      "loss": 0.008,
+      "step": 18900
+    },
+    {
+      "epoch": 0.6829673504767408,
+      "grad_norm": 0.29456032644442476,
+      "learning_rate": 7.238171399926101e-07,
+      "loss": 0.0077,
+      "step": 18910
+    },
+    {
+      "epoch": 0.6833285177694308,
+      "grad_norm": 0.3075869879586208,
+      "learning_rate": 7.223162230755935e-07,
+      "loss": 0.0082,
+      "step": 18920
+    },
+    {
+      "epoch": 0.6836896850621208,
+      "grad_norm": 0.2542280536531103,
+      "learning_rate": 7.208163702715467e-07,
+      "loss": 0.0078,
+      "step": 18930
+    },
+    {
+      "epoch": 0.6840508523548108,
+      "grad_norm": 0.2614590101859597,
+      "learning_rate": 7.193175836327356e-07,
+      "loss": 0.0073,
+      "step": 18940
+    },
+    {
+      "epoch": 0.6844120196475008,
+      "grad_norm": 0.2601751147078771,
+      "learning_rate": 7.178198652099647e-07,
+      "loss": 0.0081,
+      "step": 18950
+    },
+    {
+      "epoch": 0.6847731869401907,
+      "grad_norm": 0.3715396688789492,
+      "learning_rate": 7.16323217052579e-07,
+      "loss": 0.0082,
+      "step": 18960
+    },
+    {
+      "epoch": 0.6851343542328807,
+      "grad_norm": 0.26801880143268303,
+      "learning_rate": 7.148276412084572e-07,
+      "loss": 0.0079,
+      "step": 18970
+    },
+    {
+      "epoch": 0.6854955215255707,
+      "grad_norm": 0.3030448772248812,
+      "learning_rate": 7.133331397240126e-07,
+      "loss": 0.0073,
+      "step": 18980
+    },
+    {
+      "epoch": 0.6858566888182607,
+      "grad_norm": 0.27247726278173934,
+      "learning_rate": 7.118397146441868e-07,
+      "loss": 0.0077,
+      "step": 18990
+    },
+    {
+      "epoch": 0.6862178561109505,
+      "grad_norm": 0.3037038354190266,
+      "learning_rate": 7.103473680124495e-07,
+      "loss": 0.008,
+      "step": 19000
+    },
+    {
+      "epoch": 0.6862178561109505,
+      "eval_loss": 0.24936704337596893,
+      "eval_runtime": 40.302,
+      "eval_samples_per_second": 101.633,
+      "eval_steps_per_second": 0.199,
+      "step": 19000
+    },
+    {
+      "epoch": 0.6865790234036405,
+      "grad_norm": 0.2635569421135253,
+      "learning_rate": 7.088561018707935e-07,
+      "loss": 0.0077,
+      "step": 19010
+    },
+    {
+      "epoch": 0.6869401906963305,
+      "grad_norm": 0.3627411229014034,
+      "learning_rate": 7.073659182597354e-07,
+      "loss": 0.0078,
+      "step": 19020
+    },
+    {
+      "epoch": 0.6873013579890205,
+      "grad_norm": 0.22755039479820732,
+      "learning_rate": 7.058768192183082e-07,
+      "loss": 0.0074,
+      "step": 19030
+    },
+    {
+      "epoch": 0.6876625252817105,
+      "grad_norm": 0.3180795566316627,
+      "learning_rate": 7.04388806784063e-07,
+      "loss": 0.0079,
+      "step": 19040
+    },
+    {
+      "epoch": 0.6880236925744004,
+      "grad_norm": 0.28142515369118465,
+      "learning_rate": 7.029018829930619e-07,
+      "loss": 0.0086,
+      "step": 19050
+    },
+    {
+      "epoch": 0.6883848598670904,
+      "grad_norm": 0.2545172381667943,
+      "learning_rate": 7.014160498798795e-07,
+      "loss": 0.008,
+      "step": 19060
+    },
+    {
+      "epoch": 0.6887460271597804,
+      "grad_norm": 0.2936178530439519,
+      "learning_rate": 6.999313094775966e-07,
+      "loss": 0.0081,
+      "step": 19070
+    },
+    {
+      "epoch": 0.6891071944524704,
+      "grad_norm": 0.30789976089161764,
+      "learning_rate": 6.984476638177994e-07,
+      "loss": 0.0085,
+      "step": 19080
+    },
+    {
+      "epoch": 0.6894683617451604,
+      "grad_norm": 0.5632908788664931,
+      "learning_rate": 6.96965114930575e-07,
+      "loss": 0.008,
+      "step": 19090
+    },
+    {
+      "epoch": 0.6898295290378503,
+      "grad_norm": 0.29185541474980053,
+      "learning_rate": 6.954836648445125e-07,
+      "loss": 0.0076,
+      "step": 19100
+    },
+    {
+      "epoch": 0.6901906963305403,
+      "grad_norm": 0.25144068750214865,
+      "learning_rate": 6.940033155866942e-07,
+      "loss": 0.008,
+      "step": 19110
+    },
+    {
+      "epoch": 0.6905518636232303,
+      "grad_norm": 0.21903994093527124,
+      "learning_rate": 6.925240691826991e-07,
+      "loss": 0.0077,
+      "step": 19120
+    },
+    {
+      "epoch": 0.6909130309159203,
+      "grad_norm": 0.25841993592182544,
+      "learning_rate": 6.910459276565945e-07,
+      "loss": 0.008,
+      "step": 19130
+    },
+    {
+      "epoch": 0.6912741982086102,
+      "grad_norm": 0.31524445489475666,
+      "learning_rate": 6.895688930309383e-07,
+      "loss": 0.0078,
+      "step": 19140
+    },
+    {
+      "epoch": 0.6916353655013002,
+      "grad_norm": 0.27139440919118474,
+      "learning_rate": 6.880929673267721e-07,
+      "loss": 0.0078,
+      "step": 19150
+    },
+    {
+      "epoch": 0.6919965327939902,
+      "grad_norm": 0.33660862463061175,
+      "learning_rate": 6.866181525636206e-07,
+      "loss": 0.0076,
+      "step": 19160
+    },
+    {
+      "epoch": 0.6923577000866802,
+      "grad_norm": 0.5417517169530579,
+      "learning_rate": 6.851444507594882e-07,
+      "loss": 0.0083,
+      "step": 19170
+    },
+    {
+      "epoch": 0.6927188673793702,
+      "grad_norm": 0.36028765628201337,
+      "learning_rate": 6.836718639308578e-07,
+      "loss": 0.0085,
+      "step": 19180
+    },
+    {
+      "epoch": 0.6930800346720601,
+      "grad_norm": 0.30273169298207453,
+      "learning_rate": 6.822003940926843e-07,
+      "loss": 0.0075,
+      "step": 19190
+    },
+    {
+      "epoch": 0.6934412019647501,
+      "grad_norm": 0.4097980444519036,
+      "learning_rate": 6.807300432583967e-07,
+      "loss": 0.0079,
+      "step": 19200
+    },
+    {
+      "epoch": 0.6938023692574401,
+      "grad_norm": 0.3267057148518776,
+      "learning_rate": 6.792608134398905e-07,
+      "loss": 0.0076,
+      "step": 19210
+    },
+    {
+      "epoch": 0.69416353655013,
+      "grad_norm": 0.30457728823697994,
+      "learning_rate": 6.777927066475297e-07,
+      "loss": 0.0079,
+      "step": 19220
+    },
+    {
+      "epoch": 0.69452470384282,
+      "grad_norm": 0.24782121744823332,
+      "learning_rate": 6.763257248901399e-07,
+      "loss": 0.0072,
+      "step": 19230
+    },
+    {
+      "epoch": 0.6948858711355099,
+      "grad_norm": 0.31703934871279715,
+      "learning_rate": 6.74859870175007e-07,
+      "loss": 0.0079,
+      "step": 19240
+    },
+    {
+      "epoch": 0.6952470384281999,
+      "grad_norm": 0.25803962173832784,
+      "learning_rate": 6.733951445078769e-07,
+      "loss": 0.0073,
+      "step": 19250
+    },
+    {
+      "epoch": 0.6956082057208899,
+      "grad_norm": 0.27515315941473206,
+      "learning_rate": 6.719315498929481e-07,
+      "loss": 0.0076,
+      "step": 19260
+    },
+    {
+      "epoch": 0.6959693730135799,
+      "grad_norm": 0.21397486175260247,
+      "learning_rate": 6.704690883328738e-07,
+      "loss": 0.0071,
+      "step": 19270
+    },
+    {
+      "epoch": 0.6963305403062698,
+      "grad_norm": 0.3201514441939309,
+      "learning_rate": 6.690077618287545e-07,
+      "loss": 0.0077,
+      "step": 19280
+    },
+    {
+      "epoch": 0.6966917075989598,
+      "grad_norm": 0.33938521336945043,
+      "learning_rate": 6.6754757238014e-07,
+      "loss": 0.0077,
+      "step": 19290
+    },
+    {
+      "epoch": 0.6970528748916498,
+      "grad_norm": 0.26973764361887137,
+      "learning_rate": 6.660885219850225e-07,
+      "loss": 0.0079,
+      "step": 19300
+    },
+    {
+      "epoch": 0.6974140421843398,
+      "grad_norm": 0.25965186346706864,
+      "learning_rate": 6.646306126398358e-07,
+      "loss": 0.0074,
+      "step": 19310
+    },
+    {
+      "epoch": 0.6977752094770298,
+      "grad_norm": 0.2915533149052536,
+      "learning_rate": 6.631738463394526e-07,
+      "loss": 0.0073,
+      "step": 19320
+    },
+    {
+      "epoch": 0.6981363767697197,
+      "grad_norm": 0.3115306212811213,
+      "learning_rate": 6.617182250771828e-07,
+      "loss": 0.0076,
+      "step": 19330
+    },
+    {
+      "epoch": 0.6984975440624097,
+      "grad_norm": 0.2805236228636396,
+      "learning_rate": 6.602637508447669e-07,
+      "loss": 0.008,
+      "step": 19340
+    },
+    {
+      "epoch": 0.6988587113550997,
+      "grad_norm": 0.437646927951495,
+      "learning_rate": 6.58810425632379e-07,
+      "loss": 0.0076,
+      "step": 19350
+    },
+    {
+      "epoch": 0.6992198786477897,
+      "grad_norm": 0.22681522039870827,
+      "learning_rate": 6.573582514286181e-07,
+      "loss": 0.0076,
+      "step": 19360
+    },
+    {
+      "epoch": 0.6995810459404797,
+      "grad_norm": 0.31899048496693333,
+      "learning_rate": 6.559072302205106e-07,
+      "loss": 0.0077,
+      "step": 19370
+    },
+    {
+      "epoch": 0.6999422132331696,
+      "grad_norm": 0.3162836420100947,
+      "learning_rate": 6.544573639935041e-07,
+      "loss": 0.0082,
+      "step": 19380
+    },
+    {
+      "epoch": 0.7003033805258596,
+      "grad_norm": 0.2960588880163288,
+      "learning_rate": 6.530086547314656e-07,
+      "loss": 0.0078,
+      "step": 19390
+    },
+    {
+      "epoch": 0.7006645478185496,
+      "grad_norm": 0.23886163307321936,
+      "learning_rate": 6.515611044166787e-07,
+      "loss": 0.0076,
+      "step": 19400
+    },
+    {
+      "epoch": 0.7010257151112396,
+      "grad_norm": 0.26451849846012093,
+      "learning_rate": 6.501147150298431e-07,
+      "loss": 0.0077,
+      "step": 19410
+    },
+    {
+      "epoch": 0.7013868824039295,
+      "grad_norm": 0.24207949537942544,
+      "learning_rate": 6.486694885500677e-07,
+      "loss": 0.0076,
+      "step": 19420
+    },
+    {
+      "epoch": 0.7017480496966195,
+      "grad_norm": 0.26077069931151303,
+      "learning_rate": 6.472254269548727e-07,
+      "loss": 0.008,
+      "step": 19430
+    },
+    {
+      "epoch": 0.7021092169893095,
+      "grad_norm": 0.265309999754371,
+      "learning_rate": 6.457825322201811e-07,
+      "loss": 0.0081,
+      "step": 19440
+    },
+    {
+      "epoch": 0.7024703842819994,
+      "grad_norm": 0.5462175494527659,
+      "learning_rate": 6.443408063203226e-07,
+      "loss": 0.0075,
+      "step": 19450
+    },
+    {
+      "epoch": 0.7028315515746894,
+      "grad_norm": 0.368057232016714,
+      "learning_rate": 6.429002512280256e-07,
+      "loss": 0.0075,
+      "step": 19460
+    },
+    {
+      "epoch": 0.7031927188673793,
+      "grad_norm": 0.3942240299359542,
+      "learning_rate": 6.414608689144166e-07,
+      "loss": 0.0079,
+      "step": 19470
+    },
+    {
+      "epoch": 0.7035538861600693,
+      "grad_norm": 0.2583877662912019,
+      "learning_rate": 6.400226613490174e-07,
+      "loss": 0.0072,
+      "step": 19480
+    },
+    {
+      "epoch": 0.7039150534527593,
+      "grad_norm": 0.3540869353357415,
+      "learning_rate": 6.385856304997438e-07,
+      "loss": 0.0075,
+      "step": 19490
+    },
+    {
+      "epoch": 0.7042762207454493,
+      "grad_norm": 0.20930067059310353,
+      "learning_rate": 6.371497783328991e-07,
+      "loss": 0.0075,
+      "step": 19500
+    },
+    {
+      "epoch": 0.7046373880381392,
+      "grad_norm": 0.338618909259099,
+      "learning_rate": 6.357151068131764e-07,
+      "loss": 0.0077,
+      "step": 19510
+    },
+    {
+      "epoch": 0.7049985553308292,
+      "grad_norm": 0.2692716282155986,
+      "learning_rate": 6.342816179036506e-07,
+      "loss": 0.0076,
+      "step": 19520
+    },
+    {
+      "epoch": 0.7053597226235192,
+      "grad_norm": 0.2784206817149558,
+      "learning_rate": 6.32849313565781e-07,
+      "loss": 0.0078,
+      "step": 19530
+    },
+    {
+      "epoch": 0.7057208899162092,
+      "grad_norm": 0.2603889627324874,
+      "learning_rate": 6.314181957594045e-07,
+      "loss": 0.0075,
+      "step": 19540
+    },
+    {
+      "epoch": 0.7060820572088992,
+      "grad_norm": 0.3129203322672767,
+      "learning_rate": 6.29988266442734e-07,
+      "loss": 0.0078,
+      "step": 19550
+    },
+    {
+      "epoch": 0.7064432245015891,
+      "grad_norm": 0.2820455309079887,
+      "learning_rate": 6.28559527572358e-07,
+      "loss": 0.0077,
+      "step": 19560
+    },
+    {
+      "epoch": 0.7068043917942791,
+      "grad_norm": 0.2851787452812087,
+      "learning_rate": 6.271319811032348e-07,
+      "loss": 0.0077,
+      "step": 19570
+    },
+    {
+      "epoch": 0.7071655590869691,
+      "grad_norm": 0.35707786228643584,
+      "learning_rate": 6.257056289886907e-07,
+      "loss": 0.0078,
+      "step": 19580
+    },
+    {
+      "epoch": 0.7075267263796591,
+      "grad_norm": 0.2710538863256709,
+      "learning_rate": 6.242804731804194e-07,
+      "loss": 0.008,
+      "step": 19590
+    },
+    {
+      "epoch": 0.7078878936723491,
+      "grad_norm": 0.2564028859654098,
+      "learning_rate": 6.228565156284759e-07,
+      "loss": 0.0079,
+      "step": 19600
+    },
+    {
+      "epoch": 0.708249060965039,
+      "grad_norm": 0.7766689322357374,
+      "learning_rate": 6.214337582812775e-07,
+      "loss": 0.0082,
+      "step": 19610
+    },
+    {
+      "epoch": 0.708610228257729,
+      "grad_norm": 0.3190144564973093,
+      "learning_rate": 6.200122030855973e-07,
+      "loss": 0.0079,
+      "step": 19620
+    },
+    {
+      "epoch": 0.708971395550419,
+      "grad_norm": 0.27976506940875834,
+      "learning_rate": 6.185918519865642e-07,
+      "loss": 0.0071,
+      "step": 19630
+    },
+    {
+      "epoch": 0.709332562843109,
+      "grad_norm": 0.3681035220796888,
+      "learning_rate": 6.171727069276606e-07,
+      "loss": 0.0078,
+      "step": 19640
+    },
+    {
+      "epoch": 0.709693730135799,
+      "grad_norm": 0.3108117935229017,
+      "learning_rate": 6.157547698507168e-07,
+      "loss": 0.0073,
+      "step": 19650
+    },
+    {
+      "epoch": 0.7100548974284889,
+      "grad_norm": 0.4492972571225921,
+      "learning_rate": 6.143380426959122e-07,
+      "loss": 0.0075,
+      "step": 19660
+    },
+    {
+      "epoch": 0.7104160647211788,
+      "grad_norm": 0.26482038415083653,
+      "learning_rate": 6.129225274017684e-07,
+      "loss": 0.0081,
+      "step": 19670
+    },
+    {
+      "epoch": 0.7107772320138688,
+      "grad_norm": 0.27047207121820643,
+      "learning_rate": 6.115082259051513e-07,
+      "loss": 0.0074,
+      "step": 19680
+    },
+    {
+      "epoch": 0.7111383993065588,
+      "grad_norm": 0.6540137119341015,
+      "learning_rate": 6.100951401412639e-07,
+      "loss": 0.0078,
+      "step": 19690
+    },
+    {
+      "epoch": 0.7114995665992487,
+      "grad_norm": 0.3132600199468232,
+      "learning_rate": 6.086832720436466e-07,
+      "loss": 0.0078,
+      "step": 19700
+    },
+    {
+      "epoch": 0.7118607338919387,
+      "grad_norm": 0.23512183591813887,
+      "learning_rate": 6.072726235441729e-07,
+      "loss": 0.0076,
+      "step": 19710
+    },
+    {
+      "epoch": 0.7122219011846287,
+      "grad_norm": 0.31574518543574115,
+      "learning_rate": 6.058631965730492e-07,
+      "loss": 0.0078,
+      "step": 19720
+    },
+    {
+      "epoch": 0.7125830684773187,
+      "grad_norm": 0.293062942001639,
+      "learning_rate": 6.044549930588084e-07,
+      "loss": 0.0073,
+      "step": 19730
+    },
+    {
+      "epoch": 0.7129442357700086,
+      "grad_norm": 0.3022594709850378,
+      "learning_rate": 6.030480149283114e-07,
+      "loss": 0.0078,
+      "step": 19740
+    },
+    {
+      "epoch": 0.7133054030626986,
+      "grad_norm": 0.3468638879831703,
+      "learning_rate": 6.0164226410674e-07,
+      "loss": 0.0072,
+      "step": 19750
+    },
+    {
+      "epoch": 0.7136665703553886,
+      "grad_norm": 0.2570387636303115,
+      "learning_rate": 6.002377425175995e-07,
+      "loss": 0.0075,
+      "step": 19760
+    },
+    {
+      "epoch": 0.7140277376480786,
+      "grad_norm": 0.31858067068643076,
+      "learning_rate": 5.988344520827107e-07,
+      "loss": 0.0071,
+      "step": 19770
+    },
+    {
+      "epoch": 0.7143889049407686,
+      "grad_norm": 0.27927278902242075,
+      "learning_rate": 5.974323947222111e-07,
+      "loss": 0.0073,
+      "step": 19780
+    },
+    {
+      "epoch": 0.7147500722334585,
+      "grad_norm": 0.33634349693327237,
+      "learning_rate": 5.9603157235455e-07,
+      "loss": 0.0074,
+      "step": 19790
+    },
+    {
+      "epoch": 0.7151112395261485,
+      "grad_norm": 0.3047791152201767,
+      "learning_rate": 5.946319868964885e-07,
+      "loss": 0.0076,
+      "step": 19800
+    },
+    {
+      "epoch": 0.7154724068188385,
+      "grad_norm": 0.29612099969487793,
+      "learning_rate": 5.932336402630937e-07,
+      "loss": 0.0074,
+      "step": 19810
+    },
+    {
+      "epoch": 0.7158335741115285,
+      "grad_norm": 0.2789469842214419,
+      "learning_rate": 5.918365343677386e-07,
+      "loss": 0.0075,
+      "step": 19820
+    },
+    {
+      "epoch": 0.7161947414042185,
+      "grad_norm": 0.2883955639807882,
+      "learning_rate": 5.904406711220971e-07,
+      "loss": 0.0075,
+      "step": 19830
+    },
+    {
+      "epoch": 0.7165559086969084,
+      "grad_norm": 0.24683301014188813,
+      "learning_rate": 5.89046052436145e-07,
+      "loss": 0.0081,
+      "step": 19840
+    },
+    {
+      "epoch": 0.7169170759895984,
+      "grad_norm": 0.29526985240916565,
+      "learning_rate": 5.876526802181532e-07,
+      "loss": 0.008,
+      "step": 19850
+    },
+    {
+      "epoch": 0.7172782432822884,
+      "grad_norm": 0.3041287559972048,
+      "learning_rate": 5.862605563746876e-07,
+      "loss": 0.0076,
+      "step": 19860
+    },
+    {
+      "epoch": 0.7176394105749784,
+      "grad_norm": 0.27508769937529776,
+      "learning_rate": 5.848696828106058e-07,
+      "loss": 0.0075,
+      "step": 19870
+    },
+    {
+      "epoch": 0.7180005778676684,
+      "grad_norm": 0.3151140645118536,
+      "learning_rate": 5.834800614290555e-07,
+      "loss": 0.0075,
+      "step": 19880
+    },
+    {
+      "epoch": 0.7183617451603582,
+      "grad_norm": 0.21934535156347199,
+      "learning_rate": 5.820916941314697e-07,
+      "loss": 0.0073,
+      "step": 19890
+    },
+    {
+      "epoch": 0.7187229124530482,
+      "grad_norm": 0.27230568874103456,
+      "learning_rate": 5.807045828175671e-07,
+      "loss": 0.0082,
+      "step": 19900
+    },
+    {
+      "epoch": 0.7190840797457382,
+      "grad_norm": 0.4580465779147444,
+      "learning_rate": 5.793187293853456e-07,
+      "loss": 0.0073,
+      "step": 19910
+    },
+    {
+      "epoch": 0.7194452470384282,
+      "grad_norm": 0.2937656995853014,
+      "learning_rate": 5.779341357310845e-07,
+      "loss": 0.007,
+      "step": 19920
+    },
+    {
+      "epoch": 0.7198064143311181,
+      "grad_norm": 0.308946099263814,
+      "learning_rate": 5.765508037493371e-07,
+      "loss": 0.0075,
+      "step": 19930
+    },
+    {
+      "epoch": 0.7201675816238081,
+      "grad_norm": 0.32824497243756634,
+      "learning_rate": 5.751687353329311e-07,
+      "loss": 0.0079,
+      "step": 19940
+    },
+    {
+      "epoch": 0.7205287489164981,
+      "grad_norm": 0.2294652061335911,
+      "learning_rate": 5.737879323729663e-07,
+      "loss": 0.0074,
+      "step": 19950
+    },
+    {
+      "epoch": 0.7208899162091881,
+      "grad_norm": 0.3093832337696478,
+      "learning_rate": 5.724083967588088e-07,
+      "loss": 0.0077,
+      "step": 19960
+    },
+    {
+      "epoch": 0.721251083501878,
+      "grad_norm": 0.35177152466082473,
+      "learning_rate": 5.710301303780932e-07,
+      "loss": 0.0075,
+      "step": 19970
+    },
+    {
+      "epoch": 0.721612250794568,
+      "grad_norm": 0.31818211410765507,
+      "learning_rate": 5.696531351167155e-07,
+      "loss": 0.0074,
+      "step": 19980
+    },
+    {
+      "epoch": 0.721973418087258,
+      "grad_norm": 0.29703865033023286,
+      "learning_rate": 5.682774128588322e-07,
+      "loss": 0.0075,
+      "step": 19990
+    },
+    {
+      "epoch": 0.722334585379948,
+      "grad_norm": 0.2867183150814492,
+      "learning_rate": 5.669029654868596e-07,
+      "loss": 0.0069,
+      "step": 20000
+    },
+    {
+      "epoch": 0.722334585379948,
+      "eval_loss": 0.2589828073978424,
+      "eval_runtime": 38.9734,
+      "eval_samples_per_second": 105.097,
+      "eval_steps_per_second": 0.205,
+      "step": 20000
+    },
+    {
+      "epoch": 0.722695752672638,
+      "grad_norm": 0.25981622524461534,
+      "learning_rate": 5.655297948814685e-07,
+      "loss": 0.0069,
+      "step": 20010
+    },
+    {
+      "epoch": 0.723056919965328,
+      "grad_norm": 0.3259687372503028,
+      "learning_rate": 5.641579029215821e-07,
+      "loss": 0.0074,
+      "step": 20020
+    },
+    {
+      "epoch": 0.7234180872580179,
+      "grad_norm": 0.2884365615081812,
+      "learning_rate": 5.627872914843758e-07,
+      "loss": 0.0074,
+      "step": 20030
+    },
+    {
+      "epoch": 0.7237792545507079,
+      "grad_norm": 0.704725914948263,
+      "learning_rate": 5.614179624452707e-07,
+      "loss": 0.0075,
+      "step": 20040
+    },
+    {
+      "epoch": 0.7241404218433979,
+      "grad_norm": 0.2879345072016374,
+      "learning_rate": 5.600499176779353e-07,
+      "loss": 0.0074,
+      "step": 20050
+    },
+    {
+      "epoch": 0.7245015891360879,
+      "grad_norm": 0.20831800522859503,
+      "learning_rate": 5.58683159054279e-07,
+      "loss": 0.0078,
+      "step": 20060
+    },
+    {
+      "epoch": 0.7248627564287778,
+      "grad_norm": 0.23486901041774255,
+      "learning_rate": 5.573176884444527e-07,
+      "loss": 0.0078,
+      "step": 20070
+    },
+    {
+      "epoch": 0.7252239237214678,
+      "grad_norm": 0.2897582621993592,
+      "learning_rate": 5.559535077168443e-07,
+      "loss": 0.0074,
+      "step": 20080
+    },
+    {
+      "epoch": 0.7255850910141578,
+      "grad_norm": 0.3324862321103921,
+      "learning_rate": 5.545906187380765e-07,
+      "loss": 0.0075,
+      "step": 20090
+    },
+    {
+      "epoch": 0.7259462583068478,
+      "grad_norm": 0.2357519700580484,
+      "learning_rate": 5.532290233730042e-07,
+      "loss": 0.0073,
+      "step": 20100
+    },
+    {
+      "epoch": 0.7263074255995378,
+      "grad_norm": 0.25493606144389214,
+      "learning_rate": 5.51868723484714e-07,
+      "loss": 0.0077,
+      "step": 20110
+    },
+    {
+      "epoch": 0.7266685928922276,
+      "grad_norm": 0.35237029565025474,
+      "learning_rate": 5.505097209345175e-07,
+      "loss": 0.0076,
+      "step": 20120
+    },
+    {
+      "epoch": 0.7270297601849176,
+      "grad_norm": 0.29147566615509823,
+      "learning_rate": 5.491520175819535e-07,
+      "loss": 0.0081,
+      "step": 20130
+    },
+    {
+      "epoch": 0.7273909274776076,
+      "grad_norm": 0.3073984655511107,
+      "learning_rate": 5.477956152847809e-07,
+      "loss": 0.0077,
+      "step": 20140
+    },
+    {
+      "epoch": 0.7277520947702976,
+      "grad_norm": 0.25291978816435945,
+      "learning_rate": 5.464405158989803e-07,
+      "loss": 0.0073,
+      "step": 20150
+    },
+    {
+      "epoch": 0.7281132620629875,
+      "grad_norm": 0.32582317411731565,
+      "learning_rate": 5.45086721278748e-07,
+      "loss": 0.0078,
+      "step": 20160
+    },
+    {
+      "epoch": 0.7284744293556775,
+      "grad_norm": 0.316388557417912,
+      "learning_rate": 5.437342332764957e-07,
+      "loss": 0.0075,
+      "step": 20170
+    },
+    {
+      "epoch": 0.7288355966483675,
+      "grad_norm": 0.30659777378419867,
+      "learning_rate": 5.423830537428465e-07,
+      "loss": 0.0075,
+      "step": 20180
+    },
+    {
+      "epoch": 0.7291967639410575,
+      "grad_norm": 0.2853034085308732,
+      "learning_rate": 5.410331845266348e-07,
+      "loss": 0.0073,
+      "step": 20190
+    },
+    {
+      "epoch": 0.7295579312337475,
+      "grad_norm": 0.23327364176228504,
+      "learning_rate": 5.396846274748998e-07,
+      "loss": 0.0071,
+      "step": 20200
+    },
+    {
+      "epoch": 0.7299190985264374,
+      "grad_norm": 0.4205739839982394,
+      "learning_rate": 5.383373844328877e-07,
+      "loss": 0.0076,
+      "step": 20210
+    },
+    {
+      "epoch": 0.7302802658191274,
+      "grad_norm": 0.2546710889545267,
+      "learning_rate": 5.369914572440443e-07,
+      "loss": 0.0075,
+      "step": 20220
+    },
+    {
+      "epoch": 0.7306414331118174,
+      "grad_norm": 0.2867789937707993,
+      "learning_rate": 5.356468477500171e-07,
+      "loss": 0.0075,
+      "step": 20230
+    },
+    {
+      "epoch": 0.7310026004045074,
+      "grad_norm": 0.3523361520906528,
+      "learning_rate": 5.34303557790649e-07,
+      "loss": 0.008,
+      "step": 20240
+    },
+    {
+      "epoch": 0.7313637676971974,
+      "grad_norm": 0.310489681390172,
+      "learning_rate": 5.329615892039779e-07,
+      "loss": 0.0082,
+      "step": 20250
+    },
+    {
+      "epoch": 0.7317249349898873,
+      "grad_norm": 0.8171598419353219,
+      "learning_rate": 5.316209438262334e-07,
+      "loss": 0.008,
+      "step": 20260
+    },
+    {
+      "epoch": 0.7320861022825773,
+      "grad_norm": 0.2608283005267198,
+      "learning_rate": 5.302816234918354e-07,
+      "loss": 0.0075,
+      "step": 20270
+    },
+    {
+      "epoch": 0.7324472695752673,
+      "grad_norm": 0.32721217684353804,
+      "learning_rate": 5.289436300333895e-07,
+      "loss": 0.0079,
+      "step": 20280
+    },
+    {
+      "epoch": 0.7328084368679573,
+      "grad_norm": 0.3049759100139708,
+      "learning_rate": 5.276069652816873e-07,
+      "loss": 0.0078,
+      "step": 20290
+    },
+    {
+      "epoch": 0.7331696041606472,
+      "grad_norm": 0.3011863901130711,
+      "learning_rate": 5.262716310657002e-07,
+      "loss": 0.0075,
+      "step": 20300
+    },
+    {
+      "epoch": 0.7335307714533372,
+      "grad_norm": 0.30749386846289717,
+      "learning_rate": 5.249376292125812e-07,
+      "loss": 0.0072,
+      "step": 20310
+    },
+    {
+      "epoch": 0.7338919387460272,
+      "grad_norm": 0.4758145893072713,
+      "learning_rate": 5.236049615476588e-07,
+      "loss": 0.0074,
+      "step": 20320
+    },
+    {
+      "epoch": 0.7342531060387172,
+      "grad_norm": 0.324343690711234,
+      "learning_rate": 5.22273629894436e-07,
+      "loss": 0.0076,
+      "step": 20330
+    },
+    {
+      "epoch": 0.734614273331407,
+      "grad_norm": 0.37827923093810245,
+      "learning_rate": 5.209436360745887e-07,
+      "loss": 0.0074,
+      "step": 20340
+    },
+    {
+      "epoch": 0.734975440624097,
+      "grad_norm": 0.3193557346801481,
+      "learning_rate": 5.196149819079608e-07,
+      "loss": 0.0073,
+      "step": 20350
+    },
+    {
+      "epoch": 0.735336607916787,
+      "grad_norm": 0.2428309752099654,
+      "learning_rate": 5.182876692125648e-07,
+      "loss": 0.0073,
+      "step": 20360
+    },
+    {
+      "epoch": 0.735697775209477,
+      "grad_norm": 0.44681745443003823,
+      "learning_rate": 5.169616998045757e-07,
+      "loss": 0.0076,
+      "step": 20370
+    },
+    {
+      "epoch": 0.736058942502167,
+      "grad_norm": 0.20996693756084064,
+      "learning_rate": 5.156370754983326e-07,
+      "loss": 0.0069,
+      "step": 20380
+    },
+    {
+      "epoch": 0.736420109794857,
+      "grad_norm": 0.2574720321557167,
+      "learning_rate": 5.14313798106332e-07,
+      "loss": 0.0076,
+      "step": 20390
+    },
+    {
+      "epoch": 0.7367812770875469,
+      "grad_norm": 0.3105525343785717,
+      "learning_rate": 5.1299186943923e-07,
+      "loss": 0.0078,
+      "step": 20400
+    },
+    {
+      "epoch": 0.7371424443802369,
+      "grad_norm": 0.34515243099635207,
+      "learning_rate": 5.116712913058335e-07,
+      "loss": 0.0072,
+      "step": 20410
+    },
+    {
+      "epoch": 0.7375036116729269,
+      "grad_norm": 0.8032794338277798,
+      "learning_rate": 5.10352065513105e-07,
+      "loss": 0.0073,
+      "step": 20420
+    },
+    {
+      "epoch": 0.7378647789656169,
+      "grad_norm": 0.25027721106694595,
+      "learning_rate": 5.090341938661545e-07,
+      "loss": 0.0072,
+      "step": 20430
+    },
+    {
+      "epoch": 0.7382259462583068,
+      "grad_norm": 0.28845834017336264,
+      "learning_rate": 5.077176781682405e-07,
+      "loss": 0.0073,
+      "step": 20440
+    },
+    {
+      "epoch": 0.7385871135509968,
+      "grad_norm": 0.2405092289353493,
+      "learning_rate": 5.064025202207645e-07,
+      "loss": 0.0073,
+      "step": 20450
+    },
+    {
+      "epoch": 0.7389482808436868,
+      "grad_norm": 0.2595028649172712,
+      "learning_rate": 5.050887218232718e-07,
+      "loss": 0.007,
+      "step": 20460
+    },
+    {
+      "epoch": 0.7393094481363768,
+      "grad_norm": 0.24652985650611905,
+      "learning_rate": 5.037762847734463e-07,
+      "loss": 0.0072,
+      "step": 20470
+    },
+    {
+      "epoch": 0.7396706154290668,
+      "grad_norm": 0.3476481102138882,
+      "learning_rate": 5.024652108671106e-07,
+      "loss": 0.0073,
+      "step": 20480
+    },
+    {
+      "epoch": 0.7400317827217567,
+      "grad_norm": 0.24232972015179027,
+      "learning_rate": 5.011555018982189e-07,
+      "loss": 0.0071,
+      "step": 20490
+    },
+    {
+      "epoch": 0.7403929500144467,
+      "grad_norm": 0.23645209257749475,
+      "learning_rate": 4.99847159658862e-07,
+      "loss": 0.0071,
+      "step": 20500
+    },
+    {
+      "epoch": 0.7407541173071367,
+      "grad_norm": 0.2609779364117531,
+      "learning_rate": 4.985401859392568e-07,
+      "loss": 0.0072,
+      "step": 20510
+    },
+    {
+      "epoch": 0.7411152845998267,
+      "grad_norm": 0.33825492094481496,
+      "learning_rate": 4.97234582527751e-07,
+      "loss": 0.0072,
+      "step": 20520
+    },
+    {
+      "epoch": 0.7414764518925167,
+      "grad_norm": 0.39172348510987864,
+      "learning_rate": 4.95930351210814e-07,
+      "loss": 0.0074,
+      "step": 20530
+    },
+    {
+      "epoch": 0.7418376191852066,
+      "grad_norm": 0.30219706750319514,
+      "learning_rate": 4.946274937730407e-07,
+      "loss": 0.0076,
+      "step": 20540
+    },
+    {
+      "epoch": 0.7421987864778966,
+      "grad_norm": 0.2873461299246974,
+      "learning_rate": 4.933260119971443e-07,
+      "loss": 0.0074,
+      "step": 20550
+    },
+    {
+      "epoch": 0.7425599537705866,
+      "grad_norm": 0.8782936474922445,
+      "learning_rate": 4.920259076639557e-07,
+      "loss": 0.0074,
+      "step": 20560
+    },
+    {
+      "epoch": 0.7429211210632765,
+      "grad_norm": 0.28269843332190137,
+      "learning_rate": 4.907271825524216e-07,
+      "loss": 0.0076,
+      "step": 20570
+    },
+    {
+      "epoch": 0.7432822883559664,
+      "grad_norm": 0.3953165016196657,
+      "learning_rate": 4.89429838439602e-07,
+      "loss": 0.0071,
+      "step": 20580
+    },
+    {
+      "epoch": 0.7436434556486564,
+      "grad_norm": 0.33385834096098527,
+      "learning_rate": 4.881338771006656e-07,
+      "loss": 0.0068,
+      "step": 20590
+    },
+    {
+      "epoch": 0.7440046229413464,
+      "grad_norm": 0.3686319700247318,
+      "learning_rate": 4.868393003088912e-07,
+      "loss": 0.0076,
+      "step": 20600
+    },
+    {
+      "epoch": 0.7443657902340364,
+      "grad_norm": 0.26984412173114913,
+      "learning_rate": 4.855461098356608e-07,
+      "loss": 0.0076,
+      "step": 20610
+    },
+    {
+      "epoch": 0.7447269575267264,
+      "grad_norm": 0.29395715213266843,
+      "learning_rate": 4.84254307450462e-07,
+      "loss": 0.0075,
+      "step": 20620
+    },
+    {
+      "epoch": 0.7450881248194163,
+      "grad_norm": 0.23947242655724624,
+      "learning_rate": 4.829638949208806e-07,
+      "loss": 0.0068,
+      "step": 20630
+    },
+    {
+      "epoch": 0.7454492921121063,
+      "grad_norm": 0.18645335478537792,
+      "learning_rate": 4.81674874012602e-07,
+      "loss": 0.0071,
+      "step": 20640
+    },
+    {
+      "epoch": 0.7458104594047963,
+      "grad_norm": 0.2693453543609079,
+      "learning_rate": 4.803872464894078e-07,
+      "loss": 0.0072,
+      "step": 20650
+    },
+    {
+      "epoch": 0.7461716266974863,
+      "grad_norm": 0.2898976621313203,
+      "learning_rate": 4.791010141131719e-07,
+      "loss": 0.0079,
+      "step": 20660
+    },
+    {
+      "epoch": 0.7465327939901762,
+      "grad_norm": 0.30382236516295014,
+      "learning_rate": 4.778161786438593e-07,
+      "loss": 0.0069,
+      "step": 20670
+    },
+    {
+      "epoch": 0.7468939612828662,
+      "grad_norm": 0.3418580217532132,
+      "learning_rate": 4.765327418395251e-07,
+      "loss": 0.0075,
+      "step": 20680
+    },
+    {
+      "epoch": 0.7472551285755562,
+      "grad_norm": 0.3123298427305225,
+      "learning_rate": 4.7525070545630834e-07,
+      "loss": 0.0074,
+      "step": 20690
+    },
+    {
+      "epoch": 0.7476162958682462,
+      "grad_norm": 0.3078069916682031,
+      "learning_rate": 4.7397007124843407e-07,
+      "loss": 0.0069,
+      "step": 20700
+    },
+    {
+      "epoch": 0.7479774631609362,
+      "grad_norm": 0.24086718827052056,
+      "learning_rate": 4.7269084096820727e-07,
+      "loss": 0.007,
+      "step": 20710
+    },
+    {
+      "epoch": 0.7483386304536261,
+      "grad_norm": 0.31460408185219,
+      "learning_rate": 4.7141301636601156e-07,
+      "loss": 0.0069,
+      "step": 20720
+    },
+    {
+      "epoch": 0.7486997977463161,
+      "grad_norm": 0.25773933019458006,
+      "learning_rate": 4.7013659919030924e-07,
+      "loss": 0.0068,
+      "step": 20730
+    },
+    {
+      "epoch": 0.7490609650390061,
+      "grad_norm": 0.266823488143662,
+      "learning_rate": 4.688615911876344e-07,
+      "loss": 0.007,
+      "step": 20740
+    },
+    {
+      "epoch": 0.7494221323316961,
+      "grad_norm": 0.2886831459255116,
+      "learning_rate": 4.675879941025949e-07,
+      "loss": 0.0074,
+      "step": 20750
+    },
+    {
+      "epoch": 0.7497832996243861,
+      "grad_norm": 0.32130427120601945,
+      "learning_rate": 4.663158096778663e-07,
+      "loss": 0.0075,
+      "step": 20760
+    },
+    {
+      "epoch": 0.750144466917076,
+      "grad_norm": 0.39463481180115534,
+      "learning_rate": 4.65045039654193e-07,
+      "loss": 0.0071,
+      "step": 20770
+    },
+    {
+      "epoch": 0.750505634209766,
+      "grad_norm": 0.34191901892005433,
+      "learning_rate": 4.6377568577038203e-07,
+      "loss": 0.0079,
+      "step": 20780
+    },
+    {
+      "epoch": 0.7508668015024559,
+      "grad_norm": 0.25465883158912705,
+      "learning_rate": 4.625077497633053e-07,
+      "loss": 0.007,
+      "step": 20790
+    },
+    {
+      "epoch": 0.7512279687951459,
+      "grad_norm": 0.33143446825585154,
+      "learning_rate": 4.612412333678913e-07,
+      "loss": 0.0073,
+      "step": 20800
+    },
+    {
+      "epoch": 0.7515891360878358,
+      "grad_norm": 0.2906930083616767,
+      "learning_rate": 4.5997613831712923e-07,
+      "loss": 0.0073,
+      "step": 20810
+    },
+    {
+      "epoch": 0.7519503033805258,
+      "grad_norm": 0.28603475980161636,
+      "learning_rate": 4.587124663420611e-07,
+      "loss": 0.0068,
+      "step": 20820
+    },
+    {
+      "epoch": 0.7523114706732158,
+      "grad_norm": 0.24514029131540116,
+      "learning_rate": 4.5745021917178364e-07,
+      "loss": 0.0069,
+      "step": 20830
+    },
+    {
+      "epoch": 0.7526726379659058,
+      "grad_norm": 0.3736396962500233,
+      "learning_rate": 4.561893985334421e-07,
+      "loss": 0.0072,
+      "step": 20840
+    },
+    {
+      "epoch": 0.7530338052585958,
+      "grad_norm": 0.25957487266250107,
+      "learning_rate": 4.549300061522318e-07,
+      "loss": 0.0076,
+      "step": 20850
+    },
+    {
+      "epoch": 0.7533949725512857,
+      "grad_norm": 0.3276110633699411,
+      "learning_rate": 4.536720437513916e-07,
+      "loss": 0.0073,
+      "step": 20860
+    },
+    {
+      "epoch": 0.7537561398439757,
+      "grad_norm": 0.23958603509196402,
+      "learning_rate": 4.5241551305220645e-07,
+      "loss": 0.0073,
+      "step": 20870
+    },
+    {
+      "epoch": 0.7541173071366657,
+      "grad_norm": 0.38675913154170805,
+      "learning_rate": 4.511604157739986e-07,
+      "loss": 0.007,
+      "step": 20880
+    },
+    {
+      "epoch": 0.7544784744293557,
+      "grad_norm": 0.24950957314944372,
+      "learning_rate": 4.4990675363413234e-07,
+      "loss": 0.0067,
+      "step": 20890
+    },
+    {
+      "epoch": 0.7548396417220457,
+      "grad_norm": 0.2873390626478776,
+      "learning_rate": 4.4865452834800613e-07,
+      "loss": 0.0074,
+      "step": 20900
+    },
+    {
+      "epoch": 0.7552008090147356,
+      "grad_norm": 0.291321619433064,
+      "learning_rate": 4.4740374162905396e-07,
+      "loss": 0.0071,
+      "step": 20910
+    },
+    {
+      "epoch": 0.7555619763074256,
+      "grad_norm": 0.6687695337922507,
+      "learning_rate": 4.461543951887396e-07,
+      "loss": 0.0071,
+      "step": 20920
+    },
+    {
+      "epoch": 0.7559231436001156,
+      "grad_norm": 0.2790754089803095,
+      "learning_rate": 4.449064907365582e-07,
+      "loss": 0.0068,
+      "step": 20930
+    },
+    {
+      "epoch": 0.7562843108928056,
+      "grad_norm": 0.2335775031781843,
+      "learning_rate": 4.4366002998003e-07,
+      "loss": 0.0072,
+      "step": 20940
+    },
+    {
+      "epoch": 0.7566454781854955,
+      "grad_norm": 0.29553124059671054,
+      "learning_rate": 4.424150146247005e-07,
+      "loss": 0.0075,
+      "step": 20950
+    },
+    {
+      "epoch": 0.7570066454781855,
+      "grad_norm": 0.2713069341115668,
+      "learning_rate": 4.4117144637413697e-07,
+      "loss": 0.0072,
+      "step": 20960
+    },
+    {
+      "epoch": 0.7573678127708755,
+      "grad_norm": 0.27247607091113396,
+      "learning_rate": 4.39929326929928e-07,
+      "loss": 0.0069,
+      "step": 20970
+    },
+    {
+      "epoch": 0.7577289800635655,
+      "grad_norm": 0.206258432460031,
+      "learning_rate": 4.38688657991678e-07,
+      "loss": 0.0072,
+      "step": 20980
+    },
+    {
+      "epoch": 0.7580901473562555,
+      "grad_norm": 0.5088560934175865,
+      "learning_rate": 4.3744944125700825e-07,
+      "loss": 0.0075,
+      "step": 20990
+    },
+    {
+      "epoch": 0.7584513146489454,
+      "grad_norm": 0.3673744507609502,
+      "learning_rate": 4.3621167842155126e-07,
+      "loss": 0.0073,
+      "step": 21000
+    },
+    {
+      "epoch": 0.7584513146489454,
+      "eval_loss": 0.2710862457752228,
+      "eval_runtime": 38.4824,
+      "eval_samples_per_second": 106.438,
+      "eval_steps_per_second": 0.208,
+      "step": 21000
+    },
+    {
+      "epoch": 0.7588124819416354,
+      "grad_norm": 0.22281386452205115,
+      "learning_rate": 4.3497537117895216e-07,
+      "loss": 0.0069,
+      "step": 21010
+    },
+    {
+      "epoch": 0.7591736492343253,
+      "grad_norm": 0.428508437785164,
+      "learning_rate": 4.3374052122086247e-07,
+      "loss": 0.0072,
+      "step": 21020
+    },
+    {
+      "epoch": 0.7595348165270153,
+      "grad_norm": 0.29148738812303504,
+      "learning_rate": 4.325071302369404e-07,
+      "loss": 0.007,
+      "step": 21030
+    },
+    {
+      "epoch": 0.7598959838197052,
+      "grad_norm": 0.2519695551627036,
+      "learning_rate": 4.312751999148485e-07,
+      "loss": 0.0075,
+      "step": 21040
+    },
+    {
+      "epoch": 0.7602571511123952,
+      "grad_norm": 0.3210347468486593,
+      "learning_rate": 4.300447319402495e-07,
+      "loss": 0.0074,
+      "step": 21050
+    },
+    {
+      "epoch": 0.7606183184050852,
+      "grad_norm": 0.2992751636123447,
+      "learning_rate": 4.2881572799680654e-07,
+      "loss": 0.0071,
+      "step": 21060
+    },
+    {
+      "epoch": 0.7609794856977752,
+      "grad_norm": 0.40838987363135465,
+      "learning_rate": 4.2758818976617796e-07,
+      "loss": 0.0071,
+      "step": 21070
+    },
+    {
+      "epoch": 0.7613406529904652,
+      "grad_norm": 0.2577960760566296,
+      "learning_rate": 4.263621189280183e-07,
+      "loss": 0.0072,
+      "step": 21080
+    },
+    {
+      "epoch": 0.7617018202831551,
+      "grad_norm": 0.29821904871742916,
+      "learning_rate": 4.2513751715997265e-07,
+      "loss": 0.007,
+      "step": 21090
+    },
+    {
+      "epoch": 0.7620629875758451,
+      "grad_norm": 0.38544048909383294,
+      "learning_rate": 4.2391438613767675e-07,
+      "loss": 0.0072,
+      "step": 21100
+    },
+    {
+      "epoch": 0.7624241548685351,
+      "grad_norm": 0.2628244887464281,
+      "learning_rate": 4.226927275347534e-07,
+      "loss": 0.0075,
+      "step": 21110
+    },
+    {
+      "epoch": 0.7627853221612251,
+      "grad_norm": 0.2843151097271079,
+      "learning_rate": 4.214725430228117e-07,
+      "loss": 0.0075,
+      "step": 21120
+    },
+    {
+      "epoch": 0.7631464894539151,
+      "grad_norm": 0.2938310917871188,
+      "learning_rate": 4.2025383427144235e-07,
+      "loss": 0.0072,
+      "step": 21130
+    },
+    {
+      "epoch": 0.763507656746605,
+      "grad_norm": 0.2588528879409229,
+      "learning_rate": 4.1903660294821837e-07,
+      "loss": 0.0073,
+      "step": 21140
+    },
+    {
+      "epoch": 0.763868824039295,
+      "grad_norm": 0.26836260588203636,
+      "learning_rate": 4.1782085071868943e-07,
+      "loss": 0.007,
+      "step": 21150
+    },
+    {
+      "epoch": 0.764229991331985,
+      "grad_norm": 0.2717999069157002,
+      "learning_rate": 4.1660657924638293e-07,
+      "loss": 0.007,
+      "step": 21160
+    },
+    {
+      "epoch": 0.764591158624675,
+      "grad_norm": 0.3275711383409983,
+      "learning_rate": 4.153937901927988e-07,
+      "loss": 0.0072,
+      "step": 21170
+    },
+    {
+      "epoch": 0.764952325917365,
+      "grad_norm": 0.2646074957975533,
+      "learning_rate": 4.1418248521741026e-07,
+      "loss": 0.0073,
+      "step": 21180
+    },
+    {
+      "epoch": 0.7653134932100549,
+      "grad_norm": 0.24989224948363664,
+      "learning_rate": 4.129726659776571e-07,
+      "loss": 0.0066,
+      "step": 21190
+    },
+    {
+      "epoch": 0.7656746605027449,
+      "grad_norm": 0.27998485987311494,
+      "learning_rate": 4.1176433412894925e-07,
+      "loss": 0.0072,
+      "step": 21200
+    },
+    {
+      "epoch": 0.7660358277954349,
+      "grad_norm": 0.3333316272011483,
+      "learning_rate": 4.1055749132465904e-07,
+      "loss": 0.0065,
+      "step": 21210
+    },
+    {
+      "epoch": 0.7663969950881249,
+      "grad_norm": 0.3276695714467485,
+      "learning_rate": 4.093521392161233e-07,
+      "loss": 0.007,
+      "step": 21220
+    },
+    {
+      "epoch": 0.7667581623808148,
+      "grad_norm": 0.2636928652816062,
+      "learning_rate": 4.0814827945263695e-07,
+      "loss": 0.0074,
+      "step": 21230
+    },
+    {
+      "epoch": 0.7671193296735047,
+      "grad_norm": 0.4253967831525695,
+      "learning_rate": 4.0694591368145513e-07,
+      "loss": 0.0075,
+      "step": 21240
+    },
+    {
+      "epoch": 0.7674804969661947,
+      "grad_norm": 0.2668628639353401,
+      "learning_rate": 4.0574504354778696e-07,
+      "loss": 0.0071,
+      "step": 21250
+    },
+    {
+      "epoch": 0.7678416642588847,
+      "grad_norm": 0.2899596283498323,
+      "learning_rate": 4.045456706947969e-07,
+      "loss": 0.0073,
+      "step": 21260
+    },
+    {
+      "epoch": 0.7682028315515746,
+      "grad_norm": 0.2887122292500572,
+      "learning_rate": 4.033477967635979e-07,
+      "loss": 0.0073,
+      "step": 21270
+    },
+    {
+      "epoch": 0.7685639988442646,
+      "grad_norm": 0.3365965747129582,
+      "learning_rate": 4.0215142339325496e-07,
+      "loss": 0.007,
+      "step": 21280
+    },
+    {
+      "epoch": 0.7689251661369546,
+      "grad_norm": 0.24672179240166717,
+      "learning_rate": 4.0095655222077744e-07,
+      "loss": 0.0068,
+      "step": 21290
+    },
+    {
+      "epoch": 0.7692863334296446,
+      "grad_norm": 0.35868371614886124,
+      "learning_rate": 3.997631848811212e-07,
+      "loss": 0.0071,
+      "step": 21300
+    },
+    {
+      "epoch": 0.7696475007223346,
+      "grad_norm": 0.2708407893881428,
+      "learning_rate": 3.9857132300718266e-07,
+      "loss": 0.0074,
+      "step": 21310
+    },
+    {
+      "epoch": 0.7700086680150245,
+      "grad_norm": 0.3109436482268745,
+      "learning_rate": 3.9738096822979977e-07,
+      "loss": 0.0077,
+      "step": 21320
+    },
+    {
+      "epoch": 0.7703698353077145,
+      "grad_norm": 0.2947068317882226,
+      "learning_rate": 3.961921221777468e-07,
+      "loss": 0.007,
+      "step": 21330
+    },
+    {
+      "epoch": 0.7707310026004045,
+      "grad_norm": 0.33037652759445835,
+      "learning_rate": 3.95004786477736e-07,
+      "loss": 0.0073,
+      "step": 21340
+    },
+    {
+      "epoch": 0.7710921698930945,
+      "grad_norm": 0.277822980944687,
+      "learning_rate": 3.9381896275440916e-07,
+      "loss": 0.0068,
+      "step": 21350
+    },
+    {
+      "epoch": 0.7714533371857845,
+      "grad_norm": 0.2966251359332545,
+      "learning_rate": 3.9263465263034326e-07,
+      "loss": 0.0069,
+      "step": 21360
+    },
+    {
+      "epoch": 0.7718145044784744,
+      "grad_norm": 0.313115832588454,
+      "learning_rate": 3.9145185772604137e-07,
+      "loss": 0.0068,
+      "step": 21370
+    },
+    {
+      "epoch": 0.7721756717711644,
+      "grad_norm": 0.27704959999197054,
+      "learning_rate": 3.90270579659935e-07,
+      "loss": 0.0069,
+      "step": 21380
+    },
+    {
+      "epoch": 0.7725368390638544,
+      "grad_norm": 0.3069570672922894,
+      "learning_rate": 3.8909082004837893e-07,
+      "loss": 0.0073,
+      "step": 21390
+    },
+    {
+      "epoch": 0.7728980063565444,
+      "grad_norm": 1.2215609433264947,
+      "learning_rate": 3.879125805056515e-07,
+      "loss": 0.0068,
+      "step": 21400
+    },
+    {
+      "epoch": 0.7732591736492344,
+      "grad_norm": 0.2439087316397628,
+      "learning_rate": 3.8673586264394987e-07,
+      "loss": 0.0072,
+      "step": 21410
+    },
+    {
+      "epoch": 0.7736203409419243,
+      "grad_norm": 0.24726097380738632,
+      "learning_rate": 3.8556066807338917e-07,
+      "loss": 0.0077,
+      "step": 21420
+    },
+    {
+      "epoch": 0.7739815082346143,
+      "grad_norm": 0.28438554099156765,
+      "learning_rate": 3.843869984020016e-07,
+      "loss": 0.0069,
+      "step": 21430
+    },
+    {
+      "epoch": 0.7743426755273043,
+      "grad_norm": 0.3026309805676415,
+      "learning_rate": 3.8321485523573095e-07,
+      "loss": 0.0066,
+      "step": 21440
+    },
+    {
+      "epoch": 0.7747038428199943,
+      "grad_norm": 0.39075895421548246,
+      "learning_rate": 3.8204424017843387e-07,
+      "loss": 0.0071,
+      "step": 21450
+    },
+    {
+      "epoch": 0.7750650101126842,
+      "grad_norm": 0.3913799004484639,
+      "learning_rate": 3.8087515483187466e-07,
+      "loss": 0.0069,
+      "step": 21460
+    },
+    {
+      "epoch": 0.7754261774053741,
+      "grad_norm": 0.22504868980240947,
+      "learning_rate": 3.7970760079572597e-07,
+      "loss": 0.0075,
+      "step": 21470
+    },
+    {
+      "epoch": 0.7757873446980641,
+      "grad_norm": 0.30403189201193753,
+      "learning_rate": 3.785415796675636e-07,
+      "loss": 0.007,
+      "step": 21480
+    },
+    {
+      "epoch": 0.7761485119907541,
+      "grad_norm": 0.3491844547345964,
+      "learning_rate": 3.773770930428681e-07,
+      "loss": 0.0067,
+      "step": 21490
+    },
+    {
+      "epoch": 0.776509679283444,
+      "grad_norm": 0.2731634855690207,
+      "learning_rate": 3.762141425150168e-07,
+      "loss": 0.0068,
+      "step": 21500
+    },
+    {
+      "epoch": 0.776870846576134,
+      "grad_norm": 0.27824127191941006,
+      "learning_rate": 3.750527296752888e-07,
+      "loss": 0.0074,
+      "step": 21510
+    },
+    {
+      "epoch": 0.777232013868824,
+      "grad_norm": 0.4266078729719829,
+      "learning_rate": 3.738928561128566e-07,
+      "loss": 0.0067,
+      "step": 21520
+    },
+    {
+      "epoch": 0.777593181161514,
+      "grad_norm": 0.27825485002326533,
+      "learning_rate": 3.727345234147885e-07,
+      "loss": 0.0076,
+      "step": 21530
+    },
+    {
+      "epoch": 0.777954348454204,
+      "grad_norm": 0.2772490829274812,
+      "learning_rate": 3.715777331660425e-07,
+      "loss": 0.0065,
+      "step": 21540
+    },
+    {
+      "epoch": 0.778315515746894,
+      "grad_norm": 0.2733071651829989,
+      "learning_rate": 3.70422486949468e-07,
+      "loss": 0.0068,
+      "step": 21550
+    },
+    {
+      "epoch": 0.7786766830395839,
+      "grad_norm": 0.41922878520945733,
+      "learning_rate": 3.692687863457996e-07,
+      "loss": 0.0075,
+      "step": 21560
+    },
+    {
+      "epoch": 0.7790378503322739,
+      "grad_norm": 0.39401256951466423,
+      "learning_rate": 3.681166329336595e-07,
+      "loss": 0.0075,
+      "step": 21570
+    },
+    {
+      "epoch": 0.7793990176249639,
+      "grad_norm": 0.30508659853618225,
+      "learning_rate": 3.669660282895497e-07,
+      "loss": 0.0069,
+      "step": 21580
+    },
+    {
+      "epoch": 0.7797601849176539,
+      "grad_norm": 0.30398573064765844,
+      "learning_rate": 3.6581697398785615e-07,
+      "loss": 0.0072,
+      "step": 21590
+    },
+    {
+      "epoch": 0.7801213522103438,
+      "grad_norm": 0.29177647537740514,
+      "learning_rate": 3.6466947160084084e-07,
+      "loss": 0.0076,
+      "step": 21600
+    },
+    {
+      "epoch": 0.7804825195030338,
+      "grad_norm": 0.23475836200538605,
+      "learning_rate": 3.635235226986447e-07,
+      "loss": 0.0065,
+      "step": 21610
+    },
+    {
+      "epoch": 0.7808436867957238,
+      "grad_norm": 0.3506477812197178,
+      "learning_rate": 3.623791288492807e-07,
+      "loss": 0.0066,
+      "step": 21620
+    },
+    {
+      "epoch": 0.7812048540884138,
+      "grad_norm": 0.30270572029580317,
+      "learning_rate": 3.6123629161863614e-07,
+      "loss": 0.0072,
+      "step": 21630
+    },
+    {
+      "epoch": 0.7815660213811038,
+      "grad_norm": 0.3146013396740236,
+      "learning_rate": 3.6009501257046643e-07,
+      "loss": 0.0068,
+      "step": 21640
+    },
+    {
+      "epoch": 0.7819271886737937,
+      "grad_norm": 0.2681306325986204,
+      "learning_rate": 3.58955293266397e-07,
+      "loss": 0.0068,
+      "step": 21650
+    },
+    {
+      "epoch": 0.7822883559664837,
+      "grad_norm": 0.31694790807960155,
+      "learning_rate": 3.5781713526591625e-07,
+      "loss": 0.0072,
+      "step": 21660
+    },
+    {
+      "epoch": 0.7826495232591737,
+      "grad_norm": 0.5436086761085609,
+      "learning_rate": 3.56680540126379e-07,
+      "loss": 0.0076,
+      "step": 21670
+    },
+    {
+      "epoch": 0.7830106905518637,
+      "grad_norm": 0.3326324610105222,
+      "learning_rate": 3.555455094029999e-07,
+      "loss": 0.007,
+      "step": 21680
+    },
+    {
+      "epoch": 0.7833718578445535,
+      "grad_norm": 0.37246729828679537,
+      "learning_rate": 3.544120446488541e-07,
+      "loss": 0.0069,
+      "step": 21690
+    },
+    {
+      "epoch": 0.7837330251372435,
+      "grad_norm": 0.3477886344656612,
+      "learning_rate": 3.5328014741487256e-07,
+      "loss": 0.0064,
+      "step": 21700
+    },
+    {
+      "epoch": 0.7840941924299335,
+      "grad_norm": 0.24260460385679,
+      "learning_rate": 3.521498192498435e-07,
+      "loss": 0.0069,
+      "step": 21710
+    },
+    {
+      "epoch": 0.7844553597226235,
+      "grad_norm": 0.294785607763327,
+      "learning_rate": 3.510210617004059e-07,
+      "loss": 0.0066,
+      "step": 21720
+    },
+    {
+      "epoch": 0.7848165270153135,
+      "grad_norm": 0.31467893953509135,
+      "learning_rate": 3.4989387631105135e-07,
+      "loss": 0.007,
+      "step": 21730
+    },
+    {
+      "epoch": 0.7851776943080034,
+      "grad_norm": 0.42898987669834504,
+      "learning_rate": 3.487682646241195e-07,
+      "loss": 0.0067,
+      "step": 21740
+    },
+    {
+      "epoch": 0.7855388616006934,
+      "grad_norm": 0.30390685053464994,
+      "learning_rate": 3.4764422817979624e-07,
+      "loss": 0.0072,
+      "step": 21750
+    },
+    {
+      "epoch": 0.7859000288933834,
+      "grad_norm": 0.33252139311523743,
+      "learning_rate": 3.4652176851611334e-07,
+      "loss": 0.0074,
+      "step": 21760
+    },
+    {
+      "epoch": 0.7862611961860734,
+      "grad_norm": 0.4554883944041881,
+      "learning_rate": 3.4540088716894374e-07,
+      "loss": 0.0067,
+      "step": 21770
+    },
+    {
+      "epoch": 0.7866223634787634,
+      "grad_norm": 0.2868174401868619,
+      "learning_rate": 3.442815856720013e-07,
+      "loss": 0.0071,
+      "step": 21780
+    },
+    {
+      "epoch": 0.7869835307714533,
+      "grad_norm": 0.25098702461553946,
+      "learning_rate": 3.4316386555683856e-07,
+      "loss": 0.0067,
+      "step": 21790
+    },
+    {
+      "epoch": 0.7873446980641433,
+      "grad_norm": 0.3739249210802337,
+      "learning_rate": 3.420477283528436e-07,
+      "loss": 0.0068,
+      "step": 21800
+    },
+    {
+      "epoch": 0.7877058653568333,
+      "grad_norm": 0.2966578291478942,
+      "learning_rate": 3.409331755872383e-07,
+      "loss": 0.0069,
+      "step": 21810
+    },
+    {
+      "epoch": 0.7880670326495233,
+      "grad_norm": 0.29297824591233707,
+      "learning_rate": 3.3982020878507793e-07,
+      "loss": 0.0067,
+      "step": 21820
+    },
+    {
+      "epoch": 0.7884281999422132,
+      "grad_norm": 0.30393628682502244,
+      "learning_rate": 3.3870882946924607e-07,
+      "loss": 0.007,
+      "step": 21830
+    },
+    {
+      "epoch": 0.7887893672349032,
+      "grad_norm": 0.28758771194376437,
+      "learning_rate": 3.3759903916045537e-07,
+      "loss": 0.0071,
+      "step": 21840
+    },
+    {
+      "epoch": 0.7891505345275932,
+      "grad_norm": 0.2554830465083615,
+      "learning_rate": 3.364908393772432e-07,
+      "loss": 0.0063,
+      "step": 21850
+    },
+    {
+      "epoch": 0.7895117018202832,
+      "grad_norm": 0.3027644311567192,
+      "learning_rate": 3.353842316359717e-07,
+      "loss": 0.0076,
+      "step": 21860
+    },
+    {
+      "epoch": 0.7898728691129732,
+      "grad_norm": 0.2483580141785041,
+      "learning_rate": 3.3427921745082325e-07,
+      "loss": 0.0065,
+      "step": 21870
+    },
+    {
+      "epoch": 0.7902340364056631,
+      "grad_norm": 0.32044779674078033,
+      "learning_rate": 3.331757983338017e-07,
+      "loss": 0.0074,
+      "step": 21880
+    },
+    {
+      "epoch": 0.7905952036983531,
+      "grad_norm": 0.277883366708153,
+      "learning_rate": 3.320739757947255e-07,
+      "loss": 0.0075,
+      "step": 21890
+    },
+    {
+      "epoch": 0.7909563709910431,
+      "grad_norm": 0.30488717498581247,
+      "learning_rate": 3.3097375134123146e-07,
+      "loss": 0.0073,
+      "step": 21900
+    },
+    {
+      "epoch": 0.7913175382837331,
+      "grad_norm": 0.3064217882306154,
+      "learning_rate": 3.2987512647876747e-07,
+      "loss": 0.0071,
+      "step": 21910
+    },
+    {
+      "epoch": 0.791678705576423,
+      "grad_norm": 0.23994012193416353,
+      "learning_rate": 3.2877810271059437e-07,
+      "loss": 0.0068,
+      "step": 21920
+    },
+    {
+      "epoch": 0.7920398728691129,
+      "grad_norm": 0.29439742569965366,
+      "learning_rate": 3.2768268153778076e-07,
+      "loss": 0.0072,
+      "step": 21930
+    },
+    {
+      "epoch": 0.7924010401618029,
+      "grad_norm": 0.33323254573069916,
+      "learning_rate": 3.2658886445920374e-07,
+      "loss": 0.0072,
+      "step": 21940
+    },
+    {
+      "epoch": 0.7927622074544929,
+      "grad_norm": 0.2222115479420155,
+      "learning_rate": 3.254966529715439e-07,
+      "loss": 0.007,
+      "step": 21950
+    },
+    {
+      "epoch": 0.7931233747471829,
+      "grad_norm": 0.32478504487840587,
+      "learning_rate": 3.244060485692875e-07,
+      "loss": 0.007,
+      "step": 21960
+    },
+    {
+      "epoch": 0.7934845420398728,
+      "grad_norm": 0.35214762969720903,
+      "learning_rate": 3.23317052744718e-07,
+      "loss": 0.0071,
+      "step": 21970
+    },
+    {
+      "epoch": 0.7938457093325628,
+      "grad_norm": 0.31965219737739536,
+      "learning_rate": 3.222296669879214e-07,
+      "loss": 0.0073,
+      "step": 21980
+    },
+    {
+      "epoch": 0.7942068766252528,
+      "grad_norm": 0.25037379346288785,
+      "learning_rate": 3.211438927867785e-07,
+      "loss": 0.007,
+      "step": 21990
+    },
+    {
+      "epoch": 0.7945680439179428,
+      "grad_norm": 0.32021693094650056,
+      "learning_rate": 3.2005973162696616e-07,
+      "loss": 0.0069,
+      "step": 22000
+    },
+    {
+      "epoch": 0.7945680439179428,
+      "eval_loss": 0.2724878489971161,
+      "eval_runtime": 40.0919,
+      "eval_samples_per_second": 102.165,
+      "eval_steps_per_second": 0.2,
+      "step": 22000
+    },
+    {
+      "epoch": 0.7949292112106328,
+      "grad_norm": 0.23321399058648684,
+      "learning_rate": 3.1897718499195305e-07,
+      "loss": 0.0064,
+      "step": 22010
+    },
+    {
+      "epoch": 0.7952903785033227,
+      "grad_norm": 0.3572749956047678,
+      "learning_rate": 3.178962543629999e-07,
+      "loss": 0.0071,
+      "step": 22020
+    },
+    {
+      "epoch": 0.7956515457960127,
+      "grad_norm": 0.25038336361556274,
+      "learning_rate": 3.1681694121915476e-07,
+      "loss": 0.0067,
+      "step": 22030
+    },
+    {
+      "epoch": 0.7960127130887027,
+      "grad_norm": 0.28008251742896084,
+      "learning_rate": 3.1573924703725464e-07,
+      "loss": 0.0064,
+      "step": 22040
+    },
+    {
+      "epoch": 0.7963738803813927,
+      "grad_norm": 0.27643553922817193,
+      "learning_rate": 3.1466317329191827e-07,
+      "loss": 0.007,
+      "step": 22050
+    },
+    {
+      "epoch": 0.7967350476740827,
+      "grad_norm": 0.3120795907507818,
+      "learning_rate": 3.1358872145554973e-07,
+      "loss": 0.0072,
+      "step": 22060
+    },
+    {
+      "epoch": 0.7970962149667726,
+      "grad_norm": 0.3392090339963526,
+      "learning_rate": 3.1251589299833243e-07,
+      "loss": 0.0065,
+      "step": 22070
+    },
+    {
+      "epoch": 0.7974573822594626,
+      "grad_norm": 0.29935413994278204,
+      "learning_rate": 3.114446893882295e-07,
+      "loss": 0.0067,
+      "step": 22080
+    },
+    {
+      "epoch": 0.7978185495521526,
+      "grad_norm": 0.24082459956960195,
+      "learning_rate": 3.103751120909796e-07,
+      "loss": 0.0065,
+      "step": 22090
+    },
+    {
+      "epoch": 0.7981797168448426,
+      "grad_norm": 0.3339178757558019,
+      "learning_rate": 3.0930716257009726e-07,
+      "loss": 0.0066,
+      "step": 22100
+    },
+    {
+      "epoch": 0.7985408841375325,
+      "grad_norm": 0.5487033847253184,
+      "learning_rate": 3.0824084228686856e-07,
+      "loss": 0.0066,
+      "step": 22110
+    },
+    {
+      "epoch": 0.7989020514302225,
+      "grad_norm": 0.29466545767835944,
+      "learning_rate": 3.0717615270035166e-07,
+      "loss": 0.0065,
+      "step": 22120
+    },
+    {
+      "epoch": 0.7992632187229125,
+      "grad_norm": 0.2650759191789767,
+      "learning_rate": 3.061130952673721e-07,
+      "loss": 0.0068,
+      "step": 22130
+    },
+    {
+      "epoch": 0.7996243860156024,
+      "grad_norm": 0.2884563106167514,
+      "learning_rate": 3.0505167144252243e-07,
+      "loss": 0.0068,
+      "step": 22140
+    },
+    {
+      "epoch": 0.7999855533082924,
+      "grad_norm": 0.31186652213157534,
+      "learning_rate": 3.0399188267816063e-07,
+      "loss": 0.007,
+      "step": 22150
+    },
+    {
+      "epoch": 0.8003467206009823,
+      "grad_norm": 0.2510994692127295,
+      "learning_rate": 3.0293373042440655e-07,
+      "loss": 0.0068,
+      "step": 22160
+    },
+    {
+      "epoch": 0.8007078878936723,
+      "grad_norm": 0.3172005932684921,
+      "learning_rate": 3.018772161291416e-07,
+      "loss": 0.0066,
+      "step": 22170
+    },
+    {
+      "epoch": 0.8010690551863623,
+      "grad_norm": 0.33786832740530937,
+      "learning_rate": 3.008223412380053e-07,
+      "loss": 0.0075,
+      "step": 22180
+    },
+    {
+      "epoch": 0.8014302224790523,
+      "grad_norm": 0.27152799960348495,
+      "learning_rate": 2.9976910719439367e-07,
+      "loss": 0.007,
+      "step": 22190
+    },
+    {
+      "epoch": 0.8017913897717422,
+      "grad_norm": 0.28983004871052426,
+      "learning_rate": 2.987175154394589e-07,
+      "loss": 0.0067,
+      "step": 22200
+    },
+    {
+      "epoch": 0.8021525570644322,
+      "grad_norm": 0.30233551820635224,
+      "learning_rate": 2.976675674121047e-07,
+      "loss": 0.0074,
+      "step": 22210
+    },
+    {
+      "epoch": 0.8025137243571222,
+      "grad_norm": 0.24333533037556657,
+      "learning_rate": 2.966192645489857e-07,
+      "loss": 0.0065,
+      "step": 22220
+    },
+    {
+      "epoch": 0.8028748916498122,
+      "grad_norm": 0.2802342990690667,
+      "learning_rate": 2.9557260828450675e-07,
+      "loss": 0.0072,
+      "step": 22230
+    },
+    {
+      "epoch": 0.8032360589425022,
+      "grad_norm": 0.38687768075896806,
+      "learning_rate": 2.9452760005081785e-07,
+      "loss": 0.0069,
+      "step": 22240
+    },
+    {
+      "epoch": 0.8035972262351921,
+      "grad_norm": 0.23301943106776066,
+      "learning_rate": 2.934842412778156e-07,
+      "loss": 0.0068,
+      "step": 22250
+    },
+    {
+      "epoch": 0.8039583935278821,
+      "grad_norm": 0.27327139823916746,
+      "learning_rate": 2.924425333931383e-07,
+      "loss": 0.0065,
+      "step": 22260
+    },
+    {
+      "epoch": 0.8043195608205721,
+      "grad_norm": 0.277517171008128,
+      "learning_rate": 2.91402477822167e-07,
+      "loss": 0.0065,
+      "step": 22270
+    },
+    {
+      "epoch": 0.8046807281132621,
+      "grad_norm": 0.30970970965936484,
+      "learning_rate": 2.90364075988019e-07,
+      "loss": 0.0067,
+      "step": 22280
+    },
+    {
+      "epoch": 0.8050418954059521,
+      "grad_norm": 0.2684018912409368,
+      "learning_rate": 2.893273293115519e-07,
+      "loss": 0.0069,
+      "step": 22290
+    },
+    {
+      "epoch": 0.805403062698642,
+      "grad_norm": 0.23599620773055094,
+      "learning_rate": 2.882922392113565e-07,
+      "loss": 0.007,
+      "step": 22300
+    },
+    {
+      "epoch": 0.805764229991332,
+      "grad_norm": 0.22632316757405588,
+      "learning_rate": 2.872588071037583e-07,
+      "loss": 0.0066,
+      "step": 22310
+    },
+    {
+      "epoch": 0.806125397284022,
+      "grad_norm": 0.23620547997421587,
+      "learning_rate": 2.8622703440281253e-07,
+      "loss": 0.0071,
+      "step": 22320
+    },
+    {
+      "epoch": 0.806486564576712,
+      "grad_norm": 0.27152282144074175,
+      "learning_rate": 2.851969225203055e-07,
+      "loss": 0.0065,
+      "step": 22330
+    },
+    {
+      "epoch": 0.806847731869402,
+      "grad_norm": 0.2862203516201915,
+      "learning_rate": 2.841684728657498e-07,
+      "loss": 0.0068,
+      "step": 22340
+    },
+    {
+      "epoch": 0.8072088991620919,
+      "grad_norm": 0.26493018392084,
+      "learning_rate": 2.8314168684638466e-07,
+      "loss": 0.0066,
+      "step": 22350
+    },
+    {
+      "epoch": 0.8075700664547819,
+      "grad_norm": 0.21944473521319324,
+      "learning_rate": 2.821165658671713e-07,
+      "loss": 0.0069,
+      "step": 22360
+    },
+    {
+      "epoch": 0.8079312337474718,
+      "grad_norm": 0.2221849673666857,
+      "learning_rate": 2.810931113307943e-07,
+      "loss": 0.0064,
+      "step": 22370
+    },
+    {
+      "epoch": 0.8082924010401618,
+      "grad_norm": 0.28206113228950797,
+      "learning_rate": 2.800713246376568e-07,
+      "loss": 0.0065,
+      "step": 22380
+    },
+    {
+      "epoch": 0.8086535683328517,
+      "grad_norm": 0.25429788711839774,
+      "learning_rate": 2.7905120718588137e-07,
+      "loss": 0.0066,
+      "step": 22390
+    },
+    {
+      "epoch": 0.8090147356255417,
+      "grad_norm": 0.25949840399190227,
+      "learning_rate": 2.780327603713044e-07,
+      "loss": 0.0074,
+      "step": 22400
+    },
+    {
+      "epoch": 0.8093759029182317,
+      "grad_norm": 0.24807708305265014,
+      "learning_rate": 2.770159855874785e-07,
+      "loss": 0.0065,
+      "step": 22410
+    },
+    {
+      "epoch": 0.8097370702109217,
+      "grad_norm": 0.3024548538730322,
+      "learning_rate": 2.760008842256666e-07,
+      "loss": 0.0071,
+      "step": 22420
+    },
+    {
+      "epoch": 0.8100982375036117,
+      "grad_norm": 0.28555775061101185,
+      "learning_rate": 2.749874576748434e-07,
+      "loss": 0.0065,
+      "step": 22430
+    },
+    {
+      "epoch": 0.8104594047963016,
+      "grad_norm": 0.24502759915627004,
+      "learning_rate": 2.7397570732169084e-07,
+      "loss": 0.0071,
+      "step": 22440
+    },
+    {
+      "epoch": 0.8108205720889916,
+      "grad_norm": 0.19773322333206975,
+      "learning_rate": 2.7296563455059776e-07,
+      "loss": 0.0067,
+      "step": 22450
+    },
+    {
+      "epoch": 0.8111817393816816,
+      "grad_norm": 0.3127152203147248,
+      "learning_rate": 2.7195724074365687e-07,
+      "loss": 0.0071,
+      "step": 22460
+    },
+    {
+      "epoch": 0.8115429066743716,
+      "grad_norm": 0.2673368779901844,
+      "learning_rate": 2.709505272806651e-07,
+      "loss": 0.0066,
+      "step": 22470
+    },
+    {
+      "epoch": 0.8119040739670615,
+      "grad_norm": 0.41908754698320777,
+      "learning_rate": 2.699454955391181e-07,
+      "loss": 0.0068,
+      "step": 22480
+    },
+    {
+      "epoch": 0.8122652412597515,
+      "grad_norm": 0.3284969390297394,
+      "learning_rate": 2.689421468942127e-07,
+      "loss": 0.0065,
+      "step": 22490
+    },
+    {
+      "epoch": 0.8126264085524415,
+      "grad_norm": 0.23937622987793075,
+      "learning_rate": 2.679404827188405e-07,
+      "loss": 0.0065,
+      "step": 22500
+    },
+    {
+      "epoch": 0.8129875758451315,
+      "grad_norm": 0.346423704702232,
+      "learning_rate": 2.669405043835899e-07,
+      "loss": 0.007,
+      "step": 22510
+    },
+    {
+      "epoch": 0.8133487431378215,
+      "grad_norm": 0.34683582145817016,
+      "learning_rate": 2.659422132567415e-07,
+      "loss": 0.0068,
+      "step": 22520
+    },
+    {
+      "epoch": 0.8137099104305114,
+      "grad_norm": 0.3675797184277638,
+      "learning_rate": 2.649456107042674e-07,
+      "loss": 0.0067,
+      "step": 22530
+    },
+    {
+      "epoch": 0.8140710777232014,
+      "grad_norm": 0.3069585863307331,
+      "learning_rate": 2.639506980898302e-07,
+      "loss": 0.0066,
+      "step": 22540
+    },
+    {
+      "epoch": 0.8144322450158914,
+      "grad_norm": 0.30780594495775965,
+      "learning_rate": 2.629574767747787e-07,
+      "loss": 0.007,
+      "step": 22550
+    },
+    {
+      "epoch": 0.8147934123085814,
+      "grad_norm": 0.28492462305290256,
+      "learning_rate": 2.6196594811814874e-07,
+      "loss": 0.0065,
+      "step": 22560
+    },
+    {
+      "epoch": 0.8151545796012714,
+      "grad_norm": 0.35937690073323236,
+      "learning_rate": 2.6097611347665877e-07,
+      "loss": 0.0064,
+      "step": 22570
+    },
+    {
+      "epoch": 0.8155157468939613,
+      "grad_norm": 0.26934464791395596,
+      "learning_rate": 2.599879742047111e-07,
+      "loss": 0.0068,
+      "step": 22580
+    },
+    {
+      "epoch": 0.8158769141866512,
+      "grad_norm": 0.3246003268955504,
+      "learning_rate": 2.5900153165438645e-07,
+      "loss": 0.0071,
+      "step": 22590
+    },
+    {
+      "epoch": 0.8162380814793412,
+      "grad_norm": 0.25865558919907644,
+      "learning_rate": 2.580167871754448e-07,
+      "loss": 0.0062,
+      "step": 22600
+    },
+    {
+      "epoch": 0.8165992487720312,
+      "grad_norm": 0.3296432771711521,
+      "learning_rate": 2.57033742115322e-07,
+      "loss": 0.0067,
+      "step": 22610
+    },
+    {
+      "epoch": 0.8169604160647211,
+      "grad_norm": 0.30349612984523633,
+      "learning_rate": 2.5605239781912986e-07,
+      "loss": 0.0073,
+      "step": 22620
+    },
+    {
+      "epoch": 0.8173215833574111,
+      "grad_norm": 0.615234660083823,
+      "learning_rate": 2.5507275562965164e-07,
+      "loss": 0.0075,
+      "step": 22630
+    },
+    {
+      "epoch": 0.8176827506501011,
+      "grad_norm": 0.37828336547520613,
+      "learning_rate": 2.540948168873426e-07,
+      "loss": 0.0064,
+      "step": 22640
+    },
+    {
+      "epoch": 0.8180439179427911,
+      "grad_norm": 0.23235952598571838,
+      "learning_rate": 2.531185829303261e-07,
+      "loss": 0.0064,
+      "step": 22650
+    },
+    {
+      "epoch": 0.8184050852354811,
+      "grad_norm": 0.29312033940276794,
+      "learning_rate": 2.521440550943943e-07,
+      "loss": 0.0072,
+      "step": 22660
+    },
+    {
+      "epoch": 0.818766252528171,
+      "grad_norm": 0.26836623879018623,
+      "learning_rate": 2.511712347130033e-07,
+      "loss": 0.0064,
+      "step": 22670
+    },
+    {
+      "epoch": 0.819127419820861,
+      "grad_norm": 0.4365326198915969,
+      "learning_rate": 2.502001231172738e-07,
+      "loss": 0.0066,
+      "step": 22680
+    },
+    {
+      "epoch": 0.819488587113551,
+      "grad_norm": 0.21569705795278035,
+      "learning_rate": 2.492307216359877e-07,
+      "loss": 0.0064,
+      "step": 22690
+    },
+    {
+      "epoch": 0.819849754406241,
+      "grad_norm": 0.37243115819247297,
+      "learning_rate": 2.4826303159558795e-07,
+      "loss": 0.0065,
+      "step": 22700
+    },
+    {
+      "epoch": 0.820210921698931,
+      "grad_norm": 0.26990754330324995,
+      "learning_rate": 2.4729705432017454e-07,
+      "loss": 0.0065,
+      "step": 22710
+    },
+    {
+      "epoch": 0.8205720889916209,
+      "grad_norm": 0.26957368504115986,
+      "learning_rate": 2.463327911315051e-07,
+      "loss": 0.0067,
+      "step": 22720
+    },
+    {
+      "epoch": 0.8209332562843109,
+      "grad_norm": 0.29769413620434076,
+      "learning_rate": 2.4537024334899056e-07,
+      "loss": 0.0065,
+      "step": 22730
+    },
+    {
+      "epoch": 0.8212944235770009,
+      "grad_norm": 0.24786792994095172,
+      "learning_rate": 2.4440941228969606e-07,
+      "loss": 0.0062,
+      "step": 22740
+    },
+    {
+      "epoch": 0.8216555908696909,
+      "grad_norm": 0.28737493451121954,
+      "learning_rate": 2.434502992683358e-07,
+      "loss": 0.0069,
+      "step": 22750
+    },
+    {
+      "epoch": 0.8220167581623808,
+      "grad_norm": 0.3456577865315604,
+      "learning_rate": 2.4249290559727533e-07,
+      "loss": 0.0068,
+      "step": 22760
+    },
+    {
+      "epoch": 0.8223779254550708,
+      "grad_norm": 0.28313061946705287,
+      "learning_rate": 2.415372325865258e-07,
+      "loss": 0.0069,
+      "step": 22770
+    },
+    {
+      "epoch": 0.8227390927477608,
+      "grad_norm": 0.27349083483552966,
+      "learning_rate": 2.405832815437456e-07,
+      "loss": 0.0066,
+      "step": 22780
+    },
+    {
+      "epoch": 0.8231002600404508,
+      "grad_norm": 0.2560504609306924,
+      "learning_rate": 2.39631053774235e-07,
+      "loss": 0.0066,
+      "step": 22790
+    },
+    {
+      "epoch": 0.8234614273331408,
+      "grad_norm": 0.27352584883341696,
+      "learning_rate": 2.3868055058093855e-07,
+      "loss": 0.0065,
+      "step": 22800
+    },
+    {
+      "epoch": 0.8238225946258307,
+      "grad_norm": 0.27026877740036737,
+      "learning_rate": 2.377317732644388e-07,
+      "loss": 0.0063,
+      "step": 22810
+    },
+    {
+      "epoch": 0.8241837619185206,
+      "grad_norm": 0.26653234453902575,
+      "learning_rate": 2.3678472312295868e-07,
+      "loss": 0.0068,
+      "step": 22820
+    },
+    {
+      "epoch": 0.8245449292112106,
+      "grad_norm": 0.4793431462639471,
+      "learning_rate": 2.358394014523565e-07,
+      "loss": 0.0068,
+      "step": 22830
+    },
+    {
+      "epoch": 0.8249060965039006,
+      "grad_norm": 0.33573866492160165,
+      "learning_rate": 2.3489580954612588e-07,
+      "loss": 0.0067,
+      "step": 22840
+    },
+    {
+      "epoch": 0.8252672637965905,
+      "grad_norm": 0.2839079232211131,
+      "learning_rate": 2.3395394869539388e-07,
+      "loss": 0.0063,
+      "step": 22850
+    },
+    {
+      "epoch": 0.8256284310892805,
+      "grad_norm": 0.26089066376281617,
+      "learning_rate": 2.3301382018891877e-07,
+      "loss": 0.0065,
+      "step": 22860
+    },
+    {
+      "epoch": 0.8259895983819705,
+      "grad_norm": 0.3481488476579237,
+      "learning_rate": 2.3207542531308783e-07,
+      "loss": 0.0069,
+      "step": 22870
+    },
+    {
+      "epoch": 0.8263507656746605,
+      "grad_norm": 0.2585812471799623,
+      "learning_rate": 2.3113876535191753e-07,
+      "loss": 0.0067,
+      "step": 22880
+    },
+    {
+      "epoch": 0.8267119329673505,
+      "grad_norm": 0.24069703690638303,
+      "learning_rate": 2.3020384158704878e-07,
+      "loss": 0.0065,
+      "step": 22890
+    },
+    {
+      "epoch": 0.8270731002600404,
+      "grad_norm": 0.26582505660050126,
+      "learning_rate": 2.2927065529774842e-07,
+      "loss": 0.0066,
+      "step": 22900
+    },
+    {
+      "epoch": 0.8274342675527304,
+      "grad_norm": 0.29618626163622364,
+      "learning_rate": 2.283392077609051e-07,
+      "loss": 0.0068,
+      "step": 22910
+    },
+    {
+      "epoch": 0.8277954348454204,
+      "grad_norm": 0.2539863401630128,
+      "learning_rate": 2.2740950025102765e-07,
+      "loss": 0.0071,
+      "step": 22920
+    },
+    {
+      "epoch": 0.8281566021381104,
+      "grad_norm": 0.32198090437780236,
+      "learning_rate": 2.2648153404024559e-07,
+      "loss": 0.0065,
+      "step": 22930
+    },
+    {
+      "epoch": 0.8285177694308004,
+      "grad_norm": 0.3058162488393574,
+      "learning_rate": 2.2555531039830424e-07,
+      "loss": 0.0065,
+      "step": 22940
+    },
+    {
+      "epoch": 0.8288789367234903,
+      "grad_norm": 0.26218358392030494,
+      "learning_rate": 2.24630830592566e-07,
+      "loss": 0.0062,
+      "step": 22950
+    },
+    {
+      "epoch": 0.8292401040161803,
+      "grad_norm": 0.3337114992981248,
+      "learning_rate": 2.2370809588800566e-07,
+      "loss": 0.0066,
+      "step": 22960
+    },
+    {
+      "epoch": 0.8296012713088703,
+      "grad_norm": 0.28695206135009227,
+      "learning_rate": 2.227871075472116e-07,
+      "loss": 0.007,
+      "step": 22970
+    },
+    {
+      "epoch": 0.8299624386015603,
+      "grad_norm": 0.26921201914073906,
+      "learning_rate": 2.2186786683038158e-07,
+      "loss": 0.0064,
+      "step": 22980
+    },
+    {
+      "epoch": 0.8303236058942502,
+      "grad_norm": 0.32613947609078325,
+      "learning_rate": 2.2095037499532243e-07,
+      "loss": 0.0069,
+      "step": 22990
+    },
+    {
+      "epoch": 0.8306847731869402,
+      "grad_norm": 0.2703833384803137,
+      "learning_rate": 2.2003463329744765e-07,
+      "loss": 0.0062,
+      "step": 23000
+    },
+    {
+      "epoch": 0.8306847731869402,
+      "eval_loss": 0.2786467671394348,
+      "eval_runtime": 38.444,
+      "eval_samples_per_second": 106.545,
+      "eval_steps_per_second": 0.208,
+      "step": 23000
+    },
+    {
+      "epoch": 0.8310459404796302,
+      "grad_norm": 0.28179522666445067,
+      "learning_rate": 2.1912064298977692e-07,
+      "loss": 0.0065,
+      "step": 23010
+    },
+    {
+      "epoch": 0.8314071077723202,
+      "grad_norm": 0.22111400955485203,
+      "learning_rate": 2.1820840532293245e-07,
+      "loss": 0.0065,
+      "step": 23020
+    },
+    {
+      "epoch": 0.8317682750650102,
+      "grad_norm": 0.29434888414992566,
+      "learning_rate": 2.1729792154513918e-07,
+      "loss": 0.0068,
+      "step": 23030
+    },
+    {
+      "epoch": 0.8321294423577,
+      "grad_norm": 0.3261244362708076,
+      "learning_rate": 2.1638919290222114e-07,
+      "loss": 0.0065,
+      "step": 23040
+    },
+    {
+      "epoch": 0.83249060965039,
+      "grad_norm": 0.2729699780781145,
+      "learning_rate": 2.154822206376021e-07,
+      "loss": 0.0063,
+      "step": 23050
+    },
+    {
+      "epoch": 0.83285177694308,
+      "grad_norm": 0.22644793341481853,
+      "learning_rate": 2.1457700599230145e-07,
+      "loss": 0.0068,
+      "step": 23060
+    },
+    {
+      "epoch": 0.83321294423577,
+      "grad_norm": 0.31770621811697525,
+      "learning_rate": 2.1367355020493395e-07,
+      "loss": 0.0063,
+      "step": 23070
+    },
+    {
+      "epoch": 0.83357411152846,
+      "grad_norm": 0.2572395930131568,
+      "learning_rate": 2.1277185451170755e-07,
+      "loss": 0.0063,
+      "step": 23080
+    },
+    {
+      "epoch": 0.8339352788211499,
+      "grad_norm": 0.30891383704623504,
+      "learning_rate": 2.118719201464225e-07,
+      "loss": 0.0071,
+      "step": 23090
+    },
+    {
+      "epoch": 0.8342964461138399,
+      "grad_norm": 0.2528004824884184,
+      "learning_rate": 2.1097374834046808e-07,
+      "loss": 0.0068,
+      "step": 23100
+    },
+    {
+      "epoch": 0.8346576134065299,
+      "grad_norm": 0.29696594799897175,
+      "learning_rate": 2.100773403228229e-07,
+      "loss": 0.0068,
+      "step": 23110
+    },
+    {
+      "epoch": 0.8350187806992199,
+      "grad_norm": 0.32437638809157143,
+      "learning_rate": 2.0918269732005096e-07,
+      "loss": 0.0062,
+      "step": 23120
+    },
+    {
+      "epoch": 0.8353799479919098,
+      "grad_norm": 0.345421417494377,
+      "learning_rate": 2.0828982055630273e-07,
+      "loss": 0.0069,
+      "step": 23130
+    },
+    {
+      "epoch": 0.8357411152845998,
+      "grad_norm": 0.25889144347491005,
+      "learning_rate": 2.0739871125330962e-07,
+      "loss": 0.0066,
+      "step": 23140
+    },
+    {
+      "epoch": 0.8361022825772898,
+      "grad_norm": 0.3008662036301649,
+      "learning_rate": 2.0650937063038689e-07,
+      "loss": 0.0064,
+      "step": 23150
+    },
+    {
+      "epoch": 0.8364634498699798,
+      "grad_norm": 0.37937252852380154,
+      "learning_rate": 2.0562179990442832e-07,
+      "loss": 0.0065,
+      "step": 23160
+    },
+    {
+      "epoch": 0.8368246171626698,
+      "grad_norm": 0.24529233890478075,
+      "learning_rate": 2.0473600028990675e-07,
+      "loss": 0.0065,
+      "step": 23170
+    },
+    {
+      "epoch": 0.8371857844553597,
+      "grad_norm": 0.274403591618373,
+      "learning_rate": 2.0385197299887064e-07,
+      "loss": 0.0069,
+      "step": 23180
+    },
+    {
+      "epoch": 0.8375469517480497,
+      "grad_norm": 0.22555906596215114,
+      "learning_rate": 2.0296971924094475e-07,
+      "loss": 0.0065,
+      "step": 23190
+    },
+    {
+      "epoch": 0.8379081190407397,
+      "grad_norm": 0.28679992386840064,
+      "learning_rate": 2.0208924022332527e-07,
+      "loss": 0.0069,
+      "step": 23200
+    },
+    {
+      "epoch": 0.8382692863334297,
+      "grad_norm": 0.2839163480476032,
+      "learning_rate": 2.0121053715078193e-07,
+      "loss": 0.0061,
+      "step": 23210
+    },
+    {
+      "epoch": 0.8386304536261197,
+      "grad_norm": 0.30037732353416285,
+      "learning_rate": 2.0033361122565314e-07,
+      "loss": 0.0068,
+      "step": 23220
+    },
+    {
+      "epoch": 0.8389916209188096,
+      "grad_norm": 0.24213052397574192,
+      "learning_rate": 1.9945846364784542e-07,
+      "loss": 0.007,
+      "step": 23230
+    },
+    {
+      "epoch": 0.8393527882114996,
+      "grad_norm": 0.2544048753902333,
+      "learning_rate": 1.985850956148334e-07,
+      "loss": 0.0066,
+      "step": 23240
+    },
+    {
+      "epoch": 0.8397139555041896,
+      "grad_norm": 0.2544531768197217,
+      "learning_rate": 1.9771350832165497e-07,
+      "loss": 0.0069,
+      "step": 23250
+    },
+    {
+      "epoch": 0.8400751227968796,
+      "grad_norm": 0.33111929005481683,
+      "learning_rate": 1.9684370296091297e-07,
+      "loss": 0.0062,
+      "step": 23260
+    },
+    {
+      "epoch": 0.8404362900895694,
+      "grad_norm": 0.26052035842614485,
+      "learning_rate": 1.959756807227707e-07,
+      "loss": 0.0065,
+      "step": 23270
+    },
+    {
+      "epoch": 0.8407974573822594,
+      "grad_norm": 0.255564405391993,
+      "learning_rate": 1.9510944279495296e-07,
+      "loss": 0.0062,
+      "step": 23280
+    },
+    {
+      "epoch": 0.8411586246749494,
+      "grad_norm": 0.216937085508574,
+      "learning_rate": 1.942449903627419e-07,
+      "loss": 0.0066,
+      "step": 23290
+    },
+    {
+      "epoch": 0.8415197919676394,
+      "grad_norm": 0.25770115971611063,
+      "learning_rate": 1.9338232460897687e-07,
+      "loss": 0.0065,
+      "step": 23300
+    },
+    {
+      "epoch": 0.8418809592603294,
+      "grad_norm": 0.33620770018822593,
+      "learning_rate": 1.925214467140523e-07,
+      "loss": 0.0071,
+      "step": 23310
+    },
+    {
+      "epoch": 0.8422421265530193,
+      "grad_norm": 0.2924971387915994,
+      "learning_rate": 1.916623578559173e-07,
+      "loss": 0.0065,
+      "step": 23320
+    },
+    {
+      "epoch": 0.8426032938457093,
+      "grad_norm": 0.3157968255414696,
+      "learning_rate": 1.9080505921007153e-07,
+      "loss": 0.0064,
+      "step": 23330
+    },
+    {
+      "epoch": 0.8429644611383993,
+      "grad_norm": 0.2505243568402366,
+      "learning_rate": 1.8994955194956649e-07,
+      "loss": 0.0062,
+      "step": 23340
+    },
+    {
+      "epoch": 0.8433256284310893,
+      "grad_norm": 0.2681825876361489,
+      "learning_rate": 1.890958372450013e-07,
+      "loss": 0.0064,
+      "step": 23350
+    },
+    {
+      "epoch": 0.8436867957237792,
+      "grad_norm": 0.2674955250027987,
+      "learning_rate": 1.8824391626452337e-07,
+      "loss": 0.0065,
+      "step": 23360
+    },
+    {
+      "epoch": 0.8440479630164692,
+      "grad_norm": 0.24674418983239937,
+      "learning_rate": 1.8739379017382508e-07,
+      "loss": 0.0062,
+      "step": 23370
+    },
+    {
+      "epoch": 0.8444091303091592,
+      "grad_norm": 0.24490032072403878,
+      "learning_rate": 1.8654546013614277e-07,
+      "loss": 0.0065,
+      "step": 23380
+    },
+    {
+      "epoch": 0.8447702976018492,
+      "grad_norm": 0.3228597304281314,
+      "learning_rate": 1.8569892731225523e-07,
+      "loss": 0.0064,
+      "step": 23390
+    },
+    {
+      "epoch": 0.8451314648945392,
+      "grad_norm": 0.3503044347810507,
+      "learning_rate": 1.8485419286048293e-07,
+      "loss": 0.0071,
+      "step": 23400
+    },
+    {
+      "epoch": 0.8454926321872291,
+      "grad_norm": 0.2974832538904266,
+      "learning_rate": 1.8401125793668445e-07,
+      "loss": 0.0063,
+      "step": 23410
+    },
+    {
+      "epoch": 0.8458537994799191,
+      "grad_norm": 0.24067236063404618,
+      "learning_rate": 1.8317012369425712e-07,
+      "loss": 0.0059,
+      "step": 23420
+    },
+    {
+      "epoch": 0.8462149667726091,
+      "grad_norm": 0.31792679504208865,
+      "learning_rate": 1.8233079128413316e-07,
+      "loss": 0.0067,
+      "step": 23430
+    },
+    {
+      "epoch": 0.8465761340652991,
+      "grad_norm": 0.25133933170369027,
+      "learning_rate": 1.8149326185478072e-07,
+      "loss": 0.0067,
+      "step": 23440
+    },
+    {
+      "epoch": 0.8469373013579891,
+      "grad_norm": 0.2313826168397375,
+      "learning_rate": 1.806575365522001e-07,
+      "loss": 0.0062,
+      "step": 23450
+    },
+    {
+      "epoch": 0.847298468650679,
+      "grad_norm": 0.29267872700149566,
+      "learning_rate": 1.7982361651992269e-07,
+      "loss": 0.0066,
+      "step": 23460
+    },
+    {
+      "epoch": 0.847659635943369,
+      "grad_norm": 0.346750050171553,
+      "learning_rate": 1.789915028990103e-07,
+      "loss": 0.0067,
+      "step": 23470
+    },
+    {
+      "epoch": 0.848020803236059,
+      "grad_norm": 0.6937751304238591,
+      "learning_rate": 1.781611968280531e-07,
+      "loss": 0.0062,
+      "step": 23480
+    },
+    {
+      "epoch": 0.8483819705287489,
+      "grad_norm": 0.47638331346380614,
+      "learning_rate": 1.7733269944316715e-07,
+      "loss": 0.0063,
+      "step": 23490
+    },
+    {
+      "epoch": 0.8487431378214388,
+      "grad_norm": 0.30750560077956024,
+      "learning_rate": 1.765060118779948e-07,
+      "loss": 0.0066,
+      "step": 23500
+    },
+    {
+      "epoch": 0.8491043051141288,
+      "grad_norm": 0.3675574559542341,
+      "learning_rate": 1.7568113526370106e-07,
+      "loss": 0.0067,
+      "step": 23510
+    },
+    {
+      "epoch": 0.8494654724068188,
+      "grad_norm": 0.2779956743491352,
+      "learning_rate": 1.7485807072897369e-07,
+      "loss": 0.0064,
+      "step": 23520
+    },
+    {
+      "epoch": 0.8498266396995088,
+      "grad_norm": 0.2781368670137158,
+      "learning_rate": 1.7403681940002054e-07,
+      "loss": 0.0061,
+      "step": 23530
+    },
+    {
+      "epoch": 0.8501878069921988,
+      "grad_norm": 0.2498805966827028,
+      "learning_rate": 1.7321738240056833e-07,
+      "loss": 0.0066,
+      "step": 23540
+    },
+    {
+      "epoch": 0.8505489742848887,
+      "grad_norm": 0.24087917442322823,
+      "learning_rate": 1.723997608518612e-07,
+      "loss": 0.0067,
+      "step": 23550
+    },
+    {
+      "epoch": 0.8509101415775787,
+      "grad_norm": 0.2957994848953683,
+      "learning_rate": 1.7158395587266007e-07,
+      "loss": 0.0061,
+      "step": 23560
+    },
+    {
+      "epoch": 0.8512713088702687,
+      "grad_norm": 0.30082964852449745,
+      "learning_rate": 1.7076996857923883e-07,
+      "loss": 0.0066,
+      "step": 23570
+    },
+    {
+      "epoch": 0.8516324761629587,
+      "grad_norm": 0.42352449717989193,
+      "learning_rate": 1.699578000853857e-07,
+      "loss": 0.0063,
+      "step": 23580
+    },
+    {
+      "epoch": 0.8519936434556487,
+      "grad_norm": 0.2902695912612159,
+      "learning_rate": 1.6914745150239858e-07,
+      "loss": 0.0064,
+      "step": 23590
+    },
+    {
+      "epoch": 0.8523548107483386,
+      "grad_norm": 0.3522307718954109,
+      "learning_rate": 1.68338923939087e-07,
+      "loss": 0.0064,
+      "step": 23600
+    },
+    {
+      "epoch": 0.8527159780410286,
+      "grad_norm": 0.30495561370541246,
+      "learning_rate": 1.675322185017672e-07,
+      "loss": 0.007,
+      "step": 23610
+    },
+    {
+      "epoch": 0.8530771453337186,
+      "grad_norm": 0.28266533259251103,
+      "learning_rate": 1.6672733629426267e-07,
+      "loss": 0.0064,
+      "step": 23620
+    },
+    {
+      "epoch": 0.8534383126264086,
+      "grad_norm": 0.2747986699926616,
+      "learning_rate": 1.6592427841790282e-07,
+      "loss": 0.0066,
+      "step": 23630
+    },
+    {
+      "epoch": 0.8537994799190985,
+      "grad_norm": 0.2559307773969627,
+      "learning_rate": 1.6512304597151963e-07,
+      "loss": 0.0066,
+      "step": 23640
+    },
+    {
+      "epoch": 0.8541606472117885,
+      "grad_norm": 0.2654342797356501,
+      "learning_rate": 1.6432364005144862e-07,
+      "loss": 0.0068,
+      "step": 23650
+    },
+    {
+      "epoch": 0.8545218145044785,
+      "grad_norm": 0.2808065053454028,
+      "learning_rate": 1.6352606175152483e-07,
+      "loss": 0.0062,
+      "step": 23660
+    },
+    {
+      "epoch": 0.8548829817971685,
+      "grad_norm": 0.28570833791653316,
+      "learning_rate": 1.6273031216308364e-07,
+      "loss": 0.0069,
+      "step": 23670
+    },
+    {
+      "epoch": 0.8552441490898585,
+      "grad_norm": 0.3793963793273943,
+      "learning_rate": 1.619363923749574e-07,
+      "loss": 0.0066,
+      "step": 23680
+    },
+    {
+      "epoch": 0.8556053163825484,
+      "grad_norm": 0.3445679896104038,
+      "learning_rate": 1.6114430347347508e-07,
+      "loss": 0.0066,
+      "step": 23690
+    },
+    {
+      "epoch": 0.8559664836752384,
+      "grad_norm": 0.4184445180129593,
+      "learning_rate": 1.6035404654245977e-07,
+      "loss": 0.0065,
+      "step": 23700
+    },
+    {
+      "epoch": 0.8563276509679284,
+      "grad_norm": 0.3722265943146979,
+      "learning_rate": 1.5956562266322933e-07,
+      "loss": 0.0066,
+      "step": 23710
+    },
+    {
+      "epoch": 0.8566888182606183,
+      "grad_norm": 0.38616811273656493,
+      "learning_rate": 1.5877903291459196e-07,
+      "loss": 0.0063,
+      "step": 23720
+    },
+    {
+      "epoch": 0.8570499855533082,
+      "grad_norm": 0.48363049365552313,
+      "learning_rate": 1.5799427837284686e-07,
+      "loss": 0.0065,
+      "step": 23730
+    },
+    {
+      "epoch": 0.8574111528459982,
+      "grad_norm": 0.4740899209092282,
+      "learning_rate": 1.5721136011178193e-07,
+      "loss": 0.0065,
+      "step": 23740
+    },
+    {
+      "epoch": 0.8577723201386882,
+      "grad_norm": 0.35416930709311517,
+      "learning_rate": 1.5643027920267295e-07,
+      "loss": 0.0065,
+      "step": 23750
+    },
+    {
+      "epoch": 0.8581334874313782,
+      "grad_norm": 0.23114522483370734,
+      "learning_rate": 1.5565103671428093e-07,
+      "loss": 0.0071,
+      "step": 23760
+    },
+    {
+      "epoch": 0.8584946547240682,
+      "grad_norm": 0.31442921453737444,
+      "learning_rate": 1.5487363371285169e-07,
+      "loss": 0.006,
+      "step": 23770
+    },
+    {
+      "epoch": 0.8588558220167581,
+      "grad_norm": 0.2763114819437523,
+      "learning_rate": 1.540980712621136e-07,
+      "loss": 0.0062,
+      "step": 23780
+    },
+    {
+      "epoch": 0.8592169893094481,
+      "grad_norm": 0.2685476556121732,
+      "learning_rate": 1.533243504232777e-07,
+      "loss": 0.0064,
+      "step": 23790
+    },
+    {
+      "epoch": 0.8595781566021381,
+      "grad_norm": 0.3750638022479481,
+      "learning_rate": 1.5255247225503373e-07,
+      "loss": 0.0068,
+      "step": 23800
+    },
+    {
+      "epoch": 0.8599393238948281,
+      "grad_norm": 0.22204444049611075,
+      "learning_rate": 1.517824378135515e-07,
+      "loss": 0.006,
+      "step": 23810
+    },
+    {
+      "epoch": 0.8603004911875181,
+      "grad_norm": 0.2873087213381393,
+      "learning_rate": 1.5101424815247655e-07,
+      "loss": 0.0063,
+      "step": 23820
+    },
+    {
+      "epoch": 0.860661658480208,
+      "grad_norm": 0.352932860015409,
+      "learning_rate": 1.5024790432293174e-07,
+      "loss": 0.0063,
+      "step": 23830
+    },
+    {
+      "epoch": 0.861022825772898,
+      "grad_norm": 0.2839729180214441,
+      "learning_rate": 1.4948340737351306e-07,
+      "loss": 0.0057,
+      "step": 23840
+    },
+    {
+      "epoch": 0.861383993065588,
+      "grad_norm": 0.24281080988104536,
+      "learning_rate": 1.487207583502898e-07,
+      "loss": 0.0064,
+      "step": 23850
+    },
+    {
+      "epoch": 0.861745160358278,
+      "grad_norm": 0.314839007955314,
+      "learning_rate": 1.4795995829680253e-07,
+      "loss": 0.0063,
+      "step": 23860
+    },
+    {
+      "epoch": 0.862106327650968,
+      "grad_norm": 0.2982099021334602,
+      "learning_rate": 1.4720100825406253e-07,
+      "loss": 0.0061,
+      "step": 23870
+    },
+    {
+      "epoch": 0.8624674949436579,
+      "grad_norm": 0.28285810402450007,
+      "learning_rate": 1.4644390926054874e-07,
+      "loss": 0.0064,
+      "step": 23880
+    },
+    {
+      "epoch": 0.8628286622363479,
+      "grad_norm": 0.23734640665007467,
+      "learning_rate": 1.4568866235220802e-07,
+      "loss": 0.0063,
+      "step": 23890
+    },
+    {
+      "epoch": 0.8631898295290379,
+      "grad_norm": 0.2514902460178547,
+      "learning_rate": 1.4493526856245255e-07,
+      "loss": 0.0064,
+      "step": 23900
+    },
+    {
+      "epoch": 0.8635509968217279,
+      "grad_norm": 0.27725546307178894,
+      "learning_rate": 1.4418372892215935e-07,
+      "loss": 0.0065,
+      "step": 23910
+    },
+    {
+      "epoch": 0.8639121641144178,
+      "grad_norm": 0.2530073679667116,
+      "learning_rate": 1.4343404445966806e-07,
+      "loss": 0.0063,
+      "step": 23920
+    },
+    {
+      "epoch": 0.8642733314071078,
+      "grad_norm": 0.27471199874740204,
+      "learning_rate": 1.4268621620077942e-07,
+      "loss": 0.0059,
+      "step": 23930
+    },
+    {
+      "epoch": 0.8646344986997977,
+      "grad_norm": 0.29017533625121056,
+      "learning_rate": 1.4194024516875525e-07,
+      "loss": 0.0063,
+      "step": 23940
+    },
+    {
+      "epoch": 0.8649956659924877,
+      "grad_norm": 0.2874929139245117,
+      "learning_rate": 1.4119613238431522e-07,
+      "loss": 0.0068,
+      "step": 23950
+    },
+    {
+      "epoch": 0.8653568332851777,
+      "grad_norm": 0.3010906844304847,
+      "learning_rate": 1.4045387886563737e-07,
+      "loss": 0.0067,
+      "step": 23960
+    },
+    {
+      "epoch": 0.8657180005778676,
+      "grad_norm": 0.2023874452310962,
+      "learning_rate": 1.3971348562835478e-07,
+      "loss": 0.0067,
+      "step": 23970
+    },
+    {
+      "epoch": 0.8660791678705576,
+      "grad_norm": 0.4487727266323675,
+      "learning_rate": 1.389749536855549e-07,
+      "loss": 0.0062,
+      "step": 23980
+    },
+    {
+      "epoch": 0.8664403351632476,
+      "grad_norm": 0.2911393861388412,
+      "learning_rate": 1.3823828404777972e-07,
+      "loss": 0.0063,
+      "step": 23990
+    },
+    {
+      "epoch": 0.8668015024559376,
+      "grad_norm": 0.2781607683793593,
+      "learning_rate": 1.3750347772302141e-07,
+      "loss": 0.0066,
+      "step": 24000
+    },
+    {
+      "epoch": 0.8668015024559376,
+      "eval_loss": 0.2830405533313751,
+      "eval_runtime": 39.0281,
+      "eval_samples_per_second": 104.95,
+      "eval_steps_per_second": 0.205,
+      "step": 24000
+    },
+    {
+      "epoch": 0.8671626697486275,
+      "grad_norm": 0.2878688428747624,
+      "learning_rate": 1.3677053571672337e-07,
+      "loss": 0.0062,
+      "step": 24010
+    },
+    {
+      "epoch": 0.8675238370413175,
+      "grad_norm": 0.2350518203502992,
+      "learning_rate": 1.360394590317784e-07,
+      "loss": 0.0066,
+      "step": 24020
+    },
+    {
+      "epoch": 0.8678850043340075,
+      "grad_norm": 0.2619723277142182,
+      "learning_rate": 1.3531024866852577e-07,
+      "loss": 0.0062,
+      "step": 24030
+    },
+    {
+      "epoch": 0.8682461716266975,
+      "grad_norm": 0.2726905996099445,
+      "learning_rate": 1.345829056247525e-07,
+      "loss": 0.0066,
+      "step": 24040
+    },
+    {
+      "epoch": 0.8686073389193875,
+      "grad_norm": 0.33629943865551115,
+      "learning_rate": 1.338574308956892e-07,
+      "loss": 0.0061,
+      "step": 24050
+    },
+    {
+      "epoch": 0.8689685062120774,
+      "grad_norm": 0.3758951224553846,
+      "learning_rate": 1.3313382547401126e-07,
+      "loss": 0.0064,
+      "step": 24060
+    },
+    {
+      "epoch": 0.8693296735047674,
+      "grad_norm": 0.2560782228221128,
+      "learning_rate": 1.324120903498352e-07,
+      "loss": 0.0067,
+      "step": 24070
+    },
+    {
+      "epoch": 0.8696908407974574,
+      "grad_norm": 0.2994616528504239,
+      "learning_rate": 1.3169222651071887e-07,
+      "loss": 0.0062,
+      "step": 24080
+    },
+    {
+      "epoch": 0.8700520080901474,
+      "grad_norm": 0.4208314684333138,
+      "learning_rate": 1.3097423494165954e-07,
+      "loss": 0.006,
+      "step": 24090
+    },
+    {
+      "epoch": 0.8704131753828374,
+      "grad_norm": 0.2912735417785884,
+      "learning_rate": 1.302581166250932e-07,
+      "loss": 0.0068,
+      "step": 24100
+    },
+    {
+      "epoch": 0.8707743426755273,
+      "grad_norm": 0.2642238158201568,
+      "learning_rate": 1.2954387254089144e-07,
+      "loss": 0.0064,
+      "step": 24110
+    },
+    {
+      "epoch": 0.8711355099682173,
+      "grad_norm": 0.2606782593955475,
+      "learning_rate": 1.2883150366636275e-07,
+      "loss": 0.0067,
+      "step": 24120
+    },
+    {
+      "epoch": 0.8714966772609073,
+      "grad_norm": 0.2261638348532553,
+      "learning_rate": 1.2812101097624824e-07,
+      "loss": 0.0061,
+      "step": 24130
+    },
+    {
+      "epoch": 0.8718578445535973,
+      "grad_norm": 0.24396330850931278,
+      "learning_rate": 1.2741239544272353e-07,
+      "loss": 0.0063,
+      "step": 24140
+    },
+    {
+      "epoch": 0.8722190118462873,
+      "grad_norm": 0.3802817919674417,
+      "learning_rate": 1.2670565803539407e-07,
+      "loss": 0.0062,
+      "step": 24150
+    },
+    {
+      "epoch": 0.8725801791389771,
+      "grad_norm": 0.3358925390141907,
+      "learning_rate": 1.2600079972129647e-07,
+      "loss": 0.0064,
+      "step": 24160
+    },
+    {
+      "epoch": 0.8729413464316671,
+      "grad_norm": 0.25740079688023104,
+      "learning_rate": 1.2529782146489538e-07,
+      "loss": 0.0064,
+      "step": 24170
+    },
+    {
+      "epoch": 0.8733025137243571,
+      "grad_norm": 0.28593351801716554,
+      "learning_rate": 1.245967242280842e-07,
+      "loss": 0.0067,
+      "step": 24180
+    },
+    {
+      "epoch": 0.8736636810170471,
+      "grad_norm": 0.3230124231559693,
+      "learning_rate": 1.2389750897018093e-07,
+      "loss": 0.0064,
+      "step": 24190
+    },
+    {
+      "epoch": 0.874024848309737,
+      "grad_norm": 0.28590502892166025,
+      "learning_rate": 1.2320017664792988e-07,
+      "loss": 0.0066,
+      "step": 24200
+    },
+    {
+      "epoch": 0.874386015602427,
+      "grad_norm": 0.2834331128837407,
+      "learning_rate": 1.225047282154978e-07,
+      "loss": 0.0066,
+      "step": 24210
+    },
+    {
+      "epoch": 0.874747182895117,
+      "grad_norm": 0.2291410472685357,
+      "learning_rate": 1.2181116462447466e-07,
+      "loss": 0.0066,
+      "step": 24220
+    },
+    {
+      "epoch": 0.875108350187807,
+      "grad_norm": 0.2822186269343579,
+      "learning_rate": 1.2111948682387052e-07,
+      "loss": 0.0062,
+      "step": 24230
+    },
+    {
+      "epoch": 0.875469517480497,
+      "grad_norm": 0.314818158492125,
+      "learning_rate": 1.204296957601155e-07,
+      "loss": 0.0066,
+      "step": 24240
+    },
+    {
+      "epoch": 0.8758306847731869,
+      "grad_norm": 0.25412272171609784,
+      "learning_rate": 1.1974179237705786e-07,
+      "loss": 0.0065,
+      "step": 24250
+    },
+    {
+      "epoch": 0.8761918520658769,
+      "grad_norm": 0.23249337857344402,
+      "learning_rate": 1.1905577761596354e-07,
+      "loss": 0.0062,
+      "step": 24260
+    },
+    {
+      "epoch": 0.8765530193585669,
+      "grad_norm": 0.2752101833070256,
+      "learning_rate": 1.1837165241551334e-07,
+      "loss": 0.0063,
+      "step": 24270
+    },
+    {
+      "epoch": 0.8769141866512569,
+      "grad_norm": 0.33725871405814806,
+      "learning_rate": 1.1768941771180364e-07,
+      "loss": 0.0065,
+      "step": 24280
+    },
+    {
+      "epoch": 0.8772753539439468,
+      "grad_norm": 0.29021994031720244,
+      "learning_rate": 1.1700907443834269e-07,
+      "loss": 0.0063,
+      "step": 24290
+    },
+    {
+      "epoch": 0.8776365212366368,
+      "grad_norm": 0.286175563143652,
+      "learning_rate": 1.1633062352605217e-07,
+      "loss": 0.0063,
+      "step": 24300
+    },
+    {
+      "epoch": 0.8779976885293268,
+      "grad_norm": 0.30484979527486866,
+      "learning_rate": 1.1565406590326316e-07,
+      "loss": 0.0062,
+      "step": 24310
+    },
+    {
+      "epoch": 0.8783588558220168,
+      "grad_norm": 0.2715300585294683,
+      "learning_rate": 1.1497940249571665e-07,
+      "loss": 0.0063,
+      "step": 24320
+    },
+    {
+      "epoch": 0.8787200231147068,
+      "grad_norm": 0.26122238156320904,
+      "learning_rate": 1.1430663422656218e-07,
+      "loss": 0.0063,
+      "step": 24330
+    },
+    {
+      "epoch": 0.8790811904073967,
+      "grad_norm": 0.2728562830194229,
+      "learning_rate": 1.1363576201635506e-07,
+      "loss": 0.0063,
+      "step": 24340
+    },
+    {
+      "epoch": 0.8794423577000867,
+      "grad_norm": 0.3007483486810962,
+      "learning_rate": 1.129667867830575e-07,
+      "loss": 0.0065,
+      "step": 24350
+    },
+    {
+      "epoch": 0.8798035249927767,
+      "grad_norm": 0.30233718972760315,
+      "learning_rate": 1.1229970944203494e-07,
+      "loss": 0.0063,
+      "step": 24360
+    },
+    {
+      "epoch": 0.8801646922854667,
+      "grad_norm": 0.26512969476287773,
+      "learning_rate": 1.1163453090605691e-07,
+      "loss": 0.0066,
+      "step": 24370
+    },
+    {
+      "epoch": 0.8805258595781567,
+      "grad_norm": 0.27062757258577974,
+      "learning_rate": 1.1097125208529369e-07,
+      "loss": 0.0062,
+      "step": 24380
+    },
+    {
+      "epoch": 0.8808870268708465,
+      "grad_norm": 0.30013951410590833,
+      "learning_rate": 1.1030987388731711e-07,
+      "loss": 0.0061,
+      "step": 24390
+    },
+    {
+      "epoch": 0.8812481941635365,
+      "grad_norm": 0.3128903355048033,
+      "learning_rate": 1.0965039721709763e-07,
+      "loss": 0.0062,
+      "step": 24400
+    },
+    {
+      "epoch": 0.8816093614562265,
+      "grad_norm": 0.30174640702871525,
+      "learning_rate": 1.0899282297700474e-07,
+      "loss": 0.006,
+      "step": 24410
+    },
+    {
+      "epoch": 0.8819705287489165,
+      "grad_norm": 0.2350001157404734,
+      "learning_rate": 1.0833715206680355e-07,
+      "loss": 0.0061,
+      "step": 24420
+    },
+    {
+      "epoch": 0.8823316960416064,
+      "grad_norm": 0.34318312689843805,
+      "learning_rate": 1.0768338538365608e-07,
+      "loss": 0.0067,
+      "step": 24430
+    },
+    {
+      "epoch": 0.8826928633342964,
+      "grad_norm": 0.28201065328713903,
+      "learning_rate": 1.0703152382211789e-07,
+      "loss": 0.0064,
+      "step": 24440
+    },
+    {
+      "epoch": 0.8830540306269864,
+      "grad_norm": 0.4634941495273692,
+      "learning_rate": 1.0638156827413853e-07,
+      "loss": 0.0065,
+      "step": 24450
+    },
+    {
+      "epoch": 0.8834151979196764,
+      "grad_norm": 0.26526323285608777,
+      "learning_rate": 1.0573351962905875e-07,
+      "loss": 0.0067,
+      "step": 24460
+    },
+    {
+      "epoch": 0.8837763652123664,
+      "grad_norm": 0.23112014994617813,
+      "learning_rate": 1.0508737877361041e-07,
+      "loss": 0.0056,
+      "step": 24470
+    },
+    {
+      "epoch": 0.8841375325050563,
+      "grad_norm": 0.308454039517672,
+      "learning_rate": 1.0444314659191462e-07,
+      "loss": 0.0064,
+      "step": 24480
+    },
+    {
+      "epoch": 0.8844986997977463,
+      "grad_norm": 0.30519837743643713,
+      "learning_rate": 1.0380082396548174e-07,
+      "loss": 0.0069,
+      "step": 24490
+    },
+    {
+      "epoch": 0.8848598670904363,
+      "grad_norm": 0.27890812265922776,
+      "learning_rate": 1.0316041177320806e-07,
+      "loss": 0.0068,
+      "step": 24500
+    },
+    {
+      "epoch": 0.8852210343831263,
+      "grad_norm": 0.2468036834241297,
+      "learning_rate": 1.025219108913768e-07,
+      "loss": 0.006,
+      "step": 24510
+    },
+    {
+      "epoch": 0.8855822016758162,
+      "grad_norm": 0.24434243758535235,
+      "learning_rate": 1.0188532219365526e-07,
+      "loss": 0.006,
+      "step": 24520
+    },
+    {
+      "epoch": 0.8859433689685062,
+      "grad_norm": 0.28679317986018965,
+      "learning_rate": 1.0125064655109467e-07,
+      "loss": 0.0064,
+      "step": 24530
+    },
+    {
+      "epoch": 0.8863045362611962,
+      "grad_norm": 0.28667811956831135,
+      "learning_rate": 1.0061788483212853e-07,
+      "loss": 0.0063,
+      "step": 24540
+    },
+    {
+      "epoch": 0.8866657035538862,
+      "grad_norm": 0.29171903246679326,
+      "learning_rate": 9.998703790257108e-08,
+      "loss": 0.0063,
+      "step": 24550
+    },
+    {
+      "epoch": 0.8870268708465762,
+      "grad_norm": 0.24810140880344947,
+      "learning_rate": 9.935810662561684e-08,
+      "loss": 0.0062,
+      "step": 24560
+    },
+    {
+      "epoch": 0.8873880381392661,
+      "grad_norm": 0.38513673461075226,
+      "learning_rate": 9.873109186183976e-08,
+      "loss": 0.0062,
+      "step": 24570
+    },
+    {
+      "epoch": 0.8877492054319561,
+      "grad_norm": 0.3742216525511732,
+      "learning_rate": 9.810599446919005e-08,
+      "loss": 0.0065,
+      "step": 24580
+    },
+    {
+      "epoch": 0.8881103727246461,
+      "grad_norm": 0.27969806907437683,
+      "learning_rate": 9.748281530299569e-08,
+      "loss": 0.0061,
+      "step": 24590
+    },
+    {
+      "epoch": 0.8884715400173361,
+      "grad_norm": 0.2642047568389191,
+      "learning_rate": 9.686155521595907e-08,
+      "loss": 0.0061,
+      "step": 24600
+    },
+    {
+      "epoch": 0.888832707310026,
+      "grad_norm": 0.3288173320539464,
+      "learning_rate": 9.624221505815705e-08,
+      "loss": 0.0061,
+      "step": 24610
+    },
+    {
+      "epoch": 0.8891938746027159,
+      "grad_norm": 0.2905514303611689,
+      "learning_rate": 9.562479567703941e-08,
+      "loss": 0.0065,
+      "step": 24620
+    },
+    {
+      "epoch": 0.8895550418954059,
+      "grad_norm": 0.24453027154756984,
+      "learning_rate": 9.500929791742735e-08,
+      "loss": 0.0067,
+      "step": 24630
+    },
+    {
+      "epoch": 0.8899162091880959,
+      "grad_norm": 0.25543605968999433,
+      "learning_rate": 9.439572262151335e-08,
+      "loss": 0.0065,
+      "step": 24640
+    },
+    {
+      "epoch": 0.8902773764807859,
+      "grad_norm": 0.3113966978894074,
+      "learning_rate": 9.378407062885902e-08,
+      "loss": 0.0065,
+      "step": 24650
+    },
+    {
+      "epoch": 0.8906385437734758,
+      "grad_norm": 0.36238923052803645,
+      "learning_rate": 9.317434277639369e-08,
+      "loss": 0.0062,
+      "step": 24660
+    },
+    {
+      "epoch": 0.8909997110661658,
+      "grad_norm": 0.32669050253103155,
+      "learning_rate": 9.256653989841518e-08,
+      "loss": 0.0061,
+      "step": 24670
+    },
+    {
+      "epoch": 0.8913608783588558,
+      "grad_norm": 0.36049944806919554,
+      "learning_rate": 9.196066282658605e-08,
+      "loss": 0.0068,
+      "step": 24680
+    },
+    {
+      "epoch": 0.8917220456515458,
+      "grad_norm": 0.2886312041335618,
+      "learning_rate": 9.135671238993498e-08,
+      "loss": 0.0057,
+      "step": 24690
+    },
+    {
+      "epoch": 0.8920832129442358,
+      "grad_norm": 0.25267746441470645,
+      "learning_rate": 9.075468941485326e-08,
+      "loss": 0.0062,
+      "step": 24700
+    },
+    {
+      "epoch": 0.8924443802369257,
+      "grad_norm": 0.21894843699561678,
+      "learning_rate": 9.015459472509513e-08,
+      "loss": 0.0062,
+      "step": 24710
+    },
+    {
+      "epoch": 0.8928055475296157,
+      "grad_norm": 0.2522504719444055,
+      "learning_rate": 8.955642914177725e-08,
+      "loss": 0.0064,
+      "step": 24720
+    },
+    {
+      "epoch": 0.8931667148223057,
+      "grad_norm": 0.2595295517962483,
+      "learning_rate": 8.896019348337526e-08,
+      "loss": 0.0062,
+      "step": 24730
+    },
+    {
+      "epoch": 0.8935278821149957,
+      "grad_norm": 0.3254737114177489,
+      "learning_rate": 8.83658885657252e-08,
+      "loss": 0.0061,
+      "step": 24740
+    },
+    {
+      "epoch": 0.8938890494076857,
+      "grad_norm": 0.35213627747254966,
+      "learning_rate": 8.777351520202026e-08,
+      "loss": 0.0065,
+      "step": 24750
+    },
+    {
+      "epoch": 0.8942502167003756,
+      "grad_norm": 0.30303255026025583,
+      "learning_rate": 8.718307420281191e-08,
+      "loss": 0.0069,
+      "step": 24760
+    },
+    {
+      "epoch": 0.8946113839930656,
+      "grad_norm": 0.20979110643749657,
+      "learning_rate": 8.659456637600639e-08,
+      "loss": 0.0063,
+      "step": 24770
+    },
+    {
+      "epoch": 0.8949725512857556,
+      "grad_norm": 0.2610802039413238,
+      "learning_rate": 8.600799252686553e-08,
+      "loss": 0.0064,
+      "step": 24780
+    },
+    {
+      "epoch": 0.8953337185784456,
+      "grad_norm": 0.33690345226421986,
+      "learning_rate": 8.542335345800401e-08,
+      "loss": 0.0063,
+      "step": 24790
+    },
+    {
+      "epoch": 0.8956948858711355,
+      "grad_norm": 0.25039503378868627,
+      "learning_rate": 8.484064996939023e-08,
+      "loss": 0.0062,
+      "step": 24800
+    },
+    {
+      "epoch": 0.8960560531638255,
+      "grad_norm": 0.29678283107981496,
+      "learning_rate": 8.425988285834307e-08,
+      "loss": 0.0062,
+      "step": 24810
+    },
+    {
+      "epoch": 0.8964172204565155,
+      "grad_norm": 0.3094808099699057,
+      "learning_rate": 8.368105291953288e-08,
+      "loss": 0.0056,
+      "step": 24820
+    },
+    {
+      "epoch": 0.8967783877492055,
+      "grad_norm": 0.24004980473465307,
+      "learning_rate": 8.310416094497825e-08,
+      "loss": 0.0061,
+      "step": 24830
+    },
+    {
+      "epoch": 0.8971395550418954,
+      "grad_norm": 0.28815675931245244,
+      "learning_rate": 8.252920772404709e-08,
+      "loss": 0.0059,
+      "step": 24840
+    },
+    {
+      "epoch": 0.8975007223345853,
+      "grad_norm": 0.2611063174611941,
+      "learning_rate": 8.195619404345345e-08,
+      "loss": 0.006,
+      "step": 24850
+    },
+    {
+      "epoch": 0.8978618896272753,
+      "grad_norm": 0.2678979474525723,
+      "learning_rate": 8.138512068725896e-08,
+      "loss": 0.0068,
+      "step": 24860
+    },
+    {
+      "epoch": 0.8982230569199653,
+      "grad_norm": 0.2616199705267589,
+      "learning_rate": 8.081598843686821e-08,
+      "loss": 0.0065,
+      "step": 24870
+    },
+    {
+      "epoch": 0.8985842242126553,
+      "grad_norm": 0.24589862931888226,
+      "learning_rate": 8.02487980710318e-08,
+      "loss": 0.0064,
+      "step": 24880
+    },
+    {
+      "epoch": 0.8989453915053452,
+      "grad_norm": 0.3147904397769794,
+      "learning_rate": 7.968355036584175e-08,
+      "loss": 0.0066,
+      "step": 24890
+    },
+    {
+      "epoch": 0.8993065587980352,
+      "grad_norm": 0.25106081265658453,
+      "learning_rate": 7.912024609473284e-08,
+      "loss": 0.0061,
+      "step": 24900
+    },
+    {
+      "epoch": 0.8996677260907252,
+      "grad_norm": 0.24971497547091706,
+      "learning_rate": 7.855888602848005e-08,
+      "loss": 0.0062,
+      "step": 24910
+    },
+    {
+      "epoch": 0.9000288933834152,
+      "grad_norm": 0.28605064560358484,
+      "learning_rate": 7.799947093519877e-08,
+      "loss": 0.006,
+      "step": 24920
+    },
+    {
+      "epoch": 0.9003900606761052,
+      "grad_norm": 0.3008061744816294,
+      "learning_rate": 7.74420015803422e-08,
+      "loss": 0.0065,
+      "step": 24930
+    },
+    {
+      "epoch": 0.9007512279687951,
+      "grad_norm": 0.19746453720608906,
+      "learning_rate": 7.688647872670168e-08,
+      "loss": 0.0061,
+      "step": 24940
+    },
+    {
+      "epoch": 0.9011123952614851,
+      "grad_norm": 0.3086932200145029,
+      "learning_rate": 7.633290313440467e-08,
+      "loss": 0.0061,
+      "step": 24950
+    },
+    {
+      "epoch": 0.9014735625541751,
+      "grad_norm": 0.3169348585332674,
+      "learning_rate": 7.578127556091524e-08,
+      "loss": 0.0067,
+      "step": 24960
+    },
+    {
+      "epoch": 0.9018347298468651,
+      "grad_norm": 0.28354708836434095,
+      "learning_rate": 7.523159676103065e-08,
+      "loss": 0.0058,
+      "step": 24970
+    },
+    {
+      "epoch": 0.9021958971395551,
+      "grad_norm": 0.2816824590837959,
+      "learning_rate": 7.468386748688272e-08,
+      "loss": 0.0058,
+      "step": 24980
+    },
+    {
+      "epoch": 0.902557064432245,
+      "grad_norm": 0.4431880268540727,
+      "learning_rate": 7.41380884879348e-08,
+      "loss": 0.0068,
+      "step": 24990
+    },
+    {
+      "epoch": 0.902918231724935,
+      "grad_norm": 0.3045334632360396,
+      "learning_rate": 7.35942605109825e-08,
+      "loss": 0.0059,
+      "step": 25000
+    },
+    {
+      "epoch": 0.902918231724935,
+      "eval_loss": 0.28611651062965393,
+      "eval_runtime": 38.0226,
+      "eval_samples_per_second": 107.726,
+      "eval_steps_per_second": 0.21,
+      "step": 25000
+    },
+    {
+      "epoch": 0.903279399017625,
+      "grad_norm": 0.28923489733296603,
+      "learning_rate": 7.30523843001511e-08,
+      "loss": 0.0064,
+      "step": 25010
+    },
+    {
+      "epoch": 0.903640566310315,
+      "grad_norm": 0.2617828628575408,
+      "learning_rate": 7.251246059689564e-08,
+      "loss": 0.0063,
+      "step": 25020
+    },
+    {
+      "epoch": 0.904001733603005,
+      "grad_norm": 0.2844770133223371,
+      "learning_rate": 7.197449013999968e-08,
+      "loss": 0.0058,
+      "step": 25030
+    },
+    {
+      "epoch": 0.9043629008956949,
+      "grad_norm": 0.3183480924995839,
+      "learning_rate": 7.143847366557337e-08,
+      "loss": 0.0063,
+      "step": 25040
+    },
+    {
+      "epoch": 0.9047240681883849,
+      "grad_norm": 0.2414843044062039,
+      "learning_rate": 7.090441190705438e-08,
+      "loss": 0.0065,
+      "step": 25050
+    },
+    {
+      "epoch": 0.9050852354810748,
+      "grad_norm": 0.3136118320992289,
+      "learning_rate": 7.037230559520497e-08,
+      "loss": 0.0064,
+      "step": 25060
+    },
+    {
+      "epoch": 0.9054464027737648,
+      "grad_norm": 0.3181315231392275,
+      "learning_rate": 6.984215545811124e-08,
+      "loss": 0.0067,
+      "step": 25070
+    },
+    {
+      "epoch": 0.9058075700664547,
+      "grad_norm": 0.2616345513454167,
+      "learning_rate": 6.931396222118407e-08,
+      "loss": 0.0068,
+      "step": 25080
+    },
+    {
+      "epoch": 0.9061687373591447,
+      "grad_norm": 0.2622159751814772,
+      "learning_rate": 6.878772660715554e-08,
+      "loss": 0.0057,
+      "step": 25090
+    },
+    {
+      "epoch": 0.9065299046518347,
+      "grad_norm": 0.2978487851692376,
+      "learning_rate": 6.826344933607898e-08,
+      "loss": 0.0066,
+      "step": 25100
+    },
+    {
+      "epoch": 0.9068910719445247,
+      "grad_norm": 0.2782939041583027,
+      "learning_rate": 6.774113112532937e-08,
+      "loss": 0.0063,
+      "step": 25110
+    },
+    {
+      "epoch": 0.9072522392372147,
+      "grad_norm": 0.29380793465878907,
+      "learning_rate": 6.72207726895997e-08,
+      "loss": 0.0065,
+      "step": 25120
+    },
+    {
+      "epoch": 0.9076134065299046,
+      "grad_norm": 0.24769368441067482,
+      "learning_rate": 6.670237474090208e-08,
+      "loss": 0.0061,
+      "step": 25130
+    },
+    {
+      "epoch": 0.9079745738225946,
+      "grad_norm": 0.33676339925185683,
+      "learning_rate": 6.618593798856592e-08,
+      "loss": 0.006,
+      "step": 25140
+    },
+    {
+      "epoch": 0.9083357411152846,
+      "grad_norm": 0.2535670045949507,
+      "learning_rate": 6.567146313923733e-08,
+      "loss": 0.0063,
+      "step": 25150
+    },
+    {
+      "epoch": 0.9086969084079746,
+      "grad_norm": 0.2711888032264537,
+      "learning_rate": 6.515895089687734e-08,
+      "loss": 0.0061,
+      "step": 25160
+    },
+    {
+      "epoch": 0.9090580757006645,
+      "grad_norm": 0.32482293028069714,
+      "learning_rate": 6.46484019627625e-08,
+      "loss": 0.0065,
+      "step": 25170
+    },
+    {
+      "epoch": 0.9094192429933545,
+      "grad_norm": 0.2453735713366907,
+      "learning_rate": 6.41398170354815e-08,
+      "loss": 0.006,
+      "step": 25180
+    },
+    {
+      "epoch": 0.9097804102860445,
+      "grad_norm": 0.2964398578808616,
+      "learning_rate": 6.363319681093699e-08,
+      "loss": 0.0066,
+      "step": 25190
+    },
+    {
+      "epoch": 0.9101415775787345,
+      "grad_norm": 0.22667633611280041,
+      "learning_rate": 6.312854198234263e-08,
+      "loss": 0.0062,
+      "step": 25200
+    },
+    {
+      "epoch": 0.9105027448714245,
+      "grad_norm": 0.288259474556547,
+      "learning_rate": 6.262585324022325e-08,
+      "loss": 0.0061,
+      "step": 25210
+    },
+    {
+      "epoch": 0.9108639121641144,
+      "grad_norm": 0.26104045618691013,
+      "learning_rate": 6.212513127241292e-08,
+      "loss": 0.0062,
+      "step": 25220
+    },
+    {
+      "epoch": 0.9112250794568044,
+      "grad_norm": 0.2758499589679277,
+      "learning_rate": 6.162637676405514e-08,
+      "loss": 0.0061,
+      "step": 25230
+    },
+    {
+      "epoch": 0.9115862467494944,
+      "grad_norm": 0.28155297785037,
+      "learning_rate": 6.112959039760047e-08,
+      "loss": 0.0064,
+      "step": 25240
+    },
+    {
+      "epoch": 0.9119474140421844,
+      "grad_norm": 0.28146474869147,
+      "learning_rate": 6.063477285280767e-08,
+      "loss": 0.0063,
+      "step": 25250
+    },
+    {
+      "epoch": 0.9123085813348744,
+      "grad_norm": 0.34103586239081246,
+      "learning_rate": 6.014192480674013e-08,
+      "loss": 0.0064,
+      "step": 25260
+    },
+    {
+      "epoch": 0.9126697486275643,
+      "grad_norm": 0.29994074104839735,
+      "learning_rate": 5.96510469337676e-08,
+      "loss": 0.0063,
+      "step": 25270
+    },
+    {
+      "epoch": 0.9130309159202543,
+      "grad_norm": 0.2921073404223388,
+      "learning_rate": 5.916213990556313e-08,
+      "loss": 0.0067,
+      "step": 25280
+    },
+    {
+      "epoch": 0.9133920832129442,
+      "grad_norm": 0.30979192478599926,
+      "learning_rate": 5.8675204391103745e-08,
+      "loss": 0.006,
+      "step": 25290
+    },
+    {
+      "epoch": 0.9137532505056342,
+      "grad_norm": 0.2522636590916905,
+      "learning_rate": 5.8190241056668146e-08,
+      "loss": 0.0061,
+      "step": 25300
+    },
+    {
+      "epoch": 0.9141144177983241,
+      "grad_norm": 0.3316844893062915,
+      "learning_rate": 5.7707250565837036e-08,
+      "loss": 0.0069,
+      "step": 25310
+    },
+    {
+      "epoch": 0.9144755850910141,
+      "grad_norm": 0.27853733322438123,
+      "learning_rate": 5.722623357949125e-08,
+      "loss": 0.0064,
+      "step": 25320
+    },
+    {
+      "epoch": 0.9148367523837041,
+      "grad_norm": 0.28631163836601925,
+      "learning_rate": 5.6747190755811806e-08,
+      "loss": 0.0064,
+      "step": 25330
+    },
+    {
+      "epoch": 0.9151979196763941,
+      "grad_norm": 0.24092554017145806,
+      "learning_rate": 5.627012275027737e-08,
+      "loss": 0.0056,
+      "step": 25340
+    },
+    {
+      "epoch": 0.9155590869690841,
+      "grad_norm": 0.27715062372590465,
+      "learning_rate": 5.579503021566595e-08,
+      "loss": 0.006,
+      "step": 25350
+    },
+    {
+      "epoch": 0.915920254261774,
+      "grad_norm": 0.30464824500767607,
+      "learning_rate": 5.53219138020512e-08,
+      "loss": 0.0063,
+      "step": 25360
+    },
+    {
+      "epoch": 0.916281421554464,
+      "grad_norm": 0.42812550633739327,
+      "learning_rate": 5.485077415680362e-08,
+      "loss": 0.0065,
+      "step": 25370
+    },
+    {
+      "epoch": 0.916642588847154,
+      "grad_norm": 0.2688002687287283,
+      "learning_rate": 5.438161192458835e-08,
+      "loss": 0.0064,
+      "step": 25380
+    },
+    {
+      "epoch": 0.917003756139844,
+      "grad_norm": 0.3768693293440074,
+      "learning_rate": 5.391442774736538e-08,
+      "loss": 0.0067,
+      "step": 25390
+    },
+    {
+      "epoch": 0.917364923432534,
+      "grad_norm": 0.3020781326888371,
+      "learning_rate": 5.344922226438753e-08,
+      "loss": 0.0062,
+      "step": 25400
+    },
+    {
+      "epoch": 0.9177260907252239,
+      "grad_norm": 0.24771453906619692,
+      "learning_rate": 5.298599611220029e-08,
+      "loss": 0.0064,
+      "step": 25410
+    },
+    {
+      "epoch": 0.9180872580179139,
+      "grad_norm": 0.19985207370113273,
+      "learning_rate": 5.252474992464112e-08,
+      "loss": 0.0058,
+      "step": 25420
+    },
+    {
+      "epoch": 0.9184484253106039,
+      "grad_norm": 0.3102580629247061,
+      "learning_rate": 5.206548433283803e-08,
+      "loss": 0.0063,
+      "step": 25430
+    },
+    {
+      "epoch": 0.9188095926032939,
+      "grad_norm": 0.2844818776029681,
+      "learning_rate": 5.160819996520916e-08,
+      "loss": 0.0065,
+      "step": 25440
+    },
+    {
+      "epoch": 0.9191707598959838,
+      "grad_norm": 0.3426046535066642,
+      "learning_rate": 5.1152897447461154e-08,
+      "loss": 0.0067,
+      "step": 25450
+    },
+    {
+      "epoch": 0.9195319271886738,
+      "grad_norm": 1.4258957378201729,
+      "learning_rate": 5.069957740258985e-08,
+      "loss": 0.0063,
+      "step": 25460
+    },
+    {
+      "epoch": 0.9198930944813638,
+      "grad_norm": 0.26689309814887613,
+      "learning_rate": 5.024824045087739e-08,
+      "loss": 0.0062,
+      "step": 25470
+    },
+    {
+      "epoch": 0.9202542617740538,
+      "grad_norm": 0.29938058883069985,
+      "learning_rate": 4.979888720989345e-08,
+      "loss": 0.0061,
+      "step": 25480
+    },
+    {
+      "epoch": 0.9206154290667438,
+      "grad_norm": 0.29433819673511946,
+      "learning_rate": 4.9351518294492024e-08,
+      "loss": 0.0059,
+      "step": 25490
+    },
+    {
+      "epoch": 0.9209765963594337,
+      "grad_norm": 0.3533991760145781,
+      "learning_rate": 4.890613431681362e-08,
+      "loss": 0.0061,
+      "step": 25500
+    },
+    {
+      "epoch": 0.9213377636521236,
+      "grad_norm": 0.37010073092641493,
+      "learning_rate": 4.8462735886281086e-08,
+      "loss": 0.0063,
+      "step": 25510
+    },
+    {
+      "epoch": 0.9216989309448136,
+      "grad_norm": 0.25736261758147916,
+      "learning_rate": 4.8021323609601933e-08,
+      "loss": 0.0062,
+      "step": 25520
+    },
+    {
+      "epoch": 0.9220600982375036,
+      "grad_norm": 0.3685809551714689,
+      "learning_rate": 4.7581898090764685e-08,
+      "loss": 0.0061,
+      "step": 25530
+    },
+    {
+      "epoch": 0.9224212655301935,
+      "grad_norm": 0.276515021030957,
+      "learning_rate": 4.714445993104022e-08,
+      "loss": 0.0063,
+      "step": 25540
+    },
+    {
+      "epoch": 0.9227824328228835,
+      "grad_norm": 0.30497915075656584,
+      "learning_rate": 4.670900972897973e-08,
+      "loss": 0.0063,
+      "step": 25550
+    },
+    {
+      "epoch": 0.9231436001155735,
+      "grad_norm": 0.2964540749061863,
+      "learning_rate": 4.6275548080414785e-08,
+      "loss": 0.0063,
+      "step": 25560
+    },
+    {
+      "epoch": 0.9235047674082635,
+      "grad_norm": 0.253840105858815,
+      "learning_rate": 4.584407557845477e-08,
+      "loss": 0.0065,
+      "step": 25570
+    },
+    {
+      "epoch": 0.9238659347009535,
+      "grad_norm": 0.27817049552332546,
+      "learning_rate": 4.541459281348875e-08,
+      "loss": 0.0062,
+      "step": 25580
+    },
+    {
+      "epoch": 0.9242271019936434,
+      "grad_norm": 0.48678437272879904,
+      "learning_rate": 4.498710037318199e-08,
+      "loss": 0.0062,
+      "step": 25590
+    },
+    {
+      "epoch": 0.9245882692863334,
+      "grad_norm": 0.28379233592009406,
+      "learning_rate": 4.456159884247757e-08,
+      "loss": 0.0062,
+      "step": 25600
+    },
+    {
+      "epoch": 0.9249494365790234,
+      "grad_norm": 0.2916730068275228,
+      "learning_rate": 4.4138088803593104e-08,
+      "loss": 0.0066,
+      "step": 25610
+    },
+    {
+      "epoch": 0.9253106038717134,
+      "grad_norm": 0.36277708313470053,
+      "learning_rate": 4.371657083602254e-08,
+      "loss": 0.0067,
+      "step": 25620
+    },
+    {
+      "epoch": 0.9256717711644034,
+      "grad_norm": 0.29880629509154133,
+      "learning_rate": 4.329704551653269e-08,
+      "loss": 0.0065,
+      "step": 25630
+    },
+    {
+      "epoch": 0.9260329384570933,
+      "grad_norm": 0.2748395627855796,
+      "learning_rate": 4.287951341916535e-08,
+      "loss": 0.0057,
+      "step": 25640
+    },
+    {
+      "epoch": 0.9263941057497833,
+      "grad_norm": 0.30283480185209505,
+      "learning_rate": 4.24639751152332e-08,
+      "loss": 0.0064,
+      "step": 25650
+    },
+    {
+      "epoch": 0.9267552730424733,
+      "grad_norm": 0.25213399408682163,
+      "learning_rate": 4.2050431173322414e-08,
+      "loss": 0.0061,
+      "step": 25660
+    },
+    {
+      "epoch": 0.9271164403351633,
+      "grad_norm": 0.25614995737304946,
+      "learning_rate": 4.163888215928918e-08,
+      "loss": 0.0063,
+      "step": 25670
+    },
+    {
+      "epoch": 0.9274776076278533,
+      "grad_norm": 0.20574483616521158,
+      "learning_rate": 4.1229328636260554e-08,
+      "loss": 0.0063,
+      "step": 25680
+    },
+    {
+      "epoch": 0.9278387749205432,
+      "grad_norm": 0.2558624285713857,
+      "learning_rate": 4.082177116463276e-08,
+      "loss": 0.0061,
+      "step": 25690
+    },
+    {
+      "epoch": 0.9281999422132332,
+      "grad_norm": 0.2353427611382282,
+      "learning_rate": 4.0416210302071375e-08,
+      "loss": 0.006,
+      "step": 25700
+    },
+    {
+      "epoch": 0.9285611095059232,
+      "grad_norm": 0.2821021717716629,
+      "learning_rate": 4.001264660350917e-08,
+      "loss": 0.0066,
+      "step": 25710
+    },
+    {
+      "epoch": 0.9289222767986132,
+      "grad_norm": 0.28718795640620054,
+      "learning_rate": 3.961108062114727e-08,
+      "loss": 0.0061,
+      "step": 25720
+    },
+    {
+      "epoch": 0.9292834440913031,
+      "grad_norm": 0.3407241538796185,
+      "learning_rate": 3.921151290445213e-08,
+      "loss": 0.0064,
+      "step": 25730
+    },
+    {
+      "epoch": 0.929644611383993,
+      "grad_norm": 0.2795539806498048,
+      "learning_rate": 3.8813944000156755e-08,
+      "loss": 0.0064,
+      "step": 25740
+    },
+    {
+      "epoch": 0.930005778676683,
+      "grad_norm": 0.283391026905123,
+      "learning_rate": 3.8418374452258486e-08,
+      "loss": 0.0064,
+      "step": 25750
+    },
+    {
+      "epoch": 0.930366945969373,
+      "grad_norm": 0.27584645761656323,
+      "learning_rate": 3.802480480202003e-08,
+      "loss": 0.0064,
+      "step": 25760
+    },
+    {
+      "epoch": 0.930728113262063,
+      "grad_norm": 0.2742928151257981,
+      "learning_rate": 3.7633235587966275e-08,
+      "loss": 0.006,
+      "step": 25770
+    },
+    {
+      "epoch": 0.9310892805547529,
+      "grad_norm": 0.26727409818232584,
+      "learning_rate": 3.724366734588613e-08,
+      "loss": 0.0062,
+      "step": 25780
+    },
+    {
+      "epoch": 0.9314504478474429,
+      "grad_norm": 0.2840487269373605,
+      "learning_rate": 3.685610060882955e-08,
+      "loss": 0.006,
+      "step": 25790
+    },
+    {
+      "epoch": 0.9318116151401329,
+      "grad_norm": 0.32672536145362346,
+      "learning_rate": 3.647053590710814e-08,
+      "loss": 0.0063,
+      "step": 25800
+    },
+    {
+      "epoch": 0.9321727824328229,
+      "grad_norm": 0.2938466442387466,
+      "learning_rate": 3.608697376829456e-08,
+      "loss": 0.0062,
+      "step": 25810
+    },
+    {
+      "epoch": 0.9325339497255128,
+      "grad_norm": 0.26534025311136755,
+      "learning_rate": 3.57054147172205e-08,
+      "loss": 0.0061,
+      "step": 25820
+    },
+    {
+      "epoch": 0.9328951170182028,
+      "grad_norm": 0.4950064913739883,
+      "learning_rate": 3.532585927597765e-08,
+      "loss": 0.0061,
+      "step": 25830
+    },
+    {
+      "epoch": 0.9332562843108928,
+      "grad_norm": 0.28227912507414915,
+      "learning_rate": 3.494830796391557e-08,
+      "loss": 0.006,
+      "step": 25840
+    },
+    {
+      "epoch": 0.9336174516035828,
+      "grad_norm": 0.2609433213077335,
+      "learning_rate": 3.4572761297641867e-08,
+      "loss": 0.0062,
+      "step": 25850
+    },
+    {
+      "epoch": 0.9339786188962728,
+      "grad_norm": 0.3264611187954517,
+      "learning_rate": 3.419921979102081e-08,
+      "loss": 0.0064,
+      "step": 25860
+    },
+    {
+      "epoch": 0.9343397861889627,
+      "grad_norm": 0.2461503763693965,
+      "learning_rate": 3.382768395517355e-08,
+      "loss": 0.0061,
+      "step": 25870
+    },
+    {
+      "epoch": 0.9347009534816527,
+      "grad_norm": 0.2783517841983462,
+      "learning_rate": 3.34581542984761e-08,
+      "loss": 0.006,
+      "step": 25880
+    },
+    {
+      "epoch": 0.9350621207743427,
+      "grad_norm": 0.25300819019489734,
+      "learning_rate": 3.309063132655999e-08,
+      "loss": 0.006,
+      "step": 25890
+    },
+    {
+      "epoch": 0.9354232880670327,
+      "grad_norm": 0.31658814233613025,
+      "learning_rate": 3.272511554231078e-08,
+      "loss": 0.0059,
+      "step": 25900
+    },
+    {
+      "epoch": 0.9357844553597227,
+      "grad_norm": 0.3214057761033852,
+      "learning_rate": 3.236160744586791e-08,
+      "loss": 0.006,
+      "step": 25910
+    },
+    {
+      "epoch": 0.9361456226524126,
+      "grad_norm": 0.30696829877368115,
+      "learning_rate": 3.2000107534622844e-08,
+      "loss": 0.0058,
+      "step": 25920
+    },
+    {
+      "epoch": 0.9365067899451026,
+      "grad_norm": 0.3267997979793133,
+      "learning_rate": 3.164061630322024e-08,
+      "loss": 0.006,
+      "step": 25930
+    },
+    {
+      "epoch": 0.9368679572377926,
+      "grad_norm": 0.3170632894157784,
+      "learning_rate": 3.128313424355528e-08,
+      "loss": 0.0057,
+      "step": 25940
+    },
+    {
+      "epoch": 0.9372291245304826,
+      "grad_norm": 0.23266139441347825,
+      "learning_rate": 3.092766184477502e-08,
+      "loss": 0.0063,
+      "step": 25950
+    },
+    {
+      "epoch": 0.9375902918231724,
+      "grad_norm": 0.2660023446493081,
+      "learning_rate": 3.057419959327556e-08,
+      "loss": 0.0062,
+      "step": 25960
+    },
+    {
+      "epoch": 0.9379514591158624,
+      "grad_norm": 0.27872260176424324,
+      "learning_rate": 3.022274797270352e-08,
+      "loss": 0.0061,
+      "step": 25970
+    },
+    {
+      "epoch": 0.9383126264085524,
+      "grad_norm": 0.3266953092358667,
+      "learning_rate": 2.987330746395339e-08,
+      "loss": 0.0062,
+      "step": 25980
+    },
+    {
+      "epoch": 0.9386737937012424,
+      "grad_norm": 0.33305721650792336,
+      "learning_rate": 2.952587854516886e-08,
+      "loss": 0.006,
+      "step": 25990
+    },
+    {
+      "epoch": 0.9390349609939324,
+      "grad_norm": 0.9882648679373008,
+      "learning_rate": 2.9180461691740324e-08,
+      "loss": 0.006,
+      "step": 26000
+    },
+    {
+      "epoch": 0.9390349609939324,
+      "eval_loss": 0.28821802139282227,
+      "eval_runtime": 38.612,
+      "eval_samples_per_second": 106.081,
+      "eval_steps_per_second": 0.207,
+      "step": 26000
+    },
+    {
+      "epoch": 0.9393961282866223,
+      "grad_norm": 0.27126590390219224,
+      "learning_rate": 2.8837057376305375e-08,
+      "loss": 0.0062,
+      "step": 26010
+    },
+    {
+      "epoch": 0.9397572955793123,
+      "grad_norm": 0.3065642749520776,
+      "learning_rate": 2.849566606874765e-08,
+      "loss": 0.0064,
+      "step": 26020
+    },
+    {
+      "epoch": 0.9401184628720023,
+      "grad_norm": 0.3017803675821032,
+      "learning_rate": 2.8156288236196815e-08,
+      "loss": 0.0059,
+      "step": 26030
+    },
+    {
+      "epoch": 0.9404796301646923,
+      "grad_norm": 0.4141693173197063,
+      "learning_rate": 2.7818924343026908e-08,
+      "loss": 0.0065,
+      "step": 26040
+    },
+    {
+      "epoch": 0.9408407974573822,
+      "grad_norm": 0.2715726418478876,
+      "learning_rate": 2.7483574850856507e-08,
+      "loss": 0.0064,
+      "step": 26050
+    },
+    {
+      "epoch": 0.9412019647500722,
+      "grad_norm": 0.25374833613571285,
+      "learning_rate": 2.715024021854806e-08,
+      "loss": 0.0064,
+      "step": 26060
+    },
+    {
+      "epoch": 0.9415631320427622,
+      "grad_norm": 0.2331812358397951,
+      "learning_rate": 2.6818920902206888e-08,
+      "loss": 0.0066,
+      "step": 26070
+    },
+    {
+      "epoch": 0.9419242993354522,
+      "grad_norm": 0.29356462950654993,
+      "learning_rate": 2.648961735518035e-08,
+      "loss": 0.0062,
+      "step": 26080
+    },
+    {
+      "epoch": 0.9422854666281422,
+      "grad_norm": 0.2662020035320196,
+      "learning_rate": 2.616233002805851e-08,
+      "loss": 0.0061,
+      "step": 26090
+    },
+    {
+      "epoch": 0.9426466339208321,
+      "grad_norm": 0.25539077277200795,
+      "learning_rate": 2.5837059368671812e-08,
+      "loss": 0.0059,
+      "step": 26100
+    },
+    {
+      "epoch": 0.9430078012135221,
+      "grad_norm": 0.27043870030992123,
+      "learning_rate": 2.5513805822091563e-08,
+      "loss": 0.0063,
+      "step": 26110
+    },
+    {
+      "epoch": 0.9433689685062121,
+      "grad_norm": 0.366142716794767,
+      "learning_rate": 2.519256983062912e-08,
+      "loss": 0.0066,
+      "step": 26120
+    },
+    {
+      "epoch": 0.9437301357989021,
+      "grad_norm": 0.3382764735876802,
+      "learning_rate": 2.4873351833835045e-08,
+      "loss": 0.0061,
+      "step": 26130
+    },
+    {
+      "epoch": 0.9440913030915921,
+      "grad_norm": 0.24275772784416666,
+      "learning_rate": 2.4556152268498612e-08,
+      "loss": 0.006,
+      "step": 26140
+    },
+    {
+      "epoch": 0.944452470384282,
+      "grad_norm": 0.23249788216982545,
+      "learning_rate": 2.4240971568647462e-08,
+      "loss": 0.0062,
+      "step": 26150
+    },
+    {
+      "epoch": 0.944813637676972,
+      "grad_norm": 0.2919262471049549,
+      "learning_rate": 2.3927810165546793e-08,
+      "loss": 0.0066,
+      "step": 26160
+    },
+    {
+      "epoch": 0.945174804969662,
+      "grad_norm": 0.315988002770747,
+      "learning_rate": 2.3616668487698667e-08,
+      "loss": 0.0066,
+      "step": 26170
+    },
+    {
+      "epoch": 0.945535972262352,
+      "grad_norm": 0.4255666534508519,
+      "learning_rate": 2.3307546960841363e-08,
+      "loss": 0.0062,
+      "step": 26180
+    },
+    {
+      "epoch": 0.9458971395550418,
+      "grad_norm": 0.30151036385524305,
+      "learning_rate": 2.3000446007949038e-08,
+      "loss": 0.0067,
+      "step": 26190
+    },
+    {
+      "epoch": 0.9462583068477318,
+      "grad_norm": 0.24867030209945565,
+      "learning_rate": 2.269536604923156e-08,
+      "loss": 0.0061,
+      "step": 26200
+    },
+    {
+      "epoch": 0.9466194741404218,
+      "grad_norm": 0.28576470808622073,
+      "learning_rate": 2.2392307502132836e-08,
+      "loss": 0.0061,
+      "step": 26210
+    },
+    {
+      "epoch": 0.9469806414331118,
+      "grad_norm": 0.33373731514715926,
+      "learning_rate": 2.209127078133133e-08,
+      "loss": 0.0059,
+      "step": 26220
+    },
+    {
+      "epoch": 0.9473418087258018,
+      "grad_norm": 0.2546504557913718,
+      "learning_rate": 2.179225629873838e-08,
+      "loss": 0.006,
+      "step": 26230
+    },
+    {
+      "epoch": 0.9477029760184917,
+      "grad_norm": 0.3385664640040468,
+      "learning_rate": 2.1495264463499032e-08,
+      "loss": 0.0065,
+      "step": 26240
+    },
+    {
+      "epoch": 0.9480641433111817,
+      "grad_norm": 0.31637592611814624,
+      "learning_rate": 2.1200295681990056e-08,
+      "loss": 0.0061,
+      "step": 26250
+    },
+    {
+      "epoch": 0.9484253106038717,
+      "grad_norm": 0.2540614871145518,
+      "learning_rate": 2.090735035782093e-08,
+      "loss": 0.0065,
+      "step": 26260
+    },
+    {
+      "epoch": 0.9487864778965617,
+      "grad_norm": 0.28373784588155637,
+      "learning_rate": 2.0616428891831185e-08,
+      "loss": 0.0062,
+      "step": 26270
+    },
+    {
+      "epoch": 0.9491476451892517,
+      "grad_norm": 0.35366095720007057,
+      "learning_rate": 2.0327531682092228e-08,
+      "loss": 0.0059,
+      "step": 26280
+    },
+    {
+      "epoch": 0.9495088124819416,
+      "grad_norm": 0.2783701279465717,
+      "learning_rate": 2.0040659123904847e-08,
+      "loss": 0.0061,
+      "step": 26290
+    },
+    {
+      "epoch": 0.9498699797746316,
+      "grad_norm": 0.29506885703629276,
+      "learning_rate": 1.9755811609800057e-08,
+      "loss": 0.0063,
+      "step": 26300
+    },
+    {
+      "epoch": 0.9502311470673216,
+      "grad_norm": 0.3283021638090425,
+      "learning_rate": 1.9472989529537577e-08,
+      "loss": 0.0059,
+      "step": 26310
+    },
+    {
+      "epoch": 0.9505923143600116,
+      "grad_norm": 0.2660262972359055,
+      "learning_rate": 1.9192193270106018e-08,
+      "loss": 0.0057,
+      "step": 26320
+    },
+    {
+      "epoch": 0.9509534816527015,
+      "grad_norm": 0.25079142987096364,
+      "learning_rate": 1.8913423215721538e-08,
+      "loss": 0.0066,
+      "step": 26330
+    },
+    {
+      "epoch": 0.9513146489453915,
+      "grad_norm": 0.29867410650392845,
+      "learning_rate": 1.8636679747828846e-08,
+      "loss": 0.0061,
+      "step": 26340
+    },
+    {
+      "epoch": 0.9516758162380815,
+      "grad_norm": 0.26424318520655576,
+      "learning_rate": 1.8361963245098202e-08,
+      "loss": 0.0066,
+      "step": 26350
+    },
+    {
+      "epoch": 0.9520369835307715,
+      "grad_norm": 0.2694281419657487,
+      "learning_rate": 1.8089274083427586e-08,
+      "loss": 0.006,
+      "step": 26360
+    },
+    {
+      "epoch": 0.9523981508234615,
+      "grad_norm": 0.2487465807310416,
+      "learning_rate": 1.7818612635940367e-08,
+      "loss": 0.0064,
+      "step": 26370
+    },
+    {
+      "epoch": 0.9527593181161514,
+      "grad_norm": 0.26070185784229827,
+      "learning_rate": 1.7549979272985294e-08,
+      "loss": 0.0056,
+      "step": 26380
+    },
+    {
+      "epoch": 0.9531204854088414,
+      "grad_norm": 0.25137353157191555,
+      "learning_rate": 1.7283374362136506e-08,
+      "loss": 0.0065,
+      "step": 26390
+    },
+    {
+      "epoch": 0.9534816527015314,
+      "grad_norm": 0.29840605325425507,
+      "learning_rate": 1.701879826819236e-08,
+      "loss": 0.0068,
+      "step": 26400
+    },
+    {
+      "epoch": 0.9538428199942213,
+      "grad_norm": 0.3044388036573936,
+      "learning_rate": 1.6756251353174767e-08,
+      "loss": 0.0062,
+      "step": 26410
+    },
+    {
+      "epoch": 0.9542039872869112,
+      "grad_norm": 0.27190691376349096,
+      "learning_rate": 1.649573397633003e-08,
+      "loss": 0.0065,
+      "step": 26420
+    },
+    {
+      "epoch": 0.9545651545796012,
+      "grad_norm": 0.4999485810588351,
+      "learning_rate": 1.6237246494126167e-08,
+      "loss": 0.0061,
+      "step": 26430
+    },
+    {
+      "epoch": 0.9549263218722912,
+      "grad_norm": 0.3002941127156946,
+      "learning_rate": 1.5980789260254924e-08,
+      "loss": 0.0063,
+      "step": 26440
+    },
+    {
+      "epoch": 0.9552874891649812,
+      "grad_norm": 0.2807980518743483,
+      "learning_rate": 1.5726362625629097e-08,
+      "loss": 0.0061,
+      "step": 26450
+    },
+    {
+      "epoch": 0.9556486564576712,
+      "grad_norm": 0.5109084387850991,
+      "learning_rate": 1.5473966938383544e-08,
+      "loss": 0.0064,
+      "step": 26460
+    },
+    {
+      "epoch": 0.9560098237503611,
+      "grad_norm": 0.37801394325018606,
+      "learning_rate": 1.5223602543873837e-08,
+      "loss": 0.006,
+      "step": 26470
+    },
+    {
+      "epoch": 0.9563709910430511,
+      "grad_norm": 0.2355496114016863,
+      "learning_rate": 1.497526978467628e-08,
+      "loss": 0.0061,
+      "step": 26480
+    },
+    {
+      "epoch": 0.9567321583357411,
+      "grad_norm": 0.34209447999152176,
+      "learning_rate": 1.4728969000587233e-08,
+      "loss": 0.0061,
+      "step": 26490
+    },
+    {
+      "epoch": 0.9570933256284311,
+      "grad_norm": 0.3775868476279125,
+      "learning_rate": 1.4484700528622774e-08,
+      "loss": 0.0062,
+      "step": 26500
+    },
+    {
+      "epoch": 0.9574544929211211,
+      "grad_norm": 0.31563392114735367,
+      "learning_rate": 1.4242464703018043e-08,
+      "loss": 0.0061,
+      "step": 26510
+    },
+    {
+      "epoch": 0.957815660213811,
+      "grad_norm": 0.36471612064098213,
+      "learning_rate": 1.4002261855226572e-08,
+      "loss": 0.006,
+      "step": 26520
+    },
+    {
+      "epoch": 0.958176827506501,
+      "grad_norm": 0.2806512150638093,
+      "learning_rate": 1.3764092313920951e-08,
+      "loss": 0.0065,
+      "step": 26530
+    },
+    {
+      "epoch": 0.958537994799191,
+      "grad_norm": 0.3001935286952102,
+      "learning_rate": 1.352795640499066e-08,
+      "loss": 0.0063,
+      "step": 26540
+    },
+    {
+      "epoch": 0.958899162091881,
+      "grad_norm": 0.2736395549729549,
+      "learning_rate": 1.3293854451543242e-08,
+      "loss": 0.0064,
+      "step": 26550
+    },
+    {
+      "epoch": 0.959260329384571,
+      "grad_norm": 0.27642380285690965,
+      "learning_rate": 1.3061786773902795e-08,
+      "loss": 0.006,
+      "step": 26560
+    },
+    {
+      "epoch": 0.9596214966772609,
+      "grad_norm": 0.23859261290420694,
+      "learning_rate": 1.2831753689609982e-08,
+      "loss": 0.0059,
+      "step": 26570
+    },
+    {
+      "epoch": 0.9599826639699509,
+      "grad_norm": 0.39672835985380644,
+      "learning_rate": 1.2603755513421521e-08,
+      "loss": 0.006,
+      "step": 26580
+    },
+    {
+      "epoch": 0.9603438312626409,
+      "grad_norm": 0.25607381268474605,
+      "learning_rate": 1.2377792557309864e-08,
+      "loss": 0.0057,
+      "step": 26590
+    },
+    {
+      "epoch": 0.9607049985553309,
+      "grad_norm": 0.27318880326056916,
+      "learning_rate": 1.2153865130462183e-08,
+      "loss": 0.0064,
+      "step": 26600
+    },
+    {
+      "epoch": 0.9610661658480208,
+      "grad_norm": 0.2252711758880034,
+      "learning_rate": 1.1931973539281049e-08,
+      "loss": 0.0064,
+      "step": 26610
+    },
+    {
+      "epoch": 0.9614273331407108,
+      "grad_norm": 0.311531321905567,
+      "learning_rate": 1.1712118087383094e-08,
+      "loss": 0.006,
+      "step": 26620
+    },
+    {
+      "epoch": 0.9617885004334008,
+      "grad_norm": 0.28691853658212785,
+      "learning_rate": 1.149429907559868e-08,
+      "loss": 0.0061,
+      "step": 26630
+    },
+    {
+      "epoch": 0.9621496677260907,
+      "grad_norm": 0.27960391107125865,
+      "learning_rate": 1.1278516801971894e-08,
+      "loss": 0.0061,
+      "step": 26640
+    },
+    {
+      "epoch": 0.9625108350187807,
+      "grad_norm": 0.30333252097969426,
+      "learning_rate": 1.1064771561760223e-08,
+      "loss": 0.0064,
+      "step": 26650
+    },
+    {
+      "epoch": 0.9628720023114706,
+      "grad_norm": 0.26455010666043544,
+      "learning_rate": 1.0853063647433215e-08,
+      "loss": 0.0059,
+      "step": 26660
+    },
+    {
+      "epoch": 0.9632331696041606,
+      "grad_norm": 0.2584788443869586,
+      "learning_rate": 1.0643393348673148e-08,
+      "loss": 0.0063,
+      "step": 26670
+    },
+    {
+      "epoch": 0.9635943368968506,
+      "grad_norm": 0.28077809560897543,
+      "learning_rate": 1.0435760952374196e-08,
+      "loss": 0.0064,
+      "step": 26680
+    },
+    {
+      "epoch": 0.9639555041895406,
+      "grad_norm": 0.2526185974981328,
+      "learning_rate": 1.0230166742641766e-08,
+      "loss": 0.006,
+      "step": 26690
+    },
+    {
+      "epoch": 0.9643166714822305,
+      "grad_norm": 0.2898573722579625,
+      "learning_rate": 1.0026611000792829e-08,
+      "loss": 0.0059,
+      "step": 26700
+    },
+    {
+      "epoch": 0.9646778387749205,
+      "grad_norm": 0.24176425738521945,
+      "learning_rate": 9.825094005354751e-09,
+      "loss": 0.0064,
+      "step": 26710
+    },
+    {
+      "epoch": 0.9650390060676105,
+      "grad_norm": 0.2996567423742234,
+      "learning_rate": 9.625616032065466e-09,
+      "loss": 0.0062,
+      "step": 26720
+    },
+    {
+      "epoch": 0.9654001733603005,
+      "grad_norm": 0.371295312764522,
+      "learning_rate": 9.428177353872803e-09,
+      "loss": 0.0063,
+      "step": 26730
+    },
+    {
+      "epoch": 0.9657613406529905,
+      "grad_norm": 0.23891988476512807,
+      "learning_rate": 9.232778240933992e-09,
+      "loss": 0.0063,
+      "step": 26740
+    },
+    {
+      "epoch": 0.9661225079456804,
+      "grad_norm": 0.33408515280143036,
+      "learning_rate": 9.039418960615997e-09,
+      "loss": 0.0063,
+      "step": 26750
+    },
+    {
+      "epoch": 0.9664836752383704,
+      "grad_norm": 0.27363662161991614,
+      "learning_rate": 8.848099777494012e-09,
+      "loss": 0.0061,
+      "step": 26760
+    },
+    {
+      "epoch": 0.9668448425310604,
+      "grad_norm": 0.2885757094550822,
+      "learning_rate": 8.658820953352464e-09,
+      "loss": 0.006,
+      "step": 26770
+    },
+    {
+      "epoch": 0.9672060098237504,
+      "grad_norm": 0.36127778146007394,
+      "learning_rate": 8.471582747183514e-09,
+      "loss": 0.0061,
+      "step": 26780
+    },
+    {
+      "epoch": 0.9675671771164404,
+      "grad_norm": 0.3154546786185918,
+      "learning_rate": 8.286385415187059e-09,
+      "loss": 0.0063,
+      "step": 26790
+    },
+    {
+      "epoch": 0.9679283444091303,
+      "grad_norm": 0.23823024367360593,
+      "learning_rate": 8.103229210770724e-09,
+      "loss": 0.0062,
+      "step": 26800
+    },
+    {
+      "epoch": 0.9682895117018203,
+      "grad_norm": 0.30192159050764245,
+      "learning_rate": 7.922114384548873e-09,
+      "loss": 0.0063,
+      "step": 26810
+    },
+    {
+      "epoch": 0.9686506789945103,
+      "grad_norm": 0.27930153021582516,
+      "learning_rate": 7.743041184343102e-09,
+      "loss": 0.0066,
+      "step": 26820
+    },
+    {
+      "epoch": 0.9690118462872003,
+      "grad_norm": 0.22995442348807116,
+      "learning_rate": 7.56600985518091e-09,
+      "loss": 0.006,
+      "step": 26830
+    },
+    {
+      "epoch": 0.9693730135798903,
+      "grad_norm": 0.27823297505395034,
+      "learning_rate": 7.3910206392963645e-09,
+      "loss": 0.0058,
+      "step": 26840
+    },
+    {
+      "epoch": 0.9697341808725802,
+      "grad_norm": 0.2953067338950638,
+      "learning_rate": 7.218073776129097e-09,
+      "loss": 0.0062,
+      "step": 26850
+    },
+    {
+      "epoch": 0.9700953481652701,
+      "grad_norm": 0.3101966735910379,
+      "learning_rate": 7.047169502323814e-09,
+      "loss": 0.007,
+      "step": 26860
+    },
+    {
+      "epoch": 0.9704565154579601,
+      "grad_norm": 0.24474528775957657,
+      "learning_rate": 6.8783080517309526e-09,
+      "loss": 0.0065,
+      "step": 26870
+    },
+    {
+      "epoch": 0.9708176827506501,
+      "grad_norm": 0.25911992049295823,
+      "learning_rate": 6.711489655405356e-09,
+      "loss": 0.0064,
+      "step": 26880
+    },
+    {
+      "epoch": 0.97117885004334,
+      "grad_norm": 0.40041961597869286,
+      "learning_rate": 6.546714541606435e-09,
+      "loss": 0.0066,
+      "step": 26890
+    },
+    {
+      "epoch": 0.97154001733603,
+      "grad_norm": 0.36010964365214954,
+      "learning_rate": 6.383982935797672e-09,
+      "loss": 0.0061,
+      "step": 26900
+    },
+    {
+      "epoch": 0.97190118462872,
+      "grad_norm": 0.2727693848146648,
+      "learning_rate": 6.223295060646284e-09,
+      "loss": 0.006,
+      "step": 26910
+    },
+    {
+      "epoch": 0.97226235192141,
+      "grad_norm": 0.32071100246803325,
+      "learning_rate": 6.064651136023558e-09,
+      "loss": 0.0062,
+      "step": 26920
+    },
+    {
+      "epoch": 0.9726235192141,
+      "grad_norm": 0.4605193209788745,
+      "learning_rate": 5.908051379003354e-09,
+      "loss": 0.0061,
+      "step": 26930
+    },
+    {
+      "epoch": 0.9729846865067899,
+      "grad_norm": 0.32119753331336587,
+      "learning_rate": 5.753496003862935e-09,
+      "loss": 0.0063,
+      "step": 26940
+    },
+    {
+      "epoch": 0.9733458537994799,
+      "grad_norm": 0.36603596741056854,
+      "learning_rate": 5.6009852220821335e-09,
+      "loss": 0.0063,
+      "step": 26950
+    },
+    {
+      "epoch": 0.9737070210921699,
+      "grad_norm": 0.296787401573609,
+      "learning_rate": 5.450519242342855e-09,
+      "loss": 0.0065,
+      "step": 26960
+    },
+    {
+      "epoch": 0.9740681883848599,
+      "grad_norm": 0.2770939084508547,
+      "learning_rate": 5.302098270529409e-09,
+      "loss": 0.0061,
+      "step": 26970
+    },
+    {
+      "epoch": 0.9744293556775498,
+      "grad_norm": 0.21108740495666783,
+      "learning_rate": 5.155722509728011e-09,
+      "loss": 0.0062,
+      "step": 26980
+    },
+    {
+      "epoch": 0.9747905229702398,
+      "grad_norm": 0.29340383570801604,
+      "learning_rate": 5.0113921602256144e-09,
+      "loss": 0.0063,
+      "step": 26990
+    },
+    {
+      "epoch": 0.9751516902629298,
+      "grad_norm": 0.2542348747810611,
+      "learning_rate": 4.869107419511576e-09,
+      "loss": 0.0063,
+      "step": 27000
+    },
+    {
+      "epoch": 0.9751516902629298,
+      "eval_loss": 0.2879107594490051,
+      "eval_runtime": 38.0371,
+      "eval_samples_per_second": 107.684,
+      "eval_steps_per_second": 0.21,
+      "step": 27000
+    },
+    {
+      "epoch": 0.9755128575556198,
+      "grad_norm": 0.25764466308927286,
+      "learning_rate": 4.728868482275161e-09,
+      "loss": 0.0059,
+      "step": 27010
+    },
+    {
+      "epoch": 0.9758740248483098,
+      "grad_norm": 0.23036334657715019,
+      "learning_rate": 4.590675540407041e-09,
+      "loss": 0.0059,
+      "step": 27020
+    },
+    {
+      "epoch": 0.9762351921409997,
+      "grad_norm": 0.37739018527020496,
+      "learning_rate": 4.454528782997957e-09,
+      "loss": 0.0059,
+      "step": 27030
+    },
+    {
+      "epoch": 0.9765963594336897,
+      "grad_norm": 0.23724892856575996,
+      "learning_rate": 4.320428396339226e-09,
+      "loss": 0.006,
+      "step": 27040
+    },
+    {
+      "epoch": 0.9769575267263797,
+      "grad_norm": 0.2754941817964582,
+      "learning_rate": 4.188374563921571e-09,
+      "loss": 0.006,
+      "step": 27050
+    },
+    {
+      "epoch": 0.9773186940190697,
+      "grad_norm": 0.3193886494514938,
+      "learning_rate": 4.05836746643562e-09,
+      "loss": 0.0061,
+      "step": 27060
+    },
+    {
+      "epoch": 0.9776798613117597,
+      "grad_norm": 0.35338992303307715,
+      "learning_rate": 3.930407281771575e-09,
+      "loss": 0.0064,
+      "step": 27070
+    },
+    {
+      "epoch": 0.9780410286044496,
+      "grad_norm": 0.31313663521359125,
+      "learning_rate": 3.8044941850187144e-09,
+      "loss": 0.0063,
+      "step": 27080
+    },
+    {
+      "epoch": 0.9784021958971395,
+      "grad_norm": 0.26413072690609013,
+      "learning_rate": 3.680628348465387e-09,
+      "loss": 0.006,
+      "step": 27090
+    },
+    {
+      "epoch": 0.9787633631898295,
+      "grad_norm": 0.2953276902191807,
+      "learning_rate": 3.5588099415981847e-09,
+      "loss": 0.0063,
+      "step": 27100
+    },
+    {
+      "epoch": 0.9791245304825195,
+      "grad_norm": 0.3263557425314221,
+      "learning_rate": 3.439039131102939e-09,
+      "loss": 0.0061,
+      "step": 27110
+    },
+    {
+      "epoch": 0.9794856977752094,
+      "grad_norm": 0.25968166795522196,
+      "learning_rate": 3.3213160808630573e-09,
+      "loss": 0.0059,
+      "step": 27120
+    },
+    {
+      "epoch": 0.9798468650678994,
+      "grad_norm": 0.25820698619691657,
+      "learning_rate": 3.205640951960187e-09,
+      "loss": 0.0059,
+      "step": 27130
+    },
+    {
+      "epoch": 0.9802080323605894,
+      "grad_norm": 0.24485375570092602,
+      "learning_rate": 3.092013902674218e-09,
+      "loss": 0.0062,
+      "step": 27140
+    },
+    {
+      "epoch": 0.9805691996532794,
+      "grad_norm": 0.2870640182619394,
+      "learning_rate": 2.980435088481781e-09,
+      "loss": 0.0063,
+      "step": 27150
+    },
+    {
+      "epoch": 0.9809303669459694,
+      "grad_norm": 0.26934438711180547,
+      "learning_rate": 2.870904662057583e-09,
+      "loss": 0.0062,
+      "step": 27160
+    },
+    {
+      "epoch": 0.9812915342386593,
+      "grad_norm": 0.29849074194942216,
+      "learning_rate": 2.7634227732732386e-09,
+      "loss": 0.0064,
+      "step": 27170
+    },
+    {
+      "epoch": 0.9816527015313493,
+      "grad_norm": 0.3750188409684099,
+      "learning_rate": 2.6579895691974385e-09,
+      "loss": 0.0064,
+      "step": 27180
+    },
+    {
+      "epoch": 0.9820138688240393,
+      "grad_norm": 0.23807013477079472,
+      "learning_rate": 2.5546051940954494e-09,
+      "loss": 0.0063,
+      "step": 27190
+    },
+    {
+      "epoch": 0.9823750361167293,
+      "grad_norm": 0.19466251978266672,
+      "learning_rate": 2.4532697894291133e-09,
+      "loss": 0.006,
+      "step": 27200
+    },
+    {
+      "epoch": 0.9827362034094193,
+      "grad_norm": 0.2936000992609088,
+      "learning_rate": 2.3539834938570148e-09,
+      "loss": 0.0062,
+      "step": 27210
+    },
+    {
+      "epoch": 0.9830973707021092,
+      "grad_norm": 0.2930293795651421,
+      "learning_rate": 2.2567464432334816e-09,
+      "loss": 0.0061,
+      "step": 27220
+    },
+    {
+      "epoch": 0.9834585379947992,
+      "grad_norm": 0.2435920136526976,
+      "learning_rate": 2.1615587706090843e-09,
+      "loss": 0.006,
+      "step": 27230
+    },
+    {
+      "epoch": 0.9838197052874892,
+      "grad_norm": 0.2317523584884122,
+      "learning_rate": 2.0684206062299703e-09,
+      "loss": 0.0059,
+      "step": 27240
+    },
+    {
+      "epoch": 0.9841808725801792,
+      "grad_norm": 0.30495249051065004,
+      "learning_rate": 1.9773320775381963e-09,
+      "loss": 0.0062,
+      "step": 27250
+    },
+    {
+      "epoch": 0.9845420398728691,
+      "grad_norm": 0.2541561889630849,
+      "learning_rate": 1.888293309171063e-09,
+      "loss": 0.0063,
+      "step": 27260
+    },
+    {
+      "epoch": 0.9849032071655591,
+      "grad_norm": 0.31704233070439763,
+      "learning_rate": 1.8013044229614472e-09,
+      "loss": 0.0061,
+      "step": 27270
+    },
+    {
+      "epoch": 0.9852643744582491,
+      "grad_norm": 0.3043590860702212,
+      "learning_rate": 1.7163655379369703e-09,
+      "loss": 0.0063,
+      "step": 27280
+    },
+    {
+      "epoch": 0.9856255417509391,
+      "grad_norm": 0.2861333369314964,
+      "learning_rate": 1.6334767703206633e-09,
+      "loss": 0.0057,
+      "step": 27290
+    },
+    {
+      "epoch": 0.9859867090436291,
+      "grad_norm": 0.23033516670300275,
+      "learning_rate": 1.5526382335298017e-09,
+      "loss": 0.0061,
+      "step": 27300
+    },
+    {
+      "epoch": 0.9863478763363189,
+      "grad_norm": 0.33865484832512566,
+      "learning_rate": 1.4738500381767384e-09,
+      "loss": 0.0068,
+      "step": 27310
+    },
+    {
+      "epoch": 0.9867090436290089,
+      "grad_norm": 0.35894436053392437,
+      "learning_rate": 1.3971122920682367e-09,
+      "loss": 0.0061,
+      "step": 27320
+    },
+    {
+      "epoch": 0.9870702109216989,
+      "grad_norm": 0.27375652323454114,
+      "learning_rate": 1.3224251002053044e-09,
+      "loss": 0.0062,
+      "step": 27330
+    },
+    {
+      "epoch": 0.9874313782143889,
+      "grad_norm": 0.23085466890453998,
+      "learning_rate": 1.2497885647833608e-09,
+      "loss": 0.0055,
+      "step": 27340
+    },
+    {
+      "epoch": 0.9877925455070788,
+      "grad_norm": 0.2385584040045275,
+      "learning_rate": 1.1792027851915687e-09,
+      "loss": 0.0057,
+      "step": 27350
+    },
+    {
+      "epoch": 0.9881537127997688,
+      "grad_norm": 0.27744659053253545,
+      "learning_rate": 1.1106678580133368e-09,
+      "loss": 0.006,
+      "step": 27360
+    },
+    {
+      "epoch": 0.9885148800924588,
+      "grad_norm": 0.2578337452073481,
+      "learning_rate": 1.0441838770256508e-09,
+      "loss": 0.006,
+      "step": 27370
+    },
+    {
+      "epoch": 0.9888760473851488,
+      "grad_norm": 0.34775475755904656,
+      "learning_rate": 9.797509331992416e-10,
+      "loss": 0.0061,
+      "step": 27380
+    },
+    {
+      "epoch": 0.9892372146778388,
+      "grad_norm": 0.28713869693512956,
+      "learning_rate": 9.173691146984186e-10,
+      "loss": 0.0059,
+      "step": 27390
+    },
+    {
+      "epoch": 0.9895983819705287,
+      "grad_norm": 0.33610215505013186,
+      "learning_rate": 8.570385068810693e-10,
+      "loss": 0.0066,
+      "step": 27400
+    },
+    {
+      "epoch": 0.9899595492632187,
+      "grad_norm": 0.32368146949710996,
+      "learning_rate": 7.987591922978265e-10,
+      "loss": 0.0063,
+      "step": 27410
+    },
+    {
+      "epoch": 0.9903207165559087,
+      "grad_norm": 0.4191928285798203,
+      "learning_rate": 7.425312506930682e-10,
+      "loss": 0.0066,
+      "step": 27420
+    },
+    {
+      "epoch": 0.9906818838485987,
+      "grad_norm": 0.29350261366870306,
+      "learning_rate": 6.883547590042505e-10,
+      "loss": 0.0061,
+      "step": 27430
+    },
+    {
+      "epoch": 0.9910430511412887,
+      "grad_norm": 0.2724927235562543,
+      "learning_rate": 6.36229791361409e-10,
+      "loss": 0.0059,
+      "step": 27440
+    },
+    {
+      "epoch": 0.9914042184339786,
+      "grad_norm": 0.3102234720137091,
+      "learning_rate": 5.861564190878244e-10,
+      "loss": 0.0062,
+      "step": 27450
+    },
+    {
+      "epoch": 0.9917653857266686,
+      "grad_norm": 0.26742780489071954,
+      "learning_rate": 5.381347106995227e-10,
+      "loss": 0.0064,
+      "step": 27460
+    },
+    {
+      "epoch": 0.9921265530193586,
+      "grad_norm": 0.555835624701209,
+      "learning_rate": 4.921647319047762e-10,
+      "loss": 0.0062,
+      "step": 27470
+    },
+    {
+      "epoch": 0.9924877203120486,
+      "grad_norm": 0.29284120191196367,
+      "learning_rate": 4.4824654560510215e-10,
+      "loss": 0.0059,
+      "step": 27480
+    },
+    {
+      "epoch": 0.9928488876047386,
+      "grad_norm": 0.2927165042276575,
+      "learning_rate": 4.0638021189409735e-10,
+      "loss": 0.0063,
+      "step": 27490
+    },
+    {
+      "epoch": 0.9932100548974285,
+      "grad_norm": 0.26925222058034637,
+      "learning_rate": 3.665657880577711e-10,
+      "loss": 0.0064,
+      "step": 27500
+    },
+    {
+      "epoch": 0.9935712221901185,
+      "grad_norm": 0.3308436653516931,
+      "learning_rate": 3.288033285750447e-10,
+      "loss": 0.0064,
+      "step": 27510
+    },
+    {
+      "epoch": 0.9939323894828085,
+      "grad_norm": 0.267767776056638,
+      "learning_rate": 2.930928851162529e-10,
+      "loss": 0.0066,
+      "step": 27520
+    },
+    {
+      "epoch": 0.9942935567754985,
+      "grad_norm": 0.25752337290476124,
+      "learning_rate": 2.594345065446424e-10,
+      "loss": 0.0058,
+      "step": 27530
+    },
+    {
+      "epoch": 0.9946547240681883,
+      "grad_norm": 0.2391574952205953,
+      "learning_rate": 2.2782823891520644e-10,
+      "loss": 0.0061,
+      "step": 27540
+    },
+    {
+      "epoch": 0.9950158913608783,
+      "grad_norm": 0.3946297832016757,
+      "learning_rate": 1.9827412547501757e-10,
+      "loss": 0.0062,
+      "step": 27550
+    },
+    {
+      "epoch": 0.9953770586535683,
+      "grad_norm": 0.43167343106215705,
+      "learning_rate": 1.7077220666356086e-10,
+      "loss": 0.0063,
+      "step": 27560
+    },
+    {
+      "epoch": 0.9957382259462583,
+      "grad_norm": 0.29501813312344083,
+      "learning_rate": 1.4532252011190128e-10,
+      "loss": 0.006,
+      "step": 27570
+    },
+    {
+      "epoch": 0.9960993932389482,
+      "grad_norm": 0.26783465825372643,
+      "learning_rate": 1.2192510064318318e-10,
+      "loss": 0.0066,
+      "step": 27580
+    },
+    {
+      "epoch": 0.9964605605316382,
+      "grad_norm": 0.2944539759000121,
+      "learning_rate": 1.0057998027213078e-10,
+      "loss": 0.0063,
+      "step": 27590
+    },
+    {
+      "epoch": 0.9968217278243282,
+      "grad_norm": 0.3615616193630821,
+      "learning_rate": 8.12871882055477e-11,
+      "loss": 0.0061,
+      "step": 27600
+    },
+    {
+      "epoch": 0.9971828951170182,
+      "grad_norm": 0.2461908385757508,
+      "learning_rate": 6.40467508421505e-11,
+      "loss": 0.0058,
+      "step": 27610
+    },
+    {
+      "epoch": 0.9975440624097082,
+      "grad_norm": 0.28485635752486593,
+      "learning_rate": 4.885869177206903e-11,
+      "loss": 0.0065,
+      "step": 27620
+    },
+    {
+      "epoch": 0.9979052297023981,
+      "grad_norm": 0.29603954527977433,
+      "learning_rate": 3.5723031777346036e-11,
+      "loss": 0.0069,
+      "step": 27630
+    },
+    {
+      "epoch": 0.9982663969950881,
+      "grad_norm": 0.20904298664543156,
+      "learning_rate": 2.4639788831604115e-11,
+      "loss": 0.0062,
+      "step": 27640
+    },
+    {
+      "epoch": 0.9986275642877781,
+      "grad_norm": 0.6523198761223433,
+      "learning_rate": 1.5608978100212223e-11,
+      "loss": 0.0063,
+      "step": 27650
+    },
+    {
+      "epoch": 0.9989887315804681,
+      "grad_norm": 0.3097053400119001,
+      "learning_rate": 8.630611939952626e-12,
+      "loss": 0.0063,
+      "step": 27660
+    },
+    {
+      "epoch": 0.9993498988731581,
+      "grad_norm": 0.2623921087140999,
+      "learning_rate": 3.704699899687025e-12,
+      "loss": 0.0061,
+      "step": 27670
+    },
+    {
+      "epoch": 0.999711066165848,
+      "grad_norm": 0.24690038783500173,
+      "learning_rate": 8.312487193573582e-13,
+      "loss": 0.0064,
+      "step": 27680
+    },
+    {
+      "epoch": 1.0,
+      "full_eval_loss": 0.28296777606010437,
+      "full_eval_runtime": 8273.1503,
+      "full_eval_samples_per_second": 187.878,
+      "full_eval_steps_per_second": 0.367,
+      "step": 27688
+    },
+    {
+      "epoch": 1.0,
+      "step": 27688,
+      "total_flos": 1.0350945330266112e+17,
+      "train_loss": 0.01862134516755282,
+      "train_runtime": 213167.2265,
+      "train_samples_per_second": 66.502,
+      "train_steps_per_second": 0.13
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 27688,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 5000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.0350945330266112e+17,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}