diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,47480 @@
+{
+  "best_metric": 0.9356,
+  "best_model_checkpoint": "checkpoint/vit-large/checkpoint-11970",
+  "epoch": 100.0,
+  "eval_steps": 500,
+  "global_step": 66500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7020020484924316,
+      "learning_rate": 9.998496240601504e-06,
+      "loss": 4.6119,
+      "step": 10
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.892003059387207,
+      "learning_rate": 9.996992481203008e-06,
+      "loss": 4.5669,
+      "step": 20
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7313311100006104,
+      "learning_rate": 9.995488721804511e-06,
+      "loss": 4.5263,
+      "step": 30
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.659421443939209,
+      "learning_rate": 9.993984962406017e-06,
+      "loss": 4.4729,
+      "step": 40
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.846480131149292,
+      "learning_rate": 9.992481203007518e-06,
+      "loss": 4.3985,
+      "step": 50
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0880706310272217,
+      "learning_rate": 9.990977443609024e-06,
+      "loss": 4.3247,
+      "step": 60
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.202064037322998,
+      "learning_rate": 9.989473684210527e-06,
+      "loss": 4.233,
+      "step": 70
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.243076324462891,
+      "learning_rate": 9.98796992481203e-06,
+      "loss": 4.1598,
+      "step": 80
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.47965669631958,
+      "learning_rate": 9.986466165413534e-06,
+      "loss": 4.0929,
+      "step": 90
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.431342363357544,
+      "learning_rate": 9.984962406015038e-06,
+      "loss": 3.9735,
+      "step": 100
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.915376901626587,
+      "learning_rate": 9.983458646616541e-06,
+      "loss": 3.9051,
+      "step": 110
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6235947608947754,
+      "learning_rate": 9.981954887218046e-06,
+      "loss": 3.7795,
+      "step": 120
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.0705485343933105,
+      "learning_rate": 9.98045112781955e-06,
+      "loss": 3.6517,
+      "step": 130
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.916447162628174,
+      "learning_rate": 9.978947368421053e-06,
+      "loss": 3.5197,
+      "step": 140
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.508981227874756,
+      "learning_rate": 9.977443609022557e-06,
+      "loss": 3.4313,
+      "step": 150
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.508138179779053,
+      "learning_rate": 9.97593984962406e-06,
+      "loss": 3.3823,
+      "step": 160
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.037680625915527,
+      "learning_rate": 9.974436090225564e-06,
+      "loss": 3.2748,
+      "step": 170
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.304414749145508,
+      "learning_rate": 9.97293233082707e-06,
+      "loss": 3.1054,
+      "step": 180
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.1311421394348145,
+      "learning_rate": 9.971428571428571e-06,
+      "loss": 3.0703,
+      "step": 190
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.8540120124816895,
+      "learning_rate": 9.969924812030076e-06,
+      "loss": 2.9789,
+      "step": 200
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.647185802459717,
+      "learning_rate": 9.96842105263158e-06,
+      "loss": 2.8568,
+      "step": 210
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.943347454071045,
+      "learning_rate": 9.966917293233083e-06,
+      "loss": 2.8175,
+      "step": 220
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 17.15534210205078,
+      "learning_rate": 9.965413533834587e-06,
+      "loss": 2.7049,
+      "step": 230
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.327253341674805,
+      "learning_rate": 9.963909774436092e-06,
+      "loss": 2.6185,
+      "step": 240
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.56367301940918,
+      "learning_rate": 9.962406015037594e-06,
+      "loss": 2.4704,
+      "step": 250
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.038745880126953,
+      "learning_rate": 9.960902255639099e-06,
+      "loss": 2.4862,
+      "step": 260
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.661726474761963,
+      "learning_rate": 9.959398496240603e-06,
+      "loss": 2.4405,
+      "step": 270
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.565981864929199,
+      "learning_rate": 9.957894736842106e-06,
+      "loss": 2.3698,
+      "step": 280
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.349733829498291,
+      "learning_rate": 9.95639097744361e-06,
+      "loss": 2.3338,
+      "step": 290
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.526618957519531,
+      "learning_rate": 9.954887218045113e-06,
+      "loss": 2.2445,
+      "step": 300
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.558746337890625,
+      "learning_rate": 9.953383458646617e-06,
+      "loss": 2.202,
+      "step": 310
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.4157633781433105,
+      "learning_rate": 9.951879699248122e-06,
+      "loss": 2.1578,
+      "step": 320
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.635522842407227,
+      "learning_rate": 9.950375939849625e-06,
+      "loss": 2.099,
+      "step": 330
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.216004371643066,
+      "learning_rate": 9.948872180451129e-06,
+      "loss": 2.0472,
+      "step": 340
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.348927021026611,
+      "learning_rate": 9.947368421052632e-06,
+      "loss": 2.0272,
+      "step": 350
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.218992710113525,
+      "learning_rate": 9.945864661654136e-06,
+      "loss": 1.9022,
+      "step": 360
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 12.379638671875,
+      "learning_rate": 9.94436090225564e-06,
+      "loss": 1.9707,
+      "step": 370
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.454248905181885,
+      "learning_rate": 9.942857142857145e-06,
+      "loss": 1.9612,
+      "step": 380
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.975961685180664,
+      "learning_rate": 9.941353383458647e-06,
+      "loss": 1.8519,
+      "step": 390
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 11.839798927307129,
+      "learning_rate": 9.939849624060152e-06,
+      "loss": 1.8246,
+      "step": 400
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 21.542709350585938,
+      "learning_rate": 9.938345864661655e-06,
+      "loss": 1.7591,
+      "step": 410
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.640402793884277,
+      "learning_rate": 9.936842105263159e-06,
+      "loss": 1.7803,
+      "step": 420
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.312070369720459,
+      "learning_rate": 9.935338345864662e-06,
+      "loss": 1.6758,
+      "step": 430
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.633362770080566,
+      "learning_rate": 9.933834586466168e-06,
+      "loss": 1.7095,
+      "step": 440
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.548731803894043,
+      "learning_rate": 9.93233082706767e-06,
+      "loss": 1.6934,
+      "step": 450
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.280405044555664,
+      "learning_rate": 9.930827067669175e-06,
+      "loss": 1.74,
+      "step": 460
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.683008670806885,
+      "learning_rate": 9.929323308270678e-06,
+      "loss": 1.5803,
+      "step": 470
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.887816905975342,
+      "learning_rate": 9.927819548872182e-06,
+      "loss": 1.6039,
+      "step": 480
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.674352645874023,
+      "learning_rate": 9.926315789473685e-06,
+      "loss": 1.6009,
+      "step": 490
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.968412399291992,
+      "learning_rate": 9.924812030075189e-06,
+      "loss": 1.5391,
+      "step": 500
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.869007587432861,
+      "learning_rate": 9.923308270676692e-06,
+      "loss": 1.5226,
+      "step": 510
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.70728874206543,
+      "learning_rate": 9.921804511278196e-06,
+      "loss": 1.6173,
+      "step": 520
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.701499938964844,
+      "learning_rate": 9.920300751879701e-06,
+      "loss": 1.5042,
+      "step": 530
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.880636692047119,
+      "learning_rate": 9.918796992481203e-06,
+      "loss": 1.456,
+      "step": 540
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.021149635314941,
+      "learning_rate": 9.917293233082708e-06,
+      "loss": 1.5698,
+      "step": 550
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.1271138191223145,
+      "learning_rate": 9.915789473684211e-06,
+      "loss": 1.3838,
+      "step": 560
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.8424482345581055,
+      "learning_rate": 9.914285714285715e-06,
+      "loss": 1.3997,
+      "step": 570
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.099345207214355,
+      "learning_rate": 9.912781954887218e-06,
+      "loss": 1.4359,
+      "step": 580
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.261110782623291,
+      "learning_rate": 9.911278195488722e-06,
+      "loss": 1.454,
+      "step": 590
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.30597448348999,
+      "learning_rate": 9.909774436090226e-06,
+      "loss": 1.3907,
+      "step": 600
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.466990947723389,
+      "learning_rate": 9.90827067669173e-06,
+      "loss": 1.329,
+      "step": 610
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.6469316482543945,
+      "learning_rate": 9.906766917293234e-06,
+      "loss": 1.2848,
+      "step": 620
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.2480244636535645,
+      "learning_rate": 9.905263157894738e-06,
+      "loss": 1.4246,
+      "step": 630
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.738135814666748,
+      "learning_rate": 9.903759398496241e-06,
+      "loss": 1.3669,
+      "step": 640
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.298168182373047,
+      "learning_rate": 9.902255639097745e-06,
+      "loss": 1.2883,
+      "step": 650
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.747694969177246,
+      "learning_rate": 9.900751879699248e-06,
+      "loss": 1.2884,
+      "step": 660
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.8834,
+      "eval_loss": 0.8751662373542786,
+      "eval_runtime": 85.9745,
+      "eval_samples_per_second": 116.314,
+      "eval_steps_per_second": 0.465,
+      "step": 665
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 7.908563137054443,
+      "learning_rate": 9.899248120300754e-06,
+      "loss": 1.2429,
+      "step": 670
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.439415454864502,
+      "learning_rate": 9.897744360902255e-06,
+      "loss": 1.2716,
+      "step": 680
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 7.2306599617004395,
+      "learning_rate": 9.89624060150376e-06,
+      "loss": 1.1676,
+      "step": 690
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 6.287716388702393,
+      "learning_rate": 9.894736842105264e-06,
+      "loss": 1.2212,
+      "step": 700
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.363208770751953,
+      "learning_rate": 9.893233082706768e-06,
+      "loss": 1.259,
+      "step": 710
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.556034564971924,
+      "learning_rate": 9.891729323308271e-06,
+      "loss": 1.1383,
+      "step": 720
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 5.640949726104736,
+      "learning_rate": 9.890225563909776e-06,
+      "loss": 1.2454,
+      "step": 730
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 7.140594959259033,
+      "learning_rate": 9.888721804511278e-06,
+      "loss": 1.1363,
+      "step": 740
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.780942440032959,
+      "learning_rate": 9.887218045112783e-06,
+      "loss": 1.1806,
+      "step": 750
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 6.301698684692383,
+      "learning_rate": 9.885714285714287e-06,
+      "loss": 1.1416,
+      "step": 760
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.256317138671875,
+      "learning_rate": 9.88421052631579e-06,
+      "loss": 1.1947,
+      "step": 770
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.818164825439453,
+      "learning_rate": 9.882706766917294e-06,
+      "loss": 1.1865,
+      "step": 780
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 11.159587860107422,
+      "learning_rate": 9.881203007518797e-06,
+      "loss": 1.0911,
+      "step": 790
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.432153701782227,
+      "learning_rate": 9.879699248120301e-06,
+      "loss": 1.1486,
+      "step": 800
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.031283378601074,
+      "learning_rate": 9.878195488721806e-06,
+      "loss": 1.0838,
+      "step": 810
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 7.850508689880371,
+      "learning_rate": 9.87669172932331e-06,
+      "loss": 1.0953,
+      "step": 820
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 6.112914562225342,
+      "learning_rate": 9.875187969924813e-06,
+      "loss": 1.1159,
+      "step": 830
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 6.945899486541748,
+      "learning_rate": 9.873684210526317e-06,
+      "loss": 0.9613,
+      "step": 840
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 7.430254936218262,
+      "learning_rate": 9.87218045112782e-06,
+      "loss": 1.086,
+      "step": 850
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 6.00394868850708,
+      "learning_rate": 9.870676691729324e-06,
+      "loss": 1.0873,
+      "step": 860
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 5.864083290100098,
+      "learning_rate": 9.869172932330829e-06,
+      "loss": 1.055,
+      "step": 870
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.37553882598877,
+      "learning_rate": 9.86766917293233e-06,
+      "loss": 1.0894,
+      "step": 880
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 8.302824974060059,
+      "learning_rate": 9.866165413533836e-06,
+      "loss": 1.1037,
+      "step": 890
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 6.968749046325684,
+      "learning_rate": 9.86466165413534e-06,
+      "loss": 1.0071,
+      "step": 900
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 5.924696922302246,
+      "learning_rate": 9.863157894736843e-06,
+      "loss": 1.0498,
+      "step": 910
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.433680534362793,
+      "learning_rate": 9.861654135338347e-06,
+      "loss": 1.0456,
+      "step": 920
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.614802360534668,
+      "learning_rate": 9.86015037593985e-06,
+      "loss": 0.9659,
+      "step": 930
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.205560684204102,
+      "learning_rate": 9.858646616541354e-06,
+      "loss": 0.9692,
+      "step": 940
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 6.741930961608887,
+      "learning_rate": 9.857142857142859e-06,
+      "loss": 1.0157,
+      "step": 950
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.176901817321777,
+      "learning_rate": 9.855639097744362e-06,
+      "loss": 1.0014,
+      "step": 960
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 5.335792541503906,
+      "learning_rate": 9.854135338345866e-06,
+      "loss": 0.9361,
+      "step": 970
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 6.488396644592285,
+      "learning_rate": 9.85263157894737e-06,
+      "loss": 1.0858,
+      "step": 980
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 6.725528240203857,
+      "learning_rate": 9.851127819548873e-06,
+      "loss": 1.0027,
+      "step": 990
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.740559577941895,
+      "learning_rate": 9.849624060150376e-06,
+      "loss": 1.0121,
+      "step": 1000
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.048620223999023,
+      "learning_rate": 9.84812030075188e-06,
+      "loss": 0.967,
+      "step": 1010
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.225861072540283,
+      "learning_rate": 9.846616541353383e-06,
+      "loss": 1.0329,
+      "step": 1020
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 7.12366247177124,
+      "learning_rate": 9.845112781954887e-06,
+      "loss": 1.0117,
+      "step": 1030
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.5486273765563965,
+      "learning_rate": 9.843609022556392e-06,
+      "loss": 0.969,
+      "step": 1040
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.591785430908203,
+      "learning_rate": 9.842105263157896e-06,
+      "loss": 0.9415,
+      "step": 1050
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 7.203570365905762,
+      "learning_rate": 9.8406015037594e-06,
+      "loss": 1.0409,
+      "step": 1060
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.412242889404297,
+      "learning_rate": 9.839097744360903e-06,
+      "loss": 1.0352,
+      "step": 1070
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.688934326171875,
+      "learning_rate": 9.837593984962406e-06,
+      "loss": 1.0524,
+      "step": 1080
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 6.9523844718933105,
+      "learning_rate": 9.83609022556391e-06,
+      "loss": 0.9546,
+      "step": 1090
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.400866508483887,
+      "learning_rate": 9.834586466165415e-06,
+      "loss": 0.9361,
+      "step": 1100
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.09070110321045,
+      "learning_rate": 9.833082706766917e-06,
+      "loss": 0.9797,
+      "step": 1110
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 6.927423000335693,
+      "learning_rate": 9.831578947368422e-06,
+      "loss": 0.9439,
+      "step": 1120
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.291294574737549,
+      "learning_rate": 9.830075187969926e-06,
+      "loss": 0.9753,
+      "step": 1130
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 7.919008731842041,
+      "learning_rate": 9.828571428571429e-06,
+      "loss": 0.9697,
+      "step": 1140
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.660476684570312,
+      "learning_rate": 9.827067669172933e-06,
+      "loss": 0.8882,
+      "step": 1150
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.102679252624512,
+      "learning_rate": 9.825563909774438e-06,
+      "loss": 0.8959,
+      "step": 1160
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.902896404266357,
+      "learning_rate": 9.82406015037594e-06,
+      "loss": 0.887,
+      "step": 1170
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.6904778480529785,
+      "learning_rate": 9.822556390977445e-06,
+      "loss": 0.9352,
+      "step": 1180
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 6.770270824432373,
+      "learning_rate": 9.821052631578948e-06,
+      "loss": 0.8916,
+      "step": 1190
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.353099822998047,
+      "learning_rate": 9.819548872180452e-06,
+      "loss": 0.9523,
+      "step": 1200
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 6.385773658752441,
+      "learning_rate": 9.818045112781955e-06,
+      "loss": 0.9175,
+      "step": 1210
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 13.28996467590332,
+      "learning_rate": 9.816541353383459e-06,
+      "loss": 0.9375,
+      "step": 1220
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.252169609069824,
+      "learning_rate": 9.815037593984962e-06,
+      "loss": 1.0589,
+      "step": 1230
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 6.009567737579346,
+      "learning_rate": 9.813533834586468e-06,
+      "loss": 0.9608,
+      "step": 1240
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 4.863635063171387,
+      "learning_rate": 9.812030075187971e-06,
+      "loss": 0.9724,
+      "step": 1250
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.548372268676758,
+      "learning_rate": 9.810526315789475e-06,
+      "loss": 0.9651,
+      "step": 1260
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 8.277862548828125,
+      "learning_rate": 9.809022556390978e-06,
+      "loss": 0.8868,
+      "step": 1270
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.657036304473877,
+      "learning_rate": 9.807518796992482e-06,
+      "loss": 0.9133,
+      "step": 1280
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 7.065949440002441,
+      "learning_rate": 9.806015037593985e-06,
+      "loss": 0.9602,
+      "step": 1290
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.187036514282227,
+      "learning_rate": 9.80451127819549e-06,
+      "loss": 0.9626,
+      "step": 1300
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 7.014963150024414,
+      "learning_rate": 9.803007518796992e-06,
+      "loss": 0.8653,
+      "step": 1310
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.665754318237305,
+      "learning_rate": 9.801503759398498e-06,
+      "loss": 0.8795,
+      "step": 1320
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.92686939239502,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 0.7958,
+      "step": 1330
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.9142,
+      "eval_loss": 0.4723776876926422,
+      "eval_runtime": 84.6568,
+      "eval_samples_per_second": 118.124,
+      "eval_steps_per_second": 0.472,
+      "step": 1330
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 6.881080627441406,
+      "learning_rate": 9.798496240601505e-06,
+      "loss": 0.9416,
+      "step": 1340
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.449582099914551,
+      "learning_rate": 9.796992481203008e-06,
+      "loss": 0.8897,
+      "step": 1350
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 6.535237789154053,
+      "learning_rate": 9.795488721804513e-06,
+      "loss": 0.8241,
+      "step": 1360
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 7.858671188354492,
+      "learning_rate": 9.793984962406015e-06,
+      "loss": 0.7834,
+      "step": 1370
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 6.7106852531433105,
+      "learning_rate": 9.79248120300752e-06,
+      "loss": 0.836,
+      "step": 1380
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 6.440729141235352,
+      "learning_rate": 9.790977443609024e-06,
+      "loss": 0.775,
+      "step": 1390
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 8.935519218444824,
+      "learning_rate": 9.789473684210527e-06,
+      "loss": 0.7826,
+      "step": 1400
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 7.244870662689209,
+      "learning_rate": 9.787969924812031e-06,
+      "loss": 0.8917,
+      "step": 1410
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 9.125387191772461,
+      "learning_rate": 9.786466165413534e-06,
+      "loss": 0.8167,
+      "step": 1420
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 6.8634114265441895,
+      "learning_rate": 9.784962406015038e-06,
+      "loss": 0.8052,
+      "step": 1430
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 7.519056797027588,
+      "learning_rate": 9.783458646616543e-06,
+      "loss": 0.8204,
+      "step": 1440
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 7.861953258514404,
+      "learning_rate": 9.781954887218047e-06,
+      "loss": 0.8404,
+      "step": 1450
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 10.940001487731934,
+      "learning_rate": 9.78045112781955e-06,
+      "loss": 0.7934,
+      "step": 1460
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.963690757751465,
+      "learning_rate": 9.778947368421054e-06,
+      "loss": 0.7549,
+      "step": 1470
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 6.734865665435791,
+      "learning_rate": 9.777443609022557e-06,
+      "loss": 0.8112,
+      "step": 1480
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 7.971401214599609,
+      "learning_rate": 9.77593984962406e-06,
+      "loss": 0.8323,
+      "step": 1490
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 9.728713989257812,
+      "learning_rate": 9.774436090225564e-06,
+      "loss": 0.7441,
+      "step": 1500
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 9.506553649902344,
+      "learning_rate": 9.772932330827068e-06,
+      "loss": 0.8018,
+      "step": 1510
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 7.3224663734436035,
+      "learning_rate": 9.771428571428571e-06,
+      "loss": 0.7869,
+      "step": 1520
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 7.251104831695557,
+      "learning_rate": 9.769924812030077e-06,
+      "loss": 0.8472,
+      "step": 1530
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 7.3015055656433105,
+      "learning_rate": 9.76842105263158e-06,
+      "loss": 0.7224,
+      "step": 1540
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 9.096901893615723,
+      "learning_rate": 9.766917293233084e-06,
+      "loss": 0.8195,
+      "step": 1550
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 7.366261959075928,
+      "learning_rate": 9.765413533834587e-06,
+      "loss": 0.8322,
+      "step": 1560
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 9.955854415893555,
+      "learning_rate": 9.76390977443609e-06,
+      "loss": 0.8315,
+      "step": 1570
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.852784633636475,
+      "learning_rate": 9.762406015037594e-06,
+      "loss": 0.7113,
+      "step": 1580
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 9.9766263961792,
+      "learning_rate": 9.7609022556391e-06,
+      "loss": 0.8024,
+      "step": 1590
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 9.095175743103027,
+      "learning_rate": 9.759398496240601e-06,
+      "loss": 0.8774,
+      "step": 1600
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.913175106048584,
+      "learning_rate": 9.757894736842106e-06,
+      "loss": 0.7608,
+      "step": 1610
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 9.544361114501953,
+      "learning_rate": 9.75639097744361e-06,
+      "loss": 0.7984,
+      "step": 1620
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 6.991225242614746,
+      "learning_rate": 9.754887218045113e-06,
+      "loss": 0.7942,
+      "step": 1630
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 7.531531810760498,
+      "learning_rate": 9.753383458646617e-06,
+      "loss": 0.8005,
+      "step": 1640
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.945763111114502,
+      "learning_rate": 9.751879699248122e-06,
+      "loss": 0.7673,
+      "step": 1650
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 8.382121086120605,
+      "learning_rate": 9.750375939849624e-06,
+      "loss": 0.7966,
+      "step": 1660
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.387685775756836,
+      "learning_rate": 9.74887218045113e-06,
+      "loss": 0.7892,
+      "step": 1670
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 7.867427349090576,
+      "learning_rate": 9.747368421052633e-06,
+      "loss": 0.8002,
+      "step": 1680
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 7.549880027770996,
+      "learning_rate": 9.745864661654136e-06,
+      "loss": 0.8879,
+      "step": 1690
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 7.67978572845459,
+      "learning_rate": 9.74436090225564e-06,
+      "loss": 0.7849,
+      "step": 1700
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 8.076873779296875,
+      "learning_rate": 9.742857142857143e-06,
+      "loss": 0.6503,
+      "step": 1710
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 10.748533248901367,
+      "learning_rate": 9.741353383458647e-06,
+      "loss": 0.7868,
+      "step": 1720
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 8.428750991821289,
+      "learning_rate": 9.739849624060152e-06,
+      "loss": 0.8195,
+      "step": 1730
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 7.678562164306641,
+      "learning_rate": 9.738345864661655e-06,
+      "loss": 0.8428,
+      "step": 1740
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 7.171645164489746,
+      "learning_rate": 9.736842105263159e-06,
+      "loss": 0.809,
+      "step": 1750
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 7.041049003601074,
+      "learning_rate": 9.735338345864663e-06,
+      "loss": 0.7417,
+      "step": 1760
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 9.66743278503418,
+      "learning_rate": 9.733834586466166e-06,
+      "loss": 0.7952,
+      "step": 1770
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 9.864920616149902,
+      "learning_rate": 9.73233082706767e-06,
+      "loss": 0.8427,
+      "step": 1780
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 10.242929458618164,
+      "learning_rate": 9.730827067669175e-06,
+      "loss": 0.7599,
+      "step": 1790
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 8.138999938964844,
+      "learning_rate": 9.729323308270677e-06,
+      "loss": 0.7517,
+      "step": 1800
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 7.668764114379883,
+      "learning_rate": 9.727819548872182e-06,
+      "loss": 0.7766,
+      "step": 1810
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 6.978646278381348,
+      "learning_rate": 9.726315789473685e-06,
+      "loss": 0.7323,
+      "step": 1820
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 8.794787406921387,
+      "learning_rate": 9.724812030075189e-06,
+      "loss": 0.8644,
+      "step": 1830
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 10.154306411743164,
+      "learning_rate": 9.723308270676692e-06,
+      "loss": 0.8235,
+      "step": 1840
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 9.513362884521484,
+      "learning_rate": 9.721804511278196e-06,
+      "loss": 0.792,
+      "step": 1850
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.891651630401611,
+      "learning_rate": 9.7203007518797e-06,
+      "loss": 0.7343,
+      "step": 1860
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 6.595260143280029,
+      "learning_rate": 9.718796992481205e-06,
+      "loss": 0.6451,
+      "step": 1870
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 11.115670204162598,
+      "learning_rate": 9.717293233082708e-06,
+      "loss": 0.7841,
+      "step": 1880
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 7.82785701751709,
+      "learning_rate": 9.715789473684212e-06,
+      "loss": 0.8396,
+      "step": 1890
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 5.41937780380249,
+      "learning_rate": 9.714285714285715e-06,
+      "loss": 0.772,
+      "step": 1900
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 8.092954635620117,
+      "learning_rate": 9.712781954887219e-06,
+      "loss": 0.7124,
+      "step": 1910
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.913546562194824,
+      "learning_rate": 9.711278195488722e-06,
+      "loss": 0.7824,
+      "step": 1920
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 6.090660572052002,
+      "learning_rate": 9.709774436090227e-06,
+      "loss": 0.7911,
+      "step": 1930
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 5.547027111053467,
+      "learning_rate": 9.70827067669173e-06,
+      "loss": 0.7818,
+      "step": 1940
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 8.583475112915039,
+      "learning_rate": 9.706766917293234e-06,
+      "loss": 0.7272,
+      "step": 1950
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 8.129578590393066,
+      "learning_rate": 9.705263157894738e-06,
+      "loss": 0.7352,
+      "step": 1960
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 9.513014793395996,
+      "learning_rate": 9.703759398496242e-06,
+      "loss": 0.8126,
+      "step": 1970
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 5.819597244262695,
+      "learning_rate": 9.702255639097745e-06,
+      "loss": 0.7599,
+      "step": 1980
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 7.391184329986572,
+      "learning_rate": 9.700751879699249e-06,
+      "loss": 0.743,
+      "step": 1990
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.9207,
+      "eval_loss": 0.3750178813934326,
+      "eval_runtime": 84.5377,
+      "eval_samples_per_second": 118.29,
+      "eval_steps_per_second": 0.473,
+      "step": 1995
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 6.347775459289551,
+      "learning_rate": 9.699248120300752e-06,
+      "loss": 0.6685,
+      "step": 2000
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 6.109332084655762,
+      "learning_rate": 9.697744360902256e-06,
+      "loss": 0.7398,
+      "step": 2010
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 4.770040512084961,
+      "learning_rate": 9.69624060150376e-06,
+      "loss": 0.7587,
+      "step": 2020
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 7.322962760925293,
+      "learning_rate": 9.694736842105263e-06,
+      "loss": 0.7917,
+      "step": 2030
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 9.097600936889648,
+      "learning_rate": 9.693233082706768e-06,
+      "loss": 0.6967,
+      "step": 2040
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 10.507075309753418,
+      "learning_rate": 9.691729323308271e-06,
+      "loss": 0.7229,
+      "step": 2050
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 6.249164581298828,
+      "learning_rate": 9.690225563909775e-06,
+      "loss": 0.6491,
+      "step": 2060
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 7.525278568267822,
+      "learning_rate": 9.688721804511278e-06,
+      "loss": 0.6476,
+      "step": 2070
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 9.564391136169434,
+      "learning_rate": 9.687218045112784e-06,
+      "loss": 0.7133,
+      "step": 2080
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 6.4053955078125,
+      "learning_rate": 9.685714285714285e-06,
+      "loss": 0.8012,
+      "step": 2090
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 6.6741251945495605,
+      "learning_rate": 9.68421052631579e-06,
+      "loss": 0.7507,
+      "step": 2100
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 12.398811340332031,
+      "learning_rate": 9.682706766917294e-06,
+      "loss": 0.7138,
+      "step": 2110
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 8.505881309509277,
+      "learning_rate": 9.681203007518798e-06,
+      "loss": 0.7136,
+      "step": 2120
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 5.51025915145874,
+      "learning_rate": 9.679699248120301e-06,
+      "loss": 0.6037,
+      "step": 2130
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 6.39398193359375,
+      "learning_rate": 9.678195488721805e-06,
+      "loss": 0.6799,
+      "step": 2140
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 6.508944511413574,
+      "learning_rate": 9.676691729323308e-06,
+      "loss": 0.6587,
+      "step": 2150
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 7.027959823608398,
+      "learning_rate": 9.675187969924813e-06,
+      "loss": 0.7096,
+      "step": 2160
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 8.690469741821289,
+      "learning_rate": 9.673684210526317e-06,
+      "loss": 0.6169,
+      "step": 2170
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 9.489269256591797,
+      "learning_rate": 9.67218045112782e-06,
+      "loss": 0.6643,
+      "step": 2180
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 12.934528350830078,
+      "learning_rate": 9.670676691729324e-06,
+      "loss": 0.7485,
+      "step": 2190
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 9.072815895080566,
+      "learning_rate": 9.669172932330828e-06,
+      "loss": 0.704,
+      "step": 2200
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 7.934593200683594,
+      "learning_rate": 9.667669172932331e-06,
+      "loss": 0.7134,
+      "step": 2210
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 6.846796989440918,
+      "learning_rate": 9.666165413533836e-06,
+      "loss": 0.7372,
+      "step": 2220
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 6.8362250328063965,
+      "learning_rate": 9.664661654135338e-06,
+      "loss": 0.6854,
+      "step": 2230
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 8.184903144836426,
+      "learning_rate": 9.663157894736843e-06,
+      "loss": 0.6662,
+      "step": 2240
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 8.972626686096191,
+      "learning_rate": 9.661654135338347e-06,
+      "loss": 0.7895,
+      "step": 2250
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 6.524502754211426,
+      "learning_rate": 9.66015037593985e-06,
+      "loss": 0.8453,
+      "step": 2260
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 6.215096950531006,
+      "learning_rate": 9.658646616541354e-06,
+      "loss": 0.716,
+      "step": 2270
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 6.9516401290893555,
+      "learning_rate": 9.657142857142859e-06,
+      "loss": 0.7453,
+      "step": 2280
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 9.131119728088379,
+      "learning_rate": 9.655639097744361e-06,
+      "loss": 0.6474,
+      "step": 2290
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 7.063914775848389,
+      "learning_rate": 9.654135338345866e-06,
+      "loss": 0.6535,
+      "step": 2300
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 9.410021781921387,
+      "learning_rate": 9.65263157894737e-06,
+      "loss": 0.708,
+      "step": 2310
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 7.179042816162109,
+      "learning_rate": 9.651127819548873e-06,
+      "loss": 0.7098,
+      "step": 2320
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 8.097248077392578,
+      "learning_rate": 9.649624060150377e-06,
+      "loss": 0.7052,
+      "step": 2330
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 4.258429050445557,
+      "learning_rate": 9.64812030075188e-06,
+      "loss": 0.6803,
+      "step": 2340
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 7.451633930206299,
+      "learning_rate": 9.646616541353384e-06,
+      "loss": 0.6569,
+      "step": 2350
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 9.551535606384277,
+      "learning_rate": 9.645112781954889e-06,
+      "loss": 0.6411,
+      "step": 2360
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 6.456385612487793,
+      "learning_rate": 9.643609022556392e-06,
+      "loss": 0.5938,
+      "step": 2370
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 5.947078704833984,
+      "learning_rate": 9.642105263157896e-06,
+      "loss": 0.6175,
+      "step": 2380
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 5.279054164886475,
+      "learning_rate": 9.6406015037594e-06,
+      "loss": 0.6788,
+      "step": 2390
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 7.593316555023193,
+      "learning_rate": 9.639097744360903e-06,
+      "loss": 0.6255,
+      "step": 2400
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 7.478080749511719,
+      "learning_rate": 9.637593984962407e-06,
+      "loss": 0.7908,
+      "step": 2410
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 9.64027214050293,
+      "learning_rate": 9.636090225563912e-06,
+      "loss": 0.7241,
+      "step": 2420
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 14.006696701049805,
+      "learning_rate": 9.634586466165414e-06,
+      "loss": 0.686,
+      "step": 2430
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 5.0339789390563965,
+      "learning_rate": 9.633082706766919e-06,
+      "loss": 0.7082,
+      "step": 2440
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 8.148447036743164,
+      "learning_rate": 9.631578947368422e-06,
+      "loss": 0.6859,
+      "step": 2450
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 7.614720344543457,
+      "learning_rate": 9.630075187969926e-06,
+      "loss": 0.7117,
+      "step": 2460
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 9.017003059387207,
+      "learning_rate": 9.62857142857143e-06,
+      "loss": 0.6505,
+      "step": 2470
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 6.4466986656188965,
+      "learning_rate": 9.627067669172933e-06,
+      "loss": 0.71,
+      "step": 2480
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 7.973327159881592,
+      "learning_rate": 9.625563909774436e-06,
+      "loss": 0.6466,
+      "step": 2490
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 6.712606906890869,
+      "learning_rate": 9.62406015037594e-06,
+      "loss": 0.6767,
+      "step": 2500
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 8.149372100830078,
+      "learning_rate": 9.622556390977445e-06,
+      "loss": 0.783,
+      "step": 2510
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 8.645270347595215,
+      "learning_rate": 9.621052631578947e-06,
+      "loss": 0.6978,
+      "step": 2520
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 9.347142219543457,
+      "learning_rate": 9.619548872180452e-06,
+      "loss": 0.583,
+      "step": 2530
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 7.905392169952393,
+      "learning_rate": 9.618045112781956e-06,
+      "loss": 0.6884,
+      "step": 2540
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 8.783331871032715,
+      "learning_rate": 9.61654135338346e-06,
+      "loss": 0.691,
+      "step": 2550
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 8.456209182739258,
+      "learning_rate": 9.615037593984963e-06,
+      "loss": 0.7281,
+      "step": 2560
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 6.667693138122559,
+      "learning_rate": 9.613533834586468e-06,
+      "loss": 0.6912,
+      "step": 2570
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 8.541569709777832,
+      "learning_rate": 9.61203007518797e-06,
+      "loss": 0.6853,
+      "step": 2580
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 4.732927322387695,
+      "learning_rate": 9.610526315789475e-06,
+      "loss": 0.6647,
+      "step": 2590
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 7.604156017303467,
+      "learning_rate": 9.609022556390978e-06,
+      "loss": 0.7526,
+      "step": 2600
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 8.218050956726074,
+      "learning_rate": 9.607518796992482e-06,
+      "loss": 0.6828,
+      "step": 2610
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 5.613206386566162,
+      "learning_rate": 9.606015037593985e-06,
+      "loss": 0.6964,
+      "step": 2620
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 9.644120216369629,
+      "learning_rate": 9.604511278195489e-06,
+      "loss": 0.6912,
+      "step": 2630
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 8.14504337310791,
+      "learning_rate": 9.603007518796993e-06,
+      "loss": 0.7527,
+      "step": 2640
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 6.1560468673706055,
+      "learning_rate": 9.601503759398498e-06,
+      "loss": 0.6145,
+      "step": 2650
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 20.564706802368164,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 0.6935,
+      "step": 2660
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.9236,
+      "eval_loss": 0.319810152053833,
+      "eval_runtime": 84.5508,
+      "eval_samples_per_second": 118.272,
+      "eval_steps_per_second": 0.473,
+      "step": 2660
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 9.448854446411133,
+      "learning_rate": 9.598496240601505e-06,
+      "loss": 0.631,
+      "step": 2670
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 10.194000244140625,
+      "learning_rate": 9.596992481203008e-06,
+      "loss": 0.6046,
+      "step": 2680
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 8.277205467224121,
+      "learning_rate": 9.595488721804512e-06,
+      "loss": 0.5853,
+      "step": 2690
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 7.616865158081055,
+      "learning_rate": 9.593984962406015e-06,
+      "loss": 0.6168,
+      "step": 2700
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 5.158208847045898,
+      "learning_rate": 9.59248120300752e-06,
+      "loss": 0.5893,
+      "step": 2710
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 10.609253883361816,
+      "learning_rate": 9.590977443609022e-06,
+      "loss": 0.5819,
+      "step": 2720
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 7.288332462310791,
+      "learning_rate": 9.589473684210528e-06,
+      "loss": 0.6384,
+      "step": 2730
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 6.625866889953613,
+      "learning_rate": 9.587969924812031e-06,
+      "loss": 0.6841,
+      "step": 2740
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 8.38702392578125,
+      "learning_rate": 9.586466165413535e-06,
+      "loss": 0.5815,
+      "step": 2750
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 6.58852481842041,
+      "learning_rate": 9.584962406015038e-06,
+      "loss": 0.7091,
+      "step": 2760
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 5.776881217956543,
+      "learning_rate": 9.583458646616542e-06,
+      "loss": 0.6426,
+      "step": 2770
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 7.806540489196777,
+      "learning_rate": 9.581954887218045e-06,
+      "loss": 0.6459,
+      "step": 2780
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 7.378940582275391,
+      "learning_rate": 9.58045112781955e-06,
+      "loss": 0.6553,
+      "step": 2790
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 8.37366008758545,
+      "learning_rate": 9.578947368421054e-06,
+      "loss": 0.6024,
+      "step": 2800
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 5.783264636993408,
+      "learning_rate": 9.577443609022557e-06,
+      "loss": 0.5959,
+      "step": 2810
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 6.4687676429748535,
+      "learning_rate": 9.575939849624061e-06,
+      "loss": 0.5669,
+      "step": 2820
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 10.575803756713867,
+      "learning_rate": 9.574436090225564e-06,
+      "loss": 0.6461,
+      "step": 2830
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 9.703124046325684,
+      "learning_rate": 9.572932330827068e-06,
+      "loss": 0.5784,
+      "step": 2840
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 9.454757690429688,
+      "learning_rate": 9.571428571428573e-06,
+      "loss": 0.7381,
+      "step": 2850
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 7.034806728363037,
+      "learning_rate": 9.569924812030075e-06,
+      "loss": 0.6277,
+      "step": 2860
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 10.1060791015625,
+      "learning_rate": 9.56842105263158e-06,
+      "loss": 0.6973,
+      "step": 2870
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 7.225138187408447,
+      "learning_rate": 9.566917293233084e-06,
+      "loss": 0.587,
+      "step": 2880
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 6.221525192260742,
+      "learning_rate": 9.565413533834587e-06,
+      "loss": 0.6006,
+      "step": 2890
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 6.329552173614502,
+      "learning_rate": 9.56390977443609e-06,
+      "loss": 0.6474,
+      "step": 2900
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 8.411649703979492,
+      "learning_rate": 9.562406015037596e-06,
+      "loss": 0.5376,
+      "step": 2910
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 8.27790355682373,
+      "learning_rate": 9.560902255639098e-06,
+      "loss": 0.6324,
+      "step": 2920
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 6.995235443115234,
+      "learning_rate": 9.559398496240603e-06,
+      "loss": 0.6178,
+      "step": 2930
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 8.169748306274414,
+      "learning_rate": 9.557894736842107e-06,
+      "loss": 0.5546,
+      "step": 2940
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 7.832982063293457,
+      "learning_rate": 9.55639097744361e-06,
+      "loss": 0.5873,
+      "step": 2950
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 7.024545192718506,
+      "learning_rate": 9.554887218045114e-06,
+      "loss": 0.5919,
+      "step": 2960
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 8.610920906066895,
+      "learning_rate": 9.553383458646617e-06,
+      "loss": 0.6324,
+      "step": 2970
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 8.49885368347168,
+      "learning_rate": 9.55187969924812e-06,
+      "loss": 0.6392,
+      "step": 2980
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 6.013737678527832,
+      "learning_rate": 9.550375939849624e-06,
+      "loss": 0.6267,
+      "step": 2990
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 9.457529067993164,
+      "learning_rate": 9.54887218045113e-06,
+      "loss": 0.633,
+      "step": 3000
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 7.126248359680176,
+      "learning_rate": 9.547368421052631e-06,
+      "loss": 0.5527,
+      "step": 3010
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 8.481447219848633,
+      "learning_rate": 9.545864661654136e-06,
+      "loss": 0.7163,
+      "step": 3020
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 10.300518035888672,
+      "learning_rate": 9.54436090225564e-06,
+      "loss": 0.5921,
+      "step": 3030
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 8.265804290771484,
+      "learning_rate": 9.542857142857143e-06,
+      "loss": 0.5952,
+      "step": 3040
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 5.028606414794922,
+      "learning_rate": 9.541353383458647e-06,
+      "loss": 0.6269,
+      "step": 3050
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 8.997878074645996,
+      "learning_rate": 9.53984962406015e-06,
+      "loss": 0.5761,
+      "step": 3060
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 9.585675239562988,
+      "learning_rate": 9.538345864661654e-06,
+      "loss": 0.5851,
+      "step": 3070
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 5.83755350112915,
+      "learning_rate": 9.53684210526316e-06,
+      "loss": 0.6105,
+      "step": 3080
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 5.208207607269287,
+      "learning_rate": 9.535338345864663e-06,
+      "loss": 0.5786,
+      "step": 3090
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 9.895461082458496,
+      "learning_rate": 9.533834586466166e-06,
+      "loss": 0.623,
+      "step": 3100
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 8.958138465881348,
+      "learning_rate": 9.53233082706767e-06,
+      "loss": 0.5708,
+      "step": 3110
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 10.452126502990723,
+      "learning_rate": 9.530827067669173e-06,
+      "loss": 0.5694,
+      "step": 3120
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 7.20021915435791,
+      "learning_rate": 9.529323308270677e-06,
+      "loss": 0.6267,
+      "step": 3130
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 7.995909690856934,
+      "learning_rate": 9.527819548872182e-06,
+      "loss": 0.6989,
+      "step": 3140
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 6.9314985275268555,
+      "learning_rate": 9.526315789473684e-06,
+      "loss": 0.6093,
+      "step": 3150
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 10.158616065979004,
+      "learning_rate": 9.524812030075189e-06,
+      "loss": 0.5696,
+      "step": 3160
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 7.637181758880615,
+      "learning_rate": 9.523308270676693e-06,
+      "loss": 0.7494,
+      "step": 3170
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 7.443474769592285,
+      "learning_rate": 9.521804511278196e-06,
+      "loss": 0.7636,
+      "step": 3180
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 6.130582809448242,
+      "learning_rate": 9.5203007518797e-06,
+      "loss": 0.7397,
+      "step": 3190
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 8.699774742126465,
+      "learning_rate": 9.518796992481205e-06,
+      "loss": 0.5461,
+      "step": 3200
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 8.060851097106934,
+      "learning_rate": 9.517293233082707e-06,
+      "loss": 0.5424,
+      "step": 3210
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 6.084632396697998,
+      "learning_rate": 9.515789473684212e-06,
+      "loss": 0.6181,
+      "step": 3220
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 8.804571151733398,
+      "learning_rate": 9.514285714285715e-06,
+      "loss": 0.6696,
+      "step": 3230
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 8.552626609802246,
+      "learning_rate": 9.512781954887219e-06,
+      "loss": 0.6237,
+      "step": 3240
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 8.930567741394043,
+      "learning_rate": 9.511278195488722e-06,
+      "loss": 0.6427,
+      "step": 3250
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 8.916244506835938,
+      "learning_rate": 9.509774436090226e-06,
+      "loss": 0.608,
+      "step": 3260
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 12.679169654846191,
+      "learning_rate": 9.50827067669173e-06,
+      "loss": 0.622,
+      "step": 3270
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 9.21071720123291,
+      "learning_rate": 9.506766917293235e-06,
+      "loss": 0.6153,
+      "step": 3280
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 8.040297508239746,
+      "learning_rate": 9.505263157894738e-06,
+      "loss": 0.654,
+      "step": 3290
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 6.395382404327393,
+      "learning_rate": 9.503759398496242e-06,
+      "loss": 0.6729,
+      "step": 3300
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 8.437057495117188,
+      "learning_rate": 9.502255639097745e-06,
+      "loss": 0.6457,
+      "step": 3310
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 7.987279415130615,
+      "learning_rate": 9.500751879699249e-06,
+      "loss": 0.6159,
+      "step": 3320
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.9289,
+      "eval_loss": 0.2945062816143036,
+      "eval_runtime": 84.9367,
+      "eval_samples_per_second": 117.735,
+      "eval_steps_per_second": 0.471,
+      "step": 3325
+    },
+    {
+      "epoch": 5.01,
+      "grad_norm": 6.12150239944458,
+      "learning_rate": 9.499248120300752e-06,
+      "loss": 0.4871,
+      "step": 3330
+    },
+    {
+      "epoch": 5.02,
+      "grad_norm": 8.20666217803955,
+      "learning_rate": 9.497744360902257e-06,
+      "loss": 0.6011,
+      "step": 3340
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 8.818642616271973,
+      "learning_rate": 9.49624060150376e-06,
+      "loss": 0.5681,
+      "step": 3350
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 5.606151103973389,
+      "learning_rate": 9.494736842105265e-06,
+      "loss": 0.5494,
+      "step": 3360
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 6.230663299560547,
+      "learning_rate": 9.493233082706768e-06,
+      "loss": 0.6138,
+      "step": 3370
+    },
+    {
+      "epoch": 5.08,
+      "grad_norm": 6.923035621643066,
+      "learning_rate": 9.491729323308272e-06,
+      "loss": 0.6398,
+      "step": 3380
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 8.464035034179688,
+      "learning_rate": 9.490225563909775e-06,
+      "loss": 0.533,
+      "step": 3390
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 5.440852165222168,
+      "learning_rate": 9.488721804511279e-06,
+      "loss": 0.5901,
+      "step": 3400
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 6.880829334259033,
+      "learning_rate": 9.487218045112782e-06,
+      "loss": 0.5699,
+      "step": 3410
+    },
+    {
+      "epoch": 5.14,
+      "grad_norm": 6.773617267608643,
+      "learning_rate": 9.485714285714287e-06,
+      "loss": 0.5164,
+      "step": 3420
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 6.794729232788086,
+      "learning_rate": 9.484210526315791e-06,
+      "loss": 0.4939,
+      "step": 3430
+    },
+    {
+      "epoch": 5.17,
+      "grad_norm": 8.347722053527832,
+      "learning_rate": 9.482706766917294e-06,
+      "loss": 0.6138,
+      "step": 3440
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 6.279055595397949,
+      "learning_rate": 9.481203007518798e-06,
+      "loss": 0.4792,
+      "step": 3450
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 7.93798828125,
+      "learning_rate": 9.479699248120301e-06,
+      "loss": 0.5945,
+      "step": 3460
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 6.767178535461426,
+      "learning_rate": 9.478195488721805e-06,
+      "loss": 0.5878,
+      "step": 3470
+    },
+    {
+      "epoch": 5.23,
+      "grad_norm": 6.87293004989624,
+      "learning_rate": 9.476691729323308e-06,
+      "loss": 0.566,
+      "step": 3480
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 2.407437562942505,
+      "learning_rate": 9.475187969924814e-06,
+      "loss": 0.5014,
+      "step": 3490
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 8.233712196350098,
+      "learning_rate": 9.473684210526315e-06,
+      "loss": 0.5879,
+      "step": 3500
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 7.905375003814697,
+      "learning_rate": 9.47218045112782e-06,
+      "loss": 0.6127,
+      "step": 3510
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 5.8037238121032715,
+      "learning_rate": 9.470676691729324e-06,
+      "loss": 0.6048,
+      "step": 3520
+    },
+    {
+      "epoch": 5.31,
+      "grad_norm": 9.2665433883667,
+      "learning_rate": 9.469172932330828e-06,
+      "loss": 0.6233,
+      "step": 3530
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 5.650614261627197,
+      "learning_rate": 9.467669172932331e-06,
+      "loss": 0.5703,
+      "step": 3540
+    },
+    {
+      "epoch": 5.34,
+      "grad_norm": 5.246155738830566,
+      "learning_rate": 9.466165413533835e-06,
+      "loss": 0.5108,
+      "step": 3550
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 8.701322555541992,
+      "learning_rate": 9.464661654135338e-06,
+      "loss": 0.5783,
+      "step": 3560
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 5.870892524719238,
+      "learning_rate": 9.463157894736844e-06,
+      "loss": 0.5546,
+      "step": 3570
+    },
+    {
+      "epoch": 5.38,
+      "grad_norm": 8.061163902282715,
+      "learning_rate": 9.461654135338347e-06,
+      "loss": 0.5973,
+      "step": 3580
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 4.166900157928467,
+      "learning_rate": 9.46015037593985e-06,
+      "loss": 0.6042,
+      "step": 3590
+    },
+    {
+      "epoch": 5.41,
+      "grad_norm": 5.8524346351623535,
+      "learning_rate": 9.458646616541354e-06,
+      "loss": 0.5307,
+      "step": 3600
+    },
+    {
+      "epoch": 5.43,
+      "grad_norm": 7.229081153869629,
+      "learning_rate": 9.457142857142858e-06,
+      "loss": 0.6533,
+      "step": 3610
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 5.403026580810547,
+      "learning_rate": 9.455639097744361e-06,
+      "loss": 0.5059,
+      "step": 3620
+    },
+    {
+      "epoch": 5.46,
+      "grad_norm": 7.53814697265625,
+      "learning_rate": 9.454135338345866e-06,
+      "loss": 0.6576,
+      "step": 3630
+    },
+    {
+      "epoch": 5.47,
+      "grad_norm": 8.08530330657959,
+      "learning_rate": 9.452631578947368e-06,
+      "loss": 0.5881,
+      "step": 3640
+    },
+    {
+      "epoch": 5.49,
+      "grad_norm": 7.80808687210083,
+      "learning_rate": 9.451127819548873e-06,
+      "loss": 0.5725,
+      "step": 3650
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 4.368475914001465,
+      "learning_rate": 9.449624060150377e-06,
+      "loss": 0.5538,
+      "step": 3660
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 7.806415557861328,
+      "learning_rate": 9.44812030075188e-06,
+      "loss": 0.6739,
+      "step": 3670
+    },
+    {
+      "epoch": 5.53,
+      "grad_norm": 8.047362327575684,
+      "learning_rate": 9.446616541353384e-06,
+      "loss": 0.5229,
+      "step": 3680
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 5.89243221282959,
+      "learning_rate": 9.445112781954887e-06,
+      "loss": 0.599,
+      "step": 3690
+    },
+    {
+      "epoch": 5.56,
+      "grad_norm": 6.812804222106934,
+      "learning_rate": 9.443609022556391e-06,
+      "loss": 0.6213,
+      "step": 3700
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 7.177376747131348,
+      "learning_rate": 9.442105263157896e-06,
+      "loss": 0.5568,
+      "step": 3710
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 4.684938907623291,
+      "learning_rate": 9.4406015037594e-06,
+      "loss": 0.5266,
+      "step": 3720
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 9.440715789794922,
+      "learning_rate": 9.439097744360903e-06,
+      "loss": 0.6431,
+      "step": 3730
+    },
+    {
+      "epoch": 5.62,
+      "grad_norm": 6.564364910125732,
+      "learning_rate": 9.437593984962407e-06,
+      "loss": 0.4702,
+      "step": 3740
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 7.499642372131348,
+      "learning_rate": 9.43609022556391e-06,
+      "loss": 0.5249,
+      "step": 3750
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 5.27653169631958,
+      "learning_rate": 9.434586466165414e-06,
+      "loss": 0.4517,
+      "step": 3760
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 4.881272792816162,
+      "learning_rate": 9.433082706766919e-06,
+      "loss": 0.4682,
+      "step": 3770
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 9.005290985107422,
+      "learning_rate": 9.43157894736842e-06,
+      "loss": 0.5642,
+      "step": 3780
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 6.421182155609131,
+      "learning_rate": 9.430075187969926e-06,
+      "loss": 0.4855,
+      "step": 3790
+    },
+    {
+      "epoch": 5.71,
+      "grad_norm": 7.035130023956299,
+      "learning_rate": 9.42857142857143e-06,
+      "loss": 0.6362,
+      "step": 3800
+    },
+    {
+      "epoch": 5.73,
+      "grad_norm": 5.77438497543335,
+      "learning_rate": 9.427067669172933e-06,
+      "loss": 0.5876,
+      "step": 3810
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 8.328043937683105,
+      "learning_rate": 9.425563909774437e-06,
+      "loss": 0.4768,
+      "step": 3820
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 5.7907586097717285,
+      "learning_rate": 9.424060150375942e-06,
+      "loss": 0.5941,
+      "step": 3830
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 8.730267524719238,
+      "learning_rate": 9.422556390977444e-06,
+      "loss": 0.6426,
+      "step": 3840
+    },
+    {
+      "epoch": 5.79,
+      "grad_norm": 8.710532188415527,
+      "learning_rate": 9.421052631578949e-06,
+      "loss": 0.5909,
+      "step": 3850
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 8.74202823638916,
+      "learning_rate": 9.419548872180452e-06,
+      "loss": 0.5735,
+      "step": 3860
+    },
+    {
+      "epoch": 5.82,
+      "grad_norm": 7.489967346191406,
+      "learning_rate": 9.418045112781956e-06,
+      "loss": 0.6072,
+      "step": 3870
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 6.942547798156738,
+      "learning_rate": 9.41654135338346e-06,
+      "loss": 0.5513,
+      "step": 3880
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 5.517817497253418,
+      "learning_rate": 9.415037593984963e-06,
+      "loss": 0.5317,
+      "step": 3890
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 6.26224946975708,
+      "learning_rate": 9.413533834586466e-06,
+      "loss": 0.6014,
+      "step": 3900
+    },
+    {
+      "epoch": 5.88,
+      "grad_norm": 4.848892688751221,
+      "learning_rate": 9.412030075187972e-06,
+      "loss": 0.5628,
+      "step": 3910
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 7.279343128204346,
+      "learning_rate": 9.410526315789475e-06,
+      "loss": 0.631,
+      "step": 3920
+    },
+    {
+      "epoch": 5.91,
+      "grad_norm": 5.791496753692627,
+      "learning_rate": 9.409022556390979e-06,
+      "loss": 0.5312,
+      "step": 3930
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 4.935235977172852,
+      "learning_rate": 9.407518796992482e-06,
+      "loss": 0.4694,
+      "step": 3940
+    },
+    {
+      "epoch": 5.94,
+      "grad_norm": 5.741876125335693,
+      "learning_rate": 9.406015037593986e-06,
+      "loss": 0.6073,
+      "step": 3950
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 5.398350715637207,
+      "learning_rate": 9.40451127819549e-06,
+      "loss": 0.6009,
+      "step": 3960
+    },
+    {
+      "epoch": 5.97,
+      "grad_norm": 6.093377590179443,
+      "learning_rate": 9.403007518796994e-06,
+      "loss": 0.5845,
+      "step": 3970
+    },
+    {
+      "epoch": 5.98,
+      "grad_norm": 8.6488676071167,
+      "learning_rate": 9.401503759398496e-06,
+      "loss": 0.5932,
+      "step": 3980
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 3.0173494815826416,
+      "learning_rate": 9.4e-06,
+      "loss": 0.4423,
+      "step": 3990
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.925,
+      "eval_loss": 0.2876322865486145,
+      "eval_runtime": 85.2695,
+      "eval_samples_per_second": 117.275,
+      "eval_steps_per_second": 0.469,
+      "step": 3990
+    },
+    {
+      "epoch": 6.02,
+      "grad_norm": 4.103921890258789,
+      "learning_rate": 9.398496240601505e-06,
+      "loss": 0.5346,
+      "step": 4000
+    },
+    {
+      "epoch": 6.03,
+      "grad_norm": 4.625704288482666,
+      "learning_rate": 9.396992481203009e-06,
+      "loss": 0.5522,
+      "step": 4010
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 7.251491546630859,
+      "learning_rate": 9.395488721804512e-06,
+      "loss": 0.6501,
+      "step": 4020
+    },
+    {
+      "epoch": 6.06,
+      "grad_norm": 8.962389945983887,
+      "learning_rate": 9.393984962406016e-06,
+      "loss": 0.5402,
+      "step": 4030
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 6.207771301269531,
+      "learning_rate": 9.392481203007519e-06,
+      "loss": 0.5545,
+      "step": 4040
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 5.253688812255859,
+      "learning_rate": 9.390977443609023e-06,
+      "loss": 0.5084,
+      "step": 4050
+    },
+    {
+      "epoch": 6.11,
+      "grad_norm": 7.277046203613281,
+      "learning_rate": 9.389473684210528e-06,
+      "loss": 0.5115,
+      "step": 4060
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 5.671750068664551,
+      "learning_rate": 9.38796992481203e-06,
+      "loss": 0.5508,
+      "step": 4070
+    },
+    {
+      "epoch": 6.14,
+      "grad_norm": 3.9672038555145264,
+      "learning_rate": 9.386466165413535e-06,
+      "loss": 0.494,
+      "step": 4080
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 6.129919052124023,
+      "learning_rate": 9.384962406015038e-06,
+      "loss": 0.5898,
+      "step": 4090
+    },
+    {
+      "epoch": 6.17,
+      "grad_norm": 6.198451519012451,
+      "learning_rate": 9.383458646616542e-06,
+      "loss": 0.5378,
+      "step": 4100
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 9.286908149719238,
+      "learning_rate": 9.381954887218045e-06,
+      "loss": 0.5656,
+      "step": 4110
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 6.862420082092285,
+      "learning_rate": 9.38045112781955e-06,
+      "loss": 0.5457,
+      "step": 4120
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 5.948605537414551,
+      "learning_rate": 9.378947368421052e-06,
+      "loss": 0.5564,
+      "step": 4130
+    },
+    {
+      "epoch": 6.23,
+      "grad_norm": 7.1652116775512695,
+      "learning_rate": 9.377443609022558e-06,
+      "loss": 0.5624,
+      "step": 4140
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 7.091752052307129,
+      "learning_rate": 9.375939849624061e-06,
+      "loss": 0.532,
+      "step": 4150
+    },
+    {
+      "epoch": 6.26,
+      "grad_norm": 4.2119903564453125,
+      "learning_rate": 9.374436090225565e-06,
+      "loss": 0.385,
+      "step": 4160
+    },
+    {
+      "epoch": 6.27,
+      "grad_norm": 9.477155685424805,
+      "learning_rate": 9.372932330827068e-06,
+      "loss": 0.5019,
+      "step": 4170
+    },
+    {
+      "epoch": 6.29,
+      "grad_norm": 7.294814109802246,
+      "learning_rate": 9.371428571428572e-06,
+      "loss": 0.556,
+      "step": 4180
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 8.124314308166504,
+      "learning_rate": 9.369924812030075e-06,
+      "loss": 0.548,
+      "step": 4190
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 6.1076483726501465,
+      "learning_rate": 9.36842105263158e-06,
+      "loss": 0.5228,
+      "step": 4200
+    },
+    {
+      "epoch": 6.33,
+      "grad_norm": 7.360411643981934,
+      "learning_rate": 9.366917293233084e-06,
+      "loss": 0.6335,
+      "step": 4210
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 6.000509738922119,
+      "learning_rate": 9.365413533834588e-06,
+      "loss": 0.5727,
+      "step": 4220
+    },
+    {
+      "epoch": 6.36,
+      "grad_norm": 8.538400650024414,
+      "learning_rate": 9.363909774436091e-06,
+      "loss": 0.5135,
+      "step": 4230
+    },
+    {
+      "epoch": 6.38,
+      "grad_norm": 6.543038845062256,
+      "learning_rate": 9.362406015037595e-06,
+      "loss": 0.4859,
+      "step": 4240
+    },
+    {
+      "epoch": 6.39,
+      "grad_norm": 7.515405178070068,
+      "learning_rate": 9.360902255639098e-06,
+      "loss": 0.6172,
+      "step": 4250
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 4.859574317932129,
+      "learning_rate": 9.359398496240603e-06,
+      "loss": 0.407,
+      "step": 4260
+    },
+    {
+      "epoch": 6.42,
+      "grad_norm": 5.684931755065918,
+      "learning_rate": 9.357894736842105e-06,
+      "loss": 0.5748,
+      "step": 4270
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 5.063128471374512,
+      "learning_rate": 9.35639097744361e-06,
+      "loss": 0.5363,
+      "step": 4280
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 9.34011173248291,
+      "learning_rate": 9.354887218045114e-06,
+      "loss": 0.5072,
+      "step": 4290
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 6.302648544311523,
+      "learning_rate": 9.353383458646617e-06,
+      "loss": 0.4891,
+      "step": 4300
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 6.268799781799316,
+      "learning_rate": 9.351879699248121e-06,
+      "loss": 0.6215,
+      "step": 4310
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 5.54179573059082,
+      "learning_rate": 9.350375939849624e-06,
+      "loss": 0.5827,
+      "step": 4320
+    },
+    {
+      "epoch": 6.51,
+      "grad_norm": 6.042153835296631,
+      "learning_rate": 9.348872180451128e-06,
+      "loss": 0.5205,
+      "step": 4330
+    },
+    {
+      "epoch": 6.53,
+      "grad_norm": 7.558413982391357,
+      "learning_rate": 9.347368421052633e-06,
+      "loss": 0.5357,
+      "step": 4340
+    },
+    {
+      "epoch": 6.54,
+      "grad_norm": 7.838019847869873,
+      "learning_rate": 9.345864661654137e-06,
+      "loss": 0.5719,
+      "step": 4350
+    },
+    {
+      "epoch": 6.56,
+      "grad_norm": 10.056818008422852,
+      "learning_rate": 9.34436090225564e-06,
+      "loss": 0.542,
+      "step": 4360
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 7.325047492980957,
+      "learning_rate": 9.342857142857144e-06,
+      "loss": 0.5564,
+      "step": 4370
+    },
+    {
+      "epoch": 6.59,
+      "grad_norm": 8.13595199584961,
+      "learning_rate": 9.341353383458647e-06,
+      "loss": 0.5106,
+      "step": 4380
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 7.225549221038818,
+      "learning_rate": 9.33984962406015e-06,
+      "loss": 0.5714,
+      "step": 4390
+    },
+    {
+      "epoch": 6.62,
+      "grad_norm": 9.103632926940918,
+      "learning_rate": 9.338345864661656e-06,
+      "loss": 0.559,
+      "step": 4400
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 5.579386234283447,
+      "learning_rate": 9.336842105263158e-06,
+      "loss": 0.5789,
+      "step": 4410
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 9.875541687011719,
+      "learning_rate": 9.335338345864663e-06,
+      "loss": 0.5435,
+      "step": 4420
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 10.33945083618164,
+      "learning_rate": 9.333834586466166e-06,
+      "loss": 0.4678,
+      "step": 4430
+    },
+    {
+      "epoch": 6.68,
+      "grad_norm": 8.180964469909668,
+      "learning_rate": 9.33233082706767e-06,
+      "loss": 0.5476,
+      "step": 4440
+    },
+    {
+      "epoch": 6.69,
+      "grad_norm": 8.327938079833984,
+      "learning_rate": 9.330827067669174e-06,
+      "loss": 0.5605,
+      "step": 4450
+    },
+    {
+      "epoch": 6.71,
+      "grad_norm": 6.9362311363220215,
+      "learning_rate": 9.329323308270679e-06,
+      "loss": 0.6099,
+      "step": 4460
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 4.627447128295898,
+      "learning_rate": 9.32781954887218e-06,
+      "loss": 0.4521,
+      "step": 4470
+    },
+    {
+      "epoch": 6.74,
+      "grad_norm": 5.713562488555908,
+      "learning_rate": 9.326315789473684e-06,
+      "loss": 0.4532,
+      "step": 4480
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 6.893897533416748,
+      "learning_rate": 9.32481203007519e-06,
+      "loss": 0.5298,
+      "step": 4490
+    },
+    {
+      "epoch": 6.77,
+      "grad_norm": 6.630578994750977,
+      "learning_rate": 9.323308270676693e-06,
+      "loss": 0.5226,
+      "step": 4500
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 7.725119113922119,
+      "learning_rate": 9.321804511278196e-06,
+      "loss": 0.4581,
+      "step": 4510
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 8.243720054626465,
+      "learning_rate": 9.3203007518797e-06,
+      "loss": 0.4821,
+      "step": 4520
+    },
+    {
+      "epoch": 6.81,
+      "grad_norm": 7.254865646362305,
+      "learning_rate": 9.318796992481203e-06,
+      "loss": 0.4623,
+      "step": 4530
+    },
+    {
+      "epoch": 6.83,
+      "grad_norm": 6.510406494140625,
+      "learning_rate": 9.317293233082707e-06,
+      "loss": 0.5073,
+      "step": 4540
+    },
+    {
+      "epoch": 6.84,
+      "grad_norm": 5.531012058258057,
+      "learning_rate": 9.315789473684212e-06,
+      "loss": 0.4842,
+      "step": 4550
+    },
+    {
+      "epoch": 6.86,
+      "grad_norm": 9.50185489654541,
+      "learning_rate": 9.314285714285714e-06,
+      "loss": 0.5485,
+      "step": 4560
+    },
+    {
+      "epoch": 6.87,
+      "grad_norm": 5.591551780700684,
+      "learning_rate": 9.312781954887219e-06,
+      "loss": 0.5397,
+      "step": 4570
+    },
+    {
+      "epoch": 6.89,
+      "grad_norm": 3.7708208560943604,
+      "learning_rate": 9.311278195488723e-06,
+      "loss": 0.4746,
+      "step": 4580
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 5.826446533203125,
+      "learning_rate": 9.309774436090226e-06,
+      "loss": 0.503,
+      "step": 4590
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 14.129280090332031,
+      "learning_rate": 9.30827067669173e-06,
+      "loss": 0.5729,
+      "step": 4600
+    },
+    {
+      "epoch": 6.93,
+      "grad_norm": 5.19706392288208,
+      "learning_rate": 9.306766917293233e-06,
+      "loss": 0.4967,
+      "step": 4610
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 6.513811111450195,
+      "learning_rate": 9.305263157894737e-06,
+      "loss": 0.587,
+      "step": 4620
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 7.2199506759643555,
+      "learning_rate": 9.303759398496242e-06,
+      "loss": 0.5733,
+      "step": 4630
+    },
+    {
+      "epoch": 6.98,
+      "grad_norm": 6.173489570617676,
+      "learning_rate": 9.302255639097745e-06,
+      "loss": 0.6391,
+      "step": 4640
+    },
+    {
+      "epoch": 6.99,
+      "grad_norm": 4.977587699890137,
+      "learning_rate": 9.300751879699249e-06,
+      "loss": 0.5506,
+      "step": 4650
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.9302,
+      "eval_loss": 0.2616922855377197,
+      "eval_runtime": 85.0513,
+      "eval_samples_per_second": 117.576,
+      "eval_steps_per_second": 0.47,
+      "step": 4655
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": 7.2303266525268555,
+      "learning_rate": 9.299248120300753e-06,
+      "loss": 0.4889,
+      "step": 4660
+    },
+    {
+      "epoch": 7.02,
+      "grad_norm": 5.8191914558410645,
+      "learning_rate": 9.297744360902256e-06,
+      "loss": 0.6557,
+      "step": 4670
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 7.453029632568359,
+      "learning_rate": 9.29624060150376e-06,
+      "loss": 0.5975,
+      "step": 4680
+    },
+    {
+      "epoch": 7.05,
+      "grad_norm": 9.434555053710938,
+      "learning_rate": 9.294736842105265e-06,
+      "loss": 0.4769,
+      "step": 4690
+    },
+    {
+      "epoch": 7.07,
+      "grad_norm": 9.096846580505371,
+      "learning_rate": 9.293233082706767e-06,
+      "loss": 0.5521,
+      "step": 4700
+    },
+    {
+      "epoch": 7.08,
+      "grad_norm": 6.146598815917969,
+      "learning_rate": 9.291729323308272e-06,
+      "loss": 0.5191,
+      "step": 4710
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 5.138683319091797,
+      "learning_rate": 9.290225563909775e-06,
+      "loss": 0.4883,
+      "step": 4720
+    },
+    {
+      "epoch": 7.11,
+      "grad_norm": 5.840444564819336,
+      "learning_rate": 9.288721804511279e-06,
+      "loss": 0.4901,
+      "step": 4730
+    },
+    {
+      "epoch": 7.13,
+      "grad_norm": 5.589585781097412,
+      "learning_rate": 9.287218045112782e-06,
+      "loss": 0.4574,
+      "step": 4740
+    },
+    {
+      "epoch": 7.14,
+      "grad_norm": 7.447097301483154,
+      "learning_rate": 9.285714285714288e-06,
+      "loss": 0.4672,
+      "step": 4750
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 6.8820295333862305,
+      "learning_rate": 9.28421052631579e-06,
+      "loss": 0.5641,
+      "step": 4760
+    },
+    {
+      "epoch": 7.17,
+      "grad_norm": 5.46907901763916,
+      "learning_rate": 9.282706766917295e-06,
+      "loss": 0.4948,
+      "step": 4770
+    },
+    {
+      "epoch": 7.19,
+      "grad_norm": 5.4233527183532715,
+      "learning_rate": 9.281203007518798e-06,
+      "loss": 0.5507,
+      "step": 4780
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 6.316089153289795,
+      "learning_rate": 9.279699248120302e-06,
+      "loss": 0.5791,
+      "step": 4790
+    },
+    {
+      "epoch": 7.22,
+      "grad_norm": 3.7618801593780518,
+      "learning_rate": 9.278195488721805e-06,
+      "loss": 0.4846,
+      "step": 4800
+    },
+    {
+      "epoch": 7.23,
+      "grad_norm": 6.426711082458496,
+      "learning_rate": 9.276691729323309e-06,
+      "loss": 0.4886,
+      "step": 4810
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 6.98826265335083,
+      "learning_rate": 9.275187969924812e-06,
+      "loss": 0.4627,
+      "step": 4820
+    },
+    {
+      "epoch": 7.26,
+      "grad_norm": 6.147061824798584,
+      "learning_rate": 9.273684210526317e-06,
+      "loss": 0.4577,
+      "step": 4830
+    },
+    {
+      "epoch": 7.28,
+      "grad_norm": 7.308942794799805,
+      "learning_rate": 9.272180451127821e-06,
+      "loss": 0.4893,
+      "step": 4840
+    },
+    {
+      "epoch": 7.29,
+      "grad_norm": 8.406046867370605,
+      "learning_rate": 9.270676691729324e-06,
+      "loss": 0.4968,
+      "step": 4850
+    },
+    {
+      "epoch": 7.31,
+      "grad_norm": 4.631737232208252,
+      "learning_rate": 9.269172932330828e-06,
+      "loss": 0.4654,
+      "step": 4860
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 3.802255868911743,
+      "learning_rate": 9.267669172932331e-06,
+      "loss": 0.5018,
+      "step": 4870
+    },
+    {
+      "epoch": 7.34,
+      "grad_norm": 7.958065986633301,
+      "learning_rate": 9.266165413533835e-06,
+      "loss": 0.5358,
+      "step": 4880
+    },
+    {
+      "epoch": 7.35,
+      "grad_norm": 4.825588703155518,
+      "learning_rate": 9.26466165413534e-06,
+      "loss": 0.5201,
+      "step": 4890
+    },
+    {
+      "epoch": 7.37,
+      "grad_norm": 4.964457035064697,
+      "learning_rate": 9.263157894736842e-06,
+      "loss": 0.4819,
+      "step": 4900
+    },
+    {
+      "epoch": 7.38,
+      "grad_norm": 11.642394065856934,
+      "learning_rate": 9.261654135338347e-06,
+      "loss": 0.5989,
+      "step": 4910
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 9.31828498840332,
+      "learning_rate": 9.26015037593985e-06,
+      "loss": 0.5455,
+      "step": 4920
+    },
+    {
+      "epoch": 7.41,
+      "grad_norm": 7.8129963874816895,
+      "learning_rate": 9.258646616541354e-06,
+      "loss": 0.5542,
+      "step": 4930
+    },
+    {
+      "epoch": 7.43,
+      "grad_norm": 7.043788909912109,
+      "learning_rate": 9.257142857142858e-06,
+      "loss": 0.5153,
+      "step": 4940
+    },
+    {
+      "epoch": 7.44,
+      "grad_norm": 8.311758995056152,
+      "learning_rate": 9.255639097744363e-06,
+      "loss": 0.4802,
+      "step": 4950
+    },
+    {
+      "epoch": 7.46,
+      "grad_norm": 10.970717430114746,
+      "learning_rate": 9.254135338345865e-06,
+      "loss": 0.4495,
+      "step": 4960
+    },
+    {
+      "epoch": 7.47,
+      "grad_norm": 5.547107219696045,
+      "learning_rate": 9.252631578947368e-06,
+      "loss": 0.4824,
+      "step": 4970
+    },
+    {
+      "epoch": 7.49,
+      "grad_norm": 5.658668518066406,
+      "learning_rate": 9.251127819548874e-06,
+      "loss": 0.5722,
+      "step": 4980
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 4.896615982055664,
+      "learning_rate": 9.249624060150375e-06,
+      "loss": 0.4936,
+      "step": 4990
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 7.777392864227295,
+      "learning_rate": 9.24812030075188e-06,
+      "loss": 0.5078,
+      "step": 5000
+    },
+    {
+      "epoch": 7.53,
+      "grad_norm": 11.22333812713623,
+      "learning_rate": 9.246616541353384e-06,
+      "loss": 0.575,
+      "step": 5010
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 6.031052589416504,
+      "learning_rate": 9.245112781954888e-06,
+      "loss": 0.4742,
+      "step": 5020
+    },
+    {
+      "epoch": 7.56,
+      "grad_norm": 11.427336692810059,
+      "learning_rate": 9.243609022556391e-06,
+      "loss": 0.5519,
+      "step": 5030
+    },
+    {
+      "epoch": 7.58,
+      "grad_norm": 6.76407527923584,
+      "learning_rate": 9.242105263157896e-06,
+      "loss": 0.5202,
+      "step": 5040
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 7.091256618499756,
+      "learning_rate": 9.240601503759398e-06,
+      "loss": 0.5255,
+      "step": 5050
+    },
+    {
+      "epoch": 7.61,
+      "grad_norm": 6.818326473236084,
+      "learning_rate": 9.239097744360903e-06,
+      "loss": 0.442,
+      "step": 5060
+    },
+    {
+      "epoch": 7.62,
+      "grad_norm": 7.494906902313232,
+      "learning_rate": 9.237593984962407e-06,
+      "loss": 0.5126,
+      "step": 5070
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 5.982577800750732,
+      "learning_rate": 9.23609022556391e-06,
+      "loss": 0.5056,
+      "step": 5080
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 4.815781116485596,
+      "learning_rate": 9.234586466165414e-06,
+      "loss": 0.4896,
+      "step": 5090
+    },
+    {
+      "epoch": 7.67,
+      "grad_norm": 10.674721717834473,
+      "learning_rate": 9.233082706766918e-06,
+      "loss": 0.5209,
+      "step": 5100
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 5.937568187713623,
+      "learning_rate": 9.231578947368421e-06,
+      "loss": 0.4772,
+      "step": 5110
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 5.146367073059082,
+      "learning_rate": 9.230075187969926e-06,
+      "loss": 0.5032,
+      "step": 5120
+    },
+    {
+      "epoch": 7.71,
+      "grad_norm": 8.03272819519043,
+      "learning_rate": 9.22857142857143e-06,
+      "loss": 0.5072,
+      "step": 5130
+    },
+    {
+      "epoch": 7.73,
+      "grad_norm": 5.361180782318115,
+      "learning_rate": 9.227067669172933e-06,
+      "loss": 0.5057,
+      "step": 5140
+    },
+    {
+      "epoch": 7.74,
+      "grad_norm": 5.487973213195801,
+      "learning_rate": 9.225563909774437e-06,
+      "loss": 0.5253,
+      "step": 5150
+    },
+    {
+      "epoch": 7.76,
+      "grad_norm": 6.845251560211182,
+      "learning_rate": 9.22406015037594e-06,
+      "loss": 0.5436,
+      "step": 5160
+    },
+    {
+      "epoch": 7.77,
+      "grad_norm": 4.931974411010742,
+      "learning_rate": 9.222556390977444e-06,
+      "loss": 0.4227,
+      "step": 5170
+    },
+    {
+      "epoch": 7.79,
+      "grad_norm": 7.382147312164307,
+      "learning_rate": 9.221052631578949e-06,
+      "loss": 0.5022,
+      "step": 5180
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 8.380685806274414,
+      "learning_rate": 9.219548872180451e-06,
+      "loss": 0.5259,
+      "step": 5190
+    },
+    {
+      "epoch": 7.82,
+      "grad_norm": 6.625802993774414,
+      "learning_rate": 9.218045112781956e-06,
+      "loss": 0.5213,
+      "step": 5200
+    },
+    {
+      "epoch": 7.83,
+      "grad_norm": 6.7276692390441895,
+      "learning_rate": 9.21654135338346e-06,
+      "loss": 0.4831,
+      "step": 5210
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 5.930064678192139,
+      "learning_rate": 9.215037593984963e-06,
+      "loss": 0.4278,
+      "step": 5220
+    },
+    {
+      "epoch": 7.86,
+      "grad_norm": 5.958808422088623,
+      "learning_rate": 9.213533834586467e-06,
+      "loss": 0.5039,
+      "step": 5230
+    },
+    {
+      "epoch": 7.88,
+      "grad_norm": 8.592114448547363,
+      "learning_rate": 9.21203007518797e-06,
+      "loss": 0.48,
+      "step": 5240
+    },
+    {
+      "epoch": 7.89,
+      "grad_norm": 7.286666393280029,
+      "learning_rate": 9.210526315789474e-06,
+      "loss": 0.5513,
+      "step": 5250
+    },
+    {
+      "epoch": 7.91,
+      "grad_norm": 3.8542234897613525,
+      "learning_rate": 9.209022556390979e-06,
+      "loss": 0.4452,
+      "step": 5260
+    },
+    {
+      "epoch": 7.92,
+      "grad_norm": 8.812358856201172,
+      "learning_rate": 9.207518796992482e-06,
+      "loss": 0.4796,
+      "step": 5270
+    },
+    {
+      "epoch": 7.94,
+      "grad_norm": 7.729457378387451,
+      "learning_rate": 9.206015037593986e-06,
+      "loss": 0.5283,
+      "step": 5280
+    },
+    {
+      "epoch": 7.95,
+      "grad_norm": 5.370766639709473,
+      "learning_rate": 9.20451127819549e-06,
+      "loss": 0.557,
+      "step": 5290
+    },
+    {
+      "epoch": 7.97,
+      "grad_norm": 7.00390625,
+      "learning_rate": 9.203007518796993e-06,
+      "loss": 0.5266,
+      "step": 5300
+    },
+    {
+      "epoch": 7.98,
+      "grad_norm": 5.945902347564697,
+      "learning_rate": 9.201503759398496e-06,
+      "loss": 0.505,
+      "step": 5310
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 39.0330924987793,
+      "learning_rate": 9.200000000000002e-06,
+      "loss": 0.5673,
+      "step": 5320
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.9324,
+      "eval_loss": 0.2575855553150177,
+      "eval_runtime": 85.1441,
+      "eval_samples_per_second": 117.448,
+      "eval_steps_per_second": 0.47,
+      "step": 5320
+    },
+    {
+      "epoch": 8.02,
+      "grad_norm": 7.23183012008667,
+      "learning_rate": 9.198496240601504e-06,
+      "loss": 0.4209,
+      "step": 5330
+    },
+    {
+      "epoch": 8.03,
+      "grad_norm": 5.357458591461182,
+      "learning_rate": 9.196992481203009e-06,
+      "loss": 0.4701,
+      "step": 5340
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 9.471860885620117,
+      "learning_rate": 9.195488721804512e-06,
+      "loss": 0.4924,
+      "step": 5350
+    },
+    {
+      "epoch": 8.06,
+      "grad_norm": 7.7437214851379395,
+      "learning_rate": 9.193984962406016e-06,
+      "loss": 0.4997,
+      "step": 5360
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 6.697991371154785,
+      "learning_rate": 9.19248120300752e-06,
+      "loss": 0.4365,
+      "step": 5370
+    },
+    {
+      "epoch": 8.09,
+      "grad_norm": 9.78630542755127,
+      "learning_rate": 9.190977443609025e-06,
+      "loss": 0.5572,
+      "step": 5380
+    },
+    {
+      "epoch": 8.11,
+      "grad_norm": 6.197582244873047,
+      "learning_rate": 9.189473684210526e-06,
+      "loss": 0.4581,
+      "step": 5390
+    },
+    {
+      "epoch": 8.12,
+      "grad_norm": 5.565506458282471,
+      "learning_rate": 9.187969924812032e-06,
+      "loss": 0.5219,
+      "step": 5400
+    },
+    {
+      "epoch": 8.14,
+      "grad_norm": 5.3856353759765625,
+      "learning_rate": 9.186466165413535e-06,
+      "loss": 0.5744,
+      "step": 5410
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 6.119697093963623,
+      "learning_rate": 9.184962406015039e-06,
+      "loss": 0.3964,
+      "step": 5420
+    },
+    {
+      "epoch": 8.17,
+      "grad_norm": 5.344573497772217,
+      "learning_rate": 9.183458646616542e-06,
+      "loss": 0.4816,
+      "step": 5430
+    },
+    {
+      "epoch": 8.18,
+      "grad_norm": 5.624027729034424,
+      "learning_rate": 9.181954887218046e-06,
+      "loss": 0.4341,
+      "step": 5440
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 4.847854137420654,
+      "learning_rate": 9.180451127819549e-06,
+      "loss": 0.5471,
+      "step": 5450
+    },
+    {
+      "epoch": 8.21,
+      "grad_norm": 6.051102638244629,
+      "learning_rate": 9.178947368421053e-06,
+      "loss": 0.5593,
+      "step": 5460
+    },
+    {
+      "epoch": 8.23,
+      "grad_norm": 8.345075607299805,
+      "learning_rate": 9.177443609022558e-06,
+      "loss": 0.4132,
+      "step": 5470
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 5.369195938110352,
+      "learning_rate": 9.17593984962406e-06,
+      "loss": 0.4844,
+      "step": 5480
+    },
+    {
+      "epoch": 8.26,
+      "grad_norm": 6.0165228843688965,
+      "learning_rate": 9.174436090225565e-06,
+      "loss": 0.4712,
+      "step": 5490
+    },
+    {
+      "epoch": 8.27,
+      "grad_norm": 4.54939079284668,
+      "learning_rate": 9.172932330827068e-06,
+      "loss": 0.4956,
+      "step": 5500
+    },
+    {
+      "epoch": 8.29,
+      "grad_norm": 5.021441459655762,
+      "learning_rate": 9.171428571428572e-06,
+      "loss": 0.4428,
+      "step": 5510
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 7.0452446937561035,
+      "learning_rate": 9.169924812030075e-06,
+      "loss": 0.5036,
+      "step": 5520
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 8.039512634277344,
+      "learning_rate": 9.168421052631579e-06,
+      "loss": 0.478,
+      "step": 5530
+    },
+    {
+      "epoch": 8.33,
+      "grad_norm": 6.608015537261963,
+      "learning_rate": 9.166917293233083e-06,
+      "loss": 0.4338,
+      "step": 5540
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 6.659971714019775,
+      "learning_rate": 9.165413533834588e-06,
+      "loss": 0.5033,
+      "step": 5550
+    },
+    {
+      "epoch": 8.36,
+      "grad_norm": 5.663388252258301,
+      "learning_rate": 9.163909774436091e-06,
+      "loss": 0.4779,
+      "step": 5560
+    },
+    {
+      "epoch": 8.38,
+      "grad_norm": 5.978389739990234,
+      "learning_rate": 9.162406015037595e-06,
+      "loss": 0.4615,
+      "step": 5570
+    },
+    {
+      "epoch": 8.39,
+      "grad_norm": 5.943080425262451,
+      "learning_rate": 9.160902255639098e-06,
+      "loss": 0.4282,
+      "step": 5580
+    },
+    {
+      "epoch": 8.41,
+      "grad_norm": 3.245058536529541,
+      "learning_rate": 9.159398496240602e-06,
+      "loss": 0.4423,
+      "step": 5590
+    },
+    {
+      "epoch": 8.42,
+      "grad_norm": 5.73254919052124,
+      "learning_rate": 9.157894736842105e-06,
+      "loss": 0.4777,
+      "step": 5600
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 6.487976551055908,
+      "learning_rate": 9.15639097744361e-06,
+      "loss": 0.5132,
+      "step": 5610
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 4.1268463134765625,
+      "learning_rate": 9.154887218045112e-06,
+      "loss": 0.3942,
+      "step": 5620
+    },
+    {
+      "epoch": 8.47,
+      "grad_norm": 7.593535900115967,
+      "learning_rate": 9.153383458646618e-06,
+      "loss": 0.5109,
+      "step": 5630
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 4.127936840057373,
+      "learning_rate": 9.151879699248121e-06,
+      "loss": 0.4554,
+      "step": 5640
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 12.721508026123047,
+      "learning_rate": 9.150375939849625e-06,
+      "loss": 0.5204,
+      "step": 5650
+    },
+    {
+      "epoch": 8.51,
+      "grad_norm": 7.618612289428711,
+      "learning_rate": 9.148872180451128e-06,
+      "loss": 0.5095,
+      "step": 5660
+    },
+    {
+      "epoch": 8.53,
+      "grad_norm": 5.089692115783691,
+      "learning_rate": 9.147368421052633e-06,
+      "loss": 0.478,
+      "step": 5670
+    },
+    {
+      "epoch": 8.54,
+      "grad_norm": 6.890159606933594,
+      "learning_rate": 9.145864661654135e-06,
+      "loss": 0.4634,
+      "step": 5680
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 9.275102615356445,
+      "learning_rate": 9.14436090225564e-06,
+      "loss": 0.528,
+      "step": 5690
+    },
+    {
+      "epoch": 8.57,
+      "grad_norm": 4.839653015136719,
+      "learning_rate": 9.142857142857144e-06,
+      "loss": 0.4558,
+      "step": 5700
+    },
+    {
+      "epoch": 8.59,
+      "grad_norm": 7.7605791091918945,
+      "learning_rate": 9.141353383458647e-06,
+      "loss": 0.5086,
+      "step": 5710
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 6.753016948699951,
+      "learning_rate": 9.139849624060151e-06,
+      "loss": 0.4953,
+      "step": 5720
+    },
+    {
+      "epoch": 8.62,
+      "grad_norm": 10.297369003295898,
+      "learning_rate": 9.138345864661654e-06,
+      "loss": 0.454,
+      "step": 5730
+    },
+    {
+      "epoch": 8.63,
+      "grad_norm": 5.704922676086426,
+      "learning_rate": 9.136842105263158e-06,
+      "loss": 0.4504,
+      "step": 5740
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 5.4977030754089355,
+      "learning_rate": 9.135338345864663e-06,
+      "loss": 0.5203,
+      "step": 5750
+    },
+    {
+      "epoch": 8.66,
+      "grad_norm": 7.531189918518066,
+      "learning_rate": 9.133834586466167e-06,
+      "loss": 0.4534,
+      "step": 5760
+    },
+    {
+      "epoch": 8.68,
+      "grad_norm": 6.660569667816162,
+      "learning_rate": 9.13233082706767e-06,
+      "loss": 0.577,
+      "step": 5770
+    },
+    {
+      "epoch": 8.69,
+      "grad_norm": 6.752721309661865,
+      "learning_rate": 9.130827067669174e-06,
+      "loss": 0.4574,
+      "step": 5780
+    },
+    {
+      "epoch": 8.71,
+      "grad_norm": 7.526786804199219,
+      "learning_rate": 9.129323308270677e-06,
+      "loss": 0.4511,
+      "step": 5790
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 6.5963239669799805,
+      "learning_rate": 9.12781954887218e-06,
+      "loss": 0.4917,
+      "step": 5800
+    },
+    {
+      "epoch": 8.74,
+      "grad_norm": 4.203681468963623,
+      "learning_rate": 9.126315789473686e-06,
+      "loss": 0.5218,
+      "step": 5810
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 4.367255210876465,
+      "learning_rate": 9.124812030075188e-06,
+      "loss": 0.4162,
+      "step": 5820
+    },
+    {
+      "epoch": 8.77,
+      "grad_norm": 3.8670730590820312,
+      "learning_rate": 9.123308270676693e-06,
+      "loss": 0.3913,
+      "step": 5830
+    },
+    {
+      "epoch": 8.78,
+      "grad_norm": 9.634267807006836,
+      "learning_rate": 9.121804511278197e-06,
+      "loss": 0.5062,
+      "step": 5840
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 2.509295701980591,
+      "learning_rate": 9.1203007518797e-06,
+      "loss": 0.4422,
+      "step": 5850
+    },
+    {
+      "epoch": 8.81,
+      "grad_norm": 7.6244659423828125,
+      "learning_rate": 9.118796992481204e-06,
+      "loss": 0.4566,
+      "step": 5860
+    },
+    {
+      "epoch": 8.83,
+      "grad_norm": 6.837118625640869,
+      "learning_rate": 9.117293233082709e-06,
+      "loss": 0.5095,
+      "step": 5870
+    },
+    {
+      "epoch": 8.84,
+      "grad_norm": 4.819979667663574,
+      "learning_rate": 9.11578947368421e-06,
+      "loss": 0.4471,
+      "step": 5880
+    },
+    {
+      "epoch": 8.86,
+      "grad_norm": 6.350512504577637,
+      "learning_rate": 9.114285714285716e-06,
+      "loss": 0.4751,
+      "step": 5890
+    },
+    {
+      "epoch": 8.87,
+      "grad_norm": 3.4793074131011963,
+      "learning_rate": 9.11278195488722e-06,
+      "loss": 0.4089,
+      "step": 5900
+    },
+    {
+      "epoch": 8.89,
+      "grad_norm": 5.1062774658203125,
+      "learning_rate": 9.111278195488723e-06,
+      "loss": 0.5624,
+      "step": 5910
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 8.126543998718262,
+      "learning_rate": 9.109774436090226e-06,
+      "loss": 0.5146,
+      "step": 5920
+    },
+    {
+      "epoch": 8.92,
+      "grad_norm": 7.661808967590332,
+      "learning_rate": 9.10827067669173e-06,
+      "loss": 0.4602,
+      "step": 5930
+    },
+    {
+      "epoch": 8.93,
+      "grad_norm": 6.820888996124268,
+      "learning_rate": 9.106766917293233e-06,
+      "loss": 0.4827,
+      "step": 5940
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 24.169485092163086,
+      "learning_rate": 9.105263157894739e-06,
+      "loss": 0.5302,
+      "step": 5950
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 5.068043231964111,
+      "learning_rate": 9.103759398496242e-06,
+      "loss": 0.5453,
+      "step": 5960
+    },
+    {
+      "epoch": 8.98,
+      "grad_norm": 5.819450378417969,
+      "learning_rate": 9.102255639097744e-06,
+      "loss": 0.4811,
+      "step": 5970
+    },
+    {
+      "epoch": 8.99,
+      "grad_norm": 4.129781723022461,
+      "learning_rate": 9.10075187969925e-06,
+      "loss": 0.4613,
+      "step": 5980
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.9311,
+      "eval_loss": 0.25862327218055725,
+      "eval_runtime": 84.8621,
+      "eval_samples_per_second": 117.838,
+      "eval_steps_per_second": 0.471,
+      "step": 5985
+    },
+    {
+      "epoch": 9.01,
+      "grad_norm": 6.844424247741699,
+      "learning_rate": 9.099248120300753e-06,
+      "loss": 0.4154,
+      "step": 5990
+    },
+    {
+      "epoch": 9.02,
+      "grad_norm": 5.792689323425293,
+      "learning_rate": 9.097744360902256e-06,
+      "loss": 0.4345,
+      "step": 6000
+    },
+    {
+      "epoch": 9.04,
+      "grad_norm": 5.300471305847168,
+      "learning_rate": 9.09624060150376e-06,
+      "loss": 0.3986,
+      "step": 6010
+    },
+    {
+      "epoch": 9.05,
+      "grad_norm": 11.9564208984375,
+      "learning_rate": 9.094736842105263e-06,
+      "loss": 0.4255,
+      "step": 6020
+    },
+    {
+      "epoch": 9.07,
+      "grad_norm": 6.798022270202637,
+      "learning_rate": 9.093233082706767e-06,
+      "loss": 0.4296,
+      "step": 6030
+    },
+    {
+      "epoch": 9.08,
+      "grad_norm": 7.78212308883667,
+      "learning_rate": 9.091729323308272e-06,
+      "loss": 0.4373,
+      "step": 6040
+    },
+    {
+      "epoch": 9.1,
+      "grad_norm": 6.719583988189697,
+      "learning_rate": 9.090225563909776e-06,
+      "loss": 0.4943,
+      "step": 6050
+    },
+    {
+      "epoch": 9.11,
+      "grad_norm": 8.298834800720215,
+      "learning_rate": 9.088721804511279e-06,
+      "loss": 0.3937,
+      "step": 6060
+    },
+    {
+      "epoch": 9.13,
+      "grad_norm": 4.731727600097656,
+      "learning_rate": 9.087218045112783e-06,
+      "loss": 0.4711,
+      "step": 6070
+    },
+    {
+      "epoch": 9.14,
+      "grad_norm": 6.207810878753662,
+      "learning_rate": 9.085714285714286e-06,
+      "loss": 0.4565,
+      "step": 6080
+    },
+    {
+      "epoch": 9.16,
+      "grad_norm": 6.939966678619385,
+      "learning_rate": 9.08421052631579e-06,
+      "loss": 0.5276,
+      "step": 6090
+    },
+    {
+      "epoch": 9.17,
+      "grad_norm": 9.00831127166748,
+      "learning_rate": 9.082706766917295e-06,
+      "loss": 0.5558,
+      "step": 6100
+    },
+    {
+      "epoch": 9.19,
+      "grad_norm": 4.730199813842773,
+      "learning_rate": 9.081203007518797e-06,
+      "loss": 0.4826,
+      "step": 6110
+    },
+    {
+      "epoch": 9.2,
+      "grad_norm": 4.198337078094482,
+      "learning_rate": 9.079699248120302e-06,
+      "loss": 0.3978,
+      "step": 6120
+    },
+    {
+      "epoch": 9.22,
+      "grad_norm": 5.722704887390137,
+      "learning_rate": 9.078195488721805e-06,
+      "loss": 0.4622,
+      "step": 6130
+    },
+    {
+      "epoch": 9.23,
+      "grad_norm": 8.497228622436523,
+      "learning_rate": 9.076691729323309e-06,
+      "loss": 0.4726,
+      "step": 6140
+    },
+    {
+      "epoch": 9.25,
+      "grad_norm": 7.046009063720703,
+      "learning_rate": 9.075187969924812e-06,
+      "loss": 0.527,
+      "step": 6150
+    },
+    {
+      "epoch": 9.26,
+      "grad_norm": 7.972896099090576,
+      "learning_rate": 9.073684210526316e-06,
+      "loss": 0.3072,
+      "step": 6160
+    },
+    {
+      "epoch": 9.28,
+      "grad_norm": 8.850788116455078,
+      "learning_rate": 9.07218045112782e-06,
+      "loss": 0.4753,
+      "step": 6170
+    },
+    {
+      "epoch": 9.29,
+      "grad_norm": 6.064061641693115,
+      "learning_rate": 9.070676691729325e-06,
+      "loss": 0.4009,
+      "step": 6180
+    },
+    {
+      "epoch": 9.31,
+      "grad_norm": 6.12713098526001,
+      "learning_rate": 9.069172932330828e-06,
+      "loss": 0.4786,
+      "step": 6190
+    },
+    {
+      "epoch": 9.32,
+      "grad_norm": 8.643204689025879,
+      "learning_rate": 9.067669172932332e-06,
+      "loss": 0.5134,
+      "step": 6200
+    },
+    {
+      "epoch": 9.34,
+      "grad_norm": 7.257277488708496,
+      "learning_rate": 9.066165413533835e-06,
+      "loss": 0.4474,
+      "step": 6210
+    },
+    {
+      "epoch": 9.35,
+      "grad_norm": 7.62333869934082,
+      "learning_rate": 9.064661654135339e-06,
+      "loss": 0.4579,
+      "step": 6220
+    },
+    {
+      "epoch": 9.37,
+      "grad_norm": 9.346735954284668,
+      "learning_rate": 9.063157894736842e-06,
+      "loss": 0.4714,
+      "step": 6230
+    },
+    {
+      "epoch": 9.38,
+      "grad_norm": 3.8007750511169434,
+      "learning_rate": 9.061654135338347e-06,
+      "loss": 0.4334,
+      "step": 6240
+    },
+    {
+      "epoch": 9.4,
+      "grad_norm": 6.266302108764648,
+      "learning_rate": 9.06015037593985e-06,
+      "loss": 0.4704,
+      "step": 6250
+    },
+    {
+      "epoch": 9.41,
+      "grad_norm": 6.959786891937256,
+      "learning_rate": 9.058646616541355e-06,
+      "loss": 0.5353,
+      "step": 6260
+    },
+    {
+      "epoch": 9.43,
+      "grad_norm": 6.572616100311279,
+      "learning_rate": 9.057142857142858e-06,
+      "loss": 0.4629,
+      "step": 6270
+    },
+    {
+      "epoch": 9.44,
+      "grad_norm": 5.961916446685791,
+      "learning_rate": 9.055639097744362e-06,
+      "loss": 0.5119,
+      "step": 6280
+    },
+    {
+      "epoch": 9.46,
+      "grad_norm": 6.547915935516357,
+      "learning_rate": 9.054135338345865e-06,
+      "loss": 0.4576,
+      "step": 6290
+    },
+    {
+      "epoch": 9.47,
+      "grad_norm": 6.359402179718018,
+      "learning_rate": 9.05263157894737e-06,
+      "loss": 0.3692,
+      "step": 6300
+    },
+    {
+      "epoch": 9.49,
+      "grad_norm": 7.048614501953125,
+      "learning_rate": 9.051127819548872e-06,
+      "loss": 0.3911,
+      "step": 6310
+    },
+    {
+      "epoch": 9.5,
+      "grad_norm": 5.198198318481445,
+      "learning_rate": 9.049624060150377e-06,
+      "loss": 0.475,
+      "step": 6320
+    },
+    {
+      "epoch": 9.52,
+      "grad_norm": 5.797221660614014,
+      "learning_rate": 9.04812030075188e-06,
+      "loss": 0.3771,
+      "step": 6330
+    },
+    {
+      "epoch": 9.53,
+      "grad_norm": 5.751585483551025,
+      "learning_rate": 9.046616541353384e-06,
+      "loss": 0.497,
+      "step": 6340
+    },
+    {
+      "epoch": 9.55,
+      "grad_norm": 9.54306697845459,
+      "learning_rate": 9.045112781954888e-06,
+      "loss": 0.3976,
+      "step": 6350
+    },
+    {
+      "epoch": 9.56,
+      "grad_norm": 8.968032836914062,
+      "learning_rate": 9.043609022556391e-06,
+      "loss": 0.4252,
+      "step": 6360
+    },
+    {
+      "epoch": 9.58,
+      "grad_norm": 13.218304634094238,
+      "learning_rate": 9.042105263157895e-06,
+      "loss": 0.5191,
+      "step": 6370
+    },
+    {
+      "epoch": 9.59,
+      "grad_norm": 4.405686855316162,
+      "learning_rate": 9.0406015037594e-06,
+      "loss": 0.4149,
+      "step": 6380
+    },
+    {
+      "epoch": 9.61,
+      "grad_norm": 4.863158702850342,
+      "learning_rate": 9.039097744360904e-06,
+      "loss": 0.4875,
+      "step": 6390
+    },
+    {
+      "epoch": 9.62,
+      "grad_norm": 6.247385501861572,
+      "learning_rate": 9.037593984962407e-06,
+      "loss": 0.4106,
+      "step": 6400
+    },
+    {
+      "epoch": 9.64,
+      "grad_norm": 6.554888725280762,
+      "learning_rate": 9.03609022556391e-06,
+      "loss": 0.4645,
+      "step": 6410
+    },
+    {
+      "epoch": 9.65,
+      "grad_norm": 7.249465465545654,
+      "learning_rate": 9.034586466165414e-06,
+      "loss": 0.4291,
+      "step": 6420
+    },
+    {
+      "epoch": 9.67,
+      "grad_norm": 6.810882568359375,
+      "learning_rate": 9.033082706766918e-06,
+      "loss": 0.4637,
+      "step": 6430
+    },
+    {
+      "epoch": 9.68,
+      "grad_norm": 2.65733003616333,
+      "learning_rate": 9.031578947368423e-06,
+      "loss": 0.4585,
+      "step": 6440
+    },
+    {
+      "epoch": 9.7,
+      "grad_norm": 8.63343334197998,
+      "learning_rate": 9.030075187969925e-06,
+      "loss": 0.5187,
+      "step": 6450
+    },
+    {
+      "epoch": 9.71,
+      "grad_norm": 5.558303356170654,
+      "learning_rate": 9.028571428571428e-06,
+      "loss": 0.4819,
+      "step": 6460
+    },
+    {
+      "epoch": 9.73,
+      "grad_norm": 8.711833000183105,
+      "learning_rate": 9.027067669172933e-06,
+      "loss": 0.4585,
+      "step": 6470
+    },
+    {
+      "epoch": 9.74,
+      "grad_norm": 7.882017135620117,
+      "learning_rate": 9.025563909774437e-06,
+      "loss": 0.5154,
+      "step": 6480
+    },
+    {
+      "epoch": 9.76,
+      "grad_norm": 5.849830150604248,
+      "learning_rate": 9.02406015037594e-06,
+      "loss": 0.5093,
+      "step": 6490
+    },
+    {
+      "epoch": 9.77,
+      "grad_norm": 7.546263217926025,
+      "learning_rate": 9.022556390977444e-06,
+      "loss": 0.3969,
+      "step": 6500
+    },
+    {
+      "epoch": 9.79,
+      "grad_norm": 7.114614963531494,
+      "learning_rate": 9.021052631578948e-06,
+      "loss": 0.4485,
+      "step": 6510
+    },
+    {
+      "epoch": 9.8,
+      "grad_norm": 7.6169209480285645,
+      "learning_rate": 9.019548872180451e-06,
+      "loss": 0.425,
+      "step": 6520
+    },
+    {
+      "epoch": 9.82,
+      "grad_norm": 5.7843403816223145,
+      "learning_rate": 9.018045112781956e-06,
+      "loss": 0.4171,
+      "step": 6530
+    },
+    {
+      "epoch": 9.83,
+      "grad_norm": 6.0503082275390625,
+      "learning_rate": 9.016541353383458e-06,
+      "loss": 0.483,
+      "step": 6540
+    },
+    {
+      "epoch": 9.85,
+      "grad_norm": 7.677584648132324,
+      "learning_rate": 9.015037593984963e-06,
+      "loss": 0.4747,
+      "step": 6550
+    },
+    {
+      "epoch": 9.86,
+      "grad_norm": 5.793139934539795,
+      "learning_rate": 9.013533834586467e-06,
+      "loss": 0.3621,
+      "step": 6560
+    },
+    {
+      "epoch": 9.88,
+      "grad_norm": 6.399969577789307,
+      "learning_rate": 9.01203007518797e-06,
+      "loss": 0.4373,
+      "step": 6570
+    },
+    {
+      "epoch": 9.89,
+      "grad_norm": 10.296338081359863,
+      "learning_rate": 9.010526315789474e-06,
+      "loss": 0.417,
+      "step": 6580
+    },
+    {
+      "epoch": 9.91,
+      "grad_norm": 6.193917274475098,
+      "learning_rate": 9.009022556390979e-06,
+      "loss": 0.4419,
+      "step": 6590
+    },
+    {
+      "epoch": 9.92,
+      "grad_norm": 3.921016216278076,
+      "learning_rate": 9.007518796992481e-06,
+      "loss": 0.3981,
+      "step": 6600
+    },
+    {
+      "epoch": 9.94,
+      "grad_norm": 6.30132532119751,
+      "learning_rate": 9.006015037593986e-06,
+      "loss": 0.4699,
+      "step": 6610
+    },
+    {
+      "epoch": 9.95,
+      "grad_norm": 8.901771545410156,
+      "learning_rate": 9.00451127819549e-06,
+      "loss": 0.4308,
+      "step": 6620
+    },
+    {
+      "epoch": 9.97,
+      "grad_norm": 5.031552314758301,
+      "learning_rate": 9.003007518796993e-06,
+      "loss": 0.48,
+      "step": 6630
+    },
+    {
+      "epoch": 9.98,
+      "grad_norm": 5.636510372161865,
+      "learning_rate": 9.001503759398497e-06,
+      "loss": 0.403,
+      "step": 6640
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 3.5294342041015625,
+      "learning_rate": 9e-06,
+      "loss": 0.4179,
+      "step": 6650
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.9285,
+      "eval_loss": 0.2555387318134308,
+      "eval_runtime": 84.7099,
+      "eval_samples_per_second": 118.05,
+      "eval_steps_per_second": 0.472,
+      "step": 6650
+    },
+    {
+      "epoch": 10.02,
+      "grad_norm": 6.522907257080078,
+      "learning_rate": 8.998496240601504e-06,
+      "loss": 0.4525,
+      "step": 6660
+    },
+    {
+      "epoch": 10.03,
+      "grad_norm": 6.142210006713867,
+      "learning_rate": 8.996992481203009e-06,
+      "loss": 0.3998,
+      "step": 6670
+    },
+    {
+      "epoch": 10.05,
+      "grad_norm": 7.781100749969482,
+      "learning_rate": 8.995488721804512e-06,
+      "loss": 0.4122,
+      "step": 6680
+    },
+    {
+      "epoch": 10.06,
+      "grad_norm": 5.448252201080322,
+      "learning_rate": 8.993984962406016e-06,
+      "loss": 0.461,
+      "step": 6690
+    },
+    {
+      "epoch": 10.08,
+      "grad_norm": 7.063671588897705,
+      "learning_rate": 8.99248120300752e-06,
+      "loss": 0.432,
+      "step": 6700
+    },
+    {
+      "epoch": 10.09,
+      "grad_norm": 6.696626663208008,
+      "learning_rate": 8.990977443609023e-06,
+      "loss": 0.4208,
+      "step": 6710
+    },
+    {
+      "epoch": 10.11,
+      "grad_norm": 6.5666656494140625,
+      "learning_rate": 8.989473684210527e-06,
+      "loss": 0.4527,
+      "step": 6720
+    },
+    {
+      "epoch": 10.12,
+      "grad_norm": 8.801324844360352,
+      "learning_rate": 8.987969924812032e-06,
+      "loss": 0.442,
+      "step": 6730
+    },
+    {
+      "epoch": 10.14,
+      "grad_norm": 6.743152141571045,
+      "learning_rate": 8.986466165413534e-06,
+      "loss": 0.4424,
+      "step": 6740
+    },
+    {
+      "epoch": 10.15,
+      "grad_norm": 5.408703327178955,
+      "learning_rate": 8.984962406015039e-06,
+      "loss": 0.4548,
+      "step": 6750
+    },
+    {
+      "epoch": 10.17,
+      "grad_norm": 8.466784477233887,
+      "learning_rate": 8.983458646616542e-06,
+      "loss": 0.4355,
+      "step": 6760
+    },
+    {
+      "epoch": 10.18,
+      "grad_norm": 5.309767723083496,
+      "learning_rate": 8.981954887218046e-06,
+      "loss": 0.4023,
+      "step": 6770
+    },
+    {
+      "epoch": 10.2,
+      "grad_norm": 3.3604421615600586,
+      "learning_rate": 8.98045112781955e-06,
+      "loss": 0.4367,
+      "step": 6780
+    },
+    {
+      "epoch": 10.21,
+      "grad_norm": 6.275347709655762,
+      "learning_rate": 8.978947368421055e-06,
+      "loss": 0.4764,
+      "step": 6790
+    },
+    {
+      "epoch": 10.23,
+      "grad_norm": 6.770579814910889,
+      "learning_rate": 8.977443609022556e-06,
+      "loss": 0.4414,
+      "step": 6800
+    },
+    {
+      "epoch": 10.24,
+      "grad_norm": 8.56733512878418,
+      "learning_rate": 8.975939849624062e-06,
+      "loss": 0.4426,
+      "step": 6810
+    },
+    {
+      "epoch": 10.26,
+      "grad_norm": 6.006712436676025,
+      "learning_rate": 8.974436090225565e-06,
+      "loss": 0.3702,
+      "step": 6820
+    },
+    {
+      "epoch": 10.27,
+      "grad_norm": 4.649052143096924,
+      "learning_rate": 8.972932330827069e-06,
+      "loss": 0.5371,
+      "step": 6830
+    },
+    {
+      "epoch": 10.29,
+      "grad_norm": 9.080769538879395,
+      "learning_rate": 8.971428571428572e-06,
+      "loss": 0.4866,
+      "step": 6840
+    },
+    {
+      "epoch": 10.3,
+      "grad_norm": 5.778624534606934,
+      "learning_rate": 8.969924812030076e-06,
+      "loss": 0.381,
+      "step": 6850
+    },
+    {
+      "epoch": 10.32,
+      "grad_norm": 7.814187049865723,
+      "learning_rate": 8.96842105263158e-06,
+      "loss": 0.5162,
+      "step": 6860
+    },
+    {
+      "epoch": 10.33,
+      "grad_norm": 5.049838542938232,
+      "learning_rate": 8.966917293233084e-06,
+      "loss": 0.4879,
+      "step": 6870
+    },
+    {
+      "epoch": 10.35,
+      "grad_norm": 8.096096992492676,
+      "learning_rate": 8.965413533834588e-06,
+      "loss": 0.4726,
+      "step": 6880
+    },
+    {
+      "epoch": 10.36,
+      "grad_norm": 7.028320789337158,
+      "learning_rate": 8.963909774436091e-06,
+      "loss": 0.4424,
+      "step": 6890
+    },
+    {
+      "epoch": 10.38,
+      "grad_norm": 4.826821804046631,
+      "learning_rate": 8.962406015037595e-06,
+      "loss": 0.4552,
+      "step": 6900
+    },
+    {
+      "epoch": 10.39,
+      "grad_norm": 8.392495155334473,
+      "learning_rate": 8.960902255639098e-06,
+      "loss": 0.4378,
+      "step": 6910
+    },
+    {
+      "epoch": 10.41,
+      "grad_norm": 4.868290424346924,
+      "learning_rate": 8.959398496240602e-06,
+      "loss": 0.4151,
+      "step": 6920
+    },
+    {
+      "epoch": 10.42,
+      "grad_norm": 6.117234230041504,
+      "learning_rate": 8.957894736842107e-06,
+      "loss": 0.5149,
+      "step": 6930
+    },
+    {
+      "epoch": 10.44,
+      "grad_norm": 9.33238697052002,
+      "learning_rate": 8.956390977443609e-06,
+      "loss": 0.3984,
+      "step": 6940
+    },
+    {
+      "epoch": 10.45,
+      "grad_norm": 9.559886932373047,
+      "learning_rate": 8.954887218045113e-06,
+      "loss": 0.4171,
+      "step": 6950
+    },
+    {
+      "epoch": 10.47,
+      "grad_norm": 4.344634056091309,
+      "learning_rate": 8.953383458646618e-06,
+      "loss": 0.4419,
+      "step": 6960
+    },
+    {
+      "epoch": 10.48,
+      "grad_norm": 5.508487701416016,
+      "learning_rate": 8.951879699248121e-06,
+      "loss": 0.4605,
+      "step": 6970
+    },
+    {
+      "epoch": 10.5,
+      "grad_norm": 5.529686450958252,
+      "learning_rate": 8.950375939849625e-06,
+      "loss": 0.4004,
+      "step": 6980
+    },
+    {
+      "epoch": 10.51,
+      "grad_norm": 5.424170970916748,
+      "learning_rate": 8.948872180451128e-06,
+      "loss": 0.4351,
+      "step": 6990
+    },
+    {
+      "epoch": 10.53,
+      "grad_norm": 6.121506690979004,
+      "learning_rate": 8.947368421052632e-06,
+      "loss": 0.4198,
+      "step": 7000
+    },
+    {
+      "epoch": 10.54,
+      "grad_norm": 4.664872169494629,
+      "learning_rate": 8.945864661654135e-06,
+      "loss": 0.3917,
+      "step": 7010
+    },
+    {
+      "epoch": 10.56,
+      "grad_norm": 5.378602027893066,
+      "learning_rate": 8.94436090225564e-06,
+      "loss": 0.47,
+      "step": 7020
+    },
+    {
+      "epoch": 10.57,
+      "grad_norm": 8.281057357788086,
+      "learning_rate": 8.942857142857142e-06,
+      "loss": 0.3779,
+      "step": 7030
+    },
+    {
+      "epoch": 10.59,
+      "grad_norm": 5.378328800201416,
+      "learning_rate": 8.941353383458648e-06,
+      "loss": 0.4878,
+      "step": 7040
+    },
+    {
+      "epoch": 10.6,
+      "grad_norm": 4.809008598327637,
+      "learning_rate": 8.939849624060151e-06,
+      "loss": 0.4409,
+      "step": 7050
+    },
+    {
+      "epoch": 10.62,
+      "grad_norm": 6.703794002532959,
+      "learning_rate": 8.938345864661655e-06,
+      "loss": 0.45,
+      "step": 7060
+    },
+    {
+      "epoch": 10.63,
+      "grad_norm": 10.097111701965332,
+      "learning_rate": 8.936842105263158e-06,
+      "loss": 0.4442,
+      "step": 7070
+    },
+    {
+      "epoch": 10.65,
+      "grad_norm": 5.404522895812988,
+      "learning_rate": 8.935338345864662e-06,
+      "loss": 0.4289,
+      "step": 7080
+    },
+    {
+      "epoch": 10.66,
+      "grad_norm": 2.983161449432373,
+      "learning_rate": 8.933834586466165e-06,
+      "loss": 0.4072,
+      "step": 7090
+    },
+    {
+      "epoch": 10.68,
+      "grad_norm": 6.501340389251709,
+      "learning_rate": 8.93233082706767e-06,
+      "loss": 0.4344,
+      "step": 7100
+    },
+    {
+      "epoch": 10.69,
+      "grad_norm": 7.439212322235107,
+      "learning_rate": 8.930827067669174e-06,
+      "loss": 0.4208,
+      "step": 7110
+    },
+    {
+      "epoch": 10.71,
+      "grad_norm": 6.9180192947387695,
+      "learning_rate": 8.929323308270677e-06,
+      "loss": 0.4901,
+      "step": 7120
+    },
+    {
+      "epoch": 10.72,
+      "grad_norm": 4.9598212242126465,
+      "learning_rate": 8.927819548872181e-06,
+      "loss": 0.4551,
+      "step": 7130
+    },
+    {
+      "epoch": 10.74,
+      "grad_norm": 7.020519256591797,
+      "learning_rate": 8.926315789473685e-06,
+      "loss": 0.4469,
+      "step": 7140
+    },
+    {
+      "epoch": 10.75,
+      "grad_norm": 6.747496604919434,
+      "learning_rate": 8.924812030075188e-06,
+      "loss": 0.3626,
+      "step": 7150
+    },
+    {
+      "epoch": 10.77,
+      "grad_norm": 2.869495153427124,
+      "learning_rate": 8.923308270676693e-06,
+      "loss": 0.3794,
+      "step": 7160
+    },
+    {
+      "epoch": 10.78,
+      "grad_norm": 7.156761169433594,
+      "learning_rate": 8.921804511278195e-06,
+      "loss": 0.4909,
+      "step": 7170
+    },
+    {
+      "epoch": 10.8,
+      "grad_norm": 9.461006164550781,
+      "learning_rate": 8.9203007518797e-06,
+      "loss": 0.4487,
+      "step": 7180
+    },
+    {
+      "epoch": 10.81,
+      "grad_norm": 5.75421142578125,
+      "learning_rate": 8.918796992481204e-06,
+      "loss": 0.4953,
+      "step": 7190
+    },
+    {
+      "epoch": 10.83,
+      "grad_norm": 4.186371326446533,
+      "learning_rate": 8.917293233082707e-06,
+      "loss": 0.3788,
+      "step": 7200
+    },
+    {
+      "epoch": 10.84,
+      "grad_norm": 6.402685165405273,
+      "learning_rate": 8.915789473684211e-06,
+      "loss": 0.5009,
+      "step": 7210
+    },
+    {
+      "epoch": 10.86,
+      "grad_norm": 10.709757804870605,
+      "learning_rate": 8.914285714285716e-06,
+      "loss": 0.5308,
+      "step": 7220
+    },
+    {
+      "epoch": 10.87,
+      "grad_norm": 8.926152229309082,
+      "learning_rate": 8.912781954887218e-06,
+      "loss": 0.4461,
+      "step": 7230
+    },
+    {
+      "epoch": 10.89,
+      "grad_norm": 6.41901969909668,
+      "learning_rate": 8.911278195488723e-06,
+      "loss": 0.4188,
+      "step": 7240
+    },
+    {
+      "epoch": 10.9,
+      "grad_norm": 4.931794166564941,
+      "learning_rate": 8.909774436090227e-06,
+      "loss": 0.4089,
+      "step": 7250
+    },
+    {
+      "epoch": 10.92,
+      "grad_norm": 7.75593376159668,
+      "learning_rate": 8.90827067669173e-06,
+      "loss": 0.5261,
+      "step": 7260
+    },
+    {
+      "epoch": 10.93,
+      "grad_norm": 9.013036727905273,
+      "learning_rate": 8.906766917293234e-06,
+      "loss": 0.3274,
+      "step": 7270
+    },
+    {
+      "epoch": 10.95,
+      "grad_norm": 6.653579235076904,
+      "learning_rate": 8.905263157894737e-06,
+      "loss": 0.4646,
+      "step": 7280
+    },
+    {
+      "epoch": 10.96,
+      "grad_norm": 5.304203987121582,
+      "learning_rate": 8.90375939849624e-06,
+      "loss": 0.4399,
+      "step": 7290
+    },
+    {
+      "epoch": 10.98,
+      "grad_norm": 40.1646842956543,
+      "learning_rate": 8.902255639097746e-06,
+      "loss": 0.3007,
+      "step": 7300
+    },
+    {
+      "epoch": 10.99,
+      "grad_norm": 5.538785934448242,
+      "learning_rate": 8.90075187969925e-06,
+      "loss": 0.4438,
+      "step": 7310
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.9316,
+      "eval_loss": 0.25541195273399353,
+      "eval_runtime": 84.8107,
+      "eval_samples_per_second": 117.91,
+      "eval_steps_per_second": 0.472,
+      "step": 7315
+    },
+    {
+      "epoch": 11.01,
+      "grad_norm": 8.498943328857422,
+      "learning_rate": 8.899248120300753e-06,
+      "loss": 0.4009,
+      "step": 7320
+    },
+    {
+      "epoch": 11.02,
+      "grad_norm": 6.2147040367126465,
+      "learning_rate": 8.897744360902256e-06,
+      "loss": 0.4283,
+      "step": 7330
+    },
+    {
+      "epoch": 11.04,
+      "grad_norm": 5.028774261474609,
+      "learning_rate": 8.89624060150376e-06,
+      "loss": 0.4145,
+      "step": 7340
+    },
+    {
+      "epoch": 11.05,
+      "grad_norm": 7.040588855743408,
+      "learning_rate": 8.894736842105264e-06,
+      "loss": 0.3753,
+      "step": 7350
+    },
+    {
+      "epoch": 11.07,
+      "grad_norm": 4.658559322357178,
+      "learning_rate": 8.893233082706769e-06,
+      "loss": 0.4139,
+      "step": 7360
+    },
+    {
+      "epoch": 11.08,
+      "grad_norm": 7.867548942565918,
+      "learning_rate": 8.89172932330827e-06,
+      "loss": 0.4854,
+      "step": 7370
+    },
+    {
+      "epoch": 11.1,
+      "grad_norm": 8.354945182800293,
+      "learning_rate": 8.890225563909776e-06,
+      "loss": 0.4186,
+      "step": 7380
+    },
+    {
+      "epoch": 11.11,
+      "grad_norm": 6.198273658752441,
+      "learning_rate": 8.88872180451128e-06,
+      "loss": 0.4486,
+      "step": 7390
+    },
+    {
+      "epoch": 11.13,
+      "grad_norm": 7.401607990264893,
+      "learning_rate": 8.887218045112783e-06,
+      "loss": 0.4015,
+      "step": 7400
+    },
+    {
+      "epoch": 11.14,
+      "grad_norm": 5.412950038909912,
+      "learning_rate": 8.885714285714286e-06,
+      "loss": 0.3654,
+      "step": 7410
+    },
+    {
+      "epoch": 11.16,
+      "grad_norm": 3.7357654571533203,
+      "learning_rate": 8.884210526315792e-06,
+      "loss": 0.4165,
+      "step": 7420
+    },
+    {
+      "epoch": 11.17,
+      "grad_norm": 7.468185901641846,
+      "learning_rate": 8.882706766917293e-06,
+      "loss": 0.3241,
+      "step": 7430
+    },
+    {
+      "epoch": 11.19,
+      "grad_norm": 5.967494487762451,
+      "learning_rate": 8.881203007518799e-06,
+      "loss": 0.4364,
+      "step": 7440
+    },
+    {
+      "epoch": 11.2,
+      "grad_norm": 8.94781494140625,
+      "learning_rate": 8.879699248120302e-06,
+      "loss": 0.4927,
+      "step": 7450
+    },
+    {
+      "epoch": 11.22,
+      "grad_norm": 7.6195969581604,
+      "learning_rate": 8.878195488721804e-06,
+      "loss": 0.3722,
+      "step": 7460
+    },
+    {
+      "epoch": 11.23,
+      "grad_norm": 9.522473335266113,
+      "learning_rate": 8.876691729323309e-06,
+      "loss": 0.3974,
+      "step": 7470
+    },
+    {
+      "epoch": 11.25,
+      "grad_norm": 9.590860366821289,
+      "learning_rate": 8.875187969924813e-06,
+      "loss": 0.3842,
+      "step": 7480
+    },
+    {
+      "epoch": 11.26,
+      "grad_norm": 6.479350566864014,
+      "learning_rate": 8.873684210526316e-06,
+      "loss": 0.4135,
+      "step": 7490
+    },
+    {
+      "epoch": 11.28,
+      "grad_norm": 8.100231170654297,
+      "learning_rate": 8.87218045112782e-06,
+      "loss": 0.4262,
+      "step": 7500
+    },
+    {
+      "epoch": 11.29,
+      "grad_norm": 9.401702880859375,
+      "learning_rate": 8.870676691729325e-06,
+      "loss": 0.3899,
+      "step": 7510
+    },
+    {
+      "epoch": 11.31,
+      "grad_norm": 7.8885626792907715,
+      "learning_rate": 8.869172932330827e-06,
+      "loss": 0.4738,
+      "step": 7520
+    },
+    {
+      "epoch": 11.32,
+      "grad_norm": 7.2377753257751465,
+      "learning_rate": 8.867669172932332e-06,
+      "loss": 0.3686,
+      "step": 7530
+    },
+    {
+      "epoch": 11.34,
+      "grad_norm": 5.0235209465026855,
+      "learning_rate": 8.866165413533835e-06,
+      "loss": 0.3939,
+      "step": 7540
+    },
+    {
+      "epoch": 11.35,
+      "grad_norm": 6.832250595092773,
+      "learning_rate": 8.864661654135339e-06,
+      "loss": 0.4485,
+      "step": 7550
+    },
+    {
+      "epoch": 11.37,
+      "grad_norm": 8.186062812805176,
+      "learning_rate": 8.863157894736842e-06,
+      "loss": 0.4242,
+      "step": 7560
+    },
+    {
+      "epoch": 11.38,
+      "grad_norm": 5.467780113220215,
+      "learning_rate": 8.861654135338346e-06,
+      "loss": 0.4599,
+      "step": 7570
+    },
+    {
+      "epoch": 11.4,
+      "grad_norm": 6.155720233917236,
+      "learning_rate": 8.86015037593985e-06,
+      "loss": 0.5285,
+      "step": 7580
+    },
+    {
+      "epoch": 11.41,
+      "grad_norm": 6.44677734375,
+      "learning_rate": 8.858646616541355e-06,
+      "loss": 0.4931,
+      "step": 7590
+    },
+    {
+      "epoch": 11.43,
+      "grad_norm": 15.308818817138672,
+      "learning_rate": 8.857142857142858e-06,
+      "loss": 0.3899,
+      "step": 7600
+    },
+    {
+      "epoch": 11.44,
+      "grad_norm": 6.691050052642822,
+      "learning_rate": 8.855639097744362e-06,
+      "loss": 0.4313,
+      "step": 7610
+    },
+    {
+      "epoch": 11.46,
+      "grad_norm": 5.215397357940674,
+      "learning_rate": 8.854135338345865e-06,
+      "loss": 0.3528,
+      "step": 7620
+    },
+    {
+      "epoch": 11.47,
+      "grad_norm": 7.355811595916748,
+      "learning_rate": 8.852631578947369e-06,
+      "loss": 0.4402,
+      "step": 7630
+    },
+    {
+      "epoch": 11.49,
+      "grad_norm": 4.864825248718262,
+      "learning_rate": 8.851127819548872e-06,
+      "loss": 0.3485,
+      "step": 7640
+    },
+    {
+      "epoch": 11.5,
+      "grad_norm": 7.4907755851745605,
+      "learning_rate": 8.849624060150378e-06,
+      "loss": 0.4522,
+      "step": 7650
+    },
+    {
+      "epoch": 11.52,
+      "grad_norm": 6.480433464050293,
+      "learning_rate": 8.84812030075188e-06,
+      "loss": 0.4655,
+      "step": 7660
+    },
+    {
+      "epoch": 11.53,
+      "grad_norm": 5.072092056274414,
+      "learning_rate": 8.846616541353385e-06,
+      "loss": 0.3735,
+      "step": 7670
+    },
+    {
+      "epoch": 11.55,
+      "grad_norm": 10.207109451293945,
+      "learning_rate": 8.845112781954888e-06,
+      "loss": 0.3884,
+      "step": 7680
+    },
+    {
+      "epoch": 11.56,
+      "grad_norm": 5.795559883117676,
+      "learning_rate": 8.843609022556392e-06,
+      "loss": 0.4115,
+      "step": 7690
+    },
+    {
+      "epoch": 11.58,
+      "grad_norm": 7.781355381011963,
+      "learning_rate": 8.842105263157895e-06,
+      "loss": 0.4617,
+      "step": 7700
+    },
+    {
+      "epoch": 11.59,
+      "grad_norm": 6.770030975341797,
+      "learning_rate": 8.8406015037594e-06,
+      "loss": 0.4218,
+      "step": 7710
+    },
+    {
+      "epoch": 11.61,
+      "grad_norm": 7.052707672119141,
+      "learning_rate": 8.839097744360902e-06,
+      "loss": 0.3897,
+      "step": 7720
+    },
+    {
+      "epoch": 11.62,
+      "grad_norm": 8.882899284362793,
+      "learning_rate": 8.837593984962407e-06,
+      "loss": 0.4546,
+      "step": 7730
+    },
+    {
+      "epoch": 11.64,
+      "grad_norm": 7.858944892883301,
+      "learning_rate": 8.836090225563911e-06,
+      "loss": 0.4287,
+      "step": 7740
+    },
+    {
+      "epoch": 11.65,
+      "grad_norm": 6.74614953994751,
+      "learning_rate": 8.834586466165414e-06,
+      "loss": 0.3326,
+      "step": 7750
+    },
+    {
+      "epoch": 11.67,
+      "grad_norm": 8.970141410827637,
+      "learning_rate": 8.833082706766918e-06,
+      "loss": 0.4863,
+      "step": 7760
+    },
+    {
+      "epoch": 11.68,
+      "grad_norm": 6.568352699279785,
+      "learning_rate": 8.831578947368421e-06,
+      "loss": 0.4248,
+      "step": 7770
+    },
+    {
+      "epoch": 11.7,
+      "grad_norm": 6.05830717086792,
+      "learning_rate": 8.830075187969925e-06,
+      "loss": 0.4829,
+      "step": 7780
+    },
+    {
+      "epoch": 11.71,
+      "grad_norm": 7.666469097137451,
+      "learning_rate": 8.82857142857143e-06,
+      "loss": 0.5319,
+      "step": 7790
+    },
+    {
+      "epoch": 11.73,
+      "grad_norm": 5.955508708953857,
+      "learning_rate": 8.827067669172934e-06,
+      "loss": 0.4309,
+      "step": 7800
+    },
+    {
+      "epoch": 11.74,
+      "grad_norm": 6.9883270263671875,
+      "learning_rate": 8.825563909774437e-06,
+      "loss": 0.4157,
+      "step": 7810
+    },
+    {
+      "epoch": 11.76,
+      "grad_norm": 6.703571319580078,
+      "learning_rate": 8.82406015037594e-06,
+      "loss": 0.3219,
+      "step": 7820
+    },
+    {
+      "epoch": 11.77,
+      "grad_norm": 7.131542682647705,
+      "learning_rate": 8.822556390977444e-06,
+      "loss": 0.4492,
+      "step": 7830
+    },
+    {
+      "epoch": 11.79,
+      "grad_norm": 5.014946460723877,
+      "learning_rate": 8.821052631578948e-06,
+      "loss": 0.4189,
+      "step": 7840
+    },
+    {
+      "epoch": 11.8,
+      "grad_norm": 4.254874229431152,
+      "learning_rate": 8.819548872180453e-06,
+      "loss": 0.484,
+      "step": 7850
+    },
+    {
+      "epoch": 11.82,
+      "grad_norm": 4.319407939910889,
+      "learning_rate": 8.818045112781955e-06,
+      "loss": 0.3861,
+      "step": 7860
+    },
+    {
+      "epoch": 11.83,
+      "grad_norm": 7.9686408042907715,
+      "learning_rate": 8.81654135338346e-06,
+      "loss": 0.4264,
+      "step": 7870
+    },
+    {
+      "epoch": 11.85,
+      "grad_norm": 5.5855326652526855,
+      "learning_rate": 8.815037593984964e-06,
+      "loss": 0.4532,
+      "step": 7880
+    },
+    {
+      "epoch": 11.86,
+      "grad_norm": 6.914451599121094,
+      "learning_rate": 8.813533834586467e-06,
+      "loss": 0.4355,
+      "step": 7890
+    },
+    {
+      "epoch": 11.88,
+      "grad_norm": 7.542539596557617,
+      "learning_rate": 8.81203007518797e-06,
+      "loss": 0.3671,
+      "step": 7900
+    },
+    {
+      "epoch": 11.89,
+      "grad_norm": 7.947263717651367,
+      "learning_rate": 8.810526315789474e-06,
+      "loss": 0.373,
+      "step": 7910
+    },
+    {
+      "epoch": 11.91,
+      "grad_norm": 7.884321689605713,
+      "learning_rate": 8.809022556390978e-06,
+      "loss": 0.4827,
+      "step": 7920
+    },
+    {
+      "epoch": 11.92,
+      "grad_norm": 5.361155986785889,
+      "learning_rate": 8.807518796992483e-06,
+      "loss": 0.4485,
+      "step": 7930
+    },
+    {
+      "epoch": 11.94,
+      "grad_norm": 7.507490158081055,
+      "learning_rate": 8.806015037593986e-06,
+      "loss": 0.4446,
+      "step": 7940
+    },
+    {
+      "epoch": 11.95,
+      "grad_norm": 7.053649425506592,
+      "learning_rate": 8.804511278195488e-06,
+      "loss": 0.4112,
+      "step": 7950
+    },
+    {
+      "epoch": 11.97,
+      "grad_norm": 8.394134521484375,
+      "learning_rate": 8.803007518796993e-06,
+      "loss": 0.4221,
+      "step": 7960
+    },
+    {
+      "epoch": 11.98,
+      "grad_norm": 9.852388381958008,
+      "learning_rate": 8.801503759398497e-06,
+      "loss": 0.4178,
+      "step": 7970
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 17.5406551361084,
+      "learning_rate": 8.8e-06,
+      "loss": 0.4869,
+      "step": 7980
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.9298,
+      "eval_loss": 0.2563527822494507,
+      "eval_runtime": 84.9156,
+      "eval_samples_per_second": 117.764,
+      "eval_steps_per_second": 0.471,
+      "step": 7980
+    },
+    {
+      "epoch": 12.02,
+      "grad_norm": 4.651547908782959,
+      "learning_rate": 8.798496240601504e-06,
+      "loss": 0.4307,
+      "step": 7990
+    },
+    {
+      "epoch": 12.03,
+      "grad_norm": 5.47507905960083,
+      "learning_rate": 8.796992481203007e-06,
+      "loss": 0.3957,
+      "step": 8000
+    },
+    {
+      "epoch": 12.05,
+      "grad_norm": 7.309173583984375,
+      "learning_rate": 8.795488721804511e-06,
+      "loss": 0.3848,
+      "step": 8010
+    },
+    {
+      "epoch": 12.06,
+      "grad_norm": 5.4073591232299805,
+      "learning_rate": 8.793984962406016e-06,
+      "loss": 0.347,
+      "step": 8020
+    },
+    {
+      "epoch": 12.08,
+      "grad_norm": 9.495542526245117,
+      "learning_rate": 8.79248120300752e-06,
+      "loss": 0.3963,
+      "step": 8030
+    },
+    {
+      "epoch": 12.09,
+      "grad_norm": 7.175304412841797,
+      "learning_rate": 8.790977443609023e-06,
+      "loss": 0.4028,
+      "step": 8040
+    },
+    {
+      "epoch": 12.11,
+      "grad_norm": 5.7672624588012695,
+      "learning_rate": 8.789473684210527e-06,
+      "loss": 0.4336,
+      "step": 8050
+    },
+    {
+      "epoch": 12.12,
+      "grad_norm": 5.373271942138672,
+      "learning_rate": 8.78796992481203e-06,
+      "loss": 0.4214,
+      "step": 8060
+    },
+    {
+      "epoch": 12.14,
+      "grad_norm": 7.81503963470459,
+      "learning_rate": 8.786466165413534e-06,
+      "loss": 0.3362,
+      "step": 8070
+    },
+    {
+      "epoch": 12.15,
+      "grad_norm": 5.352240085601807,
+      "learning_rate": 8.784962406015039e-06,
+      "loss": 0.438,
+      "step": 8080
+    },
+    {
+      "epoch": 12.17,
+      "grad_norm": 4.825592994689941,
+      "learning_rate": 8.783458646616541e-06,
+      "loss": 0.3996,
+      "step": 8090
+    },
+    {
+      "epoch": 12.18,
+      "grad_norm": 4.875209808349609,
+      "learning_rate": 8.781954887218046e-06,
+      "loss": 0.4056,
+      "step": 8100
+    },
+    {
+      "epoch": 12.2,
+      "grad_norm": 6.405061721801758,
+      "learning_rate": 8.78045112781955e-06,
+      "loss": 0.404,
+      "step": 8110
+    },
+    {
+      "epoch": 12.21,
+      "grad_norm": 5.762337684631348,
+      "learning_rate": 8.778947368421053e-06,
+      "loss": 0.3609,
+      "step": 8120
+    },
+    {
+      "epoch": 12.23,
+      "grad_norm": 8.700191497802734,
+      "learning_rate": 8.777443609022557e-06,
+      "loss": 0.4316,
+      "step": 8130
+    },
+    {
+      "epoch": 12.24,
+      "grad_norm": 5.509273052215576,
+      "learning_rate": 8.775939849624062e-06,
+      "loss": 0.3814,
+      "step": 8140
+    },
+    {
+      "epoch": 12.26,
+      "grad_norm": 6.949098587036133,
+      "learning_rate": 8.774436090225564e-06,
+      "loss": 0.3588,
+      "step": 8150
+    },
+    {
+      "epoch": 12.27,
+      "grad_norm": 5.564908981323242,
+      "learning_rate": 8.772932330827069e-06,
+      "loss": 0.353,
+      "step": 8160
+    },
+    {
+      "epoch": 12.29,
+      "grad_norm": 6.935297012329102,
+      "learning_rate": 8.771428571428572e-06,
+      "loss": 0.4212,
+      "step": 8170
+    },
+    {
+      "epoch": 12.3,
+      "grad_norm": 4.811358451843262,
+      "learning_rate": 8.769924812030076e-06,
+      "loss": 0.3865,
+      "step": 8180
+    },
+    {
+      "epoch": 12.32,
+      "grad_norm": 6.4804368019104,
+      "learning_rate": 8.76842105263158e-06,
+      "loss": 0.3656,
+      "step": 8190
+    },
+    {
+      "epoch": 12.33,
+      "grad_norm": 3.9236013889312744,
+      "learning_rate": 8.766917293233083e-06,
+      "loss": 0.4885,
+      "step": 8200
+    },
+    {
+      "epoch": 12.35,
+      "grad_norm": 7.50891637802124,
+      "learning_rate": 8.765413533834586e-06,
+      "loss": 0.3962,
+      "step": 8210
+    },
+    {
+      "epoch": 12.36,
+      "grad_norm": 4.313982963562012,
+      "learning_rate": 8.763909774436092e-06,
+      "loss": 0.4023,
+      "step": 8220
+    },
+    {
+      "epoch": 12.38,
+      "grad_norm": 4.385167121887207,
+      "learning_rate": 8.762406015037595e-06,
+      "loss": 0.4841,
+      "step": 8230
+    },
+    {
+      "epoch": 12.39,
+      "grad_norm": 5.977277755737305,
+      "learning_rate": 8.760902255639099e-06,
+      "loss": 0.399,
+      "step": 8240
+    },
+    {
+      "epoch": 12.41,
+      "grad_norm": 8.858118057250977,
+      "learning_rate": 8.759398496240602e-06,
+      "loss": 0.451,
+      "step": 8250
+    },
+    {
+      "epoch": 12.42,
+      "grad_norm": 6.294662952423096,
+      "learning_rate": 8.757894736842106e-06,
+      "loss": 0.416,
+      "step": 8260
+    },
+    {
+      "epoch": 12.44,
+      "grad_norm": 4.536668300628662,
+      "learning_rate": 8.75639097744361e-06,
+      "loss": 0.407,
+      "step": 8270
+    },
+    {
+      "epoch": 12.45,
+      "grad_norm": 5.644812107086182,
+      "learning_rate": 8.754887218045114e-06,
+      "loss": 0.3685,
+      "step": 8280
+    },
+    {
+      "epoch": 12.47,
+      "grad_norm": 5.488842010498047,
+      "learning_rate": 8.753383458646616e-06,
+      "loss": 0.4136,
+      "step": 8290
+    },
+    {
+      "epoch": 12.48,
+      "grad_norm": 4.548142910003662,
+      "learning_rate": 8.751879699248122e-06,
+      "loss": 0.4502,
+      "step": 8300
+    },
+    {
+      "epoch": 12.5,
+      "grad_norm": 3.41457200050354,
+      "learning_rate": 8.750375939849625e-06,
+      "loss": 0.3598,
+      "step": 8310
+    },
+    {
+      "epoch": 12.51,
+      "grad_norm": 6.259812831878662,
+      "learning_rate": 8.748872180451129e-06,
+      "loss": 0.3843,
+      "step": 8320
+    },
+    {
+      "epoch": 12.53,
+      "grad_norm": 5.301551342010498,
+      "learning_rate": 8.747368421052632e-06,
+      "loss": 0.4038,
+      "step": 8330
+    },
+    {
+      "epoch": 12.54,
+      "grad_norm": 14.684255599975586,
+      "learning_rate": 8.745864661654137e-06,
+      "loss": 0.4115,
+      "step": 8340
+    },
+    {
+      "epoch": 12.56,
+      "grad_norm": 6.711531162261963,
+      "learning_rate": 8.744360902255639e-06,
+      "loss": 0.4112,
+      "step": 8350
+    },
+    {
+      "epoch": 12.57,
+      "grad_norm": 8.990388870239258,
+      "learning_rate": 8.742857142857144e-06,
+      "loss": 0.3547,
+      "step": 8360
+    },
+    {
+      "epoch": 12.59,
+      "grad_norm": 4.513948440551758,
+      "learning_rate": 8.741353383458648e-06,
+      "loss": 0.3776,
+      "step": 8370
+    },
+    {
+      "epoch": 12.6,
+      "grad_norm": 6.088433742523193,
+      "learning_rate": 8.739849624060151e-06,
+      "loss": 0.4116,
+      "step": 8380
+    },
+    {
+      "epoch": 12.62,
+      "grad_norm": 7.882970809936523,
+      "learning_rate": 8.738345864661655e-06,
+      "loss": 0.3712,
+      "step": 8390
+    },
+    {
+      "epoch": 12.63,
+      "grad_norm": 6.829627990722656,
+      "learning_rate": 8.736842105263158e-06,
+      "loss": 0.3842,
+      "step": 8400
+    },
+    {
+      "epoch": 12.65,
+      "grad_norm": 6.185722351074219,
+      "learning_rate": 8.735338345864662e-06,
+      "loss": 0.424,
+      "step": 8410
+    },
+    {
+      "epoch": 12.66,
+      "grad_norm": 4.945958137512207,
+      "learning_rate": 8.733834586466167e-06,
+      "loss": 0.377,
+      "step": 8420
+    },
+    {
+      "epoch": 12.68,
+      "grad_norm": 6.356648921966553,
+      "learning_rate": 8.73233082706767e-06,
+      "loss": 0.4256,
+      "step": 8430
+    },
+    {
+      "epoch": 12.69,
+      "grad_norm": 6.276622295379639,
+      "learning_rate": 8.730827067669172e-06,
+      "loss": 0.3733,
+      "step": 8440
+    },
+    {
+      "epoch": 12.71,
+      "grad_norm": 7.50572395324707,
+      "learning_rate": 8.729323308270678e-06,
+      "loss": 0.4407,
+      "step": 8450
+    },
+    {
+      "epoch": 12.72,
+      "grad_norm": 7.089003086090088,
+      "learning_rate": 8.727819548872181e-06,
+      "loss": 0.3948,
+      "step": 8460
+    },
+    {
+      "epoch": 12.74,
+      "grad_norm": 6.90725564956665,
+      "learning_rate": 8.726315789473685e-06,
+      "loss": 0.4511,
+      "step": 8470
+    },
+    {
+      "epoch": 12.75,
+      "grad_norm": 4.369374752044678,
+      "learning_rate": 8.724812030075188e-06,
+      "loss": 0.3559,
+      "step": 8480
+    },
+    {
+      "epoch": 12.77,
+      "grad_norm": 2.895493507385254,
+      "learning_rate": 8.723308270676692e-06,
+      "loss": 0.349,
+      "step": 8490
+    },
+    {
+      "epoch": 12.78,
+      "grad_norm": 8.638984680175781,
+      "learning_rate": 8.721804511278195e-06,
+      "loss": 0.3406,
+      "step": 8500
+    },
+    {
+      "epoch": 12.8,
+      "grad_norm": 7.664207458496094,
+      "learning_rate": 8.7203007518797e-06,
+      "loss": 0.3619,
+      "step": 8510
+    },
+    {
+      "epoch": 12.81,
+      "grad_norm": 4.544347286224365,
+      "learning_rate": 8.718796992481204e-06,
+      "loss": 0.3109,
+      "step": 8520
+    },
+    {
+      "epoch": 12.83,
+      "grad_norm": 6.640614032745361,
+      "learning_rate": 8.717293233082708e-06,
+      "loss": 0.4116,
+      "step": 8530
+    },
+    {
+      "epoch": 12.84,
+      "grad_norm": 7.840051174163818,
+      "learning_rate": 8.715789473684211e-06,
+      "loss": 0.4027,
+      "step": 8540
+    },
+    {
+      "epoch": 12.86,
+      "grad_norm": 10.355204582214355,
+      "learning_rate": 8.714285714285715e-06,
+      "loss": 0.4013,
+      "step": 8550
+    },
+    {
+      "epoch": 12.87,
+      "grad_norm": 7.472030162811279,
+      "learning_rate": 8.712781954887218e-06,
+      "loss": 0.4119,
+      "step": 8560
+    },
+    {
+      "epoch": 12.89,
+      "grad_norm": 7.9360246658325195,
+      "learning_rate": 8.711278195488723e-06,
+      "loss": 0.3472,
+      "step": 8570
+    },
+    {
+      "epoch": 12.9,
+      "grad_norm": 5.889431953430176,
+      "learning_rate": 8.709774436090225e-06,
+      "loss": 0.4009,
+      "step": 8580
+    },
+    {
+      "epoch": 12.92,
+      "grad_norm": 5.548401355743408,
+      "learning_rate": 8.70827067669173e-06,
+      "loss": 0.4261,
+      "step": 8590
+    },
+    {
+      "epoch": 12.93,
+      "grad_norm": 5.590747833251953,
+      "learning_rate": 8.706766917293234e-06,
+      "loss": 0.4437,
+      "step": 8600
+    },
+    {
+      "epoch": 12.95,
+      "grad_norm": 6.401696681976318,
+      "learning_rate": 8.705263157894737e-06,
+      "loss": 0.3746,
+      "step": 8610
+    },
+    {
+      "epoch": 12.96,
+      "grad_norm": 9.315383911132812,
+      "learning_rate": 8.703759398496241e-06,
+      "loss": 0.3803,
+      "step": 8620
+    },
+    {
+      "epoch": 12.98,
+      "grad_norm": 3.9589388370513916,
+      "learning_rate": 8.702255639097746e-06,
+      "loss": 0.4448,
+      "step": 8630
+    },
+    {
+      "epoch": 12.99,
+      "grad_norm": 4.445014953613281,
+      "learning_rate": 8.700751879699248e-06,
+      "loss": 0.4289,
+      "step": 8640
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.9288,
+      "eval_loss": 0.2712935507297516,
+      "eval_runtime": 84.8607,
+      "eval_samples_per_second": 117.84,
+      "eval_steps_per_second": 0.471,
+      "step": 8645
+    },
+    {
+      "epoch": 13.01,
+      "grad_norm": 5.444362163543701,
+      "learning_rate": 8.699248120300753e-06,
+      "loss": 0.3994,
+      "step": 8650
+    },
+    {
+      "epoch": 13.02,
+      "grad_norm": 5.8952178955078125,
+      "learning_rate": 8.697744360902257e-06,
+      "loss": 0.3819,
+      "step": 8660
+    },
+    {
+      "epoch": 13.04,
+      "grad_norm": 5.363025188446045,
+      "learning_rate": 8.69624060150376e-06,
+      "loss": 0.4251,
+      "step": 8670
+    },
+    {
+      "epoch": 13.05,
+      "grad_norm": 6.1266961097717285,
+      "learning_rate": 8.694736842105264e-06,
+      "loss": 0.4236,
+      "step": 8680
+    },
+    {
+      "epoch": 13.07,
+      "grad_norm": 6.096094131469727,
+      "learning_rate": 8.693233082706767e-06,
+      "loss": 0.4411,
+      "step": 8690
+    },
+    {
+      "epoch": 13.08,
+      "grad_norm": 6.0483293533325195,
+      "learning_rate": 8.69172932330827e-06,
+      "loss": 0.3538,
+      "step": 8700
+    },
+    {
+      "epoch": 13.1,
+      "grad_norm": 8.619955062866211,
+      "learning_rate": 8.690225563909776e-06,
+      "loss": 0.4698,
+      "step": 8710
+    },
+    {
+      "epoch": 13.11,
+      "grad_norm": 5.028072834014893,
+      "learning_rate": 8.68872180451128e-06,
+      "loss": 0.3883,
+      "step": 8720
+    },
+    {
+      "epoch": 13.13,
+      "grad_norm": 7.43666934967041,
+      "learning_rate": 8.687218045112783e-06,
+      "loss": 0.3552,
+      "step": 8730
+    },
+    {
+      "epoch": 13.14,
+      "grad_norm": 9.520151138305664,
+      "learning_rate": 8.685714285714287e-06,
+      "loss": 0.4079,
+      "step": 8740
+    },
+    {
+      "epoch": 13.16,
+      "grad_norm": 7.852067947387695,
+      "learning_rate": 8.68421052631579e-06,
+      "loss": 0.3607,
+      "step": 8750
+    },
+    {
+      "epoch": 13.17,
+      "grad_norm": 5.92877721786499,
+      "learning_rate": 8.682706766917294e-06,
+      "loss": 0.3739,
+      "step": 8760
+    },
+    {
+      "epoch": 13.19,
+      "grad_norm": 4.25166130065918,
+      "learning_rate": 8.681203007518799e-06,
+      "loss": 0.4621,
+      "step": 8770
+    },
+    {
+      "epoch": 13.2,
+      "grad_norm": 7.073912143707275,
+      "learning_rate": 8.6796992481203e-06,
+      "loss": 0.4465,
+      "step": 8780
+    },
+    {
+      "epoch": 13.22,
+      "grad_norm": 7.39524507522583,
+      "learning_rate": 8.678195488721806e-06,
+      "loss": 0.4303,
+      "step": 8790
+    },
+    {
+      "epoch": 13.23,
+      "grad_norm": 6.938388824462891,
+      "learning_rate": 8.67669172932331e-06,
+      "loss": 0.3535,
+      "step": 8800
+    },
+    {
+      "epoch": 13.25,
+      "grad_norm": 5.0067524909973145,
+      "learning_rate": 8.675187969924813e-06,
+      "loss": 0.4399,
+      "step": 8810
+    },
+    {
+      "epoch": 13.26,
+      "grad_norm": 6.340808391571045,
+      "learning_rate": 8.673684210526316e-06,
+      "loss": 0.4199,
+      "step": 8820
+    },
+    {
+      "epoch": 13.28,
+      "grad_norm": 4.246801853179932,
+      "learning_rate": 8.67218045112782e-06,
+      "loss": 0.4039,
+      "step": 8830
+    },
+    {
+      "epoch": 13.29,
+      "grad_norm": 4.85552453994751,
+      "learning_rate": 8.670676691729323e-06,
+      "loss": 0.3753,
+      "step": 8840
+    },
+    {
+      "epoch": 13.31,
+      "grad_norm": 6.020550727844238,
+      "learning_rate": 8.669172932330829e-06,
+      "loss": 0.4039,
+      "step": 8850
+    },
+    {
+      "epoch": 13.32,
+      "grad_norm": 3.4875411987304688,
+      "learning_rate": 8.667669172932332e-06,
+      "loss": 0.3829,
+      "step": 8860
+    },
+    {
+      "epoch": 13.34,
+      "grad_norm": 6.239095211029053,
+      "learning_rate": 8.666165413533836e-06,
+      "loss": 0.3511,
+      "step": 8870
+    },
+    {
+      "epoch": 13.35,
+      "grad_norm": 4.244966983795166,
+      "learning_rate": 8.66466165413534e-06,
+      "loss": 0.4268,
+      "step": 8880
+    },
+    {
+      "epoch": 13.37,
+      "grad_norm": 12.684317588806152,
+      "learning_rate": 8.663157894736843e-06,
+      "loss": 0.3471,
+      "step": 8890
+    },
+    {
+      "epoch": 13.38,
+      "grad_norm": 8.664961814880371,
+      "learning_rate": 8.661654135338346e-06,
+      "loss": 0.3822,
+      "step": 8900
+    },
+    {
+      "epoch": 13.4,
+      "grad_norm": 5.7766804695129395,
+      "learning_rate": 8.660150375939851e-06,
+      "loss": 0.4199,
+      "step": 8910
+    },
+    {
+      "epoch": 13.41,
+      "grad_norm": 4.019351959228516,
+      "learning_rate": 8.658646616541353e-06,
+      "loss": 0.4137,
+      "step": 8920
+    },
+    {
+      "epoch": 13.43,
+      "grad_norm": 6.156152248382568,
+      "learning_rate": 8.657142857142858e-06,
+      "loss": 0.4606,
+      "step": 8930
+    },
+    {
+      "epoch": 13.44,
+      "grad_norm": 5.74890661239624,
+      "learning_rate": 8.655639097744362e-06,
+      "loss": 0.3221,
+      "step": 8940
+    },
+    {
+      "epoch": 13.46,
+      "grad_norm": 6.321985721588135,
+      "learning_rate": 8.654135338345866e-06,
+      "loss": 0.415,
+      "step": 8950
+    },
+    {
+      "epoch": 13.47,
+      "grad_norm": 8.88508129119873,
+      "learning_rate": 8.652631578947369e-06,
+      "loss": 0.3714,
+      "step": 8960
+    },
+    {
+      "epoch": 13.49,
+      "grad_norm": 6.999327659606934,
+      "learning_rate": 8.651127819548873e-06,
+      "loss": 0.3576,
+      "step": 8970
+    },
+    {
+      "epoch": 13.5,
+      "grad_norm": 7.313613414764404,
+      "learning_rate": 8.649624060150376e-06,
+      "loss": 0.4096,
+      "step": 8980
+    },
+    {
+      "epoch": 13.52,
+      "grad_norm": 6.363276958465576,
+      "learning_rate": 8.64812030075188e-06,
+      "loss": 0.3356,
+      "step": 8990
+    },
+    {
+      "epoch": 13.53,
+      "grad_norm": 7.81085729598999,
+      "learning_rate": 8.646616541353385e-06,
+      "loss": 0.4216,
+      "step": 9000
+    },
+    {
+      "epoch": 13.55,
+      "grad_norm": 8.093158721923828,
+      "learning_rate": 8.645112781954887e-06,
+      "loss": 0.5105,
+      "step": 9010
+    },
+    {
+      "epoch": 13.56,
+      "grad_norm": 3.801630735397339,
+      "learning_rate": 8.643609022556392e-06,
+      "loss": 0.439,
+      "step": 9020
+    },
+    {
+      "epoch": 13.58,
+      "grad_norm": 5.564939975738525,
+      "learning_rate": 8.642105263157895e-06,
+      "loss": 0.3854,
+      "step": 9030
+    },
+    {
+      "epoch": 13.59,
+      "grad_norm": 9.847439765930176,
+      "learning_rate": 8.640601503759399e-06,
+      "loss": 0.4034,
+      "step": 9040
+    },
+    {
+      "epoch": 13.61,
+      "grad_norm": 9.21834659576416,
+      "learning_rate": 8.639097744360902e-06,
+      "loss": 0.4448,
+      "step": 9050
+    },
+    {
+      "epoch": 13.62,
+      "grad_norm": 4.98524808883667,
+      "learning_rate": 8.637593984962408e-06,
+      "loss": 0.3646,
+      "step": 9060
+    },
+    {
+      "epoch": 13.64,
+      "grad_norm": 6.707414150238037,
+      "learning_rate": 8.63609022556391e-06,
+      "loss": 0.3618,
+      "step": 9070
+    },
+    {
+      "epoch": 13.65,
+      "grad_norm": 5.5840840339660645,
+      "learning_rate": 8.634586466165415e-06,
+      "loss": 0.3628,
+      "step": 9080
+    },
+    {
+      "epoch": 13.67,
+      "grad_norm": 4.939608097076416,
+      "learning_rate": 8.633082706766918e-06,
+      "loss": 0.3785,
+      "step": 9090
+    },
+    {
+      "epoch": 13.68,
+      "grad_norm": 7.449197769165039,
+      "learning_rate": 8.631578947368422e-06,
+      "loss": 0.4354,
+      "step": 9100
+    },
+    {
+      "epoch": 13.7,
+      "grad_norm": 9.470358848571777,
+      "learning_rate": 8.630075187969925e-06,
+      "loss": 0.4075,
+      "step": 9110
+    },
+    {
+      "epoch": 13.71,
+      "grad_norm": 7.6183085441589355,
+      "learning_rate": 8.628571428571429e-06,
+      "loss": 0.4067,
+      "step": 9120
+    },
+    {
+      "epoch": 13.73,
+      "grad_norm": 3.0916943550109863,
+      "learning_rate": 8.627067669172932e-06,
+      "loss": 0.365,
+      "step": 9130
+    },
+    {
+      "epoch": 13.74,
+      "grad_norm": 4.251070499420166,
+      "learning_rate": 8.625563909774437e-06,
+      "loss": 0.4255,
+      "step": 9140
+    },
+    {
+      "epoch": 13.76,
+      "grad_norm": 6.8059282302856445,
+      "learning_rate": 8.624060150375941e-06,
+      "loss": 0.32,
+      "step": 9150
+    },
+    {
+      "epoch": 13.77,
+      "grad_norm": 7.302189826965332,
+      "learning_rate": 8.622556390977444e-06,
+      "loss": 0.4131,
+      "step": 9160
+    },
+    {
+      "epoch": 13.79,
+      "grad_norm": 6.402463436126709,
+      "learning_rate": 8.621052631578948e-06,
+      "loss": 0.3848,
+      "step": 9170
+    },
+    {
+      "epoch": 13.8,
+      "grad_norm": 4.343325138092041,
+      "learning_rate": 8.619548872180452e-06,
+      "loss": 0.4014,
+      "step": 9180
+    },
+    {
+      "epoch": 13.82,
+      "grad_norm": 9.013459205627441,
+      "learning_rate": 8.618045112781955e-06,
+      "loss": 0.3739,
+      "step": 9190
+    },
+    {
+      "epoch": 13.83,
+      "grad_norm": 7.037381172180176,
+      "learning_rate": 8.61654135338346e-06,
+      "loss": 0.4189,
+      "step": 9200
+    },
+    {
+      "epoch": 13.85,
+      "grad_norm": 4.7024760246276855,
+      "learning_rate": 8.615037593984962e-06,
+      "loss": 0.38,
+      "step": 9210
+    },
+    {
+      "epoch": 13.86,
+      "grad_norm": 4.808414936065674,
+      "learning_rate": 8.613533834586467e-06,
+      "loss": 0.4414,
+      "step": 9220
+    },
+    {
+      "epoch": 13.88,
+      "grad_norm": 8.237750053405762,
+      "learning_rate": 8.61203007518797e-06,
+      "loss": 0.4215,
+      "step": 9230
+    },
+    {
+      "epoch": 13.89,
+      "grad_norm": 7.862570285797119,
+      "learning_rate": 8.610526315789474e-06,
+      "loss": 0.4727,
+      "step": 9240
+    },
+    {
+      "epoch": 13.91,
+      "grad_norm": 7.045783519744873,
+      "learning_rate": 8.609022556390978e-06,
+      "loss": 0.4109,
+      "step": 9250
+    },
+    {
+      "epoch": 13.92,
+      "grad_norm": 5.3544135093688965,
+      "learning_rate": 8.607518796992483e-06,
+      "loss": 0.3824,
+      "step": 9260
+    },
+    {
+      "epoch": 13.94,
+      "grad_norm": 14.21022891998291,
+      "learning_rate": 8.606015037593985e-06,
+      "loss": 0.3659,
+      "step": 9270
+    },
+    {
+      "epoch": 13.95,
+      "grad_norm": 7.408153533935547,
+      "learning_rate": 8.60451127819549e-06,
+      "loss": 0.3508,
+      "step": 9280
+    },
+    {
+      "epoch": 13.97,
+      "grad_norm": 3.206442356109619,
+      "learning_rate": 8.603007518796994e-06,
+      "loss": 0.4451,
+      "step": 9290
+    },
+    {
+      "epoch": 13.98,
+      "grad_norm": 4.974185466766357,
+      "learning_rate": 8.601503759398497e-06,
+      "loss": 0.3888,
+      "step": 9300
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 15.39065170288086,
+      "learning_rate": 8.6e-06,
+      "loss": 0.4003,
+      "step": 9310
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.932,
+      "eval_loss": 0.2616865336894989,
+      "eval_runtime": 84.8808,
+      "eval_samples_per_second": 117.812,
+      "eval_steps_per_second": 0.471,
+      "step": 9310
+    },
+    {
+      "epoch": 14.02,
+      "grad_norm": 5.1268534660339355,
+      "learning_rate": 8.598496240601504e-06,
+      "loss": 0.4195,
+      "step": 9320
+    },
+    {
+      "epoch": 14.03,
+      "grad_norm": 6.874084949493408,
+      "learning_rate": 8.596992481203008e-06,
+      "loss": 0.3901,
+      "step": 9330
+    },
+    {
+      "epoch": 14.05,
+      "grad_norm": 11.405204772949219,
+      "learning_rate": 8.595488721804513e-06,
+      "loss": 0.3234,
+      "step": 9340
+    },
+    {
+      "epoch": 14.06,
+      "grad_norm": 4.844882965087891,
+      "learning_rate": 8.593984962406016e-06,
+      "loss": 0.3437,
+      "step": 9350
+    },
+    {
+      "epoch": 14.08,
+      "grad_norm": 7.187948226928711,
+      "learning_rate": 8.59248120300752e-06,
+      "loss": 0.3895,
+      "step": 9360
+    },
+    {
+      "epoch": 14.09,
+      "grad_norm": 3.7594106197357178,
+      "learning_rate": 8.590977443609023e-06,
+      "loss": 0.3329,
+      "step": 9370
+    },
+    {
+      "epoch": 14.11,
+      "grad_norm": 4.245199203491211,
+      "learning_rate": 8.589473684210527e-06,
+      "loss": 0.3644,
+      "step": 9380
+    },
+    {
+      "epoch": 14.12,
+      "grad_norm": 6.302145004272461,
+      "learning_rate": 8.58796992481203e-06,
+      "loss": 0.4615,
+      "step": 9390
+    },
+    {
+      "epoch": 14.14,
+      "grad_norm": 6.26497220993042,
+      "learning_rate": 8.586466165413536e-06,
+      "loss": 0.3983,
+      "step": 9400
+    },
+    {
+      "epoch": 14.15,
+      "grad_norm": 7.315799236297607,
+      "learning_rate": 8.584962406015038e-06,
+      "loss": 0.3474,
+      "step": 9410
+    },
+    {
+      "epoch": 14.17,
+      "grad_norm": 8.584407806396484,
+      "learning_rate": 8.583458646616543e-06,
+      "loss": 0.3858,
+      "step": 9420
+    },
+    {
+      "epoch": 14.18,
+      "grad_norm": 6.192986488342285,
+      "learning_rate": 8.581954887218046e-06,
+      "loss": 0.4653,
+      "step": 9430
+    },
+    {
+      "epoch": 14.2,
+      "grad_norm": 6.261072635650635,
+      "learning_rate": 8.58045112781955e-06,
+      "loss": 0.3686,
+      "step": 9440
+    },
+    {
+      "epoch": 14.21,
+      "grad_norm": 6.7162017822265625,
+      "learning_rate": 8.578947368421053e-06,
+      "loss": 0.4143,
+      "step": 9450
+    },
+    {
+      "epoch": 14.23,
+      "grad_norm": 5.550053119659424,
+      "learning_rate": 8.577443609022557e-06,
+      "loss": 0.4262,
+      "step": 9460
+    },
+    {
+      "epoch": 14.24,
+      "grad_norm": 6.601341247558594,
+      "learning_rate": 8.57593984962406e-06,
+      "loss": 0.3615,
+      "step": 9470
+    },
+    {
+      "epoch": 14.26,
+      "grad_norm": 6.859097957611084,
+      "learning_rate": 8.574436090225564e-06,
+      "loss": 0.3584,
+      "step": 9480
+    },
+    {
+      "epoch": 14.27,
+      "grad_norm": 3.824615478515625,
+      "learning_rate": 8.572932330827069e-06,
+      "loss": 0.3941,
+      "step": 9490
+    },
+    {
+      "epoch": 14.29,
+      "grad_norm": 6.923836708068848,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 0.4016,
+      "step": 9500
+    },
+    {
+      "epoch": 14.3,
+      "grad_norm": 6.395806789398193,
+      "learning_rate": 8.569924812030076e-06,
+      "loss": 0.368,
+      "step": 9510
+    },
+    {
+      "epoch": 14.32,
+      "grad_norm": 6.522418022155762,
+      "learning_rate": 8.56842105263158e-06,
+      "loss": 0.377,
+      "step": 9520
+    },
+    {
+      "epoch": 14.33,
+      "grad_norm": 7.502889633178711,
+      "learning_rate": 8.566917293233083e-06,
+      "loss": 0.4179,
+      "step": 9530
+    },
+    {
+      "epoch": 14.35,
+      "grad_norm": 6.025669574737549,
+      "learning_rate": 8.565413533834587e-06,
+      "loss": 0.3395,
+      "step": 9540
+    },
+    {
+      "epoch": 14.36,
+      "grad_norm": 7.751435279846191,
+      "learning_rate": 8.563909774436092e-06,
+      "loss": 0.3531,
+      "step": 9550
+    },
+    {
+      "epoch": 14.38,
+      "grad_norm": 6.964672088623047,
+      "learning_rate": 8.562406015037594e-06,
+      "loss": 0.3905,
+      "step": 9560
+    },
+    {
+      "epoch": 14.39,
+      "grad_norm": 2.502666473388672,
+      "learning_rate": 8.560902255639099e-06,
+      "loss": 0.3172,
+      "step": 9570
+    },
+    {
+      "epoch": 14.41,
+      "grad_norm": 7.133659839630127,
+      "learning_rate": 8.559398496240602e-06,
+      "loss": 0.442,
+      "step": 9580
+    },
+    {
+      "epoch": 14.42,
+      "grad_norm": 4.32753324508667,
+      "learning_rate": 8.557894736842106e-06,
+      "loss": 0.3458,
+      "step": 9590
+    },
+    {
+      "epoch": 14.44,
+      "grad_norm": 6.268803119659424,
+      "learning_rate": 8.55639097744361e-06,
+      "loss": 0.3431,
+      "step": 9600
+    },
+    {
+      "epoch": 14.45,
+      "grad_norm": 7.018800735473633,
+      "learning_rate": 8.554887218045113e-06,
+      "loss": 0.4369,
+      "step": 9610
+    },
+    {
+      "epoch": 14.47,
+      "grad_norm": 7.435917377471924,
+      "learning_rate": 8.553383458646617e-06,
+      "loss": 0.3861,
+      "step": 9620
+    },
+    {
+      "epoch": 14.48,
+      "grad_norm": 7.8388752937316895,
+      "learning_rate": 8.551879699248122e-06,
+      "loss": 0.4239,
+      "step": 9630
+    },
+    {
+      "epoch": 14.5,
+      "grad_norm": 7.880455493927002,
+      "learning_rate": 8.550375939849625e-06,
+      "loss": 0.3829,
+      "step": 9640
+    },
+    {
+      "epoch": 14.51,
+      "grad_norm": 5.568830490112305,
+      "learning_rate": 8.548872180451129e-06,
+      "loss": 0.4011,
+      "step": 9650
+    },
+    {
+      "epoch": 14.53,
+      "grad_norm": 5.6487274169921875,
+      "learning_rate": 8.547368421052632e-06,
+      "loss": 0.4091,
+      "step": 9660
+    },
+    {
+      "epoch": 14.54,
+      "grad_norm": 4.51718807220459,
+      "learning_rate": 8.545864661654136e-06,
+      "loss": 0.4087,
+      "step": 9670
+    },
+    {
+      "epoch": 14.56,
+      "grad_norm": 7.874798774719238,
+      "learning_rate": 8.54436090225564e-06,
+      "loss": 0.3494,
+      "step": 9680
+    },
+    {
+      "epoch": 14.57,
+      "grad_norm": 4.497681617736816,
+      "learning_rate": 8.542857142857145e-06,
+      "loss": 0.3783,
+      "step": 9690
+    },
+    {
+      "epoch": 14.59,
+      "grad_norm": 5.411101818084717,
+      "learning_rate": 8.541353383458646e-06,
+      "loss": 0.3395,
+      "step": 9700
+    },
+    {
+      "epoch": 14.6,
+      "grad_norm": 5.714541435241699,
+      "learning_rate": 8.539849624060152e-06,
+      "loss": 0.4507,
+      "step": 9710
+    },
+    {
+      "epoch": 14.62,
+      "grad_norm": 7.042336940765381,
+      "learning_rate": 8.538345864661655e-06,
+      "loss": 0.3363,
+      "step": 9720
+    },
+    {
+      "epoch": 14.63,
+      "grad_norm": 6.70949649810791,
+      "learning_rate": 8.536842105263159e-06,
+      "loss": 0.3559,
+      "step": 9730
+    },
+    {
+      "epoch": 14.65,
+      "grad_norm": 5.84644889831543,
+      "learning_rate": 8.535338345864662e-06,
+      "loss": 0.3936,
+      "step": 9740
+    },
+    {
+      "epoch": 14.66,
+      "grad_norm": 5.070087432861328,
+      "learning_rate": 8.533834586466166e-06,
+      "loss": 0.4014,
+      "step": 9750
+    },
+    {
+      "epoch": 14.68,
+      "grad_norm": 5.353463172912598,
+      "learning_rate": 8.53233082706767e-06,
+      "loss": 0.4091,
+      "step": 9760
+    },
+    {
+      "epoch": 14.69,
+      "grad_norm": 9.118497848510742,
+      "learning_rate": 8.530827067669174e-06,
+      "loss": 0.3922,
+      "step": 9770
+    },
+    {
+      "epoch": 14.71,
+      "grad_norm": 7.667191982269287,
+      "learning_rate": 8.529323308270678e-06,
+      "loss": 0.4087,
+      "step": 9780
+    },
+    {
+      "epoch": 14.72,
+      "grad_norm": 7.210267066955566,
+      "learning_rate": 8.527819548872181e-06,
+      "loss": 0.3458,
+      "step": 9790
+    },
+    {
+      "epoch": 14.74,
+      "grad_norm": 5.242373466491699,
+      "learning_rate": 8.526315789473685e-06,
+      "loss": 0.3293,
+      "step": 9800
+    },
+    {
+      "epoch": 14.75,
+      "grad_norm": 7.6933393478393555,
+      "learning_rate": 8.524812030075188e-06,
+      "loss": 0.4406,
+      "step": 9810
+    },
+    {
+      "epoch": 14.77,
+      "grad_norm": 5.179628372192383,
+      "learning_rate": 8.523308270676692e-06,
+      "loss": 0.3827,
+      "step": 9820
+    },
+    {
+      "epoch": 14.78,
+      "grad_norm": 9.525907516479492,
+      "learning_rate": 8.521804511278197e-06,
+      "loss": 0.4095,
+      "step": 9830
+    },
+    {
+      "epoch": 14.8,
+      "grad_norm": 6.132147789001465,
+      "learning_rate": 8.520300751879699e-06,
+      "loss": 0.3778,
+      "step": 9840
+    },
+    {
+      "epoch": 14.81,
+      "grad_norm": 8.04976749420166,
+      "learning_rate": 8.518796992481204e-06,
+      "loss": 0.3996,
+      "step": 9850
+    },
+    {
+      "epoch": 14.83,
+      "grad_norm": 9.131913185119629,
+      "learning_rate": 8.517293233082708e-06,
+      "loss": 0.4103,
+      "step": 9860
+    },
+    {
+      "epoch": 14.84,
+      "grad_norm": 5.724211692810059,
+      "learning_rate": 8.515789473684211e-06,
+      "loss": 0.4166,
+      "step": 9870
+    },
+    {
+      "epoch": 14.86,
+      "grad_norm": 4.537842750549316,
+      "learning_rate": 8.514285714285715e-06,
+      "loss": 0.4357,
+      "step": 9880
+    },
+    {
+      "epoch": 14.87,
+      "grad_norm": 9.75554370880127,
+      "learning_rate": 8.51278195488722e-06,
+      "loss": 0.351,
+      "step": 9890
+    },
+    {
+      "epoch": 14.89,
+      "grad_norm": 5.427340030670166,
+      "learning_rate": 8.511278195488722e-06,
+      "loss": 0.3488,
+      "step": 9900
+    },
+    {
+      "epoch": 14.9,
+      "grad_norm": 4.465277671813965,
+      "learning_rate": 8.509774436090227e-06,
+      "loss": 0.32,
+      "step": 9910
+    },
+    {
+      "epoch": 14.92,
+      "grad_norm": 4.331689834594727,
+      "learning_rate": 8.50827067669173e-06,
+      "loss": 0.4442,
+      "step": 9920
+    },
+    {
+      "epoch": 14.93,
+      "grad_norm": 5.798705577850342,
+      "learning_rate": 8.506766917293232e-06,
+      "loss": 0.4003,
+      "step": 9930
+    },
+    {
+      "epoch": 14.95,
+      "grad_norm": 9.353456497192383,
+      "learning_rate": 8.505263157894738e-06,
+      "loss": 0.3969,
+      "step": 9940
+    },
+    {
+      "epoch": 14.96,
+      "grad_norm": 7.088143825531006,
+      "learning_rate": 8.503759398496241e-06,
+      "loss": 0.3749,
+      "step": 9950
+    },
+    {
+      "epoch": 14.98,
+      "grad_norm": 6.825297832489014,
+      "learning_rate": 8.502255639097745e-06,
+      "loss": 0.3532,
+      "step": 9960
+    },
+    {
+      "epoch": 14.99,
+      "grad_norm": 3.3977503776550293,
+      "learning_rate": 8.500751879699248e-06,
+      "loss": 0.3227,
+      "step": 9970
+    },
+    {
+      "epoch": 15.0,
+      "eval_accuracy": 0.9335,
+      "eval_loss": 0.2566547989845276,
+      "eval_runtime": 84.4818,
+      "eval_samples_per_second": 118.369,
+      "eval_steps_per_second": 0.473,
+      "step": 9975
+    },
+    {
+      "epoch": 15.01,
+      "grad_norm": 3.752779483795166,
+      "learning_rate": 8.499248120300753e-06,
+      "loss": 0.2741,
+      "step": 9980
+    },
+    {
+      "epoch": 15.02,
+      "grad_norm": 5.626667499542236,
+      "learning_rate": 8.497744360902255e-06,
+      "loss": 0.3584,
+      "step": 9990
+    },
+    {
+      "epoch": 15.04,
+      "grad_norm": 5.8750834465026855,
+      "learning_rate": 8.49624060150376e-06,
+      "loss": 0.3563,
+      "step": 10000
+    },
+    {
+      "epoch": 15.05,
+      "grad_norm": 6.653073787689209,
+      "learning_rate": 8.494736842105264e-06,
+      "loss": 0.3808,
+      "step": 10010
+    },
+    {
+      "epoch": 15.07,
+      "grad_norm": 6.017663955688477,
+      "learning_rate": 8.493233082706767e-06,
+      "loss": 0.3165,
+      "step": 10020
+    },
+    {
+      "epoch": 15.08,
+      "grad_norm": 4.427550792694092,
+      "learning_rate": 8.491729323308271e-06,
+      "loss": 0.4417,
+      "step": 10030
+    },
+    {
+      "epoch": 15.1,
+      "grad_norm": 7.968047142028809,
+      "learning_rate": 8.490225563909775e-06,
+      "loss": 0.3807,
+      "step": 10040
+    },
+    {
+      "epoch": 15.11,
+      "grad_norm": 6.646710395812988,
+      "learning_rate": 8.488721804511278e-06,
+      "loss": 0.3014,
+      "step": 10050
+    },
+    {
+      "epoch": 15.13,
+      "grad_norm": 5.180790424346924,
+      "learning_rate": 8.487218045112783e-06,
+      "loss": 0.3737,
+      "step": 10060
+    },
+    {
+      "epoch": 15.14,
+      "grad_norm": 8.142125129699707,
+      "learning_rate": 8.485714285714287e-06,
+      "loss": 0.3953,
+      "step": 10070
+    },
+    {
+      "epoch": 15.16,
+      "grad_norm": 5.258510589599609,
+      "learning_rate": 8.48421052631579e-06,
+      "loss": 0.2742,
+      "step": 10080
+    },
+    {
+      "epoch": 15.17,
+      "grad_norm": 7.299388408660889,
+      "learning_rate": 8.482706766917294e-06,
+      "loss": 0.388,
+      "step": 10090
+    },
+    {
+      "epoch": 15.19,
+      "grad_norm": 9.05027961730957,
+      "learning_rate": 8.481203007518797e-06,
+      "loss": 0.3364,
+      "step": 10100
+    },
+    {
+      "epoch": 15.2,
+      "grad_norm": 8.713417053222656,
+      "learning_rate": 8.4796992481203e-06,
+      "loss": 0.2847,
+      "step": 10110
+    },
+    {
+      "epoch": 15.22,
+      "grad_norm": 4.598002910614014,
+      "learning_rate": 8.478195488721806e-06,
+      "loss": 0.3536,
+      "step": 10120
+    },
+    {
+      "epoch": 15.23,
+      "grad_norm": 8.883098602294922,
+      "learning_rate": 8.476691729323308e-06,
+      "loss": 0.3917,
+      "step": 10130
+    },
+    {
+      "epoch": 15.25,
+      "grad_norm": 9.343592643737793,
+      "learning_rate": 8.475187969924813e-06,
+      "loss": 0.3912,
+      "step": 10140
+    },
+    {
+      "epoch": 15.26,
+      "grad_norm": 11.456267356872559,
+      "learning_rate": 8.473684210526317e-06,
+      "loss": 0.3171,
+      "step": 10150
+    },
+    {
+      "epoch": 15.28,
+      "grad_norm": 7.874906539916992,
+      "learning_rate": 8.47218045112782e-06,
+      "loss": 0.4443,
+      "step": 10160
+    },
+    {
+      "epoch": 15.29,
+      "grad_norm": 7.280247211456299,
+      "learning_rate": 8.470676691729324e-06,
+      "loss": 0.4121,
+      "step": 10170
+    },
+    {
+      "epoch": 15.31,
+      "grad_norm": 7.839987754821777,
+      "learning_rate": 8.469172932330829e-06,
+      "loss": 0.4284,
+      "step": 10180
+    },
+    {
+      "epoch": 15.32,
+      "grad_norm": 5.39335298538208,
+      "learning_rate": 8.46766917293233e-06,
+      "loss": 0.4311,
+      "step": 10190
+    },
+    {
+      "epoch": 15.34,
+      "grad_norm": 8.08191204071045,
+      "learning_rate": 8.466165413533836e-06,
+      "loss": 0.3789,
+      "step": 10200
+    },
+    {
+      "epoch": 15.35,
+      "grad_norm": 3.493443250656128,
+      "learning_rate": 8.46466165413534e-06,
+      "loss": 0.3526,
+      "step": 10210
+    },
+    {
+      "epoch": 15.37,
+      "grad_norm": 7.41270637512207,
+      "learning_rate": 8.463157894736843e-06,
+      "loss": 0.4283,
+      "step": 10220
+    },
+    {
+      "epoch": 15.38,
+      "grad_norm": 6.4891486167907715,
+      "learning_rate": 8.461654135338346e-06,
+      "loss": 0.298,
+      "step": 10230
+    },
+    {
+      "epoch": 15.4,
+      "grad_norm": 6.028573989868164,
+      "learning_rate": 8.46015037593985e-06,
+      "loss": 0.3996,
+      "step": 10240
+    },
+    {
+      "epoch": 15.41,
+      "grad_norm": 3.365438938140869,
+      "learning_rate": 8.458646616541353e-06,
+      "loss": 0.334,
+      "step": 10250
+    },
+    {
+      "epoch": 15.43,
+      "grad_norm": 5.2097015380859375,
+      "learning_rate": 8.457142857142859e-06,
+      "loss": 0.3555,
+      "step": 10260
+    },
+    {
+      "epoch": 15.44,
+      "grad_norm": 4.656721591949463,
+      "learning_rate": 8.455639097744362e-06,
+      "loss": 0.398,
+      "step": 10270
+    },
+    {
+      "epoch": 15.46,
+      "grad_norm": 11.353671073913574,
+      "learning_rate": 8.454135338345866e-06,
+      "loss": 0.3341,
+      "step": 10280
+    },
+    {
+      "epoch": 15.47,
+      "grad_norm": 6.971073627471924,
+      "learning_rate": 8.45263157894737e-06,
+      "loss": 0.3569,
+      "step": 10290
+    },
+    {
+      "epoch": 15.49,
+      "grad_norm": 5.809013843536377,
+      "learning_rate": 8.451127819548873e-06,
+      "loss": 0.3642,
+      "step": 10300
+    },
+    {
+      "epoch": 15.5,
+      "grad_norm": 7.3322834968566895,
+      "learning_rate": 8.449624060150376e-06,
+      "loss": 0.3814,
+      "step": 10310
+    },
+    {
+      "epoch": 15.52,
+      "grad_norm": 9.322681427001953,
+      "learning_rate": 8.448120300751882e-06,
+      "loss": 0.3455,
+      "step": 10320
+    },
+    {
+      "epoch": 15.53,
+      "grad_norm": 7.197205066680908,
+      "learning_rate": 8.446616541353383e-06,
+      "loss": 0.3652,
+      "step": 10330
+    },
+    {
+      "epoch": 15.55,
+      "grad_norm": 7.1502766609191895,
+      "learning_rate": 8.445112781954889e-06,
+      "loss": 0.3882,
+      "step": 10340
+    },
+    {
+      "epoch": 15.56,
+      "grad_norm": 4.826005935668945,
+      "learning_rate": 8.443609022556392e-06,
+      "loss": 0.3878,
+      "step": 10350
+    },
+    {
+      "epoch": 15.58,
+      "grad_norm": 8.432343482971191,
+      "learning_rate": 8.442105263157896e-06,
+      "loss": 0.3778,
+      "step": 10360
+    },
+    {
+      "epoch": 15.59,
+      "grad_norm": 6.166329383850098,
+      "learning_rate": 8.440601503759399e-06,
+      "loss": 0.3675,
+      "step": 10370
+    },
+    {
+      "epoch": 15.61,
+      "grad_norm": 9.84304428100586,
+      "learning_rate": 8.439097744360903e-06,
+      "loss": 0.385,
+      "step": 10380
+    },
+    {
+      "epoch": 15.62,
+      "grad_norm": 4.937039375305176,
+      "learning_rate": 8.437593984962406e-06,
+      "loss": 0.3558,
+      "step": 10390
+    },
+    {
+      "epoch": 15.64,
+      "grad_norm": 5.817636966705322,
+      "learning_rate": 8.436090225563911e-06,
+      "loss": 0.3605,
+      "step": 10400
+    },
+    {
+      "epoch": 15.65,
+      "grad_norm": 6.982740879058838,
+      "learning_rate": 8.434586466165415e-06,
+      "loss": 0.2966,
+      "step": 10410
+    },
+    {
+      "epoch": 15.67,
+      "grad_norm": 5.2945098876953125,
+      "learning_rate": 8.433082706766918e-06,
+      "loss": 0.3279,
+      "step": 10420
+    },
+    {
+      "epoch": 15.68,
+      "grad_norm": 7.526950359344482,
+      "learning_rate": 8.431578947368422e-06,
+      "loss": 0.414,
+      "step": 10430
+    },
+    {
+      "epoch": 15.7,
+      "grad_norm": 8.325518608093262,
+      "learning_rate": 8.430075187969925e-06,
+      "loss": 0.314,
+      "step": 10440
+    },
+    {
+      "epoch": 15.71,
+      "grad_norm": 2.876897096633911,
+      "learning_rate": 8.428571428571429e-06,
+      "loss": 0.354,
+      "step": 10450
+    },
+    {
+      "epoch": 15.73,
+      "grad_norm": 6.776325702667236,
+      "learning_rate": 8.427067669172932e-06,
+      "loss": 0.3462,
+      "step": 10460
+    },
+    {
+      "epoch": 15.74,
+      "grad_norm": 8.158499717712402,
+      "learning_rate": 8.425563909774438e-06,
+      "loss": 0.3307,
+      "step": 10470
+    },
+    {
+      "epoch": 15.76,
+      "grad_norm": 10.01845645904541,
+      "learning_rate": 8.42406015037594e-06,
+      "loss": 0.3196,
+      "step": 10480
+    },
+    {
+      "epoch": 15.77,
+      "grad_norm": 4.778624534606934,
+      "learning_rate": 8.422556390977445e-06,
+      "loss": 0.3412,
+      "step": 10490
+    },
+    {
+      "epoch": 15.79,
+      "grad_norm": 5.8504157066345215,
+      "learning_rate": 8.421052631578948e-06,
+      "loss": 0.4183,
+      "step": 10500
+    },
+    {
+      "epoch": 15.8,
+      "grad_norm": 5.351130962371826,
+      "learning_rate": 8.419548872180452e-06,
+      "loss": 0.2639,
+      "step": 10510
+    },
+    {
+      "epoch": 15.82,
+      "grad_norm": 7.211291313171387,
+      "learning_rate": 8.418045112781955e-06,
+      "loss": 0.3021,
+      "step": 10520
+    },
+    {
+      "epoch": 15.83,
+      "grad_norm": 6.899810791015625,
+      "learning_rate": 8.416541353383459e-06,
+      "loss": 0.4069,
+      "step": 10530
+    },
+    {
+      "epoch": 15.85,
+      "grad_norm": 4.64746618270874,
+      "learning_rate": 8.415037593984962e-06,
+      "loss": 0.2798,
+      "step": 10540
+    },
+    {
+      "epoch": 15.86,
+      "grad_norm": 9.508644104003906,
+      "learning_rate": 8.413533834586468e-06,
+      "loss": 0.3791,
+      "step": 10550
+    },
+    {
+      "epoch": 15.88,
+      "grad_norm": 6.956771373748779,
+      "learning_rate": 8.412030075187971e-06,
+      "loss": 0.4188,
+      "step": 10560
+    },
+    {
+      "epoch": 15.89,
+      "grad_norm": 6.4203667640686035,
+      "learning_rate": 8.410526315789475e-06,
+      "loss": 0.3749,
+      "step": 10570
+    },
+    {
+      "epoch": 15.91,
+      "grad_norm": 8.310030937194824,
+      "learning_rate": 8.409022556390978e-06,
+      "loss": 0.4605,
+      "step": 10580
+    },
+    {
+      "epoch": 15.92,
+      "grad_norm": 8.788355827331543,
+      "learning_rate": 8.407518796992482e-06,
+      "loss": 0.3811,
+      "step": 10590
+    },
+    {
+      "epoch": 15.94,
+      "grad_norm": 5.931136131286621,
+      "learning_rate": 8.406015037593985e-06,
+      "loss": 0.4129,
+      "step": 10600
+    },
+    {
+      "epoch": 15.95,
+      "grad_norm": 7.846260070800781,
+      "learning_rate": 8.40451127819549e-06,
+      "loss": 0.3889,
+      "step": 10610
+    },
+    {
+      "epoch": 15.97,
+      "grad_norm": 6.834481239318848,
+      "learning_rate": 8.403007518796992e-06,
+      "loss": 0.4303,
+      "step": 10620
+    },
+    {
+      "epoch": 15.98,
+      "grad_norm": 3.6619720458984375,
+      "learning_rate": 8.401503759398497e-06,
+      "loss": 0.3483,
+      "step": 10630
+    },
+    {
+      "epoch": 16.0,
+      "grad_norm": 12.63433837890625,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 0.386,
+      "step": 10640
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.931,
+      "eval_loss": 0.25709524750709534,
+      "eval_runtime": 84.9959,
+      "eval_samples_per_second": 117.653,
+      "eval_steps_per_second": 0.471,
+      "step": 10640
+    },
+    {
+      "epoch": 16.02,
+      "grad_norm": 7.7511796951293945,
+      "learning_rate": 8.398496240601504e-06,
+      "loss": 0.3916,
+      "step": 10650
+    },
+    {
+      "epoch": 16.03,
+      "grad_norm": 4.640677452087402,
+      "learning_rate": 8.396992481203008e-06,
+      "loss": 0.395,
+      "step": 10660
+    },
+    {
+      "epoch": 16.05,
+      "grad_norm": 7.270589828491211,
+      "learning_rate": 8.395488721804511e-06,
+      "loss": 0.4299,
+      "step": 10670
+    },
+    {
+      "epoch": 16.06,
+      "grad_norm": 5.698379993438721,
+      "learning_rate": 8.393984962406015e-06,
+      "loss": 0.3792,
+      "step": 10680
+    },
+    {
+      "epoch": 16.08,
+      "grad_norm": 29.69732093811035,
+      "learning_rate": 8.39248120300752e-06,
+      "loss": 0.3564,
+      "step": 10690
+    },
+    {
+      "epoch": 16.09,
+      "grad_norm": 5.11942720413208,
+      "learning_rate": 8.390977443609024e-06,
+      "loss": 0.3284,
+      "step": 10700
+    },
+    {
+      "epoch": 16.11,
+      "grad_norm": 6.748551368713379,
+      "learning_rate": 8.389473684210527e-06,
+      "loss": 0.3582,
+      "step": 10710
+    },
+    {
+      "epoch": 16.12,
+      "grad_norm": 6.748464584350586,
+      "learning_rate": 8.38796992481203e-06,
+      "loss": 0.3707,
+      "step": 10720
+    },
+    {
+      "epoch": 16.14,
+      "grad_norm": 6.992805480957031,
+      "learning_rate": 8.386466165413534e-06,
+      "loss": 0.4388,
+      "step": 10730
+    },
+    {
+      "epoch": 16.15,
+      "grad_norm": 4.689752578735352,
+      "learning_rate": 8.384962406015038e-06,
+      "loss": 0.3394,
+      "step": 10740
+    },
+    {
+      "epoch": 16.17,
+      "grad_norm": 7.47608757019043,
+      "learning_rate": 8.383458646616543e-06,
+      "loss": 0.3662,
+      "step": 10750
+    },
+    {
+      "epoch": 16.18,
+      "grad_norm": 8.161937713623047,
+      "learning_rate": 8.381954887218045e-06,
+      "loss": 0.3729,
+      "step": 10760
+    },
+    {
+      "epoch": 16.2,
+      "grad_norm": 4.649080276489258,
+      "learning_rate": 8.38045112781955e-06,
+      "loss": 0.4012,
+      "step": 10770
+    },
+    {
+      "epoch": 16.21,
+      "grad_norm": 5.3081512451171875,
+      "learning_rate": 8.378947368421054e-06,
+      "loss": 0.3275,
+      "step": 10780
+    },
+    {
+      "epoch": 16.23,
+      "grad_norm": 8.424674034118652,
+      "learning_rate": 8.377443609022557e-06,
+      "loss": 0.3169,
+      "step": 10790
+    },
+    {
+      "epoch": 16.24,
+      "grad_norm": 7.213728427886963,
+      "learning_rate": 8.37593984962406e-06,
+      "loss": 0.4158,
+      "step": 10800
+    },
+    {
+      "epoch": 16.26,
+      "grad_norm": 4.228058815002441,
+      "learning_rate": 8.374436090225566e-06,
+      "loss": 0.2967,
+      "step": 10810
+    },
+    {
+      "epoch": 16.27,
+      "grad_norm": 9.091151237487793,
+      "learning_rate": 8.372932330827068e-06,
+      "loss": 0.3631,
+      "step": 10820
+    },
+    {
+      "epoch": 16.29,
+      "grad_norm": 7.325952053070068,
+      "learning_rate": 8.371428571428573e-06,
+      "loss": 0.4045,
+      "step": 10830
+    },
+    {
+      "epoch": 16.3,
+      "grad_norm": 8.557323455810547,
+      "learning_rate": 8.369924812030076e-06,
+      "loss": 0.3657,
+      "step": 10840
+    },
+    {
+      "epoch": 16.32,
+      "grad_norm": 8.98796558380127,
+      "learning_rate": 8.36842105263158e-06,
+      "loss": 0.3812,
+      "step": 10850
+    },
+    {
+      "epoch": 16.33,
+      "grad_norm": 14.74909496307373,
+      "learning_rate": 8.366917293233083e-06,
+      "loss": 0.3966,
+      "step": 10860
+    },
+    {
+      "epoch": 16.35,
+      "grad_norm": 6.872273921966553,
+      "learning_rate": 8.365413533834587e-06,
+      "loss": 0.3797,
+      "step": 10870
+    },
+    {
+      "epoch": 16.36,
+      "grad_norm": 8.392032623291016,
+      "learning_rate": 8.36390977443609e-06,
+      "loss": 0.3864,
+      "step": 10880
+    },
+    {
+      "epoch": 16.38,
+      "grad_norm": 8.61768627166748,
+      "learning_rate": 8.362406015037596e-06,
+      "loss": 0.3434,
+      "step": 10890
+    },
+    {
+      "epoch": 16.39,
+      "grad_norm": 6.622392177581787,
+      "learning_rate": 8.3609022556391e-06,
+      "loss": 0.3751,
+      "step": 10900
+    },
+    {
+      "epoch": 16.41,
+      "grad_norm": 7.108027458190918,
+      "learning_rate": 8.359398496240603e-06,
+      "loss": 0.3542,
+      "step": 10910
+    },
+    {
+      "epoch": 16.42,
+      "grad_norm": 7.145939826965332,
+      "learning_rate": 8.357894736842106e-06,
+      "loss": 0.3586,
+      "step": 10920
+    },
+    {
+      "epoch": 16.44,
+      "grad_norm": 16.715761184692383,
+      "learning_rate": 8.35639097744361e-06,
+      "loss": 0.4365,
+      "step": 10930
+    },
+    {
+      "epoch": 16.45,
+      "grad_norm": 3.3526134490966797,
+      "learning_rate": 8.354887218045113e-06,
+      "loss": 0.38,
+      "step": 10940
+    },
+    {
+      "epoch": 16.47,
+      "grad_norm": 4.425145149230957,
+      "learning_rate": 8.353383458646617e-06,
+      "loss": 0.2944,
+      "step": 10950
+    },
+    {
+      "epoch": 16.48,
+      "grad_norm": 4.0147552490234375,
+      "learning_rate": 8.35187969924812e-06,
+      "loss": 0.3836,
+      "step": 10960
+    },
+    {
+      "epoch": 16.5,
+      "grad_norm": 6.398830413818359,
+      "learning_rate": 8.350375939849624e-06,
+      "loss": 0.3402,
+      "step": 10970
+    },
+    {
+      "epoch": 16.51,
+      "grad_norm": 5.556189060211182,
+      "learning_rate": 8.348872180451129e-06,
+      "loss": 0.3728,
+      "step": 10980
+    },
+    {
+      "epoch": 16.53,
+      "grad_norm": 4.928891658782959,
+      "learning_rate": 8.347368421052633e-06,
+      "loss": 0.2993,
+      "step": 10990
+    },
+    {
+      "epoch": 16.54,
+      "grad_norm": 8.429544448852539,
+      "learning_rate": 8.345864661654136e-06,
+      "loss": 0.3896,
+      "step": 11000
+    },
+    {
+      "epoch": 16.56,
+      "grad_norm": 9.285510063171387,
+      "learning_rate": 8.34436090225564e-06,
+      "loss": 0.3428,
+      "step": 11010
+    },
+    {
+      "epoch": 16.57,
+      "grad_norm": 5.118491172790527,
+      "learning_rate": 8.342857142857143e-06,
+      "loss": 0.3583,
+      "step": 11020
+    },
+    {
+      "epoch": 16.59,
+      "grad_norm": 9.391587257385254,
+      "learning_rate": 8.341353383458647e-06,
+      "loss": 0.312,
+      "step": 11030
+    },
+    {
+      "epoch": 16.6,
+      "grad_norm": 22.18227767944336,
+      "learning_rate": 8.339849624060152e-06,
+      "loss": 0.3549,
+      "step": 11040
+    },
+    {
+      "epoch": 16.62,
+      "grad_norm": 5.834057331085205,
+      "learning_rate": 8.338345864661654e-06,
+      "loss": 0.4242,
+      "step": 11050
+    },
+    {
+      "epoch": 16.63,
+      "grad_norm": 5.136647701263428,
+      "learning_rate": 8.336842105263159e-06,
+      "loss": 0.3623,
+      "step": 11060
+    },
+    {
+      "epoch": 16.65,
+      "grad_norm": 5.481499195098877,
+      "learning_rate": 8.335338345864662e-06,
+      "loss": 0.3745,
+      "step": 11070
+    },
+    {
+      "epoch": 16.66,
+      "grad_norm": 3.6383814811706543,
+      "learning_rate": 8.333834586466166e-06,
+      "loss": 0.3526,
+      "step": 11080
+    },
+    {
+      "epoch": 16.68,
+      "grad_norm": 4.198364734649658,
+      "learning_rate": 8.33233082706767e-06,
+      "loss": 0.3247,
+      "step": 11090
+    },
+    {
+      "epoch": 16.69,
+      "grad_norm": 6.066871166229248,
+      "learning_rate": 8.330827067669175e-06,
+      "loss": 0.346,
+      "step": 11100
+    },
+    {
+      "epoch": 16.71,
+      "grad_norm": 4.822031497955322,
+      "learning_rate": 8.329323308270676e-06,
+      "loss": 0.3421,
+      "step": 11110
+    },
+    {
+      "epoch": 16.72,
+      "grad_norm": 8.0927152633667,
+      "learning_rate": 8.327819548872182e-06,
+      "loss": 0.3212,
+      "step": 11120
+    },
+    {
+      "epoch": 16.74,
+      "grad_norm": 5.737279415130615,
+      "learning_rate": 8.326315789473685e-06,
+      "loss": 0.338,
+      "step": 11130
+    },
+    {
+      "epoch": 16.75,
+      "grad_norm": 6.6013994216918945,
+      "learning_rate": 8.324812030075189e-06,
+      "loss": 0.4221,
+      "step": 11140
+    },
+    {
+      "epoch": 16.77,
+      "grad_norm": 3.22804856300354,
+      "learning_rate": 8.323308270676692e-06,
+      "loss": 0.2824,
+      "step": 11150
+    },
+    {
+      "epoch": 16.78,
+      "grad_norm": 7.42767333984375,
+      "learning_rate": 8.321804511278196e-06,
+      "loss": 0.3654,
+      "step": 11160
+    },
+    {
+      "epoch": 16.8,
+      "grad_norm": 6.393558979034424,
+      "learning_rate": 8.3203007518797e-06,
+      "loss": 0.3939,
+      "step": 11170
+    },
+    {
+      "epoch": 16.81,
+      "grad_norm": 6.1491241455078125,
+      "learning_rate": 8.318796992481204e-06,
+      "loss": 0.3454,
+      "step": 11180
+    },
+    {
+      "epoch": 16.83,
+      "grad_norm": 4.844079971313477,
+      "learning_rate": 8.317293233082708e-06,
+      "loss": 0.3744,
+      "step": 11190
+    },
+    {
+      "epoch": 16.84,
+      "grad_norm": 7.577675819396973,
+      "learning_rate": 8.315789473684212e-06,
+      "loss": 0.4259,
+      "step": 11200
+    },
+    {
+      "epoch": 16.86,
+      "grad_norm": 5.073234558105469,
+      "learning_rate": 8.314285714285715e-06,
+      "loss": 0.343,
+      "step": 11210
+    },
+    {
+      "epoch": 16.87,
+      "grad_norm": 4.934657096862793,
+      "learning_rate": 8.312781954887219e-06,
+      "loss": 0.3834,
+      "step": 11220
+    },
+    {
+      "epoch": 16.89,
+      "grad_norm": 4.744530200958252,
+      "learning_rate": 8.311278195488722e-06,
+      "loss": 0.2937,
+      "step": 11230
+    },
+    {
+      "epoch": 16.9,
+      "grad_norm": 7.569250583648682,
+      "learning_rate": 8.309774436090227e-06,
+      "loss": 0.3719,
+      "step": 11240
+    },
+    {
+      "epoch": 16.92,
+      "grad_norm": 7.076653480529785,
+      "learning_rate": 8.308270676691729e-06,
+      "loss": 0.3489,
+      "step": 11250
+    },
+    {
+      "epoch": 16.93,
+      "grad_norm": 7.1391520500183105,
+      "learning_rate": 8.306766917293234e-06,
+      "loss": 0.3325,
+      "step": 11260
+    },
+    {
+      "epoch": 16.95,
+      "grad_norm": 4.94738245010376,
+      "learning_rate": 8.305263157894738e-06,
+      "loss": 0.3421,
+      "step": 11270
+    },
+    {
+      "epoch": 16.96,
+      "grad_norm": 6.052053451538086,
+      "learning_rate": 8.303759398496241e-06,
+      "loss": 0.4267,
+      "step": 11280
+    },
+    {
+      "epoch": 16.98,
+      "grad_norm": 6.822144985198975,
+      "learning_rate": 8.302255639097745e-06,
+      "loss": 0.4932,
+      "step": 11290
+    },
+    {
+      "epoch": 16.99,
+      "grad_norm": 7.0719218254089355,
+      "learning_rate": 8.300751879699248e-06,
+      "loss": 0.3688,
+      "step": 11300
+    },
+    {
+      "epoch": 17.0,
+      "eval_accuracy": 0.9346,
+      "eval_loss": 0.25758126378059387,
+      "eval_runtime": 84.7562,
+      "eval_samples_per_second": 117.986,
+      "eval_steps_per_second": 0.472,
+      "step": 11305
+    },
+    {
+      "epoch": 17.01,
+      "grad_norm": 4.741504669189453,
+      "learning_rate": 8.299248120300752e-06,
+      "loss": 0.3837,
+      "step": 11310
+    },
+    {
+      "epoch": 17.02,
+      "grad_norm": 6.912674427032471,
+      "learning_rate": 8.297744360902257e-06,
+      "loss": 0.2938,
+      "step": 11320
+    },
+    {
+      "epoch": 17.04,
+      "grad_norm": 5.881788730621338,
+      "learning_rate": 8.29624060150376e-06,
+      "loss": 0.3984,
+      "step": 11330
+    },
+    {
+      "epoch": 17.05,
+      "grad_norm": 5.705871105194092,
+      "learning_rate": 8.294736842105264e-06,
+      "loss": 0.3793,
+      "step": 11340
+    },
+    {
+      "epoch": 17.07,
+      "grad_norm": 5.036585330963135,
+      "learning_rate": 8.293233082706768e-06,
+      "loss": 0.365,
+      "step": 11350
+    },
+    {
+      "epoch": 17.08,
+      "grad_norm": 5.040714740753174,
+      "learning_rate": 8.291729323308271e-06,
+      "loss": 0.3662,
+      "step": 11360
+    },
+    {
+      "epoch": 17.1,
+      "grad_norm": 7.933087348937988,
+      "learning_rate": 8.290225563909775e-06,
+      "loss": 0.3204,
+      "step": 11370
+    },
+    {
+      "epoch": 17.11,
+      "grad_norm": 8.096324920654297,
+      "learning_rate": 8.28872180451128e-06,
+      "loss": 0.3383,
+      "step": 11380
+    },
+    {
+      "epoch": 17.13,
+      "grad_norm": 5.693844318389893,
+      "learning_rate": 8.287218045112782e-06,
+      "loss": 0.3455,
+      "step": 11390
+    },
+    {
+      "epoch": 17.14,
+      "grad_norm": 5.274537086486816,
+      "learning_rate": 8.285714285714287e-06,
+      "loss": 0.3695,
+      "step": 11400
+    },
+    {
+      "epoch": 17.16,
+      "grad_norm": 5.249573230743408,
+      "learning_rate": 8.28421052631579e-06,
+      "loss": 0.3801,
+      "step": 11410
+    },
+    {
+      "epoch": 17.17,
+      "grad_norm": 6.644190311431885,
+      "learning_rate": 8.282706766917294e-06,
+      "loss": 0.3355,
+      "step": 11420
+    },
+    {
+      "epoch": 17.19,
+      "grad_norm": 8.79143238067627,
+      "learning_rate": 8.281203007518798e-06,
+      "loss": 0.3857,
+      "step": 11430
+    },
+    {
+      "epoch": 17.2,
+      "grad_norm": 4.0490851402282715,
+      "learning_rate": 8.279699248120301e-06,
+      "loss": 0.3224,
+      "step": 11440
+    },
+    {
+      "epoch": 17.22,
+      "grad_norm": 8.35059928894043,
+      "learning_rate": 8.278195488721805e-06,
+      "loss": 0.372,
+      "step": 11450
+    },
+    {
+      "epoch": 17.23,
+      "grad_norm": 5.733313083648682,
+      "learning_rate": 8.276691729323308e-06,
+      "loss": 0.3511,
+      "step": 11460
+    },
+    {
+      "epoch": 17.25,
+      "grad_norm": 4.974298477172852,
+      "learning_rate": 8.275187969924813e-06,
+      "loss": 0.332,
+      "step": 11470
+    },
+    {
+      "epoch": 17.26,
+      "grad_norm": 6.859165191650391,
+      "learning_rate": 8.273684210526317e-06,
+      "loss": 0.3053,
+      "step": 11480
+    },
+    {
+      "epoch": 17.28,
+      "grad_norm": 7.4038591384887695,
+      "learning_rate": 8.27218045112782e-06,
+      "loss": 0.3419,
+      "step": 11490
+    },
+    {
+      "epoch": 17.29,
+      "grad_norm": 4.045393943786621,
+      "learning_rate": 8.270676691729324e-06,
+      "loss": 0.3656,
+      "step": 11500
+    },
+    {
+      "epoch": 17.31,
+      "grad_norm": 2.965898036956787,
+      "learning_rate": 8.269172932330827e-06,
+      "loss": 0.4222,
+      "step": 11510
+    },
+    {
+      "epoch": 17.32,
+      "grad_norm": 6.6445746421813965,
+      "learning_rate": 8.267669172932331e-06,
+      "loss": 0.4351,
+      "step": 11520
+    },
+    {
+      "epoch": 17.34,
+      "grad_norm": 11.220673561096191,
+      "learning_rate": 8.266165413533836e-06,
+      "loss": 0.3738,
+      "step": 11530
+    },
+    {
+      "epoch": 17.35,
+      "grad_norm": 3.7990376949310303,
+      "learning_rate": 8.264661654135338e-06,
+      "loss": 0.2779,
+      "step": 11540
+    },
+    {
+      "epoch": 17.37,
+      "grad_norm": 6.193857669830322,
+      "learning_rate": 8.263157894736843e-06,
+      "loss": 0.3864,
+      "step": 11550
+    },
+    {
+      "epoch": 17.38,
+      "grad_norm": 7.089908123016357,
+      "learning_rate": 8.261654135338347e-06,
+      "loss": 0.3243,
+      "step": 11560
+    },
+    {
+      "epoch": 17.4,
+      "grad_norm": 10.148313522338867,
+      "learning_rate": 8.26015037593985e-06,
+      "loss": 0.3848,
+      "step": 11570
+    },
+    {
+      "epoch": 17.41,
+      "grad_norm": 7.47261905670166,
+      "learning_rate": 8.258646616541354e-06,
+      "loss": 0.3958,
+      "step": 11580
+    },
+    {
+      "epoch": 17.43,
+      "grad_norm": 8.237654685974121,
+      "learning_rate": 8.257142857142857e-06,
+      "loss": 0.4104,
+      "step": 11590
+    },
+    {
+      "epoch": 17.44,
+      "grad_norm": 7.028960227966309,
+      "learning_rate": 8.25563909774436e-06,
+      "loss": 0.3515,
+      "step": 11600
+    },
+    {
+      "epoch": 17.46,
+      "grad_norm": 6.804955959320068,
+      "learning_rate": 8.254135338345866e-06,
+      "loss": 0.3555,
+      "step": 11610
+    },
+    {
+      "epoch": 17.47,
+      "grad_norm": 8.740710258483887,
+      "learning_rate": 8.25263157894737e-06,
+      "loss": 0.328,
+      "step": 11620
+    },
+    {
+      "epoch": 17.49,
+      "grad_norm": 6.383413314819336,
+      "learning_rate": 8.251127819548873e-06,
+      "loss": 0.3352,
+      "step": 11630
+    },
+    {
+      "epoch": 17.5,
+      "grad_norm": 8.289705276489258,
+      "learning_rate": 8.249624060150377e-06,
+      "loss": 0.4067,
+      "step": 11640
+    },
+    {
+      "epoch": 17.52,
+      "grad_norm": 6.0566911697387695,
+      "learning_rate": 8.24812030075188e-06,
+      "loss": 0.3556,
+      "step": 11650
+    },
+    {
+      "epoch": 17.53,
+      "grad_norm": 9.813027381896973,
+      "learning_rate": 8.246616541353384e-06,
+      "loss": 0.3856,
+      "step": 11660
+    },
+    {
+      "epoch": 17.55,
+      "grad_norm": 6.0970988273620605,
+      "learning_rate": 8.245112781954889e-06,
+      "loss": 0.396,
+      "step": 11670
+    },
+    {
+      "epoch": 17.56,
+      "grad_norm": 4.837037086486816,
+      "learning_rate": 8.24360902255639e-06,
+      "loss": 0.3194,
+      "step": 11680
+    },
+    {
+      "epoch": 17.58,
+      "grad_norm": 10.178328514099121,
+      "learning_rate": 8.242105263157896e-06,
+      "loss": 0.4108,
+      "step": 11690
+    },
+    {
+      "epoch": 17.59,
+      "grad_norm": 3.696746826171875,
+      "learning_rate": 8.2406015037594e-06,
+      "loss": 0.3198,
+      "step": 11700
+    },
+    {
+      "epoch": 17.61,
+      "grad_norm": 6.4773993492126465,
+      "learning_rate": 8.239097744360903e-06,
+      "loss": 0.2889,
+      "step": 11710
+    },
+    {
+      "epoch": 17.62,
+      "grad_norm": 9.036526679992676,
+      "learning_rate": 8.237593984962406e-06,
+      "loss": 0.4616,
+      "step": 11720
+    },
+    {
+      "epoch": 17.64,
+      "grad_norm": 5.1061320304870605,
+      "learning_rate": 8.236090225563912e-06,
+      "loss": 0.3941,
+      "step": 11730
+    },
+    {
+      "epoch": 17.65,
+      "grad_norm": 5.17496919631958,
+      "learning_rate": 8.234586466165413e-06,
+      "loss": 0.3883,
+      "step": 11740
+    },
+    {
+      "epoch": 17.67,
+      "grad_norm": 4.007594585418701,
+      "learning_rate": 8.233082706766919e-06,
+      "loss": 0.3097,
+      "step": 11750
+    },
+    {
+      "epoch": 17.68,
+      "grad_norm": 3.903956174850464,
+      "learning_rate": 8.231578947368422e-06,
+      "loss": 0.3473,
+      "step": 11760
+    },
+    {
+      "epoch": 17.7,
+      "grad_norm": 9.270066261291504,
+      "learning_rate": 8.230075187969926e-06,
+      "loss": 0.3255,
+      "step": 11770
+    },
+    {
+      "epoch": 17.71,
+      "grad_norm": 4.118042469024658,
+      "learning_rate": 8.22857142857143e-06,
+      "loss": 0.335,
+      "step": 11780
+    },
+    {
+      "epoch": 17.73,
+      "grad_norm": 5.715611457824707,
+      "learning_rate": 8.227067669172933e-06,
+      "loss": 0.422,
+      "step": 11790
+    },
+    {
+      "epoch": 17.74,
+      "grad_norm": 5.848507881164551,
+      "learning_rate": 8.225563909774436e-06,
+      "loss": 0.3465,
+      "step": 11800
+    },
+    {
+      "epoch": 17.76,
+      "grad_norm": 5.273082733154297,
+      "learning_rate": 8.224060150375941e-06,
+      "loss": 0.3479,
+      "step": 11810
+    },
+    {
+      "epoch": 17.77,
+      "grad_norm": 4.500287055969238,
+      "learning_rate": 8.222556390977445e-06,
+      "loss": 0.3813,
+      "step": 11820
+    },
+    {
+      "epoch": 17.79,
+      "grad_norm": 7.676726341247559,
+      "learning_rate": 8.221052631578948e-06,
+      "loss": 0.3973,
+      "step": 11830
+    },
+    {
+      "epoch": 17.8,
+      "grad_norm": 6.1550211906433105,
+      "learning_rate": 8.219548872180452e-06,
+      "loss": 0.3209,
+      "step": 11840
+    },
+    {
+      "epoch": 17.82,
+      "grad_norm": 3.796853542327881,
+      "learning_rate": 8.218045112781955e-06,
+      "loss": 0.3241,
+      "step": 11850
+    },
+    {
+      "epoch": 17.83,
+      "grad_norm": 7.188779354095459,
+      "learning_rate": 8.216541353383459e-06,
+      "loss": 0.3923,
+      "step": 11860
+    },
+    {
+      "epoch": 17.85,
+      "grad_norm": 5.088048934936523,
+      "learning_rate": 8.215037593984964e-06,
+      "loss": 0.3528,
+      "step": 11870
+    },
+    {
+      "epoch": 17.86,
+      "grad_norm": 6.49263334274292,
+      "learning_rate": 8.213533834586466e-06,
+      "loss": 0.3724,
+      "step": 11880
+    },
+    {
+      "epoch": 17.88,
+      "grad_norm": 8.370095252990723,
+      "learning_rate": 8.212030075187971e-06,
+      "loss": 0.3697,
+      "step": 11890
+    },
+    {
+      "epoch": 17.89,
+      "grad_norm": 7.578341007232666,
+      "learning_rate": 8.210526315789475e-06,
+      "loss": 0.3317,
+      "step": 11900
+    },
+    {
+      "epoch": 17.91,
+      "grad_norm": 5.1709723472595215,
+      "learning_rate": 8.209022556390978e-06,
+      "loss": 0.3304,
+      "step": 11910
+    },
+    {
+      "epoch": 17.92,
+      "grad_norm": 4.586398124694824,
+      "learning_rate": 8.207518796992482e-06,
+      "loss": 0.3942,
+      "step": 11920
+    },
+    {
+      "epoch": 17.94,
+      "grad_norm": 3.6240298748016357,
+      "learning_rate": 8.206015037593985e-06,
+      "loss": 0.2839,
+      "step": 11930
+    },
+    {
+      "epoch": 17.95,
+      "grad_norm": 7.672499179840088,
+      "learning_rate": 8.204511278195489e-06,
+      "loss": 0.3539,
+      "step": 11940
+    },
+    {
+      "epoch": 17.97,
+      "grad_norm": 5.807362079620361,
+      "learning_rate": 8.203007518796992e-06,
+      "loss": 0.3006,
+      "step": 11950
+    },
+    {
+      "epoch": 17.98,
+      "grad_norm": 5.875560283660889,
+      "learning_rate": 8.201503759398498e-06,
+      "loss": 0.3857,
+      "step": 11960
+    },
+    {
+      "epoch": 18.0,
+      "grad_norm": 47.7349967956543,
+      "learning_rate": 8.2e-06,
+      "loss": 0.3985,
+      "step": 11970
+    },
+    {
+      "epoch": 18.0,
+      "eval_accuracy": 0.9356,
+      "eval_loss": 0.25322866439819336,
+      "eval_runtime": 84.6682,
+      "eval_samples_per_second": 118.108,
+      "eval_steps_per_second": 0.472,
+      "step": 11970
+    },
+    {
+      "epoch": 18.02,
+      "grad_norm": 5.787768363952637,
+      "learning_rate": 8.198496240601505e-06,
+      "loss": 0.4037,
+      "step": 11980
+    },
+    {
+      "epoch": 18.03,
+      "grad_norm": 9.899946212768555,
+      "learning_rate": 8.196992481203008e-06,
+      "loss": 0.3305,
+      "step": 11990
+    },
+    {
+      "epoch": 18.05,
+      "grad_norm": 7.430878639221191,
+      "learning_rate": 8.195488721804512e-06,
+      "loss": 0.3072,
+      "step": 12000
+    },
+    {
+      "epoch": 18.06,
+      "grad_norm": 4.635775566101074,
+      "learning_rate": 8.193984962406015e-06,
+      "loss": 0.3824,
+      "step": 12010
+    },
+    {
+      "epoch": 18.08,
+      "grad_norm": 11.533434867858887,
+      "learning_rate": 8.19248120300752e-06,
+      "loss": 0.3494,
+      "step": 12020
+    },
+    {
+      "epoch": 18.09,
+      "grad_norm": 9.492911338806152,
+      "learning_rate": 8.190977443609022e-06,
+      "loss": 0.2719,
+      "step": 12030
+    },
+    {
+      "epoch": 18.11,
+      "grad_norm": 4.291653156280518,
+      "learning_rate": 8.189473684210527e-06,
+      "loss": 0.3766,
+      "step": 12040
+    },
+    {
+      "epoch": 18.12,
+      "grad_norm": 6.437026023864746,
+      "learning_rate": 8.187969924812031e-06,
+      "loss": 0.3757,
+      "step": 12050
+    },
+    {
+      "epoch": 18.14,
+      "grad_norm": 6.730251789093018,
+      "learning_rate": 8.186466165413534e-06,
+      "loss": 0.3426,
+      "step": 12060
+    },
+    {
+      "epoch": 18.15,
+      "grad_norm": 4.403579235076904,
+      "learning_rate": 8.184962406015038e-06,
+      "loss": 0.4402,
+      "step": 12070
+    },
+    {
+      "epoch": 18.17,
+      "grad_norm": 4.053240776062012,
+      "learning_rate": 8.183458646616542e-06,
+      "loss": 0.4017,
+      "step": 12080
+    },
+    {
+      "epoch": 18.18,
+      "grad_norm": 9.355215072631836,
+      "learning_rate": 8.181954887218045e-06,
+      "loss": 0.3034,
+      "step": 12090
+    },
+    {
+      "epoch": 18.2,
+      "grad_norm": 6.016571521759033,
+      "learning_rate": 8.18045112781955e-06,
+      "loss": 0.2966,
+      "step": 12100
+    },
+    {
+      "epoch": 18.21,
+      "grad_norm": 3.7155065536499023,
+      "learning_rate": 8.178947368421054e-06,
+      "loss": 0.325,
+      "step": 12110
+    },
+    {
+      "epoch": 18.23,
+      "grad_norm": 4.735385417938232,
+      "learning_rate": 8.177443609022557e-06,
+      "loss": 0.3916,
+      "step": 12120
+    },
+    {
+      "epoch": 18.24,
+      "grad_norm": 5.728586673736572,
+      "learning_rate": 8.17593984962406e-06,
+      "loss": 0.3307,
+      "step": 12130
+    },
+    {
+      "epoch": 18.26,
+      "grad_norm": 3.7535057067871094,
+      "learning_rate": 8.174436090225564e-06,
+      "loss": 0.2576,
+      "step": 12140
+    },
+    {
+      "epoch": 18.27,
+      "grad_norm": 7.422114849090576,
+      "learning_rate": 8.172932330827068e-06,
+      "loss": 0.405,
+      "step": 12150
+    },
+    {
+      "epoch": 18.29,
+      "grad_norm": 7.051729679107666,
+      "learning_rate": 8.171428571428573e-06,
+      "loss": 0.3504,
+      "step": 12160
+    },
+    {
+      "epoch": 18.3,
+      "grad_norm": 5.593104362487793,
+      "learning_rate": 8.169924812030075e-06,
+      "loss": 0.3071,
+      "step": 12170
+    },
+    {
+      "epoch": 18.32,
+      "grad_norm": 6.640134811401367,
+      "learning_rate": 8.16842105263158e-06,
+      "loss": 0.38,
+      "step": 12180
+    },
+    {
+      "epoch": 18.33,
+      "grad_norm": 7.922089099884033,
+      "learning_rate": 8.166917293233084e-06,
+      "loss": 0.3383,
+      "step": 12190
+    },
+    {
+      "epoch": 18.35,
+      "grad_norm": 6.431313514709473,
+      "learning_rate": 8.165413533834587e-06,
+      "loss": 0.2918,
+      "step": 12200
+    },
+    {
+      "epoch": 18.36,
+      "grad_norm": 5.122470855712891,
+      "learning_rate": 8.16390977443609e-06,
+      "loss": 0.3129,
+      "step": 12210
+    },
+    {
+      "epoch": 18.38,
+      "grad_norm": 5.576056957244873,
+      "learning_rate": 8.162406015037594e-06,
+      "loss": 0.3316,
+      "step": 12220
+    },
+    {
+      "epoch": 18.39,
+      "grad_norm": 5.235702037811279,
+      "learning_rate": 8.160902255639098e-06,
+      "loss": 0.2955,
+      "step": 12230
+    },
+    {
+      "epoch": 18.41,
+      "grad_norm": 5.954382419586182,
+      "learning_rate": 8.159398496240603e-06,
+      "loss": 0.3256,
+      "step": 12240
+    },
+    {
+      "epoch": 18.42,
+      "grad_norm": 5.545419216156006,
+      "learning_rate": 8.157894736842106e-06,
+      "loss": 0.358,
+      "step": 12250
+    },
+    {
+      "epoch": 18.44,
+      "grad_norm": 12.001044273376465,
+      "learning_rate": 8.15639097744361e-06,
+      "loss": 0.3185,
+      "step": 12260
+    },
+    {
+      "epoch": 18.45,
+      "grad_norm": 10.437068939208984,
+      "learning_rate": 8.154887218045113e-06,
+      "loss": 0.3234,
+      "step": 12270
+    },
+    {
+      "epoch": 18.47,
+      "grad_norm": 5.405700206756592,
+      "learning_rate": 8.153383458646617e-06,
+      "loss": 0.251,
+      "step": 12280
+    },
+    {
+      "epoch": 18.48,
+      "grad_norm": 6.865713119506836,
+      "learning_rate": 8.15187969924812e-06,
+      "loss": 0.4126,
+      "step": 12290
+    },
+    {
+      "epoch": 18.5,
+      "grad_norm": 6.4221038818359375,
+      "learning_rate": 8.150375939849626e-06,
+      "loss": 0.3338,
+      "step": 12300
+    },
+    {
+      "epoch": 18.51,
+      "grad_norm": 5.260381698608398,
+      "learning_rate": 8.148872180451128e-06,
+      "loss": 0.3482,
+      "step": 12310
+    },
+    {
+      "epoch": 18.53,
+      "grad_norm": 8.89275074005127,
+      "learning_rate": 8.147368421052633e-06,
+      "loss": 0.3684,
+      "step": 12320
+    },
+    {
+      "epoch": 18.54,
+      "grad_norm": 4.887394428253174,
+      "learning_rate": 8.145864661654136e-06,
+      "loss": 0.3673,
+      "step": 12330
+    },
+    {
+      "epoch": 18.56,
+      "grad_norm": 3.8077456951141357,
+      "learning_rate": 8.14436090225564e-06,
+      "loss": 0.3478,
+      "step": 12340
+    },
+    {
+      "epoch": 18.57,
+      "grad_norm": 5.575742721557617,
+      "learning_rate": 8.142857142857143e-06,
+      "loss": 0.332,
+      "step": 12350
+    },
+    {
+      "epoch": 18.59,
+      "grad_norm": 6.4698333740234375,
+      "learning_rate": 8.141353383458649e-06,
+      "loss": 0.4422,
+      "step": 12360
+    },
+    {
+      "epoch": 18.6,
+      "grad_norm": 5.90009880065918,
+      "learning_rate": 8.13984962406015e-06,
+      "loss": 0.4025,
+      "step": 12370
+    },
+    {
+      "epoch": 18.62,
+      "grad_norm": 7.366026401519775,
+      "learning_rate": 8.138345864661656e-06,
+      "loss": 0.3364,
+      "step": 12380
+    },
+    {
+      "epoch": 18.63,
+      "grad_norm": 5.042256832122803,
+      "learning_rate": 8.136842105263159e-06,
+      "loss": 0.2686,
+      "step": 12390
+    },
+    {
+      "epoch": 18.65,
+      "grad_norm": 5.483407020568848,
+      "learning_rate": 8.135338345864663e-06,
+      "loss": 0.2765,
+      "step": 12400
+    },
+    {
+      "epoch": 18.66,
+      "grad_norm": 8.821917533874512,
+      "learning_rate": 8.133834586466166e-06,
+      "loss": 0.3899,
+      "step": 12410
+    },
+    {
+      "epoch": 18.68,
+      "grad_norm": 5.6711297035217285,
+      "learning_rate": 8.13233082706767e-06,
+      "loss": 0.4293,
+      "step": 12420
+    },
+    {
+      "epoch": 18.69,
+      "grad_norm": 6.44101619720459,
+      "learning_rate": 8.130827067669173e-06,
+      "loss": 0.3191,
+      "step": 12430
+    },
+    {
+      "epoch": 18.71,
+      "grad_norm": 4.042926788330078,
+      "learning_rate": 8.129323308270677e-06,
+      "loss": 0.3414,
+      "step": 12440
+    },
+    {
+      "epoch": 18.72,
+      "grad_norm": 7.297868728637695,
+      "learning_rate": 8.127819548872182e-06,
+      "loss": 0.2944,
+      "step": 12450
+    },
+    {
+      "epoch": 18.74,
+      "grad_norm": 7.283233642578125,
+      "learning_rate": 8.126315789473684e-06,
+      "loss": 0.3262,
+      "step": 12460
+    },
+    {
+      "epoch": 18.75,
+      "grad_norm": 7.256216526031494,
+      "learning_rate": 8.124812030075189e-06,
+      "loss": 0.3714,
+      "step": 12470
+    },
+    {
+      "epoch": 18.77,
+      "grad_norm": 5.120934009552002,
+      "learning_rate": 8.123308270676692e-06,
+      "loss": 0.3963,
+      "step": 12480
+    },
+    {
+      "epoch": 18.78,
+      "grad_norm": 5.541153907775879,
+      "learning_rate": 8.121804511278196e-06,
+      "loss": 0.3602,
+      "step": 12490
+    },
+    {
+      "epoch": 18.8,
+      "grad_norm": 5.731190204620361,
+      "learning_rate": 8.1203007518797e-06,
+      "loss": 0.4258,
+      "step": 12500
+    },
+    {
+      "epoch": 18.81,
+      "grad_norm": 5.054348468780518,
+      "learning_rate": 8.118796992481203e-06,
+      "loss": 0.2603,
+      "step": 12510
+    },
+    {
+      "epoch": 18.83,
+      "grad_norm": 4.947210311889648,
+      "learning_rate": 8.117293233082707e-06,
+      "loss": 0.332,
+      "step": 12520
+    },
+    {
+      "epoch": 18.84,
+      "grad_norm": 7.019644737243652,
+      "learning_rate": 8.115789473684212e-06,
+      "loss": 0.3752,
+      "step": 12530
+    },
+    {
+      "epoch": 18.86,
+      "grad_norm": 6.0113325119018555,
+      "learning_rate": 8.114285714285715e-06,
+      "loss": 0.2853,
+      "step": 12540
+    },
+    {
+      "epoch": 18.87,
+      "grad_norm": 10.12833309173584,
+      "learning_rate": 8.112781954887219e-06,
+      "loss": 0.2721,
+      "step": 12550
+    },
+    {
+      "epoch": 18.89,
+      "grad_norm": 3.987274408340454,
+      "learning_rate": 8.111278195488722e-06,
+      "loss": 0.3351,
+      "step": 12560
+    },
+    {
+      "epoch": 18.9,
+      "grad_norm": 6.421701431274414,
+      "learning_rate": 8.109774436090226e-06,
+      "loss": 0.3202,
+      "step": 12570
+    },
+    {
+      "epoch": 18.92,
+      "grad_norm": 6.776229381561279,
+      "learning_rate": 8.10827067669173e-06,
+      "loss": 0.3123,
+      "step": 12580
+    },
+    {
+      "epoch": 18.93,
+      "grad_norm": 8.584750175476074,
+      "learning_rate": 8.106766917293235e-06,
+      "loss": 0.4226,
+      "step": 12590
+    },
+    {
+      "epoch": 18.95,
+      "grad_norm": 4.636366844177246,
+      "learning_rate": 8.105263157894736e-06,
+      "loss": 0.4647,
+      "step": 12600
+    },
+    {
+      "epoch": 18.96,
+      "grad_norm": 5.144723415374756,
+      "learning_rate": 8.103759398496242e-06,
+      "loss": 0.287,
+      "step": 12610
+    },
+    {
+      "epoch": 18.98,
+      "grad_norm": 7.949281692504883,
+      "learning_rate": 8.102255639097745e-06,
+      "loss": 0.3361,
+      "step": 12620
+    },
+    {
+      "epoch": 18.99,
+      "grad_norm": 6.716331958770752,
+      "learning_rate": 8.100751879699249e-06,
+      "loss": 0.3213,
+      "step": 12630
+    },
+    {
+      "epoch": 19.0,
+      "eval_accuracy": 0.9321,
+      "eval_loss": 0.27284207940101624,
+      "eval_runtime": 84.4705,
+      "eval_samples_per_second": 118.384,
+      "eval_steps_per_second": 0.474,
+      "step": 12635
+    },
+    {
+      "epoch": 19.01,
+      "grad_norm": 5.840856075286865,
+      "learning_rate": 8.099248120300752e-06,
+      "loss": 0.3154,
+      "step": 12640
+    },
+    {
+      "epoch": 19.02,
+      "grad_norm": 6.531993389129639,
+      "learning_rate": 8.097744360902257e-06,
+      "loss": 0.3244,
+      "step": 12650
+    },
+    {
+      "epoch": 19.04,
+      "grad_norm": 8.1339111328125,
+      "learning_rate": 8.09624060150376e-06,
+      "loss": 0.3372,
+      "step": 12660
+    },
+    {
+      "epoch": 19.05,
+      "grad_norm": 5.9135518074035645,
+      "learning_rate": 8.094736842105264e-06,
+      "loss": 0.3514,
+      "step": 12670
+    },
+    {
+      "epoch": 19.07,
+      "grad_norm": 4.900998592376709,
+      "learning_rate": 8.093233082706768e-06,
+      "loss": 0.3248,
+      "step": 12680
+    },
+    {
+      "epoch": 19.08,
+      "grad_norm": 6.988156795501709,
+      "learning_rate": 8.091729323308271e-06,
+      "loss": 0.3665,
+      "step": 12690
+    },
+    {
+      "epoch": 19.1,
+      "grad_norm": 6.229936599731445,
+      "learning_rate": 8.090225563909775e-06,
+      "loss": 0.2717,
+      "step": 12700
+    },
+    {
+      "epoch": 19.11,
+      "grad_norm": 7.66952657699585,
+      "learning_rate": 8.088721804511278e-06,
+      "loss": 0.4116,
+      "step": 12710
+    },
+    {
+      "epoch": 19.13,
+      "grad_norm": 5.332541465759277,
+      "learning_rate": 8.087218045112782e-06,
+      "loss": 0.2748,
+      "step": 12720
+    },
+    {
+      "epoch": 19.14,
+      "grad_norm": 6.3354268074035645,
+      "learning_rate": 8.085714285714287e-06,
+      "loss": 0.3508,
+      "step": 12730
+    },
+    {
+      "epoch": 19.16,
+      "grad_norm": 2.598149538040161,
+      "learning_rate": 8.08421052631579e-06,
+      "loss": 0.2887,
+      "step": 12740
+    },
+    {
+      "epoch": 19.17,
+      "grad_norm": 3.951981544494629,
+      "learning_rate": 8.082706766917294e-06,
+      "loss": 0.3201,
+      "step": 12750
+    },
+    {
+      "epoch": 19.19,
+      "grad_norm": 6.856266021728516,
+      "learning_rate": 8.081203007518798e-06,
+      "loss": 0.3015,
+      "step": 12760
+    },
+    {
+      "epoch": 19.2,
+      "grad_norm": 8.871770858764648,
+      "learning_rate": 8.079699248120301e-06,
+      "loss": 0.4151,
+      "step": 12770
+    },
+    {
+      "epoch": 19.22,
+      "grad_norm": 7.119663238525391,
+      "learning_rate": 8.078195488721805e-06,
+      "loss": 0.2968,
+      "step": 12780
+    },
+    {
+      "epoch": 19.23,
+      "grad_norm": 5.254119396209717,
+      "learning_rate": 8.07669172932331e-06,
+      "loss": 0.2823,
+      "step": 12790
+    },
+    {
+      "epoch": 19.25,
+      "grad_norm": 8.958704948425293,
+      "learning_rate": 8.075187969924812e-06,
+      "loss": 0.3453,
+      "step": 12800
+    },
+    {
+      "epoch": 19.26,
+      "grad_norm": 8.45573616027832,
+      "learning_rate": 8.073684210526317e-06,
+      "loss": 0.3827,
+      "step": 12810
+    },
+    {
+      "epoch": 19.28,
+      "grad_norm": 3.3245856761932373,
+      "learning_rate": 8.07218045112782e-06,
+      "loss": 0.3515,
+      "step": 12820
+    },
+    {
+      "epoch": 19.29,
+      "grad_norm": 4.837510585784912,
+      "learning_rate": 8.070676691729324e-06,
+      "loss": 0.3397,
+      "step": 12830
+    },
+    {
+      "epoch": 19.31,
+      "grad_norm": 3.3870489597320557,
+      "learning_rate": 8.069172932330828e-06,
+      "loss": 0.3151,
+      "step": 12840
+    },
+    {
+      "epoch": 19.32,
+      "grad_norm": 6.1325201988220215,
+      "learning_rate": 8.067669172932333e-06,
+      "loss": 0.3491,
+      "step": 12850
+    },
+    {
+      "epoch": 19.34,
+      "grad_norm": 6.618387699127197,
+      "learning_rate": 8.066165413533835e-06,
+      "loss": 0.3685,
+      "step": 12860
+    },
+    {
+      "epoch": 19.35,
+      "grad_norm": 6.2998223304748535,
+      "learning_rate": 8.06466165413534e-06,
+      "loss": 0.4341,
+      "step": 12870
+    },
+    {
+      "epoch": 19.37,
+      "grad_norm": 5.405755043029785,
+      "learning_rate": 8.063157894736843e-06,
+      "loss": 0.3056,
+      "step": 12880
+    },
+    {
+      "epoch": 19.38,
+      "grad_norm": 5.74697732925415,
+      "learning_rate": 8.061654135338347e-06,
+      "loss": 0.4015,
+      "step": 12890
+    },
+    {
+      "epoch": 19.4,
+      "grad_norm": 4.222403526306152,
+      "learning_rate": 8.06015037593985e-06,
+      "loss": 0.2818,
+      "step": 12900
+    },
+    {
+      "epoch": 19.41,
+      "grad_norm": 11.104449272155762,
+      "learning_rate": 8.058646616541354e-06,
+      "loss": 0.3528,
+      "step": 12910
+    },
+    {
+      "epoch": 19.43,
+      "grad_norm": 5.580349445343018,
+      "learning_rate": 8.057142857142857e-06,
+      "loss": 0.3511,
+      "step": 12920
+    },
+    {
+      "epoch": 19.44,
+      "grad_norm": 4.909841060638428,
+      "learning_rate": 8.055639097744361e-06,
+      "loss": 0.2696,
+      "step": 12930
+    },
+    {
+      "epoch": 19.46,
+      "grad_norm": 4.32835054397583,
+      "learning_rate": 8.054135338345866e-06,
+      "loss": 0.2714,
+      "step": 12940
+    },
+    {
+      "epoch": 19.47,
+      "grad_norm": 4.742302894592285,
+      "learning_rate": 8.052631578947368e-06,
+      "loss": 0.3427,
+      "step": 12950
+    },
+    {
+      "epoch": 19.49,
+      "grad_norm": 6.302316665649414,
+      "learning_rate": 8.051127819548873e-06,
+      "loss": 0.406,
+      "step": 12960
+    },
+    {
+      "epoch": 19.5,
+      "grad_norm": 6.898422718048096,
+      "learning_rate": 8.049624060150377e-06,
+      "loss": 0.3048,
+      "step": 12970
+    },
+    {
+      "epoch": 19.52,
+      "grad_norm": 9.071293830871582,
+      "learning_rate": 8.04812030075188e-06,
+      "loss": 0.3603,
+      "step": 12980
+    },
+    {
+      "epoch": 19.53,
+      "grad_norm": 5.120814323425293,
+      "learning_rate": 8.046616541353384e-06,
+      "loss": 0.3555,
+      "step": 12990
+    },
+    {
+      "epoch": 19.55,
+      "grad_norm": 8.293437957763672,
+      "learning_rate": 8.045112781954887e-06,
+      "loss": 0.3061,
+      "step": 13000
+    },
+    {
+      "epoch": 19.56,
+      "grad_norm": 6.051889419555664,
+      "learning_rate": 8.04360902255639e-06,
+      "loss": 0.3308,
+      "step": 13010
+    },
+    {
+      "epoch": 19.58,
+      "grad_norm": 6.6684889793396,
+      "learning_rate": 8.042105263157896e-06,
+      "loss": 0.3497,
+      "step": 13020
+    },
+    {
+      "epoch": 19.59,
+      "grad_norm": 6.767400741577148,
+      "learning_rate": 8.0406015037594e-06,
+      "loss": 0.3037,
+      "step": 13030
+    },
+    {
+      "epoch": 19.61,
+      "grad_norm": 6.514301300048828,
+      "learning_rate": 8.039097744360903e-06,
+      "loss": 0.266,
+      "step": 13040
+    },
+    {
+      "epoch": 19.62,
+      "grad_norm": 4.410576820373535,
+      "learning_rate": 8.037593984962407e-06,
+      "loss": 0.3086,
+      "step": 13050
+    },
+    {
+      "epoch": 19.64,
+      "grad_norm": 8.338872909545898,
+      "learning_rate": 8.03609022556391e-06,
+      "loss": 0.3505,
+      "step": 13060
+    },
+    {
+      "epoch": 19.65,
+      "grad_norm": 7.640175819396973,
+      "learning_rate": 8.034586466165414e-06,
+      "loss": 0.3573,
+      "step": 13070
+    },
+    {
+      "epoch": 19.67,
+      "grad_norm": 5.462586402893066,
+      "learning_rate": 8.033082706766919e-06,
+      "loss": 0.3053,
+      "step": 13080
+    },
+    {
+      "epoch": 19.68,
+      "grad_norm": 4.294394016265869,
+      "learning_rate": 8.03157894736842e-06,
+      "loss": 0.3741,
+      "step": 13090
+    },
+    {
+      "epoch": 19.7,
+      "grad_norm": 4.442080497741699,
+      "learning_rate": 8.030075187969926e-06,
+      "loss": 0.2855,
+      "step": 13100
+    },
+    {
+      "epoch": 19.71,
+      "grad_norm": 3.8873209953308105,
+      "learning_rate": 8.02857142857143e-06,
+      "loss": 0.2914,
+      "step": 13110
+    },
+    {
+      "epoch": 19.73,
+      "grad_norm": 9.044189453125,
+      "learning_rate": 8.027067669172933e-06,
+      "loss": 0.3345,
+      "step": 13120
+    },
+    {
+      "epoch": 19.74,
+      "grad_norm": 9.596213340759277,
+      "learning_rate": 8.025563909774436e-06,
+      "loss": 0.3249,
+      "step": 13130
+    },
+    {
+      "epoch": 19.76,
+      "grad_norm": 6.889251708984375,
+      "learning_rate": 8.02406015037594e-06,
+      "loss": 0.3135,
+      "step": 13140
+    },
+    {
+      "epoch": 19.77,
+      "grad_norm": 6.394194602966309,
+      "learning_rate": 8.022556390977443e-06,
+      "loss": 0.2785,
+      "step": 13150
+    },
+    {
+      "epoch": 19.79,
+      "grad_norm": 8.294524192810059,
+      "learning_rate": 8.021052631578949e-06,
+      "loss": 0.3854,
+      "step": 13160
+    },
+    {
+      "epoch": 19.8,
+      "grad_norm": 9.557373046875,
+      "learning_rate": 8.019548872180452e-06,
+      "loss": 0.3031,
+      "step": 13170
+    },
+    {
+      "epoch": 19.82,
+      "grad_norm": 5.370851516723633,
+      "learning_rate": 8.018045112781956e-06,
+      "loss": 0.3836,
+      "step": 13180
+    },
+    {
+      "epoch": 19.83,
+      "grad_norm": 7.893476963043213,
+      "learning_rate": 8.01654135338346e-06,
+      "loss": 0.4143,
+      "step": 13190
+    },
+    {
+      "epoch": 19.85,
+      "grad_norm": 6.872874736785889,
+      "learning_rate": 8.015037593984963e-06,
+      "loss": 0.3131,
+      "step": 13200
+    },
+    {
+      "epoch": 19.86,
+      "grad_norm": 8.026068687438965,
+      "learning_rate": 8.013533834586466e-06,
+      "loss": 0.3558,
+      "step": 13210
+    },
+    {
+      "epoch": 19.88,
+      "grad_norm": 5.648526668548584,
+      "learning_rate": 8.012030075187971e-06,
+      "loss": 0.3377,
+      "step": 13220
+    },
+    {
+      "epoch": 19.89,
+      "grad_norm": 9.711199760437012,
+      "learning_rate": 8.010526315789473e-06,
+      "loss": 0.3421,
+      "step": 13230
+    },
+    {
+      "epoch": 19.91,
+      "grad_norm": 5.787407875061035,
+      "learning_rate": 8.009022556390979e-06,
+      "loss": 0.3919,
+      "step": 13240
+    },
+    {
+      "epoch": 19.92,
+      "grad_norm": 6.500585079193115,
+      "learning_rate": 8.007518796992482e-06,
+      "loss": 0.3383,
+      "step": 13250
+    },
+    {
+      "epoch": 19.94,
+      "grad_norm": 6.25896692276001,
+      "learning_rate": 8.006015037593986e-06,
+      "loss": 0.3031,
+      "step": 13260
+    },
+    {
+      "epoch": 19.95,
+      "grad_norm": 4.565229415893555,
+      "learning_rate": 8.004511278195489e-06,
+      "loss": 0.379,
+      "step": 13270
+    },
+    {
+      "epoch": 19.97,
+      "grad_norm": 5.765014171600342,
+      "learning_rate": 8.003007518796994e-06,
+      "loss": 0.2949,
+      "step": 13280
+    },
+    {
+      "epoch": 19.98,
+      "grad_norm": 5.820449352264404,
+      "learning_rate": 8.001503759398496e-06,
+      "loss": 0.2912,
+      "step": 13290
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 0.44211289286613464,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.3046,
+      "step": 13300
+    },
+    {
+      "epoch": 20.0,
+      "eval_accuracy": 0.9334,
+      "eval_loss": 0.27018824219703674,
+      "eval_runtime": 84.9396,
+      "eval_samples_per_second": 117.731,
+      "eval_steps_per_second": 0.471,
+      "step": 13300
+    },
+    {
+      "epoch": 20.02,
+      "grad_norm": 6.310322284698486,
+      "learning_rate": 7.998496240601505e-06,
+      "loss": 0.3525,
+      "step": 13310
+    },
+    {
+      "epoch": 20.03,
+      "grad_norm": 4.756250381469727,
+      "learning_rate": 7.996992481203008e-06,
+      "loss": 0.2493,
+      "step": 13320
+    },
+    {
+      "epoch": 20.05,
+      "grad_norm": 6.389369487762451,
+      "learning_rate": 7.995488721804512e-06,
+      "loss": 0.3425,
+      "step": 13330
+    },
+    {
+      "epoch": 20.06,
+      "grad_norm": 6.042266368865967,
+      "learning_rate": 7.993984962406015e-06,
+      "loss": 0.2894,
+      "step": 13340
+    },
+    {
+      "epoch": 20.08,
+      "grad_norm": 5.721156597137451,
+      "learning_rate": 7.992481203007519e-06,
+      "loss": 0.3251,
+      "step": 13350
+    },
+    {
+      "epoch": 20.09,
+      "grad_norm": 8.065733909606934,
+      "learning_rate": 7.990977443609024e-06,
+      "loss": 0.2898,
+      "step": 13360
+    },
+    {
+      "epoch": 20.11,
+      "grad_norm": 5.018655300140381,
+      "learning_rate": 7.989473684210528e-06,
+      "loss": 0.3525,
+      "step": 13370
+    },
+    {
+      "epoch": 20.12,
+      "grad_norm": 7.109757900238037,
+      "learning_rate": 7.987969924812031e-06,
+      "loss": 0.2977,
+      "step": 13380
+    },
+    {
+      "epoch": 20.14,
+      "grad_norm": 5.0374274253845215,
+      "learning_rate": 7.986466165413535e-06,
+      "loss": 0.2893,
+      "step": 13390
+    },
+    {
+      "epoch": 20.15,
+      "grad_norm": 6.046887397766113,
+      "learning_rate": 7.984962406015038e-06,
+      "loss": 0.2753,
+      "step": 13400
+    },
+    {
+      "epoch": 20.17,
+      "grad_norm": 4.841386795043945,
+      "learning_rate": 7.983458646616542e-06,
+      "loss": 0.3502,
+      "step": 13410
+    },
+    {
+      "epoch": 20.18,
+      "grad_norm": 6.6681036949157715,
+      "learning_rate": 7.981954887218045e-06,
+      "loss": 0.3064,
+      "step": 13420
+    },
+    {
+      "epoch": 20.2,
+      "grad_norm": 7.210595607757568,
+      "learning_rate": 7.980451127819549e-06,
+      "loss": 0.3887,
+      "step": 13430
+    },
+    {
+      "epoch": 20.21,
+      "grad_norm": 7.28894567489624,
+      "learning_rate": 7.978947368421052e-06,
+      "loss": 0.3476,
+      "step": 13440
+    },
+    {
+      "epoch": 20.23,
+      "grad_norm": 14.942082405090332,
+      "learning_rate": 7.977443609022558e-06,
+      "loss": 0.3151,
+      "step": 13450
+    },
+    {
+      "epoch": 20.24,
+      "grad_norm": 6.0876054763793945,
+      "learning_rate": 7.975939849624061e-06,
+      "loss": 0.3542,
+      "step": 13460
+    },
+    {
+      "epoch": 20.26,
+      "grad_norm": 8.294320106506348,
+      "learning_rate": 7.974436090225565e-06,
+      "loss": 0.3312,
+      "step": 13470
+    },
+    {
+      "epoch": 20.27,
+      "grad_norm": 5.090022087097168,
+      "learning_rate": 7.972932330827068e-06,
+      "loss": 0.323,
+      "step": 13480
+    },
+    {
+      "epoch": 20.29,
+      "grad_norm": 2.9050402641296387,
+      "learning_rate": 7.971428571428572e-06,
+      "loss": 0.2753,
+      "step": 13490
+    },
+    {
+      "epoch": 20.3,
+      "grad_norm": 7.342869758605957,
+      "learning_rate": 7.969924812030075e-06,
+      "loss": 0.2639,
+      "step": 13500
+    },
+    {
+      "epoch": 20.32,
+      "grad_norm": 5.894117832183838,
+      "learning_rate": 7.96842105263158e-06,
+      "loss": 0.2893,
+      "step": 13510
+    },
+    {
+      "epoch": 20.33,
+      "grad_norm": 5.4692792892456055,
+      "learning_rate": 7.966917293233082e-06,
+      "loss": 0.3384,
+      "step": 13520
+    },
+    {
+      "epoch": 20.35,
+      "grad_norm": 7.577072620391846,
+      "learning_rate": 7.965413533834587e-06,
+      "loss": 0.3865,
+      "step": 13530
+    },
+    {
+      "epoch": 20.36,
+      "grad_norm": 22.644380569458008,
+      "learning_rate": 7.963909774436091e-06,
+      "loss": 0.3332,
+      "step": 13540
+    },
+    {
+      "epoch": 20.38,
+      "grad_norm": 6.491385459899902,
+      "learning_rate": 7.962406015037594e-06,
+      "loss": 0.3662,
+      "step": 13550
+    },
+    {
+      "epoch": 20.39,
+      "grad_norm": 5.722530841827393,
+      "learning_rate": 7.960902255639098e-06,
+      "loss": 0.3452,
+      "step": 13560
+    },
+    {
+      "epoch": 20.41,
+      "grad_norm": 5.905157566070557,
+      "learning_rate": 7.959398496240603e-06,
+      "loss": 0.3424,
+      "step": 13570
+    },
+    {
+      "epoch": 20.42,
+      "grad_norm": 8.227418899536133,
+      "learning_rate": 7.957894736842105e-06,
+      "loss": 0.3635,
+      "step": 13580
+    },
+    {
+      "epoch": 20.44,
+      "grad_norm": 3.507075309753418,
+      "learning_rate": 7.95639097744361e-06,
+      "loss": 0.2959,
+      "step": 13590
+    },
+    {
+      "epoch": 20.45,
+      "grad_norm": 10.020979881286621,
+      "learning_rate": 7.954887218045114e-06,
+      "loss": 0.325,
+      "step": 13600
+    },
+    {
+      "epoch": 20.47,
+      "grad_norm": 6.452096939086914,
+      "learning_rate": 7.953383458646617e-06,
+      "loss": 0.2817,
+      "step": 13610
+    },
+    {
+      "epoch": 20.48,
+      "grad_norm": 5.931590557098389,
+      "learning_rate": 7.95187969924812e-06,
+      "loss": 0.3138,
+      "step": 13620
+    },
+    {
+      "epoch": 20.5,
+      "grad_norm": 11.746089935302734,
+      "learning_rate": 7.950375939849624e-06,
+      "loss": 0.3263,
+      "step": 13630
+    },
+    {
+      "epoch": 20.51,
+      "grad_norm": 5.600671291351318,
+      "learning_rate": 7.948872180451128e-06,
+      "loss": 0.2635,
+      "step": 13640
+    },
+    {
+      "epoch": 20.53,
+      "grad_norm": 4.570649147033691,
+      "learning_rate": 7.947368421052633e-06,
+      "loss": 0.3372,
+      "step": 13650
+    },
+    {
+      "epoch": 20.54,
+      "grad_norm": 10.411113739013672,
+      "learning_rate": 7.945864661654136e-06,
+      "loss": 0.3403,
+      "step": 13660
+    },
+    {
+      "epoch": 20.56,
+      "grad_norm": 7.8483710289001465,
+      "learning_rate": 7.94436090225564e-06,
+      "loss": 0.4401,
+      "step": 13670
+    },
+    {
+      "epoch": 20.57,
+      "grad_norm": 6.16978645324707,
+      "learning_rate": 7.942857142857144e-06,
+      "loss": 0.2905,
+      "step": 13680
+    },
+    {
+      "epoch": 20.59,
+      "grad_norm": 5.427683353424072,
+      "learning_rate": 7.941353383458647e-06,
+      "loss": 0.3853,
+      "step": 13690
+    },
+    {
+      "epoch": 20.6,
+      "grad_norm": 10.262079238891602,
+      "learning_rate": 7.93984962406015e-06,
+      "loss": 0.3126,
+      "step": 13700
+    },
+    {
+      "epoch": 20.62,
+      "grad_norm": 2.678656578063965,
+      "learning_rate": 7.938345864661656e-06,
+      "loss": 0.2864,
+      "step": 13710
+    },
+    {
+      "epoch": 20.63,
+      "grad_norm": 4.784215450286865,
+      "learning_rate": 7.936842105263158e-06,
+      "loss": 0.2532,
+      "step": 13720
+    },
+    {
+      "epoch": 20.65,
+      "grad_norm": 6.748628616333008,
+      "learning_rate": 7.935338345864663e-06,
+      "loss": 0.3338,
+      "step": 13730
+    },
+    {
+      "epoch": 20.66,
+      "grad_norm": 6.877127647399902,
+      "learning_rate": 7.933834586466166e-06,
+      "loss": 0.2662,
+      "step": 13740
+    },
+    {
+      "epoch": 20.68,
+      "grad_norm": 7.904510974884033,
+      "learning_rate": 7.93233082706767e-06,
+      "loss": 0.3532,
+      "step": 13750
+    },
+    {
+      "epoch": 20.69,
+      "grad_norm": 3.242637872695923,
+      "learning_rate": 7.930827067669173e-06,
+      "loss": 0.3012,
+      "step": 13760
+    },
+    {
+      "epoch": 20.71,
+      "grad_norm": 3.0852160453796387,
+      "learning_rate": 7.929323308270679e-06,
+      "loss": 0.2948,
+      "step": 13770
+    },
+    {
+      "epoch": 20.72,
+      "grad_norm": 10.306611061096191,
+      "learning_rate": 7.92781954887218e-06,
+      "loss": 0.2837,
+      "step": 13780
+    },
+    {
+      "epoch": 20.74,
+      "grad_norm": 3.293861150741577,
+      "learning_rate": 7.926315789473686e-06,
+      "loss": 0.3023,
+      "step": 13790
+    },
+    {
+      "epoch": 20.75,
+      "grad_norm": 6.937739849090576,
+      "learning_rate": 7.924812030075189e-06,
+      "loss": 0.3147,
+      "step": 13800
+    },
+    {
+      "epoch": 20.77,
+      "grad_norm": 4.617713928222656,
+      "learning_rate": 7.923308270676693e-06,
+      "loss": 0.3503,
+      "step": 13810
+    },
+    {
+      "epoch": 20.78,
+      "grad_norm": 4.726008892059326,
+      "learning_rate": 7.921804511278196e-06,
+      "loss": 0.3526,
+      "step": 13820
+    },
+    {
+      "epoch": 20.8,
+      "grad_norm": 6.7416558265686035,
+      "learning_rate": 7.9203007518797e-06,
+      "loss": 0.2887,
+      "step": 13830
+    },
+    {
+      "epoch": 20.81,
+      "grad_norm": 6.563324451446533,
+      "learning_rate": 7.918796992481203e-06,
+      "loss": 0.2874,
+      "step": 13840
+    },
+    {
+      "epoch": 20.83,
+      "grad_norm": 5.6465840339660645,
+      "learning_rate": 7.917293233082708e-06,
+      "loss": 0.3598,
+      "step": 13850
+    },
+    {
+      "epoch": 20.84,
+      "grad_norm": 5.625797748565674,
+      "learning_rate": 7.915789473684212e-06,
+      "loss": 0.3195,
+      "step": 13860
+    },
+    {
+      "epoch": 20.86,
+      "grad_norm": 7.804876327514648,
+      "learning_rate": 7.914285714285715e-06,
+      "loss": 0.3369,
+      "step": 13870
+    },
+    {
+      "epoch": 20.87,
+      "grad_norm": 8.359640121459961,
+      "learning_rate": 7.912781954887219e-06,
+      "loss": 0.3332,
+      "step": 13880
+    },
+    {
+      "epoch": 20.89,
+      "grad_norm": 5.867996692657471,
+      "learning_rate": 7.911278195488723e-06,
+      "loss": 0.4172,
+      "step": 13890
+    },
+    {
+      "epoch": 20.9,
+      "grad_norm": 5.789632320404053,
+      "learning_rate": 7.909774436090226e-06,
+      "loss": 0.3705,
+      "step": 13900
+    },
+    {
+      "epoch": 20.92,
+      "grad_norm": 4.992484092712402,
+      "learning_rate": 7.90827067669173e-06,
+      "loss": 0.3287,
+      "step": 13910
+    },
+    {
+      "epoch": 20.93,
+      "grad_norm": 8.009930610656738,
+      "learning_rate": 7.906766917293233e-06,
+      "loss": 0.3065,
+      "step": 13920
+    },
+    {
+      "epoch": 20.95,
+      "grad_norm": 5.598669528961182,
+      "learning_rate": 7.905263157894737e-06,
+      "loss": 0.3115,
+      "step": 13930
+    },
+    {
+      "epoch": 20.96,
+      "grad_norm": 5.0005364418029785,
+      "learning_rate": 7.903759398496242e-06,
+      "loss": 0.3635,
+      "step": 13940
+    },
+    {
+      "epoch": 20.98,
+      "grad_norm": 6.9587788581848145,
+      "learning_rate": 7.902255639097745e-06,
+      "loss": 0.3386,
+      "step": 13950
+    },
+    {
+      "epoch": 20.99,
+      "grad_norm": 6.115482330322266,
+      "learning_rate": 7.900751879699249e-06,
+      "loss": 0.3676,
+      "step": 13960
+    },
+    {
+      "epoch": 21.0,
+      "eval_accuracy": 0.9319,
+      "eval_loss": 0.2700176537036896,
+      "eval_runtime": 84.5019,
+      "eval_samples_per_second": 118.341,
+      "eval_steps_per_second": 0.473,
+      "step": 13965
+    },
+    {
+      "epoch": 21.01,
+      "grad_norm": 3.4582622051239014,
+      "learning_rate": 7.899248120300752e-06,
+      "loss": 0.6324,
+      "step": 13970
+    },
+    {
+      "epoch": 21.02,
+      "grad_norm": 5.3081512451171875,
+      "learning_rate": 7.897744360902256e-06,
+      "loss": 0.3913,
+      "step": 13980
+    },
+    {
+      "epoch": 21.04,
+      "grad_norm": 7.172489166259766,
+      "learning_rate": 7.89624060150376e-06,
+      "loss": 0.3863,
+      "step": 13990
+    },
+    {
+      "epoch": 21.05,
+      "grad_norm": 6.308826446533203,
+      "learning_rate": 7.894736842105265e-06,
+      "loss": 0.3214,
+      "step": 14000
+    },
+    {
+      "epoch": 21.07,
+      "grad_norm": 5.6395111083984375,
+      "learning_rate": 7.893233082706766e-06,
+      "loss": 0.2801,
+      "step": 14010
+    },
+    {
+      "epoch": 21.08,
+      "grad_norm": 2.734475612640381,
+      "learning_rate": 7.891729323308272e-06,
+      "loss": 0.3212,
+      "step": 14020
+    },
+    {
+      "epoch": 21.1,
+      "grad_norm": 4.720375061035156,
+      "learning_rate": 7.890225563909775e-06,
+      "loss": 0.3271,
+      "step": 14030
+    },
+    {
+      "epoch": 21.11,
+      "grad_norm": 3.2389378547668457,
+      "learning_rate": 7.888721804511279e-06,
+      "loss": 0.3118,
+      "step": 14040
+    },
+    {
+      "epoch": 21.13,
+      "grad_norm": 3.7250821590423584,
+      "learning_rate": 7.887218045112782e-06,
+      "loss": 0.337,
+      "step": 14050
+    },
+    {
+      "epoch": 21.14,
+      "grad_norm": 8.08519458770752,
+      "learning_rate": 7.885714285714286e-06,
+      "loss": 0.3126,
+      "step": 14060
+    },
+    {
+      "epoch": 21.16,
+      "grad_norm": 5.902373313903809,
+      "learning_rate": 7.88421052631579e-06,
+      "loss": 0.3088,
+      "step": 14070
+    },
+    {
+      "epoch": 21.17,
+      "grad_norm": 4.858761787414551,
+      "learning_rate": 7.882706766917294e-06,
+      "loss": 0.3258,
+      "step": 14080
+    },
+    {
+      "epoch": 21.19,
+      "grad_norm": 6.120957374572754,
+      "learning_rate": 7.881203007518798e-06,
+      "loss": 0.2718,
+      "step": 14090
+    },
+    {
+      "epoch": 21.2,
+      "grad_norm": 3.7284491062164307,
+      "learning_rate": 7.879699248120301e-06,
+      "loss": 0.3521,
+      "step": 14100
+    },
+    {
+      "epoch": 21.22,
+      "grad_norm": 6.68389892578125,
+      "learning_rate": 7.878195488721805e-06,
+      "loss": 0.345,
+      "step": 14110
+    },
+    {
+      "epoch": 21.23,
+      "grad_norm": 6.401340484619141,
+      "learning_rate": 7.876691729323309e-06,
+      "loss": 0.302,
+      "step": 14120
+    },
+    {
+      "epoch": 21.25,
+      "grad_norm": 3.490999698638916,
+      "learning_rate": 7.875187969924812e-06,
+      "loss": 0.3658,
+      "step": 14130
+    },
+    {
+      "epoch": 21.26,
+      "grad_norm": 7.066733360290527,
+      "learning_rate": 7.873684210526317e-06,
+      "loss": 0.3089,
+      "step": 14140
+    },
+    {
+      "epoch": 21.28,
+      "grad_norm": 5.916280269622803,
+      "learning_rate": 7.872180451127819e-06,
+      "loss": 0.2996,
+      "step": 14150
+    },
+    {
+      "epoch": 21.29,
+      "grad_norm": 5.348523139953613,
+      "learning_rate": 7.870676691729324e-06,
+      "loss": 0.3389,
+      "step": 14160
+    },
+    {
+      "epoch": 21.31,
+      "grad_norm": 5.067401885986328,
+      "learning_rate": 7.869172932330828e-06,
+      "loss": 0.334,
+      "step": 14170
+    },
+    {
+      "epoch": 21.32,
+      "grad_norm": 4.67525053024292,
+      "learning_rate": 7.867669172932331e-06,
+      "loss": 0.3098,
+      "step": 14180
+    },
+    {
+      "epoch": 21.34,
+      "grad_norm": 7.282690525054932,
+      "learning_rate": 7.866165413533835e-06,
+      "loss": 0.3779,
+      "step": 14190
+    },
+    {
+      "epoch": 21.35,
+      "grad_norm": 6.01992654800415,
+      "learning_rate": 7.86466165413534e-06,
+      "loss": 0.2547,
+      "step": 14200
+    },
+    {
+      "epoch": 21.37,
+      "grad_norm": 3.902848720550537,
+      "learning_rate": 7.863157894736842e-06,
+      "loss": 0.2964,
+      "step": 14210
+    },
+    {
+      "epoch": 21.38,
+      "grad_norm": 5.350156784057617,
+      "learning_rate": 7.861654135338347e-06,
+      "loss": 0.325,
+      "step": 14220
+    },
+    {
+      "epoch": 21.4,
+      "grad_norm": 6.831077575683594,
+      "learning_rate": 7.86015037593985e-06,
+      "loss": 0.3629,
+      "step": 14230
+    },
+    {
+      "epoch": 21.41,
+      "grad_norm": 6.829043865203857,
+      "learning_rate": 7.858646616541354e-06,
+      "loss": 0.364,
+      "step": 14240
+    },
+    {
+      "epoch": 21.43,
+      "grad_norm": 4.683376789093018,
+      "learning_rate": 7.857142857142858e-06,
+      "loss": 0.3197,
+      "step": 14250
+    },
+    {
+      "epoch": 21.44,
+      "grad_norm": 5.582406520843506,
+      "learning_rate": 7.855639097744361e-06,
+      "loss": 0.2911,
+      "step": 14260
+    },
+    {
+      "epoch": 21.46,
+      "grad_norm": 3.7415030002593994,
+      "learning_rate": 7.854135338345865e-06,
+      "loss": 0.321,
+      "step": 14270
+    },
+    {
+      "epoch": 21.47,
+      "grad_norm": 5.881400108337402,
+      "learning_rate": 7.85263157894737e-06,
+      "loss": 0.2524,
+      "step": 14280
+    },
+    {
+      "epoch": 21.49,
+      "grad_norm": 8.578994750976562,
+      "learning_rate": 7.851127819548873e-06,
+      "loss": 0.2918,
+      "step": 14290
+    },
+    {
+      "epoch": 21.5,
+      "grad_norm": 8.9381742477417,
+      "learning_rate": 7.849624060150377e-06,
+      "loss": 0.4393,
+      "step": 14300
+    },
+    {
+      "epoch": 21.52,
+      "grad_norm": 4.9871602058410645,
+      "learning_rate": 7.84812030075188e-06,
+      "loss": 0.3624,
+      "step": 14310
+    },
+    {
+      "epoch": 21.53,
+      "grad_norm": 6.881855487823486,
+      "learning_rate": 7.846616541353384e-06,
+      "loss": 0.3561,
+      "step": 14320
+    },
+    {
+      "epoch": 21.55,
+      "grad_norm": 4.248598575592041,
+      "learning_rate": 7.845112781954888e-06,
+      "loss": 0.2925,
+      "step": 14330
+    },
+    {
+      "epoch": 21.56,
+      "grad_norm": 3.864591121673584,
+      "learning_rate": 7.843609022556393e-06,
+      "loss": 0.327,
+      "step": 14340
+    },
+    {
+      "epoch": 21.58,
+      "grad_norm": 6.71173095703125,
+      "learning_rate": 7.842105263157895e-06,
+      "loss": 0.3397,
+      "step": 14350
+    },
+    {
+      "epoch": 21.59,
+      "grad_norm": 3.9306325912475586,
+      "learning_rate": 7.8406015037594e-06,
+      "loss": 0.2788,
+      "step": 14360
+    },
+    {
+      "epoch": 21.61,
+      "grad_norm": 5.838838577270508,
+      "learning_rate": 7.839097744360903e-06,
+      "loss": 0.3116,
+      "step": 14370
+    },
+    {
+      "epoch": 21.62,
+      "grad_norm": 13.80683708190918,
+      "learning_rate": 7.837593984962407e-06,
+      "loss": 0.2882,
+      "step": 14380
+    },
+    {
+      "epoch": 21.64,
+      "grad_norm": 5.687681198120117,
+      "learning_rate": 7.83609022556391e-06,
+      "loss": 0.3521,
+      "step": 14390
+    },
+    {
+      "epoch": 21.65,
+      "grad_norm": 6.886200904846191,
+      "learning_rate": 7.834586466165414e-06,
+      "loss": 0.3544,
+      "step": 14400
+    },
+    {
+      "epoch": 21.67,
+      "grad_norm": 7.935529708862305,
+      "learning_rate": 7.833082706766917e-06,
+      "loss": 0.3214,
+      "step": 14410
+    },
+    {
+      "epoch": 21.68,
+      "grad_norm": 8.129573822021484,
+      "learning_rate": 7.831578947368421e-06,
+      "loss": 0.3341,
+      "step": 14420
+    },
+    {
+      "epoch": 21.7,
+      "grad_norm": 6.135876178741455,
+      "learning_rate": 7.830075187969926e-06,
+      "loss": 0.2829,
+      "step": 14430
+    },
+    {
+      "epoch": 21.71,
+      "grad_norm": 8.596725463867188,
+      "learning_rate": 7.828571428571428e-06,
+      "loss": 0.3856,
+      "step": 14440
+    },
+    {
+      "epoch": 21.73,
+      "grad_norm": 3.492475748062134,
+      "learning_rate": 7.827067669172933e-06,
+      "loss": 0.4355,
+      "step": 14450
+    },
+    {
+      "epoch": 21.74,
+      "grad_norm": 8.090415000915527,
+      "learning_rate": 7.825563909774437e-06,
+      "loss": 0.2981,
+      "step": 14460
+    },
+    {
+      "epoch": 21.76,
+      "grad_norm": 6.346333026885986,
+      "learning_rate": 7.82406015037594e-06,
+      "loss": 0.3715,
+      "step": 14470
+    },
+    {
+      "epoch": 21.77,
+      "grad_norm": 4.496158123016357,
+      "learning_rate": 7.822556390977444e-06,
+      "loss": 0.3275,
+      "step": 14480
+    },
+    {
+      "epoch": 21.79,
+      "grad_norm": 5.410872459411621,
+      "learning_rate": 7.821052631578949e-06,
+      "loss": 0.3401,
+      "step": 14490
+    },
+    {
+      "epoch": 21.8,
+      "grad_norm": 6.038545608520508,
+      "learning_rate": 7.81954887218045e-06,
+      "loss": 0.374,
+      "step": 14500
+    },
+    {
+      "epoch": 21.82,
+      "grad_norm": 6.986004829406738,
+      "learning_rate": 7.818045112781956e-06,
+      "loss": 0.3294,
+      "step": 14510
+    },
+    {
+      "epoch": 21.83,
+      "grad_norm": 6.893804550170898,
+      "learning_rate": 7.81654135338346e-06,
+      "loss": 0.3189,
+      "step": 14520
+    },
+    {
+      "epoch": 21.85,
+      "grad_norm": 3.2566752433776855,
+      "learning_rate": 7.815037593984963e-06,
+      "loss": 0.3837,
+      "step": 14530
+    },
+    {
+      "epoch": 21.86,
+      "grad_norm": 18.923295974731445,
+      "learning_rate": 7.813533834586466e-06,
+      "loss": 0.3319,
+      "step": 14540
+    },
+    {
+      "epoch": 21.88,
+      "grad_norm": 4.527941703796387,
+      "learning_rate": 7.81203007518797e-06,
+      "loss": 0.2917,
+      "step": 14550
+    },
+    {
+      "epoch": 21.89,
+      "grad_norm": 4.466981410980225,
+      "learning_rate": 7.810526315789474e-06,
+      "loss": 0.3239,
+      "step": 14560
+    },
+    {
+      "epoch": 21.91,
+      "grad_norm": 7.884292125701904,
+      "learning_rate": 7.809022556390979e-06,
+      "loss": 0.3068,
+      "step": 14570
+    },
+    {
+      "epoch": 21.92,
+      "grad_norm": 5.265153408050537,
+      "learning_rate": 7.807518796992482e-06,
+      "loss": 0.3472,
+      "step": 14580
+    },
+    {
+      "epoch": 21.94,
+      "grad_norm": 5.18987512588501,
+      "learning_rate": 7.806015037593986e-06,
+      "loss": 0.3452,
+      "step": 14590
+    },
+    {
+      "epoch": 21.95,
+      "grad_norm": 2.9694736003875732,
+      "learning_rate": 7.80451127819549e-06,
+      "loss": 0.3544,
+      "step": 14600
+    },
+    {
+      "epoch": 21.97,
+      "grad_norm": 5.018842697143555,
+      "learning_rate": 7.803007518796993e-06,
+      "loss": 0.2905,
+      "step": 14610
+    },
+    {
+      "epoch": 21.98,
+      "grad_norm": 7.495065212249756,
+      "learning_rate": 7.801503759398496e-06,
+      "loss": 0.3675,
+      "step": 14620
+    },
+    {
+      "epoch": 22.0,
+      "grad_norm": 4.347672939300537,
+      "learning_rate": 7.800000000000002e-06,
+      "loss": 0.3329,
+      "step": 14630
+    },
+    {
+      "epoch": 22.0,
+      "eval_accuracy": 0.9333,
+      "eval_loss": 0.272014319896698,
+      "eval_runtime": 84.1987,
+      "eval_samples_per_second": 118.767,
+      "eval_steps_per_second": 0.475,
+      "step": 14630
+    },
+    {
+      "epoch": 22.02,
+      "grad_norm": 6.415873050689697,
+      "learning_rate": 7.798496240601503e-06,
+      "loss": 0.3666,
+      "step": 14640
+    },
+    {
+      "epoch": 22.03,
+      "grad_norm": 3.398313522338867,
+      "learning_rate": 7.796992481203009e-06,
+      "loss": 0.2848,
+      "step": 14650
+    },
+    {
+      "epoch": 22.05,
+      "grad_norm": 6.848686218261719,
+      "learning_rate": 7.795488721804512e-06,
+      "loss": 0.3241,
+      "step": 14660
+    },
+    {
+      "epoch": 22.06,
+      "grad_norm": 4.208133697509766,
+      "learning_rate": 7.793984962406016e-06,
+      "loss": 0.3478,
+      "step": 14670
+    },
+    {
+      "epoch": 22.08,
+      "grad_norm": 7.1864213943481445,
+      "learning_rate": 7.792481203007519e-06,
+      "loss": 0.3515,
+      "step": 14680
+    },
+    {
+      "epoch": 22.09,
+      "grad_norm": 6.161532878875732,
+      "learning_rate": 7.790977443609024e-06,
+      "loss": 0.3254,
+      "step": 14690
+    },
+    {
+      "epoch": 22.11,
+      "grad_norm": 5.770120143890381,
+      "learning_rate": 7.789473684210526e-06,
+      "loss": 0.3253,
+      "step": 14700
+    },
+    {
+      "epoch": 22.12,
+      "grad_norm": 6.7618021965026855,
+      "learning_rate": 7.787969924812031e-06,
+      "loss": 0.2816,
+      "step": 14710
+    },
+    {
+      "epoch": 22.14,
+      "grad_norm": 8.137096405029297,
+      "learning_rate": 7.786466165413535e-06,
+      "loss": 0.3144,
+      "step": 14720
+    },
+    {
+      "epoch": 22.15,
+      "grad_norm": 10.209376335144043,
+      "learning_rate": 7.784962406015038e-06,
+      "loss": 0.3009,
+      "step": 14730
+    },
+    {
+      "epoch": 22.17,
+      "grad_norm": 4.7329792976379395,
+      "learning_rate": 7.783458646616542e-06,
+      "loss": 0.2923,
+      "step": 14740
+    },
+    {
+      "epoch": 22.18,
+      "grad_norm": 7.629988670349121,
+      "learning_rate": 7.781954887218045e-06,
+      "loss": 0.2598,
+      "step": 14750
+    },
+    {
+      "epoch": 22.2,
+      "grad_norm": 3.4664206504821777,
+      "learning_rate": 7.780451127819549e-06,
+      "loss": 0.2701,
+      "step": 14760
+    },
+    {
+      "epoch": 22.21,
+      "grad_norm": 7.403353691101074,
+      "learning_rate": 7.778947368421054e-06,
+      "loss": 0.2668,
+      "step": 14770
+    },
+    {
+      "epoch": 22.23,
+      "grad_norm": 7.961404800415039,
+      "learning_rate": 7.777443609022558e-06,
+      "loss": 0.2531,
+      "step": 14780
+    },
+    {
+      "epoch": 22.24,
+      "grad_norm": 6.001272201538086,
+      "learning_rate": 7.775939849624061e-06,
+      "loss": 0.3847,
+      "step": 14790
+    },
+    {
+      "epoch": 22.26,
+      "grad_norm": 9.084465026855469,
+      "learning_rate": 7.774436090225565e-06,
+      "loss": 0.2646,
+      "step": 14800
+    },
+    {
+      "epoch": 22.27,
+      "grad_norm": 7.2393012046813965,
+      "learning_rate": 7.772932330827068e-06,
+      "loss": 0.3203,
+      "step": 14810
+    },
+    {
+      "epoch": 22.29,
+      "grad_norm": 8.683479309082031,
+      "learning_rate": 7.771428571428572e-06,
+      "loss": 0.285,
+      "step": 14820
+    },
+    {
+      "epoch": 22.3,
+      "grad_norm": 5.802537441253662,
+      "learning_rate": 7.769924812030077e-06,
+      "loss": 0.3638,
+      "step": 14830
+    },
+    {
+      "epoch": 22.32,
+      "grad_norm": 2.3165462017059326,
+      "learning_rate": 7.768421052631579e-06,
+      "loss": 0.3051,
+      "step": 14840
+    },
+    {
+      "epoch": 22.33,
+      "grad_norm": 7.184248924255371,
+      "learning_rate": 7.766917293233084e-06,
+      "loss": 0.3008,
+      "step": 14850
+    },
+    {
+      "epoch": 22.35,
+      "grad_norm": 4.384416580200195,
+      "learning_rate": 7.765413533834588e-06,
+      "loss": 0.3827,
+      "step": 14860
+    },
+    {
+      "epoch": 22.36,
+      "grad_norm": 4.507965087890625,
+      "learning_rate": 7.763909774436091e-06,
+      "loss": 0.2811,
+      "step": 14870
+    },
+    {
+      "epoch": 22.38,
+      "grad_norm": 6.374339580535889,
+      "learning_rate": 7.762406015037595e-06,
+      "loss": 0.2874,
+      "step": 14880
+    },
+    {
+      "epoch": 22.39,
+      "grad_norm": 6.437895774841309,
+      "learning_rate": 7.760902255639098e-06,
+      "loss": 0.3101,
+      "step": 14890
+    },
+    {
+      "epoch": 22.41,
+      "grad_norm": 7.283292770385742,
+      "learning_rate": 7.759398496240602e-06,
+      "loss": 0.3169,
+      "step": 14900
+    },
+    {
+      "epoch": 22.42,
+      "grad_norm": 7.374083042144775,
+      "learning_rate": 7.757894736842105e-06,
+      "loss": 0.334,
+      "step": 14910
+    },
+    {
+      "epoch": 22.44,
+      "grad_norm": 8.147022247314453,
+      "learning_rate": 7.75639097744361e-06,
+      "loss": 0.382,
+      "step": 14920
+    },
+    {
+      "epoch": 22.45,
+      "grad_norm": 4.416728496551514,
+      "learning_rate": 7.754887218045112e-06,
+      "loss": 0.329,
+      "step": 14930
+    },
+    {
+      "epoch": 22.47,
+      "grad_norm": 4.376534938812256,
+      "learning_rate": 7.753383458646617e-06,
+      "loss": 0.4032,
+      "step": 14940
+    },
+    {
+      "epoch": 22.48,
+      "grad_norm": 9.247624397277832,
+      "learning_rate": 7.751879699248121e-06,
+      "loss": 0.3748,
+      "step": 14950
+    },
+    {
+      "epoch": 22.5,
+      "grad_norm": 6.153301239013672,
+      "learning_rate": 7.750375939849624e-06,
+      "loss": 0.295,
+      "step": 14960
+    },
+    {
+      "epoch": 22.51,
+      "grad_norm": 9.856022834777832,
+      "learning_rate": 7.748872180451128e-06,
+      "loss": 0.3033,
+      "step": 14970
+    },
+    {
+      "epoch": 22.53,
+      "grad_norm": 6.767205715179443,
+      "learning_rate": 7.747368421052631e-06,
+      "loss": 0.3249,
+      "step": 14980
+    },
+    {
+      "epoch": 22.54,
+      "grad_norm": 7.620680809020996,
+      "learning_rate": 7.745864661654135e-06,
+      "loss": 0.3775,
+      "step": 14990
+    },
+    {
+      "epoch": 22.56,
+      "grad_norm": 4.250925064086914,
+      "learning_rate": 7.74436090225564e-06,
+      "loss": 0.3464,
+      "step": 15000
+    },
+    {
+      "epoch": 22.57,
+      "grad_norm": 5.965568542480469,
+      "learning_rate": 7.742857142857144e-06,
+      "loss": 0.287,
+      "step": 15010
+    },
+    {
+      "epoch": 22.59,
+      "grad_norm": 7.024303913116455,
+      "learning_rate": 7.741353383458647e-06,
+      "loss": 0.3145,
+      "step": 15020
+    },
+    {
+      "epoch": 22.6,
+      "grad_norm": 5.691739559173584,
+      "learning_rate": 7.73984962406015e-06,
+      "loss": 0.2593,
+      "step": 15030
+    },
+    {
+      "epoch": 22.62,
+      "grad_norm": 4.950546741485596,
+      "learning_rate": 7.738345864661654e-06,
+      "loss": 0.2312,
+      "step": 15040
+    },
+    {
+      "epoch": 22.63,
+      "grad_norm": 8.352819442749023,
+      "learning_rate": 7.736842105263158e-06,
+      "loss": 0.3138,
+      "step": 15050
+    },
+    {
+      "epoch": 22.65,
+      "grad_norm": 7.059927940368652,
+      "learning_rate": 7.735338345864663e-06,
+      "loss": 0.3281,
+      "step": 15060
+    },
+    {
+      "epoch": 22.66,
+      "grad_norm": 8.698114395141602,
+      "learning_rate": 7.733834586466165e-06,
+      "loss": 0.3882,
+      "step": 15070
+    },
+    {
+      "epoch": 22.68,
+      "grad_norm": 9.284893989562988,
+      "learning_rate": 7.73233082706767e-06,
+      "loss": 0.3517,
+      "step": 15080
+    },
+    {
+      "epoch": 22.69,
+      "grad_norm": 10.998963356018066,
+      "learning_rate": 7.730827067669174e-06,
+      "loss": 0.3258,
+      "step": 15090
+    },
+    {
+      "epoch": 22.71,
+      "grad_norm": 6.2202277183532715,
+      "learning_rate": 7.729323308270677e-06,
+      "loss": 0.3219,
+      "step": 15100
+    },
+    {
+      "epoch": 22.72,
+      "grad_norm": 5.044315814971924,
+      "learning_rate": 7.72781954887218e-06,
+      "loss": 0.3523,
+      "step": 15110
+    },
+    {
+      "epoch": 22.74,
+      "grad_norm": 7.376669883728027,
+      "learning_rate": 7.726315789473686e-06,
+      "loss": 0.3644,
+      "step": 15120
+    },
+    {
+      "epoch": 22.75,
+      "grad_norm": 4.171261787414551,
+      "learning_rate": 7.724812030075188e-06,
+      "loss": 0.356,
+      "step": 15130
+    },
+    {
+      "epoch": 22.77,
+      "grad_norm": 4.358337879180908,
+      "learning_rate": 7.723308270676693e-06,
+      "loss": 0.3316,
+      "step": 15140
+    },
+    {
+      "epoch": 22.78,
+      "grad_norm": 7.347938060760498,
+      "learning_rate": 7.721804511278196e-06,
+      "loss": 0.3134,
+      "step": 15150
+    },
+    {
+      "epoch": 22.8,
+      "grad_norm": 6.51662540435791,
+      "learning_rate": 7.7203007518797e-06,
+      "loss": 0.3304,
+      "step": 15160
+    },
+    {
+      "epoch": 22.81,
+      "grad_norm": 4.554670810699463,
+      "learning_rate": 7.718796992481203e-06,
+      "loss": 0.3568,
+      "step": 15170
+    },
+    {
+      "epoch": 22.83,
+      "grad_norm": 5.587891101837158,
+      "learning_rate": 7.717293233082707e-06,
+      "loss": 0.2347,
+      "step": 15180
+    },
+    {
+      "epoch": 22.84,
+      "grad_norm": 4.167472839355469,
+      "learning_rate": 7.71578947368421e-06,
+      "loss": 0.2926,
+      "step": 15190
+    },
+    {
+      "epoch": 22.86,
+      "grad_norm": 7.967960834503174,
+      "learning_rate": 7.714285714285716e-06,
+      "loss": 0.3413,
+      "step": 15200
+    },
+    {
+      "epoch": 22.87,
+      "grad_norm": 4.611883640289307,
+      "learning_rate": 7.71278195488722e-06,
+      "loss": 0.328,
+      "step": 15210
+    },
+    {
+      "epoch": 22.89,
+      "grad_norm": 6.755552291870117,
+      "learning_rate": 7.711278195488723e-06,
+      "loss": 0.3418,
+      "step": 15220
+    },
+    {
+      "epoch": 22.9,
+      "grad_norm": 7.961472034454346,
+      "learning_rate": 7.709774436090226e-06,
+      "loss": 0.3717,
+      "step": 15230
+    },
+    {
+      "epoch": 22.92,
+      "grad_norm": 6.6936187744140625,
+      "learning_rate": 7.70827067669173e-06,
+      "loss": 0.318,
+      "step": 15240
+    },
+    {
+      "epoch": 22.93,
+      "grad_norm": 5.8512444496154785,
+      "learning_rate": 7.706766917293233e-06,
+      "loss": 0.3481,
+      "step": 15250
+    },
+    {
+      "epoch": 22.95,
+      "grad_norm": 6.354591369628906,
+      "learning_rate": 7.705263157894738e-06,
+      "loss": 0.3909,
+      "step": 15260
+    },
+    {
+      "epoch": 22.96,
+      "grad_norm": 5.42380428314209,
+      "learning_rate": 7.70375939849624e-06,
+      "loss": 0.34,
+      "step": 15270
+    },
+    {
+      "epoch": 22.98,
+      "grad_norm": 9.55280876159668,
+      "learning_rate": 7.702255639097746e-06,
+      "loss": 0.3557,
+      "step": 15280
+    },
+    {
+      "epoch": 22.99,
+      "grad_norm": 5.095829963684082,
+      "learning_rate": 7.700751879699249e-06,
+      "loss": 0.4089,
+      "step": 15290
+    },
+    {
+      "epoch": 23.0,
+      "eval_accuracy": 0.9325,
+      "eval_loss": 0.27643856406211853,
+      "eval_runtime": 84.9045,
+      "eval_samples_per_second": 117.779,
+      "eval_steps_per_second": 0.471,
+      "step": 15295
+    },
+    {
+      "epoch": 23.01,
+      "grad_norm": 5.0213518142700195,
+      "learning_rate": 7.699248120300753e-06,
+      "loss": 0.3049,
+      "step": 15300
+    },
+    {
+      "epoch": 23.02,
+      "grad_norm": 7.806211948394775,
+      "learning_rate": 7.697744360902256e-06,
+      "loss": 0.3641,
+      "step": 15310
+    },
+    {
+      "epoch": 23.04,
+      "grad_norm": 3.328399896621704,
+      "learning_rate": 7.696240601503761e-06,
+      "loss": 0.3035,
+      "step": 15320
+    },
+    {
+      "epoch": 23.05,
+      "grad_norm": 4.0139875411987305,
+      "learning_rate": 7.694736842105263e-06,
+      "loss": 0.3657,
+      "step": 15330
+    },
+    {
+      "epoch": 23.07,
+      "grad_norm": 5.296818256378174,
+      "learning_rate": 7.693233082706768e-06,
+      "loss": 0.3067,
+      "step": 15340
+    },
+    {
+      "epoch": 23.08,
+      "grad_norm": 6.957002639770508,
+      "learning_rate": 7.691729323308272e-06,
+      "loss": 0.3746,
+      "step": 15350
+    },
+    {
+      "epoch": 23.1,
+      "grad_norm": 5.104499816894531,
+      "learning_rate": 7.690225563909775e-06,
+      "loss": 0.3475,
+      "step": 15360
+    },
+    {
+      "epoch": 23.11,
+      "grad_norm": 4.252979755401611,
+      "learning_rate": 7.688721804511279e-06,
+      "loss": 0.3913,
+      "step": 15370
+    },
+    {
+      "epoch": 23.13,
+      "grad_norm": 6.049491882324219,
+      "learning_rate": 7.687218045112782e-06,
+      "loss": 0.2903,
+      "step": 15380
+    },
+    {
+      "epoch": 23.14,
+      "grad_norm": 5.293207168579102,
+      "learning_rate": 7.685714285714286e-06,
+      "loss": 0.2508,
+      "step": 15390
+    },
+    {
+      "epoch": 23.16,
+      "grad_norm": 5.338700771331787,
+      "learning_rate": 7.68421052631579e-06,
+      "loss": 0.2695,
+      "step": 15400
+    },
+    {
+      "epoch": 23.17,
+      "grad_norm": 3.670703172683716,
+      "learning_rate": 7.682706766917295e-06,
+      "loss": 0.3534,
+      "step": 15410
+    },
+    {
+      "epoch": 23.19,
+      "grad_norm": 5.252980709075928,
+      "learning_rate": 7.681203007518796e-06,
+      "loss": 0.3555,
+      "step": 15420
+    },
+    {
+      "epoch": 23.2,
+      "grad_norm": 8.209909439086914,
+      "learning_rate": 7.679699248120302e-06,
+      "loss": 0.3796,
+      "step": 15430
+    },
+    {
+      "epoch": 23.22,
+      "grad_norm": 4.704070091247559,
+      "learning_rate": 7.678195488721805e-06,
+      "loss": 0.3024,
+      "step": 15440
+    },
+    {
+      "epoch": 23.23,
+      "grad_norm": 5.453511714935303,
+      "learning_rate": 7.676691729323309e-06,
+      "loss": 0.2894,
+      "step": 15450
+    },
+    {
+      "epoch": 23.25,
+      "grad_norm": 7.019674777984619,
+      "learning_rate": 7.675187969924812e-06,
+      "loss": 0.3011,
+      "step": 15460
+    },
+    {
+      "epoch": 23.26,
+      "grad_norm": 16.94940185546875,
+      "learning_rate": 7.673684210526316e-06,
+      "loss": 0.3526,
+      "step": 15470
+    },
+    {
+      "epoch": 23.28,
+      "grad_norm": 5.086287498474121,
+      "learning_rate": 7.67218045112782e-06,
+      "loss": 0.3219,
+      "step": 15480
+    },
+    {
+      "epoch": 23.29,
+      "grad_norm": 3.3719613552093506,
+      "learning_rate": 7.670676691729325e-06,
+      "loss": 0.2706,
+      "step": 15490
+    },
+    {
+      "epoch": 23.31,
+      "grad_norm": 6.172971248626709,
+      "learning_rate": 7.669172932330828e-06,
+      "loss": 0.2633,
+      "step": 15500
+    },
+    {
+      "epoch": 23.32,
+      "grad_norm": 6.304495811462402,
+      "learning_rate": 7.667669172932332e-06,
+      "loss": 0.3932,
+      "step": 15510
+    },
+    {
+      "epoch": 23.34,
+      "grad_norm": 7.454652309417725,
+      "learning_rate": 7.666165413533835e-06,
+      "loss": 0.3278,
+      "step": 15520
+    },
+    {
+      "epoch": 23.35,
+      "grad_norm": 4.993924617767334,
+      "learning_rate": 7.664661654135339e-06,
+      "loss": 0.3158,
+      "step": 15530
+    },
+    {
+      "epoch": 23.37,
+      "grad_norm": 8.269258499145508,
+      "learning_rate": 7.663157894736842e-06,
+      "loss": 0.2965,
+      "step": 15540
+    },
+    {
+      "epoch": 23.38,
+      "grad_norm": 9.472188949584961,
+      "learning_rate": 7.661654135338347e-06,
+      "loss": 0.3321,
+      "step": 15550
+    },
+    {
+      "epoch": 23.4,
+      "grad_norm": 4.879781246185303,
+      "learning_rate": 7.66015037593985e-06,
+      "loss": 0.3012,
+      "step": 15560
+    },
+    {
+      "epoch": 23.41,
+      "grad_norm": 3.718254566192627,
+      "learning_rate": 7.658646616541354e-06,
+      "loss": 0.3498,
+      "step": 15570
+    },
+    {
+      "epoch": 23.43,
+      "grad_norm": 3.531419038772583,
+      "learning_rate": 7.657142857142858e-06,
+      "loss": 0.2622,
+      "step": 15580
+    },
+    {
+      "epoch": 23.44,
+      "grad_norm": 4.4930949211120605,
+      "learning_rate": 7.655639097744361e-06,
+      "loss": 0.2778,
+      "step": 15590
+    },
+    {
+      "epoch": 23.46,
+      "grad_norm": 4.008451461791992,
+      "learning_rate": 7.654135338345865e-06,
+      "loss": 0.3077,
+      "step": 15600
+    },
+    {
+      "epoch": 23.47,
+      "grad_norm": 6.081947326660156,
+      "learning_rate": 7.65263157894737e-06,
+      "loss": 0.2566,
+      "step": 15610
+    },
+    {
+      "epoch": 23.49,
+      "grad_norm": 3.446821689605713,
+      "learning_rate": 7.651127819548872e-06,
+      "loss": 0.3096,
+      "step": 15620
+    },
+    {
+      "epoch": 23.5,
+      "grad_norm": 10.16897201538086,
+      "learning_rate": 7.649624060150377e-06,
+      "loss": 0.3254,
+      "step": 15630
+    },
+    {
+      "epoch": 23.52,
+      "grad_norm": 1.691789150238037,
+      "learning_rate": 7.64812030075188e-06,
+      "loss": 0.3094,
+      "step": 15640
+    },
+    {
+      "epoch": 23.53,
+      "grad_norm": 4.911680698394775,
+      "learning_rate": 7.646616541353384e-06,
+      "loss": 0.2983,
+      "step": 15650
+    },
+    {
+      "epoch": 23.55,
+      "grad_norm": 6.379064559936523,
+      "learning_rate": 7.645112781954888e-06,
+      "loss": 0.2172,
+      "step": 15660
+    },
+    {
+      "epoch": 23.56,
+      "grad_norm": 4.125355243682861,
+      "learning_rate": 7.643609022556391e-06,
+      "loss": 0.2838,
+      "step": 15670
+    },
+    {
+      "epoch": 23.58,
+      "grad_norm": 6.153583526611328,
+      "learning_rate": 7.642105263157895e-06,
+      "loss": 0.3016,
+      "step": 15680
+    },
+    {
+      "epoch": 23.59,
+      "grad_norm": 7.581343173980713,
+      "learning_rate": 7.6406015037594e-06,
+      "loss": 0.3684,
+      "step": 15690
+    },
+    {
+      "epoch": 23.61,
+      "grad_norm": 8.388538360595703,
+      "learning_rate": 7.639097744360904e-06,
+      "loss": 0.3272,
+      "step": 15700
+    },
+    {
+      "epoch": 23.62,
+      "grad_norm": 7.5239362716674805,
+      "learning_rate": 7.637593984962407e-06,
+      "loss": 0.3148,
+      "step": 15710
+    },
+    {
+      "epoch": 23.64,
+      "grad_norm": 7.033330917358398,
+      "learning_rate": 7.63609022556391e-06,
+      "loss": 0.3008,
+      "step": 15720
+    },
+    {
+      "epoch": 23.65,
+      "grad_norm": 5.851361274719238,
+      "learning_rate": 7.634586466165414e-06,
+      "loss": 0.3205,
+      "step": 15730
+    },
+    {
+      "epoch": 23.67,
+      "grad_norm": 2.9884681701660156,
+      "learning_rate": 7.633082706766918e-06,
+      "loss": 0.2622,
+      "step": 15740
+    },
+    {
+      "epoch": 23.68,
+      "grad_norm": 7.287815570831299,
+      "learning_rate": 7.631578947368423e-06,
+      "loss": 0.3984,
+      "step": 15750
+    },
+    {
+      "epoch": 23.7,
+      "grad_norm": 4.209038734436035,
+      "learning_rate": 7.630075187969925e-06,
+      "loss": 0.2849,
+      "step": 15760
+    },
+    {
+      "epoch": 23.71,
+      "grad_norm": 3.485328197479248,
+      "learning_rate": 7.62857142857143e-06,
+      "loss": 0.3395,
+      "step": 15770
+    },
+    {
+      "epoch": 23.73,
+      "grad_norm": 4.945652484893799,
+      "learning_rate": 7.6270676691729325e-06,
+      "loss": 0.3045,
+      "step": 15780
+    },
+    {
+      "epoch": 23.74,
+      "grad_norm": 9.309234619140625,
+      "learning_rate": 7.625563909774437e-06,
+      "loss": 0.3023,
+      "step": 15790
+    },
+    {
+      "epoch": 23.76,
+      "grad_norm": 4.634711265563965,
+      "learning_rate": 7.62406015037594e-06,
+      "loss": 0.371,
+      "step": 15800
+    },
+    {
+      "epoch": 23.77,
+      "grad_norm": 5.632133483886719,
+      "learning_rate": 7.622556390977445e-06,
+      "loss": 0.3813,
+      "step": 15810
+    },
+    {
+      "epoch": 23.79,
+      "grad_norm": 3.4830102920532227,
+      "learning_rate": 7.621052631578948e-06,
+      "loss": 0.2663,
+      "step": 15820
+    },
+    {
+      "epoch": 23.8,
+      "grad_norm": 4.266145706176758,
+      "learning_rate": 7.619548872180453e-06,
+      "loss": 0.3295,
+      "step": 15830
+    },
+    {
+      "epoch": 23.82,
+      "grad_norm": 6.466432571411133,
+      "learning_rate": 7.618045112781955e-06,
+      "loss": 0.3109,
+      "step": 15840
+    },
+    {
+      "epoch": 23.83,
+      "grad_norm": 7.617573261260986,
+      "learning_rate": 7.61654135338346e-06,
+      "loss": 0.3228,
+      "step": 15850
+    },
+    {
+      "epoch": 23.85,
+      "grad_norm": 6.581226825714111,
+      "learning_rate": 7.615037593984963e-06,
+      "loss": 0.293,
+      "step": 15860
+    },
+    {
+      "epoch": 23.86,
+      "grad_norm": 3.765594720840454,
+      "learning_rate": 7.6135338345864676e-06,
+      "loss": 0.2756,
+      "step": 15870
+    },
+    {
+      "epoch": 23.88,
+      "grad_norm": 5.244023323059082,
+      "learning_rate": 7.61203007518797e-06,
+      "loss": 0.3237,
+      "step": 15880
+    },
+    {
+      "epoch": 23.89,
+      "grad_norm": 4.341951370239258,
+      "learning_rate": 7.610526315789474e-06,
+      "loss": 0.3115,
+      "step": 15890
+    },
+    {
+      "epoch": 23.91,
+      "grad_norm": 7.612969875335693,
+      "learning_rate": 7.609022556390978e-06,
+      "loss": 0.3097,
+      "step": 15900
+    },
+    {
+      "epoch": 23.92,
+      "grad_norm": 4.730962753295898,
+      "learning_rate": 7.607518796992482e-06,
+      "loss": 0.381,
+      "step": 15910
+    },
+    {
+      "epoch": 23.94,
+      "grad_norm": 6.929048538208008,
+      "learning_rate": 7.606015037593986e-06,
+      "loss": 0.4364,
+      "step": 15920
+    },
+    {
+      "epoch": 23.95,
+      "grad_norm": 8.269315719604492,
+      "learning_rate": 7.604511278195489e-06,
+      "loss": 0.3312,
+      "step": 15930
+    },
+    {
+      "epoch": 23.97,
+      "grad_norm": 5.69378662109375,
+      "learning_rate": 7.603007518796993e-06,
+      "loss": 0.3105,
+      "step": 15940
+    },
+    {
+      "epoch": 23.98,
+      "grad_norm": 2.1282405853271484,
+      "learning_rate": 7.6015037593984966e-06,
+      "loss": 0.3473,
+      "step": 15950
+    },
+    {
+      "epoch": 24.0,
+      "grad_norm": 64.98310089111328,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 0.3196,
+      "step": 15960
+    },
+    {
+      "epoch": 24.0,
+      "eval_accuracy": 0.9305,
+      "eval_loss": 0.27353373169898987,
+      "eval_runtime": 84.6965,
+      "eval_samples_per_second": 118.069,
+      "eval_steps_per_second": 0.472,
+      "step": 15960
+    },
+    {
+      "epoch": 24.02,
+      "grad_norm": 6.7349653244018555,
+      "learning_rate": 7.598496240601504e-06,
+      "loss": 0.2695,
+      "step": 15970
+    },
+    {
+      "epoch": 24.03,
+      "grad_norm": 3.1804494857788086,
+      "learning_rate": 7.596992481203008e-06,
+      "loss": 0.2994,
+      "step": 15980
+    },
+    {
+      "epoch": 24.05,
+      "grad_norm": 5.326109886169434,
+      "learning_rate": 7.5954887218045115e-06,
+      "loss": 0.3565,
+      "step": 15990
+    },
+    {
+      "epoch": 24.06,
+      "grad_norm": 5.608499526977539,
+      "learning_rate": 7.593984962406016e-06,
+      "loss": 0.2904,
+      "step": 16000
+    },
+    {
+      "epoch": 24.08,
+      "grad_norm": 3.598764181137085,
+      "learning_rate": 7.592481203007519e-06,
+      "loss": 0.2959,
+      "step": 16010
+    },
+    {
+      "epoch": 24.09,
+      "grad_norm": 11.522496223449707,
+      "learning_rate": 7.590977443609024e-06,
+      "loss": 0.2509,
+      "step": 16020
+    },
+    {
+      "epoch": 24.11,
+      "grad_norm": 5.142250061035156,
+      "learning_rate": 7.589473684210526e-06,
+      "loss": 0.2768,
+      "step": 16030
+    },
+    {
+      "epoch": 24.12,
+      "grad_norm": 4.173882484436035,
+      "learning_rate": 7.587969924812031e-06,
+      "loss": 0.3249,
+      "step": 16040
+    },
+    {
+      "epoch": 24.14,
+      "grad_norm": 5.356581211090088,
+      "learning_rate": 7.586466165413534e-06,
+      "loss": 0.342,
+      "step": 16050
+    },
+    {
+      "epoch": 24.15,
+      "grad_norm": 5.133426666259766,
+      "learning_rate": 7.584962406015039e-06,
+      "loss": 0.372,
+      "step": 16060
+    },
+    {
+      "epoch": 24.17,
+      "grad_norm": 6.920117378234863,
+      "learning_rate": 7.583458646616541e-06,
+      "loss": 0.2469,
+      "step": 16070
+    },
+    {
+      "epoch": 24.18,
+      "grad_norm": 5.305706024169922,
+      "learning_rate": 7.581954887218046e-06,
+      "loss": 0.3355,
+      "step": 16080
+    },
+    {
+      "epoch": 24.2,
+      "grad_norm": 5.647949695587158,
+      "learning_rate": 7.580451127819549e-06,
+      "loss": 0.3394,
+      "step": 16090
+    },
+    {
+      "epoch": 24.21,
+      "grad_norm": 8.31951904296875,
+      "learning_rate": 7.578947368421054e-06,
+      "loss": 0.2804,
+      "step": 16100
+    },
+    {
+      "epoch": 24.23,
+      "grad_norm": 7.295516490936279,
+      "learning_rate": 7.577443609022557e-06,
+      "loss": 0.2522,
+      "step": 16110
+    },
+    {
+      "epoch": 24.24,
+      "grad_norm": 4.9929022789001465,
+      "learning_rate": 7.575939849624061e-06,
+      "loss": 0.2747,
+      "step": 16120
+    },
+    {
+      "epoch": 24.26,
+      "grad_norm": 4.699337482452393,
+      "learning_rate": 7.574436090225564e-06,
+      "loss": 0.2968,
+      "step": 16130
+    },
+    {
+      "epoch": 24.27,
+      "grad_norm": 8.267194747924805,
+      "learning_rate": 7.5729323308270685e-06,
+      "loss": 0.396,
+      "step": 16140
+    },
+    {
+      "epoch": 24.29,
+      "grad_norm": 5.007888317108154,
+      "learning_rate": 7.571428571428572e-06,
+      "loss": 0.2286,
+      "step": 16150
+    },
+    {
+      "epoch": 24.3,
+      "grad_norm": 4.469249725341797,
+      "learning_rate": 7.569924812030076e-06,
+      "loss": 0.2907,
+      "step": 16160
+    },
+    {
+      "epoch": 24.32,
+      "grad_norm": 5.973073482513428,
+      "learning_rate": 7.568421052631579e-06,
+      "loss": 0.2583,
+      "step": 16170
+    },
+    {
+      "epoch": 24.33,
+      "grad_norm": 7.025624752044678,
+      "learning_rate": 7.5669172932330834e-06,
+      "loss": 0.332,
+      "step": 16180
+    },
+    {
+      "epoch": 24.35,
+      "grad_norm": 6.508031368255615,
+      "learning_rate": 7.565413533834587e-06,
+      "loss": 0.3216,
+      "step": 16190
+    },
+    {
+      "epoch": 24.36,
+      "grad_norm": 5.830172538757324,
+      "learning_rate": 7.563909774436091e-06,
+      "loss": 0.2935,
+      "step": 16200
+    },
+    {
+      "epoch": 24.38,
+      "grad_norm": 8.18553638458252,
+      "learning_rate": 7.562406015037595e-06,
+      "loss": 0.309,
+      "step": 16210
+    },
+    {
+      "epoch": 24.39,
+      "grad_norm": 19.962448120117188,
+      "learning_rate": 7.560902255639098e-06,
+      "loss": 0.3435,
+      "step": 16220
+    },
+    {
+      "epoch": 24.41,
+      "grad_norm": 4.706577301025391,
+      "learning_rate": 7.559398496240602e-06,
+      "loss": 0.3436,
+      "step": 16230
+    },
+    {
+      "epoch": 24.42,
+      "grad_norm": 5.416121959686279,
+      "learning_rate": 7.557894736842106e-06,
+      "loss": 0.3253,
+      "step": 16240
+    },
+    {
+      "epoch": 24.44,
+      "grad_norm": 7.063791275024414,
+      "learning_rate": 7.55639097744361e-06,
+      "loss": 0.3762,
+      "step": 16250
+    },
+    {
+      "epoch": 24.45,
+      "grad_norm": 4.680507659912109,
+      "learning_rate": 7.554887218045114e-06,
+      "loss": 0.2883,
+      "step": 16260
+    },
+    {
+      "epoch": 24.47,
+      "grad_norm": 8.150341987609863,
+      "learning_rate": 7.553383458646617e-06,
+      "loss": 0.3714,
+      "step": 16270
+    },
+    {
+      "epoch": 24.48,
+      "grad_norm": 3.5707345008850098,
+      "learning_rate": 7.551879699248121e-06,
+      "loss": 0.2659,
+      "step": 16280
+    },
+    {
+      "epoch": 24.5,
+      "grad_norm": 7.740113735198975,
+      "learning_rate": 7.550375939849625e-06,
+      "loss": 0.3134,
+      "step": 16290
+    },
+    {
+      "epoch": 24.51,
+      "grad_norm": 28.682573318481445,
+      "learning_rate": 7.548872180451129e-06,
+      "loss": 0.3294,
+      "step": 16300
+    },
+    {
+      "epoch": 24.53,
+      "grad_norm": 5.2670793533325195,
+      "learning_rate": 7.547368421052632e-06,
+      "loss": 0.272,
+      "step": 16310
+    },
+    {
+      "epoch": 24.54,
+      "grad_norm": 6.7343339920043945,
+      "learning_rate": 7.545864661654136e-06,
+      "loss": 0.2836,
+      "step": 16320
+    },
+    {
+      "epoch": 24.56,
+      "grad_norm": 5.1457672119140625,
+      "learning_rate": 7.54436090225564e-06,
+      "loss": 0.222,
+      "step": 16330
+    },
+    {
+      "epoch": 24.57,
+      "grad_norm": 5.616130828857422,
+      "learning_rate": 7.542857142857144e-06,
+      "loss": 0.2684,
+      "step": 16340
+    },
+    {
+      "epoch": 24.59,
+      "grad_norm": 6.380970478057861,
+      "learning_rate": 7.5413533834586475e-06,
+      "loss": 0.3899,
+      "step": 16350
+    },
+    {
+      "epoch": 24.6,
+      "grad_norm": 9.467977523803711,
+      "learning_rate": 7.539849624060152e-06,
+      "loss": 0.2699,
+      "step": 16360
+    },
+    {
+      "epoch": 24.62,
+      "grad_norm": 5.1410136222839355,
+      "learning_rate": 7.5383458646616545e-06,
+      "loss": 0.2859,
+      "step": 16370
+    },
+    {
+      "epoch": 24.63,
+      "grad_norm": 7.608031749725342,
+      "learning_rate": 7.536842105263158e-06,
+      "loss": 0.342,
+      "step": 16380
+    },
+    {
+      "epoch": 24.65,
+      "grad_norm": 3.234297513961792,
+      "learning_rate": 7.535338345864662e-06,
+      "loss": 0.3229,
+      "step": 16390
+    },
+    {
+      "epoch": 24.66,
+      "grad_norm": 5.070233345031738,
+      "learning_rate": 7.533834586466165e-06,
+      "loss": 0.3494,
+      "step": 16400
+    },
+    {
+      "epoch": 24.68,
+      "grad_norm": 5.208074569702148,
+      "learning_rate": 7.5323308270676694e-06,
+      "loss": 0.3288,
+      "step": 16410
+    },
+    {
+      "epoch": 24.69,
+      "grad_norm": 3.5855026245117188,
+      "learning_rate": 7.530827067669173e-06,
+      "loss": 0.3241,
+      "step": 16420
+    },
+    {
+      "epoch": 24.71,
+      "grad_norm": 5.344216823577881,
+      "learning_rate": 7.529323308270677e-06,
+      "loss": 0.2997,
+      "step": 16430
+    },
+    {
+      "epoch": 24.72,
+      "grad_norm": 7.309630393981934,
+      "learning_rate": 7.527819548872181e-06,
+      "loss": 0.304,
+      "step": 16440
+    },
+    {
+      "epoch": 24.74,
+      "grad_norm": 4.991174697875977,
+      "learning_rate": 7.526315789473685e-06,
+      "loss": 0.3107,
+      "step": 16450
+    },
+    {
+      "epoch": 24.75,
+      "grad_norm": 6.519332408905029,
+      "learning_rate": 7.524812030075188e-06,
+      "loss": 0.2857,
+      "step": 16460
+    },
+    {
+      "epoch": 24.77,
+      "grad_norm": 6.599658012390137,
+      "learning_rate": 7.523308270676692e-06,
+      "loss": 0.3334,
+      "step": 16470
+    },
+    {
+      "epoch": 24.78,
+      "grad_norm": 6.142018795013428,
+      "learning_rate": 7.521804511278196e-06,
+      "loss": 0.3121,
+      "step": 16480
+    },
+    {
+      "epoch": 24.8,
+      "grad_norm": 5.483290195465088,
+      "learning_rate": 7.5203007518797e-06,
+      "loss": 0.3155,
+      "step": 16490
+    },
+    {
+      "epoch": 24.81,
+      "grad_norm": 10.318331718444824,
+      "learning_rate": 7.518796992481203e-06,
+      "loss": 0.3272,
+      "step": 16500
+    },
+    {
+      "epoch": 24.83,
+      "grad_norm": 4.9072771072387695,
+      "learning_rate": 7.517293233082707e-06,
+      "loss": 0.3807,
+      "step": 16510
+    },
+    {
+      "epoch": 24.84,
+      "grad_norm": 5.348799228668213,
+      "learning_rate": 7.515789473684211e-06,
+      "loss": 0.3662,
+      "step": 16520
+    },
+    {
+      "epoch": 24.86,
+      "grad_norm": 4.338939189910889,
+      "learning_rate": 7.514285714285715e-06,
+      "loss": 0.3022,
+      "step": 16530
+    },
+    {
+      "epoch": 24.87,
+      "grad_norm": 7.069118499755859,
+      "learning_rate": 7.512781954887219e-06,
+      "loss": 0.3111,
+      "step": 16540
+    },
+    {
+      "epoch": 24.89,
+      "grad_norm": 7.8967790603637695,
+      "learning_rate": 7.511278195488723e-06,
+      "loss": 0.2683,
+      "step": 16550
+    },
+    {
+      "epoch": 24.9,
+      "grad_norm": 6.323410987854004,
+      "learning_rate": 7.509774436090226e-06,
+      "loss": 0.3274,
+      "step": 16560
+    },
+    {
+      "epoch": 24.92,
+      "grad_norm": 6.77821683883667,
+      "learning_rate": 7.50827067669173e-06,
+      "loss": 0.2803,
+      "step": 16570
+    },
+    {
+      "epoch": 24.93,
+      "grad_norm": 6.189555644989014,
+      "learning_rate": 7.5067669172932335e-06,
+      "loss": 0.3104,
+      "step": 16580
+    },
+    {
+      "epoch": 24.95,
+      "grad_norm": 3.698765277862549,
+      "learning_rate": 7.505263157894738e-06,
+      "loss": 0.3238,
+      "step": 16590
+    },
+    {
+      "epoch": 24.96,
+      "grad_norm": 5.7407355308532715,
+      "learning_rate": 7.5037593984962405e-06,
+      "loss": 0.3003,
+      "step": 16600
+    },
+    {
+      "epoch": 24.98,
+      "grad_norm": 4.214737892150879,
+      "learning_rate": 7.502255639097745e-06,
+      "loss": 0.3016,
+      "step": 16610
+    },
+    {
+      "epoch": 24.99,
+      "grad_norm": 6.830847263336182,
+      "learning_rate": 7.5007518796992484e-06,
+      "loss": 0.2982,
+      "step": 16620
+    },
+    {
+      "epoch": 25.0,
+      "eval_accuracy": 0.9312,
+      "eval_loss": 0.27709507942199707,
+      "eval_runtime": 84.7229,
+      "eval_samples_per_second": 118.032,
+      "eval_steps_per_second": 0.472,
+      "step": 16625
+    },
+    {
+      "epoch": 25.01,
+      "grad_norm": 8.3620023727417,
+      "learning_rate": 7.499248120300753e-06,
+      "loss": 0.2671,
+      "step": 16630
+    },
+    {
+      "epoch": 25.02,
+      "grad_norm": 6.188384532928467,
+      "learning_rate": 7.497744360902256e-06,
+      "loss": 0.3315,
+      "step": 16640
+    },
+    {
+      "epoch": 25.04,
+      "grad_norm": 4.457509994506836,
+      "learning_rate": 7.496240601503761e-06,
+      "loss": 0.2999,
+      "step": 16650
+    },
+    {
+      "epoch": 25.05,
+      "grad_norm": 3.5126192569732666,
+      "learning_rate": 7.494736842105263e-06,
+      "loss": 0.2304,
+      "step": 16660
+    },
+    {
+      "epoch": 25.07,
+      "grad_norm": 8.772309303283691,
+      "learning_rate": 7.493233082706768e-06,
+      "loss": 0.3678,
+      "step": 16670
+    },
+    {
+      "epoch": 25.08,
+      "grad_norm": 6.1839189529418945,
+      "learning_rate": 7.491729323308271e-06,
+      "loss": 0.3029,
+      "step": 16680
+    },
+    {
+      "epoch": 25.1,
+      "grad_norm": 3.8670859336853027,
+      "learning_rate": 7.490225563909776e-06,
+      "loss": 0.2816,
+      "step": 16690
+    },
+    {
+      "epoch": 25.11,
+      "grad_norm": 6.036872863769531,
+      "learning_rate": 7.488721804511278e-06,
+      "loss": 0.3167,
+      "step": 16700
+    },
+    {
+      "epoch": 25.13,
+      "grad_norm": 5.029440879821777,
+      "learning_rate": 7.487218045112783e-06,
+      "loss": 0.3886,
+      "step": 16710
+    },
+    {
+      "epoch": 25.14,
+      "grad_norm": 6.317779541015625,
+      "learning_rate": 7.485714285714286e-06,
+      "loss": 0.3374,
+      "step": 16720
+    },
+    {
+      "epoch": 25.16,
+      "grad_norm": 4.726473808288574,
+      "learning_rate": 7.4842105263157905e-06,
+      "loss": 0.2748,
+      "step": 16730
+    },
+    {
+      "epoch": 25.17,
+      "grad_norm": 7.509809494018555,
+      "learning_rate": 7.482706766917294e-06,
+      "loss": 0.3124,
+      "step": 16740
+    },
+    {
+      "epoch": 25.19,
+      "grad_norm": 6.474162578582764,
+      "learning_rate": 7.481203007518798e-06,
+      "loss": 0.3249,
+      "step": 16750
+    },
+    {
+      "epoch": 25.2,
+      "grad_norm": 5.663277626037598,
+      "learning_rate": 7.479699248120301e-06,
+      "loss": 0.3175,
+      "step": 16760
+    },
+    {
+      "epoch": 25.22,
+      "grad_norm": 5.958733558654785,
+      "learning_rate": 7.4781954887218055e-06,
+      "loss": 0.2671,
+      "step": 16770
+    },
+    {
+      "epoch": 25.23,
+      "grad_norm": 3.8807878494262695,
+      "learning_rate": 7.476691729323309e-06,
+      "loss": 0.2347,
+      "step": 16780
+    },
+    {
+      "epoch": 25.25,
+      "grad_norm": 8.403759956359863,
+      "learning_rate": 7.475187969924813e-06,
+      "loss": 0.2834,
+      "step": 16790
+    },
+    {
+      "epoch": 25.26,
+      "grad_norm": 4.284714698791504,
+      "learning_rate": 7.473684210526316e-06,
+      "loss": 0.2763,
+      "step": 16800
+    },
+    {
+      "epoch": 25.28,
+      "grad_norm": 5.796482563018799,
+      "learning_rate": 7.47218045112782e-06,
+      "loss": 0.3044,
+      "step": 16810
+    },
+    {
+      "epoch": 25.29,
+      "grad_norm": 4.833478927612305,
+      "learning_rate": 7.470676691729324e-06,
+      "loss": 0.3536,
+      "step": 16820
+    },
+    {
+      "epoch": 25.31,
+      "grad_norm": 6.935545921325684,
+      "learning_rate": 7.469172932330828e-06,
+      "loss": 0.3231,
+      "step": 16830
+    },
+    {
+      "epoch": 25.32,
+      "grad_norm": 7.3656229972839355,
+      "learning_rate": 7.467669172932332e-06,
+      "loss": 0.2747,
+      "step": 16840
+    },
+    {
+      "epoch": 25.34,
+      "grad_norm": 6.294661998748779,
+      "learning_rate": 7.466165413533836e-06,
+      "loss": 0.3172,
+      "step": 16850
+    },
+    {
+      "epoch": 25.35,
+      "grad_norm": 5.849190711975098,
+      "learning_rate": 7.464661654135339e-06,
+      "loss": 0.307,
+      "step": 16860
+    },
+    {
+      "epoch": 25.37,
+      "grad_norm": 3.9488296508789062,
+      "learning_rate": 7.463157894736843e-06,
+      "loss": 0.2835,
+      "step": 16870
+    },
+    {
+      "epoch": 25.38,
+      "grad_norm": 5.17800760269165,
+      "learning_rate": 7.461654135338347e-06,
+      "loss": 0.3037,
+      "step": 16880
+    },
+    {
+      "epoch": 25.4,
+      "grad_norm": 5.4995436668396,
+      "learning_rate": 7.460150375939849e-06,
+      "loss": 0.3726,
+      "step": 16890
+    },
+    {
+      "epoch": 25.41,
+      "grad_norm": 5.913205623626709,
+      "learning_rate": 7.458646616541354e-06,
+      "loss": 0.244,
+      "step": 16900
+    },
+    {
+      "epoch": 25.43,
+      "grad_norm": 7.40120267868042,
+      "learning_rate": 7.457142857142857e-06,
+      "loss": 0.3658,
+      "step": 16910
+    },
+    {
+      "epoch": 25.44,
+      "grad_norm": 6.155134677886963,
+      "learning_rate": 7.455639097744362e-06,
+      "loss": 0.2988,
+      "step": 16920
+    },
+    {
+      "epoch": 25.46,
+      "grad_norm": 7.154820919036865,
+      "learning_rate": 7.454135338345865e-06,
+      "loss": 0.2828,
+      "step": 16930
+    },
+    {
+      "epoch": 25.47,
+      "grad_norm": 15.181897163391113,
+      "learning_rate": 7.4526315789473695e-06,
+      "loss": 0.2919,
+      "step": 16940
+    },
+    {
+      "epoch": 25.49,
+      "grad_norm": 15.616399765014648,
+      "learning_rate": 7.451127819548872e-06,
+      "loss": 0.2602,
+      "step": 16950
+    },
+    {
+      "epoch": 25.5,
+      "grad_norm": 3.783785104751587,
+      "learning_rate": 7.4496240601503765e-06,
+      "loss": 0.2663,
+      "step": 16960
+    },
+    {
+      "epoch": 25.52,
+      "grad_norm": 3.7820167541503906,
+      "learning_rate": 7.44812030075188e-06,
+      "loss": 0.3069,
+      "step": 16970
+    },
+    {
+      "epoch": 25.53,
+      "grad_norm": 22.630653381347656,
+      "learning_rate": 7.4466165413533844e-06,
+      "loss": 0.3701,
+      "step": 16980
+    },
+    {
+      "epoch": 25.55,
+      "grad_norm": 7.006454944610596,
+      "learning_rate": 7.445112781954887e-06,
+      "loss": 0.2762,
+      "step": 16990
+    },
+    {
+      "epoch": 25.56,
+      "grad_norm": 3.897531747817993,
+      "learning_rate": 7.4436090225563915e-06,
+      "loss": 0.2871,
+      "step": 17000
+    },
+    {
+      "epoch": 25.58,
+      "grad_norm": 9.737010955810547,
+      "learning_rate": 7.442105263157895e-06,
+      "loss": 0.3115,
+      "step": 17010
+    },
+    {
+      "epoch": 25.59,
+      "grad_norm": 9.285514831542969,
+      "learning_rate": 7.440601503759399e-06,
+      "loss": 0.3061,
+      "step": 17020
+    },
+    {
+      "epoch": 25.61,
+      "grad_norm": 6.54390811920166,
+      "learning_rate": 7.439097744360903e-06,
+      "loss": 0.4048,
+      "step": 17030
+    },
+    {
+      "epoch": 25.62,
+      "grad_norm": 7.407090663909912,
+      "learning_rate": 7.437593984962406e-06,
+      "loss": 0.335,
+      "step": 17040
+    },
+    {
+      "epoch": 25.64,
+      "grad_norm": 5.179807186126709,
+      "learning_rate": 7.43609022556391e-06,
+      "loss": 0.3313,
+      "step": 17050
+    },
+    {
+      "epoch": 25.65,
+      "grad_norm": 3.106466770172119,
+      "learning_rate": 7.434586466165414e-06,
+      "loss": 0.3702,
+      "step": 17060
+    },
+    {
+      "epoch": 25.67,
+      "grad_norm": 4.030908584594727,
+      "learning_rate": 7.433082706766918e-06,
+      "loss": 0.262,
+      "step": 17070
+    },
+    {
+      "epoch": 25.68,
+      "grad_norm": 5.041976451873779,
+      "learning_rate": 7.431578947368422e-06,
+      "loss": 0.3317,
+      "step": 17080
+    },
+    {
+      "epoch": 25.7,
+      "grad_norm": 7.29601526260376,
+      "learning_rate": 7.430075187969925e-06,
+      "loss": 0.295,
+      "step": 17090
+    },
+    {
+      "epoch": 25.71,
+      "grad_norm": 6.026291370391846,
+      "learning_rate": 7.428571428571429e-06,
+      "loss": 0.3131,
+      "step": 17100
+    },
+    {
+      "epoch": 25.73,
+      "grad_norm": 6.111357688903809,
+      "learning_rate": 7.427067669172933e-06,
+      "loss": 0.2887,
+      "step": 17110
+    },
+    {
+      "epoch": 25.74,
+      "grad_norm": 6.1451520919799805,
+      "learning_rate": 7.425563909774437e-06,
+      "loss": 0.2974,
+      "step": 17120
+    },
+    {
+      "epoch": 25.76,
+      "grad_norm": 4.193663120269775,
+      "learning_rate": 7.424060150375941e-06,
+      "loss": 0.2106,
+      "step": 17130
+    },
+    {
+      "epoch": 25.77,
+      "grad_norm": 5.288703441619873,
+      "learning_rate": 7.422556390977444e-06,
+      "loss": 0.2647,
+      "step": 17140
+    },
+    {
+      "epoch": 25.79,
+      "grad_norm": 9.568791389465332,
+      "learning_rate": 7.421052631578948e-06,
+      "loss": 0.383,
+      "step": 17150
+    },
+    {
+      "epoch": 25.8,
+      "grad_norm": 4.507370471954346,
+      "learning_rate": 7.419548872180452e-06,
+      "loss": 0.3109,
+      "step": 17160
+    },
+    {
+      "epoch": 25.82,
+      "grad_norm": 4.298795223236084,
+      "learning_rate": 7.4180451127819555e-06,
+      "loss": 0.3037,
+      "step": 17170
+    },
+    {
+      "epoch": 25.83,
+      "grad_norm": 7.694872856140137,
+      "learning_rate": 7.41654135338346e-06,
+      "loss": 0.353,
+      "step": 17180
+    },
+    {
+      "epoch": 25.85,
+      "grad_norm": 4.656931400299072,
+      "learning_rate": 7.4150375939849626e-06,
+      "loss": 0.3047,
+      "step": 17190
+    },
+    {
+      "epoch": 25.86,
+      "grad_norm": 6.448072910308838,
+      "learning_rate": 7.413533834586467e-06,
+      "loss": 0.3562,
+      "step": 17200
+    },
+    {
+      "epoch": 25.88,
+      "grad_norm": 7.515552043914795,
+      "learning_rate": 7.4120300751879705e-06,
+      "loss": 0.3391,
+      "step": 17210
+    },
+    {
+      "epoch": 25.89,
+      "grad_norm": 9.326400756835938,
+      "learning_rate": 7.410526315789475e-06,
+      "loss": 0.2554,
+      "step": 17220
+    },
+    {
+      "epoch": 25.91,
+      "grad_norm": 4.4977498054504395,
+      "learning_rate": 7.4090225563909775e-06,
+      "loss": 0.3115,
+      "step": 17230
+    },
+    {
+      "epoch": 25.92,
+      "grad_norm": 5.859455585479736,
+      "learning_rate": 7.407518796992482e-06,
+      "loss": 0.2923,
+      "step": 17240
+    },
+    {
+      "epoch": 25.94,
+      "grad_norm": 3.0939574241638184,
+      "learning_rate": 7.406015037593985e-06,
+      "loss": 0.3387,
+      "step": 17250
+    },
+    {
+      "epoch": 25.95,
+      "grad_norm": 6.690404415130615,
+      "learning_rate": 7.40451127819549e-06,
+      "loss": 0.2812,
+      "step": 17260
+    },
+    {
+      "epoch": 25.97,
+      "grad_norm": 3.768836736679077,
+      "learning_rate": 7.403007518796993e-06,
+      "loss": 0.2778,
+      "step": 17270
+    },
+    {
+      "epoch": 25.98,
+      "grad_norm": 4.476847171783447,
+      "learning_rate": 7.401503759398498e-06,
+      "loss": 0.3684,
+      "step": 17280
+    },
+    {
+      "epoch": 26.0,
+      "grad_norm": 0.6643197536468506,
+      "learning_rate": 7.4e-06,
+      "loss": 0.1884,
+      "step": 17290
+    },
+    {
+      "epoch": 26.0,
+      "eval_accuracy": 0.9304,
+      "eval_loss": 0.2943102717399597,
+      "eval_runtime": 84.682,
+      "eval_samples_per_second": 118.089,
+      "eval_steps_per_second": 0.472,
+      "step": 17290
+    },
+    {
+      "epoch": 26.02,
+      "grad_norm": 6.030354976654053,
+      "learning_rate": 7.398496240601505e-06,
+      "loss": 0.2949,
+      "step": 17300
+    },
+    {
+      "epoch": 26.03,
+      "grad_norm": 6.9479193687438965,
+      "learning_rate": 7.396992481203008e-06,
+      "loss": 0.3083,
+      "step": 17310
+    },
+    {
+      "epoch": 26.05,
+      "grad_norm": 5.633476257324219,
+      "learning_rate": 7.3954887218045126e-06,
+      "loss": 0.2906,
+      "step": 17320
+    },
+    {
+      "epoch": 26.06,
+      "grad_norm": 4.715734481811523,
+      "learning_rate": 7.393984962406015e-06,
+      "loss": 0.3034,
+      "step": 17330
+    },
+    {
+      "epoch": 26.08,
+      "grad_norm": 5.452293395996094,
+      "learning_rate": 7.39248120300752e-06,
+      "loss": 0.3242,
+      "step": 17340
+    },
+    {
+      "epoch": 26.09,
+      "grad_norm": 5.496037006378174,
+      "learning_rate": 7.390977443609023e-06,
+      "loss": 0.2861,
+      "step": 17350
+    },
+    {
+      "epoch": 26.11,
+      "grad_norm": 8.563288688659668,
+      "learning_rate": 7.3894736842105275e-06,
+      "loss": 0.3451,
+      "step": 17360
+    },
+    {
+      "epoch": 26.12,
+      "grad_norm": 3.828183889389038,
+      "learning_rate": 7.387969924812031e-06,
+      "loss": 0.3346,
+      "step": 17370
+    },
+    {
+      "epoch": 26.14,
+      "grad_norm": 6.062718391418457,
+      "learning_rate": 7.386466165413534e-06,
+      "loss": 0.2778,
+      "step": 17380
+    },
+    {
+      "epoch": 26.15,
+      "grad_norm": 5.245468616485596,
+      "learning_rate": 7.384962406015038e-06,
+      "loss": 0.2623,
+      "step": 17390
+    },
+    {
+      "epoch": 26.17,
+      "grad_norm": 5.32465934753418,
+      "learning_rate": 7.3834586466165416e-06,
+      "loss": 0.2771,
+      "step": 17400
+    },
+    {
+      "epoch": 26.18,
+      "grad_norm": 6.1053009033203125,
+      "learning_rate": 7.381954887218046e-06,
+      "loss": 0.2738,
+      "step": 17410
+    },
+    {
+      "epoch": 26.2,
+      "grad_norm": 4.224456310272217,
+      "learning_rate": 7.380451127819549e-06,
+      "loss": 0.2576,
+      "step": 17420
+    },
+    {
+      "epoch": 26.21,
+      "grad_norm": 5.59401798248291,
+      "learning_rate": 7.378947368421053e-06,
+      "loss": 0.2733,
+      "step": 17430
+    },
+    {
+      "epoch": 26.23,
+      "grad_norm": 9.04089641571045,
+      "learning_rate": 7.3774436090225565e-06,
+      "loss": 0.294,
+      "step": 17440
+    },
+    {
+      "epoch": 26.24,
+      "grad_norm": 5.4782233238220215,
+      "learning_rate": 7.375939849624061e-06,
+      "loss": 0.2776,
+      "step": 17450
+    },
+    {
+      "epoch": 26.26,
+      "grad_norm": 3.4573464393615723,
+      "learning_rate": 7.374436090225564e-06,
+      "loss": 0.2913,
+      "step": 17460
+    },
+    {
+      "epoch": 26.27,
+      "grad_norm": 4.618410587310791,
+      "learning_rate": 7.372932330827069e-06,
+      "loss": 0.3398,
+      "step": 17470
+    },
+    {
+      "epoch": 26.29,
+      "grad_norm": 3.339372396469116,
+      "learning_rate": 7.371428571428571e-06,
+      "loss": 0.2932,
+      "step": 17480
+    },
+    {
+      "epoch": 26.3,
+      "grad_norm": 4.173111438751221,
+      "learning_rate": 7.369924812030076e-06,
+      "loss": 0.3159,
+      "step": 17490
+    },
+    {
+      "epoch": 26.32,
+      "grad_norm": 5.077332973480225,
+      "learning_rate": 7.368421052631579e-06,
+      "loss": 0.2747,
+      "step": 17500
+    },
+    {
+      "epoch": 26.33,
+      "grad_norm": 3.467737913131714,
+      "learning_rate": 7.366917293233084e-06,
+      "loss": 0.3015,
+      "step": 17510
+    },
+    {
+      "epoch": 26.35,
+      "grad_norm": 4.046982288360596,
+      "learning_rate": 7.365413533834586e-06,
+      "loss": 0.2847,
+      "step": 17520
+    },
+    {
+      "epoch": 26.36,
+      "grad_norm": 5.868749141693115,
+      "learning_rate": 7.363909774436091e-06,
+      "loss": 0.2743,
+      "step": 17530
+    },
+    {
+      "epoch": 26.38,
+      "grad_norm": 4.371850967407227,
+      "learning_rate": 7.362406015037594e-06,
+      "loss": 0.2951,
+      "step": 17540
+    },
+    {
+      "epoch": 26.39,
+      "grad_norm": 5.016296863555908,
+      "learning_rate": 7.3609022556390986e-06,
+      "loss": 0.3517,
+      "step": 17550
+    },
+    {
+      "epoch": 26.41,
+      "grad_norm": 4.3615946769714355,
+      "learning_rate": 7.359398496240602e-06,
+      "loss": 0.4145,
+      "step": 17560
+    },
+    {
+      "epoch": 26.42,
+      "grad_norm": 5.725963115692139,
+      "learning_rate": 7.3578947368421065e-06,
+      "loss": 0.2518,
+      "step": 17570
+    },
+    {
+      "epoch": 26.44,
+      "grad_norm": 9.736555099487305,
+      "learning_rate": 7.356390977443609e-06,
+      "loss": 0.3154,
+      "step": 17580
+    },
+    {
+      "epoch": 26.45,
+      "grad_norm": 7.777997016906738,
+      "learning_rate": 7.3548872180451135e-06,
+      "loss": 0.325,
+      "step": 17590
+    },
+    {
+      "epoch": 26.47,
+      "grad_norm": 4.7586894035339355,
+      "learning_rate": 7.353383458646617e-06,
+      "loss": 0.3155,
+      "step": 17600
+    },
+    {
+      "epoch": 26.48,
+      "grad_norm": 5.5998430252075195,
+      "learning_rate": 7.351879699248121e-06,
+      "loss": 0.3358,
+      "step": 17610
+    },
+    {
+      "epoch": 26.5,
+      "grad_norm": 6.661489486694336,
+      "learning_rate": 7.350375939849624e-06,
+      "loss": 0.2973,
+      "step": 17620
+    },
+    {
+      "epoch": 26.51,
+      "grad_norm": 8.244695663452148,
+      "learning_rate": 7.348872180451128e-06,
+      "loss": 0.2739,
+      "step": 17630
+    },
+    {
+      "epoch": 26.53,
+      "grad_norm": 7.346087455749512,
+      "learning_rate": 7.347368421052632e-06,
+      "loss": 0.3281,
+      "step": 17640
+    },
+    {
+      "epoch": 26.54,
+      "grad_norm": 8.243110656738281,
+      "learning_rate": 7.345864661654136e-06,
+      "loss": 0.2994,
+      "step": 17650
+    },
+    {
+      "epoch": 26.56,
+      "grad_norm": 3.7893612384796143,
+      "learning_rate": 7.34436090225564e-06,
+      "loss": 0.3569,
+      "step": 17660
+    },
+    {
+      "epoch": 26.57,
+      "grad_norm": 4.714302062988281,
+      "learning_rate": 7.342857142857144e-06,
+      "loss": 0.2856,
+      "step": 17670
+    },
+    {
+      "epoch": 26.59,
+      "grad_norm": 7.72084903717041,
+      "learning_rate": 7.341353383458647e-06,
+      "loss": 0.2402,
+      "step": 17680
+    },
+    {
+      "epoch": 26.6,
+      "grad_norm": 4.239202499389648,
+      "learning_rate": 7.339849624060151e-06,
+      "loss": 0.2312,
+      "step": 17690
+    },
+    {
+      "epoch": 26.62,
+      "grad_norm": 6.118954658508301,
+      "learning_rate": 7.338345864661655e-06,
+      "loss": 0.3843,
+      "step": 17700
+    },
+    {
+      "epoch": 26.63,
+      "grad_norm": 6.067955017089844,
+      "learning_rate": 7.336842105263159e-06,
+      "loss": 0.3087,
+      "step": 17710
+    },
+    {
+      "epoch": 26.65,
+      "grad_norm": 6.024227619171143,
+      "learning_rate": 7.335338345864662e-06,
+      "loss": 0.2224,
+      "step": 17720
+    },
+    {
+      "epoch": 26.66,
+      "grad_norm": 3.350494146347046,
+      "learning_rate": 7.333834586466166e-06,
+      "loss": 0.2977,
+      "step": 17730
+    },
+    {
+      "epoch": 26.68,
+      "grad_norm": 5.893447399139404,
+      "learning_rate": 7.33233082706767e-06,
+      "loss": 0.2636,
+      "step": 17740
+    },
+    {
+      "epoch": 26.69,
+      "grad_norm": 5.77102518081665,
+      "learning_rate": 7.330827067669174e-06,
+      "loss": 0.2412,
+      "step": 17750
+    },
+    {
+      "epoch": 26.71,
+      "grad_norm": 4.75433349609375,
+      "learning_rate": 7.3293233082706776e-06,
+      "loss": 0.3262,
+      "step": 17760
+    },
+    {
+      "epoch": 26.72,
+      "grad_norm": 9.549809455871582,
+      "learning_rate": 7.327819548872182e-06,
+      "loss": 0.2673,
+      "step": 17770
+    },
+    {
+      "epoch": 26.74,
+      "grad_norm": 7.770995616912842,
+      "learning_rate": 7.326315789473685e-06,
+      "loss": 0.3227,
+      "step": 17780
+    },
+    {
+      "epoch": 26.75,
+      "grad_norm": 2.771538257598877,
+      "learning_rate": 7.324812030075189e-06,
+      "loss": 0.2894,
+      "step": 17790
+    },
+    {
+      "epoch": 26.77,
+      "grad_norm": 5.751589298248291,
+      "learning_rate": 7.3233082706766925e-06,
+      "loss": 0.281,
+      "step": 17800
+    },
+    {
+      "epoch": 26.78,
+      "grad_norm": 6.880566596984863,
+      "learning_rate": 7.321804511278197e-06,
+      "loss": 0.3191,
+      "step": 17810
+    },
+    {
+      "epoch": 26.8,
+      "grad_norm": 5.7594218254089355,
+      "learning_rate": 7.3203007518796995e-06,
+      "loss": 0.2659,
+      "step": 17820
+    },
+    {
+      "epoch": 26.81,
+      "grad_norm": 3.8072729110717773,
+      "learning_rate": 7.318796992481204e-06,
+      "loss": 0.2825,
+      "step": 17830
+    },
+    {
+      "epoch": 26.83,
+      "grad_norm": 7.311132431030273,
+      "learning_rate": 7.317293233082707e-06,
+      "loss": 0.3758,
+      "step": 17840
+    },
+    {
+      "epoch": 26.84,
+      "grad_norm": 6.675231456756592,
+      "learning_rate": 7.315789473684212e-06,
+      "loss": 0.2856,
+      "step": 17850
+    },
+    {
+      "epoch": 26.86,
+      "grad_norm": 9.13115119934082,
+      "learning_rate": 7.314285714285715e-06,
+      "loss": 0.266,
+      "step": 17860
+    },
+    {
+      "epoch": 26.87,
+      "grad_norm": 5.860391616821289,
+      "learning_rate": 7.312781954887218e-06,
+      "loss": 0.3354,
+      "step": 17870
+    },
+    {
+      "epoch": 26.89,
+      "grad_norm": 5.685858249664307,
+      "learning_rate": 7.311278195488722e-06,
+      "loss": 0.3235,
+      "step": 17880
+    },
+    {
+      "epoch": 26.9,
+      "grad_norm": 6.518139839172363,
+      "learning_rate": 7.309774436090226e-06,
+      "loss": 0.313,
+      "step": 17890
+    },
+    {
+      "epoch": 26.92,
+      "grad_norm": 14.772072792053223,
+      "learning_rate": 7.30827067669173e-06,
+      "loss": 0.3201,
+      "step": 17900
+    },
+    {
+      "epoch": 26.93,
+      "grad_norm": 4.325756072998047,
+      "learning_rate": 7.306766917293233e-06,
+      "loss": 0.2681,
+      "step": 17910
+    },
+    {
+      "epoch": 26.95,
+      "grad_norm": 4.291186809539795,
+      "learning_rate": 7.305263157894737e-06,
+      "loss": 0.3831,
+      "step": 17920
+    },
+    {
+      "epoch": 26.96,
+      "grad_norm": 8.093550682067871,
+      "learning_rate": 7.303759398496241e-06,
+      "loss": 0.2855,
+      "step": 17930
+    },
+    {
+      "epoch": 26.98,
+      "grad_norm": 4.962594985961914,
+      "learning_rate": 7.302255639097745e-06,
+      "loss": 0.2369,
+      "step": 17940
+    },
+    {
+      "epoch": 26.99,
+      "grad_norm": 5.591581344604492,
+      "learning_rate": 7.300751879699249e-06,
+      "loss": 0.3624,
+      "step": 17950
+    },
+    {
+      "epoch": 27.0,
+      "eval_accuracy": 0.9316,
+      "eval_loss": 0.2865539491176605,
+      "eval_runtime": 84.8448,
+      "eval_samples_per_second": 117.862,
+      "eval_steps_per_second": 0.471,
+      "step": 17955
+    },
+    {
+      "epoch": 27.01,
+      "grad_norm": 4.0642547607421875,
+      "learning_rate": 7.299248120300752e-06,
+      "loss": 0.2308,
+      "step": 17960
+    },
+    {
+      "epoch": 27.02,
+      "grad_norm": 5.927849292755127,
+      "learning_rate": 7.297744360902256e-06,
+      "loss": 0.2986,
+      "step": 17970
+    },
+    {
+      "epoch": 27.04,
+      "grad_norm": 6.369534492492676,
+      "learning_rate": 7.29624060150376e-06,
+      "loss": 0.2482,
+      "step": 17980
+    },
+    {
+      "epoch": 27.05,
+      "grad_norm": 4.340782165527344,
+      "learning_rate": 7.2947368421052636e-06,
+      "loss": 0.2583,
+      "step": 17990
+    },
+    {
+      "epoch": 27.07,
+      "grad_norm": 4.993666648864746,
+      "learning_rate": 7.293233082706768e-06,
+      "loss": 0.237,
+      "step": 18000
+    },
+    {
+      "epoch": 27.08,
+      "grad_norm": 5.864019870758057,
+      "learning_rate": 7.291729323308271e-06,
+      "loss": 0.3072,
+      "step": 18010
+    },
+    {
+      "epoch": 27.1,
+      "grad_norm": 6.993382930755615,
+      "learning_rate": 7.290225563909775e-06,
+      "loss": 0.2972,
+      "step": 18020
+    },
+    {
+      "epoch": 27.11,
+      "grad_norm": 5.060710430145264,
+      "learning_rate": 7.2887218045112785e-06,
+      "loss": 0.269,
+      "step": 18030
+    },
+    {
+      "epoch": 27.13,
+      "grad_norm": 3.3099091053009033,
+      "learning_rate": 7.287218045112783e-06,
+      "loss": 0.3175,
+      "step": 18040
+    },
+    {
+      "epoch": 27.14,
+      "grad_norm": 5.775263786315918,
+      "learning_rate": 7.285714285714286e-06,
+      "loss": 0.2844,
+      "step": 18050
+    },
+    {
+      "epoch": 27.16,
+      "grad_norm": 5.468781471252441,
+      "learning_rate": 7.28421052631579e-06,
+      "loss": 0.2273,
+      "step": 18060
+    },
+    {
+      "epoch": 27.17,
+      "grad_norm": 6.351372241973877,
+      "learning_rate": 7.282706766917293e-06,
+      "loss": 0.3525,
+      "step": 18070
+    },
+    {
+      "epoch": 27.19,
+      "grad_norm": 7.59521484375,
+      "learning_rate": 7.281203007518798e-06,
+      "loss": 0.3068,
+      "step": 18080
+    },
+    {
+      "epoch": 27.2,
+      "grad_norm": 3.0641674995422363,
+      "learning_rate": 7.279699248120301e-06,
+      "loss": 0.2865,
+      "step": 18090
+    },
+    {
+      "epoch": 27.22,
+      "grad_norm": 6.9775238037109375,
+      "learning_rate": 7.278195488721806e-06,
+      "loss": 0.3022,
+      "step": 18100
+    },
+    {
+      "epoch": 27.23,
+      "grad_norm": 4.060088634490967,
+      "learning_rate": 7.276691729323308e-06,
+      "loss": 0.3148,
+      "step": 18110
+    },
+    {
+      "epoch": 27.25,
+      "grad_norm": 7.767538070678711,
+      "learning_rate": 7.275187969924813e-06,
+      "loss": 0.229,
+      "step": 18120
+    },
+    {
+      "epoch": 27.26,
+      "grad_norm": 4.900406360626221,
+      "learning_rate": 7.273684210526316e-06,
+      "loss": 0.2672,
+      "step": 18130
+    },
+    {
+      "epoch": 27.28,
+      "grad_norm": 6.7180962562561035,
+      "learning_rate": 7.272180451127821e-06,
+      "loss": 0.24,
+      "step": 18140
+    },
+    {
+      "epoch": 27.29,
+      "grad_norm": 6.126708507537842,
+      "learning_rate": 7.270676691729323e-06,
+      "loss": 0.284,
+      "step": 18150
+    },
+    {
+      "epoch": 27.31,
+      "grad_norm": 3.9110794067382812,
+      "learning_rate": 7.269172932330828e-06,
+      "loss": 0.2719,
+      "step": 18160
+    },
+    {
+      "epoch": 27.32,
+      "grad_norm": 6.212155342102051,
+      "learning_rate": 7.267669172932331e-06,
+      "loss": 0.3151,
+      "step": 18170
+    },
+    {
+      "epoch": 27.34,
+      "grad_norm": 3.0043139457702637,
+      "learning_rate": 7.2661654135338355e-06,
+      "loss": 0.2592,
+      "step": 18180
+    },
+    {
+      "epoch": 27.35,
+      "grad_norm": 3.2263989448547363,
+      "learning_rate": 7.264661654135339e-06,
+      "loss": 0.3309,
+      "step": 18190
+    },
+    {
+      "epoch": 27.37,
+      "grad_norm": 4.7933197021484375,
+      "learning_rate": 7.263157894736843e-06,
+      "loss": 0.3262,
+      "step": 18200
+    },
+    {
+      "epoch": 27.38,
+      "grad_norm": 2.227823495864868,
+      "learning_rate": 7.261654135338346e-06,
+      "loss": 0.2645,
+      "step": 18210
+    },
+    {
+      "epoch": 27.4,
+      "grad_norm": 3.900000810623169,
+      "learning_rate": 7.2601503759398504e-06,
+      "loss": 0.3359,
+      "step": 18220
+    },
+    {
+      "epoch": 27.41,
+      "grad_norm": 3.5033364295959473,
+      "learning_rate": 7.258646616541354e-06,
+      "loss": 0.3467,
+      "step": 18230
+    },
+    {
+      "epoch": 27.43,
+      "grad_norm": 6.635146617889404,
+      "learning_rate": 7.257142857142858e-06,
+      "loss": 0.3276,
+      "step": 18240
+    },
+    {
+      "epoch": 27.44,
+      "grad_norm": 4.950628757476807,
+      "learning_rate": 7.255639097744361e-06,
+      "loss": 0.2459,
+      "step": 18250
+    },
+    {
+      "epoch": 27.46,
+      "grad_norm": 8.33364486694336,
+      "learning_rate": 7.254135338345865e-06,
+      "loss": 0.3154,
+      "step": 18260
+    },
+    {
+      "epoch": 27.47,
+      "grad_norm": 6.362850666046143,
+      "learning_rate": 7.252631578947369e-06,
+      "loss": 0.3349,
+      "step": 18270
+    },
+    {
+      "epoch": 27.49,
+      "grad_norm": 2.358121395111084,
+      "learning_rate": 7.251127819548873e-06,
+      "loss": 0.247,
+      "step": 18280
+    },
+    {
+      "epoch": 27.5,
+      "grad_norm": 7.640157699584961,
+      "learning_rate": 7.249624060150377e-06,
+      "loss": 0.259,
+      "step": 18290
+    },
+    {
+      "epoch": 27.52,
+      "grad_norm": 4.780099868774414,
+      "learning_rate": 7.248120300751881e-06,
+      "loss": 0.2818,
+      "step": 18300
+    },
+    {
+      "epoch": 27.53,
+      "grad_norm": 4.6383256912231445,
+      "learning_rate": 7.246616541353384e-06,
+      "loss": 0.2534,
+      "step": 18310
+    },
+    {
+      "epoch": 27.55,
+      "grad_norm": 4.298056125640869,
+      "learning_rate": 7.245112781954888e-06,
+      "loss": 0.2586,
+      "step": 18320
+    },
+    {
+      "epoch": 27.56,
+      "grad_norm": 7.920955181121826,
+      "learning_rate": 7.243609022556392e-06,
+      "loss": 0.2583,
+      "step": 18330
+    },
+    {
+      "epoch": 27.58,
+      "grad_norm": 2.8801300525665283,
+      "learning_rate": 7.242105263157896e-06,
+      "loss": 0.2478,
+      "step": 18340
+    },
+    {
+      "epoch": 27.59,
+      "grad_norm": 5.475898265838623,
+      "learning_rate": 7.240601503759399e-06,
+      "loss": 0.3193,
+      "step": 18350
+    },
+    {
+      "epoch": 27.61,
+      "grad_norm": 26.629573822021484,
+      "learning_rate": 7.239097744360903e-06,
+      "loss": 0.2725,
+      "step": 18360
+    },
+    {
+      "epoch": 27.62,
+      "grad_norm": 5.835658073425293,
+      "learning_rate": 7.237593984962407e-06,
+      "loss": 0.274,
+      "step": 18370
+    },
+    {
+      "epoch": 27.64,
+      "grad_norm": 7.282139301300049,
+      "learning_rate": 7.23609022556391e-06,
+      "loss": 0.288,
+      "step": 18380
+    },
+    {
+      "epoch": 27.65,
+      "grad_norm": 11.25385856628418,
+      "learning_rate": 7.2345864661654145e-06,
+      "loss": 0.2611,
+      "step": 18390
+    },
+    {
+      "epoch": 27.67,
+      "grad_norm": 6.181314945220947,
+      "learning_rate": 7.233082706766917e-06,
+      "loss": 0.3109,
+      "step": 18400
+    },
+    {
+      "epoch": 27.68,
+      "grad_norm": 5.535643100738525,
+      "learning_rate": 7.2315789473684215e-06,
+      "loss": 0.2892,
+      "step": 18410
+    },
+    {
+      "epoch": 27.7,
+      "grad_norm": 7.809751510620117,
+      "learning_rate": 7.230075187969925e-06,
+      "loss": 0.3626,
+      "step": 18420
+    },
+    {
+      "epoch": 27.71,
+      "grad_norm": 5.477492332458496,
+      "learning_rate": 7.2285714285714294e-06,
+      "loss": 0.3423,
+      "step": 18430
+    },
+    {
+      "epoch": 27.73,
+      "grad_norm": 5.311155796051025,
+      "learning_rate": 7.227067669172932e-06,
+      "loss": 0.2859,
+      "step": 18440
+    },
+    {
+      "epoch": 27.74,
+      "grad_norm": 6.58491325378418,
+      "learning_rate": 7.2255639097744365e-06,
+      "loss": 0.3323,
+      "step": 18450
+    },
+    {
+      "epoch": 27.76,
+      "grad_norm": 10.060500144958496,
+      "learning_rate": 7.22406015037594e-06,
+      "loss": 0.2803,
+      "step": 18460
+    },
+    {
+      "epoch": 27.77,
+      "grad_norm": 3.332087278366089,
+      "learning_rate": 7.222556390977444e-06,
+      "loss": 0.3011,
+      "step": 18470
+    },
+    {
+      "epoch": 27.79,
+      "grad_norm": 3.494025945663452,
+      "learning_rate": 7.221052631578948e-06,
+      "loss": 0.2756,
+      "step": 18480
+    },
+    {
+      "epoch": 27.8,
+      "grad_norm": 6.302395820617676,
+      "learning_rate": 7.219548872180452e-06,
+      "loss": 0.2454,
+      "step": 18490
+    },
+    {
+      "epoch": 27.82,
+      "grad_norm": 4.8681793212890625,
+      "learning_rate": 7.218045112781955e-06,
+      "loss": 0.349,
+      "step": 18500
+    },
+    {
+      "epoch": 27.83,
+      "grad_norm": 5.359793663024902,
+      "learning_rate": 7.216541353383459e-06,
+      "loss": 0.3316,
+      "step": 18510
+    },
+    {
+      "epoch": 27.85,
+      "grad_norm": 7.901219367980957,
+      "learning_rate": 7.215037593984963e-06,
+      "loss": 0.255,
+      "step": 18520
+    },
+    {
+      "epoch": 27.86,
+      "grad_norm": 6.270127773284912,
+      "learning_rate": 7.213533834586467e-06,
+      "loss": 0.3856,
+      "step": 18530
+    },
+    {
+      "epoch": 27.88,
+      "grad_norm": 3.3462986946105957,
+      "learning_rate": 7.21203007518797e-06,
+      "loss": 0.2656,
+      "step": 18540
+    },
+    {
+      "epoch": 27.89,
+      "grad_norm": 5.023731708526611,
+      "learning_rate": 7.210526315789474e-06,
+      "loss": 0.3367,
+      "step": 18550
+    },
+    {
+      "epoch": 27.91,
+      "grad_norm": 7.475311756134033,
+      "learning_rate": 7.209022556390978e-06,
+      "loss": 0.3652,
+      "step": 18560
+    },
+    {
+      "epoch": 27.92,
+      "grad_norm": 5.593987464904785,
+      "learning_rate": 7.207518796992482e-06,
+      "loss": 0.2514,
+      "step": 18570
+    },
+    {
+      "epoch": 27.94,
+      "grad_norm": 52.27437973022461,
+      "learning_rate": 7.206015037593986e-06,
+      "loss": 0.3338,
+      "step": 18580
+    },
+    {
+      "epoch": 27.95,
+      "grad_norm": 6.0754499435424805,
+      "learning_rate": 7.20451127819549e-06,
+      "loss": 0.2033,
+      "step": 18590
+    },
+    {
+      "epoch": 27.97,
+      "grad_norm": 13.490704536437988,
+      "learning_rate": 7.203007518796993e-06,
+      "loss": 0.328,
+      "step": 18600
+    },
+    {
+      "epoch": 27.98,
+      "grad_norm": 8.996991157531738,
+      "learning_rate": 7.201503759398497e-06,
+      "loss": 0.3082,
+      "step": 18610
+    },
+    {
+      "epoch": 28.0,
+      "grad_norm": 1.09871506690979,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 0.2957,
+      "step": 18620
+    },
+    {
+      "epoch": 28.0,
+      "eval_accuracy": 0.932,
+      "eval_loss": 0.2707752585411072,
+      "eval_runtime": 84.4192,
+      "eval_samples_per_second": 118.456,
+      "eval_steps_per_second": 0.474,
+      "step": 18620
+    },
+    {
+      "epoch": 28.02,
+      "grad_norm": 4.388790130615234,
+      "learning_rate": 7.198496240601505e-06,
+      "loss": 0.2982,
+      "step": 18630
+    },
+    {
+      "epoch": 28.03,
+      "grad_norm": 11.981380462646484,
+      "learning_rate": 7.1969924812030076e-06,
+      "loss": 0.2254,
+      "step": 18640
+    },
+    {
+      "epoch": 28.05,
+      "grad_norm": 5.74471378326416,
+      "learning_rate": 7.195488721804512e-06,
+      "loss": 0.2806,
+      "step": 18650
+    },
+    {
+      "epoch": 28.06,
+      "grad_norm": 6.635989189147949,
+      "learning_rate": 7.1939849624060154e-06,
+      "loss": 0.3081,
+      "step": 18660
+    },
+    {
+      "epoch": 28.08,
+      "grad_norm": 8.036993026733398,
+      "learning_rate": 7.19248120300752e-06,
+      "loss": 0.2684,
+      "step": 18670
+    },
+    {
+      "epoch": 28.09,
+      "grad_norm": 5.605617523193359,
+      "learning_rate": 7.190977443609023e-06,
+      "loss": 0.3037,
+      "step": 18680
+    },
+    {
+      "epoch": 28.11,
+      "grad_norm": 5.046873569488525,
+      "learning_rate": 7.189473684210527e-06,
+      "loss": 0.2489,
+      "step": 18690
+    },
+    {
+      "epoch": 28.12,
+      "grad_norm": 7.5858154296875,
+      "learning_rate": 7.18796992481203e-06,
+      "loss": 0.2868,
+      "step": 18700
+    },
+    {
+      "epoch": 28.14,
+      "grad_norm": 5.2180399894714355,
+      "learning_rate": 7.186466165413535e-06,
+      "loss": 0.282,
+      "step": 18710
+    },
+    {
+      "epoch": 28.15,
+      "grad_norm": 5.189599990844727,
+      "learning_rate": 7.184962406015038e-06,
+      "loss": 0.3185,
+      "step": 18720
+    },
+    {
+      "epoch": 28.17,
+      "grad_norm": 3.309946298599243,
+      "learning_rate": 7.183458646616543e-06,
+      "loss": 0.3342,
+      "step": 18730
+    },
+    {
+      "epoch": 28.18,
+      "grad_norm": 4.891626834869385,
+      "learning_rate": 7.181954887218045e-06,
+      "loss": 0.2972,
+      "step": 18740
+    },
+    {
+      "epoch": 28.2,
+      "grad_norm": 4.263134002685547,
+      "learning_rate": 7.18045112781955e-06,
+      "loss": 0.2384,
+      "step": 18750
+    },
+    {
+      "epoch": 28.21,
+      "grad_norm": 4.385184288024902,
+      "learning_rate": 7.178947368421053e-06,
+      "loss": 0.2699,
+      "step": 18760
+    },
+    {
+      "epoch": 28.23,
+      "grad_norm": 10.802248001098633,
+      "learning_rate": 7.1774436090225575e-06,
+      "loss": 0.3162,
+      "step": 18770
+    },
+    {
+      "epoch": 28.24,
+      "grad_norm": 4.073564052581787,
+      "learning_rate": 7.175939849624061e-06,
+      "loss": 0.2886,
+      "step": 18780
+    },
+    {
+      "epoch": 28.26,
+      "grad_norm": 4.669020175933838,
+      "learning_rate": 7.174436090225565e-06,
+      "loss": 0.3056,
+      "step": 18790
+    },
+    {
+      "epoch": 28.27,
+      "grad_norm": 7.317276477813721,
+      "learning_rate": 7.172932330827068e-06,
+      "loss": 0.3272,
+      "step": 18800
+    },
+    {
+      "epoch": 28.29,
+      "grad_norm": 2.6165945529937744,
+      "learning_rate": 7.1714285714285725e-06,
+      "loss": 0.2903,
+      "step": 18810
+    },
+    {
+      "epoch": 28.3,
+      "grad_norm": 4.007689476013184,
+      "learning_rate": 7.169924812030076e-06,
+      "loss": 0.3343,
+      "step": 18820
+    },
+    {
+      "epoch": 28.32,
+      "grad_norm": 4.415470123291016,
+      "learning_rate": 7.16842105263158e-06,
+      "loss": 0.2953,
+      "step": 18830
+    },
+    {
+      "epoch": 28.33,
+      "grad_norm": 3.293377637863159,
+      "learning_rate": 7.166917293233083e-06,
+      "loss": 0.2722,
+      "step": 18840
+    },
+    {
+      "epoch": 28.35,
+      "grad_norm": 5.47039270401001,
+      "learning_rate": 7.165413533834587e-06,
+      "loss": 0.2798,
+      "step": 18850
+    },
+    {
+      "epoch": 28.36,
+      "grad_norm": 5.256561756134033,
+      "learning_rate": 7.163909774436091e-06,
+      "loss": 0.2649,
+      "step": 18860
+    },
+    {
+      "epoch": 28.38,
+      "grad_norm": 3.793679714202881,
+      "learning_rate": 7.1624060150375944e-06,
+      "loss": 0.2919,
+      "step": 18870
+    },
+    {
+      "epoch": 28.39,
+      "grad_norm": 3.6800131797790527,
+      "learning_rate": 7.160902255639098e-06,
+      "loss": 0.2941,
+      "step": 18880
+    },
+    {
+      "epoch": 28.41,
+      "grad_norm": 6.140032768249512,
+      "learning_rate": 7.1593984962406015e-06,
+      "loss": 0.307,
+      "step": 18890
+    },
+    {
+      "epoch": 28.42,
+      "grad_norm": 5.978692531585693,
+      "learning_rate": 7.157894736842106e-06,
+      "loss": 0.2973,
+      "step": 18900
+    },
+    {
+      "epoch": 28.44,
+      "grad_norm": 6.257145404815674,
+      "learning_rate": 7.156390977443609e-06,
+      "loss": 0.3043,
+      "step": 18910
+    },
+    {
+      "epoch": 28.45,
+      "grad_norm": 6.124258518218994,
+      "learning_rate": 7.154887218045114e-06,
+      "loss": 0.2877,
+      "step": 18920
+    },
+    {
+      "epoch": 28.47,
+      "grad_norm": 4.110513687133789,
+      "learning_rate": 7.153383458646616e-06,
+      "loss": 0.2723,
+      "step": 18930
+    },
+    {
+      "epoch": 28.48,
+      "grad_norm": 8.390369415283203,
+      "learning_rate": 7.151879699248121e-06,
+      "loss": 0.2854,
+      "step": 18940
+    },
+    {
+      "epoch": 28.5,
+      "grad_norm": 4.641889572143555,
+      "learning_rate": 7.150375939849624e-06,
+      "loss": 0.2325,
+      "step": 18950
+    },
+    {
+      "epoch": 28.51,
+      "grad_norm": 9.678013801574707,
+      "learning_rate": 7.148872180451129e-06,
+      "loss": 0.2325,
+      "step": 18960
+    },
+    {
+      "epoch": 28.53,
+      "grad_norm": 4.375498294830322,
+      "learning_rate": 7.147368421052631e-06,
+      "loss": 0.2814,
+      "step": 18970
+    },
+    {
+      "epoch": 28.54,
+      "grad_norm": 11.331188201904297,
+      "learning_rate": 7.145864661654136e-06,
+      "loss": 0.2659,
+      "step": 18980
+    },
+    {
+      "epoch": 28.56,
+      "grad_norm": 4.929275035858154,
+      "learning_rate": 7.144360902255639e-06,
+      "loss": 0.3207,
+      "step": 18990
+    },
+    {
+      "epoch": 28.57,
+      "grad_norm": 6.941195487976074,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 0.2677,
+      "step": 19000
+    },
+    {
+      "epoch": 28.59,
+      "grad_norm": 7.25919771194458,
+      "learning_rate": 7.141353383458647e-06,
+      "loss": 0.265,
+      "step": 19010
+    },
+    {
+      "epoch": 28.6,
+      "grad_norm": 4.009887218475342,
+      "learning_rate": 7.1398496240601514e-06,
+      "loss": 0.289,
+      "step": 19020
+    },
+    {
+      "epoch": 28.62,
+      "grad_norm": 7.65335750579834,
+      "learning_rate": 7.138345864661654e-06,
+      "loss": 0.2977,
+      "step": 19030
+    },
+    {
+      "epoch": 28.63,
+      "grad_norm": 7.200645923614502,
+      "learning_rate": 7.1368421052631585e-06,
+      "loss": 0.3497,
+      "step": 19040
+    },
+    {
+      "epoch": 28.65,
+      "grad_norm": 4.218509674072266,
+      "learning_rate": 7.135338345864662e-06,
+      "loss": 0.2505,
+      "step": 19050
+    },
+    {
+      "epoch": 28.66,
+      "grad_norm": 5.385000228881836,
+      "learning_rate": 7.133834586466166e-06,
+      "loss": 0.2603,
+      "step": 19060
+    },
+    {
+      "epoch": 28.68,
+      "grad_norm": 6.183909893035889,
+      "learning_rate": 7.132330827067669e-06,
+      "loss": 0.3193,
+      "step": 19070
+    },
+    {
+      "epoch": 28.69,
+      "grad_norm": 7.682650089263916,
+      "learning_rate": 7.130827067669173e-06,
+      "loss": 0.3485,
+      "step": 19080
+    },
+    {
+      "epoch": 28.71,
+      "grad_norm": 6.737081527709961,
+      "learning_rate": 7.129323308270677e-06,
+      "loss": 0.4199,
+      "step": 19090
+    },
+    {
+      "epoch": 28.72,
+      "grad_norm": 6.185298442840576,
+      "learning_rate": 7.127819548872181e-06,
+      "loss": 0.3144,
+      "step": 19100
+    },
+    {
+      "epoch": 28.74,
+      "grad_norm": 3.423515558242798,
+      "learning_rate": 7.126315789473685e-06,
+      "loss": 0.2768,
+      "step": 19110
+    },
+    {
+      "epoch": 28.75,
+      "grad_norm": 7.198062896728516,
+      "learning_rate": 7.124812030075189e-06,
+      "loss": 0.3282,
+      "step": 19120
+    },
+    {
+      "epoch": 28.77,
+      "grad_norm": 5.3529372215271,
+      "learning_rate": 7.123308270676692e-06,
+      "loss": 0.2467,
+      "step": 19130
+    },
+    {
+      "epoch": 28.78,
+      "grad_norm": 7.850579261779785,
+      "learning_rate": 7.121804511278196e-06,
+      "loss": 0.3204,
+      "step": 19140
+    },
+    {
+      "epoch": 28.8,
+      "grad_norm": 8.140227317810059,
+      "learning_rate": 7.1203007518797e-06,
+      "loss": 0.3087,
+      "step": 19150
+    },
+    {
+      "epoch": 28.81,
+      "grad_norm": 6.9892778396606445,
+      "learning_rate": 7.118796992481204e-06,
+      "loss": 0.3148,
+      "step": 19160
+    },
+    {
+      "epoch": 28.83,
+      "grad_norm": 6.189218521118164,
+      "learning_rate": 7.117293233082707e-06,
+      "loss": 0.2779,
+      "step": 19170
+    },
+    {
+      "epoch": 28.84,
+      "grad_norm": 4.910006999969482,
+      "learning_rate": 7.115789473684211e-06,
+      "loss": 0.2469,
+      "step": 19180
+    },
+    {
+      "epoch": 28.86,
+      "grad_norm": 6.6792449951171875,
+      "learning_rate": 7.114285714285715e-06,
+      "loss": 0.3604,
+      "step": 19190
+    },
+    {
+      "epoch": 28.87,
+      "grad_norm": 8.408853530883789,
+      "learning_rate": 7.112781954887219e-06,
+      "loss": 0.3493,
+      "step": 19200
+    },
+    {
+      "epoch": 28.89,
+      "grad_norm": 5.114132404327393,
+      "learning_rate": 7.1112781954887225e-06,
+      "loss": 0.2874,
+      "step": 19210
+    },
+    {
+      "epoch": 28.9,
+      "grad_norm": 7.833899021148682,
+      "learning_rate": 7.109774436090227e-06,
+      "loss": 0.2365,
+      "step": 19220
+    },
+    {
+      "epoch": 28.92,
+      "grad_norm": 3.8841164112091064,
+      "learning_rate": 7.10827067669173e-06,
+      "loss": 0.2937,
+      "step": 19230
+    },
+    {
+      "epoch": 28.93,
+      "grad_norm": 11.486865997314453,
+      "learning_rate": 7.106766917293234e-06,
+      "loss": 0.2851,
+      "step": 19240
+    },
+    {
+      "epoch": 28.95,
+      "grad_norm": 9.058024406433105,
+      "learning_rate": 7.1052631578947375e-06,
+      "loss": 0.2317,
+      "step": 19250
+    },
+    {
+      "epoch": 28.96,
+      "grad_norm": 9.031519889831543,
+      "learning_rate": 7.103759398496242e-06,
+      "loss": 0.4064,
+      "step": 19260
+    },
+    {
+      "epoch": 28.98,
+      "grad_norm": 5.929012298583984,
+      "learning_rate": 7.1022556390977445e-06,
+      "loss": 0.2661,
+      "step": 19270
+    },
+    {
+      "epoch": 28.99,
+      "grad_norm": 2.968751907348633,
+      "learning_rate": 7.100751879699249e-06,
+      "loss": 0.3013,
+      "step": 19280
+    },
+    {
+      "epoch": 29.0,
+      "eval_accuracy": 0.932,
+      "eval_loss": 0.28813642263412476,
+      "eval_runtime": 84.4456,
+      "eval_samples_per_second": 118.419,
+      "eval_steps_per_second": 0.474,
+      "step": 19285
+    },
+    {
+      "epoch": 29.01,
+      "grad_norm": 5.468835830688477,
+      "learning_rate": 7.099248120300752e-06,
+      "loss": 0.2397,
+      "step": 19290
+    },
+    {
+      "epoch": 29.02,
+      "grad_norm": 7.077803134918213,
+      "learning_rate": 7.097744360902257e-06,
+      "loss": 0.3068,
+      "step": 19300
+    },
+    {
+      "epoch": 29.04,
+      "grad_norm": 5.543747425079346,
+      "learning_rate": 7.09624060150376e-06,
+      "loss": 0.2652,
+      "step": 19310
+    },
+    {
+      "epoch": 29.05,
+      "grad_norm": 7.105257034301758,
+      "learning_rate": 7.094736842105265e-06,
+      "loss": 0.3016,
+      "step": 19320
+    },
+    {
+      "epoch": 29.07,
+      "grad_norm": 7.57663106918335,
+      "learning_rate": 7.093233082706767e-06,
+      "loss": 0.3195,
+      "step": 19330
+    },
+    {
+      "epoch": 29.08,
+      "grad_norm": 3.0557851791381836,
+      "learning_rate": 7.091729323308272e-06,
+      "loss": 0.2686,
+      "step": 19340
+    },
+    {
+      "epoch": 29.1,
+      "grad_norm": 6.048924446105957,
+      "learning_rate": 7.090225563909775e-06,
+      "loss": 0.2936,
+      "step": 19350
+    },
+    {
+      "epoch": 29.11,
+      "grad_norm": 5.691340446472168,
+      "learning_rate": 7.088721804511278e-06,
+      "loss": 0.2832,
+      "step": 19360
+    },
+    {
+      "epoch": 29.13,
+      "grad_norm": 0.4570540189743042,
+      "learning_rate": 7.087218045112782e-06,
+      "loss": 0.2204,
+      "step": 19370
+    },
+    {
+      "epoch": 29.14,
+      "grad_norm": 7.902672290802002,
+      "learning_rate": 7.085714285714286e-06,
+      "loss": 0.2519,
+      "step": 19380
+    },
+    {
+      "epoch": 29.16,
+      "grad_norm": 6.772716045379639,
+      "learning_rate": 7.08421052631579e-06,
+      "loss": 0.3449,
+      "step": 19390
+    },
+    {
+      "epoch": 29.17,
+      "grad_norm": 5.602826118469238,
+      "learning_rate": 7.082706766917294e-06,
+      "loss": 0.2434,
+      "step": 19400
+    },
+    {
+      "epoch": 29.19,
+      "grad_norm": 8.90966510772705,
+      "learning_rate": 7.081203007518798e-06,
+      "loss": 0.3094,
+      "step": 19410
+    },
+    {
+      "epoch": 29.2,
+      "grad_norm": 4.287824630737305,
+      "learning_rate": 7.079699248120301e-06,
+      "loss": 0.2961,
+      "step": 19420
+    },
+    {
+      "epoch": 29.22,
+      "grad_norm": 5.467939853668213,
+      "learning_rate": 7.078195488721805e-06,
+      "loss": 0.2937,
+      "step": 19430
+    },
+    {
+      "epoch": 29.23,
+      "grad_norm": 4.856724739074707,
+      "learning_rate": 7.0766917293233086e-06,
+      "loss": 0.3074,
+      "step": 19440
+    },
+    {
+      "epoch": 29.25,
+      "grad_norm": 4.881340980529785,
+      "learning_rate": 7.075187969924813e-06,
+      "loss": 0.3334,
+      "step": 19450
+    },
+    {
+      "epoch": 29.26,
+      "grad_norm": 6.680235862731934,
+      "learning_rate": 7.073684210526316e-06,
+      "loss": 0.2404,
+      "step": 19460
+    },
+    {
+      "epoch": 29.28,
+      "grad_norm": 3.575651168823242,
+      "learning_rate": 7.07218045112782e-06,
+      "loss": 0.2544,
+      "step": 19470
+    },
+    {
+      "epoch": 29.29,
+      "grad_norm": 3.3242225646972656,
+      "learning_rate": 7.0706766917293235e-06,
+      "loss": 0.2732,
+      "step": 19480
+    },
+    {
+      "epoch": 29.31,
+      "grad_norm": 3.5619566440582275,
+      "learning_rate": 7.069172932330828e-06,
+      "loss": 0.3299,
+      "step": 19490
+    },
+    {
+      "epoch": 29.32,
+      "grad_norm": 6.326242923736572,
+      "learning_rate": 7.067669172932331e-06,
+      "loss": 0.3078,
+      "step": 19500
+    },
+    {
+      "epoch": 29.34,
+      "grad_norm": 3.983919620513916,
+      "learning_rate": 7.066165413533836e-06,
+      "loss": 0.2805,
+      "step": 19510
+    },
+    {
+      "epoch": 29.35,
+      "grad_norm": 7.784421443939209,
+      "learning_rate": 7.064661654135338e-06,
+      "loss": 0.3146,
+      "step": 19520
+    },
+    {
+      "epoch": 29.37,
+      "grad_norm": 7.162094593048096,
+      "learning_rate": 7.063157894736843e-06,
+      "loss": 0.2942,
+      "step": 19530
+    },
+    {
+      "epoch": 29.38,
+      "grad_norm": 4.964542388916016,
+      "learning_rate": 7.061654135338346e-06,
+      "loss": 0.3356,
+      "step": 19540
+    },
+    {
+      "epoch": 29.4,
+      "grad_norm": 6.946933269500732,
+      "learning_rate": 7.060150375939851e-06,
+      "loss": 0.312,
+      "step": 19550
+    },
+    {
+      "epoch": 29.41,
+      "grad_norm": 6.954648494720459,
+      "learning_rate": 7.058646616541353e-06,
+      "loss": 0.3188,
+      "step": 19560
+    },
+    {
+      "epoch": 29.43,
+      "grad_norm": 6.528570175170898,
+      "learning_rate": 7.057142857142858e-06,
+      "loss": 0.2966,
+      "step": 19570
+    },
+    {
+      "epoch": 29.44,
+      "grad_norm": 5.95659875869751,
+      "learning_rate": 7.055639097744361e-06,
+      "loss": 0.3398,
+      "step": 19580
+    },
+    {
+      "epoch": 29.46,
+      "grad_norm": 2.7002809047698975,
+      "learning_rate": 7.054135338345866e-06,
+      "loss": 0.2737,
+      "step": 19590
+    },
+    {
+      "epoch": 29.47,
+      "grad_norm": 4.33442497253418,
+      "learning_rate": 7.052631578947369e-06,
+      "loss": 0.3002,
+      "step": 19600
+    },
+    {
+      "epoch": 29.49,
+      "grad_norm": 6.4437713623046875,
+      "learning_rate": 7.051127819548873e-06,
+      "loss": 0.28,
+      "step": 19610
+    },
+    {
+      "epoch": 29.5,
+      "grad_norm": 3.9194564819335938,
+      "learning_rate": 7.049624060150376e-06,
+      "loss": 0.3069,
+      "step": 19620
+    },
+    {
+      "epoch": 29.52,
+      "grad_norm": 5.887040615081787,
+      "learning_rate": 7.0481203007518805e-06,
+      "loss": 0.2634,
+      "step": 19630
+    },
+    {
+      "epoch": 29.53,
+      "grad_norm": 4.538432598114014,
+      "learning_rate": 7.046616541353384e-06,
+      "loss": 0.2684,
+      "step": 19640
+    },
+    {
+      "epoch": 29.55,
+      "grad_norm": 7.666621208190918,
+      "learning_rate": 7.045112781954888e-06,
+      "loss": 0.2344,
+      "step": 19650
+    },
+    {
+      "epoch": 29.56,
+      "grad_norm": 5.944088935852051,
+      "learning_rate": 7.043609022556391e-06,
+      "loss": 0.2358,
+      "step": 19660
+    },
+    {
+      "epoch": 29.58,
+      "grad_norm": 4.063729286193848,
+      "learning_rate": 7.0421052631578954e-06,
+      "loss": 0.2532,
+      "step": 19670
+    },
+    {
+      "epoch": 29.59,
+      "grad_norm": 3.9652276039123535,
+      "learning_rate": 7.040601503759399e-06,
+      "loss": 0.3471,
+      "step": 19680
+    },
+    {
+      "epoch": 29.61,
+      "grad_norm": 5.847537040710449,
+      "learning_rate": 7.039097744360903e-06,
+      "loss": 0.2433,
+      "step": 19690
+    },
+    {
+      "epoch": 29.62,
+      "grad_norm": 4.780336380004883,
+      "learning_rate": 7.037593984962407e-06,
+      "loss": 0.3358,
+      "step": 19700
+    },
+    {
+      "epoch": 29.64,
+      "grad_norm": 3.0303971767425537,
+      "learning_rate": 7.03609022556391e-06,
+      "loss": 0.2168,
+      "step": 19710
+    },
+    {
+      "epoch": 29.65,
+      "grad_norm": 3.0998427867889404,
+      "learning_rate": 7.034586466165414e-06,
+      "loss": 0.2995,
+      "step": 19720
+    },
+    {
+      "epoch": 29.67,
+      "grad_norm": 2.832582473754883,
+      "learning_rate": 7.033082706766918e-06,
+      "loss": 0.2413,
+      "step": 19730
+    },
+    {
+      "epoch": 29.68,
+      "grad_norm": 7.660305023193359,
+      "learning_rate": 7.031578947368422e-06,
+      "loss": 0.2962,
+      "step": 19740
+    },
+    {
+      "epoch": 29.7,
+      "grad_norm": 5.547909259796143,
+      "learning_rate": 7.030075187969926e-06,
+      "loss": 0.3422,
+      "step": 19750
+    },
+    {
+      "epoch": 29.71,
+      "grad_norm": 8.43567180633545,
+      "learning_rate": 7.028571428571429e-06,
+      "loss": 0.3488,
+      "step": 19760
+    },
+    {
+      "epoch": 29.73,
+      "grad_norm": 8.269464492797852,
+      "learning_rate": 7.027067669172933e-06,
+      "loss": 0.2961,
+      "step": 19770
+    },
+    {
+      "epoch": 29.74,
+      "grad_norm": 10.270258903503418,
+      "learning_rate": 7.025563909774437e-06,
+      "loss": 0.2522,
+      "step": 19780
+    },
+    {
+      "epoch": 29.76,
+      "grad_norm": 5.3482537269592285,
+      "learning_rate": 7.024060150375941e-06,
+      "loss": 0.2841,
+      "step": 19790
+    },
+    {
+      "epoch": 29.77,
+      "grad_norm": 5.868763446807861,
+      "learning_rate": 7.022556390977444e-06,
+      "loss": 0.2728,
+      "step": 19800
+    },
+    {
+      "epoch": 29.79,
+      "grad_norm": 12.772025108337402,
+      "learning_rate": 7.021052631578948e-06,
+      "loss": 0.2202,
+      "step": 19810
+    },
+    {
+      "epoch": 29.8,
+      "grad_norm": 9.140650749206543,
+      "learning_rate": 7.019548872180452e-06,
+      "loss": 0.3589,
+      "step": 19820
+    },
+    {
+      "epoch": 29.82,
+      "grad_norm": 5.013607978820801,
+      "learning_rate": 7.018045112781956e-06,
+      "loss": 0.2229,
+      "step": 19830
+    },
+    {
+      "epoch": 29.83,
+      "grad_norm": 8.760464668273926,
+      "learning_rate": 7.0165413533834595e-06,
+      "loss": 0.3337,
+      "step": 19840
+    },
+    {
+      "epoch": 29.85,
+      "grad_norm": 7.938315391540527,
+      "learning_rate": 7.015037593984964e-06,
+      "loss": 0.3052,
+      "step": 19850
+    },
+    {
+      "epoch": 29.86,
+      "grad_norm": 7.728393077850342,
+      "learning_rate": 7.0135338345864665e-06,
+      "loss": 0.2861,
+      "step": 19860
+    },
+    {
+      "epoch": 29.88,
+      "grad_norm": 6.810389995574951,
+      "learning_rate": 7.01203007518797e-06,
+      "loss": 0.3451,
+      "step": 19870
+    },
+    {
+      "epoch": 29.89,
+      "grad_norm": 3.8391294479370117,
+      "learning_rate": 7.010526315789474e-06,
+      "loss": 0.2302,
+      "step": 19880
+    },
+    {
+      "epoch": 29.91,
+      "grad_norm": 8.415011405944824,
+      "learning_rate": 7.009022556390977e-06,
+      "loss": 0.3055,
+      "step": 19890
+    },
+    {
+      "epoch": 29.92,
+      "grad_norm": 3.6795806884765625,
+      "learning_rate": 7.0075187969924815e-06,
+      "loss": 0.3007,
+      "step": 19900
+    },
+    {
+      "epoch": 29.94,
+      "grad_norm": 6.91939640045166,
+      "learning_rate": 7.006015037593985e-06,
+      "loss": 0.296,
+      "step": 19910
+    },
+    {
+      "epoch": 29.95,
+      "grad_norm": 4.2215681076049805,
+      "learning_rate": 7.004511278195489e-06,
+      "loss": 0.26,
+      "step": 19920
+    },
+    {
+      "epoch": 29.97,
+      "grad_norm": 3.9634735584259033,
+      "learning_rate": 7.003007518796993e-06,
+      "loss": 0.2929,
+      "step": 19930
+    },
+    {
+      "epoch": 29.98,
+      "grad_norm": 4.356061935424805,
+      "learning_rate": 7.001503759398497e-06,
+      "loss": 0.2259,
+      "step": 19940
+    },
+    {
+      "epoch": 30.0,
+      "grad_norm": 1.2935045957565308,
+      "learning_rate": 7e-06,
+      "loss": 0.2811,
+      "step": 19950
+    },
+    {
+      "epoch": 30.0,
+      "eval_accuracy": 0.9304,
+      "eval_loss": 0.2939961552619934,
+      "eval_runtime": 85.2279,
+      "eval_samples_per_second": 117.333,
+      "eval_steps_per_second": 0.469,
+      "step": 19950
+    },
+    {
+      "epoch": 30.02,
+      "grad_norm": 4.659907817840576,
+      "learning_rate": 6.998496240601504e-06,
+      "loss": 0.2815,
+      "step": 19960
+    },
+    {
+      "epoch": 30.03,
+      "grad_norm": 7.678098201751709,
+      "learning_rate": 6.996992481203008e-06,
+      "loss": 0.2224,
+      "step": 19970
+    },
+    {
+      "epoch": 30.05,
+      "grad_norm": 3.063901662826538,
+      "learning_rate": 6.995488721804512e-06,
+      "loss": 0.2971,
+      "step": 19980
+    },
+    {
+      "epoch": 30.06,
+      "grad_norm": 5.704662322998047,
+      "learning_rate": 6.993984962406015e-06,
+      "loss": 0.276,
+      "step": 19990
+    },
+    {
+      "epoch": 30.08,
+      "grad_norm": 5.150294780731201,
+      "learning_rate": 6.992481203007519e-06,
+      "loss": 0.3514,
+      "step": 20000
+    },
+    {
+      "epoch": 30.09,
+      "grad_norm": 6.354630947113037,
+      "learning_rate": 6.990977443609023e-06,
+      "loss": 0.2391,
+      "step": 20010
+    },
+    {
+      "epoch": 30.11,
+      "grad_norm": 3.2037034034729004,
+      "learning_rate": 6.989473684210527e-06,
+      "loss": 0.2754,
+      "step": 20020
+    },
+    {
+      "epoch": 30.12,
+      "grad_norm": 9.49938678741455,
+      "learning_rate": 6.987969924812031e-06,
+      "loss": 0.3042,
+      "step": 20030
+    },
+    {
+      "epoch": 30.14,
+      "grad_norm": 7.142296314239502,
+      "learning_rate": 6.986466165413535e-06,
+      "loss": 0.2314,
+      "step": 20040
+    },
+    {
+      "epoch": 30.15,
+      "grad_norm": 7.547910690307617,
+      "learning_rate": 6.984962406015038e-06,
+      "loss": 0.2545,
+      "step": 20050
+    },
+    {
+      "epoch": 30.17,
+      "grad_norm": 3.509197235107422,
+      "learning_rate": 6.983458646616542e-06,
+      "loss": 0.2241,
+      "step": 20060
+    },
+    {
+      "epoch": 30.18,
+      "grad_norm": 5.678248882293701,
+      "learning_rate": 6.9819548872180455e-06,
+      "loss": 0.2485,
+      "step": 20070
+    },
+    {
+      "epoch": 30.2,
+      "grad_norm": 6.966419696807861,
+      "learning_rate": 6.98045112781955e-06,
+      "loss": 0.2342,
+      "step": 20080
+    },
+    {
+      "epoch": 30.21,
+      "grad_norm": 5.888791084289551,
+      "learning_rate": 6.9789473684210525e-06,
+      "loss": 0.2433,
+      "step": 20090
+    },
+    {
+      "epoch": 30.23,
+      "grad_norm": 6.074886322021484,
+      "learning_rate": 6.977443609022557e-06,
+      "loss": 0.3245,
+      "step": 20100
+    },
+    {
+      "epoch": 30.24,
+      "grad_norm": 8.158697128295898,
+      "learning_rate": 6.9759398496240604e-06,
+      "loss": 0.2656,
+      "step": 20110
+    },
+    {
+      "epoch": 30.26,
+      "grad_norm": 4.1151909828186035,
+      "learning_rate": 6.974436090225565e-06,
+      "loss": 0.2862,
+      "step": 20120
+    },
+    {
+      "epoch": 30.27,
+      "grad_norm": 5.835092067718506,
+      "learning_rate": 6.972932330827068e-06,
+      "loss": 0.3154,
+      "step": 20130
+    },
+    {
+      "epoch": 30.29,
+      "grad_norm": 7.01699161529541,
+      "learning_rate": 6.971428571428573e-06,
+      "loss": 0.3146,
+      "step": 20140
+    },
+    {
+      "epoch": 30.3,
+      "grad_norm": 4.701117515563965,
+      "learning_rate": 6.969924812030075e-06,
+      "loss": 0.2476,
+      "step": 20150
+    },
+    {
+      "epoch": 30.32,
+      "grad_norm": 7.054307460784912,
+      "learning_rate": 6.96842105263158e-06,
+      "loss": 0.2467,
+      "step": 20160
+    },
+    {
+      "epoch": 30.33,
+      "grad_norm": 6.54818868637085,
+      "learning_rate": 6.966917293233083e-06,
+      "loss": 0.2588,
+      "step": 20170
+    },
+    {
+      "epoch": 30.35,
+      "grad_norm": 4.79921817779541,
+      "learning_rate": 6.965413533834588e-06,
+      "loss": 0.2452,
+      "step": 20180
+    },
+    {
+      "epoch": 30.36,
+      "grad_norm": 7.635383129119873,
+      "learning_rate": 6.96390977443609e-06,
+      "loss": 0.2971,
+      "step": 20190
+    },
+    {
+      "epoch": 30.38,
+      "grad_norm": 6.174952507019043,
+      "learning_rate": 6.962406015037595e-06,
+      "loss": 0.2438,
+      "step": 20200
+    },
+    {
+      "epoch": 30.39,
+      "grad_norm": 8.053534507751465,
+      "learning_rate": 6.960902255639098e-06,
+      "loss": 0.3147,
+      "step": 20210
+    },
+    {
+      "epoch": 30.41,
+      "grad_norm": 5.930545806884766,
+      "learning_rate": 6.9593984962406025e-06,
+      "loss": 0.3246,
+      "step": 20220
+    },
+    {
+      "epoch": 30.42,
+      "grad_norm": 5.0360493659973145,
+      "learning_rate": 6.957894736842106e-06,
+      "loss": 0.3028,
+      "step": 20230
+    },
+    {
+      "epoch": 30.44,
+      "grad_norm": 5.536324977874756,
+      "learning_rate": 6.95639097744361e-06,
+      "loss": 0.2762,
+      "step": 20240
+    },
+    {
+      "epoch": 30.45,
+      "grad_norm": 3.6296212673187256,
+      "learning_rate": 6.954887218045113e-06,
+      "loss": 0.2799,
+      "step": 20250
+    },
+    {
+      "epoch": 30.47,
+      "grad_norm": 5.655179977416992,
+      "learning_rate": 6.9533834586466175e-06,
+      "loss": 0.3003,
+      "step": 20260
+    },
+    {
+      "epoch": 30.48,
+      "grad_norm": 5.260562419891357,
+      "learning_rate": 6.951879699248121e-06,
+      "loss": 0.2221,
+      "step": 20270
+    },
+    {
+      "epoch": 30.5,
+      "grad_norm": 4.750830173492432,
+      "learning_rate": 6.950375939849625e-06,
+      "loss": 0.2439,
+      "step": 20280
+    },
+    {
+      "epoch": 30.51,
+      "grad_norm": 6.058375835418701,
+      "learning_rate": 6.948872180451128e-06,
+      "loss": 0.2716,
+      "step": 20290
+    },
+    {
+      "epoch": 30.53,
+      "grad_norm": 4.154491901397705,
+      "learning_rate": 6.947368421052632e-06,
+      "loss": 0.2311,
+      "step": 20300
+    },
+    {
+      "epoch": 30.54,
+      "grad_norm": 4.645699977874756,
+      "learning_rate": 6.945864661654136e-06,
+      "loss": 0.3293,
+      "step": 20310
+    },
+    {
+      "epoch": 30.56,
+      "grad_norm": 4.276103973388672,
+      "learning_rate": 6.94436090225564e-06,
+      "loss": 0.2236,
+      "step": 20320
+    },
+    {
+      "epoch": 30.57,
+      "grad_norm": 5.338918685913086,
+      "learning_rate": 6.942857142857144e-06,
+      "loss": 0.3162,
+      "step": 20330
+    },
+    {
+      "epoch": 30.59,
+      "grad_norm": 4.5537285804748535,
+      "learning_rate": 6.941353383458648e-06,
+      "loss": 0.2801,
+      "step": 20340
+    },
+    {
+      "epoch": 30.6,
+      "grad_norm": 4.682258605957031,
+      "learning_rate": 6.939849624060151e-06,
+      "loss": 0.2674,
+      "step": 20350
+    },
+    {
+      "epoch": 30.62,
+      "grad_norm": 5.651388168334961,
+      "learning_rate": 6.938345864661654e-06,
+      "loss": 0.3353,
+      "step": 20360
+    },
+    {
+      "epoch": 30.63,
+      "grad_norm": 4.2494797706604,
+      "learning_rate": 6.936842105263159e-06,
+      "loss": 0.3508,
+      "step": 20370
+    },
+    {
+      "epoch": 30.65,
+      "grad_norm": 5.229772090911865,
+      "learning_rate": 6.935338345864661e-06,
+      "loss": 0.2578,
+      "step": 20380
+    },
+    {
+      "epoch": 30.66,
+      "grad_norm": 1.8404473066329956,
+      "learning_rate": 6.933834586466166e-06,
+      "loss": 0.2398,
+      "step": 20390
+    },
+    {
+      "epoch": 30.68,
+      "grad_norm": 3.122826337814331,
+      "learning_rate": 6.932330827067669e-06,
+      "loss": 0.2732,
+      "step": 20400
+    },
+    {
+      "epoch": 30.69,
+      "grad_norm": 3.0387120246887207,
+      "learning_rate": 6.930827067669174e-06,
+      "loss": 0.2722,
+      "step": 20410
+    },
+    {
+      "epoch": 30.71,
+      "grad_norm": 3.9582316875457764,
+      "learning_rate": 6.929323308270677e-06,
+      "loss": 0.2502,
+      "step": 20420
+    },
+    {
+      "epoch": 30.72,
+      "grad_norm": 4.217288494110107,
+      "learning_rate": 6.9278195488721815e-06,
+      "loss": 0.2763,
+      "step": 20430
+    },
+    {
+      "epoch": 30.74,
+      "grad_norm": 5.593020915985107,
+      "learning_rate": 6.926315789473684e-06,
+      "loss": 0.3234,
+      "step": 20440
+    },
+    {
+      "epoch": 30.75,
+      "grad_norm": 6.763814926147461,
+      "learning_rate": 6.9248120300751886e-06,
+      "loss": 0.3275,
+      "step": 20450
+    },
+    {
+      "epoch": 30.77,
+      "grad_norm": 4.887606620788574,
+      "learning_rate": 6.923308270676692e-06,
+      "loss": 0.266,
+      "step": 20460
+    },
+    {
+      "epoch": 30.78,
+      "grad_norm": 2.9521090984344482,
+      "learning_rate": 6.9218045112781964e-06,
+      "loss": 0.2542,
+      "step": 20470
+    },
+    {
+      "epoch": 30.8,
+      "grad_norm": 6.818195819854736,
+      "learning_rate": 6.920300751879699e-06,
+      "loss": 0.3018,
+      "step": 20480
+    },
+    {
+      "epoch": 30.81,
+      "grad_norm": 9.507380485534668,
+      "learning_rate": 6.9187969924812035e-06,
+      "loss": 0.2462,
+      "step": 20490
+    },
+    {
+      "epoch": 30.83,
+      "grad_norm": 6.438077926635742,
+      "learning_rate": 6.917293233082707e-06,
+      "loss": 0.2951,
+      "step": 20500
+    },
+    {
+      "epoch": 30.84,
+      "grad_norm": 2.8363983631134033,
+      "learning_rate": 6.915789473684211e-06,
+      "loss": 0.3104,
+      "step": 20510
+    },
+    {
+      "epoch": 30.86,
+      "grad_norm": 4.733820915222168,
+      "learning_rate": 6.914285714285715e-06,
+      "loss": 0.2756,
+      "step": 20520
+    },
+    {
+      "epoch": 30.87,
+      "grad_norm": 6.546677112579346,
+      "learning_rate": 6.912781954887218e-06,
+      "loss": 0.3449,
+      "step": 20530
+    },
+    {
+      "epoch": 30.89,
+      "grad_norm": 7.878147125244141,
+      "learning_rate": 6.911278195488722e-06,
+      "loss": 0.2393,
+      "step": 20540
+    },
+    {
+      "epoch": 30.9,
+      "grad_norm": 4.541097164154053,
+      "learning_rate": 6.909774436090226e-06,
+      "loss": 0.3316,
+      "step": 20550
+    },
+    {
+      "epoch": 30.92,
+      "grad_norm": 6.208898544311523,
+      "learning_rate": 6.90827067669173e-06,
+      "loss": 0.3414,
+      "step": 20560
+    },
+    {
+      "epoch": 30.93,
+      "grad_norm": 5.853148460388184,
+      "learning_rate": 6.906766917293234e-06,
+      "loss": 0.356,
+      "step": 20570
+    },
+    {
+      "epoch": 30.95,
+      "grad_norm": 5.502560615539551,
+      "learning_rate": 6.905263157894737e-06,
+      "loss": 0.2357,
+      "step": 20580
+    },
+    {
+      "epoch": 30.96,
+      "grad_norm": 3.3541431427001953,
+      "learning_rate": 6.903759398496241e-06,
+      "loss": 0.2527,
+      "step": 20590
+    },
+    {
+      "epoch": 30.98,
+      "grad_norm": 20.790668487548828,
+      "learning_rate": 6.902255639097745e-06,
+      "loss": 0.3085,
+      "step": 20600
+    },
+    {
+      "epoch": 30.99,
+      "grad_norm": 6.3316216468811035,
+      "learning_rate": 6.900751879699249e-06,
+      "loss": 0.2031,
+      "step": 20610
+    },
+    {
+      "epoch": 31.0,
+      "eval_accuracy": 0.9335,
+      "eval_loss": 0.2801915109157562,
+      "eval_runtime": 84.9003,
+      "eval_samples_per_second": 117.785,
+      "eval_steps_per_second": 0.471,
+      "step": 20615
+    },
+    {
+      "epoch": 31.01,
+      "grad_norm": 7.1735148429870605,
+      "learning_rate": 6.899248120300753e-06,
+      "loss": 0.2616,
+      "step": 20620
+    },
+    {
+      "epoch": 31.02,
+      "grad_norm": 3.6667120456695557,
+      "learning_rate": 6.897744360902256e-06,
+      "loss": 0.354,
+      "step": 20630
+    },
+    {
+      "epoch": 31.04,
+      "grad_norm": 5.409661293029785,
+      "learning_rate": 6.89624060150376e-06,
+      "loss": 0.323,
+      "step": 20640
+    },
+    {
+      "epoch": 31.05,
+      "grad_norm": 4.91942834854126,
+      "learning_rate": 6.894736842105264e-06,
+      "loss": 0.3082,
+      "step": 20650
+    },
+    {
+      "epoch": 31.07,
+      "grad_norm": 7.898626804351807,
+      "learning_rate": 6.8932330827067675e-06,
+      "loss": 0.2154,
+      "step": 20660
+    },
+    {
+      "epoch": 31.08,
+      "grad_norm": 5.627191543579102,
+      "learning_rate": 6.891729323308272e-06,
+      "loss": 0.3336,
+      "step": 20670
+    },
+    {
+      "epoch": 31.1,
+      "grad_norm": 6.512294769287109,
+      "learning_rate": 6.8902255639097746e-06,
+      "loss": 0.2925,
+      "step": 20680
+    },
+    {
+      "epoch": 31.11,
+      "grad_norm": 5.055330276489258,
+      "learning_rate": 6.888721804511279e-06,
+      "loss": 0.195,
+      "step": 20690
+    },
+    {
+      "epoch": 31.13,
+      "grad_norm": 4.006707191467285,
+      "learning_rate": 6.8872180451127825e-06,
+      "loss": 0.2374,
+      "step": 20700
+    },
+    {
+      "epoch": 31.14,
+      "grad_norm": 6.585967063903809,
+      "learning_rate": 6.885714285714287e-06,
+      "loss": 0.2938,
+      "step": 20710
+    },
+    {
+      "epoch": 31.16,
+      "grad_norm": 7.993644714355469,
+      "learning_rate": 6.8842105263157895e-06,
+      "loss": 0.2862,
+      "step": 20720
+    },
+    {
+      "epoch": 31.17,
+      "grad_norm": 6.300648212432861,
+      "learning_rate": 6.882706766917294e-06,
+      "loss": 0.3122,
+      "step": 20730
+    },
+    {
+      "epoch": 31.19,
+      "grad_norm": 6.135032653808594,
+      "learning_rate": 6.881203007518797e-06,
+      "loss": 0.2494,
+      "step": 20740
+    },
+    {
+      "epoch": 31.2,
+      "grad_norm": 3.280155658721924,
+      "learning_rate": 6.879699248120302e-06,
+      "loss": 0.3043,
+      "step": 20750
+    },
+    {
+      "epoch": 31.22,
+      "grad_norm": 6.118671417236328,
+      "learning_rate": 6.878195488721805e-06,
+      "loss": 0.2779,
+      "step": 20760
+    },
+    {
+      "epoch": 31.23,
+      "grad_norm": 8.142518043518066,
+      "learning_rate": 6.87669172932331e-06,
+      "loss": 0.3617,
+      "step": 20770
+    },
+    {
+      "epoch": 31.25,
+      "grad_norm": 5.192366123199463,
+      "learning_rate": 6.875187969924812e-06,
+      "loss": 0.3242,
+      "step": 20780
+    },
+    {
+      "epoch": 31.26,
+      "grad_norm": 5.72282075881958,
+      "learning_rate": 6.873684210526317e-06,
+      "loss": 0.2499,
+      "step": 20790
+    },
+    {
+      "epoch": 31.28,
+      "grad_norm": 6.699811935424805,
+      "learning_rate": 6.87218045112782e-06,
+      "loss": 0.2588,
+      "step": 20800
+    },
+    {
+      "epoch": 31.29,
+      "grad_norm": 5.5423173904418945,
+      "learning_rate": 6.8706766917293246e-06,
+      "loss": 0.3162,
+      "step": 20810
+    },
+    {
+      "epoch": 31.31,
+      "grad_norm": 6.422053813934326,
+      "learning_rate": 6.869172932330827e-06,
+      "loss": 0.2665,
+      "step": 20820
+    },
+    {
+      "epoch": 31.32,
+      "grad_norm": 6.014066219329834,
+      "learning_rate": 6.867669172932332e-06,
+      "loss": 0.2328,
+      "step": 20830
+    },
+    {
+      "epoch": 31.34,
+      "grad_norm": 13.063108444213867,
+      "learning_rate": 6.866165413533835e-06,
+      "loss": 0.2635,
+      "step": 20840
+    },
+    {
+      "epoch": 31.35,
+      "grad_norm": 5.6524882316589355,
+      "learning_rate": 6.864661654135339e-06,
+      "loss": 0.3334,
+      "step": 20850
+    },
+    {
+      "epoch": 31.37,
+      "grad_norm": 4.057956218719482,
+      "learning_rate": 6.863157894736843e-06,
+      "loss": 0.3186,
+      "step": 20860
+    },
+    {
+      "epoch": 31.38,
+      "grad_norm": 4.220489025115967,
+      "learning_rate": 6.861654135338346e-06,
+      "loss": 0.2551,
+      "step": 20870
+    },
+    {
+      "epoch": 31.4,
+      "grad_norm": 7.07074499130249,
+      "learning_rate": 6.86015037593985e-06,
+      "loss": 0.2236,
+      "step": 20880
+    },
+    {
+      "epoch": 31.41,
+      "grad_norm": 4.640635967254639,
+      "learning_rate": 6.8586466165413536e-06,
+      "loss": 0.2972,
+      "step": 20890
+    },
+    {
+      "epoch": 31.43,
+      "grad_norm": 11.102641105651855,
+      "learning_rate": 6.857142857142858e-06,
+      "loss": 0.2724,
+      "step": 20900
+    },
+    {
+      "epoch": 31.44,
+      "grad_norm": 8.299273490905762,
+      "learning_rate": 6.855639097744361e-06,
+      "loss": 0.267,
+      "step": 20910
+    },
+    {
+      "epoch": 31.46,
+      "grad_norm": 6.167851448059082,
+      "learning_rate": 6.854135338345865e-06,
+      "loss": 0.2695,
+      "step": 20920
+    },
+    {
+      "epoch": 31.47,
+      "grad_norm": 6.481257438659668,
+      "learning_rate": 6.8526315789473685e-06,
+      "loss": 0.2886,
+      "step": 20930
+    },
+    {
+      "epoch": 31.49,
+      "grad_norm": 6.720365047454834,
+      "learning_rate": 6.851127819548873e-06,
+      "loss": 0.2833,
+      "step": 20940
+    },
+    {
+      "epoch": 31.5,
+      "grad_norm": 6.032981872558594,
+      "learning_rate": 6.849624060150376e-06,
+      "loss": 0.2235,
+      "step": 20950
+    },
+    {
+      "epoch": 31.52,
+      "grad_norm": 7.538634300231934,
+      "learning_rate": 6.848120300751881e-06,
+      "loss": 0.261,
+      "step": 20960
+    },
+    {
+      "epoch": 31.53,
+      "grad_norm": 3.033374071121216,
+      "learning_rate": 6.846616541353383e-06,
+      "loss": 0.2572,
+      "step": 20970
+    },
+    {
+      "epoch": 31.55,
+      "grad_norm": 4.783783435821533,
+      "learning_rate": 6.845112781954888e-06,
+      "loss": 0.268,
+      "step": 20980
+    },
+    {
+      "epoch": 31.56,
+      "grad_norm": 5.95822811126709,
+      "learning_rate": 6.843609022556391e-06,
+      "loss": 0.265,
+      "step": 20990
+    },
+    {
+      "epoch": 31.58,
+      "grad_norm": 8.217907905578613,
+      "learning_rate": 6.842105263157896e-06,
+      "loss": 0.3141,
+      "step": 21000
+    },
+    {
+      "epoch": 31.59,
+      "grad_norm": 9.597149848937988,
+      "learning_rate": 6.840601503759398e-06,
+      "loss": 0.2925,
+      "step": 21010
+    },
+    {
+      "epoch": 31.61,
+      "grad_norm": 7.296209812164307,
+      "learning_rate": 6.839097744360903e-06,
+      "loss": 0.3045,
+      "step": 21020
+    },
+    {
+      "epoch": 31.62,
+      "grad_norm": 5.84061336517334,
+      "learning_rate": 6.837593984962406e-06,
+      "loss": 0.3293,
+      "step": 21030
+    },
+    {
+      "epoch": 31.64,
+      "grad_norm": 4.444825172424316,
+      "learning_rate": 6.8360902255639106e-06,
+      "loss": 0.2902,
+      "step": 21040
+    },
+    {
+      "epoch": 31.65,
+      "grad_norm": 12.595419883728027,
+      "learning_rate": 6.834586466165414e-06,
+      "loss": 0.253,
+      "step": 21050
+    },
+    {
+      "epoch": 31.67,
+      "grad_norm": 8.716811180114746,
+      "learning_rate": 6.8330827067669185e-06,
+      "loss": 0.2794,
+      "step": 21060
+    },
+    {
+      "epoch": 31.68,
+      "grad_norm": 6.067722320556641,
+      "learning_rate": 6.831578947368421e-06,
+      "loss": 0.2841,
+      "step": 21070
+    },
+    {
+      "epoch": 31.7,
+      "grad_norm": 4.765297889709473,
+      "learning_rate": 6.8300751879699255e-06,
+      "loss": 0.2732,
+      "step": 21080
+    },
+    {
+      "epoch": 31.71,
+      "grad_norm": 7.085923671722412,
+      "learning_rate": 6.828571428571429e-06,
+      "loss": 0.2453,
+      "step": 21090
+    },
+    {
+      "epoch": 31.73,
+      "grad_norm": 4.44352912902832,
+      "learning_rate": 6.827067669172933e-06,
+      "loss": 0.3153,
+      "step": 21100
+    },
+    {
+      "epoch": 31.74,
+      "grad_norm": 6.792245864868164,
+      "learning_rate": 6.825563909774436e-06,
+      "loss": 0.2952,
+      "step": 21110
+    },
+    {
+      "epoch": 31.76,
+      "grad_norm": 4.471166133880615,
+      "learning_rate": 6.82406015037594e-06,
+      "loss": 0.282,
+      "step": 21120
+    },
+    {
+      "epoch": 31.77,
+      "grad_norm": 57.37477493286133,
+      "learning_rate": 6.822556390977444e-06,
+      "loss": 0.2059,
+      "step": 21130
+    },
+    {
+      "epoch": 31.79,
+      "grad_norm": 4.992650985717773,
+      "learning_rate": 6.821052631578948e-06,
+      "loss": 0.2746,
+      "step": 21140
+    },
+    {
+      "epoch": 31.8,
+      "grad_norm": 4.329148292541504,
+      "learning_rate": 6.819548872180452e-06,
+      "loss": 0.2674,
+      "step": 21150
+    },
+    {
+      "epoch": 31.82,
+      "grad_norm": 4.691008567810059,
+      "learning_rate": 6.818045112781956e-06,
+      "loss": 0.3004,
+      "step": 21160
+    },
+    {
+      "epoch": 31.83,
+      "grad_norm": 7.85280704498291,
+      "learning_rate": 6.816541353383459e-06,
+      "loss": 0.2769,
+      "step": 21170
+    },
+    {
+      "epoch": 31.85,
+      "grad_norm": 7.473185062408447,
+      "learning_rate": 6.815037593984963e-06,
+      "loss": 0.3217,
+      "step": 21180
+    },
+    {
+      "epoch": 31.86,
+      "grad_norm": 5.13551664352417,
+      "learning_rate": 6.813533834586467e-06,
+      "loss": 0.2425,
+      "step": 21190
+    },
+    {
+      "epoch": 31.88,
+      "grad_norm": 4.801725387573242,
+      "learning_rate": 6.812030075187971e-06,
+      "loss": 0.3008,
+      "step": 21200
+    },
+    {
+      "epoch": 31.89,
+      "grad_norm": 6.320078372955322,
+      "learning_rate": 6.810526315789474e-06,
+      "loss": 0.2616,
+      "step": 21210
+    },
+    {
+      "epoch": 31.91,
+      "grad_norm": 7.002920150756836,
+      "learning_rate": 6.809022556390978e-06,
+      "loss": 0.2647,
+      "step": 21220
+    },
+    {
+      "epoch": 31.92,
+      "grad_norm": 7.263726711273193,
+      "learning_rate": 6.807518796992482e-06,
+      "loss": 0.3434,
+      "step": 21230
+    },
+    {
+      "epoch": 31.94,
+      "grad_norm": 6.2287139892578125,
+      "learning_rate": 6.806015037593986e-06,
+      "loss": 0.2775,
+      "step": 21240
+    },
+    {
+      "epoch": 31.95,
+      "grad_norm": 3.209961175918579,
+      "learning_rate": 6.8045112781954896e-06,
+      "loss": 0.2689,
+      "step": 21250
+    },
+    {
+      "epoch": 31.97,
+      "grad_norm": 5.191007614135742,
+      "learning_rate": 6.803007518796994e-06,
+      "loss": 0.2824,
+      "step": 21260
+    },
+    {
+      "epoch": 31.98,
+      "grad_norm": 7.311644077301025,
+      "learning_rate": 6.801503759398497e-06,
+      "loss": 0.3207,
+      "step": 21270
+    },
+    {
+      "epoch": 32.0,
+      "grad_norm": 33.7657356262207,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 0.3268,
+      "step": 21280
+    },
+    {
+      "epoch": 32.0,
+      "eval_accuracy": 0.9312,
+      "eval_loss": 0.2803204655647278,
+      "eval_runtime": 84.7157,
+      "eval_samples_per_second": 118.042,
+      "eval_steps_per_second": 0.472,
+      "step": 21280
+    },
+    {
+      "epoch": 32.02,
+      "grad_norm": 8.19914722442627,
+      "learning_rate": 6.7984962406015045e-06,
+      "loss": 0.3492,
+      "step": 21290
+    },
+    {
+      "epoch": 32.03,
+      "grad_norm": 5.234302043914795,
+      "learning_rate": 6.796992481203009e-06,
+      "loss": 0.2722,
+      "step": 21300
+    },
+    {
+      "epoch": 32.05,
+      "grad_norm": 5.019562244415283,
+      "learning_rate": 6.7954887218045115e-06,
+      "loss": 0.245,
+      "step": 21310
+    },
+    {
+      "epoch": 32.06,
+      "grad_norm": 4.851130962371826,
+      "learning_rate": 6.793984962406016e-06,
+      "loss": 0.3073,
+      "step": 21320
+    },
+    {
+      "epoch": 32.08,
+      "grad_norm": 4.4718241691589355,
+      "learning_rate": 6.792481203007519e-06,
+      "loss": 0.2929,
+      "step": 21330
+    },
+    {
+      "epoch": 32.09,
+      "grad_norm": 4.94941520690918,
+      "learning_rate": 6.790977443609023e-06,
+      "loss": 0.2941,
+      "step": 21340
+    },
+    {
+      "epoch": 32.11,
+      "grad_norm": 5.517939567565918,
+      "learning_rate": 6.789473684210527e-06,
+      "loss": 0.2528,
+      "step": 21350
+    },
+    {
+      "epoch": 32.12,
+      "grad_norm": 6.751891136169434,
+      "learning_rate": 6.78796992481203e-06,
+      "loss": 0.3063,
+      "step": 21360
+    },
+    {
+      "epoch": 32.14,
+      "grad_norm": 6.197683811187744,
+      "learning_rate": 6.786466165413534e-06,
+      "loss": 0.2686,
+      "step": 21370
+    },
+    {
+      "epoch": 32.15,
+      "grad_norm": 5.212826728820801,
+      "learning_rate": 6.784962406015038e-06,
+      "loss": 0.2153,
+      "step": 21380
+    },
+    {
+      "epoch": 32.17,
+      "grad_norm": 6.147881984710693,
+      "learning_rate": 6.783458646616542e-06,
+      "loss": 0.3603,
+      "step": 21390
+    },
+    {
+      "epoch": 32.18,
+      "grad_norm": 6.068096160888672,
+      "learning_rate": 6.781954887218045e-06,
+      "loss": 0.3144,
+      "step": 21400
+    },
+    {
+      "epoch": 32.2,
+      "grad_norm": 6.553431987762451,
+      "learning_rate": 6.780451127819549e-06,
+      "loss": 0.2902,
+      "step": 21410
+    },
+    {
+      "epoch": 32.21,
+      "grad_norm": 6.4035515785217285,
+      "learning_rate": 6.778947368421053e-06,
+      "loss": 0.2156,
+      "step": 21420
+    },
+    {
+      "epoch": 32.23,
+      "grad_norm": 4.220142841339111,
+      "learning_rate": 6.777443609022557e-06,
+      "loss": 0.2905,
+      "step": 21430
+    },
+    {
+      "epoch": 32.24,
+      "grad_norm": 4.17386531829834,
+      "learning_rate": 6.775939849624061e-06,
+      "loss": 0.3002,
+      "step": 21440
+    },
+    {
+      "epoch": 32.26,
+      "grad_norm": 3.2023849487304688,
+      "learning_rate": 6.774436090225564e-06,
+      "loss": 0.3396,
+      "step": 21450
+    },
+    {
+      "epoch": 32.27,
+      "grad_norm": 2.8168628215789795,
+      "learning_rate": 6.772932330827068e-06,
+      "loss": 0.248,
+      "step": 21460
+    },
+    {
+      "epoch": 32.29,
+      "grad_norm": 4.326303482055664,
+      "learning_rate": 6.771428571428572e-06,
+      "loss": 0.2821,
+      "step": 21470
+    },
+    {
+      "epoch": 32.3,
+      "grad_norm": 3.3244757652282715,
+      "learning_rate": 6.769924812030076e-06,
+      "loss": 0.2907,
+      "step": 21480
+    },
+    {
+      "epoch": 32.32,
+      "grad_norm": 5.797328948974609,
+      "learning_rate": 6.76842105263158e-06,
+      "loss": 0.2961,
+      "step": 21490
+    },
+    {
+      "epoch": 32.33,
+      "grad_norm": 3.571841239929199,
+      "learning_rate": 6.766917293233083e-06,
+      "loss": 0.1967,
+      "step": 21500
+    },
+    {
+      "epoch": 32.35,
+      "grad_norm": 5.261970520019531,
+      "learning_rate": 6.765413533834587e-06,
+      "loss": 0.241,
+      "step": 21510
+    },
+    {
+      "epoch": 32.36,
+      "grad_norm": 6.464574813842773,
+      "learning_rate": 6.7639097744360905e-06,
+      "loss": 0.2727,
+      "step": 21520
+    },
+    {
+      "epoch": 32.38,
+      "grad_norm": 4.337974548339844,
+      "learning_rate": 6.762406015037595e-06,
+      "loss": 0.2508,
+      "step": 21530
+    },
+    {
+      "epoch": 32.39,
+      "grad_norm": 3.5019898414611816,
+      "learning_rate": 6.760902255639098e-06,
+      "loss": 0.2173,
+      "step": 21540
+    },
+    {
+      "epoch": 32.41,
+      "grad_norm": 7.147461891174316,
+      "learning_rate": 6.759398496240602e-06,
+      "loss": 0.3561,
+      "step": 21550
+    },
+    {
+      "epoch": 32.42,
+      "grad_norm": 5.347845077514648,
+      "learning_rate": 6.7578947368421054e-06,
+      "loss": 0.2572,
+      "step": 21560
+    },
+    {
+      "epoch": 32.44,
+      "grad_norm": 3.541206121444702,
+      "learning_rate": 6.75639097744361e-06,
+      "loss": 0.266,
+      "step": 21570
+    },
+    {
+      "epoch": 32.45,
+      "grad_norm": 4.880330562591553,
+      "learning_rate": 6.754887218045113e-06,
+      "loss": 0.2968,
+      "step": 21580
+    },
+    {
+      "epoch": 32.47,
+      "grad_norm": 6.621898174285889,
+      "learning_rate": 6.753383458646618e-06,
+      "loss": 0.1805,
+      "step": 21590
+    },
+    {
+      "epoch": 32.48,
+      "grad_norm": 7.193774700164795,
+      "learning_rate": 6.75187969924812e-06,
+      "loss": 0.2282,
+      "step": 21600
+    },
+    {
+      "epoch": 32.5,
+      "grad_norm": 5.0164408683776855,
+      "learning_rate": 6.750375939849625e-06,
+      "loss": 0.3077,
+      "step": 21610
+    },
+    {
+      "epoch": 32.51,
+      "grad_norm": 8.061211585998535,
+      "learning_rate": 6.748872180451128e-06,
+      "loss": 0.2758,
+      "step": 21620
+    },
+    {
+      "epoch": 32.53,
+      "grad_norm": 6.3904852867126465,
+      "learning_rate": 6.747368421052633e-06,
+      "loss": 0.2499,
+      "step": 21630
+    },
+    {
+      "epoch": 32.54,
+      "grad_norm": 8.771563529968262,
+      "learning_rate": 6.745864661654135e-06,
+      "loss": 0.3302,
+      "step": 21640
+    },
+    {
+      "epoch": 32.56,
+      "grad_norm": 7.61922550201416,
+      "learning_rate": 6.74436090225564e-06,
+      "loss": 0.2691,
+      "step": 21650
+    },
+    {
+      "epoch": 32.57,
+      "grad_norm": 7.187370300292969,
+      "learning_rate": 6.742857142857143e-06,
+      "loss": 0.3071,
+      "step": 21660
+    },
+    {
+      "epoch": 32.59,
+      "grad_norm": 3.4237334728240967,
+      "learning_rate": 6.7413533834586475e-06,
+      "loss": 0.296,
+      "step": 21670
+    },
+    {
+      "epoch": 32.6,
+      "grad_norm": 4.828934192657471,
+      "learning_rate": 6.739849624060151e-06,
+      "loss": 0.3158,
+      "step": 21680
+    },
+    {
+      "epoch": 32.62,
+      "grad_norm": 5.043696880340576,
+      "learning_rate": 6.738345864661655e-06,
+      "loss": 0.2656,
+      "step": 21690
+    },
+    {
+      "epoch": 32.63,
+      "grad_norm": 6.4835004806518555,
+      "learning_rate": 6.736842105263158e-06,
+      "loss": 0.2796,
+      "step": 21700
+    },
+    {
+      "epoch": 32.65,
+      "grad_norm": 2.3352560997009277,
+      "learning_rate": 6.7353383458646624e-06,
+      "loss": 0.294,
+      "step": 21710
+    },
+    {
+      "epoch": 32.66,
+      "grad_norm": 7.649822235107422,
+      "learning_rate": 6.733834586466166e-06,
+      "loss": 0.2857,
+      "step": 21720
+    },
+    {
+      "epoch": 32.68,
+      "grad_norm": 5.9862565994262695,
+      "learning_rate": 6.73233082706767e-06,
+      "loss": 0.2401,
+      "step": 21730
+    },
+    {
+      "epoch": 32.69,
+      "grad_norm": 5.723448276519775,
+      "learning_rate": 6.730827067669173e-06,
+      "loss": 0.3426,
+      "step": 21740
+    },
+    {
+      "epoch": 32.71,
+      "grad_norm": 2.8246076107025146,
+      "learning_rate": 6.729323308270677e-06,
+      "loss": 0.2431,
+      "step": 21750
+    },
+    {
+      "epoch": 32.72,
+      "grad_norm": 7.292664527893066,
+      "learning_rate": 6.727819548872181e-06,
+      "loss": 0.2624,
+      "step": 21760
+    },
+    {
+      "epoch": 32.74,
+      "grad_norm": 4.674510478973389,
+      "learning_rate": 6.726315789473685e-06,
+      "loss": 0.3025,
+      "step": 21770
+    },
+    {
+      "epoch": 32.75,
+      "grad_norm": 4.479643821716309,
+      "learning_rate": 6.724812030075189e-06,
+      "loss": 0.2175,
+      "step": 21780
+    },
+    {
+      "epoch": 32.77,
+      "grad_norm": 4.182172775268555,
+      "learning_rate": 6.723308270676693e-06,
+      "loss": 0.3039,
+      "step": 21790
+    },
+    {
+      "epoch": 32.78,
+      "grad_norm": 5.148652076721191,
+      "learning_rate": 6.721804511278196e-06,
+      "loss": 0.2735,
+      "step": 21800
+    },
+    {
+      "epoch": 32.8,
+      "grad_norm": 4.162076950073242,
+      "learning_rate": 6.7203007518797e-06,
+      "loss": 0.2652,
+      "step": 21810
+    },
+    {
+      "epoch": 32.81,
+      "grad_norm": 5.020504474639893,
+      "learning_rate": 6.718796992481204e-06,
+      "loss": 0.3151,
+      "step": 21820
+    },
+    {
+      "epoch": 32.83,
+      "grad_norm": 3.3114843368530273,
+      "learning_rate": 6.717293233082708e-06,
+      "loss": 0.3088,
+      "step": 21830
+    },
+    {
+      "epoch": 32.84,
+      "grad_norm": 7.768105983734131,
+      "learning_rate": 6.715789473684211e-06,
+      "loss": 0.2574,
+      "step": 21840
+    },
+    {
+      "epoch": 32.86,
+      "grad_norm": 5.251821517944336,
+      "learning_rate": 6.714285714285714e-06,
+      "loss": 0.2308,
+      "step": 21850
+    },
+    {
+      "epoch": 32.87,
+      "grad_norm": 3.617433786392212,
+      "learning_rate": 6.712781954887219e-06,
+      "loss": 0.2609,
+      "step": 21860
+    },
+    {
+      "epoch": 32.89,
+      "grad_norm": 7.9688615798950195,
+      "learning_rate": 6.711278195488722e-06,
+      "loss": 0.2844,
+      "step": 21870
+    },
+    {
+      "epoch": 32.9,
+      "grad_norm": 3.2969768047332764,
+      "learning_rate": 6.7097744360902265e-06,
+      "loss": 0.2347,
+      "step": 21880
+    },
+    {
+      "epoch": 32.92,
+      "grad_norm": 7.453803539276123,
+      "learning_rate": 6.708270676691729e-06,
+      "loss": 0.2808,
+      "step": 21890
+    },
+    {
+      "epoch": 32.93,
+      "grad_norm": 4.9906744956970215,
+      "learning_rate": 6.7067669172932335e-06,
+      "loss": 0.2362,
+      "step": 21900
+    },
+    {
+      "epoch": 32.95,
+      "grad_norm": 3.7684195041656494,
+      "learning_rate": 6.705263157894737e-06,
+      "loss": 0.2816,
+      "step": 21910
+    },
+    {
+      "epoch": 32.96,
+      "grad_norm": 4.246355056762695,
+      "learning_rate": 6.7037593984962414e-06,
+      "loss": 0.2608,
+      "step": 21920
+    },
+    {
+      "epoch": 32.98,
+      "grad_norm": 10.584210395812988,
+      "learning_rate": 6.702255639097744e-06,
+      "loss": 0.3571,
+      "step": 21930
+    },
+    {
+      "epoch": 32.99,
+      "grad_norm": 7.687515735626221,
+      "learning_rate": 6.7007518796992485e-06,
+      "loss": 0.218,
+      "step": 21940
+    },
+    {
+      "epoch": 33.0,
+      "eval_accuracy": 0.9307,
+      "eval_loss": 0.28834185004234314,
+      "eval_runtime": 84.8678,
+      "eval_samples_per_second": 117.83,
+      "eval_steps_per_second": 0.471,
+      "step": 21945
+    },
+    {
+      "epoch": 33.01,
+      "grad_norm": 2.6484274864196777,
+      "learning_rate": 6.699248120300752e-06,
+      "loss": 0.2635,
+      "step": 21950
+    },
+    {
+      "epoch": 33.02,
+      "grad_norm": 6.136809825897217,
+      "learning_rate": 6.697744360902256e-06,
+      "loss": 0.2438,
+      "step": 21960
+    },
+    {
+      "epoch": 33.04,
+      "grad_norm": 2.4499707221984863,
+      "learning_rate": 6.69624060150376e-06,
+      "loss": 0.2448,
+      "step": 21970
+    },
+    {
+      "epoch": 33.05,
+      "grad_norm": 8.61922550201416,
+      "learning_rate": 6.694736842105264e-06,
+      "loss": 0.2895,
+      "step": 21980
+    },
+    {
+      "epoch": 33.07,
+      "grad_norm": 4.760676383972168,
+      "learning_rate": 6.693233082706767e-06,
+      "loss": 0.2138,
+      "step": 21990
+    },
+    {
+      "epoch": 33.08,
+      "grad_norm": 4.3553853034973145,
+      "learning_rate": 6.691729323308271e-06,
+      "loss": 0.3182,
+      "step": 22000
+    },
+    {
+      "epoch": 33.1,
+      "grad_norm": 6.192696571350098,
+      "learning_rate": 6.690225563909775e-06,
+      "loss": 0.2446,
+      "step": 22010
+    },
+    {
+      "epoch": 33.11,
+      "grad_norm": 8.466830253601074,
+      "learning_rate": 6.688721804511279e-06,
+      "loss": 0.3202,
+      "step": 22020
+    },
+    {
+      "epoch": 33.13,
+      "grad_norm": 5.688916206359863,
+      "learning_rate": 6.687218045112782e-06,
+      "loss": 0.3136,
+      "step": 22030
+    },
+    {
+      "epoch": 33.14,
+      "grad_norm": 7.089521408081055,
+      "learning_rate": 6.685714285714286e-06,
+      "loss": 0.3003,
+      "step": 22040
+    },
+    {
+      "epoch": 33.16,
+      "grad_norm": 5.240951061248779,
+      "learning_rate": 6.68421052631579e-06,
+      "loss": 0.2772,
+      "step": 22050
+    },
+    {
+      "epoch": 33.17,
+      "grad_norm": 4.90128755569458,
+      "learning_rate": 6.682706766917294e-06,
+      "loss": 0.2949,
+      "step": 22060
+    },
+    {
+      "epoch": 33.19,
+      "grad_norm": 1.9220607280731201,
+      "learning_rate": 6.681203007518798e-06,
+      "loss": 0.2716,
+      "step": 22070
+    },
+    {
+      "epoch": 33.2,
+      "grad_norm": 11.761991500854492,
+      "learning_rate": 6.679699248120302e-06,
+      "loss": 0.2519,
+      "step": 22080
+    },
+    {
+      "epoch": 33.22,
+      "grad_norm": 3.534703016281128,
+      "learning_rate": 6.678195488721805e-06,
+      "loss": 0.2398,
+      "step": 22090
+    },
+    {
+      "epoch": 33.23,
+      "grad_norm": 5.2441887855529785,
+      "learning_rate": 6.676691729323309e-06,
+      "loss": 0.2638,
+      "step": 22100
+    },
+    {
+      "epoch": 33.25,
+      "grad_norm": 5.488763809204102,
+      "learning_rate": 6.6751879699248125e-06,
+      "loss": 0.2381,
+      "step": 22110
+    },
+    {
+      "epoch": 33.26,
+      "grad_norm": 4.989120006561279,
+      "learning_rate": 6.673684210526317e-06,
+      "loss": 0.2582,
+      "step": 22120
+    },
+    {
+      "epoch": 33.28,
+      "grad_norm": 5.641976833343506,
+      "learning_rate": 6.6721804511278196e-06,
+      "loss": 0.2608,
+      "step": 22130
+    },
+    {
+      "epoch": 33.29,
+      "grad_norm": 5.41449499130249,
+      "learning_rate": 6.670676691729324e-06,
+      "loss": 0.3076,
+      "step": 22140
+    },
+    {
+      "epoch": 33.31,
+      "grad_norm": 5.977665424346924,
+      "learning_rate": 6.6691729323308274e-06,
+      "loss": 0.2189,
+      "step": 22150
+    },
+    {
+      "epoch": 33.32,
+      "grad_norm": 6.3149213790893555,
+      "learning_rate": 6.667669172932332e-06,
+      "loss": 0.3281,
+      "step": 22160
+    },
+    {
+      "epoch": 33.34,
+      "grad_norm": 2.910731077194214,
+      "learning_rate": 6.666165413533835e-06,
+      "loss": 0.2563,
+      "step": 22170
+    },
+    {
+      "epoch": 33.35,
+      "grad_norm": 5.496181011199951,
+      "learning_rate": 6.664661654135339e-06,
+      "loss": 0.2087,
+      "step": 22180
+    },
+    {
+      "epoch": 33.37,
+      "grad_norm": 4.933781623840332,
+      "learning_rate": 6.663157894736842e-06,
+      "loss": 0.2249,
+      "step": 22190
+    },
+    {
+      "epoch": 33.38,
+      "grad_norm": 6.685660362243652,
+      "learning_rate": 6.661654135338347e-06,
+      "loss": 0.3751,
+      "step": 22200
+    },
+    {
+      "epoch": 33.4,
+      "grad_norm": 3.7350828647613525,
+      "learning_rate": 6.66015037593985e-06,
+      "loss": 0.2126,
+      "step": 22210
+    },
+    {
+      "epoch": 33.41,
+      "grad_norm": 7.765669822692871,
+      "learning_rate": 6.658646616541355e-06,
+      "loss": 0.3009,
+      "step": 22220
+    },
+    {
+      "epoch": 33.43,
+      "grad_norm": 5.381826877593994,
+      "learning_rate": 6.657142857142857e-06,
+      "loss": 0.3344,
+      "step": 22230
+    },
+    {
+      "epoch": 33.44,
+      "grad_norm": 3.3078157901763916,
+      "learning_rate": 6.655639097744362e-06,
+      "loss": 0.3077,
+      "step": 22240
+    },
+    {
+      "epoch": 33.46,
+      "grad_norm": 3.342693328857422,
+      "learning_rate": 6.654135338345865e-06,
+      "loss": 0.2235,
+      "step": 22250
+    },
+    {
+      "epoch": 33.47,
+      "grad_norm": 5.725019454956055,
+      "learning_rate": 6.6526315789473695e-06,
+      "loss": 0.3253,
+      "step": 22260
+    },
+    {
+      "epoch": 33.49,
+      "grad_norm": 7.685015678405762,
+      "learning_rate": 6.651127819548873e-06,
+      "loss": 0.2343,
+      "step": 22270
+    },
+    {
+      "epoch": 33.5,
+      "grad_norm": 3.291060447692871,
+      "learning_rate": 6.649624060150377e-06,
+      "loss": 0.2907,
+      "step": 22280
+    },
+    {
+      "epoch": 33.52,
+      "grad_norm": 5.01746940612793,
+      "learning_rate": 6.64812030075188e-06,
+      "loss": 0.2124,
+      "step": 22290
+    },
+    {
+      "epoch": 33.53,
+      "grad_norm": 8.099120140075684,
+      "learning_rate": 6.6466165413533845e-06,
+      "loss": 0.2862,
+      "step": 22300
+    },
+    {
+      "epoch": 33.55,
+      "grad_norm": 6.425881862640381,
+      "learning_rate": 6.645112781954888e-06,
+      "loss": 0.216,
+      "step": 22310
+    },
+    {
+      "epoch": 33.56,
+      "grad_norm": 6.16859769821167,
+      "learning_rate": 6.643609022556392e-06,
+      "loss": 0.3214,
+      "step": 22320
+    },
+    {
+      "epoch": 33.58,
+      "grad_norm": 4.426242828369141,
+      "learning_rate": 6.642105263157895e-06,
+      "loss": 0.2471,
+      "step": 22330
+    },
+    {
+      "epoch": 33.59,
+      "grad_norm": 4.199456691741943,
+      "learning_rate": 6.6406015037593985e-06,
+      "loss": 0.2696,
+      "step": 22340
+    },
+    {
+      "epoch": 33.61,
+      "grad_norm": 4.16933536529541,
+      "learning_rate": 6.639097744360903e-06,
+      "loss": 0.1926,
+      "step": 22350
+    },
+    {
+      "epoch": 33.62,
+      "grad_norm": 7.852532863616943,
+      "learning_rate": 6.6375939849624064e-06,
+      "loss": 0.2983,
+      "step": 22360
+    },
+    {
+      "epoch": 33.64,
+      "grad_norm": 5.3104119300842285,
+      "learning_rate": 6.63609022556391e-06,
+      "loss": 0.2412,
+      "step": 22370
+    },
+    {
+      "epoch": 33.65,
+      "grad_norm": 4.502700328826904,
+      "learning_rate": 6.6345864661654135e-06,
+      "loss": 0.3023,
+      "step": 22380
+    },
+    {
+      "epoch": 33.67,
+      "grad_norm": 4.96920108795166,
+      "learning_rate": 6.633082706766918e-06,
+      "loss": 0.2634,
+      "step": 22390
+    },
+    {
+      "epoch": 33.68,
+      "grad_norm": 7.31613826751709,
+      "learning_rate": 6.631578947368421e-06,
+      "loss": 0.3551,
+      "step": 22400
+    },
+    {
+      "epoch": 33.7,
+      "grad_norm": 4.635395526885986,
+      "learning_rate": 6.630075187969926e-06,
+      "loss": 0.2552,
+      "step": 22410
+    },
+    {
+      "epoch": 33.71,
+      "grad_norm": 7.699193477630615,
+      "learning_rate": 6.628571428571428e-06,
+      "loss": 0.2955,
+      "step": 22420
+    },
+    {
+      "epoch": 33.73,
+      "grad_norm": 5.194084644317627,
+      "learning_rate": 6.627067669172933e-06,
+      "loss": 0.2743,
+      "step": 22430
+    },
+    {
+      "epoch": 33.74,
+      "grad_norm": 9.023811340332031,
+      "learning_rate": 6.625563909774436e-06,
+      "loss": 0.2817,
+      "step": 22440
+    },
+    {
+      "epoch": 33.76,
+      "grad_norm": 4.306658744812012,
+      "learning_rate": 6.624060150375941e-06,
+      "loss": 0.3008,
+      "step": 22450
+    },
+    {
+      "epoch": 33.77,
+      "grad_norm": 2.301297187805176,
+      "learning_rate": 6.622556390977443e-06,
+      "loss": 0.2354,
+      "step": 22460
+    },
+    {
+      "epoch": 33.79,
+      "grad_norm": 4.28262186050415,
+      "learning_rate": 6.621052631578948e-06,
+      "loss": 0.2656,
+      "step": 22470
+    },
+    {
+      "epoch": 33.8,
+      "grad_norm": 6.531603813171387,
+      "learning_rate": 6.619548872180451e-06,
+      "loss": 0.3753,
+      "step": 22480
+    },
+    {
+      "epoch": 33.82,
+      "grad_norm": 8.290949821472168,
+      "learning_rate": 6.6180451127819556e-06,
+      "loss": 0.2842,
+      "step": 22490
+    },
+    {
+      "epoch": 33.83,
+      "grad_norm": 3.5719752311706543,
+      "learning_rate": 6.616541353383459e-06,
+      "loss": 0.2617,
+      "step": 22500
+    },
+    {
+      "epoch": 33.85,
+      "grad_norm": 6.209147930145264,
+      "learning_rate": 6.6150375939849635e-06,
+      "loss": 0.3299,
+      "step": 22510
+    },
+    {
+      "epoch": 33.86,
+      "grad_norm": 3.604736566543579,
+      "learning_rate": 6.613533834586466e-06,
+      "loss": 0.1786,
+      "step": 22520
+    },
+    {
+      "epoch": 33.88,
+      "grad_norm": 8.47739028930664,
+      "learning_rate": 6.6120300751879705e-06,
+      "loss": 0.3399,
+      "step": 22530
+    },
+    {
+      "epoch": 33.89,
+      "grad_norm": 4.806243896484375,
+      "learning_rate": 6.610526315789474e-06,
+      "loss": 0.2474,
+      "step": 22540
+    },
+    {
+      "epoch": 33.91,
+      "grad_norm": 4.551697731018066,
+      "learning_rate": 6.609022556390978e-06,
+      "loss": 0.2451,
+      "step": 22550
+    },
+    {
+      "epoch": 33.92,
+      "grad_norm": 4.965244293212891,
+      "learning_rate": 6.607518796992481e-06,
+      "loss": 0.2971,
+      "step": 22560
+    },
+    {
+      "epoch": 33.94,
+      "grad_norm": 4.888105392456055,
+      "learning_rate": 6.606015037593985e-06,
+      "loss": 0.3368,
+      "step": 22570
+    },
+    {
+      "epoch": 33.95,
+      "grad_norm": 7.72111701965332,
+      "learning_rate": 6.604511278195489e-06,
+      "loss": 0.2713,
+      "step": 22580
+    },
+    {
+      "epoch": 33.97,
+      "grad_norm": 4.930686950683594,
+      "learning_rate": 6.603007518796993e-06,
+      "loss": 0.2445,
+      "step": 22590
+    },
+    {
+      "epoch": 33.98,
+      "grad_norm": 6.420777797698975,
+      "learning_rate": 6.601503759398497e-06,
+      "loss": 0.2183,
+      "step": 22600
+    },
+    {
+      "epoch": 34.0,
+      "grad_norm": 0.24922990798950195,
+      "learning_rate": 6.600000000000001e-06,
+      "loss": 0.217,
+      "step": 22610
+    },
+    {
+      "epoch": 34.0,
+      "eval_accuracy": 0.9356,
+      "eval_loss": 0.2865731716156006,
+      "eval_runtime": 84.2017,
+      "eval_samples_per_second": 118.763,
+      "eval_steps_per_second": 0.475,
+      "step": 22610
+    },
+    {
+      "epoch": 34.02,
+      "grad_norm": 5.616089820861816,
+      "learning_rate": 6.598496240601504e-06,
+      "loss": 0.3432,
+      "step": 22620
+    },
+    {
+      "epoch": 34.03,
+      "grad_norm": 1.8118317127227783,
+      "learning_rate": 6.596992481203008e-06,
+      "loss": 0.2664,
+      "step": 22630
+    },
+    {
+      "epoch": 34.05,
+      "grad_norm": 5.037443161010742,
+      "learning_rate": 6.595488721804512e-06,
+      "loss": 0.27,
+      "step": 22640
+    },
+    {
+      "epoch": 34.06,
+      "grad_norm": 3.1154747009277344,
+      "learning_rate": 6.593984962406016e-06,
+      "loss": 0.2335,
+      "step": 22650
+    },
+    {
+      "epoch": 34.08,
+      "grad_norm": 5.144960880279541,
+      "learning_rate": 6.592481203007519e-06,
+      "loss": 0.3204,
+      "step": 22660
+    },
+    {
+      "epoch": 34.09,
+      "grad_norm": 5.48844575881958,
+      "learning_rate": 6.590977443609023e-06,
+      "loss": 0.3072,
+      "step": 22670
+    },
+    {
+      "epoch": 34.11,
+      "grad_norm": 8.889720916748047,
+      "learning_rate": 6.589473684210527e-06,
+      "loss": 0.2398,
+      "step": 22680
+    },
+    {
+      "epoch": 34.12,
+      "grad_norm": 4.211205959320068,
+      "learning_rate": 6.587969924812031e-06,
+      "loss": 0.2425,
+      "step": 22690
+    },
+    {
+      "epoch": 34.14,
+      "grad_norm": 6.309680938720703,
+      "learning_rate": 6.5864661654135345e-06,
+      "loss": 0.2828,
+      "step": 22700
+    },
+    {
+      "epoch": 34.15,
+      "grad_norm": 5.038494110107422,
+      "learning_rate": 6.584962406015039e-06,
+      "loss": 0.2633,
+      "step": 22710
+    },
+    {
+      "epoch": 34.17,
+      "grad_norm": 1.1278971433639526,
+      "learning_rate": 6.583458646616542e-06,
+      "loss": 0.2935,
+      "step": 22720
+    },
+    {
+      "epoch": 34.18,
+      "grad_norm": 5.858100891113281,
+      "learning_rate": 6.581954887218046e-06,
+      "loss": 0.2771,
+      "step": 22730
+    },
+    {
+      "epoch": 34.2,
+      "grad_norm": 6.1508402824401855,
+      "learning_rate": 6.5804511278195495e-06,
+      "loss": 0.3256,
+      "step": 22740
+    },
+    {
+      "epoch": 34.21,
+      "grad_norm": 5.791116237640381,
+      "learning_rate": 6.578947368421054e-06,
+      "loss": 0.2354,
+      "step": 22750
+    },
+    {
+      "epoch": 34.23,
+      "grad_norm": 6.087039470672607,
+      "learning_rate": 6.5774436090225565e-06,
+      "loss": 0.2098,
+      "step": 22760
+    },
+    {
+      "epoch": 34.24,
+      "grad_norm": 6.835604667663574,
+      "learning_rate": 6.575939849624061e-06,
+      "loss": 0.285,
+      "step": 22770
+    },
+    {
+      "epoch": 34.26,
+      "grad_norm": 6.213393688201904,
+      "learning_rate": 6.574436090225564e-06,
+      "loss": 0.2488,
+      "step": 22780
+    },
+    {
+      "epoch": 34.27,
+      "grad_norm": 6.763580322265625,
+      "learning_rate": 6.572932330827069e-06,
+      "loss": 0.2043,
+      "step": 22790
+    },
+    {
+      "epoch": 34.29,
+      "grad_norm": 5.975349426269531,
+      "learning_rate": 6.571428571428572e-06,
+      "loss": 0.3282,
+      "step": 22800
+    },
+    {
+      "epoch": 34.3,
+      "grad_norm": 5.7898406982421875,
+      "learning_rate": 6.569924812030077e-06,
+      "loss": 0.2369,
+      "step": 22810
+    },
+    {
+      "epoch": 34.32,
+      "grad_norm": 3.7875592708587646,
+      "learning_rate": 6.568421052631579e-06,
+      "loss": 0.3036,
+      "step": 22820
+    },
+    {
+      "epoch": 34.33,
+      "grad_norm": 7.54277229309082,
+      "learning_rate": 6.566917293233083e-06,
+      "loss": 0.2441,
+      "step": 22830
+    },
+    {
+      "epoch": 34.35,
+      "grad_norm": 5.216723442077637,
+      "learning_rate": 6.565413533834587e-06,
+      "loss": 0.2537,
+      "step": 22840
+    },
+    {
+      "epoch": 34.36,
+      "grad_norm": 5.119565486907959,
+      "learning_rate": 6.56390977443609e-06,
+      "loss": 0.3228,
+      "step": 22850
+    },
+    {
+      "epoch": 34.38,
+      "grad_norm": 10.754874229431152,
+      "learning_rate": 6.562406015037594e-06,
+      "loss": 0.1848,
+      "step": 22860
+    },
+    {
+      "epoch": 34.39,
+      "grad_norm": 8.560099601745605,
+      "learning_rate": 6.560902255639098e-06,
+      "loss": 0.2648,
+      "step": 22870
+    },
+    {
+      "epoch": 34.41,
+      "grad_norm": 6.255738258361816,
+      "learning_rate": 6.559398496240602e-06,
+      "loss": 0.2562,
+      "step": 22880
+    },
+    {
+      "epoch": 34.42,
+      "grad_norm": 4.35892391204834,
+      "learning_rate": 6.557894736842106e-06,
+      "loss": 0.2838,
+      "step": 22890
+    },
+    {
+      "epoch": 34.44,
+      "grad_norm": 6.469654083251953,
+      "learning_rate": 6.55639097744361e-06,
+      "loss": 0.2885,
+      "step": 22900
+    },
+    {
+      "epoch": 34.45,
+      "grad_norm": 7.00548791885376,
+      "learning_rate": 6.554887218045113e-06,
+      "loss": 0.327,
+      "step": 22910
+    },
+    {
+      "epoch": 34.47,
+      "grad_norm": 6.044936180114746,
+      "learning_rate": 6.553383458646617e-06,
+      "loss": 0.2582,
+      "step": 22920
+    },
+    {
+      "epoch": 34.48,
+      "grad_norm": 6.903324127197266,
+      "learning_rate": 6.5518796992481206e-06,
+      "loss": 0.2673,
+      "step": 22930
+    },
+    {
+      "epoch": 34.5,
+      "grad_norm": 4.936728477478027,
+      "learning_rate": 6.550375939849625e-06,
+      "loss": 0.2469,
+      "step": 22940
+    },
+    {
+      "epoch": 34.51,
+      "grad_norm": 5.086190223693848,
+      "learning_rate": 6.548872180451128e-06,
+      "loss": 0.2323,
+      "step": 22950
+    },
+    {
+      "epoch": 34.53,
+      "grad_norm": 2.871715784072876,
+      "learning_rate": 6.547368421052632e-06,
+      "loss": 0.2158,
+      "step": 22960
+    },
+    {
+      "epoch": 34.54,
+      "grad_norm": 4.698546886444092,
+      "learning_rate": 6.5458646616541355e-06,
+      "loss": 0.257,
+      "step": 22970
+    },
+    {
+      "epoch": 34.56,
+      "grad_norm": 6.885629177093506,
+      "learning_rate": 6.54436090225564e-06,
+      "loss": 0.2187,
+      "step": 22980
+    },
+    {
+      "epoch": 34.57,
+      "grad_norm": 3.128361463546753,
+      "learning_rate": 6.542857142857143e-06,
+      "loss": 0.2876,
+      "step": 22990
+    },
+    {
+      "epoch": 34.59,
+      "grad_norm": 3.115068197250366,
+      "learning_rate": 6.541353383458648e-06,
+      "loss": 0.2565,
+      "step": 23000
+    },
+    {
+      "epoch": 34.6,
+      "grad_norm": 4.966819763183594,
+      "learning_rate": 6.53984962406015e-06,
+      "loss": 0.3301,
+      "step": 23010
+    },
+    {
+      "epoch": 34.62,
+      "grad_norm": 3.8072173595428467,
+      "learning_rate": 6.538345864661655e-06,
+      "loss": 0.2815,
+      "step": 23020
+    },
+    {
+      "epoch": 34.63,
+      "grad_norm": 7.346180438995361,
+      "learning_rate": 6.536842105263158e-06,
+      "loss": 0.2456,
+      "step": 23030
+    },
+    {
+      "epoch": 34.65,
+      "grad_norm": 5.6631317138671875,
+      "learning_rate": 6.535338345864663e-06,
+      "loss": 0.2629,
+      "step": 23040
+    },
+    {
+      "epoch": 34.66,
+      "grad_norm": 4.316861629486084,
+      "learning_rate": 6.533834586466165e-06,
+      "loss": 0.218,
+      "step": 23050
+    },
+    {
+      "epoch": 34.68,
+      "grad_norm": 5.202853202819824,
+      "learning_rate": 6.53233082706767e-06,
+      "loss": 0.3226,
+      "step": 23060
+    },
+    {
+      "epoch": 34.69,
+      "grad_norm": 5.017209529876709,
+      "learning_rate": 6.530827067669173e-06,
+      "loss": 0.3087,
+      "step": 23070
+    },
+    {
+      "epoch": 34.71,
+      "grad_norm": 9.907183647155762,
+      "learning_rate": 6.529323308270678e-06,
+      "loss": 0.2096,
+      "step": 23080
+    },
+    {
+      "epoch": 34.72,
+      "grad_norm": 5.446949005126953,
+      "learning_rate": 6.527819548872181e-06,
+      "loss": 0.2725,
+      "step": 23090
+    },
+    {
+      "epoch": 34.74,
+      "grad_norm": 5.2780537605285645,
+      "learning_rate": 6.526315789473685e-06,
+      "loss": 0.2429,
+      "step": 23100
+    },
+    {
+      "epoch": 34.75,
+      "grad_norm": 3.5522546768188477,
+      "learning_rate": 6.524812030075188e-06,
+      "loss": 0.257,
+      "step": 23110
+    },
+    {
+      "epoch": 34.77,
+      "grad_norm": 6.132695198059082,
+      "learning_rate": 6.5233082706766925e-06,
+      "loss": 0.3128,
+      "step": 23120
+    },
+    {
+      "epoch": 34.78,
+      "grad_norm": 7.575121879577637,
+      "learning_rate": 6.521804511278196e-06,
+      "loss": 0.2529,
+      "step": 23130
+    },
+    {
+      "epoch": 34.8,
+      "grad_norm": 6.6957244873046875,
+      "learning_rate": 6.5203007518797e-06,
+      "loss": 0.2612,
+      "step": 23140
+    },
+    {
+      "epoch": 34.81,
+      "grad_norm": 4.827077865600586,
+      "learning_rate": 6.518796992481203e-06,
+      "loss": 0.307,
+      "step": 23150
+    },
+    {
+      "epoch": 34.83,
+      "grad_norm": 7.132809638977051,
+      "learning_rate": 6.5172932330827074e-06,
+      "loss": 0.2111,
+      "step": 23160
+    },
+    {
+      "epoch": 34.84,
+      "grad_norm": 3.0563974380493164,
+      "learning_rate": 6.515789473684211e-06,
+      "loss": 0.2332,
+      "step": 23170
+    },
+    {
+      "epoch": 34.86,
+      "grad_norm": 5.9987592697143555,
+      "learning_rate": 6.514285714285715e-06,
+      "loss": 0.3418,
+      "step": 23180
+    },
+    {
+      "epoch": 34.87,
+      "grad_norm": 4.280242443084717,
+      "learning_rate": 6.512781954887219e-06,
+      "loss": 0.2447,
+      "step": 23190
+    },
+    {
+      "epoch": 34.89,
+      "grad_norm": 4.552205562591553,
+      "learning_rate": 6.511278195488722e-06,
+      "loss": 0.2697,
+      "step": 23200
+    },
+    {
+      "epoch": 34.9,
+      "grad_norm": 5.720810890197754,
+      "learning_rate": 6.509774436090226e-06,
+      "loss": 0.3134,
+      "step": 23210
+    },
+    {
+      "epoch": 34.92,
+      "grad_norm": 5.941717147827148,
+      "learning_rate": 6.50827067669173e-06,
+      "loss": 0.2535,
+      "step": 23220
+    },
+    {
+      "epoch": 34.93,
+      "grad_norm": 5.756423473358154,
+      "learning_rate": 6.506766917293234e-06,
+      "loss": 0.2717,
+      "step": 23230
+    },
+    {
+      "epoch": 34.95,
+      "grad_norm": 6.748666763305664,
+      "learning_rate": 6.505263157894738e-06,
+      "loss": 0.2522,
+      "step": 23240
+    },
+    {
+      "epoch": 34.96,
+      "grad_norm": 3.929311990737915,
+      "learning_rate": 6.503759398496241e-06,
+      "loss": 0.3033,
+      "step": 23250
+    },
+    {
+      "epoch": 34.98,
+      "grad_norm": 5.266348838806152,
+      "learning_rate": 6.502255639097745e-06,
+      "loss": 0.2428,
+      "step": 23260
+    },
+    {
+      "epoch": 34.99,
+      "grad_norm": 6.238527774810791,
+      "learning_rate": 6.500751879699249e-06,
+      "loss": 0.2032,
+      "step": 23270
+    },
+    {
+      "epoch": 35.0,
+      "eval_accuracy": 0.9317,
+      "eval_loss": 0.29047271609306335,
+      "eval_runtime": 84.6472,
+      "eval_samples_per_second": 118.137,
+      "eval_steps_per_second": 0.473,
+      "step": 23275
+    },
+    {
+      "epoch": 35.01,
+      "grad_norm": 4.792929172515869,
+      "learning_rate": 6.499248120300753e-06,
+      "loss": 0.2373,
+      "step": 23280
+    },
+    {
+      "epoch": 35.02,
+      "grad_norm": 5.885370254516602,
+      "learning_rate": 6.497744360902256e-06,
+      "loss": 0.2296,
+      "step": 23290
+    },
+    {
+      "epoch": 35.04,
+      "grad_norm": 3.2437541484832764,
+      "learning_rate": 6.49624060150376e-06,
+      "loss": 0.3242,
+      "step": 23300
+    },
+    {
+      "epoch": 35.05,
+      "grad_norm": 4.990633964538574,
+      "learning_rate": 6.494736842105264e-06,
+      "loss": 0.3038,
+      "step": 23310
+    },
+    {
+      "epoch": 35.07,
+      "grad_norm": 4.416833877563477,
+      "learning_rate": 6.493233082706768e-06,
+      "loss": 0.2001,
+      "step": 23320
+    },
+    {
+      "epoch": 35.08,
+      "grad_norm": 8.694164276123047,
+      "learning_rate": 6.4917293233082715e-06,
+      "loss": 0.2442,
+      "step": 23330
+    },
+    {
+      "epoch": 35.1,
+      "grad_norm": 9.367162704467773,
+      "learning_rate": 6.490225563909774e-06,
+      "loss": 0.3088,
+      "step": 23340
+    },
+    {
+      "epoch": 35.11,
+      "grad_norm": 3.4011213779449463,
+      "learning_rate": 6.4887218045112785e-06,
+      "loss": 0.2324,
+      "step": 23350
+    },
+    {
+      "epoch": 35.13,
+      "grad_norm": 5.543578624725342,
+      "learning_rate": 6.487218045112782e-06,
+      "loss": 0.2571,
+      "step": 23360
+    },
+    {
+      "epoch": 35.14,
+      "grad_norm": 6.542150974273682,
+      "learning_rate": 6.485714285714286e-06,
+      "loss": 0.233,
+      "step": 23370
+    },
+    {
+      "epoch": 35.16,
+      "grad_norm": 3.572125196456909,
+      "learning_rate": 6.484210526315789e-06,
+      "loss": 0.317,
+      "step": 23380
+    },
+    {
+      "epoch": 35.17,
+      "grad_norm": 6.729075908660889,
+      "learning_rate": 6.4827067669172935e-06,
+      "loss": 0.1965,
+      "step": 23390
+    },
+    {
+      "epoch": 35.19,
+      "grad_norm": 5.777679920196533,
+      "learning_rate": 6.481203007518797e-06,
+      "loss": 0.2543,
+      "step": 23400
+    },
+    {
+      "epoch": 35.2,
+      "grad_norm": 4.606062412261963,
+      "learning_rate": 6.479699248120301e-06,
+      "loss": 0.2478,
+      "step": 23410
+    },
+    {
+      "epoch": 35.22,
+      "grad_norm": 4.347212791442871,
+      "learning_rate": 6.478195488721805e-06,
+      "loss": 0.2716,
+      "step": 23420
+    },
+    {
+      "epoch": 35.23,
+      "grad_norm": 6.0750732421875,
+      "learning_rate": 6.476691729323309e-06,
+      "loss": 0.3095,
+      "step": 23430
+    },
+    {
+      "epoch": 35.25,
+      "grad_norm": 3.2727339267730713,
+      "learning_rate": 6.475187969924812e-06,
+      "loss": 0.2904,
+      "step": 23440
+    },
+    {
+      "epoch": 35.26,
+      "grad_norm": 10.178025245666504,
+      "learning_rate": 6.473684210526316e-06,
+      "loss": 0.2603,
+      "step": 23450
+    },
+    {
+      "epoch": 35.28,
+      "grad_norm": 6.457759380340576,
+      "learning_rate": 6.47218045112782e-06,
+      "loss": 0.3092,
+      "step": 23460
+    },
+    {
+      "epoch": 35.29,
+      "grad_norm": 1.357625126838684,
+      "learning_rate": 6.470676691729324e-06,
+      "loss": 0.1796,
+      "step": 23470
+    },
+    {
+      "epoch": 35.31,
+      "grad_norm": 6.197760105133057,
+      "learning_rate": 6.469172932330827e-06,
+      "loss": 0.331,
+      "step": 23480
+    },
+    {
+      "epoch": 35.32,
+      "grad_norm": 3.364699602127075,
+      "learning_rate": 6.467669172932331e-06,
+      "loss": 0.2559,
+      "step": 23490
+    },
+    {
+      "epoch": 35.34,
+      "grad_norm": 6.3956618309021,
+      "learning_rate": 6.466165413533835e-06,
+      "loss": 0.2026,
+      "step": 23500
+    },
+    {
+      "epoch": 35.35,
+      "grad_norm": 6.079551696777344,
+      "learning_rate": 6.464661654135339e-06,
+      "loss": 0.2687,
+      "step": 23510
+    },
+    {
+      "epoch": 35.37,
+      "grad_norm": 5.305349349975586,
+      "learning_rate": 6.463157894736843e-06,
+      "loss": 0.2628,
+      "step": 23520
+    },
+    {
+      "epoch": 35.38,
+      "grad_norm": 3.124565362930298,
+      "learning_rate": 6.461654135338347e-06,
+      "loss": 0.2969,
+      "step": 23530
+    },
+    {
+      "epoch": 35.4,
+      "grad_norm": 5.130784511566162,
+      "learning_rate": 6.46015037593985e-06,
+      "loss": 0.2696,
+      "step": 23540
+    },
+    {
+      "epoch": 35.41,
+      "grad_norm": 5.65298318862915,
+      "learning_rate": 6.458646616541354e-06,
+      "loss": 0.2035,
+      "step": 23550
+    },
+    {
+      "epoch": 35.43,
+      "grad_norm": 5.994067192077637,
+      "learning_rate": 6.4571428571428575e-06,
+      "loss": 0.3008,
+      "step": 23560
+    },
+    {
+      "epoch": 35.44,
+      "grad_norm": 4.849883556365967,
+      "learning_rate": 6.455639097744362e-06,
+      "loss": 0.2681,
+      "step": 23570
+    },
+    {
+      "epoch": 35.46,
+      "grad_norm": 7.175223350524902,
+      "learning_rate": 6.4541353383458646e-06,
+      "loss": 0.3364,
+      "step": 23580
+    },
+    {
+      "epoch": 35.47,
+      "grad_norm": 5.904483318328857,
+      "learning_rate": 6.452631578947369e-06,
+      "loss": 0.2725,
+      "step": 23590
+    },
+    {
+      "epoch": 35.49,
+      "grad_norm": 6.1249260902404785,
+      "learning_rate": 6.4511278195488724e-06,
+      "loss": 0.3174,
+      "step": 23600
+    },
+    {
+      "epoch": 35.5,
+      "grad_norm": 5.8022332191467285,
+      "learning_rate": 6.449624060150377e-06,
+      "loss": 0.2296,
+      "step": 23610
+    },
+    {
+      "epoch": 35.52,
+      "grad_norm": 4.012286186218262,
+      "learning_rate": 6.44812030075188e-06,
+      "loss": 0.2594,
+      "step": 23620
+    },
+    {
+      "epoch": 35.53,
+      "grad_norm": 5.536571502685547,
+      "learning_rate": 6.446616541353385e-06,
+      "loss": 0.2567,
+      "step": 23630
+    },
+    {
+      "epoch": 35.55,
+      "grad_norm": 5.743587017059326,
+      "learning_rate": 6.445112781954887e-06,
+      "loss": 0.2928,
+      "step": 23640
+    },
+    {
+      "epoch": 35.56,
+      "grad_norm": 9.780457496643066,
+      "learning_rate": 6.443609022556392e-06,
+      "loss": 0.2897,
+      "step": 23650
+    },
+    {
+      "epoch": 35.58,
+      "grad_norm": 4.584736347198486,
+      "learning_rate": 6.442105263157895e-06,
+      "loss": 0.1786,
+      "step": 23660
+    },
+    {
+      "epoch": 35.59,
+      "grad_norm": 1.9408100843429565,
+      "learning_rate": 6.4406015037594e-06,
+      "loss": 0.3392,
+      "step": 23670
+    },
+    {
+      "epoch": 35.61,
+      "grad_norm": 6.093573570251465,
+      "learning_rate": 6.439097744360902e-06,
+      "loss": 0.286,
+      "step": 23680
+    },
+    {
+      "epoch": 35.62,
+      "grad_norm": 7.220097541809082,
+      "learning_rate": 6.437593984962407e-06,
+      "loss": 0.297,
+      "step": 23690
+    },
+    {
+      "epoch": 35.64,
+      "grad_norm": 5.479585647583008,
+      "learning_rate": 6.43609022556391e-06,
+      "loss": 0.269,
+      "step": 23700
+    },
+    {
+      "epoch": 35.65,
+      "grad_norm": 6.430034160614014,
+      "learning_rate": 6.4345864661654145e-06,
+      "loss": 0.2266,
+      "step": 23710
+    },
+    {
+      "epoch": 35.67,
+      "grad_norm": 2.1765003204345703,
+      "learning_rate": 6.433082706766918e-06,
+      "loss": 0.2206,
+      "step": 23720
+    },
+    {
+      "epoch": 35.68,
+      "grad_norm": 4.447315216064453,
+      "learning_rate": 6.431578947368422e-06,
+      "loss": 0.235,
+      "step": 23730
+    },
+    {
+      "epoch": 35.7,
+      "grad_norm": 8.542555809020996,
+      "learning_rate": 6.430075187969925e-06,
+      "loss": 0.2793,
+      "step": 23740
+    },
+    {
+      "epoch": 35.71,
+      "grad_norm": 7.156379699707031,
+      "learning_rate": 6.4285714285714295e-06,
+      "loss": 0.258,
+      "step": 23750
+    },
+    {
+      "epoch": 35.73,
+      "grad_norm": 5.014525413513184,
+      "learning_rate": 6.427067669172933e-06,
+      "loss": 0.2489,
+      "step": 23760
+    },
+    {
+      "epoch": 35.74,
+      "grad_norm": 3.7913882732391357,
+      "learning_rate": 6.425563909774437e-06,
+      "loss": 0.2875,
+      "step": 23770
+    },
+    {
+      "epoch": 35.76,
+      "grad_norm": 5.546550750732422,
+      "learning_rate": 6.42406015037594e-06,
+      "loss": 0.2528,
+      "step": 23780
+    },
+    {
+      "epoch": 35.77,
+      "grad_norm": 5.281984329223633,
+      "learning_rate": 6.422556390977444e-06,
+      "loss": 0.2684,
+      "step": 23790
+    },
+    {
+      "epoch": 35.79,
+      "grad_norm": 4.275721073150635,
+      "learning_rate": 6.421052631578948e-06,
+      "loss": 0.2681,
+      "step": 23800
+    },
+    {
+      "epoch": 35.8,
+      "grad_norm": 5.552488803863525,
+      "learning_rate": 6.419548872180452e-06,
+      "loss": 0.2607,
+      "step": 23810
+    },
+    {
+      "epoch": 35.82,
+      "grad_norm": 4.155028343200684,
+      "learning_rate": 6.418045112781956e-06,
+      "loss": 0.246,
+      "step": 23820
+    },
+    {
+      "epoch": 35.83,
+      "grad_norm": 10.015910148620605,
+      "learning_rate": 6.4165413533834585e-06,
+      "loss": 0.3312,
+      "step": 23830
+    },
+    {
+      "epoch": 35.85,
+      "grad_norm": 6.853920936584473,
+      "learning_rate": 6.415037593984963e-06,
+      "loss": 0.3212,
+      "step": 23840
+    },
+    {
+      "epoch": 35.86,
+      "grad_norm": 7.707338809967041,
+      "learning_rate": 6.413533834586466e-06,
+      "loss": 0.2773,
+      "step": 23850
+    },
+    {
+      "epoch": 35.88,
+      "grad_norm": 6.877796649932861,
+      "learning_rate": 6.412030075187971e-06,
+      "loss": 0.306,
+      "step": 23860
+    },
+    {
+      "epoch": 35.89,
+      "grad_norm": 6.9555487632751465,
+      "learning_rate": 6.410526315789473e-06,
+      "loss": 0.3005,
+      "step": 23870
+    },
+    {
+      "epoch": 35.91,
+      "grad_norm": 5.553525924682617,
+      "learning_rate": 6.409022556390978e-06,
+      "loss": 0.3166,
+      "step": 23880
+    },
+    {
+      "epoch": 35.92,
+      "grad_norm": 3.6754038333892822,
+      "learning_rate": 6.407518796992481e-06,
+      "loss": 0.2731,
+      "step": 23890
+    },
+    {
+      "epoch": 35.94,
+      "grad_norm": 11.875730514526367,
+      "learning_rate": 6.406015037593986e-06,
+      "loss": 0.3046,
+      "step": 23900
+    },
+    {
+      "epoch": 35.95,
+      "grad_norm": 5.181977272033691,
+      "learning_rate": 6.404511278195489e-06,
+      "loss": 0.2291,
+      "step": 23910
+    },
+    {
+      "epoch": 35.97,
+      "grad_norm": 4.224656105041504,
+      "learning_rate": 6.4030075187969935e-06,
+      "loss": 0.2335,
+      "step": 23920
+    },
+    {
+      "epoch": 35.98,
+      "grad_norm": 3.9968817234039307,
+      "learning_rate": 6.401503759398496e-06,
+      "loss": 0.3294,
+      "step": 23930
+    },
+    {
+      "epoch": 36.0,
+      "grad_norm": 0.03734013810753822,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 0.2539,
+      "step": 23940
+    },
+    {
+      "epoch": 36.0,
+      "eval_accuracy": 0.9313,
+      "eval_loss": 0.28184354305267334,
+      "eval_runtime": 84.4602,
+      "eval_samples_per_second": 118.399,
+      "eval_steps_per_second": 0.474,
+      "step": 23940
+    },
+    {
+      "epoch": 36.02,
+      "grad_norm": 4.844832897186279,
+      "learning_rate": 6.398496240601504e-06,
+      "loss": 0.247,
+      "step": 23950
+    },
+    {
+      "epoch": 36.03,
+      "grad_norm": 8.93395709991455,
+      "learning_rate": 6.3969924812030084e-06,
+      "loss": 0.2341,
+      "step": 23960
+    },
+    {
+      "epoch": 36.05,
+      "grad_norm": 6.799169063568115,
+      "learning_rate": 6.395488721804511e-06,
+      "loss": 0.1804,
+      "step": 23970
+    },
+    {
+      "epoch": 36.06,
+      "grad_norm": 6.8090901374816895,
+      "learning_rate": 6.3939849624060155e-06,
+      "loss": 0.3,
+      "step": 23980
+    },
+    {
+      "epoch": 36.08,
+      "grad_norm": 4.476424217224121,
+      "learning_rate": 6.392481203007519e-06,
+      "loss": 0.2273,
+      "step": 23990
+    },
+    {
+      "epoch": 36.09,
+      "grad_norm": 5.188058853149414,
+      "learning_rate": 6.390977443609023e-06,
+      "loss": 0.3322,
+      "step": 24000
+    },
+    {
+      "epoch": 36.11,
+      "grad_norm": 6.895328998565674,
+      "learning_rate": 6.389473684210527e-06,
+      "loss": 0.2552,
+      "step": 24010
+    },
+    {
+      "epoch": 36.12,
+      "grad_norm": 6.052617073059082,
+      "learning_rate": 6.38796992481203e-06,
+      "loss": 0.3099,
+      "step": 24020
+    },
+    {
+      "epoch": 36.14,
+      "grad_norm": 4.1569366455078125,
+      "learning_rate": 6.386466165413534e-06,
+      "loss": 0.2408,
+      "step": 24030
+    },
+    {
+      "epoch": 36.15,
+      "grad_norm": 10.491902351379395,
+      "learning_rate": 6.384962406015038e-06,
+      "loss": 0.3039,
+      "step": 24040
+    },
+    {
+      "epoch": 36.17,
+      "grad_norm": 1.8690141439437866,
+      "learning_rate": 6.383458646616542e-06,
+      "loss": 0.2466,
+      "step": 24050
+    },
+    {
+      "epoch": 36.18,
+      "grad_norm": 6.9134650230407715,
+      "learning_rate": 6.381954887218046e-06,
+      "loss": 0.2627,
+      "step": 24060
+    },
+    {
+      "epoch": 36.2,
+      "grad_norm": 6.333374977111816,
+      "learning_rate": 6.380451127819549e-06,
+      "loss": 0.2438,
+      "step": 24070
+    },
+    {
+      "epoch": 36.21,
+      "grad_norm": 5.581048488616943,
+      "learning_rate": 6.378947368421053e-06,
+      "loss": 0.2583,
+      "step": 24080
+    },
+    {
+      "epoch": 36.23,
+      "grad_norm": 14.36394214630127,
+      "learning_rate": 6.377443609022557e-06,
+      "loss": 0.29,
+      "step": 24090
+    },
+    {
+      "epoch": 36.24,
+      "grad_norm": 3.045477867126465,
+      "learning_rate": 6.375939849624061e-06,
+      "loss": 0.238,
+      "step": 24100
+    },
+    {
+      "epoch": 36.26,
+      "grad_norm": 7.110077857971191,
+      "learning_rate": 6.374436090225565e-06,
+      "loss": 0.3271,
+      "step": 24110
+    },
+    {
+      "epoch": 36.27,
+      "grad_norm": 5.218198299407959,
+      "learning_rate": 6.372932330827068e-06,
+      "loss": 0.2952,
+      "step": 24120
+    },
+    {
+      "epoch": 36.29,
+      "grad_norm": 10.754549026489258,
+      "learning_rate": 6.371428571428572e-06,
+      "loss": 0.2661,
+      "step": 24130
+    },
+    {
+      "epoch": 36.3,
+      "grad_norm": 3.623293876647949,
+      "learning_rate": 6.369924812030076e-06,
+      "loss": 0.189,
+      "step": 24140
+    },
+    {
+      "epoch": 36.32,
+      "grad_norm": 5.730505466461182,
+      "learning_rate": 6.3684210526315795e-06,
+      "loss": 0.2611,
+      "step": 24150
+    },
+    {
+      "epoch": 36.33,
+      "grad_norm": 6.075179100036621,
+      "learning_rate": 6.366917293233084e-06,
+      "loss": 0.3031,
+      "step": 24160
+    },
+    {
+      "epoch": 36.35,
+      "grad_norm": 5.460084915161133,
+      "learning_rate": 6.365413533834587e-06,
+      "loss": 0.2325,
+      "step": 24170
+    },
+    {
+      "epoch": 36.36,
+      "grad_norm": 5.883102893829346,
+      "learning_rate": 6.363909774436091e-06,
+      "loss": 0.3088,
+      "step": 24180
+    },
+    {
+      "epoch": 36.38,
+      "grad_norm": 2.4602198600769043,
+      "learning_rate": 6.3624060150375945e-06,
+      "loss": 0.2467,
+      "step": 24190
+    },
+    {
+      "epoch": 36.39,
+      "grad_norm": 7.747447490692139,
+      "learning_rate": 6.360902255639099e-06,
+      "loss": 0.2818,
+      "step": 24200
+    },
+    {
+      "epoch": 36.41,
+      "grad_norm": 5.207604885101318,
+      "learning_rate": 6.3593984962406015e-06,
+      "loss": 0.2313,
+      "step": 24210
+    },
+    {
+      "epoch": 36.42,
+      "grad_norm": 3.6121327877044678,
+      "learning_rate": 6.357894736842106e-06,
+      "loss": 0.2216,
+      "step": 24220
+    },
+    {
+      "epoch": 36.44,
+      "grad_norm": 8.649584770202637,
+      "learning_rate": 6.356390977443609e-06,
+      "loss": 0.2514,
+      "step": 24230
+    },
+    {
+      "epoch": 36.45,
+      "grad_norm": 8.641491889953613,
+      "learning_rate": 6.354887218045114e-06,
+      "loss": 0.2535,
+      "step": 24240
+    },
+    {
+      "epoch": 36.47,
+      "grad_norm": 5.185530662536621,
+      "learning_rate": 6.353383458646617e-06,
+      "loss": 0.1882,
+      "step": 24250
+    },
+    {
+      "epoch": 36.48,
+      "grad_norm": 6.481791019439697,
+      "learning_rate": 6.351879699248122e-06,
+      "loss": 0.2954,
+      "step": 24260
+    },
+    {
+      "epoch": 36.5,
+      "grad_norm": 3.8771073818206787,
+      "learning_rate": 6.350375939849624e-06,
+      "loss": 0.2594,
+      "step": 24270
+    },
+    {
+      "epoch": 36.51,
+      "grad_norm": 3.2521779537200928,
+      "learning_rate": 6.348872180451129e-06,
+      "loss": 0.1955,
+      "step": 24280
+    },
+    {
+      "epoch": 36.53,
+      "grad_norm": 4.331993103027344,
+      "learning_rate": 6.347368421052632e-06,
+      "loss": 0.2964,
+      "step": 24290
+    },
+    {
+      "epoch": 36.54,
+      "grad_norm": 5.428518295288086,
+      "learning_rate": 6.3458646616541366e-06,
+      "loss": 0.3339,
+      "step": 24300
+    },
+    {
+      "epoch": 36.56,
+      "grad_norm": 5.912198543548584,
+      "learning_rate": 6.344360902255639e-06,
+      "loss": 0.2986,
+      "step": 24310
+    },
+    {
+      "epoch": 36.57,
+      "grad_norm": 6.707327365875244,
+      "learning_rate": 6.342857142857143e-06,
+      "loss": 0.2508,
+      "step": 24320
+    },
+    {
+      "epoch": 36.59,
+      "grad_norm": 5.767425060272217,
+      "learning_rate": 6.341353383458647e-06,
+      "loss": 0.2468,
+      "step": 24330
+    },
+    {
+      "epoch": 36.6,
+      "grad_norm": 6.399289608001709,
+      "learning_rate": 6.339849624060151e-06,
+      "loss": 0.2553,
+      "step": 24340
+    },
+    {
+      "epoch": 36.62,
+      "grad_norm": 4.582364082336426,
+      "learning_rate": 6.338345864661655e-06,
+      "loss": 0.2534,
+      "step": 24350
+    },
+    {
+      "epoch": 36.63,
+      "grad_norm": 3.3339309692382812,
+      "learning_rate": 6.336842105263158e-06,
+      "loss": 0.2576,
+      "step": 24360
+    },
+    {
+      "epoch": 36.65,
+      "grad_norm": 4.761340141296387,
+      "learning_rate": 6.335338345864662e-06,
+      "loss": 0.2606,
+      "step": 24370
+    },
+    {
+      "epoch": 36.66,
+      "grad_norm": 4.896874904632568,
+      "learning_rate": 6.3338345864661656e-06,
+      "loss": 0.2359,
+      "step": 24380
+    },
+    {
+      "epoch": 36.68,
+      "grad_norm": 8.0689697265625,
+      "learning_rate": 6.33233082706767e-06,
+      "loss": 0.2927,
+      "step": 24390
+    },
+    {
+      "epoch": 36.69,
+      "grad_norm": 7.064751148223877,
+      "learning_rate": 6.330827067669173e-06,
+      "loss": 0.2416,
+      "step": 24400
+    },
+    {
+      "epoch": 36.71,
+      "grad_norm": 4.654138565063477,
+      "learning_rate": 6.329323308270677e-06,
+      "loss": 0.2719,
+      "step": 24410
+    },
+    {
+      "epoch": 36.72,
+      "grad_norm": 3.553248167037964,
+      "learning_rate": 6.3278195488721805e-06,
+      "loss": 0.2545,
+      "step": 24420
+    },
+    {
+      "epoch": 36.74,
+      "grad_norm": 4.331087589263916,
+      "learning_rate": 6.326315789473685e-06,
+      "loss": 0.2351,
+      "step": 24430
+    },
+    {
+      "epoch": 36.75,
+      "grad_norm": 5.628582954406738,
+      "learning_rate": 6.324812030075188e-06,
+      "loss": 0.3051,
+      "step": 24440
+    },
+    {
+      "epoch": 36.77,
+      "grad_norm": 6.611728191375732,
+      "learning_rate": 6.323308270676693e-06,
+      "loss": 0.24,
+      "step": 24450
+    },
+    {
+      "epoch": 36.78,
+      "grad_norm": 5.588087558746338,
+      "learning_rate": 6.321804511278195e-06,
+      "loss": 0.2702,
+      "step": 24460
+    },
+    {
+      "epoch": 36.8,
+      "grad_norm": 2.716883420944214,
+      "learning_rate": 6.3203007518797e-06,
+      "loss": 0.2452,
+      "step": 24470
+    },
+    {
+      "epoch": 36.81,
+      "grad_norm": 3.0164976119995117,
+      "learning_rate": 6.318796992481203e-06,
+      "loss": 0.2998,
+      "step": 24480
+    },
+    {
+      "epoch": 36.83,
+      "grad_norm": 9.347999572753906,
+      "learning_rate": 6.317293233082708e-06,
+      "loss": 0.2446,
+      "step": 24490
+    },
+    {
+      "epoch": 36.84,
+      "grad_norm": 13.030348777770996,
+      "learning_rate": 6.31578947368421e-06,
+      "loss": 0.2017,
+      "step": 24500
+    },
+    {
+      "epoch": 36.86,
+      "grad_norm": 4.36257791519165,
+      "learning_rate": 6.314285714285715e-06,
+      "loss": 0.331,
+      "step": 24510
+    },
+    {
+      "epoch": 36.87,
+      "grad_norm": 6.986863136291504,
+      "learning_rate": 6.312781954887218e-06,
+      "loss": 0.289,
+      "step": 24520
+    },
+    {
+      "epoch": 36.89,
+      "grad_norm": 2.9986495971679688,
+      "learning_rate": 6.311278195488723e-06,
+      "loss": 0.2632,
+      "step": 24530
+    },
+    {
+      "epoch": 36.9,
+      "grad_norm": 6.963717460632324,
+      "learning_rate": 6.309774436090226e-06,
+      "loss": 0.283,
+      "step": 24540
+    },
+    {
+      "epoch": 36.92,
+      "grad_norm": 2.926877975463867,
+      "learning_rate": 6.3082706766917305e-06,
+      "loss": 0.2462,
+      "step": 24550
+    },
+    {
+      "epoch": 36.93,
+      "grad_norm": 2.751361131668091,
+      "learning_rate": 6.306766917293233e-06,
+      "loss": 0.2986,
+      "step": 24560
+    },
+    {
+      "epoch": 36.95,
+      "grad_norm": 7.566164493560791,
+      "learning_rate": 6.3052631578947375e-06,
+      "loss": 0.2437,
+      "step": 24570
+    },
+    {
+      "epoch": 36.96,
+      "grad_norm": 10.196669578552246,
+      "learning_rate": 6.303759398496241e-06,
+      "loss": 0.3045,
+      "step": 24580
+    },
+    {
+      "epoch": 36.98,
+      "grad_norm": 3.9259514808654785,
+      "learning_rate": 6.302255639097745e-06,
+      "loss": 0.2523,
+      "step": 24590
+    },
+    {
+      "epoch": 36.99,
+      "grad_norm": 3.9328627586364746,
+      "learning_rate": 6.300751879699248e-06,
+      "loss": 0.2104,
+      "step": 24600
+    },
+    {
+      "epoch": 37.0,
+      "eval_accuracy": 0.9329,
+      "eval_loss": 0.2907086908817291,
+      "eval_runtime": 84.7416,
+      "eval_samples_per_second": 118.006,
+      "eval_steps_per_second": 0.472,
+      "step": 24605
+    },
+    {
+      "epoch": 37.01,
+      "grad_norm": 5.213517189025879,
+      "learning_rate": 6.2992481203007524e-06,
+      "loss": 0.2175,
+      "step": 24610
+    },
+    {
+      "epoch": 37.02,
+      "grad_norm": 5.479267120361328,
+      "learning_rate": 6.297744360902256e-06,
+      "loss": 0.2723,
+      "step": 24620
+    },
+    {
+      "epoch": 37.04,
+      "grad_norm": 5.910623073577881,
+      "learning_rate": 6.29624060150376e-06,
+      "loss": 0.2277,
+      "step": 24630
+    },
+    {
+      "epoch": 37.05,
+      "grad_norm": 3.9250035285949707,
+      "learning_rate": 6.294736842105264e-06,
+      "loss": 0.2364,
+      "step": 24640
+    },
+    {
+      "epoch": 37.07,
+      "grad_norm": 5.7687201499938965,
+      "learning_rate": 6.293233082706768e-06,
+      "loss": 0.2434,
+      "step": 24650
+    },
+    {
+      "epoch": 37.08,
+      "grad_norm": 6.585549354553223,
+      "learning_rate": 6.291729323308271e-06,
+      "loss": 0.2501,
+      "step": 24660
+    },
+    {
+      "epoch": 37.1,
+      "grad_norm": 9.130806922912598,
+      "learning_rate": 6.290225563909775e-06,
+      "loss": 0.3176,
+      "step": 24670
+    },
+    {
+      "epoch": 37.11,
+      "grad_norm": 5.826327800750732,
+      "learning_rate": 6.288721804511279e-06,
+      "loss": 0.259,
+      "step": 24680
+    },
+    {
+      "epoch": 37.13,
+      "grad_norm": 3.6899287700653076,
+      "learning_rate": 6.287218045112783e-06,
+      "loss": 0.225,
+      "step": 24690
+    },
+    {
+      "epoch": 37.14,
+      "grad_norm": 7.567675590515137,
+      "learning_rate": 6.285714285714286e-06,
+      "loss": 0.3511,
+      "step": 24700
+    },
+    {
+      "epoch": 37.16,
+      "grad_norm": 4.670961856842041,
+      "learning_rate": 6.28421052631579e-06,
+      "loss": 0.2718,
+      "step": 24710
+    },
+    {
+      "epoch": 37.17,
+      "grad_norm": 6.8157057762146,
+      "learning_rate": 6.282706766917294e-06,
+      "loss": 0.2676,
+      "step": 24720
+    },
+    {
+      "epoch": 37.19,
+      "grad_norm": 9.021499633789062,
+      "learning_rate": 6.281203007518798e-06,
+      "loss": 0.2615,
+      "step": 24730
+    },
+    {
+      "epoch": 37.2,
+      "grad_norm": 5.826773166656494,
+      "learning_rate": 6.2796992481203016e-06,
+      "loss": 0.2652,
+      "step": 24740
+    },
+    {
+      "epoch": 37.22,
+      "grad_norm": 6.296904563903809,
+      "learning_rate": 6.278195488721806e-06,
+      "loss": 0.2632,
+      "step": 24750
+    },
+    {
+      "epoch": 37.23,
+      "grad_norm": 3.506654739379883,
+      "learning_rate": 6.276691729323309e-06,
+      "loss": 0.2946,
+      "step": 24760
+    },
+    {
+      "epoch": 37.25,
+      "grad_norm": 6.164925575256348,
+      "learning_rate": 6.275187969924813e-06,
+      "loss": 0.2214,
+      "step": 24770
+    },
+    {
+      "epoch": 37.26,
+      "grad_norm": 5.420670986175537,
+      "learning_rate": 6.2736842105263165e-06,
+      "loss": 0.2709,
+      "step": 24780
+    },
+    {
+      "epoch": 37.28,
+      "grad_norm": 5.200439453125,
+      "learning_rate": 6.272180451127821e-06,
+      "loss": 0.2599,
+      "step": 24790
+    },
+    {
+      "epoch": 37.29,
+      "grad_norm": 5.116550445556641,
+      "learning_rate": 6.2706766917293235e-06,
+      "loss": 0.2643,
+      "step": 24800
+    },
+    {
+      "epoch": 37.31,
+      "grad_norm": 4.102775573730469,
+      "learning_rate": 6.269172932330828e-06,
+      "loss": 0.2768,
+      "step": 24810
+    },
+    {
+      "epoch": 37.32,
+      "grad_norm": 5.1978254318237305,
+      "learning_rate": 6.267669172932331e-06,
+      "loss": 0.2132,
+      "step": 24820
+    },
+    {
+      "epoch": 37.34,
+      "grad_norm": 7.613659858703613,
+      "learning_rate": 6.266165413533835e-06,
+      "loss": 0.3066,
+      "step": 24830
+    },
+    {
+      "epoch": 37.35,
+      "grad_norm": 7.9959917068481445,
+      "learning_rate": 6.264661654135339e-06,
+      "loss": 0.2219,
+      "step": 24840
+    },
+    {
+      "epoch": 37.37,
+      "grad_norm": 7.0650811195373535,
+      "learning_rate": 6.263157894736842e-06,
+      "loss": 0.2497,
+      "step": 24850
+    },
+    {
+      "epoch": 37.38,
+      "grad_norm": 4.512048721313477,
+      "learning_rate": 6.261654135338346e-06,
+      "loss": 0.243,
+      "step": 24860
+    },
+    {
+      "epoch": 37.4,
+      "grad_norm": 3.7339611053466797,
+      "learning_rate": 6.26015037593985e-06,
+      "loss": 0.1984,
+      "step": 24870
+    },
+    {
+      "epoch": 37.41,
+      "grad_norm": 5.475858211517334,
+      "learning_rate": 6.258646616541354e-06,
+      "loss": 0.2657,
+      "step": 24880
+    },
+    {
+      "epoch": 37.43,
+      "grad_norm": 5.605356693267822,
+      "learning_rate": 6.257142857142857e-06,
+      "loss": 0.2905,
+      "step": 24890
+    },
+    {
+      "epoch": 37.44,
+      "grad_norm": 9.095086097717285,
+      "learning_rate": 6.255639097744361e-06,
+      "loss": 0.3472,
+      "step": 24900
+    },
+    {
+      "epoch": 37.46,
+      "grad_norm": 6.019258975982666,
+      "learning_rate": 6.254135338345865e-06,
+      "loss": 0.3156,
+      "step": 24910
+    },
+    {
+      "epoch": 37.47,
+      "grad_norm": 1.5480716228485107,
+      "learning_rate": 6.252631578947369e-06,
+      "loss": 0.244,
+      "step": 24920
+    },
+    {
+      "epoch": 37.49,
+      "grad_norm": 5.178092002868652,
+      "learning_rate": 6.251127819548873e-06,
+      "loss": 0.2058,
+      "step": 24930
+    },
+    {
+      "epoch": 37.5,
+      "grad_norm": 5.011317729949951,
+      "learning_rate": 6.249624060150376e-06,
+      "loss": 0.2563,
+      "step": 24940
+    },
+    {
+      "epoch": 37.52,
+      "grad_norm": 6.064964771270752,
+      "learning_rate": 6.24812030075188e-06,
+      "loss": 0.2321,
+      "step": 24950
+    },
+    {
+      "epoch": 37.53,
+      "grad_norm": 4.71431827545166,
+      "learning_rate": 6.246616541353384e-06,
+      "loss": 0.2923,
+      "step": 24960
+    },
+    {
+      "epoch": 37.55,
+      "grad_norm": 7.876862525939941,
+      "learning_rate": 6.245112781954888e-06,
+      "loss": 0.3188,
+      "step": 24970
+    },
+    {
+      "epoch": 37.56,
+      "grad_norm": 5.848691940307617,
+      "learning_rate": 6.243609022556392e-06,
+      "loss": 0.2144,
+      "step": 24980
+    },
+    {
+      "epoch": 37.58,
+      "grad_norm": 10.56225299835205,
+      "learning_rate": 6.242105263157895e-06,
+      "loss": 0.2761,
+      "step": 24990
+    },
+    {
+      "epoch": 37.59,
+      "grad_norm": 4.970668792724609,
+      "learning_rate": 6.240601503759399e-06,
+      "loss": 0.2745,
+      "step": 25000
+    },
+    {
+      "epoch": 37.61,
+      "grad_norm": 9.470555305480957,
+      "learning_rate": 6.2390977443609025e-06,
+      "loss": 0.2858,
+      "step": 25010
+    },
+    {
+      "epoch": 37.62,
+      "grad_norm": 7.466954708099365,
+      "learning_rate": 6.237593984962407e-06,
+      "loss": 0.2461,
+      "step": 25020
+    },
+    {
+      "epoch": 37.64,
+      "grad_norm": 7.034529685974121,
+      "learning_rate": 6.23609022556391e-06,
+      "loss": 0.2675,
+      "step": 25030
+    },
+    {
+      "epoch": 37.65,
+      "grad_norm": 11.616925239562988,
+      "learning_rate": 6.234586466165414e-06,
+      "loss": 0.2296,
+      "step": 25040
+    },
+    {
+      "epoch": 37.67,
+      "grad_norm": 4.31875467300415,
+      "learning_rate": 6.2330827067669174e-06,
+      "loss": 0.2634,
+      "step": 25050
+    },
+    {
+      "epoch": 37.68,
+      "grad_norm": 5.886952877044678,
+      "learning_rate": 6.231578947368422e-06,
+      "loss": 0.2379,
+      "step": 25060
+    },
+    {
+      "epoch": 37.7,
+      "grad_norm": 4.664031982421875,
+      "learning_rate": 6.230075187969925e-06,
+      "loss": 0.2202,
+      "step": 25070
+    },
+    {
+      "epoch": 37.71,
+      "grad_norm": 4.743470191955566,
+      "learning_rate": 6.22857142857143e-06,
+      "loss": 0.2281,
+      "step": 25080
+    },
+    {
+      "epoch": 37.73,
+      "grad_norm": 2.948579788208008,
+      "learning_rate": 6.227067669172932e-06,
+      "loss": 0.225,
+      "step": 25090
+    },
+    {
+      "epoch": 37.74,
+      "grad_norm": 4.467003345489502,
+      "learning_rate": 6.225563909774437e-06,
+      "loss": 0.241,
+      "step": 25100
+    },
+    {
+      "epoch": 37.76,
+      "grad_norm": 5.358434677124023,
+      "learning_rate": 6.22406015037594e-06,
+      "loss": 0.2336,
+      "step": 25110
+    },
+    {
+      "epoch": 37.77,
+      "grad_norm": 3.714233875274658,
+      "learning_rate": 6.222556390977445e-06,
+      "loss": 0.2112,
+      "step": 25120
+    },
+    {
+      "epoch": 37.79,
+      "grad_norm": 6.145929336547852,
+      "learning_rate": 6.221052631578947e-06,
+      "loss": 0.3233,
+      "step": 25130
+    },
+    {
+      "epoch": 37.8,
+      "grad_norm": 5.789137840270996,
+      "learning_rate": 6.219548872180452e-06,
+      "loss": 0.2396,
+      "step": 25140
+    },
+    {
+      "epoch": 37.82,
+      "grad_norm": 7.418402671813965,
+      "learning_rate": 6.218045112781955e-06,
+      "loss": 0.2444,
+      "step": 25150
+    },
+    {
+      "epoch": 37.83,
+      "grad_norm": 4.341475486755371,
+      "learning_rate": 6.2165413533834595e-06,
+      "loss": 0.2258,
+      "step": 25160
+    },
+    {
+      "epoch": 37.85,
+      "grad_norm": 5.642782211303711,
+      "learning_rate": 6.215037593984963e-06,
+      "loss": 0.2946,
+      "step": 25170
+    },
+    {
+      "epoch": 37.86,
+      "grad_norm": 6.9304680824279785,
+      "learning_rate": 6.213533834586467e-06,
+      "loss": 0.2476,
+      "step": 25180
+    },
+    {
+      "epoch": 37.88,
+      "grad_norm": 5.448428153991699,
+      "learning_rate": 6.21203007518797e-06,
+      "loss": 0.2455,
+      "step": 25190
+    },
+    {
+      "epoch": 37.89,
+      "grad_norm": 5.857348918914795,
+      "learning_rate": 6.2105263157894745e-06,
+      "loss": 0.2786,
+      "step": 25200
+    },
+    {
+      "epoch": 37.91,
+      "grad_norm": 3.1548938751220703,
+      "learning_rate": 6.209022556390978e-06,
+      "loss": 0.2123,
+      "step": 25210
+    },
+    {
+      "epoch": 37.92,
+      "grad_norm": 5.268825531005859,
+      "learning_rate": 6.207518796992482e-06,
+      "loss": 0.1931,
+      "step": 25220
+    },
+    {
+      "epoch": 37.94,
+      "grad_norm": 5.20718240737915,
+      "learning_rate": 6.206015037593985e-06,
+      "loss": 0.2774,
+      "step": 25230
+    },
+    {
+      "epoch": 37.95,
+      "grad_norm": 7.931674957275391,
+      "learning_rate": 6.204511278195489e-06,
+      "loss": 0.2652,
+      "step": 25240
+    },
+    {
+      "epoch": 37.97,
+      "grad_norm": 4.287258625030518,
+      "learning_rate": 6.203007518796993e-06,
+      "loss": 0.3186,
+      "step": 25250
+    },
+    {
+      "epoch": 37.98,
+      "grad_norm": 8.485033988952637,
+      "learning_rate": 6.201503759398497e-06,
+      "loss": 0.2543,
+      "step": 25260
+    },
+    {
+      "epoch": 38.0,
+      "grad_norm": 0.33357012271881104,
+      "learning_rate": 6.200000000000001e-06,
+      "loss": 0.264,
+      "step": 25270
+    },
+    {
+      "epoch": 38.0,
+      "eval_accuracy": 0.9298,
+      "eval_loss": 0.3029746413230896,
+      "eval_runtime": 84.4594,
+      "eval_samples_per_second": 118.4,
+      "eval_steps_per_second": 0.474,
+      "step": 25270
+    },
+    {
+      "epoch": 38.02,
+      "grad_norm": 5.222786903381348,
+      "learning_rate": 6.198496240601505e-06,
+      "loss": 0.2575,
+      "step": 25280
+    },
+    {
+      "epoch": 38.03,
+      "grad_norm": 2.386619806289673,
+      "learning_rate": 6.196992481203008e-06,
+      "loss": 0.2419,
+      "step": 25290
+    },
+    {
+      "epoch": 38.05,
+      "grad_norm": 3.8002138137817383,
+      "learning_rate": 6.195488721804512e-06,
+      "loss": 0.2632,
+      "step": 25300
+    },
+    {
+      "epoch": 38.06,
+      "grad_norm": 4.710455417633057,
+      "learning_rate": 6.193984962406016e-06,
+      "loss": 0.1841,
+      "step": 25310
+    },
+    {
+      "epoch": 38.08,
+      "grad_norm": 3.5531041622161865,
+      "learning_rate": 6.192481203007518e-06,
+      "loss": 0.2501,
+      "step": 25320
+    },
+    {
+      "epoch": 38.09,
+      "grad_norm": 4.9958062171936035,
+      "learning_rate": 6.190977443609023e-06,
+      "loss": 0.2728,
+      "step": 25330
+    },
+    {
+      "epoch": 38.11,
+      "grad_norm": 3.301966428756714,
+      "learning_rate": 6.189473684210526e-06,
+      "loss": 0.2564,
+      "step": 25340
+    },
+    {
+      "epoch": 38.12,
+      "grad_norm": 4.437223434448242,
+      "learning_rate": 6.187969924812031e-06,
+      "loss": 0.3406,
+      "step": 25350
+    },
+    {
+      "epoch": 38.14,
+      "grad_norm": 3.1734423637390137,
+      "learning_rate": 6.186466165413534e-06,
+      "loss": 0.2296,
+      "step": 25360
+    },
+    {
+      "epoch": 38.15,
+      "grad_norm": 8.440972328186035,
+      "learning_rate": 6.1849624060150385e-06,
+      "loss": 0.2089,
+      "step": 25370
+    },
+    {
+      "epoch": 38.17,
+      "grad_norm": 3.5104334354400635,
+      "learning_rate": 6.183458646616541e-06,
+      "loss": 0.3243,
+      "step": 25380
+    },
+    {
+      "epoch": 38.18,
+      "grad_norm": 5.661349296569824,
+      "learning_rate": 6.1819548872180455e-06,
+      "loss": 0.2984,
+      "step": 25390
+    },
+    {
+      "epoch": 38.2,
+      "grad_norm": 4.282464981079102,
+      "learning_rate": 6.180451127819549e-06,
+      "loss": 0.2493,
+      "step": 25400
+    },
+    {
+      "epoch": 38.21,
+      "grad_norm": 4.8217926025390625,
+      "learning_rate": 6.1789473684210534e-06,
+      "loss": 0.289,
+      "step": 25410
+    },
+    {
+      "epoch": 38.23,
+      "grad_norm": 5.01459264755249,
+      "learning_rate": 6.177443609022556e-06,
+      "loss": 0.3262,
+      "step": 25420
+    },
+    {
+      "epoch": 38.24,
+      "grad_norm": 5.024478435516357,
+      "learning_rate": 6.1759398496240605e-06,
+      "loss": 0.297,
+      "step": 25430
+    },
+    {
+      "epoch": 38.26,
+      "grad_norm": 9.888729095458984,
+      "learning_rate": 6.174436090225564e-06,
+      "loss": 0.2414,
+      "step": 25440
+    },
+    {
+      "epoch": 38.27,
+      "grad_norm": 6.393784046173096,
+      "learning_rate": 6.172932330827068e-06,
+      "loss": 0.2329,
+      "step": 25450
+    },
+    {
+      "epoch": 38.29,
+      "grad_norm": 4.191177845001221,
+      "learning_rate": 6.171428571428572e-06,
+      "loss": 0.303,
+      "step": 25460
+    },
+    {
+      "epoch": 38.3,
+      "grad_norm": 5.586199760437012,
+      "learning_rate": 6.169924812030076e-06,
+      "loss": 0.2868,
+      "step": 25470
+    },
+    {
+      "epoch": 38.32,
+      "grad_norm": 4.708113670349121,
+      "learning_rate": 6.168421052631579e-06,
+      "loss": 0.2418,
+      "step": 25480
+    },
+    {
+      "epoch": 38.33,
+      "grad_norm": 8.40892505645752,
+      "learning_rate": 6.166917293233083e-06,
+      "loss": 0.2396,
+      "step": 25490
+    },
+    {
+      "epoch": 38.35,
+      "grad_norm": 6.81309175491333,
+      "learning_rate": 6.165413533834587e-06,
+      "loss": 0.2496,
+      "step": 25500
+    },
+    {
+      "epoch": 38.36,
+      "grad_norm": 8.350881576538086,
+      "learning_rate": 6.163909774436091e-06,
+      "loss": 0.2002,
+      "step": 25510
+    },
+    {
+      "epoch": 38.38,
+      "grad_norm": 3.867704153060913,
+      "learning_rate": 6.162406015037594e-06,
+      "loss": 0.2544,
+      "step": 25520
+    },
+    {
+      "epoch": 38.39,
+      "grad_norm": 2.9632339477539062,
+      "learning_rate": 6.160902255639098e-06,
+      "loss": 0.2444,
+      "step": 25530
+    },
+    {
+      "epoch": 38.41,
+      "grad_norm": 2.540008544921875,
+      "learning_rate": 6.159398496240602e-06,
+      "loss": 0.1792,
+      "step": 25540
+    },
+    {
+      "epoch": 38.42,
+      "grad_norm": 5.204866886138916,
+      "learning_rate": 6.157894736842106e-06,
+      "loss": 0.2785,
+      "step": 25550
+    },
+    {
+      "epoch": 38.44,
+      "grad_norm": 5.661520957946777,
+      "learning_rate": 6.15639097744361e-06,
+      "loss": 0.2656,
+      "step": 25560
+    },
+    {
+      "epoch": 38.45,
+      "grad_norm": 6.089839458465576,
+      "learning_rate": 6.154887218045114e-06,
+      "loss": 0.2698,
+      "step": 25570
+    },
+    {
+      "epoch": 38.47,
+      "grad_norm": 6.78479528427124,
+      "learning_rate": 6.153383458646617e-06,
+      "loss": 0.2754,
+      "step": 25580
+    },
+    {
+      "epoch": 38.48,
+      "grad_norm": 5.978816032409668,
+      "learning_rate": 6.151879699248121e-06,
+      "loss": 0.2906,
+      "step": 25590
+    },
+    {
+      "epoch": 38.5,
+      "grad_norm": 6.391565799713135,
+      "learning_rate": 6.1503759398496245e-06,
+      "loss": 0.2373,
+      "step": 25600
+    },
+    {
+      "epoch": 38.51,
+      "grad_norm": 6.544304847717285,
+      "learning_rate": 6.148872180451129e-06,
+      "loss": 0.256,
+      "step": 25610
+    },
+    {
+      "epoch": 38.53,
+      "grad_norm": 3.7540414333343506,
+      "learning_rate": 6.1473684210526316e-06,
+      "loss": 0.2462,
+      "step": 25620
+    },
+    {
+      "epoch": 38.54,
+      "grad_norm": 4.831808567047119,
+      "learning_rate": 6.145864661654136e-06,
+      "loss": 0.2307,
+      "step": 25630
+    },
+    {
+      "epoch": 38.56,
+      "grad_norm": 3.9876370429992676,
+      "learning_rate": 6.1443609022556395e-06,
+      "loss": 0.182,
+      "step": 25640
+    },
+    {
+      "epoch": 38.57,
+      "grad_norm": 4.689981460571289,
+      "learning_rate": 6.142857142857144e-06,
+      "loss": 0.2577,
+      "step": 25650
+    },
+    {
+      "epoch": 38.59,
+      "grad_norm": 6.713221073150635,
+      "learning_rate": 6.141353383458647e-06,
+      "loss": 0.2902,
+      "step": 25660
+    },
+    {
+      "epoch": 38.6,
+      "grad_norm": 5.03303861618042,
+      "learning_rate": 6.139849624060152e-06,
+      "loss": 0.1913,
+      "step": 25670
+    },
+    {
+      "epoch": 38.62,
+      "grad_norm": 6.612110137939453,
+      "learning_rate": 6.138345864661654e-06,
+      "loss": 0.2724,
+      "step": 25680
+    },
+    {
+      "epoch": 38.63,
+      "grad_norm": 0.28649356961250305,
+      "learning_rate": 6.136842105263159e-06,
+      "loss": 0.2571,
+      "step": 25690
+    },
+    {
+      "epoch": 38.65,
+      "grad_norm": 3.211318016052246,
+      "learning_rate": 6.135338345864662e-06,
+      "loss": 0.2778,
+      "step": 25700
+    },
+    {
+      "epoch": 38.66,
+      "grad_norm": 2.4958066940307617,
+      "learning_rate": 6.133834586466167e-06,
+      "loss": 0.2794,
+      "step": 25710
+    },
+    {
+      "epoch": 38.68,
+      "grad_norm": 7.281109809875488,
+      "learning_rate": 6.132330827067669e-06,
+      "loss": 0.2804,
+      "step": 25720
+    },
+    {
+      "epoch": 38.69,
+      "grad_norm": 5.622795581817627,
+      "learning_rate": 6.130827067669174e-06,
+      "loss": 0.2422,
+      "step": 25730
+    },
+    {
+      "epoch": 38.71,
+      "grad_norm": 6.729783058166504,
+      "learning_rate": 6.129323308270677e-06,
+      "loss": 0.252,
+      "step": 25740
+    },
+    {
+      "epoch": 38.72,
+      "grad_norm": 8.58928108215332,
+      "learning_rate": 6.1278195488721816e-06,
+      "loss": 0.2239,
+      "step": 25750
+    },
+    {
+      "epoch": 38.74,
+      "grad_norm": 4.566402912139893,
+      "learning_rate": 6.126315789473685e-06,
+      "loss": 0.2211,
+      "step": 25760
+    },
+    {
+      "epoch": 38.75,
+      "grad_norm": 13.606522560119629,
+      "learning_rate": 6.124812030075189e-06,
+      "loss": 0.2715,
+      "step": 25770
+    },
+    {
+      "epoch": 38.77,
+      "grad_norm": 7.567572593688965,
+      "learning_rate": 6.123308270676692e-06,
+      "loss": 0.223,
+      "step": 25780
+    },
+    {
+      "epoch": 38.78,
+      "grad_norm": 5.111462116241455,
+      "learning_rate": 6.1218045112781965e-06,
+      "loss": 0.2691,
+      "step": 25790
+    },
+    {
+      "epoch": 38.8,
+      "grad_norm": 3.890907049179077,
+      "learning_rate": 6.1203007518797e-06,
+      "loss": 0.2056,
+      "step": 25800
+    },
+    {
+      "epoch": 38.81,
+      "grad_norm": 8.117431640625,
+      "learning_rate": 6.118796992481203e-06,
+      "loss": 0.2391,
+      "step": 25810
+    },
+    {
+      "epoch": 38.83,
+      "grad_norm": 3.895110845565796,
+      "learning_rate": 6.117293233082707e-06,
+      "loss": 0.2405,
+      "step": 25820
+    },
+    {
+      "epoch": 38.84,
+      "grad_norm": 5.332907676696777,
+      "learning_rate": 6.1157894736842106e-06,
+      "loss": 0.2132,
+      "step": 25830
+    },
+    {
+      "epoch": 38.86,
+      "grad_norm": 5.767151355743408,
+      "learning_rate": 6.114285714285715e-06,
+      "loss": 0.2321,
+      "step": 25840
+    },
+    {
+      "epoch": 38.87,
+      "grad_norm": 6.4265899658203125,
+      "learning_rate": 6.1127819548872184e-06,
+      "loss": 0.265,
+      "step": 25850
+    },
+    {
+      "epoch": 38.89,
+      "grad_norm": 5.724833011627197,
+      "learning_rate": 6.111278195488722e-06,
+      "loss": 0.2308,
+      "step": 25860
+    },
+    {
+      "epoch": 38.9,
+      "grad_norm": 4.019651412963867,
+      "learning_rate": 6.1097744360902255e-06,
+      "loss": 0.2451,
+      "step": 25870
+    },
+    {
+      "epoch": 38.92,
+      "grad_norm": 6.344007968902588,
+      "learning_rate": 6.10827067669173e-06,
+      "loss": 0.2445,
+      "step": 25880
+    },
+    {
+      "epoch": 38.93,
+      "grad_norm": 4.17384672164917,
+      "learning_rate": 6.106766917293233e-06,
+      "loss": 0.2344,
+      "step": 25890
+    },
+    {
+      "epoch": 38.95,
+      "grad_norm": 7.522851467132568,
+      "learning_rate": 6.105263157894738e-06,
+      "loss": 0.2443,
+      "step": 25900
+    },
+    {
+      "epoch": 38.96,
+      "grad_norm": 5.275865077972412,
+      "learning_rate": 6.10375939849624e-06,
+      "loss": 0.2393,
+      "step": 25910
+    },
+    {
+      "epoch": 38.98,
+      "grad_norm": 12.723134994506836,
+      "learning_rate": 6.102255639097745e-06,
+      "loss": 0.2157,
+      "step": 25920
+    },
+    {
+      "epoch": 38.99,
+      "grad_norm": 6.258028507232666,
+      "learning_rate": 6.100751879699248e-06,
+      "loss": 0.3343,
+      "step": 25930
+    },
+    {
+      "epoch": 39.0,
+      "eval_accuracy": 0.9299,
+      "eval_loss": 0.30296048521995544,
+      "eval_runtime": 85.0044,
+      "eval_samples_per_second": 117.641,
+      "eval_steps_per_second": 0.471,
+      "step": 25935
+    },
+    {
+      "epoch": 39.01,
+      "grad_norm": 7.032735347747803,
+      "learning_rate": 6.099248120300753e-06,
+      "loss": 0.2688,
+      "step": 25940
+    },
+    {
+      "epoch": 39.02,
+      "grad_norm": 4.064879417419434,
+      "learning_rate": 6.097744360902256e-06,
+      "loss": 0.2388,
+      "step": 25950
+    },
+    {
+      "epoch": 39.04,
+      "grad_norm": 10.355484008789062,
+      "learning_rate": 6.09624060150376e-06,
+      "loss": 0.236,
+      "step": 25960
+    },
+    {
+      "epoch": 39.05,
+      "grad_norm": 5.589707374572754,
+      "learning_rate": 6.094736842105263e-06,
+      "loss": 0.284,
+      "step": 25970
+    },
+    {
+      "epoch": 39.07,
+      "grad_norm": 5.585343360900879,
+      "learning_rate": 6.0932330827067676e-06,
+      "loss": 0.2678,
+      "step": 25980
+    },
+    {
+      "epoch": 39.08,
+      "grad_norm": 5.8805670738220215,
+      "learning_rate": 6.091729323308271e-06,
+      "loss": 0.2468,
+      "step": 25990
+    },
+    {
+      "epoch": 39.1,
+      "grad_norm": 3.8114516735076904,
+      "learning_rate": 6.0902255639097755e-06,
+      "loss": 0.2651,
+      "step": 26000
+    },
+    {
+      "epoch": 39.11,
+      "grad_norm": 2.4582362174987793,
+      "learning_rate": 6.088721804511278e-06,
+      "loss": 0.2564,
+      "step": 26010
+    },
+    {
+      "epoch": 39.13,
+      "grad_norm": 7.29220724105835,
+      "learning_rate": 6.0872180451127825e-06,
+      "loss": 0.2869,
+      "step": 26020
+    },
+    {
+      "epoch": 39.14,
+      "grad_norm": 5.818304061889648,
+      "learning_rate": 6.085714285714286e-06,
+      "loss": 0.1954,
+      "step": 26030
+    },
+    {
+      "epoch": 39.16,
+      "grad_norm": 3.6301677227020264,
+      "learning_rate": 6.08421052631579e-06,
+      "loss": 0.2483,
+      "step": 26040
+    },
+    {
+      "epoch": 39.17,
+      "grad_norm": 8.798738479614258,
+      "learning_rate": 6.082706766917293e-06,
+      "loss": 0.2372,
+      "step": 26050
+    },
+    {
+      "epoch": 39.19,
+      "grad_norm": 4.8981404304504395,
+      "learning_rate": 6.081203007518797e-06,
+      "loss": 0.1533,
+      "step": 26060
+    },
+    {
+      "epoch": 39.2,
+      "grad_norm": 3.82854962348938,
+      "learning_rate": 6.079699248120301e-06,
+      "loss": 0.1997,
+      "step": 26070
+    },
+    {
+      "epoch": 39.22,
+      "grad_norm": 5.643593788146973,
+      "learning_rate": 6.078195488721805e-06,
+      "loss": 0.195,
+      "step": 26080
+    },
+    {
+      "epoch": 39.23,
+      "grad_norm": 5.393759727478027,
+      "learning_rate": 6.076691729323309e-06,
+      "loss": 0.2615,
+      "step": 26090
+    },
+    {
+      "epoch": 39.25,
+      "grad_norm": 4.524438381195068,
+      "learning_rate": 6.075187969924813e-06,
+      "loss": 0.2547,
+      "step": 26100
+    },
+    {
+      "epoch": 39.26,
+      "grad_norm": 4.402938365936279,
+      "learning_rate": 6.073684210526316e-06,
+      "loss": 0.3235,
+      "step": 26110
+    },
+    {
+      "epoch": 39.28,
+      "grad_norm": 4.850139141082764,
+      "learning_rate": 6.07218045112782e-06,
+      "loss": 0.3356,
+      "step": 26120
+    },
+    {
+      "epoch": 39.29,
+      "grad_norm": 6.940186023712158,
+      "learning_rate": 6.070676691729324e-06,
+      "loss": 0.2929,
+      "step": 26130
+    },
+    {
+      "epoch": 39.31,
+      "grad_norm": 5.906674861907959,
+      "learning_rate": 6.069172932330828e-06,
+      "loss": 0.2722,
+      "step": 26140
+    },
+    {
+      "epoch": 39.32,
+      "grad_norm": 6.132667541503906,
+      "learning_rate": 6.067669172932331e-06,
+      "loss": 0.2593,
+      "step": 26150
+    },
+    {
+      "epoch": 39.34,
+      "grad_norm": 4.581514835357666,
+      "learning_rate": 6.066165413533835e-06,
+      "loss": 0.2535,
+      "step": 26160
+    },
+    {
+      "epoch": 39.35,
+      "grad_norm": 7.306211471557617,
+      "learning_rate": 6.064661654135339e-06,
+      "loss": 0.2108,
+      "step": 26170
+    },
+    {
+      "epoch": 39.37,
+      "grad_norm": 5.0160040855407715,
+      "learning_rate": 6.063157894736843e-06,
+      "loss": 0.2747,
+      "step": 26180
+    },
+    {
+      "epoch": 39.38,
+      "grad_norm": 6.838048458099365,
+      "learning_rate": 6.0616541353383466e-06,
+      "loss": 0.2339,
+      "step": 26190
+    },
+    {
+      "epoch": 39.4,
+      "grad_norm": 7.156051158905029,
+      "learning_rate": 6.060150375939851e-06,
+      "loss": 0.2565,
+      "step": 26200
+    },
+    {
+      "epoch": 39.41,
+      "grad_norm": 3.844694137573242,
+      "learning_rate": 6.058646616541354e-06,
+      "loss": 0.2644,
+      "step": 26210
+    },
+    {
+      "epoch": 39.43,
+      "grad_norm": 9.594381332397461,
+      "learning_rate": 6.057142857142858e-06,
+      "loss": 0.2381,
+      "step": 26220
+    },
+    {
+      "epoch": 39.44,
+      "grad_norm": 6.174018383026123,
+      "learning_rate": 6.0556390977443615e-06,
+      "loss": 0.2974,
+      "step": 26230
+    },
+    {
+      "epoch": 39.46,
+      "grad_norm": 4.462780952453613,
+      "learning_rate": 6.054135338345866e-06,
+      "loss": 0.3042,
+      "step": 26240
+    },
+    {
+      "epoch": 39.47,
+      "grad_norm": 5.761167049407959,
+      "learning_rate": 6.0526315789473685e-06,
+      "loss": 0.2494,
+      "step": 26250
+    },
+    {
+      "epoch": 39.49,
+      "grad_norm": 6.0231733322143555,
+      "learning_rate": 6.051127819548873e-06,
+      "loss": 0.2232,
+      "step": 26260
+    },
+    {
+      "epoch": 39.5,
+      "grad_norm": 6.92008113861084,
+      "learning_rate": 6.049624060150376e-06,
+      "loss": 0.2763,
+      "step": 26270
+    },
+    {
+      "epoch": 39.52,
+      "grad_norm": 8.816802978515625,
+      "learning_rate": 6.048120300751881e-06,
+      "loss": 0.2386,
+      "step": 26280
+    },
+    {
+      "epoch": 39.53,
+      "grad_norm": 9.278828620910645,
+      "learning_rate": 6.046616541353384e-06,
+      "loss": 0.305,
+      "step": 26290
+    },
+    {
+      "epoch": 39.55,
+      "grad_norm": 6.65963888168335,
+      "learning_rate": 6.045112781954889e-06,
+      "loss": 0.2616,
+      "step": 26300
+    },
+    {
+      "epoch": 39.56,
+      "grad_norm": 6.531272888183594,
+      "learning_rate": 6.043609022556391e-06,
+      "loss": 0.2339,
+      "step": 26310
+    },
+    {
+      "epoch": 39.58,
+      "grad_norm": 5.144484043121338,
+      "learning_rate": 6.042105263157895e-06,
+      "loss": 0.3106,
+      "step": 26320
+    },
+    {
+      "epoch": 39.59,
+      "grad_norm": 5.5032243728637695,
+      "learning_rate": 6.040601503759399e-06,
+      "loss": 0.2968,
+      "step": 26330
+    },
+    {
+      "epoch": 39.61,
+      "grad_norm": 3.399604320526123,
+      "learning_rate": 6.039097744360902e-06,
+      "loss": 0.2153,
+      "step": 26340
+    },
+    {
+      "epoch": 39.62,
+      "grad_norm": 6.184195041656494,
+      "learning_rate": 6.037593984962406e-06,
+      "loss": 0.306,
+      "step": 26350
+    },
+    {
+      "epoch": 39.64,
+      "grad_norm": 5.967001914978027,
+      "learning_rate": 6.03609022556391e-06,
+      "loss": 0.3201,
+      "step": 26360
+    },
+    {
+      "epoch": 39.65,
+      "grad_norm": 3.1237454414367676,
+      "learning_rate": 6.034586466165414e-06,
+      "loss": 0.2411,
+      "step": 26370
+    },
+    {
+      "epoch": 39.67,
+      "grad_norm": 5.807510852813721,
+      "learning_rate": 6.033082706766918e-06,
+      "loss": 0.2651,
+      "step": 26380
+    },
+    {
+      "epoch": 39.68,
+      "grad_norm": 3.013021230697632,
+      "learning_rate": 6.031578947368422e-06,
+      "loss": 0.2835,
+      "step": 26390
+    },
+    {
+      "epoch": 39.7,
+      "grad_norm": 6.245032787322998,
+      "learning_rate": 6.030075187969925e-06,
+      "loss": 0.3153,
+      "step": 26400
+    },
+    {
+      "epoch": 39.71,
+      "grad_norm": 2.45760440826416,
+      "learning_rate": 6.028571428571429e-06,
+      "loss": 0.2501,
+      "step": 26410
+    },
+    {
+      "epoch": 39.73,
+      "grad_norm": 6.795231342315674,
+      "learning_rate": 6.0270676691729326e-06,
+      "loss": 0.2885,
+      "step": 26420
+    },
+    {
+      "epoch": 39.74,
+      "grad_norm": 4.70608377456665,
+      "learning_rate": 6.025563909774437e-06,
+      "loss": 0.2741,
+      "step": 26430
+    },
+    {
+      "epoch": 39.76,
+      "grad_norm": 3.3417322635650635,
+      "learning_rate": 6.02406015037594e-06,
+      "loss": 0.31,
+      "step": 26440
+    },
+    {
+      "epoch": 39.77,
+      "grad_norm": 4.526313781738281,
+      "learning_rate": 6.022556390977444e-06,
+      "loss": 0.3074,
+      "step": 26450
+    },
+    {
+      "epoch": 39.79,
+      "grad_norm": 4.290112495422363,
+      "learning_rate": 6.0210526315789475e-06,
+      "loss": 0.2311,
+      "step": 26460
+    },
+    {
+      "epoch": 39.8,
+      "grad_norm": 7.921751499176025,
+      "learning_rate": 6.019548872180452e-06,
+      "loss": 0.2454,
+      "step": 26470
+    },
+    {
+      "epoch": 39.82,
+      "grad_norm": 5.810790538787842,
+      "learning_rate": 6.018045112781955e-06,
+      "loss": 0.2412,
+      "step": 26480
+    },
+    {
+      "epoch": 39.83,
+      "grad_norm": 7.158168792724609,
+      "learning_rate": 6.01654135338346e-06,
+      "loss": 0.2827,
+      "step": 26490
+    },
+    {
+      "epoch": 39.85,
+      "grad_norm": 5.13280725479126,
+      "learning_rate": 6.015037593984962e-06,
+      "loss": 0.2417,
+      "step": 26500
+    },
+    {
+      "epoch": 39.86,
+      "grad_norm": 5.112004280090332,
+      "learning_rate": 6.013533834586467e-06,
+      "loss": 0.2607,
+      "step": 26510
+    },
+    {
+      "epoch": 39.88,
+      "grad_norm": 3.5853915214538574,
+      "learning_rate": 6.01203007518797e-06,
+      "loss": 0.2105,
+      "step": 26520
+    },
+    {
+      "epoch": 39.89,
+      "grad_norm": 7.726737976074219,
+      "learning_rate": 6.010526315789475e-06,
+      "loss": 0.2266,
+      "step": 26530
+    },
+    {
+      "epoch": 39.91,
+      "grad_norm": 4.511539936065674,
+      "learning_rate": 6.009022556390977e-06,
+      "loss": 0.2388,
+      "step": 26540
+    },
+    {
+      "epoch": 39.92,
+      "grad_norm": 6.07355260848999,
+      "learning_rate": 6.007518796992482e-06,
+      "loss": 0.2129,
+      "step": 26550
+    },
+    {
+      "epoch": 39.94,
+      "grad_norm": 4.063088417053223,
+      "learning_rate": 6.006015037593985e-06,
+      "loss": 0.2759,
+      "step": 26560
+    },
+    {
+      "epoch": 39.95,
+      "grad_norm": 8.524380683898926,
+      "learning_rate": 6.00451127819549e-06,
+      "loss": 0.2703,
+      "step": 26570
+    },
+    {
+      "epoch": 39.97,
+      "grad_norm": 5.095054626464844,
+      "learning_rate": 6.003007518796993e-06,
+      "loss": 0.2418,
+      "step": 26580
+    },
+    {
+      "epoch": 39.98,
+      "grad_norm": 4.896398067474365,
+      "learning_rate": 6.001503759398497e-06,
+      "loss": 0.294,
+      "step": 26590
+    },
+    {
+      "epoch": 40.0,
+      "grad_norm": 41.88473892211914,
+      "learning_rate": 6e-06,
+      "loss": 0.2252,
+      "step": 26600
+    },
+    {
+      "epoch": 40.0,
+      "eval_accuracy": 0.9313,
+      "eval_loss": 0.2959575057029724,
+      "eval_runtime": 84.9779,
+      "eval_samples_per_second": 117.678,
+      "eval_steps_per_second": 0.471,
+      "step": 26600
+    },
+    {
+      "epoch": 40.02,
+      "grad_norm": 11.272649765014648,
+      "learning_rate": 5.9984962406015045e-06,
+      "loss": 0.2401,
+      "step": 26610
+    },
+    {
+      "epoch": 40.03,
+      "grad_norm": 6.342247486114502,
+      "learning_rate": 5.996992481203008e-06,
+      "loss": 0.1853,
+      "step": 26620
+    },
+    {
+      "epoch": 40.05,
+      "grad_norm": 8.127535820007324,
+      "learning_rate": 5.995488721804512e-06,
+      "loss": 0.2091,
+      "step": 26630
+    },
+    {
+      "epoch": 40.06,
+      "grad_norm": 7.700329303741455,
+      "learning_rate": 5.993984962406015e-06,
+      "loss": 0.2537,
+      "step": 26640
+    },
+    {
+      "epoch": 40.08,
+      "grad_norm": 4.249181747436523,
+      "learning_rate": 5.9924812030075194e-06,
+      "loss": 0.2576,
+      "step": 26650
+    },
+    {
+      "epoch": 40.09,
+      "grad_norm": 6.067375659942627,
+      "learning_rate": 5.990977443609023e-06,
+      "loss": 0.2815,
+      "step": 26660
+    },
+    {
+      "epoch": 40.11,
+      "grad_norm": 5.851858139038086,
+      "learning_rate": 5.989473684210527e-06,
+      "loss": 0.2741,
+      "step": 26670
+    },
+    {
+      "epoch": 40.12,
+      "grad_norm": 2.2680578231811523,
+      "learning_rate": 5.987969924812031e-06,
+      "loss": 0.2571,
+      "step": 26680
+    },
+    {
+      "epoch": 40.14,
+      "grad_norm": 6.329833984375,
+      "learning_rate": 5.986466165413534e-06,
+      "loss": 0.1918,
+      "step": 26690
+    },
+    {
+      "epoch": 40.15,
+      "grad_norm": 3.3135337829589844,
+      "learning_rate": 5.984962406015038e-06,
+      "loss": 0.2999,
+      "step": 26700
+    },
+    {
+      "epoch": 40.17,
+      "grad_norm": 5.360442161560059,
+      "learning_rate": 5.983458646616542e-06,
+      "loss": 0.248,
+      "step": 26710
+    },
+    {
+      "epoch": 40.18,
+      "grad_norm": 5.619331359863281,
+      "learning_rate": 5.981954887218046e-06,
+      "loss": 0.2665,
+      "step": 26720
+    },
+    {
+      "epoch": 40.2,
+      "grad_norm": 6.200700283050537,
+      "learning_rate": 5.98045112781955e-06,
+      "loss": 0.2415,
+      "step": 26730
+    },
+    {
+      "epoch": 40.21,
+      "grad_norm": 6.159794330596924,
+      "learning_rate": 5.978947368421053e-06,
+      "loss": 0.3265,
+      "step": 26740
+    },
+    {
+      "epoch": 40.23,
+      "grad_norm": 4.464012622833252,
+      "learning_rate": 5.977443609022557e-06,
+      "loss": 0.2426,
+      "step": 26750
+    },
+    {
+      "epoch": 40.24,
+      "grad_norm": 6.896475791931152,
+      "learning_rate": 5.975939849624061e-06,
+      "loss": 0.2842,
+      "step": 26760
+    },
+    {
+      "epoch": 40.26,
+      "grad_norm": 5.030261993408203,
+      "learning_rate": 5.974436090225565e-06,
+      "loss": 0.2208,
+      "step": 26770
+    },
+    {
+      "epoch": 40.27,
+      "grad_norm": 7.4109392166137695,
+      "learning_rate": 5.972932330827068e-06,
+      "loss": 0.3284,
+      "step": 26780
+    },
+    {
+      "epoch": 40.29,
+      "grad_norm": 3.889845132827759,
+      "learning_rate": 5.971428571428572e-06,
+      "loss": 0.2688,
+      "step": 26790
+    },
+    {
+      "epoch": 40.3,
+      "grad_norm": 2.1321589946746826,
+      "learning_rate": 5.969924812030076e-06,
+      "loss": 0.1939,
+      "step": 26800
+    },
+    {
+      "epoch": 40.32,
+      "grad_norm": 10.229703903198242,
+      "learning_rate": 5.968421052631579e-06,
+      "loss": 0.2851,
+      "step": 26810
+    },
+    {
+      "epoch": 40.33,
+      "grad_norm": 2.8729848861694336,
+      "learning_rate": 5.9669172932330835e-06,
+      "loss": 0.2451,
+      "step": 26820
+    },
+    {
+      "epoch": 40.35,
+      "grad_norm": 5.045018672943115,
+      "learning_rate": 5.965413533834586e-06,
+      "loss": 0.2655,
+      "step": 26830
+    },
+    {
+      "epoch": 40.36,
+      "grad_norm": 6.061784744262695,
+      "learning_rate": 5.9639097744360905e-06,
+      "loss": 0.2106,
+      "step": 26840
+    },
+    {
+      "epoch": 40.38,
+      "grad_norm": 4.3670654296875,
+      "learning_rate": 5.962406015037594e-06,
+      "loss": 0.2508,
+      "step": 26850
+    },
+    {
+      "epoch": 40.39,
+      "grad_norm": 4.169200897216797,
+      "learning_rate": 5.9609022556390984e-06,
+      "loss": 0.2374,
+      "step": 26860
+    },
+    {
+      "epoch": 40.41,
+      "grad_norm": 9.870522499084473,
+      "learning_rate": 5.959398496240601e-06,
+      "loss": 0.2112,
+      "step": 26870
+    },
+    {
+      "epoch": 40.42,
+      "grad_norm": 4.387085914611816,
+      "learning_rate": 5.9578947368421055e-06,
+      "loss": 0.2245,
+      "step": 26880
+    },
+    {
+      "epoch": 40.44,
+      "grad_norm": 4.728475093841553,
+      "learning_rate": 5.956390977443609e-06,
+      "loss": 0.2252,
+      "step": 26890
+    },
+    {
+      "epoch": 40.45,
+      "grad_norm": 3.2639694213867188,
+      "learning_rate": 5.954887218045113e-06,
+      "loss": 0.2637,
+      "step": 26900
+    },
+    {
+      "epoch": 40.47,
+      "grad_norm": 3.462743043899536,
+      "learning_rate": 5.953383458646617e-06,
+      "loss": 0.2298,
+      "step": 26910
+    },
+    {
+      "epoch": 40.48,
+      "grad_norm": 7.873288631439209,
+      "learning_rate": 5.951879699248121e-06,
+      "loss": 0.2781,
+      "step": 26920
+    },
+    {
+      "epoch": 40.5,
+      "grad_norm": 5.6623334884643555,
+      "learning_rate": 5.950375939849624e-06,
+      "loss": 0.3083,
+      "step": 26930
+    },
+    {
+      "epoch": 40.51,
+      "grad_norm": 4.990388870239258,
+      "learning_rate": 5.948872180451128e-06,
+      "loss": 0.2423,
+      "step": 26940
+    },
+    {
+      "epoch": 40.53,
+      "grad_norm": 4.813294887542725,
+      "learning_rate": 5.947368421052632e-06,
+      "loss": 0.3243,
+      "step": 26950
+    },
+    {
+      "epoch": 40.54,
+      "grad_norm": 4.974578857421875,
+      "learning_rate": 5.945864661654136e-06,
+      "loss": 0.2494,
+      "step": 26960
+    },
+    {
+      "epoch": 40.56,
+      "grad_norm": 5.326529502868652,
+      "learning_rate": 5.944360902255639e-06,
+      "loss": 0.3272,
+      "step": 26970
+    },
+    {
+      "epoch": 40.57,
+      "grad_norm": 6.8446245193481445,
+      "learning_rate": 5.942857142857143e-06,
+      "loss": 0.3175,
+      "step": 26980
+    },
+    {
+      "epoch": 40.59,
+      "grad_norm": 7.015409469604492,
+      "learning_rate": 5.941353383458647e-06,
+      "loss": 0.2259,
+      "step": 26990
+    },
+    {
+      "epoch": 40.6,
+      "grad_norm": 5.787068843841553,
+      "learning_rate": 5.939849624060151e-06,
+      "loss": 0.2027,
+      "step": 27000
+    },
+    {
+      "epoch": 40.62,
+      "grad_norm": 8.761579513549805,
+      "learning_rate": 5.938345864661655e-06,
+      "loss": 0.229,
+      "step": 27010
+    },
+    {
+      "epoch": 40.63,
+      "grad_norm": 6.765686988830566,
+      "learning_rate": 5.936842105263159e-06,
+      "loss": 0.2945,
+      "step": 27020
+    },
+    {
+      "epoch": 40.65,
+      "grad_norm": 8.93405532836914,
+      "learning_rate": 5.935338345864662e-06,
+      "loss": 0.3073,
+      "step": 27030
+    },
+    {
+      "epoch": 40.66,
+      "grad_norm": 6.20574951171875,
+      "learning_rate": 5.933834586466166e-06,
+      "loss": 0.2785,
+      "step": 27040
+    },
+    {
+      "epoch": 40.68,
+      "grad_norm": 7.853849411010742,
+      "learning_rate": 5.9323308270676695e-06,
+      "loss": 0.2294,
+      "step": 27050
+    },
+    {
+      "epoch": 40.69,
+      "grad_norm": 4.646696090698242,
+      "learning_rate": 5.930827067669174e-06,
+      "loss": 0.2995,
+      "step": 27060
+    },
+    {
+      "epoch": 40.71,
+      "grad_norm": 7.266605377197266,
+      "learning_rate": 5.9293233082706766e-06,
+      "loss": 0.259,
+      "step": 27070
+    },
+    {
+      "epoch": 40.72,
+      "grad_norm": 6.365235805511475,
+      "learning_rate": 5.927819548872181e-06,
+      "loss": 0.2905,
+      "step": 27080
+    },
+    {
+      "epoch": 40.74,
+      "grad_norm": 4.6850457191467285,
+      "learning_rate": 5.9263157894736844e-06,
+      "loss": 0.2313,
+      "step": 27090
+    },
+    {
+      "epoch": 40.75,
+      "grad_norm": 2.9510128498077393,
+      "learning_rate": 5.924812030075189e-06,
+      "loss": 0.3019,
+      "step": 27100
+    },
+    {
+      "epoch": 40.77,
+      "grad_norm": 6.478631973266602,
+      "learning_rate": 5.923308270676692e-06,
+      "loss": 0.2318,
+      "step": 27110
+    },
+    {
+      "epoch": 40.78,
+      "grad_norm": 4.466511249542236,
+      "learning_rate": 5.921804511278197e-06,
+      "loss": 0.2816,
+      "step": 27120
+    },
+    {
+      "epoch": 40.8,
+      "grad_norm": 3.0053863525390625,
+      "learning_rate": 5.920300751879699e-06,
+      "loss": 0.2526,
+      "step": 27130
+    },
+    {
+      "epoch": 40.81,
+      "grad_norm": 4.911371231079102,
+      "learning_rate": 5.918796992481204e-06,
+      "loss": 0.2546,
+      "step": 27140
+    },
+    {
+      "epoch": 40.83,
+      "grad_norm": 7.452986240386963,
+      "learning_rate": 5.917293233082707e-06,
+      "loss": 0.2378,
+      "step": 27150
+    },
+    {
+      "epoch": 40.84,
+      "grad_norm": 6.792994499206543,
+      "learning_rate": 5.915789473684212e-06,
+      "loss": 0.3151,
+      "step": 27160
+    },
+    {
+      "epoch": 40.86,
+      "grad_norm": 5.100854396820068,
+      "learning_rate": 5.914285714285714e-06,
+      "loss": 0.253,
+      "step": 27170
+    },
+    {
+      "epoch": 40.87,
+      "grad_norm": 3.16489577293396,
+      "learning_rate": 5.912781954887219e-06,
+      "loss": 0.2431,
+      "step": 27180
+    },
+    {
+      "epoch": 40.89,
+      "grad_norm": 9.801965713500977,
+      "learning_rate": 5.911278195488722e-06,
+      "loss": 0.2013,
+      "step": 27190
+    },
+    {
+      "epoch": 40.9,
+      "grad_norm": 3.110069990158081,
+      "learning_rate": 5.9097744360902265e-06,
+      "loss": 0.2555,
+      "step": 27200
+    },
+    {
+      "epoch": 40.92,
+      "grad_norm": 4.8249006271362305,
+      "learning_rate": 5.90827067669173e-06,
+      "loss": 0.1823,
+      "step": 27210
+    },
+    {
+      "epoch": 40.93,
+      "grad_norm": 2.37221622467041,
+      "learning_rate": 5.9067669172932344e-06,
+      "loss": 0.2177,
+      "step": 27220
+    },
+    {
+      "epoch": 40.95,
+      "grad_norm": 4.915449142456055,
+      "learning_rate": 5.905263157894737e-06,
+      "loss": 0.2638,
+      "step": 27230
+    },
+    {
+      "epoch": 40.96,
+      "grad_norm": 4.170838832855225,
+      "learning_rate": 5.9037593984962415e-06,
+      "loss": 0.2424,
+      "step": 27240
+    },
+    {
+      "epoch": 40.98,
+      "grad_norm": 6.760472297668457,
+      "learning_rate": 5.902255639097745e-06,
+      "loss": 0.237,
+      "step": 27250
+    },
+    {
+      "epoch": 40.99,
+      "grad_norm": 4.23274564743042,
+      "learning_rate": 5.900751879699249e-06,
+      "loss": 0.2453,
+      "step": 27260
+    },
+    {
+      "epoch": 41.0,
+      "eval_accuracy": 0.9302,
+      "eval_loss": 0.2976926863193512,
+      "eval_runtime": 84.6737,
+      "eval_samples_per_second": 118.1,
+      "eval_steps_per_second": 0.472,
+      "step": 27265
+    },
+    {
+      "epoch": 41.01,
+      "grad_norm": 7.621011734008789,
+      "learning_rate": 5.899248120300752e-06,
+      "loss": 0.2292,
+      "step": 27270
+    },
+    {
+      "epoch": 41.02,
+      "grad_norm": 3.8886141777038574,
+      "learning_rate": 5.897744360902256e-06,
+      "loss": 0.1851,
+      "step": 27280
+    },
+    {
+      "epoch": 41.04,
+      "grad_norm": 5.533069133758545,
+      "learning_rate": 5.89624060150376e-06,
+      "loss": 0.2613,
+      "step": 27290
+    },
+    {
+      "epoch": 41.05,
+      "grad_norm": 4.005669116973877,
+      "learning_rate": 5.8947368421052634e-06,
+      "loss": 0.1644,
+      "step": 27300
+    },
+    {
+      "epoch": 41.07,
+      "grad_norm": 3.7740938663482666,
+      "learning_rate": 5.893233082706768e-06,
+      "loss": 0.1893,
+      "step": 27310
+    },
+    {
+      "epoch": 41.08,
+      "grad_norm": 8.077632904052734,
+      "learning_rate": 5.8917293233082705e-06,
+      "loss": 0.2166,
+      "step": 27320
+    },
+    {
+      "epoch": 41.1,
+      "grad_norm": 2.8428125381469727,
+      "learning_rate": 5.890225563909775e-06,
+      "loss": 0.2274,
+      "step": 27330
+    },
+    {
+      "epoch": 41.11,
+      "grad_norm": 2.748422861099243,
+      "learning_rate": 5.888721804511278e-06,
+      "loss": 0.1786,
+      "step": 27340
+    },
+    {
+      "epoch": 41.13,
+      "grad_norm": 6.080495834350586,
+      "learning_rate": 5.887218045112783e-06,
+      "loss": 0.3044,
+      "step": 27350
+    },
+    {
+      "epoch": 41.14,
+      "grad_norm": 6.971614837646484,
+      "learning_rate": 5.885714285714285e-06,
+      "loss": 0.2141,
+      "step": 27360
+    },
+    {
+      "epoch": 41.16,
+      "grad_norm": 3.363379716873169,
+      "learning_rate": 5.88421052631579e-06,
+      "loss": 0.3337,
+      "step": 27370
+    },
+    {
+      "epoch": 41.17,
+      "grad_norm": 4.116064548492432,
+      "learning_rate": 5.882706766917293e-06,
+      "loss": 0.2026,
+      "step": 27380
+    },
+    {
+      "epoch": 41.19,
+      "grad_norm": 2.9727118015289307,
+      "learning_rate": 5.881203007518798e-06,
+      "loss": 0.2342,
+      "step": 27390
+    },
+    {
+      "epoch": 41.2,
+      "grad_norm": 4.2843337059021,
+      "learning_rate": 5.879699248120301e-06,
+      "loss": 0.2033,
+      "step": 27400
+    },
+    {
+      "epoch": 41.22,
+      "grad_norm": 11.28203010559082,
+      "learning_rate": 5.8781954887218055e-06,
+      "loss": 0.2623,
+      "step": 27410
+    },
+    {
+      "epoch": 41.23,
+      "grad_norm": 1.7853152751922607,
+      "learning_rate": 5.876691729323308e-06,
+      "loss": 0.1762,
+      "step": 27420
+    },
+    {
+      "epoch": 41.25,
+      "grad_norm": 2.5484893321990967,
+      "learning_rate": 5.8751879699248126e-06,
+      "loss": 0.1423,
+      "step": 27430
+    },
+    {
+      "epoch": 41.26,
+      "grad_norm": 5.5011186599731445,
+      "learning_rate": 5.873684210526316e-06,
+      "loss": 0.2804,
+      "step": 27440
+    },
+    {
+      "epoch": 41.28,
+      "grad_norm": 4.6855645179748535,
+      "learning_rate": 5.8721804511278204e-06,
+      "loss": 0.2465,
+      "step": 27450
+    },
+    {
+      "epoch": 41.29,
+      "grad_norm": 7.020749092102051,
+      "learning_rate": 5.870676691729323e-06,
+      "loss": 0.1858,
+      "step": 27460
+    },
+    {
+      "epoch": 41.31,
+      "grad_norm": 3.362569808959961,
+      "learning_rate": 5.8691729323308275e-06,
+      "loss": 0.2039,
+      "step": 27470
+    },
+    {
+      "epoch": 41.32,
+      "grad_norm": 7.235138893127441,
+      "learning_rate": 5.867669172932331e-06,
+      "loss": 0.2206,
+      "step": 27480
+    },
+    {
+      "epoch": 41.34,
+      "grad_norm": 8.273979187011719,
+      "learning_rate": 5.866165413533835e-06,
+      "loss": 0.2561,
+      "step": 27490
+    },
+    {
+      "epoch": 41.35,
+      "grad_norm": 9.396625518798828,
+      "learning_rate": 5.864661654135339e-06,
+      "loss": 0.2517,
+      "step": 27500
+    },
+    {
+      "epoch": 41.37,
+      "grad_norm": 6.119128704071045,
+      "learning_rate": 5.863157894736842e-06,
+      "loss": 0.2438,
+      "step": 27510
+    },
+    {
+      "epoch": 41.38,
+      "grad_norm": 6.962316989898682,
+      "learning_rate": 5.861654135338346e-06,
+      "loss": 0.267,
+      "step": 27520
+    },
+    {
+      "epoch": 41.4,
+      "grad_norm": 2.296393632888794,
+      "learning_rate": 5.86015037593985e-06,
+      "loss": 0.2358,
+      "step": 27530
+    },
+    {
+      "epoch": 41.41,
+      "grad_norm": 6.2318830490112305,
+      "learning_rate": 5.858646616541354e-06,
+      "loss": 0.2647,
+      "step": 27540
+    },
+    {
+      "epoch": 41.43,
+      "grad_norm": 5.737059116363525,
+      "learning_rate": 5.857142857142858e-06,
+      "loss": 0.2347,
+      "step": 27550
+    },
+    {
+      "epoch": 41.44,
+      "grad_norm": 6.697840213775635,
+      "learning_rate": 5.855639097744361e-06,
+      "loss": 0.1956,
+      "step": 27560
+    },
+    {
+      "epoch": 41.46,
+      "grad_norm": 4.799352169036865,
+      "learning_rate": 5.854135338345865e-06,
+      "loss": 0.2545,
+      "step": 27570
+    },
+    {
+      "epoch": 41.47,
+      "grad_norm": 9.115018844604492,
+      "learning_rate": 5.852631578947369e-06,
+      "loss": 0.2221,
+      "step": 27580
+    },
+    {
+      "epoch": 41.49,
+      "grad_norm": 3.4361791610717773,
+      "learning_rate": 5.851127819548873e-06,
+      "loss": 0.2736,
+      "step": 27590
+    },
+    {
+      "epoch": 41.5,
+      "grad_norm": 6.11599588394165,
+      "learning_rate": 5.849624060150377e-06,
+      "loss": 0.2768,
+      "step": 27600
+    },
+    {
+      "epoch": 41.52,
+      "grad_norm": 17.091064453125,
+      "learning_rate": 5.84812030075188e-06,
+      "loss": 0.2645,
+      "step": 27610
+    },
+    {
+      "epoch": 41.53,
+      "grad_norm": 2.6468470096588135,
+      "learning_rate": 5.846616541353384e-06,
+      "loss": 0.2765,
+      "step": 27620
+    },
+    {
+      "epoch": 41.55,
+      "grad_norm": 8.783407211303711,
+      "learning_rate": 5.845112781954888e-06,
+      "loss": 0.2277,
+      "step": 27630
+    },
+    {
+      "epoch": 41.56,
+      "grad_norm": 9.349631309509277,
+      "learning_rate": 5.8436090225563915e-06,
+      "loss": 0.2197,
+      "step": 27640
+    },
+    {
+      "epoch": 41.58,
+      "grad_norm": 5.436890125274658,
+      "learning_rate": 5.842105263157896e-06,
+      "loss": 0.2632,
+      "step": 27650
+    },
+    {
+      "epoch": 41.59,
+      "grad_norm": 2.644437313079834,
+      "learning_rate": 5.840601503759399e-06,
+      "loss": 0.2113,
+      "step": 27660
+    },
+    {
+      "epoch": 41.61,
+      "grad_norm": 7.14797306060791,
+      "learning_rate": 5.839097744360903e-06,
+      "loss": 0.2395,
+      "step": 27670
+    },
+    {
+      "epoch": 41.62,
+      "grad_norm": 5.408485412597656,
+      "learning_rate": 5.8375939849624065e-06,
+      "loss": 0.23,
+      "step": 27680
+    },
+    {
+      "epoch": 41.64,
+      "grad_norm": 6.316678524017334,
+      "learning_rate": 5.836090225563911e-06,
+      "loss": 0.2506,
+      "step": 27690
+    },
+    {
+      "epoch": 41.65,
+      "grad_norm": 4.856528282165527,
+      "learning_rate": 5.8345864661654135e-06,
+      "loss": 0.2565,
+      "step": 27700
+    },
+    {
+      "epoch": 41.67,
+      "grad_norm": 3.731872797012329,
+      "learning_rate": 5.833082706766918e-06,
+      "loss": 0.2733,
+      "step": 27710
+    },
+    {
+      "epoch": 41.68,
+      "grad_norm": 10.921002388000488,
+      "learning_rate": 5.831578947368421e-06,
+      "loss": 0.233,
+      "step": 27720
+    },
+    {
+      "epoch": 41.7,
+      "grad_norm": 8.051673889160156,
+      "learning_rate": 5.830075187969926e-06,
+      "loss": 0.2357,
+      "step": 27730
+    },
+    {
+      "epoch": 41.71,
+      "grad_norm": 5.143133640289307,
+      "learning_rate": 5.828571428571429e-06,
+      "loss": 0.2399,
+      "step": 27740
+    },
+    {
+      "epoch": 41.73,
+      "grad_norm": 6.688822269439697,
+      "learning_rate": 5.827067669172934e-06,
+      "loss": 0.2199,
+      "step": 27750
+    },
+    {
+      "epoch": 41.74,
+      "grad_norm": 5.77666711807251,
+      "learning_rate": 5.825563909774436e-06,
+      "loss": 0.2227,
+      "step": 27760
+    },
+    {
+      "epoch": 41.76,
+      "grad_norm": 5.851940155029297,
+      "learning_rate": 5.824060150375941e-06,
+      "loss": 0.2161,
+      "step": 27770
+    },
+    {
+      "epoch": 41.77,
+      "grad_norm": 3.7167811393737793,
+      "learning_rate": 5.822556390977444e-06,
+      "loss": 0.2276,
+      "step": 27780
+    },
+    {
+      "epoch": 41.79,
+      "grad_norm": 1.5249907970428467,
+      "learning_rate": 5.8210526315789486e-06,
+      "loss": 0.2857,
+      "step": 27790
+    },
+    {
+      "epoch": 41.8,
+      "grad_norm": 3.788996458053589,
+      "learning_rate": 5.819548872180451e-06,
+      "loss": 0.2901,
+      "step": 27800
+    },
+    {
+      "epoch": 41.82,
+      "grad_norm": 3.4979562759399414,
+      "learning_rate": 5.818045112781955e-06,
+      "loss": 0.2643,
+      "step": 27810
+    },
+    {
+      "epoch": 41.83,
+      "grad_norm": 5.074125289916992,
+      "learning_rate": 5.816541353383459e-06,
+      "loss": 0.2909,
+      "step": 27820
+    },
+    {
+      "epoch": 41.85,
+      "grad_norm": 7.354587078094482,
+      "learning_rate": 5.815037593984963e-06,
+      "loss": 0.2756,
+      "step": 27830
+    },
+    {
+      "epoch": 41.86,
+      "grad_norm": 5.332225322723389,
+      "learning_rate": 5.813533834586467e-06,
+      "loss": 0.2369,
+      "step": 27840
+    },
+    {
+      "epoch": 41.88,
+      "grad_norm": 3.5564475059509277,
+      "learning_rate": 5.81203007518797e-06,
+      "loss": 0.3292,
+      "step": 27850
+    },
+    {
+      "epoch": 41.89,
+      "grad_norm": 5.977893829345703,
+      "learning_rate": 5.810526315789474e-06,
+      "loss": 0.2441,
+      "step": 27860
+    },
+    {
+      "epoch": 41.91,
+      "grad_norm": 7.151272773742676,
+      "learning_rate": 5.8090225563909776e-06,
+      "loss": 0.2375,
+      "step": 27870
+    },
+    {
+      "epoch": 41.92,
+      "grad_norm": 3.628330707550049,
+      "learning_rate": 5.807518796992482e-06,
+      "loss": 0.2102,
+      "step": 27880
+    },
+    {
+      "epoch": 41.94,
+      "grad_norm": 6.060537338256836,
+      "learning_rate": 5.806015037593985e-06,
+      "loss": 0.2511,
+      "step": 27890
+    },
+    {
+      "epoch": 41.95,
+      "grad_norm": 6.281295299530029,
+      "learning_rate": 5.804511278195489e-06,
+      "loss": 0.2389,
+      "step": 27900
+    },
+    {
+      "epoch": 41.97,
+      "grad_norm": 6.898097515106201,
+      "learning_rate": 5.8030075187969925e-06,
+      "loss": 0.282,
+      "step": 27910
+    },
+    {
+      "epoch": 41.98,
+      "grad_norm": 4.405340671539307,
+      "learning_rate": 5.801503759398497e-06,
+      "loss": 0.2674,
+      "step": 27920
+    },
+    {
+      "epoch": 42.0,
+      "grad_norm": 4.30500602722168,
+      "learning_rate": 5.8e-06,
+      "loss": 0.2467,
+      "step": 27930
+    },
+    {
+      "epoch": 42.0,
+      "eval_accuracy": 0.9293,
+      "eval_loss": 0.30341148376464844,
+      "eval_runtime": 84.726,
+      "eval_samples_per_second": 118.028,
+      "eval_steps_per_second": 0.472,
+      "step": 27930
+    },
+    {
+      "epoch": 42.02,
+      "grad_norm": 6.07786750793457,
+      "learning_rate": 5.798496240601505e-06,
+      "loss": 0.2712,
+      "step": 27940
+    },
+    {
+      "epoch": 42.03,
+      "grad_norm": 6.053528308868408,
+      "learning_rate": 5.796992481203007e-06,
+      "loss": 0.2482,
+      "step": 27950
+    },
+    {
+      "epoch": 42.05,
+      "grad_norm": 5.752837657928467,
+      "learning_rate": 5.795488721804512e-06,
+      "loss": 0.2579,
+      "step": 27960
+    },
+    {
+      "epoch": 42.06,
+      "grad_norm": 8.296350479125977,
+      "learning_rate": 5.793984962406015e-06,
+      "loss": 0.2481,
+      "step": 27970
+    },
+    {
+      "epoch": 42.08,
+      "grad_norm": 4.709738254547119,
+      "learning_rate": 5.79248120300752e-06,
+      "loss": 0.1771,
+      "step": 27980
+    },
+    {
+      "epoch": 42.09,
+      "grad_norm": 3.9645566940307617,
+      "learning_rate": 5.790977443609022e-06,
+      "loss": 0.218,
+      "step": 27990
+    },
+    {
+      "epoch": 42.11,
+      "grad_norm": 5.714948654174805,
+      "learning_rate": 5.789473684210527e-06,
+      "loss": 0.2734,
+      "step": 28000
+    },
+    {
+      "epoch": 42.12,
+      "grad_norm": 3.882260799407959,
+      "learning_rate": 5.78796992481203e-06,
+      "loss": 0.2747,
+      "step": 28010
+    },
+    {
+      "epoch": 42.14,
+      "grad_norm": 4.841667175292969,
+      "learning_rate": 5.786466165413535e-06,
+      "loss": 0.2728,
+      "step": 28020
+    },
+    {
+      "epoch": 42.15,
+      "grad_norm": 7.8732523918151855,
+      "learning_rate": 5.784962406015038e-06,
+      "loss": 0.2417,
+      "step": 28030
+    },
+    {
+      "epoch": 42.17,
+      "grad_norm": 6.003359317779541,
+      "learning_rate": 5.7834586466165425e-06,
+      "loss": 0.2442,
+      "step": 28040
+    },
+    {
+      "epoch": 42.18,
+      "grad_norm": 4.850991249084473,
+      "learning_rate": 5.781954887218045e-06,
+      "loss": 0.2731,
+      "step": 28050
+    },
+    {
+      "epoch": 42.2,
+      "grad_norm": 5.1097025871276855,
+      "learning_rate": 5.7804511278195495e-06,
+      "loss": 0.1836,
+      "step": 28060
+    },
+    {
+      "epoch": 42.21,
+      "grad_norm": 6.158971309661865,
+      "learning_rate": 5.778947368421053e-06,
+      "loss": 0.2186,
+      "step": 28070
+    },
+    {
+      "epoch": 42.23,
+      "grad_norm": 4.05385160446167,
+      "learning_rate": 5.777443609022557e-06,
+      "loss": 0.2081,
+      "step": 28080
+    },
+    {
+      "epoch": 42.24,
+      "grad_norm": 6.728810787200928,
+      "learning_rate": 5.77593984962406e-06,
+      "loss": 0.2637,
+      "step": 28090
+    },
+    {
+      "epoch": 42.26,
+      "grad_norm": 4.585230350494385,
+      "learning_rate": 5.7744360902255644e-06,
+      "loss": 0.2556,
+      "step": 28100
+    },
+    {
+      "epoch": 42.27,
+      "grad_norm": 4.899889945983887,
+      "learning_rate": 5.772932330827068e-06,
+      "loss": 0.2692,
+      "step": 28110
+    },
+    {
+      "epoch": 42.29,
+      "grad_norm": 8.909112930297852,
+      "learning_rate": 5.771428571428572e-06,
+      "loss": 0.1978,
+      "step": 28120
+    },
+    {
+      "epoch": 42.3,
+      "grad_norm": 5.728882312774658,
+      "learning_rate": 5.769924812030076e-06,
+      "loss": 0.2635,
+      "step": 28130
+    },
+    {
+      "epoch": 42.32,
+      "grad_norm": 4.432710647583008,
+      "learning_rate": 5.76842105263158e-06,
+      "loss": 0.2616,
+      "step": 28140
+    },
+    {
+      "epoch": 42.33,
+      "grad_norm": 3.3793838024139404,
+      "learning_rate": 5.766917293233083e-06,
+      "loss": 0.2387,
+      "step": 28150
+    },
+    {
+      "epoch": 42.35,
+      "grad_norm": 10.177301406860352,
+      "learning_rate": 5.765413533834587e-06,
+      "loss": 0.2494,
+      "step": 28160
+    },
+    {
+      "epoch": 42.36,
+      "grad_norm": 5.089664459228516,
+      "learning_rate": 5.763909774436091e-06,
+      "loss": 0.2588,
+      "step": 28170
+    },
+    {
+      "epoch": 42.38,
+      "grad_norm": 3.203890800476074,
+      "learning_rate": 5.762406015037595e-06,
+      "loss": 0.3037,
+      "step": 28180
+    },
+    {
+      "epoch": 42.39,
+      "grad_norm": 5.658225059509277,
+      "learning_rate": 5.760902255639098e-06,
+      "loss": 0.2522,
+      "step": 28190
+    },
+    {
+      "epoch": 42.41,
+      "grad_norm": 3.704941987991333,
+      "learning_rate": 5.759398496240602e-06,
+      "loss": 0.2382,
+      "step": 28200
+    },
+    {
+      "epoch": 42.42,
+      "grad_norm": 4.1319780349731445,
+      "learning_rate": 5.757894736842106e-06,
+      "loss": 0.2014,
+      "step": 28210
+    },
+    {
+      "epoch": 42.44,
+      "grad_norm": 4.998769760131836,
+      "learning_rate": 5.75639097744361e-06,
+      "loss": 0.2205,
+      "step": 28220
+    },
+    {
+      "epoch": 42.45,
+      "grad_norm": 2.3330812454223633,
+      "learning_rate": 5.7548872180451136e-06,
+      "loss": 0.2242,
+      "step": 28230
+    },
+    {
+      "epoch": 42.47,
+      "grad_norm": 3.1000936031341553,
+      "learning_rate": 5.753383458646618e-06,
+      "loss": 0.2238,
+      "step": 28240
+    },
+    {
+      "epoch": 42.48,
+      "grad_norm": 5.262942790985107,
+      "learning_rate": 5.751879699248121e-06,
+      "loss": 0.2225,
+      "step": 28250
+    },
+    {
+      "epoch": 42.5,
+      "grad_norm": 7.2910475730896,
+      "learning_rate": 5.750375939849625e-06,
+      "loss": 0.2829,
+      "step": 28260
+    },
+    {
+      "epoch": 42.51,
+      "grad_norm": 2.4421637058258057,
+      "learning_rate": 5.7488721804511285e-06,
+      "loss": 0.284,
+      "step": 28270
+    },
+    {
+      "epoch": 42.53,
+      "grad_norm": 2.7164719104766846,
+      "learning_rate": 5.747368421052633e-06,
+      "loss": 0.2569,
+      "step": 28280
+    },
+    {
+      "epoch": 42.54,
+      "grad_norm": 5.2328877449035645,
+      "learning_rate": 5.7458646616541355e-06,
+      "loss": 0.2374,
+      "step": 28290
+    },
+    {
+      "epoch": 42.56,
+      "grad_norm": 3.4275119304656982,
+      "learning_rate": 5.744360902255639e-06,
+      "loss": 0.2283,
+      "step": 28300
+    },
+    {
+      "epoch": 42.57,
+      "grad_norm": 3.65972900390625,
+      "learning_rate": 5.742857142857143e-06,
+      "loss": 0.307,
+      "step": 28310
+    },
+    {
+      "epoch": 42.59,
+      "grad_norm": 4.684558391571045,
+      "learning_rate": 5.741353383458647e-06,
+      "loss": 0.2323,
+      "step": 28320
+    },
+    {
+      "epoch": 42.6,
+      "grad_norm": 7.553560733795166,
+      "learning_rate": 5.739849624060151e-06,
+      "loss": 0.1697,
+      "step": 28330
+    },
+    {
+      "epoch": 42.62,
+      "grad_norm": 5.401693820953369,
+      "learning_rate": 5.738345864661654e-06,
+      "loss": 0.2497,
+      "step": 28340
+    },
+    {
+      "epoch": 42.63,
+      "grad_norm": 4.120377540588379,
+      "learning_rate": 5.736842105263158e-06,
+      "loss": 0.285,
+      "step": 28350
+    },
+    {
+      "epoch": 42.65,
+      "grad_norm": 6.782698154449463,
+      "learning_rate": 5.735338345864662e-06,
+      "loss": 0.2505,
+      "step": 28360
+    },
+    {
+      "epoch": 42.66,
+      "grad_norm": 10.522151947021484,
+      "learning_rate": 5.733834586466166e-06,
+      "loss": 0.234,
+      "step": 28370
+    },
+    {
+      "epoch": 42.68,
+      "grad_norm": 6.4431233406066895,
+      "learning_rate": 5.732330827067669e-06,
+      "loss": 0.2463,
+      "step": 28380
+    },
+    {
+      "epoch": 42.69,
+      "grad_norm": 5.447474479675293,
+      "learning_rate": 5.730827067669173e-06,
+      "loss": 0.334,
+      "step": 28390
+    },
+    {
+      "epoch": 42.71,
+      "grad_norm": 7.193316459655762,
+      "learning_rate": 5.729323308270677e-06,
+      "loss": 0.1952,
+      "step": 28400
+    },
+    {
+      "epoch": 42.72,
+      "grad_norm": 3.5891122817993164,
+      "learning_rate": 5.727819548872181e-06,
+      "loss": 0.222,
+      "step": 28410
+    },
+    {
+      "epoch": 42.74,
+      "grad_norm": 2.9756336212158203,
+      "learning_rate": 5.726315789473685e-06,
+      "loss": 0.2087,
+      "step": 28420
+    },
+    {
+      "epoch": 42.75,
+      "grad_norm": 5.7030720710754395,
+      "learning_rate": 5.724812030075188e-06,
+      "loss": 0.2127,
+      "step": 28430
+    },
+    {
+      "epoch": 42.77,
+      "grad_norm": 4.419146537780762,
+      "learning_rate": 5.723308270676692e-06,
+      "loss": 0.1878,
+      "step": 28440
+    },
+    {
+      "epoch": 42.78,
+      "grad_norm": 5.313589096069336,
+      "learning_rate": 5.721804511278196e-06,
+      "loss": 0.2445,
+      "step": 28450
+    },
+    {
+      "epoch": 42.8,
+      "grad_norm": 5.905307769775391,
+      "learning_rate": 5.7203007518797e-06,
+      "loss": 0.2381,
+      "step": 28460
+    },
+    {
+      "epoch": 42.81,
+      "grad_norm": 5.510866641998291,
+      "learning_rate": 5.718796992481204e-06,
+      "loss": 0.2788,
+      "step": 28470
+    },
+    {
+      "epoch": 42.83,
+      "grad_norm": 5.849859237670898,
+      "learning_rate": 5.717293233082707e-06,
+      "loss": 0.2436,
+      "step": 28480
+    },
+    {
+      "epoch": 42.84,
+      "grad_norm": 5.243779182434082,
+      "learning_rate": 5.715789473684211e-06,
+      "loss": 0.1705,
+      "step": 28490
+    },
+    {
+      "epoch": 42.86,
+      "grad_norm": 3.643617868423462,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.3002,
+      "step": 28500
+    },
+    {
+      "epoch": 42.87,
+      "grad_norm": 3.605794906616211,
+      "learning_rate": 5.712781954887219e-06,
+      "loss": 0.1879,
+      "step": 28510
+    },
+    {
+      "epoch": 42.89,
+      "grad_norm": 4.405106067657471,
+      "learning_rate": 5.711278195488722e-06,
+      "loss": 0.222,
+      "step": 28520
+    },
+    {
+      "epoch": 42.9,
+      "grad_norm": 7.1118597984313965,
+      "learning_rate": 5.709774436090226e-06,
+      "loss": 0.3624,
+      "step": 28530
+    },
+    {
+      "epoch": 42.92,
+      "grad_norm": 5.815327167510986,
+      "learning_rate": 5.7082706766917294e-06,
+      "loss": 0.2162,
+      "step": 28540
+    },
+    {
+      "epoch": 42.93,
+      "grad_norm": 5.0367021560668945,
+      "learning_rate": 5.706766917293234e-06,
+      "loss": 0.2481,
+      "step": 28550
+    },
+    {
+      "epoch": 42.95,
+      "grad_norm": 6.6812005043029785,
+      "learning_rate": 5.705263157894737e-06,
+      "loss": 0.2977,
+      "step": 28560
+    },
+    {
+      "epoch": 42.96,
+      "grad_norm": 3.0472841262817383,
+      "learning_rate": 5.703759398496242e-06,
+      "loss": 0.225,
+      "step": 28570
+    },
+    {
+      "epoch": 42.98,
+      "grad_norm": 5.001314163208008,
+      "learning_rate": 5.702255639097744e-06,
+      "loss": 0.1936,
+      "step": 28580
+    },
+    {
+      "epoch": 42.99,
+      "grad_norm": 7.2746124267578125,
+      "learning_rate": 5.700751879699249e-06,
+      "loss": 0.2208,
+      "step": 28590
+    },
+    {
+      "epoch": 43.0,
+      "eval_accuracy": 0.9316,
+      "eval_loss": 0.30223846435546875,
+      "eval_runtime": 84.2035,
+      "eval_samples_per_second": 118.76,
+      "eval_steps_per_second": 0.475,
+      "step": 28595
+    },
+    {
+      "epoch": 43.01,
+      "grad_norm": 5.429502010345459,
+      "learning_rate": 5.699248120300752e-06,
+      "loss": 0.2629,
+      "step": 28600
+    },
+    {
+      "epoch": 43.02,
+      "grad_norm": 8.030689239501953,
+      "learning_rate": 5.697744360902257e-06,
+      "loss": 0.1887,
+      "step": 28610
+    },
+    {
+      "epoch": 43.04,
+      "grad_norm": 4.963591575622559,
+      "learning_rate": 5.696240601503759e-06,
+      "loss": 0.2608,
+      "step": 28620
+    },
+    {
+      "epoch": 43.05,
+      "grad_norm": 7.090287685394287,
+      "learning_rate": 5.694736842105264e-06,
+      "loss": 0.248,
+      "step": 28630
+    },
+    {
+      "epoch": 43.07,
+      "grad_norm": 5.958043575286865,
+      "learning_rate": 5.693233082706767e-06,
+      "loss": 0.3033,
+      "step": 28640
+    },
+    {
+      "epoch": 43.08,
+      "grad_norm": 6.555380821228027,
+      "learning_rate": 5.6917293233082715e-06,
+      "loss": 0.2399,
+      "step": 28650
+    },
+    {
+      "epoch": 43.1,
+      "grad_norm": 7.416017055511475,
+      "learning_rate": 5.690225563909775e-06,
+      "loss": 0.2513,
+      "step": 28660
+    },
+    {
+      "epoch": 43.11,
+      "grad_norm": 4.798736095428467,
+      "learning_rate": 5.688721804511279e-06,
+      "loss": 0.2721,
+      "step": 28670
+    },
+    {
+      "epoch": 43.13,
+      "grad_norm": 6.545383930206299,
+      "learning_rate": 5.687218045112782e-06,
+      "loss": 0.2962,
+      "step": 28680
+    },
+    {
+      "epoch": 43.14,
+      "grad_norm": 7.702247142791748,
+      "learning_rate": 5.6857142857142865e-06,
+      "loss": 0.2098,
+      "step": 28690
+    },
+    {
+      "epoch": 43.16,
+      "grad_norm": 2.8116698265075684,
+      "learning_rate": 5.68421052631579e-06,
+      "loss": 0.1915,
+      "step": 28700
+    },
+    {
+      "epoch": 43.17,
+      "grad_norm": 8.904080390930176,
+      "learning_rate": 5.682706766917294e-06,
+      "loss": 0.3113,
+      "step": 28710
+    },
+    {
+      "epoch": 43.19,
+      "grad_norm": 3.884316921234131,
+      "learning_rate": 5.681203007518797e-06,
+      "loss": 0.2209,
+      "step": 28720
+    },
+    {
+      "epoch": 43.2,
+      "grad_norm": 3.413797616958618,
+      "learning_rate": 5.679699248120301e-06,
+      "loss": 0.1643,
+      "step": 28730
+    },
+    {
+      "epoch": 43.22,
+      "grad_norm": 7.109938144683838,
+      "learning_rate": 5.678195488721805e-06,
+      "loss": 0.2486,
+      "step": 28740
+    },
+    {
+      "epoch": 43.23,
+      "grad_norm": 4.222686290740967,
+      "learning_rate": 5.676691729323309e-06,
+      "loss": 0.2572,
+      "step": 28750
+    },
+    {
+      "epoch": 43.25,
+      "grad_norm": 4.973549842834473,
+      "learning_rate": 5.675187969924813e-06,
+      "loss": 0.2302,
+      "step": 28760
+    },
+    {
+      "epoch": 43.26,
+      "grad_norm": 4.342545986175537,
+      "learning_rate": 5.673684210526317e-06,
+      "loss": 0.2825,
+      "step": 28770
+    },
+    {
+      "epoch": 43.28,
+      "grad_norm": 5.619150638580322,
+      "learning_rate": 5.67218045112782e-06,
+      "loss": 0.2352,
+      "step": 28780
+    },
+    {
+      "epoch": 43.29,
+      "grad_norm": 6.378529071807861,
+      "learning_rate": 5.670676691729323e-06,
+      "loss": 0.231,
+      "step": 28790
+    },
+    {
+      "epoch": 43.31,
+      "grad_norm": 3.949969530105591,
+      "learning_rate": 5.669172932330828e-06,
+      "loss": 0.2399,
+      "step": 28800
+    },
+    {
+      "epoch": 43.32,
+      "grad_norm": 3.982423782348633,
+      "learning_rate": 5.66766917293233e-06,
+      "loss": 0.1841,
+      "step": 28810
+    },
+    {
+      "epoch": 43.34,
+      "grad_norm": 3.037238359451294,
+      "learning_rate": 5.666165413533835e-06,
+      "loss": 0.2084,
+      "step": 28820
+    },
+    {
+      "epoch": 43.35,
+      "grad_norm": 5.667693614959717,
+      "learning_rate": 5.664661654135338e-06,
+      "loss": 0.2061,
+      "step": 28830
+    },
+    {
+      "epoch": 43.37,
+      "grad_norm": 5.317113876342773,
+      "learning_rate": 5.663157894736843e-06,
+      "loss": 0.2213,
+      "step": 28840
+    },
+    {
+      "epoch": 43.38,
+      "grad_norm": 7.288003921508789,
+      "learning_rate": 5.661654135338346e-06,
+      "loss": 0.3024,
+      "step": 28850
+    },
+    {
+      "epoch": 43.4,
+      "grad_norm": 6.291513919830322,
+      "learning_rate": 5.6601503759398505e-06,
+      "loss": 0.2002,
+      "step": 28860
+    },
+    {
+      "epoch": 43.41,
+      "grad_norm": 4.533992767333984,
+      "learning_rate": 5.658646616541353e-06,
+      "loss": 0.2687,
+      "step": 28870
+    },
+    {
+      "epoch": 43.43,
+      "grad_norm": 5.96859073638916,
+      "learning_rate": 5.6571428571428576e-06,
+      "loss": 0.2142,
+      "step": 28880
+    },
+    {
+      "epoch": 43.44,
+      "grad_norm": 5.588306903839111,
+      "learning_rate": 5.655639097744361e-06,
+      "loss": 0.2045,
+      "step": 28890
+    },
+    {
+      "epoch": 43.46,
+      "grad_norm": 3.9208312034606934,
+      "learning_rate": 5.6541353383458654e-06,
+      "loss": 0.3291,
+      "step": 28900
+    },
+    {
+      "epoch": 43.47,
+      "grad_norm": 4.575656414031982,
+      "learning_rate": 5.652631578947368e-06,
+      "loss": 0.203,
+      "step": 28910
+    },
+    {
+      "epoch": 43.49,
+      "grad_norm": 4.715184211730957,
+      "learning_rate": 5.6511278195488725e-06,
+      "loss": 0.2584,
+      "step": 28920
+    },
+    {
+      "epoch": 43.5,
+      "grad_norm": 7.345559597015381,
+      "learning_rate": 5.649624060150376e-06,
+      "loss": 0.2279,
+      "step": 28930
+    },
+    {
+      "epoch": 43.52,
+      "grad_norm": 8.647790908813477,
+      "learning_rate": 5.64812030075188e-06,
+      "loss": 0.286,
+      "step": 28940
+    },
+    {
+      "epoch": 43.53,
+      "grad_norm": 3.186676025390625,
+      "learning_rate": 5.646616541353384e-06,
+      "loss": 0.2159,
+      "step": 28950
+    },
+    {
+      "epoch": 43.55,
+      "grad_norm": 5.287961006164551,
+      "learning_rate": 5.645112781954888e-06,
+      "loss": 0.2934,
+      "step": 28960
+    },
+    {
+      "epoch": 43.56,
+      "grad_norm": 1.5112135410308838,
+      "learning_rate": 5.643609022556391e-06,
+      "loss": 0.2194,
+      "step": 28970
+    },
+    {
+      "epoch": 43.58,
+      "grad_norm": 4.162924289703369,
+      "learning_rate": 5.642105263157895e-06,
+      "loss": 0.1885,
+      "step": 28980
+    },
+    {
+      "epoch": 43.59,
+      "grad_norm": 6.321695327758789,
+      "learning_rate": 5.640601503759399e-06,
+      "loss": 0.1848,
+      "step": 28990
+    },
+    {
+      "epoch": 43.61,
+      "grad_norm": 7.07379674911499,
+      "learning_rate": 5.639097744360903e-06,
+      "loss": 0.2176,
+      "step": 29000
+    },
+    {
+      "epoch": 43.62,
+      "grad_norm": 2.475600242614746,
+      "learning_rate": 5.637593984962406e-06,
+      "loss": 0.2473,
+      "step": 29010
+    },
+    {
+      "epoch": 43.64,
+      "grad_norm": 2.369236946105957,
+      "learning_rate": 5.63609022556391e-06,
+      "loss": 0.2338,
+      "step": 29020
+    },
+    {
+      "epoch": 43.65,
+      "grad_norm": 8.563982963562012,
+      "learning_rate": 5.634586466165414e-06,
+      "loss": 0.2339,
+      "step": 29030
+    },
+    {
+      "epoch": 43.67,
+      "grad_norm": 11.370944023132324,
+      "learning_rate": 5.633082706766918e-06,
+      "loss": 0.2524,
+      "step": 29040
+    },
+    {
+      "epoch": 43.68,
+      "grad_norm": 8.391033172607422,
+      "learning_rate": 5.631578947368422e-06,
+      "loss": 0.2441,
+      "step": 29050
+    },
+    {
+      "epoch": 43.7,
+      "grad_norm": 4.559657096862793,
+      "learning_rate": 5.630075187969926e-06,
+      "loss": 0.2515,
+      "step": 29060
+    },
+    {
+      "epoch": 43.71,
+      "grad_norm": 4.1660637855529785,
+      "learning_rate": 5.628571428571429e-06,
+      "loss": 0.2495,
+      "step": 29070
+    },
+    {
+      "epoch": 43.73,
+      "grad_norm": 8.0914306640625,
+      "learning_rate": 5.627067669172933e-06,
+      "loss": 0.2166,
+      "step": 29080
+    },
+    {
+      "epoch": 43.74,
+      "grad_norm": 5.4428019523620605,
+      "learning_rate": 5.6255639097744365e-06,
+      "loss": 0.2574,
+      "step": 29090
+    },
+    {
+      "epoch": 43.76,
+      "grad_norm": 4.377140045166016,
+      "learning_rate": 5.624060150375941e-06,
+      "loss": 0.2215,
+      "step": 29100
+    },
+    {
+      "epoch": 43.77,
+      "grad_norm": 5.647352695465088,
+      "learning_rate": 5.6225563909774436e-06,
+      "loss": 0.2947,
+      "step": 29110
+    },
+    {
+      "epoch": 43.79,
+      "grad_norm": 4.882458209991455,
+      "learning_rate": 5.621052631578948e-06,
+      "loss": 0.2522,
+      "step": 29120
+    },
+    {
+      "epoch": 43.8,
+      "grad_norm": 6.157678604125977,
+      "learning_rate": 5.6195488721804515e-06,
+      "loss": 0.221,
+      "step": 29130
+    },
+    {
+      "epoch": 43.82,
+      "grad_norm": 6.887104511260986,
+      "learning_rate": 5.618045112781956e-06,
+      "loss": 0.2797,
+      "step": 29140
+    },
+    {
+      "epoch": 43.83,
+      "grad_norm": 3.868295907974243,
+      "learning_rate": 5.616541353383459e-06,
+      "loss": 0.2551,
+      "step": 29150
+    },
+    {
+      "epoch": 43.85,
+      "grad_norm": 9.585166931152344,
+      "learning_rate": 5.615037593984964e-06,
+      "loss": 0.2301,
+      "step": 29160
+    },
+    {
+      "epoch": 43.86,
+      "grad_norm": 5.948240756988525,
+      "learning_rate": 5.613533834586466e-06,
+      "loss": 0.2388,
+      "step": 29170
+    },
+    {
+      "epoch": 43.88,
+      "grad_norm": 5.998847961425781,
+      "learning_rate": 5.612030075187971e-06,
+      "loss": 0.2077,
+      "step": 29180
+    },
+    {
+      "epoch": 43.89,
+      "grad_norm": 5.52202844619751,
+      "learning_rate": 5.610526315789474e-06,
+      "loss": 0.1981,
+      "step": 29190
+    },
+    {
+      "epoch": 43.91,
+      "grad_norm": 3.2842657566070557,
+      "learning_rate": 5.609022556390979e-06,
+      "loss": 0.2744,
+      "step": 29200
+    },
+    {
+      "epoch": 43.92,
+      "grad_norm": 6.214591026306152,
+      "learning_rate": 5.607518796992481e-06,
+      "loss": 0.1858,
+      "step": 29210
+    },
+    {
+      "epoch": 43.94,
+      "grad_norm": 5.361661911010742,
+      "learning_rate": 5.606015037593986e-06,
+      "loss": 0.2499,
+      "step": 29220
+    },
+    {
+      "epoch": 43.95,
+      "grad_norm": 5.610089302062988,
+      "learning_rate": 5.604511278195489e-06,
+      "loss": 0.2287,
+      "step": 29230
+    },
+    {
+      "epoch": 43.97,
+      "grad_norm": 5.380805492401123,
+      "learning_rate": 5.6030075187969936e-06,
+      "loss": 0.2702,
+      "step": 29240
+    },
+    {
+      "epoch": 43.98,
+      "grad_norm": 3.860438823699951,
+      "learning_rate": 5.601503759398497e-06,
+      "loss": 0.2509,
+      "step": 29250
+    },
+    {
+      "epoch": 44.0,
+      "grad_norm": 0.10728010535240173,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 0.1808,
+      "step": 29260
+    },
+    {
+      "epoch": 44.0,
+      "eval_accuracy": 0.9304,
+      "eval_loss": 0.30674317479133606,
+      "eval_runtime": 85.2945,
+      "eval_samples_per_second": 117.241,
+      "eval_steps_per_second": 0.469,
+      "step": 29260
+    },
+    {
+      "epoch": 44.02,
+      "grad_norm": 7.593850612640381,
+      "learning_rate": 5.598496240601504e-06,
+      "loss": 0.2738,
+      "step": 29270
+    },
+    {
+      "epoch": 44.03,
+      "grad_norm": 6.34241247177124,
+      "learning_rate": 5.596992481203008e-06,
+      "loss": 0.2215,
+      "step": 29280
+    },
+    {
+      "epoch": 44.05,
+      "grad_norm": 5.706809997558594,
+      "learning_rate": 5.595488721804512e-06,
+      "loss": 0.277,
+      "step": 29290
+    },
+    {
+      "epoch": 44.06,
+      "grad_norm": 4.330235481262207,
+      "learning_rate": 5.593984962406015e-06,
+      "loss": 0.2421,
+      "step": 29300
+    },
+    {
+      "epoch": 44.08,
+      "grad_norm": 6.897051811218262,
+      "learning_rate": 5.592481203007519e-06,
+      "loss": 0.2199,
+      "step": 29310
+    },
+    {
+      "epoch": 44.09,
+      "grad_norm": 3.312359571456909,
+      "learning_rate": 5.5909774436090226e-06,
+      "loss": 0.1884,
+      "step": 29320
+    },
+    {
+      "epoch": 44.11,
+      "grad_norm": 4.363166809082031,
+      "learning_rate": 5.589473684210527e-06,
+      "loss": 0.255,
+      "step": 29330
+    },
+    {
+      "epoch": 44.12,
+      "grad_norm": 7.3857269287109375,
+      "learning_rate": 5.5879699248120304e-06,
+      "loss": 0.298,
+      "step": 29340
+    },
+    {
+      "epoch": 44.14,
+      "grad_norm": 4.645081520080566,
+      "learning_rate": 5.586466165413534e-06,
+      "loss": 0.318,
+      "step": 29350
+    },
+    {
+      "epoch": 44.15,
+      "grad_norm": 5.76023006439209,
+      "learning_rate": 5.5849624060150375e-06,
+      "loss": 0.2442,
+      "step": 29360
+    },
+    {
+      "epoch": 44.17,
+      "grad_norm": 3.3517050743103027,
+      "learning_rate": 5.583458646616542e-06,
+      "loss": 0.2442,
+      "step": 29370
+    },
+    {
+      "epoch": 44.18,
+      "grad_norm": 4.758605480194092,
+      "learning_rate": 5.581954887218045e-06,
+      "loss": 0.2328,
+      "step": 29380
+    },
+    {
+      "epoch": 44.2,
+      "grad_norm": 5.125278949737549,
+      "learning_rate": 5.58045112781955e-06,
+      "loss": 0.2316,
+      "step": 29390
+    },
+    {
+      "epoch": 44.21,
+      "grad_norm": 5.346681594848633,
+      "learning_rate": 5.578947368421052e-06,
+      "loss": 0.2119,
+      "step": 29400
+    },
+    {
+      "epoch": 44.23,
+      "grad_norm": 5.317344665527344,
+      "learning_rate": 5.577443609022557e-06,
+      "loss": 0.1829,
+      "step": 29410
+    },
+    {
+      "epoch": 44.24,
+      "grad_norm": 6.973268985748291,
+      "learning_rate": 5.57593984962406e-06,
+      "loss": 0.2207,
+      "step": 29420
+    },
+    {
+      "epoch": 44.26,
+      "grad_norm": 2.5319881439208984,
+      "learning_rate": 5.574436090225565e-06,
+      "loss": 0.2677,
+      "step": 29430
+    },
+    {
+      "epoch": 44.27,
+      "grad_norm": 1.6542171239852905,
+      "learning_rate": 5.572932330827068e-06,
+      "loss": 0.2092,
+      "step": 29440
+    },
+    {
+      "epoch": 44.29,
+      "grad_norm": 5.905990123748779,
+      "learning_rate": 5.571428571428572e-06,
+      "loss": 0.2059,
+      "step": 29450
+    },
+    {
+      "epoch": 44.3,
+      "grad_norm": 3.364076614379883,
+      "learning_rate": 5.569924812030075e-06,
+      "loss": 0.2847,
+      "step": 29460
+    },
+    {
+      "epoch": 44.32,
+      "grad_norm": 2.6462037563323975,
+      "learning_rate": 5.5684210526315796e-06,
+      "loss": 0.2227,
+      "step": 29470
+    },
+    {
+      "epoch": 44.33,
+      "grad_norm": 4.426711559295654,
+      "learning_rate": 5.566917293233083e-06,
+      "loss": 0.2758,
+      "step": 29480
+    },
+    {
+      "epoch": 44.35,
+      "grad_norm": 6.097959518432617,
+      "learning_rate": 5.5654135338345875e-06,
+      "loss": 0.256,
+      "step": 29490
+    },
+    {
+      "epoch": 44.36,
+      "grad_norm": 3.295834541320801,
+      "learning_rate": 5.56390977443609e-06,
+      "loss": 0.2586,
+      "step": 29500
+    },
+    {
+      "epoch": 44.38,
+      "grad_norm": 6.391618251800537,
+      "learning_rate": 5.5624060150375945e-06,
+      "loss": 0.2186,
+      "step": 29510
+    },
+    {
+      "epoch": 44.39,
+      "grad_norm": 5.297180652618408,
+      "learning_rate": 5.560902255639098e-06,
+      "loss": 0.2465,
+      "step": 29520
+    },
+    {
+      "epoch": 44.41,
+      "grad_norm": 3.471672773361206,
+      "learning_rate": 5.559398496240602e-06,
+      "loss": 0.2363,
+      "step": 29530
+    },
+    {
+      "epoch": 44.42,
+      "grad_norm": 3.7030515670776367,
+      "learning_rate": 5.557894736842105e-06,
+      "loss": 0.2308,
+      "step": 29540
+    },
+    {
+      "epoch": 44.44,
+      "grad_norm": 4.913259029388428,
+      "learning_rate": 5.556390977443609e-06,
+      "loss": 0.2459,
+      "step": 29550
+    },
+    {
+      "epoch": 44.45,
+      "grad_norm": 7.469844818115234,
+      "learning_rate": 5.554887218045113e-06,
+      "loss": 0.1795,
+      "step": 29560
+    },
+    {
+      "epoch": 44.47,
+      "grad_norm": 6.876651763916016,
+      "learning_rate": 5.553383458646617e-06,
+      "loss": 0.2296,
+      "step": 29570
+    },
+    {
+      "epoch": 44.48,
+      "grad_norm": 3.4335014820098877,
+      "learning_rate": 5.551879699248121e-06,
+      "loss": 0.2769,
+      "step": 29580
+    },
+    {
+      "epoch": 44.5,
+      "grad_norm": 5.788983345031738,
+      "learning_rate": 5.550375939849625e-06,
+      "loss": 0.2428,
+      "step": 29590
+    },
+    {
+      "epoch": 44.51,
+      "grad_norm": 4.837071895599365,
+      "learning_rate": 5.548872180451128e-06,
+      "loss": 0.2378,
+      "step": 29600
+    },
+    {
+      "epoch": 44.53,
+      "grad_norm": 4.5005202293396,
+      "learning_rate": 5.547368421052632e-06,
+      "loss": 0.2397,
+      "step": 29610
+    },
+    {
+      "epoch": 44.54,
+      "grad_norm": 4.948091983795166,
+      "learning_rate": 5.545864661654136e-06,
+      "loss": 0.2521,
+      "step": 29620
+    },
+    {
+      "epoch": 44.56,
+      "grad_norm": 7.149682998657227,
+      "learning_rate": 5.54436090225564e-06,
+      "loss": 0.2811,
+      "step": 29630
+    },
+    {
+      "epoch": 44.57,
+      "grad_norm": 4.339492321014404,
+      "learning_rate": 5.542857142857143e-06,
+      "loss": 0.19,
+      "step": 29640
+    },
+    {
+      "epoch": 44.59,
+      "grad_norm": 5.427370548248291,
+      "learning_rate": 5.541353383458647e-06,
+      "loss": 0.2379,
+      "step": 29650
+    },
+    {
+      "epoch": 44.6,
+      "grad_norm": 3.6944634914398193,
+      "learning_rate": 5.539849624060151e-06,
+      "loss": 0.183,
+      "step": 29660
+    },
+    {
+      "epoch": 44.62,
+      "grad_norm": 5.690896034240723,
+      "learning_rate": 5.538345864661655e-06,
+      "loss": 0.1527,
+      "step": 29670
+    },
+    {
+      "epoch": 44.63,
+      "grad_norm": 5.7550048828125,
+      "learning_rate": 5.5368421052631586e-06,
+      "loss": 0.1855,
+      "step": 29680
+    },
+    {
+      "epoch": 44.65,
+      "grad_norm": 2.685657024383545,
+      "learning_rate": 5.535338345864663e-06,
+      "loss": 0.2244,
+      "step": 29690
+    },
+    {
+      "epoch": 44.66,
+      "grad_norm": 3.263343334197998,
+      "learning_rate": 5.533834586466166e-06,
+      "loss": 0.2051,
+      "step": 29700
+    },
+    {
+      "epoch": 44.68,
+      "grad_norm": 5.3920183181762695,
+      "learning_rate": 5.53233082706767e-06,
+      "loss": 0.2987,
+      "step": 29710
+    },
+    {
+      "epoch": 44.69,
+      "grad_norm": 9.512574195861816,
+      "learning_rate": 5.5308270676691735e-06,
+      "loss": 0.2528,
+      "step": 29720
+    },
+    {
+      "epoch": 44.71,
+      "grad_norm": 8.392505645751953,
+      "learning_rate": 5.529323308270678e-06,
+      "loss": 0.2906,
+      "step": 29730
+    },
+    {
+      "epoch": 44.72,
+      "grad_norm": 5.764971733093262,
+      "learning_rate": 5.5278195488721805e-06,
+      "loss": 0.2531,
+      "step": 29740
+    },
+    {
+      "epoch": 44.74,
+      "grad_norm": 4.783633232116699,
+      "learning_rate": 5.526315789473685e-06,
+      "loss": 0.3142,
+      "step": 29750
+    },
+    {
+      "epoch": 44.75,
+      "grad_norm": 7.186029434204102,
+      "learning_rate": 5.524812030075188e-06,
+      "loss": 0.2501,
+      "step": 29760
+    },
+    {
+      "epoch": 44.77,
+      "grad_norm": 1.497518539428711,
+      "learning_rate": 5.523308270676693e-06,
+      "loss": 0.2384,
+      "step": 29770
+    },
+    {
+      "epoch": 44.78,
+      "grad_norm": 8.040397644042969,
+      "learning_rate": 5.521804511278196e-06,
+      "loss": 0.1935,
+      "step": 29780
+    },
+    {
+      "epoch": 44.8,
+      "grad_norm": 5.375740051269531,
+      "learning_rate": 5.520300751879699e-06,
+      "loss": 0.2086,
+      "step": 29790
+    },
+    {
+      "epoch": 44.81,
+      "grad_norm": 4.856134414672852,
+      "learning_rate": 5.518796992481203e-06,
+      "loss": 0.2516,
+      "step": 29800
+    },
+    {
+      "epoch": 44.83,
+      "grad_norm": 5.769529819488525,
+      "learning_rate": 5.517293233082707e-06,
+      "loss": 0.2667,
+      "step": 29810
+    },
+    {
+      "epoch": 44.84,
+      "grad_norm": 3.9014487266540527,
+      "learning_rate": 5.515789473684211e-06,
+      "loss": 0.2218,
+      "step": 29820
+    },
+    {
+      "epoch": 44.86,
+      "grad_norm": 3.6634175777435303,
+      "learning_rate": 5.514285714285714e-06,
+      "loss": 0.2317,
+      "step": 29830
+    },
+    {
+      "epoch": 44.87,
+      "grad_norm": 3.2795653343200684,
+      "learning_rate": 5.512781954887218e-06,
+      "loss": 0.2201,
+      "step": 29840
+    },
+    {
+      "epoch": 44.89,
+      "grad_norm": 3.049172878265381,
+      "learning_rate": 5.511278195488722e-06,
+      "loss": 0.1521,
+      "step": 29850
+    },
+    {
+      "epoch": 44.9,
+      "grad_norm": 3.333717107772827,
+      "learning_rate": 5.509774436090226e-06,
+      "loss": 0.272,
+      "step": 29860
+    },
+    {
+      "epoch": 44.92,
+      "grad_norm": 6.023979187011719,
+      "learning_rate": 5.50827067669173e-06,
+      "loss": 0.2422,
+      "step": 29870
+    },
+    {
+      "epoch": 44.93,
+      "grad_norm": 2.427889347076416,
+      "learning_rate": 5.506766917293234e-06,
+      "loss": 0.219,
+      "step": 29880
+    },
+    {
+      "epoch": 44.95,
+      "grad_norm": 7.4705729484558105,
+      "learning_rate": 5.505263157894737e-06,
+      "loss": 0.2594,
+      "step": 29890
+    },
+    {
+      "epoch": 44.96,
+      "grad_norm": 3.8348019123077393,
+      "learning_rate": 5.503759398496241e-06,
+      "loss": 0.2171,
+      "step": 29900
+    },
+    {
+      "epoch": 44.98,
+      "grad_norm": 6.202742576599121,
+      "learning_rate": 5.502255639097745e-06,
+      "loss": 0.3124,
+      "step": 29910
+    },
+    {
+      "epoch": 44.99,
+      "grad_norm": 6.531280517578125,
+      "learning_rate": 5.500751879699249e-06,
+      "loss": 0.2477,
+      "step": 29920
+    },
+    {
+      "epoch": 45.0,
+      "eval_accuracy": 0.9289,
+      "eval_loss": 0.30727890133857727,
+      "eval_runtime": 84.3468,
+      "eval_samples_per_second": 118.558,
+      "eval_steps_per_second": 0.474,
+      "step": 29925
+    },
+    {
+      "epoch": 45.01,
+      "grad_norm": 6.2103376388549805,
+      "learning_rate": 5.499248120300752e-06,
+      "loss": 0.3184,
+      "step": 29930
+    },
+    {
+      "epoch": 45.02,
+      "grad_norm": 4.898624897003174,
+      "learning_rate": 5.497744360902256e-06,
+      "loss": 0.2263,
+      "step": 29940
+    },
+    {
+      "epoch": 45.04,
+      "grad_norm": 5.601806640625,
+      "learning_rate": 5.4962406015037595e-06,
+      "loss": 0.1897,
+      "step": 29950
+    },
+    {
+      "epoch": 45.05,
+      "grad_norm": 3.2715799808502197,
+      "learning_rate": 5.494736842105264e-06,
+      "loss": 0.2605,
+      "step": 29960
+    },
+    {
+      "epoch": 45.07,
+      "grad_norm": 4.670425891876221,
+      "learning_rate": 5.493233082706767e-06,
+      "loss": 0.2316,
+      "step": 29970
+    },
+    {
+      "epoch": 45.08,
+      "grad_norm": 3.2674357891082764,
+      "learning_rate": 5.491729323308272e-06,
+      "loss": 0.1808,
+      "step": 29980
+    },
+    {
+      "epoch": 45.1,
+      "grad_norm": 4.537569522857666,
+      "learning_rate": 5.4902255639097744e-06,
+      "loss": 0.2665,
+      "step": 29990
+    },
+    {
+      "epoch": 45.11,
+      "grad_norm": 4.6132378578186035,
+      "learning_rate": 5.488721804511279e-06,
+      "loss": 0.2267,
+      "step": 30000
+    },
+    {
+      "epoch": 45.13,
+      "grad_norm": 6.702213764190674,
+      "learning_rate": 5.487218045112782e-06,
+      "loss": 0.2676,
+      "step": 30010
+    },
+    {
+      "epoch": 45.14,
+      "grad_norm": 5.990668296813965,
+      "learning_rate": 5.485714285714287e-06,
+      "loss": 0.265,
+      "step": 30020
+    },
+    {
+      "epoch": 45.16,
+      "grad_norm": 4.7114787101745605,
+      "learning_rate": 5.484210526315789e-06,
+      "loss": 0.204,
+      "step": 30030
+    },
+    {
+      "epoch": 45.17,
+      "grad_norm": 2.577725648880005,
+      "learning_rate": 5.482706766917294e-06,
+      "loss": 0.2601,
+      "step": 30040
+    },
+    {
+      "epoch": 45.19,
+      "grad_norm": 7.188079833984375,
+      "learning_rate": 5.481203007518797e-06,
+      "loss": 0.2447,
+      "step": 30050
+    },
+    {
+      "epoch": 45.2,
+      "grad_norm": 8.671704292297363,
+      "learning_rate": 5.479699248120302e-06,
+      "loss": 0.202,
+      "step": 30060
+    },
+    {
+      "epoch": 45.22,
+      "grad_norm": 1.6407883167266846,
+      "learning_rate": 5.478195488721805e-06,
+      "loss": 0.2504,
+      "step": 30070
+    },
+    {
+      "epoch": 45.23,
+      "grad_norm": 6.040456295013428,
+      "learning_rate": 5.476691729323309e-06,
+      "loss": 0.2701,
+      "step": 30080
+    },
+    {
+      "epoch": 45.25,
+      "grad_norm": 4.857377529144287,
+      "learning_rate": 5.475187969924812e-06,
+      "loss": 0.1846,
+      "step": 30090
+    },
+    {
+      "epoch": 45.26,
+      "grad_norm": 4.847814083099365,
+      "learning_rate": 5.4736842105263165e-06,
+      "loss": 0.2355,
+      "step": 30100
+    },
+    {
+      "epoch": 45.28,
+      "grad_norm": 5.133174896240234,
+      "learning_rate": 5.47218045112782e-06,
+      "loss": 0.2244,
+      "step": 30110
+    },
+    {
+      "epoch": 45.29,
+      "grad_norm": 4.437646389007568,
+      "learning_rate": 5.470676691729324e-06,
+      "loss": 0.2366,
+      "step": 30120
+    },
+    {
+      "epoch": 45.31,
+      "grad_norm": 9.677255630493164,
+      "learning_rate": 5.469172932330827e-06,
+      "loss": 0.2258,
+      "step": 30130
+    },
+    {
+      "epoch": 45.32,
+      "grad_norm": 10.691313743591309,
+      "learning_rate": 5.4676691729323314e-06,
+      "loss": 0.2529,
+      "step": 30140
+    },
+    {
+      "epoch": 45.34,
+      "grad_norm": 4.892699718475342,
+      "learning_rate": 5.466165413533835e-06,
+      "loss": 0.212,
+      "step": 30150
+    },
+    {
+      "epoch": 45.35,
+      "grad_norm": 8.477937698364258,
+      "learning_rate": 5.464661654135339e-06,
+      "loss": 0.2291,
+      "step": 30160
+    },
+    {
+      "epoch": 45.37,
+      "grad_norm": 4.592231273651123,
+      "learning_rate": 5.463157894736843e-06,
+      "loss": 0.2567,
+      "step": 30170
+    },
+    {
+      "epoch": 45.38,
+      "grad_norm": 4.459981918334961,
+      "learning_rate": 5.461654135338346e-06,
+      "loss": 0.2596,
+      "step": 30180
+    },
+    {
+      "epoch": 45.4,
+      "grad_norm": 6.071181774139404,
+      "learning_rate": 5.46015037593985e-06,
+      "loss": 0.2445,
+      "step": 30190
+    },
+    {
+      "epoch": 45.41,
+      "grad_norm": 5.071913242340088,
+      "learning_rate": 5.458646616541354e-06,
+      "loss": 0.208,
+      "step": 30200
+    },
+    {
+      "epoch": 45.43,
+      "grad_norm": 7.438906669616699,
+      "learning_rate": 5.457142857142858e-06,
+      "loss": 0.2321,
+      "step": 30210
+    },
+    {
+      "epoch": 45.44,
+      "grad_norm": 5.6083292961120605,
+      "learning_rate": 5.455639097744362e-06,
+      "loss": 0.1798,
+      "step": 30220
+    },
+    {
+      "epoch": 45.46,
+      "grad_norm": 5.303403377532959,
+      "learning_rate": 5.454135338345865e-06,
+      "loss": 0.2718,
+      "step": 30230
+    },
+    {
+      "epoch": 45.47,
+      "grad_norm": 4.844057083129883,
+      "learning_rate": 5.452631578947369e-06,
+      "loss": 0.1896,
+      "step": 30240
+    },
+    {
+      "epoch": 45.49,
+      "grad_norm": 4.131418228149414,
+      "learning_rate": 5.451127819548873e-06,
+      "loss": 0.2972,
+      "step": 30250
+    },
+    {
+      "epoch": 45.5,
+      "grad_norm": 4.064949035644531,
+      "learning_rate": 5.449624060150377e-06,
+      "loss": 0.2298,
+      "step": 30260
+    },
+    {
+      "epoch": 45.52,
+      "grad_norm": 4.775879859924316,
+      "learning_rate": 5.44812030075188e-06,
+      "loss": 0.2578,
+      "step": 30270
+    },
+    {
+      "epoch": 45.53,
+      "grad_norm": 8.634382247924805,
+      "learning_rate": 5.446616541353383e-06,
+      "loss": 0.2508,
+      "step": 30280
+    },
+    {
+      "epoch": 45.55,
+      "grad_norm": 2.0520710945129395,
+      "learning_rate": 5.445112781954888e-06,
+      "loss": 0.2033,
+      "step": 30290
+    },
+    {
+      "epoch": 45.56,
+      "grad_norm": 7.116695880889893,
+      "learning_rate": 5.443609022556391e-06,
+      "loss": 0.2528,
+      "step": 30300
+    },
+    {
+      "epoch": 45.58,
+      "grad_norm": 5.326464653015137,
+      "learning_rate": 5.4421052631578955e-06,
+      "loss": 0.2536,
+      "step": 30310
+    },
+    {
+      "epoch": 45.59,
+      "grad_norm": 10.817031860351562,
+      "learning_rate": 5.440601503759398e-06,
+      "loss": 0.2231,
+      "step": 30320
+    },
+    {
+      "epoch": 45.61,
+      "grad_norm": 4.3931965827941895,
+      "learning_rate": 5.4390977443609025e-06,
+      "loss": 0.261,
+      "step": 30330
+    },
+    {
+      "epoch": 45.62,
+      "grad_norm": 5.171022891998291,
+      "learning_rate": 5.437593984962406e-06,
+      "loss": 0.2064,
+      "step": 30340
+    },
+    {
+      "epoch": 45.64,
+      "grad_norm": 6.176837921142578,
+      "learning_rate": 5.4360902255639104e-06,
+      "loss": 0.2539,
+      "step": 30350
+    },
+    {
+      "epoch": 45.65,
+      "grad_norm": 5.6013078689575195,
+      "learning_rate": 5.434586466165413e-06,
+      "loss": 0.229,
+      "step": 30360
+    },
+    {
+      "epoch": 45.67,
+      "grad_norm": 7.163257122039795,
+      "learning_rate": 5.4330827067669175e-06,
+      "loss": 0.2193,
+      "step": 30370
+    },
+    {
+      "epoch": 45.68,
+      "grad_norm": 5.970539093017578,
+      "learning_rate": 5.431578947368421e-06,
+      "loss": 0.2238,
+      "step": 30380
+    },
+    {
+      "epoch": 45.7,
+      "grad_norm": 4.074548721313477,
+      "learning_rate": 5.430075187969925e-06,
+      "loss": 0.1859,
+      "step": 30390
+    },
+    {
+      "epoch": 45.71,
+      "grad_norm": 6.107884883880615,
+      "learning_rate": 5.428571428571429e-06,
+      "loss": 0.2406,
+      "step": 30400
+    },
+    {
+      "epoch": 45.73,
+      "grad_norm": 6.470170497894287,
+      "learning_rate": 5.427067669172933e-06,
+      "loss": 0.2258,
+      "step": 30410
+    },
+    {
+      "epoch": 45.74,
+      "grad_norm": 5.55424165725708,
+      "learning_rate": 5.425563909774436e-06,
+      "loss": 0.2494,
+      "step": 30420
+    },
+    {
+      "epoch": 45.76,
+      "grad_norm": 7.8290910720825195,
+      "learning_rate": 5.42406015037594e-06,
+      "loss": 0.1809,
+      "step": 30430
+    },
+    {
+      "epoch": 45.77,
+      "grad_norm": 3.04001784324646,
+      "learning_rate": 5.422556390977444e-06,
+      "loss": 0.2146,
+      "step": 30440
+    },
+    {
+      "epoch": 45.79,
+      "grad_norm": 4.655550479888916,
+      "learning_rate": 5.421052631578948e-06,
+      "loss": 0.2232,
+      "step": 30450
+    },
+    {
+      "epoch": 45.8,
+      "grad_norm": 5.950998306274414,
+      "learning_rate": 5.419548872180451e-06,
+      "loss": 0.275,
+      "step": 30460
+    },
+    {
+      "epoch": 45.82,
+      "grad_norm": 6.617697238922119,
+      "learning_rate": 5.418045112781955e-06,
+      "loss": 0.1975,
+      "step": 30470
+    },
+    {
+      "epoch": 45.83,
+      "grad_norm": 4.946775913238525,
+      "learning_rate": 5.416541353383459e-06,
+      "loss": 0.2631,
+      "step": 30480
+    },
+    {
+      "epoch": 45.85,
+      "grad_norm": 8.360345840454102,
+      "learning_rate": 5.415037593984963e-06,
+      "loss": 0.2695,
+      "step": 30490
+    },
+    {
+      "epoch": 45.86,
+      "grad_norm": 5.876543998718262,
+      "learning_rate": 5.413533834586467e-06,
+      "loss": 0.253,
+      "step": 30500
+    },
+    {
+      "epoch": 45.88,
+      "grad_norm": 5.3216047286987305,
+      "learning_rate": 5.412030075187971e-06,
+      "loss": 0.2594,
+      "step": 30510
+    },
+    {
+      "epoch": 45.89,
+      "grad_norm": 3.109912633895874,
+      "learning_rate": 5.410526315789474e-06,
+      "loss": 0.2005,
+      "step": 30520
+    },
+    {
+      "epoch": 45.91,
+      "grad_norm": 16.188390731811523,
+      "learning_rate": 5.409022556390978e-06,
+      "loss": 0.1955,
+      "step": 30530
+    },
+    {
+      "epoch": 45.92,
+      "grad_norm": 4.817110061645508,
+      "learning_rate": 5.4075187969924815e-06,
+      "loss": 0.2133,
+      "step": 30540
+    },
+    {
+      "epoch": 45.94,
+      "grad_norm": 6.045027732849121,
+      "learning_rate": 5.406015037593986e-06,
+      "loss": 0.2392,
+      "step": 30550
+    },
+    {
+      "epoch": 45.95,
+      "grad_norm": 3.955580234527588,
+      "learning_rate": 5.4045112781954886e-06,
+      "loss": 0.224,
+      "step": 30560
+    },
+    {
+      "epoch": 45.97,
+      "grad_norm": 2.896059036254883,
+      "learning_rate": 5.403007518796993e-06,
+      "loss": 0.2249,
+      "step": 30570
+    },
+    {
+      "epoch": 45.98,
+      "grad_norm": 5.959082126617432,
+      "learning_rate": 5.4015037593984964e-06,
+      "loss": 0.2038,
+      "step": 30580
+    },
+    {
+      "epoch": 46.0,
+      "grad_norm": 23.55235481262207,
+      "learning_rate": 5.400000000000001e-06,
+      "loss": 0.2059,
+      "step": 30590
+    },
+    {
+      "epoch": 46.0,
+      "eval_accuracy": 0.931,
+      "eval_loss": 0.30095645785331726,
+      "eval_runtime": 84.8003,
+      "eval_samples_per_second": 117.924,
+      "eval_steps_per_second": 0.472,
+      "step": 30590
+    },
+    {
+      "epoch": 46.02,
+      "grad_norm": 14.315238952636719,
+      "learning_rate": 5.398496240601504e-06,
+      "loss": 0.2133,
+      "step": 30600
+    },
+    {
+      "epoch": 46.03,
+      "grad_norm": 3.028569221496582,
+      "learning_rate": 5.396992481203009e-06,
+      "loss": 0.2082,
+      "step": 30610
+    },
+    {
+      "epoch": 46.05,
+      "grad_norm": 4.44739294052124,
+      "learning_rate": 5.395488721804511e-06,
+      "loss": 0.3014,
+      "step": 30620
+    },
+    {
+      "epoch": 46.06,
+      "grad_norm": 5.786952018737793,
+      "learning_rate": 5.393984962406016e-06,
+      "loss": 0.196,
+      "step": 30630
+    },
+    {
+      "epoch": 46.08,
+      "grad_norm": 6.879361152648926,
+      "learning_rate": 5.392481203007519e-06,
+      "loss": 0.2238,
+      "step": 30640
+    },
+    {
+      "epoch": 46.09,
+      "grad_norm": 2.7093942165374756,
+      "learning_rate": 5.390977443609024e-06,
+      "loss": 0.236,
+      "step": 30650
+    },
+    {
+      "epoch": 46.11,
+      "grad_norm": 4.412895679473877,
+      "learning_rate": 5.389473684210526e-06,
+      "loss": 0.214,
+      "step": 30660
+    },
+    {
+      "epoch": 46.12,
+      "grad_norm": 2.3530495166778564,
+      "learning_rate": 5.387969924812031e-06,
+      "loss": 0.2996,
+      "step": 30670
+    },
+    {
+      "epoch": 46.14,
+      "grad_norm": 5.1143107414245605,
+      "learning_rate": 5.386466165413534e-06,
+      "loss": 0.2465,
+      "step": 30680
+    },
+    {
+      "epoch": 46.15,
+      "grad_norm": 5.444381237030029,
+      "learning_rate": 5.3849624060150385e-06,
+      "loss": 0.2921,
+      "step": 30690
+    },
+    {
+      "epoch": 46.17,
+      "grad_norm": 6.554684162139893,
+      "learning_rate": 5.383458646616542e-06,
+      "loss": 0.2532,
+      "step": 30700
+    },
+    {
+      "epoch": 46.18,
+      "grad_norm": 6.188846588134766,
+      "learning_rate": 5.3819548872180464e-06,
+      "loss": 0.1924,
+      "step": 30710
+    },
+    {
+      "epoch": 46.2,
+      "grad_norm": 4.2928361892700195,
+      "learning_rate": 5.380451127819549e-06,
+      "loss": 0.2115,
+      "step": 30720
+    },
+    {
+      "epoch": 46.21,
+      "grad_norm": 1.6907163858413696,
+      "learning_rate": 5.3789473684210535e-06,
+      "loss": 0.3012,
+      "step": 30730
+    },
+    {
+      "epoch": 46.23,
+      "grad_norm": 8.911703109741211,
+      "learning_rate": 5.377443609022557e-06,
+      "loss": 0.2386,
+      "step": 30740
+    },
+    {
+      "epoch": 46.24,
+      "grad_norm": 12.857329368591309,
+      "learning_rate": 5.375939849624061e-06,
+      "loss": 0.2622,
+      "step": 30750
+    },
+    {
+      "epoch": 46.26,
+      "grad_norm": 4.299805641174316,
+      "learning_rate": 5.374436090225564e-06,
+      "loss": 0.2495,
+      "step": 30760
+    },
+    {
+      "epoch": 46.27,
+      "grad_norm": 4.013417720794678,
+      "learning_rate": 5.3729323308270675e-06,
+      "loss": 0.2414,
+      "step": 30770
+    },
+    {
+      "epoch": 46.29,
+      "grad_norm": 5.057952404022217,
+      "learning_rate": 5.371428571428572e-06,
+      "loss": 0.3043,
+      "step": 30780
+    },
+    {
+      "epoch": 46.3,
+      "grad_norm": 2.708482265472412,
+      "learning_rate": 5.3699248120300754e-06,
+      "loss": 0.2512,
+      "step": 30790
+    },
+    {
+      "epoch": 46.32,
+      "grad_norm": 4.2768988609313965,
+      "learning_rate": 5.36842105263158e-06,
+      "loss": 0.2791,
+      "step": 30800
+    },
+    {
+      "epoch": 46.33,
+      "grad_norm": 10.092656135559082,
+      "learning_rate": 5.3669172932330825e-06,
+      "loss": 0.2439,
+      "step": 30810
+    },
+    {
+      "epoch": 46.35,
+      "grad_norm": 7.980869293212891,
+      "learning_rate": 5.365413533834587e-06,
+      "loss": 0.2967,
+      "step": 30820
+    },
+    {
+      "epoch": 46.36,
+      "grad_norm": 6.0706000328063965,
+      "learning_rate": 5.36390977443609e-06,
+      "loss": 0.2448,
+      "step": 30830
+    },
+    {
+      "epoch": 46.38,
+      "grad_norm": 5.074240207672119,
+      "learning_rate": 5.362406015037595e-06,
+      "loss": 0.2367,
+      "step": 30840
+    },
+    {
+      "epoch": 46.39,
+      "grad_norm": 4.816855430603027,
+      "learning_rate": 5.360902255639097e-06,
+      "loss": 0.2318,
+      "step": 30850
+    },
+    {
+      "epoch": 46.41,
+      "grad_norm": 3.8746345043182373,
+      "learning_rate": 5.359398496240602e-06,
+      "loss": 0.2336,
+      "step": 30860
+    },
+    {
+      "epoch": 46.42,
+      "grad_norm": 5.22980260848999,
+      "learning_rate": 5.357894736842105e-06,
+      "loss": 0.2148,
+      "step": 30870
+    },
+    {
+      "epoch": 46.44,
+      "grad_norm": 6.546250343322754,
+      "learning_rate": 5.35639097744361e-06,
+      "loss": 0.2513,
+      "step": 30880
+    },
+    {
+      "epoch": 46.45,
+      "grad_norm": 3.120495319366455,
+      "learning_rate": 5.354887218045113e-06,
+      "loss": 0.2673,
+      "step": 30890
+    },
+    {
+      "epoch": 46.47,
+      "grad_norm": 4.755849838256836,
+      "learning_rate": 5.3533834586466175e-06,
+      "loss": 0.2706,
+      "step": 30900
+    },
+    {
+      "epoch": 46.48,
+      "grad_norm": 3.598883628845215,
+      "learning_rate": 5.35187969924812e-06,
+      "loss": 0.1639,
+      "step": 30910
+    },
+    {
+      "epoch": 46.5,
+      "grad_norm": 4.807301044464111,
+      "learning_rate": 5.3503759398496246e-06,
+      "loss": 0.2063,
+      "step": 30920
+    },
+    {
+      "epoch": 46.51,
+      "grad_norm": 4.723282337188721,
+      "learning_rate": 5.348872180451128e-06,
+      "loss": 0.1954,
+      "step": 30930
+    },
+    {
+      "epoch": 46.53,
+      "grad_norm": 6.898508071899414,
+      "learning_rate": 5.3473684210526325e-06,
+      "loss": 0.274,
+      "step": 30940
+    },
+    {
+      "epoch": 46.54,
+      "grad_norm": 1.9944006204605103,
+      "learning_rate": 5.345864661654135e-06,
+      "loss": 0.2548,
+      "step": 30950
+    },
+    {
+      "epoch": 46.56,
+      "grad_norm": 4.713189125061035,
+      "learning_rate": 5.3443609022556395e-06,
+      "loss": 0.2342,
+      "step": 30960
+    },
+    {
+      "epoch": 46.57,
+      "grad_norm": 5.438635349273682,
+      "learning_rate": 5.342857142857143e-06,
+      "loss": 0.2282,
+      "step": 30970
+    },
+    {
+      "epoch": 46.59,
+      "grad_norm": 4.57274055480957,
+      "learning_rate": 5.341353383458647e-06,
+      "loss": 0.2339,
+      "step": 30980
+    },
+    {
+      "epoch": 46.6,
+      "grad_norm": 6.225501537322998,
+      "learning_rate": 5.339849624060151e-06,
+      "loss": 0.2293,
+      "step": 30990
+    },
+    {
+      "epoch": 46.62,
+      "grad_norm": 10.704837799072266,
+      "learning_rate": 5.338345864661654e-06,
+      "loss": 0.2327,
+      "step": 31000
+    },
+    {
+      "epoch": 46.63,
+      "grad_norm": 7.582201957702637,
+      "learning_rate": 5.336842105263158e-06,
+      "loss": 0.2563,
+      "step": 31010
+    },
+    {
+      "epoch": 46.65,
+      "grad_norm": 6.811306476593018,
+      "learning_rate": 5.335338345864662e-06,
+      "loss": 0.198,
+      "step": 31020
+    },
+    {
+      "epoch": 46.66,
+      "grad_norm": 3.75140118598938,
+      "learning_rate": 5.333834586466166e-06,
+      "loss": 0.2295,
+      "step": 31030
+    },
+    {
+      "epoch": 46.68,
+      "grad_norm": 7.410831928253174,
+      "learning_rate": 5.33233082706767e-06,
+      "loss": 0.2624,
+      "step": 31040
+    },
+    {
+      "epoch": 46.69,
+      "grad_norm": 4.584934711456299,
+      "learning_rate": 5.330827067669173e-06,
+      "loss": 0.2611,
+      "step": 31050
+    },
+    {
+      "epoch": 46.71,
+      "grad_norm": 2.435068130493164,
+      "learning_rate": 5.329323308270677e-06,
+      "loss": 0.2045,
+      "step": 31060
+    },
+    {
+      "epoch": 46.72,
+      "grad_norm": 6.5712056159973145,
+      "learning_rate": 5.327819548872181e-06,
+      "loss": 0.1974,
+      "step": 31070
+    },
+    {
+      "epoch": 46.74,
+      "grad_norm": 6.2264862060546875,
+      "learning_rate": 5.326315789473685e-06,
+      "loss": 0.3578,
+      "step": 31080
+    },
+    {
+      "epoch": 46.75,
+      "grad_norm": 4.949437141418457,
+      "learning_rate": 5.324812030075189e-06,
+      "loss": 0.2336,
+      "step": 31090
+    },
+    {
+      "epoch": 46.77,
+      "grad_norm": 8.773530006408691,
+      "learning_rate": 5.323308270676692e-06,
+      "loss": 0.2793,
+      "step": 31100
+    },
+    {
+      "epoch": 46.78,
+      "grad_norm": 4.285604000091553,
+      "learning_rate": 5.321804511278196e-06,
+      "loss": 0.2367,
+      "step": 31110
+    },
+    {
+      "epoch": 46.8,
+      "grad_norm": 5.313040256500244,
+      "learning_rate": 5.3203007518797e-06,
+      "loss": 0.2804,
+      "step": 31120
+    },
+    {
+      "epoch": 46.81,
+      "grad_norm": 4.113051414489746,
+      "learning_rate": 5.3187969924812035e-06,
+      "loss": 0.2158,
+      "step": 31130
+    },
+    {
+      "epoch": 46.83,
+      "grad_norm": 3.403592824935913,
+      "learning_rate": 5.317293233082708e-06,
+      "loss": 0.2477,
+      "step": 31140
+    },
+    {
+      "epoch": 46.84,
+      "grad_norm": 9.54311466217041,
+      "learning_rate": 5.315789473684211e-06,
+      "loss": 0.2816,
+      "step": 31150
+    },
+    {
+      "epoch": 46.86,
+      "grad_norm": 11.368732452392578,
+      "learning_rate": 5.314285714285715e-06,
+      "loss": 0.2512,
+      "step": 31160
+    },
+    {
+      "epoch": 46.87,
+      "grad_norm": 8.554758071899414,
+      "learning_rate": 5.3127819548872185e-06,
+      "loss": 0.2527,
+      "step": 31170
+    },
+    {
+      "epoch": 46.89,
+      "grad_norm": 4.945856094360352,
+      "learning_rate": 5.311278195488723e-06,
+      "loss": 0.2287,
+      "step": 31180
+    },
+    {
+      "epoch": 46.9,
+      "grad_norm": 5.310131072998047,
+      "learning_rate": 5.3097744360902255e-06,
+      "loss": 0.2594,
+      "step": 31190
+    },
+    {
+      "epoch": 46.92,
+      "grad_norm": 5.681679725646973,
+      "learning_rate": 5.30827067669173e-06,
+      "loss": 0.2152,
+      "step": 31200
+    },
+    {
+      "epoch": 46.93,
+      "grad_norm": 7.65255069732666,
+      "learning_rate": 5.306766917293233e-06,
+      "loss": 0.2017,
+      "step": 31210
+    },
+    {
+      "epoch": 46.95,
+      "grad_norm": 7.68698787689209,
+      "learning_rate": 5.305263157894738e-06,
+      "loss": 0.2355,
+      "step": 31220
+    },
+    {
+      "epoch": 46.96,
+      "grad_norm": 6.632236003875732,
+      "learning_rate": 5.303759398496241e-06,
+      "loss": 0.2626,
+      "step": 31230
+    },
+    {
+      "epoch": 46.98,
+      "grad_norm": 3.1763651371002197,
+      "learning_rate": 5.302255639097746e-06,
+      "loss": 0.2013,
+      "step": 31240
+    },
+    {
+      "epoch": 46.99,
+      "grad_norm": 5.487941741943359,
+      "learning_rate": 5.300751879699248e-06,
+      "loss": 0.2156,
+      "step": 31250
+    },
+    {
+      "epoch": 47.0,
+      "eval_accuracy": 0.9318,
+      "eval_loss": 0.2919594645500183,
+      "eval_runtime": 84.7462,
+      "eval_samples_per_second": 117.999,
+      "eval_steps_per_second": 0.472,
+      "step": 31255
+    },
+    {
+      "epoch": 47.01,
+      "grad_norm": 4.854064464569092,
+      "learning_rate": 5.299248120300753e-06,
+      "loss": 0.2527,
+      "step": 31260
+    },
+    {
+      "epoch": 47.02,
+      "grad_norm": 6.271472454071045,
+      "learning_rate": 5.297744360902256e-06,
+      "loss": 0.191,
+      "step": 31270
+    },
+    {
+      "epoch": 47.04,
+      "grad_norm": 6.080266952514648,
+      "learning_rate": 5.296240601503759e-06,
+      "loss": 0.2479,
+      "step": 31280
+    },
+    {
+      "epoch": 47.05,
+      "grad_norm": 7.085132122039795,
+      "learning_rate": 5.294736842105263e-06,
+      "loss": 0.2624,
+      "step": 31290
+    },
+    {
+      "epoch": 47.07,
+      "grad_norm": 4.636129856109619,
+      "learning_rate": 5.293233082706767e-06,
+      "loss": 0.1893,
+      "step": 31300
+    },
+    {
+      "epoch": 47.08,
+      "grad_norm": 6.469307899475098,
+      "learning_rate": 5.291729323308271e-06,
+      "loss": 0.1865,
+      "step": 31310
+    },
+    {
+      "epoch": 47.1,
+      "grad_norm": 8.775065422058105,
+      "learning_rate": 5.290225563909775e-06,
+      "loss": 0.2459,
+      "step": 31320
+    },
+    {
+      "epoch": 47.11,
+      "grad_norm": 3.9796481132507324,
+      "learning_rate": 5.288721804511279e-06,
+      "loss": 0.257,
+      "step": 31330
+    },
+    {
+      "epoch": 47.13,
+      "grad_norm": 5.34701681137085,
+      "learning_rate": 5.287218045112782e-06,
+      "loss": 0.199,
+      "step": 31340
+    },
+    {
+      "epoch": 47.14,
+      "grad_norm": 6.631876468658447,
+      "learning_rate": 5.285714285714286e-06,
+      "loss": 0.2502,
+      "step": 31350
+    },
+    {
+      "epoch": 47.16,
+      "grad_norm": 4.332927227020264,
+      "learning_rate": 5.2842105263157896e-06,
+      "loss": 0.2145,
+      "step": 31360
+    },
+    {
+      "epoch": 47.17,
+      "grad_norm": 7.87563419342041,
+      "learning_rate": 5.282706766917294e-06,
+      "loss": 0.2484,
+      "step": 31370
+    },
+    {
+      "epoch": 47.19,
+      "grad_norm": 4.990164279937744,
+      "learning_rate": 5.281203007518797e-06,
+      "loss": 0.2739,
+      "step": 31380
+    },
+    {
+      "epoch": 47.2,
+      "grad_norm": 5.960522174835205,
+      "learning_rate": 5.279699248120301e-06,
+      "loss": 0.2388,
+      "step": 31390
+    },
+    {
+      "epoch": 47.22,
+      "grad_norm": 8.016762733459473,
+      "learning_rate": 5.2781954887218045e-06,
+      "loss": 0.2013,
+      "step": 31400
+    },
+    {
+      "epoch": 47.23,
+      "grad_norm": 4.126075744628906,
+      "learning_rate": 5.276691729323309e-06,
+      "loss": 0.2994,
+      "step": 31410
+    },
+    {
+      "epoch": 47.25,
+      "grad_norm": 4.315171241760254,
+      "learning_rate": 5.275187969924812e-06,
+      "loss": 0.3082,
+      "step": 31420
+    },
+    {
+      "epoch": 47.26,
+      "grad_norm": 6.842277526855469,
+      "learning_rate": 5.273684210526317e-06,
+      "loss": 0.2019,
+      "step": 31430
+    },
+    {
+      "epoch": 47.28,
+      "grad_norm": 16.402454376220703,
+      "learning_rate": 5.272180451127819e-06,
+      "loss": 0.3009,
+      "step": 31440
+    },
+    {
+      "epoch": 47.29,
+      "grad_norm": 4.734708309173584,
+      "learning_rate": 5.270676691729324e-06,
+      "loss": 0.1851,
+      "step": 31450
+    },
+    {
+      "epoch": 47.31,
+      "grad_norm": 7.005868434906006,
+      "learning_rate": 5.269172932330827e-06,
+      "loss": 0.2021,
+      "step": 31460
+    },
+    {
+      "epoch": 47.32,
+      "grad_norm": 4.445467948913574,
+      "learning_rate": 5.267669172932332e-06,
+      "loss": 0.2028,
+      "step": 31470
+    },
+    {
+      "epoch": 47.34,
+      "grad_norm": 3.294844627380371,
+      "learning_rate": 5.266165413533834e-06,
+      "loss": 0.2901,
+      "step": 31480
+    },
+    {
+      "epoch": 47.35,
+      "grad_norm": 5.847194194793701,
+      "learning_rate": 5.264661654135339e-06,
+      "loss": 0.2215,
+      "step": 31490
+    },
+    {
+      "epoch": 47.37,
+      "grad_norm": 3.2134320735931396,
+      "learning_rate": 5.263157894736842e-06,
+      "loss": 0.213,
+      "step": 31500
+    },
+    {
+      "epoch": 47.38,
+      "grad_norm": 7.5339035987854,
+      "learning_rate": 5.261654135338347e-06,
+      "loss": 0.2378,
+      "step": 31510
+    },
+    {
+      "epoch": 47.4,
+      "grad_norm": 3.7223424911499023,
+      "learning_rate": 5.26015037593985e-06,
+      "loss": 0.3221,
+      "step": 31520
+    },
+    {
+      "epoch": 47.41,
+      "grad_norm": 4.247413158416748,
+      "learning_rate": 5.2586466165413545e-06,
+      "loss": 0.274,
+      "step": 31530
+    },
+    {
+      "epoch": 47.43,
+      "grad_norm": 4.432199478149414,
+      "learning_rate": 5.257142857142857e-06,
+      "loss": 0.2672,
+      "step": 31540
+    },
+    {
+      "epoch": 47.44,
+      "grad_norm": 4.548630714416504,
+      "learning_rate": 5.2556390977443615e-06,
+      "loss": 0.2413,
+      "step": 31550
+    },
+    {
+      "epoch": 47.46,
+      "grad_norm": 5.084230899810791,
+      "learning_rate": 5.254135338345865e-06,
+      "loss": 0.233,
+      "step": 31560
+    },
+    {
+      "epoch": 47.47,
+      "grad_norm": 2.7033839225769043,
+      "learning_rate": 5.252631578947369e-06,
+      "loss": 0.2344,
+      "step": 31570
+    },
+    {
+      "epoch": 47.49,
+      "grad_norm": 6.172457218170166,
+      "learning_rate": 5.251127819548872e-06,
+      "loss": 0.2259,
+      "step": 31580
+    },
+    {
+      "epoch": 47.5,
+      "grad_norm": 4.346304893493652,
+      "learning_rate": 5.2496240601503764e-06,
+      "loss": 0.2755,
+      "step": 31590
+    },
+    {
+      "epoch": 47.52,
+      "grad_norm": 5.721127986907959,
+      "learning_rate": 5.24812030075188e-06,
+      "loss": 0.2132,
+      "step": 31600
+    },
+    {
+      "epoch": 47.53,
+      "grad_norm": 4.425881862640381,
+      "learning_rate": 5.246616541353384e-06,
+      "loss": 0.249,
+      "step": 31610
+    },
+    {
+      "epoch": 47.55,
+      "grad_norm": 5.007637977600098,
+      "learning_rate": 5.245112781954888e-06,
+      "loss": 0.2291,
+      "step": 31620
+    },
+    {
+      "epoch": 47.56,
+      "grad_norm": 5.64668083190918,
+      "learning_rate": 5.243609022556392e-06,
+      "loss": 0.2487,
+      "step": 31630
+    },
+    {
+      "epoch": 47.58,
+      "grad_norm": 4.099886417388916,
+      "learning_rate": 5.242105263157895e-06,
+      "loss": 0.2259,
+      "step": 31640
+    },
+    {
+      "epoch": 47.59,
+      "grad_norm": 4.3517584800720215,
+      "learning_rate": 5.240601503759399e-06,
+      "loss": 0.1817,
+      "step": 31650
+    },
+    {
+      "epoch": 47.61,
+      "grad_norm": 7.435219764709473,
+      "learning_rate": 5.239097744360903e-06,
+      "loss": 0.2105,
+      "step": 31660
+    },
+    {
+      "epoch": 47.62,
+      "grad_norm": 4.386735439300537,
+      "learning_rate": 5.237593984962407e-06,
+      "loss": 0.2167,
+      "step": 31670
+    },
+    {
+      "epoch": 47.64,
+      "grad_norm": 9.977165222167969,
+      "learning_rate": 5.23609022556391e-06,
+      "loss": 0.2578,
+      "step": 31680
+    },
+    {
+      "epoch": 47.65,
+      "grad_norm": 5.100271701812744,
+      "learning_rate": 5.234586466165414e-06,
+      "loss": 0.2037,
+      "step": 31690
+    },
+    {
+      "epoch": 47.67,
+      "grad_norm": 12.209273338317871,
+      "learning_rate": 5.233082706766918e-06,
+      "loss": 0.2373,
+      "step": 31700
+    },
+    {
+      "epoch": 47.68,
+      "grad_norm": 4.825228691101074,
+      "learning_rate": 5.231578947368422e-06,
+      "loss": 0.217,
+      "step": 31710
+    },
+    {
+      "epoch": 47.7,
+      "grad_norm": 4.916600227355957,
+      "learning_rate": 5.2300751879699256e-06,
+      "loss": 0.2243,
+      "step": 31720
+    },
+    {
+      "epoch": 47.71,
+      "grad_norm": 4.243865966796875,
+      "learning_rate": 5.22857142857143e-06,
+      "loss": 0.198,
+      "step": 31730
+    },
+    {
+      "epoch": 47.73,
+      "grad_norm": 4.085118293762207,
+      "learning_rate": 5.227067669172933e-06,
+      "loss": 0.2393,
+      "step": 31740
+    },
+    {
+      "epoch": 47.74,
+      "grad_norm": 3.169874906539917,
+      "learning_rate": 5.225563909774437e-06,
+      "loss": 0.227,
+      "step": 31750
+    },
+    {
+      "epoch": 47.76,
+      "grad_norm": 5.03513240814209,
+      "learning_rate": 5.2240601503759405e-06,
+      "loss": 0.1579,
+      "step": 31760
+    },
+    {
+      "epoch": 47.77,
+      "grad_norm": 4.764927864074707,
+      "learning_rate": 5.222556390977443e-06,
+      "loss": 0.2179,
+      "step": 31770
+    },
+    {
+      "epoch": 47.79,
+      "grad_norm": 3.114337205886841,
+      "learning_rate": 5.2210526315789475e-06,
+      "loss": 0.1934,
+      "step": 31780
+    },
+    {
+      "epoch": 47.8,
+      "grad_norm": 4.358081817626953,
+      "learning_rate": 5.219548872180451e-06,
+      "loss": 0.1997,
+      "step": 31790
+    },
+    {
+      "epoch": 47.82,
+      "grad_norm": 4.839754104614258,
+      "learning_rate": 5.218045112781955e-06,
+      "loss": 0.2273,
+      "step": 31800
+    },
+    {
+      "epoch": 47.83,
+      "grad_norm": 4.898540496826172,
+      "learning_rate": 5.216541353383459e-06,
+      "loss": 0.2084,
+      "step": 31810
+    },
+    {
+      "epoch": 47.85,
+      "grad_norm": 7.094751834869385,
+      "learning_rate": 5.215037593984963e-06,
+      "loss": 0.2072,
+      "step": 31820
+    },
+    {
+      "epoch": 47.86,
+      "grad_norm": 6.301358699798584,
+      "learning_rate": 5.213533834586466e-06,
+      "loss": 0.2076,
+      "step": 31830
+    },
+    {
+      "epoch": 47.88,
+      "grad_norm": 5.953322410583496,
+      "learning_rate": 5.21203007518797e-06,
+      "loss": 0.234,
+      "step": 31840
+    },
+    {
+      "epoch": 47.89,
+      "grad_norm": 4.481212615966797,
+      "learning_rate": 5.210526315789474e-06,
+      "loss": 0.2074,
+      "step": 31850
+    },
+    {
+      "epoch": 47.91,
+      "grad_norm": 7.783965587615967,
+      "learning_rate": 5.209022556390978e-06,
+      "loss": 0.2597,
+      "step": 31860
+    },
+    {
+      "epoch": 47.92,
+      "grad_norm": 10.910694122314453,
+      "learning_rate": 5.207518796992481e-06,
+      "loss": 0.2851,
+      "step": 31870
+    },
+    {
+      "epoch": 47.94,
+      "grad_norm": 1.9801486730575562,
+      "learning_rate": 5.206015037593985e-06,
+      "loss": 0.2154,
+      "step": 31880
+    },
+    {
+      "epoch": 47.95,
+      "grad_norm": 11.243240356445312,
+      "learning_rate": 5.204511278195489e-06,
+      "loss": 0.2158,
+      "step": 31890
+    },
+    {
+      "epoch": 47.97,
+      "grad_norm": 2.5445377826690674,
+      "learning_rate": 5.203007518796993e-06,
+      "loss": 0.2497,
+      "step": 31900
+    },
+    {
+      "epoch": 47.98,
+      "grad_norm": 5.46054744720459,
+      "learning_rate": 5.201503759398497e-06,
+      "loss": 0.226,
+      "step": 31910
+    },
+    {
+      "epoch": 48.0,
+      "grad_norm": 26.797359466552734,
+      "learning_rate": 5.2e-06,
+      "loss": 0.2719,
+      "step": 31920
+    },
+    {
+      "epoch": 48.0,
+      "eval_accuracy": 0.9311,
+      "eval_loss": 0.3056710362434387,
+      "eval_runtime": 84.8487,
+      "eval_samples_per_second": 117.857,
+      "eval_steps_per_second": 0.471,
+      "step": 31920
+    },
+    {
+      "epoch": 48.02,
+      "grad_norm": 3.771094799041748,
+      "learning_rate": 5.198496240601504e-06,
+      "loss": 0.197,
+      "step": 31930
+    },
+    {
+      "epoch": 48.03,
+      "grad_norm": 3.460973024368286,
+      "learning_rate": 5.196992481203008e-06,
+      "loss": 0.2012,
+      "step": 31940
+    },
+    {
+      "epoch": 48.05,
+      "grad_norm": 7.17149543762207,
+      "learning_rate": 5.195488721804512e-06,
+      "loss": 0.2799,
+      "step": 31950
+    },
+    {
+      "epoch": 48.06,
+      "grad_norm": 2.8762929439544678,
+      "learning_rate": 5.193984962406016e-06,
+      "loss": 0.2632,
+      "step": 31960
+    },
+    {
+      "epoch": 48.08,
+      "grad_norm": 5.614086627960205,
+      "learning_rate": 5.192481203007519e-06,
+      "loss": 0.205,
+      "step": 31970
+    },
+    {
+      "epoch": 48.09,
+      "grad_norm": 5.698775768280029,
+      "learning_rate": 5.190977443609023e-06,
+      "loss": 0.2031,
+      "step": 31980
+    },
+    {
+      "epoch": 48.11,
+      "grad_norm": 3.4972689151763916,
+      "learning_rate": 5.1894736842105265e-06,
+      "loss": 0.1935,
+      "step": 31990
+    },
+    {
+      "epoch": 48.12,
+      "grad_norm": 6.399228096008301,
+      "learning_rate": 5.187969924812031e-06,
+      "loss": 0.201,
+      "step": 32000
+    },
+    {
+      "epoch": 48.14,
+      "grad_norm": 6.805654048919678,
+      "learning_rate": 5.186466165413534e-06,
+      "loss": 0.1714,
+      "step": 32010
+    },
+    {
+      "epoch": 48.15,
+      "grad_norm": 4.721649646759033,
+      "learning_rate": 5.184962406015038e-06,
+      "loss": 0.2139,
+      "step": 32020
+    },
+    {
+      "epoch": 48.17,
+      "grad_norm": 5.464878082275391,
+      "learning_rate": 5.1834586466165414e-06,
+      "loss": 0.2137,
+      "step": 32030
+    },
+    {
+      "epoch": 48.18,
+      "grad_norm": 5.358616352081299,
+      "learning_rate": 5.181954887218046e-06,
+      "loss": 0.2196,
+      "step": 32040
+    },
+    {
+      "epoch": 48.2,
+      "grad_norm": 8.136012077331543,
+      "learning_rate": 5.180451127819549e-06,
+      "loss": 0.2861,
+      "step": 32050
+    },
+    {
+      "epoch": 48.21,
+      "grad_norm": 10.187384605407715,
+      "learning_rate": 5.178947368421054e-06,
+      "loss": 0.2223,
+      "step": 32060
+    },
+    {
+      "epoch": 48.23,
+      "grad_norm": 4.251842498779297,
+      "learning_rate": 5.177443609022556e-06,
+      "loss": 0.1949,
+      "step": 32070
+    },
+    {
+      "epoch": 48.24,
+      "grad_norm": 6.019373893737793,
+      "learning_rate": 5.175939849624061e-06,
+      "loss": 0.2512,
+      "step": 32080
+    },
+    {
+      "epoch": 48.26,
+      "grad_norm": 5.162561416625977,
+      "learning_rate": 5.174436090225564e-06,
+      "loss": 0.2408,
+      "step": 32090
+    },
+    {
+      "epoch": 48.27,
+      "grad_norm": 4.33914852142334,
+      "learning_rate": 5.172932330827069e-06,
+      "loss": 0.2901,
+      "step": 32100
+    },
+    {
+      "epoch": 48.29,
+      "grad_norm": 6.14069938659668,
+      "learning_rate": 5.171428571428571e-06,
+      "loss": 0.1849,
+      "step": 32110
+    },
+    {
+      "epoch": 48.3,
+      "grad_norm": 4.674006462097168,
+      "learning_rate": 5.169924812030076e-06,
+      "loss": 0.2483,
+      "step": 32120
+    },
+    {
+      "epoch": 48.32,
+      "grad_norm": 9.517473220825195,
+      "learning_rate": 5.168421052631579e-06,
+      "loss": 0.2329,
+      "step": 32130
+    },
+    {
+      "epoch": 48.33,
+      "grad_norm": 3.6502087116241455,
+      "learning_rate": 5.1669172932330835e-06,
+      "loss": 0.2357,
+      "step": 32140
+    },
+    {
+      "epoch": 48.35,
+      "grad_norm": 2.8939104080200195,
+      "learning_rate": 5.165413533834587e-06,
+      "loss": 0.265,
+      "step": 32150
+    },
+    {
+      "epoch": 48.36,
+      "grad_norm": 4.541179656982422,
+      "learning_rate": 5.163909774436091e-06,
+      "loss": 0.2007,
+      "step": 32160
+    },
+    {
+      "epoch": 48.38,
+      "grad_norm": 3.5995399951934814,
+      "learning_rate": 5.162406015037594e-06,
+      "loss": 0.1895,
+      "step": 32170
+    },
+    {
+      "epoch": 48.39,
+      "grad_norm": 4.528938293457031,
+      "learning_rate": 5.1609022556390985e-06,
+      "loss": 0.2574,
+      "step": 32180
+    },
+    {
+      "epoch": 48.41,
+      "grad_norm": 3.6735799312591553,
+      "learning_rate": 5.159398496240602e-06,
+      "loss": 0.2564,
+      "step": 32190
+    },
+    {
+      "epoch": 48.42,
+      "grad_norm": 7.794544219970703,
+      "learning_rate": 5.157894736842106e-06,
+      "loss": 0.2534,
+      "step": 32200
+    },
+    {
+      "epoch": 48.44,
+      "grad_norm": 9.559613227844238,
+      "learning_rate": 5.156390977443609e-06,
+      "loss": 0.2308,
+      "step": 32210
+    },
+    {
+      "epoch": 48.45,
+      "grad_norm": 5.519898414611816,
+      "learning_rate": 5.154887218045113e-06,
+      "loss": 0.2299,
+      "step": 32220
+    },
+    {
+      "epoch": 48.47,
+      "grad_norm": 5.717340469360352,
+      "learning_rate": 5.153383458646617e-06,
+      "loss": 0.2285,
+      "step": 32230
+    },
+    {
+      "epoch": 48.48,
+      "grad_norm": 4.257745265960693,
+      "learning_rate": 5.151879699248121e-06,
+      "loss": 0.2257,
+      "step": 32240
+    },
+    {
+      "epoch": 48.5,
+      "grad_norm": 3.741241693496704,
+      "learning_rate": 5.150375939849625e-06,
+      "loss": 0.2006,
+      "step": 32250
+    },
+    {
+      "epoch": 48.51,
+      "grad_norm": 4.753995418548584,
+      "learning_rate": 5.1488721804511275e-06,
+      "loss": 0.2896,
+      "step": 32260
+    },
+    {
+      "epoch": 48.53,
+      "grad_norm": 4.3221635818481445,
+      "learning_rate": 5.147368421052632e-06,
+      "loss": 0.2416,
+      "step": 32270
+    },
+    {
+      "epoch": 48.54,
+      "grad_norm": 6.452643394470215,
+      "learning_rate": 5.145864661654135e-06,
+      "loss": 0.2546,
+      "step": 32280
+    },
+    {
+      "epoch": 48.56,
+      "grad_norm": 10.680849075317383,
+      "learning_rate": 5.14436090225564e-06,
+      "loss": 0.2578,
+      "step": 32290
+    },
+    {
+      "epoch": 48.57,
+      "grad_norm": 10.565488815307617,
+      "learning_rate": 5.142857142857142e-06,
+      "loss": 0.2105,
+      "step": 32300
+    },
+    {
+      "epoch": 48.59,
+      "grad_norm": 6.964345455169678,
+      "learning_rate": 5.141353383458647e-06,
+      "loss": 0.2407,
+      "step": 32310
+    },
+    {
+      "epoch": 48.6,
+      "grad_norm": 5.355147838592529,
+      "learning_rate": 5.13984962406015e-06,
+      "loss": 0.122,
+      "step": 32320
+    },
+    {
+      "epoch": 48.62,
+      "grad_norm": 4.7337565422058105,
+      "learning_rate": 5.138345864661655e-06,
+      "loss": 0.2202,
+      "step": 32330
+    },
+    {
+      "epoch": 48.63,
+      "grad_norm": 3.4737660884857178,
+      "learning_rate": 5.136842105263158e-06,
+      "loss": 0.2382,
+      "step": 32340
+    },
+    {
+      "epoch": 48.65,
+      "grad_norm": 3.8572425842285156,
+      "learning_rate": 5.1353383458646625e-06,
+      "loss": 0.2735,
+      "step": 32350
+    },
+    {
+      "epoch": 48.66,
+      "grad_norm": 3.9931530952453613,
+      "learning_rate": 5.133834586466165e-06,
+      "loss": 0.2399,
+      "step": 32360
+    },
+    {
+      "epoch": 48.68,
+      "grad_norm": 6.738966941833496,
+      "learning_rate": 5.1323308270676696e-06,
+      "loss": 0.2595,
+      "step": 32370
+    },
+    {
+      "epoch": 48.69,
+      "grad_norm": 5.455244064331055,
+      "learning_rate": 5.130827067669173e-06,
+      "loss": 0.1811,
+      "step": 32380
+    },
+    {
+      "epoch": 48.71,
+      "grad_norm": 8.760961532592773,
+      "learning_rate": 5.1293233082706774e-06,
+      "loss": 0.2323,
+      "step": 32390
+    },
+    {
+      "epoch": 48.72,
+      "grad_norm": 1.5300565958023071,
+      "learning_rate": 5.12781954887218e-06,
+      "loss": 0.2225,
+      "step": 32400
+    },
+    {
+      "epoch": 48.74,
+      "grad_norm": 4.378961563110352,
+      "learning_rate": 5.1263157894736845e-06,
+      "loss": 0.1771,
+      "step": 32410
+    },
+    {
+      "epoch": 48.75,
+      "grad_norm": 5.027568340301514,
+      "learning_rate": 5.124812030075188e-06,
+      "loss": 0.1993,
+      "step": 32420
+    },
+    {
+      "epoch": 48.77,
+      "grad_norm": 3.989525556564331,
+      "learning_rate": 5.123308270676692e-06,
+      "loss": 0.2182,
+      "step": 32430
+    },
+    {
+      "epoch": 48.78,
+      "grad_norm": 5.23897123336792,
+      "learning_rate": 5.121804511278196e-06,
+      "loss": 0.1689,
+      "step": 32440
+    },
+    {
+      "epoch": 48.8,
+      "grad_norm": 6.119510173797607,
+      "learning_rate": 5.1203007518797e-06,
+      "loss": 0.1778,
+      "step": 32450
+    },
+    {
+      "epoch": 48.81,
+      "grad_norm": 5.041784286499023,
+      "learning_rate": 5.118796992481203e-06,
+      "loss": 0.2215,
+      "step": 32460
+    },
+    {
+      "epoch": 48.83,
+      "grad_norm": 4.170535564422607,
+      "learning_rate": 5.117293233082707e-06,
+      "loss": 0.1809,
+      "step": 32470
+    },
+    {
+      "epoch": 48.84,
+      "grad_norm": 5.3539204597473145,
+      "learning_rate": 5.115789473684211e-06,
+      "loss": 0.2089,
+      "step": 32480
+    },
+    {
+      "epoch": 48.86,
+      "grad_norm": 3.4489662647247314,
+      "learning_rate": 5.114285714285715e-06,
+      "loss": 0.1827,
+      "step": 32490
+    },
+    {
+      "epoch": 48.87,
+      "grad_norm": 3.438225507736206,
+      "learning_rate": 5.112781954887218e-06,
+      "loss": 0.205,
+      "step": 32500
+    },
+    {
+      "epoch": 48.89,
+      "grad_norm": 5.150193214416504,
+      "learning_rate": 5.111278195488722e-06,
+      "loss": 0.195,
+      "step": 32510
+    },
+    {
+      "epoch": 48.9,
+      "grad_norm": 7.413297176361084,
+      "learning_rate": 5.109774436090226e-06,
+      "loss": 0.2422,
+      "step": 32520
+    },
+    {
+      "epoch": 48.92,
+      "grad_norm": 4.774962902069092,
+      "learning_rate": 5.10827067669173e-06,
+      "loss": 0.2325,
+      "step": 32530
+    },
+    {
+      "epoch": 48.93,
+      "grad_norm": 6.498291969299316,
+      "learning_rate": 5.106766917293234e-06,
+      "loss": 0.257,
+      "step": 32540
+    },
+    {
+      "epoch": 48.95,
+      "grad_norm": 5.5614728927612305,
+      "learning_rate": 5.105263157894738e-06,
+      "loss": 0.18,
+      "step": 32550
+    },
+    {
+      "epoch": 48.96,
+      "grad_norm": 9.62303352355957,
+      "learning_rate": 5.103759398496241e-06,
+      "loss": 0.2386,
+      "step": 32560
+    },
+    {
+      "epoch": 48.98,
+      "grad_norm": 7.92848539352417,
+      "learning_rate": 5.102255639097745e-06,
+      "loss": 0.2292,
+      "step": 32570
+    },
+    {
+      "epoch": 48.99,
+      "grad_norm": 4.110299587249756,
+      "learning_rate": 5.1007518796992485e-06,
+      "loss": 0.2156,
+      "step": 32580
+    },
+    {
+      "epoch": 49.0,
+      "eval_accuracy": 0.9292,
+      "eval_loss": 0.31269803643226624,
+      "eval_runtime": 84.4748,
+      "eval_samples_per_second": 118.379,
+      "eval_steps_per_second": 0.474,
+      "step": 32585
+    },
+    {
+      "epoch": 49.01,
+      "grad_norm": 7.004157543182373,
+      "learning_rate": 5.099248120300753e-06,
+      "loss": 0.3448,
+      "step": 32590
+    },
+    {
+      "epoch": 49.02,
+      "grad_norm": 4.223759651184082,
+      "learning_rate": 5.097744360902256e-06,
+      "loss": 0.2868,
+      "step": 32600
+    },
+    {
+      "epoch": 49.04,
+      "grad_norm": 3.9704689979553223,
+      "learning_rate": 5.09624060150376e-06,
+      "loss": 0.2549,
+      "step": 32610
+    },
+    {
+      "epoch": 49.05,
+      "grad_norm": 3.7155163288116455,
+      "learning_rate": 5.0947368421052635e-06,
+      "loss": 0.2238,
+      "step": 32620
+    },
+    {
+      "epoch": 49.07,
+      "grad_norm": 4.680662631988525,
+      "learning_rate": 5.093233082706768e-06,
+      "loss": 0.2062,
+      "step": 32630
+    },
+    {
+      "epoch": 49.08,
+      "grad_norm": 4.773167610168457,
+      "learning_rate": 5.091729323308271e-06,
+      "loss": 0.1666,
+      "step": 32640
+    },
+    {
+      "epoch": 49.1,
+      "grad_norm": 8.678242683410645,
+      "learning_rate": 5.090225563909776e-06,
+      "loss": 0.2808,
+      "step": 32650
+    },
+    {
+      "epoch": 49.11,
+      "grad_norm": 5.068692207336426,
+      "learning_rate": 5.088721804511278e-06,
+      "loss": 0.2623,
+      "step": 32660
+    },
+    {
+      "epoch": 49.13,
+      "grad_norm": 2.034926414489746,
+      "learning_rate": 5.087218045112783e-06,
+      "loss": 0.2039,
+      "step": 32670
+    },
+    {
+      "epoch": 49.14,
+      "grad_norm": 7.186866283416748,
+      "learning_rate": 5.085714285714286e-06,
+      "loss": 0.2584,
+      "step": 32680
+    },
+    {
+      "epoch": 49.16,
+      "grad_norm": 7.276577472686768,
+      "learning_rate": 5.084210526315791e-06,
+      "loss": 0.1756,
+      "step": 32690
+    },
+    {
+      "epoch": 49.17,
+      "grad_norm": 3.856623888015747,
+      "learning_rate": 5.082706766917293e-06,
+      "loss": 0.1908,
+      "step": 32700
+    },
+    {
+      "epoch": 49.19,
+      "grad_norm": 4.449421405792236,
+      "learning_rate": 5.081203007518798e-06,
+      "loss": 0.233,
+      "step": 32710
+    },
+    {
+      "epoch": 49.2,
+      "grad_norm": 7.532309055328369,
+      "learning_rate": 5.079699248120301e-06,
+      "loss": 0.2067,
+      "step": 32720
+    },
+    {
+      "epoch": 49.22,
+      "grad_norm": 5.170470237731934,
+      "learning_rate": 5.0781954887218056e-06,
+      "loss": 0.2273,
+      "step": 32730
+    },
+    {
+      "epoch": 49.23,
+      "grad_norm": 3.9134700298309326,
+      "learning_rate": 5.076691729323309e-06,
+      "loss": 0.2791,
+      "step": 32740
+    },
+    {
+      "epoch": 49.25,
+      "grad_norm": 6.696217060089111,
+      "learning_rate": 5.075187969924813e-06,
+      "loss": 0.224,
+      "step": 32750
+    },
+    {
+      "epoch": 49.26,
+      "grad_norm": 9.15458869934082,
+      "learning_rate": 5.073684210526316e-06,
+      "loss": 0.2591,
+      "step": 32760
+    },
+    {
+      "epoch": 49.28,
+      "grad_norm": 3.095736026763916,
+      "learning_rate": 5.07218045112782e-06,
+      "loss": 0.1681,
+      "step": 32770
+    },
+    {
+      "epoch": 49.29,
+      "grad_norm": 5.608436107635498,
+      "learning_rate": 5.070676691729324e-06,
+      "loss": 0.2269,
+      "step": 32780
+    },
+    {
+      "epoch": 49.31,
+      "grad_norm": 5.287787437438965,
+      "learning_rate": 5.069172932330827e-06,
+      "loss": 0.2443,
+      "step": 32790
+    },
+    {
+      "epoch": 49.32,
+      "grad_norm": 8.475576400756836,
+      "learning_rate": 5.067669172932331e-06,
+      "loss": 0.3255,
+      "step": 32800
+    },
+    {
+      "epoch": 49.34,
+      "grad_norm": 3.967184066772461,
+      "learning_rate": 5.0661654135338346e-06,
+      "loss": 0.1712,
+      "step": 32810
+    },
+    {
+      "epoch": 49.35,
+      "grad_norm": 7.337131023406982,
+      "learning_rate": 5.064661654135339e-06,
+      "loss": 0.3166,
+      "step": 32820
+    },
+    {
+      "epoch": 49.37,
+      "grad_norm": 6.623044967651367,
+      "learning_rate": 5.0631578947368424e-06,
+      "loss": 0.3035,
+      "step": 32830
+    },
+    {
+      "epoch": 49.38,
+      "grad_norm": 5.275474548339844,
+      "learning_rate": 5.061654135338346e-06,
+      "loss": 0.2025,
+      "step": 32840
+    },
+    {
+      "epoch": 49.4,
+      "grad_norm": 6.363266468048096,
+      "learning_rate": 5.0601503759398495e-06,
+      "loss": 0.3033,
+      "step": 32850
+    },
+    {
+      "epoch": 49.41,
+      "grad_norm": 3.8887858390808105,
+      "learning_rate": 5.058646616541354e-06,
+      "loss": 0.2282,
+      "step": 32860
+    },
+    {
+      "epoch": 49.43,
+      "grad_norm": 7.158453941345215,
+      "learning_rate": 5.057142857142857e-06,
+      "loss": 0.2427,
+      "step": 32870
+    },
+    {
+      "epoch": 49.44,
+      "grad_norm": 5.235705375671387,
+      "learning_rate": 5.055639097744362e-06,
+      "loss": 0.3072,
+      "step": 32880
+    },
+    {
+      "epoch": 49.46,
+      "grad_norm": 3.328618049621582,
+      "learning_rate": 5.054135338345864e-06,
+      "loss": 0.2078,
+      "step": 32890
+    },
+    {
+      "epoch": 49.47,
+      "grad_norm": 7.1164398193359375,
+      "learning_rate": 5.052631578947369e-06,
+      "loss": 0.2707,
+      "step": 32900
+    },
+    {
+      "epoch": 49.49,
+      "grad_norm": 4.149364948272705,
+      "learning_rate": 5.051127819548872e-06,
+      "loss": 0.1759,
+      "step": 32910
+    },
+    {
+      "epoch": 49.5,
+      "grad_norm": 5.859124183654785,
+      "learning_rate": 5.049624060150377e-06,
+      "loss": 0.2588,
+      "step": 32920
+    },
+    {
+      "epoch": 49.52,
+      "grad_norm": 7.663355350494385,
+      "learning_rate": 5.04812030075188e-06,
+      "loss": 0.285,
+      "step": 32930
+    },
+    {
+      "epoch": 49.53,
+      "grad_norm": 5.694937229156494,
+      "learning_rate": 5.046616541353384e-06,
+      "loss": 0.2546,
+      "step": 32940
+    },
+    {
+      "epoch": 49.55,
+      "grad_norm": 7.81641960144043,
+      "learning_rate": 5.045112781954887e-06,
+      "loss": 0.1886,
+      "step": 32950
+    },
+    {
+      "epoch": 49.56,
+      "grad_norm": 4.236837387084961,
+      "learning_rate": 5.043609022556392e-06,
+      "loss": 0.2051,
+      "step": 32960
+    },
+    {
+      "epoch": 49.58,
+      "grad_norm": 4.034359931945801,
+      "learning_rate": 5.042105263157895e-06,
+      "loss": 0.1713,
+      "step": 32970
+    },
+    {
+      "epoch": 49.59,
+      "grad_norm": 10.613497734069824,
+      "learning_rate": 5.0406015037593995e-06,
+      "loss": 0.2747,
+      "step": 32980
+    },
+    {
+      "epoch": 49.61,
+      "grad_norm": 6.468116760253906,
+      "learning_rate": 5.039097744360902e-06,
+      "loss": 0.2624,
+      "step": 32990
+    },
+    {
+      "epoch": 49.62,
+      "grad_norm": 6.871288776397705,
+      "learning_rate": 5.0375939849624065e-06,
+      "loss": 0.2418,
+      "step": 33000
+    },
+    {
+      "epoch": 49.64,
+      "grad_norm": 5.56874942779541,
+      "learning_rate": 5.03609022556391e-06,
+      "loss": 0.2581,
+      "step": 33010
+    },
+    {
+      "epoch": 49.65,
+      "grad_norm": 4.636134147644043,
+      "learning_rate": 5.034586466165414e-06,
+      "loss": 0.2183,
+      "step": 33020
+    },
+    {
+      "epoch": 49.67,
+      "grad_norm": 3.762443780899048,
+      "learning_rate": 5.033082706766917e-06,
+      "loss": 0.2067,
+      "step": 33030
+    },
+    {
+      "epoch": 49.68,
+      "grad_norm": 2.3653151988983154,
+      "learning_rate": 5.0315789473684214e-06,
+      "loss": 0.1524,
+      "step": 33040
+    },
+    {
+      "epoch": 49.7,
+      "grad_norm": 1.23353111743927,
+      "learning_rate": 5.030075187969925e-06,
+      "loss": 0.2496,
+      "step": 33050
+    },
+    {
+      "epoch": 49.71,
+      "grad_norm": 6.81777811050415,
+      "learning_rate": 5.028571428571429e-06,
+      "loss": 0.2662,
+      "step": 33060
+    },
+    {
+      "epoch": 49.73,
+      "grad_norm": 3.0888495445251465,
+      "learning_rate": 5.027067669172933e-06,
+      "loss": 0.2782,
+      "step": 33070
+    },
+    {
+      "epoch": 49.74,
+      "grad_norm": 5.303954124450684,
+      "learning_rate": 5.025563909774437e-06,
+      "loss": 0.205,
+      "step": 33080
+    },
+    {
+      "epoch": 49.76,
+      "grad_norm": 3.523970603942871,
+      "learning_rate": 5.02406015037594e-06,
+      "loss": 0.1989,
+      "step": 33090
+    },
+    {
+      "epoch": 49.77,
+      "grad_norm": 5.378790378570557,
+      "learning_rate": 5.022556390977444e-06,
+      "loss": 0.2331,
+      "step": 33100
+    },
+    {
+      "epoch": 49.79,
+      "grad_norm": 5.677578926086426,
+      "learning_rate": 5.021052631578948e-06,
+      "loss": 0.2528,
+      "step": 33110
+    },
+    {
+      "epoch": 49.8,
+      "grad_norm": 9.107217788696289,
+      "learning_rate": 5.019548872180452e-06,
+      "loss": 0.1952,
+      "step": 33120
+    },
+    {
+      "epoch": 49.82,
+      "grad_norm": 4.993006706237793,
+      "learning_rate": 5.018045112781955e-06,
+      "loss": 0.1862,
+      "step": 33130
+    },
+    {
+      "epoch": 49.83,
+      "grad_norm": 5.515392780303955,
+      "learning_rate": 5.016541353383459e-06,
+      "loss": 0.2022,
+      "step": 33140
+    },
+    {
+      "epoch": 49.85,
+      "grad_norm": 8.87382698059082,
+      "learning_rate": 5.015037593984963e-06,
+      "loss": 0.2332,
+      "step": 33150
+    },
+    {
+      "epoch": 49.86,
+      "grad_norm": 6.892481327056885,
+      "learning_rate": 5.013533834586467e-06,
+      "loss": 0.1816,
+      "step": 33160
+    },
+    {
+      "epoch": 49.88,
+      "grad_norm": 2.5804250240325928,
+      "learning_rate": 5.0120300751879706e-06,
+      "loss": 0.223,
+      "step": 33170
+    },
+    {
+      "epoch": 49.89,
+      "grad_norm": 20.23832130432129,
+      "learning_rate": 5.010526315789475e-06,
+      "loss": 0.1953,
+      "step": 33180
+    },
+    {
+      "epoch": 49.91,
+      "grad_norm": 3.5330405235290527,
+      "learning_rate": 5.009022556390978e-06,
+      "loss": 0.2249,
+      "step": 33190
+    },
+    {
+      "epoch": 49.92,
+      "grad_norm": 7.592809677124023,
+      "learning_rate": 5.007518796992482e-06,
+      "loss": 0.3143,
+      "step": 33200
+    },
+    {
+      "epoch": 49.94,
+      "grad_norm": 5.459137916564941,
+      "learning_rate": 5.0060150375939855e-06,
+      "loss": 0.2856,
+      "step": 33210
+    },
+    {
+      "epoch": 49.95,
+      "grad_norm": 5.517751216888428,
+      "learning_rate": 5.00451127819549e-06,
+      "loss": 0.1942,
+      "step": 33220
+    },
+    {
+      "epoch": 49.97,
+      "grad_norm": 5.425515651702881,
+      "learning_rate": 5.0030075187969925e-06,
+      "loss": 0.207,
+      "step": 33230
+    },
+    {
+      "epoch": 49.98,
+      "grad_norm": 4.391520023345947,
+      "learning_rate": 5.001503759398497e-06,
+      "loss": 0.2186,
+      "step": 33240
+    },
+    {
+      "epoch": 50.0,
+      "grad_norm": 8.936457633972168,
+      "learning_rate": 5e-06,
+      "loss": 0.2562,
+      "step": 33250
+    },
+    {
+      "epoch": 50.0,
+      "eval_accuracy": 0.93,
+      "eval_loss": 0.31154853105545044,
+      "eval_runtime": 84.9326,
+      "eval_samples_per_second": 117.74,
+      "eval_steps_per_second": 0.471,
+      "step": 33250
+    },
+    {
+      "epoch": 50.02,
+      "grad_norm": 4.7076497077941895,
+      "learning_rate": 4.998496240601504e-06,
+      "loss": 0.2196,
+      "step": 33260
+    },
+    {
+      "epoch": 50.03,
+      "grad_norm": 3.4939560890197754,
+      "learning_rate": 4.996992481203008e-06,
+      "loss": 0.2139,
+      "step": 33270
+    },
+    {
+      "epoch": 50.05,
+      "grad_norm": 3.0447769165039062,
+      "learning_rate": 4.995488721804512e-06,
+      "loss": 0.1865,
+      "step": 33280
+    },
+    {
+      "epoch": 50.06,
+      "grad_norm": 2.098543405532837,
+      "learning_rate": 4.993984962406015e-06,
+      "loss": 0.2419,
+      "step": 33290
+    },
+    {
+      "epoch": 50.08,
+      "grad_norm": 5.257837772369385,
+      "learning_rate": 4.992481203007519e-06,
+      "loss": 0.2051,
+      "step": 33300
+    },
+    {
+      "epoch": 50.09,
+      "grad_norm": 4.075168609619141,
+      "learning_rate": 4.990977443609023e-06,
+      "loss": 0.2148,
+      "step": 33310
+    },
+    {
+      "epoch": 50.11,
+      "grad_norm": 3.889608144760132,
+      "learning_rate": 4.989473684210527e-06,
+      "loss": 0.2621,
+      "step": 33320
+    },
+    {
+      "epoch": 50.12,
+      "grad_norm": 5.086583614349365,
+      "learning_rate": 4.98796992481203e-06,
+      "loss": 0.2163,
+      "step": 33330
+    },
+    {
+      "epoch": 50.14,
+      "grad_norm": 4.367302894592285,
+      "learning_rate": 4.986466165413535e-06,
+      "loss": 0.2182,
+      "step": 33340
+    },
+    {
+      "epoch": 50.15,
+      "grad_norm": 4.624885559082031,
+      "learning_rate": 4.984962406015038e-06,
+      "loss": 0.2048,
+      "step": 33350
+    },
+    {
+      "epoch": 50.17,
+      "grad_norm": 5.809417247772217,
+      "learning_rate": 4.983458646616542e-06,
+      "loss": 0.2646,
+      "step": 33360
+    },
+    {
+      "epoch": 50.18,
+      "grad_norm": 2.8429479598999023,
+      "learning_rate": 4.981954887218046e-06,
+      "loss": 0.2445,
+      "step": 33370
+    },
+    {
+      "epoch": 50.2,
+      "grad_norm": 7.78208065032959,
+      "learning_rate": 4.9804511278195495e-06,
+      "loss": 0.186,
+      "step": 33380
+    },
+    {
+      "epoch": 50.21,
+      "grad_norm": 5.874340534210205,
+      "learning_rate": 4.978947368421053e-06,
+      "loss": 0.2056,
+      "step": 33390
+    },
+    {
+      "epoch": 50.23,
+      "grad_norm": 3.485530138015747,
+      "learning_rate": 4.977443609022557e-06,
+      "loss": 0.2269,
+      "step": 33400
+    },
+    {
+      "epoch": 50.24,
+      "grad_norm": 6.335367679595947,
+      "learning_rate": 4.975939849624061e-06,
+      "loss": 0.1929,
+      "step": 33410
+    },
+    {
+      "epoch": 50.26,
+      "grad_norm": 7.2522382736206055,
+      "learning_rate": 4.9744360902255645e-06,
+      "loss": 0.2561,
+      "step": 33420
+    },
+    {
+      "epoch": 50.27,
+      "grad_norm": 8.870294570922852,
+      "learning_rate": 4.972932330827068e-06,
+      "loss": 0.2507,
+      "step": 33430
+    },
+    {
+      "epoch": 50.29,
+      "grad_norm": 2.679324150085449,
+      "learning_rate": 4.971428571428572e-06,
+      "loss": 0.1811,
+      "step": 33440
+    },
+    {
+      "epoch": 50.3,
+      "grad_norm": 4.810848236083984,
+      "learning_rate": 4.969924812030076e-06,
+      "loss": 0.2152,
+      "step": 33450
+    },
+    {
+      "epoch": 50.32,
+      "grad_norm": 7.02167272567749,
+      "learning_rate": 4.968421052631579e-06,
+      "loss": 0.1854,
+      "step": 33460
+    },
+    {
+      "epoch": 50.33,
+      "grad_norm": 3.6129260063171387,
+      "learning_rate": 4.966917293233084e-06,
+      "loss": 0.234,
+      "step": 33470
+    },
+    {
+      "epoch": 50.35,
+      "grad_norm": 3.452362060546875,
+      "learning_rate": 4.965413533834587e-06,
+      "loss": 0.2666,
+      "step": 33480
+    },
+    {
+      "epoch": 50.36,
+      "grad_norm": 6.474914073944092,
+      "learning_rate": 4.963909774436091e-06,
+      "loss": 0.2418,
+      "step": 33490
+    },
+    {
+      "epoch": 50.38,
+      "grad_norm": 3.1944398880004883,
+      "learning_rate": 4.962406015037594e-06,
+      "loss": 0.2529,
+      "step": 33500
+    },
+    {
+      "epoch": 50.39,
+      "grad_norm": 4.8096699714660645,
+      "learning_rate": 4.960902255639098e-06,
+      "loss": 0.2176,
+      "step": 33510
+    },
+    {
+      "epoch": 50.41,
+      "grad_norm": 4.8568925857543945,
+      "learning_rate": 4.959398496240601e-06,
+      "loss": 0.1876,
+      "step": 33520
+    },
+    {
+      "epoch": 50.42,
+      "grad_norm": 5.153539657592773,
+      "learning_rate": 4.957894736842106e-06,
+      "loss": 0.2234,
+      "step": 33530
+    },
+    {
+      "epoch": 50.44,
+      "grad_norm": 12.457379341125488,
+      "learning_rate": 4.956390977443609e-06,
+      "loss": 0.2267,
+      "step": 33540
+    },
+    {
+      "epoch": 50.45,
+      "grad_norm": 3.840301990509033,
+      "learning_rate": 4.954887218045113e-06,
+      "loss": 0.231,
+      "step": 33550
+    },
+    {
+      "epoch": 50.47,
+      "grad_norm": 4.2671966552734375,
+      "learning_rate": 4.953383458646617e-06,
+      "loss": 0.2221,
+      "step": 33560
+    },
+    {
+      "epoch": 50.48,
+      "grad_norm": 1.5949358940124512,
+      "learning_rate": 4.951879699248121e-06,
+      "loss": 0.2137,
+      "step": 33570
+    },
+    {
+      "epoch": 50.5,
+      "grad_norm": 4.011662006378174,
+      "learning_rate": 4.950375939849624e-06,
+      "loss": 0.2406,
+      "step": 33580
+    },
+    {
+      "epoch": 50.51,
+      "grad_norm": 5.408310413360596,
+      "learning_rate": 4.948872180451128e-06,
+      "loss": 0.2115,
+      "step": 33590
+    },
+    {
+      "epoch": 50.53,
+      "grad_norm": 13.068920135498047,
+      "learning_rate": 4.947368421052632e-06,
+      "loss": 0.2343,
+      "step": 33600
+    },
+    {
+      "epoch": 50.54,
+      "grad_norm": 5.289875507354736,
+      "learning_rate": 4.9458646616541356e-06,
+      "loss": 0.212,
+      "step": 33610
+    },
+    {
+      "epoch": 50.56,
+      "grad_norm": 9.420212745666504,
+      "learning_rate": 4.944360902255639e-06,
+      "loss": 0.2435,
+      "step": 33620
+    },
+    {
+      "epoch": 50.57,
+      "grad_norm": 4.894474506378174,
+      "learning_rate": 4.9428571428571435e-06,
+      "loss": 0.1806,
+      "step": 33630
+    },
+    {
+      "epoch": 50.59,
+      "grad_norm": 5.293659687042236,
+      "learning_rate": 4.941353383458647e-06,
+      "loss": 0.2205,
+      "step": 33640
+    },
+    {
+      "epoch": 50.6,
+      "grad_norm": 3.308318853378296,
+      "learning_rate": 4.9398496240601505e-06,
+      "loss": 0.2025,
+      "step": 33650
+    },
+    {
+      "epoch": 50.62,
+      "grad_norm": 6.303762435913086,
+      "learning_rate": 4.938345864661655e-06,
+      "loss": 0.2399,
+      "step": 33660
+    },
+    {
+      "epoch": 50.63,
+      "grad_norm": 5.376105308532715,
+      "learning_rate": 4.936842105263158e-06,
+      "loss": 0.1792,
+      "step": 33670
+    },
+    {
+      "epoch": 50.65,
+      "grad_norm": 7.851215839385986,
+      "learning_rate": 4.935338345864662e-06,
+      "loss": 0.2101,
+      "step": 33680
+    },
+    {
+      "epoch": 50.66,
+      "grad_norm": 4.721893310546875,
+      "learning_rate": 4.933834586466165e-06,
+      "loss": 0.2835,
+      "step": 33690
+    },
+    {
+      "epoch": 50.68,
+      "grad_norm": 5.323988914489746,
+      "learning_rate": 4.93233082706767e-06,
+      "loss": 0.2572,
+      "step": 33700
+    },
+    {
+      "epoch": 50.69,
+      "grad_norm": 6.7880964279174805,
+      "learning_rate": 4.930827067669173e-06,
+      "loss": 0.2099,
+      "step": 33710
+    },
+    {
+      "epoch": 50.71,
+      "grad_norm": 4.452812194824219,
+      "learning_rate": 4.929323308270677e-06,
+      "loss": 0.2393,
+      "step": 33720
+    },
+    {
+      "epoch": 50.72,
+      "grad_norm": 3.799269676208496,
+      "learning_rate": 4.927819548872181e-06,
+      "loss": 0.1869,
+      "step": 33730
+    },
+    {
+      "epoch": 50.74,
+      "grad_norm": 4.9699273109436035,
+      "learning_rate": 4.926315789473685e-06,
+      "loss": 0.242,
+      "step": 33740
+    },
+    {
+      "epoch": 50.75,
+      "grad_norm": 3.6977415084838867,
+      "learning_rate": 4.924812030075188e-06,
+      "loss": 0.1882,
+      "step": 33750
+    },
+    {
+      "epoch": 50.77,
+      "grad_norm": 4.581582546234131,
+      "learning_rate": 4.923308270676692e-06,
+      "loss": 0.2257,
+      "step": 33760
+    },
+    {
+      "epoch": 50.78,
+      "grad_norm": 6.505821228027344,
+      "learning_rate": 4.921804511278196e-06,
+      "loss": 0.2212,
+      "step": 33770
+    },
+    {
+      "epoch": 50.8,
+      "grad_norm": 6.289919376373291,
+      "learning_rate": 4.9203007518797e-06,
+      "loss": 0.2553,
+      "step": 33780
+    },
+    {
+      "epoch": 50.81,
+      "grad_norm": 3.341113567352295,
+      "learning_rate": 4.918796992481203e-06,
+      "loss": 0.2567,
+      "step": 33790
+    },
+    {
+      "epoch": 50.83,
+      "grad_norm": 4.804263114929199,
+      "learning_rate": 4.9172932330827075e-06,
+      "loss": 0.2432,
+      "step": 33800
+    },
+    {
+      "epoch": 50.84,
+      "grad_norm": 6.457090854644775,
+      "learning_rate": 4.915789473684211e-06,
+      "loss": 0.223,
+      "step": 33810
+    },
+    {
+      "epoch": 50.86,
+      "grad_norm": 7.253601551055908,
+      "learning_rate": 4.9142857142857145e-06,
+      "loss": 0.2747,
+      "step": 33820
+    },
+    {
+      "epoch": 50.87,
+      "grad_norm": 4.229115962982178,
+      "learning_rate": 4.912781954887219e-06,
+      "loss": 0.2616,
+      "step": 33830
+    },
+    {
+      "epoch": 50.89,
+      "grad_norm": 3.4108991622924805,
+      "learning_rate": 4.9112781954887224e-06,
+      "loss": 0.2176,
+      "step": 33840
+    },
+    {
+      "epoch": 50.9,
+      "grad_norm": 6.427495002746582,
+      "learning_rate": 4.909774436090226e-06,
+      "loss": 0.1867,
+      "step": 33850
+    },
+    {
+      "epoch": 50.92,
+      "grad_norm": 4.823879718780518,
+      "learning_rate": 4.9082706766917295e-06,
+      "loss": 0.2758,
+      "step": 33860
+    },
+    {
+      "epoch": 50.93,
+      "grad_norm": 5.894002914428711,
+      "learning_rate": 4.906766917293234e-06,
+      "loss": 0.275,
+      "step": 33870
+    },
+    {
+      "epoch": 50.95,
+      "grad_norm": 7.533518314361572,
+      "learning_rate": 4.905263157894737e-06,
+      "loss": 0.2269,
+      "step": 33880
+    },
+    {
+      "epoch": 50.96,
+      "grad_norm": 6.328519344329834,
+      "learning_rate": 4.903759398496241e-06,
+      "loss": 0.2842,
+      "step": 33890
+    },
+    {
+      "epoch": 50.98,
+      "grad_norm": 2.4413256645202637,
+      "learning_rate": 4.902255639097745e-06,
+      "loss": 0.1786,
+      "step": 33900
+    },
+    {
+      "epoch": 50.99,
+      "grad_norm": 2.602782964706421,
+      "learning_rate": 4.900751879699249e-06,
+      "loss": 0.1847,
+      "step": 33910
+    },
+    {
+      "epoch": 51.0,
+      "eval_accuracy": 0.9311,
+      "eval_loss": 0.3058427572250366,
+      "eval_runtime": 84.9491,
+      "eval_samples_per_second": 117.718,
+      "eval_steps_per_second": 0.471,
+      "step": 33915
+    },
+    {
+      "epoch": 51.01,
+      "grad_norm": 12.357138633728027,
+      "learning_rate": 4.899248120300752e-06,
+      "loss": 0.1841,
+      "step": 33920
+    },
+    {
+      "epoch": 51.02,
+      "grad_norm": 6.953237056732178,
+      "learning_rate": 4.897744360902257e-06,
+      "loss": 0.2775,
+      "step": 33930
+    },
+    {
+      "epoch": 51.04,
+      "grad_norm": 4.370208263397217,
+      "learning_rate": 4.89624060150376e-06,
+      "loss": 0.216,
+      "step": 33940
+    },
+    {
+      "epoch": 51.05,
+      "grad_norm": 5.940341472625732,
+      "learning_rate": 4.894736842105264e-06,
+      "loss": 0.2888,
+      "step": 33950
+    },
+    {
+      "epoch": 51.07,
+      "grad_norm": 2.9402363300323486,
+      "learning_rate": 4.893233082706767e-06,
+      "loss": 0.2437,
+      "step": 33960
+    },
+    {
+      "epoch": 51.08,
+      "grad_norm": 6.118091583251953,
+      "learning_rate": 4.8917293233082716e-06,
+      "loss": 0.2558,
+      "step": 33970
+    },
+    {
+      "epoch": 51.1,
+      "grad_norm": 3.255457878112793,
+      "learning_rate": 4.890225563909775e-06,
+      "loss": 0.217,
+      "step": 33980
+    },
+    {
+      "epoch": 51.11,
+      "grad_norm": 6.439407825469971,
+      "learning_rate": 4.888721804511279e-06,
+      "loss": 0.2119,
+      "step": 33990
+    },
+    {
+      "epoch": 51.13,
+      "grad_norm": 5.628686904907227,
+      "learning_rate": 4.887218045112782e-06,
+      "loss": 0.2647,
+      "step": 34000
+    },
+    {
+      "epoch": 51.14,
+      "grad_norm": 3.908249616622925,
+      "learning_rate": 4.885714285714286e-06,
+      "loss": 0.221,
+      "step": 34010
+    },
+    {
+      "epoch": 51.16,
+      "grad_norm": 5.72270393371582,
+      "learning_rate": 4.88421052631579e-06,
+      "loss": 0.2116,
+      "step": 34020
+    },
+    {
+      "epoch": 51.17,
+      "grad_norm": 6.743198394775391,
+      "learning_rate": 4.8827067669172935e-06,
+      "loss": 0.2716,
+      "step": 34030
+    },
+    {
+      "epoch": 51.19,
+      "grad_norm": 6.313354969024658,
+      "learning_rate": 4.881203007518797e-06,
+      "loss": 0.242,
+      "step": 34040
+    },
+    {
+      "epoch": 51.2,
+      "grad_norm": 4.181947708129883,
+      "learning_rate": 4.8796992481203006e-06,
+      "loss": 0.2448,
+      "step": 34050
+    },
+    {
+      "epoch": 51.22,
+      "grad_norm": 7.440803050994873,
+      "learning_rate": 4.878195488721805e-06,
+      "loss": 0.2734,
+      "step": 34060
+    },
+    {
+      "epoch": 51.23,
+      "grad_norm": 7.924306869506836,
+      "learning_rate": 4.8766917293233085e-06,
+      "loss": 0.2438,
+      "step": 34070
+    },
+    {
+      "epoch": 51.25,
+      "grad_norm": 4.542608737945557,
+      "learning_rate": 4.875187969924812e-06,
+      "loss": 0.2291,
+      "step": 34080
+    },
+    {
+      "epoch": 51.26,
+      "grad_norm": 3.60923171043396,
+      "learning_rate": 4.873684210526316e-06,
+      "loss": 0.201,
+      "step": 34090
+    },
+    {
+      "epoch": 51.28,
+      "grad_norm": 7.753152847290039,
+      "learning_rate": 4.87218045112782e-06,
+      "loss": 0.2213,
+      "step": 34100
+    },
+    {
+      "epoch": 51.29,
+      "grad_norm": 4.770752429962158,
+      "learning_rate": 4.870676691729323e-06,
+      "loss": 0.1961,
+      "step": 34110
+    },
+    {
+      "epoch": 51.31,
+      "grad_norm": 4.9310479164123535,
+      "learning_rate": 4.869172932330828e-06,
+      "loss": 0.1574,
+      "step": 34120
+    },
+    {
+      "epoch": 51.32,
+      "grad_norm": 3.8556582927703857,
+      "learning_rate": 4.867669172932331e-06,
+      "loss": 0.2608,
+      "step": 34130
+    },
+    {
+      "epoch": 51.34,
+      "grad_norm": 6.298683166503906,
+      "learning_rate": 4.866165413533835e-06,
+      "loss": 0.2558,
+      "step": 34140
+    },
+    {
+      "epoch": 51.35,
+      "grad_norm": 3.553882122039795,
+      "learning_rate": 4.864661654135338e-06,
+      "loss": 0.152,
+      "step": 34150
+    },
+    {
+      "epoch": 51.37,
+      "grad_norm": 4.3429975509643555,
+      "learning_rate": 4.863157894736843e-06,
+      "loss": 0.2239,
+      "step": 34160
+    },
+    {
+      "epoch": 51.38,
+      "grad_norm": 3.7060399055480957,
+      "learning_rate": 4.861654135338346e-06,
+      "loss": 0.268,
+      "step": 34170
+    },
+    {
+      "epoch": 51.4,
+      "grad_norm": 4.588679313659668,
+      "learning_rate": 4.86015037593985e-06,
+      "loss": 0.2724,
+      "step": 34180
+    },
+    {
+      "epoch": 51.41,
+      "grad_norm": 3.715130567550659,
+      "learning_rate": 4.858646616541354e-06,
+      "loss": 0.1813,
+      "step": 34190
+    },
+    {
+      "epoch": 51.43,
+      "grad_norm": 6.059221267700195,
+      "learning_rate": 4.857142857142858e-06,
+      "loss": 0.201,
+      "step": 34200
+    },
+    {
+      "epoch": 51.44,
+      "grad_norm": 9.141858100891113,
+      "learning_rate": 4.855639097744361e-06,
+      "loss": 0.2786,
+      "step": 34210
+    },
+    {
+      "epoch": 51.46,
+      "grad_norm": 5.355362892150879,
+      "learning_rate": 4.854135338345865e-06,
+      "loss": 0.2175,
+      "step": 34220
+    },
+    {
+      "epoch": 51.47,
+      "grad_norm": 4.845487117767334,
+      "learning_rate": 4.852631578947369e-06,
+      "loss": 0.2597,
+      "step": 34230
+    },
+    {
+      "epoch": 51.49,
+      "grad_norm": 4.735893249511719,
+      "learning_rate": 4.8511278195488725e-06,
+      "loss": 0.2157,
+      "step": 34240
+    },
+    {
+      "epoch": 51.5,
+      "grad_norm": 5.093930244445801,
+      "learning_rate": 4.849624060150376e-06,
+      "loss": 0.1836,
+      "step": 34250
+    },
+    {
+      "epoch": 51.52,
+      "grad_norm": 1.877554178237915,
+      "learning_rate": 4.84812030075188e-06,
+      "loss": 0.1809,
+      "step": 34260
+    },
+    {
+      "epoch": 51.53,
+      "grad_norm": 5.213804721832275,
+      "learning_rate": 4.846616541353384e-06,
+      "loss": 0.2439,
+      "step": 34270
+    },
+    {
+      "epoch": 51.55,
+      "grad_norm": 2.9461185932159424,
+      "learning_rate": 4.8451127819548874e-06,
+      "loss": 0.1654,
+      "step": 34280
+    },
+    {
+      "epoch": 51.56,
+      "grad_norm": 3.598033905029297,
+      "learning_rate": 4.843609022556392e-06,
+      "loss": 0.1807,
+      "step": 34290
+    },
+    {
+      "epoch": 51.58,
+      "grad_norm": 3.246472120285034,
+      "learning_rate": 4.842105263157895e-06,
+      "loss": 0.2873,
+      "step": 34300
+    },
+    {
+      "epoch": 51.59,
+      "grad_norm": 7.0338921546936035,
+      "learning_rate": 4.840601503759399e-06,
+      "loss": 0.257,
+      "step": 34310
+    },
+    {
+      "epoch": 51.61,
+      "grad_norm": 7.394874095916748,
+      "learning_rate": 4.839097744360902e-06,
+      "loss": 0.1845,
+      "step": 34320
+    },
+    {
+      "epoch": 51.62,
+      "grad_norm": 5.90761661529541,
+      "learning_rate": 4.837593984962407e-06,
+      "loss": 0.2476,
+      "step": 34330
+    },
+    {
+      "epoch": 51.64,
+      "grad_norm": 2.5447800159454346,
+      "learning_rate": 4.83609022556391e-06,
+      "loss": 0.2252,
+      "step": 34340
+    },
+    {
+      "epoch": 51.65,
+      "grad_norm": 6.951369285583496,
+      "learning_rate": 4.834586466165414e-06,
+      "loss": 0.198,
+      "step": 34350
+    },
+    {
+      "epoch": 51.67,
+      "grad_norm": 8.201958656311035,
+      "learning_rate": 4.833082706766918e-06,
+      "loss": 0.2374,
+      "step": 34360
+    },
+    {
+      "epoch": 51.68,
+      "grad_norm": 7.837949752807617,
+      "learning_rate": 4.831578947368422e-06,
+      "loss": 0.1674,
+      "step": 34370
+    },
+    {
+      "epoch": 51.7,
+      "grad_norm": 5.410804748535156,
+      "learning_rate": 4.830075187969925e-06,
+      "loss": 0.2734,
+      "step": 34380
+    },
+    {
+      "epoch": 51.71,
+      "grad_norm": 3.385745048522949,
+      "learning_rate": 4.8285714285714295e-06,
+      "loss": 0.234,
+      "step": 34390
+    },
+    {
+      "epoch": 51.73,
+      "grad_norm": 7.365379810333252,
+      "learning_rate": 4.827067669172933e-06,
+      "loss": 0.2139,
+      "step": 34400
+    },
+    {
+      "epoch": 51.74,
+      "grad_norm": 6.774968147277832,
+      "learning_rate": 4.8255639097744366e-06,
+      "loss": 0.151,
+      "step": 34410
+    },
+    {
+      "epoch": 51.76,
+      "grad_norm": 5.092763900756836,
+      "learning_rate": 4.82406015037594e-06,
+      "loss": 0.2152,
+      "step": 34420
+    },
+    {
+      "epoch": 51.77,
+      "grad_norm": 3.2947609424591064,
+      "learning_rate": 4.8225563909774445e-06,
+      "loss": 0.2309,
+      "step": 34430
+    },
+    {
+      "epoch": 51.79,
+      "grad_norm": 4.706571102142334,
+      "learning_rate": 4.821052631578948e-06,
+      "loss": 0.2546,
+      "step": 34440
+    },
+    {
+      "epoch": 51.8,
+      "grad_norm": 8.115716934204102,
+      "learning_rate": 4.8195488721804515e-06,
+      "loss": 0.1969,
+      "step": 34450
+    },
+    {
+      "epoch": 51.82,
+      "grad_norm": 7.135550022125244,
+      "learning_rate": 4.818045112781956e-06,
+      "loss": 0.2594,
+      "step": 34460
+    },
+    {
+      "epoch": 51.83,
+      "grad_norm": 4.663880348205566,
+      "learning_rate": 4.816541353383459e-06,
+      "loss": 0.245,
+      "step": 34470
+    },
+    {
+      "epoch": 51.85,
+      "grad_norm": 5.021750450134277,
+      "learning_rate": 4.815037593984963e-06,
+      "loss": 0.2571,
+      "step": 34480
+    },
+    {
+      "epoch": 51.86,
+      "grad_norm": 3.6975669860839844,
+      "learning_rate": 4.813533834586466e-06,
+      "loss": 0.2923,
+      "step": 34490
+    },
+    {
+      "epoch": 51.88,
+      "grad_norm": 5.429826259613037,
+      "learning_rate": 4.81203007518797e-06,
+      "loss": 0.158,
+      "step": 34500
+    },
+    {
+      "epoch": 51.89,
+      "grad_norm": 5.443629264831543,
+      "learning_rate": 4.8105263157894735e-06,
+      "loss": 0.1788,
+      "step": 34510
+    },
+    {
+      "epoch": 51.91,
+      "grad_norm": 2.6010379791259766,
+      "learning_rate": 4.809022556390978e-06,
+      "loss": 0.2203,
+      "step": 34520
+    },
+    {
+      "epoch": 51.92,
+      "grad_norm": 2.72519588470459,
+      "learning_rate": 4.807518796992481e-06,
+      "loss": 0.2021,
+      "step": 34530
+    },
+    {
+      "epoch": 51.94,
+      "grad_norm": 7.415553569793701,
+      "learning_rate": 4.806015037593985e-06,
+      "loss": 0.2145,
+      "step": 34540
+    },
+    {
+      "epoch": 51.95,
+      "grad_norm": 3.181380033493042,
+      "learning_rate": 4.804511278195489e-06,
+      "loss": 0.1729,
+      "step": 34550
+    },
+    {
+      "epoch": 51.97,
+      "grad_norm": 4.509653091430664,
+      "learning_rate": 4.803007518796993e-06,
+      "loss": 0.1733,
+      "step": 34560
+    },
+    {
+      "epoch": 51.98,
+      "grad_norm": 2.9621808528900146,
+      "learning_rate": 4.801503759398496e-06,
+      "loss": 0.2322,
+      "step": 34570
+    },
+    {
+      "epoch": 52.0,
+      "grad_norm": 16.024185180664062,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.2453,
+      "step": 34580
+    },
+    {
+      "epoch": 52.0,
+      "eval_accuracy": 0.9308,
+      "eval_loss": 0.3179681897163391,
+      "eval_runtime": 84.5142,
+      "eval_samples_per_second": 118.323,
+      "eval_steps_per_second": 0.473,
+      "step": 34580
+    },
+    {
+      "epoch": 52.02,
+      "grad_norm": 4.019076347351074,
+      "learning_rate": 4.798496240601504e-06,
+      "loss": 0.1856,
+      "step": 34590
+    },
+    {
+      "epoch": 52.03,
+      "grad_norm": 4.832670211791992,
+      "learning_rate": 4.796992481203008e-06,
+      "loss": 0.1791,
+      "step": 34600
+    },
+    {
+      "epoch": 52.05,
+      "grad_norm": 3.983247756958008,
+      "learning_rate": 4.795488721804511e-06,
+      "loss": 0.2176,
+      "step": 34610
+    },
+    {
+      "epoch": 52.06,
+      "grad_norm": 5.2948832511901855,
+      "learning_rate": 4.7939849624060156e-06,
+      "loss": 0.2737,
+      "step": 34620
+    },
+    {
+      "epoch": 52.08,
+      "grad_norm": 3.9428210258483887,
+      "learning_rate": 4.792481203007519e-06,
+      "loss": 0.2457,
+      "step": 34630
+    },
+    {
+      "epoch": 52.09,
+      "grad_norm": 7.841516971588135,
+      "learning_rate": 4.790977443609023e-06,
+      "loss": 0.2125,
+      "step": 34640
+    },
+    {
+      "epoch": 52.11,
+      "grad_norm": 2.363363265991211,
+      "learning_rate": 4.789473684210527e-06,
+      "loss": 0.1784,
+      "step": 34650
+    },
+    {
+      "epoch": 52.12,
+      "grad_norm": 3.259340286254883,
+      "learning_rate": 4.7879699248120305e-06,
+      "loss": 0.2177,
+      "step": 34660
+    },
+    {
+      "epoch": 52.14,
+      "grad_norm": 2.5872132778167725,
+      "learning_rate": 4.786466165413534e-06,
+      "loss": 0.229,
+      "step": 34670
+    },
+    {
+      "epoch": 52.15,
+      "grad_norm": 6.223973274230957,
+      "learning_rate": 4.7849624060150375e-06,
+      "loss": 0.2612,
+      "step": 34680
+    },
+    {
+      "epoch": 52.17,
+      "grad_norm": 8.220977783203125,
+      "learning_rate": 4.783458646616542e-06,
+      "loss": 0.2316,
+      "step": 34690
+    },
+    {
+      "epoch": 52.18,
+      "grad_norm": 5.418039321899414,
+      "learning_rate": 4.781954887218045e-06,
+      "loss": 0.2063,
+      "step": 34700
+    },
+    {
+      "epoch": 52.2,
+      "grad_norm": 2.79838490486145,
+      "learning_rate": 4.780451127819549e-06,
+      "loss": 0.1965,
+      "step": 34710
+    },
+    {
+      "epoch": 52.21,
+      "grad_norm": 5.194334030151367,
+      "learning_rate": 4.778947368421053e-06,
+      "loss": 0.1817,
+      "step": 34720
+    },
+    {
+      "epoch": 52.23,
+      "grad_norm": 4.306404113769531,
+      "learning_rate": 4.777443609022557e-06,
+      "loss": 0.1943,
+      "step": 34730
+    },
+    {
+      "epoch": 52.24,
+      "grad_norm": 5.353394508361816,
+      "learning_rate": 4.77593984962406e-06,
+      "loss": 0.2019,
+      "step": 34740
+    },
+    {
+      "epoch": 52.26,
+      "grad_norm": 4.193369388580322,
+      "learning_rate": 4.774436090225565e-06,
+      "loss": 0.1593,
+      "step": 34750
+    },
+    {
+      "epoch": 52.27,
+      "grad_norm": 5.131227493286133,
+      "learning_rate": 4.772932330827068e-06,
+      "loss": 0.2845,
+      "step": 34760
+    },
+    {
+      "epoch": 52.29,
+      "grad_norm": 6.315420627593994,
+      "learning_rate": 4.771428571428572e-06,
+      "loss": 0.2163,
+      "step": 34770
+    },
+    {
+      "epoch": 52.3,
+      "grad_norm": 5.833438873291016,
+      "learning_rate": 4.769924812030075e-06,
+      "loss": 0.2064,
+      "step": 34780
+    },
+    {
+      "epoch": 52.32,
+      "grad_norm": 2.4818150997161865,
+      "learning_rate": 4.76842105263158e-06,
+      "loss": 0.2169,
+      "step": 34790
+    },
+    {
+      "epoch": 52.33,
+      "grad_norm": 5.627359390258789,
+      "learning_rate": 4.766917293233083e-06,
+      "loss": 0.2654,
+      "step": 34800
+    },
+    {
+      "epoch": 52.35,
+      "grad_norm": 4.044781684875488,
+      "learning_rate": 4.765413533834587e-06,
+      "loss": 0.233,
+      "step": 34810
+    },
+    {
+      "epoch": 52.36,
+      "grad_norm": 9.770264625549316,
+      "learning_rate": 4.763909774436091e-06,
+      "loss": 0.2018,
+      "step": 34820
+    },
+    {
+      "epoch": 52.38,
+      "grad_norm": 5.436629295349121,
+      "learning_rate": 4.7624060150375945e-06,
+      "loss": 0.1943,
+      "step": 34830
+    },
+    {
+      "epoch": 52.39,
+      "grad_norm": 8.624287605285645,
+      "learning_rate": 4.760902255639098e-06,
+      "loss": 0.2665,
+      "step": 34840
+    },
+    {
+      "epoch": 52.41,
+      "grad_norm": 5.233580589294434,
+      "learning_rate": 4.759398496240602e-06,
+      "loss": 0.1882,
+      "step": 34850
+    },
+    {
+      "epoch": 52.42,
+      "grad_norm": 5.307055950164795,
+      "learning_rate": 4.757894736842106e-06,
+      "loss": 0.2765,
+      "step": 34860
+    },
+    {
+      "epoch": 52.44,
+      "grad_norm": 4.9283928871154785,
+      "learning_rate": 4.7563909774436095e-06,
+      "loss": 0.2012,
+      "step": 34870
+    },
+    {
+      "epoch": 52.45,
+      "grad_norm": 7.545740604400635,
+      "learning_rate": 4.754887218045113e-06,
+      "loss": 0.1997,
+      "step": 34880
+    },
+    {
+      "epoch": 52.47,
+      "grad_norm": 4.995537757873535,
+      "learning_rate": 4.753383458646617e-06,
+      "loss": 0.2371,
+      "step": 34890
+    },
+    {
+      "epoch": 52.48,
+      "grad_norm": 4.122995376586914,
+      "learning_rate": 4.751879699248121e-06,
+      "loss": 0.2107,
+      "step": 34900
+    },
+    {
+      "epoch": 52.5,
+      "grad_norm": 6.755688667297363,
+      "learning_rate": 4.750375939849624e-06,
+      "loss": 0.1947,
+      "step": 34910
+    },
+    {
+      "epoch": 52.51,
+      "grad_norm": 3.243288278579712,
+      "learning_rate": 4.748872180451129e-06,
+      "loss": 0.2091,
+      "step": 34920
+    },
+    {
+      "epoch": 52.53,
+      "grad_norm": 4.718216896057129,
+      "learning_rate": 4.747368421052632e-06,
+      "loss": 0.2318,
+      "step": 34930
+    },
+    {
+      "epoch": 52.54,
+      "grad_norm": 6.7740936279296875,
+      "learning_rate": 4.745864661654136e-06,
+      "loss": 0.2014,
+      "step": 34940
+    },
+    {
+      "epoch": 52.56,
+      "grad_norm": 3.995626449584961,
+      "learning_rate": 4.744360902255639e-06,
+      "loss": 0.2062,
+      "step": 34950
+    },
+    {
+      "epoch": 52.57,
+      "grad_norm": 6.141635417938232,
+      "learning_rate": 4.742857142857144e-06,
+      "loss": 0.183,
+      "step": 34960
+    },
+    {
+      "epoch": 52.59,
+      "grad_norm": 7.161436557769775,
+      "learning_rate": 4.741353383458647e-06,
+      "loss": 0.2518,
+      "step": 34970
+    },
+    {
+      "epoch": 52.6,
+      "grad_norm": 2.2341339588165283,
+      "learning_rate": 4.739849624060151e-06,
+      "loss": 0.2516,
+      "step": 34980
+    },
+    {
+      "epoch": 52.62,
+      "grad_norm": 5.752125263214111,
+      "learning_rate": 4.738345864661654e-06,
+      "loss": 0.1559,
+      "step": 34990
+    },
+    {
+      "epoch": 52.63,
+      "grad_norm": 4.806941032409668,
+      "learning_rate": 4.736842105263158e-06,
+      "loss": 0.177,
+      "step": 35000
+    },
+    {
+      "epoch": 52.65,
+      "grad_norm": 2.0574851036071777,
+      "learning_rate": 4.735338345864662e-06,
+      "loss": 0.1954,
+      "step": 35010
+    },
+    {
+      "epoch": 52.66,
+      "grad_norm": 4.443716526031494,
+      "learning_rate": 4.733834586466166e-06,
+      "loss": 0.2045,
+      "step": 35020
+    },
+    {
+      "epoch": 52.68,
+      "grad_norm": 3.878716468811035,
+      "learning_rate": 4.732330827067669e-06,
+      "loss": 0.2095,
+      "step": 35030
+    },
+    {
+      "epoch": 52.69,
+      "grad_norm": 10.685707092285156,
+      "learning_rate": 4.7308270676691735e-06,
+      "loss": 0.2586,
+      "step": 35040
+    },
+    {
+      "epoch": 52.71,
+      "grad_norm": 3.904240369796753,
+      "learning_rate": 4.729323308270677e-06,
+      "loss": 0.3103,
+      "step": 35050
+    },
+    {
+      "epoch": 52.72,
+      "grad_norm": 4.6063337326049805,
+      "learning_rate": 4.7278195488721806e-06,
+      "loss": 0.1517,
+      "step": 35060
+    },
+    {
+      "epoch": 52.74,
+      "grad_norm": 4.317811489105225,
+      "learning_rate": 4.726315789473684e-06,
+      "loss": 0.216,
+      "step": 35070
+    },
+    {
+      "epoch": 52.75,
+      "grad_norm": 7.456090927124023,
+      "learning_rate": 4.7248120300751884e-06,
+      "loss": 0.2438,
+      "step": 35080
+    },
+    {
+      "epoch": 52.77,
+      "grad_norm": 7.284855842590332,
+      "learning_rate": 4.723308270676692e-06,
+      "loss": 0.2263,
+      "step": 35090
+    },
+    {
+      "epoch": 52.78,
+      "grad_norm": 4.406165599822998,
+      "learning_rate": 4.7218045112781955e-06,
+      "loss": 0.181,
+      "step": 35100
+    },
+    {
+      "epoch": 52.8,
+      "grad_norm": 2.9364101886749268,
+      "learning_rate": 4.7203007518797e-06,
+      "loss": 0.2351,
+      "step": 35110
+    },
+    {
+      "epoch": 52.81,
+      "grad_norm": 5.638926029205322,
+      "learning_rate": 4.718796992481203e-06,
+      "loss": 0.2289,
+      "step": 35120
+    },
+    {
+      "epoch": 52.83,
+      "grad_norm": 4.295120716094971,
+      "learning_rate": 4.717293233082707e-06,
+      "loss": 0.211,
+      "step": 35130
+    },
+    {
+      "epoch": 52.84,
+      "grad_norm": 4.422874450683594,
+      "learning_rate": 4.71578947368421e-06,
+      "loss": 0.1947,
+      "step": 35140
+    },
+    {
+      "epoch": 52.86,
+      "grad_norm": 0.5836885571479797,
+      "learning_rate": 4.714285714285715e-06,
+      "loss": 0.2299,
+      "step": 35150
+    },
+    {
+      "epoch": 52.87,
+      "grad_norm": 4.698862552642822,
+      "learning_rate": 4.712781954887218e-06,
+      "loss": 0.1864,
+      "step": 35160
+    },
+    {
+      "epoch": 52.89,
+      "grad_norm": 4.577108860015869,
+      "learning_rate": 4.711278195488722e-06,
+      "loss": 0.2257,
+      "step": 35170
+    },
+    {
+      "epoch": 52.9,
+      "grad_norm": 7.971583843231201,
+      "learning_rate": 4.709774436090226e-06,
+      "loss": 0.2093,
+      "step": 35180
+    },
+    {
+      "epoch": 52.92,
+      "grad_norm": 6.064690589904785,
+      "learning_rate": 4.70827067669173e-06,
+      "loss": 0.2444,
+      "step": 35190
+    },
+    {
+      "epoch": 52.93,
+      "grad_norm": 7.225642681121826,
+      "learning_rate": 4.706766917293233e-06,
+      "loss": 0.2325,
+      "step": 35200
+    },
+    {
+      "epoch": 52.95,
+      "grad_norm": 4.481410503387451,
+      "learning_rate": 4.705263157894738e-06,
+      "loss": 0.2357,
+      "step": 35210
+    },
+    {
+      "epoch": 52.96,
+      "grad_norm": 4.95440673828125,
+      "learning_rate": 4.703759398496241e-06,
+      "loss": 0.2526,
+      "step": 35220
+    },
+    {
+      "epoch": 52.98,
+      "grad_norm": 3.8696534633636475,
+      "learning_rate": 4.702255639097745e-06,
+      "loss": 0.2009,
+      "step": 35230
+    },
+    {
+      "epoch": 52.99,
+      "grad_norm": 6.271843910217285,
+      "learning_rate": 4.700751879699248e-06,
+      "loss": 0.2763,
+      "step": 35240
+    },
+    {
+      "epoch": 53.0,
+      "eval_accuracy": 0.932,
+      "eval_loss": 0.30757245421409607,
+      "eval_runtime": 84.7331,
+      "eval_samples_per_second": 118.018,
+      "eval_steps_per_second": 0.472,
+      "step": 35245
+    },
+    {
+      "epoch": 53.01,
+      "grad_norm": 8.736787796020508,
+      "learning_rate": 4.6992481203007525e-06,
+      "loss": 0.2478,
+      "step": 35250
+    },
+    {
+      "epoch": 53.02,
+      "grad_norm": 6.983564853668213,
+      "learning_rate": 4.697744360902256e-06,
+      "loss": 0.2756,
+      "step": 35260
+    },
+    {
+      "epoch": 53.04,
+      "grad_norm": 3.659640073776245,
+      "learning_rate": 4.6962406015037595e-06,
+      "loss": 0.2843,
+      "step": 35270
+    },
+    {
+      "epoch": 53.05,
+      "grad_norm": 6.107889175415039,
+      "learning_rate": 4.694736842105264e-06,
+      "loss": 0.2212,
+      "step": 35280
+    },
+    {
+      "epoch": 53.07,
+      "grad_norm": 3.515962600708008,
+      "learning_rate": 4.693233082706767e-06,
+      "loss": 0.199,
+      "step": 35290
+    },
+    {
+      "epoch": 53.08,
+      "grad_norm": 4.900839328765869,
+      "learning_rate": 4.691729323308271e-06,
+      "loss": 0.2398,
+      "step": 35300
+    },
+    {
+      "epoch": 53.1,
+      "grad_norm": 4.189992904663086,
+      "learning_rate": 4.690225563909775e-06,
+      "loss": 0.2255,
+      "step": 35310
+    },
+    {
+      "epoch": 53.11,
+      "grad_norm": 4.84755277633667,
+      "learning_rate": 4.688721804511279e-06,
+      "loss": 0.1594,
+      "step": 35320
+    },
+    {
+      "epoch": 53.13,
+      "grad_norm": 4.748797416687012,
+      "learning_rate": 4.687218045112782e-06,
+      "loss": 0.1627,
+      "step": 35330
+    },
+    {
+      "epoch": 53.14,
+      "grad_norm": 9.220650672912598,
+      "learning_rate": 4.685714285714286e-06,
+      "loss": 0.2756,
+      "step": 35340
+    },
+    {
+      "epoch": 53.16,
+      "grad_norm": 4.247942924499512,
+      "learning_rate": 4.68421052631579e-06,
+      "loss": 0.1987,
+      "step": 35350
+    },
+    {
+      "epoch": 53.17,
+      "grad_norm": 4.994369983673096,
+      "learning_rate": 4.682706766917294e-06,
+      "loss": 0.1973,
+      "step": 35360
+    },
+    {
+      "epoch": 53.19,
+      "grad_norm": 5.42191743850708,
+      "learning_rate": 4.681203007518797e-06,
+      "loss": 0.2074,
+      "step": 35370
+    },
+    {
+      "epoch": 53.2,
+      "grad_norm": 3.5923995971679688,
+      "learning_rate": 4.679699248120302e-06,
+      "loss": 0.2612,
+      "step": 35380
+    },
+    {
+      "epoch": 53.22,
+      "grad_norm": 5.668188095092773,
+      "learning_rate": 4.678195488721805e-06,
+      "loss": 0.2088,
+      "step": 35390
+    },
+    {
+      "epoch": 53.23,
+      "grad_norm": 4.510677337646484,
+      "learning_rate": 4.676691729323309e-06,
+      "loss": 0.2987,
+      "step": 35400
+    },
+    {
+      "epoch": 53.25,
+      "grad_norm": 6.8448028564453125,
+      "learning_rate": 4.675187969924812e-06,
+      "loss": 0.236,
+      "step": 35410
+    },
+    {
+      "epoch": 53.26,
+      "grad_norm": 5.6076130867004395,
+      "learning_rate": 4.6736842105263166e-06,
+      "loss": 0.183,
+      "step": 35420
+    },
+    {
+      "epoch": 53.28,
+      "grad_norm": 5.029841423034668,
+      "learning_rate": 4.67218045112782e-06,
+      "loss": 0.1961,
+      "step": 35430
+    },
+    {
+      "epoch": 53.29,
+      "grad_norm": 4.591315746307373,
+      "learning_rate": 4.670676691729324e-06,
+      "loss": 0.2227,
+      "step": 35440
+    },
+    {
+      "epoch": 53.31,
+      "grad_norm": 8.496804237365723,
+      "learning_rate": 4.669172932330828e-06,
+      "loss": 0.2237,
+      "step": 35450
+    },
+    {
+      "epoch": 53.32,
+      "grad_norm": 4.822425842285156,
+      "learning_rate": 4.6676691729323315e-06,
+      "loss": 0.2348,
+      "step": 35460
+    },
+    {
+      "epoch": 53.34,
+      "grad_norm": 3.9777705669403076,
+      "learning_rate": 4.666165413533835e-06,
+      "loss": 0.2115,
+      "step": 35470
+    },
+    {
+      "epoch": 53.35,
+      "grad_norm": 16.29306983947754,
+      "learning_rate": 4.664661654135339e-06,
+      "loss": 0.1931,
+      "step": 35480
+    },
+    {
+      "epoch": 53.37,
+      "grad_norm": 2.485229730606079,
+      "learning_rate": 4.663157894736842e-06,
+      "loss": 0.1755,
+      "step": 35490
+    },
+    {
+      "epoch": 53.38,
+      "grad_norm": 5.529495716094971,
+      "learning_rate": 4.661654135338346e-06,
+      "loss": 0.2145,
+      "step": 35500
+    },
+    {
+      "epoch": 53.4,
+      "grad_norm": 5.058990478515625,
+      "learning_rate": 4.66015037593985e-06,
+      "loss": 0.2215,
+      "step": 35510
+    },
+    {
+      "epoch": 53.41,
+      "grad_norm": 3.2668142318725586,
+      "learning_rate": 4.6586466165413534e-06,
+      "loss": 0.2098,
+      "step": 35520
+    },
+    {
+      "epoch": 53.43,
+      "grad_norm": 8.407674789428711,
+      "learning_rate": 4.657142857142857e-06,
+      "loss": 0.1849,
+      "step": 35530
+    },
+    {
+      "epoch": 53.44,
+      "grad_norm": 5.4570488929748535,
+      "learning_rate": 4.655639097744361e-06,
+      "loss": 0.2515,
+      "step": 35540
+    },
+    {
+      "epoch": 53.46,
+      "grad_norm": 4.499083042144775,
+      "learning_rate": 4.654135338345865e-06,
+      "loss": 0.2259,
+      "step": 35550
+    },
+    {
+      "epoch": 53.47,
+      "grad_norm": 9.1441068649292,
+      "learning_rate": 4.652631578947368e-06,
+      "loss": 0.2769,
+      "step": 35560
+    },
+    {
+      "epoch": 53.49,
+      "grad_norm": 5.146740436553955,
+      "learning_rate": 4.651127819548873e-06,
+      "loss": 0.2152,
+      "step": 35570
+    },
+    {
+      "epoch": 53.5,
+      "grad_norm": 7.009411811828613,
+      "learning_rate": 4.649624060150376e-06,
+      "loss": 0.2264,
+      "step": 35580
+    },
+    {
+      "epoch": 53.52,
+      "grad_norm": 6.005060195922852,
+      "learning_rate": 4.64812030075188e-06,
+      "loss": 0.1889,
+      "step": 35590
+    },
+    {
+      "epoch": 53.53,
+      "grad_norm": 8.248251914978027,
+      "learning_rate": 4.646616541353383e-06,
+      "loss": 0.2329,
+      "step": 35600
+    },
+    {
+      "epoch": 53.55,
+      "grad_norm": 6.950926303863525,
+      "learning_rate": 4.645112781954888e-06,
+      "loss": 0.2437,
+      "step": 35610
+    },
+    {
+      "epoch": 53.56,
+      "grad_norm": 6.0409836769104,
+      "learning_rate": 4.643609022556391e-06,
+      "loss": 0.1981,
+      "step": 35620
+    },
+    {
+      "epoch": 53.58,
+      "grad_norm": 6.35344123840332,
+      "learning_rate": 4.642105263157895e-06,
+      "loss": 0.2629,
+      "step": 35630
+    },
+    {
+      "epoch": 53.59,
+      "grad_norm": 4.371424674987793,
+      "learning_rate": 4.640601503759399e-06,
+      "loss": 0.2816,
+      "step": 35640
+    },
+    {
+      "epoch": 53.61,
+      "grad_norm": 6.9506659507751465,
+      "learning_rate": 4.639097744360903e-06,
+      "loss": 0.2337,
+      "step": 35650
+    },
+    {
+      "epoch": 53.62,
+      "grad_norm": 4.818626403808594,
+      "learning_rate": 4.637593984962406e-06,
+      "loss": 0.1822,
+      "step": 35660
+    },
+    {
+      "epoch": 53.64,
+      "grad_norm": 5.629130840301514,
+      "learning_rate": 4.6360902255639105e-06,
+      "loss": 0.1786,
+      "step": 35670
+    },
+    {
+      "epoch": 53.65,
+      "grad_norm": 5.757498264312744,
+      "learning_rate": 4.634586466165414e-06,
+      "loss": 0.1709,
+      "step": 35680
+    },
+    {
+      "epoch": 53.67,
+      "grad_norm": 6.2977800369262695,
+      "learning_rate": 4.6330827067669175e-06,
+      "loss": 0.2847,
+      "step": 35690
+    },
+    {
+      "epoch": 53.68,
+      "grad_norm": 4.801835536956787,
+      "learning_rate": 4.631578947368421e-06,
+      "loss": 0.2441,
+      "step": 35700
+    },
+    {
+      "epoch": 53.7,
+      "grad_norm": 4.978909015655518,
+      "learning_rate": 4.630075187969925e-06,
+      "loss": 0.2422,
+      "step": 35710
+    },
+    {
+      "epoch": 53.71,
+      "grad_norm": 2.3563828468322754,
+      "learning_rate": 4.628571428571429e-06,
+      "loss": 0.2588,
+      "step": 35720
+    },
+    {
+      "epoch": 53.73,
+      "grad_norm": 5.150228977203369,
+      "learning_rate": 4.6270676691729324e-06,
+      "loss": 0.2531,
+      "step": 35730
+    },
+    {
+      "epoch": 53.74,
+      "grad_norm": 6.409964084625244,
+      "learning_rate": 4.625563909774437e-06,
+      "loss": 0.1561,
+      "step": 35740
+    },
+    {
+      "epoch": 53.76,
+      "grad_norm": 4.846547603607178,
+      "learning_rate": 4.62406015037594e-06,
+      "loss": 0.2398,
+      "step": 35750
+    },
+    {
+      "epoch": 53.77,
+      "grad_norm": 5.468544006347656,
+      "learning_rate": 4.622556390977444e-06,
+      "loss": 0.2449,
+      "step": 35760
+    },
+    {
+      "epoch": 53.79,
+      "grad_norm": 5.312195301055908,
+      "learning_rate": 4.621052631578948e-06,
+      "loss": 0.2073,
+      "step": 35770
+    },
+    {
+      "epoch": 53.8,
+      "grad_norm": 7.502721309661865,
+      "learning_rate": 4.619548872180452e-06,
+      "loss": 0.246,
+      "step": 35780
+    },
+    {
+      "epoch": 53.82,
+      "grad_norm": 3.2691874504089355,
+      "learning_rate": 4.618045112781955e-06,
+      "loss": 0.2128,
+      "step": 35790
+    },
+    {
+      "epoch": 53.83,
+      "grad_norm": 8.228704452514648,
+      "learning_rate": 4.616541353383459e-06,
+      "loss": 0.2276,
+      "step": 35800
+    },
+    {
+      "epoch": 53.85,
+      "grad_norm": 5.924352169036865,
+      "learning_rate": 4.615037593984963e-06,
+      "loss": 0.219,
+      "step": 35810
+    },
+    {
+      "epoch": 53.86,
+      "grad_norm": 3.2145702838897705,
+      "learning_rate": 4.613533834586467e-06,
+      "loss": 0.235,
+      "step": 35820
+    },
+    {
+      "epoch": 53.88,
+      "grad_norm": 8.739072799682617,
+      "learning_rate": 4.61203007518797e-06,
+      "loss": 0.1665,
+      "step": 35830
+    },
+    {
+      "epoch": 53.89,
+      "grad_norm": 3.9934237003326416,
+      "learning_rate": 4.6105263157894745e-06,
+      "loss": 0.1886,
+      "step": 35840
+    },
+    {
+      "epoch": 53.91,
+      "grad_norm": 6.60872745513916,
+      "learning_rate": 4.609022556390978e-06,
+      "loss": 0.2398,
+      "step": 35850
+    },
+    {
+      "epoch": 53.92,
+      "grad_norm": 7.813724994659424,
+      "learning_rate": 4.6075187969924816e-06,
+      "loss": 0.2371,
+      "step": 35860
+    },
+    {
+      "epoch": 53.94,
+      "grad_norm": 5.628241539001465,
+      "learning_rate": 4.606015037593985e-06,
+      "loss": 0.238,
+      "step": 35870
+    },
+    {
+      "epoch": 53.95,
+      "grad_norm": 4.604690074920654,
+      "learning_rate": 4.6045112781954894e-06,
+      "loss": 0.2523,
+      "step": 35880
+    },
+    {
+      "epoch": 53.97,
+      "grad_norm": 3.7630534172058105,
+      "learning_rate": 4.603007518796993e-06,
+      "loss": 0.2467,
+      "step": 35890
+    },
+    {
+      "epoch": 53.98,
+      "grad_norm": 5.215484142303467,
+      "learning_rate": 4.6015037593984965e-06,
+      "loss": 0.2155,
+      "step": 35900
+    },
+    {
+      "epoch": 54.0,
+      "grad_norm": 1.31098210811615,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 0.1876,
+      "step": 35910
+    },
+    {
+      "epoch": 54.0,
+      "eval_accuracy": 0.9318,
+      "eval_loss": 0.3097255229949951,
+      "eval_runtime": 84.7816,
+      "eval_samples_per_second": 117.95,
+      "eval_steps_per_second": 0.472,
+      "step": 35910
+    },
+    {
+      "epoch": 54.02,
+      "grad_norm": 1.5609022378921509,
+      "learning_rate": 4.598496240601504e-06,
+      "loss": 0.1557,
+      "step": 35920
+    },
+    {
+      "epoch": 54.03,
+      "grad_norm": 5.817564487457275,
+      "learning_rate": 4.596992481203008e-06,
+      "loss": 0.2565,
+      "step": 35930
+    },
+    {
+      "epoch": 54.05,
+      "grad_norm": 4.572742938995361,
+      "learning_rate": 4.595488721804512e-06,
+      "loss": 0.2119,
+      "step": 35940
+    },
+    {
+      "epoch": 54.06,
+      "grad_norm": 6.757458686828613,
+      "learning_rate": 4.593984962406016e-06,
+      "loss": 0.2227,
+      "step": 35950
+    },
+    {
+      "epoch": 54.08,
+      "grad_norm": 5.000956058502197,
+      "learning_rate": 4.592481203007519e-06,
+      "loss": 0.212,
+      "step": 35960
+    },
+    {
+      "epoch": 54.09,
+      "grad_norm": 9.407301902770996,
+      "learning_rate": 4.590977443609023e-06,
+      "loss": 0.2416,
+      "step": 35970
+    },
+    {
+      "epoch": 54.11,
+      "grad_norm": 5.10260534286499,
+      "learning_rate": 4.589473684210526e-06,
+      "loss": 0.2637,
+      "step": 35980
+    },
+    {
+      "epoch": 54.12,
+      "grad_norm": 6.334413528442383,
+      "learning_rate": 4.58796992481203e-06,
+      "loss": 0.2182,
+      "step": 35990
+    },
+    {
+      "epoch": 54.14,
+      "grad_norm": 3.6286683082580566,
+      "learning_rate": 4.586466165413534e-06,
+      "loss": 0.1799,
+      "step": 36000
+    },
+    {
+      "epoch": 54.15,
+      "grad_norm": 4.822414398193359,
+      "learning_rate": 4.584962406015038e-06,
+      "loss": 0.278,
+      "step": 36010
+    },
+    {
+      "epoch": 54.17,
+      "grad_norm": 9.361292839050293,
+      "learning_rate": 4.583458646616541e-06,
+      "loss": 0.2302,
+      "step": 36020
+    },
+    {
+      "epoch": 54.18,
+      "grad_norm": 5.843459129333496,
+      "learning_rate": 4.581954887218046e-06,
+      "loss": 0.2186,
+      "step": 36030
+    },
+    {
+      "epoch": 54.2,
+      "grad_norm": 3.1918785572052,
+      "learning_rate": 4.580451127819549e-06,
+      "loss": 0.2655,
+      "step": 36040
+    },
+    {
+      "epoch": 54.21,
+      "grad_norm": 5.561599254608154,
+      "learning_rate": 4.578947368421053e-06,
+      "loss": 0.2058,
+      "step": 36050
+    },
+    {
+      "epoch": 54.23,
+      "grad_norm": 2.6801440715789795,
+      "learning_rate": 4.577443609022556e-06,
+      "loss": 0.2265,
+      "step": 36060
+    },
+    {
+      "epoch": 54.24,
+      "grad_norm": 5.820893287658691,
+      "learning_rate": 4.5759398496240605e-06,
+      "loss": 0.2562,
+      "step": 36070
+    },
+    {
+      "epoch": 54.26,
+      "grad_norm": 5.376248836517334,
+      "learning_rate": 4.574436090225564e-06,
+      "loss": 0.233,
+      "step": 36080
+    },
+    {
+      "epoch": 54.27,
+      "grad_norm": 3.018950939178467,
+      "learning_rate": 4.572932330827068e-06,
+      "loss": 0.2185,
+      "step": 36090
+    },
+    {
+      "epoch": 54.29,
+      "grad_norm": 7.458950996398926,
+      "learning_rate": 4.571428571428572e-06,
+      "loss": 0.2014,
+      "step": 36100
+    },
+    {
+      "epoch": 54.3,
+      "grad_norm": 4.388574600219727,
+      "learning_rate": 4.5699248120300755e-06,
+      "loss": 0.2267,
+      "step": 36110
+    },
+    {
+      "epoch": 54.32,
+      "grad_norm": 4.121464252471924,
+      "learning_rate": 4.568421052631579e-06,
+      "loss": 0.3017,
+      "step": 36120
+    },
+    {
+      "epoch": 54.33,
+      "grad_norm": 3.9616527557373047,
+      "learning_rate": 4.566917293233083e-06,
+      "loss": 0.2357,
+      "step": 36130
+    },
+    {
+      "epoch": 54.35,
+      "grad_norm": 7.739973545074463,
+      "learning_rate": 4.565413533834587e-06,
+      "loss": 0.2267,
+      "step": 36140
+    },
+    {
+      "epoch": 54.36,
+      "grad_norm": 5.042054176330566,
+      "learning_rate": 4.56390977443609e-06,
+      "loss": 0.2686,
+      "step": 36150
+    },
+    {
+      "epoch": 54.38,
+      "grad_norm": 6.037046909332275,
+      "learning_rate": 4.562406015037594e-06,
+      "loss": 0.1939,
+      "step": 36160
+    },
+    {
+      "epoch": 54.39,
+      "grad_norm": 6.330253601074219,
+      "learning_rate": 4.560902255639098e-06,
+      "loss": 0.2152,
+      "step": 36170
+    },
+    {
+      "epoch": 54.41,
+      "grad_norm": 5.244846820831299,
+      "learning_rate": 4.559398496240602e-06,
+      "loss": 0.1767,
+      "step": 36180
+    },
+    {
+      "epoch": 54.42,
+      "grad_norm": 4.092870712280273,
+      "learning_rate": 4.557894736842105e-06,
+      "loss": 0.2417,
+      "step": 36190
+    },
+    {
+      "epoch": 54.44,
+      "grad_norm": 4.611246585845947,
+      "learning_rate": 4.55639097744361e-06,
+      "loss": 0.2388,
+      "step": 36200
+    },
+    {
+      "epoch": 54.45,
+      "grad_norm": 5.9267377853393555,
+      "learning_rate": 4.554887218045113e-06,
+      "loss": 0.2705,
+      "step": 36210
+    },
+    {
+      "epoch": 54.47,
+      "grad_norm": 3.914341926574707,
+      "learning_rate": 4.553383458646617e-06,
+      "loss": 0.2293,
+      "step": 36220
+    },
+    {
+      "epoch": 54.48,
+      "grad_norm": 6.353694438934326,
+      "learning_rate": 4.551879699248121e-06,
+      "loss": 0.2363,
+      "step": 36230
+    },
+    {
+      "epoch": 54.5,
+      "grad_norm": 5.587188720703125,
+      "learning_rate": 4.550375939849625e-06,
+      "loss": 0.2194,
+      "step": 36240
+    },
+    {
+      "epoch": 54.51,
+      "grad_norm": 7.160801887512207,
+      "learning_rate": 4.548872180451128e-06,
+      "loss": 0.2358,
+      "step": 36250
+    },
+    {
+      "epoch": 54.53,
+      "grad_norm": 4.975449085235596,
+      "learning_rate": 4.547368421052632e-06,
+      "loss": 0.2003,
+      "step": 36260
+    },
+    {
+      "epoch": 54.54,
+      "grad_norm": 5.509252071380615,
+      "learning_rate": 4.545864661654136e-06,
+      "loss": 0.2053,
+      "step": 36270
+    },
+    {
+      "epoch": 54.56,
+      "grad_norm": 6.666650772094727,
+      "learning_rate": 4.5443609022556395e-06,
+      "loss": 0.1576,
+      "step": 36280
+    },
+    {
+      "epoch": 54.57,
+      "grad_norm": 7.263132095336914,
+      "learning_rate": 4.542857142857143e-06,
+      "loss": 0.1459,
+      "step": 36290
+    },
+    {
+      "epoch": 54.59,
+      "grad_norm": 4.0889410972595215,
+      "learning_rate": 4.541353383458647e-06,
+      "loss": 0.1864,
+      "step": 36300
+    },
+    {
+      "epoch": 54.6,
+      "grad_norm": 5.318135738372803,
+      "learning_rate": 4.539849624060151e-06,
+      "loss": 0.258,
+      "step": 36310
+    },
+    {
+      "epoch": 54.62,
+      "grad_norm": 5.723468780517578,
+      "learning_rate": 4.5383458646616544e-06,
+      "loss": 0.2035,
+      "step": 36320
+    },
+    {
+      "epoch": 54.63,
+      "grad_norm": 5.231049060821533,
+      "learning_rate": 4.536842105263158e-06,
+      "loss": 0.1719,
+      "step": 36330
+    },
+    {
+      "epoch": 54.65,
+      "grad_norm": 5.921431541442871,
+      "learning_rate": 4.535338345864662e-06,
+      "loss": 0.3196,
+      "step": 36340
+    },
+    {
+      "epoch": 54.66,
+      "grad_norm": 5.602309703826904,
+      "learning_rate": 4.533834586466166e-06,
+      "loss": 0.2069,
+      "step": 36350
+    },
+    {
+      "epoch": 54.68,
+      "grad_norm": 8.991955757141113,
+      "learning_rate": 4.532330827067669e-06,
+      "loss": 0.2646,
+      "step": 36360
+    },
+    {
+      "epoch": 54.69,
+      "grad_norm": 3.8575797080993652,
+      "learning_rate": 4.530827067669174e-06,
+      "loss": 0.1887,
+      "step": 36370
+    },
+    {
+      "epoch": 54.71,
+      "grad_norm": 3.4165918827056885,
+      "learning_rate": 4.529323308270677e-06,
+      "loss": 0.1746,
+      "step": 36380
+    },
+    {
+      "epoch": 54.72,
+      "grad_norm": 5.644247531890869,
+      "learning_rate": 4.527819548872181e-06,
+      "loss": 0.2347,
+      "step": 36390
+    },
+    {
+      "epoch": 54.74,
+      "grad_norm": 6.519909381866455,
+      "learning_rate": 4.526315789473685e-06,
+      "loss": 0.232,
+      "step": 36400
+    },
+    {
+      "epoch": 54.75,
+      "grad_norm": 3.5762217044830322,
+      "learning_rate": 4.524812030075189e-06,
+      "loss": 0.2537,
+      "step": 36410
+    },
+    {
+      "epoch": 54.77,
+      "grad_norm": 4.4489898681640625,
+      "learning_rate": 4.523308270676692e-06,
+      "loss": 0.2041,
+      "step": 36420
+    },
+    {
+      "epoch": 54.78,
+      "grad_norm": 5.16048526763916,
+      "learning_rate": 4.521804511278196e-06,
+      "loss": 0.1966,
+      "step": 36430
+    },
+    {
+      "epoch": 54.8,
+      "grad_norm": 4.274946689605713,
+      "learning_rate": 4.5203007518797e-06,
+      "loss": 0.2945,
+      "step": 36440
+    },
+    {
+      "epoch": 54.81,
+      "grad_norm": 4.338449954986572,
+      "learning_rate": 4.518796992481204e-06,
+      "loss": 0.1972,
+      "step": 36450
+    },
+    {
+      "epoch": 54.83,
+      "grad_norm": 5.914998531341553,
+      "learning_rate": 4.517293233082707e-06,
+      "loss": 0.1903,
+      "step": 36460
+    },
+    {
+      "epoch": 54.84,
+      "grad_norm": 23.506519317626953,
+      "learning_rate": 4.5157894736842115e-06,
+      "loss": 0.2072,
+      "step": 36470
+    },
+    {
+      "epoch": 54.86,
+      "grad_norm": 5.0130743980407715,
+      "learning_rate": 4.514285714285714e-06,
+      "loss": 0.2302,
+      "step": 36480
+    },
+    {
+      "epoch": 54.87,
+      "grad_norm": 3.10636568069458,
+      "learning_rate": 4.5127819548872185e-06,
+      "loss": 0.1711,
+      "step": 36490
+    },
+    {
+      "epoch": 54.89,
+      "grad_norm": 6.681315898895264,
+      "learning_rate": 4.511278195488722e-06,
+      "loss": 0.2502,
+      "step": 36500
+    },
+    {
+      "epoch": 54.9,
+      "grad_norm": 5.002006530761719,
+      "learning_rate": 4.5097744360902255e-06,
+      "loss": 0.1971,
+      "step": 36510
+    },
+    {
+      "epoch": 54.92,
+      "grad_norm": 1.6818307638168335,
+      "learning_rate": 4.508270676691729e-06,
+      "loss": 0.1604,
+      "step": 36520
+    },
+    {
+      "epoch": 54.93,
+      "grad_norm": 5.976077556610107,
+      "learning_rate": 4.5067669172932334e-06,
+      "loss": 0.2416,
+      "step": 36530
+    },
+    {
+      "epoch": 54.95,
+      "grad_norm": 4.581597805023193,
+      "learning_rate": 4.505263157894737e-06,
+      "loss": 0.2433,
+      "step": 36540
+    },
+    {
+      "epoch": 54.96,
+      "grad_norm": 6.989867210388184,
+      "learning_rate": 4.5037593984962405e-06,
+      "loss": 0.2687,
+      "step": 36550
+    },
+    {
+      "epoch": 54.98,
+      "grad_norm": 8.435478210449219,
+      "learning_rate": 4.502255639097745e-06,
+      "loss": 0.2427,
+      "step": 36560
+    },
+    {
+      "epoch": 54.99,
+      "grad_norm": 5.939743518829346,
+      "learning_rate": 4.500751879699248e-06,
+      "loss": 0.1774,
+      "step": 36570
+    },
+    {
+      "epoch": 55.0,
+      "eval_accuracy": 0.9321,
+      "eval_loss": 0.3104659616947174,
+      "eval_runtime": 84.6339,
+      "eval_samples_per_second": 118.156,
+      "eval_steps_per_second": 0.473,
+      "step": 36575
+    },
+    {
+      "epoch": 55.01,
+      "grad_norm": 4.729255676269531,
+      "learning_rate": 4.499248120300752e-06,
+      "loss": 0.4104,
+      "step": 36580
+    },
+    {
+      "epoch": 55.02,
+      "grad_norm": 4.213466644287109,
+      "learning_rate": 4.497744360902256e-06,
+      "loss": 0.24,
+      "step": 36590
+    },
+    {
+      "epoch": 55.04,
+      "grad_norm": 2.702681064605713,
+      "learning_rate": 4.49624060150376e-06,
+      "loss": 0.1929,
+      "step": 36600
+    },
+    {
+      "epoch": 55.05,
+      "grad_norm": 5.444637775421143,
+      "learning_rate": 4.494736842105263e-06,
+      "loss": 0.2168,
+      "step": 36610
+    },
+    {
+      "epoch": 55.07,
+      "grad_norm": 8.909059524536133,
+      "learning_rate": 4.493233082706767e-06,
+      "loss": 0.2456,
+      "step": 36620
+    },
+    {
+      "epoch": 55.08,
+      "grad_norm": 3.999340772628784,
+      "learning_rate": 4.491729323308271e-06,
+      "loss": 0.1867,
+      "step": 36630
+    },
+    {
+      "epoch": 55.1,
+      "grad_norm": 4.647693157196045,
+      "learning_rate": 4.490225563909775e-06,
+      "loss": 0.2448,
+      "step": 36640
+    },
+    {
+      "epoch": 55.11,
+      "grad_norm": 4.264559745788574,
+      "learning_rate": 4.488721804511278e-06,
+      "loss": 0.1637,
+      "step": 36650
+    },
+    {
+      "epoch": 55.13,
+      "grad_norm": 2.194603204727173,
+      "learning_rate": 4.4872180451127826e-06,
+      "loss": 0.2602,
+      "step": 36660
+    },
+    {
+      "epoch": 55.14,
+      "grad_norm": 3.7559118270874023,
+      "learning_rate": 4.485714285714286e-06,
+      "loss": 0.2043,
+      "step": 36670
+    },
+    {
+      "epoch": 55.16,
+      "grad_norm": 3.0917916297912598,
+      "learning_rate": 4.48421052631579e-06,
+      "loss": 0.1839,
+      "step": 36680
+    },
+    {
+      "epoch": 55.17,
+      "grad_norm": 4.948066234588623,
+      "learning_rate": 4.482706766917294e-06,
+      "loss": 0.2476,
+      "step": 36690
+    },
+    {
+      "epoch": 55.19,
+      "grad_norm": 5.164941310882568,
+      "learning_rate": 4.4812030075187975e-06,
+      "loss": 0.1977,
+      "step": 36700
+    },
+    {
+      "epoch": 55.2,
+      "grad_norm": 4.678724765777588,
+      "learning_rate": 4.479699248120301e-06,
+      "loss": 0.2576,
+      "step": 36710
+    },
+    {
+      "epoch": 55.22,
+      "grad_norm": 8.130949020385742,
+      "learning_rate": 4.4781954887218045e-06,
+      "loss": 0.232,
+      "step": 36720
+    },
+    {
+      "epoch": 55.23,
+      "grad_norm": 4.381587505340576,
+      "learning_rate": 4.476691729323309e-06,
+      "loss": 0.2102,
+      "step": 36730
+    },
+    {
+      "epoch": 55.25,
+      "grad_norm": 3.894073724746704,
+      "learning_rate": 4.475187969924812e-06,
+      "loss": 0.2271,
+      "step": 36740
+    },
+    {
+      "epoch": 55.26,
+      "grad_norm": 7.002061367034912,
+      "learning_rate": 4.473684210526316e-06,
+      "loss": 0.1887,
+      "step": 36750
+    },
+    {
+      "epoch": 55.28,
+      "grad_norm": 5.8444132804870605,
+      "learning_rate": 4.47218045112782e-06,
+      "loss": 0.2588,
+      "step": 36760
+    },
+    {
+      "epoch": 55.29,
+      "grad_norm": 3.5753517150878906,
+      "learning_rate": 4.470676691729324e-06,
+      "loss": 0.2312,
+      "step": 36770
+    },
+    {
+      "epoch": 55.31,
+      "grad_norm": 7.695021629333496,
+      "learning_rate": 4.469172932330827e-06,
+      "loss": 0.2307,
+      "step": 36780
+    },
+    {
+      "epoch": 55.32,
+      "grad_norm": 4.55169153213501,
+      "learning_rate": 4.467669172932331e-06,
+      "loss": 0.2242,
+      "step": 36790
+    },
+    {
+      "epoch": 55.34,
+      "grad_norm": 2.349003553390503,
+      "learning_rate": 4.466165413533835e-06,
+      "loss": 0.2035,
+      "step": 36800
+    },
+    {
+      "epoch": 55.35,
+      "grad_norm": 5.827582359313965,
+      "learning_rate": 4.464661654135339e-06,
+      "loss": 0.2916,
+      "step": 36810
+    },
+    {
+      "epoch": 55.37,
+      "grad_norm": 3.1232826709747314,
+      "learning_rate": 4.463157894736842e-06,
+      "loss": 0.231,
+      "step": 36820
+    },
+    {
+      "epoch": 55.38,
+      "grad_norm": 6.336644172668457,
+      "learning_rate": 4.461654135338347e-06,
+      "loss": 0.2344,
+      "step": 36830
+    },
+    {
+      "epoch": 55.4,
+      "grad_norm": 4.629255771636963,
+      "learning_rate": 4.46015037593985e-06,
+      "loss": 0.2577,
+      "step": 36840
+    },
+    {
+      "epoch": 55.41,
+      "grad_norm": 5.478154182434082,
+      "learning_rate": 4.458646616541354e-06,
+      "loss": 0.2311,
+      "step": 36850
+    },
+    {
+      "epoch": 55.43,
+      "grad_norm": 7.846320629119873,
+      "learning_rate": 4.457142857142858e-06,
+      "loss": 0.247,
+      "step": 36860
+    },
+    {
+      "epoch": 55.44,
+      "grad_norm": 3.461475133895874,
+      "learning_rate": 4.4556390977443615e-06,
+      "loss": 0.1488,
+      "step": 36870
+    },
+    {
+      "epoch": 55.46,
+      "grad_norm": 3.4155285358428955,
+      "learning_rate": 4.454135338345865e-06,
+      "loss": 0.1368,
+      "step": 36880
+    },
+    {
+      "epoch": 55.47,
+      "grad_norm": 2.114795684814453,
+      "learning_rate": 4.452631578947369e-06,
+      "loss": 0.1632,
+      "step": 36890
+    },
+    {
+      "epoch": 55.49,
+      "grad_norm": 1.200542688369751,
+      "learning_rate": 4.451127819548873e-06,
+      "loss": 0.1591,
+      "step": 36900
+    },
+    {
+      "epoch": 55.5,
+      "grad_norm": 4.999506950378418,
+      "learning_rate": 4.4496240601503765e-06,
+      "loss": 0.1666,
+      "step": 36910
+    },
+    {
+      "epoch": 55.52,
+      "grad_norm": 3.8099682331085205,
+      "learning_rate": 4.44812030075188e-06,
+      "loss": 0.1601,
+      "step": 36920
+    },
+    {
+      "epoch": 55.53,
+      "grad_norm": 7.439540386199951,
+      "learning_rate": 4.446616541353384e-06,
+      "loss": 0.2179,
+      "step": 36930
+    },
+    {
+      "epoch": 55.55,
+      "grad_norm": 4.367668151855469,
+      "learning_rate": 4.445112781954888e-06,
+      "loss": 0.1956,
+      "step": 36940
+    },
+    {
+      "epoch": 55.56,
+      "grad_norm": 5.904849529266357,
+      "learning_rate": 4.443609022556391e-06,
+      "loss": 0.2132,
+      "step": 36950
+    },
+    {
+      "epoch": 55.58,
+      "grad_norm": 9.836445808410645,
+      "learning_rate": 4.442105263157896e-06,
+      "loss": 0.2255,
+      "step": 36960
+    },
+    {
+      "epoch": 55.59,
+      "grad_norm": 2.742265224456787,
+      "learning_rate": 4.440601503759399e-06,
+      "loss": 0.2005,
+      "step": 36970
+    },
+    {
+      "epoch": 55.61,
+      "grad_norm": 6.371535778045654,
+      "learning_rate": 4.439097744360902e-06,
+      "loss": 0.1944,
+      "step": 36980
+    },
+    {
+      "epoch": 55.62,
+      "grad_norm": 3.961942195892334,
+      "learning_rate": 4.437593984962406e-06,
+      "loss": 0.1885,
+      "step": 36990
+    },
+    {
+      "epoch": 55.64,
+      "grad_norm": 7.4699320793151855,
+      "learning_rate": 4.43609022556391e-06,
+      "loss": 0.2383,
+      "step": 37000
+    },
+    {
+      "epoch": 55.65,
+      "grad_norm": 2.479861259460449,
+      "learning_rate": 4.434586466165413e-06,
+      "loss": 0.2797,
+      "step": 37010
+    },
+    {
+      "epoch": 55.67,
+      "grad_norm": 4.376903057098389,
+      "learning_rate": 4.433082706766918e-06,
+      "loss": 0.1898,
+      "step": 37020
+    },
+    {
+      "epoch": 55.68,
+      "grad_norm": 5.335219860076904,
+      "learning_rate": 4.431578947368421e-06,
+      "loss": 0.2382,
+      "step": 37030
+    },
+    {
+      "epoch": 55.7,
+      "grad_norm": 4.678366184234619,
+      "learning_rate": 4.430075187969925e-06,
+      "loss": 0.2029,
+      "step": 37040
+    },
+    {
+      "epoch": 55.71,
+      "grad_norm": 5.352789878845215,
+      "learning_rate": 4.428571428571429e-06,
+      "loss": 0.2286,
+      "step": 37050
+    },
+    {
+      "epoch": 55.73,
+      "grad_norm": 5.189038276672363,
+      "learning_rate": 4.427067669172933e-06,
+      "loss": 0.1941,
+      "step": 37060
+    },
+    {
+      "epoch": 55.74,
+      "grad_norm": 10.683388710021973,
+      "learning_rate": 4.425563909774436e-06,
+      "loss": 0.195,
+      "step": 37070
+    },
+    {
+      "epoch": 55.76,
+      "grad_norm": 6.178273677825928,
+      "learning_rate": 4.42406015037594e-06,
+      "loss": 0.2207,
+      "step": 37080
+    },
+    {
+      "epoch": 55.77,
+      "grad_norm": 3.819103479385376,
+      "learning_rate": 4.422556390977444e-06,
+      "loss": 0.2128,
+      "step": 37090
+    },
+    {
+      "epoch": 55.79,
+      "grad_norm": 6.3040452003479,
+      "learning_rate": 4.4210526315789476e-06,
+      "loss": 0.1738,
+      "step": 37100
+    },
+    {
+      "epoch": 55.8,
+      "grad_norm": 5.171235084533691,
+      "learning_rate": 4.419548872180451e-06,
+      "loss": 0.2041,
+      "step": 37110
+    },
+    {
+      "epoch": 55.82,
+      "grad_norm": 5.310617446899414,
+      "learning_rate": 4.4180451127819555e-06,
+      "loss": 0.2201,
+      "step": 37120
+    },
+    {
+      "epoch": 55.83,
+      "grad_norm": 7.987985610961914,
+      "learning_rate": 4.416541353383459e-06,
+      "loss": 0.1631,
+      "step": 37130
+    },
+    {
+      "epoch": 55.85,
+      "grad_norm": 3.634016275405884,
+      "learning_rate": 4.4150375939849625e-06,
+      "loss": 0.248,
+      "step": 37140
+    },
+    {
+      "epoch": 55.86,
+      "grad_norm": 7.253101348876953,
+      "learning_rate": 4.413533834586467e-06,
+      "loss": 0.2631,
+      "step": 37150
+    },
+    {
+      "epoch": 55.88,
+      "grad_norm": 5.290616989135742,
+      "learning_rate": 4.41203007518797e-06,
+      "loss": 0.1595,
+      "step": 37160
+    },
+    {
+      "epoch": 55.89,
+      "grad_norm": 7.694485187530518,
+      "learning_rate": 4.410526315789474e-06,
+      "loss": 0.2274,
+      "step": 37170
+    },
+    {
+      "epoch": 55.91,
+      "grad_norm": 4.015926837921143,
+      "learning_rate": 4.409022556390977e-06,
+      "loss": 0.1889,
+      "step": 37180
+    },
+    {
+      "epoch": 55.92,
+      "grad_norm": 6.907248020172119,
+      "learning_rate": 4.407518796992482e-06,
+      "loss": 0.2389,
+      "step": 37190
+    },
+    {
+      "epoch": 55.94,
+      "grad_norm": 9.580093383789062,
+      "learning_rate": 4.406015037593985e-06,
+      "loss": 0.1979,
+      "step": 37200
+    },
+    {
+      "epoch": 55.95,
+      "grad_norm": 5.89567756652832,
+      "learning_rate": 4.404511278195489e-06,
+      "loss": 0.2167,
+      "step": 37210
+    },
+    {
+      "epoch": 55.97,
+      "grad_norm": 6.980398178100586,
+      "learning_rate": 4.403007518796993e-06,
+      "loss": 0.2273,
+      "step": 37220
+    },
+    {
+      "epoch": 55.98,
+      "grad_norm": 4.1732869148254395,
+      "learning_rate": 4.401503759398497e-06,
+      "loss": 0.217,
+      "step": 37230
+    },
+    {
+      "epoch": 56.0,
+      "grad_norm": 0.04293264448642731,
+      "learning_rate": 4.4e-06,
+      "loss": 0.2011,
+      "step": 37240
+    },
+    {
+      "epoch": 56.0,
+      "eval_accuracy": 0.9337,
+      "eval_loss": 0.3107641935348511,
+      "eval_runtime": 84.6401,
+      "eval_samples_per_second": 118.147,
+      "eval_steps_per_second": 0.473,
+      "step": 37240
+    },
+    {
+      "epoch": 56.02,
+      "grad_norm": 2.9971933364868164,
+      "learning_rate": 4.398496240601504e-06,
+      "loss": 0.1546,
+      "step": 37250
+    },
+    {
+      "epoch": 56.03,
+      "grad_norm": 4.23447322845459,
+      "learning_rate": 4.396992481203008e-06,
+      "loss": 0.2525,
+      "step": 37260
+    },
+    {
+      "epoch": 56.05,
+      "grad_norm": 2.2622056007385254,
+      "learning_rate": 4.395488721804512e-06,
+      "loss": 0.192,
+      "step": 37270
+    },
+    {
+      "epoch": 56.06,
+      "grad_norm": 6.475707054138184,
+      "learning_rate": 4.393984962406015e-06,
+      "loss": 0.2287,
+      "step": 37280
+    },
+    {
+      "epoch": 56.08,
+      "grad_norm": 5.559025764465332,
+      "learning_rate": 4.3924812030075195e-06,
+      "loss": 0.2248,
+      "step": 37290
+    },
+    {
+      "epoch": 56.09,
+      "grad_norm": 8.06891918182373,
+      "learning_rate": 4.390977443609023e-06,
+      "loss": 0.1641,
+      "step": 37300
+    },
+    {
+      "epoch": 56.11,
+      "grad_norm": 3.4583325386047363,
+      "learning_rate": 4.3894736842105266e-06,
+      "loss": 0.231,
+      "step": 37310
+    },
+    {
+      "epoch": 56.12,
+      "grad_norm": 11.369089126586914,
+      "learning_rate": 4.387969924812031e-06,
+      "loss": 0.2368,
+      "step": 37320
+    },
+    {
+      "epoch": 56.14,
+      "grad_norm": 3.740903854370117,
+      "learning_rate": 4.3864661654135344e-06,
+      "loss": 0.1868,
+      "step": 37330
+    },
+    {
+      "epoch": 56.15,
+      "grad_norm": 7.389920234680176,
+      "learning_rate": 4.384962406015038e-06,
+      "loss": 0.2224,
+      "step": 37340
+    },
+    {
+      "epoch": 56.17,
+      "grad_norm": 4.443047046661377,
+      "learning_rate": 4.3834586466165415e-06,
+      "loss": 0.2308,
+      "step": 37350
+    },
+    {
+      "epoch": 56.18,
+      "grad_norm": 5.819294452667236,
+      "learning_rate": 4.381954887218046e-06,
+      "loss": 0.1937,
+      "step": 37360
+    },
+    {
+      "epoch": 56.2,
+      "grad_norm": 4.005828857421875,
+      "learning_rate": 4.380451127819549e-06,
+      "loss": 0.2269,
+      "step": 37370
+    },
+    {
+      "epoch": 56.21,
+      "grad_norm": 9.527501106262207,
+      "learning_rate": 4.378947368421053e-06,
+      "loss": 0.2721,
+      "step": 37380
+    },
+    {
+      "epoch": 56.23,
+      "grad_norm": 5.382596492767334,
+      "learning_rate": 4.377443609022557e-06,
+      "loss": 0.2143,
+      "step": 37390
+    },
+    {
+      "epoch": 56.24,
+      "grad_norm": 8.085970878601074,
+      "learning_rate": 4.375939849624061e-06,
+      "loss": 0.1549,
+      "step": 37400
+    },
+    {
+      "epoch": 56.26,
+      "grad_norm": 7.813641548156738,
+      "learning_rate": 4.374436090225564e-06,
+      "loss": 0.1926,
+      "step": 37410
+    },
+    {
+      "epoch": 56.27,
+      "grad_norm": 5.974503993988037,
+      "learning_rate": 4.372932330827069e-06,
+      "loss": 0.2217,
+      "step": 37420
+    },
+    {
+      "epoch": 56.29,
+      "grad_norm": 3.7494096755981445,
+      "learning_rate": 4.371428571428572e-06,
+      "loss": 0.1835,
+      "step": 37430
+    },
+    {
+      "epoch": 56.3,
+      "grad_norm": 8.034648895263672,
+      "learning_rate": 4.369924812030076e-06,
+      "loss": 0.2319,
+      "step": 37440
+    },
+    {
+      "epoch": 56.32,
+      "grad_norm": 5.222734451293945,
+      "learning_rate": 4.368421052631579e-06,
+      "loss": 0.2468,
+      "step": 37450
+    },
+    {
+      "epoch": 56.33,
+      "grad_norm": 2.54164981842041,
+      "learning_rate": 4.3669172932330836e-06,
+      "loss": 0.1801,
+      "step": 37460
+    },
+    {
+      "epoch": 56.35,
+      "grad_norm": 8.72768783569336,
+      "learning_rate": 4.365413533834586e-06,
+      "loss": 0.2092,
+      "step": 37470
+    },
+    {
+      "epoch": 56.36,
+      "grad_norm": 10.628849983215332,
+      "learning_rate": 4.363909774436091e-06,
+      "loss": 0.2025,
+      "step": 37480
+    },
+    {
+      "epoch": 56.38,
+      "grad_norm": 5.263908863067627,
+      "learning_rate": 4.362406015037594e-06,
+      "loss": 0.1637,
+      "step": 37490
+    },
+    {
+      "epoch": 56.39,
+      "grad_norm": 3.4017181396484375,
+      "learning_rate": 4.360902255639098e-06,
+      "loss": 0.2772,
+      "step": 37500
+    },
+    {
+      "epoch": 56.41,
+      "grad_norm": 6.786311626434326,
+      "learning_rate": 4.359398496240602e-06,
+      "loss": 0.1987,
+      "step": 37510
+    },
+    {
+      "epoch": 56.42,
+      "grad_norm": 6.281602382659912,
+      "learning_rate": 4.3578947368421055e-06,
+      "loss": 0.1707,
+      "step": 37520
+    },
+    {
+      "epoch": 56.44,
+      "grad_norm": 6.129687786102295,
+      "learning_rate": 4.356390977443609e-06,
+      "loss": 0.2591,
+      "step": 37530
+    },
+    {
+      "epoch": 56.45,
+      "grad_norm": 4.573347568511963,
+      "learning_rate": 4.3548872180451126e-06,
+      "loss": 0.2142,
+      "step": 37540
+    },
+    {
+      "epoch": 56.47,
+      "grad_norm": 4.783934593200684,
+      "learning_rate": 4.353383458646617e-06,
+      "loss": 0.1747,
+      "step": 37550
+    },
+    {
+      "epoch": 56.48,
+      "grad_norm": 3.693443775177002,
+      "learning_rate": 4.3518796992481205e-06,
+      "loss": 0.2222,
+      "step": 37560
+    },
+    {
+      "epoch": 56.5,
+      "grad_norm": 5.919257640838623,
+      "learning_rate": 4.350375939849624e-06,
+      "loss": 0.2443,
+      "step": 37570
+    },
+    {
+      "epoch": 56.51,
+      "grad_norm": 9.530694007873535,
+      "learning_rate": 4.348872180451128e-06,
+      "loss": 0.1908,
+      "step": 37580
+    },
+    {
+      "epoch": 56.53,
+      "grad_norm": 5.846238613128662,
+      "learning_rate": 4.347368421052632e-06,
+      "loss": 0.2141,
+      "step": 37590
+    },
+    {
+      "epoch": 56.54,
+      "grad_norm": 5.0612616539001465,
+      "learning_rate": 4.345864661654135e-06,
+      "loss": 0.1802,
+      "step": 37600
+    },
+    {
+      "epoch": 56.56,
+      "grad_norm": 3.5078976154327393,
+      "learning_rate": 4.34436090225564e-06,
+      "loss": 0.1856,
+      "step": 37610
+    },
+    {
+      "epoch": 56.57,
+      "grad_norm": 2.6494715213775635,
+      "learning_rate": 4.342857142857143e-06,
+      "loss": 0.2228,
+      "step": 37620
+    },
+    {
+      "epoch": 56.59,
+      "grad_norm": 7.06261682510376,
+      "learning_rate": 4.341353383458647e-06,
+      "loss": 0.1996,
+      "step": 37630
+    },
+    {
+      "epoch": 56.6,
+      "grad_norm": 4.200351238250732,
+      "learning_rate": 4.33984962406015e-06,
+      "loss": 0.1868,
+      "step": 37640
+    },
+    {
+      "epoch": 56.62,
+      "grad_norm": 3.3782460689544678,
+      "learning_rate": 4.338345864661655e-06,
+      "loss": 0.1933,
+      "step": 37650
+    },
+    {
+      "epoch": 56.63,
+      "grad_norm": 5.649930953979492,
+      "learning_rate": 4.336842105263158e-06,
+      "loss": 0.1826,
+      "step": 37660
+    },
+    {
+      "epoch": 56.65,
+      "grad_norm": 5.642304420471191,
+      "learning_rate": 4.335338345864662e-06,
+      "loss": 0.232,
+      "step": 37670
+    },
+    {
+      "epoch": 56.66,
+      "grad_norm": 4.194146156311035,
+      "learning_rate": 4.333834586466166e-06,
+      "loss": 0.232,
+      "step": 37680
+    },
+    {
+      "epoch": 56.68,
+      "grad_norm": 6.466404438018799,
+      "learning_rate": 4.33233082706767e-06,
+      "loss": 0.184,
+      "step": 37690
+    },
+    {
+      "epoch": 56.69,
+      "grad_norm": 6.861640930175781,
+      "learning_rate": 4.330827067669173e-06,
+      "loss": 0.2036,
+      "step": 37700
+    },
+    {
+      "epoch": 56.71,
+      "grad_norm": 5.0166850090026855,
+      "learning_rate": 4.329323308270677e-06,
+      "loss": 0.2537,
+      "step": 37710
+    },
+    {
+      "epoch": 56.72,
+      "grad_norm": 6.62353515625,
+      "learning_rate": 4.327819548872181e-06,
+      "loss": 0.2708,
+      "step": 37720
+    },
+    {
+      "epoch": 56.74,
+      "grad_norm": 5.707888603210449,
+      "learning_rate": 4.3263157894736845e-06,
+      "loss": 0.162,
+      "step": 37730
+    },
+    {
+      "epoch": 56.75,
+      "grad_norm": 6.166172981262207,
+      "learning_rate": 4.324812030075188e-06,
+      "loss": 0.2149,
+      "step": 37740
+    },
+    {
+      "epoch": 56.77,
+      "grad_norm": 8.481252670288086,
+      "learning_rate": 4.323308270676692e-06,
+      "loss": 0.2423,
+      "step": 37750
+    },
+    {
+      "epoch": 56.78,
+      "grad_norm": 1.8157340288162231,
+      "learning_rate": 4.321804511278196e-06,
+      "loss": 0.2519,
+      "step": 37760
+    },
+    {
+      "epoch": 56.8,
+      "grad_norm": 6.047143459320068,
+      "learning_rate": 4.3203007518796994e-06,
+      "loss": 0.3102,
+      "step": 37770
+    },
+    {
+      "epoch": 56.81,
+      "grad_norm": 3.9835569858551025,
+      "learning_rate": 4.318796992481204e-06,
+      "loss": 0.2141,
+      "step": 37780
+    },
+    {
+      "epoch": 56.83,
+      "grad_norm": 3.2040421962738037,
+      "learning_rate": 4.317293233082707e-06,
+      "loss": 0.1938,
+      "step": 37790
+    },
+    {
+      "epoch": 56.84,
+      "grad_norm": 3.5100557804107666,
+      "learning_rate": 4.315789473684211e-06,
+      "loss": 0.1956,
+      "step": 37800
+    },
+    {
+      "epoch": 56.86,
+      "grad_norm": 12.860550880432129,
+      "learning_rate": 4.314285714285714e-06,
+      "loss": 0.2704,
+      "step": 37810
+    },
+    {
+      "epoch": 56.87,
+      "grad_norm": 4.295982360839844,
+      "learning_rate": 4.312781954887219e-06,
+      "loss": 0.2478,
+      "step": 37820
+    },
+    {
+      "epoch": 56.89,
+      "grad_norm": 5.491456031799316,
+      "learning_rate": 4.311278195488722e-06,
+      "loss": 0.2482,
+      "step": 37830
+    },
+    {
+      "epoch": 56.9,
+      "grad_norm": 3.705702781677246,
+      "learning_rate": 4.309774436090226e-06,
+      "loss": 0.189,
+      "step": 37840
+    },
+    {
+      "epoch": 56.92,
+      "grad_norm": 9.076353073120117,
+      "learning_rate": 4.30827067669173e-06,
+      "loss": 0.2324,
+      "step": 37850
+    },
+    {
+      "epoch": 56.93,
+      "grad_norm": 5.757368564605713,
+      "learning_rate": 4.306766917293234e-06,
+      "loss": 0.2676,
+      "step": 37860
+    },
+    {
+      "epoch": 56.95,
+      "grad_norm": 5.485102653503418,
+      "learning_rate": 4.305263157894737e-06,
+      "loss": 0.1747,
+      "step": 37870
+    },
+    {
+      "epoch": 56.96,
+      "grad_norm": 3.859330892562866,
+      "learning_rate": 4.3037593984962415e-06,
+      "loss": 0.2029,
+      "step": 37880
+    },
+    {
+      "epoch": 56.98,
+      "grad_norm": 2.6711108684539795,
+      "learning_rate": 4.302255639097745e-06,
+      "loss": 0.2083,
+      "step": 37890
+    },
+    {
+      "epoch": 56.99,
+      "grad_norm": 5.341872692108154,
+      "learning_rate": 4.3007518796992486e-06,
+      "loss": 0.2142,
+      "step": 37900
+    },
+    {
+      "epoch": 57.0,
+      "eval_accuracy": 0.9312,
+      "eval_loss": 0.3190965950489044,
+      "eval_runtime": 84.8921,
+      "eval_samples_per_second": 117.797,
+      "eval_steps_per_second": 0.471,
+      "step": 37905
+    },
+    {
+      "epoch": 57.01,
+      "grad_norm": 3.6653482913970947,
+      "learning_rate": 4.299248120300752e-06,
+      "loss": 0.2008,
+      "step": 37910
+    },
+    {
+      "epoch": 57.02,
+      "grad_norm": 3.9756722450256348,
+      "learning_rate": 4.2977443609022565e-06,
+      "loss": 0.1872,
+      "step": 37920
+    },
+    {
+      "epoch": 57.04,
+      "grad_norm": 9.079606056213379,
+      "learning_rate": 4.29624060150376e-06,
+      "loss": 0.1913,
+      "step": 37930
+    },
+    {
+      "epoch": 57.05,
+      "grad_norm": 5.466242790222168,
+      "learning_rate": 4.2947368421052635e-06,
+      "loss": 0.2364,
+      "step": 37940
+    },
+    {
+      "epoch": 57.07,
+      "grad_norm": 1.4235931634902954,
+      "learning_rate": 4.293233082706768e-06,
+      "loss": 0.2255,
+      "step": 37950
+    },
+    {
+      "epoch": 57.08,
+      "grad_norm": 13.605971336364746,
+      "learning_rate": 4.291729323308271e-06,
+      "loss": 0.2454,
+      "step": 37960
+    },
+    {
+      "epoch": 57.1,
+      "grad_norm": 5.704050064086914,
+      "learning_rate": 4.290225563909775e-06,
+      "loss": 0.2411,
+      "step": 37970
+    },
+    {
+      "epoch": 57.11,
+      "grad_norm": 4.244338035583496,
+      "learning_rate": 4.288721804511278e-06,
+      "loss": 0.2289,
+      "step": 37980
+    },
+    {
+      "epoch": 57.13,
+      "grad_norm": 5.691931247711182,
+      "learning_rate": 4.287218045112782e-06,
+      "loss": 0.2031,
+      "step": 37990
+    },
+    {
+      "epoch": 57.14,
+      "grad_norm": 8.121678352355957,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 0.2434,
+      "step": 38000
+    },
+    {
+      "epoch": 57.16,
+      "grad_norm": 2.5804800987243652,
+      "learning_rate": 4.28421052631579e-06,
+      "loss": 0.1663,
+      "step": 38010
+    },
+    {
+      "epoch": 57.17,
+      "grad_norm": 6.460157871246338,
+      "learning_rate": 4.282706766917293e-06,
+      "loss": 0.1966,
+      "step": 38020
+    },
+    {
+      "epoch": 57.19,
+      "grad_norm": 6.010727405548096,
+      "learning_rate": 4.281203007518797e-06,
+      "loss": 0.2547,
+      "step": 38030
+    },
+    {
+      "epoch": 57.2,
+      "grad_norm": 3.152200937271118,
+      "learning_rate": 4.279699248120301e-06,
+      "loss": 0.1743,
+      "step": 38040
+    },
+    {
+      "epoch": 57.22,
+      "grad_norm": 6.710914134979248,
+      "learning_rate": 4.278195488721805e-06,
+      "loss": 0.2268,
+      "step": 38050
+    },
+    {
+      "epoch": 57.23,
+      "grad_norm": 2.6717779636383057,
+      "learning_rate": 4.276691729323308e-06,
+      "loss": 0.1521,
+      "step": 38060
+    },
+    {
+      "epoch": 57.25,
+      "grad_norm": 5.592037200927734,
+      "learning_rate": 4.275187969924813e-06,
+      "loss": 0.2159,
+      "step": 38070
+    },
+    {
+      "epoch": 57.26,
+      "grad_norm": 3.0837130546569824,
+      "learning_rate": 4.273684210526316e-06,
+      "loss": 0.2232,
+      "step": 38080
+    },
+    {
+      "epoch": 57.28,
+      "grad_norm": 5.526956558227539,
+      "learning_rate": 4.27218045112782e-06,
+      "loss": 0.1866,
+      "step": 38090
+    },
+    {
+      "epoch": 57.29,
+      "grad_norm": 5.77144718170166,
+      "learning_rate": 4.270676691729323e-06,
+      "loss": 0.2533,
+      "step": 38100
+    },
+    {
+      "epoch": 57.31,
+      "grad_norm": 3.7679331302642822,
+      "learning_rate": 4.2691729323308276e-06,
+      "loss": 0.2217,
+      "step": 38110
+    },
+    {
+      "epoch": 57.32,
+      "grad_norm": 19.891773223876953,
+      "learning_rate": 4.267669172932331e-06,
+      "loss": 0.1851,
+      "step": 38120
+    },
+    {
+      "epoch": 57.34,
+      "grad_norm": 8.34205436706543,
+      "learning_rate": 4.266165413533835e-06,
+      "loss": 0.2048,
+      "step": 38130
+    },
+    {
+      "epoch": 57.35,
+      "grad_norm": 5.428603172302246,
+      "learning_rate": 4.264661654135339e-06,
+      "loss": 0.2122,
+      "step": 38140
+    },
+    {
+      "epoch": 57.37,
+      "grad_norm": 3.5390725135803223,
+      "learning_rate": 4.2631578947368425e-06,
+      "loss": 0.2041,
+      "step": 38150
+    },
+    {
+      "epoch": 57.38,
+      "grad_norm": 4.302058219909668,
+      "learning_rate": 4.261654135338346e-06,
+      "loss": 0.1994,
+      "step": 38160
+    },
+    {
+      "epoch": 57.4,
+      "grad_norm": 16.373668670654297,
+      "learning_rate": 4.2601503759398495e-06,
+      "loss": 0.2482,
+      "step": 38170
+    },
+    {
+      "epoch": 57.41,
+      "grad_norm": 3.5577502250671387,
+      "learning_rate": 4.258646616541354e-06,
+      "loss": 0.1791,
+      "step": 38180
+    },
+    {
+      "epoch": 57.43,
+      "grad_norm": 3.964195728302002,
+      "learning_rate": 4.257142857142857e-06,
+      "loss": 0.2346,
+      "step": 38190
+    },
+    {
+      "epoch": 57.44,
+      "grad_norm": 4.708043575286865,
+      "learning_rate": 4.255639097744361e-06,
+      "loss": 0.1822,
+      "step": 38200
+    },
+    {
+      "epoch": 57.46,
+      "grad_norm": 4.237730503082275,
+      "learning_rate": 4.254135338345865e-06,
+      "loss": 0.1897,
+      "step": 38210
+    },
+    {
+      "epoch": 57.47,
+      "grad_norm": 4.254096508026123,
+      "learning_rate": 4.252631578947369e-06,
+      "loss": 0.2153,
+      "step": 38220
+    },
+    {
+      "epoch": 57.49,
+      "grad_norm": 2.041449546813965,
+      "learning_rate": 4.251127819548872e-06,
+      "loss": 0.1745,
+      "step": 38230
+    },
+    {
+      "epoch": 57.5,
+      "grad_norm": 2.9774394035339355,
+      "learning_rate": 4.249624060150377e-06,
+      "loss": 0.2481,
+      "step": 38240
+    },
+    {
+      "epoch": 57.52,
+      "grad_norm": 2.6556615829467773,
+      "learning_rate": 4.24812030075188e-06,
+      "loss": 0.1992,
+      "step": 38250
+    },
+    {
+      "epoch": 57.53,
+      "grad_norm": 4.096647262573242,
+      "learning_rate": 4.246616541353384e-06,
+      "loss": 0.2644,
+      "step": 38260
+    },
+    {
+      "epoch": 57.55,
+      "grad_norm": 4.746277809143066,
+      "learning_rate": 4.245112781954887e-06,
+      "loss": 0.1919,
+      "step": 38270
+    },
+    {
+      "epoch": 57.56,
+      "grad_norm": 5.280014514923096,
+      "learning_rate": 4.243609022556392e-06,
+      "loss": 0.2576,
+      "step": 38280
+    },
+    {
+      "epoch": 57.58,
+      "grad_norm": 6.924928188323975,
+      "learning_rate": 4.242105263157895e-06,
+      "loss": 0.1854,
+      "step": 38290
+    },
+    {
+      "epoch": 57.59,
+      "grad_norm": 3.799520969390869,
+      "learning_rate": 4.240601503759399e-06,
+      "loss": 0.2675,
+      "step": 38300
+    },
+    {
+      "epoch": 57.61,
+      "grad_norm": 5.440141677856445,
+      "learning_rate": 4.239097744360903e-06,
+      "loss": 0.1975,
+      "step": 38310
+    },
+    {
+      "epoch": 57.62,
+      "grad_norm": 5.276878356933594,
+      "learning_rate": 4.2375939849624065e-06,
+      "loss": 0.2251,
+      "step": 38320
+    },
+    {
+      "epoch": 57.64,
+      "grad_norm": 5.117806434631348,
+      "learning_rate": 4.23609022556391e-06,
+      "loss": 0.1818,
+      "step": 38330
+    },
+    {
+      "epoch": 57.65,
+      "grad_norm": 5.137388229370117,
+      "learning_rate": 4.2345864661654144e-06,
+      "loss": 0.2429,
+      "step": 38340
+    },
+    {
+      "epoch": 57.67,
+      "grad_norm": 1.6786266565322876,
+      "learning_rate": 4.233082706766918e-06,
+      "loss": 0.2236,
+      "step": 38350
+    },
+    {
+      "epoch": 57.68,
+      "grad_norm": 4.657053470611572,
+      "learning_rate": 4.2315789473684215e-06,
+      "loss": 0.2304,
+      "step": 38360
+    },
+    {
+      "epoch": 57.7,
+      "grad_norm": 6.737263202667236,
+      "learning_rate": 4.230075187969925e-06,
+      "loss": 0.1869,
+      "step": 38370
+    },
+    {
+      "epoch": 57.71,
+      "grad_norm": 8.260579109191895,
+      "learning_rate": 4.228571428571429e-06,
+      "loss": 0.2162,
+      "step": 38380
+    },
+    {
+      "epoch": 57.73,
+      "grad_norm": 4.183199405670166,
+      "learning_rate": 4.227067669172933e-06,
+      "loss": 0.2047,
+      "step": 38390
+    },
+    {
+      "epoch": 57.74,
+      "grad_norm": 6.769223213195801,
+      "learning_rate": 4.225563909774436e-06,
+      "loss": 0.1686,
+      "step": 38400
+    },
+    {
+      "epoch": 57.76,
+      "grad_norm": 6.208756923675537,
+      "learning_rate": 4.224060150375941e-06,
+      "loss": 0.2128,
+      "step": 38410
+    },
+    {
+      "epoch": 57.77,
+      "grad_norm": 5.576897144317627,
+      "learning_rate": 4.222556390977444e-06,
+      "loss": 0.1921,
+      "step": 38420
+    },
+    {
+      "epoch": 57.79,
+      "grad_norm": 6.998033046722412,
+      "learning_rate": 4.221052631578948e-06,
+      "loss": 0.2155,
+      "step": 38430
+    },
+    {
+      "epoch": 57.8,
+      "grad_norm": 4.374655246734619,
+      "learning_rate": 4.219548872180451e-06,
+      "loss": 0.2439,
+      "step": 38440
+    },
+    {
+      "epoch": 57.82,
+      "grad_norm": 3.7079293727874756,
+      "learning_rate": 4.218045112781956e-06,
+      "loss": 0.228,
+      "step": 38450
+    },
+    {
+      "epoch": 57.83,
+      "grad_norm": 7.11881685256958,
+      "learning_rate": 4.216541353383459e-06,
+      "loss": 0.2366,
+      "step": 38460
+    },
+    {
+      "epoch": 57.85,
+      "grad_norm": 11.91295051574707,
+      "learning_rate": 4.215037593984963e-06,
+      "loss": 0.2144,
+      "step": 38470
+    },
+    {
+      "epoch": 57.86,
+      "grad_norm": 4.202861309051514,
+      "learning_rate": 4.213533834586466e-06,
+      "loss": 0.2088,
+      "step": 38480
+    },
+    {
+      "epoch": 57.88,
+      "grad_norm": 2.2428081035614014,
+      "learning_rate": 4.21203007518797e-06,
+      "loss": 0.2238,
+      "step": 38490
+    },
+    {
+      "epoch": 57.89,
+      "grad_norm": 6.293793678283691,
+      "learning_rate": 4.210526315789474e-06,
+      "loss": 0.246,
+      "step": 38500
+    },
+    {
+      "epoch": 57.91,
+      "grad_norm": 3.7611231803894043,
+      "learning_rate": 4.209022556390978e-06,
+      "loss": 0.2574,
+      "step": 38510
+    },
+    {
+      "epoch": 57.92,
+      "grad_norm": 3.548879623413086,
+      "learning_rate": 4.207518796992481e-06,
+      "loss": 0.2064,
+      "step": 38520
+    },
+    {
+      "epoch": 57.94,
+      "grad_norm": 3.5769736766815186,
+      "learning_rate": 4.2060150375939855e-06,
+      "loss": 0.1324,
+      "step": 38530
+    },
+    {
+      "epoch": 57.95,
+      "grad_norm": 6.342837810516357,
+      "learning_rate": 4.204511278195489e-06,
+      "loss": 0.1634,
+      "step": 38540
+    },
+    {
+      "epoch": 57.97,
+      "grad_norm": 3.722198724746704,
+      "learning_rate": 4.2030075187969926e-06,
+      "loss": 0.2034,
+      "step": 38550
+    },
+    {
+      "epoch": 57.98,
+      "grad_norm": 6.514225482940674,
+      "learning_rate": 4.201503759398496e-06,
+      "loss": 0.2175,
+      "step": 38560
+    },
+    {
+      "epoch": 58.0,
+      "grad_norm": 0.08241364359855652,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.1931,
+      "step": 38570
+    },
+    {
+      "epoch": 58.0,
+      "eval_accuracy": 0.9299,
+      "eval_loss": 0.3219141960144043,
+      "eval_runtime": 84.8061,
+      "eval_samples_per_second": 117.916,
+      "eval_steps_per_second": 0.472,
+      "step": 38570
+    },
+    {
+      "epoch": 58.02,
+      "grad_norm": 3.8991994857788086,
+      "learning_rate": 4.198496240601504e-06,
+      "loss": 0.2143,
+      "step": 38580
+    },
+    {
+      "epoch": 58.03,
+      "grad_norm": 4.180047035217285,
+      "learning_rate": 4.1969924812030075e-06,
+      "loss": 0.1819,
+      "step": 38590
+    },
+    {
+      "epoch": 58.05,
+      "grad_norm": 9.162282943725586,
+      "learning_rate": 4.195488721804512e-06,
+      "loss": 0.2224,
+      "step": 38600
+    },
+    {
+      "epoch": 58.06,
+      "grad_norm": 3.6437289714813232,
+      "learning_rate": 4.193984962406015e-06,
+      "loss": 0.1961,
+      "step": 38610
+    },
+    {
+      "epoch": 58.08,
+      "grad_norm": 5.206214904785156,
+      "learning_rate": 4.192481203007519e-06,
+      "loss": 0.2267,
+      "step": 38620
+    },
+    {
+      "epoch": 58.09,
+      "grad_norm": 7.128195762634277,
+      "learning_rate": 4.190977443609022e-06,
+      "loss": 0.2716,
+      "step": 38630
+    },
+    {
+      "epoch": 58.11,
+      "grad_norm": 6.48333740234375,
+      "learning_rate": 4.189473684210527e-06,
+      "loss": 0.2381,
+      "step": 38640
+    },
+    {
+      "epoch": 58.12,
+      "grad_norm": 6.768774509429932,
+      "learning_rate": 4.18796992481203e-06,
+      "loss": 0.2204,
+      "step": 38650
+    },
+    {
+      "epoch": 58.14,
+      "grad_norm": 7.444092273712158,
+      "learning_rate": 4.186466165413534e-06,
+      "loss": 0.2563,
+      "step": 38660
+    },
+    {
+      "epoch": 58.15,
+      "grad_norm": 5.427529335021973,
+      "learning_rate": 4.184962406015038e-06,
+      "loss": 0.1636,
+      "step": 38670
+    },
+    {
+      "epoch": 58.17,
+      "grad_norm": 9.827468872070312,
+      "learning_rate": 4.183458646616542e-06,
+      "loss": 0.2042,
+      "step": 38680
+    },
+    {
+      "epoch": 58.18,
+      "grad_norm": 7.735494613647461,
+      "learning_rate": 4.181954887218045e-06,
+      "loss": 0.2047,
+      "step": 38690
+    },
+    {
+      "epoch": 58.2,
+      "grad_norm": 4.068072319030762,
+      "learning_rate": 4.18045112781955e-06,
+      "loss": 0.2105,
+      "step": 38700
+    },
+    {
+      "epoch": 58.21,
+      "grad_norm": 5.502895832061768,
+      "learning_rate": 4.178947368421053e-06,
+      "loss": 0.2376,
+      "step": 38710
+    },
+    {
+      "epoch": 58.23,
+      "grad_norm": 4.636054515838623,
+      "learning_rate": 4.177443609022557e-06,
+      "loss": 0.2079,
+      "step": 38720
+    },
+    {
+      "epoch": 58.24,
+      "grad_norm": 3.9590959548950195,
+      "learning_rate": 4.17593984962406e-06,
+      "loss": 0.2144,
+      "step": 38730
+    },
+    {
+      "epoch": 58.26,
+      "grad_norm": 4.759368896484375,
+      "learning_rate": 4.1744360902255645e-06,
+      "loss": 0.1763,
+      "step": 38740
+    },
+    {
+      "epoch": 58.27,
+      "grad_norm": 4.7535400390625,
+      "learning_rate": 4.172932330827068e-06,
+      "loss": 0.1848,
+      "step": 38750
+    },
+    {
+      "epoch": 58.29,
+      "grad_norm": 8.80599308013916,
+      "learning_rate": 4.1714285714285715e-06,
+      "loss": 0.2202,
+      "step": 38760
+    },
+    {
+      "epoch": 58.3,
+      "grad_norm": 5.612436771392822,
+      "learning_rate": 4.169924812030076e-06,
+      "loss": 0.1748,
+      "step": 38770
+    },
+    {
+      "epoch": 58.32,
+      "grad_norm": 5.2577972412109375,
+      "learning_rate": 4.1684210526315794e-06,
+      "loss": 0.2463,
+      "step": 38780
+    },
+    {
+      "epoch": 58.33,
+      "grad_norm": 6.162819862365723,
+      "learning_rate": 4.166917293233083e-06,
+      "loss": 0.1727,
+      "step": 38790
+    },
+    {
+      "epoch": 58.35,
+      "grad_norm": 5.589300155639648,
+      "learning_rate": 4.165413533834587e-06,
+      "loss": 0.2035,
+      "step": 38800
+    },
+    {
+      "epoch": 58.36,
+      "grad_norm": 3.496182680130005,
+      "learning_rate": 4.163909774436091e-06,
+      "loss": 0.1862,
+      "step": 38810
+    },
+    {
+      "epoch": 58.38,
+      "grad_norm": 7.816127777099609,
+      "learning_rate": 4.162406015037594e-06,
+      "loss": 0.1733,
+      "step": 38820
+    },
+    {
+      "epoch": 58.39,
+      "grad_norm": 8.98507308959961,
+      "learning_rate": 4.160902255639098e-06,
+      "loss": 0.1806,
+      "step": 38830
+    },
+    {
+      "epoch": 58.41,
+      "grad_norm": 3.1957809925079346,
+      "learning_rate": 4.159398496240602e-06,
+      "loss": 0.1946,
+      "step": 38840
+    },
+    {
+      "epoch": 58.42,
+      "grad_norm": 2.860567331314087,
+      "learning_rate": 4.157894736842106e-06,
+      "loss": 0.2619,
+      "step": 38850
+    },
+    {
+      "epoch": 58.44,
+      "grad_norm": 6.364305019378662,
+      "learning_rate": 4.156390977443609e-06,
+      "loss": 0.2311,
+      "step": 38860
+    },
+    {
+      "epoch": 58.45,
+      "grad_norm": 8.90013599395752,
+      "learning_rate": 4.154887218045114e-06,
+      "loss": 0.2202,
+      "step": 38870
+    },
+    {
+      "epoch": 58.47,
+      "grad_norm": 6.431805610656738,
+      "learning_rate": 4.153383458646617e-06,
+      "loss": 0.1814,
+      "step": 38880
+    },
+    {
+      "epoch": 58.48,
+      "grad_norm": 7.369435787200928,
+      "learning_rate": 4.151879699248121e-06,
+      "loss": 0.2696,
+      "step": 38890
+    },
+    {
+      "epoch": 58.5,
+      "grad_norm": 4.004574775695801,
+      "learning_rate": 4.150375939849624e-06,
+      "loss": 0.1867,
+      "step": 38900
+    },
+    {
+      "epoch": 58.51,
+      "grad_norm": 3.348543643951416,
+      "learning_rate": 4.1488721804511286e-06,
+      "loss": 0.1993,
+      "step": 38910
+    },
+    {
+      "epoch": 58.53,
+      "grad_norm": 3.0820324420928955,
+      "learning_rate": 4.147368421052632e-06,
+      "loss": 0.2131,
+      "step": 38920
+    },
+    {
+      "epoch": 58.54,
+      "grad_norm": 5.479022026062012,
+      "learning_rate": 4.145864661654136e-06,
+      "loss": 0.255,
+      "step": 38930
+    },
+    {
+      "epoch": 58.56,
+      "grad_norm": 5.253530979156494,
+      "learning_rate": 4.14436090225564e-06,
+      "loss": 0.1944,
+      "step": 38940
+    },
+    {
+      "epoch": 58.57,
+      "grad_norm": 3.170219659805298,
+      "learning_rate": 4.1428571428571435e-06,
+      "loss": 0.1768,
+      "step": 38950
+    },
+    {
+      "epoch": 58.59,
+      "grad_norm": 3.0692005157470703,
+      "learning_rate": 4.141353383458647e-06,
+      "loss": 0.2054,
+      "step": 38960
+    },
+    {
+      "epoch": 58.6,
+      "grad_norm": 5.293381214141846,
+      "learning_rate": 4.1398496240601505e-06,
+      "loss": 0.2317,
+      "step": 38970
+    },
+    {
+      "epoch": 58.62,
+      "grad_norm": 5.903608322143555,
+      "learning_rate": 4.138345864661654e-06,
+      "loss": 0.2098,
+      "step": 38980
+    },
+    {
+      "epoch": 58.63,
+      "grad_norm": 6.2776665687561035,
+      "learning_rate": 4.136842105263158e-06,
+      "loss": 0.2591,
+      "step": 38990
+    },
+    {
+      "epoch": 58.65,
+      "grad_norm": 1.9607787132263184,
+      "learning_rate": 4.135338345864662e-06,
+      "loss": 0.1817,
+      "step": 39000
+    },
+    {
+      "epoch": 58.66,
+      "grad_norm": 4.821249961853027,
+      "learning_rate": 4.1338345864661654e-06,
+      "loss": 0.1306,
+      "step": 39010
+    },
+    {
+      "epoch": 58.68,
+      "grad_norm": 4.712565898895264,
+      "learning_rate": 4.132330827067669e-06,
+      "loss": 0.2086,
+      "step": 39020
+    },
+    {
+      "epoch": 58.69,
+      "grad_norm": 4.460330486297607,
+      "learning_rate": 4.130827067669173e-06,
+      "loss": 0.1441,
+      "step": 39030
+    },
+    {
+      "epoch": 58.71,
+      "grad_norm": 4.247945308685303,
+      "learning_rate": 4.129323308270677e-06,
+      "loss": 0.2307,
+      "step": 39040
+    },
+    {
+      "epoch": 58.72,
+      "grad_norm": 5.8864521980285645,
+      "learning_rate": 4.12781954887218e-06,
+      "loss": 0.2461,
+      "step": 39050
+    },
+    {
+      "epoch": 58.74,
+      "grad_norm": 4.723229885101318,
+      "learning_rate": 4.126315789473685e-06,
+      "loss": 0.2235,
+      "step": 39060
+    },
+    {
+      "epoch": 58.75,
+      "grad_norm": 3.2411322593688965,
+      "learning_rate": 4.124812030075188e-06,
+      "loss": 0.1957,
+      "step": 39070
+    },
+    {
+      "epoch": 58.77,
+      "grad_norm": 7.346851348876953,
+      "learning_rate": 4.123308270676692e-06,
+      "loss": 0.2102,
+      "step": 39080
+    },
+    {
+      "epoch": 58.78,
+      "grad_norm": 5.536585330963135,
+      "learning_rate": 4.121804511278195e-06,
+      "loss": 0.2596,
+      "step": 39090
+    },
+    {
+      "epoch": 58.8,
+      "grad_norm": 2.5167481899261475,
+      "learning_rate": 4.1203007518797e-06,
+      "loss": 0.1739,
+      "step": 39100
+    },
+    {
+      "epoch": 58.81,
+      "grad_norm": 2.2827956676483154,
+      "learning_rate": 4.118796992481203e-06,
+      "loss": 0.2394,
+      "step": 39110
+    },
+    {
+      "epoch": 58.83,
+      "grad_norm": 3.863255262374878,
+      "learning_rate": 4.117293233082707e-06,
+      "loss": 0.205,
+      "step": 39120
+    },
+    {
+      "epoch": 58.84,
+      "grad_norm": 5.798130035400391,
+      "learning_rate": 4.115789473684211e-06,
+      "loss": 0.2017,
+      "step": 39130
+    },
+    {
+      "epoch": 58.86,
+      "grad_norm": 3.510673761367798,
+      "learning_rate": 4.114285714285715e-06,
+      "loss": 0.2036,
+      "step": 39140
+    },
+    {
+      "epoch": 58.87,
+      "grad_norm": 4.811961650848389,
+      "learning_rate": 4.112781954887218e-06,
+      "loss": 0.2651,
+      "step": 39150
+    },
+    {
+      "epoch": 58.89,
+      "grad_norm": 6.230356216430664,
+      "learning_rate": 4.1112781954887225e-06,
+      "loss": 0.1561,
+      "step": 39160
+    },
+    {
+      "epoch": 58.9,
+      "grad_norm": 5.7233757972717285,
+      "learning_rate": 4.109774436090226e-06,
+      "loss": 0.1541,
+      "step": 39170
+    },
+    {
+      "epoch": 58.92,
+      "grad_norm": 6.532334327697754,
+      "learning_rate": 4.1082706766917295e-06,
+      "loss": 0.2049,
+      "step": 39180
+    },
+    {
+      "epoch": 58.93,
+      "grad_norm": 4.839334011077881,
+      "learning_rate": 4.106766917293233e-06,
+      "loss": 0.1846,
+      "step": 39190
+    },
+    {
+      "epoch": 58.95,
+      "grad_norm": 8.426265716552734,
+      "learning_rate": 4.105263157894737e-06,
+      "loss": 0.2213,
+      "step": 39200
+    },
+    {
+      "epoch": 58.96,
+      "grad_norm": 3.9503211975097656,
+      "learning_rate": 4.103759398496241e-06,
+      "loss": 0.1891,
+      "step": 39210
+    },
+    {
+      "epoch": 58.98,
+      "grad_norm": 5.404723167419434,
+      "learning_rate": 4.1022556390977444e-06,
+      "loss": 0.2554,
+      "step": 39220
+    },
+    {
+      "epoch": 58.99,
+      "grad_norm": 2.1365013122558594,
+      "learning_rate": 4.100751879699249e-06,
+      "loss": 0.2328,
+      "step": 39230
+    },
+    {
+      "epoch": 59.0,
+      "eval_accuracy": 0.9316,
+      "eval_loss": 0.315520316362381,
+      "eval_runtime": 84.5641,
+      "eval_samples_per_second": 118.253,
+      "eval_steps_per_second": 0.473,
+      "step": 39235
+    },
+    {
+      "epoch": 59.01,
+      "grad_norm": 6.348912239074707,
+      "learning_rate": 4.099248120300752e-06,
+      "loss": 0.174,
+      "step": 39240
+    },
+    {
+      "epoch": 59.02,
+      "grad_norm": 7.103587627410889,
+      "learning_rate": 4.097744360902256e-06,
+      "loss": 0.2624,
+      "step": 39250
+    },
+    {
+      "epoch": 59.04,
+      "grad_norm": 3.9258017539978027,
+      "learning_rate": 4.09624060150376e-06,
+      "loss": 0.2086,
+      "step": 39260
+    },
+    {
+      "epoch": 59.05,
+      "grad_norm": 4.404173851013184,
+      "learning_rate": 4.094736842105264e-06,
+      "loss": 0.2327,
+      "step": 39270
+    },
+    {
+      "epoch": 59.07,
+      "grad_norm": 4.899421691894531,
+      "learning_rate": 4.093233082706767e-06,
+      "loss": 0.2527,
+      "step": 39280
+    },
+    {
+      "epoch": 59.08,
+      "grad_norm": 5.5442094802856445,
+      "learning_rate": 4.091729323308271e-06,
+      "loss": 0.1896,
+      "step": 39290
+    },
+    {
+      "epoch": 59.1,
+      "grad_norm": 3.816484212875366,
+      "learning_rate": 4.090225563909775e-06,
+      "loss": 0.2282,
+      "step": 39300
+    },
+    {
+      "epoch": 59.11,
+      "grad_norm": 7.135293006896973,
+      "learning_rate": 4.088721804511279e-06,
+      "loss": 0.217,
+      "step": 39310
+    },
+    {
+      "epoch": 59.13,
+      "grad_norm": 7.170220375061035,
+      "learning_rate": 4.087218045112782e-06,
+      "loss": 0.2658,
+      "step": 39320
+    },
+    {
+      "epoch": 59.14,
+      "grad_norm": 3.617730140686035,
+      "learning_rate": 4.0857142857142865e-06,
+      "loss": 0.1469,
+      "step": 39330
+    },
+    {
+      "epoch": 59.16,
+      "grad_norm": 3.9617254734039307,
+      "learning_rate": 4.08421052631579e-06,
+      "loss": 0.2034,
+      "step": 39340
+    },
+    {
+      "epoch": 59.17,
+      "grad_norm": 5.964526653289795,
+      "learning_rate": 4.0827067669172936e-06,
+      "loss": 0.2128,
+      "step": 39350
+    },
+    {
+      "epoch": 59.19,
+      "grad_norm": 5.290602684020996,
+      "learning_rate": 4.081203007518797e-06,
+      "loss": 0.2284,
+      "step": 39360
+    },
+    {
+      "epoch": 59.2,
+      "grad_norm": 9.3060941696167,
+      "learning_rate": 4.0796992481203015e-06,
+      "loss": 0.2395,
+      "step": 39370
+    },
+    {
+      "epoch": 59.22,
+      "grad_norm": 7.939350128173828,
+      "learning_rate": 4.078195488721805e-06,
+      "loss": 0.2119,
+      "step": 39380
+    },
+    {
+      "epoch": 59.23,
+      "grad_norm": 5.00067138671875,
+      "learning_rate": 4.0766917293233085e-06,
+      "loss": 0.1631,
+      "step": 39390
+    },
+    {
+      "epoch": 59.25,
+      "grad_norm": 5.9795918464660645,
+      "learning_rate": 4.075187969924813e-06,
+      "loss": 0.1553,
+      "step": 39400
+    },
+    {
+      "epoch": 59.26,
+      "grad_norm": 2.5546774864196777,
+      "learning_rate": 4.073684210526316e-06,
+      "loss": 0.159,
+      "step": 39410
+    },
+    {
+      "epoch": 59.28,
+      "grad_norm": 8.247179985046387,
+      "learning_rate": 4.07218045112782e-06,
+      "loss": 0.2191,
+      "step": 39420
+    },
+    {
+      "epoch": 59.29,
+      "grad_norm": 5.384758949279785,
+      "learning_rate": 4.070676691729324e-06,
+      "loss": 0.1586,
+      "step": 39430
+    },
+    {
+      "epoch": 59.31,
+      "grad_norm": 2.1114683151245117,
+      "learning_rate": 4.069172932330828e-06,
+      "loss": 0.2198,
+      "step": 39440
+    },
+    {
+      "epoch": 59.32,
+      "grad_norm": 7.196309566497803,
+      "learning_rate": 4.067669172932331e-06,
+      "loss": 0.2132,
+      "step": 39450
+    },
+    {
+      "epoch": 59.34,
+      "grad_norm": 12.441351890563965,
+      "learning_rate": 4.066165413533835e-06,
+      "loss": 0.1764,
+      "step": 39460
+    },
+    {
+      "epoch": 59.35,
+      "grad_norm": 4.480091094970703,
+      "learning_rate": 4.064661654135338e-06,
+      "loss": 0.2407,
+      "step": 39470
+    },
+    {
+      "epoch": 59.37,
+      "grad_norm": 4.517475605010986,
+      "learning_rate": 4.063157894736842e-06,
+      "loss": 0.2106,
+      "step": 39480
+    },
+    {
+      "epoch": 59.38,
+      "grad_norm": 6.213492393493652,
+      "learning_rate": 4.061654135338346e-06,
+      "loss": 0.2005,
+      "step": 39490
+    },
+    {
+      "epoch": 59.4,
+      "grad_norm": 3.382683515548706,
+      "learning_rate": 4.06015037593985e-06,
+      "loss": 0.17,
+      "step": 39500
+    },
+    {
+      "epoch": 59.41,
+      "grad_norm": 5.262118816375732,
+      "learning_rate": 4.058646616541353e-06,
+      "loss": 0.2167,
+      "step": 39510
+    },
+    {
+      "epoch": 59.43,
+      "grad_norm": 8.032793998718262,
+      "learning_rate": 4.057142857142858e-06,
+      "loss": 0.2287,
+      "step": 39520
+    },
+    {
+      "epoch": 59.44,
+      "grad_norm": 5.3855767250061035,
+      "learning_rate": 4.055639097744361e-06,
+      "loss": 0.1747,
+      "step": 39530
+    },
+    {
+      "epoch": 59.46,
+      "grad_norm": 8.112833976745605,
+      "learning_rate": 4.054135338345865e-06,
+      "loss": 0.1867,
+      "step": 39540
+    },
+    {
+      "epoch": 59.47,
+      "grad_norm": 4.0277910232543945,
+      "learning_rate": 4.052631578947368e-06,
+      "loss": 0.2311,
+      "step": 39550
+    },
+    {
+      "epoch": 59.49,
+      "grad_norm": 2.4710066318511963,
+      "learning_rate": 4.0511278195488725e-06,
+      "loss": 0.1935,
+      "step": 39560
+    },
+    {
+      "epoch": 59.5,
+      "grad_norm": 7.679681301116943,
+      "learning_rate": 4.049624060150376e-06,
+      "loss": 0.2504,
+      "step": 39570
+    },
+    {
+      "epoch": 59.52,
+      "grad_norm": 8.625081062316895,
+      "learning_rate": 4.04812030075188e-06,
+      "loss": 0.1737,
+      "step": 39580
+    },
+    {
+      "epoch": 59.53,
+      "grad_norm": 4.943390846252441,
+      "learning_rate": 4.046616541353384e-06,
+      "loss": 0.2392,
+      "step": 39590
+    },
+    {
+      "epoch": 59.55,
+      "grad_norm": 3.67423677444458,
+      "learning_rate": 4.0451127819548875e-06,
+      "loss": 0.1376,
+      "step": 39600
+    },
+    {
+      "epoch": 59.56,
+      "grad_norm": 5.872762203216553,
+      "learning_rate": 4.043609022556391e-06,
+      "loss": 0.22,
+      "step": 39610
+    },
+    {
+      "epoch": 59.58,
+      "grad_norm": 6.167865753173828,
+      "learning_rate": 4.042105263157895e-06,
+      "loss": 0.2386,
+      "step": 39620
+    },
+    {
+      "epoch": 59.59,
+      "grad_norm": 3.7467803955078125,
+      "learning_rate": 4.040601503759399e-06,
+      "loss": 0.2012,
+      "step": 39630
+    },
+    {
+      "epoch": 59.61,
+      "grad_norm": 4.501016139984131,
+      "learning_rate": 4.039097744360902e-06,
+      "loss": 0.2285,
+      "step": 39640
+    },
+    {
+      "epoch": 59.62,
+      "grad_norm": 5.832193851470947,
+      "learning_rate": 4.037593984962406e-06,
+      "loss": 0.1822,
+      "step": 39650
+    },
+    {
+      "epoch": 59.64,
+      "grad_norm": 3.999112844467163,
+      "learning_rate": 4.03609022556391e-06,
+      "loss": 0.2316,
+      "step": 39660
+    },
+    {
+      "epoch": 59.65,
+      "grad_norm": 9.898731231689453,
+      "learning_rate": 4.034586466165414e-06,
+      "loss": 0.2284,
+      "step": 39670
+    },
+    {
+      "epoch": 59.67,
+      "grad_norm": 4.816281318664551,
+      "learning_rate": 4.033082706766917e-06,
+      "loss": 0.168,
+      "step": 39680
+    },
+    {
+      "epoch": 59.68,
+      "grad_norm": 6.9781293869018555,
+      "learning_rate": 4.031578947368422e-06,
+      "loss": 0.2007,
+      "step": 39690
+    },
+    {
+      "epoch": 59.7,
+      "grad_norm": 4.193832874298096,
+      "learning_rate": 4.030075187969925e-06,
+      "loss": 0.2435,
+      "step": 39700
+    },
+    {
+      "epoch": 59.71,
+      "grad_norm": 3.802783966064453,
+      "learning_rate": 4.028571428571429e-06,
+      "loss": 0.2144,
+      "step": 39710
+    },
+    {
+      "epoch": 59.73,
+      "grad_norm": 2.7991650104522705,
+      "learning_rate": 4.027067669172933e-06,
+      "loss": 0.1264,
+      "step": 39720
+    },
+    {
+      "epoch": 59.74,
+      "grad_norm": 4.3173394203186035,
+      "learning_rate": 4.025563909774437e-06,
+      "loss": 0.1968,
+      "step": 39730
+    },
+    {
+      "epoch": 59.76,
+      "grad_norm": 4.434994220733643,
+      "learning_rate": 4.02406015037594e-06,
+      "loss": 0.2191,
+      "step": 39740
+    },
+    {
+      "epoch": 59.77,
+      "grad_norm": 3.0349013805389404,
+      "learning_rate": 4.022556390977444e-06,
+      "loss": 0.1819,
+      "step": 39750
+    },
+    {
+      "epoch": 59.79,
+      "grad_norm": 3.8737924098968506,
+      "learning_rate": 4.021052631578948e-06,
+      "loss": 0.2129,
+      "step": 39760
+    },
+    {
+      "epoch": 59.8,
+      "grad_norm": 6.809603214263916,
+      "learning_rate": 4.0195488721804515e-06,
+      "loss": 0.2261,
+      "step": 39770
+    },
+    {
+      "epoch": 59.82,
+      "grad_norm": 1.9871443510055542,
+      "learning_rate": 4.018045112781955e-06,
+      "loss": 0.1694,
+      "step": 39780
+    },
+    {
+      "epoch": 59.83,
+      "grad_norm": 7.0583977699279785,
+      "learning_rate": 4.016541353383459e-06,
+      "loss": 0.2135,
+      "step": 39790
+    },
+    {
+      "epoch": 59.85,
+      "grad_norm": 5.820276737213135,
+      "learning_rate": 4.015037593984963e-06,
+      "loss": 0.2536,
+      "step": 39800
+    },
+    {
+      "epoch": 59.86,
+      "grad_norm": 6.490072250366211,
+      "learning_rate": 4.0135338345864665e-06,
+      "loss": 0.1756,
+      "step": 39810
+    },
+    {
+      "epoch": 59.88,
+      "grad_norm": 7.442599296569824,
+      "learning_rate": 4.01203007518797e-06,
+      "loss": 0.1695,
+      "step": 39820
+    },
+    {
+      "epoch": 59.89,
+      "grad_norm": 5.296566486358643,
+      "learning_rate": 4.010526315789474e-06,
+      "loss": 0.2458,
+      "step": 39830
+    },
+    {
+      "epoch": 59.91,
+      "grad_norm": 13.77005672454834,
+      "learning_rate": 4.009022556390978e-06,
+      "loss": 0.218,
+      "step": 39840
+    },
+    {
+      "epoch": 59.92,
+      "grad_norm": 4.534656524658203,
+      "learning_rate": 4.007518796992481e-06,
+      "loss": 0.2051,
+      "step": 39850
+    },
+    {
+      "epoch": 59.94,
+      "grad_norm": 6.368296146392822,
+      "learning_rate": 4.006015037593986e-06,
+      "loss": 0.1526,
+      "step": 39860
+    },
+    {
+      "epoch": 59.95,
+      "grad_norm": 7.145541667938232,
+      "learning_rate": 4.004511278195489e-06,
+      "loss": 0.2447,
+      "step": 39870
+    },
+    {
+      "epoch": 59.97,
+      "grad_norm": 4.66719388961792,
+      "learning_rate": 4.003007518796993e-06,
+      "loss": 0.259,
+      "step": 39880
+    },
+    {
+      "epoch": 59.98,
+      "grad_norm": 3.6382246017456055,
+      "learning_rate": 4.001503759398497e-06,
+      "loss": 0.1915,
+      "step": 39890
+    },
+    {
+      "epoch": 60.0,
+      "grad_norm": 0.5675032138824463,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.145,
+      "step": 39900
+    },
+    {
+      "epoch": 60.0,
+      "eval_accuracy": 0.9295,
+      "eval_loss": 0.3215762674808502,
+      "eval_runtime": 85.0976,
+      "eval_samples_per_second": 117.512,
+      "eval_steps_per_second": 0.47,
+      "step": 39900
+    },
+    {
+      "epoch": 60.02,
+      "grad_norm": 2.28545880317688,
+      "learning_rate": 3.998496240601504e-06,
+      "loss": 0.1901,
+      "step": 39910
+    },
+    {
+      "epoch": 60.03,
+      "grad_norm": 15.001397132873535,
+      "learning_rate": 3.996992481203008e-06,
+      "loss": 0.1951,
+      "step": 39920
+    },
+    {
+      "epoch": 60.05,
+      "grad_norm": 6.06882905960083,
+      "learning_rate": 3.995488721804512e-06,
+      "loss": 0.1432,
+      "step": 39930
+    },
+    {
+      "epoch": 60.06,
+      "grad_norm": 3.984029769897461,
+      "learning_rate": 3.993984962406016e-06,
+      "loss": 0.2015,
+      "step": 39940
+    },
+    {
+      "epoch": 60.08,
+      "grad_norm": 17.75435447692871,
+      "learning_rate": 3.992481203007519e-06,
+      "loss": 0.2074,
+      "step": 39950
+    },
+    {
+      "epoch": 60.09,
+      "grad_norm": 6.104353427886963,
+      "learning_rate": 3.990977443609023e-06,
+      "loss": 0.2112,
+      "step": 39960
+    },
+    {
+      "epoch": 60.11,
+      "grad_norm": 13.67701530456543,
+      "learning_rate": 3.989473684210526e-06,
+      "loss": 0.226,
+      "step": 39970
+    },
+    {
+      "epoch": 60.12,
+      "grad_norm": 3.5879509449005127,
+      "learning_rate": 3.9879699248120305e-06,
+      "loss": 0.176,
+      "step": 39980
+    },
+    {
+      "epoch": 60.14,
+      "grad_norm": 4.916759490966797,
+      "learning_rate": 3.986466165413534e-06,
+      "loss": 0.1977,
+      "step": 39990
+    },
+    {
+      "epoch": 60.15,
+      "grad_norm": 7.997819423675537,
+      "learning_rate": 3.9849624060150376e-06,
+      "loss": 0.199,
+      "step": 40000
+    },
+    {
+      "epoch": 60.17,
+      "grad_norm": 3.076766014099121,
+      "learning_rate": 3.983458646616541e-06,
+      "loss": 0.1479,
+      "step": 40010
+    },
+    {
+      "epoch": 60.18,
+      "grad_norm": 6.796036720275879,
+      "learning_rate": 3.9819548872180454e-06,
+      "loss": 0.2257,
+      "step": 40020
+    },
+    {
+      "epoch": 60.2,
+      "grad_norm": 6.836452007293701,
+      "learning_rate": 3.980451127819549e-06,
+      "loss": 0.2594,
+      "step": 40030
+    },
+    {
+      "epoch": 60.21,
+      "grad_norm": 7.558508396148682,
+      "learning_rate": 3.9789473684210525e-06,
+      "loss": 0.2029,
+      "step": 40040
+    },
+    {
+      "epoch": 60.23,
+      "grad_norm": 7.763191223144531,
+      "learning_rate": 3.977443609022557e-06,
+      "loss": 0.1931,
+      "step": 40050
+    },
+    {
+      "epoch": 60.24,
+      "grad_norm": 6.8963799476623535,
+      "learning_rate": 3.97593984962406e-06,
+      "loss": 0.2369,
+      "step": 40060
+    },
+    {
+      "epoch": 60.26,
+      "grad_norm": 6.124290943145752,
+      "learning_rate": 3.974436090225564e-06,
+      "loss": 0.2332,
+      "step": 40070
+    },
+    {
+      "epoch": 60.27,
+      "grad_norm": 2.4686286449432373,
+      "learning_rate": 3.972932330827068e-06,
+      "loss": 0.1754,
+      "step": 40080
+    },
+    {
+      "epoch": 60.29,
+      "grad_norm": 4.793791770935059,
+      "learning_rate": 3.971428571428572e-06,
+      "loss": 0.2353,
+      "step": 40090
+    },
+    {
+      "epoch": 60.3,
+      "grad_norm": 2.7970879077911377,
+      "learning_rate": 3.969924812030075e-06,
+      "loss": 0.1597,
+      "step": 40100
+    },
+    {
+      "epoch": 60.32,
+      "grad_norm": 6.253334045410156,
+      "learning_rate": 3.968421052631579e-06,
+      "loss": 0.2115,
+      "step": 40110
+    },
+    {
+      "epoch": 60.33,
+      "grad_norm": 7.9991631507873535,
+      "learning_rate": 3.966917293233083e-06,
+      "loss": 0.2242,
+      "step": 40120
+    },
+    {
+      "epoch": 60.35,
+      "grad_norm": 7.119922637939453,
+      "learning_rate": 3.965413533834587e-06,
+      "loss": 0.2398,
+      "step": 40130
+    },
+    {
+      "epoch": 60.36,
+      "grad_norm": 2.1998627185821533,
+      "learning_rate": 3.96390977443609e-06,
+      "loss": 0.1425,
+      "step": 40140
+    },
+    {
+      "epoch": 60.38,
+      "grad_norm": 5.503266334533691,
+      "learning_rate": 3.9624060150375946e-06,
+      "loss": 0.1403,
+      "step": 40150
+    },
+    {
+      "epoch": 60.39,
+      "grad_norm": 4.746212959289551,
+      "learning_rate": 3.960902255639098e-06,
+      "loss": 0.2247,
+      "step": 40160
+    },
+    {
+      "epoch": 60.41,
+      "grad_norm": 3.6244451999664307,
+      "learning_rate": 3.959398496240602e-06,
+      "loss": 0.2011,
+      "step": 40170
+    },
+    {
+      "epoch": 60.42,
+      "grad_norm": 2.700594663619995,
+      "learning_rate": 3.957894736842106e-06,
+      "loss": 0.1956,
+      "step": 40180
+    },
+    {
+      "epoch": 60.44,
+      "grad_norm": 8.941556930541992,
+      "learning_rate": 3.9563909774436095e-06,
+      "loss": 0.2188,
+      "step": 40190
+    },
+    {
+      "epoch": 60.45,
+      "grad_norm": 3.507028818130493,
+      "learning_rate": 3.954887218045113e-06,
+      "loss": 0.191,
+      "step": 40200
+    },
+    {
+      "epoch": 60.47,
+      "grad_norm": 3.7977683544158936,
+      "learning_rate": 3.9533834586466165e-06,
+      "loss": 0.194,
+      "step": 40210
+    },
+    {
+      "epoch": 60.48,
+      "grad_norm": 4.61478328704834,
+      "learning_rate": 3.951879699248121e-06,
+      "loss": 0.2135,
+      "step": 40220
+    },
+    {
+      "epoch": 60.5,
+      "grad_norm": 3.350219488143921,
+      "learning_rate": 3.950375939849624e-06,
+      "loss": 0.224,
+      "step": 40230
+    },
+    {
+      "epoch": 60.51,
+      "grad_norm": 4.675546169281006,
+      "learning_rate": 3.948872180451128e-06,
+      "loss": 0.1851,
+      "step": 40240
+    },
+    {
+      "epoch": 60.53,
+      "grad_norm": 4.169593811035156,
+      "learning_rate": 3.947368421052632e-06,
+      "loss": 0.2605,
+      "step": 40250
+    },
+    {
+      "epoch": 60.54,
+      "grad_norm": 4.376767635345459,
+      "learning_rate": 3.945864661654136e-06,
+      "loss": 0.2116,
+      "step": 40260
+    },
+    {
+      "epoch": 60.56,
+      "grad_norm": 3.2285714149475098,
+      "learning_rate": 3.944360902255639e-06,
+      "loss": 0.1881,
+      "step": 40270
+    },
+    {
+      "epoch": 60.57,
+      "grad_norm": 3.8466668128967285,
+      "learning_rate": 3.942857142857143e-06,
+      "loss": 0.2534,
+      "step": 40280
+    },
+    {
+      "epoch": 60.59,
+      "grad_norm": 8.687283515930176,
+      "learning_rate": 3.941353383458647e-06,
+      "loss": 0.2319,
+      "step": 40290
+    },
+    {
+      "epoch": 60.6,
+      "grad_norm": 3.930408239364624,
+      "learning_rate": 3.939849624060151e-06,
+      "loss": 0.2331,
+      "step": 40300
+    },
+    {
+      "epoch": 60.62,
+      "grad_norm": 3.52876615524292,
+      "learning_rate": 3.938345864661654e-06,
+      "loss": 0.1579,
+      "step": 40310
+    },
+    {
+      "epoch": 60.63,
+      "grad_norm": 8.303112030029297,
+      "learning_rate": 3.936842105263159e-06,
+      "loss": 0.1893,
+      "step": 40320
+    },
+    {
+      "epoch": 60.65,
+      "grad_norm": 6.3138957023620605,
+      "learning_rate": 3.935338345864662e-06,
+      "loss": 0.192,
+      "step": 40330
+    },
+    {
+      "epoch": 60.66,
+      "grad_norm": 3.4053993225097656,
+      "learning_rate": 3.933834586466166e-06,
+      "loss": 0.1935,
+      "step": 40340
+    },
+    {
+      "epoch": 60.68,
+      "grad_norm": 9.262184143066406,
+      "learning_rate": 3.93233082706767e-06,
+      "loss": 0.1957,
+      "step": 40350
+    },
+    {
+      "epoch": 60.69,
+      "grad_norm": 5.6493754386901855,
+      "learning_rate": 3.9308270676691736e-06,
+      "loss": 0.2533,
+      "step": 40360
+    },
+    {
+      "epoch": 60.71,
+      "grad_norm": 11.651237487792969,
+      "learning_rate": 3.929323308270677e-06,
+      "loss": 0.1964,
+      "step": 40370
+    },
+    {
+      "epoch": 60.72,
+      "grad_norm": 3.903280019760132,
+      "learning_rate": 3.927819548872181e-06,
+      "loss": 0.1883,
+      "step": 40380
+    },
+    {
+      "epoch": 60.74,
+      "grad_norm": 1.6574941873550415,
+      "learning_rate": 3.926315789473685e-06,
+      "loss": 0.2372,
+      "step": 40390
+    },
+    {
+      "epoch": 60.75,
+      "grad_norm": 6.501513957977295,
+      "learning_rate": 3.9248120300751885e-06,
+      "loss": 0.13,
+      "step": 40400
+    },
+    {
+      "epoch": 60.77,
+      "grad_norm": 5.910683631896973,
+      "learning_rate": 3.923308270676692e-06,
+      "loss": 0.2353,
+      "step": 40410
+    },
+    {
+      "epoch": 60.78,
+      "grad_norm": 1.645150065422058,
+      "learning_rate": 3.921804511278196e-06,
+      "loss": 0.2189,
+      "step": 40420
+    },
+    {
+      "epoch": 60.8,
+      "grad_norm": 5.449909210205078,
+      "learning_rate": 3.9203007518797e-06,
+      "loss": 0.1899,
+      "step": 40430
+    },
+    {
+      "epoch": 60.81,
+      "grad_norm": 4.566460609436035,
+      "learning_rate": 3.918796992481203e-06,
+      "loss": 0.2828,
+      "step": 40440
+    },
+    {
+      "epoch": 60.83,
+      "grad_norm": 4.585579872131348,
+      "learning_rate": 3.917293233082707e-06,
+      "loss": 0.1729,
+      "step": 40450
+    },
+    {
+      "epoch": 60.84,
+      "grad_norm": 6.5645976066589355,
+      "learning_rate": 3.9157894736842104e-06,
+      "loss": 0.164,
+      "step": 40460
+    },
+    {
+      "epoch": 60.86,
+      "grad_norm": 9.433652877807617,
+      "learning_rate": 3.914285714285714e-06,
+      "loss": 0.219,
+      "step": 40470
+    },
+    {
+      "epoch": 60.87,
+      "grad_norm": 5.390902042388916,
+      "learning_rate": 3.912781954887218e-06,
+      "loss": 0.2299,
+      "step": 40480
+    },
+    {
+      "epoch": 60.89,
+      "grad_norm": 3.7667274475097656,
+      "learning_rate": 3.911278195488722e-06,
+      "loss": 0.2021,
+      "step": 40490
+    },
+    {
+      "epoch": 60.9,
+      "grad_norm": 4.511600017547607,
+      "learning_rate": 3.909774436090225e-06,
+      "loss": 0.2136,
+      "step": 40500
+    },
+    {
+      "epoch": 60.92,
+      "grad_norm": 3.638052225112915,
+      "learning_rate": 3.90827067669173e-06,
+      "loss": 0.1652,
+      "step": 40510
+    },
+    {
+      "epoch": 60.93,
+      "grad_norm": 13.318207740783691,
+      "learning_rate": 3.906766917293233e-06,
+      "loss": 0.1892,
+      "step": 40520
+    },
+    {
+      "epoch": 60.95,
+      "grad_norm": 4.776786804199219,
+      "learning_rate": 3.905263157894737e-06,
+      "loss": 0.2,
+      "step": 40530
+    },
+    {
+      "epoch": 60.96,
+      "grad_norm": 3.423015832901001,
+      "learning_rate": 3.903759398496241e-06,
+      "loss": 0.2601,
+      "step": 40540
+    },
+    {
+      "epoch": 60.98,
+      "grad_norm": 4.982937812805176,
+      "learning_rate": 3.902255639097745e-06,
+      "loss": 0.1368,
+      "step": 40550
+    },
+    {
+      "epoch": 60.99,
+      "grad_norm": 7.498082637786865,
+      "learning_rate": 3.900751879699248e-06,
+      "loss": 0.2804,
+      "step": 40560
+    },
+    {
+      "epoch": 61.0,
+      "eval_accuracy": 0.9298,
+      "eval_loss": 0.3252774775028229,
+      "eval_runtime": 84.231,
+      "eval_samples_per_second": 118.721,
+      "eval_steps_per_second": 0.475,
+      "step": 40565
+    },
+    {
+      "epoch": 61.01,
+      "grad_norm": 2.947619676589966,
+      "learning_rate": 3.899248120300752e-06,
+      "loss": 0.2347,
+      "step": 40570
+    },
+    {
+      "epoch": 61.02,
+      "grad_norm": 4.865323543548584,
+      "learning_rate": 3.897744360902256e-06,
+      "loss": 0.1525,
+      "step": 40580
+    },
+    {
+      "epoch": 61.04,
+      "grad_norm": 4.381734848022461,
+      "learning_rate": 3.8962406015037596e-06,
+      "loss": 0.2192,
+      "step": 40590
+    },
+    {
+      "epoch": 61.05,
+      "grad_norm": 4.15067195892334,
+      "learning_rate": 3.894736842105263e-06,
+      "loss": 0.2132,
+      "step": 40600
+    },
+    {
+      "epoch": 61.07,
+      "grad_norm": 4.439324855804443,
+      "learning_rate": 3.8932330827067675e-06,
+      "loss": 0.2321,
+      "step": 40610
+    },
+    {
+      "epoch": 61.08,
+      "grad_norm": 3.797804594039917,
+      "learning_rate": 3.891729323308271e-06,
+      "loss": 0.1685,
+      "step": 40620
+    },
+    {
+      "epoch": 61.1,
+      "grad_norm": 5.394155502319336,
+      "learning_rate": 3.8902255639097745e-06,
+      "loss": 0.1848,
+      "step": 40630
+    },
+    {
+      "epoch": 61.11,
+      "grad_norm": 6.317572116851807,
+      "learning_rate": 3.888721804511279e-06,
+      "loss": 0.2337,
+      "step": 40640
+    },
+    {
+      "epoch": 61.13,
+      "grad_norm": 4.338418483734131,
+      "learning_rate": 3.887218045112782e-06,
+      "loss": 0.1896,
+      "step": 40650
+    },
+    {
+      "epoch": 61.14,
+      "grad_norm": 4.074448585510254,
+      "learning_rate": 3.885714285714286e-06,
+      "loss": 0.2793,
+      "step": 40660
+    },
+    {
+      "epoch": 61.16,
+      "grad_norm": 7.455972671508789,
+      "learning_rate": 3.884210526315789e-06,
+      "loss": 0.2104,
+      "step": 40670
+    },
+    {
+      "epoch": 61.17,
+      "grad_norm": 6.7889204025268555,
+      "learning_rate": 3.882706766917294e-06,
+      "loss": 0.2767,
+      "step": 40680
+    },
+    {
+      "epoch": 61.19,
+      "grad_norm": 0.9154432415962219,
+      "learning_rate": 3.881203007518797e-06,
+      "loss": 0.1906,
+      "step": 40690
+    },
+    {
+      "epoch": 61.2,
+      "grad_norm": 5.013433933258057,
+      "learning_rate": 3.879699248120301e-06,
+      "loss": 0.1908,
+      "step": 40700
+    },
+    {
+      "epoch": 61.22,
+      "grad_norm": 4.357288360595703,
+      "learning_rate": 3.878195488721805e-06,
+      "loss": 0.1872,
+      "step": 40710
+    },
+    {
+      "epoch": 61.23,
+      "grad_norm": 4.3976569175720215,
+      "learning_rate": 3.876691729323309e-06,
+      "loss": 0.1649,
+      "step": 40720
+    },
+    {
+      "epoch": 61.25,
+      "grad_norm": 5.771651744842529,
+      "learning_rate": 3.875187969924812e-06,
+      "loss": 0.2027,
+      "step": 40730
+    },
+    {
+      "epoch": 61.26,
+      "grad_norm": 4.635252475738525,
+      "learning_rate": 3.873684210526316e-06,
+      "loss": 0.2115,
+      "step": 40740
+    },
+    {
+      "epoch": 61.28,
+      "grad_norm": 3.256643295288086,
+      "learning_rate": 3.87218045112782e-06,
+      "loss": 0.2126,
+      "step": 40750
+    },
+    {
+      "epoch": 61.29,
+      "grad_norm": 5.223315715789795,
+      "learning_rate": 3.870676691729324e-06,
+      "loss": 0.2077,
+      "step": 40760
+    },
+    {
+      "epoch": 61.31,
+      "grad_norm": 4.067598819732666,
+      "learning_rate": 3.869172932330827e-06,
+      "loss": 0.1596,
+      "step": 40770
+    },
+    {
+      "epoch": 61.32,
+      "grad_norm": 3.9475274085998535,
+      "learning_rate": 3.8676691729323315e-06,
+      "loss": 0.2226,
+      "step": 40780
+    },
+    {
+      "epoch": 61.34,
+      "grad_norm": 5.894591331481934,
+      "learning_rate": 3.866165413533835e-06,
+      "loss": 0.2033,
+      "step": 40790
+    },
+    {
+      "epoch": 61.35,
+      "grad_norm": 7.091558933258057,
+      "learning_rate": 3.8646616541353386e-06,
+      "loss": 0.234,
+      "step": 40800
+    },
+    {
+      "epoch": 61.37,
+      "grad_norm": 7.475659370422363,
+      "learning_rate": 3.863157894736843e-06,
+      "loss": 0.1912,
+      "step": 40810
+    },
+    {
+      "epoch": 61.38,
+      "grad_norm": 4.879849910736084,
+      "learning_rate": 3.8616541353383464e-06,
+      "loss": 0.1841,
+      "step": 40820
+    },
+    {
+      "epoch": 61.4,
+      "grad_norm": 5.782186031341553,
+      "learning_rate": 3.86015037593985e-06,
+      "loss": 0.2058,
+      "step": 40830
+    },
+    {
+      "epoch": 61.41,
+      "grad_norm": 5.1282196044921875,
+      "learning_rate": 3.8586466165413535e-06,
+      "loss": 0.2359,
+      "step": 40840
+    },
+    {
+      "epoch": 61.43,
+      "grad_norm": 5.245487213134766,
+      "learning_rate": 3.857142857142858e-06,
+      "loss": 0.2142,
+      "step": 40850
+    },
+    {
+      "epoch": 61.44,
+      "grad_norm": 5.231210708618164,
+      "learning_rate": 3.855639097744361e-06,
+      "loss": 0.2342,
+      "step": 40860
+    },
+    {
+      "epoch": 61.46,
+      "grad_norm": 2.8753762245178223,
+      "learning_rate": 3.854135338345865e-06,
+      "loss": 0.2064,
+      "step": 40870
+    },
+    {
+      "epoch": 61.47,
+      "grad_norm": 10.435833930969238,
+      "learning_rate": 3.852631578947369e-06,
+      "loss": 0.1609,
+      "step": 40880
+    },
+    {
+      "epoch": 61.49,
+      "grad_norm": 6.82497501373291,
+      "learning_rate": 3.851127819548873e-06,
+      "loss": 0.237,
+      "step": 40890
+    },
+    {
+      "epoch": 61.5,
+      "grad_norm": 3.6825904846191406,
+      "learning_rate": 3.849624060150376e-06,
+      "loss": 0.2074,
+      "step": 40900
+    },
+    {
+      "epoch": 61.52,
+      "grad_norm": 2.7955589294433594,
+      "learning_rate": 3.848120300751881e-06,
+      "loss": 0.1649,
+      "step": 40910
+    },
+    {
+      "epoch": 61.53,
+      "grad_norm": 5.636986255645752,
+      "learning_rate": 3.846616541353384e-06,
+      "loss": 0.1815,
+      "step": 40920
+    },
+    {
+      "epoch": 61.55,
+      "grad_norm": 5.697868347167969,
+      "learning_rate": 3.845112781954888e-06,
+      "loss": 0.1894,
+      "step": 40930
+    },
+    {
+      "epoch": 61.56,
+      "grad_norm": 4.8334808349609375,
+      "learning_rate": 3.843609022556391e-06,
+      "loss": 0.2018,
+      "step": 40940
+    },
+    {
+      "epoch": 61.58,
+      "grad_norm": 6.514235496520996,
+      "learning_rate": 3.842105263157895e-06,
+      "loss": 0.2102,
+      "step": 40950
+    },
+    {
+      "epoch": 61.59,
+      "grad_norm": 4.935739994049072,
+      "learning_rate": 3.840601503759398e-06,
+      "loss": 0.1594,
+      "step": 40960
+    },
+    {
+      "epoch": 61.61,
+      "grad_norm": 14.143031120300293,
+      "learning_rate": 3.839097744360903e-06,
+      "loss": 0.2094,
+      "step": 40970
+    },
+    {
+      "epoch": 61.62,
+      "grad_norm": 4.090781211853027,
+      "learning_rate": 3.837593984962406e-06,
+      "loss": 0.2485,
+      "step": 40980
+    },
+    {
+      "epoch": 61.64,
+      "grad_norm": 9.650825500488281,
+      "learning_rate": 3.83609022556391e-06,
+      "loss": 0.1868,
+      "step": 40990
+    },
+    {
+      "epoch": 61.65,
+      "grad_norm": 3.9108171463012695,
+      "learning_rate": 3.834586466165414e-06,
+      "loss": 0.2519,
+      "step": 41000
+    },
+    {
+      "epoch": 61.67,
+      "grad_norm": 3.324169874191284,
+      "learning_rate": 3.8330827067669175e-06,
+      "loss": 0.1668,
+      "step": 41010
+    },
+    {
+      "epoch": 61.68,
+      "grad_norm": 6.214301586151123,
+      "learning_rate": 3.831578947368421e-06,
+      "loss": 0.2764,
+      "step": 41020
+    },
+    {
+      "epoch": 61.7,
+      "grad_norm": 4.068943500518799,
+      "learning_rate": 3.830075187969925e-06,
+      "loss": 0.2038,
+      "step": 41030
+    },
+    {
+      "epoch": 61.71,
+      "grad_norm": 3.677924633026123,
+      "learning_rate": 3.828571428571429e-06,
+      "loss": 0.1743,
+      "step": 41040
+    },
+    {
+      "epoch": 61.73,
+      "grad_norm": 4.874640464782715,
+      "learning_rate": 3.8270676691729325e-06,
+      "loss": 0.2046,
+      "step": 41050
+    },
+    {
+      "epoch": 61.74,
+      "grad_norm": 4.538558006286621,
+      "learning_rate": 3.825563909774436e-06,
+      "loss": 0.2542,
+      "step": 41060
+    },
+    {
+      "epoch": 61.76,
+      "grad_norm": 7.119689464569092,
+      "learning_rate": 3.82406015037594e-06,
+      "loss": 0.1539,
+      "step": 41070
+    },
+    {
+      "epoch": 61.77,
+      "grad_norm": 6.170871257781982,
+      "learning_rate": 3.822556390977444e-06,
+      "loss": 0.3122,
+      "step": 41080
+    },
+    {
+      "epoch": 61.79,
+      "grad_norm": 4.091243267059326,
+      "learning_rate": 3.821052631578947e-06,
+      "loss": 0.2069,
+      "step": 41090
+    },
+    {
+      "epoch": 61.8,
+      "grad_norm": 5.074096202850342,
+      "learning_rate": 3.819548872180452e-06,
+      "loss": 0.2046,
+      "step": 41100
+    },
+    {
+      "epoch": 61.82,
+      "grad_norm": 6.762053966522217,
+      "learning_rate": 3.818045112781955e-06,
+      "loss": 0.1954,
+      "step": 41110
+    },
+    {
+      "epoch": 61.83,
+      "grad_norm": 3.479602098464966,
+      "learning_rate": 3.816541353383459e-06,
+      "loss": 0.2208,
+      "step": 41120
+    },
+    {
+      "epoch": 61.85,
+      "grad_norm": 4.815489768981934,
+      "learning_rate": 3.815037593984962e-06,
+      "loss": 0.191,
+      "step": 41130
+    },
+    {
+      "epoch": 61.86,
+      "grad_norm": 3.5152673721313477,
+      "learning_rate": 3.8135338345864663e-06,
+      "loss": 0.245,
+      "step": 41140
+    },
+    {
+      "epoch": 61.88,
+      "grad_norm": 2.4947726726531982,
+      "learning_rate": 3.81203007518797e-06,
+      "loss": 0.2218,
+      "step": 41150
+    },
+    {
+      "epoch": 61.89,
+      "grad_norm": 3.991682291030884,
+      "learning_rate": 3.810526315789474e-06,
+      "loss": 0.166,
+      "step": 41160
+    },
+    {
+      "epoch": 61.91,
+      "grad_norm": 5.845560550689697,
+      "learning_rate": 3.8090225563909777e-06,
+      "loss": 0.1897,
+      "step": 41170
+    },
+    {
+      "epoch": 61.92,
+      "grad_norm": 4.220813274383545,
+      "learning_rate": 3.8075187969924816e-06,
+      "loss": 0.1668,
+      "step": 41180
+    },
+    {
+      "epoch": 61.94,
+      "grad_norm": 6.980546951293945,
+      "learning_rate": 3.806015037593985e-06,
+      "loss": 0.185,
+      "step": 41190
+    },
+    {
+      "epoch": 61.95,
+      "grad_norm": 4.129945278167725,
+      "learning_rate": 3.804511278195489e-06,
+      "loss": 0.2469,
+      "step": 41200
+    },
+    {
+      "epoch": 61.97,
+      "grad_norm": 2.966998815536499,
+      "learning_rate": 3.803007518796993e-06,
+      "loss": 0.2203,
+      "step": 41210
+    },
+    {
+      "epoch": 61.98,
+      "grad_norm": 4.130844593048096,
+      "learning_rate": 3.8015037593984965e-06,
+      "loss": 0.2189,
+      "step": 41220
+    },
+    {
+      "epoch": 62.0,
+      "grad_norm": 19.200584411621094,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 0.1696,
+      "step": 41230
+    },
+    {
+      "epoch": 62.0,
+      "eval_accuracy": 0.9315,
+      "eval_loss": 0.30856576561927795,
+      "eval_runtime": 84.694,
+      "eval_samples_per_second": 118.072,
+      "eval_steps_per_second": 0.472,
+      "step": 41230
+    },
+    {
+      "epoch": 62.02,
+      "grad_norm": 3.957939386367798,
+      "learning_rate": 3.798496240601504e-06,
+      "loss": 0.1619,
+      "step": 41240
+    },
+    {
+      "epoch": 62.03,
+      "grad_norm": 5.774256229400635,
+      "learning_rate": 3.796992481203008e-06,
+      "loss": 0.1798,
+      "step": 41250
+    },
+    {
+      "epoch": 62.05,
+      "grad_norm": 4.3362274169921875,
+      "learning_rate": 3.795488721804512e-06,
+      "loss": 0.1837,
+      "step": 41260
+    },
+    {
+      "epoch": 62.06,
+      "grad_norm": 5.490597248077393,
+      "learning_rate": 3.7939849624060154e-06,
+      "loss": 0.168,
+      "step": 41270
+    },
+    {
+      "epoch": 62.08,
+      "grad_norm": 5.736985683441162,
+      "learning_rate": 3.7924812030075193e-06,
+      "loss": 0.2156,
+      "step": 41280
+    },
+    {
+      "epoch": 62.09,
+      "grad_norm": 5.470156192779541,
+      "learning_rate": 3.790977443609023e-06,
+      "loss": 0.1872,
+      "step": 41290
+    },
+    {
+      "epoch": 62.11,
+      "grad_norm": 6.704098701477051,
+      "learning_rate": 3.789473684210527e-06,
+      "loss": 0.2517,
+      "step": 41300
+    },
+    {
+      "epoch": 62.12,
+      "grad_norm": 3.4402177333831787,
+      "learning_rate": 3.7879699248120303e-06,
+      "loss": 0.1928,
+      "step": 41310
+    },
+    {
+      "epoch": 62.14,
+      "grad_norm": 4.056411266326904,
+      "learning_rate": 3.7864661654135343e-06,
+      "loss": 0.2138,
+      "step": 41320
+    },
+    {
+      "epoch": 62.15,
+      "grad_norm": 6.119357109069824,
+      "learning_rate": 3.784962406015038e-06,
+      "loss": 0.2381,
+      "step": 41330
+    },
+    {
+      "epoch": 62.17,
+      "grad_norm": 6.0787787437438965,
+      "learning_rate": 3.7834586466165417e-06,
+      "loss": 0.1763,
+      "step": 41340
+    },
+    {
+      "epoch": 62.18,
+      "grad_norm": 5.722217082977295,
+      "learning_rate": 3.7819548872180457e-06,
+      "loss": 0.1505,
+      "step": 41350
+    },
+    {
+      "epoch": 62.2,
+      "grad_norm": 1.9974137544631958,
+      "learning_rate": 3.780451127819549e-06,
+      "loss": 0.1912,
+      "step": 41360
+    },
+    {
+      "epoch": 62.21,
+      "grad_norm": 3.360246181488037,
+      "learning_rate": 3.778947368421053e-06,
+      "loss": 0.2063,
+      "step": 41370
+    },
+    {
+      "epoch": 62.23,
+      "grad_norm": 2.2372615337371826,
+      "learning_rate": 3.777443609022557e-06,
+      "loss": 0.1898,
+      "step": 41380
+    },
+    {
+      "epoch": 62.24,
+      "grad_norm": 7.954653263092041,
+      "learning_rate": 3.7759398496240606e-06,
+      "loss": 0.1837,
+      "step": 41390
+    },
+    {
+      "epoch": 62.26,
+      "grad_norm": 4.030917644500732,
+      "learning_rate": 3.7744360902255645e-06,
+      "loss": 0.206,
+      "step": 41400
+    },
+    {
+      "epoch": 62.27,
+      "grad_norm": 6.891762733459473,
+      "learning_rate": 3.772932330827068e-06,
+      "loss": 0.2181,
+      "step": 41410
+    },
+    {
+      "epoch": 62.29,
+      "grad_norm": 6.047714710235596,
+      "learning_rate": 3.771428571428572e-06,
+      "loss": 0.2758,
+      "step": 41420
+    },
+    {
+      "epoch": 62.3,
+      "grad_norm": 3.0663442611694336,
+      "learning_rate": 3.769924812030076e-06,
+      "loss": 0.1705,
+      "step": 41430
+    },
+    {
+      "epoch": 62.32,
+      "grad_norm": 5.566664218902588,
+      "learning_rate": 3.768421052631579e-06,
+      "loss": 0.1676,
+      "step": 41440
+    },
+    {
+      "epoch": 62.33,
+      "grad_norm": 3.831843614578247,
+      "learning_rate": 3.7669172932330825e-06,
+      "loss": 0.1544,
+      "step": 41450
+    },
+    {
+      "epoch": 62.35,
+      "grad_norm": 3.094122886657715,
+      "learning_rate": 3.7654135338345865e-06,
+      "loss": 0.2025,
+      "step": 41460
+    },
+    {
+      "epoch": 62.36,
+      "grad_norm": 10.858261108398438,
+      "learning_rate": 3.7639097744360904e-06,
+      "loss": 0.2037,
+      "step": 41470
+    },
+    {
+      "epoch": 62.38,
+      "grad_norm": 3.284749746322632,
+      "learning_rate": 3.762406015037594e-06,
+      "loss": 0.177,
+      "step": 41480
+    },
+    {
+      "epoch": 62.39,
+      "grad_norm": 5.909909248352051,
+      "learning_rate": 3.760902255639098e-06,
+      "loss": 0.1802,
+      "step": 41490
+    },
+    {
+      "epoch": 62.41,
+      "grad_norm": 2.958390474319458,
+      "learning_rate": 3.7593984962406014e-06,
+      "loss": 0.1886,
+      "step": 41500
+    },
+    {
+      "epoch": 62.42,
+      "grad_norm": 5.828741550445557,
+      "learning_rate": 3.7578947368421053e-06,
+      "loss": 0.2374,
+      "step": 41510
+    },
+    {
+      "epoch": 62.44,
+      "grad_norm": 6.119723320007324,
+      "learning_rate": 3.7563909774436093e-06,
+      "loss": 0.1787,
+      "step": 41520
+    },
+    {
+      "epoch": 62.45,
+      "grad_norm": 5.897652626037598,
+      "learning_rate": 3.754887218045113e-06,
+      "loss": 0.1704,
+      "step": 41530
+    },
+    {
+      "epoch": 62.47,
+      "grad_norm": 3.6232078075408936,
+      "learning_rate": 3.7533834586466168e-06,
+      "loss": 0.2112,
+      "step": 41540
+    },
+    {
+      "epoch": 62.48,
+      "grad_norm": 4.357728958129883,
+      "learning_rate": 3.7518796992481203e-06,
+      "loss": 0.1745,
+      "step": 41550
+    },
+    {
+      "epoch": 62.5,
+      "grad_norm": 3.9227254390716553,
+      "learning_rate": 3.7503759398496242e-06,
+      "loss": 0.1279,
+      "step": 41560
+    },
+    {
+      "epoch": 62.51,
+      "grad_norm": 4.716411113739014,
+      "learning_rate": 3.748872180451128e-06,
+      "loss": 0.1731,
+      "step": 41570
+    },
+    {
+      "epoch": 62.53,
+      "grad_norm": 5.446840763092041,
+      "learning_rate": 3.7473684210526317e-06,
+      "loss": 0.173,
+      "step": 41580
+    },
+    {
+      "epoch": 62.54,
+      "grad_norm": 9.927755355834961,
+      "learning_rate": 3.7458646616541356e-06,
+      "loss": 0.2158,
+      "step": 41590
+    },
+    {
+      "epoch": 62.56,
+      "grad_norm": 4.404961585998535,
+      "learning_rate": 3.744360902255639e-06,
+      "loss": 0.2027,
+      "step": 41600
+    },
+    {
+      "epoch": 62.57,
+      "grad_norm": 8.573219299316406,
+      "learning_rate": 3.742857142857143e-06,
+      "loss": 0.2426,
+      "step": 41610
+    },
+    {
+      "epoch": 62.59,
+      "grad_norm": 5.887312889099121,
+      "learning_rate": 3.741353383458647e-06,
+      "loss": 0.249,
+      "step": 41620
+    },
+    {
+      "epoch": 62.6,
+      "grad_norm": 8.018643379211426,
+      "learning_rate": 3.7398496240601505e-06,
+      "loss": 0.1449,
+      "step": 41630
+    },
+    {
+      "epoch": 62.62,
+      "grad_norm": 5.178997993469238,
+      "learning_rate": 3.7383458646616545e-06,
+      "loss": 0.2288,
+      "step": 41640
+    },
+    {
+      "epoch": 62.63,
+      "grad_norm": 4.393047332763672,
+      "learning_rate": 3.736842105263158e-06,
+      "loss": 0.2121,
+      "step": 41650
+    },
+    {
+      "epoch": 62.65,
+      "grad_norm": 4.626613616943359,
+      "learning_rate": 3.735338345864662e-06,
+      "loss": 0.1793,
+      "step": 41660
+    },
+    {
+      "epoch": 62.66,
+      "grad_norm": 2.9352829456329346,
+      "learning_rate": 3.733834586466166e-06,
+      "loss": 0.1641,
+      "step": 41670
+    },
+    {
+      "epoch": 62.68,
+      "grad_norm": 7.237253665924072,
+      "learning_rate": 3.7323308270676694e-06,
+      "loss": 0.1669,
+      "step": 41680
+    },
+    {
+      "epoch": 62.69,
+      "grad_norm": 7.4211835861206055,
+      "learning_rate": 3.7308270676691734e-06,
+      "loss": 0.1607,
+      "step": 41690
+    },
+    {
+      "epoch": 62.71,
+      "grad_norm": 5.203456401824951,
+      "learning_rate": 3.729323308270677e-06,
+      "loss": 0.1946,
+      "step": 41700
+    },
+    {
+      "epoch": 62.72,
+      "grad_norm": 6.299890995025635,
+      "learning_rate": 3.727819548872181e-06,
+      "loss": 0.17,
+      "step": 41710
+    },
+    {
+      "epoch": 62.74,
+      "grad_norm": 6.215569972991943,
+      "learning_rate": 3.7263157894736848e-06,
+      "loss": 0.1552,
+      "step": 41720
+    },
+    {
+      "epoch": 62.75,
+      "grad_norm": 6.16441011428833,
+      "learning_rate": 3.7248120300751883e-06,
+      "loss": 0.1511,
+      "step": 41730
+    },
+    {
+      "epoch": 62.77,
+      "grad_norm": 2.391936779022217,
+      "learning_rate": 3.7233082706766922e-06,
+      "loss": 0.2533,
+      "step": 41740
+    },
+    {
+      "epoch": 62.78,
+      "grad_norm": 5.521732807159424,
+      "learning_rate": 3.7218045112781957e-06,
+      "loss": 0.2338,
+      "step": 41750
+    },
+    {
+      "epoch": 62.8,
+      "grad_norm": 2.640953779220581,
+      "learning_rate": 3.7203007518796997e-06,
+      "loss": 0.2111,
+      "step": 41760
+    },
+    {
+      "epoch": 62.81,
+      "grad_norm": 4.677870273590088,
+      "learning_rate": 3.718796992481203e-06,
+      "loss": 0.2166,
+      "step": 41770
+    },
+    {
+      "epoch": 62.83,
+      "grad_norm": 4.7918291091918945,
+      "learning_rate": 3.717293233082707e-06,
+      "loss": 0.2068,
+      "step": 41780
+    },
+    {
+      "epoch": 62.84,
+      "grad_norm": 4.125340938568115,
+      "learning_rate": 3.715789473684211e-06,
+      "loss": 0.1522,
+      "step": 41790
+    },
+    {
+      "epoch": 62.86,
+      "grad_norm": 5.1403117179870605,
+      "learning_rate": 3.7142857142857146e-06,
+      "loss": 0.2198,
+      "step": 41800
+    },
+    {
+      "epoch": 62.87,
+      "grad_norm": 4.696916580200195,
+      "learning_rate": 3.7127819548872185e-06,
+      "loss": 0.1715,
+      "step": 41810
+    },
+    {
+      "epoch": 62.89,
+      "grad_norm": 3.3158819675445557,
+      "learning_rate": 3.711278195488722e-06,
+      "loss": 0.172,
+      "step": 41820
+    },
+    {
+      "epoch": 62.9,
+      "grad_norm": 6.180699825286865,
+      "learning_rate": 3.709774436090226e-06,
+      "loss": 0.2321,
+      "step": 41830
+    },
+    {
+      "epoch": 62.92,
+      "grad_norm": 5.126304626464844,
+      "learning_rate": 3.70827067669173e-06,
+      "loss": 0.1684,
+      "step": 41840
+    },
+    {
+      "epoch": 62.93,
+      "grad_norm": 7.246421813964844,
+      "learning_rate": 3.7067669172932335e-06,
+      "loss": 0.1933,
+      "step": 41850
+    },
+    {
+      "epoch": 62.95,
+      "grad_norm": 4.105726718902588,
+      "learning_rate": 3.7052631578947374e-06,
+      "loss": 0.2189,
+      "step": 41860
+    },
+    {
+      "epoch": 62.96,
+      "grad_norm": 3.8485569953918457,
+      "learning_rate": 3.703759398496241e-06,
+      "loss": 0.1698,
+      "step": 41870
+    },
+    {
+      "epoch": 62.98,
+      "grad_norm": 3.669081211090088,
+      "learning_rate": 3.702255639097745e-06,
+      "loss": 0.1472,
+      "step": 41880
+    },
+    {
+      "epoch": 62.99,
+      "grad_norm": 8.504827499389648,
+      "learning_rate": 3.700751879699249e-06,
+      "loss": 0.2194,
+      "step": 41890
+    },
+    {
+      "epoch": 63.0,
+      "eval_accuracy": 0.9313,
+      "eval_loss": 0.31697988510131836,
+      "eval_runtime": 84.413,
+      "eval_samples_per_second": 118.465,
+      "eval_steps_per_second": 0.474,
+      "step": 41895
+    },
+    {
+      "epoch": 63.01,
+      "grad_norm": 3.0113203525543213,
+      "learning_rate": 3.6992481203007523e-06,
+      "loss": 0.3075,
+      "step": 41900
+    },
+    {
+      "epoch": 63.02,
+      "grad_norm": 5.44133186340332,
+      "learning_rate": 3.6977443609022563e-06,
+      "loss": 0.2535,
+      "step": 41910
+    },
+    {
+      "epoch": 63.04,
+      "grad_norm": 4.054553985595703,
+      "learning_rate": 3.69624060150376e-06,
+      "loss": 0.1859,
+      "step": 41920
+    },
+    {
+      "epoch": 63.05,
+      "grad_norm": 2.3150360584259033,
+      "learning_rate": 3.6947368421052637e-06,
+      "loss": 0.1969,
+      "step": 41930
+    },
+    {
+      "epoch": 63.07,
+      "grad_norm": 5.139898777008057,
+      "learning_rate": 3.693233082706767e-06,
+      "loss": 0.1992,
+      "step": 41940
+    },
+    {
+      "epoch": 63.08,
+      "grad_norm": 3.828563928604126,
+      "learning_rate": 3.6917293233082708e-06,
+      "loss": 0.2596,
+      "step": 41950
+    },
+    {
+      "epoch": 63.1,
+      "grad_norm": 5.204103469848633,
+      "learning_rate": 3.6902255639097743e-06,
+      "loss": 0.1468,
+      "step": 41960
+    },
+    {
+      "epoch": 63.11,
+      "grad_norm": 2.3387629985809326,
+      "learning_rate": 3.6887218045112782e-06,
+      "loss": 0.1782,
+      "step": 41970
+    },
+    {
+      "epoch": 63.13,
+      "grad_norm": 5.02309513092041,
+      "learning_rate": 3.687218045112782e-06,
+      "loss": 0.2029,
+      "step": 41980
+    },
+    {
+      "epoch": 63.14,
+      "grad_norm": 6.977344512939453,
+      "learning_rate": 3.6857142857142857e-06,
+      "loss": 0.1883,
+      "step": 41990
+    },
+    {
+      "epoch": 63.16,
+      "grad_norm": 6.206759929656982,
+      "learning_rate": 3.6842105263157896e-06,
+      "loss": 0.2007,
+      "step": 42000
+    },
+    {
+      "epoch": 63.17,
+      "grad_norm": 7.947546482086182,
+      "learning_rate": 3.682706766917293e-06,
+      "loss": 0.2051,
+      "step": 42010
+    },
+    {
+      "epoch": 63.19,
+      "grad_norm": 5.88115119934082,
+      "learning_rate": 3.681203007518797e-06,
+      "loss": 0.168,
+      "step": 42020
+    },
+    {
+      "epoch": 63.2,
+      "grad_norm": 5.9370198249816895,
+      "learning_rate": 3.679699248120301e-06,
+      "loss": 0.1566,
+      "step": 42030
+    },
+    {
+      "epoch": 63.22,
+      "grad_norm": 4.835606098175049,
+      "learning_rate": 3.6781954887218046e-06,
+      "loss": 0.1499,
+      "step": 42040
+    },
+    {
+      "epoch": 63.23,
+      "grad_norm": 4.858092308044434,
+      "learning_rate": 3.6766917293233085e-06,
+      "loss": 0.1637,
+      "step": 42050
+    },
+    {
+      "epoch": 63.25,
+      "grad_norm": 3.878875970840454,
+      "learning_rate": 3.675187969924812e-06,
+      "loss": 0.2207,
+      "step": 42060
+    },
+    {
+      "epoch": 63.26,
+      "grad_norm": 4.815737724304199,
+      "learning_rate": 3.673684210526316e-06,
+      "loss": 0.1478,
+      "step": 42070
+    },
+    {
+      "epoch": 63.28,
+      "grad_norm": 5.654510498046875,
+      "learning_rate": 3.67218045112782e-06,
+      "loss": 0.1988,
+      "step": 42080
+    },
+    {
+      "epoch": 63.29,
+      "grad_norm": 7.474091053009033,
+      "learning_rate": 3.6706766917293234e-06,
+      "loss": 0.1832,
+      "step": 42090
+    },
+    {
+      "epoch": 63.31,
+      "grad_norm": 8.080961227416992,
+      "learning_rate": 3.6691729323308274e-06,
+      "loss": 0.1956,
+      "step": 42100
+    },
+    {
+      "epoch": 63.32,
+      "grad_norm": 5.121028423309326,
+      "learning_rate": 3.667669172932331e-06,
+      "loss": 0.1844,
+      "step": 42110
+    },
+    {
+      "epoch": 63.34,
+      "grad_norm": 5.617638111114502,
+      "learning_rate": 3.666165413533835e-06,
+      "loss": 0.2037,
+      "step": 42120
+    },
+    {
+      "epoch": 63.35,
+      "grad_norm": 5.1372270584106445,
+      "learning_rate": 3.6646616541353388e-06,
+      "loss": 0.171,
+      "step": 42130
+    },
+    {
+      "epoch": 63.37,
+      "grad_norm": 5.457404136657715,
+      "learning_rate": 3.6631578947368423e-06,
+      "loss": 0.1537,
+      "step": 42140
+    },
+    {
+      "epoch": 63.38,
+      "grad_norm": 9.514487266540527,
+      "learning_rate": 3.6616541353383462e-06,
+      "loss": 0.2131,
+      "step": 42150
+    },
+    {
+      "epoch": 63.4,
+      "grad_norm": 6.3367414474487305,
+      "learning_rate": 3.6601503759398498e-06,
+      "loss": 0.1886,
+      "step": 42160
+    },
+    {
+      "epoch": 63.41,
+      "grad_norm": 7.155886650085449,
+      "learning_rate": 3.6586466165413537e-06,
+      "loss": 0.2358,
+      "step": 42170
+    },
+    {
+      "epoch": 63.43,
+      "grad_norm": 4.044269561767578,
+      "learning_rate": 3.6571428571428576e-06,
+      "loss": 0.271,
+      "step": 42180
+    },
+    {
+      "epoch": 63.44,
+      "grad_norm": 7.966317653656006,
+      "learning_rate": 3.655639097744361e-06,
+      "loss": 0.182,
+      "step": 42190
+    },
+    {
+      "epoch": 63.46,
+      "grad_norm": 8.241473197937012,
+      "learning_rate": 3.654135338345865e-06,
+      "loss": 0.2768,
+      "step": 42200
+    },
+    {
+      "epoch": 63.47,
+      "grad_norm": 5.769531726837158,
+      "learning_rate": 3.6526315789473686e-06,
+      "loss": 0.2628,
+      "step": 42210
+    },
+    {
+      "epoch": 63.49,
+      "grad_norm": 4.371389865875244,
+      "learning_rate": 3.6511278195488726e-06,
+      "loss": 0.2603,
+      "step": 42220
+    },
+    {
+      "epoch": 63.5,
+      "grad_norm": 3.9458630084991455,
+      "learning_rate": 3.649624060150376e-06,
+      "loss": 0.1757,
+      "step": 42230
+    },
+    {
+      "epoch": 63.52,
+      "grad_norm": 8.014843940734863,
+      "learning_rate": 3.64812030075188e-06,
+      "loss": 0.1979,
+      "step": 42240
+    },
+    {
+      "epoch": 63.53,
+      "grad_norm": 4.720882892608643,
+      "learning_rate": 3.646616541353384e-06,
+      "loss": 0.2039,
+      "step": 42250
+    },
+    {
+      "epoch": 63.55,
+      "grad_norm": 13.893559455871582,
+      "learning_rate": 3.6451127819548875e-06,
+      "loss": 0.2065,
+      "step": 42260
+    },
+    {
+      "epoch": 63.56,
+      "grad_norm": 5.4362406730651855,
+      "learning_rate": 3.6436090225563914e-06,
+      "loss": 0.169,
+      "step": 42270
+    },
+    {
+      "epoch": 63.58,
+      "grad_norm": 4.523813247680664,
+      "learning_rate": 3.642105263157895e-06,
+      "loss": 0.2152,
+      "step": 42280
+    },
+    {
+      "epoch": 63.59,
+      "grad_norm": 2.5966227054595947,
+      "learning_rate": 3.640601503759399e-06,
+      "loss": 0.1696,
+      "step": 42290
+    },
+    {
+      "epoch": 63.61,
+      "grad_norm": 5.089998245239258,
+      "learning_rate": 3.639097744360903e-06,
+      "loss": 0.1484,
+      "step": 42300
+    },
+    {
+      "epoch": 63.62,
+      "grad_norm": 5.716888427734375,
+      "learning_rate": 3.6375939849624064e-06,
+      "loss": 0.1763,
+      "step": 42310
+    },
+    {
+      "epoch": 63.64,
+      "grad_norm": 5.487515926361084,
+      "learning_rate": 3.6360902255639103e-06,
+      "loss": 0.1831,
+      "step": 42320
+    },
+    {
+      "epoch": 63.65,
+      "grad_norm": 3.316849708557129,
+      "learning_rate": 3.634586466165414e-06,
+      "loss": 0.2312,
+      "step": 42330
+    },
+    {
+      "epoch": 63.67,
+      "grad_norm": 5.662327766418457,
+      "learning_rate": 3.6330827067669178e-06,
+      "loss": 0.2383,
+      "step": 42340
+    },
+    {
+      "epoch": 63.68,
+      "grad_norm": 6.571043968200684,
+      "learning_rate": 3.6315789473684217e-06,
+      "loss": 0.2039,
+      "step": 42350
+    },
+    {
+      "epoch": 63.7,
+      "grad_norm": 5.3168044090271,
+      "learning_rate": 3.6300751879699252e-06,
+      "loss": 0.1692,
+      "step": 42360
+    },
+    {
+      "epoch": 63.71,
+      "grad_norm": 3.067420482635498,
+      "learning_rate": 3.628571428571429e-06,
+      "loss": 0.1852,
+      "step": 42370
+    },
+    {
+      "epoch": 63.73,
+      "grad_norm": 8.511592864990234,
+      "learning_rate": 3.6270676691729327e-06,
+      "loss": 0.1966,
+      "step": 42380
+    },
+    {
+      "epoch": 63.74,
+      "grad_norm": 6.0220232009887695,
+      "learning_rate": 3.6255639097744366e-06,
+      "loss": 0.1325,
+      "step": 42390
+    },
+    {
+      "epoch": 63.76,
+      "grad_norm": 5.685608386993408,
+      "learning_rate": 3.6240601503759406e-06,
+      "loss": 0.228,
+      "step": 42400
+    },
+    {
+      "epoch": 63.77,
+      "grad_norm": 6.7746052742004395,
+      "learning_rate": 3.622556390977444e-06,
+      "loss": 0.2469,
+      "step": 42410
+    },
+    {
+      "epoch": 63.79,
+      "grad_norm": 5.100920677185059,
+      "learning_rate": 3.621052631578948e-06,
+      "loss": 0.1907,
+      "step": 42420
+    },
+    {
+      "epoch": 63.8,
+      "grad_norm": 3.674767255783081,
+      "learning_rate": 3.6195488721804515e-06,
+      "loss": 0.1922,
+      "step": 42430
+    },
+    {
+      "epoch": 63.82,
+      "grad_norm": 7.852967739105225,
+      "learning_rate": 3.618045112781955e-06,
+      "loss": 0.1973,
+      "step": 42440
+    },
+    {
+      "epoch": 63.83,
+      "grad_norm": 3.643383741378784,
+      "learning_rate": 3.6165413533834586e-06,
+      "loss": 0.1633,
+      "step": 42450
+    },
+    {
+      "epoch": 63.85,
+      "grad_norm": 5.850439071655273,
+      "learning_rate": 3.6150375939849625e-06,
+      "loss": 0.1915,
+      "step": 42460
+    },
+    {
+      "epoch": 63.86,
+      "grad_norm": 4.816995620727539,
+      "learning_rate": 3.613533834586466e-06,
+      "loss": 0.2416,
+      "step": 42470
+    },
+    {
+      "epoch": 63.88,
+      "grad_norm": 4.024276256561279,
+      "learning_rate": 3.61203007518797e-06,
+      "loss": 0.207,
+      "step": 42480
+    },
+    {
+      "epoch": 63.89,
+      "grad_norm": 4.65853214263916,
+      "learning_rate": 3.610526315789474e-06,
+      "loss": 0.1428,
+      "step": 42490
+    },
+    {
+      "epoch": 63.91,
+      "grad_norm": 1.6227343082427979,
+      "learning_rate": 3.6090225563909775e-06,
+      "loss": 0.2439,
+      "step": 42500
+    },
+    {
+      "epoch": 63.92,
+      "grad_norm": 5.130610466003418,
+      "learning_rate": 3.6075187969924814e-06,
+      "loss": 0.2491,
+      "step": 42510
+    },
+    {
+      "epoch": 63.94,
+      "grad_norm": 7.856945991516113,
+      "learning_rate": 3.606015037593985e-06,
+      "loss": 0.2532,
+      "step": 42520
+    },
+    {
+      "epoch": 63.95,
+      "grad_norm": 3.6116113662719727,
+      "learning_rate": 3.604511278195489e-06,
+      "loss": 0.258,
+      "step": 42530
+    },
+    {
+      "epoch": 63.97,
+      "grad_norm": 6.224160671234131,
+      "learning_rate": 3.603007518796993e-06,
+      "loss": 0.2729,
+      "step": 42540
+    },
+    {
+      "epoch": 63.98,
+      "grad_norm": 5.5140156745910645,
+      "learning_rate": 3.6015037593984963e-06,
+      "loss": 0.1669,
+      "step": 42550
+    },
+    {
+      "epoch": 64.0,
+      "grad_norm": 13.927995681762695,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 0.2297,
+      "step": 42560
+    },
+    {
+      "epoch": 64.0,
+      "eval_accuracy": 0.9293,
+      "eval_loss": 0.3231370151042938,
+      "eval_runtime": 85.3183,
+      "eval_samples_per_second": 117.208,
+      "eval_steps_per_second": 0.469,
+      "step": 42560
+    },
+    {
+      "epoch": 64.02,
+      "grad_norm": 8.015310287475586,
+      "learning_rate": 3.5984962406015038e-06,
+      "loss": 0.2013,
+      "step": 42570
+    },
+    {
+      "epoch": 64.03,
+      "grad_norm": 7.057667255401611,
+      "learning_rate": 3.5969924812030077e-06,
+      "loss": 0.2238,
+      "step": 42580
+    },
+    {
+      "epoch": 64.05,
+      "grad_norm": 4.543213367462158,
+      "learning_rate": 3.5954887218045117e-06,
+      "loss": 0.1907,
+      "step": 42590
+    },
+    {
+      "epoch": 64.06,
+      "grad_norm": 2.699389934539795,
+      "learning_rate": 3.593984962406015e-06,
+      "loss": 0.1875,
+      "step": 42600
+    },
+    {
+      "epoch": 64.08,
+      "grad_norm": 10.956419944763184,
+      "learning_rate": 3.592481203007519e-06,
+      "loss": 0.2209,
+      "step": 42610
+    },
+    {
+      "epoch": 64.09,
+      "grad_norm": 3.574812173843384,
+      "learning_rate": 3.5909774436090226e-06,
+      "loss": 0.2168,
+      "step": 42620
+    },
+    {
+      "epoch": 64.11,
+      "grad_norm": 4.676037311553955,
+      "learning_rate": 3.5894736842105266e-06,
+      "loss": 0.1879,
+      "step": 42630
+    },
+    {
+      "epoch": 64.12,
+      "grad_norm": 3.858914375305176,
+      "learning_rate": 3.5879699248120305e-06,
+      "loss": 0.2086,
+      "step": 42640
+    },
+    {
+      "epoch": 64.14,
+      "grad_norm": 4.243492126464844,
+      "learning_rate": 3.586466165413534e-06,
+      "loss": 0.2135,
+      "step": 42650
+    },
+    {
+      "epoch": 64.15,
+      "grad_norm": 5.530261039733887,
+      "learning_rate": 3.584962406015038e-06,
+      "loss": 0.2014,
+      "step": 42660
+    },
+    {
+      "epoch": 64.17,
+      "grad_norm": 5.095816135406494,
+      "learning_rate": 3.5834586466165415e-06,
+      "loss": 0.1919,
+      "step": 42670
+    },
+    {
+      "epoch": 64.18,
+      "grad_norm": 2.273388624191284,
+      "learning_rate": 3.5819548872180455e-06,
+      "loss": 0.15,
+      "step": 42680
+    },
+    {
+      "epoch": 64.2,
+      "grad_norm": 5.889043807983398,
+      "learning_rate": 3.580451127819549e-06,
+      "loss": 0.2731,
+      "step": 42690
+    },
+    {
+      "epoch": 64.21,
+      "grad_norm": 3.004957437515259,
+      "learning_rate": 3.578947368421053e-06,
+      "loss": 0.2121,
+      "step": 42700
+    },
+    {
+      "epoch": 64.23,
+      "grad_norm": 3.732741594314575,
+      "learning_rate": 3.577443609022557e-06,
+      "loss": 0.1757,
+      "step": 42710
+    },
+    {
+      "epoch": 64.24,
+      "grad_norm": 4.3243584632873535,
+      "learning_rate": 3.5759398496240604e-06,
+      "loss": 0.2344,
+      "step": 42720
+    },
+    {
+      "epoch": 64.26,
+      "grad_norm": 4.092397689819336,
+      "learning_rate": 3.5744360902255643e-06,
+      "loss": 0.1653,
+      "step": 42730
+    },
+    {
+      "epoch": 64.27,
+      "grad_norm": 5.2698187828063965,
+      "learning_rate": 3.572932330827068e-06,
+      "loss": 0.19,
+      "step": 42740
+    },
+    {
+      "epoch": 64.29,
+      "grad_norm": 7.864371299743652,
+      "learning_rate": 3.5714285714285718e-06,
+      "loss": 0.1977,
+      "step": 42750
+    },
+    {
+      "epoch": 64.3,
+      "grad_norm": 2.6256632804870605,
+      "learning_rate": 3.5699248120300757e-06,
+      "loss": 0.2334,
+      "step": 42760
+    },
+    {
+      "epoch": 64.32,
+      "grad_norm": 2.710421085357666,
+      "learning_rate": 3.5684210526315792e-06,
+      "loss": 0.1919,
+      "step": 42770
+    },
+    {
+      "epoch": 64.33,
+      "grad_norm": 5.321908950805664,
+      "learning_rate": 3.566917293233083e-06,
+      "loss": 0.1552,
+      "step": 42780
+    },
+    {
+      "epoch": 64.35,
+      "grad_norm": 6.268770694732666,
+      "learning_rate": 3.5654135338345867e-06,
+      "loss": 0.1363,
+      "step": 42790
+    },
+    {
+      "epoch": 64.36,
+      "grad_norm": 4.892660617828369,
+      "learning_rate": 3.5639097744360906e-06,
+      "loss": 0.1603,
+      "step": 42800
+    },
+    {
+      "epoch": 64.38,
+      "grad_norm": 5.775477409362793,
+      "learning_rate": 3.5624060150375946e-06,
+      "loss": 0.2558,
+      "step": 42810
+    },
+    {
+      "epoch": 64.39,
+      "grad_norm": 7.529040813446045,
+      "learning_rate": 3.560902255639098e-06,
+      "loss": 0.2444,
+      "step": 42820
+    },
+    {
+      "epoch": 64.41,
+      "grad_norm": 2.752485513687134,
+      "learning_rate": 3.559398496240602e-06,
+      "loss": 0.2147,
+      "step": 42830
+    },
+    {
+      "epoch": 64.42,
+      "grad_norm": 7.767008304595947,
+      "learning_rate": 3.5578947368421056e-06,
+      "loss": 0.2318,
+      "step": 42840
+    },
+    {
+      "epoch": 64.44,
+      "grad_norm": 6.706613063812256,
+      "learning_rate": 3.5563909774436095e-06,
+      "loss": 0.192,
+      "step": 42850
+    },
+    {
+      "epoch": 64.45,
+      "grad_norm": 3.0275185108184814,
+      "learning_rate": 3.5548872180451135e-06,
+      "loss": 0.1605,
+      "step": 42860
+    },
+    {
+      "epoch": 64.47,
+      "grad_norm": 4.424815654754639,
+      "learning_rate": 3.553383458646617e-06,
+      "loss": 0.1919,
+      "step": 42870
+    },
+    {
+      "epoch": 64.48,
+      "grad_norm": 4.253983020782471,
+      "learning_rate": 3.551879699248121e-06,
+      "loss": 0.2079,
+      "step": 42880
+    },
+    {
+      "epoch": 64.5,
+      "grad_norm": 4.817580699920654,
+      "learning_rate": 3.5503759398496244e-06,
+      "loss": 0.1557,
+      "step": 42890
+    },
+    {
+      "epoch": 64.51,
+      "grad_norm": 10.810401916503906,
+      "learning_rate": 3.5488721804511284e-06,
+      "loss": 0.1925,
+      "step": 42900
+    },
+    {
+      "epoch": 64.53,
+      "grad_norm": 4.242228984832764,
+      "learning_rate": 3.5473684210526323e-06,
+      "loss": 0.1776,
+      "step": 42910
+    },
+    {
+      "epoch": 64.54,
+      "grad_norm": 3.9990408420562744,
+      "learning_rate": 3.545864661654136e-06,
+      "loss": 0.1627,
+      "step": 42920
+    },
+    {
+      "epoch": 64.56,
+      "grad_norm": 5.8364105224609375,
+      "learning_rate": 3.544360902255639e-06,
+      "loss": 0.1859,
+      "step": 42930
+    },
+    {
+      "epoch": 64.57,
+      "grad_norm": 8.85914134979248,
+      "learning_rate": 3.542857142857143e-06,
+      "loss": 0.2357,
+      "step": 42940
+    },
+    {
+      "epoch": 64.59,
+      "grad_norm": 6.398134708404541,
+      "learning_rate": 3.541353383458647e-06,
+      "loss": 0.2228,
+      "step": 42950
+    },
+    {
+      "epoch": 64.6,
+      "grad_norm": 1.9391348361968994,
+      "learning_rate": 3.5398496240601503e-06,
+      "loss": 0.1869,
+      "step": 42960
+    },
+    {
+      "epoch": 64.62,
+      "grad_norm": 4.14783239364624,
+      "learning_rate": 3.5383458646616543e-06,
+      "loss": 0.2012,
+      "step": 42970
+    },
+    {
+      "epoch": 64.63,
+      "grad_norm": 2.7045633792877197,
+      "learning_rate": 3.536842105263158e-06,
+      "loss": 0.1754,
+      "step": 42980
+    },
+    {
+      "epoch": 64.65,
+      "grad_norm": 4.491335391998291,
+      "learning_rate": 3.5353383458646617e-06,
+      "loss": 0.2195,
+      "step": 42990
+    },
+    {
+      "epoch": 64.66,
+      "grad_norm": 5.1978678703308105,
+      "learning_rate": 3.5338345864661657e-06,
+      "loss": 0.1891,
+      "step": 43000
+    },
+    {
+      "epoch": 64.68,
+      "grad_norm": 1.4553602933883667,
+      "learning_rate": 3.532330827067669e-06,
+      "loss": 0.1871,
+      "step": 43010
+    },
+    {
+      "epoch": 64.69,
+      "grad_norm": 3.4015984535217285,
+      "learning_rate": 3.530827067669173e-06,
+      "loss": 0.2434,
+      "step": 43020
+    },
+    {
+      "epoch": 64.71,
+      "grad_norm": 6.651217937469482,
+      "learning_rate": 3.5293233082706767e-06,
+      "loss": 0.1889,
+      "step": 43030
+    },
+    {
+      "epoch": 64.72,
+      "grad_norm": 4.7721848487854,
+      "learning_rate": 3.5278195488721806e-06,
+      "loss": 0.222,
+      "step": 43040
+    },
+    {
+      "epoch": 64.74,
+      "grad_norm": 2.7029495239257812,
+      "learning_rate": 3.5263157894736846e-06,
+      "loss": 0.1711,
+      "step": 43050
+    },
+    {
+      "epoch": 64.75,
+      "grad_norm": 11.2900972366333,
+      "learning_rate": 3.524812030075188e-06,
+      "loss": 0.19,
+      "step": 43060
+    },
+    {
+      "epoch": 64.77,
+      "grad_norm": 2.7460989952087402,
+      "learning_rate": 3.523308270676692e-06,
+      "loss": 0.213,
+      "step": 43070
+    },
+    {
+      "epoch": 64.78,
+      "grad_norm": 3.8076796531677246,
+      "learning_rate": 3.5218045112781955e-06,
+      "loss": 0.1665,
+      "step": 43080
+    },
+    {
+      "epoch": 64.8,
+      "grad_norm": 4.072009086608887,
+      "learning_rate": 3.5203007518796995e-06,
+      "loss": 0.1745,
+      "step": 43090
+    },
+    {
+      "epoch": 64.81,
+      "grad_norm": 2.836899995803833,
+      "learning_rate": 3.5187969924812034e-06,
+      "loss": 0.1815,
+      "step": 43100
+    },
+    {
+      "epoch": 64.83,
+      "grad_norm": 6.04483699798584,
+      "learning_rate": 3.517293233082707e-06,
+      "loss": 0.1649,
+      "step": 43110
+    },
+    {
+      "epoch": 64.84,
+      "grad_norm": 3.882234811782837,
+      "learning_rate": 3.515789473684211e-06,
+      "loss": 0.1793,
+      "step": 43120
+    },
+    {
+      "epoch": 64.86,
+      "grad_norm": 6.106405735015869,
+      "learning_rate": 3.5142857142857144e-06,
+      "loss": 0.2961,
+      "step": 43130
+    },
+    {
+      "epoch": 64.87,
+      "grad_norm": 3.823786973953247,
+      "learning_rate": 3.5127819548872183e-06,
+      "loss": 0.1846,
+      "step": 43140
+    },
+    {
+      "epoch": 64.89,
+      "grad_norm": 4.050282955169678,
+      "learning_rate": 3.511278195488722e-06,
+      "loss": 0.2784,
+      "step": 43150
+    },
+    {
+      "epoch": 64.9,
+      "grad_norm": 4.701742649078369,
+      "learning_rate": 3.509774436090226e-06,
+      "loss": 0.1875,
+      "step": 43160
+    },
+    {
+      "epoch": 64.92,
+      "grad_norm": 6.7218708992004395,
+      "learning_rate": 3.5082706766917297e-06,
+      "loss": 0.1865,
+      "step": 43170
+    },
+    {
+      "epoch": 64.93,
+      "grad_norm": 4.6365132331848145,
+      "learning_rate": 3.5067669172932333e-06,
+      "loss": 0.2224,
+      "step": 43180
+    },
+    {
+      "epoch": 64.95,
+      "grad_norm": 8.348788261413574,
+      "learning_rate": 3.505263157894737e-06,
+      "loss": 0.2309,
+      "step": 43190
+    },
+    {
+      "epoch": 64.96,
+      "grad_norm": 6.804657936096191,
+      "learning_rate": 3.5037593984962407e-06,
+      "loss": 0.1802,
+      "step": 43200
+    },
+    {
+      "epoch": 64.98,
+      "grad_norm": 6.6085405349731445,
+      "learning_rate": 3.5022556390977447e-06,
+      "loss": 0.2321,
+      "step": 43210
+    },
+    {
+      "epoch": 64.99,
+      "grad_norm": 6.796005725860596,
+      "learning_rate": 3.5007518796992486e-06,
+      "loss": 0.2108,
+      "step": 43220
+    },
+    {
+      "epoch": 65.0,
+      "eval_accuracy": 0.9313,
+      "eval_loss": 0.31611478328704834,
+      "eval_runtime": 84.6549,
+      "eval_samples_per_second": 118.127,
+      "eval_steps_per_second": 0.473,
+      "step": 43225
+    },
+    {
+      "epoch": 65.01,
+      "grad_norm": 33.686763763427734,
+      "learning_rate": 3.499248120300752e-06,
+      "loss": 0.2085,
+      "step": 43230
+    },
+    {
+      "epoch": 65.02,
+      "grad_norm": 4.142588138580322,
+      "learning_rate": 3.497744360902256e-06,
+      "loss": 0.1216,
+      "step": 43240
+    },
+    {
+      "epoch": 65.04,
+      "grad_norm": 4.3597612380981445,
+      "learning_rate": 3.4962406015037596e-06,
+      "loss": 0.202,
+      "step": 43250
+    },
+    {
+      "epoch": 65.05,
+      "grad_norm": 5.954403877258301,
+      "learning_rate": 3.4947368421052635e-06,
+      "loss": 0.1543,
+      "step": 43260
+    },
+    {
+      "epoch": 65.07,
+      "grad_norm": 4.9531474113464355,
+      "learning_rate": 3.4932330827067675e-06,
+      "loss": 0.2382,
+      "step": 43270
+    },
+    {
+      "epoch": 65.08,
+      "grad_norm": 4.433995723724365,
+      "learning_rate": 3.491729323308271e-06,
+      "loss": 0.2213,
+      "step": 43280
+    },
+    {
+      "epoch": 65.1,
+      "grad_norm": 9.919368743896484,
+      "learning_rate": 3.490225563909775e-06,
+      "loss": 0.2256,
+      "step": 43290
+    },
+    {
+      "epoch": 65.11,
+      "grad_norm": 1.1881593465805054,
+      "learning_rate": 3.4887218045112785e-06,
+      "loss": 0.2056,
+      "step": 43300
+    },
+    {
+      "epoch": 65.13,
+      "grad_norm": 5.632315635681152,
+      "learning_rate": 3.4872180451127824e-06,
+      "loss": 0.2989,
+      "step": 43310
+    },
+    {
+      "epoch": 65.14,
+      "grad_norm": 7.14915132522583,
+      "learning_rate": 3.4857142857142863e-06,
+      "loss": 0.2547,
+      "step": 43320
+    },
+    {
+      "epoch": 65.16,
+      "grad_norm": 3.911741018295288,
+      "learning_rate": 3.48421052631579e-06,
+      "loss": 0.2429,
+      "step": 43330
+    },
+    {
+      "epoch": 65.17,
+      "grad_norm": 2.1838467121124268,
+      "learning_rate": 3.482706766917294e-06,
+      "loss": 0.1845,
+      "step": 43340
+    },
+    {
+      "epoch": 65.19,
+      "grad_norm": 3.118250846862793,
+      "learning_rate": 3.4812030075187973e-06,
+      "loss": 0.2137,
+      "step": 43350
+    },
+    {
+      "epoch": 65.2,
+      "grad_norm": 4.089922904968262,
+      "learning_rate": 3.4796992481203013e-06,
+      "loss": 0.1868,
+      "step": 43360
+    },
+    {
+      "epoch": 65.22,
+      "grad_norm": 5.602779388427734,
+      "learning_rate": 3.478195488721805e-06,
+      "loss": 0.17,
+      "step": 43370
+    },
+    {
+      "epoch": 65.23,
+      "grad_norm": 8.60183334350586,
+      "learning_rate": 3.4766917293233087e-06,
+      "loss": 0.2105,
+      "step": 43380
+    },
+    {
+      "epoch": 65.25,
+      "grad_norm": 5.107520580291748,
+      "learning_rate": 3.4751879699248127e-06,
+      "loss": 0.2588,
+      "step": 43390
+    },
+    {
+      "epoch": 65.26,
+      "grad_norm": 10.866013526916504,
+      "learning_rate": 3.473684210526316e-06,
+      "loss": 0.1978,
+      "step": 43400
+    },
+    {
+      "epoch": 65.28,
+      "grad_norm": 4.005733489990234,
+      "learning_rate": 3.47218045112782e-06,
+      "loss": 0.1811,
+      "step": 43410
+    },
+    {
+      "epoch": 65.29,
+      "grad_norm": 3.853426694869995,
+      "learning_rate": 3.470676691729324e-06,
+      "loss": 0.2217,
+      "step": 43420
+    },
+    {
+      "epoch": 65.31,
+      "grad_norm": 5.65993070602417,
+      "learning_rate": 3.469172932330827e-06,
+      "loss": 0.216,
+      "step": 43430
+    },
+    {
+      "epoch": 65.32,
+      "grad_norm": 3.9332337379455566,
+      "learning_rate": 3.4676691729323307e-06,
+      "loss": 0.2144,
+      "step": 43440
+    },
+    {
+      "epoch": 65.34,
+      "grad_norm": 5.378814697265625,
+      "learning_rate": 3.4661654135338346e-06,
+      "loss": 0.1719,
+      "step": 43450
+    },
+    {
+      "epoch": 65.35,
+      "grad_norm": 4.357859134674072,
+      "learning_rate": 3.4646616541353386e-06,
+      "loss": 0.1897,
+      "step": 43460
+    },
+    {
+      "epoch": 65.37,
+      "grad_norm": 3.1881186962127686,
+      "learning_rate": 3.463157894736842e-06,
+      "loss": 0.1678,
+      "step": 43470
+    },
+    {
+      "epoch": 65.38,
+      "grad_norm": 3.442866325378418,
+      "learning_rate": 3.461654135338346e-06,
+      "loss": 0.1805,
+      "step": 43480
+    },
+    {
+      "epoch": 65.4,
+      "grad_norm": 9.554397583007812,
+      "learning_rate": 3.4601503759398496e-06,
+      "loss": 0.178,
+      "step": 43490
+    },
+    {
+      "epoch": 65.41,
+      "grad_norm": 3.902580976486206,
+      "learning_rate": 3.4586466165413535e-06,
+      "loss": 0.1855,
+      "step": 43500
+    },
+    {
+      "epoch": 65.43,
+      "grad_norm": 5.2838850021362305,
+      "learning_rate": 3.4571428571428574e-06,
+      "loss": 0.2158,
+      "step": 43510
+    },
+    {
+      "epoch": 65.44,
+      "grad_norm": 3.7892587184906006,
+      "learning_rate": 3.455639097744361e-06,
+      "loss": 0.1871,
+      "step": 43520
+    },
+    {
+      "epoch": 65.46,
+      "grad_norm": 3.122152090072632,
+      "learning_rate": 3.454135338345865e-06,
+      "loss": 0.142,
+      "step": 43530
+    },
+    {
+      "epoch": 65.47,
+      "grad_norm": 5.339186191558838,
+      "learning_rate": 3.4526315789473684e-06,
+      "loss": 0.1896,
+      "step": 43540
+    },
+    {
+      "epoch": 65.49,
+      "grad_norm": 6.880535125732422,
+      "learning_rate": 3.4511278195488724e-06,
+      "loss": 0.238,
+      "step": 43550
+    },
+    {
+      "epoch": 65.5,
+      "grad_norm": 4.716340065002441,
+      "learning_rate": 3.4496240601503763e-06,
+      "loss": 0.1584,
+      "step": 43560
+    },
+    {
+      "epoch": 65.52,
+      "grad_norm": 5.00018835067749,
+      "learning_rate": 3.44812030075188e-06,
+      "loss": 0.211,
+      "step": 43570
+    },
+    {
+      "epoch": 65.53,
+      "grad_norm": 2.4130773544311523,
+      "learning_rate": 3.4466165413533838e-06,
+      "loss": 0.2621,
+      "step": 43580
+    },
+    {
+      "epoch": 65.55,
+      "grad_norm": 0.7229984998703003,
+      "learning_rate": 3.4451127819548873e-06,
+      "loss": 0.213,
+      "step": 43590
+    },
+    {
+      "epoch": 65.56,
+      "grad_norm": 7.815069675445557,
+      "learning_rate": 3.4436090225563912e-06,
+      "loss": 0.2066,
+      "step": 43600
+    },
+    {
+      "epoch": 65.58,
+      "grad_norm": 4.840019702911377,
+      "learning_rate": 3.4421052631578947e-06,
+      "loss": 0.182,
+      "step": 43610
+    },
+    {
+      "epoch": 65.59,
+      "grad_norm": 5.94028377532959,
+      "learning_rate": 3.4406015037593987e-06,
+      "loss": 0.1918,
+      "step": 43620
+    },
+    {
+      "epoch": 65.61,
+      "grad_norm": 5.4512939453125,
+      "learning_rate": 3.4390977443609026e-06,
+      "loss": 0.2402,
+      "step": 43630
+    },
+    {
+      "epoch": 65.62,
+      "grad_norm": 4.47932243347168,
+      "learning_rate": 3.437593984962406e-06,
+      "loss": 0.206,
+      "step": 43640
+    },
+    {
+      "epoch": 65.64,
+      "grad_norm": 5.5348310470581055,
+      "learning_rate": 3.43609022556391e-06,
+      "loss": 0.238,
+      "step": 43650
+    },
+    {
+      "epoch": 65.65,
+      "grad_norm": 6.362865924835205,
+      "learning_rate": 3.4345864661654136e-06,
+      "loss": 0.1752,
+      "step": 43660
+    },
+    {
+      "epoch": 65.67,
+      "grad_norm": 6.699802875518799,
+      "learning_rate": 3.4330827067669176e-06,
+      "loss": 0.2295,
+      "step": 43670
+    },
+    {
+      "epoch": 65.68,
+      "grad_norm": 2.2860307693481445,
+      "learning_rate": 3.4315789473684215e-06,
+      "loss": 0.1659,
+      "step": 43680
+    },
+    {
+      "epoch": 65.7,
+      "grad_norm": 2.297809362411499,
+      "learning_rate": 3.430075187969925e-06,
+      "loss": 0.1341,
+      "step": 43690
+    },
+    {
+      "epoch": 65.71,
+      "grad_norm": 7.806978702545166,
+      "learning_rate": 3.428571428571429e-06,
+      "loss": 0.1756,
+      "step": 43700
+    },
+    {
+      "epoch": 65.73,
+      "grad_norm": 8.060872077941895,
+      "learning_rate": 3.4270676691729325e-06,
+      "loss": 0.208,
+      "step": 43710
+    },
+    {
+      "epoch": 65.74,
+      "grad_norm": 4.351926803588867,
+      "learning_rate": 3.4255639097744364e-06,
+      "loss": 0.1765,
+      "step": 43720
+    },
+    {
+      "epoch": 65.76,
+      "grad_norm": 6.094818592071533,
+      "learning_rate": 3.4240601503759404e-06,
+      "loss": 0.219,
+      "step": 43730
+    },
+    {
+      "epoch": 65.77,
+      "grad_norm": 5.638463497161865,
+      "learning_rate": 3.422556390977444e-06,
+      "loss": 0.2396,
+      "step": 43740
+    },
+    {
+      "epoch": 65.79,
+      "grad_norm": 3.789339303970337,
+      "learning_rate": 3.421052631578948e-06,
+      "loss": 0.1845,
+      "step": 43750
+    },
+    {
+      "epoch": 65.8,
+      "grad_norm": 7.909036636352539,
+      "learning_rate": 3.4195488721804513e-06,
+      "loss": 0.2344,
+      "step": 43760
+    },
+    {
+      "epoch": 65.82,
+      "grad_norm": 6.027281761169434,
+      "learning_rate": 3.4180451127819553e-06,
+      "loss": 0.1474,
+      "step": 43770
+    },
+    {
+      "epoch": 65.83,
+      "grad_norm": 5.84333610534668,
+      "learning_rate": 3.4165413533834592e-06,
+      "loss": 0.147,
+      "step": 43780
+    },
+    {
+      "epoch": 65.85,
+      "grad_norm": 8.227249145507812,
+      "learning_rate": 3.4150375939849627e-06,
+      "loss": 0.2314,
+      "step": 43790
+    },
+    {
+      "epoch": 65.86,
+      "grad_norm": 3.2717232704162598,
+      "learning_rate": 3.4135338345864667e-06,
+      "loss": 0.2419,
+      "step": 43800
+    },
+    {
+      "epoch": 65.88,
+      "grad_norm": 5.8105926513671875,
+      "learning_rate": 3.41203007518797e-06,
+      "loss": 0.2375,
+      "step": 43810
+    },
+    {
+      "epoch": 65.89,
+      "grad_norm": 3.668238878250122,
+      "learning_rate": 3.410526315789474e-06,
+      "loss": 0.234,
+      "step": 43820
+    },
+    {
+      "epoch": 65.91,
+      "grad_norm": 5.884802341461182,
+      "learning_rate": 3.409022556390978e-06,
+      "loss": 0.1602,
+      "step": 43830
+    },
+    {
+      "epoch": 65.92,
+      "grad_norm": 3.720730781555176,
+      "learning_rate": 3.4075187969924816e-06,
+      "loss": 0.2016,
+      "step": 43840
+    },
+    {
+      "epoch": 65.94,
+      "grad_norm": 3.9387388229370117,
+      "learning_rate": 3.4060150375939856e-06,
+      "loss": 0.1904,
+      "step": 43850
+    },
+    {
+      "epoch": 65.95,
+      "grad_norm": 7.595775604248047,
+      "learning_rate": 3.404511278195489e-06,
+      "loss": 0.1867,
+      "step": 43860
+    },
+    {
+      "epoch": 65.97,
+      "grad_norm": 4.808709621429443,
+      "learning_rate": 3.403007518796993e-06,
+      "loss": 0.2034,
+      "step": 43870
+    },
+    {
+      "epoch": 65.98,
+      "grad_norm": 5.746549606323242,
+      "learning_rate": 3.401503759398497e-06,
+      "loss": 0.1736,
+      "step": 43880
+    },
+    {
+      "epoch": 66.0,
+      "grad_norm": 0.12739986181259155,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 0.1696,
+      "step": 43890
+    },
+    {
+      "epoch": 66.0,
+      "eval_accuracy": 0.929,
+      "eval_loss": 0.32690301537513733,
+      "eval_runtime": 84.7083,
+      "eval_samples_per_second": 118.052,
+      "eval_steps_per_second": 0.472,
+      "step": 43890
+    },
+    {
+      "epoch": 66.02,
+      "grad_norm": 7.90481424331665,
+      "learning_rate": 3.3984962406015044e-06,
+      "loss": 0.1554,
+      "step": 43900
+    },
+    {
+      "epoch": 66.03,
+      "grad_norm": 6.900293350219727,
+      "learning_rate": 3.396992481203008e-06,
+      "loss": 0.1998,
+      "step": 43910
+    },
+    {
+      "epoch": 66.05,
+      "grad_norm": 5.227355003356934,
+      "learning_rate": 3.3954887218045115e-06,
+      "loss": 0.1637,
+      "step": 43920
+    },
+    {
+      "epoch": 66.06,
+      "grad_norm": 3.3653359413146973,
+      "learning_rate": 3.393984962406015e-06,
+      "loss": 0.1715,
+      "step": 43930
+    },
+    {
+      "epoch": 66.08,
+      "grad_norm": 7.483799934387207,
+      "learning_rate": 3.392481203007519e-06,
+      "loss": 0.1775,
+      "step": 43940
+    },
+    {
+      "epoch": 66.09,
+      "grad_norm": 6.125503063201904,
+      "learning_rate": 3.3909774436090224e-06,
+      "loss": 0.2052,
+      "step": 43950
+    },
+    {
+      "epoch": 66.11,
+      "grad_norm": 2.947721004486084,
+      "learning_rate": 3.3894736842105264e-06,
+      "loss": 0.1686,
+      "step": 43960
+    },
+    {
+      "epoch": 66.12,
+      "grad_norm": 4.430096626281738,
+      "learning_rate": 3.3879699248120303e-06,
+      "loss": 0.1511,
+      "step": 43970
+    },
+    {
+      "epoch": 66.14,
+      "grad_norm": 4.3853349685668945,
+      "learning_rate": 3.386466165413534e-06,
+      "loss": 0.2122,
+      "step": 43980
+    },
+    {
+      "epoch": 66.15,
+      "grad_norm": 3.0524966716766357,
+      "learning_rate": 3.384962406015038e-06,
+      "loss": 0.138,
+      "step": 43990
+    },
+    {
+      "epoch": 66.17,
+      "grad_norm": 2.938044786453247,
+      "learning_rate": 3.3834586466165413e-06,
+      "loss": 0.2325,
+      "step": 44000
+    },
+    {
+      "epoch": 66.18,
+      "grad_norm": 4.645559787750244,
+      "learning_rate": 3.3819548872180453e-06,
+      "loss": 0.2521,
+      "step": 44010
+    },
+    {
+      "epoch": 66.2,
+      "grad_norm": 7.084641456604004,
+      "learning_rate": 3.380451127819549e-06,
+      "loss": 0.2237,
+      "step": 44020
+    },
+    {
+      "epoch": 66.21,
+      "grad_norm": 8.430667877197266,
+      "learning_rate": 3.3789473684210527e-06,
+      "loss": 0.2012,
+      "step": 44030
+    },
+    {
+      "epoch": 66.23,
+      "grad_norm": 2.169158697128296,
+      "learning_rate": 3.3774436090225567e-06,
+      "loss": 0.1534,
+      "step": 44040
+    },
+    {
+      "epoch": 66.24,
+      "grad_norm": 4.5548553466796875,
+      "learning_rate": 3.37593984962406e-06,
+      "loss": 0.2439,
+      "step": 44050
+    },
+    {
+      "epoch": 66.26,
+      "grad_norm": 3.5269806385040283,
+      "learning_rate": 3.374436090225564e-06,
+      "loss": 0.2261,
+      "step": 44060
+    },
+    {
+      "epoch": 66.27,
+      "grad_norm": 4.3473334312438965,
+      "learning_rate": 3.3729323308270676e-06,
+      "loss": 0.1643,
+      "step": 44070
+    },
+    {
+      "epoch": 66.29,
+      "grad_norm": 1.9924315214157104,
+      "learning_rate": 3.3714285714285716e-06,
+      "loss": 0.1739,
+      "step": 44080
+    },
+    {
+      "epoch": 66.3,
+      "grad_norm": 4.501370906829834,
+      "learning_rate": 3.3699248120300755e-06,
+      "loss": 0.2115,
+      "step": 44090
+    },
+    {
+      "epoch": 66.32,
+      "grad_norm": 9.435755729675293,
+      "learning_rate": 3.368421052631579e-06,
+      "loss": 0.1906,
+      "step": 44100
+    },
+    {
+      "epoch": 66.33,
+      "grad_norm": 5.539870738983154,
+      "learning_rate": 3.366917293233083e-06,
+      "loss": 0.1738,
+      "step": 44110
+    },
+    {
+      "epoch": 66.35,
+      "grad_norm": 6.538062572479248,
+      "learning_rate": 3.3654135338345865e-06,
+      "loss": 0.1997,
+      "step": 44120
+    },
+    {
+      "epoch": 66.36,
+      "grad_norm": 3.478639602661133,
+      "learning_rate": 3.3639097744360904e-06,
+      "loss": 0.2232,
+      "step": 44130
+    },
+    {
+      "epoch": 66.38,
+      "grad_norm": 3.3659911155700684,
+      "learning_rate": 3.3624060150375944e-06,
+      "loss": 0.1294,
+      "step": 44140
+    },
+    {
+      "epoch": 66.39,
+      "grad_norm": 1.8332750797271729,
+      "learning_rate": 3.360902255639098e-06,
+      "loss": 0.171,
+      "step": 44150
+    },
+    {
+      "epoch": 66.41,
+      "grad_norm": 6.612475872039795,
+      "learning_rate": 3.359398496240602e-06,
+      "loss": 0.2437,
+      "step": 44160
+    },
+    {
+      "epoch": 66.42,
+      "grad_norm": 4.539286136627197,
+      "learning_rate": 3.3578947368421054e-06,
+      "loss": 0.1853,
+      "step": 44170
+    },
+    {
+      "epoch": 66.44,
+      "grad_norm": 6.16352653503418,
+      "learning_rate": 3.3563909774436093e-06,
+      "loss": 0.2118,
+      "step": 44180
+    },
+    {
+      "epoch": 66.45,
+      "grad_norm": 6.134418487548828,
+      "learning_rate": 3.3548872180451133e-06,
+      "loss": 0.2087,
+      "step": 44190
+    },
+    {
+      "epoch": 66.47,
+      "grad_norm": 3.634429454803467,
+      "learning_rate": 3.3533834586466168e-06,
+      "loss": 0.1724,
+      "step": 44200
+    },
+    {
+      "epoch": 66.48,
+      "grad_norm": 3.4543323516845703,
+      "learning_rate": 3.3518796992481207e-06,
+      "loss": 0.1575,
+      "step": 44210
+    },
+    {
+      "epoch": 66.5,
+      "grad_norm": 4.991121768951416,
+      "learning_rate": 3.3503759398496242e-06,
+      "loss": 0.2694,
+      "step": 44220
+    },
+    {
+      "epoch": 66.51,
+      "grad_norm": 5.8151936531066895,
+      "learning_rate": 3.348872180451128e-06,
+      "loss": 0.2191,
+      "step": 44230
+    },
+    {
+      "epoch": 66.53,
+      "grad_norm": 4.985229969024658,
+      "learning_rate": 3.347368421052632e-06,
+      "loss": 0.2083,
+      "step": 44240
+    },
+    {
+      "epoch": 66.54,
+      "grad_norm": 4.52158260345459,
+      "learning_rate": 3.3458646616541356e-06,
+      "loss": 0.2116,
+      "step": 44250
+    },
+    {
+      "epoch": 66.56,
+      "grad_norm": 5.498661518096924,
+      "learning_rate": 3.3443609022556396e-06,
+      "loss": 0.2242,
+      "step": 44260
+    },
+    {
+      "epoch": 66.57,
+      "grad_norm": 4.108017444610596,
+      "learning_rate": 3.342857142857143e-06,
+      "loss": 0.2415,
+      "step": 44270
+    },
+    {
+      "epoch": 66.59,
+      "grad_norm": 7.99500846862793,
+      "learning_rate": 3.341353383458647e-06,
+      "loss": 0.2235,
+      "step": 44280
+    },
+    {
+      "epoch": 66.6,
+      "grad_norm": 3.367448329925537,
+      "learning_rate": 3.339849624060151e-06,
+      "loss": 0.1975,
+      "step": 44290
+    },
+    {
+      "epoch": 66.62,
+      "grad_norm": 3.148461103439331,
+      "learning_rate": 3.3383458646616545e-06,
+      "loss": 0.1706,
+      "step": 44300
+    },
+    {
+      "epoch": 66.63,
+      "grad_norm": 5.184333324432373,
+      "learning_rate": 3.3368421052631584e-06,
+      "loss": 0.158,
+      "step": 44310
+    },
+    {
+      "epoch": 66.65,
+      "grad_norm": 8.876405715942383,
+      "learning_rate": 3.335338345864662e-06,
+      "loss": 0.2529,
+      "step": 44320
+    },
+    {
+      "epoch": 66.66,
+      "grad_norm": 1.495922565460205,
+      "learning_rate": 3.333834586466166e-06,
+      "loss": 0.1659,
+      "step": 44330
+    },
+    {
+      "epoch": 66.68,
+      "grad_norm": 4.976419925689697,
+      "learning_rate": 3.3323308270676694e-06,
+      "loss": 0.2705,
+      "step": 44340
+    },
+    {
+      "epoch": 66.69,
+      "grad_norm": 8.409575462341309,
+      "learning_rate": 3.3308270676691734e-06,
+      "loss": 0.237,
+      "step": 44350
+    },
+    {
+      "epoch": 66.71,
+      "grad_norm": 11.252154350280762,
+      "learning_rate": 3.3293233082706773e-06,
+      "loss": 0.225,
+      "step": 44360
+    },
+    {
+      "epoch": 66.72,
+      "grad_norm": 6.9660820960998535,
+      "learning_rate": 3.327819548872181e-06,
+      "loss": 0.1612,
+      "step": 44370
+    },
+    {
+      "epoch": 66.74,
+      "grad_norm": 7.809027671813965,
+      "learning_rate": 3.3263157894736848e-06,
+      "loss": 0.1657,
+      "step": 44380
+    },
+    {
+      "epoch": 66.75,
+      "grad_norm": 4.254601955413818,
+      "learning_rate": 3.3248120300751883e-06,
+      "loss": 0.1746,
+      "step": 44390
+    },
+    {
+      "epoch": 66.77,
+      "grad_norm": 7.9498291015625,
+      "learning_rate": 3.3233082706766922e-06,
+      "loss": 0.1704,
+      "step": 44400
+    },
+    {
+      "epoch": 66.78,
+      "grad_norm": 2.0477476119995117,
+      "learning_rate": 3.321804511278196e-06,
+      "loss": 0.1805,
+      "step": 44410
+    },
+    {
+      "epoch": 66.8,
+      "grad_norm": 4.7251715660095215,
+      "learning_rate": 3.3203007518796993e-06,
+      "loss": 0.1856,
+      "step": 44420
+    },
+    {
+      "epoch": 66.81,
+      "grad_norm": 3.7293553352355957,
+      "learning_rate": 3.3187969924812032e-06,
+      "loss": 0.1233,
+      "step": 44430
+    },
+    {
+      "epoch": 66.83,
+      "grad_norm": 6.081485748291016,
+      "learning_rate": 3.3172932330827067e-06,
+      "loss": 0.1719,
+      "step": 44440
+    },
+    {
+      "epoch": 66.84,
+      "grad_norm": 5.9813079833984375,
+      "learning_rate": 3.3157894736842107e-06,
+      "loss": 0.186,
+      "step": 44450
+    },
+    {
+      "epoch": 66.86,
+      "grad_norm": 6.273433208465576,
+      "learning_rate": 3.314285714285714e-06,
+      "loss": 0.236,
+      "step": 44460
+    },
+    {
+      "epoch": 66.87,
+      "grad_norm": 6.025407314300537,
+      "learning_rate": 3.312781954887218e-06,
+      "loss": 0.2297,
+      "step": 44470
+    },
+    {
+      "epoch": 66.89,
+      "grad_norm": 4.0798659324646,
+      "learning_rate": 3.3112781954887217e-06,
+      "loss": 0.1626,
+      "step": 44480
+    },
+    {
+      "epoch": 66.9,
+      "grad_norm": 1.6640973091125488,
+      "learning_rate": 3.3097744360902256e-06,
+      "loss": 0.1482,
+      "step": 44490
+    },
+    {
+      "epoch": 66.92,
+      "grad_norm": 5.056227207183838,
+      "learning_rate": 3.3082706766917295e-06,
+      "loss": 0.2078,
+      "step": 44500
+    },
+    {
+      "epoch": 66.93,
+      "grad_norm": 4.699299335479736,
+      "learning_rate": 3.306766917293233e-06,
+      "loss": 0.2021,
+      "step": 44510
+    },
+    {
+      "epoch": 66.95,
+      "grad_norm": 2.141939163208008,
+      "learning_rate": 3.305263157894737e-06,
+      "loss": 0.1897,
+      "step": 44520
+    },
+    {
+      "epoch": 66.96,
+      "grad_norm": 4.3747687339782715,
+      "learning_rate": 3.3037593984962405e-06,
+      "loss": 0.2014,
+      "step": 44530
+    },
+    {
+      "epoch": 66.98,
+      "grad_norm": 5.673415660858154,
+      "learning_rate": 3.3022556390977445e-06,
+      "loss": 0.2335,
+      "step": 44540
+    },
+    {
+      "epoch": 66.99,
+      "grad_norm": 2.832839250564575,
+      "learning_rate": 3.3007518796992484e-06,
+      "loss": 0.1946,
+      "step": 44550
+    },
+    {
+      "epoch": 67.0,
+      "eval_accuracy": 0.9302,
+      "eval_loss": 0.330706924200058,
+      "eval_runtime": 84.1739,
+      "eval_samples_per_second": 118.802,
+      "eval_steps_per_second": 0.475,
+      "step": 44555
+    },
+    {
+      "epoch": 67.01,
+      "grad_norm": 5.633582592010498,
+      "learning_rate": 3.299248120300752e-06,
+      "loss": 0.1437,
+      "step": 44560
+    },
+    {
+      "epoch": 67.02,
+      "grad_norm": 3.9651787281036377,
+      "learning_rate": 3.297744360902256e-06,
+      "loss": 0.157,
+      "step": 44570
+    },
+    {
+      "epoch": 67.04,
+      "grad_norm": 3.220197916030884,
+      "learning_rate": 3.2962406015037594e-06,
+      "loss": 0.1883,
+      "step": 44580
+    },
+    {
+      "epoch": 67.05,
+      "grad_norm": 6.965760707855225,
+      "learning_rate": 3.2947368421052633e-06,
+      "loss": 0.2133,
+      "step": 44590
+    },
+    {
+      "epoch": 67.07,
+      "grad_norm": 7.967811107635498,
+      "learning_rate": 3.2932330827067673e-06,
+      "loss": 0.1972,
+      "step": 44600
+    },
+    {
+      "epoch": 67.08,
+      "grad_norm": 8.779460906982422,
+      "learning_rate": 3.291729323308271e-06,
+      "loss": 0.2013,
+      "step": 44610
+    },
+    {
+      "epoch": 67.1,
+      "grad_norm": 4.243956565856934,
+      "learning_rate": 3.2902255639097747e-06,
+      "loss": 0.1606,
+      "step": 44620
+    },
+    {
+      "epoch": 67.11,
+      "grad_norm": 6.18154764175415,
+      "learning_rate": 3.2887218045112783e-06,
+      "loss": 0.231,
+      "step": 44630
+    },
+    {
+      "epoch": 67.13,
+      "grad_norm": 4.825216293334961,
+      "learning_rate": 3.287218045112782e-06,
+      "loss": 0.2034,
+      "step": 44640
+    },
+    {
+      "epoch": 67.14,
+      "grad_norm": 4.287315845489502,
+      "learning_rate": 3.285714285714286e-06,
+      "loss": 0.1463,
+      "step": 44650
+    },
+    {
+      "epoch": 67.16,
+      "grad_norm": 4.091511249542236,
+      "learning_rate": 3.2842105263157897e-06,
+      "loss": 0.1423,
+      "step": 44660
+    },
+    {
+      "epoch": 67.17,
+      "grad_norm": 4.6609978675842285,
+      "learning_rate": 3.2827067669172936e-06,
+      "loss": 0.2448,
+      "step": 44670
+    },
+    {
+      "epoch": 67.19,
+      "grad_norm": 7.273349761962891,
+      "learning_rate": 3.281203007518797e-06,
+      "loss": 0.1618,
+      "step": 44680
+    },
+    {
+      "epoch": 67.2,
+      "grad_norm": 6.079645156860352,
+      "learning_rate": 3.279699248120301e-06,
+      "loss": 0.2034,
+      "step": 44690
+    },
+    {
+      "epoch": 67.22,
+      "grad_norm": 4.420186996459961,
+      "learning_rate": 3.278195488721805e-06,
+      "loss": 0.1605,
+      "step": 44700
+    },
+    {
+      "epoch": 67.23,
+      "grad_norm": 2.5898821353912354,
+      "learning_rate": 3.2766917293233085e-06,
+      "loss": 0.1425,
+      "step": 44710
+    },
+    {
+      "epoch": 67.25,
+      "grad_norm": 5.7673468589782715,
+      "learning_rate": 3.2751879699248125e-06,
+      "loss": 0.1487,
+      "step": 44720
+    },
+    {
+      "epoch": 67.26,
+      "grad_norm": 4.787065505981445,
+      "learning_rate": 3.273684210526316e-06,
+      "loss": 0.2058,
+      "step": 44730
+    },
+    {
+      "epoch": 67.28,
+      "grad_norm": 8.583816528320312,
+      "learning_rate": 3.27218045112782e-06,
+      "loss": 0.194,
+      "step": 44740
+    },
+    {
+      "epoch": 67.29,
+      "grad_norm": 6.30866813659668,
+      "learning_rate": 3.270676691729324e-06,
+      "loss": 0.2036,
+      "step": 44750
+    },
+    {
+      "epoch": 67.31,
+      "grad_norm": 3.9293291568756104,
+      "learning_rate": 3.2691729323308274e-06,
+      "loss": 0.1755,
+      "step": 44760
+    },
+    {
+      "epoch": 67.32,
+      "grad_norm": 6.662651538848877,
+      "learning_rate": 3.2676691729323313e-06,
+      "loss": 0.1482,
+      "step": 44770
+    },
+    {
+      "epoch": 67.34,
+      "grad_norm": 4.203287124633789,
+      "learning_rate": 3.266165413533835e-06,
+      "loss": 0.263,
+      "step": 44780
+    },
+    {
+      "epoch": 67.35,
+      "grad_norm": 5.568451881408691,
+      "learning_rate": 3.264661654135339e-06,
+      "loss": 0.1302,
+      "step": 44790
+    },
+    {
+      "epoch": 67.37,
+      "grad_norm": 11.907421112060547,
+      "learning_rate": 3.2631578947368423e-06,
+      "loss": 0.2896,
+      "step": 44800
+    },
+    {
+      "epoch": 67.38,
+      "grad_norm": 7.300894737243652,
+      "learning_rate": 3.2616541353383463e-06,
+      "loss": 0.2324,
+      "step": 44810
+    },
+    {
+      "epoch": 67.4,
+      "grad_norm": 0.4205104112625122,
+      "learning_rate": 3.26015037593985e-06,
+      "loss": 0.1475,
+      "step": 44820
+    },
+    {
+      "epoch": 67.41,
+      "grad_norm": 2.742250919342041,
+      "learning_rate": 3.2586466165413537e-06,
+      "loss": 0.223,
+      "step": 44830
+    },
+    {
+      "epoch": 67.43,
+      "grad_norm": 4.337477684020996,
+      "learning_rate": 3.2571428571428577e-06,
+      "loss": 0.1887,
+      "step": 44840
+    },
+    {
+      "epoch": 67.44,
+      "grad_norm": 6.059717178344727,
+      "learning_rate": 3.255639097744361e-06,
+      "loss": 0.2286,
+      "step": 44850
+    },
+    {
+      "epoch": 67.46,
+      "grad_norm": 8.457310676574707,
+      "learning_rate": 3.254135338345865e-06,
+      "loss": 0.1812,
+      "step": 44860
+    },
+    {
+      "epoch": 67.47,
+      "grad_norm": 2.976374626159668,
+      "learning_rate": 3.252631578947369e-06,
+      "loss": 0.2129,
+      "step": 44870
+    },
+    {
+      "epoch": 67.49,
+      "grad_norm": 5.421755790710449,
+      "learning_rate": 3.2511278195488726e-06,
+      "loss": 0.1674,
+      "step": 44880
+    },
+    {
+      "epoch": 67.5,
+      "grad_norm": 7.67271614074707,
+      "learning_rate": 3.2496240601503765e-06,
+      "loss": 0.1708,
+      "step": 44890
+    },
+    {
+      "epoch": 67.52,
+      "grad_norm": 7.604538440704346,
+      "learning_rate": 3.24812030075188e-06,
+      "loss": 0.205,
+      "step": 44900
+    },
+    {
+      "epoch": 67.53,
+      "grad_norm": 5.754815578460693,
+      "learning_rate": 3.246616541353384e-06,
+      "loss": 0.1922,
+      "step": 44910
+    },
+    {
+      "epoch": 67.55,
+      "grad_norm": 5.512721061706543,
+      "learning_rate": 3.245112781954887e-06,
+      "loss": 0.1671,
+      "step": 44920
+    },
+    {
+      "epoch": 67.56,
+      "grad_norm": 5.614846706390381,
+      "learning_rate": 3.243609022556391e-06,
+      "loss": 0.1837,
+      "step": 44930
+    },
+    {
+      "epoch": 67.58,
+      "grad_norm": 6.0121612548828125,
+      "learning_rate": 3.2421052631578945e-06,
+      "loss": 0.221,
+      "step": 44940
+    },
+    {
+      "epoch": 67.59,
+      "grad_norm": 4.721285820007324,
+      "learning_rate": 3.2406015037593985e-06,
+      "loss": 0.1704,
+      "step": 44950
+    },
+    {
+      "epoch": 67.61,
+      "grad_norm": 5.494776248931885,
+      "learning_rate": 3.2390977443609024e-06,
+      "loss": 0.2032,
+      "step": 44960
+    },
+    {
+      "epoch": 67.62,
+      "grad_norm": 3.387678861618042,
+      "learning_rate": 3.237593984962406e-06,
+      "loss": 0.2162,
+      "step": 44970
+    },
+    {
+      "epoch": 67.64,
+      "grad_norm": 5.2446417808532715,
+      "learning_rate": 3.23609022556391e-06,
+      "loss": 0.138,
+      "step": 44980
+    },
+    {
+      "epoch": 67.65,
+      "grad_norm": 6.1764349937438965,
+      "learning_rate": 3.2345864661654134e-06,
+      "loss": 0.1532,
+      "step": 44990
+    },
+    {
+      "epoch": 67.67,
+      "grad_norm": 4.519616603851318,
+      "learning_rate": 3.2330827067669174e-06,
+      "loss": 0.1948,
+      "step": 45000
+    },
+    {
+      "epoch": 67.68,
+      "grad_norm": 7.864477634429932,
+      "learning_rate": 3.2315789473684213e-06,
+      "loss": 0.1927,
+      "step": 45010
+    },
+    {
+      "epoch": 67.7,
+      "grad_norm": 2.657796859741211,
+      "learning_rate": 3.230075187969925e-06,
+      "loss": 0.1907,
+      "step": 45020
+    },
+    {
+      "epoch": 67.71,
+      "grad_norm": 6.408271312713623,
+      "learning_rate": 3.2285714285714288e-06,
+      "loss": 0.1843,
+      "step": 45030
+    },
+    {
+      "epoch": 67.73,
+      "grad_norm": 4.096652984619141,
+      "learning_rate": 3.2270676691729323e-06,
+      "loss": 0.174,
+      "step": 45040
+    },
+    {
+      "epoch": 67.74,
+      "grad_norm": 4.657118797302246,
+      "learning_rate": 3.2255639097744362e-06,
+      "loss": 0.1888,
+      "step": 45050
+    },
+    {
+      "epoch": 67.76,
+      "grad_norm": 4.9201860427856445,
+      "learning_rate": 3.22406015037594e-06,
+      "loss": 0.2123,
+      "step": 45060
+    },
+    {
+      "epoch": 67.77,
+      "grad_norm": 5.590874195098877,
+      "learning_rate": 3.2225563909774437e-06,
+      "loss": 0.1487,
+      "step": 45070
+    },
+    {
+      "epoch": 67.79,
+      "grad_norm": 7.1157050132751465,
+      "learning_rate": 3.2210526315789476e-06,
+      "loss": 0.193,
+      "step": 45080
+    },
+    {
+      "epoch": 67.8,
+      "grad_norm": 2.877906560897827,
+      "learning_rate": 3.219548872180451e-06,
+      "loss": 0.1419,
+      "step": 45090
+    },
+    {
+      "epoch": 67.82,
+      "grad_norm": 5.331236839294434,
+      "learning_rate": 3.218045112781955e-06,
+      "loss": 0.1581,
+      "step": 45100
+    },
+    {
+      "epoch": 67.83,
+      "grad_norm": 6.2138991355896,
+      "learning_rate": 3.216541353383459e-06,
+      "loss": 0.201,
+      "step": 45110
+    },
+    {
+      "epoch": 67.85,
+      "grad_norm": 3.055180072784424,
+      "learning_rate": 3.2150375939849625e-06,
+      "loss": 0.1932,
+      "step": 45120
+    },
+    {
+      "epoch": 67.86,
+      "grad_norm": 5.881607532501221,
+      "learning_rate": 3.2135338345864665e-06,
+      "loss": 0.2327,
+      "step": 45130
+    },
+    {
+      "epoch": 67.88,
+      "grad_norm": 7.188892841339111,
+      "learning_rate": 3.21203007518797e-06,
+      "loss": 0.1651,
+      "step": 45140
+    },
+    {
+      "epoch": 67.89,
+      "grad_norm": 3.549654960632324,
+      "learning_rate": 3.210526315789474e-06,
+      "loss": 0.2572,
+      "step": 45150
+    },
+    {
+      "epoch": 67.91,
+      "grad_norm": 9.034676551818848,
+      "learning_rate": 3.209022556390978e-06,
+      "loss": 0.1358,
+      "step": 45160
+    },
+    {
+      "epoch": 67.92,
+      "grad_norm": 4.870662689208984,
+      "learning_rate": 3.2075187969924814e-06,
+      "loss": 0.207,
+      "step": 45170
+    },
+    {
+      "epoch": 67.94,
+      "grad_norm": 1.617182731628418,
+      "learning_rate": 3.2060150375939854e-06,
+      "loss": 0.2525,
+      "step": 45180
+    },
+    {
+      "epoch": 67.95,
+      "grad_norm": 2.620441198348999,
+      "learning_rate": 3.204511278195489e-06,
+      "loss": 0.1688,
+      "step": 45190
+    },
+    {
+      "epoch": 67.97,
+      "grad_norm": 9.695724487304688,
+      "learning_rate": 3.203007518796993e-06,
+      "loss": 0.233,
+      "step": 45200
+    },
+    {
+      "epoch": 67.98,
+      "grad_norm": 6.005362033843994,
+      "learning_rate": 3.2015037593984968e-06,
+      "loss": 0.1813,
+      "step": 45210
+    },
+    {
+      "epoch": 68.0,
+      "grad_norm": 0.01271333172917366,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.1492,
+      "step": 45220
+    },
+    {
+      "epoch": 68.0,
+      "eval_accuracy": 0.9296,
+      "eval_loss": 0.32478708028793335,
+      "eval_runtime": 84.6456,
+      "eval_samples_per_second": 118.14,
+      "eval_steps_per_second": 0.473,
+      "step": 45220
+    },
+    {
+      "epoch": 68.02,
+      "grad_norm": 9.229415893554688,
+      "learning_rate": 3.1984962406015042e-06,
+      "loss": 0.2163,
+      "step": 45230
+    },
+    {
+      "epoch": 68.03,
+      "grad_norm": 5.0831298828125,
+      "learning_rate": 3.1969924812030077e-06,
+      "loss": 0.2027,
+      "step": 45240
+    },
+    {
+      "epoch": 68.05,
+      "grad_norm": 4.618161678314209,
+      "learning_rate": 3.1954887218045117e-06,
+      "loss": 0.2085,
+      "step": 45250
+    },
+    {
+      "epoch": 68.06,
+      "grad_norm": 2.447524070739746,
+      "learning_rate": 3.193984962406015e-06,
+      "loss": 0.1593,
+      "step": 45260
+    },
+    {
+      "epoch": 68.08,
+      "grad_norm": 5.367290496826172,
+      "learning_rate": 3.192481203007519e-06,
+      "loss": 0.1527,
+      "step": 45270
+    },
+    {
+      "epoch": 68.09,
+      "grad_norm": 3.5726702213287354,
+      "learning_rate": 3.190977443609023e-06,
+      "loss": 0.1963,
+      "step": 45280
+    },
+    {
+      "epoch": 68.11,
+      "grad_norm": 6.074667930603027,
+      "learning_rate": 3.1894736842105266e-06,
+      "loss": 0.2226,
+      "step": 45290
+    },
+    {
+      "epoch": 68.12,
+      "grad_norm": 6.223268508911133,
+      "learning_rate": 3.1879699248120305e-06,
+      "loss": 0.277,
+      "step": 45300
+    },
+    {
+      "epoch": 68.14,
+      "grad_norm": 4.411190509796143,
+      "learning_rate": 3.186466165413534e-06,
+      "loss": 0.1662,
+      "step": 45310
+    },
+    {
+      "epoch": 68.15,
+      "grad_norm": 5.630961894989014,
+      "learning_rate": 3.184962406015038e-06,
+      "loss": 0.2317,
+      "step": 45320
+    },
+    {
+      "epoch": 68.17,
+      "grad_norm": 6.117316246032715,
+      "learning_rate": 3.183458646616542e-06,
+      "loss": 0.2044,
+      "step": 45330
+    },
+    {
+      "epoch": 68.18,
+      "grad_norm": 6.058127403259277,
+      "learning_rate": 3.1819548872180455e-06,
+      "loss": 0.1925,
+      "step": 45340
+    },
+    {
+      "epoch": 68.2,
+      "grad_norm": 4.601078033447266,
+      "learning_rate": 3.1804511278195494e-06,
+      "loss": 0.2433,
+      "step": 45350
+    },
+    {
+      "epoch": 68.21,
+      "grad_norm": 11.11681842803955,
+      "learning_rate": 3.178947368421053e-06,
+      "loss": 0.2035,
+      "step": 45360
+    },
+    {
+      "epoch": 68.23,
+      "grad_norm": 4.8060503005981445,
+      "learning_rate": 3.177443609022557e-06,
+      "loss": 0.1838,
+      "step": 45370
+    },
+    {
+      "epoch": 68.24,
+      "grad_norm": 3.692517042160034,
+      "learning_rate": 3.175939849624061e-06,
+      "loss": 0.1765,
+      "step": 45380
+    },
+    {
+      "epoch": 68.26,
+      "grad_norm": 7.551518440246582,
+      "learning_rate": 3.1744360902255643e-06,
+      "loss": 0.1897,
+      "step": 45390
+    },
+    {
+      "epoch": 68.27,
+      "grad_norm": 4.985076427459717,
+      "learning_rate": 3.1729323308270683e-06,
+      "loss": 0.216,
+      "step": 45400
+    },
+    {
+      "epoch": 68.29,
+      "grad_norm": 8.353325843811035,
+      "learning_rate": 3.1714285714285714e-06,
+      "loss": 0.196,
+      "step": 45410
+    },
+    {
+      "epoch": 68.3,
+      "grad_norm": 6.800642490386963,
+      "learning_rate": 3.1699248120300753e-06,
+      "loss": 0.1798,
+      "step": 45420
+    },
+    {
+      "epoch": 68.32,
+      "grad_norm": 7.222949028015137,
+      "learning_rate": 3.168421052631579e-06,
+      "loss": 0.1988,
+      "step": 45430
+    },
+    {
+      "epoch": 68.33,
+      "grad_norm": 7.158170700073242,
+      "learning_rate": 3.1669172932330828e-06,
+      "loss": 0.2191,
+      "step": 45440
+    },
+    {
+      "epoch": 68.35,
+      "grad_norm": 5.271254062652588,
+      "learning_rate": 3.1654135338345863e-06,
+      "loss": 0.2169,
+      "step": 45450
+    },
+    {
+      "epoch": 68.36,
+      "grad_norm": 2.8725428581237793,
+      "learning_rate": 3.1639097744360902e-06,
+      "loss": 0.2091,
+      "step": 45460
+    },
+    {
+      "epoch": 68.38,
+      "grad_norm": 6.836148262023926,
+      "learning_rate": 3.162406015037594e-06,
+      "loss": 0.2319,
+      "step": 45470
+    },
+    {
+      "epoch": 68.39,
+      "grad_norm": 3.735705614089966,
+      "learning_rate": 3.1609022556390977e-06,
+      "loss": 0.2037,
+      "step": 45480
+    },
+    {
+      "epoch": 68.41,
+      "grad_norm": 6.039109706878662,
+      "learning_rate": 3.1593984962406016e-06,
+      "loss": 0.2016,
+      "step": 45490
+    },
+    {
+      "epoch": 68.42,
+      "grad_norm": 6.477905750274658,
+      "learning_rate": 3.157894736842105e-06,
+      "loss": 0.185,
+      "step": 45500
+    },
+    {
+      "epoch": 68.44,
+      "grad_norm": 5.0428900718688965,
+      "learning_rate": 3.156390977443609e-06,
+      "loss": 0.192,
+      "step": 45510
+    },
+    {
+      "epoch": 68.45,
+      "grad_norm": 4.29931640625,
+      "learning_rate": 3.154887218045113e-06,
+      "loss": 0.1585,
+      "step": 45520
+    },
+    {
+      "epoch": 68.47,
+      "grad_norm": 4.26162052154541,
+      "learning_rate": 3.1533834586466166e-06,
+      "loss": 0.1979,
+      "step": 45530
+    },
+    {
+      "epoch": 68.48,
+      "grad_norm": 6.1737380027771,
+      "learning_rate": 3.1518796992481205e-06,
+      "loss": 0.1783,
+      "step": 45540
+    },
+    {
+      "epoch": 68.5,
+      "grad_norm": 3.1564579010009766,
+      "learning_rate": 3.150375939849624e-06,
+      "loss": 0.1443,
+      "step": 45550
+    },
+    {
+      "epoch": 68.51,
+      "grad_norm": 2.920849323272705,
+      "learning_rate": 3.148872180451128e-06,
+      "loss": 0.2208,
+      "step": 45560
+    },
+    {
+      "epoch": 68.53,
+      "grad_norm": 4.952502727508545,
+      "learning_rate": 3.147368421052632e-06,
+      "loss": 0.1836,
+      "step": 45570
+    },
+    {
+      "epoch": 68.54,
+      "grad_norm": 2.473483085632324,
+      "learning_rate": 3.1458646616541354e-06,
+      "loss": 0.1324,
+      "step": 45580
+    },
+    {
+      "epoch": 68.56,
+      "grad_norm": 5.73502779006958,
+      "learning_rate": 3.1443609022556394e-06,
+      "loss": 0.2821,
+      "step": 45590
+    },
+    {
+      "epoch": 68.57,
+      "grad_norm": 3.3458425998687744,
+      "learning_rate": 3.142857142857143e-06,
+      "loss": 0.1789,
+      "step": 45600
+    },
+    {
+      "epoch": 68.59,
+      "grad_norm": 5.300927639007568,
+      "learning_rate": 3.141353383458647e-06,
+      "loss": 0.2018,
+      "step": 45610
+    },
+    {
+      "epoch": 68.6,
+      "grad_norm": 6.269550323486328,
+      "learning_rate": 3.1398496240601508e-06,
+      "loss": 0.1991,
+      "step": 45620
+    },
+    {
+      "epoch": 68.62,
+      "grad_norm": 6.622176170349121,
+      "learning_rate": 3.1383458646616543e-06,
+      "loss": 0.1805,
+      "step": 45630
+    },
+    {
+      "epoch": 68.63,
+      "grad_norm": 7.759119987487793,
+      "learning_rate": 3.1368421052631582e-06,
+      "loss": 0.2455,
+      "step": 45640
+    },
+    {
+      "epoch": 68.65,
+      "grad_norm": 5.007221221923828,
+      "learning_rate": 3.1353383458646618e-06,
+      "loss": 0.1756,
+      "step": 45650
+    },
+    {
+      "epoch": 68.66,
+      "grad_norm": 5.451401710510254,
+      "learning_rate": 3.1338345864661657e-06,
+      "loss": 0.2058,
+      "step": 45660
+    },
+    {
+      "epoch": 68.68,
+      "grad_norm": 0.3662077784538269,
+      "learning_rate": 3.1323308270676696e-06,
+      "loss": 0.2024,
+      "step": 45670
+    },
+    {
+      "epoch": 68.69,
+      "grad_norm": 7.444847583770752,
+      "learning_rate": 3.130827067669173e-06,
+      "loss": 0.2119,
+      "step": 45680
+    },
+    {
+      "epoch": 68.71,
+      "grad_norm": 4.96779203414917,
+      "learning_rate": 3.129323308270677e-06,
+      "loss": 0.2708,
+      "step": 45690
+    },
+    {
+      "epoch": 68.72,
+      "grad_norm": 5.92564582824707,
+      "learning_rate": 3.1278195488721806e-06,
+      "loss": 0.1653,
+      "step": 45700
+    },
+    {
+      "epoch": 68.74,
+      "grad_norm": 5.052578449249268,
+      "learning_rate": 3.1263157894736846e-06,
+      "loss": 0.1841,
+      "step": 45710
+    },
+    {
+      "epoch": 68.75,
+      "grad_norm": 3.4545114040374756,
+      "learning_rate": 3.124812030075188e-06,
+      "loss": 0.2163,
+      "step": 45720
+    },
+    {
+      "epoch": 68.77,
+      "grad_norm": 5.480828762054443,
+      "learning_rate": 3.123308270676692e-06,
+      "loss": 0.1242,
+      "step": 45730
+    },
+    {
+      "epoch": 68.78,
+      "grad_norm": 4.088112831115723,
+      "learning_rate": 3.121804511278196e-06,
+      "loss": 0.1112,
+      "step": 45740
+    },
+    {
+      "epoch": 68.8,
+      "grad_norm": 4.465242862701416,
+      "learning_rate": 3.1203007518796995e-06,
+      "loss": 0.2095,
+      "step": 45750
+    },
+    {
+      "epoch": 68.81,
+      "grad_norm": 7.828632831573486,
+      "learning_rate": 3.1187969924812034e-06,
+      "loss": 0.246,
+      "step": 45760
+    },
+    {
+      "epoch": 68.83,
+      "grad_norm": 4.115504741668701,
+      "learning_rate": 3.117293233082707e-06,
+      "loss": 0.1527,
+      "step": 45770
+    },
+    {
+      "epoch": 68.84,
+      "grad_norm": 4.330153465270996,
+      "learning_rate": 3.115789473684211e-06,
+      "loss": 0.1638,
+      "step": 45780
+    },
+    {
+      "epoch": 68.86,
+      "grad_norm": 5.463741302490234,
+      "learning_rate": 3.114285714285715e-06,
+      "loss": 0.2238,
+      "step": 45790
+    },
+    {
+      "epoch": 68.87,
+      "grad_norm": 7.970693111419678,
+      "learning_rate": 3.1127819548872184e-06,
+      "loss": 0.1802,
+      "step": 45800
+    },
+    {
+      "epoch": 68.89,
+      "grad_norm": 4.3561577796936035,
+      "learning_rate": 3.1112781954887223e-06,
+      "loss": 0.2137,
+      "step": 45810
+    },
+    {
+      "epoch": 68.9,
+      "grad_norm": 5.0850958824157715,
+      "learning_rate": 3.109774436090226e-06,
+      "loss": 0.1676,
+      "step": 45820
+    },
+    {
+      "epoch": 68.92,
+      "grad_norm": 4.8231730461120605,
+      "learning_rate": 3.1082706766917298e-06,
+      "loss": 0.1825,
+      "step": 45830
+    },
+    {
+      "epoch": 68.93,
+      "grad_norm": 3.689068555831909,
+      "learning_rate": 3.1067669172932337e-06,
+      "loss": 0.1906,
+      "step": 45840
+    },
+    {
+      "epoch": 68.95,
+      "grad_norm": 3.177267074584961,
+      "learning_rate": 3.1052631578947372e-06,
+      "loss": 0.1653,
+      "step": 45850
+    },
+    {
+      "epoch": 68.96,
+      "grad_norm": 3.5612058639526367,
+      "learning_rate": 3.103759398496241e-06,
+      "loss": 0.1703,
+      "step": 45860
+    },
+    {
+      "epoch": 68.98,
+      "grad_norm": 7.400056838989258,
+      "learning_rate": 3.1022556390977447e-06,
+      "loss": 0.1732,
+      "step": 45870
+    },
+    {
+      "epoch": 68.99,
+      "grad_norm": 8.14755916595459,
+      "learning_rate": 3.1007518796992486e-06,
+      "loss": 0.223,
+      "step": 45880
+    },
+    {
+      "epoch": 69.0,
+      "eval_accuracy": 0.9293,
+      "eval_loss": 0.33156681060791016,
+      "eval_runtime": 84.519,
+      "eval_samples_per_second": 118.317,
+      "eval_steps_per_second": 0.473,
+      "step": 45885
+    },
+    {
+      "epoch": 69.01,
+      "grad_norm": 4.614025592803955,
+      "learning_rate": 3.0992481203007526e-06,
+      "loss": 0.1577,
+      "step": 45890
+    },
+    {
+      "epoch": 69.02,
+      "grad_norm": 5.442816257476807,
+      "learning_rate": 3.097744360902256e-06,
+      "loss": 0.2265,
+      "step": 45900
+    },
+    {
+      "epoch": 69.04,
+      "grad_norm": 5.297948837280273,
+      "learning_rate": 3.096240601503759e-06,
+      "loss": 0.199,
+      "step": 45910
+    },
+    {
+      "epoch": 69.05,
+      "grad_norm": 5.042107582092285,
+      "learning_rate": 3.094736842105263e-06,
+      "loss": 0.1949,
+      "step": 45920
+    },
+    {
+      "epoch": 69.07,
+      "grad_norm": 6.529865264892578,
+      "learning_rate": 3.093233082706767e-06,
+      "loss": 0.2015,
+      "step": 45930
+    },
+    {
+      "epoch": 69.08,
+      "grad_norm": 6.556023120880127,
+      "learning_rate": 3.0917293233082706e-06,
+      "loss": 0.2051,
+      "step": 45940
+    },
+    {
+      "epoch": 69.1,
+      "grad_norm": 5.027461051940918,
+      "learning_rate": 3.0902255639097745e-06,
+      "loss": 0.1894,
+      "step": 45950
+    },
+    {
+      "epoch": 69.11,
+      "grad_norm": 3.367723226547241,
+      "learning_rate": 3.088721804511278e-06,
+      "loss": 0.1466,
+      "step": 45960
+    },
+    {
+      "epoch": 69.13,
+      "grad_norm": 4.521204471588135,
+      "learning_rate": 3.087218045112782e-06,
+      "loss": 0.1885,
+      "step": 45970
+    },
+    {
+      "epoch": 69.14,
+      "grad_norm": 5.166673183441162,
+      "learning_rate": 3.085714285714286e-06,
+      "loss": 0.1593,
+      "step": 45980
+    },
+    {
+      "epoch": 69.16,
+      "grad_norm": 4.3414812088012695,
+      "learning_rate": 3.0842105263157895e-06,
+      "loss": 0.1804,
+      "step": 45990
+    },
+    {
+      "epoch": 69.17,
+      "grad_norm": 6.790599822998047,
+      "learning_rate": 3.0827067669172934e-06,
+      "loss": 0.177,
+      "step": 46000
+    },
+    {
+      "epoch": 69.19,
+      "grad_norm": 5.932426452636719,
+      "learning_rate": 3.081203007518797e-06,
+      "loss": 0.2503,
+      "step": 46010
+    },
+    {
+      "epoch": 69.2,
+      "grad_norm": 3.4542813301086426,
+      "learning_rate": 3.079699248120301e-06,
+      "loss": 0.1616,
+      "step": 46020
+    },
+    {
+      "epoch": 69.22,
+      "grad_norm": 3.6076695919036865,
+      "learning_rate": 3.078195488721805e-06,
+      "loss": 0.1794,
+      "step": 46030
+    },
+    {
+      "epoch": 69.23,
+      "grad_norm": 6.4292378425598145,
+      "learning_rate": 3.0766917293233083e-06,
+      "loss": 0.1487,
+      "step": 46040
+    },
+    {
+      "epoch": 69.25,
+      "grad_norm": 7.210880279541016,
+      "learning_rate": 3.0751879699248123e-06,
+      "loss": 0.1703,
+      "step": 46050
+    },
+    {
+      "epoch": 69.26,
+      "grad_norm": 2.0395233631134033,
+      "learning_rate": 3.0736842105263158e-06,
+      "loss": 0.1636,
+      "step": 46060
+    },
+    {
+      "epoch": 69.28,
+      "grad_norm": 9.782295227050781,
+      "learning_rate": 3.0721804511278197e-06,
+      "loss": 0.1979,
+      "step": 46070
+    },
+    {
+      "epoch": 69.29,
+      "grad_norm": 3.0649471282958984,
+      "learning_rate": 3.0706766917293237e-06,
+      "loss": 0.2242,
+      "step": 46080
+    },
+    {
+      "epoch": 69.31,
+      "grad_norm": 7.512526035308838,
+      "learning_rate": 3.069172932330827e-06,
+      "loss": 0.2796,
+      "step": 46090
+    },
+    {
+      "epoch": 69.32,
+      "grad_norm": 3.8779592514038086,
+      "learning_rate": 3.067669172932331e-06,
+      "loss": 0.1872,
+      "step": 46100
+    },
+    {
+      "epoch": 69.34,
+      "grad_norm": 4.834461212158203,
+      "learning_rate": 3.0661654135338346e-06,
+      "loss": 0.1784,
+      "step": 46110
+    },
+    {
+      "epoch": 69.35,
+      "grad_norm": 4.335732460021973,
+      "learning_rate": 3.0646616541353386e-06,
+      "loss": 0.1715,
+      "step": 46120
+    },
+    {
+      "epoch": 69.37,
+      "grad_norm": 8.79086971282959,
+      "learning_rate": 3.0631578947368425e-06,
+      "loss": 0.2193,
+      "step": 46130
+    },
+    {
+      "epoch": 69.38,
+      "grad_norm": 3.5167338848114014,
+      "learning_rate": 3.061654135338346e-06,
+      "loss": 0.1857,
+      "step": 46140
+    },
+    {
+      "epoch": 69.4,
+      "grad_norm": 3.6358115673065186,
+      "learning_rate": 3.06015037593985e-06,
+      "loss": 0.1648,
+      "step": 46150
+    },
+    {
+      "epoch": 69.41,
+      "grad_norm": 4.419709205627441,
+      "learning_rate": 3.0586466165413535e-06,
+      "loss": 0.1463,
+      "step": 46160
+    },
+    {
+      "epoch": 69.43,
+      "grad_norm": 12.194480895996094,
+      "learning_rate": 3.0571428571428575e-06,
+      "loss": 0.2125,
+      "step": 46170
+    },
+    {
+      "epoch": 69.44,
+      "grad_norm": 6.218225002288818,
+      "learning_rate": 3.055639097744361e-06,
+      "loss": 0.1879,
+      "step": 46180
+    },
+    {
+      "epoch": 69.46,
+      "grad_norm": 4.367229461669922,
+      "learning_rate": 3.054135338345865e-06,
+      "loss": 0.1527,
+      "step": 46190
+    },
+    {
+      "epoch": 69.47,
+      "grad_norm": 4.397371768951416,
+      "learning_rate": 3.052631578947369e-06,
+      "loss": 0.214,
+      "step": 46200
+    },
+    {
+      "epoch": 69.49,
+      "grad_norm": 9.216353416442871,
+      "learning_rate": 3.0511278195488724e-06,
+      "loss": 0.2037,
+      "step": 46210
+    },
+    {
+      "epoch": 69.5,
+      "grad_norm": 4.541748523712158,
+      "learning_rate": 3.0496240601503763e-06,
+      "loss": 0.1943,
+      "step": 46220
+    },
+    {
+      "epoch": 69.52,
+      "grad_norm": 2.6720430850982666,
+      "learning_rate": 3.04812030075188e-06,
+      "loss": 0.1699,
+      "step": 46230
+    },
+    {
+      "epoch": 69.53,
+      "grad_norm": 3.862180471420288,
+      "learning_rate": 3.0466165413533838e-06,
+      "loss": 0.1799,
+      "step": 46240
+    },
+    {
+      "epoch": 69.55,
+      "grad_norm": 3.2997727394104004,
+      "learning_rate": 3.0451127819548877e-06,
+      "loss": 0.1758,
+      "step": 46250
+    },
+    {
+      "epoch": 69.56,
+      "grad_norm": 6.742743015289307,
+      "learning_rate": 3.0436090225563912e-06,
+      "loss": 0.1675,
+      "step": 46260
+    },
+    {
+      "epoch": 69.58,
+      "grad_norm": 3.4689953327178955,
+      "learning_rate": 3.042105263157895e-06,
+      "loss": 0.1566,
+      "step": 46270
+    },
+    {
+      "epoch": 69.59,
+      "grad_norm": 4.367392063140869,
+      "learning_rate": 3.0406015037593987e-06,
+      "loss": 0.1839,
+      "step": 46280
+    },
+    {
+      "epoch": 69.61,
+      "grad_norm": 2.263700008392334,
+      "learning_rate": 3.0390977443609027e-06,
+      "loss": 0.1933,
+      "step": 46290
+    },
+    {
+      "epoch": 69.62,
+      "grad_norm": 7.104362487792969,
+      "learning_rate": 3.0375939849624066e-06,
+      "loss": 0.1841,
+      "step": 46300
+    },
+    {
+      "epoch": 69.64,
+      "grad_norm": 7.7885236740112305,
+      "learning_rate": 3.03609022556391e-06,
+      "loss": 0.2721,
+      "step": 46310
+    },
+    {
+      "epoch": 69.65,
+      "grad_norm": 5.928685665130615,
+      "learning_rate": 3.034586466165414e-06,
+      "loss": 0.2017,
+      "step": 46320
+    },
+    {
+      "epoch": 69.67,
+      "grad_norm": 3.4155795574188232,
+      "learning_rate": 3.0330827067669176e-06,
+      "loss": 0.1821,
+      "step": 46330
+    },
+    {
+      "epoch": 69.68,
+      "grad_norm": 4.222193717956543,
+      "learning_rate": 3.0315789473684215e-06,
+      "loss": 0.2059,
+      "step": 46340
+    },
+    {
+      "epoch": 69.7,
+      "grad_norm": 6.882046222686768,
+      "learning_rate": 3.0300751879699255e-06,
+      "loss": 0.1614,
+      "step": 46350
+    },
+    {
+      "epoch": 69.71,
+      "grad_norm": 4.649013996124268,
+      "learning_rate": 3.028571428571429e-06,
+      "loss": 0.2039,
+      "step": 46360
+    },
+    {
+      "epoch": 69.73,
+      "grad_norm": 3.898035764694214,
+      "learning_rate": 3.027067669172933e-06,
+      "loss": 0.1677,
+      "step": 46370
+    },
+    {
+      "epoch": 69.74,
+      "grad_norm": 4.432783603668213,
+      "learning_rate": 3.0255639097744364e-06,
+      "loss": 0.1762,
+      "step": 46380
+    },
+    {
+      "epoch": 69.76,
+      "grad_norm": 8.104802131652832,
+      "learning_rate": 3.0240601503759404e-06,
+      "loss": 0.1619,
+      "step": 46390
+    },
+    {
+      "epoch": 69.77,
+      "grad_norm": 4.149446487426758,
+      "learning_rate": 3.0225563909774443e-06,
+      "loss": 0.1904,
+      "step": 46400
+    },
+    {
+      "epoch": 69.79,
+      "grad_norm": 6.308215141296387,
+      "learning_rate": 3.0210526315789474e-06,
+      "loss": 0.1996,
+      "step": 46410
+    },
+    {
+      "epoch": 69.8,
+      "grad_norm": 8.179604530334473,
+      "learning_rate": 3.019548872180451e-06,
+      "loss": 0.2209,
+      "step": 46420
+    },
+    {
+      "epoch": 69.82,
+      "grad_norm": 5.956761837005615,
+      "learning_rate": 3.018045112781955e-06,
+      "loss": 0.1432,
+      "step": 46430
+    },
+    {
+      "epoch": 69.83,
+      "grad_norm": 5.909665584564209,
+      "learning_rate": 3.016541353383459e-06,
+      "loss": 0.2528,
+      "step": 46440
+    },
+    {
+      "epoch": 69.85,
+      "grad_norm": 6.1077399253845215,
+      "learning_rate": 3.0150375939849623e-06,
+      "loss": 0.2055,
+      "step": 46450
+    },
+    {
+      "epoch": 69.86,
+      "grad_norm": 4.353606700897217,
+      "learning_rate": 3.0135338345864663e-06,
+      "loss": 0.2351,
+      "step": 46460
+    },
+    {
+      "epoch": 69.88,
+      "grad_norm": 5.350019931793213,
+      "learning_rate": 3.01203007518797e-06,
+      "loss": 0.1732,
+      "step": 46470
+    },
+    {
+      "epoch": 69.89,
+      "grad_norm": 3.917721748352051,
+      "learning_rate": 3.0105263157894737e-06,
+      "loss": 0.1791,
+      "step": 46480
+    },
+    {
+      "epoch": 69.91,
+      "grad_norm": 10.580587387084961,
+      "learning_rate": 3.0090225563909777e-06,
+      "loss": 0.1899,
+      "step": 46490
+    },
+    {
+      "epoch": 69.92,
+      "grad_norm": 4.388562202453613,
+      "learning_rate": 3.007518796992481e-06,
+      "loss": 0.2287,
+      "step": 46500
+    },
+    {
+      "epoch": 69.94,
+      "grad_norm": 5.693699359893799,
+      "learning_rate": 3.006015037593985e-06,
+      "loss": 0.2259,
+      "step": 46510
+    },
+    {
+      "epoch": 69.95,
+      "grad_norm": 3.3817298412323,
+      "learning_rate": 3.0045112781954887e-06,
+      "loss": 0.2412,
+      "step": 46520
+    },
+    {
+      "epoch": 69.97,
+      "grad_norm": 1.847433090209961,
+      "learning_rate": 3.0030075187969926e-06,
+      "loss": 0.1723,
+      "step": 46530
+    },
+    {
+      "epoch": 69.98,
+      "grad_norm": 6.4717254638671875,
+      "learning_rate": 3.0015037593984966e-06,
+      "loss": 0.2032,
+      "step": 46540
+    },
+    {
+      "epoch": 70.0,
+      "grad_norm": 0.9074813723564148,
+      "learning_rate": 3e-06,
+      "loss": 0.1738,
+      "step": 46550
+    },
+    {
+      "epoch": 70.0,
+      "eval_accuracy": 0.9295,
+      "eval_loss": 0.3248044550418854,
+      "eval_runtime": 84.7278,
+      "eval_samples_per_second": 118.025,
+      "eval_steps_per_second": 0.472,
+      "step": 46550
+    },
+    {
+      "epoch": 70.02,
+      "grad_norm": 5.463174819946289,
+      "learning_rate": 2.998496240601504e-06,
+      "loss": 0.2036,
+      "step": 46560
+    },
+    {
+      "epoch": 70.03,
+      "grad_norm": 2.8487234115600586,
+      "learning_rate": 2.9969924812030075e-06,
+      "loss": 0.1619,
+      "step": 46570
+    },
+    {
+      "epoch": 70.05,
+      "grad_norm": 4.683668613433838,
+      "learning_rate": 2.9954887218045115e-06,
+      "loss": 0.2471,
+      "step": 46580
+    },
+    {
+      "epoch": 70.06,
+      "grad_norm": 5.568604946136475,
+      "learning_rate": 2.9939849624060154e-06,
+      "loss": 0.2199,
+      "step": 46590
+    },
+    {
+      "epoch": 70.08,
+      "grad_norm": 8.11655330657959,
+      "learning_rate": 2.992481203007519e-06,
+      "loss": 0.1753,
+      "step": 46600
+    },
+    {
+      "epoch": 70.09,
+      "grad_norm": 7.468130588531494,
+      "learning_rate": 2.990977443609023e-06,
+      "loss": 0.1809,
+      "step": 46610
+    },
+    {
+      "epoch": 70.11,
+      "grad_norm": 6.2125115394592285,
+      "learning_rate": 2.9894736842105264e-06,
+      "loss": 0.2751,
+      "step": 46620
+    },
+    {
+      "epoch": 70.12,
+      "grad_norm": 3.6229348182678223,
+      "learning_rate": 2.9879699248120303e-06,
+      "loss": 0.1723,
+      "step": 46630
+    },
+    {
+      "epoch": 70.14,
+      "grad_norm": 4.364720344543457,
+      "learning_rate": 2.986466165413534e-06,
+      "loss": 0.1679,
+      "step": 46640
+    },
+    {
+      "epoch": 70.15,
+      "grad_norm": 6.14267110824585,
+      "learning_rate": 2.984962406015038e-06,
+      "loss": 0.2083,
+      "step": 46650
+    },
+    {
+      "epoch": 70.17,
+      "grad_norm": 3.9673428535461426,
+      "learning_rate": 2.9834586466165418e-06,
+      "loss": 0.1551,
+      "step": 46660
+    },
+    {
+      "epoch": 70.18,
+      "grad_norm": 4.621110916137695,
+      "learning_rate": 2.9819548872180453e-06,
+      "loss": 0.1569,
+      "step": 46670
+    },
+    {
+      "epoch": 70.2,
+      "grad_norm": 8.371132850646973,
+      "learning_rate": 2.9804511278195492e-06,
+      "loss": 0.1963,
+      "step": 46680
+    },
+    {
+      "epoch": 70.21,
+      "grad_norm": 5.156669616699219,
+      "learning_rate": 2.9789473684210527e-06,
+      "loss": 0.2096,
+      "step": 46690
+    },
+    {
+      "epoch": 70.23,
+      "grad_norm": 4.981484413146973,
+      "learning_rate": 2.9774436090225567e-06,
+      "loss": 0.1313,
+      "step": 46700
+    },
+    {
+      "epoch": 70.24,
+      "grad_norm": 5.359274387359619,
+      "learning_rate": 2.9759398496240606e-06,
+      "loss": 0.183,
+      "step": 46710
+    },
+    {
+      "epoch": 70.26,
+      "grad_norm": 6.65374231338501,
+      "learning_rate": 2.974436090225564e-06,
+      "loss": 0.2083,
+      "step": 46720
+    },
+    {
+      "epoch": 70.27,
+      "grad_norm": 7.133225917816162,
+      "learning_rate": 2.972932330827068e-06,
+      "loss": 0.2509,
+      "step": 46730
+    },
+    {
+      "epoch": 70.29,
+      "grad_norm": 4.245519638061523,
+      "learning_rate": 2.9714285714285716e-06,
+      "loss": 0.1871,
+      "step": 46740
+    },
+    {
+      "epoch": 70.3,
+      "grad_norm": 5.218483924865723,
+      "learning_rate": 2.9699248120300755e-06,
+      "loss": 0.1595,
+      "step": 46750
+    },
+    {
+      "epoch": 70.32,
+      "grad_norm": 6.059344291687012,
+      "learning_rate": 2.9684210526315795e-06,
+      "loss": 0.1771,
+      "step": 46760
+    },
+    {
+      "epoch": 70.33,
+      "grad_norm": 4.6386237144470215,
+      "learning_rate": 2.966917293233083e-06,
+      "loss": 0.1868,
+      "step": 46770
+    },
+    {
+      "epoch": 70.35,
+      "grad_norm": 4.3146257400512695,
+      "learning_rate": 2.965413533834587e-06,
+      "loss": 0.1941,
+      "step": 46780
+    },
+    {
+      "epoch": 70.36,
+      "grad_norm": 4.2247395515441895,
+      "learning_rate": 2.9639097744360905e-06,
+      "loss": 0.2796,
+      "step": 46790
+    },
+    {
+      "epoch": 70.38,
+      "grad_norm": 4.537256717681885,
+      "learning_rate": 2.9624060150375944e-06,
+      "loss": 0.1675,
+      "step": 46800
+    },
+    {
+      "epoch": 70.39,
+      "grad_norm": 10.793876647949219,
+      "learning_rate": 2.9609022556390983e-06,
+      "loss": 0.1738,
+      "step": 46810
+    },
+    {
+      "epoch": 70.41,
+      "grad_norm": 3.7752058506011963,
+      "learning_rate": 2.959398496240602e-06,
+      "loss": 0.1864,
+      "step": 46820
+    },
+    {
+      "epoch": 70.42,
+      "grad_norm": 5.952646732330322,
+      "learning_rate": 2.957894736842106e-06,
+      "loss": 0.1994,
+      "step": 46830
+    },
+    {
+      "epoch": 70.44,
+      "grad_norm": 2.576373815536499,
+      "learning_rate": 2.9563909774436093e-06,
+      "loss": 0.1891,
+      "step": 46840
+    },
+    {
+      "epoch": 70.45,
+      "grad_norm": 3.7196900844573975,
+      "learning_rate": 2.9548872180451133e-06,
+      "loss": 0.2391,
+      "step": 46850
+    },
+    {
+      "epoch": 70.47,
+      "grad_norm": 4.4083571434021,
+      "learning_rate": 2.9533834586466172e-06,
+      "loss": 0.2241,
+      "step": 46860
+    },
+    {
+      "epoch": 70.48,
+      "grad_norm": 4.769841194152832,
+      "learning_rate": 2.9518796992481207e-06,
+      "loss": 0.1591,
+      "step": 46870
+    },
+    {
+      "epoch": 70.5,
+      "grad_norm": 4.998278617858887,
+      "learning_rate": 2.9503759398496247e-06,
+      "loss": 0.1785,
+      "step": 46880
+    },
+    {
+      "epoch": 70.51,
+      "grad_norm": 4.37769889831543,
+      "learning_rate": 2.948872180451128e-06,
+      "loss": 0.16,
+      "step": 46890
+    },
+    {
+      "epoch": 70.53,
+      "grad_norm": 7.736121654510498,
+      "learning_rate": 2.9473684210526317e-06,
+      "loss": 0.2141,
+      "step": 46900
+    },
+    {
+      "epoch": 70.54,
+      "grad_norm": 8.18359375,
+      "learning_rate": 2.9458646616541352e-06,
+      "loss": 0.2537,
+      "step": 46910
+    },
+    {
+      "epoch": 70.56,
+      "grad_norm": 7.136547088623047,
+      "learning_rate": 2.944360902255639e-06,
+      "loss": 0.1986,
+      "step": 46920
+    },
+    {
+      "epoch": 70.57,
+      "grad_norm": 2.9580116271972656,
+      "learning_rate": 2.9428571428571427e-06,
+      "loss": 0.1481,
+      "step": 46930
+    },
+    {
+      "epoch": 70.59,
+      "grad_norm": 5.023125171661377,
+      "learning_rate": 2.9413533834586466e-06,
+      "loss": 0.2149,
+      "step": 46940
+    },
+    {
+      "epoch": 70.6,
+      "grad_norm": 5.102907657623291,
+      "learning_rate": 2.9398496240601506e-06,
+      "loss": 0.1215,
+      "step": 46950
+    },
+    {
+      "epoch": 70.62,
+      "grad_norm": 3.1318459510803223,
+      "learning_rate": 2.938345864661654e-06,
+      "loss": 0.1812,
+      "step": 46960
+    },
+    {
+      "epoch": 70.63,
+      "grad_norm": 3.511244058609009,
+      "learning_rate": 2.936842105263158e-06,
+      "loss": 0.1735,
+      "step": 46970
+    },
+    {
+      "epoch": 70.65,
+      "grad_norm": 8.335837364196777,
+      "learning_rate": 2.9353383458646616e-06,
+      "loss": 0.1951,
+      "step": 46980
+    },
+    {
+      "epoch": 70.66,
+      "grad_norm": 3.936018228530884,
+      "learning_rate": 2.9338345864661655e-06,
+      "loss": 0.2294,
+      "step": 46990
+    },
+    {
+      "epoch": 70.68,
+      "grad_norm": 6.189547538757324,
+      "learning_rate": 2.9323308270676694e-06,
+      "loss": 0.2205,
+      "step": 47000
+    },
+    {
+      "epoch": 70.69,
+      "grad_norm": 3.3733437061309814,
+      "learning_rate": 2.930827067669173e-06,
+      "loss": 0.1982,
+      "step": 47010
+    },
+    {
+      "epoch": 70.71,
+      "grad_norm": 3.5287833213806152,
+      "learning_rate": 2.929323308270677e-06,
+      "loss": 0.1972,
+      "step": 47020
+    },
+    {
+      "epoch": 70.72,
+      "grad_norm": 21.65901756286621,
+      "learning_rate": 2.9278195488721804e-06,
+      "loss": 0.2244,
+      "step": 47030
+    },
+    {
+      "epoch": 70.74,
+      "grad_norm": 7.8496904373168945,
+      "learning_rate": 2.9263157894736844e-06,
+      "loss": 0.114,
+      "step": 47040
+    },
+    {
+      "epoch": 70.75,
+      "grad_norm": 3.459660053253174,
+      "learning_rate": 2.9248120300751883e-06,
+      "loss": 0.2043,
+      "step": 47050
+    },
+    {
+      "epoch": 70.77,
+      "grad_norm": 6.128682613372803,
+      "learning_rate": 2.923308270676692e-06,
+      "loss": 0.1774,
+      "step": 47060
+    },
+    {
+      "epoch": 70.78,
+      "grad_norm": 6.694797515869141,
+      "learning_rate": 2.9218045112781958e-06,
+      "loss": 0.1901,
+      "step": 47070
+    },
+    {
+      "epoch": 70.8,
+      "grad_norm": 7.470389366149902,
+      "learning_rate": 2.9203007518796993e-06,
+      "loss": 0.1957,
+      "step": 47080
+    },
+    {
+      "epoch": 70.81,
+      "grad_norm": 5.193848133087158,
+      "learning_rate": 2.9187969924812032e-06,
+      "loss": 0.1428,
+      "step": 47090
+    },
+    {
+      "epoch": 70.83,
+      "grad_norm": 5.4394049644470215,
+      "learning_rate": 2.9172932330827068e-06,
+      "loss": 0.1873,
+      "step": 47100
+    },
+    {
+      "epoch": 70.84,
+      "grad_norm": 5.622503280639648,
+      "learning_rate": 2.9157894736842107e-06,
+      "loss": 0.2147,
+      "step": 47110
+    },
+    {
+      "epoch": 70.86,
+      "grad_norm": 9.180618286132812,
+      "learning_rate": 2.9142857142857146e-06,
+      "loss": 0.1662,
+      "step": 47120
+    },
+    {
+      "epoch": 70.87,
+      "grad_norm": 3.7462944984436035,
+      "learning_rate": 2.912781954887218e-06,
+      "loss": 0.1925,
+      "step": 47130
+    },
+    {
+      "epoch": 70.89,
+      "grad_norm": 6.142997741699219,
+      "learning_rate": 2.911278195488722e-06,
+      "loss": 0.1398,
+      "step": 47140
+    },
+    {
+      "epoch": 70.9,
+      "grad_norm": 7.028275966644287,
+      "learning_rate": 2.9097744360902256e-06,
+      "loss": 0.2052,
+      "step": 47150
+    },
+    {
+      "epoch": 70.92,
+      "grad_norm": 4.399960517883301,
+      "learning_rate": 2.9082706766917296e-06,
+      "loss": 0.2089,
+      "step": 47160
+    },
+    {
+      "epoch": 70.93,
+      "grad_norm": 2.562202215194702,
+      "learning_rate": 2.9067669172932335e-06,
+      "loss": 0.1564,
+      "step": 47170
+    },
+    {
+      "epoch": 70.95,
+      "grad_norm": 9.725334167480469,
+      "learning_rate": 2.905263157894737e-06,
+      "loss": 0.1702,
+      "step": 47180
+    },
+    {
+      "epoch": 70.96,
+      "grad_norm": 4.6251654624938965,
+      "learning_rate": 2.903759398496241e-06,
+      "loss": 0.247,
+      "step": 47190
+    },
+    {
+      "epoch": 70.98,
+      "grad_norm": 5.6048760414123535,
+      "learning_rate": 2.9022556390977445e-06,
+      "loss": 0.2587,
+      "step": 47200
+    },
+    {
+      "epoch": 70.99,
+      "grad_norm": 2.857663631439209,
+      "learning_rate": 2.9007518796992484e-06,
+      "loss": 0.2251,
+      "step": 47210
+    },
+    {
+      "epoch": 71.0,
+      "eval_accuracy": 0.9305,
+      "eval_loss": 0.32967016100883484,
+      "eval_runtime": 84.8297,
+      "eval_samples_per_second": 117.883,
+      "eval_steps_per_second": 0.472,
+      "step": 47215
+    },
+    {
+      "epoch": 71.01,
+      "grad_norm": 3.354151487350464,
+      "learning_rate": 2.8992481203007524e-06,
+      "loss": 0.1378,
+      "step": 47220
+    },
+    {
+      "epoch": 71.02,
+      "grad_norm": 14.883716583251953,
+      "learning_rate": 2.897744360902256e-06,
+      "loss": 0.1888,
+      "step": 47230
+    },
+    {
+      "epoch": 71.04,
+      "grad_norm": 3.659508228302002,
+      "learning_rate": 2.89624060150376e-06,
+      "loss": 0.1611,
+      "step": 47240
+    },
+    {
+      "epoch": 71.05,
+      "grad_norm": 4.597053527832031,
+      "learning_rate": 2.8947368421052634e-06,
+      "loss": 0.1918,
+      "step": 47250
+    },
+    {
+      "epoch": 71.07,
+      "grad_norm": 4.022745132446289,
+      "learning_rate": 2.8932330827067673e-06,
+      "loss": 0.2231,
+      "step": 47260
+    },
+    {
+      "epoch": 71.08,
+      "grad_norm": 3.9845337867736816,
+      "learning_rate": 2.8917293233082712e-06,
+      "loss": 0.2327,
+      "step": 47270
+    },
+    {
+      "epoch": 71.1,
+      "grad_norm": 6.1547698974609375,
+      "learning_rate": 2.8902255639097748e-06,
+      "loss": 0.2559,
+      "step": 47280
+    },
+    {
+      "epoch": 71.11,
+      "grad_norm": 10.55440616607666,
+      "learning_rate": 2.8887218045112787e-06,
+      "loss": 0.166,
+      "step": 47290
+    },
+    {
+      "epoch": 71.13,
+      "grad_norm": 4.419111251831055,
+      "learning_rate": 2.8872180451127822e-06,
+      "loss": 0.1644,
+      "step": 47300
+    },
+    {
+      "epoch": 71.14,
+      "grad_norm": 2.8730411529541016,
+      "learning_rate": 2.885714285714286e-06,
+      "loss": 0.1481,
+      "step": 47310
+    },
+    {
+      "epoch": 71.16,
+      "grad_norm": 4.056236267089844,
+      "learning_rate": 2.88421052631579e-06,
+      "loss": 0.2006,
+      "step": 47320
+    },
+    {
+      "epoch": 71.17,
+      "grad_norm": 5.246281623840332,
+      "learning_rate": 2.8827067669172936e-06,
+      "loss": 0.2015,
+      "step": 47330
+    },
+    {
+      "epoch": 71.19,
+      "grad_norm": 4.46673059463501,
+      "learning_rate": 2.8812030075187976e-06,
+      "loss": 0.1138,
+      "step": 47340
+    },
+    {
+      "epoch": 71.2,
+      "grad_norm": 9.421850204467773,
+      "learning_rate": 2.879699248120301e-06,
+      "loss": 0.2566,
+      "step": 47350
+    },
+    {
+      "epoch": 71.22,
+      "grad_norm": 5.661787033081055,
+      "learning_rate": 2.878195488721805e-06,
+      "loss": 0.1804,
+      "step": 47360
+    },
+    {
+      "epoch": 71.23,
+      "grad_norm": 4.693019390106201,
+      "learning_rate": 2.876691729323309e-06,
+      "loss": 0.1765,
+      "step": 47370
+    },
+    {
+      "epoch": 71.25,
+      "grad_norm": 6.069882392883301,
+      "learning_rate": 2.8751879699248125e-06,
+      "loss": 0.173,
+      "step": 47380
+    },
+    {
+      "epoch": 71.26,
+      "grad_norm": 8.419574737548828,
+      "learning_rate": 2.8736842105263164e-06,
+      "loss": 0.1686,
+      "step": 47390
+    },
+    {
+      "epoch": 71.28,
+      "grad_norm": 2.8149406909942627,
+      "learning_rate": 2.8721804511278195e-06,
+      "loss": 0.2335,
+      "step": 47400
+    },
+    {
+      "epoch": 71.29,
+      "grad_norm": 5.66928243637085,
+      "learning_rate": 2.8706766917293235e-06,
+      "loss": 0.2031,
+      "step": 47410
+    },
+    {
+      "epoch": 71.31,
+      "grad_norm": 3.403899908065796,
+      "learning_rate": 2.869172932330827e-06,
+      "loss": 0.2124,
+      "step": 47420
+    },
+    {
+      "epoch": 71.32,
+      "grad_norm": 5.2177910804748535,
+      "learning_rate": 2.867669172932331e-06,
+      "loss": 0.1582,
+      "step": 47430
+    },
+    {
+      "epoch": 71.34,
+      "grad_norm": 5.742595195770264,
+      "learning_rate": 2.8661654135338344e-06,
+      "loss": 0.186,
+      "step": 47440
+    },
+    {
+      "epoch": 71.35,
+      "grad_norm": 4.816529273986816,
+      "learning_rate": 2.8646616541353384e-06,
+      "loss": 0.1572,
+      "step": 47450
+    },
+    {
+      "epoch": 71.37,
+      "grad_norm": 4.075284004211426,
+      "learning_rate": 2.8631578947368423e-06,
+      "loss": 0.1586,
+      "step": 47460
+    },
+    {
+      "epoch": 71.38,
+      "grad_norm": 5.9553446769714355,
+      "learning_rate": 2.861654135338346e-06,
+      "loss": 0.1871,
+      "step": 47470
+    },
+    {
+      "epoch": 71.4,
+      "grad_norm": 5.976673126220703,
+      "learning_rate": 2.86015037593985e-06,
+      "loss": 0.1897,
+      "step": 47480
+    },
+    {
+      "epoch": 71.41,
+      "grad_norm": 3.2416350841522217,
+      "learning_rate": 2.8586466165413533e-06,
+      "loss": 0.1402,
+      "step": 47490
+    },
+    {
+      "epoch": 71.43,
+      "grad_norm": 3.216466188430786,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 0.1308,
+      "step": 47500
+    },
+    {
+      "epoch": 71.44,
+      "grad_norm": 3.74507737159729,
+      "learning_rate": 2.855639097744361e-06,
+      "loss": 0.2048,
+      "step": 47510
+    },
+    {
+      "epoch": 71.46,
+      "grad_norm": 2.735323905944824,
+      "learning_rate": 2.8541353383458647e-06,
+      "loss": 0.2007,
+      "step": 47520
+    },
+    {
+      "epoch": 71.47,
+      "grad_norm": 4.694336414337158,
+      "learning_rate": 2.8526315789473687e-06,
+      "loss": 0.1947,
+      "step": 47530
+    },
+    {
+      "epoch": 71.49,
+      "grad_norm": 4.027325630187988,
+      "learning_rate": 2.851127819548872e-06,
+      "loss": 0.1442,
+      "step": 47540
+    },
+    {
+      "epoch": 71.5,
+      "grad_norm": 4.482545852661133,
+      "learning_rate": 2.849624060150376e-06,
+      "loss": 0.1949,
+      "step": 47550
+    },
+    {
+      "epoch": 71.52,
+      "grad_norm": 21.274612426757812,
+      "learning_rate": 2.8481203007518796e-06,
+      "loss": 0.2099,
+      "step": 47560
+    },
+    {
+      "epoch": 71.53,
+      "grad_norm": 7.941802978515625,
+      "learning_rate": 2.8466165413533836e-06,
+      "loss": 0.1769,
+      "step": 47570
+    },
+    {
+      "epoch": 71.55,
+      "grad_norm": 11.027046203613281,
+      "learning_rate": 2.8451127819548875e-06,
+      "loss": 0.2497,
+      "step": 47580
+    },
+    {
+      "epoch": 71.56,
+      "grad_norm": 1.5711476802825928,
+      "learning_rate": 2.843609022556391e-06,
+      "loss": 0.1957,
+      "step": 47590
+    },
+    {
+      "epoch": 71.58,
+      "grad_norm": 8.369176864624023,
+      "learning_rate": 2.842105263157895e-06,
+      "loss": 0.2131,
+      "step": 47600
+    },
+    {
+      "epoch": 71.59,
+      "grad_norm": 4.709144592285156,
+      "learning_rate": 2.8406015037593985e-06,
+      "loss": 0.2316,
+      "step": 47610
+    },
+    {
+      "epoch": 71.61,
+      "grad_norm": 4.679172992706299,
+      "learning_rate": 2.8390977443609024e-06,
+      "loss": 0.1916,
+      "step": 47620
+    },
+    {
+      "epoch": 71.62,
+      "grad_norm": 7.5052924156188965,
+      "learning_rate": 2.8375939849624064e-06,
+      "loss": 0.1613,
+      "step": 47630
+    },
+    {
+      "epoch": 71.64,
+      "grad_norm": 6.649926662445068,
+      "learning_rate": 2.83609022556391e-06,
+      "loss": 0.1848,
+      "step": 47640
+    },
+    {
+      "epoch": 71.65,
+      "grad_norm": 5.853745937347412,
+      "learning_rate": 2.834586466165414e-06,
+      "loss": 0.1762,
+      "step": 47650
+    },
+    {
+      "epoch": 71.67,
+      "grad_norm": 6.767156600952148,
+      "learning_rate": 2.8330827067669174e-06,
+      "loss": 0.1358,
+      "step": 47660
+    },
+    {
+      "epoch": 71.68,
+      "grad_norm": 2.320805311203003,
+      "learning_rate": 2.8315789473684213e-06,
+      "loss": 0.1576,
+      "step": 47670
+    },
+    {
+      "epoch": 71.7,
+      "grad_norm": 6.782162666320801,
+      "learning_rate": 2.8300751879699253e-06,
+      "loss": 0.2299,
+      "step": 47680
+    },
+    {
+      "epoch": 71.71,
+      "grad_norm": 5.48624324798584,
+      "learning_rate": 2.8285714285714288e-06,
+      "loss": 0.1981,
+      "step": 47690
+    },
+    {
+      "epoch": 71.73,
+      "grad_norm": 4.4011335372924805,
+      "learning_rate": 2.8270676691729327e-06,
+      "loss": 0.1662,
+      "step": 47700
+    },
+    {
+      "epoch": 71.74,
+      "grad_norm": 2.359865188598633,
+      "learning_rate": 2.8255639097744362e-06,
+      "loss": 0.1419,
+      "step": 47710
+    },
+    {
+      "epoch": 71.76,
+      "grad_norm": 6.685113430023193,
+      "learning_rate": 2.82406015037594e-06,
+      "loss": 0.2208,
+      "step": 47720
+    },
+    {
+      "epoch": 71.77,
+      "grad_norm": 4.598508358001709,
+      "learning_rate": 2.822556390977444e-06,
+      "loss": 0.1872,
+      "step": 47730
+    },
+    {
+      "epoch": 71.79,
+      "grad_norm": 4.389933109283447,
+      "learning_rate": 2.8210526315789476e-06,
+      "loss": 0.196,
+      "step": 47740
+    },
+    {
+      "epoch": 71.8,
+      "grad_norm": 3.9596188068389893,
+      "learning_rate": 2.8195488721804516e-06,
+      "loss": 0.121,
+      "step": 47750
+    },
+    {
+      "epoch": 71.82,
+      "grad_norm": 1.1457980871200562,
+      "learning_rate": 2.818045112781955e-06,
+      "loss": 0.2009,
+      "step": 47760
+    },
+    {
+      "epoch": 71.83,
+      "grad_norm": 6.121349811553955,
+      "learning_rate": 2.816541353383459e-06,
+      "loss": 0.1835,
+      "step": 47770
+    },
+    {
+      "epoch": 71.85,
+      "grad_norm": 4.898597717285156,
+      "learning_rate": 2.815037593984963e-06,
+      "loss": 0.1345,
+      "step": 47780
+    },
+    {
+      "epoch": 71.86,
+      "grad_norm": 10.096803665161133,
+      "learning_rate": 2.8135338345864665e-06,
+      "loss": 0.1851,
+      "step": 47790
+    },
+    {
+      "epoch": 71.88,
+      "grad_norm": 5.386133193969727,
+      "learning_rate": 2.8120300751879705e-06,
+      "loss": 0.1939,
+      "step": 47800
+    },
+    {
+      "epoch": 71.89,
+      "grad_norm": 4.9606428146362305,
+      "learning_rate": 2.810526315789474e-06,
+      "loss": 0.1756,
+      "step": 47810
+    },
+    {
+      "epoch": 71.91,
+      "grad_norm": 3.927384853363037,
+      "learning_rate": 2.809022556390978e-06,
+      "loss": 0.1958,
+      "step": 47820
+    },
+    {
+      "epoch": 71.92,
+      "grad_norm": 1.8455919027328491,
+      "learning_rate": 2.807518796992482e-06,
+      "loss": 0.1686,
+      "step": 47830
+    },
+    {
+      "epoch": 71.94,
+      "grad_norm": 4.488922119140625,
+      "learning_rate": 2.8060150375939854e-06,
+      "loss": 0.1532,
+      "step": 47840
+    },
+    {
+      "epoch": 71.95,
+      "grad_norm": 5.330088138580322,
+      "learning_rate": 2.8045112781954893e-06,
+      "loss": 0.1574,
+      "step": 47850
+    },
+    {
+      "epoch": 71.97,
+      "grad_norm": 5.4682207107543945,
+      "learning_rate": 2.803007518796993e-06,
+      "loss": 0.1734,
+      "step": 47860
+    },
+    {
+      "epoch": 71.98,
+      "grad_norm": 6.902002811431885,
+      "learning_rate": 2.8015037593984968e-06,
+      "loss": 0.1673,
+      "step": 47870
+    },
+    {
+      "epoch": 72.0,
+      "grad_norm": 0.017349636182188988,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.1518,
+      "step": 47880
+    },
+    {
+      "epoch": 72.0,
+      "eval_accuracy": 0.9311,
+      "eval_loss": 0.3322134017944336,
+      "eval_runtime": 84.7025,
+      "eval_samples_per_second": 118.06,
+      "eval_steps_per_second": 0.472,
+      "step": 47880
+    },
+    {
+      "epoch": 72.02,
+      "grad_norm": 2.0250072479248047,
+      "learning_rate": 2.798496240601504e-06,
+      "loss": 0.2221,
+      "step": 47890
+    },
+    {
+      "epoch": 72.03,
+      "grad_norm": 3.505276679992676,
+      "learning_rate": 2.7969924812030073e-06,
+      "loss": 0.1402,
+      "step": 47900
+    },
+    {
+      "epoch": 72.05,
+      "grad_norm": 4.723272323608398,
+      "learning_rate": 2.7954887218045113e-06,
+      "loss": 0.1832,
+      "step": 47910
+    },
+    {
+      "epoch": 72.06,
+      "grad_norm": 11.436491012573242,
+      "learning_rate": 2.7939849624060152e-06,
+      "loss": 0.1595,
+      "step": 47920
+    },
+    {
+      "epoch": 72.08,
+      "grad_norm": 8.044598579406738,
+      "learning_rate": 2.7924812030075187e-06,
+      "loss": 0.2156,
+      "step": 47930
+    },
+    {
+      "epoch": 72.09,
+      "grad_norm": 5.603691577911377,
+      "learning_rate": 2.7909774436090227e-06,
+      "loss": 0.1677,
+      "step": 47940
+    },
+    {
+      "epoch": 72.11,
+      "grad_norm": 3.3360657691955566,
+      "learning_rate": 2.789473684210526e-06,
+      "loss": 0.1723,
+      "step": 47950
+    },
+    {
+      "epoch": 72.12,
+      "grad_norm": 5.758209228515625,
+      "learning_rate": 2.78796992481203e-06,
+      "loss": 0.1905,
+      "step": 47960
+    },
+    {
+      "epoch": 72.14,
+      "grad_norm": 6.080915451049805,
+      "learning_rate": 2.786466165413534e-06,
+      "loss": 0.2027,
+      "step": 47970
+    },
+    {
+      "epoch": 72.15,
+      "grad_norm": 4.1375017166137695,
+      "learning_rate": 2.7849624060150376e-06,
+      "loss": 0.1241,
+      "step": 47980
+    },
+    {
+      "epoch": 72.17,
+      "grad_norm": 4.435542106628418,
+      "learning_rate": 2.7834586466165415e-06,
+      "loss": 0.2314,
+      "step": 47990
+    },
+    {
+      "epoch": 72.18,
+      "grad_norm": 3.8110594749450684,
+      "learning_rate": 2.781954887218045e-06,
+      "loss": 0.1544,
+      "step": 48000
+    },
+    {
+      "epoch": 72.2,
+      "grad_norm": 3.3765087127685547,
+      "learning_rate": 2.780451127819549e-06,
+      "loss": 0.2048,
+      "step": 48010
+    },
+    {
+      "epoch": 72.21,
+      "grad_norm": 6.244101047515869,
+      "learning_rate": 2.7789473684210525e-06,
+      "loss": 0.1541,
+      "step": 48020
+    },
+    {
+      "epoch": 72.23,
+      "grad_norm": 8.180159568786621,
+      "learning_rate": 2.7774436090225565e-06,
+      "loss": 0.1783,
+      "step": 48030
+    },
+    {
+      "epoch": 72.24,
+      "grad_norm": 2.5639536380767822,
+      "learning_rate": 2.7759398496240604e-06,
+      "loss": 0.1643,
+      "step": 48040
+    },
+    {
+      "epoch": 72.26,
+      "grad_norm": 3.273380756378174,
+      "learning_rate": 2.774436090225564e-06,
+      "loss": 0.1452,
+      "step": 48050
+    },
+    {
+      "epoch": 72.27,
+      "grad_norm": 5.514414310455322,
+      "learning_rate": 2.772932330827068e-06,
+      "loss": 0.1401,
+      "step": 48060
+    },
+    {
+      "epoch": 72.29,
+      "grad_norm": 6.009876251220703,
+      "learning_rate": 2.7714285714285714e-06,
+      "loss": 0.1812,
+      "step": 48070
+    },
+    {
+      "epoch": 72.3,
+      "grad_norm": 1.5807594060897827,
+      "learning_rate": 2.7699248120300753e-06,
+      "loss": 0.1392,
+      "step": 48080
+    },
+    {
+      "epoch": 72.32,
+      "grad_norm": 4.397557258605957,
+      "learning_rate": 2.7684210526315793e-06,
+      "loss": 0.1882,
+      "step": 48090
+    },
+    {
+      "epoch": 72.33,
+      "grad_norm": 5.294480323791504,
+      "learning_rate": 2.766917293233083e-06,
+      "loss": 0.2203,
+      "step": 48100
+    },
+    {
+      "epoch": 72.35,
+      "grad_norm": 2.36199688911438,
+      "learning_rate": 2.7654135338345867e-06,
+      "loss": 0.2092,
+      "step": 48110
+    },
+    {
+      "epoch": 72.36,
+      "grad_norm": 6.017412185668945,
+      "learning_rate": 2.7639097744360903e-06,
+      "loss": 0.1849,
+      "step": 48120
+    },
+    {
+      "epoch": 72.38,
+      "grad_norm": 6.914893627166748,
+      "learning_rate": 2.762406015037594e-06,
+      "loss": 0.1619,
+      "step": 48130
+    },
+    {
+      "epoch": 72.39,
+      "grad_norm": 3.912938117980957,
+      "learning_rate": 2.760902255639098e-06,
+      "loss": 0.2099,
+      "step": 48140
+    },
+    {
+      "epoch": 72.41,
+      "grad_norm": 4.893369197845459,
+      "learning_rate": 2.7593984962406017e-06,
+      "loss": 0.245,
+      "step": 48150
+    },
+    {
+      "epoch": 72.42,
+      "grad_norm": 3.4922471046447754,
+      "learning_rate": 2.7578947368421056e-06,
+      "loss": 0.1452,
+      "step": 48160
+    },
+    {
+      "epoch": 72.44,
+      "grad_norm": 10.81135082244873,
+      "learning_rate": 2.756390977443609e-06,
+      "loss": 0.1703,
+      "step": 48170
+    },
+    {
+      "epoch": 72.45,
+      "grad_norm": 5.498162269592285,
+      "learning_rate": 2.754887218045113e-06,
+      "loss": 0.1733,
+      "step": 48180
+    },
+    {
+      "epoch": 72.47,
+      "grad_norm": 3.2235593795776367,
+      "learning_rate": 2.753383458646617e-06,
+      "loss": 0.2167,
+      "step": 48190
+    },
+    {
+      "epoch": 72.48,
+      "grad_norm": 7.933281421661377,
+      "learning_rate": 2.7518796992481205e-06,
+      "loss": 0.2088,
+      "step": 48200
+    },
+    {
+      "epoch": 72.5,
+      "grad_norm": 3.2224316596984863,
+      "learning_rate": 2.7503759398496245e-06,
+      "loss": 0.1841,
+      "step": 48210
+    },
+    {
+      "epoch": 72.51,
+      "grad_norm": 5.602688312530518,
+      "learning_rate": 2.748872180451128e-06,
+      "loss": 0.1924,
+      "step": 48220
+    },
+    {
+      "epoch": 72.53,
+      "grad_norm": 4.668603897094727,
+      "learning_rate": 2.747368421052632e-06,
+      "loss": 0.2254,
+      "step": 48230
+    },
+    {
+      "epoch": 72.54,
+      "grad_norm": 5.575139045715332,
+      "learning_rate": 2.745864661654136e-06,
+      "loss": 0.1475,
+      "step": 48240
+    },
+    {
+      "epoch": 72.56,
+      "grad_norm": 5.640378475189209,
+      "learning_rate": 2.7443609022556394e-06,
+      "loss": 0.2488,
+      "step": 48250
+    },
+    {
+      "epoch": 72.57,
+      "grad_norm": 5.799993515014648,
+      "learning_rate": 2.7428571428571433e-06,
+      "loss": 0.206,
+      "step": 48260
+    },
+    {
+      "epoch": 72.59,
+      "grad_norm": 4.525552272796631,
+      "learning_rate": 2.741353383458647e-06,
+      "loss": 0.1442,
+      "step": 48270
+    },
+    {
+      "epoch": 72.6,
+      "grad_norm": 6.230297565460205,
+      "learning_rate": 2.739849624060151e-06,
+      "loss": 0.2157,
+      "step": 48280
+    },
+    {
+      "epoch": 72.62,
+      "grad_norm": 4.961235046386719,
+      "learning_rate": 2.7383458646616543e-06,
+      "loss": 0.1658,
+      "step": 48290
+    },
+    {
+      "epoch": 72.63,
+      "grad_norm": 6.751160621643066,
+      "learning_rate": 2.7368421052631583e-06,
+      "loss": 0.2332,
+      "step": 48300
+    },
+    {
+      "epoch": 72.65,
+      "grad_norm": 4.099827766418457,
+      "learning_rate": 2.735338345864662e-06,
+      "loss": 0.185,
+      "step": 48310
+    },
+    {
+      "epoch": 72.66,
+      "grad_norm": 6.973018646240234,
+      "learning_rate": 2.7338345864661657e-06,
+      "loss": 0.1865,
+      "step": 48320
+    },
+    {
+      "epoch": 72.68,
+      "grad_norm": 4.904829978942871,
+      "learning_rate": 2.7323308270676697e-06,
+      "loss": 0.1717,
+      "step": 48330
+    },
+    {
+      "epoch": 72.69,
+      "grad_norm": 7.860152721405029,
+      "learning_rate": 2.730827067669173e-06,
+      "loss": 0.1862,
+      "step": 48340
+    },
+    {
+      "epoch": 72.71,
+      "grad_norm": 5.617115020751953,
+      "learning_rate": 2.729323308270677e-06,
+      "loss": 0.2077,
+      "step": 48350
+    },
+    {
+      "epoch": 72.72,
+      "grad_norm": 4.507359504699707,
+      "learning_rate": 2.727819548872181e-06,
+      "loss": 0.2003,
+      "step": 48360
+    },
+    {
+      "epoch": 72.74,
+      "grad_norm": 4.982139587402344,
+      "learning_rate": 2.7263157894736846e-06,
+      "loss": 0.237,
+      "step": 48370
+    },
+    {
+      "epoch": 72.75,
+      "grad_norm": 4.0217604637146,
+      "learning_rate": 2.7248120300751885e-06,
+      "loss": 0.1453,
+      "step": 48380
+    },
+    {
+      "epoch": 72.77,
+      "grad_norm": 7.618423938751221,
+      "learning_rate": 2.7233082706766916e-06,
+      "loss": 0.1914,
+      "step": 48390
+    },
+    {
+      "epoch": 72.78,
+      "grad_norm": 6.787216663360596,
+      "learning_rate": 2.7218045112781956e-06,
+      "loss": 0.1451,
+      "step": 48400
+    },
+    {
+      "epoch": 72.8,
+      "grad_norm": 4.273850440979004,
+      "learning_rate": 2.720300751879699e-06,
+      "loss": 0.2088,
+      "step": 48410
+    },
+    {
+      "epoch": 72.81,
+      "grad_norm": 5.214813709259033,
+      "learning_rate": 2.718796992481203e-06,
+      "loss": 0.1597,
+      "step": 48420
+    },
+    {
+      "epoch": 72.83,
+      "grad_norm": 4.873306751251221,
+      "learning_rate": 2.7172932330827066e-06,
+      "loss": 0.1827,
+      "step": 48430
+    },
+    {
+      "epoch": 72.84,
+      "grad_norm": 5.74524450302124,
+      "learning_rate": 2.7157894736842105e-06,
+      "loss": 0.1795,
+      "step": 48440
+    },
+    {
+      "epoch": 72.86,
+      "grad_norm": 6.21406364440918,
+      "learning_rate": 2.7142857142857144e-06,
+      "loss": 0.2171,
+      "step": 48450
+    },
+    {
+      "epoch": 72.87,
+      "grad_norm": 3.394968032836914,
+      "learning_rate": 2.712781954887218e-06,
+      "loss": 0.1758,
+      "step": 48460
+    },
+    {
+      "epoch": 72.89,
+      "grad_norm": 2.936285972595215,
+      "learning_rate": 2.711278195488722e-06,
+      "loss": 0.2162,
+      "step": 48470
+    },
+    {
+      "epoch": 72.9,
+      "grad_norm": 2.2436184883117676,
+      "learning_rate": 2.7097744360902254e-06,
+      "loss": 0.1331,
+      "step": 48480
+    },
+    {
+      "epoch": 72.92,
+      "grad_norm": 4.331397533416748,
+      "learning_rate": 2.7082706766917294e-06,
+      "loss": 0.1934,
+      "step": 48490
+    },
+    {
+      "epoch": 72.93,
+      "grad_norm": 8.824311256408691,
+      "learning_rate": 2.7067669172932333e-06,
+      "loss": 0.2538,
+      "step": 48500
+    },
+    {
+      "epoch": 72.95,
+      "grad_norm": 7.463961124420166,
+      "learning_rate": 2.705263157894737e-06,
+      "loss": 0.212,
+      "step": 48510
+    },
+    {
+      "epoch": 72.96,
+      "grad_norm": 9.955108642578125,
+      "learning_rate": 2.7037593984962408e-06,
+      "loss": 0.2191,
+      "step": 48520
+    },
+    {
+      "epoch": 72.98,
+      "grad_norm": 3.3155603408813477,
+      "learning_rate": 2.7022556390977443e-06,
+      "loss": 0.1842,
+      "step": 48530
+    },
+    {
+      "epoch": 72.99,
+      "grad_norm": 3.515291929244995,
+      "learning_rate": 2.7007518796992482e-06,
+      "loss": 0.1914,
+      "step": 48540
+    },
+    {
+      "epoch": 73.0,
+      "eval_accuracy": 0.931,
+      "eval_loss": 0.3263280391693115,
+      "eval_runtime": 84.3768,
+      "eval_samples_per_second": 118.516,
+      "eval_steps_per_second": 0.474,
+      "step": 48545
+    },
+    {
+      "epoch": 73.01,
+      "grad_norm": 5.640552520751953,
+      "learning_rate": 2.699248120300752e-06,
+      "loss": 0.1428,
+      "step": 48550
+    },
+    {
+      "epoch": 73.02,
+      "grad_norm": 4.772902965545654,
+      "learning_rate": 2.6977443609022557e-06,
+      "loss": 0.1821,
+      "step": 48560
+    },
+    {
+      "epoch": 73.04,
+      "grad_norm": 3.3709876537323,
+      "learning_rate": 2.6962406015037596e-06,
+      "loss": 0.1939,
+      "step": 48570
+    },
+    {
+      "epoch": 73.05,
+      "grad_norm": 8.709583282470703,
+      "learning_rate": 2.694736842105263e-06,
+      "loss": 0.24,
+      "step": 48580
+    },
+    {
+      "epoch": 73.07,
+      "grad_norm": 3.556995391845703,
+      "learning_rate": 2.693233082706767e-06,
+      "loss": 0.1939,
+      "step": 48590
+    },
+    {
+      "epoch": 73.08,
+      "grad_norm": 4.488801002502441,
+      "learning_rate": 2.691729323308271e-06,
+      "loss": 0.1767,
+      "step": 48600
+    },
+    {
+      "epoch": 73.1,
+      "grad_norm": 3.0920093059539795,
+      "learning_rate": 2.6902255639097746e-06,
+      "loss": 0.148,
+      "step": 48610
+    },
+    {
+      "epoch": 73.11,
+      "grad_norm": 7.643060207366943,
+      "learning_rate": 2.6887218045112785e-06,
+      "loss": 0.1924,
+      "step": 48620
+    },
+    {
+      "epoch": 73.13,
+      "grad_norm": 11.217390060424805,
+      "learning_rate": 2.687218045112782e-06,
+      "loss": 0.2318,
+      "step": 48630
+    },
+    {
+      "epoch": 73.14,
+      "grad_norm": 8.313175201416016,
+      "learning_rate": 2.685714285714286e-06,
+      "loss": 0.1629,
+      "step": 48640
+    },
+    {
+      "epoch": 73.16,
+      "grad_norm": 4.154397010803223,
+      "learning_rate": 2.68421052631579e-06,
+      "loss": 0.1958,
+      "step": 48650
+    },
+    {
+      "epoch": 73.17,
+      "grad_norm": 3.842160224914551,
+      "learning_rate": 2.6827067669172934e-06,
+      "loss": 0.1451,
+      "step": 48660
+    },
+    {
+      "epoch": 73.19,
+      "grad_norm": 6.915648460388184,
+      "learning_rate": 2.6812030075187974e-06,
+      "loss": 0.1452,
+      "step": 48670
+    },
+    {
+      "epoch": 73.2,
+      "grad_norm": 4.855998992919922,
+      "learning_rate": 2.679699248120301e-06,
+      "loss": 0.1455,
+      "step": 48680
+    },
+    {
+      "epoch": 73.22,
+      "grad_norm": 1.8087424039840698,
+      "learning_rate": 2.678195488721805e-06,
+      "loss": 0.1554,
+      "step": 48690
+    },
+    {
+      "epoch": 73.23,
+      "grad_norm": 1.4605119228363037,
+      "learning_rate": 2.6766917293233088e-06,
+      "loss": 0.2174,
+      "step": 48700
+    },
+    {
+      "epoch": 73.25,
+      "grad_norm": 6.26193380355835,
+      "learning_rate": 2.6751879699248123e-06,
+      "loss": 0.2441,
+      "step": 48710
+    },
+    {
+      "epoch": 73.26,
+      "grad_norm": 6.614605903625488,
+      "learning_rate": 2.6736842105263162e-06,
+      "loss": 0.2272,
+      "step": 48720
+    },
+    {
+      "epoch": 73.28,
+      "grad_norm": 3.6685211658477783,
+      "learning_rate": 2.6721804511278197e-06,
+      "loss": 0.1562,
+      "step": 48730
+    },
+    {
+      "epoch": 73.29,
+      "grad_norm": 4.120753288269043,
+      "learning_rate": 2.6706766917293237e-06,
+      "loss": 0.1567,
+      "step": 48740
+    },
+    {
+      "epoch": 73.31,
+      "grad_norm": 16.986003875732422,
+      "learning_rate": 2.669172932330827e-06,
+      "loss": 0.2109,
+      "step": 48750
+    },
+    {
+      "epoch": 73.32,
+      "grad_norm": 6.729957580566406,
+      "learning_rate": 2.667669172932331e-06,
+      "loss": 0.2327,
+      "step": 48760
+    },
+    {
+      "epoch": 73.34,
+      "grad_norm": 3.7136998176574707,
+      "learning_rate": 2.666165413533835e-06,
+      "loss": 0.2161,
+      "step": 48770
+    },
+    {
+      "epoch": 73.35,
+      "grad_norm": 3.925506591796875,
+      "learning_rate": 2.6646616541353386e-06,
+      "loss": 0.2049,
+      "step": 48780
+    },
+    {
+      "epoch": 73.37,
+      "grad_norm": 11.8861665725708,
+      "learning_rate": 2.6631578947368426e-06,
+      "loss": 0.1432,
+      "step": 48790
+    },
+    {
+      "epoch": 73.38,
+      "grad_norm": 4.450233459472656,
+      "learning_rate": 2.661654135338346e-06,
+      "loss": 0.1958,
+      "step": 48800
+    },
+    {
+      "epoch": 73.4,
+      "grad_norm": 7.333054542541504,
+      "learning_rate": 2.66015037593985e-06,
+      "loss": 0.2231,
+      "step": 48810
+    },
+    {
+      "epoch": 73.41,
+      "grad_norm": 5.88569974899292,
+      "learning_rate": 2.658646616541354e-06,
+      "loss": 0.1502,
+      "step": 48820
+    },
+    {
+      "epoch": 73.43,
+      "grad_norm": 4.52285099029541,
+      "learning_rate": 2.6571428571428575e-06,
+      "loss": 0.2144,
+      "step": 48830
+    },
+    {
+      "epoch": 73.44,
+      "grad_norm": 6.34915828704834,
+      "learning_rate": 2.6556390977443614e-06,
+      "loss": 0.221,
+      "step": 48840
+    },
+    {
+      "epoch": 73.46,
+      "grad_norm": 2.9954657554626465,
+      "learning_rate": 2.654135338345865e-06,
+      "loss": 0.1306,
+      "step": 48850
+    },
+    {
+      "epoch": 73.47,
+      "grad_norm": 4.723332405090332,
+      "learning_rate": 2.652631578947369e-06,
+      "loss": 0.1725,
+      "step": 48860
+    },
+    {
+      "epoch": 73.49,
+      "grad_norm": 5.571535587310791,
+      "learning_rate": 2.651127819548873e-06,
+      "loss": 0.2305,
+      "step": 48870
+    },
+    {
+      "epoch": 73.5,
+      "grad_norm": 6.174119472503662,
+      "learning_rate": 2.6496240601503763e-06,
+      "loss": 0.2203,
+      "step": 48880
+    },
+    {
+      "epoch": 73.52,
+      "grad_norm": 3.709364175796509,
+      "learning_rate": 2.6481203007518794e-06,
+      "loss": 0.1775,
+      "step": 48890
+    },
+    {
+      "epoch": 73.53,
+      "grad_norm": 7.268237590789795,
+      "learning_rate": 2.6466165413533834e-06,
+      "loss": 0.212,
+      "step": 48900
+    },
+    {
+      "epoch": 73.55,
+      "grad_norm": 1.238796591758728,
+      "learning_rate": 2.6451127819548873e-06,
+      "loss": 0.1747,
+      "step": 48910
+    },
+    {
+      "epoch": 73.56,
+      "grad_norm": 3.121783971786499,
+      "learning_rate": 2.643609022556391e-06,
+      "loss": 0.1938,
+      "step": 48920
+    },
+    {
+      "epoch": 73.58,
+      "grad_norm": 4.7542500495910645,
+      "learning_rate": 2.6421052631578948e-06,
+      "loss": 0.2038,
+      "step": 48930
+    },
+    {
+      "epoch": 73.59,
+      "grad_norm": 5.120677471160889,
+      "learning_rate": 2.6406015037593983e-06,
+      "loss": 0.2355,
+      "step": 48940
+    },
+    {
+      "epoch": 73.61,
+      "grad_norm": 6.299097537994385,
+      "learning_rate": 2.6390977443609022e-06,
+      "loss": 0.2069,
+      "step": 48950
+    },
+    {
+      "epoch": 73.62,
+      "grad_norm": 4.0084428787231445,
+      "learning_rate": 2.637593984962406e-06,
+      "loss": 0.1697,
+      "step": 48960
+    },
+    {
+      "epoch": 73.64,
+      "grad_norm": 3.608576536178589,
+      "learning_rate": 2.6360902255639097e-06,
+      "loss": 0.1726,
+      "step": 48970
+    },
+    {
+      "epoch": 73.65,
+      "grad_norm": 5.771646499633789,
+      "learning_rate": 2.6345864661654137e-06,
+      "loss": 0.2716,
+      "step": 48980
+    },
+    {
+      "epoch": 73.67,
+      "grad_norm": 8.164382934570312,
+      "learning_rate": 2.633082706766917e-06,
+      "loss": 0.1862,
+      "step": 48990
+    },
+    {
+      "epoch": 73.68,
+      "grad_norm": 5.360281467437744,
+      "learning_rate": 2.631578947368421e-06,
+      "loss": 0.163,
+      "step": 49000
+    },
+    {
+      "epoch": 73.7,
+      "grad_norm": 5.156652450561523,
+      "learning_rate": 2.630075187969925e-06,
+      "loss": 0.1928,
+      "step": 49010
+    },
+    {
+      "epoch": 73.71,
+      "grad_norm": 3.7978944778442383,
+      "learning_rate": 2.6285714285714286e-06,
+      "loss": 0.1318,
+      "step": 49020
+    },
+    {
+      "epoch": 73.73,
+      "grad_norm": 7.672252655029297,
+      "learning_rate": 2.6270676691729325e-06,
+      "loss": 0.2079,
+      "step": 49030
+    },
+    {
+      "epoch": 73.74,
+      "grad_norm": 4.497259616851807,
+      "learning_rate": 2.625563909774436e-06,
+      "loss": 0.2074,
+      "step": 49040
+    },
+    {
+      "epoch": 73.76,
+      "grad_norm": 4.420539855957031,
+      "learning_rate": 2.62406015037594e-06,
+      "loss": 0.1372,
+      "step": 49050
+    },
+    {
+      "epoch": 73.77,
+      "grad_norm": 4.705691814422607,
+      "learning_rate": 2.622556390977444e-06,
+      "loss": 0.2247,
+      "step": 49060
+    },
+    {
+      "epoch": 73.79,
+      "grad_norm": 4.325020790100098,
+      "learning_rate": 2.6210526315789474e-06,
+      "loss": 0.1768,
+      "step": 49070
+    },
+    {
+      "epoch": 73.8,
+      "grad_norm": 4.0322442054748535,
+      "learning_rate": 2.6195488721804514e-06,
+      "loss": 0.1707,
+      "step": 49080
+    },
+    {
+      "epoch": 73.82,
+      "grad_norm": 7.42966890335083,
+      "learning_rate": 2.618045112781955e-06,
+      "loss": 0.2521,
+      "step": 49090
+    },
+    {
+      "epoch": 73.83,
+      "grad_norm": 4.630422115325928,
+      "learning_rate": 2.616541353383459e-06,
+      "loss": 0.2168,
+      "step": 49100
+    },
+    {
+      "epoch": 73.85,
+      "grad_norm": 5.983313083648682,
+      "learning_rate": 2.6150375939849628e-06,
+      "loss": 0.2102,
+      "step": 49110
+    },
+    {
+      "epoch": 73.86,
+      "grad_norm": 2.551830291748047,
+      "learning_rate": 2.6135338345864663e-06,
+      "loss": 0.1586,
+      "step": 49120
+    },
+    {
+      "epoch": 73.88,
+      "grad_norm": 6.646661281585693,
+      "learning_rate": 2.6120300751879702e-06,
+      "loss": 0.2198,
+      "step": 49130
+    },
+    {
+      "epoch": 73.89,
+      "grad_norm": 5.7124176025390625,
+      "learning_rate": 2.6105263157894738e-06,
+      "loss": 0.2081,
+      "step": 49140
+    },
+    {
+      "epoch": 73.91,
+      "grad_norm": 15.754138946533203,
+      "learning_rate": 2.6090225563909777e-06,
+      "loss": 0.1661,
+      "step": 49150
+    },
+    {
+      "epoch": 73.92,
+      "grad_norm": 5.884487152099609,
+      "learning_rate": 2.6075187969924817e-06,
+      "loss": 0.1548,
+      "step": 49160
+    },
+    {
+      "epoch": 73.94,
+      "grad_norm": 1.3139573335647583,
+      "learning_rate": 2.606015037593985e-06,
+      "loss": 0.1773,
+      "step": 49170
+    },
+    {
+      "epoch": 73.95,
+      "grad_norm": 4.792730808258057,
+      "learning_rate": 2.604511278195489e-06,
+      "loss": 0.1666,
+      "step": 49180
+    },
+    {
+      "epoch": 73.97,
+      "grad_norm": 1.9457217454910278,
+      "learning_rate": 2.6030075187969926e-06,
+      "loss": 0.1575,
+      "step": 49190
+    },
+    {
+      "epoch": 73.98,
+      "grad_norm": 6.806911468505859,
+      "learning_rate": 2.6015037593984966e-06,
+      "loss": 0.1794,
+      "step": 49200
+    },
+    {
+      "epoch": 74.0,
+      "grad_norm": 0.026680290699005127,
+      "learning_rate": 2.6e-06,
+      "loss": 0.2097,
+      "step": 49210
+    },
+    {
+      "epoch": 74.0,
+      "eval_accuracy": 0.9294,
+      "eval_loss": 0.33669278025627136,
+      "eval_runtime": 84.9311,
+      "eval_samples_per_second": 117.743,
+      "eval_steps_per_second": 0.471,
+      "step": 49210
+    },
+    {
+      "epoch": 74.02,
+      "grad_norm": 6.460815906524658,
+      "learning_rate": 2.598496240601504e-06,
+      "loss": 0.1717,
+      "step": 49220
+    },
+    {
+      "epoch": 74.03,
+      "grad_norm": 5.829145908355713,
+      "learning_rate": 2.596992481203008e-06,
+      "loss": 0.2016,
+      "step": 49230
+    },
+    {
+      "epoch": 74.05,
+      "grad_norm": 11.389643669128418,
+      "learning_rate": 2.5954887218045115e-06,
+      "loss": 0.2195,
+      "step": 49240
+    },
+    {
+      "epoch": 74.06,
+      "grad_norm": 2.792567253112793,
+      "learning_rate": 2.5939849624060154e-06,
+      "loss": 0.1553,
+      "step": 49250
+    },
+    {
+      "epoch": 74.08,
+      "grad_norm": 3.7517435550689697,
+      "learning_rate": 2.592481203007519e-06,
+      "loss": 0.1597,
+      "step": 49260
+    },
+    {
+      "epoch": 74.09,
+      "grad_norm": 4.992012023925781,
+      "learning_rate": 2.590977443609023e-06,
+      "loss": 0.1487,
+      "step": 49270
+    },
+    {
+      "epoch": 74.11,
+      "grad_norm": 4.339962482452393,
+      "learning_rate": 2.589473684210527e-06,
+      "loss": 0.1671,
+      "step": 49280
+    },
+    {
+      "epoch": 74.12,
+      "grad_norm": 2.0285701751708984,
+      "learning_rate": 2.5879699248120304e-06,
+      "loss": 0.1598,
+      "step": 49290
+    },
+    {
+      "epoch": 74.14,
+      "grad_norm": 3.2112770080566406,
+      "learning_rate": 2.5864661654135343e-06,
+      "loss": 0.1299,
+      "step": 49300
+    },
+    {
+      "epoch": 74.15,
+      "grad_norm": 3.4635910987854004,
+      "learning_rate": 2.584962406015038e-06,
+      "loss": 0.1812,
+      "step": 49310
+    },
+    {
+      "epoch": 74.17,
+      "grad_norm": 8.604695320129395,
+      "learning_rate": 2.5834586466165418e-06,
+      "loss": 0.2481,
+      "step": 49320
+    },
+    {
+      "epoch": 74.18,
+      "grad_norm": 3.5391902923583984,
+      "learning_rate": 2.5819548872180457e-06,
+      "loss": 0.1636,
+      "step": 49330
+    },
+    {
+      "epoch": 74.2,
+      "grad_norm": 1.695279836654663,
+      "learning_rate": 2.5804511278195492e-06,
+      "loss": 0.1739,
+      "step": 49340
+    },
+    {
+      "epoch": 74.21,
+      "grad_norm": 4.957259178161621,
+      "learning_rate": 2.578947368421053e-06,
+      "loss": 0.1369,
+      "step": 49350
+    },
+    {
+      "epoch": 74.23,
+      "grad_norm": 6.202530860900879,
+      "learning_rate": 2.5774436090225567e-06,
+      "loss": 0.2257,
+      "step": 49360
+    },
+    {
+      "epoch": 74.24,
+      "grad_norm": 7.947466850280762,
+      "learning_rate": 2.5759398496240606e-06,
+      "loss": 0.1961,
+      "step": 49370
+    },
+    {
+      "epoch": 74.26,
+      "grad_norm": 6.451323509216309,
+      "learning_rate": 2.5744360902255637e-06,
+      "loss": 0.2195,
+      "step": 49380
+    },
+    {
+      "epoch": 74.27,
+      "grad_norm": 5.297338008880615,
+      "learning_rate": 2.5729323308270677e-06,
+      "loss": 0.1799,
+      "step": 49390
+    },
+    {
+      "epoch": 74.29,
+      "grad_norm": 4.922117233276367,
+      "learning_rate": 2.571428571428571e-06,
+      "loss": 0.2173,
+      "step": 49400
+    },
+    {
+      "epoch": 74.3,
+      "grad_norm": 1.2803465127944946,
+      "learning_rate": 2.569924812030075e-06,
+      "loss": 0.1658,
+      "step": 49410
+    },
+    {
+      "epoch": 74.32,
+      "grad_norm": 2.5603795051574707,
+      "learning_rate": 2.568421052631579e-06,
+      "loss": 0.1427,
+      "step": 49420
+    },
+    {
+      "epoch": 74.33,
+      "grad_norm": 5.846049785614014,
+      "learning_rate": 2.5669172932330826e-06,
+      "loss": 0.2481,
+      "step": 49430
+    },
+    {
+      "epoch": 74.35,
+      "grad_norm": 6.60819149017334,
+      "learning_rate": 2.5654135338345865e-06,
+      "loss": 0.1853,
+      "step": 49440
+    },
+    {
+      "epoch": 74.36,
+      "grad_norm": 3.9946815967559814,
+      "learning_rate": 2.56390977443609e-06,
+      "loss": 0.2012,
+      "step": 49450
+    },
+    {
+      "epoch": 74.38,
+      "grad_norm": 9.012832641601562,
+      "learning_rate": 2.562406015037594e-06,
+      "loss": 0.2052,
+      "step": 49460
+    },
+    {
+      "epoch": 74.39,
+      "grad_norm": 4.631398677825928,
+      "learning_rate": 2.560902255639098e-06,
+      "loss": 0.1995,
+      "step": 49470
+    },
+    {
+      "epoch": 74.41,
+      "grad_norm": 3.723080635070801,
+      "learning_rate": 2.5593984962406015e-06,
+      "loss": 0.1542,
+      "step": 49480
+    },
+    {
+      "epoch": 74.42,
+      "grad_norm": 5.1829633712768555,
+      "learning_rate": 2.5578947368421054e-06,
+      "loss": 0.2244,
+      "step": 49490
+    },
+    {
+      "epoch": 74.44,
+      "grad_norm": 6.623849391937256,
+      "learning_rate": 2.556390977443609e-06,
+      "loss": 0.2161,
+      "step": 49500
+    },
+    {
+      "epoch": 74.45,
+      "grad_norm": 7.301265239715576,
+      "learning_rate": 2.554887218045113e-06,
+      "loss": 0.1301,
+      "step": 49510
+    },
+    {
+      "epoch": 74.47,
+      "grad_norm": 3.6955883502960205,
+      "learning_rate": 2.553383458646617e-06,
+      "loss": 0.1758,
+      "step": 49520
+    },
+    {
+      "epoch": 74.48,
+      "grad_norm": 9.507878303527832,
+      "learning_rate": 2.5518796992481203e-06,
+      "loss": 0.1833,
+      "step": 49530
+    },
+    {
+      "epoch": 74.5,
+      "grad_norm": 3.5736141204833984,
+      "learning_rate": 2.5503759398496243e-06,
+      "loss": 0.1582,
+      "step": 49540
+    },
+    {
+      "epoch": 74.51,
+      "grad_norm": 4.5012006759643555,
+      "learning_rate": 2.548872180451128e-06,
+      "loss": 0.1849,
+      "step": 49550
+    },
+    {
+      "epoch": 74.53,
+      "grad_norm": 1.9890389442443848,
+      "learning_rate": 2.5473684210526317e-06,
+      "loss": 0.2109,
+      "step": 49560
+    },
+    {
+      "epoch": 74.54,
+      "grad_norm": 6.087845325469971,
+      "learning_rate": 2.5458646616541357e-06,
+      "loss": 0.2116,
+      "step": 49570
+    },
+    {
+      "epoch": 74.56,
+      "grad_norm": 4.9256157875061035,
+      "learning_rate": 2.544360902255639e-06,
+      "loss": 0.1751,
+      "step": 49580
+    },
+    {
+      "epoch": 74.57,
+      "grad_norm": 7.381353855133057,
+      "learning_rate": 2.542857142857143e-06,
+      "loss": 0.1588,
+      "step": 49590
+    },
+    {
+      "epoch": 74.59,
+      "grad_norm": 2.9443299770355225,
+      "learning_rate": 2.5413533834586467e-06,
+      "loss": 0.1763,
+      "step": 49600
+    },
+    {
+      "epoch": 74.6,
+      "grad_norm": 4.252871513366699,
+      "learning_rate": 2.5398496240601506e-06,
+      "loss": 0.1819,
+      "step": 49610
+    },
+    {
+      "epoch": 74.62,
+      "grad_norm": 4.36507511138916,
+      "learning_rate": 2.5383458646616545e-06,
+      "loss": 0.1788,
+      "step": 49620
+    },
+    {
+      "epoch": 74.63,
+      "grad_norm": 5.069661617279053,
+      "learning_rate": 2.536842105263158e-06,
+      "loss": 0.2279,
+      "step": 49630
+    },
+    {
+      "epoch": 74.65,
+      "grad_norm": 1.5738500356674194,
+      "learning_rate": 2.535338345864662e-06,
+      "loss": 0.1562,
+      "step": 49640
+    },
+    {
+      "epoch": 74.66,
+      "grad_norm": 3.425536632537842,
+      "learning_rate": 2.5338345864661655e-06,
+      "loss": 0.1367,
+      "step": 49650
+    },
+    {
+      "epoch": 74.68,
+      "grad_norm": 7.227283954620361,
+      "learning_rate": 2.5323308270676695e-06,
+      "loss": 0.1524,
+      "step": 49660
+    },
+    {
+      "epoch": 74.69,
+      "grad_norm": 4.1548075675964355,
+      "learning_rate": 2.530827067669173e-06,
+      "loss": 0.1361,
+      "step": 49670
+    },
+    {
+      "epoch": 74.71,
+      "grad_norm": 5.1794610023498535,
+      "learning_rate": 2.529323308270677e-06,
+      "loss": 0.1373,
+      "step": 49680
+    },
+    {
+      "epoch": 74.72,
+      "grad_norm": 7.747637748718262,
+      "learning_rate": 2.527819548872181e-06,
+      "loss": 0.1758,
+      "step": 49690
+    },
+    {
+      "epoch": 74.74,
+      "grad_norm": 3.845520496368408,
+      "learning_rate": 2.5263157894736844e-06,
+      "loss": 0.2025,
+      "step": 49700
+    },
+    {
+      "epoch": 74.75,
+      "grad_norm": 8.356101989746094,
+      "learning_rate": 2.5248120300751883e-06,
+      "loss": 0.1597,
+      "step": 49710
+    },
+    {
+      "epoch": 74.77,
+      "grad_norm": 3.1833062171936035,
+      "learning_rate": 2.523308270676692e-06,
+      "loss": 0.1411,
+      "step": 49720
+    },
+    {
+      "epoch": 74.78,
+      "grad_norm": 6.692790508270264,
+      "learning_rate": 2.521804511278196e-06,
+      "loss": 0.1809,
+      "step": 49730
+    },
+    {
+      "epoch": 74.8,
+      "grad_norm": 6.7796454429626465,
+      "learning_rate": 2.5203007518796997e-06,
+      "loss": 0.1418,
+      "step": 49740
+    },
+    {
+      "epoch": 74.81,
+      "grad_norm": 4.885359287261963,
+      "learning_rate": 2.5187969924812033e-06,
+      "loss": 0.2639,
+      "step": 49750
+    },
+    {
+      "epoch": 74.83,
+      "grad_norm": 6.379741668701172,
+      "learning_rate": 2.517293233082707e-06,
+      "loss": 0.2007,
+      "step": 49760
+    },
+    {
+      "epoch": 74.84,
+      "grad_norm": 7.056103706359863,
+      "learning_rate": 2.5157894736842107e-06,
+      "loss": 0.2262,
+      "step": 49770
+    },
+    {
+      "epoch": 74.86,
+      "grad_norm": 7.777561664581299,
+      "learning_rate": 2.5142857142857147e-06,
+      "loss": 0.2332,
+      "step": 49780
+    },
+    {
+      "epoch": 74.87,
+      "grad_norm": 5.086285591125488,
+      "learning_rate": 2.5127819548872186e-06,
+      "loss": 0.2878,
+      "step": 49790
+    },
+    {
+      "epoch": 74.89,
+      "grad_norm": 2.857342481613159,
+      "learning_rate": 2.511278195488722e-06,
+      "loss": 0.1854,
+      "step": 49800
+    },
+    {
+      "epoch": 74.9,
+      "grad_norm": 2.6614773273468018,
+      "learning_rate": 2.509774436090226e-06,
+      "loss": 0.1806,
+      "step": 49810
+    },
+    {
+      "epoch": 74.92,
+      "grad_norm": 6.267194747924805,
+      "learning_rate": 2.5082706766917296e-06,
+      "loss": 0.1506,
+      "step": 49820
+    },
+    {
+      "epoch": 74.93,
+      "grad_norm": 7.167734622955322,
+      "learning_rate": 2.5067669172932335e-06,
+      "loss": 0.1308,
+      "step": 49830
+    },
+    {
+      "epoch": 74.95,
+      "grad_norm": 3.8216676712036133,
+      "learning_rate": 2.5052631578947375e-06,
+      "loss": 0.1914,
+      "step": 49840
+    },
+    {
+      "epoch": 74.96,
+      "grad_norm": 4.203033447265625,
+      "learning_rate": 2.503759398496241e-06,
+      "loss": 0.1303,
+      "step": 49850
+    },
+    {
+      "epoch": 74.98,
+      "grad_norm": 2.4965991973876953,
+      "learning_rate": 2.502255639097745e-06,
+      "loss": 0.1785,
+      "step": 49860
+    },
+    {
+      "epoch": 74.99,
+      "grad_norm": 1.2828130722045898,
+      "learning_rate": 2.5007518796992484e-06,
+      "loss": 0.1423,
+      "step": 49870
+    },
+    {
+      "epoch": 75.0,
+      "eval_accuracy": 0.9299,
+      "eval_loss": 0.3285817503929138,
+      "eval_runtime": 84.9307,
+      "eval_samples_per_second": 117.743,
+      "eval_steps_per_second": 0.471,
+      "step": 49875
+    },
+    {
+      "epoch": 75.01,
+      "grad_norm": 3.812678337097168,
+      "learning_rate": 2.499248120300752e-06,
+      "loss": 0.1716,
+      "step": 49880
+    },
+    {
+      "epoch": 75.02,
+      "grad_norm": 1.7645814418792725,
+      "learning_rate": 2.497744360902256e-06,
+      "loss": 0.1242,
+      "step": 49890
+    },
+    {
+      "epoch": 75.04,
+      "grad_norm": 3.069406509399414,
+      "learning_rate": 2.4962406015037594e-06,
+      "loss": 0.1708,
+      "step": 49900
+    },
+    {
+      "epoch": 75.05,
+      "grad_norm": 4.357547283172607,
+      "learning_rate": 2.4947368421052634e-06,
+      "loss": 0.1933,
+      "step": 49910
+    },
+    {
+      "epoch": 75.07,
+      "grad_norm": 6.222829818725586,
+      "learning_rate": 2.4932330827067673e-06,
+      "loss": 0.1914,
+      "step": 49920
+    },
+    {
+      "epoch": 75.08,
+      "grad_norm": 5.128503322601318,
+      "learning_rate": 2.491729323308271e-06,
+      "loss": 0.2163,
+      "step": 49930
+    },
+    {
+      "epoch": 75.1,
+      "grad_norm": 4.528236389160156,
+      "learning_rate": 2.4902255639097748e-06,
+      "loss": 0.1979,
+      "step": 49940
+    },
+    {
+      "epoch": 75.11,
+      "grad_norm": 3.4620494842529297,
+      "learning_rate": 2.4887218045112783e-06,
+      "loss": 0.1922,
+      "step": 49950
+    },
+    {
+      "epoch": 75.13,
+      "grad_norm": 7.054966449737549,
+      "learning_rate": 2.4872180451127822e-06,
+      "loss": 0.242,
+      "step": 49960
+    },
+    {
+      "epoch": 75.14,
+      "grad_norm": 5.699782371520996,
+      "learning_rate": 2.485714285714286e-06,
+      "loss": 0.1489,
+      "step": 49970
+    },
+    {
+      "epoch": 75.16,
+      "grad_norm": 6.1800408363342285,
+      "learning_rate": 2.4842105263157897e-06,
+      "loss": 0.1506,
+      "step": 49980
+    },
+    {
+      "epoch": 75.17,
+      "grad_norm": 2.10766863822937,
+      "learning_rate": 2.4827067669172936e-06,
+      "loss": 0.1478,
+      "step": 49990
+    },
+    {
+      "epoch": 75.19,
+      "grad_norm": 3.423696279525757,
+      "learning_rate": 2.481203007518797e-06,
+      "loss": 0.1526,
+      "step": 50000
+    },
+    {
+      "epoch": 75.2,
+      "grad_norm": 3.7721095085144043,
+      "learning_rate": 2.4796992481203007e-06,
+      "loss": 0.2094,
+      "step": 50010
+    },
+    {
+      "epoch": 75.22,
+      "grad_norm": 5.331075668334961,
+      "learning_rate": 2.4781954887218046e-06,
+      "loss": 0.1597,
+      "step": 50020
+    },
+    {
+      "epoch": 75.23,
+      "grad_norm": 6.622517108917236,
+      "learning_rate": 2.4766917293233086e-06,
+      "loss": 0.2322,
+      "step": 50030
+    },
+    {
+      "epoch": 75.25,
+      "grad_norm": 6.970418930053711,
+      "learning_rate": 2.475187969924812e-06,
+      "loss": 0.1812,
+      "step": 50040
+    },
+    {
+      "epoch": 75.26,
+      "grad_norm": 7.753775119781494,
+      "learning_rate": 2.473684210526316e-06,
+      "loss": 0.1515,
+      "step": 50050
+    },
+    {
+      "epoch": 75.28,
+      "grad_norm": 13.116275787353516,
+      "learning_rate": 2.4721804511278195e-06,
+      "loss": 0.1887,
+      "step": 50060
+    },
+    {
+      "epoch": 75.29,
+      "grad_norm": 4.448867321014404,
+      "learning_rate": 2.4706766917293235e-06,
+      "loss": 0.1433,
+      "step": 50070
+    },
+    {
+      "epoch": 75.31,
+      "grad_norm": 3.7794928550720215,
+      "learning_rate": 2.4691729323308274e-06,
+      "loss": 0.1493,
+      "step": 50080
+    },
+    {
+      "epoch": 75.32,
+      "grad_norm": 5.0668416023254395,
+      "learning_rate": 2.467669172932331e-06,
+      "loss": 0.1536,
+      "step": 50090
+    },
+    {
+      "epoch": 75.34,
+      "grad_norm": 3.405148983001709,
+      "learning_rate": 2.466165413533835e-06,
+      "loss": 0.222,
+      "step": 50100
+    },
+    {
+      "epoch": 75.35,
+      "grad_norm": 5.900206565856934,
+      "learning_rate": 2.4646616541353384e-06,
+      "loss": 0.1388,
+      "step": 50110
+    },
+    {
+      "epoch": 75.37,
+      "grad_norm": 4.416190147399902,
+      "learning_rate": 2.4631578947368424e-06,
+      "loss": 0.1744,
+      "step": 50120
+    },
+    {
+      "epoch": 75.38,
+      "grad_norm": 7.959252834320068,
+      "learning_rate": 2.461654135338346e-06,
+      "loss": 0.2386,
+      "step": 50130
+    },
+    {
+      "epoch": 75.4,
+      "grad_norm": 4.957620143890381,
+      "learning_rate": 2.46015037593985e-06,
+      "loss": 0.1717,
+      "step": 50140
+    },
+    {
+      "epoch": 75.41,
+      "grad_norm": 6.785633087158203,
+      "learning_rate": 2.4586466165413538e-06,
+      "loss": 0.1783,
+      "step": 50150
+    },
+    {
+      "epoch": 75.43,
+      "grad_norm": 8.179920196533203,
+      "learning_rate": 2.4571428571428573e-06,
+      "loss": 0.2003,
+      "step": 50160
+    },
+    {
+      "epoch": 75.44,
+      "grad_norm": 2.4973435401916504,
+      "learning_rate": 2.4556390977443612e-06,
+      "loss": 0.1877,
+      "step": 50170
+    },
+    {
+      "epoch": 75.46,
+      "grad_norm": 4.815390586853027,
+      "learning_rate": 2.4541353383458647e-06,
+      "loss": 0.26,
+      "step": 50180
+    },
+    {
+      "epoch": 75.47,
+      "grad_norm": 6.23560905456543,
+      "learning_rate": 2.4526315789473687e-06,
+      "loss": 0.1683,
+      "step": 50190
+    },
+    {
+      "epoch": 75.49,
+      "grad_norm": 0.35644736886024475,
+      "learning_rate": 2.4511278195488726e-06,
+      "loss": 0.1489,
+      "step": 50200
+    },
+    {
+      "epoch": 75.5,
+      "grad_norm": 5.562798500061035,
+      "learning_rate": 2.449624060150376e-06,
+      "loss": 0.2343,
+      "step": 50210
+    },
+    {
+      "epoch": 75.52,
+      "grad_norm": 6.464638710021973,
+      "learning_rate": 2.44812030075188e-06,
+      "loss": 0.2042,
+      "step": 50220
+    },
+    {
+      "epoch": 75.53,
+      "grad_norm": 3.009085178375244,
+      "learning_rate": 2.4466165413533836e-06,
+      "loss": 0.1565,
+      "step": 50230
+    },
+    {
+      "epoch": 75.55,
+      "grad_norm": 3.3092687129974365,
+      "learning_rate": 2.4451127819548875e-06,
+      "loss": 0.1583,
+      "step": 50240
+    },
+    {
+      "epoch": 75.56,
+      "grad_norm": 1.3774584531784058,
+      "learning_rate": 2.443609022556391e-06,
+      "loss": 0.1763,
+      "step": 50250
+    },
+    {
+      "epoch": 75.58,
+      "grad_norm": 4.398240089416504,
+      "learning_rate": 2.442105263157895e-06,
+      "loss": 0.2157,
+      "step": 50260
+    },
+    {
+      "epoch": 75.59,
+      "grad_norm": 2.5172154903411865,
+      "learning_rate": 2.4406015037593985e-06,
+      "loss": 0.2076,
+      "step": 50270
+    },
+    {
+      "epoch": 75.61,
+      "grad_norm": 5.356011867523193,
+      "learning_rate": 2.4390977443609025e-06,
+      "loss": 0.1441,
+      "step": 50280
+    },
+    {
+      "epoch": 75.62,
+      "grad_norm": 5.028241157531738,
+      "learning_rate": 2.437593984962406e-06,
+      "loss": 0.1921,
+      "step": 50290
+    },
+    {
+      "epoch": 75.64,
+      "grad_norm": 5.889922142028809,
+      "learning_rate": 2.43609022556391e-06,
+      "loss": 0.1315,
+      "step": 50300
+    },
+    {
+      "epoch": 75.65,
+      "grad_norm": 6.243462562561035,
+      "learning_rate": 2.434586466165414e-06,
+      "loss": 0.2106,
+      "step": 50310
+    },
+    {
+      "epoch": 75.67,
+      "grad_norm": 6.075836181640625,
+      "learning_rate": 2.4330827067669174e-06,
+      "loss": 0.2055,
+      "step": 50320
+    },
+    {
+      "epoch": 75.68,
+      "grad_norm": 5.1835150718688965,
+      "learning_rate": 2.4315789473684213e-06,
+      "loss": 0.2298,
+      "step": 50330
+    },
+    {
+      "epoch": 75.7,
+      "grad_norm": 4.660686016082764,
+      "learning_rate": 2.430075187969925e-06,
+      "loss": 0.173,
+      "step": 50340
+    },
+    {
+      "epoch": 75.71,
+      "grad_norm": 3.6795620918273926,
+      "learning_rate": 2.428571428571429e-06,
+      "loss": 0.1852,
+      "step": 50350
+    },
+    {
+      "epoch": 75.73,
+      "grad_norm": 4.784815788269043,
+      "learning_rate": 2.4270676691729323e-06,
+      "loss": 0.1739,
+      "step": 50360
+    },
+    {
+      "epoch": 75.74,
+      "grad_norm": 6.113933563232422,
+      "learning_rate": 2.4255639097744363e-06,
+      "loss": 0.2325,
+      "step": 50370
+    },
+    {
+      "epoch": 75.76,
+      "grad_norm": 3.9142940044403076,
+      "learning_rate": 2.42406015037594e-06,
+      "loss": 0.1929,
+      "step": 50380
+    },
+    {
+      "epoch": 75.77,
+      "grad_norm": 3.883265256881714,
+      "learning_rate": 2.4225563909774437e-06,
+      "loss": 0.2146,
+      "step": 50390
+    },
+    {
+      "epoch": 75.79,
+      "grad_norm": 10.637557029724121,
+      "learning_rate": 2.4210526315789477e-06,
+      "loss": 0.2488,
+      "step": 50400
+    },
+    {
+      "epoch": 75.8,
+      "grad_norm": 3.309053659439087,
+      "learning_rate": 2.419548872180451e-06,
+      "loss": 0.2057,
+      "step": 50410
+    },
+    {
+      "epoch": 75.82,
+      "grad_norm": 7.711801052093506,
+      "learning_rate": 2.418045112781955e-06,
+      "loss": 0.2042,
+      "step": 50420
+    },
+    {
+      "epoch": 75.83,
+      "grad_norm": 3.687312602996826,
+      "learning_rate": 2.416541353383459e-06,
+      "loss": 0.1693,
+      "step": 50430
+    },
+    {
+      "epoch": 75.85,
+      "grad_norm": 5.525257587432861,
+      "learning_rate": 2.4150375939849626e-06,
+      "loss": 0.2097,
+      "step": 50440
+    },
+    {
+      "epoch": 75.86,
+      "grad_norm": 4.86057186126709,
+      "learning_rate": 2.4135338345864665e-06,
+      "loss": 0.1844,
+      "step": 50450
+    },
+    {
+      "epoch": 75.88,
+      "grad_norm": 2.9280784130096436,
+      "learning_rate": 2.41203007518797e-06,
+      "loss": 0.13,
+      "step": 50460
+    },
+    {
+      "epoch": 75.89,
+      "grad_norm": 2.978746175765991,
+      "learning_rate": 2.410526315789474e-06,
+      "loss": 0.1435,
+      "step": 50470
+    },
+    {
+      "epoch": 75.91,
+      "grad_norm": 7.612166404724121,
+      "learning_rate": 2.409022556390978e-06,
+      "loss": 0.1784,
+      "step": 50480
+    },
+    {
+      "epoch": 75.92,
+      "grad_norm": 5.406614303588867,
+      "learning_rate": 2.4075187969924814e-06,
+      "loss": 0.2408,
+      "step": 50490
+    },
+    {
+      "epoch": 75.94,
+      "grad_norm": 7.032416820526123,
+      "learning_rate": 2.406015037593985e-06,
+      "loss": 0.1511,
+      "step": 50500
+    },
+    {
+      "epoch": 75.95,
+      "grad_norm": 6.901172161102295,
+      "learning_rate": 2.404511278195489e-06,
+      "loss": 0.2211,
+      "step": 50510
+    },
+    {
+      "epoch": 75.97,
+      "grad_norm": 5.923864841461182,
+      "learning_rate": 2.4030075187969924e-06,
+      "loss": 0.18,
+      "step": 50520
+    },
+    {
+      "epoch": 75.98,
+      "grad_norm": 7.222829341888428,
+      "learning_rate": 2.4015037593984964e-06,
+      "loss": 0.2284,
+      "step": 50530
+    },
+    {
+      "epoch": 76.0,
+      "grad_norm": 0.01305259671062231,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.1953,
+      "step": 50540
+    },
+    {
+      "epoch": 76.0,
+      "eval_accuracy": 0.9307,
+      "eval_loss": 0.3337118625640869,
+      "eval_runtime": 84.2661,
+      "eval_samples_per_second": 118.672,
+      "eval_steps_per_second": 0.475,
+      "step": 50540
+    },
+    {
+      "epoch": 76.02,
+      "grad_norm": 5.751012802124023,
+      "learning_rate": 2.398496240601504e-06,
+      "loss": 0.1967,
+      "step": 50550
+    },
+    {
+      "epoch": 76.03,
+      "grad_norm": 4.3412089347839355,
+      "learning_rate": 2.3969924812030078e-06,
+      "loss": 0.1634,
+      "step": 50560
+    },
+    {
+      "epoch": 76.05,
+      "grad_norm": 5.5192155838012695,
+      "learning_rate": 2.3954887218045113e-06,
+      "loss": 0.1627,
+      "step": 50570
+    },
+    {
+      "epoch": 76.06,
+      "grad_norm": 12.006901741027832,
+      "learning_rate": 2.3939849624060152e-06,
+      "loss": 0.2048,
+      "step": 50580
+    },
+    {
+      "epoch": 76.08,
+      "grad_norm": 3.821943521499634,
+      "learning_rate": 2.3924812030075188e-06,
+      "loss": 0.1681,
+      "step": 50590
+    },
+    {
+      "epoch": 76.09,
+      "grad_norm": 7.498505115509033,
+      "learning_rate": 2.3909774436090227e-06,
+      "loss": 0.2294,
+      "step": 50600
+    },
+    {
+      "epoch": 76.11,
+      "grad_norm": 1.650476336479187,
+      "learning_rate": 2.3894736842105266e-06,
+      "loss": 0.1929,
+      "step": 50610
+    },
+    {
+      "epoch": 76.12,
+      "grad_norm": 3.873107671737671,
+      "learning_rate": 2.38796992481203e-06,
+      "loss": 0.1953,
+      "step": 50620
+    },
+    {
+      "epoch": 76.14,
+      "grad_norm": 5.848465442657471,
+      "learning_rate": 2.386466165413534e-06,
+      "loss": 0.2391,
+      "step": 50630
+    },
+    {
+      "epoch": 76.15,
+      "grad_norm": 10.972421646118164,
+      "learning_rate": 2.3849624060150376e-06,
+      "loss": 0.1747,
+      "step": 50640
+    },
+    {
+      "epoch": 76.17,
+      "grad_norm": 3.4140889644622803,
+      "learning_rate": 2.3834586466165416e-06,
+      "loss": 0.1541,
+      "step": 50650
+    },
+    {
+      "epoch": 76.18,
+      "grad_norm": 4.576592922210693,
+      "learning_rate": 2.3819548872180455e-06,
+      "loss": 0.2146,
+      "step": 50660
+    },
+    {
+      "epoch": 76.2,
+      "grad_norm": 7.381382465362549,
+      "learning_rate": 2.380451127819549e-06,
+      "loss": 0.1912,
+      "step": 50670
+    },
+    {
+      "epoch": 76.21,
+      "grad_norm": 7.457504749298096,
+      "learning_rate": 2.378947368421053e-06,
+      "loss": 0.1681,
+      "step": 50680
+    },
+    {
+      "epoch": 76.23,
+      "grad_norm": 10.005772590637207,
+      "learning_rate": 2.3774436090225565e-06,
+      "loss": 0.2002,
+      "step": 50690
+    },
+    {
+      "epoch": 76.24,
+      "grad_norm": 8.001717567443848,
+      "learning_rate": 2.3759398496240604e-06,
+      "loss": 0.2091,
+      "step": 50700
+    },
+    {
+      "epoch": 76.26,
+      "grad_norm": 4.301360130310059,
+      "learning_rate": 2.3744360902255644e-06,
+      "loss": 0.1585,
+      "step": 50710
+    },
+    {
+      "epoch": 76.27,
+      "grad_norm": 2.692290782928467,
+      "learning_rate": 2.372932330827068e-06,
+      "loss": 0.1992,
+      "step": 50720
+    },
+    {
+      "epoch": 76.29,
+      "grad_norm": 7.287367820739746,
+      "learning_rate": 2.371428571428572e-06,
+      "loss": 0.2296,
+      "step": 50730
+    },
+    {
+      "epoch": 76.3,
+      "grad_norm": 5.782092094421387,
+      "learning_rate": 2.3699248120300754e-06,
+      "loss": 0.1614,
+      "step": 50740
+    },
+    {
+      "epoch": 76.32,
+      "grad_norm": 1.4290227890014648,
+      "learning_rate": 2.368421052631579e-06,
+      "loss": 0.1392,
+      "step": 50750
+    },
+    {
+      "epoch": 76.33,
+      "grad_norm": 5.241323471069336,
+      "learning_rate": 2.366917293233083e-06,
+      "loss": 0.119,
+      "step": 50760
+    },
+    {
+      "epoch": 76.35,
+      "grad_norm": 5.544551849365234,
+      "learning_rate": 2.3654135338345868e-06,
+      "loss": 0.1617,
+      "step": 50770
+    },
+    {
+      "epoch": 76.36,
+      "grad_norm": 4.271751880645752,
+      "learning_rate": 2.3639097744360903e-06,
+      "loss": 0.1435,
+      "step": 50780
+    },
+    {
+      "epoch": 76.38,
+      "grad_norm": 3.508246660232544,
+      "learning_rate": 2.3624060150375942e-06,
+      "loss": 0.1811,
+      "step": 50790
+    },
+    {
+      "epoch": 76.39,
+      "grad_norm": 6.839121341705322,
+      "learning_rate": 2.3609022556390977e-06,
+      "loss": 0.1943,
+      "step": 50800
+    },
+    {
+      "epoch": 76.41,
+      "grad_norm": 6.446866512298584,
+      "learning_rate": 2.3593984962406017e-06,
+      "loss": 0.2103,
+      "step": 50810
+    },
+    {
+      "epoch": 76.42,
+      "grad_norm": 3.7222402095794678,
+      "learning_rate": 2.357894736842105e-06,
+      "loss": 0.1546,
+      "step": 50820
+    },
+    {
+      "epoch": 76.44,
+      "grad_norm": 9.640228271484375,
+      "learning_rate": 2.356390977443609e-06,
+      "loss": 0.2003,
+      "step": 50830
+    },
+    {
+      "epoch": 76.45,
+      "grad_norm": 5.100734710693359,
+      "learning_rate": 2.354887218045113e-06,
+      "loss": 0.1579,
+      "step": 50840
+    },
+    {
+      "epoch": 76.47,
+      "grad_norm": 2.1371684074401855,
+      "learning_rate": 2.3533834586466166e-06,
+      "loss": 0.1618,
+      "step": 50850
+    },
+    {
+      "epoch": 76.48,
+      "grad_norm": 7.091713905334473,
+      "learning_rate": 2.3518796992481205e-06,
+      "loss": 0.1943,
+      "step": 50860
+    },
+    {
+      "epoch": 76.5,
+      "grad_norm": 5.84473180770874,
+      "learning_rate": 2.350375939849624e-06,
+      "loss": 0.15,
+      "step": 50870
+    },
+    {
+      "epoch": 76.51,
+      "grad_norm": 4.024866104125977,
+      "learning_rate": 2.348872180451128e-06,
+      "loss": 0.166,
+      "step": 50880
+    },
+    {
+      "epoch": 76.53,
+      "grad_norm": 5.06535530090332,
+      "learning_rate": 2.347368421052632e-06,
+      "loss": 0.1813,
+      "step": 50890
+    },
+    {
+      "epoch": 76.54,
+      "grad_norm": 5.515821933746338,
+      "learning_rate": 2.3458646616541355e-06,
+      "loss": 0.1672,
+      "step": 50900
+    },
+    {
+      "epoch": 76.56,
+      "grad_norm": 5.713620185852051,
+      "learning_rate": 2.3443609022556394e-06,
+      "loss": 0.2043,
+      "step": 50910
+    },
+    {
+      "epoch": 76.57,
+      "grad_norm": 4.078904151916504,
+      "learning_rate": 2.342857142857143e-06,
+      "loss": 0.1606,
+      "step": 50920
+    },
+    {
+      "epoch": 76.59,
+      "grad_norm": 5.5044660568237305,
+      "learning_rate": 2.341353383458647e-06,
+      "loss": 0.1721,
+      "step": 50930
+    },
+    {
+      "epoch": 76.6,
+      "grad_norm": 6.557012557983398,
+      "learning_rate": 2.339849624060151e-06,
+      "loss": 0.1614,
+      "step": 50940
+    },
+    {
+      "epoch": 76.62,
+      "grad_norm": 2.52939772605896,
+      "learning_rate": 2.3383458646616543e-06,
+      "loss": 0.1872,
+      "step": 50950
+    },
+    {
+      "epoch": 76.63,
+      "grad_norm": 1.881983757019043,
+      "learning_rate": 2.3368421052631583e-06,
+      "loss": 0.1311,
+      "step": 50960
+    },
+    {
+      "epoch": 76.65,
+      "grad_norm": 3.7978851795196533,
+      "learning_rate": 2.335338345864662e-06,
+      "loss": 0.213,
+      "step": 50970
+    },
+    {
+      "epoch": 76.66,
+      "grad_norm": 6.723447799682617,
+      "learning_rate": 2.3338345864661657e-06,
+      "loss": 0.1595,
+      "step": 50980
+    },
+    {
+      "epoch": 76.68,
+      "grad_norm": 4.094229698181152,
+      "learning_rate": 2.3323308270676697e-06,
+      "loss": 0.1888,
+      "step": 50990
+    },
+    {
+      "epoch": 76.69,
+      "grad_norm": 5.497591495513916,
+      "learning_rate": 2.330827067669173e-06,
+      "loss": 0.1733,
+      "step": 51000
+    },
+    {
+      "epoch": 76.71,
+      "grad_norm": 6.945542812347412,
+      "learning_rate": 2.3293233082706767e-06,
+      "loss": 0.225,
+      "step": 51010
+    },
+    {
+      "epoch": 76.72,
+      "grad_norm": 7.769118309020996,
+      "learning_rate": 2.3278195488721807e-06,
+      "loss": 0.2114,
+      "step": 51020
+    },
+    {
+      "epoch": 76.74,
+      "grad_norm": 5.106688976287842,
+      "learning_rate": 2.326315789473684e-06,
+      "loss": 0.1166,
+      "step": 51030
+    },
+    {
+      "epoch": 76.75,
+      "grad_norm": 7.118248462677002,
+      "learning_rate": 2.324812030075188e-06,
+      "loss": 0.1807,
+      "step": 51040
+    },
+    {
+      "epoch": 76.77,
+      "grad_norm": 4.859402656555176,
+      "learning_rate": 2.3233082706766916e-06,
+      "loss": 0.217,
+      "step": 51050
+    },
+    {
+      "epoch": 76.78,
+      "grad_norm": 4.8244829177856445,
+      "learning_rate": 2.3218045112781956e-06,
+      "loss": 0.1698,
+      "step": 51060
+    },
+    {
+      "epoch": 76.8,
+      "grad_norm": 4.1632914543151855,
+      "learning_rate": 2.3203007518796995e-06,
+      "loss": 0.1531,
+      "step": 51070
+    },
+    {
+      "epoch": 76.81,
+      "grad_norm": 5.575042247772217,
+      "learning_rate": 2.318796992481203e-06,
+      "loss": 0.1539,
+      "step": 51080
+    },
+    {
+      "epoch": 76.83,
+      "grad_norm": 4.370471000671387,
+      "learning_rate": 2.317293233082707e-06,
+      "loss": 0.1313,
+      "step": 51090
+    },
+    {
+      "epoch": 76.84,
+      "grad_norm": 4.721175193786621,
+      "learning_rate": 2.3157894736842105e-06,
+      "loss": 0.1698,
+      "step": 51100
+    },
+    {
+      "epoch": 76.86,
+      "grad_norm": 6.6235857009887695,
+      "learning_rate": 2.3142857142857145e-06,
+      "loss": 0.2313,
+      "step": 51110
+    },
+    {
+      "epoch": 76.87,
+      "grad_norm": 15.119837760925293,
+      "learning_rate": 2.3127819548872184e-06,
+      "loss": 0.163,
+      "step": 51120
+    },
+    {
+      "epoch": 76.89,
+      "grad_norm": 2.5521862506866455,
+      "learning_rate": 2.311278195488722e-06,
+      "loss": 0.2031,
+      "step": 51130
+    },
+    {
+      "epoch": 76.9,
+      "grad_norm": 8.849630355834961,
+      "learning_rate": 2.309774436090226e-06,
+      "loss": 0.182,
+      "step": 51140
+    },
+    {
+      "epoch": 76.92,
+      "grad_norm": 6.023241996765137,
+      "learning_rate": 2.3082706766917294e-06,
+      "loss": 0.1807,
+      "step": 51150
+    },
+    {
+      "epoch": 76.93,
+      "grad_norm": 6.313910007476807,
+      "learning_rate": 2.3067669172932333e-06,
+      "loss": 0.1622,
+      "step": 51160
+    },
+    {
+      "epoch": 76.95,
+      "grad_norm": 7.0026116371154785,
+      "learning_rate": 2.3052631578947373e-06,
+      "loss": 0.1677,
+      "step": 51170
+    },
+    {
+      "epoch": 76.96,
+      "grad_norm": 5.539087772369385,
+      "learning_rate": 2.3037593984962408e-06,
+      "loss": 0.1771,
+      "step": 51180
+    },
+    {
+      "epoch": 76.98,
+      "grad_norm": 2.8735461235046387,
+      "learning_rate": 2.3022556390977447e-06,
+      "loss": 0.1641,
+      "step": 51190
+    },
+    {
+      "epoch": 76.99,
+      "grad_norm": 4.759913921356201,
+      "learning_rate": 2.3007518796992482e-06,
+      "loss": 0.1599,
+      "step": 51200
+    },
+    {
+      "epoch": 77.0,
+      "eval_accuracy": 0.9313,
+      "eval_loss": 0.3295079469680786,
+      "eval_runtime": 84.3369,
+      "eval_samples_per_second": 118.572,
+      "eval_steps_per_second": 0.474,
+      "step": 51205
+    },
+    {
+      "epoch": 77.01,
+      "grad_norm": 3.568769693374634,
+      "learning_rate": 2.299248120300752e-06,
+      "loss": 0.1626,
+      "step": 51210
+    },
+    {
+      "epoch": 77.02,
+      "grad_norm": 6.137712478637695,
+      "learning_rate": 2.297744360902256e-06,
+      "loss": 0.1651,
+      "step": 51220
+    },
+    {
+      "epoch": 77.04,
+      "grad_norm": 2.2381041049957275,
+      "learning_rate": 2.2962406015037596e-06,
+      "loss": 0.1933,
+      "step": 51230
+    },
+    {
+      "epoch": 77.05,
+      "grad_norm": 6.399120330810547,
+      "learning_rate": 2.294736842105263e-06,
+      "loss": 0.1864,
+      "step": 51240
+    },
+    {
+      "epoch": 77.07,
+      "grad_norm": 5.5201334953308105,
+      "learning_rate": 2.293233082706767e-06,
+      "loss": 0.2318,
+      "step": 51250
+    },
+    {
+      "epoch": 77.08,
+      "grad_norm": 6.623318195343018,
+      "learning_rate": 2.2917293233082706e-06,
+      "loss": 0.2449,
+      "step": 51260
+    },
+    {
+      "epoch": 77.1,
+      "grad_norm": 5.644393444061279,
+      "learning_rate": 2.2902255639097746e-06,
+      "loss": 0.2073,
+      "step": 51270
+    },
+    {
+      "epoch": 77.11,
+      "grad_norm": 5.483371257781982,
+      "learning_rate": 2.288721804511278e-06,
+      "loss": 0.185,
+      "step": 51280
+    },
+    {
+      "epoch": 77.13,
+      "grad_norm": 5.581244468688965,
+      "learning_rate": 2.287218045112782e-06,
+      "loss": 0.2171,
+      "step": 51290
+    },
+    {
+      "epoch": 77.14,
+      "grad_norm": 7.645112991333008,
+      "learning_rate": 2.285714285714286e-06,
+      "loss": 0.2034,
+      "step": 51300
+    },
+    {
+      "epoch": 77.16,
+      "grad_norm": 5.600865364074707,
+      "learning_rate": 2.2842105263157895e-06,
+      "loss": 0.2298,
+      "step": 51310
+    },
+    {
+      "epoch": 77.17,
+      "grad_norm": 8.792081832885742,
+      "learning_rate": 2.2827067669172934e-06,
+      "loss": 0.1311,
+      "step": 51320
+    },
+    {
+      "epoch": 77.19,
+      "grad_norm": 4.581193447113037,
+      "learning_rate": 2.281203007518797e-06,
+      "loss": 0.1835,
+      "step": 51330
+    },
+    {
+      "epoch": 77.2,
+      "grad_norm": 3.9883158206939697,
+      "learning_rate": 2.279699248120301e-06,
+      "loss": 0.1439,
+      "step": 51340
+    },
+    {
+      "epoch": 77.22,
+      "grad_norm": 3.542494535446167,
+      "learning_rate": 2.278195488721805e-06,
+      "loss": 0.1538,
+      "step": 51350
+    },
+    {
+      "epoch": 77.23,
+      "grad_norm": 3.4795663356781006,
+      "learning_rate": 2.2766917293233084e-06,
+      "loss": 0.1669,
+      "step": 51360
+    },
+    {
+      "epoch": 77.25,
+      "grad_norm": 5.9982194900512695,
+      "learning_rate": 2.2751879699248123e-06,
+      "loss": 0.1479,
+      "step": 51370
+    },
+    {
+      "epoch": 77.26,
+      "grad_norm": 4.513314247131348,
+      "learning_rate": 2.273684210526316e-06,
+      "loss": 0.2073,
+      "step": 51380
+    },
+    {
+      "epoch": 77.28,
+      "grad_norm": 3.709369421005249,
+      "learning_rate": 2.2721804511278198e-06,
+      "loss": 0.1449,
+      "step": 51390
+    },
+    {
+      "epoch": 77.29,
+      "grad_norm": 4.331890106201172,
+      "learning_rate": 2.2706766917293237e-06,
+      "loss": 0.1677,
+      "step": 51400
+    },
+    {
+      "epoch": 77.31,
+      "grad_norm": 3.835942506790161,
+      "learning_rate": 2.2691729323308272e-06,
+      "loss": 0.1495,
+      "step": 51410
+    },
+    {
+      "epoch": 77.32,
+      "grad_norm": 4.771125316619873,
+      "learning_rate": 2.267669172932331e-06,
+      "loss": 0.1771,
+      "step": 51420
+    },
+    {
+      "epoch": 77.34,
+      "grad_norm": 2.807570695877075,
+      "learning_rate": 2.2661654135338347e-06,
+      "loss": 0.1593,
+      "step": 51430
+    },
+    {
+      "epoch": 77.35,
+      "grad_norm": 2.986487627029419,
+      "learning_rate": 2.2646616541353386e-06,
+      "loss": 0.1418,
+      "step": 51440
+    },
+    {
+      "epoch": 77.37,
+      "grad_norm": 3.4816768169403076,
+      "learning_rate": 2.2631578947368426e-06,
+      "loss": 0.1787,
+      "step": 51450
+    },
+    {
+      "epoch": 77.38,
+      "grad_norm": 4.014631271362305,
+      "learning_rate": 2.261654135338346e-06,
+      "loss": 0.1266,
+      "step": 51460
+    },
+    {
+      "epoch": 77.4,
+      "grad_norm": 6.388309955596924,
+      "learning_rate": 2.26015037593985e-06,
+      "loss": 0.2472,
+      "step": 51470
+    },
+    {
+      "epoch": 77.41,
+      "grad_norm": 3.9746556282043457,
+      "learning_rate": 2.2586466165413536e-06,
+      "loss": 0.2174,
+      "step": 51480
+    },
+    {
+      "epoch": 77.43,
+      "grad_norm": 5.2674455642700195,
+      "learning_rate": 2.257142857142857e-06,
+      "loss": 0.1419,
+      "step": 51490
+    },
+    {
+      "epoch": 77.44,
+      "grad_norm": 4.513287544250488,
+      "learning_rate": 2.255639097744361e-06,
+      "loss": 0.1721,
+      "step": 51500
+    },
+    {
+      "epoch": 77.46,
+      "grad_norm": 7.286464691162109,
+      "learning_rate": 2.2541353383458645e-06,
+      "loss": 0.1816,
+      "step": 51510
+    },
+    {
+      "epoch": 77.47,
+      "grad_norm": 3.867978572845459,
+      "learning_rate": 2.2526315789473685e-06,
+      "loss": 0.1384,
+      "step": 51520
+    },
+    {
+      "epoch": 77.49,
+      "grad_norm": 0.8837634921073914,
+      "learning_rate": 2.2511278195488724e-06,
+      "loss": 0.1681,
+      "step": 51530
+    },
+    {
+      "epoch": 77.5,
+      "grad_norm": 8.97939395904541,
+      "learning_rate": 2.249624060150376e-06,
+      "loss": 0.1419,
+      "step": 51540
+    },
+    {
+      "epoch": 77.52,
+      "grad_norm": 3.708096742630005,
+      "learning_rate": 2.24812030075188e-06,
+      "loss": 0.1476,
+      "step": 51550
+    },
+    {
+      "epoch": 77.53,
+      "grad_norm": 3.5969924926757812,
+      "learning_rate": 2.2466165413533834e-06,
+      "loss": 0.188,
+      "step": 51560
+    },
+    {
+      "epoch": 77.55,
+      "grad_norm": 2.7595226764678955,
+      "learning_rate": 2.2451127819548873e-06,
+      "loss": 0.2098,
+      "step": 51570
+    },
+    {
+      "epoch": 77.56,
+      "grad_norm": 5.377525329589844,
+      "learning_rate": 2.2436090225563913e-06,
+      "loss": 0.2238,
+      "step": 51580
+    },
+    {
+      "epoch": 77.58,
+      "grad_norm": 8.016778945922852,
+      "learning_rate": 2.242105263157895e-06,
+      "loss": 0.1233,
+      "step": 51590
+    },
+    {
+      "epoch": 77.59,
+      "grad_norm": 4.480692386627197,
+      "learning_rate": 2.2406015037593987e-06,
+      "loss": 0.2026,
+      "step": 51600
+    },
+    {
+      "epoch": 77.61,
+      "grad_norm": 6.609382629394531,
+      "learning_rate": 2.2390977443609023e-06,
+      "loss": 0.1878,
+      "step": 51610
+    },
+    {
+      "epoch": 77.62,
+      "grad_norm": 6.064496994018555,
+      "learning_rate": 2.237593984962406e-06,
+      "loss": 0.2227,
+      "step": 51620
+    },
+    {
+      "epoch": 77.64,
+      "grad_norm": 2.538266181945801,
+      "learning_rate": 2.23609022556391e-06,
+      "loss": 0.1967,
+      "step": 51630
+    },
+    {
+      "epoch": 77.65,
+      "grad_norm": 5.404106140136719,
+      "learning_rate": 2.2345864661654137e-06,
+      "loss": 0.254,
+      "step": 51640
+    },
+    {
+      "epoch": 77.67,
+      "grad_norm": 5.271445274353027,
+      "learning_rate": 2.2330827067669176e-06,
+      "loss": 0.132,
+      "step": 51650
+    },
+    {
+      "epoch": 77.68,
+      "grad_norm": 4.448204040527344,
+      "learning_rate": 2.231578947368421e-06,
+      "loss": 0.1674,
+      "step": 51660
+    },
+    {
+      "epoch": 77.7,
+      "grad_norm": 8.482146263122559,
+      "learning_rate": 2.230075187969925e-06,
+      "loss": 0.2233,
+      "step": 51670
+    },
+    {
+      "epoch": 77.71,
+      "grad_norm": 3.992495536804199,
+      "learning_rate": 2.228571428571429e-06,
+      "loss": 0.141,
+      "step": 51680
+    },
+    {
+      "epoch": 77.73,
+      "grad_norm": 9.992817878723145,
+      "learning_rate": 2.2270676691729325e-06,
+      "loss": 0.2356,
+      "step": 51690
+    },
+    {
+      "epoch": 77.74,
+      "grad_norm": 3.759432554244995,
+      "learning_rate": 2.2255639097744365e-06,
+      "loss": 0.1199,
+      "step": 51700
+    },
+    {
+      "epoch": 77.76,
+      "grad_norm": 4.334221363067627,
+      "learning_rate": 2.22406015037594e-06,
+      "loss": 0.1971,
+      "step": 51710
+    },
+    {
+      "epoch": 77.77,
+      "grad_norm": 1.510976791381836,
+      "learning_rate": 2.222556390977444e-06,
+      "loss": 0.1708,
+      "step": 51720
+    },
+    {
+      "epoch": 77.79,
+      "grad_norm": 5.777114391326904,
+      "learning_rate": 2.221052631578948e-06,
+      "loss": 0.2107,
+      "step": 51730
+    },
+    {
+      "epoch": 77.8,
+      "grad_norm": 2.896449327468872,
+      "learning_rate": 2.219548872180451e-06,
+      "loss": 0.1942,
+      "step": 51740
+    },
+    {
+      "epoch": 77.82,
+      "grad_norm": 17.983856201171875,
+      "learning_rate": 2.218045112781955e-06,
+      "loss": 0.1846,
+      "step": 51750
+    },
+    {
+      "epoch": 77.83,
+      "grad_norm": 7.573366165161133,
+      "learning_rate": 2.216541353383459e-06,
+      "loss": 0.1972,
+      "step": 51760
+    },
+    {
+      "epoch": 77.85,
+      "grad_norm": 6.8880109786987305,
+      "learning_rate": 2.2150375939849624e-06,
+      "loss": 0.18,
+      "step": 51770
+    },
+    {
+      "epoch": 77.86,
+      "grad_norm": 2.3339192867279053,
+      "learning_rate": 2.2135338345864663e-06,
+      "loss": 0.1842,
+      "step": 51780
+    },
+    {
+      "epoch": 77.88,
+      "grad_norm": 3.973484754562378,
+      "learning_rate": 2.21203007518797e-06,
+      "loss": 0.2218,
+      "step": 51790
+    },
+    {
+      "epoch": 77.89,
+      "grad_norm": 3.132448673248291,
+      "learning_rate": 2.2105263157894738e-06,
+      "loss": 0.2052,
+      "step": 51800
+    },
+    {
+      "epoch": 77.91,
+      "grad_norm": 5.713258743286133,
+      "learning_rate": 2.2090225563909777e-06,
+      "loss": 0.1334,
+      "step": 51810
+    },
+    {
+      "epoch": 77.92,
+      "grad_norm": 4.339270114898682,
+      "learning_rate": 2.2075187969924812e-06,
+      "loss": 0.2167,
+      "step": 51820
+    },
+    {
+      "epoch": 77.94,
+      "grad_norm": 1.3507264852523804,
+      "learning_rate": 2.206015037593985e-06,
+      "loss": 0.2518,
+      "step": 51830
+    },
+    {
+      "epoch": 77.95,
+      "grad_norm": 7.650753974914551,
+      "learning_rate": 2.2045112781954887e-06,
+      "loss": 0.2293,
+      "step": 51840
+    },
+    {
+      "epoch": 77.97,
+      "grad_norm": 5.412219524383545,
+      "learning_rate": 2.2030075187969927e-06,
+      "loss": 0.1966,
+      "step": 51850
+    },
+    {
+      "epoch": 77.98,
+      "grad_norm": 7.676565647125244,
+      "learning_rate": 2.2015037593984966e-06,
+      "loss": 0.2069,
+      "step": 51860
+    },
+    {
+      "epoch": 78.0,
+      "grad_norm": 0.12734095752239227,
+      "learning_rate": 2.2e-06,
+      "loss": 0.2077,
+      "step": 51870
+    },
+    {
+      "epoch": 78.0,
+      "eval_accuracy": 0.9312,
+      "eval_loss": 0.32849153876304626,
+      "eval_runtime": 85.6428,
+      "eval_samples_per_second": 116.764,
+      "eval_steps_per_second": 0.467,
+      "step": 51870
+    },
+    {
+      "epoch": 78.02,
+      "grad_norm": 3.1705241203308105,
+      "learning_rate": 2.198496240601504e-06,
+      "loss": 0.1415,
+      "step": 51880
+    },
+    {
+      "epoch": 78.03,
+      "grad_norm": 4.243330955505371,
+      "learning_rate": 2.1969924812030076e-06,
+      "loss": 0.1765,
+      "step": 51890
+    },
+    {
+      "epoch": 78.05,
+      "grad_norm": 7.999730110168457,
+      "learning_rate": 2.1954887218045115e-06,
+      "loss": 0.2041,
+      "step": 51900
+    },
+    {
+      "epoch": 78.06,
+      "grad_norm": 4.726391792297363,
+      "learning_rate": 2.1939849624060155e-06,
+      "loss": 0.2337,
+      "step": 51910
+    },
+    {
+      "epoch": 78.08,
+      "grad_norm": 4.67405891418457,
+      "learning_rate": 2.192481203007519e-06,
+      "loss": 0.1231,
+      "step": 51920
+    },
+    {
+      "epoch": 78.09,
+      "grad_norm": 6.686806678771973,
+      "learning_rate": 2.190977443609023e-06,
+      "loss": 0.1047,
+      "step": 51930
+    },
+    {
+      "epoch": 78.11,
+      "grad_norm": 2.4821736812591553,
+      "learning_rate": 2.1894736842105264e-06,
+      "loss": 0.1818,
+      "step": 51940
+    },
+    {
+      "epoch": 78.12,
+      "grad_norm": 5.056117057800293,
+      "learning_rate": 2.1879699248120304e-06,
+      "loss": 0.1644,
+      "step": 51950
+    },
+    {
+      "epoch": 78.14,
+      "grad_norm": 5.994955062866211,
+      "learning_rate": 2.1864661654135343e-06,
+      "loss": 0.1509,
+      "step": 51960
+    },
+    {
+      "epoch": 78.15,
+      "grad_norm": 4.456724643707275,
+      "learning_rate": 2.184962406015038e-06,
+      "loss": 0.1468,
+      "step": 51970
+    },
+    {
+      "epoch": 78.17,
+      "grad_norm": 3.952169895172119,
+      "learning_rate": 2.1834586466165418e-06,
+      "loss": 0.1785,
+      "step": 51980
+    },
+    {
+      "epoch": 78.18,
+      "grad_norm": 3.9694976806640625,
+      "learning_rate": 2.1819548872180453e-06,
+      "loss": 0.2603,
+      "step": 51990
+    },
+    {
+      "epoch": 78.2,
+      "grad_norm": 4.302563190460205,
+      "learning_rate": 2.180451127819549e-06,
+      "loss": 0.1793,
+      "step": 52000
+    },
+    {
+      "epoch": 78.21,
+      "grad_norm": 1.5013707876205444,
+      "learning_rate": 2.1789473684210528e-06,
+      "loss": 0.2053,
+      "step": 52010
+    },
+    {
+      "epoch": 78.23,
+      "grad_norm": 4.439993858337402,
+      "learning_rate": 2.1774436090225563e-06,
+      "loss": 0.1417,
+      "step": 52020
+    },
+    {
+      "epoch": 78.24,
+      "grad_norm": 4.53973388671875,
+      "learning_rate": 2.1759398496240602e-06,
+      "loss": 0.1573,
+      "step": 52030
+    },
+    {
+      "epoch": 78.26,
+      "grad_norm": 6.344988822937012,
+      "learning_rate": 2.174436090225564e-06,
+      "loss": 0.1905,
+      "step": 52040
+    },
+    {
+      "epoch": 78.27,
+      "grad_norm": 3.4215738773345947,
+      "learning_rate": 2.1729323308270677e-06,
+      "loss": 0.1587,
+      "step": 52050
+    },
+    {
+      "epoch": 78.29,
+      "grad_norm": 5.529082298278809,
+      "learning_rate": 2.1714285714285716e-06,
+      "loss": 0.1935,
+      "step": 52060
+    },
+    {
+      "epoch": 78.3,
+      "grad_norm": 2.7385475635528564,
+      "learning_rate": 2.169924812030075e-06,
+      "loss": 0.1677,
+      "step": 52070
+    },
+    {
+      "epoch": 78.32,
+      "grad_norm": 2.511803150177002,
+      "learning_rate": 2.168421052631579e-06,
+      "loss": 0.2112,
+      "step": 52080
+    },
+    {
+      "epoch": 78.33,
+      "grad_norm": 4.103121757507324,
+      "learning_rate": 2.166917293233083e-06,
+      "loss": 0.1629,
+      "step": 52090
+    },
+    {
+      "epoch": 78.35,
+      "grad_norm": 4.167917728424072,
+      "learning_rate": 2.1654135338345866e-06,
+      "loss": 0.1722,
+      "step": 52100
+    },
+    {
+      "epoch": 78.36,
+      "grad_norm": 3.2801246643066406,
+      "learning_rate": 2.1639097744360905e-06,
+      "loss": 0.2039,
+      "step": 52110
+    },
+    {
+      "epoch": 78.38,
+      "grad_norm": 4.19163703918457,
+      "learning_rate": 2.162406015037594e-06,
+      "loss": 0.181,
+      "step": 52120
+    },
+    {
+      "epoch": 78.39,
+      "grad_norm": 3.487081527709961,
+      "learning_rate": 2.160902255639098e-06,
+      "loss": 0.1752,
+      "step": 52130
+    },
+    {
+      "epoch": 78.41,
+      "grad_norm": 2.9936952590942383,
+      "learning_rate": 2.159398496240602e-06,
+      "loss": 0.2173,
+      "step": 52140
+    },
+    {
+      "epoch": 78.42,
+      "grad_norm": 4.499261379241943,
+      "learning_rate": 2.1578947368421054e-06,
+      "loss": 0.1907,
+      "step": 52150
+    },
+    {
+      "epoch": 78.44,
+      "grad_norm": 5.857420921325684,
+      "learning_rate": 2.1563909774436094e-06,
+      "loss": 0.1689,
+      "step": 52160
+    },
+    {
+      "epoch": 78.45,
+      "grad_norm": 4.8537468910217285,
+      "learning_rate": 2.154887218045113e-06,
+      "loss": 0.1216,
+      "step": 52170
+    },
+    {
+      "epoch": 78.47,
+      "grad_norm": 6.32722806930542,
+      "learning_rate": 2.153383458646617e-06,
+      "loss": 0.2262,
+      "step": 52180
+    },
+    {
+      "epoch": 78.48,
+      "grad_norm": 11.071094512939453,
+      "learning_rate": 2.1518796992481208e-06,
+      "loss": 0.1113,
+      "step": 52190
+    },
+    {
+      "epoch": 78.5,
+      "grad_norm": 5.603166580200195,
+      "learning_rate": 2.1503759398496243e-06,
+      "loss": 0.167,
+      "step": 52200
+    },
+    {
+      "epoch": 78.51,
+      "grad_norm": 5.128636837005615,
+      "learning_rate": 2.1488721804511282e-06,
+      "loss": 0.1974,
+      "step": 52210
+    },
+    {
+      "epoch": 78.53,
+      "grad_norm": 2.067934036254883,
+      "learning_rate": 2.1473684210526317e-06,
+      "loss": 0.1654,
+      "step": 52220
+    },
+    {
+      "epoch": 78.54,
+      "grad_norm": 4.146607398986816,
+      "learning_rate": 2.1458646616541357e-06,
+      "loss": 0.1807,
+      "step": 52230
+    },
+    {
+      "epoch": 78.56,
+      "grad_norm": 2.6241300106048584,
+      "learning_rate": 2.144360902255639e-06,
+      "loss": 0.1445,
+      "step": 52240
+    },
+    {
+      "epoch": 78.57,
+      "grad_norm": 4.428233623504639,
+      "learning_rate": 2.1428571428571427e-06,
+      "loss": 0.1427,
+      "step": 52250
+    },
+    {
+      "epoch": 78.59,
+      "grad_norm": 4.425225734710693,
+      "learning_rate": 2.1413533834586467e-06,
+      "loss": 0.1506,
+      "step": 52260
+    },
+    {
+      "epoch": 78.6,
+      "grad_norm": 2.6560704708099365,
+      "learning_rate": 2.1398496240601506e-06,
+      "loss": 0.1788,
+      "step": 52270
+    },
+    {
+      "epoch": 78.62,
+      "grad_norm": 3.350560188293457,
+      "learning_rate": 2.138345864661654e-06,
+      "loss": 0.1513,
+      "step": 52280
+    },
+    {
+      "epoch": 78.63,
+      "grad_norm": 3.5324301719665527,
+      "learning_rate": 2.136842105263158e-06,
+      "loss": 0.1701,
+      "step": 52290
+    },
+    {
+      "epoch": 78.65,
+      "grad_norm": 4.827585697174072,
+      "learning_rate": 2.1353383458646616e-06,
+      "loss": 0.2025,
+      "step": 52300
+    },
+    {
+      "epoch": 78.66,
+      "grad_norm": 0.8045023679733276,
+      "learning_rate": 2.1338345864661655e-06,
+      "loss": 0.1426,
+      "step": 52310
+    },
+    {
+      "epoch": 78.68,
+      "grad_norm": 4.572910308837891,
+      "learning_rate": 2.1323308270676695e-06,
+      "loss": 0.1788,
+      "step": 52320
+    },
+    {
+      "epoch": 78.69,
+      "grad_norm": 5.141964435577393,
+      "learning_rate": 2.130827067669173e-06,
+      "loss": 0.191,
+      "step": 52330
+    },
+    {
+      "epoch": 78.71,
+      "grad_norm": 16.9417667388916,
+      "learning_rate": 2.129323308270677e-06,
+      "loss": 0.1658,
+      "step": 52340
+    },
+    {
+      "epoch": 78.72,
+      "grad_norm": 7.1351399421691895,
+      "learning_rate": 2.1278195488721805e-06,
+      "loss": 0.1802,
+      "step": 52350
+    },
+    {
+      "epoch": 78.74,
+      "grad_norm": 5.015010833740234,
+      "learning_rate": 2.1263157894736844e-06,
+      "loss": 0.227,
+      "step": 52360
+    },
+    {
+      "epoch": 78.75,
+      "grad_norm": 4.296870231628418,
+      "learning_rate": 2.1248120300751883e-06,
+      "loss": 0.1983,
+      "step": 52370
+    },
+    {
+      "epoch": 78.77,
+      "grad_norm": 6.891214370727539,
+      "learning_rate": 2.123308270676692e-06,
+      "loss": 0.2017,
+      "step": 52380
+    },
+    {
+      "epoch": 78.78,
+      "grad_norm": 4.369607925415039,
+      "learning_rate": 2.121804511278196e-06,
+      "loss": 0.1861,
+      "step": 52390
+    },
+    {
+      "epoch": 78.8,
+      "grad_norm": 7.993550777435303,
+      "learning_rate": 2.1203007518796993e-06,
+      "loss": 0.1986,
+      "step": 52400
+    },
+    {
+      "epoch": 78.81,
+      "grad_norm": 5.03591251373291,
+      "learning_rate": 2.1187969924812033e-06,
+      "loss": 0.1679,
+      "step": 52410
+    },
+    {
+      "epoch": 78.83,
+      "grad_norm": 3.6212151050567627,
+      "learning_rate": 2.1172932330827072e-06,
+      "loss": 0.2262,
+      "step": 52420
+    },
+    {
+      "epoch": 78.84,
+      "grad_norm": 6.3097310066223145,
+      "learning_rate": 2.1157894736842107e-06,
+      "loss": 0.2434,
+      "step": 52430
+    },
+    {
+      "epoch": 78.86,
+      "grad_norm": 8.02677059173584,
+      "learning_rate": 2.1142857142857147e-06,
+      "loss": 0.1698,
+      "step": 52440
+    },
+    {
+      "epoch": 78.87,
+      "grad_norm": 6.799673080444336,
+      "learning_rate": 2.112781954887218e-06,
+      "loss": 0.2062,
+      "step": 52450
+    },
+    {
+      "epoch": 78.89,
+      "grad_norm": 3.5356316566467285,
+      "learning_rate": 2.111278195488722e-06,
+      "loss": 0.1969,
+      "step": 52460
+    },
+    {
+      "epoch": 78.9,
+      "grad_norm": 4.078927993774414,
+      "learning_rate": 2.1097744360902257e-06,
+      "loss": 0.1694,
+      "step": 52470
+    },
+    {
+      "epoch": 78.92,
+      "grad_norm": 4.125448226928711,
+      "learning_rate": 2.1082706766917296e-06,
+      "loss": 0.1752,
+      "step": 52480
+    },
+    {
+      "epoch": 78.93,
+      "grad_norm": 5.614774703979492,
+      "learning_rate": 2.106766917293233e-06,
+      "loss": 0.1573,
+      "step": 52490
+    },
+    {
+      "epoch": 78.95,
+      "grad_norm": 3.6688623428344727,
+      "learning_rate": 2.105263157894737e-06,
+      "loss": 0.1792,
+      "step": 52500
+    },
+    {
+      "epoch": 78.96,
+      "grad_norm": 3.9358396530151367,
+      "learning_rate": 2.1037593984962406e-06,
+      "loss": 0.1619,
+      "step": 52510
+    },
+    {
+      "epoch": 78.98,
+      "grad_norm": 3.6908819675445557,
+      "learning_rate": 2.1022556390977445e-06,
+      "loss": 0.1755,
+      "step": 52520
+    },
+    {
+      "epoch": 78.99,
+      "grad_norm": 6.504554271697998,
+      "learning_rate": 2.100751879699248e-06,
+      "loss": 0.2053,
+      "step": 52530
+    },
+    {
+      "epoch": 79.0,
+      "eval_accuracy": 0.9309,
+      "eval_loss": 0.32775041460990906,
+      "eval_runtime": 84.345,
+      "eval_samples_per_second": 118.561,
+      "eval_steps_per_second": 0.474,
+      "step": 52535
+    },
+    {
+      "epoch": 79.01,
+      "grad_norm": 4.1648664474487305,
+      "learning_rate": 2.099248120300752e-06,
+      "loss": 0.2267,
+      "step": 52540
+    },
+    {
+      "epoch": 79.02,
+      "grad_norm": 7.5626020431518555,
+      "learning_rate": 2.097744360902256e-06,
+      "loss": 0.2043,
+      "step": 52550
+    },
+    {
+      "epoch": 79.04,
+      "grad_norm": 4.533354759216309,
+      "learning_rate": 2.0962406015037594e-06,
+      "loss": 0.1998,
+      "step": 52560
+    },
+    {
+      "epoch": 79.05,
+      "grad_norm": 2.196251630783081,
+      "learning_rate": 2.0947368421052634e-06,
+      "loss": 0.1536,
+      "step": 52570
+    },
+    {
+      "epoch": 79.07,
+      "grad_norm": 8.702381134033203,
+      "learning_rate": 2.093233082706767e-06,
+      "loss": 0.1829,
+      "step": 52580
+    },
+    {
+      "epoch": 79.08,
+      "grad_norm": 6.189172267913818,
+      "learning_rate": 2.091729323308271e-06,
+      "loss": 0.2646,
+      "step": 52590
+    },
+    {
+      "epoch": 79.1,
+      "grad_norm": 14.350298881530762,
+      "learning_rate": 2.090225563909775e-06,
+      "loss": 0.1486,
+      "step": 52600
+    },
+    {
+      "epoch": 79.11,
+      "grad_norm": 5.265256404876709,
+      "learning_rate": 2.0887218045112783e-06,
+      "loss": 0.1852,
+      "step": 52610
+    },
+    {
+      "epoch": 79.13,
+      "grad_norm": 5.7867350578308105,
+      "learning_rate": 2.0872180451127823e-06,
+      "loss": 0.1778,
+      "step": 52620
+    },
+    {
+      "epoch": 79.14,
+      "grad_norm": 3.8867268562316895,
+      "learning_rate": 2.0857142857142858e-06,
+      "loss": 0.1716,
+      "step": 52630
+    },
+    {
+      "epoch": 79.16,
+      "grad_norm": 10.908556938171387,
+      "learning_rate": 2.0842105263157897e-06,
+      "loss": 0.2087,
+      "step": 52640
+    },
+    {
+      "epoch": 79.17,
+      "grad_norm": 8.96177864074707,
+      "learning_rate": 2.0827067669172937e-06,
+      "loss": 0.1587,
+      "step": 52650
+    },
+    {
+      "epoch": 79.19,
+      "grad_norm": 4.244720935821533,
+      "learning_rate": 2.081203007518797e-06,
+      "loss": 0.1488,
+      "step": 52660
+    },
+    {
+      "epoch": 79.2,
+      "grad_norm": 3.8978004455566406,
+      "learning_rate": 2.079699248120301e-06,
+      "loss": 0.1361,
+      "step": 52670
+    },
+    {
+      "epoch": 79.22,
+      "grad_norm": 1.49534010887146,
+      "learning_rate": 2.0781954887218046e-06,
+      "loss": 0.1506,
+      "step": 52680
+    },
+    {
+      "epoch": 79.23,
+      "grad_norm": 6.862305164337158,
+      "learning_rate": 2.0766917293233086e-06,
+      "loss": 0.2312,
+      "step": 52690
+    },
+    {
+      "epoch": 79.25,
+      "grad_norm": 5.542629718780518,
+      "learning_rate": 2.075187969924812e-06,
+      "loss": 0.1923,
+      "step": 52700
+    },
+    {
+      "epoch": 79.26,
+      "grad_norm": 5.134133338928223,
+      "learning_rate": 2.073684210526316e-06,
+      "loss": 0.1765,
+      "step": 52710
+    },
+    {
+      "epoch": 79.28,
+      "grad_norm": 4.834011554718018,
+      "learning_rate": 2.07218045112782e-06,
+      "loss": 0.1291,
+      "step": 52720
+    },
+    {
+      "epoch": 79.29,
+      "grad_norm": 9.411290168762207,
+      "learning_rate": 2.0706766917293235e-06,
+      "loss": 0.1661,
+      "step": 52730
+    },
+    {
+      "epoch": 79.31,
+      "grad_norm": 4.970304489135742,
+      "learning_rate": 2.069172932330827e-06,
+      "loss": 0.1876,
+      "step": 52740
+    },
+    {
+      "epoch": 79.32,
+      "grad_norm": 4.562291622161865,
+      "learning_rate": 2.067669172932331e-06,
+      "loss": 0.1942,
+      "step": 52750
+    },
+    {
+      "epoch": 79.34,
+      "grad_norm": 5.195777416229248,
+      "learning_rate": 2.0661654135338345e-06,
+      "loss": 0.1821,
+      "step": 52760
+    },
+    {
+      "epoch": 79.35,
+      "grad_norm": 6.937760829925537,
+      "learning_rate": 2.0646616541353384e-06,
+      "loss": 0.2023,
+      "step": 52770
+    },
+    {
+      "epoch": 79.37,
+      "grad_norm": 6.619980812072754,
+      "learning_rate": 2.0631578947368424e-06,
+      "loss": 0.1777,
+      "step": 52780
+    },
+    {
+      "epoch": 79.38,
+      "grad_norm": 2.9072442054748535,
+      "learning_rate": 2.061654135338346e-06,
+      "loss": 0.1435,
+      "step": 52790
+    },
+    {
+      "epoch": 79.4,
+      "grad_norm": 1.8150124549865723,
+      "learning_rate": 2.06015037593985e-06,
+      "loss": 0.1548,
+      "step": 52800
+    },
+    {
+      "epoch": 79.41,
+      "grad_norm": 4.043828010559082,
+      "learning_rate": 2.0586466165413533e-06,
+      "loss": 0.1577,
+      "step": 52810
+    },
+    {
+      "epoch": 79.43,
+      "grad_norm": 4.577513694763184,
+      "learning_rate": 2.0571428571428573e-06,
+      "loss": 0.2405,
+      "step": 52820
+    },
+    {
+      "epoch": 79.44,
+      "grad_norm": 4.094361782073975,
+      "learning_rate": 2.0556390977443612e-06,
+      "loss": 0.1409,
+      "step": 52830
+    },
+    {
+      "epoch": 79.46,
+      "grad_norm": 3.072739362716675,
+      "learning_rate": 2.0541353383458648e-06,
+      "loss": 0.1992,
+      "step": 52840
+    },
+    {
+      "epoch": 79.47,
+      "grad_norm": 5.828579425811768,
+      "learning_rate": 2.0526315789473687e-06,
+      "loss": 0.1278,
+      "step": 52850
+    },
+    {
+      "epoch": 79.49,
+      "grad_norm": 5.253977298736572,
+      "learning_rate": 2.0511278195488722e-06,
+      "loss": 0.1984,
+      "step": 52860
+    },
+    {
+      "epoch": 79.5,
+      "grad_norm": 3.553118944168091,
+      "learning_rate": 2.049624060150376e-06,
+      "loss": 0.1384,
+      "step": 52870
+    },
+    {
+      "epoch": 79.52,
+      "grad_norm": 7.722125053405762,
+      "learning_rate": 2.04812030075188e-06,
+      "loss": 0.2084,
+      "step": 52880
+    },
+    {
+      "epoch": 79.53,
+      "grad_norm": 3.237384557723999,
+      "learning_rate": 2.0466165413533836e-06,
+      "loss": 0.163,
+      "step": 52890
+    },
+    {
+      "epoch": 79.55,
+      "grad_norm": 11.586670875549316,
+      "learning_rate": 2.0451127819548876e-06,
+      "loss": 0.2027,
+      "step": 52900
+    },
+    {
+      "epoch": 79.56,
+      "grad_norm": 5.43867301940918,
+      "learning_rate": 2.043609022556391e-06,
+      "loss": 0.1672,
+      "step": 52910
+    },
+    {
+      "epoch": 79.58,
+      "grad_norm": 6.87939977645874,
+      "learning_rate": 2.042105263157895e-06,
+      "loss": 0.1675,
+      "step": 52920
+    },
+    {
+      "epoch": 79.59,
+      "grad_norm": 7.075170993804932,
+      "learning_rate": 2.0406015037593985e-06,
+      "loss": 0.2382,
+      "step": 52930
+    },
+    {
+      "epoch": 79.61,
+      "grad_norm": 4.605086326599121,
+      "learning_rate": 2.0390977443609025e-06,
+      "loss": 0.1238,
+      "step": 52940
+    },
+    {
+      "epoch": 79.62,
+      "grad_norm": 4.68049955368042,
+      "learning_rate": 2.0375939849624064e-06,
+      "loss": 0.1679,
+      "step": 52950
+    },
+    {
+      "epoch": 79.64,
+      "grad_norm": 4.511661529541016,
+      "learning_rate": 2.03609022556391e-06,
+      "loss": 0.1604,
+      "step": 52960
+    },
+    {
+      "epoch": 79.65,
+      "grad_norm": 5.428410530090332,
+      "learning_rate": 2.034586466165414e-06,
+      "loss": 0.1244,
+      "step": 52970
+    },
+    {
+      "epoch": 79.67,
+      "grad_norm": 4.728567600250244,
+      "learning_rate": 2.0330827067669174e-06,
+      "loss": 0.1375,
+      "step": 52980
+    },
+    {
+      "epoch": 79.68,
+      "grad_norm": 4.785933017730713,
+      "learning_rate": 2.031578947368421e-06,
+      "loss": 0.1517,
+      "step": 52990
+    },
+    {
+      "epoch": 79.7,
+      "grad_norm": 6.504793167114258,
+      "learning_rate": 2.030075187969925e-06,
+      "loss": 0.1692,
+      "step": 53000
+    },
+    {
+      "epoch": 79.71,
+      "grad_norm": 4.675958156585693,
+      "learning_rate": 2.028571428571429e-06,
+      "loss": 0.1915,
+      "step": 53010
+    },
+    {
+      "epoch": 79.73,
+      "grad_norm": 5.776334285736084,
+      "learning_rate": 2.0270676691729323e-06,
+      "loss": 0.1526,
+      "step": 53020
+    },
+    {
+      "epoch": 79.74,
+      "grad_norm": 11.183079719543457,
+      "learning_rate": 2.0255639097744363e-06,
+      "loss": 0.1586,
+      "step": 53030
+    },
+    {
+      "epoch": 79.76,
+      "grad_norm": 8.511357307434082,
+      "learning_rate": 2.02406015037594e-06,
+      "loss": 0.1941,
+      "step": 53040
+    },
+    {
+      "epoch": 79.77,
+      "grad_norm": 22.351648330688477,
+      "learning_rate": 2.0225563909774437e-06,
+      "loss": 0.1433,
+      "step": 53050
+    },
+    {
+      "epoch": 79.79,
+      "grad_norm": 7.769514083862305,
+      "learning_rate": 2.0210526315789477e-06,
+      "loss": 0.1788,
+      "step": 53060
+    },
+    {
+      "epoch": 79.8,
+      "grad_norm": 5.4393415451049805,
+      "learning_rate": 2.019548872180451e-06,
+      "loss": 0.2059,
+      "step": 53070
+    },
+    {
+      "epoch": 79.82,
+      "grad_norm": 5.8692145347595215,
+      "learning_rate": 2.018045112781955e-06,
+      "loss": 0.2085,
+      "step": 53080
+    },
+    {
+      "epoch": 79.83,
+      "grad_norm": 7.843011379241943,
+      "learning_rate": 2.0165413533834587e-06,
+      "loss": 0.2123,
+      "step": 53090
+    },
+    {
+      "epoch": 79.85,
+      "grad_norm": 3.590719223022461,
+      "learning_rate": 2.0150375939849626e-06,
+      "loss": 0.1455,
+      "step": 53100
+    },
+    {
+      "epoch": 79.86,
+      "grad_norm": 2.503929853439331,
+      "learning_rate": 2.0135338345864665e-06,
+      "loss": 0.1364,
+      "step": 53110
+    },
+    {
+      "epoch": 79.88,
+      "grad_norm": 4.913735866546631,
+      "learning_rate": 2.01203007518797e-06,
+      "loss": 0.159,
+      "step": 53120
+    },
+    {
+      "epoch": 79.89,
+      "grad_norm": 7.748683929443359,
+      "learning_rate": 2.010526315789474e-06,
+      "loss": 0.1912,
+      "step": 53130
+    },
+    {
+      "epoch": 79.91,
+      "grad_norm": 5.099629878997803,
+      "learning_rate": 2.0090225563909775e-06,
+      "loss": 0.2093,
+      "step": 53140
+    },
+    {
+      "epoch": 79.92,
+      "grad_norm": 8.263333320617676,
+      "learning_rate": 2.0075187969924815e-06,
+      "loss": 0.1407,
+      "step": 53150
+    },
+    {
+      "epoch": 79.94,
+      "grad_norm": 9.621796607971191,
+      "learning_rate": 2.006015037593985e-06,
+      "loss": 0.1398,
+      "step": 53160
+    },
+    {
+      "epoch": 79.95,
+      "grad_norm": 4.470506191253662,
+      "learning_rate": 2.004511278195489e-06,
+      "loss": 0.1811,
+      "step": 53170
+    },
+    {
+      "epoch": 79.97,
+      "grad_norm": 1.4590644836425781,
+      "learning_rate": 2.003007518796993e-06,
+      "loss": 0.2257,
+      "step": 53180
+    },
+    {
+      "epoch": 79.98,
+      "grad_norm": 5.552732944488525,
+      "learning_rate": 2.0015037593984964e-06,
+      "loss": 0.2092,
+      "step": 53190
+    },
+    {
+      "epoch": 80.0,
+      "grad_norm": 0.11614461988210678,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.1846,
+      "step": 53200
+    },
+    {
+      "epoch": 80.0,
+      "eval_accuracy": 0.9307,
+      "eval_loss": 0.3290669918060303,
+      "eval_runtime": 85.0743,
+      "eval_samples_per_second": 117.544,
+      "eval_steps_per_second": 0.47,
+      "step": 53200
+    },
+    {
+      "epoch": 80.02,
+      "grad_norm": 7.221068859100342,
+      "learning_rate": 1.998496240601504e-06,
+      "loss": 0.2097,
+      "step": 53210
+    },
+    {
+      "epoch": 80.03,
+      "grad_norm": 8.729702949523926,
+      "learning_rate": 1.996992481203008e-06,
+      "loss": 0.1546,
+      "step": 53220
+    },
+    {
+      "epoch": 80.05,
+      "grad_norm": 6.764802932739258,
+      "learning_rate": 1.9954887218045113e-06,
+      "loss": 0.1736,
+      "step": 53230
+    },
+    {
+      "epoch": 80.06,
+      "grad_norm": 6.79409646987915,
+      "learning_rate": 1.9939849624060153e-06,
+      "loss": 0.2107,
+      "step": 53240
+    },
+    {
+      "epoch": 80.08,
+      "grad_norm": 6.066366672515869,
+      "learning_rate": 1.9924812030075188e-06,
+      "loss": 0.2023,
+      "step": 53250
+    },
+    {
+      "epoch": 80.09,
+      "grad_norm": 6.539670944213867,
+      "learning_rate": 1.9909774436090227e-06,
+      "loss": 0.1947,
+      "step": 53260
+    },
+    {
+      "epoch": 80.11,
+      "grad_norm": 4.000879287719727,
+      "learning_rate": 1.9894736842105262e-06,
+      "loss": 0.1613,
+      "step": 53270
+    },
+    {
+      "epoch": 80.12,
+      "grad_norm": 6.228635787963867,
+      "learning_rate": 1.98796992481203e-06,
+      "loss": 0.1893,
+      "step": 53280
+    },
+    {
+      "epoch": 80.14,
+      "grad_norm": 3.983610153198242,
+      "learning_rate": 1.986466165413534e-06,
+      "loss": 0.176,
+      "step": 53290
+    },
+    {
+      "epoch": 80.15,
+      "grad_norm": 4.762825965881348,
+      "learning_rate": 1.9849624060150376e-06,
+      "loss": 0.2192,
+      "step": 53300
+    },
+    {
+      "epoch": 80.17,
+      "grad_norm": 5.903609275817871,
+      "learning_rate": 1.9834586466165416e-06,
+      "loss": 0.169,
+      "step": 53310
+    },
+    {
+      "epoch": 80.18,
+      "grad_norm": 5.4333176612854,
+      "learning_rate": 1.981954887218045e-06,
+      "loss": 0.1943,
+      "step": 53320
+    },
+    {
+      "epoch": 80.2,
+      "grad_norm": 6.44717264175415,
+      "learning_rate": 1.980451127819549e-06,
+      "loss": 0.2043,
+      "step": 53330
+    },
+    {
+      "epoch": 80.21,
+      "grad_norm": 8.601734161376953,
+      "learning_rate": 1.978947368421053e-06,
+      "loss": 0.1577,
+      "step": 53340
+    },
+    {
+      "epoch": 80.23,
+      "grad_norm": 3.8428754806518555,
+      "learning_rate": 1.9774436090225565e-06,
+      "loss": 0.2007,
+      "step": 53350
+    },
+    {
+      "epoch": 80.24,
+      "grad_norm": 6.386511325836182,
+      "learning_rate": 1.9759398496240604e-06,
+      "loss": 0.2227,
+      "step": 53360
+    },
+    {
+      "epoch": 80.26,
+      "grad_norm": 4.776179790496826,
+      "learning_rate": 1.974436090225564e-06,
+      "loss": 0.1553,
+      "step": 53370
+    },
+    {
+      "epoch": 80.27,
+      "grad_norm": 4.775459289550781,
+      "learning_rate": 1.972932330827068e-06,
+      "loss": 0.1948,
+      "step": 53380
+    },
+    {
+      "epoch": 80.29,
+      "grad_norm": 7.2186760902404785,
+      "learning_rate": 1.9714285714285714e-06,
+      "loss": 0.164,
+      "step": 53390
+    },
+    {
+      "epoch": 80.3,
+      "grad_norm": 12.228165626525879,
+      "learning_rate": 1.9699248120300754e-06,
+      "loss": 0.17,
+      "step": 53400
+    },
+    {
+      "epoch": 80.32,
+      "grad_norm": 3.7471702098846436,
+      "learning_rate": 1.9684210526315793e-06,
+      "loss": 0.1536,
+      "step": 53410
+    },
+    {
+      "epoch": 80.33,
+      "grad_norm": 4.057961940765381,
+      "learning_rate": 1.966917293233083e-06,
+      "loss": 0.1399,
+      "step": 53420
+    },
+    {
+      "epoch": 80.35,
+      "grad_norm": 5.822246074676514,
+      "learning_rate": 1.9654135338345868e-06,
+      "loss": 0.1627,
+      "step": 53430
+    },
+    {
+      "epoch": 80.36,
+      "grad_norm": 2.5543007850646973,
+      "learning_rate": 1.9639097744360903e-06,
+      "loss": 0.1273,
+      "step": 53440
+    },
+    {
+      "epoch": 80.38,
+      "grad_norm": 5.922774314880371,
+      "learning_rate": 1.9624060150375942e-06,
+      "loss": 0.212,
+      "step": 53450
+    },
+    {
+      "epoch": 80.39,
+      "grad_norm": 4.007785320281982,
+      "learning_rate": 1.960902255639098e-06,
+      "loss": 0.1515,
+      "step": 53460
+    },
+    {
+      "epoch": 80.41,
+      "grad_norm": 2.3833940029144287,
+      "learning_rate": 1.9593984962406017e-06,
+      "loss": 0.1828,
+      "step": 53470
+    },
+    {
+      "epoch": 80.42,
+      "grad_norm": 7.1434431076049805,
+      "learning_rate": 1.9578947368421052e-06,
+      "loss": 0.2086,
+      "step": 53480
+    },
+    {
+      "epoch": 80.44,
+      "grad_norm": 5.60752534866333,
+      "learning_rate": 1.956390977443609e-06,
+      "loss": 0.1768,
+      "step": 53490
+    },
+    {
+      "epoch": 80.45,
+      "grad_norm": 9.981124877929688,
+      "learning_rate": 1.9548872180451127e-06,
+      "loss": 0.1785,
+      "step": 53500
+    },
+    {
+      "epoch": 80.47,
+      "grad_norm": 2.1084699630737305,
+      "learning_rate": 1.9533834586466166e-06,
+      "loss": 0.2514,
+      "step": 53510
+    },
+    {
+      "epoch": 80.48,
+      "grad_norm": 3.638641834259033,
+      "learning_rate": 1.9518796992481206e-06,
+      "loss": 0.1162,
+      "step": 53520
+    },
+    {
+      "epoch": 80.5,
+      "grad_norm": 4.928158760070801,
+      "learning_rate": 1.950375939849624e-06,
+      "loss": 0.1602,
+      "step": 53530
+    },
+    {
+      "epoch": 80.51,
+      "grad_norm": 4.468033790588379,
+      "learning_rate": 1.948872180451128e-06,
+      "loss": 0.0994,
+      "step": 53540
+    },
+    {
+      "epoch": 80.53,
+      "grad_norm": 6.67031717300415,
+      "learning_rate": 1.9473684210526315e-06,
+      "loss": 0.1815,
+      "step": 53550
+    },
+    {
+      "epoch": 80.54,
+      "grad_norm": 4.985138893127441,
+      "learning_rate": 1.9458646616541355e-06,
+      "loss": 0.2014,
+      "step": 53560
+    },
+    {
+      "epoch": 80.56,
+      "grad_norm": 4.438155174255371,
+      "learning_rate": 1.9443609022556394e-06,
+      "loss": 0.1987,
+      "step": 53570
+    },
+    {
+      "epoch": 80.57,
+      "grad_norm": 6.645019054412842,
+      "learning_rate": 1.942857142857143e-06,
+      "loss": 0.1801,
+      "step": 53580
+    },
+    {
+      "epoch": 80.59,
+      "grad_norm": 2.9220592975616455,
+      "learning_rate": 1.941353383458647e-06,
+      "loss": 0.15,
+      "step": 53590
+    },
+    {
+      "epoch": 80.6,
+      "grad_norm": 3.196962356567383,
+      "learning_rate": 1.9398496240601504e-06,
+      "loss": 0.1512,
+      "step": 53600
+    },
+    {
+      "epoch": 80.62,
+      "grad_norm": 6.56658935546875,
+      "learning_rate": 1.9383458646616544e-06,
+      "loss": 0.1686,
+      "step": 53610
+    },
+    {
+      "epoch": 80.63,
+      "grad_norm": 8.974480628967285,
+      "learning_rate": 1.936842105263158e-06,
+      "loss": 0.1591,
+      "step": 53620
+    },
+    {
+      "epoch": 80.65,
+      "grad_norm": 9.335490226745605,
+      "learning_rate": 1.935338345864662e-06,
+      "loss": 0.1731,
+      "step": 53630
+    },
+    {
+      "epoch": 80.66,
+      "grad_norm": 8.802009582519531,
+      "learning_rate": 1.9338345864661658e-06,
+      "loss": 0.1393,
+      "step": 53640
+    },
+    {
+      "epoch": 80.68,
+      "grad_norm": 2.8996877670288086,
+      "learning_rate": 1.9323308270676693e-06,
+      "loss": 0.1955,
+      "step": 53650
+    },
+    {
+      "epoch": 80.69,
+      "grad_norm": 8.079301834106445,
+      "learning_rate": 1.9308270676691732e-06,
+      "loss": 0.2178,
+      "step": 53660
+    },
+    {
+      "epoch": 80.71,
+      "grad_norm": 4.58353853225708,
+      "learning_rate": 1.9293233082706767e-06,
+      "loss": 0.252,
+      "step": 53670
+    },
+    {
+      "epoch": 80.72,
+      "grad_norm": 1.9493281841278076,
+      "learning_rate": 1.9278195488721807e-06,
+      "loss": 0.15,
+      "step": 53680
+    },
+    {
+      "epoch": 80.74,
+      "grad_norm": 2.8279833793640137,
+      "learning_rate": 1.9263157894736846e-06,
+      "loss": 0.1748,
+      "step": 53690
+    },
+    {
+      "epoch": 80.75,
+      "grad_norm": 6.2720818519592285,
+      "learning_rate": 1.924812030075188e-06,
+      "loss": 0.1943,
+      "step": 53700
+    },
+    {
+      "epoch": 80.77,
+      "grad_norm": 5.162914752960205,
+      "learning_rate": 1.923308270676692e-06,
+      "loss": 0.1387,
+      "step": 53710
+    },
+    {
+      "epoch": 80.78,
+      "grad_norm": 5.782294273376465,
+      "learning_rate": 1.9218045112781956e-06,
+      "loss": 0.1838,
+      "step": 53720
+    },
+    {
+      "epoch": 80.8,
+      "grad_norm": 5.94707727432251,
+      "learning_rate": 1.920300751879699e-06,
+      "loss": 0.1411,
+      "step": 53730
+    },
+    {
+      "epoch": 80.81,
+      "grad_norm": 7.4608073234558105,
+      "learning_rate": 1.918796992481203e-06,
+      "loss": 0.1491,
+      "step": 53740
+    },
+    {
+      "epoch": 80.83,
+      "grad_norm": 3.1741888523101807,
+      "learning_rate": 1.917293233082707e-06,
+      "loss": 0.1506,
+      "step": 53750
+    },
+    {
+      "epoch": 80.84,
+      "grad_norm": 5.37153959274292,
+      "learning_rate": 1.9157894736842105e-06,
+      "loss": 0.2074,
+      "step": 53760
+    },
+    {
+      "epoch": 80.86,
+      "grad_norm": 2.308962821960449,
+      "learning_rate": 1.9142857142857145e-06,
+      "loss": 0.17,
+      "step": 53770
+    },
+    {
+      "epoch": 80.87,
+      "grad_norm": 2.749995470046997,
+      "learning_rate": 1.912781954887218e-06,
+      "loss": 0.1656,
+      "step": 53780
+    },
+    {
+      "epoch": 80.89,
+      "grad_norm": 4.785193920135498,
+      "learning_rate": 1.911278195488722e-06,
+      "loss": 0.1801,
+      "step": 53790
+    },
+    {
+      "epoch": 80.9,
+      "grad_norm": 5.815794467926025,
+      "learning_rate": 1.909774436090226e-06,
+      "loss": 0.1982,
+      "step": 53800
+    },
+    {
+      "epoch": 80.92,
+      "grad_norm": 4.500646591186523,
+      "learning_rate": 1.9082706766917294e-06,
+      "loss": 0.1509,
+      "step": 53810
+    },
+    {
+      "epoch": 80.93,
+      "grad_norm": 3.7033073902130127,
+      "learning_rate": 1.9067669172932331e-06,
+      "loss": 0.1613,
+      "step": 53820
+    },
+    {
+      "epoch": 80.95,
+      "grad_norm": 5.228376865386963,
+      "learning_rate": 1.905263157894737e-06,
+      "loss": 0.1766,
+      "step": 53830
+    },
+    {
+      "epoch": 80.96,
+      "grad_norm": 5.924206256866455,
+      "learning_rate": 1.9037593984962408e-06,
+      "loss": 0.1858,
+      "step": 53840
+    },
+    {
+      "epoch": 80.98,
+      "grad_norm": 2.352541923522949,
+      "learning_rate": 1.9022556390977445e-06,
+      "loss": 0.1322,
+      "step": 53850
+    },
+    {
+      "epoch": 80.99,
+      "grad_norm": 5.281574249267578,
+      "learning_rate": 1.9007518796992483e-06,
+      "loss": 0.1909,
+      "step": 53860
+    },
+    {
+      "epoch": 81.0,
+      "eval_accuracy": 0.9291,
+      "eval_loss": 0.34166744351387024,
+      "eval_runtime": 84.595,
+      "eval_samples_per_second": 118.21,
+      "eval_steps_per_second": 0.473,
+      "step": 53865
+    },
+    {
+      "epoch": 81.01,
+      "grad_norm": 6.562623977661133,
+      "learning_rate": 1.899248120300752e-06,
+      "loss": 0.1835,
+      "step": 53870
+    },
+    {
+      "epoch": 81.02,
+      "grad_norm": 6.715619087219238,
+      "learning_rate": 1.897744360902256e-06,
+      "loss": 0.2188,
+      "step": 53880
+    },
+    {
+      "epoch": 81.04,
+      "grad_norm": 3.6184804439544678,
+      "learning_rate": 1.8962406015037597e-06,
+      "loss": 0.1587,
+      "step": 53890
+    },
+    {
+      "epoch": 81.05,
+      "grad_norm": 7.747524261474609,
+      "learning_rate": 1.8947368421052634e-06,
+      "loss": 0.1861,
+      "step": 53900
+    },
+    {
+      "epoch": 81.07,
+      "grad_norm": 4.222412586212158,
+      "learning_rate": 1.8932330827067671e-06,
+      "loss": 0.1427,
+      "step": 53910
+    },
+    {
+      "epoch": 81.08,
+      "grad_norm": 7.423925399780273,
+      "learning_rate": 1.8917293233082709e-06,
+      "loss": 0.2178,
+      "step": 53920
+    },
+    {
+      "epoch": 81.1,
+      "grad_norm": 6.593136310577393,
+      "learning_rate": 1.8902255639097746e-06,
+      "loss": 0.2083,
+      "step": 53930
+    },
+    {
+      "epoch": 81.11,
+      "grad_norm": 5.739638805389404,
+      "learning_rate": 1.8887218045112785e-06,
+      "loss": 0.2182,
+      "step": 53940
+    },
+    {
+      "epoch": 81.13,
+      "grad_norm": 4.201329708099365,
+      "learning_rate": 1.8872180451127823e-06,
+      "loss": 0.1661,
+      "step": 53950
+    },
+    {
+      "epoch": 81.14,
+      "grad_norm": 4.125972270965576,
+      "learning_rate": 1.885714285714286e-06,
+      "loss": 0.177,
+      "step": 53960
+    },
+    {
+      "epoch": 81.16,
+      "grad_norm": 9.847689628601074,
+      "learning_rate": 1.8842105263157895e-06,
+      "loss": 0.1991,
+      "step": 53970
+    },
+    {
+      "epoch": 81.17,
+      "grad_norm": 2.6463875770568848,
+      "learning_rate": 1.8827067669172932e-06,
+      "loss": 0.1789,
+      "step": 53980
+    },
+    {
+      "epoch": 81.19,
+      "grad_norm": 5.124504089355469,
+      "learning_rate": 1.881203007518797e-06,
+      "loss": 0.1799,
+      "step": 53990
+    },
+    {
+      "epoch": 81.2,
+      "grad_norm": 6.498737335205078,
+      "learning_rate": 1.8796992481203007e-06,
+      "loss": 0.1637,
+      "step": 54000
+    },
+    {
+      "epoch": 81.22,
+      "grad_norm": 5.662612438201904,
+      "learning_rate": 1.8781954887218046e-06,
+      "loss": 0.1571,
+      "step": 54010
+    },
+    {
+      "epoch": 81.23,
+      "grad_norm": 4.6632232666015625,
+      "learning_rate": 1.8766917293233084e-06,
+      "loss": 0.1476,
+      "step": 54020
+    },
+    {
+      "epoch": 81.25,
+      "grad_norm": 3.081489324569702,
+      "learning_rate": 1.8751879699248121e-06,
+      "loss": 0.1544,
+      "step": 54030
+    },
+    {
+      "epoch": 81.26,
+      "grad_norm": 2.8134820461273193,
+      "learning_rate": 1.8736842105263158e-06,
+      "loss": 0.173,
+      "step": 54040
+    },
+    {
+      "epoch": 81.28,
+      "grad_norm": 6.176023483276367,
+      "learning_rate": 1.8721804511278196e-06,
+      "loss": 0.1662,
+      "step": 54050
+    },
+    {
+      "epoch": 81.29,
+      "grad_norm": 5.997969150543213,
+      "learning_rate": 1.8706766917293235e-06,
+      "loss": 0.1763,
+      "step": 54060
+    },
+    {
+      "epoch": 81.31,
+      "grad_norm": 7.557049751281738,
+      "learning_rate": 1.8691729323308272e-06,
+      "loss": 0.2356,
+      "step": 54070
+    },
+    {
+      "epoch": 81.32,
+      "grad_norm": 1.760933518409729,
+      "learning_rate": 1.867669172932331e-06,
+      "loss": 0.1678,
+      "step": 54080
+    },
+    {
+      "epoch": 81.34,
+      "grad_norm": 5.9524993896484375,
+      "learning_rate": 1.8661654135338347e-06,
+      "loss": 0.1923,
+      "step": 54090
+    },
+    {
+      "epoch": 81.35,
+      "grad_norm": 3.9762470722198486,
+      "learning_rate": 1.8646616541353384e-06,
+      "loss": 0.1556,
+      "step": 54100
+    },
+    {
+      "epoch": 81.37,
+      "grad_norm": 3.218543767929077,
+      "learning_rate": 1.8631578947368424e-06,
+      "loss": 0.1857,
+      "step": 54110
+    },
+    {
+      "epoch": 81.38,
+      "grad_norm": 6.120358467102051,
+      "learning_rate": 1.8616541353383461e-06,
+      "loss": 0.2167,
+      "step": 54120
+    },
+    {
+      "epoch": 81.4,
+      "grad_norm": 4.884064674377441,
+      "learning_rate": 1.8601503759398498e-06,
+      "loss": 0.1552,
+      "step": 54130
+    },
+    {
+      "epoch": 81.41,
+      "grad_norm": 5.395893573760986,
+      "learning_rate": 1.8586466165413536e-06,
+      "loss": 0.1528,
+      "step": 54140
+    },
+    {
+      "epoch": 81.43,
+      "grad_norm": 2.572039842605591,
+      "learning_rate": 1.8571428571428573e-06,
+      "loss": 0.1606,
+      "step": 54150
+    },
+    {
+      "epoch": 81.44,
+      "grad_norm": 2.791781425476074,
+      "learning_rate": 1.855639097744361e-06,
+      "loss": 0.1705,
+      "step": 54160
+    },
+    {
+      "epoch": 81.46,
+      "grad_norm": 3.3535261154174805,
+      "learning_rate": 1.854135338345865e-06,
+      "loss": 0.2432,
+      "step": 54170
+    },
+    {
+      "epoch": 81.47,
+      "grad_norm": 4.468529224395752,
+      "learning_rate": 1.8526315789473687e-06,
+      "loss": 0.1785,
+      "step": 54180
+    },
+    {
+      "epoch": 81.49,
+      "grad_norm": 3.004258871078491,
+      "learning_rate": 1.8511278195488724e-06,
+      "loss": 0.2144,
+      "step": 54190
+    },
+    {
+      "epoch": 81.5,
+      "grad_norm": 6.375463485717773,
+      "learning_rate": 1.8496240601503762e-06,
+      "loss": 0.1407,
+      "step": 54200
+    },
+    {
+      "epoch": 81.52,
+      "grad_norm": 7.171412467956543,
+      "learning_rate": 1.84812030075188e-06,
+      "loss": 0.1834,
+      "step": 54210
+    },
+    {
+      "epoch": 81.53,
+      "grad_norm": 4.493435859680176,
+      "learning_rate": 1.8466165413533834e-06,
+      "loss": 0.1702,
+      "step": 54220
+    },
+    {
+      "epoch": 81.55,
+      "grad_norm": 3.168029308319092,
+      "learning_rate": 1.8451127819548871e-06,
+      "loss": 0.1886,
+      "step": 54230
+    },
+    {
+      "epoch": 81.56,
+      "grad_norm": 6.606454372406006,
+      "learning_rate": 1.843609022556391e-06,
+      "loss": 0.1678,
+      "step": 54240
+    },
+    {
+      "epoch": 81.58,
+      "grad_norm": 7.317569255828857,
+      "learning_rate": 1.8421052631578948e-06,
+      "loss": 0.2086,
+      "step": 54250
+    },
+    {
+      "epoch": 81.59,
+      "grad_norm": 5.867735385894775,
+      "learning_rate": 1.8406015037593986e-06,
+      "loss": 0.2158,
+      "step": 54260
+    },
+    {
+      "epoch": 81.61,
+      "grad_norm": 6.016121864318848,
+      "learning_rate": 1.8390977443609023e-06,
+      "loss": 0.1706,
+      "step": 54270
+    },
+    {
+      "epoch": 81.62,
+      "grad_norm": 2.902163028717041,
+      "learning_rate": 1.837593984962406e-06,
+      "loss": 0.1605,
+      "step": 54280
+    },
+    {
+      "epoch": 81.64,
+      "grad_norm": 4.931081771850586,
+      "learning_rate": 1.83609022556391e-06,
+      "loss": 0.1675,
+      "step": 54290
+    },
+    {
+      "epoch": 81.65,
+      "grad_norm": 6.6242241859436035,
+      "learning_rate": 1.8345864661654137e-06,
+      "loss": 0.218,
+      "step": 54300
+    },
+    {
+      "epoch": 81.67,
+      "grad_norm": 4.538776397705078,
+      "learning_rate": 1.8330827067669174e-06,
+      "loss": 0.1534,
+      "step": 54310
+    },
+    {
+      "epoch": 81.68,
+      "grad_norm": 5.3194708824157715,
+      "learning_rate": 1.8315789473684211e-06,
+      "loss": 0.1789,
+      "step": 54320
+    },
+    {
+      "epoch": 81.7,
+      "grad_norm": 6.646446704864502,
+      "learning_rate": 1.8300751879699249e-06,
+      "loss": 0.1519,
+      "step": 54330
+    },
+    {
+      "epoch": 81.71,
+      "grad_norm": 7.2585129737854,
+      "learning_rate": 1.8285714285714288e-06,
+      "loss": 0.1952,
+      "step": 54340
+    },
+    {
+      "epoch": 81.73,
+      "grad_norm": 6.1801323890686035,
+      "learning_rate": 1.8270676691729326e-06,
+      "loss": 0.1726,
+      "step": 54350
+    },
+    {
+      "epoch": 81.74,
+      "grad_norm": 5.834466457366943,
+      "learning_rate": 1.8255639097744363e-06,
+      "loss": 0.1767,
+      "step": 54360
+    },
+    {
+      "epoch": 81.76,
+      "grad_norm": 5.508495807647705,
+      "learning_rate": 1.82406015037594e-06,
+      "loss": 0.1731,
+      "step": 54370
+    },
+    {
+      "epoch": 81.77,
+      "grad_norm": 1.7921135425567627,
+      "learning_rate": 1.8225563909774437e-06,
+      "loss": 0.2335,
+      "step": 54380
+    },
+    {
+      "epoch": 81.79,
+      "grad_norm": 5.268810749053955,
+      "learning_rate": 1.8210526315789475e-06,
+      "loss": 0.2098,
+      "step": 54390
+    },
+    {
+      "epoch": 81.8,
+      "grad_norm": 5.814802169799805,
+      "learning_rate": 1.8195488721804514e-06,
+      "loss": 0.1643,
+      "step": 54400
+    },
+    {
+      "epoch": 81.82,
+      "grad_norm": 4.461940765380859,
+      "learning_rate": 1.8180451127819551e-06,
+      "loss": 0.1374,
+      "step": 54410
+    },
+    {
+      "epoch": 81.83,
+      "grad_norm": 3.247183322906494,
+      "learning_rate": 1.8165413533834589e-06,
+      "loss": 0.1355,
+      "step": 54420
+    },
+    {
+      "epoch": 81.85,
+      "grad_norm": 1.9381792545318604,
+      "learning_rate": 1.8150375939849626e-06,
+      "loss": 0.1723,
+      "step": 54430
+    },
+    {
+      "epoch": 81.86,
+      "grad_norm": 7.478023529052734,
+      "learning_rate": 1.8135338345864663e-06,
+      "loss": 0.1846,
+      "step": 54440
+    },
+    {
+      "epoch": 81.88,
+      "grad_norm": 5.697751522064209,
+      "learning_rate": 1.8120300751879703e-06,
+      "loss": 0.1653,
+      "step": 54450
+    },
+    {
+      "epoch": 81.89,
+      "grad_norm": 5.588354587554932,
+      "learning_rate": 1.810526315789474e-06,
+      "loss": 0.1838,
+      "step": 54460
+    },
+    {
+      "epoch": 81.91,
+      "grad_norm": 6.156862258911133,
+      "learning_rate": 1.8090225563909775e-06,
+      "loss": 0.211,
+      "step": 54470
+    },
+    {
+      "epoch": 81.92,
+      "grad_norm": 7.52768087387085,
+      "learning_rate": 1.8075187969924813e-06,
+      "loss": 0.2083,
+      "step": 54480
+    },
+    {
+      "epoch": 81.94,
+      "grad_norm": 3.558300018310547,
+      "learning_rate": 1.806015037593985e-06,
+      "loss": 0.1508,
+      "step": 54490
+    },
+    {
+      "epoch": 81.95,
+      "grad_norm": 5.417634010314941,
+      "learning_rate": 1.8045112781954887e-06,
+      "loss": 0.1776,
+      "step": 54500
+    },
+    {
+      "epoch": 81.97,
+      "grad_norm": 5.08277702331543,
+      "learning_rate": 1.8030075187969925e-06,
+      "loss": 0.1513,
+      "step": 54510
+    },
+    {
+      "epoch": 81.98,
+      "grad_norm": 5.362541675567627,
+      "learning_rate": 1.8015037593984964e-06,
+      "loss": 0.1559,
+      "step": 54520
+    },
+    {
+      "epoch": 82.0,
+      "grad_norm": 1.1385769844055176,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 0.1971,
+      "step": 54530
+    },
+    {
+      "epoch": 82.0,
+      "eval_accuracy": 0.9289,
+      "eval_loss": 0.33228030800819397,
+      "eval_runtime": 85.2251,
+      "eval_samples_per_second": 117.336,
+      "eval_steps_per_second": 0.469,
+      "step": 54530
+    },
+    {
+      "epoch": 82.02,
+      "grad_norm": 2.4561429023742676,
+      "learning_rate": 1.7984962406015039e-06,
+      "loss": 0.1383,
+      "step": 54540
+    },
+    {
+      "epoch": 82.03,
+      "grad_norm": 4.4165425300598145,
+      "learning_rate": 1.7969924812030076e-06,
+      "loss": 0.1265,
+      "step": 54550
+    },
+    {
+      "epoch": 82.05,
+      "grad_norm": 8.235807418823242,
+      "learning_rate": 1.7954887218045113e-06,
+      "loss": 0.1736,
+      "step": 54560
+    },
+    {
+      "epoch": 82.06,
+      "grad_norm": 5.088337421417236,
+      "learning_rate": 1.7939849624060153e-06,
+      "loss": 0.1613,
+      "step": 54570
+    },
+    {
+      "epoch": 82.08,
+      "grad_norm": 4.968101501464844,
+      "learning_rate": 1.792481203007519e-06,
+      "loss": 0.1572,
+      "step": 54580
+    },
+    {
+      "epoch": 82.09,
+      "grad_norm": 3.9068446159362793,
+      "learning_rate": 1.7909774436090227e-06,
+      "loss": 0.2085,
+      "step": 54590
+    },
+    {
+      "epoch": 82.11,
+      "grad_norm": 4.162015914916992,
+      "learning_rate": 1.7894736842105265e-06,
+      "loss": 0.1924,
+      "step": 54600
+    },
+    {
+      "epoch": 82.12,
+      "grad_norm": 5.8883256912231445,
+      "learning_rate": 1.7879699248120302e-06,
+      "loss": 0.1482,
+      "step": 54610
+    },
+    {
+      "epoch": 82.14,
+      "grad_norm": 3.305724859237671,
+      "learning_rate": 1.786466165413534e-06,
+      "loss": 0.134,
+      "step": 54620
+    },
+    {
+      "epoch": 82.15,
+      "grad_norm": 3.4198994636535645,
+      "learning_rate": 1.7849624060150379e-06,
+      "loss": 0.122,
+      "step": 54630
+    },
+    {
+      "epoch": 82.17,
+      "grad_norm": 4.999546051025391,
+      "learning_rate": 1.7834586466165416e-06,
+      "loss": 0.2511,
+      "step": 54640
+    },
+    {
+      "epoch": 82.18,
+      "grad_norm": 3.644953489303589,
+      "learning_rate": 1.7819548872180453e-06,
+      "loss": 0.1224,
+      "step": 54650
+    },
+    {
+      "epoch": 82.2,
+      "grad_norm": 2.552194833755493,
+      "learning_rate": 1.780451127819549e-06,
+      "loss": 0.1872,
+      "step": 54660
+    },
+    {
+      "epoch": 82.21,
+      "grad_norm": 6.016075134277344,
+      "learning_rate": 1.7789473684210528e-06,
+      "loss": 0.1731,
+      "step": 54670
+    },
+    {
+      "epoch": 82.23,
+      "grad_norm": 3.725740671157837,
+      "learning_rate": 1.7774436090225567e-06,
+      "loss": 0.1697,
+      "step": 54680
+    },
+    {
+      "epoch": 82.24,
+      "grad_norm": 4.1656999588012695,
+      "learning_rate": 1.7759398496240605e-06,
+      "loss": 0.1599,
+      "step": 54690
+    },
+    {
+      "epoch": 82.26,
+      "grad_norm": 5.3262481689453125,
+      "learning_rate": 1.7744360902255642e-06,
+      "loss": 0.2012,
+      "step": 54700
+    },
+    {
+      "epoch": 82.27,
+      "grad_norm": 4.335447788238525,
+      "learning_rate": 1.772932330827068e-06,
+      "loss": 0.1789,
+      "step": 54710
+    },
+    {
+      "epoch": 82.29,
+      "grad_norm": 5.814763069152832,
+      "learning_rate": 1.7714285714285714e-06,
+      "loss": 0.1574,
+      "step": 54720
+    },
+    {
+      "epoch": 82.3,
+      "grad_norm": 8.759305000305176,
+      "learning_rate": 1.7699248120300752e-06,
+      "loss": 0.1821,
+      "step": 54730
+    },
+    {
+      "epoch": 82.32,
+      "grad_norm": 4.221161842346191,
+      "learning_rate": 1.768421052631579e-06,
+      "loss": 0.1759,
+      "step": 54740
+    },
+    {
+      "epoch": 82.33,
+      "grad_norm": 4.3325910568237305,
+      "learning_rate": 1.7669172932330828e-06,
+      "loss": 0.258,
+      "step": 54750
+    },
+    {
+      "epoch": 82.35,
+      "grad_norm": 4.424845218658447,
+      "learning_rate": 1.7654135338345866e-06,
+      "loss": 0.2111,
+      "step": 54760
+    },
+    {
+      "epoch": 82.36,
+      "grad_norm": 6.010334014892578,
+      "learning_rate": 1.7639097744360903e-06,
+      "loss": 0.2227,
+      "step": 54770
+    },
+    {
+      "epoch": 82.38,
+      "grad_norm": 4.39691162109375,
+      "learning_rate": 1.762406015037594e-06,
+      "loss": 0.1556,
+      "step": 54780
+    },
+    {
+      "epoch": 82.39,
+      "grad_norm": 4.894425868988037,
+      "learning_rate": 1.7609022556390978e-06,
+      "loss": 0.1579,
+      "step": 54790
+    },
+    {
+      "epoch": 82.41,
+      "grad_norm": 5.324883937835693,
+      "learning_rate": 1.7593984962406017e-06,
+      "loss": 0.2109,
+      "step": 54800
+    },
+    {
+      "epoch": 82.42,
+      "grad_norm": 6.122479438781738,
+      "learning_rate": 1.7578947368421054e-06,
+      "loss": 0.2117,
+      "step": 54810
+    },
+    {
+      "epoch": 82.44,
+      "grad_norm": 5.467765808105469,
+      "learning_rate": 1.7563909774436092e-06,
+      "loss": 0.1589,
+      "step": 54820
+    },
+    {
+      "epoch": 82.45,
+      "grad_norm": 8.767003059387207,
+      "learning_rate": 1.754887218045113e-06,
+      "loss": 0.1147,
+      "step": 54830
+    },
+    {
+      "epoch": 82.47,
+      "grad_norm": 9.435832977294922,
+      "learning_rate": 1.7533834586466166e-06,
+      "loss": 0.1292,
+      "step": 54840
+    },
+    {
+      "epoch": 82.48,
+      "grad_norm": 5.217706203460693,
+      "learning_rate": 1.7518796992481204e-06,
+      "loss": 0.148,
+      "step": 54850
+    },
+    {
+      "epoch": 82.5,
+      "grad_norm": 2.389409303665161,
+      "learning_rate": 1.7503759398496243e-06,
+      "loss": 0.1228,
+      "step": 54860
+    },
+    {
+      "epoch": 82.51,
+      "grad_norm": 1.7925937175750732,
+      "learning_rate": 1.748872180451128e-06,
+      "loss": 0.1766,
+      "step": 54870
+    },
+    {
+      "epoch": 82.53,
+      "grad_norm": 2.3438923358917236,
+      "learning_rate": 1.7473684210526318e-06,
+      "loss": 0.1621,
+      "step": 54880
+    },
+    {
+      "epoch": 82.54,
+      "grad_norm": 5.975131511688232,
+      "learning_rate": 1.7458646616541355e-06,
+      "loss": 0.1248,
+      "step": 54890
+    },
+    {
+      "epoch": 82.56,
+      "grad_norm": 12.676349639892578,
+      "learning_rate": 1.7443609022556392e-06,
+      "loss": 0.1825,
+      "step": 54900
+    },
+    {
+      "epoch": 82.57,
+      "grad_norm": 9.043052673339844,
+      "learning_rate": 1.7428571428571432e-06,
+      "loss": 0.1963,
+      "step": 54910
+    },
+    {
+      "epoch": 82.59,
+      "grad_norm": 2.8596065044403076,
+      "learning_rate": 1.741353383458647e-06,
+      "loss": 0.1685,
+      "step": 54920
+    },
+    {
+      "epoch": 82.6,
+      "grad_norm": 4.849795341491699,
+      "learning_rate": 1.7398496240601506e-06,
+      "loss": 0.1839,
+      "step": 54930
+    },
+    {
+      "epoch": 82.62,
+      "grad_norm": 7.803761959075928,
+      "learning_rate": 1.7383458646616544e-06,
+      "loss": 0.1732,
+      "step": 54940
+    },
+    {
+      "epoch": 82.63,
+      "grad_norm": 6.521337985992432,
+      "learning_rate": 1.736842105263158e-06,
+      "loss": 0.1562,
+      "step": 54950
+    },
+    {
+      "epoch": 82.65,
+      "grad_norm": 12.948254585266113,
+      "learning_rate": 1.735338345864662e-06,
+      "loss": 0.1772,
+      "step": 54960
+    },
+    {
+      "epoch": 82.66,
+      "grad_norm": 3.319334030151367,
+      "learning_rate": 1.7338345864661653e-06,
+      "loss": 0.2134,
+      "step": 54970
+    },
+    {
+      "epoch": 82.68,
+      "grad_norm": 2.6258463859558105,
+      "learning_rate": 1.7323308270676693e-06,
+      "loss": 0.1354,
+      "step": 54980
+    },
+    {
+      "epoch": 82.69,
+      "grad_norm": 5.683926105499268,
+      "learning_rate": 1.730827067669173e-06,
+      "loss": 0.1092,
+      "step": 54990
+    },
+    {
+      "epoch": 82.71,
+      "grad_norm": 6.205657958984375,
+      "learning_rate": 1.7293233082706767e-06,
+      "loss": 0.1635,
+      "step": 55000
+    },
+    {
+      "epoch": 82.72,
+      "grad_norm": 2.16133713722229,
+      "learning_rate": 1.7278195488721805e-06,
+      "loss": 0.1746,
+      "step": 55010
+    },
+    {
+      "epoch": 82.74,
+      "grad_norm": 5.119611740112305,
+      "learning_rate": 1.7263157894736842e-06,
+      "loss": 0.1608,
+      "step": 55020
+    },
+    {
+      "epoch": 82.75,
+      "grad_norm": 5.32130241394043,
+      "learning_rate": 1.7248120300751882e-06,
+      "loss": 0.1386,
+      "step": 55030
+    },
+    {
+      "epoch": 82.77,
+      "grad_norm": 5.220561981201172,
+      "learning_rate": 1.7233082706766919e-06,
+      "loss": 0.192,
+      "step": 55040
+    },
+    {
+      "epoch": 82.78,
+      "grad_norm": 5.081111431121826,
+      "learning_rate": 1.7218045112781956e-06,
+      "loss": 0.1744,
+      "step": 55050
+    },
+    {
+      "epoch": 82.8,
+      "grad_norm": 7.443409442901611,
+      "learning_rate": 1.7203007518796993e-06,
+      "loss": 0.2045,
+      "step": 55060
+    },
+    {
+      "epoch": 82.81,
+      "grad_norm": 5.770323276519775,
+      "learning_rate": 1.718796992481203e-06,
+      "loss": 0.1719,
+      "step": 55070
+    },
+    {
+      "epoch": 82.83,
+      "grad_norm": 5.356820106506348,
+      "learning_rate": 1.7172932330827068e-06,
+      "loss": 0.2003,
+      "step": 55080
+    },
+    {
+      "epoch": 82.84,
+      "grad_norm": 4.85361385345459,
+      "learning_rate": 1.7157894736842107e-06,
+      "loss": 0.1951,
+      "step": 55090
+    },
+    {
+      "epoch": 82.86,
+      "grad_norm": 4.393242359161377,
+      "learning_rate": 1.7142857142857145e-06,
+      "loss": 0.198,
+      "step": 55100
+    },
+    {
+      "epoch": 82.87,
+      "grad_norm": 2.715772867202759,
+      "learning_rate": 1.7127819548872182e-06,
+      "loss": 0.131,
+      "step": 55110
+    },
+    {
+      "epoch": 82.89,
+      "grad_norm": 5.954652786254883,
+      "learning_rate": 1.711278195488722e-06,
+      "loss": 0.2029,
+      "step": 55120
+    },
+    {
+      "epoch": 82.9,
+      "grad_norm": 3.8711795806884766,
+      "learning_rate": 1.7097744360902257e-06,
+      "loss": 0.1509,
+      "step": 55130
+    },
+    {
+      "epoch": 82.92,
+      "grad_norm": 4.712859630584717,
+      "learning_rate": 1.7082706766917296e-06,
+      "loss": 0.1485,
+      "step": 55140
+    },
+    {
+      "epoch": 82.93,
+      "grad_norm": 6.353842735290527,
+      "learning_rate": 1.7067669172932333e-06,
+      "loss": 0.1959,
+      "step": 55150
+    },
+    {
+      "epoch": 82.95,
+      "grad_norm": 4.206510066986084,
+      "learning_rate": 1.705263157894737e-06,
+      "loss": 0.1318,
+      "step": 55160
+    },
+    {
+      "epoch": 82.96,
+      "grad_norm": 4.608467102050781,
+      "learning_rate": 1.7037593984962408e-06,
+      "loss": 0.2017,
+      "step": 55170
+    },
+    {
+      "epoch": 82.98,
+      "grad_norm": 5.684893608093262,
+      "learning_rate": 1.7022556390977445e-06,
+      "loss": 0.1917,
+      "step": 55180
+    },
+    {
+      "epoch": 82.99,
+      "grad_norm": 4.197262763977051,
+      "learning_rate": 1.7007518796992485e-06,
+      "loss": 0.1739,
+      "step": 55190
+    },
+    {
+      "epoch": 83.0,
+      "eval_accuracy": 0.9323,
+      "eval_loss": 0.3265763223171234,
+      "eval_runtime": 84.4148,
+      "eval_samples_per_second": 118.463,
+      "eval_steps_per_second": 0.474,
+      "step": 55195
+    },
+    {
+      "epoch": 83.01,
+      "grad_norm": 4.629222869873047,
+      "learning_rate": 1.6992481203007522e-06,
+      "loss": 0.1463,
+      "step": 55200
+    },
+    {
+      "epoch": 83.02,
+      "grad_norm": 5.791532516479492,
+      "learning_rate": 1.6977443609022557e-06,
+      "loss": 0.1871,
+      "step": 55210
+    },
+    {
+      "epoch": 83.04,
+      "grad_norm": 6.979246616363525,
+      "learning_rate": 1.6962406015037595e-06,
+      "loss": 0.1502,
+      "step": 55220
+    },
+    {
+      "epoch": 83.05,
+      "grad_norm": 1.3585783243179321,
+      "learning_rate": 1.6947368421052632e-06,
+      "loss": 0.1733,
+      "step": 55230
+    },
+    {
+      "epoch": 83.07,
+      "grad_norm": 1.0612996816635132,
+      "learning_rate": 1.693233082706767e-06,
+      "loss": 0.1586,
+      "step": 55240
+    },
+    {
+      "epoch": 83.08,
+      "grad_norm": 3.546823740005493,
+      "learning_rate": 1.6917293233082707e-06,
+      "loss": 0.1586,
+      "step": 55250
+    },
+    {
+      "epoch": 83.1,
+      "grad_norm": 4.408979892730713,
+      "learning_rate": 1.6902255639097746e-06,
+      "loss": 0.2008,
+      "step": 55260
+    },
+    {
+      "epoch": 83.11,
+      "grad_norm": 4.838244438171387,
+      "learning_rate": 1.6887218045112783e-06,
+      "loss": 0.1753,
+      "step": 55270
+    },
+    {
+      "epoch": 83.13,
+      "grad_norm": 2.8008768558502197,
+      "learning_rate": 1.687218045112782e-06,
+      "loss": 0.1515,
+      "step": 55280
+    },
+    {
+      "epoch": 83.14,
+      "grad_norm": 1.7678091526031494,
+      "learning_rate": 1.6857142857142858e-06,
+      "loss": 0.1716,
+      "step": 55290
+    },
+    {
+      "epoch": 83.16,
+      "grad_norm": 4.2498087882995605,
+      "learning_rate": 1.6842105263157895e-06,
+      "loss": 0.2125,
+      "step": 55300
+    },
+    {
+      "epoch": 83.17,
+      "grad_norm": 4.752126216888428,
+      "learning_rate": 1.6827067669172933e-06,
+      "loss": 0.195,
+      "step": 55310
+    },
+    {
+      "epoch": 83.19,
+      "grad_norm": 2.648486375808716,
+      "learning_rate": 1.6812030075187972e-06,
+      "loss": 0.1554,
+      "step": 55320
+    },
+    {
+      "epoch": 83.2,
+      "grad_norm": 5.214283466339111,
+      "learning_rate": 1.679699248120301e-06,
+      "loss": 0.2508,
+      "step": 55330
+    },
+    {
+      "epoch": 83.22,
+      "grad_norm": 7.676024913787842,
+      "learning_rate": 1.6781954887218047e-06,
+      "loss": 0.1883,
+      "step": 55340
+    },
+    {
+      "epoch": 83.23,
+      "grad_norm": 5.258351802825928,
+      "learning_rate": 1.6766917293233084e-06,
+      "loss": 0.1941,
+      "step": 55350
+    },
+    {
+      "epoch": 83.25,
+      "grad_norm": 5.472609043121338,
+      "learning_rate": 1.6751879699248121e-06,
+      "loss": 0.1522,
+      "step": 55360
+    },
+    {
+      "epoch": 83.26,
+      "grad_norm": 3.57065749168396,
+      "learning_rate": 1.673684210526316e-06,
+      "loss": 0.1801,
+      "step": 55370
+    },
+    {
+      "epoch": 83.28,
+      "grad_norm": 4.2197651863098145,
+      "learning_rate": 1.6721804511278198e-06,
+      "loss": 0.1963,
+      "step": 55380
+    },
+    {
+      "epoch": 83.29,
+      "grad_norm": 7.963762283325195,
+      "learning_rate": 1.6706766917293235e-06,
+      "loss": 0.2155,
+      "step": 55390
+    },
+    {
+      "epoch": 83.31,
+      "grad_norm": 4.304222106933594,
+      "learning_rate": 1.6691729323308273e-06,
+      "loss": 0.225,
+      "step": 55400
+    },
+    {
+      "epoch": 83.32,
+      "grad_norm": 4.590184211730957,
+      "learning_rate": 1.667669172932331e-06,
+      "loss": 0.174,
+      "step": 55410
+    },
+    {
+      "epoch": 83.34,
+      "grad_norm": 2.6624538898468018,
+      "learning_rate": 1.6661654135338347e-06,
+      "loss": 0.1555,
+      "step": 55420
+    },
+    {
+      "epoch": 83.35,
+      "grad_norm": 6.655846118927002,
+      "learning_rate": 1.6646616541353387e-06,
+      "loss": 0.1236,
+      "step": 55430
+    },
+    {
+      "epoch": 83.37,
+      "grad_norm": 5.1804633140563965,
+      "learning_rate": 1.6631578947368424e-06,
+      "loss": 0.2144,
+      "step": 55440
+    },
+    {
+      "epoch": 83.38,
+      "grad_norm": 6.537972450256348,
+      "learning_rate": 1.6616541353383461e-06,
+      "loss": 0.1807,
+      "step": 55450
+    },
+    {
+      "epoch": 83.4,
+      "grad_norm": 3.086345911026001,
+      "learning_rate": 1.6601503759398496e-06,
+      "loss": 0.1505,
+      "step": 55460
+    },
+    {
+      "epoch": 83.41,
+      "grad_norm": 6.289426326751709,
+      "learning_rate": 1.6586466165413534e-06,
+      "loss": 0.1967,
+      "step": 55470
+    },
+    {
+      "epoch": 83.43,
+      "grad_norm": 4.445835113525391,
+      "learning_rate": 1.657142857142857e-06,
+      "loss": 0.153,
+      "step": 55480
+    },
+    {
+      "epoch": 83.44,
+      "grad_norm": 5.393485069274902,
+      "learning_rate": 1.6556390977443608e-06,
+      "loss": 0.1946,
+      "step": 55490
+    },
+    {
+      "epoch": 83.46,
+      "grad_norm": 4.735557556152344,
+      "learning_rate": 1.6541353383458648e-06,
+      "loss": 0.1649,
+      "step": 55500
+    },
+    {
+      "epoch": 83.47,
+      "grad_norm": 2.2363061904907227,
+      "learning_rate": 1.6526315789473685e-06,
+      "loss": 0.2182,
+      "step": 55510
+    },
+    {
+      "epoch": 83.49,
+      "grad_norm": 1.6198318004608154,
+      "learning_rate": 1.6511278195488722e-06,
+      "loss": 0.2171,
+      "step": 55520
+    },
+    {
+      "epoch": 83.5,
+      "grad_norm": 3.38696551322937,
+      "learning_rate": 1.649624060150376e-06,
+      "loss": 0.1405,
+      "step": 55530
+    },
+    {
+      "epoch": 83.52,
+      "grad_norm": 1.2411551475524902,
+      "learning_rate": 1.6481203007518797e-06,
+      "loss": 0.1632,
+      "step": 55540
+    },
+    {
+      "epoch": 83.53,
+      "grad_norm": 3.3203301429748535,
+      "learning_rate": 1.6466165413533836e-06,
+      "loss": 0.1592,
+      "step": 55550
+    },
+    {
+      "epoch": 83.55,
+      "grad_norm": 13.928276062011719,
+      "learning_rate": 1.6451127819548874e-06,
+      "loss": 0.1434,
+      "step": 55560
+    },
+    {
+      "epoch": 83.56,
+      "grad_norm": 3.75093150138855,
+      "learning_rate": 1.643609022556391e-06,
+      "loss": 0.2471,
+      "step": 55570
+    },
+    {
+      "epoch": 83.58,
+      "grad_norm": 6.298830032348633,
+      "learning_rate": 1.6421052631578948e-06,
+      "loss": 0.1624,
+      "step": 55580
+    },
+    {
+      "epoch": 83.59,
+      "grad_norm": 4.140471935272217,
+      "learning_rate": 1.6406015037593986e-06,
+      "loss": 0.1668,
+      "step": 55590
+    },
+    {
+      "epoch": 83.61,
+      "grad_norm": 7.234468936920166,
+      "learning_rate": 1.6390977443609025e-06,
+      "loss": 0.1933,
+      "step": 55600
+    },
+    {
+      "epoch": 83.62,
+      "grad_norm": 5.53890323638916,
+      "learning_rate": 1.6375939849624062e-06,
+      "loss": 0.1686,
+      "step": 55610
+    },
+    {
+      "epoch": 83.64,
+      "grad_norm": 7.111166000366211,
+      "learning_rate": 1.63609022556391e-06,
+      "loss": 0.1447,
+      "step": 55620
+    },
+    {
+      "epoch": 83.65,
+      "grad_norm": 4.891413688659668,
+      "learning_rate": 1.6345864661654137e-06,
+      "loss": 0.1894,
+      "step": 55630
+    },
+    {
+      "epoch": 83.67,
+      "grad_norm": 5.022510051727295,
+      "learning_rate": 1.6330827067669174e-06,
+      "loss": 0.1572,
+      "step": 55640
+    },
+    {
+      "epoch": 83.68,
+      "grad_norm": 6.799849033355713,
+      "learning_rate": 1.6315789473684212e-06,
+      "loss": 0.2438,
+      "step": 55650
+    },
+    {
+      "epoch": 83.7,
+      "grad_norm": 6.336165904998779,
+      "learning_rate": 1.630075187969925e-06,
+      "loss": 0.1738,
+      "step": 55660
+    },
+    {
+      "epoch": 83.71,
+      "grad_norm": 4.65312385559082,
+      "learning_rate": 1.6285714285714288e-06,
+      "loss": 0.1406,
+      "step": 55670
+    },
+    {
+      "epoch": 83.73,
+      "grad_norm": 6.864182472229004,
+      "learning_rate": 1.6270676691729326e-06,
+      "loss": 0.1773,
+      "step": 55680
+    },
+    {
+      "epoch": 83.74,
+      "grad_norm": 3.7218823432922363,
+      "learning_rate": 1.6255639097744363e-06,
+      "loss": 0.1292,
+      "step": 55690
+    },
+    {
+      "epoch": 83.76,
+      "grad_norm": 6.442845821380615,
+      "learning_rate": 1.62406015037594e-06,
+      "loss": 0.1512,
+      "step": 55700
+    },
+    {
+      "epoch": 83.77,
+      "grad_norm": 4.894304275512695,
+      "learning_rate": 1.6225563909774435e-06,
+      "loss": 0.162,
+      "step": 55710
+    },
+    {
+      "epoch": 83.79,
+      "grad_norm": 5.067921161651611,
+      "learning_rate": 1.6210526315789473e-06,
+      "loss": 0.1662,
+      "step": 55720
+    },
+    {
+      "epoch": 83.8,
+      "grad_norm": 10.18139362335205,
+      "learning_rate": 1.6195488721804512e-06,
+      "loss": 0.2123,
+      "step": 55730
+    },
+    {
+      "epoch": 83.82,
+      "grad_norm": 6.891566753387451,
+      "learning_rate": 1.618045112781955e-06,
+      "loss": 0.1701,
+      "step": 55740
+    },
+    {
+      "epoch": 83.83,
+      "grad_norm": 3.438103437423706,
+      "learning_rate": 1.6165413533834587e-06,
+      "loss": 0.1598,
+      "step": 55750
+    },
+    {
+      "epoch": 83.85,
+      "grad_norm": 3.0639941692352295,
+      "learning_rate": 1.6150375939849624e-06,
+      "loss": 0.2099,
+      "step": 55760
+    },
+    {
+      "epoch": 83.86,
+      "grad_norm": 5.10312557220459,
+      "learning_rate": 1.6135338345864661e-06,
+      "loss": 0.1678,
+      "step": 55770
+    },
+    {
+      "epoch": 83.88,
+      "grad_norm": 3.48979115486145,
+      "learning_rate": 1.61203007518797e-06,
+      "loss": 0.1578,
+      "step": 55780
+    },
+    {
+      "epoch": 83.89,
+      "grad_norm": 7.103468418121338,
+      "learning_rate": 1.6105263157894738e-06,
+      "loss": 0.1769,
+      "step": 55790
+    },
+    {
+      "epoch": 83.91,
+      "grad_norm": 4.037766456604004,
+      "learning_rate": 1.6090225563909775e-06,
+      "loss": 0.2101,
+      "step": 55800
+    },
+    {
+      "epoch": 83.92,
+      "grad_norm": 3.874589204788208,
+      "learning_rate": 1.6075187969924813e-06,
+      "loss": 0.2364,
+      "step": 55810
+    },
+    {
+      "epoch": 83.94,
+      "grad_norm": 4.208032608032227,
+      "learning_rate": 1.606015037593985e-06,
+      "loss": 0.1712,
+      "step": 55820
+    },
+    {
+      "epoch": 83.95,
+      "grad_norm": 4.456460475921631,
+      "learning_rate": 1.604511278195489e-06,
+      "loss": 0.2004,
+      "step": 55830
+    },
+    {
+      "epoch": 83.97,
+      "grad_norm": 6.270118236541748,
+      "learning_rate": 1.6030075187969927e-06,
+      "loss": 0.156,
+      "step": 55840
+    },
+    {
+      "epoch": 83.98,
+      "grad_norm": 5.951742172241211,
+      "learning_rate": 1.6015037593984964e-06,
+      "loss": 0.2439,
+      "step": 55850
+    },
+    {
+      "epoch": 84.0,
+      "grad_norm": 0.09190316498279572,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 0.1537,
+      "step": 55860
+    },
+    {
+      "epoch": 84.0,
+      "eval_accuracy": 0.9294,
+      "eval_loss": 0.33126822113990784,
+      "eval_runtime": 84.9554,
+      "eval_samples_per_second": 117.709,
+      "eval_steps_per_second": 0.471,
+      "step": 55860
+    },
+    {
+      "epoch": 84.02,
+      "grad_norm": 4.058501243591309,
+      "learning_rate": 1.5984962406015039e-06,
+      "loss": 0.238,
+      "step": 55870
+    },
+    {
+      "epoch": 84.03,
+      "grad_norm": 3.7976138591766357,
+      "learning_rate": 1.5969924812030076e-06,
+      "loss": 0.1423,
+      "step": 55880
+    },
+    {
+      "epoch": 84.05,
+      "grad_norm": 4.4287614822387695,
+      "learning_rate": 1.5954887218045115e-06,
+      "loss": 0.1356,
+      "step": 55890
+    },
+    {
+      "epoch": 84.06,
+      "grad_norm": 5.282792568206787,
+      "learning_rate": 1.5939849624060153e-06,
+      "loss": 0.1902,
+      "step": 55900
+    },
+    {
+      "epoch": 84.08,
+      "grad_norm": 5.178953170776367,
+      "learning_rate": 1.592481203007519e-06,
+      "loss": 0.1712,
+      "step": 55910
+    },
+    {
+      "epoch": 84.09,
+      "grad_norm": 0.7578862905502319,
+      "learning_rate": 1.5909774436090227e-06,
+      "loss": 0.1615,
+      "step": 55920
+    },
+    {
+      "epoch": 84.11,
+      "grad_norm": 6.100287437438965,
+      "learning_rate": 1.5894736842105265e-06,
+      "loss": 0.144,
+      "step": 55930
+    },
+    {
+      "epoch": 84.12,
+      "grad_norm": 10.830288887023926,
+      "learning_rate": 1.5879699248120304e-06,
+      "loss": 0.1449,
+      "step": 55940
+    },
+    {
+      "epoch": 84.14,
+      "grad_norm": 6.316339015960693,
+      "learning_rate": 1.5864661654135341e-06,
+      "loss": 0.202,
+      "step": 55950
+    },
+    {
+      "epoch": 84.15,
+      "grad_norm": 4.204655647277832,
+      "learning_rate": 1.5849624060150377e-06,
+      "loss": 0.1957,
+      "step": 55960
+    },
+    {
+      "epoch": 84.17,
+      "grad_norm": 5.749233722686768,
+      "learning_rate": 1.5834586466165414e-06,
+      "loss": 0.2423,
+      "step": 55970
+    },
+    {
+      "epoch": 84.18,
+      "grad_norm": 2.607123851776123,
+      "learning_rate": 1.5819548872180451e-06,
+      "loss": 0.1752,
+      "step": 55980
+    },
+    {
+      "epoch": 84.2,
+      "grad_norm": 3.5867726802825928,
+      "learning_rate": 1.5804511278195489e-06,
+      "loss": 0.2333,
+      "step": 55990
+    },
+    {
+      "epoch": 84.21,
+      "grad_norm": 3.1847939491271973,
+      "learning_rate": 1.5789473684210526e-06,
+      "loss": 0.1552,
+      "step": 56000
+    },
+    {
+      "epoch": 84.23,
+      "grad_norm": 5.968934535980225,
+      "learning_rate": 1.5774436090225565e-06,
+      "loss": 0.155,
+      "step": 56010
+    },
+    {
+      "epoch": 84.24,
+      "grad_norm": 3.6189496517181396,
+      "learning_rate": 1.5759398496240603e-06,
+      "loss": 0.1612,
+      "step": 56020
+    },
+    {
+      "epoch": 84.26,
+      "grad_norm": 1.3396971225738525,
+      "learning_rate": 1.574436090225564e-06,
+      "loss": 0.1614,
+      "step": 56030
+    },
+    {
+      "epoch": 84.27,
+      "grad_norm": 2.855109930038452,
+      "learning_rate": 1.5729323308270677e-06,
+      "loss": 0.1615,
+      "step": 56040
+    },
+    {
+      "epoch": 84.29,
+      "grad_norm": 4.582769393920898,
+      "learning_rate": 1.5714285714285714e-06,
+      "loss": 0.2042,
+      "step": 56050
+    },
+    {
+      "epoch": 84.3,
+      "grad_norm": 5.339018821716309,
+      "learning_rate": 1.5699248120300754e-06,
+      "loss": 0.1094,
+      "step": 56060
+    },
+    {
+      "epoch": 84.32,
+      "grad_norm": 7.760763168334961,
+      "learning_rate": 1.5684210526315791e-06,
+      "loss": 0.2093,
+      "step": 56070
+    },
+    {
+      "epoch": 84.33,
+      "grad_norm": 7.610210418701172,
+      "learning_rate": 1.5669172932330829e-06,
+      "loss": 0.1975,
+      "step": 56080
+    },
+    {
+      "epoch": 84.35,
+      "grad_norm": 3.7673017978668213,
+      "learning_rate": 1.5654135338345866e-06,
+      "loss": 0.1645,
+      "step": 56090
+    },
+    {
+      "epoch": 84.36,
+      "grad_norm": 2.311070203781128,
+      "learning_rate": 1.5639097744360903e-06,
+      "loss": 0.1574,
+      "step": 56100
+    },
+    {
+      "epoch": 84.38,
+      "grad_norm": 5.664913177490234,
+      "learning_rate": 1.562406015037594e-06,
+      "loss": 0.1414,
+      "step": 56110
+    },
+    {
+      "epoch": 84.39,
+      "grad_norm": 6.023291110992432,
+      "learning_rate": 1.560902255639098e-06,
+      "loss": 0.1766,
+      "step": 56120
+    },
+    {
+      "epoch": 84.41,
+      "grad_norm": 4.557356357574463,
+      "learning_rate": 1.5593984962406017e-06,
+      "loss": 0.1912,
+      "step": 56130
+    },
+    {
+      "epoch": 84.42,
+      "grad_norm": 3.221834421157837,
+      "learning_rate": 1.5578947368421054e-06,
+      "loss": 0.1643,
+      "step": 56140
+    },
+    {
+      "epoch": 84.44,
+      "grad_norm": 7.018657684326172,
+      "learning_rate": 1.5563909774436092e-06,
+      "loss": 0.2211,
+      "step": 56150
+    },
+    {
+      "epoch": 84.45,
+      "grad_norm": 5.987710475921631,
+      "learning_rate": 1.554887218045113e-06,
+      "loss": 0.1389,
+      "step": 56160
+    },
+    {
+      "epoch": 84.47,
+      "grad_norm": 3.3997559547424316,
+      "learning_rate": 1.5533834586466169e-06,
+      "loss": 0.1328,
+      "step": 56170
+    },
+    {
+      "epoch": 84.48,
+      "grad_norm": 3.822932243347168,
+      "learning_rate": 1.5518796992481206e-06,
+      "loss": 0.2355,
+      "step": 56180
+    },
+    {
+      "epoch": 84.5,
+      "grad_norm": 4.0138349533081055,
+      "learning_rate": 1.5503759398496243e-06,
+      "loss": 0.1783,
+      "step": 56190
+    },
+    {
+      "epoch": 84.51,
+      "grad_norm": 8.904444694519043,
+      "learning_rate": 1.548872180451128e-06,
+      "loss": 0.2122,
+      "step": 56200
+    },
+    {
+      "epoch": 84.53,
+      "grad_norm": 7.065569877624512,
+      "learning_rate": 1.5473684210526316e-06,
+      "loss": 0.1961,
+      "step": 56210
+    },
+    {
+      "epoch": 84.54,
+      "grad_norm": 8.500096321105957,
+      "learning_rate": 1.5458646616541353e-06,
+      "loss": 0.1663,
+      "step": 56220
+    },
+    {
+      "epoch": 84.56,
+      "grad_norm": 4.012298583984375,
+      "learning_rate": 1.544360902255639e-06,
+      "loss": 0.2012,
+      "step": 56230
+    },
+    {
+      "epoch": 84.57,
+      "grad_norm": 7.489271640777588,
+      "learning_rate": 1.542857142857143e-06,
+      "loss": 0.1539,
+      "step": 56240
+    },
+    {
+      "epoch": 84.59,
+      "grad_norm": 3.202110767364502,
+      "learning_rate": 1.5413533834586467e-06,
+      "loss": 0.2089,
+      "step": 56250
+    },
+    {
+      "epoch": 84.6,
+      "grad_norm": 1.4287196397781372,
+      "learning_rate": 1.5398496240601504e-06,
+      "loss": 0.1811,
+      "step": 56260
+    },
+    {
+      "epoch": 84.62,
+      "grad_norm": 5.019970417022705,
+      "learning_rate": 1.5383458646616542e-06,
+      "loss": 0.184,
+      "step": 56270
+    },
+    {
+      "epoch": 84.63,
+      "grad_norm": 2.368131637573242,
+      "learning_rate": 1.5368421052631579e-06,
+      "loss": 0.1327,
+      "step": 56280
+    },
+    {
+      "epoch": 84.65,
+      "grad_norm": 3.924480676651001,
+      "learning_rate": 1.5353383458646618e-06,
+      "loss": 0.1848,
+      "step": 56290
+    },
+    {
+      "epoch": 84.66,
+      "grad_norm": 8.339082717895508,
+      "learning_rate": 1.5338345864661656e-06,
+      "loss": 0.2257,
+      "step": 56300
+    },
+    {
+      "epoch": 84.68,
+      "grad_norm": 4.868645668029785,
+      "learning_rate": 1.5323308270676693e-06,
+      "loss": 0.1838,
+      "step": 56310
+    },
+    {
+      "epoch": 84.69,
+      "grad_norm": 5.948423862457275,
+      "learning_rate": 1.530827067669173e-06,
+      "loss": 0.1669,
+      "step": 56320
+    },
+    {
+      "epoch": 84.71,
+      "grad_norm": 3.8026175498962402,
+      "learning_rate": 1.5293233082706768e-06,
+      "loss": 0.1501,
+      "step": 56330
+    },
+    {
+      "epoch": 84.72,
+      "grad_norm": 6.754022598266602,
+      "learning_rate": 1.5278195488721805e-06,
+      "loss": 0.1706,
+      "step": 56340
+    },
+    {
+      "epoch": 84.74,
+      "grad_norm": 4.1264495849609375,
+      "learning_rate": 1.5263157894736844e-06,
+      "loss": 0.2263,
+      "step": 56350
+    },
+    {
+      "epoch": 84.75,
+      "grad_norm": 7.627476692199707,
+      "learning_rate": 1.5248120300751882e-06,
+      "loss": 0.1672,
+      "step": 56360
+    },
+    {
+      "epoch": 84.77,
+      "grad_norm": 3.8254830837249756,
+      "learning_rate": 1.5233082706766919e-06,
+      "loss": 0.2137,
+      "step": 56370
+    },
+    {
+      "epoch": 84.78,
+      "grad_norm": 4.202638149261475,
+      "learning_rate": 1.5218045112781956e-06,
+      "loss": 0.1923,
+      "step": 56380
+    },
+    {
+      "epoch": 84.8,
+      "grad_norm": 4.201712608337402,
+      "learning_rate": 1.5203007518796994e-06,
+      "loss": 0.1436,
+      "step": 56390
+    },
+    {
+      "epoch": 84.81,
+      "grad_norm": 4.552309989929199,
+      "learning_rate": 1.5187969924812033e-06,
+      "loss": 0.1477,
+      "step": 56400
+    },
+    {
+      "epoch": 84.83,
+      "grad_norm": 5.196654319763184,
+      "learning_rate": 1.517293233082707e-06,
+      "loss": 0.2116,
+      "step": 56410
+    },
+    {
+      "epoch": 84.84,
+      "grad_norm": 6.609741687774658,
+      "learning_rate": 1.5157894736842108e-06,
+      "loss": 0.2068,
+      "step": 56420
+    },
+    {
+      "epoch": 84.86,
+      "grad_norm": 3.7197117805480957,
+      "learning_rate": 1.5142857142857145e-06,
+      "loss": 0.2105,
+      "step": 56430
+    },
+    {
+      "epoch": 84.87,
+      "grad_norm": 1.3698334693908691,
+      "learning_rate": 1.5127819548872182e-06,
+      "loss": 0.2096,
+      "step": 56440
+    },
+    {
+      "epoch": 84.89,
+      "grad_norm": 5.72020959854126,
+      "learning_rate": 1.5112781954887222e-06,
+      "loss": 0.1488,
+      "step": 56450
+    },
+    {
+      "epoch": 84.9,
+      "grad_norm": 2.5910472869873047,
+      "learning_rate": 1.5097744360902255e-06,
+      "loss": 0.1275,
+      "step": 56460
+    },
+    {
+      "epoch": 84.92,
+      "grad_norm": 6.082350254058838,
+      "learning_rate": 1.5082706766917294e-06,
+      "loss": 0.1829,
+      "step": 56470
+    },
+    {
+      "epoch": 84.93,
+      "grad_norm": 3.298006296157837,
+      "learning_rate": 1.5067669172932331e-06,
+      "loss": 0.1375,
+      "step": 56480
+    },
+    {
+      "epoch": 84.95,
+      "grad_norm": 5.607242107391357,
+      "learning_rate": 1.5052631578947369e-06,
+      "loss": 0.155,
+      "step": 56490
+    },
+    {
+      "epoch": 84.96,
+      "grad_norm": 3.7076168060302734,
+      "learning_rate": 1.5037593984962406e-06,
+      "loss": 0.1593,
+      "step": 56500
+    },
+    {
+      "epoch": 84.98,
+      "grad_norm": 2.4990506172180176,
+      "learning_rate": 1.5022556390977443e-06,
+      "loss": 0.1322,
+      "step": 56510
+    },
+    {
+      "epoch": 84.99,
+      "grad_norm": 2.6433887481689453,
+      "learning_rate": 1.5007518796992483e-06,
+      "loss": 0.1706,
+      "step": 56520
+    },
+    {
+      "epoch": 85.0,
+      "eval_accuracy": 0.928,
+      "eval_loss": 0.3395210802555084,
+      "eval_runtime": 84.3204,
+      "eval_samples_per_second": 118.595,
+      "eval_steps_per_second": 0.474,
+      "step": 56525
+    },
+    {
+      "epoch": 85.01,
+      "grad_norm": 4.445404529571533,
+      "learning_rate": 1.499248120300752e-06,
+      "loss": 0.162,
+      "step": 56530
+    },
+    {
+      "epoch": 85.02,
+      "grad_norm": 9.234296798706055,
+      "learning_rate": 1.4977443609022557e-06,
+      "loss": 0.1999,
+      "step": 56540
+    },
+    {
+      "epoch": 85.04,
+      "grad_norm": 4.423348426818848,
+      "learning_rate": 1.4962406015037595e-06,
+      "loss": 0.1403,
+      "step": 56550
+    },
+    {
+      "epoch": 85.05,
+      "grad_norm": 2.3429605960845947,
+      "learning_rate": 1.4947368421052632e-06,
+      "loss": 0.199,
+      "step": 56560
+    },
+    {
+      "epoch": 85.07,
+      "grad_norm": 5.699217319488525,
+      "learning_rate": 1.493233082706767e-06,
+      "loss": 0.2032,
+      "step": 56570
+    },
+    {
+      "epoch": 85.08,
+      "grad_norm": 3.4555978775024414,
+      "learning_rate": 1.4917293233082709e-06,
+      "loss": 0.1305,
+      "step": 56580
+    },
+    {
+      "epoch": 85.1,
+      "grad_norm": 4.182602882385254,
+      "learning_rate": 1.4902255639097746e-06,
+      "loss": 0.2042,
+      "step": 56590
+    },
+    {
+      "epoch": 85.11,
+      "grad_norm": 7.7609686851501465,
+      "learning_rate": 1.4887218045112783e-06,
+      "loss": 0.2182,
+      "step": 56600
+    },
+    {
+      "epoch": 85.13,
+      "grad_norm": 4.991521835327148,
+      "learning_rate": 1.487218045112782e-06,
+      "loss": 0.1919,
+      "step": 56610
+    },
+    {
+      "epoch": 85.14,
+      "grad_norm": 5.988854885101318,
+      "learning_rate": 1.4857142857142858e-06,
+      "loss": 0.2383,
+      "step": 56620
+    },
+    {
+      "epoch": 85.16,
+      "grad_norm": 3.479478597640991,
+      "learning_rate": 1.4842105263157897e-06,
+      "loss": 0.2143,
+      "step": 56630
+    },
+    {
+      "epoch": 85.17,
+      "grad_norm": 4.183614253997803,
+      "learning_rate": 1.4827067669172935e-06,
+      "loss": 0.2017,
+      "step": 56640
+    },
+    {
+      "epoch": 85.19,
+      "grad_norm": 4.1281418800354,
+      "learning_rate": 1.4812030075187972e-06,
+      "loss": 0.1597,
+      "step": 56650
+    },
+    {
+      "epoch": 85.2,
+      "grad_norm": 4.900537014007568,
+      "learning_rate": 1.479699248120301e-06,
+      "loss": 0.1677,
+      "step": 56660
+    },
+    {
+      "epoch": 85.22,
+      "grad_norm": 2.788635492324829,
+      "learning_rate": 1.4781954887218047e-06,
+      "loss": 0.1888,
+      "step": 56670
+    },
+    {
+      "epoch": 85.23,
+      "grad_norm": 8.006430625915527,
+      "learning_rate": 1.4766917293233086e-06,
+      "loss": 0.1544,
+      "step": 56680
+    },
+    {
+      "epoch": 85.25,
+      "grad_norm": 4.344766616821289,
+      "learning_rate": 1.4751879699248123e-06,
+      "loss": 0.147,
+      "step": 56690
+    },
+    {
+      "epoch": 85.26,
+      "grad_norm": 5.809935092926025,
+      "learning_rate": 1.4736842105263159e-06,
+      "loss": 0.1912,
+      "step": 56700
+    },
+    {
+      "epoch": 85.28,
+      "grad_norm": 6.294594764709473,
+      "learning_rate": 1.4721804511278196e-06,
+      "loss": 0.1366,
+      "step": 56710
+    },
+    {
+      "epoch": 85.29,
+      "grad_norm": 5.718160152435303,
+      "learning_rate": 1.4706766917293233e-06,
+      "loss": 0.1554,
+      "step": 56720
+    },
+    {
+      "epoch": 85.31,
+      "grad_norm": 4.895095348358154,
+      "learning_rate": 1.469172932330827e-06,
+      "loss": 0.2095,
+      "step": 56730
+    },
+    {
+      "epoch": 85.32,
+      "grad_norm": 4.996368885040283,
+      "learning_rate": 1.4676691729323308e-06,
+      "loss": 0.2099,
+      "step": 56740
+    },
+    {
+      "epoch": 85.34,
+      "grad_norm": 4.071951866149902,
+      "learning_rate": 1.4661654135338347e-06,
+      "loss": 0.1789,
+      "step": 56750
+    },
+    {
+      "epoch": 85.35,
+      "grad_norm": 4.790773391723633,
+      "learning_rate": 1.4646616541353385e-06,
+      "loss": 0.2203,
+      "step": 56760
+    },
+    {
+      "epoch": 85.37,
+      "grad_norm": 3.539854049682617,
+      "learning_rate": 1.4631578947368422e-06,
+      "loss": 0.1297,
+      "step": 56770
+    },
+    {
+      "epoch": 85.38,
+      "grad_norm": 8.937888145446777,
+      "learning_rate": 1.461654135338346e-06,
+      "loss": 0.2044,
+      "step": 56780
+    },
+    {
+      "epoch": 85.4,
+      "grad_norm": 5.165222644805908,
+      "learning_rate": 1.4601503759398496e-06,
+      "loss": 0.1447,
+      "step": 56790
+    },
+    {
+      "epoch": 85.41,
+      "grad_norm": 7.947958946228027,
+      "learning_rate": 1.4586466165413534e-06,
+      "loss": 0.2182,
+      "step": 56800
+    },
+    {
+      "epoch": 85.43,
+      "grad_norm": 3.317690849304199,
+      "learning_rate": 1.4571428571428573e-06,
+      "loss": 0.2299,
+      "step": 56810
+    },
+    {
+      "epoch": 85.44,
+      "grad_norm": 3.5233917236328125,
+      "learning_rate": 1.455639097744361e-06,
+      "loss": 0.2098,
+      "step": 56820
+    },
+    {
+      "epoch": 85.46,
+      "grad_norm": 3.6351675987243652,
+      "learning_rate": 1.4541353383458648e-06,
+      "loss": 0.1576,
+      "step": 56830
+    },
+    {
+      "epoch": 85.47,
+      "grad_norm": 5.808753490447998,
+      "learning_rate": 1.4526315789473685e-06,
+      "loss": 0.1761,
+      "step": 56840
+    },
+    {
+      "epoch": 85.49,
+      "grad_norm": 4.585028171539307,
+      "learning_rate": 1.4511278195488722e-06,
+      "loss": 0.1982,
+      "step": 56850
+    },
+    {
+      "epoch": 85.5,
+      "grad_norm": 5.568889617919922,
+      "learning_rate": 1.4496240601503762e-06,
+      "loss": 0.1516,
+      "step": 56860
+    },
+    {
+      "epoch": 85.52,
+      "grad_norm": 9.540451049804688,
+      "learning_rate": 1.44812030075188e-06,
+      "loss": 0.1922,
+      "step": 56870
+    },
+    {
+      "epoch": 85.53,
+      "grad_norm": 3.1969432830810547,
+      "learning_rate": 1.4466165413533836e-06,
+      "loss": 0.1595,
+      "step": 56880
+    },
+    {
+      "epoch": 85.55,
+      "grad_norm": 5.823395252227783,
+      "learning_rate": 1.4451127819548874e-06,
+      "loss": 0.2047,
+      "step": 56890
+    },
+    {
+      "epoch": 85.56,
+      "grad_norm": 5.089601039886475,
+      "learning_rate": 1.4436090225563911e-06,
+      "loss": 0.2095,
+      "step": 56900
+    },
+    {
+      "epoch": 85.58,
+      "grad_norm": 6.277038097381592,
+      "learning_rate": 1.442105263157895e-06,
+      "loss": 0.154,
+      "step": 56910
+    },
+    {
+      "epoch": 85.59,
+      "grad_norm": 4.296693801879883,
+      "learning_rate": 1.4406015037593988e-06,
+      "loss": 0.1405,
+      "step": 56920
+    },
+    {
+      "epoch": 85.61,
+      "grad_norm": 4.177289962768555,
+      "learning_rate": 1.4390977443609025e-06,
+      "loss": 0.1869,
+      "step": 56930
+    },
+    {
+      "epoch": 85.62,
+      "grad_norm": 5.647763729095459,
+      "learning_rate": 1.4375939849624062e-06,
+      "loss": 0.2275,
+      "step": 56940
+    },
+    {
+      "epoch": 85.64,
+      "grad_norm": 6.22650146484375,
+      "learning_rate": 1.4360902255639098e-06,
+      "loss": 0.1854,
+      "step": 56950
+    },
+    {
+      "epoch": 85.65,
+      "grad_norm": 1.4636141061782837,
+      "learning_rate": 1.4345864661654135e-06,
+      "loss": 0.1481,
+      "step": 56960
+    },
+    {
+      "epoch": 85.67,
+      "grad_norm": 5.929630279541016,
+      "learning_rate": 1.4330827067669172e-06,
+      "loss": 0.1671,
+      "step": 56970
+    },
+    {
+      "epoch": 85.68,
+      "grad_norm": 6.411832809448242,
+      "learning_rate": 1.4315789473684212e-06,
+      "loss": 0.1598,
+      "step": 56980
+    },
+    {
+      "epoch": 85.7,
+      "grad_norm": 4.0760602951049805,
+      "learning_rate": 1.430075187969925e-06,
+      "loss": 0.1493,
+      "step": 56990
+    },
+    {
+      "epoch": 85.71,
+      "grad_norm": 5.515470027923584,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 0.133,
+      "step": 57000
+    },
+    {
+      "epoch": 85.73,
+      "grad_norm": 4.49883508682251,
+      "learning_rate": 1.4270676691729324e-06,
+      "loss": 0.2057,
+      "step": 57010
+    },
+    {
+      "epoch": 85.74,
+      "grad_norm": 6.55890417098999,
+      "learning_rate": 1.425563909774436e-06,
+      "loss": 0.1498,
+      "step": 57020
+    },
+    {
+      "epoch": 85.76,
+      "grad_norm": 6.194241523742676,
+      "learning_rate": 1.4240601503759398e-06,
+      "loss": 0.1963,
+      "step": 57030
+    },
+    {
+      "epoch": 85.77,
+      "grad_norm": 5.173956394195557,
+      "learning_rate": 1.4225563909774438e-06,
+      "loss": 0.1443,
+      "step": 57040
+    },
+    {
+      "epoch": 85.79,
+      "grad_norm": 4.395569801330566,
+      "learning_rate": 1.4210526315789475e-06,
+      "loss": 0.1895,
+      "step": 57050
+    },
+    {
+      "epoch": 85.8,
+      "grad_norm": 6.59354305267334,
+      "learning_rate": 1.4195488721804512e-06,
+      "loss": 0.1671,
+      "step": 57060
+    },
+    {
+      "epoch": 85.82,
+      "grad_norm": 4.2721028327941895,
+      "learning_rate": 1.418045112781955e-06,
+      "loss": 0.1004,
+      "step": 57070
+    },
+    {
+      "epoch": 85.83,
+      "grad_norm": 6.9995436668396,
+      "learning_rate": 1.4165413533834587e-06,
+      "loss": 0.1782,
+      "step": 57080
+    },
+    {
+      "epoch": 85.85,
+      "grad_norm": 4.45127010345459,
+      "learning_rate": 1.4150375939849626e-06,
+      "loss": 0.2011,
+      "step": 57090
+    },
+    {
+      "epoch": 85.86,
+      "grad_norm": 7.801600456237793,
+      "learning_rate": 1.4135338345864664e-06,
+      "loss": 0.1099,
+      "step": 57100
+    },
+    {
+      "epoch": 85.88,
+      "grad_norm": 5.012579917907715,
+      "learning_rate": 1.41203007518797e-06,
+      "loss": 0.1656,
+      "step": 57110
+    },
+    {
+      "epoch": 85.89,
+      "grad_norm": 6.178292751312256,
+      "learning_rate": 1.4105263157894738e-06,
+      "loss": 0.1934,
+      "step": 57120
+    },
+    {
+      "epoch": 85.91,
+      "grad_norm": 3.175626039505005,
+      "learning_rate": 1.4090225563909776e-06,
+      "loss": 0.1558,
+      "step": 57130
+    },
+    {
+      "epoch": 85.92,
+      "grad_norm": 7.903815746307373,
+      "learning_rate": 1.4075187969924815e-06,
+      "loss": 0.1989,
+      "step": 57140
+    },
+    {
+      "epoch": 85.94,
+      "grad_norm": 5.2723774909973145,
+      "learning_rate": 1.4060150375939852e-06,
+      "loss": 0.1617,
+      "step": 57150
+    },
+    {
+      "epoch": 85.95,
+      "grad_norm": 7.170572280883789,
+      "learning_rate": 1.404511278195489e-06,
+      "loss": 0.1336,
+      "step": 57160
+    },
+    {
+      "epoch": 85.97,
+      "grad_norm": 8.153802871704102,
+      "learning_rate": 1.4030075187969927e-06,
+      "loss": 0.1318,
+      "step": 57170
+    },
+    {
+      "epoch": 85.98,
+      "grad_norm": 1.0930061340332031,
+      "learning_rate": 1.4015037593984964e-06,
+      "loss": 0.1619,
+      "step": 57180
+    },
+    {
+      "epoch": 86.0,
+      "grad_norm": 1.082709789276123,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 0.199,
+      "step": 57190
+    },
+    {
+      "epoch": 86.0,
+      "eval_accuracy": 0.9303,
+      "eval_loss": 0.33443817496299744,
+      "eval_runtime": 84.4787,
+      "eval_samples_per_second": 118.373,
+      "eval_steps_per_second": 0.473,
+      "step": 57190
+    },
+    {
+      "epoch": 86.02,
+      "grad_norm": 4.792495250701904,
+      "learning_rate": 1.3984962406015037e-06,
+      "loss": 0.1635,
+      "step": 57200
+    },
+    {
+      "epoch": 86.03,
+      "grad_norm": 6.919180870056152,
+      "learning_rate": 1.3969924812030076e-06,
+      "loss": 0.1994,
+      "step": 57210
+    },
+    {
+      "epoch": 86.05,
+      "grad_norm": 3.3340089321136475,
+      "learning_rate": 1.3954887218045113e-06,
+      "loss": 0.1505,
+      "step": 57220
+    },
+    {
+      "epoch": 86.06,
+      "grad_norm": 4.591991424560547,
+      "learning_rate": 1.393984962406015e-06,
+      "loss": 0.1473,
+      "step": 57230
+    },
+    {
+      "epoch": 86.08,
+      "grad_norm": 5.2993550300598145,
+      "learning_rate": 1.3924812030075188e-06,
+      "loss": 0.1736,
+      "step": 57240
+    },
+    {
+      "epoch": 86.09,
+      "grad_norm": 3.2715325355529785,
+      "learning_rate": 1.3909774436090225e-06,
+      "loss": 0.1788,
+      "step": 57250
+    },
+    {
+      "epoch": 86.11,
+      "grad_norm": 5.598659515380859,
+      "learning_rate": 1.3894736842105263e-06,
+      "loss": 0.1715,
+      "step": 57260
+    },
+    {
+      "epoch": 86.12,
+      "grad_norm": 5.6875,
+      "learning_rate": 1.3879699248120302e-06,
+      "loss": 0.1439,
+      "step": 57270
+    },
+    {
+      "epoch": 86.14,
+      "grad_norm": 3.4048891067504883,
+      "learning_rate": 1.386466165413534e-06,
+      "loss": 0.1514,
+      "step": 57280
+    },
+    {
+      "epoch": 86.15,
+      "grad_norm": 6.038202285766602,
+      "learning_rate": 1.3849624060150377e-06,
+      "loss": 0.1718,
+      "step": 57290
+    },
+    {
+      "epoch": 86.17,
+      "grad_norm": 4.095704078674316,
+      "learning_rate": 1.3834586466165414e-06,
+      "loss": 0.159,
+      "step": 57300
+    },
+    {
+      "epoch": 86.18,
+      "grad_norm": 6.841707706451416,
+      "learning_rate": 1.3819548872180451e-06,
+      "loss": 0.1779,
+      "step": 57310
+    },
+    {
+      "epoch": 86.2,
+      "grad_norm": 2.8215293884277344,
+      "learning_rate": 1.380451127819549e-06,
+      "loss": 0.1868,
+      "step": 57320
+    },
+    {
+      "epoch": 86.21,
+      "grad_norm": 6.776797294616699,
+      "learning_rate": 1.3789473684210528e-06,
+      "loss": 0.183,
+      "step": 57330
+    },
+    {
+      "epoch": 86.23,
+      "grad_norm": 9.345000267028809,
+      "learning_rate": 1.3774436090225565e-06,
+      "loss": 0.1724,
+      "step": 57340
+    },
+    {
+      "epoch": 86.24,
+      "grad_norm": 5.53676700592041,
+      "learning_rate": 1.3759398496240603e-06,
+      "loss": 0.186,
+      "step": 57350
+    },
+    {
+      "epoch": 86.26,
+      "grad_norm": 6.88644552230835,
+      "learning_rate": 1.374436090225564e-06,
+      "loss": 0.1801,
+      "step": 57360
+    },
+    {
+      "epoch": 86.27,
+      "grad_norm": 10.067435264587402,
+      "learning_rate": 1.372932330827068e-06,
+      "loss": 0.1977,
+      "step": 57370
+    },
+    {
+      "epoch": 86.29,
+      "grad_norm": 3.2209675312042236,
+      "learning_rate": 1.3714285714285717e-06,
+      "loss": 0.1715,
+      "step": 57380
+    },
+    {
+      "epoch": 86.3,
+      "grad_norm": 5.24599027633667,
+      "learning_rate": 1.3699248120300754e-06,
+      "loss": 0.1733,
+      "step": 57390
+    },
+    {
+      "epoch": 86.32,
+      "grad_norm": 4.074387550354004,
+      "learning_rate": 1.3684210526315791e-06,
+      "loss": 0.1588,
+      "step": 57400
+    },
+    {
+      "epoch": 86.33,
+      "grad_norm": 3.2896995544433594,
+      "learning_rate": 1.3669172932330829e-06,
+      "loss": 0.1384,
+      "step": 57410
+    },
+    {
+      "epoch": 86.35,
+      "grad_norm": 6.851515293121338,
+      "learning_rate": 1.3654135338345866e-06,
+      "loss": 0.2287,
+      "step": 57420
+    },
+    {
+      "epoch": 86.36,
+      "grad_norm": 3.76686954498291,
+      "learning_rate": 1.3639097744360905e-06,
+      "loss": 0.1623,
+      "step": 57430
+    },
+    {
+      "epoch": 86.38,
+      "grad_norm": 7.297066688537598,
+      "learning_rate": 1.3624060150375943e-06,
+      "loss": 0.1177,
+      "step": 57440
+    },
+    {
+      "epoch": 86.39,
+      "grad_norm": 4.633073329925537,
+      "learning_rate": 1.3609022556390978e-06,
+      "loss": 0.1634,
+      "step": 57450
+    },
+    {
+      "epoch": 86.41,
+      "grad_norm": 5.141488552093506,
+      "learning_rate": 1.3593984962406015e-06,
+      "loss": 0.2001,
+      "step": 57460
+    },
+    {
+      "epoch": 86.42,
+      "grad_norm": 4.2999467849731445,
+      "learning_rate": 1.3578947368421052e-06,
+      "loss": 0.1434,
+      "step": 57470
+    },
+    {
+      "epoch": 86.44,
+      "grad_norm": 1.721150517463684,
+      "learning_rate": 1.356390977443609e-06,
+      "loss": 0.2166,
+      "step": 57480
+    },
+    {
+      "epoch": 86.45,
+      "grad_norm": 6.083622455596924,
+      "learning_rate": 1.3548872180451127e-06,
+      "loss": 0.1603,
+      "step": 57490
+    },
+    {
+      "epoch": 86.47,
+      "grad_norm": 8.409628868103027,
+      "learning_rate": 1.3533834586466167e-06,
+      "loss": 0.1633,
+      "step": 57500
+    },
+    {
+      "epoch": 86.48,
+      "grad_norm": 5.203959941864014,
+      "learning_rate": 1.3518796992481204e-06,
+      "loss": 0.191,
+      "step": 57510
+    },
+    {
+      "epoch": 86.5,
+      "grad_norm": 3.071166515350342,
+      "learning_rate": 1.3503759398496241e-06,
+      "loss": 0.1879,
+      "step": 57520
+    },
+    {
+      "epoch": 86.51,
+      "grad_norm": 3.929776191711426,
+      "learning_rate": 1.3488721804511278e-06,
+      "loss": 0.1791,
+      "step": 57530
+    },
+    {
+      "epoch": 86.53,
+      "grad_norm": 4.9922709465026855,
+      "learning_rate": 1.3473684210526316e-06,
+      "loss": 0.1235,
+      "step": 57540
+    },
+    {
+      "epoch": 86.54,
+      "grad_norm": 3.8462400436401367,
+      "learning_rate": 1.3458646616541355e-06,
+      "loss": 0.1654,
+      "step": 57550
+    },
+    {
+      "epoch": 86.56,
+      "grad_norm": 3.1140220165252686,
+      "learning_rate": 1.3443609022556392e-06,
+      "loss": 0.2052,
+      "step": 57560
+    },
+    {
+      "epoch": 86.57,
+      "grad_norm": 8.724045753479004,
+      "learning_rate": 1.342857142857143e-06,
+      "loss": 0.1703,
+      "step": 57570
+    },
+    {
+      "epoch": 86.59,
+      "grad_norm": 4.744037628173828,
+      "learning_rate": 1.3413533834586467e-06,
+      "loss": 0.2456,
+      "step": 57580
+    },
+    {
+      "epoch": 86.6,
+      "grad_norm": 2.585832118988037,
+      "learning_rate": 1.3398496240601504e-06,
+      "loss": 0.1464,
+      "step": 57590
+    },
+    {
+      "epoch": 86.62,
+      "grad_norm": 5.518332004547119,
+      "learning_rate": 1.3383458646616544e-06,
+      "loss": 0.1899,
+      "step": 57600
+    },
+    {
+      "epoch": 86.63,
+      "grad_norm": 12.576923370361328,
+      "learning_rate": 1.3368421052631581e-06,
+      "loss": 0.1964,
+      "step": 57610
+    },
+    {
+      "epoch": 86.65,
+      "grad_norm": 2.010282516479492,
+      "learning_rate": 1.3353383458646618e-06,
+      "loss": 0.1554,
+      "step": 57620
+    },
+    {
+      "epoch": 86.66,
+      "grad_norm": 2.587585687637329,
+      "learning_rate": 1.3338345864661656e-06,
+      "loss": 0.1413,
+      "step": 57630
+    },
+    {
+      "epoch": 86.68,
+      "grad_norm": 4.673181533813477,
+      "learning_rate": 1.3323308270676693e-06,
+      "loss": 0.209,
+      "step": 57640
+    },
+    {
+      "epoch": 86.69,
+      "grad_norm": 3.9027647972106934,
+      "learning_rate": 1.330827067669173e-06,
+      "loss": 0.176,
+      "step": 57650
+    },
+    {
+      "epoch": 86.71,
+      "grad_norm": 7.965885639190674,
+      "learning_rate": 1.329323308270677e-06,
+      "loss": 0.1014,
+      "step": 57660
+    },
+    {
+      "epoch": 86.72,
+      "grad_norm": 4.3658881187438965,
+      "learning_rate": 1.3278195488721807e-06,
+      "loss": 0.1876,
+      "step": 57670
+    },
+    {
+      "epoch": 86.74,
+      "grad_norm": 8.340523719787598,
+      "learning_rate": 1.3263157894736844e-06,
+      "loss": 0.1785,
+      "step": 57680
+    },
+    {
+      "epoch": 86.75,
+      "grad_norm": 4.654315948486328,
+      "learning_rate": 1.3248120300751882e-06,
+      "loss": 0.1522,
+      "step": 57690
+    },
+    {
+      "epoch": 86.77,
+      "grad_norm": 7.557748317718506,
+      "learning_rate": 1.3233082706766917e-06,
+      "loss": 0.2559,
+      "step": 57700
+    },
+    {
+      "epoch": 86.78,
+      "grad_norm": 5.0106329917907715,
+      "learning_rate": 1.3218045112781954e-06,
+      "loss": 0.1944,
+      "step": 57710
+    },
+    {
+      "epoch": 86.8,
+      "grad_norm": 5.76795768737793,
+      "learning_rate": 1.3203007518796992e-06,
+      "loss": 0.1632,
+      "step": 57720
+    },
+    {
+      "epoch": 86.81,
+      "grad_norm": 4.331972599029541,
+      "learning_rate": 1.318796992481203e-06,
+      "loss": 0.1461,
+      "step": 57730
+    },
+    {
+      "epoch": 86.83,
+      "grad_norm": 5.625306606292725,
+      "learning_rate": 1.3172932330827068e-06,
+      "loss": 0.0988,
+      "step": 57740
+    },
+    {
+      "epoch": 86.84,
+      "grad_norm": 3.771822452545166,
+      "learning_rate": 1.3157894736842106e-06,
+      "loss": 0.139,
+      "step": 57750
+    },
+    {
+      "epoch": 86.86,
+      "grad_norm": 2.988506555557251,
+      "learning_rate": 1.3142857142857143e-06,
+      "loss": 0.1482,
+      "step": 57760
+    },
+    {
+      "epoch": 86.87,
+      "grad_norm": 6.377354145050049,
+      "learning_rate": 1.312781954887218e-06,
+      "loss": 0.1966,
+      "step": 57770
+    },
+    {
+      "epoch": 86.89,
+      "grad_norm": 26.265287399291992,
+      "learning_rate": 1.311278195488722e-06,
+      "loss": 0.2471,
+      "step": 57780
+    },
+    {
+      "epoch": 86.9,
+      "grad_norm": 7.3494062423706055,
+      "learning_rate": 1.3097744360902257e-06,
+      "loss": 0.1744,
+      "step": 57790
+    },
+    {
+      "epoch": 86.92,
+      "grad_norm": 6.284552574157715,
+      "learning_rate": 1.3082706766917294e-06,
+      "loss": 0.1936,
+      "step": 57800
+    },
+    {
+      "epoch": 86.93,
+      "grad_norm": 2.87434720993042,
+      "learning_rate": 1.3067669172932332e-06,
+      "loss": 0.1445,
+      "step": 57810
+    },
+    {
+      "epoch": 86.95,
+      "grad_norm": 3.9225971698760986,
+      "learning_rate": 1.3052631578947369e-06,
+      "loss": 0.1282,
+      "step": 57820
+    },
+    {
+      "epoch": 86.96,
+      "grad_norm": 6.310343265533447,
+      "learning_rate": 1.3037593984962408e-06,
+      "loss": 0.1504,
+      "step": 57830
+    },
+    {
+      "epoch": 86.98,
+      "grad_norm": 2.322134494781494,
+      "learning_rate": 1.3022556390977446e-06,
+      "loss": 0.1774,
+      "step": 57840
+    },
+    {
+      "epoch": 86.99,
+      "grad_norm": 6.937019348144531,
+      "learning_rate": 1.3007518796992483e-06,
+      "loss": 0.2013,
+      "step": 57850
+    },
+    {
+      "epoch": 87.0,
+      "eval_accuracy": 0.9294,
+      "eval_loss": 0.33600765466690063,
+      "eval_runtime": 84.8466,
+      "eval_samples_per_second": 117.86,
+      "eval_steps_per_second": 0.471,
+      "step": 57855
+    },
+    {
+      "epoch": 87.01,
+      "grad_norm": 6.285330772399902,
+      "learning_rate": 1.299248120300752e-06,
+      "loss": 0.1726,
+      "step": 57860
+    },
+    {
+      "epoch": 87.02,
+      "grad_norm": 5.90061616897583,
+      "learning_rate": 1.2977443609022557e-06,
+      "loss": 0.1748,
+      "step": 57870
+    },
+    {
+      "epoch": 87.04,
+      "grad_norm": 4.586902141571045,
+      "learning_rate": 1.2962406015037595e-06,
+      "loss": 0.1611,
+      "step": 57880
+    },
+    {
+      "epoch": 87.05,
+      "grad_norm": 6.66023063659668,
+      "learning_rate": 1.2947368421052634e-06,
+      "loss": 0.1986,
+      "step": 57890
+    },
+    {
+      "epoch": 87.07,
+      "grad_norm": 5.385473728179932,
+      "learning_rate": 1.2932330827067672e-06,
+      "loss": 0.1067,
+      "step": 57900
+    },
+    {
+      "epoch": 87.08,
+      "grad_norm": 6.819510459899902,
+      "learning_rate": 1.2917293233082709e-06,
+      "loss": 0.2073,
+      "step": 57910
+    },
+    {
+      "epoch": 87.1,
+      "grad_norm": 6.095653533935547,
+      "learning_rate": 1.2902255639097746e-06,
+      "loss": 0.182,
+      "step": 57920
+    },
+    {
+      "epoch": 87.11,
+      "grad_norm": 5.213226318359375,
+      "learning_rate": 1.2887218045112783e-06,
+      "loss": 0.1669,
+      "step": 57930
+    },
+    {
+      "epoch": 87.13,
+      "grad_norm": 6.771183013916016,
+      "learning_rate": 1.2872180451127819e-06,
+      "loss": 0.1516,
+      "step": 57940
+    },
+    {
+      "epoch": 87.14,
+      "grad_norm": 4.856118202209473,
+      "learning_rate": 1.2857142857142856e-06,
+      "loss": 0.1819,
+      "step": 57950
+    },
+    {
+      "epoch": 87.16,
+      "grad_norm": 4.535681247711182,
+      "learning_rate": 1.2842105263157895e-06,
+      "loss": 0.14,
+      "step": 57960
+    },
+    {
+      "epoch": 87.17,
+      "grad_norm": 4.036566734313965,
+      "learning_rate": 1.2827067669172933e-06,
+      "loss": 0.136,
+      "step": 57970
+    },
+    {
+      "epoch": 87.19,
+      "grad_norm": 4.947079181671143,
+      "learning_rate": 1.281203007518797e-06,
+      "loss": 0.1617,
+      "step": 57980
+    },
+    {
+      "epoch": 87.2,
+      "grad_norm": 2.912419557571411,
+      "learning_rate": 1.2796992481203007e-06,
+      "loss": 0.1863,
+      "step": 57990
+    },
+    {
+      "epoch": 87.22,
+      "grad_norm": 5.682669639587402,
+      "learning_rate": 1.2781954887218045e-06,
+      "loss": 0.1317,
+      "step": 58000
+    },
+    {
+      "epoch": 87.23,
+      "grad_norm": 5.527998447418213,
+      "learning_rate": 1.2766917293233084e-06,
+      "loss": 0.1618,
+      "step": 58010
+    },
+    {
+      "epoch": 87.25,
+      "grad_norm": 6.747208595275879,
+      "learning_rate": 1.2751879699248121e-06,
+      "loss": 0.1686,
+      "step": 58020
+    },
+    {
+      "epoch": 87.26,
+      "grad_norm": 5.503549098968506,
+      "learning_rate": 1.2736842105263159e-06,
+      "loss": 0.1881,
+      "step": 58030
+    },
+    {
+      "epoch": 87.28,
+      "grad_norm": 5.4713029861450195,
+      "learning_rate": 1.2721804511278196e-06,
+      "loss": 0.184,
+      "step": 58040
+    },
+    {
+      "epoch": 87.29,
+      "grad_norm": 2.895097494125366,
+      "learning_rate": 1.2706766917293233e-06,
+      "loss": 0.1819,
+      "step": 58050
+    },
+    {
+      "epoch": 87.31,
+      "grad_norm": 6.448250770568848,
+      "learning_rate": 1.2691729323308273e-06,
+      "loss": 0.1763,
+      "step": 58060
+    },
+    {
+      "epoch": 87.32,
+      "grad_norm": 3.6069977283477783,
+      "learning_rate": 1.267669172932331e-06,
+      "loss": 0.1478,
+      "step": 58070
+    },
+    {
+      "epoch": 87.34,
+      "grad_norm": 13.077876091003418,
+      "learning_rate": 1.2661654135338347e-06,
+      "loss": 0.1803,
+      "step": 58080
+    },
+    {
+      "epoch": 87.35,
+      "grad_norm": 2.1468687057495117,
+      "learning_rate": 1.2646616541353385e-06,
+      "loss": 0.1337,
+      "step": 58090
+    },
+    {
+      "epoch": 87.37,
+      "grad_norm": 4.480571269989014,
+      "learning_rate": 1.2631578947368422e-06,
+      "loss": 0.1444,
+      "step": 58100
+    },
+    {
+      "epoch": 87.38,
+      "grad_norm": 8.891602516174316,
+      "learning_rate": 1.261654135338346e-06,
+      "loss": 0.1664,
+      "step": 58110
+    },
+    {
+      "epoch": 87.4,
+      "grad_norm": 8.419784545898438,
+      "learning_rate": 1.2601503759398499e-06,
+      "loss": 0.1594,
+      "step": 58120
+    },
+    {
+      "epoch": 87.41,
+      "grad_norm": 3.2222964763641357,
+      "learning_rate": 1.2586466165413536e-06,
+      "loss": 0.1197,
+      "step": 58130
+    },
+    {
+      "epoch": 87.43,
+      "grad_norm": 6.81358528137207,
+      "learning_rate": 1.2571428571428573e-06,
+      "loss": 0.1934,
+      "step": 58140
+    },
+    {
+      "epoch": 87.44,
+      "grad_norm": 8.876537322998047,
+      "learning_rate": 1.255639097744361e-06,
+      "loss": 0.1814,
+      "step": 58150
+    },
+    {
+      "epoch": 87.46,
+      "grad_norm": 7.15813684463501,
+      "learning_rate": 1.2541353383458648e-06,
+      "loss": 0.1791,
+      "step": 58160
+    },
+    {
+      "epoch": 87.47,
+      "grad_norm": 7.933776378631592,
+      "learning_rate": 1.2526315789473687e-06,
+      "loss": 0.1833,
+      "step": 58170
+    },
+    {
+      "epoch": 87.49,
+      "grad_norm": 4.0624284744262695,
+      "learning_rate": 1.2511278195488725e-06,
+      "loss": 0.1686,
+      "step": 58180
+    },
+    {
+      "epoch": 87.5,
+      "grad_norm": 6.974576950073242,
+      "learning_rate": 1.249624060150376e-06,
+      "loss": 0.151,
+      "step": 58190
+    },
+    {
+      "epoch": 87.52,
+      "grad_norm": 5.813340663909912,
+      "learning_rate": 1.2481203007518797e-06,
+      "loss": 0.158,
+      "step": 58200
+    },
+    {
+      "epoch": 87.53,
+      "grad_norm": 4.3520073890686035,
+      "learning_rate": 1.2466165413533837e-06,
+      "loss": 0.1725,
+      "step": 58210
+    },
+    {
+      "epoch": 87.55,
+      "grad_norm": 5.577434539794922,
+      "learning_rate": 1.2451127819548874e-06,
+      "loss": 0.178,
+      "step": 58220
+    },
+    {
+      "epoch": 87.56,
+      "grad_norm": 1.6522976160049438,
+      "learning_rate": 1.2436090225563911e-06,
+      "loss": 0.17,
+      "step": 58230
+    },
+    {
+      "epoch": 87.58,
+      "grad_norm": 8.239200592041016,
+      "learning_rate": 1.2421052631578948e-06,
+      "loss": 0.1651,
+      "step": 58240
+    },
+    {
+      "epoch": 87.59,
+      "grad_norm": 3.1411798000335693,
+      "learning_rate": 1.2406015037593986e-06,
+      "loss": 0.1839,
+      "step": 58250
+    },
+    {
+      "epoch": 87.61,
+      "grad_norm": 8.265312194824219,
+      "learning_rate": 1.2390977443609023e-06,
+      "loss": 0.1692,
+      "step": 58260
+    },
+    {
+      "epoch": 87.62,
+      "grad_norm": 7.520175457000732,
+      "learning_rate": 1.237593984962406e-06,
+      "loss": 0.1909,
+      "step": 58270
+    },
+    {
+      "epoch": 87.64,
+      "grad_norm": 9.53011703491211,
+      "learning_rate": 1.2360902255639098e-06,
+      "loss": 0.2348,
+      "step": 58280
+    },
+    {
+      "epoch": 87.65,
+      "grad_norm": 5.072795391082764,
+      "learning_rate": 1.2345864661654137e-06,
+      "loss": 0.1431,
+      "step": 58290
+    },
+    {
+      "epoch": 87.67,
+      "grad_norm": 4.908288955688477,
+      "learning_rate": 1.2330827067669174e-06,
+      "loss": 0.2498,
+      "step": 58300
+    },
+    {
+      "epoch": 87.68,
+      "grad_norm": 5.3725762367248535,
+      "learning_rate": 1.2315789473684212e-06,
+      "loss": 0.1617,
+      "step": 58310
+    },
+    {
+      "epoch": 87.7,
+      "grad_norm": 5.9502034187316895,
+      "learning_rate": 1.230075187969925e-06,
+      "loss": 0.0983,
+      "step": 58320
+    },
+    {
+      "epoch": 87.71,
+      "grad_norm": 7.810244083404541,
+      "learning_rate": 1.2285714285714286e-06,
+      "loss": 0.1523,
+      "step": 58330
+    },
+    {
+      "epoch": 87.73,
+      "grad_norm": 3.383789539337158,
+      "learning_rate": 1.2270676691729324e-06,
+      "loss": 0.1778,
+      "step": 58340
+    },
+    {
+      "epoch": 87.74,
+      "grad_norm": 3.804659366607666,
+      "learning_rate": 1.2255639097744363e-06,
+      "loss": 0.1638,
+      "step": 58350
+    },
+    {
+      "epoch": 87.76,
+      "grad_norm": 6.533144950866699,
+      "learning_rate": 1.22406015037594e-06,
+      "loss": 0.1497,
+      "step": 58360
+    },
+    {
+      "epoch": 87.77,
+      "grad_norm": 5.189228057861328,
+      "learning_rate": 1.2225563909774438e-06,
+      "loss": 0.15,
+      "step": 58370
+    },
+    {
+      "epoch": 87.79,
+      "grad_norm": 7.200681209564209,
+      "learning_rate": 1.2210526315789475e-06,
+      "loss": 0.151,
+      "step": 58380
+    },
+    {
+      "epoch": 87.8,
+      "grad_norm": 5.005568027496338,
+      "learning_rate": 1.2195488721804512e-06,
+      "loss": 0.1967,
+      "step": 58390
+    },
+    {
+      "epoch": 87.82,
+      "grad_norm": 7.443274974822998,
+      "learning_rate": 1.218045112781955e-06,
+      "loss": 0.2047,
+      "step": 58400
+    },
+    {
+      "epoch": 87.83,
+      "grad_norm": 8.02253532409668,
+      "learning_rate": 1.2165413533834587e-06,
+      "loss": 0.1725,
+      "step": 58410
+    },
+    {
+      "epoch": 87.85,
+      "grad_norm": 4.530572414398193,
+      "learning_rate": 1.2150375939849624e-06,
+      "loss": 0.2269,
+      "step": 58420
+    },
+    {
+      "epoch": 87.86,
+      "grad_norm": 6.738208293914795,
+      "learning_rate": 1.2135338345864662e-06,
+      "loss": 0.1799,
+      "step": 58430
+    },
+    {
+      "epoch": 87.88,
+      "grad_norm": 4.552396297454834,
+      "learning_rate": 1.21203007518797e-06,
+      "loss": 0.1505,
+      "step": 58440
+    },
+    {
+      "epoch": 87.89,
+      "grad_norm": 3.540454387664795,
+      "learning_rate": 1.2105263157894738e-06,
+      "loss": 0.1816,
+      "step": 58450
+    },
+    {
+      "epoch": 87.91,
+      "grad_norm": 4.805243015289307,
+      "learning_rate": 1.2090225563909776e-06,
+      "loss": 0.1457,
+      "step": 58460
+    },
+    {
+      "epoch": 87.92,
+      "grad_norm": 6.110500812530518,
+      "learning_rate": 1.2075187969924813e-06,
+      "loss": 0.1308,
+      "step": 58470
+    },
+    {
+      "epoch": 87.94,
+      "grad_norm": 2.2697036266326904,
+      "learning_rate": 1.206015037593985e-06,
+      "loss": 0.1508,
+      "step": 58480
+    },
+    {
+      "epoch": 87.95,
+      "grad_norm": 1.7307939529418945,
+      "learning_rate": 1.204511278195489e-06,
+      "loss": 0.1436,
+      "step": 58490
+    },
+    {
+      "epoch": 87.97,
+      "grad_norm": 4.120843410491943,
+      "learning_rate": 1.2030075187969925e-06,
+      "loss": 0.1766,
+      "step": 58500
+    },
+    {
+      "epoch": 87.98,
+      "grad_norm": 5.4482927322387695,
+      "learning_rate": 1.2015037593984962e-06,
+      "loss": 0.2227,
+      "step": 58510
+    },
+    {
+      "epoch": 88.0,
+      "grad_norm": 0.018793661147356033,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 0.1495,
+      "step": 58520
+    },
+    {
+      "epoch": 88.0,
+      "eval_accuracy": 0.9307,
+      "eval_loss": 0.33708590269088745,
+      "eval_runtime": 84.8059,
+      "eval_samples_per_second": 117.916,
+      "eval_steps_per_second": 0.472,
+      "step": 58520
+    },
+    {
+      "epoch": 88.02,
+      "grad_norm": 4.496594429016113,
+      "learning_rate": 1.1984962406015039e-06,
+      "loss": 0.2237,
+      "step": 58530
+    },
+    {
+      "epoch": 88.03,
+      "grad_norm": 4.44071102142334,
+      "learning_rate": 1.1969924812030076e-06,
+      "loss": 0.2014,
+      "step": 58540
+    },
+    {
+      "epoch": 88.05,
+      "grad_norm": 4.390169620513916,
+      "learning_rate": 1.1954887218045113e-06,
+      "loss": 0.1108,
+      "step": 58550
+    },
+    {
+      "epoch": 88.06,
+      "grad_norm": 5.913539409637451,
+      "learning_rate": 1.193984962406015e-06,
+      "loss": 0.1861,
+      "step": 58560
+    },
+    {
+      "epoch": 88.08,
+      "grad_norm": 4.491770267486572,
+      "learning_rate": 1.1924812030075188e-06,
+      "loss": 0.1935,
+      "step": 58570
+    },
+    {
+      "epoch": 88.09,
+      "grad_norm": 5.476345539093018,
+      "learning_rate": 1.1909774436090228e-06,
+      "loss": 0.1692,
+      "step": 58580
+    },
+    {
+      "epoch": 88.11,
+      "grad_norm": 5.3462347984313965,
+      "learning_rate": 1.1894736842105265e-06,
+      "loss": 0.1797,
+      "step": 58590
+    },
+    {
+      "epoch": 88.12,
+      "grad_norm": 2.1195619106292725,
+      "learning_rate": 1.1879699248120302e-06,
+      "loss": 0.1823,
+      "step": 58600
+    },
+    {
+      "epoch": 88.14,
+      "grad_norm": 8.027178764343262,
+      "learning_rate": 1.186466165413534e-06,
+      "loss": 0.2147,
+      "step": 58610
+    },
+    {
+      "epoch": 88.15,
+      "grad_norm": 6.265287399291992,
+      "learning_rate": 1.1849624060150377e-06,
+      "loss": 0.1477,
+      "step": 58620
+    },
+    {
+      "epoch": 88.17,
+      "grad_norm": 4.6563191413879395,
+      "learning_rate": 1.1834586466165414e-06,
+      "loss": 0.1371,
+      "step": 58630
+    },
+    {
+      "epoch": 88.18,
+      "grad_norm": 3.4797229766845703,
+      "learning_rate": 1.1819548872180451e-06,
+      "loss": 0.1688,
+      "step": 58640
+    },
+    {
+      "epoch": 88.2,
+      "grad_norm": 12.818514823913574,
+      "learning_rate": 1.1804511278195489e-06,
+      "loss": 0.1646,
+      "step": 58650
+    },
+    {
+      "epoch": 88.21,
+      "grad_norm": 6.959251880645752,
+      "learning_rate": 1.1789473684210526e-06,
+      "loss": 0.1892,
+      "step": 58660
+    },
+    {
+      "epoch": 88.23,
+      "grad_norm": 5.956885814666748,
+      "learning_rate": 1.1774436090225565e-06,
+      "loss": 0.191,
+      "step": 58670
+    },
+    {
+      "epoch": 88.24,
+      "grad_norm": 3.1932976245880127,
+      "learning_rate": 1.1759398496240603e-06,
+      "loss": 0.187,
+      "step": 58680
+    },
+    {
+      "epoch": 88.26,
+      "grad_norm": 2.2420973777770996,
+      "learning_rate": 1.174436090225564e-06,
+      "loss": 0.1762,
+      "step": 58690
+    },
+    {
+      "epoch": 88.27,
+      "grad_norm": 3.847198486328125,
+      "learning_rate": 1.1729323308270677e-06,
+      "loss": 0.1542,
+      "step": 58700
+    },
+    {
+      "epoch": 88.29,
+      "grad_norm": 7.925451755523682,
+      "learning_rate": 1.1714285714285715e-06,
+      "loss": 0.1671,
+      "step": 58710
+    },
+    {
+      "epoch": 88.3,
+      "grad_norm": 4.539743423461914,
+      "learning_rate": 1.1699248120300754e-06,
+      "loss": 0.1702,
+      "step": 58720
+    },
+    {
+      "epoch": 88.32,
+      "grad_norm": 0.4386990964412689,
+      "learning_rate": 1.1684210526315791e-06,
+      "loss": 0.1528,
+      "step": 58730
+    },
+    {
+      "epoch": 88.33,
+      "grad_norm": 0.1927802413702011,
+      "learning_rate": 1.1669172932330829e-06,
+      "loss": 0.1165,
+      "step": 58740
+    },
+    {
+      "epoch": 88.35,
+      "grad_norm": 4.060121059417725,
+      "learning_rate": 1.1654135338345866e-06,
+      "loss": 0.1751,
+      "step": 58750
+    },
+    {
+      "epoch": 88.36,
+      "grad_norm": 6.014774799346924,
+      "learning_rate": 1.1639097744360903e-06,
+      "loss": 0.1487,
+      "step": 58760
+    },
+    {
+      "epoch": 88.38,
+      "grad_norm": 2.9679887294769287,
+      "learning_rate": 1.162406015037594e-06,
+      "loss": 0.1832,
+      "step": 58770
+    },
+    {
+      "epoch": 88.39,
+      "grad_norm": 3.4474730491638184,
+      "learning_rate": 1.1609022556390978e-06,
+      "loss": 0.181,
+      "step": 58780
+    },
+    {
+      "epoch": 88.41,
+      "grad_norm": 5.39216947555542,
+      "learning_rate": 1.1593984962406015e-06,
+      "loss": 0.1964,
+      "step": 58790
+    },
+    {
+      "epoch": 88.42,
+      "grad_norm": 3.7178587913513184,
+      "learning_rate": 1.1578947368421053e-06,
+      "loss": 0.1406,
+      "step": 58800
+    },
+    {
+      "epoch": 88.44,
+      "grad_norm": 2.4193215370178223,
+      "learning_rate": 1.1563909774436092e-06,
+      "loss": 0.1928,
+      "step": 58810
+    },
+    {
+      "epoch": 88.45,
+      "grad_norm": 5.395383834838867,
+      "learning_rate": 1.154887218045113e-06,
+      "loss": 0.2289,
+      "step": 58820
+    },
+    {
+      "epoch": 88.47,
+      "grad_norm": 3.297912120819092,
+      "learning_rate": 1.1533834586466167e-06,
+      "loss": 0.1687,
+      "step": 58830
+    },
+    {
+      "epoch": 88.48,
+      "grad_norm": 6.842408180236816,
+      "learning_rate": 1.1518796992481204e-06,
+      "loss": 0.1959,
+      "step": 58840
+    },
+    {
+      "epoch": 88.5,
+      "grad_norm": 2.9778201580047607,
+      "learning_rate": 1.1503759398496241e-06,
+      "loss": 0.1847,
+      "step": 58850
+    },
+    {
+      "epoch": 88.51,
+      "grad_norm": 4.785979270935059,
+      "learning_rate": 1.148872180451128e-06,
+      "loss": 0.1789,
+      "step": 58860
+    },
+    {
+      "epoch": 88.53,
+      "grad_norm": 2.7606077194213867,
+      "learning_rate": 1.1473684210526316e-06,
+      "loss": 0.1995,
+      "step": 58870
+    },
+    {
+      "epoch": 88.54,
+      "grad_norm": 6.0019097328186035,
+      "learning_rate": 1.1458646616541353e-06,
+      "loss": 0.1979,
+      "step": 58880
+    },
+    {
+      "epoch": 88.56,
+      "grad_norm": 3.31396746635437,
+      "learning_rate": 1.144360902255639e-06,
+      "loss": 0.1191,
+      "step": 58890
+    },
+    {
+      "epoch": 88.57,
+      "grad_norm": 5.867440223693848,
+      "learning_rate": 1.142857142857143e-06,
+      "loss": 0.1948,
+      "step": 58900
+    },
+    {
+      "epoch": 88.59,
+      "grad_norm": 6.043757438659668,
+      "learning_rate": 1.1413533834586467e-06,
+      "loss": 0.145,
+      "step": 58910
+    },
+    {
+      "epoch": 88.6,
+      "grad_norm": 5.402288913726807,
+      "learning_rate": 1.1398496240601504e-06,
+      "loss": 0.166,
+      "step": 58920
+    },
+    {
+      "epoch": 88.62,
+      "grad_norm": 6.221014976501465,
+      "learning_rate": 1.1383458646616542e-06,
+      "loss": 0.2292,
+      "step": 58930
+    },
+    {
+      "epoch": 88.63,
+      "grad_norm": 6.4407572746276855,
+      "learning_rate": 1.136842105263158e-06,
+      "loss": 0.1624,
+      "step": 58940
+    },
+    {
+      "epoch": 88.65,
+      "grad_norm": 2.440732955932617,
+      "learning_rate": 1.1353383458646619e-06,
+      "loss": 0.1674,
+      "step": 58950
+    },
+    {
+      "epoch": 88.66,
+      "grad_norm": 4.425859451293945,
+      "learning_rate": 1.1338345864661656e-06,
+      "loss": 0.212,
+      "step": 58960
+    },
+    {
+      "epoch": 88.68,
+      "grad_norm": 5.0957818031311035,
+      "learning_rate": 1.1323308270676693e-06,
+      "loss": 0.2014,
+      "step": 58970
+    },
+    {
+      "epoch": 88.69,
+      "grad_norm": 3.4791855812072754,
+      "learning_rate": 1.130827067669173e-06,
+      "loss": 0.2379,
+      "step": 58980
+    },
+    {
+      "epoch": 88.71,
+      "grad_norm": 8.51900577545166,
+      "learning_rate": 1.1293233082706768e-06,
+      "loss": 0.1812,
+      "step": 58990
+    },
+    {
+      "epoch": 88.72,
+      "grad_norm": 3.879889965057373,
+      "learning_rate": 1.1278195488721805e-06,
+      "loss": 0.1396,
+      "step": 59000
+    },
+    {
+      "epoch": 88.74,
+      "grad_norm": 4.356947422027588,
+      "learning_rate": 1.1263157894736842e-06,
+      "loss": 0.1634,
+      "step": 59010
+    },
+    {
+      "epoch": 88.75,
+      "grad_norm": 3.2641735076904297,
+      "learning_rate": 1.124812030075188e-06,
+      "loss": 0.1805,
+      "step": 59020
+    },
+    {
+      "epoch": 88.77,
+      "grad_norm": 7.944269180297852,
+      "learning_rate": 1.1233082706766917e-06,
+      "loss": 0.1875,
+      "step": 59030
+    },
+    {
+      "epoch": 88.78,
+      "grad_norm": 3.0705432891845703,
+      "learning_rate": 1.1218045112781956e-06,
+      "loss": 0.1884,
+      "step": 59040
+    },
+    {
+      "epoch": 88.8,
+      "grad_norm": 5.9474196434021,
+      "learning_rate": 1.1203007518796994e-06,
+      "loss": 0.1614,
+      "step": 59050
+    },
+    {
+      "epoch": 88.81,
+      "grad_norm": 8.783775329589844,
+      "learning_rate": 1.118796992481203e-06,
+      "loss": 0.2516,
+      "step": 59060
+    },
+    {
+      "epoch": 88.83,
+      "grad_norm": 4.90156888961792,
+      "learning_rate": 1.1172932330827068e-06,
+      "loss": 0.1655,
+      "step": 59070
+    },
+    {
+      "epoch": 88.84,
+      "grad_norm": 4.247372627258301,
+      "learning_rate": 1.1157894736842106e-06,
+      "loss": 0.2167,
+      "step": 59080
+    },
+    {
+      "epoch": 88.86,
+      "grad_norm": 4.955306053161621,
+      "learning_rate": 1.1142857142857145e-06,
+      "loss": 0.159,
+      "step": 59090
+    },
+    {
+      "epoch": 88.87,
+      "grad_norm": 4.56100606918335,
+      "learning_rate": 1.1127819548872182e-06,
+      "loss": 0.1346,
+      "step": 59100
+    },
+    {
+      "epoch": 88.89,
+      "grad_norm": 4.857382297515869,
+      "learning_rate": 1.111278195488722e-06,
+      "loss": 0.1887,
+      "step": 59110
+    },
+    {
+      "epoch": 88.9,
+      "grad_norm": 4.279732704162598,
+      "learning_rate": 1.1097744360902255e-06,
+      "loss": 0.1426,
+      "step": 59120
+    },
+    {
+      "epoch": 88.92,
+      "grad_norm": 4.930658340454102,
+      "learning_rate": 1.1082706766917294e-06,
+      "loss": 0.2294,
+      "step": 59130
+    },
+    {
+      "epoch": 88.93,
+      "grad_norm": 5.308464050292969,
+      "learning_rate": 1.1067669172932332e-06,
+      "loss": 0.1437,
+      "step": 59140
+    },
+    {
+      "epoch": 88.95,
+      "grad_norm": 5.982918739318848,
+      "learning_rate": 1.1052631578947369e-06,
+      "loss": 0.2073,
+      "step": 59150
+    },
+    {
+      "epoch": 88.96,
+      "grad_norm": 6.234976291656494,
+      "learning_rate": 1.1037593984962406e-06,
+      "loss": 0.1346,
+      "step": 59160
+    },
+    {
+      "epoch": 88.98,
+      "grad_norm": 12.91275691986084,
+      "learning_rate": 1.1022556390977444e-06,
+      "loss": 0.2041,
+      "step": 59170
+    },
+    {
+      "epoch": 88.99,
+      "grad_norm": 2.407877206802368,
+      "learning_rate": 1.1007518796992483e-06,
+      "loss": 0.1042,
+      "step": 59180
+    },
+    {
+      "epoch": 89.0,
+      "eval_accuracy": 0.9316,
+      "eval_loss": 0.33024507761001587,
+      "eval_runtime": 84.4728,
+      "eval_samples_per_second": 118.381,
+      "eval_steps_per_second": 0.474,
+      "step": 59185
+    },
+    {
+      "epoch": 89.01,
+      "grad_norm": 9.109488487243652,
+      "learning_rate": 1.099248120300752e-06,
+      "loss": 0.187,
+      "step": 59190
+    },
+    {
+      "epoch": 89.02,
+      "grad_norm": 2.8222405910491943,
+      "learning_rate": 1.0977443609022558e-06,
+      "loss": 0.1887,
+      "step": 59200
+    },
+    {
+      "epoch": 89.04,
+      "grad_norm": 3.990464448928833,
+      "learning_rate": 1.0962406015037595e-06,
+      "loss": 0.1464,
+      "step": 59210
+    },
+    {
+      "epoch": 89.05,
+      "grad_norm": 3.481008529663086,
+      "learning_rate": 1.0947368421052632e-06,
+      "loss": 0.1941,
+      "step": 59220
+    },
+    {
+      "epoch": 89.07,
+      "grad_norm": 4.007563591003418,
+      "learning_rate": 1.0932330827067672e-06,
+      "loss": 0.2157,
+      "step": 59230
+    },
+    {
+      "epoch": 89.08,
+      "grad_norm": 4.087463855743408,
+      "learning_rate": 1.0917293233082709e-06,
+      "loss": 0.1788,
+      "step": 59240
+    },
+    {
+      "epoch": 89.1,
+      "grad_norm": 7.297598838806152,
+      "learning_rate": 1.0902255639097744e-06,
+      "loss": 0.1653,
+      "step": 59250
+    },
+    {
+      "epoch": 89.11,
+      "grad_norm": 4.800395488739014,
+      "learning_rate": 1.0887218045112781e-06,
+      "loss": 0.2075,
+      "step": 59260
+    },
+    {
+      "epoch": 89.13,
+      "grad_norm": 3.8315725326538086,
+      "learning_rate": 1.087218045112782e-06,
+      "loss": 0.1383,
+      "step": 59270
+    },
+    {
+      "epoch": 89.14,
+      "grad_norm": 6.9601898193359375,
+      "learning_rate": 1.0857142857142858e-06,
+      "loss": 0.1993,
+      "step": 59280
+    },
+    {
+      "epoch": 89.16,
+      "grad_norm": 6.42357063293457,
+      "learning_rate": 1.0842105263157895e-06,
+      "loss": 0.1716,
+      "step": 59290
+    },
+    {
+      "epoch": 89.17,
+      "grad_norm": 4.080452919006348,
+      "learning_rate": 1.0827067669172933e-06,
+      "loss": 0.198,
+      "step": 59300
+    },
+    {
+      "epoch": 89.19,
+      "grad_norm": 4.082363605499268,
+      "learning_rate": 1.081203007518797e-06,
+      "loss": 0.1456,
+      "step": 59310
+    },
+    {
+      "epoch": 89.2,
+      "grad_norm": 2.618720531463623,
+      "learning_rate": 1.079699248120301e-06,
+      "loss": 0.2309,
+      "step": 59320
+    },
+    {
+      "epoch": 89.22,
+      "grad_norm": 6.8665080070495605,
+      "learning_rate": 1.0781954887218047e-06,
+      "loss": 0.1487,
+      "step": 59330
+    },
+    {
+      "epoch": 89.23,
+      "grad_norm": 6.956995487213135,
+      "learning_rate": 1.0766917293233084e-06,
+      "loss": 0.2283,
+      "step": 59340
+    },
+    {
+      "epoch": 89.25,
+      "grad_norm": 4.275158405303955,
+      "learning_rate": 1.0751879699248121e-06,
+      "loss": 0.1732,
+      "step": 59350
+    },
+    {
+      "epoch": 89.26,
+      "grad_norm": 6.3027024269104,
+      "learning_rate": 1.0736842105263159e-06,
+      "loss": 0.1888,
+      "step": 59360
+    },
+    {
+      "epoch": 89.28,
+      "grad_norm": 6.413710594177246,
+      "learning_rate": 1.0721804511278196e-06,
+      "loss": 0.2427,
+      "step": 59370
+    },
+    {
+      "epoch": 89.29,
+      "grad_norm": 5.2090888023376465,
+      "learning_rate": 1.0706766917293233e-06,
+      "loss": 0.1906,
+      "step": 59380
+    },
+    {
+      "epoch": 89.31,
+      "grad_norm": 5.576053142547607,
+      "learning_rate": 1.069172932330827e-06,
+      "loss": 0.1398,
+      "step": 59390
+    },
+    {
+      "epoch": 89.32,
+      "grad_norm": 3.6044418811798096,
+      "learning_rate": 1.0676691729323308e-06,
+      "loss": 0.1232,
+      "step": 59400
+    },
+    {
+      "epoch": 89.34,
+      "grad_norm": 3.0662240982055664,
+      "learning_rate": 1.0661654135338347e-06,
+      "loss": 0.189,
+      "step": 59410
+    },
+    {
+      "epoch": 89.35,
+      "grad_norm": 3.142246961593628,
+      "learning_rate": 1.0646616541353385e-06,
+      "loss": 0.1447,
+      "step": 59420
+    },
+    {
+      "epoch": 89.37,
+      "grad_norm": 3.3175511360168457,
+      "learning_rate": 1.0631578947368422e-06,
+      "loss": 0.127,
+      "step": 59430
+    },
+    {
+      "epoch": 89.38,
+      "grad_norm": 4.238184928894043,
+      "learning_rate": 1.061654135338346e-06,
+      "loss": 0.1321,
+      "step": 59440
+    },
+    {
+      "epoch": 89.4,
+      "grad_norm": 7.08275842666626,
+      "learning_rate": 1.0601503759398497e-06,
+      "loss": 0.1795,
+      "step": 59450
+    },
+    {
+      "epoch": 89.41,
+      "grad_norm": 7.184491157531738,
+      "learning_rate": 1.0586466165413536e-06,
+      "loss": 0.2363,
+      "step": 59460
+    },
+    {
+      "epoch": 89.43,
+      "grad_norm": 4.729690074920654,
+      "learning_rate": 1.0571428571428573e-06,
+      "loss": 0.1752,
+      "step": 59470
+    },
+    {
+      "epoch": 89.44,
+      "grad_norm": 5.500245094299316,
+      "learning_rate": 1.055639097744361e-06,
+      "loss": 0.1708,
+      "step": 59480
+    },
+    {
+      "epoch": 89.46,
+      "grad_norm": 4.5759406089782715,
+      "learning_rate": 1.0541353383458648e-06,
+      "loss": 0.1884,
+      "step": 59490
+    },
+    {
+      "epoch": 89.47,
+      "grad_norm": 2.6944706439971924,
+      "learning_rate": 1.0526315789473685e-06,
+      "loss": 0.1609,
+      "step": 59500
+    },
+    {
+      "epoch": 89.49,
+      "grad_norm": 3.369946002960205,
+      "learning_rate": 1.0511278195488723e-06,
+      "loss": 0.1926,
+      "step": 59510
+    },
+    {
+      "epoch": 89.5,
+      "grad_norm": 4.653051376342773,
+      "learning_rate": 1.049624060150376e-06,
+      "loss": 0.1455,
+      "step": 59520
+    },
+    {
+      "epoch": 89.52,
+      "grad_norm": 0.8550413846969604,
+      "learning_rate": 1.0481203007518797e-06,
+      "loss": 0.1589,
+      "step": 59530
+    },
+    {
+      "epoch": 89.53,
+      "grad_norm": 4.918313026428223,
+      "learning_rate": 1.0466165413533835e-06,
+      "loss": 0.0954,
+      "step": 59540
+    },
+    {
+      "epoch": 89.55,
+      "grad_norm": 6.69185209274292,
+      "learning_rate": 1.0451127819548874e-06,
+      "loss": 0.1798,
+      "step": 59550
+    },
+    {
+      "epoch": 89.56,
+      "grad_norm": 6.381489276885986,
+      "learning_rate": 1.0436090225563911e-06,
+      "loss": 0.1493,
+      "step": 59560
+    },
+    {
+      "epoch": 89.58,
+      "grad_norm": 2.553143262863159,
+      "learning_rate": 1.0421052631578949e-06,
+      "loss": 0.1313,
+      "step": 59570
+    },
+    {
+      "epoch": 89.59,
+      "grad_norm": 2.599313974380493,
+      "learning_rate": 1.0406015037593986e-06,
+      "loss": 0.171,
+      "step": 59580
+    },
+    {
+      "epoch": 89.61,
+      "grad_norm": 6.928783893585205,
+      "learning_rate": 1.0390977443609023e-06,
+      "loss": 0.2035,
+      "step": 59590
+    },
+    {
+      "epoch": 89.62,
+      "grad_norm": 4.362387657165527,
+      "learning_rate": 1.037593984962406e-06,
+      "loss": 0.1783,
+      "step": 59600
+    },
+    {
+      "epoch": 89.64,
+      "grad_norm": 2.9104678630828857,
+      "learning_rate": 1.03609022556391e-06,
+      "loss": 0.1895,
+      "step": 59610
+    },
+    {
+      "epoch": 89.65,
+      "grad_norm": 6.155645370483398,
+      "learning_rate": 1.0345864661654135e-06,
+      "loss": 0.1856,
+      "step": 59620
+    },
+    {
+      "epoch": 89.67,
+      "grad_norm": 5.490917205810547,
+      "learning_rate": 1.0330827067669172e-06,
+      "loss": 0.151,
+      "step": 59630
+    },
+    {
+      "epoch": 89.68,
+      "grad_norm": 0.3577495515346527,
+      "learning_rate": 1.0315789473684212e-06,
+      "loss": 0.1647,
+      "step": 59640
+    },
+    {
+      "epoch": 89.7,
+      "grad_norm": 3.6894378662109375,
+      "learning_rate": 1.030075187969925e-06,
+      "loss": 0.1578,
+      "step": 59650
+    },
+    {
+      "epoch": 89.71,
+      "grad_norm": 7.005927562713623,
+      "learning_rate": 1.0285714285714286e-06,
+      "loss": 0.1644,
+      "step": 59660
+    },
+    {
+      "epoch": 89.73,
+      "grad_norm": 2.355557680130005,
+      "learning_rate": 1.0270676691729324e-06,
+      "loss": 0.1616,
+      "step": 59670
+    },
+    {
+      "epoch": 89.74,
+      "grad_norm": 5.249048233032227,
+      "learning_rate": 1.0255639097744361e-06,
+      "loss": 0.1285,
+      "step": 59680
+    },
+    {
+      "epoch": 89.76,
+      "grad_norm": 9.33356761932373,
+      "learning_rate": 1.02406015037594e-06,
+      "loss": 0.1326,
+      "step": 59690
+    },
+    {
+      "epoch": 89.77,
+      "grad_norm": 3.254373550415039,
+      "learning_rate": 1.0225563909774438e-06,
+      "loss": 0.1191,
+      "step": 59700
+    },
+    {
+      "epoch": 89.79,
+      "grad_norm": 4.756227493286133,
+      "learning_rate": 1.0210526315789475e-06,
+      "loss": 0.1221,
+      "step": 59710
+    },
+    {
+      "epoch": 89.8,
+      "grad_norm": 4.242854118347168,
+      "learning_rate": 1.0195488721804512e-06,
+      "loss": 0.1932,
+      "step": 59720
+    },
+    {
+      "epoch": 89.82,
+      "grad_norm": 6.5166425704956055,
+      "learning_rate": 1.018045112781955e-06,
+      "loss": 0.1373,
+      "step": 59730
+    },
+    {
+      "epoch": 89.83,
+      "grad_norm": 2.136077880859375,
+      "learning_rate": 1.0165413533834587e-06,
+      "loss": 0.1699,
+      "step": 59740
+    },
+    {
+      "epoch": 89.85,
+      "grad_norm": 2.9859583377838135,
+      "learning_rate": 1.0150375939849624e-06,
+      "loss": 0.1677,
+      "step": 59750
+    },
+    {
+      "epoch": 89.86,
+      "grad_norm": 5.3651933670043945,
+      "learning_rate": 1.0135338345864662e-06,
+      "loss": 0.176,
+      "step": 59760
+    },
+    {
+      "epoch": 89.88,
+      "grad_norm": 4.6981520652771,
+      "learning_rate": 1.01203007518797e-06,
+      "loss": 0.1921,
+      "step": 59770
+    },
+    {
+      "epoch": 89.89,
+      "grad_norm": 4.584779262542725,
+      "learning_rate": 1.0105263157894738e-06,
+      "loss": 0.1803,
+      "step": 59780
+    },
+    {
+      "epoch": 89.91,
+      "grad_norm": 7.521881580352783,
+      "learning_rate": 1.0090225563909776e-06,
+      "loss": 0.1587,
+      "step": 59790
+    },
+    {
+      "epoch": 89.92,
+      "grad_norm": 7.652347564697266,
+      "learning_rate": 1.0075187969924813e-06,
+      "loss": 0.1852,
+      "step": 59800
+    },
+    {
+      "epoch": 89.94,
+      "grad_norm": 1.188881516456604,
+      "learning_rate": 1.006015037593985e-06,
+      "loss": 0.1649,
+      "step": 59810
+    },
+    {
+      "epoch": 89.95,
+      "grad_norm": 3.557762622833252,
+      "learning_rate": 1.0045112781954888e-06,
+      "loss": 0.1391,
+      "step": 59820
+    },
+    {
+      "epoch": 89.97,
+      "grad_norm": 1.8400624990463257,
+      "learning_rate": 1.0030075187969925e-06,
+      "loss": 0.2029,
+      "step": 59830
+    },
+    {
+      "epoch": 89.98,
+      "grad_norm": 5.2794575691223145,
+      "learning_rate": 1.0015037593984964e-06,
+      "loss": 0.1536,
+      "step": 59840
+    },
+    {
+      "epoch": 90.0,
+      "grad_norm": 43.79523849487305,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.1681,
+      "step": 59850
+    },
+    {
+      "epoch": 90.0,
+      "eval_accuracy": 0.9295,
+      "eval_loss": 0.3303840756416321,
+      "eval_runtime": 84.4298,
+      "eval_samples_per_second": 118.442,
+      "eval_steps_per_second": 0.474,
+      "step": 59850
+    },
+    {
+      "epoch": 90.02,
+      "grad_norm": 4.819916725158691,
+      "learning_rate": 9.98496240601504e-07,
+      "loss": 0.1482,
+      "step": 59860
+    },
+    {
+      "epoch": 90.03,
+      "grad_norm": 8.389203071594238,
+      "learning_rate": 9.969924812030076e-07,
+      "loss": 0.204,
+      "step": 59870
+    },
+    {
+      "epoch": 90.05,
+      "grad_norm": 4.255865097045898,
+      "learning_rate": 9.954887218045114e-07,
+      "loss": 0.2107,
+      "step": 59880
+    },
+    {
+      "epoch": 90.06,
+      "grad_norm": 2.4185428619384766,
+      "learning_rate": 9.93984962406015e-07,
+      "loss": 0.2002,
+      "step": 59890
+    },
+    {
+      "epoch": 90.08,
+      "grad_norm": 6.136438846588135,
+      "learning_rate": 9.924812030075188e-07,
+      "loss": 0.1432,
+      "step": 59900
+    },
+    {
+      "epoch": 90.09,
+      "grad_norm": 4.596153259277344,
+      "learning_rate": 9.909774436090226e-07,
+      "loss": 0.2041,
+      "step": 59910
+    },
+    {
+      "epoch": 90.11,
+      "grad_norm": 6.638514518737793,
+      "learning_rate": 9.894736842105265e-07,
+      "loss": 0.1594,
+      "step": 59920
+    },
+    {
+      "epoch": 90.12,
+      "grad_norm": 5.299813270568848,
+      "learning_rate": 9.879699248120302e-07,
+      "loss": 0.1713,
+      "step": 59930
+    },
+    {
+      "epoch": 90.14,
+      "grad_norm": 6.116607666015625,
+      "learning_rate": 9.86466165413534e-07,
+      "loss": 0.1566,
+      "step": 59940
+    },
+    {
+      "epoch": 90.15,
+      "grad_norm": 6.558013916015625,
+      "learning_rate": 9.849624060150377e-07,
+      "loss": 0.203,
+      "step": 59950
+    },
+    {
+      "epoch": 90.17,
+      "grad_norm": 2.691136598587036,
+      "learning_rate": 9.834586466165414e-07,
+      "loss": 0.1761,
+      "step": 59960
+    },
+    {
+      "epoch": 90.18,
+      "grad_norm": 6.008741855621338,
+      "learning_rate": 9.819548872180451e-07,
+      "loss": 0.2017,
+      "step": 59970
+    },
+    {
+      "epoch": 90.2,
+      "grad_norm": 7.76501989364624,
+      "learning_rate": 9.80451127819549e-07,
+      "loss": 0.188,
+      "step": 59980
+    },
+    {
+      "epoch": 90.21,
+      "grad_norm": 3.065690517425537,
+      "learning_rate": 9.789473684210526e-07,
+      "loss": 0.1711,
+      "step": 59990
+    },
+    {
+      "epoch": 90.23,
+      "grad_norm": 6.988502502441406,
+      "learning_rate": 9.774436090225563e-07,
+      "loss": 0.1323,
+      "step": 60000
+    },
+    {
+      "epoch": 90.24,
+      "grad_norm": 4.135454177856445,
+      "learning_rate": 9.759398496240603e-07,
+      "loss": 0.1522,
+      "step": 60010
+    },
+    {
+      "epoch": 90.26,
+      "grad_norm": 5.1034016609191895,
+      "learning_rate": 9.74436090225564e-07,
+      "loss": 0.1998,
+      "step": 60020
+    },
+    {
+      "epoch": 90.27,
+      "grad_norm": 2.8112096786499023,
+      "learning_rate": 9.729323308270677e-07,
+      "loss": 0.1588,
+      "step": 60030
+    },
+    {
+      "epoch": 90.29,
+      "grad_norm": 2.276792049407959,
+      "learning_rate": 9.714285714285715e-07,
+      "loss": 0.152,
+      "step": 60040
+    },
+    {
+      "epoch": 90.3,
+      "grad_norm": 16.462791442871094,
+      "learning_rate": 9.699248120300752e-07,
+      "loss": 0.1913,
+      "step": 60050
+    },
+    {
+      "epoch": 90.32,
+      "grad_norm": 7.889447212219238,
+      "learning_rate": 9.68421052631579e-07,
+      "loss": 0.1824,
+      "step": 60060
+    },
+    {
+      "epoch": 90.33,
+      "grad_norm": 5.8780837059021,
+      "learning_rate": 9.669172932330829e-07,
+      "loss": 0.1683,
+      "step": 60070
+    },
+    {
+      "epoch": 90.35,
+      "grad_norm": 1.407557487487793,
+      "learning_rate": 9.654135338345866e-07,
+      "loss": 0.1696,
+      "step": 60080
+    },
+    {
+      "epoch": 90.36,
+      "grad_norm": 2.4125938415527344,
+      "learning_rate": 9.639097744360903e-07,
+      "loss": 0.1517,
+      "step": 60090
+    },
+    {
+      "epoch": 90.38,
+      "grad_norm": 6.825364112854004,
+      "learning_rate": 9.62406015037594e-07,
+      "loss": 0.217,
+      "step": 60100
+    },
+    {
+      "epoch": 90.39,
+      "grad_norm": 5.1629743576049805,
+      "learning_rate": 9.609022556390978e-07,
+      "loss": 0.1541,
+      "step": 60110
+    },
+    {
+      "epoch": 90.41,
+      "grad_norm": 6.873205184936523,
+      "learning_rate": 9.593984962406015e-07,
+      "loss": 0.1416,
+      "step": 60120
+    },
+    {
+      "epoch": 90.42,
+      "grad_norm": 5.59041690826416,
+      "learning_rate": 9.578947368421053e-07,
+      "loss": 0.153,
+      "step": 60130
+    },
+    {
+      "epoch": 90.44,
+      "grad_norm": 3.045102834701538,
+      "learning_rate": 9.56390977443609e-07,
+      "loss": 0.1723,
+      "step": 60140
+    },
+    {
+      "epoch": 90.45,
+      "grad_norm": 3.393899917602539,
+      "learning_rate": 9.54887218045113e-07,
+      "loss": 0.1852,
+      "step": 60150
+    },
+    {
+      "epoch": 90.47,
+      "grad_norm": 5.689571380615234,
+      "learning_rate": 9.533834586466166e-07,
+      "loss": 0.2115,
+      "step": 60160
+    },
+    {
+      "epoch": 90.48,
+      "grad_norm": 5.110867023468018,
+      "learning_rate": 9.518796992481204e-07,
+      "loss": 0.1877,
+      "step": 60170
+    },
+    {
+      "epoch": 90.5,
+      "grad_norm": 2.7676916122436523,
+      "learning_rate": 9.503759398496241e-07,
+      "loss": 0.1966,
+      "step": 60180
+    },
+    {
+      "epoch": 90.51,
+      "grad_norm": 0.4056377410888672,
+      "learning_rate": 9.48872180451128e-07,
+      "loss": 0.1819,
+      "step": 60190
+    },
+    {
+      "epoch": 90.53,
+      "grad_norm": 15.291031837463379,
+      "learning_rate": 9.473684210526317e-07,
+      "loss": 0.1564,
+      "step": 60200
+    },
+    {
+      "epoch": 90.54,
+      "grad_norm": 4.443658828735352,
+      "learning_rate": 9.458646616541354e-07,
+      "loss": 0.1336,
+      "step": 60210
+    },
+    {
+      "epoch": 90.56,
+      "grad_norm": 8.213922500610352,
+      "learning_rate": 9.443609022556393e-07,
+      "loss": 0.1578,
+      "step": 60220
+    },
+    {
+      "epoch": 90.57,
+      "grad_norm": 5.976234436035156,
+      "learning_rate": 9.42857142857143e-07,
+      "loss": 0.1815,
+      "step": 60230
+    },
+    {
+      "epoch": 90.59,
+      "grad_norm": 3.436178684234619,
+      "learning_rate": 9.413533834586466e-07,
+      "loss": 0.1836,
+      "step": 60240
+    },
+    {
+      "epoch": 90.6,
+      "grad_norm": 7.1853203773498535,
+      "learning_rate": 9.398496240601504e-07,
+      "loss": 0.1733,
+      "step": 60250
+    },
+    {
+      "epoch": 90.62,
+      "grad_norm": 3.7736470699310303,
+      "learning_rate": 9.383458646616542e-07,
+      "loss": 0.0901,
+      "step": 60260
+    },
+    {
+      "epoch": 90.63,
+      "grad_norm": 6.868650436401367,
+      "learning_rate": 9.368421052631579e-07,
+      "loss": 0.1495,
+      "step": 60270
+    },
+    {
+      "epoch": 90.65,
+      "grad_norm": 6.130927085876465,
+      "learning_rate": 9.353383458646618e-07,
+      "loss": 0.175,
+      "step": 60280
+    },
+    {
+      "epoch": 90.66,
+      "grad_norm": 2.877918004989624,
+      "learning_rate": 9.338345864661655e-07,
+      "loss": 0.1292,
+      "step": 60290
+    },
+    {
+      "epoch": 90.68,
+      "grad_norm": 6.6926493644714355,
+      "learning_rate": 9.323308270676692e-07,
+      "loss": 0.1629,
+      "step": 60300
+    },
+    {
+      "epoch": 90.69,
+      "grad_norm": 4.85024356842041,
+      "learning_rate": 9.308270676691731e-07,
+      "loss": 0.1462,
+      "step": 60310
+    },
+    {
+      "epoch": 90.71,
+      "grad_norm": 4.398621559143066,
+      "learning_rate": 9.293233082706768e-07,
+      "loss": 0.1805,
+      "step": 60320
+    },
+    {
+      "epoch": 90.72,
+      "grad_norm": 4.464656829833984,
+      "learning_rate": 9.278195488721805e-07,
+      "loss": 0.1624,
+      "step": 60330
+    },
+    {
+      "epoch": 90.74,
+      "grad_norm": 2.839735984802246,
+      "learning_rate": 9.263157894736844e-07,
+      "loss": 0.1813,
+      "step": 60340
+    },
+    {
+      "epoch": 90.75,
+      "grad_norm": 3.697030544281006,
+      "learning_rate": 9.248120300751881e-07,
+      "loss": 0.1472,
+      "step": 60350
+    },
+    {
+      "epoch": 90.77,
+      "grad_norm": 4.30585241317749,
+      "learning_rate": 9.233082706766917e-07,
+      "loss": 0.1624,
+      "step": 60360
+    },
+    {
+      "epoch": 90.78,
+      "grad_norm": 5.344664096832275,
+      "learning_rate": 9.218045112781955e-07,
+      "loss": 0.1486,
+      "step": 60370
+    },
+    {
+      "epoch": 90.8,
+      "grad_norm": 3.9240634441375732,
+      "learning_rate": 9.203007518796993e-07,
+      "loss": 0.1843,
+      "step": 60380
+    },
+    {
+      "epoch": 90.81,
+      "grad_norm": 2.0143322944641113,
+      "learning_rate": 9.18796992481203e-07,
+      "loss": 0.1147,
+      "step": 60390
+    },
+    {
+      "epoch": 90.83,
+      "grad_norm": 6.274852275848389,
+      "learning_rate": 9.172932330827068e-07,
+      "loss": 0.169,
+      "step": 60400
+    },
+    {
+      "epoch": 90.84,
+      "grad_norm": 5.181715488433838,
+      "learning_rate": 9.157894736842106e-07,
+      "loss": 0.1845,
+      "step": 60410
+    },
+    {
+      "epoch": 90.86,
+      "grad_norm": 0.9931633472442627,
+      "learning_rate": 9.142857142857144e-07,
+      "loss": 0.1365,
+      "step": 60420
+    },
+    {
+      "epoch": 90.87,
+      "grad_norm": 5.815258502960205,
+      "learning_rate": 9.127819548872181e-07,
+      "loss": 0.1571,
+      "step": 60430
+    },
+    {
+      "epoch": 90.89,
+      "grad_norm": 4.237338542938232,
+      "learning_rate": 9.112781954887219e-07,
+      "loss": 0.1783,
+      "step": 60440
+    },
+    {
+      "epoch": 90.9,
+      "grad_norm": 4.938007354736328,
+      "learning_rate": 9.097744360902257e-07,
+      "loss": 0.1615,
+      "step": 60450
+    },
+    {
+      "epoch": 90.92,
+      "grad_norm": 7.971113204956055,
+      "learning_rate": 9.082706766917294e-07,
+      "loss": 0.1811,
+      "step": 60460
+    },
+    {
+      "epoch": 90.93,
+      "grad_norm": 6.53290319442749,
+      "learning_rate": 9.067669172932332e-07,
+      "loss": 0.1816,
+      "step": 60470
+    },
+    {
+      "epoch": 90.95,
+      "grad_norm": 5.972143173217773,
+      "learning_rate": 9.05263157894737e-07,
+      "loss": 0.2215,
+      "step": 60480
+    },
+    {
+      "epoch": 90.96,
+      "grad_norm": 6.27506685256958,
+      "learning_rate": 9.037593984962406e-07,
+      "loss": 0.1527,
+      "step": 60490
+    },
+    {
+      "epoch": 90.98,
+      "grad_norm": 3.595836877822876,
+      "learning_rate": 9.022556390977444e-07,
+      "loss": 0.1536,
+      "step": 60500
+    },
+    {
+      "epoch": 90.99,
+      "grad_norm": 4.775961399078369,
+      "learning_rate": 9.007518796992482e-07,
+      "loss": 0.1802,
+      "step": 60510
+    },
+    {
+      "epoch": 91.0,
+      "eval_accuracy": 0.9298,
+      "eval_loss": 0.3351175785064697,
+      "eval_runtime": 84.5299,
+      "eval_samples_per_second": 118.301,
+      "eval_steps_per_second": 0.473,
+      "step": 60515
+    },
+    {
+      "epoch": 91.01,
+      "grad_norm": 7.370502948760986,
+      "learning_rate": 8.992481203007519e-07,
+      "loss": 0.2365,
+      "step": 60520
+    },
+    {
+      "epoch": 91.02,
+      "grad_norm": 8.751474380493164,
+      "learning_rate": 8.977443609022557e-07,
+      "loss": 0.2011,
+      "step": 60530
+    },
+    {
+      "epoch": 91.04,
+      "grad_norm": 3.2370705604553223,
+      "learning_rate": 8.962406015037595e-07,
+      "loss": 0.1858,
+      "step": 60540
+    },
+    {
+      "epoch": 91.05,
+      "grad_norm": 5.268539905548096,
+      "learning_rate": 8.947368421052632e-07,
+      "loss": 0.2265,
+      "step": 60550
+    },
+    {
+      "epoch": 91.07,
+      "grad_norm": 5.824682235717773,
+      "learning_rate": 8.93233082706767e-07,
+      "loss": 0.17,
+      "step": 60560
+    },
+    {
+      "epoch": 91.08,
+      "grad_norm": 4.85299015045166,
+      "learning_rate": 8.917293233082708e-07,
+      "loss": 0.1715,
+      "step": 60570
+    },
+    {
+      "epoch": 91.1,
+      "grad_norm": 3.4943690299987793,
+      "learning_rate": 8.902255639097745e-07,
+      "loss": 0.1739,
+      "step": 60580
+    },
+    {
+      "epoch": 91.11,
+      "grad_norm": 5.409356594085693,
+      "learning_rate": 8.887218045112784e-07,
+      "loss": 0.1693,
+      "step": 60590
+    },
+    {
+      "epoch": 91.13,
+      "grad_norm": 6.857398986816406,
+      "learning_rate": 8.872180451127821e-07,
+      "loss": 0.1883,
+      "step": 60600
+    },
+    {
+      "epoch": 91.14,
+      "grad_norm": 4.355463027954102,
+      "learning_rate": 8.857142857142857e-07,
+      "loss": 0.1444,
+      "step": 60610
+    },
+    {
+      "epoch": 91.16,
+      "grad_norm": 5.453840255737305,
+      "learning_rate": 8.842105263157895e-07,
+      "loss": 0.1129,
+      "step": 60620
+    },
+    {
+      "epoch": 91.17,
+      "grad_norm": 2.509838581085205,
+      "learning_rate": 8.827067669172933e-07,
+      "loss": 0.1524,
+      "step": 60630
+    },
+    {
+      "epoch": 91.19,
+      "grad_norm": 4.419145107269287,
+      "learning_rate": 8.81203007518797e-07,
+      "loss": 0.1007,
+      "step": 60640
+    },
+    {
+      "epoch": 91.2,
+      "grad_norm": 2.9368622303009033,
+      "learning_rate": 8.796992481203009e-07,
+      "loss": 0.1373,
+      "step": 60650
+    },
+    {
+      "epoch": 91.22,
+      "grad_norm": 5.6778082847595215,
+      "learning_rate": 8.781954887218046e-07,
+      "loss": 0.1914,
+      "step": 60660
+    },
+    {
+      "epoch": 91.23,
+      "grad_norm": 4.565642833709717,
+      "learning_rate": 8.766917293233083e-07,
+      "loss": 0.1885,
+      "step": 60670
+    },
+    {
+      "epoch": 91.25,
+      "grad_norm": 12.5923433303833,
+      "learning_rate": 8.751879699248122e-07,
+      "loss": 0.1841,
+      "step": 60680
+    },
+    {
+      "epoch": 91.26,
+      "grad_norm": 3.8963727951049805,
+      "learning_rate": 8.736842105263159e-07,
+      "loss": 0.2082,
+      "step": 60690
+    },
+    {
+      "epoch": 91.28,
+      "grad_norm": 7.847858428955078,
+      "learning_rate": 8.721804511278196e-07,
+      "loss": 0.2347,
+      "step": 60700
+    },
+    {
+      "epoch": 91.29,
+      "grad_norm": 6.093050003051758,
+      "learning_rate": 8.706766917293235e-07,
+      "loss": 0.2005,
+      "step": 60710
+    },
+    {
+      "epoch": 91.31,
+      "grad_norm": 6.70009708404541,
+      "learning_rate": 8.691729323308272e-07,
+      "loss": 0.1581,
+      "step": 60720
+    },
+    {
+      "epoch": 91.32,
+      "grad_norm": 5.4743852615356445,
+      "learning_rate": 8.67669172932331e-07,
+      "loss": 0.1938,
+      "step": 60730
+    },
+    {
+      "epoch": 91.34,
+      "grad_norm": 6.605056285858154,
+      "learning_rate": 8.661654135338346e-07,
+      "loss": 0.2517,
+      "step": 60740
+    },
+    {
+      "epoch": 91.35,
+      "grad_norm": 5.257135391235352,
+      "learning_rate": 8.646616541353384e-07,
+      "loss": 0.1354,
+      "step": 60750
+    },
+    {
+      "epoch": 91.37,
+      "grad_norm": 4.14031982421875,
+      "learning_rate": 8.631578947368421e-07,
+      "loss": 0.2008,
+      "step": 60760
+    },
+    {
+      "epoch": 91.38,
+      "grad_norm": 4.664227485656738,
+      "learning_rate": 8.616541353383459e-07,
+      "loss": 0.166,
+      "step": 60770
+    },
+    {
+      "epoch": 91.4,
+      "grad_norm": 7.249124526977539,
+      "learning_rate": 8.601503759398497e-07,
+      "loss": 0.1729,
+      "step": 60780
+    },
+    {
+      "epoch": 91.41,
+      "grad_norm": 1.5819720029830933,
+      "learning_rate": 8.586466165413534e-07,
+      "loss": 0.1673,
+      "step": 60790
+    },
+    {
+      "epoch": 91.43,
+      "grad_norm": 8.89696216583252,
+      "learning_rate": 8.571428571428572e-07,
+      "loss": 0.1627,
+      "step": 60800
+    },
+    {
+      "epoch": 91.44,
+      "grad_norm": 6.409267425537109,
+      "learning_rate": 8.55639097744361e-07,
+      "loss": 0.1917,
+      "step": 60810
+    },
+    {
+      "epoch": 91.46,
+      "grad_norm": 13.142133712768555,
+      "learning_rate": 8.541353383458648e-07,
+      "loss": 0.2156,
+      "step": 60820
+    },
+    {
+      "epoch": 91.47,
+      "grad_norm": 3.5235838890075684,
+      "learning_rate": 8.526315789473685e-07,
+      "loss": 0.1644,
+      "step": 60830
+    },
+    {
+      "epoch": 91.49,
+      "grad_norm": 5.23291015625,
+      "learning_rate": 8.511278195488723e-07,
+      "loss": 0.2202,
+      "step": 60840
+    },
+    {
+      "epoch": 91.5,
+      "grad_norm": 8.659149169921875,
+      "learning_rate": 8.496240601503761e-07,
+      "loss": 0.1695,
+      "step": 60850
+    },
+    {
+      "epoch": 91.52,
+      "grad_norm": 5.686830043792725,
+      "learning_rate": 8.481203007518797e-07,
+      "loss": 0.1524,
+      "step": 60860
+    },
+    {
+      "epoch": 91.53,
+      "grad_norm": 5.107367992401123,
+      "learning_rate": 8.466165413533835e-07,
+      "loss": 0.1736,
+      "step": 60870
+    },
+    {
+      "epoch": 91.55,
+      "grad_norm": 8.266801834106445,
+      "learning_rate": 8.451127819548873e-07,
+      "loss": 0.1509,
+      "step": 60880
+    },
+    {
+      "epoch": 91.56,
+      "grad_norm": 8.239102363586426,
+      "learning_rate": 8.43609022556391e-07,
+      "loss": 0.1974,
+      "step": 60890
+    },
+    {
+      "epoch": 91.58,
+      "grad_norm": 4.153801918029785,
+      "learning_rate": 8.421052631578948e-07,
+      "loss": 0.2202,
+      "step": 60900
+    },
+    {
+      "epoch": 91.59,
+      "grad_norm": 4.520059585571289,
+      "learning_rate": 8.406015037593986e-07,
+      "loss": 0.1404,
+      "step": 60910
+    },
+    {
+      "epoch": 91.61,
+      "grad_norm": 3.584502935409546,
+      "learning_rate": 8.390977443609023e-07,
+      "loss": 0.1914,
+      "step": 60920
+    },
+    {
+      "epoch": 91.62,
+      "grad_norm": 5.777580261230469,
+      "learning_rate": 8.375939849624061e-07,
+      "loss": 0.1338,
+      "step": 60930
+    },
+    {
+      "epoch": 91.64,
+      "grad_norm": 4.816416263580322,
+      "learning_rate": 8.360902255639099e-07,
+      "loss": 0.1983,
+      "step": 60940
+    },
+    {
+      "epoch": 91.65,
+      "grad_norm": 1.8106648921966553,
+      "learning_rate": 8.345864661654136e-07,
+      "loss": 0.1728,
+      "step": 60950
+    },
+    {
+      "epoch": 91.67,
+      "grad_norm": 6.586467266082764,
+      "learning_rate": 8.330827067669174e-07,
+      "loss": 0.1635,
+      "step": 60960
+    },
+    {
+      "epoch": 91.68,
+      "grad_norm": 4.0947136878967285,
+      "learning_rate": 8.315789473684212e-07,
+      "loss": 0.1938,
+      "step": 60970
+    },
+    {
+      "epoch": 91.7,
+      "grad_norm": 6.031853199005127,
+      "learning_rate": 8.300751879699248e-07,
+      "loss": 0.18,
+      "step": 60980
+    },
+    {
+      "epoch": 91.71,
+      "grad_norm": 5.872425079345703,
+      "learning_rate": 8.285714285714285e-07,
+      "loss": 0.1693,
+      "step": 60990
+    },
+    {
+      "epoch": 91.73,
+      "grad_norm": 9.427947998046875,
+      "learning_rate": 8.270676691729324e-07,
+      "loss": 0.187,
+      "step": 61000
+    },
+    {
+      "epoch": 91.74,
+      "grad_norm": 5.179823398590088,
+      "learning_rate": 8.255639097744361e-07,
+      "loss": 0.1372,
+      "step": 61010
+    },
+    {
+      "epoch": 91.76,
+      "grad_norm": 5.486731052398682,
+      "learning_rate": 8.240601503759398e-07,
+      "loss": 0.1654,
+      "step": 61020
+    },
+    {
+      "epoch": 91.77,
+      "grad_norm": 3.1775834560394287,
+      "learning_rate": 8.225563909774437e-07,
+      "loss": 0.176,
+      "step": 61030
+    },
+    {
+      "epoch": 91.79,
+      "grad_norm": 4.589211463928223,
+      "learning_rate": 8.210526315789474e-07,
+      "loss": 0.1452,
+      "step": 61040
+    },
+    {
+      "epoch": 91.8,
+      "grad_norm": 3.6685729026794434,
+      "learning_rate": 8.195488721804513e-07,
+      "loss": 0.1883,
+      "step": 61050
+    },
+    {
+      "epoch": 91.82,
+      "grad_norm": 4.512982368469238,
+      "learning_rate": 8.18045112781955e-07,
+      "loss": 0.1589,
+      "step": 61060
+    },
+    {
+      "epoch": 91.83,
+      "grad_norm": 9.20313549041748,
+      "learning_rate": 8.165413533834587e-07,
+      "loss": 0.1399,
+      "step": 61070
+    },
+    {
+      "epoch": 91.85,
+      "grad_norm": 5.7398905754089355,
+      "learning_rate": 8.150375939849625e-07,
+      "loss": 0.1689,
+      "step": 61080
+    },
+    {
+      "epoch": 91.86,
+      "grad_norm": 4.950761795043945,
+      "learning_rate": 8.135338345864663e-07,
+      "loss": 0.1259,
+      "step": 61090
+    },
+    {
+      "epoch": 91.88,
+      "grad_norm": 7.066133499145508,
+      "learning_rate": 8.1203007518797e-07,
+      "loss": 0.1673,
+      "step": 61100
+    },
+    {
+      "epoch": 91.89,
+      "grad_norm": 4.5306596755981445,
+      "learning_rate": 8.105263157894736e-07,
+      "loss": 0.173,
+      "step": 61110
+    },
+    {
+      "epoch": 91.91,
+      "grad_norm": 6.547530651092529,
+      "learning_rate": 8.090225563909775e-07,
+      "loss": 0.1469,
+      "step": 61120
+    },
+    {
+      "epoch": 91.92,
+      "grad_norm": 6.574265003204346,
+      "learning_rate": 8.075187969924812e-07,
+      "loss": 0.1464,
+      "step": 61130
+    },
+    {
+      "epoch": 91.94,
+      "grad_norm": 2.8120083808898926,
+      "learning_rate": 8.06015037593985e-07,
+      "loss": 0.1422,
+      "step": 61140
+    },
+    {
+      "epoch": 91.95,
+      "grad_norm": 3.274015188217163,
+      "learning_rate": 8.045112781954888e-07,
+      "loss": 0.187,
+      "step": 61150
+    },
+    {
+      "epoch": 91.97,
+      "grad_norm": 16.76445198059082,
+      "learning_rate": 8.030075187969925e-07,
+      "loss": 0.1365,
+      "step": 61160
+    },
+    {
+      "epoch": 91.98,
+      "grad_norm": 3.863375663757324,
+      "learning_rate": 8.015037593984963e-07,
+      "loss": 0.1375,
+      "step": 61170
+    },
+    {
+      "epoch": 92.0,
+      "grad_norm": 23.6773738861084,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 0.268,
+      "step": 61180
+    },
+    {
+      "epoch": 92.0,
+      "eval_accuracy": 0.9305,
+      "eval_loss": 0.33316439390182495,
+      "eval_runtime": 84.8569,
+      "eval_samples_per_second": 117.845,
+      "eval_steps_per_second": 0.471,
+      "step": 61180
+    },
+    {
+      "epoch": 92.02,
+      "grad_norm": 6.592239856719971,
+      "learning_rate": 7.984962406015038e-07,
+      "loss": 0.1531,
+      "step": 61190
+    },
+    {
+      "epoch": 92.03,
+      "grad_norm": 3.465878963470459,
+      "learning_rate": 7.969924812030076e-07,
+      "loss": 0.1796,
+      "step": 61200
+    },
+    {
+      "epoch": 92.05,
+      "grad_norm": 5.356743812561035,
+      "learning_rate": 7.954887218045114e-07,
+      "loss": 0.1227,
+      "step": 61210
+    },
+    {
+      "epoch": 92.06,
+      "grad_norm": 9.6128511428833,
+      "learning_rate": 7.939849624060152e-07,
+      "loss": 0.1512,
+      "step": 61220
+    },
+    {
+      "epoch": 92.08,
+      "grad_norm": 5.72312068939209,
+      "learning_rate": 7.924812030075188e-07,
+      "loss": 0.1641,
+      "step": 61230
+    },
+    {
+      "epoch": 92.09,
+      "grad_norm": 2.734785556793213,
+      "learning_rate": 7.909774436090226e-07,
+      "loss": 0.2045,
+      "step": 61240
+    },
+    {
+      "epoch": 92.11,
+      "grad_norm": 5.764986991882324,
+      "learning_rate": 7.894736842105263e-07,
+      "loss": 0.1482,
+      "step": 61250
+    },
+    {
+      "epoch": 92.12,
+      "grad_norm": 6.9798173904418945,
+      "learning_rate": 7.879699248120301e-07,
+      "loss": 0.2035,
+      "step": 61260
+    },
+    {
+      "epoch": 92.14,
+      "grad_norm": 6.008689880371094,
+      "learning_rate": 7.864661654135339e-07,
+      "loss": 0.1314,
+      "step": 61270
+    },
+    {
+      "epoch": 92.15,
+      "grad_norm": 4.28978157043457,
+      "learning_rate": 7.849624060150377e-07,
+      "loss": 0.1305,
+      "step": 61280
+    },
+    {
+      "epoch": 92.17,
+      "grad_norm": 2.4776744842529297,
+      "learning_rate": 7.834586466165414e-07,
+      "loss": 0.1741,
+      "step": 61290
+    },
+    {
+      "epoch": 92.18,
+      "grad_norm": 6.749001502990723,
+      "learning_rate": 7.819548872180452e-07,
+      "loss": 0.1177,
+      "step": 61300
+    },
+    {
+      "epoch": 92.2,
+      "grad_norm": 5.488044738769531,
+      "learning_rate": 7.80451127819549e-07,
+      "loss": 0.2004,
+      "step": 61310
+    },
+    {
+      "epoch": 92.21,
+      "grad_norm": 6.313375949859619,
+      "learning_rate": 7.789473684210527e-07,
+      "loss": 0.2079,
+      "step": 61320
+    },
+    {
+      "epoch": 92.23,
+      "grad_norm": 4.431061267852783,
+      "learning_rate": 7.774436090225565e-07,
+      "loss": 0.135,
+      "step": 61330
+    },
+    {
+      "epoch": 92.24,
+      "grad_norm": 4.987079620361328,
+      "learning_rate": 7.759398496240603e-07,
+      "loss": 0.1677,
+      "step": 61340
+    },
+    {
+      "epoch": 92.26,
+      "grad_norm": 6.565682411193848,
+      "learning_rate": 7.74436090225564e-07,
+      "loss": 0.1398,
+      "step": 61350
+    },
+    {
+      "epoch": 92.27,
+      "grad_norm": 8.182050704956055,
+      "learning_rate": 7.729323308270676e-07,
+      "loss": 0.1806,
+      "step": 61360
+    },
+    {
+      "epoch": 92.29,
+      "grad_norm": 1.415570855140686,
+      "learning_rate": 7.714285714285715e-07,
+      "loss": 0.1538,
+      "step": 61370
+    },
+    {
+      "epoch": 92.3,
+      "grad_norm": 7.348145961761475,
+      "learning_rate": 7.699248120300752e-07,
+      "loss": 0.152,
+      "step": 61380
+    },
+    {
+      "epoch": 92.32,
+      "grad_norm": 2.7620201110839844,
+      "learning_rate": 7.684210526315789e-07,
+      "loss": 0.2018,
+      "step": 61390
+    },
+    {
+      "epoch": 92.33,
+      "grad_norm": 6.486240863800049,
+      "learning_rate": 7.669172932330828e-07,
+      "loss": 0.257,
+      "step": 61400
+    },
+    {
+      "epoch": 92.35,
+      "grad_norm": 4.345780372619629,
+      "learning_rate": 7.654135338345865e-07,
+      "loss": 0.2216,
+      "step": 61410
+    },
+    {
+      "epoch": 92.36,
+      "grad_norm": 3.1555426120758057,
+      "learning_rate": 7.639097744360902e-07,
+      "loss": 0.1257,
+      "step": 61420
+    },
+    {
+      "epoch": 92.38,
+      "grad_norm": 4.688722610473633,
+      "learning_rate": 7.624060150375941e-07,
+      "loss": 0.1059,
+      "step": 61430
+    },
+    {
+      "epoch": 92.39,
+      "grad_norm": 4.647052764892578,
+      "learning_rate": 7.609022556390978e-07,
+      "loss": 0.1449,
+      "step": 61440
+    },
+    {
+      "epoch": 92.41,
+      "grad_norm": 7.654082775115967,
+      "learning_rate": 7.593984962406016e-07,
+      "loss": 0.2042,
+      "step": 61450
+    },
+    {
+      "epoch": 92.42,
+      "grad_norm": 12.962769508361816,
+      "learning_rate": 7.578947368421054e-07,
+      "loss": 0.1656,
+      "step": 61460
+    },
+    {
+      "epoch": 92.44,
+      "grad_norm": 4.620476245880127,
+      "learning_rate": 7.563909774436091e-07,
+      "loss": 0.1846,
+      "step": 61470
+    },
+    {
+      "epoch": 92.45,
+      "grad_norm": 6.064635753631592,
+      "learning_rate": 7.548872180451127e-07,
+      "loss": 0.1734,
+      "step": 61480
+    },
+    {
+      "epoch": 92.47,
+      "grad_norm": 4.810222625732422,
+      "learning_rate": 7.533834586466166e-07,
+      "loss": 0.1596,
+      "step": 61490
+    },
+    {
+      "epoch": 92.48,
+      "grad_norm": 6.1294755935668945,
+      "learning_rate": 7.518796992481203e-07,
+      "loss": 0.1405,
+      "step": 61500
+    },
+    {
+      "epoch": 92.5,
+      "grad_norm": 6.559119701385498,
+      "learning_rate": 7.503759398496241e-07,
+      "loss": 0.1389,
+      "step": 61510
+    },
+    {
+      "epoch": 92.51,
+      "grad_norm": 4.981884479522705,
+      "learning_rate": 7.488721804511279e-07,
+      "loss": 0.1999,
+      "step": 61520
+    },
+    {
+      "epoch": 92.53,
+      "grad_norm": 6.0175604820251465,
+      "learning_rate": 7.473684210526316e-07,
+      "loss": 0.194,
+      "step": 61530
+    },
+    {
+      "epoch": 92.54,
+      "grad_norm": 6.699028968811035,
+      "learning_rate": 7.458646616541354e-07,
+      "loss": 0.2085,
+      "step": 61540
+    },
+    {
+      "epoch": 92.56,
+      "grad_norm": 9.893536567687988,
+      "learning_rate": 7.443609022556392e-07,
+      "loss": 0.1569,
+      "step": 61550
+    },
+    {
+      "epoch": 92.57,
+      "grad_norm": 3.370506525039673,
+      "learning_rate": 7.428571428571429e-07,
+      "loss": 0.1606,
+      "step": 61560
+    },
+    {
+      "epoch": 92.59,
+      "grad_norm": 4.382241725921631,
+      "learning_rate": 7.413533834586467e-07,
+      "loss": 0.1473,
+      "step": 61570
+    },
+    {
+      "epoch": 92.6,
+      "grad_norm": 1.784295916557312,
+      "learning_rate": 7.398496240601505e-07,
+      "loss": 0.1692,
+      "step": 61580
+    },
+    {
+      "epoch": 92.62,
+      "grad_norm": 2.770347833633423,
+      "learning_rate": 7.383458646616543e-07,
+      "loss": 0.1704,
+      "step": 61590
+    },
+    {
+      "epoch": 92.63,
+      "grad_norm": 3.2620766162872314,
+      "learning_rate": 7.368421052631579e-07,
+      "loss": 0.1742,
+      "step": 61600
+    },
+    {
+      "epoch": 92.65,
+      "grad_norm": 8.29354476928711,
+      "learning_rate": 7.353383458646617e-07,
+      "loss": 0.1271,
+      "step": 61610
+    },
+    {
+      "epoch": 92.66,
+      "grad_norm": 3.8892619609832764,
+      "learning_rate": 7.338345864661654e-07,
+      "loss": 0.2461,
+      "step": 61620
+    },
+    {
+      "epoch": 92.68,
+      "grad_norm": 7.872462272644043,
+      "learning_rate": 7.323308270676692e-07,
+      "loss": 0.1877,
+      "step": 61630
+    },
+    {
+      "epoch": 92.69,
+      "grad_norm": 4.30030632019043,
+      "learning_rate": 7.30827067669173e-07,
+      "loss": 0.167,
+      "step": 61640
+    },
+    {
+      "epoch": 92.71,
+      "grad_norm": 3.3043644428253174,
+      "learning_rate": 7.293233082706767e-07,
+      "loss": 0.2256,
+      "step": 61650
+    },
+    {
+      "epoch": 92.72,
+      "grad_norm": 6.955836296081543,
+      "learning_rate": 7.278195488721805e-07,
+      "loss": 0.2138,
+      "step": 61660
+    },
+    {
+      "epoch": 92.74,
+      "grad_norm": 2.1110410690307617,
+      "learning_rate": 7.263157894736843e-07,
+      "loss": 0.1773,
+      "step": 61670
+    },
+    {
+      "epoch": 92.75,
+      "grad_norm": 6.610522747039795,
+      "learning_rate": 7.248120300751881e-07,
+      "loss": 0.1779,
+      "step": 61680
+    },
+    {
+      "epoch": 92.77,
+      "grad_norm": 4.267979145050049,
+      "learning_rate": 7.233082706766918e-07,
+      "loss": 0.1249,
+      "step": 61690
+    },
+    {
+      "epoch": 92.78,
+      "grad_norm": 5.974915027618408,
+      "learning_rate": 7.218045112781956e-07,
+      "loss": 0.1679,
+      "step": 61700
+    },
+    {
+      "epoch": 92.8,
+      "grad_norm": 2.948686122894287,
+      "learning_rate": 7.203007518796994e-07,
+      "loss": 0.1572,
+      "step": 61710
+    },
+    {
+      "epoch": 92.81,
+      "grad_norm": 4.3910040855407715,
+      "learning_rate": 7.187969924812031e-07,
+      "loss": 0.1469,
+      "step": 61720
+    },
+    {
+      "epoch": 92.83,
+      "grad_norm": 5.496255874633789,
+      "learning_rate": 7.172932330827067e-07,
+      "loss": 0.1574,
+      "step": 61730
+    },
+    {
+      "epoch": 92.84,
+      "grad_norm": 6.861441135406494,
+      "learning_rate": 7.157894736842106e-07,
+      "loss": 0.161,
+      "step": 61740
+    },
+    {
+      "epoch": 92.86,
+      "grad_norm": 6.682887554168701,
+      "learning_rate": 7.142857142857143e-07,
+      "loss": 0.1741,
+      "step": 61750
+    },
+    {
+      "epoch": 92.87,
+      "grad_norm": 5.851301670074463,
+      "learning_rate": 7.12781954887218e-07,
+      "loss": 0.2264,
+      "step": 61760
+    },
+    {
+      "epoch": 92.89,
+      "grad_norm": 6.062599182128906,
+      "learning_rate": 7.112781954887219e-07,
+      "loss": 0.1853,
+      "step": 61770
+    },
+    {
+      "epoch": 92.9,
+      "grad_norm": 3.691725969314575,
+      "learning_rate": 7.097744360902256e-07,
+      "loss": 0.1196,
+      "step": 61780
+    },
+    {
+      "epoch": 92.92,
+      "grad_norm": 6.096661567687988,
+      "learning_rate": 7.082706766917293e-07,
+      "loss": 0.1816,
+      "step": 61790
+    },
+    {
+      "epoch": 92.93,
+      "grad_norm": 0.75389564037323,
+      "learning_rate": 7.067669172932332e-07,
+      "loss": 0.1673,
+      "step": 61800
+    },
+    {
+      "epoch": 92.95,
+      "grad_norm": 8.029284477233887,
+      "learning_rate": 7.052631578947369e-07,
+      "loss": 0.1728,
+      "step": 61810
+    },
+    {
+      "epoch": 92.96,
+      "grad_norm": 9.124761581420898,
+      "learning_rate": 7.037593984962407e-07,
+      "loss": 0.2143,
+      "step": 61820
+    },
+    {
+      "epoch": 92.98,
+      "grad_norm": 6.918426036834717,
+      "learning_rate": 7.022556390977445e-07,
+      "loss": 0.1845,
+      "step": 61830
+    },
+    {
+      "epoch": 92.99,
+      "grad_norm": 6.886169910430908,
+      "learning_rate": 7.007518796992482e-07,
+      "loss": 0.1807,
+      "step": 61840
+    },
+    {
+      "epoch": 93.0,
+      "eval_accuracy": 0.9307,
+      "eval_loss": 0.3299960494041443,
+      "eval_runtime": 84.6466,
+      "eval_samples_per_second": 118.138,
+      "eval_steps_per_second": 0.473,
+      "step": 61845
+    },
+    {
+      "epoch": 93.01,
+      "grad_norm": 5.7528886795043945,
+      "learning_rate": 6.992481203007518e-07,
+      "loss": 0.1765,
+      "step": 61850
+    },
+    {
+      "epoch": 93.02,
+      "grad_norm": 6.218891620635986,
+      "learning_rate": 6.977443609022557e-07,
+      "loss": 0.1465,
+      "step": 61860
+    },
+    {
+      "epoch": 93.04,
+      "grad_norm": 3.1827406883239746,
+      "learning_rate": 6.962406015037594e-07,
+      "loss": 0.1436,
+      "step": 61870
+    },
+    {
+      "epoch": 93.05,
+      "grad_norm": 2.038198232650757,
+      "learning_rate": 6.947368421052631e-07,
+      "loss": 0.1519,
+      "step": 61880
+    },
+    {
+      "epoch": 93.07,
+      "grad_norm": 4.038834095001221,
+      "learning_rate": 6.93233082706767e-07,
+      "loss": 0.1381,
+      "step": 61890
+    },
+    {
+      "epoch": 93.08,
+      "grad_norm": 8.468280792236328,
+      "learning_rate": 6.917293233082707e-07,
+      "loss": 0.1598,
+      "step": 61900
+    },
+    {
+      "epoch": 93.1,
+      "grad_norm": 5.541633129119873,
+      "learning_rate": 6.902255639097745e-07,
+      "loss": 0.1888,
+      "step": 61910
+    },
+    {
+      "epoch": 93.11,
+      "grad_norm": 6.266165733337402,
+      "learning_rate": 6.887218045112783e-07,
+      "loss": 0.1338,
+      "step": 61920
+    },
+    {
+      "epoch": 93.13,
+      "grad_norm": 2.7548935413360596,
+      "learning_rate": 6.87218045112782e-07,
+      "loss": 0.1435,
+      "step": 61930
+    },
+    {
+      "epoch": 93.14,
+      "grad_norm": 6.2727131843566895,
+      "learning_rate": 6.857142857142858e-07,
+      "loss": 0.1858,
+      "step": 61940
+    },
+    {
+      "epoch": 93.16,
+      "grad_norm": 8.00541877746582,
+      "learning_rate": 6.842105263157896e-07,
+      "loss": 0.1809,
+      "step": 61950
+    },
+    {
+      "epoch": 93.17,
+      "grad_norm": 4.776533126831055,
+      "learning_rate": 6.827067669172933e-07,
+      "loss": 0.1105,
+      "step": 61960
+    },
+    {
+      "epoch": 93.19,
+      "grad_norm": 0.5957467555999756,
+      "learning_rate": 6.812030075187971e-07,
+      "loss": 0.1535,
+      "step": 61970
+    },
+    {
+      "epoch": 93.2,
+      "grad_norm": 2.1154375076293945,
+      "learning_rate": 6.796992481203008e-07,
+      "loss": 0.1391,
+      "step": 61980
+    },
+    {
+      "epoch": 93.22,
+      "grad_norm": 3.501743793487549,
+      "learning_rate": 6.781954887218045e-07,
+      "loss": 0.1809,
+      "step": 61990
+    },
+    {
+      "epoch": 93.23,
+      "grad_norm": 19.21070098876953,
+      "learning_rate": 6.766917293233083e-07,
+      "loss": 0.2236,
+      "step": 62000
+    },
+    {
+      "epoch": 93.25,
+      "grad_norm": 8.024877548217773,
+      "learning_rate": 6.751879699248121e-07,
+      "loss": 0.1522,
+      "step": 62010
+    },
+    {
+      "epoch": 93.26,
+      "grad_norm": 3.7269623279571533,
+      "learning_rate": 6.736842105263158e-07,
+      "loss": 0.1898,
+      "step": 62020
+    },
+    {
+      "epoch": 93.28,
+      "grad_norm": 3.2718605995178223,
+      "learning_rate": 6.721804511278196e-07,
+      "loss": 0.1798,
+      "step": 62030
+    },
+    {
+      "epoch": 93.29,
+      "grad_norm": 5.502279281616211,
+      "learning_rate": 6.706766917293234e-07,
+      "loss": 0.1604,
+      "step": 62040
+    },
+    {
+      "epoch": 93.31,
+      "grad_norm": 5.675206661224365,
+      "learning_rate": 6.691729323308272e-07,
+      "loss": 0.1656,
+      "step": 62050
+    },
+    {
+      "epoch": 93.32,
+      "grad_norm": 5.183054447174072,
+      "learning_rate": 6.676691729323309e-07,
+      "loss": 0.2103,
+      "step": 62060
+    },
+    {
+      "epoch": 93.34,
+      "grad_norm": 1.9447113275527954,
+      "learning_rate": 6.661654135338347e-07,
+      "loss": 0.1621,
+      "step": 62070
+    },
+    {
+      "epoch": 93.35,
+      "grad_norm": 3.6219539642333984,
+      "learning_rate": 6.646616541353385e-07,
+      "loss": 0.126,
+      "step": 62080
+    },
+    {
+      "epoch": 93.37,
+      "grad_norm": 4.979419708251953,
+      "learning_rate": 6.631578947368422e-07,
+      "loss": 0.1749,
+      "step": 62090
+    },
+    {
+      "epoch": 93.38,
+      "grad_norm": 7.001568794250488,
+      "learning_rate": 6.616541353383458e-07,
+      "loss": 0.1297,
+      "step": 62100
+    },
+    {
+      "epoch": 93.4,
+      "grad_norm": 7.3389387130737305,
+      "learning_rate": 6.601503759398496e-07,
+      "loss": 0.1794,
+      "step": 62110
+    },
+    {
+      "epoch": 93.41,
+      "grad_norm": 4.73595666885376,
+      "learning_rate": 6.586466165413534e-07,
+      "loss": 0.1852,
+      "step": 62120
+    },
+    {
+      "epoch": 93.43,
+      "grad_norm": 5.426881790161133,
+      "learning_rate": 6.571428571428571e-07,
+      "loss": 0.2275,
+      "step": 62130
+    },
+    {
+      "epoch": 93.44,
+      "grad_norm": 7.103865146636963,
+      "learning_rate": 6.55639097744361e-07,
+      "loss": 0.1765,
+      "step": 62140
+    },
+    {
+      "epoch": 93.46,
+      "grad_norm": 5.244080066680908,
+      "learning_rate": 6.541353383458647e-07,
+      "loss": 0.1559,
+      "step": 62150
+    },
+    {
+      "epoch": 93.47,
+      "grad_norm": 11.155570030212402,
+      "learning_rate": 6.526315789473684e-07,
+      "loss": 0.191,
+      "step": 62160
+    },
+    {
+      "epoch": 93.49,
+      "grad_norm": 3.5410802364349365,
+      "learning_rate": 6.511278195488723e-07,
+      "loss": 0.1308,
+      "step": 62170
+    },
+    {
+      "epoch": 93.5,
+      "grad_norm": 5.664346218109131,
+      "learning_rate": 6.49624060150376e-07,
+      "loss": 0.1734,
+      "step": 62180
+    },
+    {
+      "epoch": 93.52,
+      "grad_norm": 3.049161434173584,
+      "learning_rate": 6.481203007518797e-07,
+      "loss": 0.1595,
+      "step": 62190
+    },
+    {
+      "epoch": 93.53,
+      "grad_norm": 3.4412500858306885,
+      "learning_rate": 6.466165413533836e-07,
+      "loss": 0.1986,
+      "step": 62200
+    },
+    {
+      "epoch": 93.55,
+      "grad_norm": 5.76616096496582,
+      "learning_rate": 6.451127819548873e-07,
+      "loss": 0.2203,
+      "step": 62210
+    },
+    {
+      "epoch": 93.56,
+      "grad_norm": 3.9202961921691895,
+      "learning_rate": 6.436090225563909e-07,
+      "loss": 0.1739,
+      "step": 62220
+    },
+    {
+      "epoch": 93.58,
+      "grad_norm": 5.559597969055176,
+      "learning_rate": 6.421052631578948e-07,
+      "loss": 0.1472,
+      "step": 62230
+    },
+    {
+      "epoch": 93.59,
+      "grad_norm": 6.698195934295654,
+      "learning_rate": 6.406015037593985e-07,
+      "loss": 0.1822,
+      "step": 62240
+    },
+    {
+      "epoch": 93.61,
+      "grad_norm": 7.657590866088867,
+      "learning_rate": 6.390977443609022e-07,
+      "loss": 0.1469,
+      "step": 62250
+    },
+    {
+      "epoch": 93.62,
+      "grad_norm": 9.91557788848877,
+      "learning_rate": 6.375939849624061e-07,
+      "loss": 0.1382,
+      "step": 62260
+    },
+    {
+      "epoch": 93.64,
+      "grad_norm": 4.059289932250977,
+      "learning_rate": 6.360902255639098e-07,
+      "loss": 0.1408,
+      "step": 62270
+    },
+    {
+      "epoch": 93.65,
+      "grad_norm": 3.5155844688415527,
+      "learning_rate": 6.345864661654136e-07,
+      "loss": 0.153,
+      "step": 62280
+    },
+    {
+      "epoch": 93.67,
+      "grad_norm": 2.9289627075195312,
+      "learning_rate": 6.330827067669174e-07,
+      "loss": 0.133,
+      "step": 62290
+    },
+    {
+      "epoch": 93.68,
+      "grad_norm": 1.8064905405044556,
+      "learning_rate": 6.315789473684211e-07,
+      "loss": 0.1543,
+      "step": 62300
+    },
+    {
+      "epoch": 93.7,
+      "grad_norm": 5.917304039001465,
+      "learning_rate": 6.300751879699249e-07,
+      "loss": 0.1764,
+      "step": 62310
+    },
+    {
+      "epoch": 93.71,
+      "grad_norm": 9.10999584197998,
+      "learning_rate": 6.285714285714287e-07,
+      "loss": 0.1601,
+      "step": 62320
+    },
+    {
+      "epoch": 93.73,
+      "grad_norm": 9.873708724975586,
+      "learning_rate": 6.270676691729324e-07,
+      "loss": 0.2098,
+      "step": 62330
+    },
+    {
+      "epoch": 93.74,
+      "grad_norm": 4.124228000640869,
+      "learning_rate": 6.255639097744362e-07,
+      "loss": 0.1395,
+      "step": 62340
+    },
+    {
+      "epoch": 93.76,
+      "grad_norm": 5.858664035797119,
+      "learning_rate": 6.240601503759399e-07,
+      "loss": 0.2124,
+      "step": 62350
+    },
+    {
+      "epoch": 93.77,
+      "grad_norm": 5.991408348083496,
+      "learning_rate": 6.225563909774437e-07,
+      "loss": 0.1532,
+      "step": 62360
+    },
+    {
+      "epoch": 93.79,
+      "grad_norm": 3.827173948287964,
+      "learning_rate": 6.210526315789474e-07,
+      "loss": 0.1139,
+      "step": 62370
+    },
+    {
+      "epoch": 93.8,
+      "grad_norm": 3.3806679248809814,
+      "learning_rate": 6.195488721804512e-07,
+      "loss": 0.1869,
+      "step": 62380
+    },
+    {
+      "epoch": 93.82,
+      "grad_norm": 5.624974250793457,
+      "learning_rate": 6.180451127819549e-07,
+      "loss": 0.2374,
+      "step": 62390
+    },
+    {
+      "epoch": 93.83,
+      "grad_norm": 3.611284017562866,
+      "learning_rate": 6.165413533834587e-07,
+      "loss": 0.2001,
+      "step": 62400
+    },
+    {
+      "epoch": 93.85,
+      "grad_norm": 3.4336488246917725,
+      "learning_rate": 6.150375939849625e-07,
+      "loss": 0.2229,
+      "step": 62410
+    },
+    {
+      "epoch": 93.86,
+      "grad_norm": 5.959348201751709,
+      "learning_rate": 6.135338345864662e-07,
+      "loss": 0.1408,
+      "step": 62420
+    },
+    {
+      "epoch": 93.88,
+      "grad_norm": 4.848119258880615,
+      "learning_rate": 6.1203007518797e-07,
+      "loss": 0.1566,
+      "step": 62430
+    },
+    {
+      "epoch": 93.89,
+      "grad_norm": 4.178713798522949,
+      "learning_rate": 6.105263157894738e-07,
+      "loss": 0.1452,
+      "step": 62440
+    },
+    {
+      "epoch": 93.91,
+      "grad_norm": 7.14044713973999,
+      "learning_rate": 6.090225563909775e-07,
+      "loss": 0.15,
+      "step": 62450
+    },
+    {
+      "epoch": 93.92,
+      "grad_norm": 6.114139556884766,
+      "learning_rate": 6.075187969924812e-07,
+      "loss": 0.1249,
+      "step": 62460
+    },
+    {
+      "epoch": 93.94,
+      "grad_norm": 2.516326904296875,
+      "learning_rate": 6.06015037593985e-07,
+      "loss": 0.112,
+      "step": 62470
+    },
+    {
+      "epoch": 93.95,
+      "grad_norm": 6.2728705406188965,
+      "learning_rate": 6.045112781954888e-07,
+      "loss": 0.1425,
+      "step": 62480
+    },
+    {
+      "epoch": 93.97,
+      "grad_norm": 5.063921928405762,
+      "learning_rate": 6.030075187969925e-07,
+      "loss": 0.1956,
+      "step": 62490
+    },
+    {
+      "epoch": 93.98,
+      "grad_norm": 1.3252296447753906,
+      "learning_rate": 6.015037593984962e-07,
+      "loss": 0.1332,
+      "step": 62500
+    },
+    {
+      "epoch": 94.0,
+      "grad_norm": 0.03227702155709267,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 0.1855,
+      "step": 62510
+    },
+    {
+      "epoch": 94.0,
+      "eval_accuracy": 0.9303,
+      "eval_loss": 0.33146244287490845,
+      "eval_runtime": 85.2753,
+      "eval_samples_per_second": 117.267,
+      "eval_steps_per_second": 0.469,
+      "step": 62510
+    },
+    {
+      "epoch": 94.02,
+      "grad_norm": 4.525468826293945,
+      "learning_rate": 5.984962406015038e-07,
+      "loss": 0.1478,
+      "step": 62520
+    },
+    {
+      "epoch": 94.03,
+      "grad_norm": 3.999218702316284,
+      "learning_rate": 5.969924812030075e-07,
+      "loss": 0.1595,
+      "step": 62530
+    },
+    {
+      "epoch": 94.05,
+      "grad_norm": 5.01107120513916,
+      "learning_rate": 5.954887218045114e-07,
+      "loss": 0.1798,
+      "step": 62540
+    },
+    {
+      "epoch": 94.06,
+      "grad_norm": 3.2505600452423096,
+      "learning_rate": 5.939849624060151e-07,
+      "loss": 0.1772,
+      "step": 62550
+    },
+    {
+      "epoch": 94.08,
+      "grad_norm": 5.565808296203613,
+      "learning_rate": 5.924812030075188e-07,
+      "loss": 0.163,
+      "step": 62560
+    },
+    {
+      "epoch": 94.09,
+      "grad_norm": 5.079768180847168,
+      "learning_rate": 5.909774436090226e-07,
+      "loss": 0.1596,
+      "step": 62570
+    },
+    {
+      "epoch": 94.11,
+      "grad_norm": 6.390072345733643,
+      "learning_rate": 5.894736842105263e-07,
+      "loss": 0.1867,
+      "step": 62580
+    },
+    {
+      "epoch": 94.12,
+      "grad_norm": 3.9409000873565674,
+      "learning_rate": 5.879699248120301e-07,
+      "loss": 0.1428,
+      "step": 62590
+    },
+    {
+      "epoch": 94.14,
+      "grad_norm": 7.87678337097168,
+      "learning_rate": 5.864661654135339e-07,
+      "loss": 0.139,
+      "step": 62600
+    },
+    {
+      "epoch": 94.15,
+      "grad_norm": 3.9159417152404785,
+      "learning_rate": 5.849624060150377e-07,
+      "loss": 0.1449,
+      "step": 62610
+    },
+    {
+      "epoch": 94.17,
+      "grad_norm": 5.255906581878662,
+      "learning_rate": 5.834586466165414e-07,
+      "loss": 0.1411,
+      "step": 62620
+    },
+    {
+      "epoch": 94.18,
+      "grad_norm": 4.5999436378479,
+      "learning_rate": 5.819548872180452e-07,
+      "loss": 0.1697,
+      "step": 62630
+    },
+    {
+      "epoch": 94.2,
+      "grad_norm": 10.575860977172852,
+      "learning_rate": 5.804511278195489e-07,
+      "loss": 0.1787,
+      "step": 62640
+    },
+    {
+      "epoch": 94.21,
+      "grad_norm": 8.932710647583008,
+      "learning_rate": 5.789473684210526e-07,
+      "loss": 0.1589,
+      "step": 62650
+    },
+    {
+      "epoch": 94.23,
+      "grad_norm": 8.139424324035645,
+      "learning_rate": 5.774436090225565e-07,
+      "loss": 0.1776,
+      "step": 62660
+    },
+    {
+      "epoch": 94.24,
+      "grad_norm": 1.6099450588226318,
+      "learning_rate": 5.759398496240602e-07,
+      "loss": 0.1939,
+      "step": 62670
+    },
+    {
+      "epoch": 94.26,
+      "grad_norm": 2.6869306564331055,
+      "learning_rate": 5.74436090225564e-07,
+      "loss": 0.1988,
+      "step": 62680
+    },
+    {
+      "epoch": 94.27,
+      "grad_norm": 5.661433219909668,
+      "learning_rate": 5.729323308270677e-07,
+      "loss": 0.2013,
+      "step": 62690
+    },
+    {
+      "epoch": 94.29,
+      "grad_norm": 6.6180315017700195,
+      "learning_rate": 5.714285714285715e-07,
+      "loss": 0.188,
+      "step": 62700
+    },
+    {
+      "epoch": 94.3,
+      "grad_norm": 4.5534563064575195,
+      "learning_rate": 5.699248120300752e-07,
+      "loss": 0.1996,
+      "step": 62710
+    },
+    {
+      "epoch": 94.32,
+      "grad_norm": 5.5353193283081055,
+      "learning_rate": 5.68421052631579e-07,
+      "loss": 0.1214,
+      "step": 62720
+    },
+    {
+      "epoch": 94.33,
+      "grad_norm": 3.558548927307129,
+      "learning_rate": 5.669172932330828e-07,
+      "loss": 0.1451,
+      "step": 62730
+    },
+    {
+      "epoch": 94.35,
+      "grad_norm": 4.777967929840088,
+      "learning_rate": 5.654135338345865e-07,
+      "loss": 0.1993,
+      "step": 62740
+    },
+    {
+      "epoch": 94.36,
+      "grad_norm": 5.252827167510986,
+      "learning_rate": 5.639097744360903e-07,
+      "loss": 0.1442,
+      "step": 62750
+    },
+    {
+      "epoch": 94.38,
+      "grad_norm": 6.862250804901123,
+      "learning_rate": 5.62406015037594e-07,
+      "loss": 0.191,
+      "step": 62760
+    },
+    {
+      "epoch": 94.39,
+      "grad_norm": 7.284124851226807,
+      "learning_rate": 5.609022556390978e-07,
+      "loss": 0.1813,
+      "step": 62770
+    },
+    {
+      "epoch": 94.41,
+      "grad_norm": 4.488624572753906,
+      "learning_rate": 5.593984962406016e-07,
+      "loss": 0.1546,
+      "step": 62780
+    },
+    {
+      "epoch": 94.42,
+      "grad_norm": 4.856323719024658,
+      "learning_rate": 5.578947368421053e-07,
+      "loss": 0.1553,
+      "step": 62790
+    },
+    {
+      "epoch": 94.44,
+      "grad_norm": 5.90994119644165,
+      "learning_rate": 5.563909774436091e-07,
+      "loss": 0.168,
+      "step": 62800
+    },
+    {
+      "epoch": 94.45,
+      "grad_norm": 4.328577041625977,
+      "learning_rate": 5.548872180451127e-07,
+      "loss": 0.1539,
+      "step": 62810
+    },
+    {
+      "epoch": 94.47,
+      "grad_norm": 2.8183932304382324,
+      "learning_rate": 5.533834586466166e-07,
+      "loss": 0.1973,
+      "step": 62820
+    },
+    {
+      "epoch": 94.48,
+      "grad_norm": 5.438328266143799,
+      "learning_rate": 5.518796992481203e-07,
+      "loss": 0.17,
+      "step": 62830
+    },
+    {
+      "epoch": 94.5,
+      "grad_norm": 4.676559925079346,
+      "learning_rate": 5.503759398496241e-07,
+      "loss": 0.131,
+      "step": 62840
+    },
+    {
+      "epoch": 94.51,
+      "grad_norm": 8.041285514831543,
+      "learning_rate": 5.488721804511279e-07,
+      "loss": 0.1931,
+      "step": 62850
+    },
+    {
+      "epoch": 94.53,
+      "grad_norm": 6.983100414276123,
+      "learning_rate": 5.473684210526316e-07,
+      "loss": 0.1621,
+      "step": 62860
+    },
+    {
+      "epoch": 94.54,
+      "grad_norm": 8.173839569091797,
+      "learning_rate": 5.458646616541354e-07,
+      "loss": 0.1366,
+      "step": 62870
+    },
+    {
+      "epoch": 94.56,
+      "grad_norm": 4.067495346069336,
+      "learning_rate": 5.443609022556391e-07,
+      "loss": 0.137,
+      "step": 62880
+    },
+    {
+      "epoch": 94.57,
+      "grad_norm": 1.778714656829834,
+      "learning_rate": 5.428571428571429e-07,
+      "loss": 0.1684,
+      "step": 62890
+    },
+    {
+      "epoch": 94.59,
+      "grad_norm": 4.966789722442627,
+      "learning_rate": 5.413533834586466e-07,
+      "loss": 0.1576,
+      "step": 62900
+    },
+    {
+      "epoch": 94.6,
+      "grad_norm": 6.58605432510376,
+      "learning_rate": 5.398496240601505e-07,
+      "loss": 0.1515,
+      "step": 62910
+    },
+    {
+      "epoch": 94.62,
+      "grad_norm": 5.37683629989624,
+      "learning_rate": 5.383458646616542e-07,
+      "loss": 0.2039,
+      "step": 62920
+    },
+    {
+      "epoch": 94.63,
+      "grad_norm": 4.260867595672607,
+      "learning_rate": 5.368421052631579e-07,
+      "loss": 0.169,
+      "step": 62930
+    },
+    {
+      "epoch": 94.65,
+      "grad_norm": 8.92151927947998,
+      "learning_rate": 5.353383458646617e-07,
+      "loss": 0.1674,
+      "step": 62940
+    },
+    {
+      "epoch": 94.66,
+      "grad_norm": 5.447167873382568,
+      "learning_rate": 5.338345864661654e-07,
+      "loss": 0.1623,
+      "step": 62950
+    },
+    {
+      "epoch": 94.68,
+      "grad_norm": 5.827322483062744,
+      "learning_rate": 5.323308270676692e-07,
+      "loss": 0.1627,
+      "step": 62960
+    },
+    {
+      "epoch": 94.69,
+      "grad_norm": 3.478543281555176,
+      "learning_rate": 5.30827067669173e-07,
+      "loss": 0.1619,
+      "step": 62970
+    },
+    {
+      "epoch": 94.71,
+      "grad_norm": 2.6951584815979004,
+      "learning_rate": 5.293233082706768e-07,
+      "loss": 0.1155,
+      "step": 62980
+    },
+    {
+      "epoch": 94.72,
+      "grad_norm": 4.290323734283447,
+      "learning_rate": 5.278195488721805e-07,
+      "loss": 0.2299,
+      "step": 62990
+    },
+    {
+      "epoch": 94.74,
+      "grad_norm": 2.3207318782806396,
+      "learning_rate": 5.263157894736843e-07,
+      "loss": 0.1377,
+      "step": 63000
+    },
+    {
+      "epoch": 94.75,
+      "grad_norm": 4.391146659851074,
+      "learning_rate": 5.24812030075188e-07,
+      "loss": 0.1383,
+      "step": 63010
+    },
+    {
+      "epoch": 94.77,
+      "grad_norm": 8.625882148742676,
+      "learning_rate": 5.233082706766917e-07,
+      "loss": 0.1567,
+      "step": 63020
+    },
+    {
+      "epoch": 94.78,
+      "grad_norm": 5.556321620941162,
+      "learning_rate": 5.218045112781956e-07,
+      "loss": 0.1483,
+      "step": 63030
+    },
+    {
+      "epoch": 94.8,
+      "grad_norm": 3.1565639972686768,
+      "learning_rate": 5.203007518796993e-07,
+      "loss": 0.1173,
+      "step": 63040
+    },
+    {
+      "epoch": 94.81,
+      "grad_norm": 1.9764903783798218,
+      "learning_rate": 5.18796992481203e-07,
+      "loss": 0.1378,
+      "step": 63050
+    },
+    {
+      "epoch": 94.83,
+      "grad_norm": 6.802098274230957,
+      "learning_rate": 5.172932330827068e-07,
+      "loss": 0.1888,
+      "step": 63060
+    },
+    {
+      "epoch": 94.84,
+      "grad_norm": 10.17620849609375,
+      "learning_rate": 5.157894736842106e-07,
+      "loss": 0.1397,
+      "step": 63070
+    },
+    {
+      "epoch": 94.86,
+      "grad_norm": 2.024324893951416,
+      "learning_rate": 5.142857142857143e-07,
+      "loss": 0.1553,
+      "step": 63080
+    },
+    {
+      "epoch": 94.87,
+      "grad_norm": 4.072267532348633,
+      "learning_rate": 5.127819548872181e-07,
+      "loss": 0.1659,
+      "step": 63090
+    },
+    {
+      "epoch": 94.89,
+      "grad_norm": 5.023469924926758,
+      "learning_rate": 5.112781954887219e-07,
+      "loss": 0.1639,
+      "step": 63100
+    },
+    {
+      "epoch": 94.9,
+      "grad_norm": 8.388982772827148,
+      "learning_rate": 5.097744360902256e-07,
+      "loss": 0.1692,
+      "step": 63110
+    },
+    {
+      "epoch": 94.92,
+      "grad_norm": 3.7290244102478027,
+      "learning_rate": 5.082706766917294e-07,
+      "loss": 0.1179,
+      "step": 63120
+    },
+    {
+      "epoch": 94.93,
+      "grad_norm": 7.133794784545898,
+      "learning_rate": 5.067669172932331e-07,
+      "loss": 0.1777,
+      "step": 63130
+    },
+    {
+      "epoch": 94.95,
+      "grad_norm": 1.3478121757507324,
+      "learning_rate": 5.052631578947369e-07,
+      "loss": 0.1594,
+      "step": 63140
+    },
+    {
+      "epoch": 94.96,
+      "grad_norm": 2.427905321121216,
+      "learning_rate": 5.037593984962407e-07,
+      "loss": 0.1275,
+      "step": 63150
+    },
+    {
+      "epoch": 94.98,
+      "grad_norm": 4.035090446472168,
+      "learning_rate": 5.022556390977444e-07,
+      "loss": 0.1687,
+      "step": 63160
+    },
+    {
+      "epoch": 94.99,
+      "grad_norm": 3.4016542434692383,
+      "learning_rate": 5.007518796992482e-07,
+      "loss": 0.1747,
+      "step": 63170
+    },
+    {
+      "epoch": 95.0,
+      "eval_accuracy": 0.9295,
+      "eval_loss": 0.33236411213874817,
+      "eval_runtime": 84.9502,
+      "eval_samples_per_second": 117.716,
+      "eval_steps_per_second": 0.471,
+      "step": 63175
+    },
+    {
+      "epoch": 95.01,
+      "grad_norm": 6.205942630767822,
+      "learning_rate": 4.99248120300752e-07,
+      "loss": 0.1474,
+      "step": 63180
+    },
+    {
+      "epoch": 95.02,
+      "grad_norm": 6.6230788230896,
+      "learning_rate": 4.977443609022557e-07,
+      "loss": 0.1549,
+      "step": 63190
+    },
+    {
+      "epoch": 95.04,
+      "grad_norm": 8.845964431762695,
+      "learning_rate": 4.962406015037594e-07,
+      "loss": 0.1962,
+      "step": 63200
+    },
+    {
+      "epoch": 95.05,
+      "grad_norm": 3.8196167945861816,
+      "learning_rate": 4.947368421052632e-07,
+      "loss": 0.1145,
+      "step": 63210
+    },
+    {
+      "epoch": 95.07,
+      "grad_norm": 4.0402607917785645,
+      "learning_rate": 4.93233082706767e-07,
+      "loss": 0.1538,
+      "step": 63220
+    },
+    {
+      "epoch": 95.08,
+      "grad_norm": 0.3398207724094391,
+      "learning_rate": 4.917293233082707e-07,
+      "loss": 0.1827,
+      "step": 63230
+    },
+    {
+      "epoch": 95.1,
+      "grad_norm": 5.126389503479004,
+      "learning_rate": 4.902255639097745e-07,
+      "loss": 0.1691,
+      "step": 63240
+    },
+    {
+      "epoch": 95.11,
+      "grad_norm": 5.045697212219238,
+      "learning_rate": 4.887218045112782e-07,
+      "loss": 0.1115,
+      "step": 63250
+    },
+    {
+      "epoch": 95.13,
+      "grad_norm": 6.75160551071167,
+      "learning_rate": 4.87218045112782e-07,
+      "loss": 0.1862,
+      "step": 63260
+    },
+    {
+      "epoch": 95.14,
+      "grad_norm": 3.8719749450683594,
+      "learning_rate": 4.857142857142857e-07,
+      "loss": 0.1878,
+      "step": 63270
+    },
+    {
+      "epoch": 95.16,
+      "grad_norm": 6.269413471221924,
+      "learning_rate": 4.842105263157895e-07,
+      "loss": 0.1563,
+      "step": 63280
+    },
+    {
+      "epoch": 95.17,
+      "grad_norm": 4.618571758270264,
+      "learning_rate": 4.827067669172933e-07,
+      "loss": 0.1733,
+      "step": 63290
+    },
+    {
+      "epoch": 95.19,
+      "grad_norm": 4.135458469390869,
+      "learning_rate": 4.81203007518797e-07,
+      "loss": 0.2016,
+      "step": 63300
+    },
+    {
+      "epoch": 95.2,
+      "grad_norm": 7.51487398147583,
+      "learning_rate": 4.796992481203008e-07,
+      "loss": 0.1643,
+      "step": 63310
+    },
+    {
+      "epoch": 95.22,
+      "grad_norm": 4.156686305999756,
+      "learning_rate": 4.781954887218045e-07,
+      "loss": 0.1999,
+      "step": 63320
+    },
+    {
+      "epoch": 95.23,
+      "grad_norm": 8.415448188781738,
+      "learning_rate": 4.766917293233083e-07,
+      "loss": 0.2051,
+      "step": 63330
+    },
+    {
+      "epoch": 95.25,
+      "grad_norm": 3.9293386936187744,
+      "learning_rate": 4.7518796992481207e-07,
+      "loss": 0.1811,
+      "step": 63340
+    },
+    {
+      "epoch": 95.26,
+      "grad_norm": 5.730102062225342,
+      "learning_rate": 4.7368421052631585e-07,
+      "loss": 0.1573,
+      "step": 63350
+    },
+    {
+      "epoch": 95.28,
+      "grad_norm": 5.62534236907959,
+      "learning_rate": 4.7218045112781963e-07,
+      "loss": 0.1802,
+      "step": 63360
+    },
+    {
+      "epoch": 95.29,
+      "grad_norm": 5.367505073547363,
+      "learning_rate": 4.706766917293233e-07,
+      "loss": 0.1676,
+      "step": 63370
+    },
+    {
+      "epoch": 95.31,
+      "grad_norm": 5.48174524307251,
+      "learning_rate": 4.691729323308271e-07,
+      "loss": 0.1394,
+      "step": 63380
+    },
+    {
+      "epoch": 95.32,
+      "grad_norm": 5.894782543182373,
+      "learning_rate": 4.676691729323309e-07,
+      "loss": 0.1703,
+      "step": 63390
+    },
+    {
+      "epoch": 95.34,
+      "grad_norm": 3.4429852962493896,
+      "learning_rate": 4.661654135338346e-07,
+      "loss": 0.1782,
+      "step": 63400
+    },
+    {
+      "epoch": 95.35,
+      "grad_norm": 5.609094142913818,
+      "learning_rate": 4.646616541353384e-07,
+      "loss": 0.2212,
+      "step": 63410
+    },
+    {
+      "epoch": 95.37,
+      "grad_norm": 6.826117992401123,
+      "learning_rate": 4.631578947368422e-07,
+      "loss": 0.1549,
+      "step": 63420
+    },
+    {
+      "epoch": 95.38,
+      "grad_norm": 6.199655532836914,
+      "learning_rate": 4.6165413533834585e-07,
+      "loss": 0.2035,
+      "step": 63430
+    },
+    {
+      "epoch": 95.4,
+      "grad_norm": 3.159210443496704,
+      "learning_rate": 4.6015037593984964e-07,
+      "loss": 0.168,
+      "step": 63440
+    },
+    {
+      "epoch": 95.41,
+      "grad_norm": 4.340883255004883,
+      "learning_rate": 4.586466165413534e-07,
+      "loss": 0.1506,
+      "step": 63450
+    },
+    {
+      "epoch": 95.43,
+      "grad_norm": 6.630001068115234,
+      "learning_rate": 4.571428571428572e-07,
+      "loss": 0.1342,
+      "step": 63460
+    },
+    {
+      "epoch": 95.44,
+      "grad_norm": 6.111904144287109,
+      "learning_rate": 4.5563909774436094e-07,
+      "loss": 0.1454,
+      "step": 63470
+    },
+    {
+      "epoch": 95.46,
+      "grad_norm": 8.334033012390137,
+      "learning_rate": 4.541353383458647e-07,
+      "loss": 0.1583,
+      "step": 63480
+    },
+    {
+      "epoch": 95.47,
+      "grad_norm": 3.537055730819702,
+      "learning_rate": 4.526315789473685e-07,
+      "loss": 0.1506,
+      "step": 63490
+    },
+    {
+      "epoch": 95.49,
+      "grad_norm": 4.5531792640686035,
+      "learning_rate": 4.511278195488722e-07,
+      "loss": 0.1923,
+      "step": 63500
+    },
+    {
+      "epoch": 95.5,
+      "grad_norm": 4.920889854431152,
+      "learning_rate": 4.4962406015037597e-07,
+      "loss": 0.1556,
+      "step": 63510
+    },
+    {
+      "epoch": 95.52,
+      "grad_norm": 4.547194004058838,
+      "learning_rate": 4.4812030075187975e-07,
+      "loss": 0.1974,
+      "step": 63520
+    },
+    {
+      "epoch": 95.53,
+      "grad_norm": 3.0299079418182373,
+      "learning_rate": 4.466165413533835e-07,
+      "loss": 0.1801,
+      "step": 63530
+    },
+    {
+      "epoch": 95.55,
+      "grad_norm": 5.605795383453369,
+      "learning_rate": 4.4511278195488726e-07,
+      "loss": 0.1776,
+      "step": 63540
+    },
+    {
+      "epoch": 95.56,
+      "grad_norm": 8.868003845214844,
+      "learning_rate": 4.4360902255639105e-07,
+      "loss": 0.1175,
+      "step": 63550
+    },
+    {
+      "epoch": 95.58,
+      "grad_norm": 5.188925266265869,
+      "learning_rate": 4.421052631578947e-07,
+      "loss": 0.1546,
+      "step": 63560
+    },
+    {
+      "epoch": 95.59,
+      "grad_norm": 3.829498529434204,
+      "learning_rate": 4.406015037593985e-07,
+      "loss": 0.2129,
+      "step": 63570
+    },
+    {
+      "epoch": 95.61,
+      "grad_norm": 7.33263635635376,
+      "learning_rate": 4.390977443609023e-07,
+      "loss": 0.2169,
+      "step": 63580
+    },
+    {
+      "epoch": 95.62,
+      "grad_norm": 6.37952995300293,
+      "learning_rate": 4.375939849624061e-07,
+      "loss": 0.1823,
+      "step": 63590
+    },
+    {
+      "epoch": 95.64,
+      "grad_norm": 4.5436201095581055,
+      "learning_rate": 4.360902255639098e-07,
+      "loss": 0.1293,
+      "step": 63600
+    },
+    {
+      "epoch": 95.65,
+      "grad_norm": 4.836911201477051,
+      "learning_rate": 4.345864661654136e-07,
+      "loss": 0.1519,
+      "step": 63610
+    },
+    {
+      "epoch": 95.67,
+      "grad_norm": 3.8880579471588135,
+      "learning_rate": 4.330827067669173e-07,
+      "loss": 0.1459,
+      "step": 63620
+    },
+    {
+      "epoch": 95.68,
+      "grad_norm": 5.021939277648926,
+      "learning_rate": 4.3157894736842105e-07,
+      "loss": 0.2607,
+      "step": 63630
+    },
+    {
+      "epoch": 95.7,
+      "grad_norm": 5.220824241638184,
+      "learning_rate": 4.3007518796992484e-07,
+      "loss": 0.2041,
+      "step": 63640
+    },
+    {
+      "epoch": 95.71,
+      "grad_norm": 6.0544514656066895,
+      "learning_rate": 4.285714285714286e-07,
+      "loss": 0.1999,
+      "step": 63650
+    },
+    {
+      "epoch": 95.73,
+      "grad_norm": 3.8796777725219727,
+      "learning_rate": 4.270676691729324e-07,
+      "loss": 0.1769,
+      "step": 63660
+    },
+    {
+      "epoch": 95.74,
+      "grad_norm": 3.2138679027557373,
+      "learning_rate": 4.2556390977443613e-07,
+      "loss": 0.1565,
+      "step": 63670
+    },
+    {
+      "epoch": 95.76,
+      "grad_norm": 6.524970054626465,
+      "learning_rate": 4.2406015037593987e-07,
+      "loss": 0.1141,
+      "step": 63680
+    },
+    {
+      "epoch": 95.77,
+      "grad_norm": 5.798688888549805,
+      "learning_rate": 4.2255639097744365e-07,
+      "loss": 0.182,
+      "step": 63690
+    },
+    {
+      "epoch": 95.79,
+      "grad_norm": 4.808920383453369,
+      "learning_rate": 4.210526315789474e-07,
+      "loss": 0.1496,
+      "step": 63700
+    },
+    {
+      "epoch": 95.8,
+      "grad_norm": 2.4658915996551514,
+      "learning_rate": 4.1954887218045116e-07,
+      "loss": 0.1157,
+      "step": 63710
+    },
+    {
+      "epoch": 95.82,
+      "grad_norm": 3.5132291316986084,
+      "learning_rate": 4.1804511278195495e-07,
+      "loss": 0.1528,
+      "step": 63720
+    },
+    {
+      "epoch": 95.83,
+      "grad_norm": 8.57204818725586,
+      "learning_rate": 4.165413533834587e-07,
+      "loss": 0.2021,
+      "step": 63730
+    },
+    {
+      "epoch": 95.85,
+      "grad_norm": 6.247355937957764,
+      "learning_rate": 4.150375939849624e-07,
+      "loss": 0.1266,
+      "step": 63740
+    },
+    {
+      "epoch": 95.86,
+      "grad_norm": 6.308277606964111,
+      "learning_rate": 4.135338345864662e-07,
+      "loss": 0.1494,
+      "step": 63750
+    },
+    {
+      "epoch": 95.88,
+      "grad_norm": 4.962404251098633,
+      "learning_rate": 4.120300751879699e-07,
+      "loss": 0.1847,
+      "step": 63760
+    },
+    {
+      "epoch": 95.89,
+      "grad_norm": 3.808089017868042,
+      "learning_rate": 4.105263157894737e-07,
+      "loss": 0.1727,
+      "step": 63770
+    },
+    {
+      "epoch": 95.91,
+      "grad_norm": 4.107529640197754,
+      "learning_rate": 4.090225563909775e-07,
+      "loss": 0.1562,
+      "step": 63780
+    },
+    {
+      "epoch": 95.92,
+      "grad_norm": 6.463099956512451,
+      "learning_rate": 4.075187969924813e-07,
+      "loss": 0.145,
+      "step": 63790
+    },
+    {
+      "epoch": 95.94,
+      "grad_norm": 9.739261627197266,
+      "learning_rate": 4.06015037593985e-07,
+      "loss": 0.1749,
+      "step": 63800
+    },
+    {
+      "epoch": 95.95,
+      "grad_norm": 4.032473087310791,
+      "learning_rate": 4.0451127819548874e-07,
+      "loss": 0.1542,
+      "step": 63810
+    },
+    {
+      "epoch": 95.97,
+      "grad_norm": 4.938335418701172,
+      "learning_rate": 4.030075187969925e-07,
+      "loss": 0.1901,
+      "step": 63820
+    },
+    {
+      "epoch": 95.98,
+      "grad_norm": 5.375936985015869,
+      "learning_rate": 4.0150375939849625e-07,
+      "loss": 0.2043,
+      "step": 63830
+    },
+    {
+      "epoch": 96.0,
+      "grad_norm": 15.463907241821289,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 0.1783,
+      "step": 63840
+    },
+    {
+      "epoch": 96.0,
+      "eval_accuracy": 0.9315,
+      "eval_loss": 0.3313089907169342,
+      "eval_runtime": 85.2264,
+      "eval_samples_per_second": 117.334,
+      "eval_steps_per_second": 0.469,
+      "step": 63840
+    },
+    {
+      "epoch": 96.02,
+      "grad_norm": 3.8098232746124268,
+      "learning_rate": 3.984962406015038e-07,
+      "loss": 0.177,
+      "step": 63850
+    },
+    {
+      "epoch": 96.03,
+      "grad_norm": 2.248171806335449,
+      "learning_rate": 3.969924812030076e-07,
+      "loss": 0.1043,
+      "step": 63860
+    },
+    {
+      "epoch": 96.05,
+      "grad_norm": 5.304133415222168,
+      "learning_rate": 3.954887218045113e-07,
+      "loss": 0.1221,
+      "step": 63870
+    },
+    {
+      "epoch": 96.06,
+      "grad_norm": 9.222192764282227,
+      "learning_rate": 3.9398496240601506e-07,
+      "loss": 0.1829,
+      "step": 63880
+    },
+    {
+      "epoch": 96.08,
+      "grad_norm": 5.404617786407471,
+      "learning_rate": 3.9248120300751885e-07,
+      "loss": 0.1773,
+      "step": 63890
+    },
+    {
+      "epoch": 96.09,
+      "grad_norm": 4.069046974182129,
+      "learning_rate": 3.909774436090226e-07,
+      "loss": 0.1894,
+      "step": 63900
+    },
+    {
+      "epoch": 96.11,
+      "grad_norm": 5.911056995391846,
+      "learning_rate": 3.8947368421052636e-07,
+      "loss": 0.1733,
+      "step": 63910
+    },
+    {
+      "epoch": 96.12,
+      "grad_norm": 5.389181137084961,
+      "learning_rate": 3.8796992481203015e-07,
+      "loss": 0.1594,
+      "step": 63920
+    },
+    {
+      "epoch": 96.14,
+      "grad_norm": 5.494384288787842,
+      "learning_rate": 3.864661654135338e-07,
+      "loss": 0.1526,
+      "step": 63930
+    },
+    {
+      "epoch": 96.15,
+      "grad_norm": 3.113043785095215,
+      "learning_rate": 3.849624060150376e-07,
+      "loss": 0.1355,
+      "step": 63940
+    },
+    {
+      "epoch": 96.17,
+      "grad_norm": 6.085422039031982,
+      "learning_rate": 3.834586466165414e-07,
+      "loss": 0.1163,
+      "step": 63950
+    },
+    {
+      "epoch": 96.18,
+      "grad_norm": 4.339461326599121,
+      "learning_rate": 3.819548872180451e-07,
+      "loss": 0.1856,
+      "step": 63960
+    },
+    {
+      "epoch": 96.2,
+      "grad_norm": 5.448220729827881,
+      "learning_rate": 3.804511278195489e-07,
+      "loss": 0.1677,
+      "step": 63970
+    },
+    {
+      "epoch": 96.21,
+      "grad_norm": 8.586206436157227,
+      "learning_rate": 3.789473684210527e-07,
+      "loss": 0.1541,
+      "step": 63980
+    },
+    {
+      "epoch": 96.23,
+      "grad_norm": 7.591394424438477,
+      "learning_rate": 3.7744360902255637e-07,
+      "loss": 0.1415,
+      "step": 63990
+    },
+    {
+      "epoch": 96.24,
+      "grad_norm": 6.6830153465271,
+      "learning_rate": 3.7593984962406015e-07,
+      "loss": 0.1787,
+      "step": 64000
+    },
+    {
+      "epoch": 96.26,
+      "grad_norm": 4.420645713806152,
+      "learning_rate": 3.7443609022556394e-07,
+      "loss": 0.1118,
+      "step": 64010
+    },
+    {
+      "epoch": 96.27,
+      "grad_norm": 6.308743000030518,
+      "learning_rate": 3.729323308270677e-07,
+      "loss": 0.1742,
+      "step": 64020
+    },
+    {
+      "epoch": 96.29,
+      "grad_norm": 5.491370677947998,
+      "learning_rate": 3.7142857142857145e-07,
+      "loss": 0.1964,
+      "step": 64030
+    },
+    {
+      "epoch": 96.3,
+      "grad_norm": 3.9227075576782227,
+      "learning_rate": 3.6992481203007523e-07,
+      "loss": 0.1793,
+      "step": 64040
+    },
+    {
+      "epoch": 96.32,
+      "grad_norm": 2.4654548168182373,
+      "learning_rate": 3.6842105263157896e-07,
+      "loss": 0.1948,
+      "step": 64050
+    },
+    {
+      "epoch": 96.33,
+      "grad_norm": 4.7875590324401855,
+      "learning_rate": 3.669172932330827e-07,
+      "loss": 0.1703,
+      "step": 64060
+    },
+    {
+      "epoch": 96.35,
+      "grad_norm": 5.792588233947754,
+      "learning_rate": 3.654135338345865e-07,
+      "loss": 0.2115,
+      "step": 64070
+    },
+    {
+      "epoch": 96.36,
+      "grad_norm": 3.1025025844573975,
+      "learning_rate": 3.6390977443609026e-07,
+      "loss": 0.1557,
+      "step": 64080
+    },
+    {
+      "epoch": 96.38,
+      "grad_norm": 7.800917625427246,
+      "learning_rate": 3.6240601503759405e-07,
+      "loss": 0.1808,
+      "step": 64090
+    },
+    {
+      "epoch": 96.39,
+      "grad_norm": 5.239587306976318,
+      "learning_rate": 3.609022556390978e-07,
+      "loss": 0.1319,
+      "step": 64100
+    },
+    {
+      "epoch": 96.41,
+      "grad_norm": 4.564650058746338,
+      "learning_rate": 3.5939849624060156e-07,
+      "loss": 0.1893,
+      "step": 64110
+    },
+    {
+      "epoch": 96.42,
+      "grad_norm": 5.047513484954834,
+      "learning_rate": 3.578947368421053e-07,
+      "loss": 0.1427,
+      "step": 64120
+    },
+    {
+      "epoch": 96.44,
+      "grad_norm": 5.4176249504089355,
+      "learning_rate": 3.56390977443609e-07,
+      "loss": 0.1932,
+      "step": 64130
+    },
+    {
+      "epoch": 96.45,
+      "grad_norm": 5.930634021759033,
+      "learning_rate": 3.548872180451128e-07,
+      "loss": 0.1444,
+      "step": 64140
+    },
+    {
+      "epoch": 96.47,
+      "grad_norm": 4.005468368530273,
+      "learning_rate": 3.533834586466166e-07,
+      "loss": 0.163,
+      "step": 64150
+    },
+    {
+      "epoch": 96.48,
+      "grad_norm": 6.028830528259277,
+      "learning_rate": 3.518796992481204e-07,
+      "loss": 0.1771,
+      "step": 64160
+    },
+    {
+      "epoch": 96.5,
+      "grad_norm": 2.6578989028930664,
+      "learning_rate": 3.503759398496241e-07,
+      "loss": 0.1031,
+      "step": 64170
+    },
+    {
+      "epoch": 96.51,
+      "grad_norm": 5.434506416320801,
+      "learning_rate": 3.4887218045112784e-07,
+      "loss": 0.1739,
+      "step": 64180
+    },
+    {
+      "epoch": 96.53,
+      "grad_norm": 5.756526470184326,
+      "learning_rate": 3.4736842105263157e-07,
+      "loss": 0.1191,
+      "step": 64190
+    },
+    {
+      "epoch": 96.54,
+      "grad_norm": 4.299066543579102,
+      "learning_rate": 3.4586466165413535e-07,
+      "loss": 0.1522,
+      "step": 64200
+    },
+    {
+      "epoch": 96.56,
+      "grad_norm": 4.130573272705078,
+      "learning_rate": 3.4436090225563913e-07,
+      "loss": 0.1381,
+      "step": 64210
+    },
+    {
+      "epoch": 96.57,
+      "grad_norm": 5.61849308013916,
+      "learning_rate": 3.428571428571429e-07,
+      "loss": 0.1467,
+      "step": 64220
+    },
+    {
+      "epoch": 96.59,
+      "grad_norm": 4.367659568786621,
+      "learning_rate": 3.4135338345864665e-07,
+      "loss": 0.1557,
+      "step": 64230
+    },
+    {
+      "epoch": 96.6,
+      "grad_norm": 2.8992245197296143,
+      "learning_rate": 3.398496240601504e-07,
+      "loss": 0.1537,
+      "step": 64240
+    },
+    {
+      "epoch": 96.62,
+      "grad_norm": 3.538975715637207,
+      "learning_rate": 3.3834586466165416e-07,
+      "loss": 0.1396,
+      "step": 64250
+    },
+    {
+      "epoch": 96.63,
+      "grad_norm": 8.194727897644043,
+      "learning_rate": 3.368421052631579e-07,
+      "loss": 0.1698,
+      "step": 64260
+    },
+    {
+      "epoch": 96.65,
+      "grad_norm": 10.290156364440918,
+      "learning_rate": 3.353383458646617e-07,
+      "loss": 0.1561,
+      "step": 64270
+    },
+    {
+      "epoch": 96.66,
+      "grad_norm": 5.7417073249816895,
+      "learning_rate": 3.3383458646616546e-07,
+      "loss": 0.1215,
+      "step": 64280
+    },
+    {
+      "epoch": 96.68,
+      "grad_norm": 6.722829818725586,
+      "learning_rate": 3.3233082706766924e-07,
+      "loss": 0.2179,
+      "step": 64290
+    },
+    {
+      "epoch": 96.69,
+      "grad_norm": 8.941481590270996,
+      "learning_rate": 3.308270676691729e-07,
+      "loss": 0.1763,
+      "step": 64300
+    },
+    {
+      "epoch": 96.71,
+      "grad_norm": 4.601995468139648,
+      "learning_rate": 3.293233082706767e-07,
+      "loss": 0.1611,
+      "step": 64310
+    },
+    {
+      "epoch": 96.72,
+      "grad_norm": 7.5421671867370605,
+      "learning_rate": 3.278195488721805e-07,
+      "loss": 0.1459,
+      "step": 64320
+    },
+    {
+      "epoch": 96.74,
+      "grad_norm": 2.4349803924560547,
+      "learning_rate": 3.263157894736842e-07,
+      "loss": 0.1892,
+      "step": 64330
+    },
+    {
+      "epoch": 96.75,
+      "grad_norm": 1.8686342239379883,
+      "learning_rate": 3.24812030075188e-07,
+      "loss": 0.1407,
+      "step": 64340
+    },
+    {
+      "epoch": 96.77,
+      "grad_norm": 5.259500503540039,
+      "learning_rate": 3.233082706766918e-07,
+      "loss": 0.1057,
+      "step": 64350
+    },
+    {
+      "epoch": 96.78,
+      "grad_norm": 4.295063018798828,
+      "learning_rate": 3.2180451127819547e-07,
+      "loss": 0.1865,
+      "step": 64360
+    },
+    {
+      "epoch": 96.8,
+      "grad_norm": 4.434532642364502,
+      "learning_rate": 3.2030075187969925e-07,
+      "loss": 0.1993,
+      "step": 64370
+    },
+    {
+      "epoch": 96.81,
+      "grad_norm": 5.843010902404785,
+      "learning_rate": 3.1879699248120303e-07,
+      "loss": 0.166,
+      "step": 64380
+    },
+    {
+      "epoch": 96.83,
+      "grad_norm": 4.545680046081543,
+      "learning_rate": 3.172932330827068e-07,
+      "loss": 0.135,
+      "step": 64390
+    },
+    {
+      "epoch": 96.84,
+      "grad_norm": 6.0127739906311035,
+      "learning_rate": 3.1578947368421055e-07,
+      "loss": 0.201,
+      "step": 64400
+    },
+    {
+      "epoch": 96.86,
+      "grad_norm": 6.95708703994751,
+      "learning_rate": 3.1428571428571433e-07,
+      "loss": 0.1584,
+      "step": 64410
+    },
+    {
+      "epoch": 96.87,
+      "grad_norm": 0.3541220724582672,
+      "learning_rate": 3.127819548872181e-07,
+      "loss": 0.143,
+      "step": 64420
+    },
+    {
+      "epoch": 96.89,
+      "grad_norm": 4.712253570556641,
+      "learning_rate": 3.1127819548872185e-07,
+      "loss": 0.1417,
+      "step": 64430
+    },
+    {
+      "epoch": 96.9,
+      "grad_norm": 5.915809154510498,
+      "learning_rate": 3.097744360902256e-07,
+      "loss": 0.135,
+      "step": 64440
+    },
+    {
+      "epoch": 96.92,
+      "grad_norm": 7.92987060546875,
+      "learning_rate": 3.0827067669172936e-07,
+      "loss": 0.1586,
+      "step": 64450
+    },
+    {
+      "epoch": 96.93,
+      "grad_norm": 3.605088472366333,
+      "learning_rate": 3.067669172932331e-07,
+      "loss": 0.1384,
+      "step": 64460
+    },
+    {
+      "epoch": 96.95,
+      "grad_norm": 5.079803943634033,
+      "learning_rate": 3.052631578947369e-07,
+      "loss": 0.177,
+      "step": 64470
+    },
+    {
+      "epoch": 96.96,
+      "grad_norm": 5.349300384521484,
+      "learning_rate": 3.037593984962406e-07,
+      "loss": 0.2294,
+      "step": 64480
+    },
+    {
+      "epoch": 96.98,
+      "grad_norm": 9.588523864746094,
+      "learning_rate": 3.022556390977444e-07,
+      "loss": 0.1532,
+      "step": 64490
+    },
+    {
+      "epoch": 96.99,
+      "grad_norm": 3.1783509254455566,
+      "learning_rate": 3.007518796992481e-07,
+      "loss": 0.1256,
+      "step": 64500
+    },
+    {
+      "epoch": 97.0,
+      "eval_accuracy": 0.9308,
+      "eval_loss": 0.332674503326416,
+      "eval_runtime": 84.9569,
+      "eval_samples_per_second": 117.707,
+      "eval_steps_per_second": 0.471,
+      "step": 64505
+    },
+    {
+      "epoch": 97.01,
+      "grad_norm": 7.344843864440918,
+      "learning_rate": 2.992481203007519e-07,
+      "loss": 0.1059,
+      "step": 64510
+    },
+    {
+      "epoch": 97.02,
+      "grad_norm": 10.291905403137207,
+      "learning_rate": 2.977443609022557e-07,
+      "loss": 0.1662,
+      "step": 64520
+    },
+    {
+      "epoch": 97.04,
+      "grad_norm": 8.140653610229492,
+      "learning_rate": 2.962406015037594e-07,
+      "loss": 0.2008,
+      "step": 64530
+    },
+    {
+      "epoch": 97.05,
+      "grad_norm": 3.3860104084014893,
+      "learning_rate": 2.9473684210526315e-07,
+      "loss": 0.1893,
+      "step": 64540
+    },
+    {
+      "epoch": 97.07,
+      "grad_norm": 7.287869930267334,
+      "learning_rate": 2.9323308270676693e-07,
+      "loss": 0.1382,
+      "step": 64550
+    },
+    {
+      "epoch": 97.08,
+      "grad_norm": 8.065459251403809,
+      "learning_rate": 2.917293233082707e-07,
+      "loss": 0.1349,
+      "step": 64560
+    },
+    {
+      "epoch": 97.1,
+      "grad_norm": 5.940702438354492,
+      "learning_rate": 2.9022556390977445e-07,
+      "loss": 0.1896,
+      "step": 64570
+    },
+    {
+      "epoch": 97.11,
+      "grad_norm": 8.597441673278809,
+      "learning_rate": 2.8872180451127823e-07,
+      "loss": 0.1467,
+      "step": 64580
+    },
+    {
+      "epoch": 97.13,
+      "grad_norm": 4.228404998779297,
+      "learning_rate": 2.87218045112782e-07,
+      "loss": 0.1584,
+      "step": 64590
+    },
+    {
+      "epoch": 97.14,
+      "grad_norm": 8.518570899963379,
+      "learning_rate": 2.8571428571428575e-07,
+      "loss": 0.1133,
+      "step": 64600
+    },
+    {
+      "epoch": 97.16,
+      "grad_norm": 5.979735374450684,
+      "learning_rate": 2.842105263157895e-07,
+      "loss": 0.1639,
+      "step": 64610
+    },
+    {
+      "epoch": 97.17,
+      "grad_norm": 4.625314235687256,
+      "learning_rate": 2.8270676691729326e-07,
+      "loss": 0.1755,
+      "step": 64620
+    },
+    {
+      "epoch": 97.19,
+      "grad_norm": 2.0218687057495117,
+      "learning_rate": 2.81203007518797e-07,
+      "loss": 0.1521,
+      "step": 64630
+    },
+    {
+      "epoch": 97.2,
+      "grad_norm": 6.187188148498535,
+      "learning_rate": 2.796992481203008e-07,
+      "loss": 0.162,
+      "step": 64640
+    },
+    {
+      "epoch": 97.22,
+      "grad_norm": 6.775570869445801,
+      "learning_rate": 2.7819548872180456e-07,
+      "loss": 0.1703,
+      "step": 64650
+    },
+    {
+      "epoch": 97.23,
+      "grad_norm": 1.6887646913528442,
+      "learning_rate": 2.766917293233083e-07,
+      "loss": 0.2072,
+      "step": 64660
+    },
+    {
+      "epoch": 97.25,
+      "grad_norm": 5.309747695922852,
+      "learning_rate": 2.751879699248121e-07,
+      "loss": 0.1646,
+      "step": 64670
+    },
+    {
+      "epoch": 97.26,
+      "grad_norm": 5.663767337799072,
+      "learning_rate": 2.736842105263158e-07,
+      "loss": 0.166,
+      "step": 64680
+    },
+    {
+      "epoch": 97.28,
+      "grad_norm": 5.132472991943359,
+      "learning_rate": 2.7218045112781954e-07,
+      "loss": 0.1857,
+      "step": 64690
+    },
+    {
+      "epoch": 97.29,
+      "grad_norm": 5.090039253234863,
+      "learning_rate": 2.706766917293233e-07,
+      "loss": 0.195,
+      "step": 64700
+    },
+    {
+      "epoch": 97.31,
+      "grad_norm": 9.945984840393066,
+      "learning_rate": 2.691729323308271e-07,
+      "loss": 0.1869,
+      "step": 64710
+    },
+    {
+      "epoch": 97.32,
+      "grad_norm": 9.467114448547363,
+      "learning_rate": 2.6766917293233083e-07,
+      "loss": 0.1387,
+      "step": 64720
+    },
+    {
+      "epoch": 97.34,
+      "grad_norm": 2.083440065383911,
+      "learning_rate": 2.661654135338346e-07,
+      "loss": 0.1623,
+      "step": 64730
+    },
+    {
+      "epoch": 97.35,
+      "grad_norm": 3.3765485286712646,
+      "learning_rate": 2.646616541353384e-07,
+      "loss": 0.1059,
+      "step": 64740
+    },
+    {
+      "epoch": 97.37,
+      "grad_norm": 2.4343514442443848,
+      "learning_rate": 2.6315789473684213e-07,
+      "loss": 0.193,
+      "step": 64750
+    },
+    {
+      "epoch": 97.38,
+      "grad_norm": 7.881185531616211,
+      "learning_rate": 2.6165413533834586e-07,
+      "loss": 0.162,
+      "step": 64760
+    },
+    {
+      "epoch": 97.4,
+      "grad_norm": 3.8608689308166504,
+      "learning_rate": 2.6015037593984965e-07,
+      "loss": 0.1648,
+      "step": 64770
+    },
+    {
+      "epoch": 97.41,
+      "grad_norm": 3.745732069015503,
+      "learning_rate": 2.586466165413534e-07,
+      "loss": 0.1706,
+      "step": 64780
+    },
+    {
+      "epoch": 97.43,
+      "grad_norm": 1.355181336402893,
+      "learning_rate": 2.5714285714285716e-07,
+      "loss": 0.1558,
+      "step": 64790
+    },
+    {
+      "epoch": 97.44,
+      "grad_norm": 3.218841075897217,
+      "learning_rate": 2.5563909774436095e-07,
+      "loss": 0.1518,
+      "step": 64800
+    },
+    {
+      "epoch": 97.46,
+      "grad_norm": 1.6268922090530396,
+      "learning_rate": 2.541353383458647e-07,
+      "loss": 0.1573,
+      "step": 64810
+    },
+    {
+      "epoch": 97.47,
+      "grad_norm": 2.534693479537964,
+      "learning_rate": 2.5263157894736846e-07,
+      "loss": 0.1684,
+      "step": 64820
+    },
+    {
+      "epoch": 97.49,
+      "grad_norm": 5.223088264465332,
+      "learning_rate": 2.511278195488722e-07,
+      "loss": 0.1612,
+      "step": 64830
+    },
+    {
+      "epoch": 97.5,
+      "grad_norm": 4.601364612579346,
+      "learning_rate": 2.49624060150376e-07,
+      "loss": 0.1334,
+      "step": 64840
+    },
+    {
+      "epoch": 97.52,
+      "grad_norm": 3.1391961574554443,
+      "learning_rate": 2.481203007518797e-07,
+      "loss": 0.1509,
+      "step": 64850
+    },
+    {
+      "epoch": 97.53,
+      "grad_norm": 5.453991889953613,
+      "learning_rate": 2.466165413533835e-07,
+      "loss": 0.1266,
+      "step": 64860
+    },
+    {
+      "epoch": 97.55,
+      "grad_norm": 4.62083625793457,
+      "learning_rate": 2.4511278195488727e-07,
+      "loss": 0.158,
+      "step": 64870
+    },
+    {
+      "epoch": 97.56,
+      "grad_norm": 1.933967113494873,
+      "learning_rate": 2.43609022556391e-07,
+      "loss": 0.0955,
+      "step": 64880
+    },
+    {
+      "epoch": 97.58,
+      "grad_norm": 5.738483428955078,
+      "learning_rate": 2.4210526315789473e-07,
+      "loss": 0.1555,
+      "step": 64890
+    },
+    {
+      "epoch": 97.59,
+      "grad_norm": 4.406610012054443,
+      "learning_rate": 2.406015037593985e-07,
+      "loss": 0.1667,
+      "step": 64900
+    },
+    {
+      "epoch": 97.61,
+      "grad_norm": 4.557981491088867,
+      "learning_rate": 2.3909774436090225e-07,
+      "loss": 0.1796,
+      "step": 64910
+    },
+    {
+      "epoch": 97.62,
+      "grad_norm": 4.654937744140625,
+      "learning_rate": 2.3759398496240603e-07,
+      "loss": 0.1713,
+      "step": 64920
+    },
+    {
+      "epoch": 97.64,
+      "grad_norm": 5.671332359313965,
+      "learning_rate": 2.3609022556390982e-07,
+      "loss": 0.1292,
+      "step": 64930
+    },
+    {
+      "epoch": 97.65,
+      "grad_norm": 5.643190860748291,
+      "learning_rate": 2.3458646616541355e-07,
+      "loss": 0.1565,
+      "step": 64940
+    },
+    {
+      "epoch": 97.67,
+      "grad_norm": 2.340914487838745,
+      "learning_rate": 2.330827067669173e-07,
+      "loss": 0.128,
+      "step": 64950
+    },
+    {
+      "epoch": 97.68,
+      "grad_norm": 7.151415824890137,
+      "learning_rate": 2.315789473684211e-07,
+      "loss": 0.1618,
+      "step": 64960
+    },
+    {
+      "epoch": 97.7,
+      "grad_norm": 7.448176860809326,
+      "learning_rate": 2.3007518796992482e-07,
+      "loss": 0.1396,
+      "step": 64970
+    },
+    {
+      "epoch": 97.71,
+      "grad_norm": 5.021811008453369,
+      "learning_rate": 2.285714285714286e-07,
+      "loss": 0.1772,
+      "step": 64980
+    },
+    {
+      "epoch": 97.73,
+      "grad_norm": 0.7833675146102905,
+      "learning_rate": 2.2706766917293236e-07,
+      "loss": 0.1125,
+      "step": 64990
+    },
+    {
+      "epoch": 97.74,
+      "grad_norm": 7.163773536682129,
+      "learning_rate": 2.255639097744361e-07,
+      "loss": 0.2541,
+      "step": 65000
+    },
+    {
+      "epoch": 97.76,
+      "grad_norm": 7.621840000152588,
+      "learning_rate": 2.2406015037593987e-07,
+      "loss": 0.1367,
+      "step": 65010
+    },
+    {
+      "epoch": 97.77,
+      "grad_norm": 3.7870657444000244,
+      "learning_rate": 2.2255639097744363e-07,
+      "loss": 0.1336,
+      "step": 65020
+    },
+    {
+      "epoch": 97.79,
+      "grad_norm": 4.785586357116699,
+      "learning_rate": 2.2105263157894736e-07,
+      "loss": 0.1779,
+      "step": 65030
+    },
+    {
+      "epoch": 97.8,
+      "grad_norm": 2.2367138862609863,
+      "learning_rate": 2.1954887218045115e-07,
+      "loss": 0.1886,
+      "step": 65040
+    },
+    {
+      "epoch": 97.82,
+      "grad_norm": 4.493763446807861,
+      "learning_rate": 2.180451127819549e-07,
+      "loss": 0.1943,
+      "step": 65050
+    },
+    {
+      "epoch": 97.83,
+      "grad_norm": 6.870443344116211,
+      "learning_rate": 2.1654135338345866e-07,
+      "loss": 0.0937,
+      "step": 65060
+    },
+    {
+      "epoch": 97.85,
+      "grad_norm": 6.161388874053955,
+      "learning_rate": 2.1503759398496242e-07,
+      "loss": 0.192,
+      "step": 65070
+    },
+    {
+      "epoch": 97.86,
+      "grad_norm": 5.1376729011535645,
+      "learning_rate": 2.135338345864662e-07,
+      "loss": 0.1587,
+      "step": 65080
+    },
+    {
+      "epoch": 97.88,
+      "grad_norm": 3.9065380096435547,
+      "learning_rate": 2.1203007518796993e-07,
+      "loss": 0.1104,
+      "step": 65090
+    },
+    {
+      "epoch": 97.89,
+      "grad_norm": 5.1056227684021,
+      "learning_rate": 2.105263157894737e-07,
+      "loss": 0.157,
+      "step": 65100
+    },
+    {
+      "epoch": 97.91,
+      "grad_norm": 3.8293356895446777,
+      "learning_rate": 2.0902255639097747e-07,
+      "loss": 0.17,
+      "step": 65110
+    },
+    {
+      "epoch": 97.92,
+      "grad_norm": 4.883228302001953,
+      "learning_rate": 2.075187969924812e-07,
+      "loss": 0.1989,
+      "step": 65120
+    },
+    {
+      "epoch": 97.94,
+      "grad_norm": 6.1433210372924805,
+      "learning_rate": 2.0601503759398496e-07,
+      "loss": 0.2161,
+      "step": 65130
+    },
+    {
+      "epoch": 97.95,
+      "grad_norm": 3.8383498191833496,
+      "learning_rate": 2.0451127819548875e-07,
+      "loss": 0.2358,
+      "step": 65140
+    },
+    {
+      "epoch": 97.97,
+      "grad_norm": 3.27778959274292,
+      "learning_rate": 2.030075187969925e-07,
+      "loss": 0.1498,
+      "step": 65150
+    },
+    {
+      "epoch": 97.98,
+      "grad_norm": 5.0947265625,
+      "learning_rate": 2.0150375939849626e-07,
+      "loss": 0.1577,
+      "step": 65160
+    },
+    {
+      "epoch": 98.0,
+      "grad_norm": 0.12277551740407944,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 0.0984,
+      "step": 65170
+    },
+    {
+      "epoch": 98.0,
+      "eval_accuracy": 0.9317,
+      "eval_loss": 0.32914501428604126,
+      "eval_runtime": 84.4819,
+      "eval_samples_per_second": 118.369,
+      "eval_steps_per_second": 0.473,
+      "step": 65170
+    },
+    {
+      "epoch": 98.02,
+      "grad_norm": 8.998836517333984,
+      "learning_rate": 1.984962406015038e-07,
+      "loss": 0.1479,
+      "step": 65180
+    },
+    {
+      "epoch": 98.03,
+      "grad_norm": 4.802471160888672,
+      "learning_rate": 1.9699248120300753e-07,
+      "loss": 0.1792,
+      "step": 65190
+    },
+    {
+      "epoch": 98.05,
+      "grad_norm": 6.974554061889648,
+      "learning_rate": 1.954887218045113e-07,
+      "loss": 0.1599,
+      "step": 65200
+    },
+    {
+      "epoch": 98.06,
+      "grad_norm": 5.7014641761779785,
+      "learning_rate": 1.9398496240601507e-07,
+      "loss": 0.1527,
+      "step": 65210
+    },
+    {
+      "epoch": 98.08,
+      "grad_norm": 7.29013729095459,
+      "learning_rate": 1.924812030075188e-07,
+      "loss": 0.1443,
+      "step": 65220
+    },
+    {
+      "epoch": 98.09,
+      "grad_norm": 3.837073802947998,
+      "learning_rate": 1.9097744360902256e-07,
+      "loss": 0.1044,
+      "step": 65230
+    },
+    {
+      "epoch": 98.11,
+      "grad_norm": 9.046390533447266,
+      "learning_rate": 1.8947368421052634e-07,
+      "loss": 0.1598,
+      "step": 65240
+    },
+    {
+      "epoch": 98.12,
+      "grad_norm": 7.997859954833984,
+      "learning_rate": 1.8796992481203008e-07,
+      "loss": 0.2044,
+      "step": 65250
+    },
+    {
+      "epoch": 98.14,
+      "grad_norm": 5.434107780456543,
+      "learning_rate": 1.8646616541353386e-07,
+      "loss": 0.1655,
+      "step": 65260
+    },
+    {
+      "epoch": 98.15,
+      "grad_norm": 4.262781620025635,
+      "learning_rate": 1.8496240601503762e-07,
+      "loss": 0.1393,
+      "step": 65270
+    },
+    {
+      "epoch": 98.17,
+      "grad_norm": 6.696924209594727,
+      "learning_rate": 1.8345864661654135e-07,
+      "loss": 0.1341,
+      "step": 65280
+    },
+    {
+      "epoch": 98.18,
+      "grad_norm": 6.295987129211426,
+      "learning_rate": 1.8195488721804513e-07,
+      "loss": 0.1655,
+      "step": 65290
+    },
+    {
+      "epoch": 98.2,
+      "grad_norm": 2.8167364597320557,
+      "learning_rate": 1.804511278195489e-07,
+      "loss": 0.1448,
+      "step": 65300
+    },
+    {
+      "epoch": 98.21,
+      "grad_norm": 2.0722157955169678,
+      "learning_rate": 1.7894736842105265e-07,
+      "loss": 0.2011,
+      "step": 65310
+    },
+    {
+      "epoch": 98.23,
+      "grad_norm": 4.755069255828857,
+      "learning_rate": 1.774436090225564e-07,
+      "loss": 0.1573,
+      "step": 65320
+    },
+    {
+      "epoch": 98.24,
+      "grad_norm": 11.782632827758789,
+      "learning_rate": 1.759398496240602e-07,
+      "loss": 0.1736,
+      "step": 65330
+    },
+    {
+      "epoch": 98.26,
+      "grad_norm": 4.000039577484131,
+      "learning_rate": 1.7443609022556392e-07,
+      "loss": 0.1365,
+      "step": 65340
+    },
+    {
+      "epoch": 98.27,
+      "grad_norm": 4.4005656242370605,
+      "learning_rate": 1.7293233082706767e-07,
+      "loss": 0.1795,
+      "step": 65350
+    },
+    {
+      "epoch": 98.29,
+      "grad_norm": 7.827691555023193,
+      "learning_rate": 1.7142857142857146e-07,
+      "loss": 0.1507,
+      "step": 65360
+    },
+    {
+      "epoch": 98.3,
+      "grad_norm": 3.5540237426757812,
+      "learning_rate": 1.699248120300752e-07,
+      "loss": 0.1653,
+      "step": 65370
+    },
+    {
+      "epoch": 98.32,
+      "grad_norm": 4.963322639465332,
+      "learning_rate": 1.6842105263157895e-07,
+      "loss": 0.1742,
+      "step": 65380
+    },
+    {
+      "epoch": 98.33,
+      "grad_norm": 4.910261154174805,
+      "learning_rate": 1.6691729323308273e-07,
+      "loss": 0.1477,
+      "step": 65390
+    },
+    {
+      "epoch": 98.35,
+      "grad_norm": 7.770264148712158,
+      "learning_rate": 1.6541353383458646e-07,
+      "loss": 0.1713,
+      "step": 65400
+    },
+    {
+      "epoch": 98.36,
+      "grad_norm": 2.305351495742798,
+      "learning_rate": 1.6390977443609025e-07,
+      "loss": 0.1471,
+      "step": 65410
+    },
+    {
+      "epoch": 98.38,
+      "grad_norm": 5.237599849700928,
+      "learning_rate": 1.62406015037594e-07,
+      "loss": 0.2242,
+      "step": 65420
+    },
+    {
+      "epoch": 98.39,
+      "grad_norm": 3.8604493141174316,
+      "learning_rate": 1.6090225563909773e-07,
+      "loss": 0.1789,
+      "step": 65430
+    },
+    {
+      "epoch": 98.41,
+      "grad_norm": 4.205555438995361,
+      "learning_rate": 1.5939849624060152e-07,
+      "loss": 0.1286,
+      "step": 65440
+    },
+    {
+      "epoch": 98.42,
+      "grad_norm": 6.994670867919922,
+      "learning_rate": 1.5789473684210527e-07,
+      "loss": 0.204,
+      "step": 65450
+    },
+    {
+      "epoch": 98.44,
+      "grad_norm": 6.597128391265869,
+      "learning_rate": 1.5639097744360906e-07,
+      "loss": 0.1172,
+      "step": 65460
+    },
+    {
+      "epoch": 98.45,
+      "grad_norm": 9.189870834350586,
+      "learning_rate": 1.548872180451128e-07,
+      "loss": 0.1601,
+      "step": 65470
+    },
+    {
+      "epoch": 98.47,
+      "grad_norm": 5.771359920501709,
+      "learning_rate": 1.5338345864661655e-07,
+      "loss": 0.1457,
+      "step": 65480
+    },
+    {
+      "epoch": 98.48,
+      "grad_norm": 6.4306840896606445,
+      "learning_rate": 1.518796992481203e-07,
+      "loss": 0.1385,
+      "step": 65490
+    },
+    {
+      "epoch": 98.5,
+      "grad_norm": 3.8584532737731934,
+      "learning_rate": 1.5037593984962406e-07,
+      "loss": 0.1687,
+      "step": 65500
+    },
+    {
+      "epoch": 98.51,
+      "grad_norm": 4.318583965301514,
+      "learning_rate": 1.4887218045112784e-07,
+      "loss": 0.1136,
+      "step": 65510
+    },
+    {
+      "epoch": 98.53,
+      "grad_norm": 4.790411472320557,
+      "learning_rate": 1.4736842105263158e-07,
+      "loss": 0.176,
+      "step": 65520
+    },
+    {
+      "epoch": 98.54,
+      "grad_norm": 7.689479351043701,
+      "learning_rate": 1.4586466165413536e-07,
+      "loss": 0.17,
+      "step": 65530
+    },
+    {
+      "epoch": 98.56,
+      "grad_norm": 6.151116371154785,
+      "learning_rate": 1.4436090225563912e-07,
+      "loss": 0.1942,
+      "step": 65540
+    },
+    {
+      "epoch": 98.57,
+      "grad_norm": 5.650688648223877,
+      "learning_rate": 1.4285714285714287e-07,
+      "loss": 0.2166,
+      "step": 65550
+    },
+    {
+      "epoch": 98.59,
+      "grad_norm": 4.956643581390381,
+      "learning_rate": 1.4135338345864663e-07,
+      "loss": 0.1672,
+      "step": 65560
+    },
+    {
+      "epoch": 98.6,
+      "grad_norm": 1.9977216720581055,
+      "learning_rate": 1.398496240601504e-07,
+      "loss": 0.153,
+      "step": 65570
+    },
+    {
+      "epoch": 98.62,
+      "grad_norm": 6.504947185516357,
+      "learning_rate": 1.3834586466165415e-07,
+      "loss": 0.2146,
+      "step": 65580
+    },
+    {
+      "epoch": 98.63,
+      "grad_norm": 3.9041435718536377,
+      "learning_rate": 1.368421052631579e-07,
+      "loss": 0.2097,
+      "step": 65590
+    },
+    {
+      "epoch": 98.65,
+      "grad_norm": 6.294933795928955,
+      "learning_rate": 1.3533834586466166e-07,
+      "loss": 0.1532,
+      "step": 65600
+    },
+    {
+      "epoch": 98.66,
+      "grad_norm": 3.49302339553833,
+      "learning_rate": 1.3383458646616542e-07,
+      "loss": 0.1407,
+      "step": 65610
+    },
+    {
+      "epoch": 98.68,
+      "grad_norm": 6.846986293792725,
+      "learning_rate": 1.323308270676692e-07,
+      "loss": 0.1436,
+      "step": 65620
+    },
+    {
+      "epoch": 98.69,
+      "grad_norm": 8.050239562988281,
+      "learning_rate": 1.3082706766917293e-07,
+      "loss": 0.1993,
+      "step": 65630
+    },
+    {
+      "epoch": 98.71,
+      "grad_norm": 2.393629789352417,
+      "learning_rate": 1.293233082706767e-07,
+      "loss": 0.1998,
+      "step": 65640
+    },
+    {
+      "epoch": 98.72,
+      "grad_norm": 3.4583864212036133,
+      "learning_rate": 1.2781954887218047e-07,
+      "loss": 0.2058,
+      "step": 65650
+    },
+    {
+      "epoch": 98.74,
+      "grad_norm": 1.8351730108261108,
+      "learning_rate": 1.2631578947368423e-07,
+      "loss": 0.0967,
+      "step": 65660
+    },
+    {
+      "epoch": 98.75,
+      "grad_norm": 4.4519195556640625,
+      "learning_rate": 1.24812030075188e-07,
+      "loss": 0.1455,
+      "step": 65670
+    },
+    {
+      "epoch": 98.77,
+      "grad_norm": 3.725358009338379,
+      "learning_rate": 1.2330827067669174e-07,
+      "loss": 0.1497,
+      "step": 65680
+    },
+    {
+      "epoch": 98.78,
+      "grad_norm": 5.022524356842041,
+      "learning_rate": 1.218045112781955e-07,
+      "loss": 0.2046,
+      "step": 65690
+    },
+    {
+      "epoch": 98.8,
+      "grad_norm": 0.5539684295654297,
+      "learning_rate": 1.2030075187969926e-07,
+      "loss": 0.1661,
+      "step": 65700
+    },
+    {
+      "epoch": 98.81,
+      "grad_norm": 5.730692386627197,
+      "learning_rate": 1.1879699248120302e-07,
+      "loss": 0.1907,
+      "step": 65710
+    },
+    {
+      "epoch": 98.83,
+      "grad_norm": 4.800581455230713,
+      "learning_rate": 1.1729323308270677e-07,
+      "loss": 0.1923,
+      "step": 65720
+    },
+    {
+      "epoch": 98.84,
+      "grad_norm": 7.648699760437012,
+      "learning_rate": 1.1578947368421054e-07,
+      "loss": 0.1876,
+      "step": 65730
+    },
+    {
+      "epoch": 98.86,
+      "grad_norm": 6.9826860427856445,
+      "learning_rate": 1.142857142857143e-07,
+      "loss": 0.1962,
+      "step": 65740
+    },
+    {
+      "epoch": 98.87,
+      "grad_norm": 4.280595779418945,
+      "learning_rate": 1.1278195488721805e-07,
+      "loss": 0.1995,
+      "step": 65750
+    },
+    {
+      "epoch": 98.89,
+      "grad_norm": 9.39132308959961,
+      "learning_rate": 1.1127819548872182e-07,
+      "loss": 0.1467,
+      "step": 65760
+    },
+    {
+      "epoch": 98.9,
+      "grad_norm": 8.573371887207031,
+      "learning_rate": 1.0977443609022557e-07,
+      "loss": 0.156,
+      "step": 65770
+    },
+    {
+      "epoch": 98.92,
+      "grad_norm": 3.3703553676605225,
+      "learning_rate": 1.0827067669172933e-07,
+      "loss": 0.1639,
+      "step": 65780
+    },
+    {
+      "epoch": 98.93,
+      "grad_norm": 5.628261089324951,
+      "learning_rate": 1.067669172932331e-07,
+      "loss": 0.1439,
+      "step": 65790
+    },
+    {
+      "epoch": 98.95,
+      "grad_norm": 4.770748615264893,
+      "learning_rate": 1.0526315789473685e-07,
+      "loss": 0.1691,
+      "step": 65800
+    },
+    {
+      "epoch": 98.96,
+      "grad_norm": 3.2969655990600586,
+      "learning_rate": 1.037593984962406e-07,
+      "loss": 0.129,
+      "step": 65810
+    },
+    {
+      "epoch": 98.98,
+      "grad_norm": 5.129558563232422,
+      "learning_rate": 1.0225563909774437e-07,
+      "loss": 0.1381,
+      "step": 65820
+    },
+    {
+      "epoch": 98.99,
+      "grad_norm": 7.410007953643799,
+      "learning_rate": 1.0075187969924813e-07,
+      "loss": 0.1525,
+      "step": 65830
+    },
+    {
+      "epoch": 99.0,
+      "eval_accuracy": 0.9311,
+      "eval_loss": 0.33068838715553284,
+      "eval_runtime": 84.5618,
+      "eval_samples_per_second": 118.257,
+      "eval_steps_per_second": 0.473,
+      "step": 65835
+    },
+    {
+      "epoch": 99.01,
+      "grad_norm": 4.0409674644470215,
+      "learning_rate": 9.92481203007519e-08,
+      "loss": 0.1796,
+      "step": 65840
+    },
+    {
+      "epoch": 99.02,
+      "grad_norm": 6.996447563171387,
+      "learning_rate": 9.774436090225564e-08,
+      "loss": 0.1921,
+      "step": 65850
+    },
+    {
+      "epoch": 99.04,
+      "grad_norm": 9.530542373657227,
+      "learning_rate": 9.62406015037594e-08,
+      "loss": 0.1522,
+      "step": 65860
+    },
+    {
+      "epoch": 99.05,
+      "grad_norm": 1.6565377712249756,
+      "learning_rate": 9.473684210526317e-08,
+      "loss": 0.1283,
+      "step": 65870
+    },
+    {
+      "epoch": 99.07,
+      "grad_norm": 4.351380348205566,
+      "learning_rate": 9.323308270676693e-08,
+      "loss": 0.1222,
+      "step": 65880
+    },
+    {
+      "epoch": 99.08,
+      "grad_norm": 5.406339645385742,
+      "learning_rate": 9.172932330827067e-08,
+      "loss": 0.1445,
+      "step": 65890
+    },
+    {
+      "epoch": 99.1,
+      "grad_norm": 3.567111015319824,
+      "learning_rate": 9.022556390977444e-08,
+      "loss": 0.0886,
+      "step": 65900
+    },
+    {
+      "epoch": 99.11,
+      "grad_norm": 4.423712730407715,
+      "learning_rate": 8.87218045112782e-08,
+      "loss": 0.158,
+      "step": 65910
+    },
+    {
+      "epoch": 99.13,
+      "grad_norm": 1.147392988204956,
+      "learning_rate": 8.721804511278196e-08,
+      "loss": 0.1439,
+      "step": 65920
+    },
+    {
+      "epoch": 99.14,
+      "grad_norm": 4.6901936531066895,
+      "learning_rate": 8.571428571428573e-08,
+      "loss": 0.1749,
+      "step": 65930
+    },
+    {
+      "epoch": 99.16,
+      "grad_norm": 9.2613525390625,
+      "learning_rate": 8.421052631578947e-08,
+      "loss": 0.1984,
+      "step": 65940
+    },
+    {
+      "epoch": 99.17,
+      "grad_norm": 3.2010562419891357,
+      "learning_rate": 8.270676691729323e-08,
+      "loss": 0.115,
+      "step": 65950
+    },
+    {
+      "epoch": 99.19,
+      "grad_norm": 2.4863340854644775,
+      "learning_rate": 8.1203007518797e-08,
+      "loss": 0.1331,
+      "step": 65960
+    },
+    {
+      "epoch": 99.2,
+      "grad_norm": 2.9829752445220947,
+      "learning_rate": 7.969924812030076e-08,
+      "loss": 0.1423,
+      "step": 65970
+    },
+    {
+      "epoch": 99.22,
+      "grad_norm": 4.100430011749268,
+      "learning_rate": 7.819548872180453e-08,
+      "loss": 0.1391,
+      "step": 65980
+    },
+    {
+      "epoch": 99.23,
+      "grad_norm": 4.504250526428223,
+      "learning_rate": 7.669172932330827e-08,
+      "loss": 0.1529,
+      "step": 65990
+    },
+    {
+      "epoch": 99.25,
+      "grad_norm": 3.5403831005096436,
+      "learning_rate": 7.518796992481203e-08,
+      "loss": 0.1882,
+      "step": 66000
+    },
+    {
+      "epoch": 99.26,
+      "grad_norm": 4.3315253257751465,
+      "learning_rate": 7.368421052631579e-08,
+      "loss": 0.1514,
+      "step": 66010
+    },
+    {
+      "epoch": 99.28,
+      "grad_norm": 5.822681903839111,
+      "learning_rate": 7.218045112781956e-08,
+      "loss": 0.1829,
+      "step": 66020
+    },
+    {
+      "epoch": 99.29,
+      "grad_norm": 5.7567596435546875,
+      "learning_rate": 7.067669172932332e-08,
+      "loss": 0.1862,
+      "step": 66030
+    },
+    {
+      "epoch": 99.31,
+      "grad_norm": 7.93897819519043,
+      "learning_rate": 6.917293233082707e-08,
+      "loss": 0.1586,
+      "step": 66040
+    },
+    {
+      "epoch": 99.32,
+      "grad_norm": 2.92777156829834,
+      "learning_rate": 6.766917293233083e-08,
+      "loss": 0.1325,
+      "step": 66050
+    },
+    {
+      "epoch": 99.34,
+      "grad_norm": 5.535176753997803,
+      "learning_rate": 6.61654135338346e-08,
+      "loss": 0.192,
+      "step": 66060
+    },
+    {
+      "epoch": 99.35,
+      "grad_norm": 6.798808574676514,
+      "learning_rate": 6.466165413533834e-08,
+      "loss": 0.171,
+      "step": 66070
+    },
+    {
+      "epoch": 99.37,
+      "grad_norm": 7.453383922576904,
+      "learning_rate": 6.315789473684211e-08,
+      "loss": 0.2143,
+      "step": 66080
+    },
+    {
+      "epoch": 99.38,
+      "grad_norm": 5.353199005126953,
+      "learning_rate": 6.165413533834587e-08,
+      "loss": 0.165,
+      "step": 66090
+    },
+    {
+      "epoch": 99.4,
+      "grad_norm": 3.9845528602600098,
+      "learning_rate": 6.015037593984963e-08,
+      "loss": 0.1643,
+      "step": 66100
+    },
+    {
+      "epoch": 99.41,
+      "grad_norm": 10.434816360473633,
+      "learning_rate": 5.864661654135339e-08,
+      "loss": 0.2099,
+      "step": 66110
+    },
+    {
+      "epoch": 99.43,
+      "grad_norm": 3.1787760257720947,
+      "learning_rate": 5.714285714285715e-08,
+      "loss": 0.1539,
+      "step": 66120
+    },
+    {
+      "epoch": 99.44,
+      "grad_norm": 6.48075532913208,
+      "learning_rate": 5.563909774436091e-08,
+      "loss": 0.1741,
+      "step": 66130
+    },
+    {
+      "epoch": 99.46,
+      "grad_norm": 3.269533634185791,
+      "learning_rate": 5.4135338345864665e-08,
+      "loss": 0.1438,
+      "step": 66140
+    },
+    {
+      "epoch": 99.47,
+      "grad_norm": 8.6683988571167,
+      "learning_rate": 5.263157894736842e-08,
+      "loss": 0.1949,
+      "step": 66150
+    },
+    {
+      "epoch": 99.49,
+      "grad_norm": 4.051426887512207,
+      "learning_rate": 5.1127819548872186e-08,
+      "loss": 0.1634,
+      "step": 66160
+    },
+    {
+      "epoch": 99.5,
+      "grad_norm": 6.472201347351074,
+      "learning_rate": 4.962406015037595e-08,
+      "loss": 0.1474,
+      "step": 66170
+    },
+    {
+      "epoch": 99.52,
+      "grad_norm": 4.984071731567383,
+      "learning_rate": 4.81203007518797e-08,
+      "loss": 0.1707,
+      "step": 66180
+    },
+    {
+      "epoch": 99.53,
+      "grad_norm": 5.94837760925293,
+      "learning_rate": 4.6616541353383465e-08,
+      "loss": 0.2083,
+      "step": 66190
+    },
+    {
+      "epoch": 99.55,
+      "grad_norm": 5.263061046600342,
+      "learning_rate": 4.511278195488722e-08,
+      "loss": 0.1573,
+      "step": 66200
+    },
+    {
+      "epoch": 99.56,
+      "grad_norm": 5.428894996643066,
+      "learning_rate": 4.360902255639098e-08,
+      "loss": 0.1753,
+      "step": 66210
+    },
+    {
+      "epoch": 99.58,
+      "grad_norm": 6.092723846435547,
+      "learning_rate": 4.2105263157894737e-08,
+      "loss": 0.1658,
+      "step": 66220
+    },
+    {
+      "epoch": 99.59,
+      "grad_norm": 5.707245826721191,
+      "learning_rate": 4.06015037593985e-08,
+      "loss": 0.1583,
+      "step": 66230
+    },
+    {
+      "epoch": 99.61,
+      "grad_norm": 5.840432167053223,
+      "learning_rate": 3.9097744360902264e-08,
+      "loss": 0.2175,
+      "step": 66240
+    },
+    {
+      "epoch": 99.62,
+      "grad_norm": 5.600442409515381,
+      "learning_rate": 3.7593984962406015e-08,
+      "loss": 0.2012,
+      "step": 66250
+    },
+    {
+      "epoch": 99.64,
+      "grad_norm": 8.73216724395752,
+      "learning_rate": 3.609022556390978e-08,
+      "loss": 0.1126,
+      "step": 66260
+    },
+    {
+      "epoch": 99.65,
+      "grad_norm": 4.184023857116699,
+      "learning_rate": 3.4586466165413536e-08,
+      "loss": 0.1326,
+      "step": 66270
+    },
+    {
+      "epoch": 99.67,
+      "grad_norm": 2.4098730087280273,
+      "learning_rate": 3.30827067669173e-08,
+      "loss": 0.1219,
+      "step": 66280
+    },
+    {
+      "epoch": 99.68,
+      "grad_norm": 6.737592697143555,
+      "learning_rate": 3.157894736842106e-08,
+      "loss": 0.1788,
+      "step": 66290
+    },
+    {
+      "epoch": 99.7,
+      "grad_norm": 3.4307661056518555,
+      "learning_rate": 3.0075187969924815e-08,
+      "loss": 0.1677,
+      "step": 66300
+    },
+    {
+      "epoch": 99.71,
+      "grad_norm": 6.838919162750244,
+      "learning_rate": 2.8571428571428575e-08,
+      "loss": 0.156,
+      "step": 66310
+    },
+    {
+      "epoch": 99.73,
+      "grad_norm": 3.750256061553955,
+      "learning_rate": 2.7067669172932333e-08,
+      "loss": 0.1118,
+      "step": 66320
+    },
+    {
+      "epoch": 99.74,
+      "grad_norm": 3.541330099105835,
+      "learning_rate": 2.5563909774436093e-08,
+      "loss": 0.1467,
+      "step": 66330
+    },
+    {
+      "epoch": 99.76,
+      "grad_norm": 6.930870532989502,
+      "learning_rate": 2.406015037593985e-08,
+      "loss": 0.1469,
+      "step": 66340
+    },
+    {
+      "epoch": 99.77,
+      "grad_norm": 4.462535858154297,
+      "learning_rate": 2.255639097744361e-08,
+      "loss": 0.1693,
+      "step": 66350
+    },
+    {
+      "epoch": 99.79,
+      "grad_norm": 16.731098175048828,
+      "learning_rate": 2.1052631578947368e-08,
+      "loss": 0.1813,
+      "step": 66360
+    },
+    {
+      "epoch": 99.8,
+      "grad_norm": 4.853457450866699,
+      "learning_rate": 1.9548872180451132e-08,
+      "loss": 0.1693,
+      "step": 66370
+    },
+    {
+      "epoch": 99.82,
+      "grad_norm": 8.614503860473633,
+      "learning_rate": 1.804511278195489e-08,
+      "loss": 0.1212,
+      "step": 66380
+    },
+    {
+      "epoch": 99.83,
+      "grad_norm": 1.8390692472457886,
+      "learning_rate": 1.654135338345865e-08,
+      "loss": 0.1156,
+      "step": 66390
+    },
+    {
+      "epoch": 99.85,
+      "grad_norm": 3.2085022926330566,
+      "learning_rate": 1.5037593984962407e-08,
+      "loss": 0.1469,
+      "step": 66400
+    },
+    {
+      "epoch": 99.86,
+      "grad_norm": 4.798015594482422,
+      "learning_rate": 1.3533834586466166e-08,
+      "loss": 0.1651,
+      "step": 66410
+    },
+    {
+      "epoch": 99.88,
+      "grad_norm": 5.5648322105407715,
+      "learning_rate": 1.2030075187969925e-08,
+      "loss": 0.1682,
+      "step": 66420
+    },
+    {
+      "epoch": 99.89,
+      "grad_norm": 4.767200469970703,
+      "learning_rate": 1.0526315789473684e-08,
+      "loss": 0.2008,
+      "step": 66430
+    },
+    {
+      "epoch": 99.91,
+      "grad_norm": 2.938972234725952,
+      "learning_rate": 9.022556390977445e-09,
+      "loss": 0.1684,
+      "step": 66440
+    },
+    {
+      "epoch": 99.92,
+      "grad_norm": 7.364297389984131,
+      "learning_rate": 7.518796992481204e-09,
+      "loss": 0.1327,
+      "step": 66450
+    },
+    {
+      "epoch": 99.94,
+      "grad_norm": 4.032517433166504,
+      "learning_rate": 6.015037593984963e-09,
+      "loss": 0.1577,
+      "step": 66460
+    },
+    {
+      "epoch": 99.95,
+      "grad_norm": 5.099705219268799,
+      "learning_rate": 4.511278195488722e-09,
+      "loss": 0.1541,
+      "step": 66470
+    },
+    {
+      "epoch": 99.97,
+      "grad_norm": 5.294932842254639,
+      "learning_rate": 3.0075187969924813e-09,
+      "loss": 0.147,
+      "step": 66480
+    },
+    {
+      "epoch": 99.98,
+      "grad_norm": 5.263985633850098,
+      "learning_rate": 1.5037593984962407e-09,
+      "loss": 0.0833,
+      "step": 66490
+    },
+    {
+      "epoch": 100.0,
+      "grad_norm": 0.18724946677684784,
+      "learning_rate": 0.0,
+      "loss": 0.1471,
+      "step": 66500
+    },
+    {
+      "epoch": 100.0,
+      "eval_accuracy": 0.9309,
+      "eval_loss": 0.33010387420654297,
+      "eval_runtime": 84.7295,
+      "eval_samples_per_second": 118.023,
+      "eval_steps_per_second": 0.472,
+      "step": 66500
+    },
+    {
+      "epoch": 100.0,
+      "step": 66500,
+      "total_flos": 1.1646058381332455e+21,
+      "train_loss": 0.2947179788530321,
+      "train_runtime": 117326.6726,
+      "train_samples_per_second": 36.224,
+      "train_steps_per_second": 0.567
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 66500,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 100,
+  "save_steps": 500,
+  "total_flos": 1.1646058381332455e+21,
+  "train_batch_size": 64,
+  "trial_name": null,
+  "trial_params": null
+}