{
  "best_global_step": 1000,
  "best_metric": 0.4849427044391632,
  "best_model_checkpoint": "./openai/whisper-small-kpo-waxal/checkpoint-1000",
  "epoch": 5.543840177580466,
  "eval_steps": 500,
  "global_step": 2500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.05549389567147614,
      "grad_norm": 23.202674865722656,
      "learning_rate": 4.800000000000001e-06,
      "loss": 5.858453369140625,
      "step": 25
    },
    {
      "epoch": 0.11098779134295228,
      "grad_norm": 10.280803680419922,
      "learning_rate": 9.800000000000001e-06,
      "loss": 2.6889111328125,
      "step": 50
    },
    {
      "epoch": 0.16648168701442842,
      "grad_norm": 11.007328987121582,
      "learning_rate": 1.48e-05,
      "loss": 1.9935995483398437,
      "step": 75
    },
    {
      "epoch": 0.22197558268590456,
      "grad_norm": 8.749046325683594,
      "learning_rate": 1.9800000000000004e-05,
      "loss": 1.6322769165039062,
      "step": 100
    },
    {
      "epoch": 0.27746947835738067,
      "grad_norm": 10.869550704956055,
      "learning_rate": 2.48e-05,
      "loss": 1.5226153564453124,
      "step": 125
    },
    {
      "epoch": 0.33296337402885684,
      "grad_norm": 6.452018737792969,
      "learning_rate": 2.98e-05,
      "loss": 1.417423858642578,
      "step": 150
    },
    {
      "epoch": 0.38845726970033295,
      "grad_norm": 8.565808296203613,
      "learning_rate": 3.48e-05,
      "loss": 1.358447265625,
      "step": 175
    },
    {
      "epoch": 0.4439511653718091,
      "grad_norm": 7.219192981719971,
      "learning_rate": 3.9800000000000005e-05,
      "loss": 1.337242889404297,
      "step": 200
    },
    {
      "epoch": 0.49944506104328523,
      "grad_norm": 8.29965591430664,
      "learning_rate": 4.4800000000000005e-05,
      "loss": 1.273004684448242,
      "step": 225
    },
    {
      "epoch": 0.5549389567147613,
      "grad_norm": 7.17012357711792,
      "learning_rate": 4.9800000000000004e-05,
      "loss": 1.2329086303710937,
      "step": 250
    },
    {
      "epoch": 0.6104328523862376,
      "grad_norm": 5.866935729980469,
      "learning_rate": 5.4800000000000004e-05,
      "loss": 1.2089253234863282,
      "step": 275
    },
    {
      "epoch": 0.6659267480577137,
      "grad_norm": 6.515329360961914,
      "learning_rate": 5.9800000000000003e-05,
      "loss": 1.1282675170898437,
      "step": 300
    },
    {
      "epoch": 0.7214206437291898,
      "grad_norm": 5.666633129119873,
      "learning_rate": 6.48e-05,
      "loss": 1.148940887451172,
      "step": 325
    },
    {
      "epoch": 0.7769145394006659,
      "grad_norm": 6.579802513122559,
      "learning_rate": 6.98e-05,
      "loss": 1.1574817657470704,
      "step": 350
    },
    {
      "epoch": 0.832408435072142,
      "grad_norm": 6.159289360046387,
      "learning_rate": 7.48e-05,
      "loss": 1.1553614807128907,
      "step": 375
    },
    {
      "epoch": 0.8879023307436182,
      "grad_norm": 4.6766204833984375,
      "learning_rate": 7.98e-05,
      "loss": 1.1272061920166017,
      "step": 400
    },
    {
      "epoch": 0.9433962264150944,
      "grad_norm": 5.207922458648682,
      "learning_rate": 8.48e-05,
      "loss": 1.1430768585205078,
      "step": 425
    },
    {
      "epoch": 0.9988901220865705,
      "grad_norm": 5.042243480682373,
      "learning_rate": 8.98e-05,
      "loss": 1.1235774230957032,
      "step": 450
    },
    {
      "epoch": 1.053274139844617,
      "grad_norm": 5.123244285583496,
      "learning_rate": 9.48e-05,
      "loss": 0.9237544250488281,
      "step": 475
    },
    {
      "epoch": 1.1087680355160932,
      "grad_norm": 4.658648490905762,
      "learning_rate": 9.98e-05,
      "loss": 0.96513427734375,
      "step": 500
    },
    {
      "epoch": 1.1087680355160932,
      "eval_cer": 0.279414692054622,
      "eval_loss": 0.5450363755226135,
      "eval_runtime": 211.0102,
      "eval_samples_per_second": 8.341,
      "eval_steps_per_second": 0.133,
      "eval_wer": 0.77612952445537,
      "step": 500
    },
    {
      "epoch": 1.1642619311875695,
      "grad_norm": 5.715120792388916,
      "learning_rate": 9.981580966999233e-05,
      "loss": 1.0203121948242186,
      "step": 525
    },
    {
      "epoch": 1.2197558268590456,
      "grad_norm": 5.163341522216797,
      "learning_rate": 9.9623944742901e-05,
      "loss": 0.9690386962890625,
      "step": 550
    },
    {
      "epoch": 1.2752497225305217,
      "grad_norm": 5.062475681304932,
      "learning_rate": 9.943207981580968e-05,
      "loss": 0.9559343719482422,
      "step": 575
    },
    {
      "epoch": 1.3307436182019978,
      "grad_norm": 5.288034915924072,
      "learning_rate": 9.924021488871836e-05,
      "loss": 0.9570340728759765,
      "step": 600
    },
    {
      "epoch": 1.386237513873474,
      "grad_norm": 4.448609828948975,
      "learning_rate": 9.904834996162702e-05,
      "loss": 0.9402964782714843,
      "step": 625
    },
    {
      "epoch": 1.44173140954495,
      "grad_norm": 3.7210636138916016,
      "learning_rate": 9.885648503453569e-05,
      "loss": 0.9727059173583984,
      "step": 650
    },
    {
      "epoch": 1.4972253052164262,
      "grad_norm": 4.641828536987305,
      "learning_rate": 9.866462010744436e-05,
      "loss": 0.9270778656005859,
      "step": 675
    },
    {
      "epoch": 1.5527192008879025,
      "grad_norm": 4.52638053894043,
      "learning_rate": 9.847275518035304e-05,
      "loss": 0.9360514831542969,
      "step": 700
    },
    {
      "epoch": 1.6082130965593784,
      "grad_norm": 4.332831859588623,
      "learning_rate": 9.828089025326171e-05,
      "loss": 0.9043086242675781,
      "step": 725
    },
    {
      "epoch": 1.6637069922308547,
      "grad_norm": 4.253662109375,
      "learning_rate": 9.808902532617038e-05,
      "loss": 0.9130302429199219,
      "step": 750
    },
    {
      "epoch": 1.7192008879023306,
      "grad_norm": 3.683013916015625,
      "learning_rate": 9.789716039907906e-05,
      "loss": 0.9052154541015625,
      "step": 775
    },
    {
      "epoch": 1.774694783573807,
      "grad_norm": 3.542875051498413,
      "learning_rate": 9.770529547198772e-05,
      "loss": 0.9154473114013671,
      "step": 800
    },
    {
      "epoch": 1.830188679245283,
      "grad_norm": 3.8648242950439453,
      "learning_rate": 9.75134305448964e-05,
      "loss": 0.8997632598876953,
      "step": 825
    },
    {
      "epoch": 1.8856825749167592,
      "grad_norm": 3.9503533840179443,
      "learning_rate": 9.732156561780506e-05,
      "loss": 0.934085464477539,
      "step": 850
    },
    {
      "epoch": 1.9411764705882353,
      "grad_norm": 3.7985453605651855,
      "learning_rate": 9.712970069071374e-05,
      "loss": 0.9187238311767578,
      "step": 875
    },
    {
      "epoch": 1.9966703662597114,
      "grad_norm": 4.175370693206787,
      "learning_rate": 9.693783576362242e-05,
      "loss": 0.9017095947265625,
      "step": 900
    },
    {
      "epoch": 2.051054384017758,
      "grad_norm": 3.5287599563598633,
      "learning_rate": 9.674597083653109e-05,
      "loss": 0.699445571899414,
      "step": 925
    },
    {
      "epoch": 2.106548279689234,
      "grad_norm": 3.792794704437256,
      "learning_rate": 9.655410590943976e-05,
      "loss": 0.6688622283935547,
      "step": 950
    },
    {
      "epoch": 2.1620421753607104,
      "grad_norm": 3.698681354522705,
      "learning_rate": 9.636224098234843e-05,
      "loss": 0.7054427337646484,
      "step": 975
    },
    {
      "epoch": 2.2175360710321863,
      "grad_norm": 4.332374095916748,
      "learning_rate": 9.61703760552571e-05,
      "loss": 0.7015140533447266,
      "step": 1000
    },
    {
      "epoch": 2.2175360710321863,
      "eval_cer": 0.26578814979424464,
      "eval_loss": 0.4849427044391632,
      "eval_runtime": 197.0927,
      "eval_samples_per_second": 8.93,
      "eval_steps_per_second": 0.142,
      "eval_wer": 0.7392896499109943,
      "step": 1000
    },
    {
      "epoch": 2.2730299667036626,
      "grad_norm": 2.7401206493377686,
      "learning_rate": 9.597851112816577e-05,
      "loss": 0.6754109191894532,
      "step": 1025
    },
    {
      "epoch": 2.328523862375139,
      "grad_norm": 3.663332462310791,
      "learning_rate": 9.578664620107444e-05,
      "loss": 0.6997630310058593,
      "step": 1050
    },
    {
      "epoch": 2.384017758046615,
      "grad_norm": 3.4043989181518555,
      "learning_rate": 9.559478127398312e-05,
      "loss": 0.6894276428222657,
      "step": 1075
    },
    {
      "epoch": 2.439511653718091,
      "grad_norm": 2.929145574569702,
      "learning_rate": 9.540291634689179e-05,
      "loss": 0.7089242553710937,
      "step": 1100
    },
    {
      "epoch": 2.495005549389567,
      "grad_norm": 4.095437526702881,
      "learning_rate": 9.521105141980047e-05,
      "loss": 0.7410848236083984,
      "step": 1125
    },
    {
      "epoch": 2.5504994450610434,
      "grad_norm": 3.1726880073547363,
      "learning_rate": 9.501918649270914e-05,
      "loss": 0.7000830841064453,
      "step": 1150
    },
    {
      "epoch": 2.6059933407325193,
      "grad_norm": 3.1098122596740723,
      "learning_rate": 9.482732156561781e-05,
      "loss": 0.704638442993164,
      "step": 1175
    },
    {
      "epoch": 2.6614872364039956,
      "grad_norm": 3.6283864974975586,
      "learning_rate": 9.463545663852649e-05,
      "loss": 0.7068312072753906,
      "step": 1200
    },
    {
      "epoch": 2.7169811320754715,
      "grad_norm": 3.307800531387329,
      "learning_rate": 9.444359171143516e-05,
      "loss": 0.703956527709961,
      "step": 1225
    },
    {
      "epoch": 2.772475027746948,
      "grad_norm": 3.7725470066070557,
      "learning_rate": 9.425172678434382e-05,
      "loss": 0.7211872863769532,
      "step": 1250
    },
    {
      "epoch": 2.8279689234184238,
      "grad_norm": 3.9565186500549316,
      "learning_rate": 9.405986185725249e-05,
      "loss": 0.7074081420898437,
      "step": 1275
    },
    {
      "epoch": 2.8834628190899,
      "grad_norm": 2.8780977725982666,
      "learning_rate": 9.386799693016117e-05,
      "loss": 0.7226101684570313,
      "step": 1300
    },
    {
      "epoch": 2.9389567147613764,
      "grad_norm": 3.7520499229431152,
      "learning_rate": 9.367613200306984e-05,
      "loss": 0.7472096252441406,
      "step": 1325
    },
    {
      "epoch": 2.9944506104328523,
      "grad_norm": 3.8520190715789795,
      "learning_rate": 9.348426707597852e-05,
      "loss": 0.7333020782470703,
      "step": 1350
    },
    {
      "epoch": 3.048834628190899,
      "grad_norm": 3.3807711601257324,
      "learning_rate": 9.329240214888719e-05,
      "loss": 0.5055583190917968,
      "step": 1375
    },
    {
      "epoch": 3.104328523862375,
      "grad_norm": 2.786001205444336,
      "learning_rate": 9.310053722179586e-05,
      "loss": 0.5185424041748047,
      "step": 1400
    },
    {
      "epoch": 3.1598224195338513,
      "grad_norm": 2.996492385864258,
      "learning_rate": 9.290867229470454e-05,
      "loss": 0.48687358856201174,
      "step": 1425
    },
    {
      "epoch": 3.2153163152053272,
      "grad_norm": 3.1847450733184814,
      "learning_rate": 9.27168073676132e-05,
      "loss": 0.5263909530639649,
      "step": 1450
    },
    {
      "epoch": 3.2708102108768036,
      "grad_norm": 3.4472248554229736,
      "learning_rate": 9.252494244052187e-05,
      "loss": 0.4982139205932617,
      "step": 1475
    },
    {
      "epoch": 3.32630410654828,
      "grad_norm": 3.200753688812256,
      "learning_rate": 9.233307751343055e-05,
      "loss": 0.5312052917480469,
      "step": 1500
    },
    {
      "epoch": 3.32630410654828,
      "eval_cer": 0.25577363653143875,
      "eval_loss": 0.49039116501808167,
      "eval_runtime": 196.3736,
      "eval_samples_per_second": 8.963,
      "eval_steps_per_second": 0.143,
      "eval_wer": 0.7167754513859456,
      "step": 1500
    },
    {
      "epoch": 3.381798002219756,
      "grad_norm": 3.604504108428955,
      "learning_rate": 9.214121258633922e-05,
      "loss": 0.5276579666137695,
      "step": 1525
    },
    {
      "epoch": 3.437291897891232,
      "grad_norm": 3.206066131591797,
      "learning_rate": 9.19493476592479e-05,
      "loss": 0.5374650955200195,
      "step": 1550
    },
    {
      "epoch": 3.492785793562708,
      "grad_norm": 3.3676679134368896,
      "learning_rate": 9.175748273215656e-05,
      "loss": 0.5156414031982421,
      "step": 1575
    },
    {
      "epoch": 3.5482796892341844,
      "grad_norm": 2.790008068084717,
      "learning_rate": 9.156561780506524e-05,
      "loss": 0.5307263946533203,
      "step": 1600
    },
    {
      "epoch": 3.6037735849056602,
      "grad_norm": 3.1276090145111084,
      "learning_rate": 9.13737528779739e-05,
      "loss": 0.5411431884765625,
      "step": 1625
    },
    {
      "epoch": 3.6592674805771366,
      "grad_norm": 2.7448694705963135,
      "learning_rate": 9.118188795088259e-05,
      "loss": 0.5055676651000977,
      "step": 1650
    },
    {
      "epoch": 3.714761376248613,
      "grad_norm": 3.283252477645874,
      "learning_rate": 9.099002302379126e-05,
      "loss": 0.541312484741211,
      "step": 1675
    },
    {
      "epoch": 3.770255271920089,
      "grad_norm": 2.889359951019287,
      "learning_rate": 9.079815809669992e-05,
      "loss": 0.5509023666381836,
      "step": 1700
    },
    {
      "epoch": 3.8257491675915647,
      "grad_norm": 3.8324856758117676,
      "learning_rate": 9.06062931696086e-05,
      "loss": 0.554903221130371,
      "step": 1725
    },
    {
      "epoch": 3.881243063263041,
      "grad_norm": 2.673110008239746,
      "learning_rate": 9.041442824251727e-05,
      "loss": 0.5421103668212891,
      "step": 1750
    },
    {
      "epoch": 3.9367369589345174,
      "grad_norm": 3.1440036296844482,
      "learning_rate": 9.022256331542594e-05,
      "loss": 0.5445112609863281,
      "step": 1775
    },
    {
      "epoch": 3.9922308546059933,
      "grad_norm": 3.1932945251464844,
      "learning_rate": 9.003069838833462e-05,
      "loss": 0.5776439666748047,
      "step": 1800
    },
    {
      "epoch": 4.04661487236404,
      "grad_norm": 2.3801052570343018,
      "learning_rate": 8.983883346124329e-05,
      "loss": 0.3385904312133789,
      "step": 1825
    },
    {
      "epoch": 4.102108768035516,
      "grad_norm": 2.50673770904541,
      "learning_rate": 8.964696853415197e-05,
      "loss": 0.3545961380004883,
      "step": 1850
    },
    {
      "epoch": 4.157602663706992,
      "grad_norm": 2.9095983505249023,
      "learning_rate": 8.945510360706062e-05,
      "loss": 0.35681617736816407,
      "step": 1875
    },
    {
      "epoch": 4.213096559378468,
      "grad_norm": 2.819798707962036,
      "learning_rate": 8.92632386799693e-05,
      "loss": 0.33423614501953125,
      "step": 1900
    },
    {
      "epoch": 4.268590455049944,
      "grad_norm": 3.1799139976501465,
      "learning_rate": 8.907137375287797e-05,
      "loss": 0.34991111755371096,
      "step": 1925
    },
    {
      "epoch": 4.324084350721421,
      "grad_norm": 2.945661783218384,
      "learning_rate": 8.887950882578665e-05,
      "loss": 0.3595976257324219,
      "step": 1950
    },
    {
      "epoch": 4.379578246392897,
      "grad_norm": 3.115243673324585,
      "learning_rate": 8.868764389869532e-05,
      "loss": 0.37289775848388673,
      "step": 1975
    },
    {
      "epoch": 4.435072142064373,
      "grad_norm": 2.712531328201294,
      "learning_rate": 8.849577897160399e-05,
      "loss": 0.37404159545898436,
      "step": 2000
    },
    {
      "epoch": 4.435072142064373,
      "eval_cer": 0.26078632206414065,
      "eval_loss": 0.5362429022789001,
      "eval_runtime": 196.8231,
      "eval_samples_per_second": 8.942,
      "eval_steps_per_second": 0.142,
      "eval_wer": 0.7275578536916165,
      "step": 2000
    },
    {
      "epoch": 4.490566037735849,
      "grad_norm": 2.464198350906372,
      "learning_rate": 8.830391404451267e-05,
      "loss": 0.363293342590332,
      "step": 2025
    },
    {
      "epoch": 4.546059933407325,
      "grad_norm": 3.0920937061309814,
      "learning_rate": 8.811204911742134e-05,
      "loss": 0.3825681686401367,
      "step": 2050
    },
    {
      "epoch": 4.601553829078801,
      "grad_norm": 3.222235679626465,
      "learning_rate": 8.792018419033002e-05,
      "loss": 0.37105091094970705,
      "step": 2075
    },
    {
      "epoch": 4.657047724750278,
      "grad_norm": 2.993513584136963,
      "learning_rate": 8.772831926323869e-05,
      "loss": 0.3899434280395508,
      "step": 2100
    },
    {
      "epoch": 4.712541620421754,
      "grad_norm": 2.9703752994537354,
      "learning_rate": 8.753645433614735e-05,
      "loss": 0.3792466735839844,
      "step": 2125
    },
    {
      "epoch": 4.76803551609323,
      "grad_norm": 3.080484628677368,
      "learning_rate": 8.734458940905604e-05,
      "loss": 0.3823820877075195,
      "step": 2150
    },
    {
      "epoch": 4.823529411764706,
      "grad_norm": 2.8566219806671143,
      "learning_rate": 8.71527244819647e-05,
      "loss": 0.3843545150756836,
      "step": 2175
    },
    {
      "epoch": 4.879023307436182,
      "grad_norm": 2.913648843765259,
      "learning_rate": 8.696085955487337e-05,
      "loss": 0.3789427947998047,
      "step": 2200
    },
    {
      "epoch": 4.934517203107658,
      "grad_norm": 3.40332293510437,
      "learning_rate": 8.676899462778204e-05,
      "loss": 0.4051662063598633,
      "step": 2225
    },
    {
      "epoch": 4.990011098779134,
      "grad_norm": 2.9995148181915283,
      "learning_rate": 8.657712970069072e-05,
      "loss": 0.4019134902954102,
      "step": 2250
    },
    {
      "epoch": 5.0443951165371805,
      "grad_norm": 2.564155101776123,
      "learning_rate": 8.63852647735994e-05,
      "loss": 0.24176448822021485,
      "step": 2275
    },
    {
      "epoch": 5.099889012208657,
      "grad_norm": 2.846173048019409,
      "learning_rate": 8.619339984650806e-05,
      "loss": 0.21462602615356446,
      "step": 2300
    },
    {
      "epoch": 5.155382907880133,
      "grad_norm": 2.316955327987671,
      "learning_rate": 8.600153491941674e-05,
      "loss": 0.22590169906616211,
      "step": 2325
    },
    {
      "epoch": 5.210876803551609,
      "grad_norm": 2.8092174530029297,
      "learning_rate": 8.58096699923254e-05,
      "loss": 0.21187593460083007,
      "step": 2350
    },
    {
      "epoch": 5.266370699223086,
      "grad_norm": 2.7788586616516113,
      "learning_rate": 8.561780506523409e-05,
      "loss": 0.2337300682067871,
      "step": 2375
    },
    {
      "epoch": 5.321864594894562,
      "grad_norm": 2.503588914871216,
      "learning_rate": 8.542594013814275e-05,
      "loss": 0.23152971267700195,
      "step": 2400
    },
    {
      "epoch": 5.377358490566038,
      "grad_norm": 2.533194065093994,
      "learning_rate": 8.523407521105142e-05,
      "loss": 0.23667486190795897,
      "step": 2425
    },
    {
      "epoch": 5.4328523862375135,
      "grad_norm": 2.8751583099365234,
      "learning_rate": 8.50422102839601e-05,
      "loss": 0.22971002578735353,
      "step": 2450
    },
    {
      "epoch": 5.48834628190899,
      "grad_norm": 2.6748523712158203,
      "learning_rate": 8.485034535686877e-05,
      "loss": 0.2416807746887207,
      "step": 2475
    },
    {
      "epoch": 5.543840177580466,
      "grad_norm": 2.857468605041504,
      "learning_rate": 8.465848042977744e-05,
      "loss": 0.23854978561401366,
      "step": 2500
    },
    {
      "epoch": 5.543840177580466,
      "eval_cer": 0.2641594794045581,
      "eval_loss": 0.5750833749771118,
      "eval_runtime": 198.9608,
      "eval_samples_per_second": 8.846,
      "eval_steps_per_second": 0.141,
      "eval_wer": 0.723912859201492,
      "step": 2500
    }
  ],
  "logging_steps": 25,
  "max_steps": 13530,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 30,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 3,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 3
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.30565305585664e+19,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}