diff --git "a/checkpoint-5400/trainer_state.json" "b/checkpoint-5400/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-5400/trainer_state.json"
@@ -0,0 +1,3894 @@
+{
+  "best_metric": 0.23525514186650207,
+  "best_model_checkpoint": "./checkpoints/w2v-pa-v2/checkpoint-5400",
+  "epoch": 0.196235191511011,
+  "eval_steps": 600,
+  "global_step": 5400,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00036339850279816846,
+      "grad_norm": 12.712770462036133,
+      "learning_rate": 3.633985027981685e-08,
+      "loss": 8.1015,
+      "step": 10
+    },
+    {
+      "epoch": 0.0007267970055963369,
+      "grad_norm": 14.929631233215332,
+      "learning_rate": 7.26797005596337e-08,
+      "loss": 8.5111,
+      "step": 20
+    },
+    {
+      "epoch": 0.0010901955083945055,
+      "grad_norm": 12.774781227111816,
+      "learning_rate": 1.0901955083945056e-07,
+      "loss": 8.71,
+      "step": 30
+    },
+    {
+      "epoch": 0.0014535940111926739,
+      "grad_norm": 15.27083969116211,
+      "learning_rate": 1.417254160912857e-07,
+      "loss": 9.2894,
+      "step": 40
+    },
+    {
+      "epoch": 0.0018169925139908424,
+      "grad_norm": 31.209775924682617,
+      "learning_rate": 1.7806526637110256e-07,
+      "loss": 8.6779,
+      "step": 50
+    },
+    {
+      "epoch": 0.002180391016789011,
+      "grad_norm": 12.960335731506348,
+      "learning_rate": 2.1440511665091943e-07,
+      "loss": 8.0399,
+      "step": 60
+    },
+    {
+      "epoch": 0.002543789519587179,
+      "grad_norm": 14.452157020568848,
+      "learning_rate": 2.5074496693073626e-07,
+      "loss": 7.9357,
+      "step": 70
+    },
+    {
+      "epoch": 0.0029071880223853477,
+      "grad_norm": 12.74867057800293,
+      "learning_rate": 2.8708481721055307e-07,
+      "loss": 8.0764,
+      "step": 80
+    },
+    {
+      "epoch": 0.0032705865251835163,
+      "grad_norm": 19.905397415161133,
+      "learning_rate": 3.2342466749036993e-07,
+      "loss": 8.2522,
+      "step": 90
+    },
+    {
+      "epoch": 0.003633985027981685,
+      "grad_norm": Infinity,
+      "learning_rate": 3.5613053274220513e-07,
+      "loss": 7.8161,
+      "step": 100
+    },
+    {
+      "epoch": 0.003997383530779853,
+      "grad_norm": 15.877684593200684,
+      "learning_rate": 3.92470383022022e-07,
+      "loss": 7.4866,
+      "step": 110
+    },
+    {
+      "epoch": 0.004360782033578022,
+      "grad_norm": 19.216800689697266,
+      "learning_rate": 4.2881023330183885e-07,
+      "loss": 7.0324,
+      "step": 120
+    },
+    {
+      "epoch": 0.0047241805363761906,
+      "grad_norm": 16.937118530273438,
+      "learning_rate": 4.651500835816557e-07,
+      "loss": 6.7873,
+      "step": 130
+    },
+    {
+      "epoch": 0.005087579039174358,
+      "grad_norm": 27.858692169189453,
+      "learning_rate": 5.014899338614725e-07,
+      "loss": 7.2063,
+      "step": 140
+    },
+    {
+      "epoch": 0.005450977541972527,
+      "grad_norm": Infinity,
+      "learning_rate": 5.341957991133076e-07,
+      "loss": 6.5827,
+      "step": 150
+    },
+    {
+      "epoch": 0.005814376044770695,
+      "grad_norm": 21.252164840698242,
+      "learning_rate": 5.669016643651428e-07,
+      "loss": 7.819,
+      "step": 160
+    },
+    {
+      "epoch": 0.006177774547568864,
+      "grad_norm": 20.977886199951172,
+      "learning_rate": 6.032415146449597e-07,
+      "loss": 5.4741,
+      "step": 170
+    },
+    {
+      "epoch": 0.0065411730503670325,
+      "grad_norm": 35.25390625,
+      "learning_rate": 6.395813649247765e-07,
+      "loss": 5.7225,
+      "step": 180
+    },
+    {
+      "epoch": 0.006904571553165201,
+      "grad_norm": 7.165033340454102,
+      "learning_rate": 6.759212152045934e-07,
+      "loss": 4.7475,
+      "step": 190
+    },
+    {
+      "epoch": 0.00726797005596337,
+      "grad_norm": 14.877301216125488,
+      "learning_rate": 7.122610654844103e-07,
+      "loss": 4.6375,
+      "step": 200
+    },
+    {
+      "epoch": 0.007631368558761538,
+      "grad_norm": 5.826667785644531,
+      "learning_rate": 7.486009157642272e-07,
+      "loss": 4.356,
+      "step": 210
+    },
+    {
+      "epoch": 0.007994767061559707,
+      "grad_norm": 6.022212982177734,
+      "learning_rate": 7.84940766044044e-07,
+      "loss": 4.2138,
+      "step": 220
+    },
+    {
+      "epoch": 0.008358165564357875,
+      "grad_norm": 4.790489196777344,
+      "learning_rate": 8.212806163238608e-07,
+      "loss": 4.0662,
+      "step": 230
+    },
+    {
+      "epoch": 0.008721564067156044,
+      "grad_norm": 4.448057174682617,
+      "learning_rate": 8.576204666036777e-07,
+      "loss": 3.9507,
+      "step": 240
+    },
+    {
+      "epoch": 0.009084962569954213,
+      "grad_norm": 47.487003326416016,
+      "learning_rate": 8.939603168834945e-07,
+      "loss": 3.9576,
+      "step": 250
+    },
+    {
+      "epoch": 0.009448361072752381,
+      "grad_norm": 8.64856243133545,
+      "learning_rate": 9.303001671633114e-07,
+      "loss": 3.7102,
+      "step": 260
+    },
+    {
+      "epoch": 0.009811759575550548,
+      "grad_norm": 8.821709632873535,
+      "learning_rate": 9.66640017443128e-07,
+      "loss": 3.6644,
+      "step": 270
+    },
+    {
+      "epoch": 0.010175158078348717,
+      "grad_norm": 14.071539878845215,
+      "learning_rate": 1.002979867722945e-06,
+      "loss": 3.6909,
+      "step": 280
+    },
+    {
+      "epoch": 0.010538556581146885,
+      "grad_norm": 6.68039083480835,
+      "learning_rate": 1.0393197180027619e-06,
+      "loss": 3.6458,
+      "step": 290
+    },
+    {
+      "epoch": 0.010901955083945054,
+      "grad_norm": 20.664649963378906,
+      "learning_rate": 1.0756595682825787e-06,
+      "loss": 3.6332,
+      "step": 300
+    },
+    {
+      "epoch": 0.011265353586743222,
+      "grad_norm": 2.5272624492645264,
+      "learning_rate": 1.1119994185623955e-06,
+      "loss": 3.4835,
+      "step": 310
+    },
+    {
+      "epoch": 0.01162875208954139,
+      "grad_norm": 8.353235244750977,
+      "learning_rate": 1.1483392688422123e-06,
+      "loss": 3.4892,
+      "step": 320
+    },
+    {
+      "epoch": 0.01199215059233956,
+      "grad_norm": 7.0964531898498535,
+      "learning_rate": 1.1846791191220293e-06,
+      "loss": 3.5586,
+      "step": 330
+    },
+    {
+      "epoch": 0.012355549095137728,
+      "grad_norm": 4.734161376953125,
+      "learning_rate": 1.2210189694018461e-06,
+      "loss": 3.4996,
+      "step": 340
+    },
+    {
+      "epoch": 0.012718947597935897,
+      "grad_norm": 47.409996032714844,
+      "learning_rate": 1.257358819681663e-06,
+      "loss": 3.5565,
+      "step": 350
+    },
+    {
+      "epoch": 0.013082346100734065,
+      "grad_norm": 2.880244016647339,
+      "learning_rate": 1.2936986699614797e-06,
+      "loss": 3.4154,
+      "step": 360
+    },
+    {
+      "epoch": 0.013445744603532234,
+      "grad_norm": 6.637233734130859,
+      "learning_rate": 1.3300385202412968e-06,
+      "loss": 3.4119,
+      "step": 370
+    },
+    {
+      "epoch": 0.013809143106330402,
+      "grad_norm": 11.791736602783203,
+      "learning_rate": 1.3663783705211136e-06,
+      "loss": 3.4347,
+      "step": 380
+    },
+    {
+      "epoch": 0.01417254160912857,
+      "grad_norm": 8.274836540222168,
+      "learning_rate": 1.4027182208009304e-06,
+      "loss": 3.4253,
+      "step": 390
+    },
+    {
+      "epoch": 0.01453594011192674,
+      "grad_norm": 10.09929084777832,
+      "learning_rate": 1.4390580710807472e-06,
+      "loss": 3.4702,
+      "step": 400
+    },
+    {
+      "epoch": 0.014899338614724908,
+      "grad_norm": 6.32951545715332,
+      "learning_rate": 1.4753979213605642e-06,
+      "loss": 3.3513,
+      "step": 410
+    },
+    {
+      "epoch": 0.015262737117523077,
+      "grad_norm": 2.4888486862182617,
+      "learning_rate": 1.511737771640381e-06,
+      "loss": 3.3421,
+      "step": 420
+    },
+    {
+      "epoch": 0.015626135620321245,
+      "grad_norm": 3.02103328704834,
+      "learning_rate": 1.5480776219201978e-06,
+      "loss": 3.397,
+      "step": 430
+    },
+    {
+      "epoch": 0.015989534123119414,
+      "grad_norm": 7.464268207550049,
+      "learning_rate": 1.5844174722000146e-06,
+      "loss": 3.3582,
+      "step": 440
+    },
+    {
+      "epoch": 0.016352932625917582,
+      "grad_norm": 18.908123016357422,
+      "learning_rate": 1.6207573224798317e-06,
+      "loss": 3.4034,
+      "step": 450
+    },
+    {
+      "epoch": 0.01671633112871575,
+      "grad_norm": 2.487326145172119,
+      "learning_rate": 1.6570971727596485e-06,
+      "loss": 3.2229,
+      "step": 460
+    },
+    {
+      "epoch": 0.01707972963151392,
+      "grad_norm": 2.3999946117401123,
+      "learning_rate": 1.6934370230394653e-06,
+      "loss": 3.2185,
+      "step": 470
+    },
+    {
+      "epoch": 0.017443128134312088,
+      "grad_norm": 5.007234573364258,
+      "learning_rate": 1.729776873319282e-06,
+      "loss": 3.2069,
+      "step": 480
+    },
+    {
+      "epoch": 0.017806526637110257,
+      "grad_norm": 6.393301963806152,
+      "learning_rate": 1.766116723599099e-06,
+      "loss": 3.0687,
+      "step": 490
+    },
+    {
+      "epoch": 0.018169925139908425,
+      "grad_norm": 45.44938278198242,
+      "learning_rate": 1.802456573878916e-06,
+      "loss": 3.112,
+      "step": 500
+    },
+    {
+      "epoch": 0.018533323642706594,
+      "grad_norm": 7.32182502746582,
+      "learning_rate": 1.8387964241587327e-06,
+      "loss": 2.951,
+      "step": 510
+    },
+    {
+      "epoch": 0.018896722145504762,
+      "grad_norm": 3.3864173889160156,
+      "learning_rate": 1.8751362744385495e-06,
+      "loss": 2.8879,
+      "step": 520
+    },
+    {
+      "epoch": 0.019260120648302927,
+      "grad_norm": 5.429958343505859,
+      "learning_rate": 1.911476124718366e-06,
+      "loss": 2.7393,
+      "step": 530
+    },
+    {
+      "epoch": 0.019623519151101096,
+      "grad_norm": 5.3577985763549805,
+      "learning_rate": 1.947815974998183e-06,
+      "loss": 2.4813,
+      "step": 540
+    },
+    {
+      "epoch": 0.019986917653899264,
+      "grad_norm": 13.970659255981445,
+      "learning_rate": 1.9841558252779998e-06,
+      "loss": 2.3787,
+      "step": 550
+    },
+    {
+      "epoch": 0.020350316156697433,
+      "grad_norm": 5.2666754722595215,
+      "learning_rate": 2.0204956755578166e-06,
+      "loss": 2.207,
+      "step": 560
+    },
+    {
+      "epoch": 0.0207137146594956,
+      "grad_norm": 4.184991359710693,
+      "learning_rate": 2.0568355258376334e-06,
+      "loss": 2.0383,
+      "step": 570
+    },
+    {
+      "epoch": 0.02107711316229377,
+      "grad_norm": 6.312343597412109,
+      "learning_rate": 2.09317537611745e-06,
+      "loss": 1.8416,
+      "step": 580
+    },
+    {
+      "epoch": 0.02144051166509194,
+      "grad_norm": 4.754147529602051,
+      "learning_rate": 2.1295152263972674e-06,
+      "loss": 1.6002,
+      "step": 590
+    },
+    {
+      "epoch": 0.021803910167890107,
+      "grad_norm": 21.47913360595703,
+      "learning_rate": 2.1658550766770842e-06,
+      "loss": 1.6015,
+      "step": 600
+    },
+    {
+      "epoch": 0.021803910167890107,
+      "eval_loss": 1.5154471397399902,
+      "eval_runtime": 180.9184,
+      "eval_samples_per_second": 40.98,
+      "eval_steps_per_second": 5.124,
+      "eval_wer": 0.7997531177954872,
+      "step": 600
+    },
+    {
+      "epoch": 0.022167308670688276,
+      "grad_norm": 4.2374348640441895,
+      "learning_rate": 2.202194926956901e-06,
+      "loss": 1.4842,
+      "step": 610
+    },
+    {
+      "epoch": 0.022530707173486444,
+      "grad_norm": 4.392132759094238,
+      "learning_rate": 2.238534777236718e-06,
+      "loss": 1.3776,
+      "step": 620
+    },
+    {
+      "epoch": 0.022894105676284613,
+      "grad_norm": 4.682064533233643,
+      "learning_rate": 2.2748746275165347e-06,
+      "loss": 1.3177,
+      "step": 630
+    },
+    {
+      "epoch": 0.02325750417908278,
+      "grad_norm": 4.8396077156066895,
+      "learning_rate": 2.3112144777963515e-06,
+      "loss": 1.0737,
+      "step": 640
+    },
+    {
+      "epoch": 0.02362090268188095,
+      "grad_norm": 33.27382278442383,
+      "learning_rate": 2.3475543280761683e-06,
+      "loss": 1.3046,
+      "step": 650
+    },
+    {
+      "epoch": 0.02398430118467912,
+      "grad_norm": 5.410325050354004,
+      "learning_rate": 2.383894178355985e-06,
+      "loss": 1.1021,
+      "step": 660
+    },
+    {
+      "epoch": 0.024347699687477287,
+      "grad_norm": 3.9523680210113525,
+      "learning_rate": 2.420234028635802e-06,
+      "loss": 1.0602,
+      "step": 670
+    },
+    {
+      "epoch": 0.024711098190275456,
+      "grad_norm": 9.141073226928711,
+      "learning_rate": 2.456573878915619e-06,
+      "loss": 1.0631,
+      "step": 680
+    },
+    {
+      "epoch": 0.025074496693073624,
+      "grad_norm": 5.3534626960754395,
+      "learning_rate": 2.492913729195436e-06,
+      "loss": 0.8968,
+      "step": 690
+    },
+    {
+      "epoch": 0.025437895195871793,
+      "grad_norm": 32.30677795410156,
+      "learning_rate": 2.5292535794752527e-06,
+      "loss": 1.0439,
+      "step": 700
+    },
+    {
+      "epoch": 0.02580129369866996,
+      "grad_norm": 4.310474872589111,
+      "learning_rate": 2.5655934297550696e-06,
+      "loss": 0.954,
+      "step": 710
+    },
+    {
+      "epoch": 0.02616469220146813,
+      "grad_norm": 5.586440563201904,
+      "learning_rate": 2.6019332800348864e-06,
+      "loss": 1.0031,
+      "step": 720
+    },
+    {
+      "epoch": 0.0265280907042663,
+      "grad_norm": 3.6927313804626465,
+      "learning_rate": 2.638273130314703e-06,
+      "loss": 0.7956,
+      "step": 730
+    },
+    {
+      "epoch": 0.026891489207064467,
+      "grad_norm": 4.270529747009277,
+      "learning_rate": 2.67461298059452e-06,
+      "loss": 0.8874,
+      "step": 740
+    },
+    {
+      "epoch": 0.027254887709862636,
+      "grad_norm": 23.553489685058594,
+      "learning_rate": 2.710952830874337e-06,
+      "loss": 0.8523,
+      "step": 750
+    },
+    {
+      "epoch": 0.027618286212660804,
+      "grad_norm": 5.342041492462158,
+      "learning_rate": 2.747292681154154e-06,
+      "loss": 0.9029,
+      "step": 760
+    },
+    {
+      "epoch": 0.027981684715458973,
+      "grad_norm": 3.3802621364593506,
+      "learning_rate": 2.783632531433971e-06,
+      "loss": 0.8378,
+      "step": 770
+    },
+    {
+      "epoch": 0.02834508321825714,
+      "grad_norm": 6.378807067871094,
+      "learning_rate": 2.8199723817137876e-06,
+      "loss": 0.8085,
+      "step": 780
+    },
+    {
+      "epoch": 0.02870848172105531,
+      "grad_norm": 4.007000923156738,
+      "learning_rate": 2.8563122319936045e-06,
+      "loss": 0.8218,
+      "step": 790
+    },
+    {
+      "epoch": 0.02907188022385348,
+      "grad_norm": 68.16226196289062,
+      "learning_rate": 2.8926520822734213e-06,
+      "loss": 1.2055,
+      "step": 800
+    },
+    {
+      "epoch": 0.029435278726651647,
+      "grad_norm": 6.70043420791626,
+      "learning_rate": 2.928991932553238e-06,
+      "loss": 0.7641,
+      "step": 810
+    },
+    {
+      "epoch": 0.029798677229449816,
+      "grad_norm": 5.498161315917969,
+      "learning_rate": 2.965331782833055e-06,
+      "loss": 0.7739,
+      "step": 820
+    },
+    {
+      "epoch": 0.030162075732247984,
+      "grad_norm": 9.515852928161621,
+      "learning_rate": 3.0016716331128717e-06,
+      "loss": 0.8293,
+      "step": 830
+    },
+    {
+      "epoch": 0.030525474235046153,
+      "grad_norm": 13.3881196975708,
+      "learning_rate": 3.0380114833926885e-06,
+      "loss": 0.5597,
+      "step": 840
+    },
+    {
+      "epoch": 0.03088887273784432,
+      "grad_norm": 13.670549392700195,
+      "learning_rate": 3.0743513336725057e-06,
+      "loss": 0.7658,
+      "step": 850
+    },
+    {
+      "epoch": 0.03125227124064249,
+      "grad_norm": 3.58305287361145,
+      "learning_rate": 3.1106911839523226e-06,
+      "loss": 0.7036,
+      "step": 860
+    },
+    {
+      "epoch": 0.031615669743440655,
+      "grad_norm": 4.119450569152832,
+      "learning_rate": 3.147031034232139e-06,
+      "loss": 0.6842,
+      "step": 870
+    },
+    {
+      "epoch": 0.03197906824623883,
+      "grad_norm": 6.412299156188965,
+      "learning_rate": 3.183370884511956e-06,
+      "loss": 0.7148,
+      "step": 880
+    },
+    {
+      "epoch": 0.03234246674903699,
+      "grad_norm": 8.700023651123047,
+      "learning_rate": 3.2197107347917726e-06,
+      "loss": 1.4861,
+      "step": 890
+    },
+    {
+      "epoch": 0.032705865251835164,
+      "grad_norm": 18.78075408935547,
+      "learning_rate": 3.25605058507159e-06,
+      "loss": 0.7162,
+      "step": 900
+    },
+    {
+      "epoch": 0.03306926375463333,
+      "grad_norm": 4.078335762023926,
+      "learning_rate": 3.292390435351406e-06,
+      "loss": 0.643,
+      "step": 910
+    },
+    {
+      "epoch": 0.0334326622574315,
+      "grad_norm": 6.603452682495117,
+      "learning_rate": 3.3287302856312234e-06,
+      "loss": 0.6623,
+      "step": 920
+    },
+    {
+      "epoch": 0.03379606076022967,
+      "grad_norm": 5.817732334136963,
+      "learning_rate": 3.3650701359110402e-06,
+      "loss": 0.6265,
+      "step": 930
+    },
+    {
+      "epoch": 0.03415945926302784,
+      "grad_norm": 8.310086250305176,
+      "learning_rate": 3.4014099861908575e-06,
+      "loss": 0.5343,
+      "step": 940
+    },
+    {
+      "epoch": 0.034522857765826004,
+      "grad_norm": NaN,
+      "learning_rate": 3.4341158514426923e-06,
+      "loss": 0.6999,
+      "step": 950
+    },
+    {
+      "epoch": 0.034886256268624176,
+      "grad_norm": 4.416926860809326,
+      "learning_rate": 3.4704557017225087e-06,
+      "loss": 0.6688,
+      "step": 960
+    },
+    {
+      "epoch": 0.03524965477142234,
+      "grad_norm": 3.2407495975494385,
+      "learning_rate": 3.506795552002326e-06,
+      "loss": 0.5808,
+      "step": 970
+    },
+    {
+      "epoch": 0.03561305327422051,
+      "grad_norm": NaN,
+      "learning_rate": 3.539501417254161e-06,
+      "loss": 3.0266,
+      "step": 980
+    },
+    {
+      "epoch": 0.03597645177701868,
+      "grad_norm": 8.086112022399902,
+      "learning_rate": 3.575841267533978e-06,
+      "loss": 0.5829,
+      "step": 990
+    },
+    {
+      "epoch": 0.03633985027981685,
+      "grad_norm": 252.45077514648438,
+      "learning_rate": 3.612181117813795e-06,
+      "loss": 0.7371,
+      "step": 1000
+    },
+    {
+      "epoch": 0.036703248782615015,
+      "grad_norm": 3.5969936847686768,
+      "learning_rate": 3.648520968093612e-06,
+      "loss": 0.6632,
+      "step": 1010
+    },
+    {
+      "epoch": 0.03706664728541319,
+      "grad_norm": 3.0116841793060303,
+      "learning_rate": 3.6848608183734285e-06,
+      "loss": 0.537,
+      "step": 1020
+    },
+    {
+      "epoch": 0.03743004578821135,
+      "grad_norm": 5.494657039642334,
+      "learning_rate": 3.7212006686532457e-06,
+      "loss": 0.5422,
+      "step": 1030
+    },
+    {
+      "epoch": 0.037793444291009524,
+      "grad_norm": 21.526798248291016,
+      "learning_rate": 3.757540518933062e-06,
+      "loss": 0.5003,
+      "step": 1040
+    },
+    {
+      "epoch": 0.03815684279380769,
+      "grad_norm": 80.90055084228516,
+      "learning_rate": 3.7938803692128793e-06,
+      "loss": 0.6566,
+      "step": 1050
+    },
+    {
+      "epoch": 0.038520241296605855,
+      "grad_norm": 3.7678096294403076,
+      "learning_rate": 3.830220219492696e-06,
+      "loss": 0.5758,
+      "step": 1060
+    },
+    {
+      "epoch": 0.03888363979940403,
+      "grad_norm": 4.526616096496582,
+      "learning_rate": 3.866560069772512e-06,
+      "loss": 0.5648,
+      "step": 1070
+    },
+    {
+      "epoch": 0.03924703830220219,
+      "grad_norm": 4.571674346923828,
+      "learning_rate": 3.90289992005233e-06,
+      "loss": 0.5864,
+      "step": 1080
+    },
+    {
+      "epoch": 0.039610436805000364,
+      "grad_norm": 5.295219421386719,
+      "learning_rate": 3.939239770332146e-06,
+      "loss": 0.4476,
+      "step": 1090
+    },
+    {
+      "epoch": 0.03997383530779853,
+      "grad_norm": 16.631162643432617,
+      "learning_rate": 3.975579620611963e-06,
+      "loss": 0.6198,
+      "step": 1100
+    },
+    {
+      "epoch": 0.0403372338105967,
+      "grad_norm": 4.685397624969482,
+      "learning_rate": 4.01191947089178e-06,
+      "loss": 0.7512,
+      "step": 1110
+    },
+    {
+      "epoch": 0.040700632313394866,
+      "grad_norm": 3.333232879638672,
+      "learning_rate": 4.048259321171597e-06,
+      "loss": 0.5087,
+      "step": 1120
+    },
+    {
+      "epoch": 0.04106403081619304,
+      "grad_norm": 5.501911640167236,
+      "learning_rate": 4.084599171451414e-06,
+      "loss": 0.5772,
+      "step": 1130
+    },
+    {
+      "epoch": 0.0414274293189912,
+      "grad_norm": 8.066693305969238,
+      "learning_rate": 4.120939021731231e-06,
+      "loss": 0.4641,
+      "step": 1140
+    },
+    {
+      "epoch": 0.041790827821789375,
+      "grad_norm": 13.463829040527344,
+      "learning_rate": 4.1572788720110474e-06,
+      "loss": 0.5192,
+      "step": 1150
+    },
+    {
+      "epoch": 0.04215422632458754,
+      "grad_norm": 4.132773399353027,
+      "learning_rate": 4.193618722290864e-06,
+      "loss": 0.4696,
+      "step": 1160
+    },
+    {
+      "epoch": 0.04251762482738571,
+      "grad_norm": 6.176777362823486,
+      "learning_rate": 4.229958572570681e-06,
+      "loss": 0.4851,
+      "step": 1170
+    },
+    {
+      "epoch": 0.04288102333018388,
+      "grad_norm": 8.26610279083252,
+      "learning_rate": 4.266298422850498e-06,
+      "loss": 0.4967,
+      "step": 1180
+    },
+    {
+      "epoch": 0.04324442183298205,
+      "grad_norm": 3.9725544452667236,
+      "learning_rate": 4.302638273130315e-06,
+      "loss": 0.431,
+      "step": 1190
+    },
+    {
+      "epoch": 0.043607820335780215,
+      "grad_norm": 22.353294372558594,
+      "learning_rate": 4.338978123410132e-06,
+      "loss": 0.6523,
+      "step": 1200
+    },
+    {
+      "epoch": 0.043607820335780215,
+      "eval_loss": 0.5945897102355957,
+      "eval_runtime": 180.5168,
+      "eval_samples_per_second": 41.071,
+      "eval_steps_per_second": 5.135,
+      "eval_wer": 0.3718390908925881,
+      "step": 1200
+    },
+    {
+      "epoch": 0.04397121883857839,
+      "grad_norm": 3.7954189777374268,
+      "learning_rate": 4.375317973689948e-06,
+      "loss": 0.4511,
+      "step": 1210
+    },
+    {
+      "epoch": 0.04433461734137655,
+      "grad_norm": 5.583435535430908,
+      "learning_rate": 4.411657823969766e-06,
+      "loss": 1.4019,
+      "step": 1220
+    },
+    {
+      "epoch": 0.044698015844174724,
+      "grad_norm": 8.544243812561035,
+      "learning_rate": 4.447997674249582e-06,
+      "loss": 0.4467,
+      "step": 1230
+    },
+    {
+      "epoch": 0.04506141434697289,
+      "grad_norm": 3.8716418743133545,
+      "learning_rate": 4.4843375245293996e-06,
+      "loss": 0.4347,
+      "step": 1240
+    },
+    {
+      "epoch": 0.04542481284977106,
+      "grad_norm": 19.459606170654297,
+      "learning_rate": 4.5206773748092155e-06,
+      "loss": 0.616,
+      "step": 1250
+    },
+    {
+      "epoch": 0.045788211352569226,
+      "grad_norm": 5.474793434143066,
+      "learning_rate": 4.557017225089033e-06,
+      "loss": 0.4689,
+      "step": 1260
+    },
+    {
+      "epoch": 0.0461516098553674,
+      "grad_norm": 4.705495834350586,
+      "learning_rate": 4.593357075368849e-06,
+      "loss": 0.4623,
+      "step": 1270
+    },
+    {
+      "epoch": 0.04651500835816556,
+      "grad_norm": 6.779942035675049,
+      "learning_rate": 4.629696925648667e-06,
+      "loss": 0.4418,
+      "step": 1280
+    },
+    {
+      "epoch": 0.046878406860963735,
+      "grad_norm": 6.802936553955078,
+      "learning_rate": 4.666036775928484e-06,
+      "loss": 0.4429,
+      "step": 1290
+    },
+    {
+      "epoch": 0.0472418053637619,
+      "grad_norm": 17.47754669189453,
+      "learning_rate": 4.7023766262083004e-06,
+      "loss": 0.596,
+      "step": 1300
+    },
+    {
+      "epoch": 0.04760520386656007,
+      "grad_norm": 4.036036968231201,
+      "learning_rate": 4.738716476488117e-06,
+      "loss": 0.4362,
+      "step": 1310
+    },
+    {
+      "epoch": 0.04796860236935824,
+      "grad_norm": 6.022701740264893,
+      "learning_rate": 4.775056326767934e-06,
+      "loss": 0.5092,
+      "step": 1320
+    },
+    {
+      "epoch": 0.04833200087215641,
+      "grad_norm": 5.533923625946045,
+      "learning_rate": 4.811396177047751e-06,
+      "loss": 0.4358,
+      "step": 1330
+    },
+    {
+      "epoch": 0.048695399374954575,
+      "grad_norm": 3.4037017822265625,
+      "learning_rate": 4.847736027327568e-06,
+      "loss": 0.3684,
+      "step": 1340
+    },
+    {
+      "epoch": 0.04905879787775275,
+      "grad_norm": 13.625974655151367,
+      "learning_rate": 4.8840758776073845e-06,
+      "loss": 0.583,
+      "step": 1350
+    },
+    {
+      "epoch": 0.04942219638055091,
+      "grad_norm": 3.597294330596924,
+      "learning_rate": 4.920415727887201e-06,
+      "loss": 0.4561,
+      "step": 1360
+    },
+    {
+      "epoch": 0.049785594883349084,
+      "grad_norm": 2.8846936225891113,
+      "learning_rate": 4.956755578167018e-06,
+      "loss": 0.409,
+      "step": 1370
+    },
+    {
+      "epoch": 0.05014899338614725,
+      "grad_norm": 5.500187397003174,
+      "learning_rate": 4.993095428446836e-06,
+      "loss": 0.4531,
+      "step": 1380
+    },
+    {
+      "epoch": 0.05051239188894542,
+      "grad_norm": 3.1203413009643555,
+      "learning_rate": 5.029435278726652e-06,
+      "loss": 0.4004,
+      "step": 1390
+    },
+    {
+      "epoch": 0.050875790391743586,
+      "grad_norm": 98.18115234375,
+      "learning_rate": 5.0657751290064685e-06,
+      "loss": 0.522,
+      "step": 1400
+    },
+    {
+      "epoch": 0.05123918889454175,
+      "grad_norm": 3.901418924331665,
+      "learning_rate": 5.102114979286285e-06,
+      "loss": 0.4041,
+      "step": 1410
+    },
+    {
+      "epoch": 0.05160258739733992,
+      "grad_norm": 4.045637130737305,
+      "learning_rate": 5.138454829566102e-06,
+      "loss": 0.4051,
+      "step": 1420
+    },
+    {
+      "epoch": 0.05196598590013809,
+      "grad_norm": 6.835183143615723,
+      "learning_rate": 5.174794679845919e-06,
+      "loss": 0.4937,
+      "step": 1430
+    },
+    {
+      "epoch": 0.05232938440293626,
+      "grad_norm": 7.708272457122803,
+      "learning_rate": 5.211134530125736e-06,
+      "loss": 0.3818,
+      "step": 1440
+    },
+    {
+      "epoch": 0.052692782905734425,
+      "grad_norm": 24.2607364654541,
+      "learning_rate": 5.247474380405553e-06,
+      "loss": 0.5445,
+      "step": 1450
+    },
+    {
+      "epoch": 0.0530561814085326,
+      "grad_norm": 3.3517005443573,
+      "learning_rate": 5.283814230685369e-06,
+      "loss": 0.4079,
+      "step": 1460
+    },
+    {
+      "epoch": 0.05341957991133076,
+      "grad_norm": 12.727778434753418,
+      "learning_rate": 5.320154080965187e-06,
+      "loss": 0.4285,
+      "step": 1470
+    },
+    {
+      "epoch": 0.053782978414128935,
+      "grad_norm": 4.984294891357422,
+      "learning_rate": 5.356493931245003e-06,
+      "loss": 0.5006,
+      "step": 1480
+    },
+    {
+      "epoch": 0.0541463769169271,
+      "grad_norm": 3.3041558265686035,
+      "learning_rate": 5.392833781524821e-06,
+      "loss": 0.3729,
+      "step": 1490
+    },
+    {
+      "epoch": 0.05450977541972527,
+      "grad_norm": 38.074546813964844,
+      "learning_rate": 5.429173631804637e-06,
+      "loss": 0.5401,
+      "step": 1500
+    },
+    {
+      "epoch": 0.05487317392252344,
+      "grad_norm": 5.649720668792725,
+      "learning_rate": 5.465513482084454e-06,
+      "loss": 0.3879,
+      "step": 1510
+    },
+    {
+      "epoch": 0.05523657242532161,
+      "grad_norm": 3.107583522796631,
+      "learning_rate": 5.501853332364271e-06,
+      "loss": 0.4144,
+      "step": 1520
+    },
+    {
+      "epoch": 0.055599970928119774,
+      "grad_norm": 19.246564865112305,
+      "learning_rate": 5.538193182644088e-06,
+      "loss": 0.4314,
+      "step": 1530
+    },
+    {
+      "epoch": 0.055963369430917946,
+      "grad_norm": 4.72367525100708,
+      "learning_rate": 5.574533032923905e-06,
+      "loss": 0.3576,
+      "step": 1540
+    },
+    {
+      "epoch": 0.05632676793371611,
+      "grad_norm": 25.88886260986328,
+      "learning_rate": 5.6108728832037215e-06,
+      "loss": 0.5385,
+      "step": 1550
+    },
+    {
+      "epoch": 0.05669016643651428,
+      "grad_norm": 3.1524956226348877,
+      "learning_rate": 5.647212733483538e-06,
+      "loss": 0.4075,
+      "step": 1560
+    },
+    {
+      "epoch": 0.05705356493931245,
+      "grad_norm": 3.883281707763672,
+      "learning_rate": 5.683552583763355e-06,
+      "loss": 0.4242,
+      "step": 1570
+    },
+    {
+      "epoch": 0.05741696344211062,
+      "grad_norm": 16.935935974121094,
+      "learning_rate": 5.719892434043172e-06,
+      "loss": 0.6194,
+      "step": 1580
+    },
+    {
+      "epoch": 0.057780361944908785,
+      "grad_norm": 4.23909330368042,
+      "learning_rate": 5.756232284322989e-06,
+      "loss": 0.4206,
+      "step": 1590
+    },
+    {
+      "epoch": 0.05814376044770696,
+      "grad_norm": 16.6039981842041,
+      "learning_rate": 5.7925721346028056e-06,
+      "loss": 0.4854,
+      "step": 1600
+    },
+    {
+      "epoch": 0.05850715895050512,
+      "grad_norm": 2.5220890045166016,
+      "learning_rate": 5.828911984882622e-06,
+      "loss": 0.4186,
+      "step": 1610
+    },
+    {
+      "epoch": 0.058870557453303295,
+      "grad_norm": 3.075101613998413,
+      "learning_rate": 5.865251835162439e-06,
+      "loss": 0.8877,
+      "step": 1620
+    },
+    {
+      "epoch": 0.05923395595610146,
+      "grad_norm": 5.511383056640625,
+      "learning_rate": 5.901591685442257e-06,
+      "loss": 0.4219,
+      "step": 1630
+    },
+    {
+      "epoch": 0.05959735445889963,
+      "grad_norm": 2.9449989795684814,
+      "learning_rate": 5.937931535722073e-06,
+      "loss": 0.2992,
+      "step": 1640
+    },
+    {
+      "epoch": 0.0599607529616978,
+      "grad_norm": 31.823612213134766,
+      "learning_rate": 5.9742713860018905e-06,
+      "loss": 0.5939,
+      "step": 1650
+    },
+    {
+      "epoch": 0.06032415146449597,
+      "grad_norm": 4.240995407104492,
+      "learning_rate": 6.010611236281706e-06,
+      "loss": 0.4176,
+      "step": 1660
+    },
+    {
+      "epoch": 0.060687549967294134,
+      "grad_norm": 2.6084980964660645,
+      "learning_rate": 6.046951086561524e-06,
+      "loss": 0.3542,
+      "step": 1670
+    },
+    {
+      "epoch": 0.061050948470092306,
+      "grad_norm": 8.318774223327637,
+      "learning_rate": 6.08329093684134e-06,
+      "loss": 0.3968,
+      "step": 1680
+    },
+    {
+      "epoch": 0.06141434697289047,
+      "grad_norm": 5.18604850769043,
+      "learning_rate": 6.119630787121158e-06,
+      "loss": 0.3879,
+      "step": 1690
+    },
+    {
+      "epoch": 0.06177774547568864,
+      "grad_norm": 51.732086181640625,
+      "learning_rate": 6.1559706374009745e-06,
+      "loss": 0.5025,
+      "step": 1700
+    },
+    {
+      "epoch": 0.06214114397848681,
+      "grad_norm": 2.5876500606536865,
+      "learning_rate": 6.192310487680791e-06,
+      "loss": 0.3558,
+      "step": 1710
+    },
+    {
+      "epoch": 0.06250454248128498,
+      "grad_norm": 5.071794033050537,
+      "learning_rate": 6.228650337960608e-06,
+      "loss": 0.3534,
+      "step": 1720
+    },
+    {
+      "epoch": 0.06286794098408315,
+      "grad_norm": 11.539891242980957,
+      "learning_rate": 6.264990188240424e-06,
+      "loss": 0.4628,
+      "step": 1730
+    },
+    {
+      "epoch": 0.06323133948688131,
+      "grad_norm": 3.275383710861206,
+      "learning_rate": 6.301330038520243e-06,
+      "loss": 0.3368,
+      "step": 1740
+    },
+    {
+      "epoch": 0.06359473798967948,
+      "grad_norm": 41.4942741394043,
+      "learning_rate": 6.3376698888000586e-06,
+      "loss": 0.53,
+      "step": 1750
+    },
+    {
+      "epoch": 0.06395813649247765,
+      "grad_norm": 3.0071399211883545,
+      "learning_rate": 6.374009739079875e-06,
+      "loss": 0.3623,
+      "step": 1760
+    },
+    {
+      "epoch": 0.06432153499527582,
+      "grad_norm": 3.385955333709717,
+      "learning_rate": 6.410349589359692e-06,
+      "loss": 0.3476,
+      "step": 1770
+    },
+    {
+      "epoch": 0.06468493349807398,
+      "grad_norm": 3.872527599334717,
+      "learning_rate": 6.446689439639508e-06,
+      "loss": 0.3168,
+      "step": 1780
+    },
+    {
+      "epoch": 0.06504833200087215,
+      "grad_norm": 4.668768882751465,
+      "learning_rate": 6.483029289919327e-06,
+      "loss": 0.3813,
+      "step": 1790
+    },
+    {
+      "epoch": 0.06541173050367033,
+      "grad_norm": 69.33656311035156,
+      "learning_rate": 6.519369140199143e-06,
+      "loss": 0.4557,
+      "step": 1800
+    },
+    {
+      "epoch": 0.06541173050367033,
+      "eval_loss": 0.5579342246055603,
+      "eval_runtime": 180.31,
+      "eval_samples_per_second": 41.118,
+      "eval_steps_per_second": 5.141,
+      "eval_wer": 0.34055221740156477,
+      "step": 1800
+    },
+    {
+      "epoch": 0.0657751290064685,
+      "grad_norm": 4.2848381996154785,
+      "learning_rate": 6.555708990478959e-06,
+      "loss": 0.5481,
+      "step": 1810
+    },
+    {
+      "epoch": 0.06613852750926666,
+      "grad_norm": 11.31700325012207,
+      "learning_rate": 6.592048840758775e-06,
+      "loss": 0.3479,
+      "step": 1820
+    },
+    {
+      "epoch": 0.06650192601206482,
+      "grad_norm": 6.088991165161133,
+      "learning_rate": 6.628388691038594e-06,
+      "loss": 0.3994,
+      "step": 1830
+    },
+    {
+      "epoch": 0.066865324514863,
+      "grad_norm": 4.342681407928467,
+      "learning_rate": 6.66472854131841e-06,
+      "loss": 0.2953,
+      "step": 1840
+    },
+    {
+      "epoch": 0.06722872301766117,
+      "grad_norm": 75.97467041015625,
+      "learning_rate": 6.701068391598227e-06,
+      "loss": 0.4807,
+      "step": 1850
+    },
+    {
+      "epoch": 0.06759212152045933,
+      "grad_norm": 3.8739049434661865,
+      "learning_rate": 6.7374082418780435e-06,
+      "loss": 0.3963,
+      "step": 1860
+    },
+    {
+      "epoch": 0.0679555200232575,
+      "grad_norm": 2.6209168434143066,
+      "learning_rate": 6.773748092157861e-06,
+      "loss": 0.422,
+      "step": 1870
+    },
+    {
+      "epoch": 0.06831891852605568,
+      "grad_norm": 17.530773162841797,
+      "learning_rate": 6.810087942437678e-06,
+      "loss": 0.3939,
+      "step": 1880
+    },
+    {
+      "epoch": 0.06868231702885384,
+      "grad_norm": 3.475748300552368,
+      "learning_rate": 6.846427792717494e-06,
+      "loss": 0.2996,
+      "step": 1890
+    },
+    {
+      "epoch": 0.06904571553165201,
+      "grad_norm": 20.979995727539062,
+      "learning_rate": 6.882767642997311e-06,
+      "loss": 0.4528,
+      "step": 1900
+    },
+    {
+      "epoch": 0.06940911403445017,
+      "grad_norm": 3.8432774543762207,
+      "learning_rate": 6.919107493277128e-06,
+      "loss": 0.2959,
+      "step": 1910
+    },
+    {
+      "epoch": 0.06977251253724835,
+      "grad_norm": 7.830467700958252,
+      "learning_rate": 6.955447343556945e-06,
+      "loss": 0.3378,
+      "step": 1920
+    },
+    {
+      "epoch": 0.07013591104004652,
+      "grad_norm": 15.633039474487305,
+      "learning_rate": 6.991787193836762e-06,
+      "loss": 0.4,
+      "step": 1930
+    },
+    {
+      "epoch": 0.07049930954284468,
+      "grad_norm": 13.628314971923828,
+      "learning_rate": 7.028127044116578e-06,
+      "loss": 0.3255,
+      "step": 1940
+    },
+    {
+      "epoch": 0.07086270804564285,
+      "grad_norm": 33.001773834228516,
+      "learning_rate": 7.064466894396396e-06,
+      "loss": 0.4367,
+      "step": 1950
+    },
+    {
+      "epoch": 0.07122610654844103,
+      "grad_norm": 3.5115041732788086,
+      "learning_rate": 7.100806744676212e-06,
+      "loss": 0.3279,
+      "step": 1960
+    },
+    {
+      "epoch": 0.07158950505123919,
+      "grad_norm": 3.0497541427612305,
+      "learning_rate": 7.137146594956029e-06,
+      "loss": 2.8797,
+      "step": 1970
+    },
+    {
+      "epoch": 0.07195290355403736,
+      "grad_norm": 6.17769718170166,
+      "learning_rate": 7.173486445235845e-06,
+      "loss": 0.3534,
+      "step": 1980
+    },
+    {
+      "epoch": 0.07231630205683552,
+      "grad_norm": 5.4114789962768555,
+      "learning_rate": 7.209826295515664e-06,
+      "loss": 0.3309,
+      "step": 1990
+    },
+    {
+      "epoch": 0.0726797005596337,
+      "grad_norm": 11.600439071655273,
+      "learning_rate": 7.24616614579548e-06,
+      "loss": 0.4382,
+      "step": 2000
+    },
+    {
+      "epoch": 0.07304309906243187,
+      "grad_norm": 3.4476027488708496,
+      "learning_rate": 7.2825059960752965e-06,
+      "loss": 0.3487,
+      "step": 2010
+    },
+    {
+      "epoch": 0.07340649756523003,
+      "grad_norm": 5.642564296722412,
+      "learning_rate": 7.318845846355113e-06,
+      "loss": 0.3513,
+      "step": 2020
+    },
+    {
+      "epoch": 0.0737698960680282,
+      "grad_norm": 7.132052898406982,
+      "learning_rate": 7.355185696634931e-06,
+      "loss": 0.4564,
+      "step": 2030
+    },
+    {
+      "epoch": 0.07413329457082637,
+      "grad_norm": 6.583246231079102,
+      "learning_rate": 7.391525546914748e-06,
+      "loss": 0.3376,
+      "step": 2040
+    },
+    {
+      "epoch": 0.07449669307362454,
+      "grad_norm": 23.98805809020996,
+      "learning_rate": 7.427865397194564e-06,
+      "loss": 0.4217,
+      "step": 2050
+    },
+    {
+      "epoch": 0.0748600915764227,
+      "grad_norm": 3.9135584831237793,
+      "learning_rate": 7.4642052474743805e-06,
+      "loss": 0.3324,
+      "step": 2060
+    },
+    {
+      "epoch": 0.07522349007922087,
+      "grad_norm": 3.4022698402404785,
+      "learning_rate": 7.500545097754198e-06,
+      "loss": 0.3391,
+      "step": 2070
+    },
+    {
+      "epoch": 0.07558688858201905,
+      "grad_norm": 8.37547779083252,
+      "learning_rate": 7.536884948034015e-06,
+      "loss": 0.3119,
+      "step": 2080
+    },
+    {
+      "epoch": 0.07595028708481721,
+      "grad_norm": 6.2167558670043945,
+      "learning_rate": 7.573224798313831e-06,
+      "loss": 0.3247,
+      "step": 2090
+    },
+    {
+      "epoch": 0.07631368558761538,
+      "grad_norm": 81.76036834716797,
+      "learning_rate": 7.609564648593648e-06,
+      "loss": 0.4281,
+      "step": 2100
+    },
+    {
+      "epoch": 0.07667708409041354,
+      "grad_norm": 2.8961973190307617,
+      "learning_rate": 7.645904498873465e-06,
+      "loss": 0.3368,
+      "step": 2110
+    },
+    {
+      "epoch": 0.07704048259321171,
+      "grad_norm": 4.699477195739746,
+      "learning_rate": 7.682244349153282e-06,
+      "loss": 0.3403,
+      "step": 2120
+    },
+    {
+      "epoch": 0.07740388109600989,
+      "grad_norm": 4.429138660430908,
+      "learning_rate": 7.718584199433098e-06,
+      "loss": 0.3182,
+      "step": 2130
+    },
+    {
+      "epoch": 0.07776727959880805,
+      "grad_norm": 2.7269580364227295,
+      "learning_rate": 7.754924049712916e-06,
+      "loss": 0.2828,
+      "step": 2140
+    },
+    {
+      "epoch": 0.07813067810160622,
+      "grad_norm": 15.126232147216797,
+      "learning_rate": 7.791263899992732e-06,
+      "loss": 0.4606,
+      "step": 2150
+    },
+    {
+      "epoch": 0.07849407660440438,
+      "grad_norm": 10.14072322845459,
+      "learning_rate": 7.82760375027255e-06,
+      "loss": 0.3451,
+      "step": 2160
+    },
+    {
+      "epoch": 0.07885747510720256,
+      "grad_norm": 4.95914363861084,
+      "learning_rate": 7.863943600552365e-06,
+      "loss": 0.3612,
+      "step": 2170
+    },
+    {
+      "epoch": 0.07922087361000073,
+      "grad_norm": 4.115192413330078,
+      "learning_rate": 7.900283450832183e-06,
+      "loss": 0.3222,
+      "step": 2180
+    },
+    {
+      "epoch": 0.07958427211279889,
+      "grad_norm": 5.405594825744629,
+      "learning_rate": 7.936623301111999e-06,
+      "loss": 0.3474,
+      "step": 2190
+    },
+    {
+      "epoch": 0.07994767061559706,
+      "grad_norm": 23.328718185424805,
+      "learning_rate": 7.972963151391817e-06,
+      "loss": 0.4797,
+      "step": 2200
+    },
+    {
+      "epoch": 0.08031106911839524,
+      "grad_norm": 3.5595099925994873,
+      "learning_rate": 8.009303001671634e-06,
+      "loss": 0.3305,
+      "step": 2210
+    },
+    {
+      "epoch": 0.0806744676211934,
+      "grad_norm": 3.048445463180542,
+      "learning_rate": 8.04564285195145e-06,
+      "loss": 0.318,
+      "step": 2220
+    },
+    {
+      "epoch": 0.08103786612399157,
+      "grad_norm": 5.857702732086182,
+      "learning_rate": 8.081982702231266e-06,
+      "loss": 0.3497,
+      "step": 2230
+    },
+    {
+      "epoch": 0.08140126462678973,
+      "grad_norm": 3.0092968940734863,
+      "learning_rate": 8.118322552511084e-06,
+      "loss": 0.2995,
+      "step": 2240
+    },
+    {
+      "epoch": 0.08176466312958791,
+      "grad_norm": 9.337843894958496,
+      "learning_rate": 8.154662402790902e-06,
+      "loss": 0.4517,
+      "step": 2250
+    },
+    {
+      "epoch": 0.08212806163238608,
+      "grad_norm": 3.136950969696045,
+      "learning_rate": 8.191002253070718e-06,
+      "loss": 0.2927,
+      "step": 2260
+    },
+    {
+      "epoch": 0.08249146013518424,
+      "grad_norm": 4.228198051452637,
+      "learning_rate": 8.227342103350534e-06,
+      "loss": 1.2185,
+      "step": 2270
+    },
+    {
+      "epoch": 0.0828548586379824,
+      "grad_norm": 7.404679298400879,
+      "learning_rate": 8.263681953630351e-06,
+      "loss": 0.3448,
+      "step": 2280
+    },
+    {
+      "epoch": 0.08321825714078059,
+      "grad_norm": 7.873497009277344,
+      "learning_rate": 8.300021803910169e-06,
+      "loss": 0.2965,
+      "step": 2290
+    },
+    {
+      "epoch": 0.08358165564357875,
+      "grad_norm": 12.266081809997559,
+      "learning_rate": 8.336361654189985e-06,
+      "loss": 0.4631,
+      "step": 2300
+    },
+    {
+      "epoch": 0.08394505414637692,
+      "grad_norm": 3.3576557636260986,
+      "learning_rate": 8.3727015044698e-06,
+      "loss": 0.3339,
+      "step": 2310
+    },
+    {
+      "epoch": 0.08430845264917508,
+      "grad_norm": 3.0854902267456055,
+      "learning_rate": 8.40904135474962e-06,
+      "loss": 0.3448,
+      "step": 2320
+    },
+    {
+      "epoch": 0.08467185115197326,
+      "grad_norm": 6.1308746337890625,
+      "learning_rate": 8.445381205029436e-06,
+      "loss": 0.386,
+      "step": 2330
+    },
+    {
+      "epoch": 0.08503524965477142,
+      "grad_norm": 4.458275318145752,
+      "learning_rate": 8.481721055309252e-06,
+      "loss": 0.2916,
+      "step": 2340
+    },
+    {
+      "epoch": 0.08539864815756959,
+      "grad_norm": 25.443647384643555,
+      "learning_rate": 8.51806090558907e-06,
+      "loss": 0.4232,
+      "step": 2350
+    },
+    {
+      "epoch": 0.08576204666036776,
+      "grad_norm": 324.4353332519531,
+      "learning_rate": 8.554400755868887e-06,
+      "loss": 2.4995,
+      "step": 2360
+    },
+    {
+      "epoch": 0.08612544516316593,
+      "grad_norm": 17.593692779541016,
+      "learning_rate": 8.590740606148703e-06,
+      "loss": 0.2952,
+      "step": 2370
+    },
+    {
+      "epoch": 0.0864888436659641,
+      "grad_norm": 3.4646732807159424,
+      "learning_rate": 8.62708045642852e-06,
+      "loss": 0.2961,
+      "step": 2380
+    },
+    {
+      "epoch": 0.08685224216876226,
+      "grad_norm": 2.9895999431610107,
+      "learning_rate": 8.663420306708337e-06,
+      "loss": 0.2852,
+      "step": 2390
+    },
+    {
+      "epoch": 0.08721564067156043,
+      "grad_norm": 24.221176147460938,
+      "learning_rate": 8.699760156988155e-06,
+      "loss": 0.4343,
+      "step": 2400
+    },
+    {
+      "epoch": 0.08721564067156043,
+      "eval_loss": 0.47036415338516235,
+      "eval_runtime": 180.1154,
+      "eval_samples_per_second": 41.162,
+      "eval_steps_per_second": 5.147,
+      "eval_wer": 0.28054713453264835,
+      "step": 2400
+    },
+    {
+      "epoch": 0.08757903917435861,
+      "grad_norm": 161.69967651367188,
+      "learning_rate": 8.73610000726797e-06,
+      "loss": 1.4598,
+      "step": 2410
+    },
+    {
+      "epoch": 0.08794243767715677,
+      "grad_norm": 10.37559700012207,
+      "learning_rate": 8.772439857547786e-06,
+      "loss": 0.3042,
+      "step": 2420
+    },
+    {
+      "epoch": 0.08830583617995494,
+      "grad_norm": 5.90106725692749,
+      "learning_rate": 8.808779707827604e-06,
+      "loss": 0.3385,
+      "step": 2430
+    },
+    {
+      "epoch": 0.0886692346827531,
+      "grad_norm": 9.207955360412598,
+      "learning_rate": 8.845119558107422e-06,
+      "loss": 0.2963,
+      "step": 2440
+    },
+    {
+      "epoch": 0.08903263318555127,
+      "grad_norm": 22.280956268310547,
+      "learning_rate": 8.881459408387238e-06,
+      "loss": 0.4505,
+      "step": 2450
+    },
+    {
+      "epoch": 0.08939603168834945,
+      "grad_norm": 3.090710401535034,
+      "learning_rate": 8.917799258667055e-06,
+      "loss": 0.3114,
+      "step": 2460
+    },
+    {
+      "epoch": 0.08975943019114761,
+      "grad_norm": 4.144134044647217,
+      "learning_rate": 8.954139108946871e-06,
+      "loss": 0.2855,
+      "step": 2470
+    },
+    {
+      "epoch": 0.09012282869394578,
+      "grad_norm": 4.343112468719482,
+      "learning_rate": 8.990478959226687e-06,
+      "loss": 0.2906,
+      "step": 2480
+    },
+    {
+      "epoch": 0.09048622719674394,
+      "grad_norm": 2.6925292015075684,
+      "learning_rate": 9.026818809506505e-06,
+      "loss": 0.284,
+      "step": 2490
+    },
+    {
+      "epoch": 0.09084962569954212,
+      "grad_norm": 29.639341354370117,
+      "learning_rate": 9.063158659786323e-06,
+      "loss": 0.3411,
+      "step": 2500
+    },
+    {
+      "epoch": 0.09121302420234029,
+      "grad_norm": 4.425374984741211,
+      "learning_rate": 9.099498510066139e-06,
+      "loss": 0.3041,
+      "step": 2510
+    },
+    {
+      "epoch": 0.09157642270513845,
+      "grad_norm": 5.6643195152282715,
+      "learning_rate": 9.135838360345955e-06,
+      "loss": 0.3123,
+      "step": 2520
+    },
+    {
+      "epoch": 0.09193982120793662,
+      "grad_norm": 3.9098479747772217,
+      "learning_rate": 9.172178210625772e-06,
+      "loss": 0.3664,
+      "step": 2530
+    },
+    {
+      "epoch": 0.0923032197107348,
+      "grad_norm": 3.133389949798584,
+      "learning_rate": 9.20851806090559e-06,
+      "loss": 0.2708,
+      "step": 2540
+    },
+    {
+      "epoch": 0.09266661821353296,
+      "grad_norm": 43.00468063354492,
+      "learning_rate": 9.244857911185406e-06,
+      "loss": 0.4215,
+      "step": 2550
+    },
+    {
+      "epoch": 0.09303001671633113,
+      "grad_norm": 3.1411876678466797,
+      "learning_rate": 9.281197761465222e-06,
+      "loss": 0.2983,
+      "step": 2560
+    },
+    {
+      "epoch": 0.09339341521912929,
+      "grad_norm": 3.263828754425049,
+      "learning_rate": 9.317537611745041e-06,
+      "loss": 0.3137,
+      "step": 2570
+    },
+    {
+      "epoch": 0.09375681372192747,
+      "grad_norm": 3.618751049041748,
+      "learning_rate": 9.353877462024857e-06,
+      "loss": 0.3279,
+      "step": 2580
+    },
+    {
+      "epoch": 0.09412021222472564,
+      "grad_norm": 3.6551568508148193,
+      "learning_rate": 9.390217312304673e-06,
+      "loss": 0.2409,
+      "step": 2590
+    },
+    {
+      "epoch": 0.0944836107275238,
+      "grad_norm": 8.680901527404785,
+      "learning_rate": 9.42655716258449e-06,
+      "loss": 0.373,
+      "step": 2600
+    },
+    {
+      "epoch": 0.09484700923032197,
+      "grad_norm": 4.761026382446289,
+      "learning_rate": 9.462897012864308e-06,
+      "loss": 0.2777,
+      "step": 2610
+    },
+    {
+      "epoch": 0.09521040773312014,
+      "grad_norm": 3.142723321914673,
+      "learning_rate": 9.499236863144124e-06,
+      "loss": 0.2882,
+      "step": 2620
+    },
+    {
+      "epoch": 0.09557380623591831,
+      "grad_norm": 2.969968795776367,
+      "learning_rate": 9.53557671342394e-06,
+      "loss": 0.3086,
+      "step": 2630
+    },
+    {
+      "epoch": 0.09593720473871648,
+      "grad_norm": 3.754549264907837,
+      "learning_rate": 9.571916563703758e-06,
+      "loss": 0.259,
+      "step": 2640
+    },
+    {
+      "epoch": 0.09630060324151464,
+      "grad_norm": 23.7288761138916,
+      "learning_rate": 9.608256413983576e-06,
+      "loss": 0.4284,
+      "step": 2650
+    },
+    {
+      "epoch": 0.09666400174431282,
+      "grad_norm": 2.7727372646331787,
+      "learning_rate": 9.644596264263392e-06,
+      "loss": 0.2602,
+      "step": 2660
+    },
+    {
+      "epoch": 0.09702740024711098,
+      "grad_norm": 14.707064628601074,
+      "learning_rate": 9.680936114543208e-06,
+      "loss": 0.3059,
+      "step": 2670
+    },
+    {
+      "epoch": 0.09739079874990915,
+      "grad_norm": 3.8396642208099365,
+      "learning_rate": 9.717275964823025e-06,
+      "loss": 0.2811,
+      "step": 2680
+    },
+    {
+      "epoch": 0.09775419725270731,
+      "grad_norm": 2.9460713863372803,
+      "learning_rate": 9.753615815102843e-06,
+      "loss": 0.2686,
+      "step": 2690
+    },
+    {
+      "epoch": 0.0981175957555055,
+      "grad_norm": 20.107336044311523,
+      "learning_rate": 9.789955665382659e-06,
+      "loss": 0.4306,
+      "step": 2700
+    },
+    {
+      "epoch": 0.09848099425830366,
+      "grad_norm": 3.1286280155181885,
+      "learning_rate": 9.826295515662477e-06,
+      "loss": 0.3059,
+      "step": 2710
+    },
+    {
+      "epoch": 0.09884439276110182,
+      "grad_norm": 6.160215854644775,
+      "learning_rate": 9.862635365942292e-06,
+      "loss": 0.3046,
+      "step": 2720
+    },
+    {
+      "epoch": 0.09920779126389999,
+      "grad_norm": 6.1921186447143555,
+      "learning_rate": 9.89897521622211e-06,
+      "loss": 0.285,
+      "step": 2730
+    },
+    {
+      "epoch": 0.09957118976669817,
+      "grad_norm": 13.759759902954102,
+      "learning_rate": 9.935315066501926e-06,
+      "loss": 0.2888,
+      "step": 2740
+    },
+    {
+      "epoch": 0.09993458826949633,
+      "grad_norm": 13.92764949798584,
+      "learning_rate": 9.971654916781744e-06,
+      "loss": 0.4266,
+      "step": 2750
+    },
+    {
+      "epoch": 0.1002979867722945,
+      "grad_norm": 3.3999857902526855,
+      "learning_rate": 1.000799476706156e-05,
+      "loss": 0.2858,
+      "step": 2760
+    },
+    {
+      "epoch": 0.10066138527509266,
+      "grad_norm": 4.103928089141846,
+      "learning_rate": 1.0044334617341377e-05,
+      "loss": 0.262,
+      "step": 2770
+    },
+    {
+      "epoch": 0.10102478377789084,
+      "grad_norm": 6.15985107421875,
+      "learning_rate": 1.0080674467621195e-05,
+      "loss": 0.2866,
+      "step": 2780
+    },
+    {
+      "epoch": 0.10138818228068901,
+      "grad_norm": 4.904097557067871,
+      "learning_rate": 1.0117014317901011e-05,
+      "loss": 0.5057,
+      "step": 2790
+    },
+    {
+      "epoch": 0.10175158078348717,
+      "grad_norm": 15.2875337600708,
+      "learning_rate": 1.0153354168180827e-05,
+      "loss": 0.4345,
+      "step": 2800
+    },
+    {
+      "epoch": 0.10211497928628534,
+      "grad_norm": 2.4697763919830322,
+      "learning_rate": 1.0189694018460643e-05,
+      "loss": 0.2693,
+      "step": 2810
+    },
+    {
+      "epoch": 0.1024783777890835,
+      "grad_norm": 5.04618501663208,
+      "learning_rate": 1.0226033868740462e-05,
+      "loss": 0.2868,
+      "step": 2820
+    },
+    {
+      "epoch": 0.10284177629188168,
+      "grad_norm": 5.851120948791504,
+      "learning_rate": 1.0262373719020278e-05,
+      "loss": 0.3425,
+      "step": 2830
+    },
+    {
+      "epoch": 0.10320517479467985,
+      "grad_norm": 2.1007258892059326,
+      "learning_rate": 1.0298713569300094e-05,
+      "loss": 0.2394,
+      "step": 2840
+    },
+    {
+      "epoch": 0.10356857329747801,
+      "grad_norm": 23.411701202392578,
+      "learning_rate": 1.0335053419579912e-05,
+      "loss": 0.4125,
+      "step": 2850
+    },
+    {
+      "epoch": 0.10393197180027618,
+      "grad_norm": 4.178852558135986,
+      "learning_rate": 1.037139326985973e-05,
+      "loss": 0.2951,
+      "step": 2860
+    },
+    {
+      "epoch": 0.10429537030307436,
+      "grad_norm": 1.7873708009719849,
+      "learning_rate": 1.0407733120139545e-05,
+      "loss": 0.3272,
+      "step": 2870
+    },
+    {
+      "epoch": 0.10465876880587252,
+      "grad_norm": 7.603367328643799,
+      "learning_rate": 1.0444072970419361e-05,
+      "loss": 0.2779,
+      "step": 2880
+    },
+    {
+      "epoch": 0.10502216730867069,
+      "grad_norm": 3.468761444091797,
+      "learning_rate": 1.0480412820699179e-05,
+      "loss": 0.3007,
+      "step": 2890
+    },
+    {
+      "epoch": 0.10538556581146885,
+      "grad_norm": 16.35407829284668,
+      "learning_rate": 1.0516752670978997e-05,
+      "loss": 0.3918,
+      "step": 2900
+    },
+    {
+      "epoch": 0.10574896431426703,
+      "grad_norm": 3.4226725101470947,
+      "learning_rate": 1.0553092521258813e-05,
+      "loss": 3.7156,
+      "step": 2910
+    },
+    {
+      "epoch": 0.1061123628170652,
+      "grad_norm": 9.006295204162598,
+      "learning_rate": 1.058943237153863e-05,
+      "loss": 0.4075,
+      "step": 2920
+    },
+    {
+      "epoch": 0.10647576131986336,
+      "grad_norm": 4.993385314941406,
+      "learning_rate": 1.0625772221818446e-05,
+      "loss": 0.3588,
+      "step": 2930
+    },
+    {
+      "epoch": 0.10683915982266153,
+      "grad_norm": 3.7684736251831055,
+      "learning_rate": 1.0662112072098264e-05,
+      "loss": 0.2429,
+      "step": 2940
+    },
+    {
+      "epoch": 0.1072025583254597,
+      "grad_norm": 40.301170349121094,
+      "learning_rate": 1.069845192237808e-05,
+      "loss": 0.4739,
+      "step": 2950
+    },
+    {
+      "epoch": 0.10756595682825787,
+      "grad_norm": 3.772693157196045,
+      "learning_rate": 1.0734791772657898e-05,
+      "loss": 0.3284,
+      "step": 2960
+    },
+    {
+      "epoch": 0.10792935533105603,
+      "grad_norm": 3.0183212757110596,
+      "learning_rate": 1.0771131622937714e-05,
+      "loss": 0.38,
+      "step": 2970
+    },
+    {
+      "epoch": 0.1082927538338542,
+      "grad_norm": 6.61776876449585,
+      "learning_rate": 1.0807471473217531e-05,
+      "loss": 0.2793,
+      "step": 2980
+    },
+    {
+      "epoch": 0.10865615233665238,
+      "grad_norm": 6.112472057342529,
+      "learning_rate": 1.0843811323497347e-05,
+      "loss": 0.2447,
+      "step": 2990
+    },
+    {
+      "epoch": 0.10901955083945054,
+      "grad_norm": 10.800559997558594,
+      "learning_rate": 1.0880151173777165e-05,
+      "loss": 0.373,
+      "step": 3000
+    },
+    {
+      "epoch": 0.10901955083945054,
+      "eval_loss": 0.4652940630912781,
+      "eval_runtime": 180.0765,
+      "eval_samples_per_second": 41.171,
+      "eval_steps_per_second": 5.148,
+      "eval_wer": 0.27681667181004593,
+      "step": 3000
+    },
+    {
+      "epoch": 0.10938294934224871,
+      "grad_norm": 7.778831958770752,
+      "learning_rate": 1.091649102405698e-05,
+      "loss": 0.29,
+      "step": 3010
+    },
+    {
+      "epoch": 0.10974634784504687,
+      "grad_norm": 2.855592966079712,
+      "learning_rate": 1.0952830874336798e-05,
+      "loss": 0.2411,
+      "step": 3020
+    },
+    {
+      "epoch": 0.11010974634784505,
+      "grad_norm": 4.229335784912109,
+      "learning_rate": 1.0989170724616616e-05,
+      "loss": 0.3247,
+      "step": 3030
+    },
+    {
+      "epoch": 0.11047314485064322,
+      "grad_norm": 3.8145949840545654,
+      "learning_rate": 1.1025510574896432e-05,
+      "loss": 0.2242,
+      "step": 3040
+    },
+    {
+      "epoch": 0.11083654335344138,
+      "grad_norm": 22.571304321289062,
+      "learning_rate": 1.1061850425176248e-05,
+      "loss": 0.3959,
+      "step": 3050
+    },
+    {
+      "epoch": 0.11119994185623955,
+      "grad_norm": 2.4706461429595947,
+      "learning_rate": 1.1098190275456066e-05,
+      "loss": 0.2466,
+      "step": 3060
+    },
+    {
+      "epoch": 0.11156334035903773,
+      "grad_norm": 4.497069358825684,
+      "learning_rate": 1.1134530125735883e-05,
+      "loss": 2.1968,
+      "step": 3070
+    },
+    {
+      "epoch": 0.11192673886183589,
+      "grad_norm": 5.060062885284424,
+      "learning_rate": 1.11708699760157e-05,
+      "loss": 0.2921,
+      "step": 3080
+    },
+    {
+      "epoch": 0.11229013736463406,
+      "grad_norm": 2.7882325649261475,
+      "learning_rate": 1.1207209826295515e-05,
+      "loss": 0.2534,
+      "step": 3090
+    },
+    {
+      "epoch": 0.11265353586743222,
+      "grad_norm": 9.96241569519043,
+      "learning_rate": 1.1243549676575333e-05,
+      "loss": 0.421,
+      "step": 3100
+    },
+    {
+      "epoch": 0.1130169343702304,
+      "grad_norm": 31.262916564941406,
+      "learning_rate": 1.127988952685515e-05,
+      "loss": 0.4048,
+      "step": 3110
+    },
+    {
+      "epoch": 0.11338033287302857,
+      "grad_norm": 3.472343921661377,
+      "learning_rate": 1.1316229377134967e-05,
+      "loss": 0.2798,
+      "step": 3120
+    },
+    {
+      "epoch": 0.11374373137582673,
+      "grad_norm": 4.074085235595703,
+      "learning_rate": 1.1352569227414783e-05,
+      "loss": 0.299,
+      "step": 3130
+    },
+    {
+      "epoch": 0.1141071298786249,
+      "grad_norm": 2.879512310028076,
+      "learning_rate": 1.1388909077694602e-05,
+      "loss": 0.2137,
+      "step": 3140
+    },
+    {
+      "epoch": 0.11447052838142306,
+      "grad_norm": 125.17889404296875,
+      "learning_rate": 1.1425248927974418e-05,
+      "loss": 0.5418,
+      "step": 3150
+    },
+    {
+      "epoch": 0.11483392688422124,
+      "grad_norm": 4.171487808227539,
+      "learning_rate": 1.1461588778254234e-05,
+      "loss": 0.2685,
+      "step": 3160
+    },
+    {
+      "epoch": 0.1151973253870194,
+      "grad_norm": 2.1496529579162598,
+      "learning_rate": 1.1497928628534051e-05,
+      "loss": 0.2421,
+      "step": 3170
+    },
+    {
+      "epoch": 0.11556072388981757,
+      "grad_norm": 2.6266047954559326,
+      "learning_rate": 1.1534268478813867e-05,
+      "loss": 0.3288,
+      "step": 3180
+    },
+    {
+      "epoch": 0.11592412239261574,
+      "grad_norm": 3.7677230834960938,
+      "learning_rate": 1.1570608329093685e-05,
+      "loss": 0.3093,
+      "step": 3190
+    },
+    {
+      "epoch": 0.11628752089541392,
+      "grad_norm": 9.4945707321167,
+      "learning_rate": 1.1606948179373501e-05,
+      "loss": 0.3066,
+      "step": 3200
+    },
+    {
+      "epoch": 0.11665091939821208,
+      "grad_norm": 2.5509915351867676,
+      "learning_rate": 1.1643288029653319e-05,
+      "loss": 0.2615,
+      "step": 3210
+    },
+    {
+      "epoch": 0.11701431790101025,
+      "grad_norm": 3.066624641418457,
+      "learning_rate": 1.1679627879933135e-05,
+      "loss": 0.3224,
+      "step": 3220
+    },
+    {
+      "epoch": 0.11737771640380841,
+      "grad_norm": 6.494440078735352,
+      "learning_rate": 1.1715967730212952e-05,
+      "loss": 0.3017,
+      "step": 3230
+    },
+    {
+      "epoch": 0.11774111490660659,
+      "grad_norm": 3.4675605297088623,
+      "learning_rate": 1.1752307580492768e-05,
+      "loss": 0.2152,
+      "step": 3240
+    },
+    {
+      "epoch": 0.11810451340940475,
+      "grad_norm": 15.5110445022583,
+      "learning_rate": 1.1788647430772586e-05,
+      "loss": 0.349,
+      "step": 3250
+    },
+    {
+      "epoch": 0.11846791191220292,
+      "grad_norm": 1.972530484199524,
+      "learning_rate": 1.1824987281052402e-05,
+      "loss": 0.2728,
+      "step": 3260
+    },
+    {
+      "epoch": 0.11883131041500108,
+      "grad_norm": 4.018677711486816,
+      "learning_rate": 1.186132713133222e-05,
+      "loss": 0.254,
+      "step": 3270
+    },
+    {
+      "epoch": 0.11919470891779926,
+      "grad_norm": 4.95416784286499,
+      "learning_rate": 1.1897666981612037e-05,
+      "loss": 0.2465,
+      "step": 3280
+    },
+    {
+      "epoch": 0.11955810742059743,
+      "grad_norm": 3.165599822998047,
+      "learning_rate": 1.1934006831891853e-05,
+      "loss": 0.2537,
+      "step": 3290
+    },
+    {
+      "epoch": 0.1199215059233956,
+      "grad_norm": 8.508636474609375,
+      "learning_rate": 1.1970346682171669e-05,
+      "loss": 0.3655,
+      "step": 3300
+    },
+    {
+      "epoch": 0.12028490442619376,
+      "grad_norm": 2.3892879486083984,
+      "learning_rate": 1.2006686532451487e-05,
+      "loss": 0.252,
+      "step": 3310
+    },
+    {
+      "epoch": 0.12064830292899194,
+      "grad_norm": 3.591564178466797,
+      "learning_rate": 1.2043026382731304e-05,
+      "loss": 0.2401,
+      "step": 3320
+    },
+    {
+      "epoch": 0.1210117014317901,
+      "grad_norm": 3.891261577606201,
+      "learning_rate": 1.207936623301112e-05,
+      "loss": 0.2909,
+      "step": 3330
+    },
+    {
+      "epoch": 0.12137509993458827,
+      "grad_norm": 4.691511154174805,
+      "learning_rate": 1.2115706083290936e-05,
+      "loss": 0.2304,
+      "step": 3340
+    },
+    {
+      "epoch": 0.12173849843738643,
+      "grad_norm": 18.415170669555664,
+      "learning_rate": 1.2152045933570754e-05,
+      "loss": 0.35,
+      "step": 3350
+    },
+    {
+      "epoch": 0.12210189694018461,
+      "grad_norm": 3.9105615615844727,
+      "learning_rate": 1.2188385783850572e-05,
+      "loss": 0.3112,
+      "step": 3360
+    },
+    {
+      "epoch": 0.12246529544298278,
+      "grad_norm": 3.215313196182251,
+      "learning_rate": 1.2224725634130388e-05,
+      "loss": 0.2492,
+      "step": 3370
+    },
+    {
+      "epoch": 0.12282869394578094,
+      "grad_norm": 9.30749225616455,
+      "learning_rate": 1.2261065484410204e-05,
+      "loss": 0.2696,
+      "step": 3380
+    },
+    {
+      "epoch": 0.12319209244857911,
+      "grad_norm": 4.9797682762146,
+      "learning_rate": 1.2297405334690023e-05,
+      "loss": 0.2197,
+      "step": 3390
+    },
+    {
+      "epoch": 0.12355549095137729,
+      "grad_norm": 19.632797241210938,
+      "learning_rate": 1.2333745184969839e-05,
+      "loss": 0.3411,
+      "step": 3400
+    },
+    {
+      "epoch": 0.12391888945417545,
+      "grad_norm": 4.509830474853516,
+      "learning_rate": 1.2370085035249655e-05,
+      "loss": 0.2394,
+      "step": 3410
+    },
+    {
+      "epoch": 0.12428228795697362,
+      "grad_norm": 2.253514051437378,
+      "learning_rate": 1.2406424885529473e-05,
+      "loss": 0.266,
+      "step": 3420
+    },
+    {
+      "epoch": 0.12464568645977178,
+      "grad_norm": 3.123828172683716,
+      "learning_rate": 1.244276473580929e-05,
+      "loss": 0.2675,
+      "step": 3430
+    },
+    {
+      "epoch": 0.12500908496256996,
+      "grad_norm": 34.37680435180664,
+      "learning_rate": 1.2479104586089106e-05,
+      "loss": 0.279,
+      "step": 3440
+    },
+    {
+      "epoch": 0.1253724834653681,
+      "grad_norm": 10.051690101623535,
+      "learning_rate": 1.2515444436368922e-05,
+      "loss": 0.3364,
+      "step": 3450
+    },
+    {
+      "epoch": 0.1257358819681663,
+      "grad_norm": 2.1765711307525635,
+      "learning_rate": 1.255178428664874e-05,
+      "loss": 0.2288,
+      "step": 3460
+    },
+    {
+      "epoch": 0.12609928047096447,
+      "grad_norm": 2.4910778999328613,
+      "learning_rate": 1.2588124136928556e-05,
+      "loss": 0.2866,
+      "step": 3470
+    },
+    {
+      "epoch": 0.12646267897376262,
+      "grad_norm": 7.379613876342773,
+      "learning_rate": 1.2624463987208373e-05,
+      "loss": 0.2618,
+      "step": 3480
+    },
+    {
+      "epoch": 0.1268260774765608,
+      "grad_norm": 2.681814432144165,
+      "learning_rate": 1.266080383748819e-05,
+      "loss": 0.2405,
+      "step": 3490
+    },
+    {
+      "epoch": 0.12718947597935895,
+      "grad_norm": 83.93474578857422,
+      "learning_rate": 1.2697143687768007e-05,
+      "loss": 0.338,
+      "step": 3500
+    },
+    {
+      "epoch": 0.12755287448215713,
+      "grad_norm": 1.5564826726913452,
+      "learning_rate": 1.2733483538047825e-05,
+      "loss": 0.2305,
+      "step": 3510
+    },
+    {
+      "epoch": 0.1279162729849553,
+      "grad_norm": 2.6026437282562256,
+      "learning_rate": 1.2769823388327639e-05,
+      "loss": 0.2618,
+      "step": 3520
+    },
+    {
+      "epoch": 0.12827967148775346,
+      "grad_norm": 8.228372573852539,
+      "learning_rate": 1.2806163238607458e-05,
+      "loss": 0.2586,
+      "step": 3530
+    },
+    {
+      "epoch": 0.12864306999055164,
+      "grad_norm": 2.643139362335205,
+      "learning_rate": 1.2842503088887276e-05,
+      "loss": 0.2197,
+      "step": 3540
+    },
+    {
+      "epoch": 0.12900646849334982,
+      "grad_norm": 400.0296325683594,
+      "learning_rate": 1.287884293916709e-05,
+      "loss": 0.3586,
+      "step": 3550
+    },
+    {
+      "epoch": 0.12936986699614797,
+      "grad_norm": 1.6349281072616577,
+      "learning_rate": 1.2915182789446908e-05,
+      "loss": 0.2364,
+      "step": 3560
+    },
+    {
+      "epoch": 0.12973326549894615,
+      "grad_norm": 2.6573753356933594,
+      "learning_rate": 1.2951522639726724e-05,
+      "loss": 0.2195,
+      "step": 3570
+    },
+    {
+      "epoch": 0.1300966640017443,
+      "grad_norm": 4.2721686363220215,
+      "learning_rate": 1.2987862490006542e-05,
+      "loss": 0.3092,
+      "step": 3580
+    },
+    {
+      "epoch": 0.13046006250454248,
+      "grad_norm": 2.9982502460479736,
+      "learning_rate": 1.302420234028636e-05,
+      "loss": 0.2826,
+      "step": 3590
+    },
+    {
+      "epoch": 0.13082346100734066,
+      "grad_norm": 8.903009414672852,
+      "learning_rate": 1.3060542190566175e-05,
+      "loss": 0.3367,
+      "step": 3600
+    },
+    {
+      "epoch": 0.13082346100734066,
+      "eval_loss": 0.4490436017513275,
+      "eval_runtime": 179.8743,
+      "eval_samples_per_second": 41.218,
+      "eval_steps_per_second": 5.154,
+      "eval_wer": 0.2664058670829778,
+      "step": 3600
+    },
+    {
+      "epoch": 0.1311868595101388,
+      "grad_norm": 2.9746363162994385,
+      "learning_rate": 1.3096882040845993e-05,
+      "loss": 0.2418,
+      "step": 3610
+    },
+    {
+      "epoch": 0.131550258012937,
+      "grad_norm": 2.274872303009033,
+      "learning_rate": 1.313322189112581e-05,
+      "loss": 0.3052,
+      "step": 3620
+    },
+    {
+      "epoch": 0.13191365651573517,
+      "grad_norm": 7.114847660064697,
+      "learning_rate": 1.3169561741405625e-05,
+      "loss": 0.2821,
+      "step": 3630
+    },
+    {
+      "epoch": 0.13227705501853332,
+      "grad_norm": 3.2101128101348877,
+      "learning_rate": 1.3205901591685444e-05,
+      "loss": 0.2223,
+      "step": 3640
+    },
+    {
+      "epoch": 0.1326404535213315,
+      "grad_norm": 18.914968490600586,
+      "learning_rate": 1.3242241441965258e-05,
+      "loss": 0.3809,
+      "step": 3650
+    },
+    {
+      "epoch": 0.13300385202412965,
+      "grad_norm": 2.399569272994995,
+      "learning_rate": 1.3278581292245076e-05,
+      "loss": 0.2221,
+      "step": 3660
+    },
+    {
+      "epoch": 0.13336725052692783,
+      "grad_norm": 5.76792573928833,
+      "learning_rate": 1.3314921142524894e-05,
+      "loss": 0.2487,
+      "step": 3670
+    },
+    {
+      "epoch": 0.133730649029726,
+      "grad_norm": 3.6859967708587646,
+      "learning_rate": 1.335126099280471e-05,
+      "loss": 0.2781,
+      "step": 3680
+    },
+    {
+      "epoch": 0.13409404753252416,
+      "grad_norm": 2.9653141498565674,
+      "learning_rate": 1.3387600843084527e-05,
+      "loss": 0.2258,
+      "step": 3690
+    },
+    {
+      "epoch": 0.13445744603532234,
+      "grad_norm": 19.170753479003906,
+      "learning_rate": 1.3423940693364345e-05,
+      "loss": 0.3902,
+      "step": 3700
+    },
+    {
+      "epoch": 0.13482084453812052,
+      "grad_norm": 2.2880115509033203,
+      "learning_rate": 1.3460280543644161e-05,
+      "loss": 0.2745,
+      "step": 3710
+    },
+    {
+      "epoch": 0.13518424304091867,
+      "grad_norm": 2.5196125507354736,
+      "learning_rate": 1.3496620393923979e-05,
+      "loss": 0.2293,
+      "step": 3720
+    },
+    {
+      "epoch": 0.13554764154371685,
+      "grad_norm": 3.827986001968384,
+      "learning_rate": 1.3532960244203793e-05,
+      "loss": 0.259,
+      "step": 3730
+    },
+    {
+      "epoch": 0.135911040046515,
+      "grad_norm": 3.4211530685424805,
+      "learning_rate": 1.356930009448361e-05,
+      "loss": 0.3256,
+      "step": 3740
+    },
+    {
+      "epoch": 0.13627443854931318,
+      "grad_norm": 26.879398345947266,
+      "learning_rate": 1.360563994476343e-05,
+      "loss": 0.3208,
+      "step": 3750
+    },
+    {
+      "epoch": 0.13663783705211135,
+      "grad_norm": 2.316091775894165,
+      "learning_rate": 1.3641979795043244e-05,
+      "loss": 0.2316,
+      "step": 3760
+    },
+    {
+      "epoch": 0.1370012355549095,
+      "grad_norm": 4.098924160003662,
+      "learning_rate": 1.3678319645323062e-05,
+      "loss": 0.2399,
+      "step": 3770
+    },
+    {
+      "epoch": 0.13736463405770769,
+      "grad_norm": 6.9372687339782715,
+      "learning_rate": 1.371465949560288e-05,
+      "loss": 0.2858,
+      "step": 3780
+    },
+    {
+      "epoch": 0.13772803256050586,
+      "grad_norm": 2.509535789489746,
+      "learning_rate": 1.3750999345882695e-05,
+      "loss": 0.2113,
+      "step": 3790
+    },
+    {
+      "epoch": 0.13809143106330402,
+      "grad_norm": 7.7181077003479,
+      "learning_rate": 1.3787339196162513e-05,
+      "loss": 0.3279,
+      "step": 3800
+    },
+    {
+      "epoch": 0.1384548295661022,
+      "grad_norm": 2.6843245029449463,
+      "learning_rate": 1.3823679046442329e-05,
+      "loss": 0.2267,
+      "step": 3810
+    },
+    {
+      "epoch": 0.13881822806890035,
+      "grad_norm": 3.05159068107605,
+      "learning_rate": 1.3860018896722147e-05,
+      "loss": 0.229,
+      "step": 3820
+    },
+    {
+      "epoch": 0.13918162657169852,
+      "grad_norm": 5.029635429382324,
+      "learning_rate": 1.3896358747001964e-05,
+      "loss": 0.2627,
+      "step": 3830
+    },
+    {
+      "epoch": 0.1395450250744967,
+      "grad_norm": 2.8287103176116943,
+      "learning_rate": 1.3932698597281779e-05,
+      "loss": 0.2294,
+      "step": 3840
+    },
+    {
+      "epoch": 0.13990842357729485,
+      "grad_norm": 24.862224578857422,
+      "learning_rate": 1.3969038447561598e-05,
+      "loss": 0.3198,
+      "step": 3850
+    },
+    {
+      "epoch": 0.14027182208009303,
+      "grad_norm": 5.624647617340088,
+      "learning_rate": 1.4005378297841412e-05,
+      "loss": 0.2641,
+      "step": 3860
+    },
+    {
+      "epoch": 0.14063522058289118,
+      "grad_norm": 1.6199389696121216,
+      "learning_rate": 1.404171814812123e-05,
+      "loss": 0.2279,
+      "step": 3870
+    },
+    {
+      "epoch": 0.14099861908568936,
+      "grad_norm": 2.864058017730713,
+      "learning_rate": 1.4078057998401047e-05,
+      "loss": 0.2448,
+      "step": 3880
+    },
+    {
+      "epoch": 0.14136201758848754,
+      "grad_norm": 3.897899627685547,
+      "learning_rate": 1.4114397848680863e-05,
+      "loss": 0.2438,
+      "step": 3890
+    },
+    {
+      "epoch": 0.1417254160912857,
+      "grad_norm": 42.4840087890625,
+      "learning_rate": 1.4150737698960681e-05,
+      "loss": 0.3604,
+      "step": 3900
+    },
+    {
+      "epoch": 0.14208881459408387,
+      "grad_norm": 1.6532913446426392,
+      "learning_rate": 1.4187077549240499e-05,
+      "loss": 0.2469,
+      "step": 3910
+    },
+    {
+      "epoch": 0.14245221309688205,
+      "grad_norm": 2.3755931854248047,
+      "learning_rate": 1.4223417399520315e-05,
+      "loss": 0.2233,
+      "step": 3920
+    },
+    {
+      "epoch": 0.1428156115996802,
+      "grad_norm": 5.866461277008057,
+      "learning_rate": 1.4259757249800132e-05,
+      "loss": 0.2952,
+      "step": 3930
+    },
+    {
+      "epoch": 0.14317901010247838,
+      "grad_norm": 3.171570301055908,
+      "learning_rate": 1.4296097100079947e-05,
+      "loss": 0.2329,
+      "step": 3940
+    },
+    {
+      "epoch": 0.14354240860527653,
+      "grad_norm": 23.302635192871094,
+      "learning_rate": 1.4332436950359764e-05,
+      "loss": 0.3592,
+      "step": 3950
+    },
+    {
+      "epoch": 0.1439058071080747,
+      "grad_norm": 2.3609213829040527,
+      "learning_rate": 1.4368776800639584e-05,
+      "loss": 1.1975,
+      "step": 3960
+    },
+    {
+      "epoch": 0.1442692056108729,
+      "grad_norm": 2.857872486114502,
+      "learning_rate": 1.4405116650919398e-05,
+      "loss": 0.265,
+      "step": 3970
+    },
+    {
+      "epoch": 0.14463260411367104,
+      "grad_norm": 6.918335914611816,
+      "learning_rate": 1.4441456501199216e-05,
+      "loss": 0.4057,
+      "step": 3980
+    },
+    {
+      "epoch": 0.14499600261646922,
+      "grad_norm": 3.8019461631774902,
+      "learning_rate": 1.4477796351479033e-05,
+      "loss": 0.299,
+      "step": 3990
+    },
+    {
+      "epoch": 0.1453594011192674,
+      "grad_norm": 30.963428497314453,
+      "learning_rate": 1.451413620175885e-05,
+      "loss": 0.335,
+      "step": 4000
+    },
+    {
+      "epoch": 0.14572279962206555,
+      "grad_norm": 2.3968963623046875,
+      "learning_rate": 1.4550476052038667e-05,
+      "loss": 0.224,
+      "step": 4010
+    },
+    {
+      "epoch": 0.14608619812486373,
+      "grad_norm": 6.7229485511779785,
+      "learning_rate": 1.4586815902318481e-05,
+      "loss": 0.2657,
+      "step": 4020
+    },
+    {
+      "epoch": 0.14644959662766188,
+      "grad_norm": 17.447879791259766,
+      "learning_rate": 1.46231557525983e-05,
+      "loss": 0.2199,
+      "step": 4030
+    },
+    {
+      "epoch": 0.14681299513046006,
+      "grad_norm": 2.020756721496582,
+      "learning_rate": 1.4659495602878118e-05,
+      "loss": 0.31,
+      "step": 4040
+    },
+    {
+      "epoch": 0.14717639363325824,
+      "grad_norm": 38.28268814086914,
+      "learning_rate": 1.4695835453157932e-05,
+      "loss": 0.3861,
+      "step": 4050
+    },
+    {
+      "epoch": 0.1475397921360564,
+      "grad_norm": 2.085073232650757,
+      "learning_rate": 1.473217530343775e-05,
+      "loss": 0.2078,
+      "step": 4060
+    },
+    {
+      "epoch": 0.14790319063885457,
+      "grad_norm": 3.453597068786621,
+      "learning_rate": 1.476851515371757e-05,
+      "loss": 0.296,
+      "step": 4070
+    },
+    {
+      "epoch": 0.14826658914165275,
+      "grad_norm": 2.3039424419403076,
+      "learning_rate": 1.4804855003997384e-05,
+      "loss": 0.2346,
+      "step": 4080
+    },
+    {
+      "epoch": 0.1486299876444509,
+      "grad_norm": 3.217890977859497,
+      "learning_rate": 1.4841194854277201e-05,
+      "loss": 0.2243,
+      "step": 4090
+    },
+    {
+      "epoch": 0.14899338614724908,
+      "grad_norm": 12.48748779296875,
+      "learning_rate": 1.4877534704557017e-05,
+      "loss": 0.3378,
+      "step": 4100
+    },
+    {
+      "epoch": 0.14935678465004723,
+      "grad_norm": 2.781388282775879,
+      "learning_rate": 1.4913874554836835e-05,
+      "loss": 0.2167,
+      "step": 4110
+    },
+    {
+      "epoch": 0.1497201831528454,
+      "grad_norm": 2.564457893371582,
+      "learning_rate": 1.4950214405116653e-05,
+      "loss": 0.2187,
+      "step": 4120
+    },
+    {
+      "epoch": 0.1500835816556436,
+      "grad_norm": 9.590895652770996,
+      "learning_rate": 1.4986554255396469e-05,
+      "loss": 0.2444,
+      "step": 4130
+    },
+    {
+      "epoch": 0.15044698015844174,
+      "grad_norm": 2.8055028915405273,
+      "learning_rate": 1.5022894105676286e-05,
+      "loss": 0.2499,
+      "step": 4140
+    },
+    {
+      "epoch": 0.15081037866123992,
+      "grad_norm": 7.157045364379883,
+      "learning_rate": 1.5059233955956104e-05,
+      "loss": 0.361,
+      "step": 4150
+    },
+    {
+      "epoch": 0.1511737771640381,
+      "grad_norm": 3.369006633758545,
+      "learning_rate": 1.5095573806235918e-05,
+      "loss": 0.259,
+      "step": 4160
+    },
+    {
+      "epoch": 0.15153717566683625,
+      "grad_norm": 5.334355354309082,
+      "learning_rate": 1.5131913656515736e-05,
+      "loss": 0.2797,
+      "step": 4170
+    },
+    {
+      "epoch": 0.15190057416963443,
+      "grad_norm": 6.667120456695557,
+      "learning_rate": 1.5168253506795552e-05,
+      "loss": 0.2678,
+      "step": 4180
+    },
+    {
+      "epoch": 0.15226397267243258,
+      "grad_norm": 1.7419887781143188,
+      "learning_rate": 1.520459335707537e-05,
+      "loss": 0.2432,
+      "step": 4190
+    },
+    {
+      "epoch": 0.15262737117523076,
+      "grad_norm": 7.022573947906494,
+      "learning_rate": 1.5240933207355187e-05,
+      "loss": 0.2955,
+      "step": 4200
+    },
+    {
+      "epoch": 0.15262737117523076,
+      "eval_loss": 0.4361402690410614,
+      "eval_runtime": 180.5933,
+      "eval_samples_per_second": 41.054,
+      "eval_steps_per_second": 5.133,
+      "eval_wer": 0.2589540181894095,
+      "step": 4200
+    },
+    {
+      "epoch": 0.15299076967802894,
+      "grad_norm": 3.108078718185425,
+      "learning_rate": 1.5277273057635e-05,
+      "loss": 3.6147,
+      "step": 4210
+    },
+    {
+      "epoch": 0.1533541681808271,
+      "grad_norm": 2.6063787937164307,
+      "learning_rate": 1.531361290791482e-05,
+      "loss": 0.232,
+      "step": 4220
+    },
+    {
+      "epoch": 0.15371756668362527,
+      "grad_norm": 3.581697463989258,
+      "learning_rate": 1.5349952758194637e-05,
+      "loss": 0.2451,
+      "step": 4230
+    },
+    {
+      "epoch": 0.15408096518642342,
+      "grad_norm": 2.5910837650299072,
+      "learning_rate": 1.5386292608474453e-05,
+      "loss": 0.2283,
+      "step": 4240
+    },
+    {
+      "epoch": 0.1544443636892216,
+      "grad_norm": 70.38739013671875,
+      "learning_rate": 1.5422632458754272e-05,
+      "loss": 0.396,
+      "step": 4250
+    },
+    {
+      "epoch": 0.15480776219201978,
+      "grad_norm": 3.5658187866210938,
+      "learning_rate": 1.5458972309034088e-05,
+      "loss": 0.2116,
+      "step": 4260
+    },
+    {
+      "epoch": 0.15517116069481793,
+      "grad_norm": 5.393126487731934,
+      "learning_rate": 1.5495312159313904e-05,
+      "loss": 0.2382,
+      "step": 4270
+    },
+    {
+      "epoch": 0.1555345591976161,
+      "grad_norm": 10.135586738586426,
+      "learning_rate": 1.5531652009593723e-05,
+      "loss": 0.2485,
+      "step": 4280
+    },
+    {
+      "epoch": 0.15589795770041429,
+      "grad_norm": 2.1143031120300293,
+      "learning_rate": 1.5567991859873536e-05,
+      "loss": 0.1936,
+      "step": 4290
+    },
+    {
+      "epoch": 0.15626135620321244,
+      "grad_norm": 20.077383041381836,
+      "learning_rate": 1.5604331710153355e-05,
+      "loss": 0.3818,
+      "step": 4300
+    },
+    {
+      "epoch": 0.15662475470601062,
+      "grad_norm": 3.793126344680786,
+      "learning_rate": 1.564067156043317e-05,
+      "loss": 0.2245,
+      "step": 4310
+    },
+    {
+      "epoch": 0.15698815320880877,
+      "grad_norm": 3.2057955265045166,
+      "learning_rate": 1.5677011410712987e-05,
+      "loss": 0.2551,
+      "step": 4320
+    },
+    {
+      "epoch": 0.15735155171160695,
+      "grad_norm": 5.002716064453125,
+      "learning_rate": 1.5713351260992806e-05,
+      "loss": 0.2951,
+      "step": 4330
+    },
+    {
+      "epoch": 0.15771495021440513,
+      "grad_norm": 2.2240726947784424,
+      "learning_rate": 1.5749691111272622e-05,
+      "loss": 0.1993,
+      "step": 4340
+    },
+    {
+      "epoch": 0.15807834871720328,
+      "grad_norm": 55.30891036987305,
+      "learning_rate": 1.578603096155244e-05,
+      "loss": 0.2803,
+      "step": 4350
+    },
+    {
+      "epoch": 0.15844174722000146,
+      "grad_norm": 1.9186596870422363,
+      "learning_rate": 1.5822370811832258e-05,
+      "loss": 0.2234,
+      "step": 4360
+    },
+    {
+      "epoch": 0.15880514572279963,
+      "grad_norm": 1.7817661762237549,
+      "learning_rate": 1.5858710662112074e-05,
+      "loss": 0.2038,
+      "step": 4370
+    },
+    {
+      "epoch": 0.15916854422559779,
+      "grad_norm": 3.046330690383911,
+      "learning_rate": 1.589505051239189e-05,
+      "loss": 0.2809,
+      "step": 4380
+    },
+    {
+      "epoch": 0.15953194272839596,
+      "grad_norm": 5.43302583694458,
+      "learning_rate": 1.5931390362671706e-05,
+      "loss": 0.1896,
+      "step": 4390
+    },
+    {
+      "epoch": 0.15989534123119412,
+      "grad_norm": 12.185855865478516,
+      "learning_rate": 1.596773021295152e-05,
+      "loss": 0.2984,
+      "step": 4400
+    },
+    {
+      "epoch": 0.1602587397339923,
+      "grad_norm": 1.9507842063903809,
+      "learning_rate": 1.600407006323134e-05,
+      "loss": 0.2064,
+      "step": 4410
+    },
+    {
+      "epoch": 0.16062213823679047,
+      "grad_norm": 4.536543846130371,
+      "learning_rate": 1.6040409913511157e-05,
+      "loss": 0.2433,
+      "step": 4420
+    },
+    {
+      "epoch": 0.16098553673958862,
+      "grad_norm": 3.101174831390381,
+      "learning_rate": 1.6076749763790973e-05,
+      "loss": 0.2746,
+      "step": 4430
+    },
+    {
+      "epoch": 0.1613489352423868,
+      "grad_norm": 2.2098021507263184,
+      "learning_rate": 1.6113089614070792e-05,
+      "loss": 0.17,
+      "step": 4440
+    },
+    {
+      "epoch": 0.16171233374518498,
+      "grad_norm": 59.360809326171875,
+      "learning_rate": 1.6149429464350608e-05,
+      "loss": 0.379,
+      "step": 4450
+    },
+    {
+      "epoch": 0.16207573224798313,
+      "grad_norm": 6.364736557006836,
+      "learning_rate": 1.6185769314630424e-05,
+      "loss": 0.2224,
+      "step": 4460
+    },
+    {
+      "epoch": 0.1624391307507813,
+      "grad_norm": 3.2455356121063232,
+      "learning_rate": 1.622210916491024e-05,
+      "loss": 0.2195,
+      "step": 4470
+    },
+    {
+      "epoch": 0.16280252925357946,
+      "grad_norm": 6.399629592895508,
+      "learning_rate": 1.625844901519006e-05,
+      "loss": 0.266,
+      "step": 4480
+    },
+    {
+      "epoch": 0.16316592775637764,
+      "grad_norm": 16.19785499572754,
+      "learning_rate": 1.6294788865469875e-05,
+      "loss": 0.1836,
+      "step": 4490
+    },
+    {
+      "epoch": 0.16352932625917582,
+      "grad_norm": 7.909778594970703,
+      "learning_rate": 1.633112871574969e-05,
+      "loss": 0.6016,
+      "step": 4500
+    },
+    {
+      "epoch": 0.16389272476197397,
+      "grad_norm": 2.8134663105010986,
+      "learning_rate": 1.636746856602951e-05,
+      "loss": 0.2148,
+      "step": 4510
+    },
+    {
+      "epoch": 0.16425612326477215,
+      "grad_norm": 2.667999505996704,
+      "learning_rate": 1.6403808416309327e-05,
+      "loss": 0.2294,
+      "step": 4520
+    },
+    {
+      "epoch": 0.1646195217675703,
+      "grad_norm": 3.355242967605591,
+      "learning_rate": 1.6440148266589143e-05,
+      "loss": 0.2097,
+      "step": 4530
+    },
+    {
+      "epoch": 0.16498292027036848,
+      "grad_norm": 2.6241908073425293,
+      "learning_rate": 1.647648811686896e-05,
+      "loss": 0.2337,
+      "step": 4540
+    },
+    {
+      "epoch": 0.16534631877316666,
+      "grad_norm": 16.759428024291992,
+      "learning_rate": 1.6512827967148775e-05,
+      "loss": 0.2944,
+      "step": 4550
+    },
+    {
+      "epoch": 0.1657097172759648,
+      "grad_norm": 3.098898410797119,
+      "learning_rate": 1.6549167817428594e-05,
+      "loss": 0.1895,
+      "step": 4560
+    },
+    {
+      "epoch": 0.166073115778763,
+      "grad_norm": 4.042644023895264,
+      "learning_rate": 1.658550766770841e-05,
+      "loss": 0.2369,
+      "step": 4570
+    },
+    {
+      "epoch": 0.16643651428156117,
+      "grad_norm": 7.174807548522949,
+      "learning_rate": 1.6621847517988226e-05,
+      "loss": 0.2331,
+      "step": 4580
+    },
+    {
+      "epoch": 0.16679991278435932,
+      "grad_norm": 2.1805012226104736,
+      "learning_rate": 1.6658187368268045e-05,
+      "loss": 0.2422,
+      "step": 4590
+    },
+    {
+      "epoch": 0.1671633112871575,
+      "grad_norm": 18.097871780395508,
+      "learning_rate": 1.6694527218547858e-05,
+      "loss": 0.347,
+      "step": 4600
+    },
+    {
+      "epoch": 0.16752670978995565,
+      "grad_norm": 3.48561429977417,
+      "learning_rate": 1.6730867068827677e-05,
+      "loss": 0.2985,
+      "step": 4610
+    },
+    {
+      "epoch": 0.16789010829275383,
+      "grad_norm": 1.7519229650497437,
+      "learning_rate": 1.6767206919107496e-05,
+      "loss": 0.2204,
+      "step": 4620
+    },
+    {
+      "epoch": 0.168253506795552,
+      "grad_norm": 3.7641661167144775,
+      "learning_rate": 1.680354676938731e-05,
+      "loss": 0.2348,
+      "step": 4630
+    },
+    {
+      "epoch": 0.16861690529835016,
+      "grad_norm": 3.0688085556030273,
+      "learning_rate": 1.683988661966713e-05,
+      "loss": 0.2147,
+      "step": 4640
+    },
+    {
+      "epoch": 0.16898030380114834,
+      "grad_norm": 25.845094680786133,
+      "learning_rate": 1.6876226469946944e-05,
+      "loss": 0.3671,
+      "step": 4650
+    },
+    {
+      "epoch": 0.16934370230394652,
+      "grad_norm": 2.841994524002075,
+      "learning_rate": 1.691256632022676e-05,
+      "loss": 0.2182,
+      "step": 4660
+    },
+    {
+      "epoch": 0.16970710080674467,
+      "grad_norm": 1.0501997470855713,
+      "learning_rate": 1.694890617050658e-05,
+      "loss": 0.1791,
+      "step": 4670
+    },
+    {
+      "epoch": 0.17007049930954285,
+      "grad_norm": 3.3973441123962402,
+      "learning_rate": 1.6985246020786392e-05,
+      "loss": 0.3338,
+      "step": 4680
+    },
+    {
+      "epoch": 0.170433897812341,
+      "grad_norm": 1.8442267179489136,
+      "learning_rate": 1.702158587106621e-05,
+      "loss": 0.2528,
+      "step": 4690
+    },
+    {
+      "epoch": 0.17079729631513918,
+      "grad_norm": 42.373409271240234,
+      "learning_rate": 1.705792572134603e-05,
+      "loss": 0.2892,
+      "step": 4700
+    },
+    {
+      "epoch": 0.17116069481793736,
+      "grad_norm": 6.344671726226807,
+      "learning_rate": 1.7094265571625844e-05,
+      "loss": 0.2474,
+      "step": 4710
+    },
+    {
+      "epoch": 0.1715240933207355,
+      "grad_norm": 1.6177664995193481,
+      "learning_rate": 1.7130605421905663e-05,
+      "loss": 0.2364,
+      "step": 4720
+    },
+    {
+      "epoch": 0.1718874918235337,
+      "grad_norm": 4.98591423034668,
+      "learning_rate": 1.7166945272185482e-05,
+      "loss": 0.2046,
+      "step": 4730
+    },
+    {
+      "epoch": 0.17225089032633187,
+      "grad_norm": 7.943169116973877,
+      "learning_rate": 1.7203285122465295e-05,
+      "loss": 0.293,
+      "step": 4740
+    },
+    {
+      "epoch": 0.17261428882913002,
+      "grad_norm": 7.402034759521484,
+      "learning_rate": 1.7239624972745114e-05,
+      "loss": 0.2722,
+      "step": 4750
+    },
+    {
+      "epoch": 0.1729776873319282,
+      "grad_norm": 13.290019035339355,
+      "learning_rate": 1.727596482302493e-05,
+      "loss": 0.347,
+      "step": 4760
+    },
+    {
+      "epoch": 0.17334108583472635,
+      "grad_norm": 1.8591586351394653,
+      "learning_rate": 1.7312304673304746e-05,
+      "loss": 0.2291,
+      "step": 4770
+    },
+    {
+      "epoch": 0.17370448433752453,
+      "grad_norm": 2.5220861434936523,
+      "learning_rate": 1.7348644523584565e-05,
+      "loss": 0.2436,
+      "step": 4780
+    },
+    {
+      "epoch": 0.1740678828403227,
+      "grad_norm": 1.8692690134048462,
+      "learning_rate": 1.738498437386438e-05,
+      "loss": 0.1782,
+      "step": 4790
+    },
+    {
+      "epoch": 0.17443128134312086,
+      "grad_norm": 12.558557510375977,
+      "learning_rate": 1.7421324224144197e-05,
+      "loss": 0.3347,
+      "step": 4800
+    },
+    {
+      "epoch": 0.17443128134312086,
+      "eval_loss": 0.4148472547531128,
+      "eval_runtime": 180.0999,
+      "eval_samples_per_second": 41.166,
+      "eval_steps_per_second": 5.147,
+      "eval_wer": 0.23564543358687168,
+      "step": 4800
+    },
+    {
+      "epoch": 0.17479467984591904,
+      "grad_norm": 6.168694972991943,
+      "learning_rate": 1.7457664074424017e-05,
+      "loss": 0.2183,
+      "step": 4810
+    },
+    {
+      "epoch": 0.17515807834871722,
+      "grad_norm": 5.153416633605957,
+      "learning_rate": 1.749400392470383e-05,
+      "loss": 0.2689,
+      "step": 4820
+    },
+    {
+      "epoch": 0.17552147685151537,
+      "grad_norm": 2.8500893115997314,
+      "learning_rate": 1.753034377498365e-05,
+      "loss": 0.2848,
+      "step": 4830
+    },
+    {
+      "epoch": 0.17588487535431355,
+      "grad_norm": 17.89117431640625,
+      "learning_rate": 1.7566683625263465e-05,
+      "loss": 0.2539,
+      "step": 4840
+    },
+    {
+      "epoch": 0.1762482738571117,
+      "grad_norm": 19.455005645751953,
+      "learning_rate": 1.760302347554328e-05,
+      "loss": 0.3166,
+      "step": 4850
+    },
+    {
+      "epoch": 0.17661167235990988,
+      "grad_norm": 1.7975777387619019,
+      "learning_rate": 1.76393633258231e-05,
+      "loss": 0.1927,
+      "step": 4860
+    },
+    {
+      "epoch": 0.17697507086270806,
+      "grad_norm": 4.6790690422058105,
+      "learning_rate": 1.7675703176102916e-05,
+      "loss": 0.2248,
+      "step": 4870
+    },
+    {
+      "epoch": 0.1773384693655062,
+      "grad_norm": 3.2644243240356445,
+      "learning_rate": 1.7712043026382732e-05,
+      "loss": 0.2239,
+      "step": 4880
+    },
+    {
+      "epoch": 0.1777018678683044,
+      "grad_norm": 1.9375410079956055,
+      "learning_rate": 1.7748382876662548e-05,
+      "loss": 0.2053,
+      "step": 4890
+    },
+    {
+      "epoch": 0.17806526637110254,
+      "grad_norm": 15.435178756713867,
+      "learning_rate": 1.7784722726942367e-05,
+      "loss": 0.2903,
+      "step": 4900
+    },
+    {
+      "epoch": 0.17842866487390072,
+      "grad_norm": 2.486330270767212,
+      "learning_rate": 1.7821062577222183e-05,
+      "loss": 0.2598,
+      "step": 4910
+    },
+    {
+      "epoch": 0.1787920633766989,
+      "grad_norm": 2.5542314052581787,
+      "learning_rate": 1.7857402427502e-05,
+      "loss": 0.2305,
+      "step": 4920
+    },
+    {
+      "epoch": 0.17915546187949705,
+      "grad_norm": 3.6416103839874268,
+      "learning_rate": 1.7893742277781815e-05,
+      "loss": 1.046,
+      "step": 4930
+    },
+    {
+      "epoch": 0.17951886038229523,
+      "grad_norm": 1.9395058155059814,
+      "learning_rate": 1.7930082128061634e-05,
+      "loss": 0.2466,
+      "step": 4940
+    },
+    {
+      "epoch": 0.1798822588850934,
+      "grad_norm": 7.664824962615967,
+      "learning_rate": 1.796642197834145e-05,
+      "loss": 0.2871,
+      "step": 4950
+    },
+    {
+      "epoch": 0.18024565738789156,
+      "grad_norm": 2.0301320552825928,
+      "learning_rate": 1.8002761828621266e-05,
+      "loss": 0.1996,
+      "step": 4960
+    },
+    {
+      "epoch": 0.18060905589068973,
+      "grad_norm": 8.371182441711426,
+      "learning_rate": 1.8039101678901082e-05,
+      "loss": 0.1947,
+      "step": 4970
+    },
+    {
+      "epoch": 0.18097245439348789,
+      "grad_norm": 2.6746129989624023,
+      "learning_rate": 1.80754415291809e-05,
+      "loss": 0.2679,
+      "step": 4980
+    },
+    {
+      "epoch": 0.18133585289628606,
+      "grad_norm": 3.448202133178711,
+      "learning_rate": 1.8111781379460718e-05,
+      "loss": 0.1859,
+      "step": 4990
+    },
+    {
+      "epoch": 0.18169925139908424,
+      "grad_norm": 28.57021141052246,
+      "learning_rate": 1.8148121229740534e-05,
+      "loss": 0.3318,
+      "step": 5000
+    },
+    {
+      "epoch": 0.1820626499018824,
+      "grad_norm": 4.731750965118408,
+      "learning_rate": 1.8184461080020353e-05,
+      "loss": 0.2354,
+      "step": 5010
+    },
+    {
+      "epoch": 0.18242604840468057,
+      "grad_norm": 1.6815394163131714,
+      "learning_rate": 1.822080093030017e-05,
+      "loss": 0.2075,
+      "step": 5020
+    },
+    {
+      "epoch": 0.18278944690747875,
+      "grad_norm": 3.868263006210327,
+      "learning_rate": 1.8257140780579985e-05,
+      "loss": 0.2345,
+      "step": 5030
+    },
+    {
+      "epoch": 0.1831528454102769,
+      "grad_norm": 1.964240550994873,
+      "learning_rate": 1.82934806308598e-05,
+      "loss": 0.222,
+      "step": 5040
+    },
+    {
+      "epoch": 0.18351624391307508,
+      "grad_norm": 11.881858825683594,
+      "learning_rate": 1.8329820481139617e-05,
+      "loss": 0.3251,
+      "step": 5050
+    },
+    {
+      "epoch": 0.18387964241587323,
+      "grad_norm": 1.8463056087493896,
+      "learning_rate": 1.8366160331419436e-05,
+      "loss": 0.2255,
+      "step": 5060
+    },
+    {
+      "epoch": 0.1842430409186714,
+      "grad_norm": 2.592672348022461,
+      "learning_rate": 1.8402500181699252e-05,
+      "loss": 0.1904,
+      "step": 5070
+    },
+    {
+      "epoch": 0.1846064394214696,
+      "grad_norm": 4.0694074630737305,
+      "learning_rate": 1.8438840031979068e-05,
+      "loss": 0.2,
+      "step": 5080
+    },
+    {
+      "epoch": 0.18496983792426774,
+      "grad_norm": 2.101837396621704,
+      "learning_rate": 1.8475179882258887e-05,
+      "loss": 0.1927,
+      "step": 5090
+    },
+    {
+      "epoch": 0.18533323642706592,
+      "grad_norm": 22.162702560424805,
+      "learning_rate": 1.8511519732538703e-05,
+      "loss": 0.3481,
+      "step": 5100
+    },
+    {
+      "epoch": 0.1856966349298641,
+      "grad_norm": 2.7928340435028076,
+      "learning_rate": 1.854785958281852e-05,
+      "loss": 0.2344,
+      "step": 5110
+    },
+    {
+      "epoch": 0.18606003343266225,
+      "grad_norm": 1.8618485927581787,
+      "learning_rate": 1.858419943309834e-05,
+      "loss": 0.2139,
+      "step": 5120
+    },
+    {
+      "epoch": 0.18642343193546043,
+      "grad_norm": 2.9611120223999023,
+      "learning_rate": 1.862053928337815e-05,
+      "loss": 0.2194,
+      "step": 5130
+    },
+    {
+      "epoch": 0.18678683043825858,
+      "grad_norm": 5.181276321411133,
+      "learning_rate": 1.865687913365797e-05,
+      "loss": 0.2596,
+      "step": 5140
+    },
+    {
+      "epoch": 0.18715022894105676,
+      "grad_norm": 10.01041030883789,
+      "learning_rate": 1.8693218983937787e-05,
+      "loss": 0.3122,
+      "step": 5150
+    },
+    {
+      "epoch": 0.18751362744385494,
+      "grad_norm": 4.952126979827881,
+      "learning_rate": 1.8729558834217603e-05,
+      "loss": 0.2183,
+      "step": 5160
+    },
+    {
+      "epoch": 0.1878770259466531,
+      "grad_norm": 2.19279146194458,
+      "learning_rate": 1.8765898684497422e-05,
+      "loss": 0.2439,
+      "step": 5170
+    },
+    {
+      "epoch": 0.18824042444945127,
+      "grad_norm": 3.5189321041107178,
+      "learning_rate": 1.8802238534777238e-05,
+      "loss": 0.2343,
+      "step": 5180
+    },
+    {
+      "epoch": 0.18860382295224945,
+      "grad_norm": 2.0936787128448486,
+      "learning_rate": 1.8838578385057054e-05,
+      "loss": 0.1831,
+      "step": 5190
+    },
+    {
+      "epoch": 0.1889672214550476,
+      "grad_norm": 12.835061073303223,
+      "learning_rate": 1.8874918235336873e-05,
+      "loss": 0.2561,
+      "step": 5200
+    },
+    {
+      "epoch": 0.18933061995784578,
+      "grad_norm": 1.6738308668136597,
+      "learning_rate": 1.8911258085616686e-05,
+      "loss": 1.0257,
+      "step": 5210
+    },
+    {
+      "epoch": 0.18969401846064393,
+      "grad_norm": 2.7661142349243164,
+      "learning_rate": 1.8947597935896505e-05,
+      "loss": 0.2398,
+      "step": 5220
+    },
+    {
+      "epoch": 0.1900574169634421,
+      "grad_norm": 4.173921585083008,
+      "learning_rate": 1.8983937786176324e-05,
+      "loss": 0.2157,
+      "step": 5230
+    },
+    {
+      "epoch": 0.1904208154662403,
+      "grad_norm": 3.7037158012390137,
+      "learning_rate": 1.9020277636456137e-05,
+      "loss": 0.2182,
+      "step": 5240
+    },
+    {
+      "epoch": 0.19078421396903844,
+      "grad_norm": 16.288227081298828,
+      "learning_rate": 1.9056617486735956e-05,
+      "loss": 0.2829,
+      "step": 5250
+    },
+    {
+      "epoch": 0.19114761247183662,
+      "grad_norm": 2.0504090785980225,
+      "learning_rate": 1.9092957337015772e-05,
+      "loss": 0.201,
+      "step": 5260
+    },
+    {
+      "epoch": 0.19151101097463477,
+      "grad_norm": 1.2266415357589722,
+      "learning_rate": 1.9129297187295588e-05,
+      "loss": 0.2072,
+      "step": 5270
+    },
+    {
+      "epoch": 0.19187440947743295,
+      "grad_norm": 4.910546779632568,
+      "learning_rate": 1.9165637037575408e-05,
+      "loss": 0.1824,
+      "step": 5280
+    },
+    {
+      "epoch": 0.19223780798023113,
+      "grad_norm": 3.093318223953247,
+      "learning_rate": 1.9201976887855224e-05,
+      "loss": 0.2471,
+      "step": 5290
+    },
+    {
+      "epoch": 0.19260120648302928,
+      "grad_norm": 6.74167013168335,
+      "learning_rate": 1.923831673813504e-05,
+      "loss": 0.2912,
+      "step": 5300
+    },
+    {
+      "epoch": 0.19296460498582746,
+      "grad_norm": 2.0540058612823486,
+      "learning_rate": 1.927465658841486e-05,
+      "loss": 0.2599,
+      "step": 5310
+    },
+    {
+      "epoch": 0.19332800348862564,
+      "grad_norm": 2.407750129699707,
+      "learning_rate": 1.931099643869467e-05,
+      "loss": 0.2478,
+      "step": 5320
+    },
+    {
+      "epoch": 0.1936914019914238,
+      "grad_norm": 5.479567527770996,
+      "learning_rate": 1.934733628897449e-05,
+      "loss": 0.5936,
+      "step": 5330
+    },
+    {
+      "epoch": 0.19405480049422197,
+      "grad_norm": 1.912705659866333,
+      "learning_rate": 1.9383676139254307e-05,
+      "loss": 0.215,
+      "step": 5340
+    },
+    {
+      "epoch": 0.19441819899702012,
+      "grad_norm": 38.24689865112305,
+      "learning_rate": 1.9420015989534123e-05,
+      "loss": 0.286,
+      "step": 5350
+    },
+    {
+      "epoch": 0.1947815974998183,
+      "grad_norm": 3.4196550846099854,
+      "learning_rate": 1.9456355839813942e-05,
+      "loss": 0.4764,
+      "step": 5360
+    },
+    {
+      "epoch": 0.19514499600261648,
+      "grad_norm": 1.705702781677246,
+      "learning_rate": 1.9492695690093758e-05,
+      "loss": 0.195,
+      "step": 5370
+    },
+    {
+      "epoch": 0.19550839450541463,
+      "grad_norm": 2.7188572883605957,
+      "learning_rate": 1.9529035540373574e-05,
+      "loss": 0.2318,
+      "step": 5380
+    },
+    {
+      "epoch": 0.1958717930082128,
+      "grad_norm": 5.217918872833252,
+      "learning_rate": 1.9565375390653393e-05,
+      "loss": 0.2288,
+      "step": 5390
+    },
+    {
+      "epoch": 0.196235191511011,
+      "grad_norm": 7.094780921936035,
+      "learning_rate": 1.960171524093321e-05,
+      "loss": 0.3607,
+      "step": 5400
+    },
+    {
+      "epoch": 0.196235191511011,
+      "eval_loss": 0.3953820765018463,
+      "eval_runtime": 180.5214,
+      "eval_samples_per_second": 41.07,
+      "eval_steps_per_second": 5.135,
+      "eval_wer": 0.23525514186650207,
+      "step": 5400
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 165108,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 1800,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.7687984050549832e+19,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}