{
  "best_global_step": 2000,
  "best_metric": 2.3495261116052184,
  "best_model_checkpoint": "./SALAMA_NEW7/checkpoint-2000",
  "epoch": 0.798881565807869,
  "eval_steps": 2000,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.003994407829039345,
      "grad_norm": 3.857668399810791,
      "learning_rate": 1.8e-07,
      "loss": 0.0481,
      "step": 10
    },
    {
      "epoch": 0.00798881565807869,
      "grad_norm": 2.135991096496582,
      "learning_rate": 3.8e-07,
      "loss": 0.0317,
      "step": 20
    },
    {
      "epoch": 0.011983223487118035,
      "grad_norm": 2.663736343383789,
      "learning_rate": 5.800000000000001e-07,
      "loss": 0.0313,
      "step": 30
    },
    {
      "epoch": 0.01597763131615738,
      "grad_norm": 6.300492286682129,
      "learning_rate": 7.8e-07,
      "loss": 0.0442,
      "step": 40
    },
    {
      "epoch": 0.019972039145196723,
      "grad_norm": 3.308899164199829,
      "learning_rate": 9.800000000000001e-07,
      "loss": 0.0378,
      "step": 50
    },
    {
      "epoch": 0.02396644697423607,
      "grad_norm": 3.85744309425354,
      "learning_rate": 1.1800000000000001e-06,
      "loss": 0.0387,
      "step": 60
    },
    {
      "epoch": 0.027960854803275415,
      "grad_norm": 2.7074952125549316,
      "learning_rate": 1.3800000000000001e-06,
      "loss": 0.0411,
      "step": 70
    },
    {
      "epoch": 0.03195526263231476,
      "grad_norm": 1.499977707862854,
      "learning_rate": 1.5800000000000001e-06,
      "loss": 0.0373,
      "step": 80
    },
    {
      "epoch": 0.035949670461354104,
      "grad_norm": 3.291985273361206,
      "learning_rate": 1.7800000000000001e-06,
      "loss": 0.0291,
      "step": 90
    },
    {
      "epoch": 0.039944078290393446,
      "grad_norm": 5.130978107452393,
      "learning_rate": 1.98e-06,
      "loss": 0.035,
      "step": 100
    },
    {
      "epoch": 0.043938486119432796,
      "grad_norm": 3.7789242267608643,
      "learning_rate": 2.1800000000000003e-06,
      "loss": 0.0373,
      "step": 110
    },
    {
      "epoch": 0.04793289394847214,
      "grad_norm": 1.6493456363677979,
      "learning_rate": 2.38e-06,
      "loss": 0.0375,
      "step": 120
    },
    {
      "epoch": 0.05192730177751148,
      "grad_norm": 3.3285486698150635,
      "learning_rate": 2.5800000000000003e-06,
      "loss": 0.0358,
      "step": 130
    },
    {
      "epoch": 0.05592170960655083,
      "grad_norm": 2.768843650817871,
      "learning_rate": 2.7800000000000005e-06,
      "loss": 0.0447,
      "step": 140
    },
    {
      "epoch": 0.05991611743559017,
      "grad_norm": 4.878016471862793,
      "learning_rate": 2.9800000000000003e-06,
      "loss": 0.0388,
      "step": 150
    },
    {
      "epoch": 0.06391052526462952,
      "grad_norm": 4.028059959411621,
      "learning_rate": 3.1800000000000005e-06,
      "loss": 0.0398,
      "step": 160
    },
    {
      "epoch": 0.06790493309366886,
      "grad_norm": 3.255234956741333,
      "learning_rate": 3.3800000000000007e-06,
      "loss": 0.0355,
      "step": 170
    },
    {
      "epoch": 0.07189934092270821,
      "grad_norm": 3.8220138549804688,
      "learning_rate": 3.58e-06,
      "loss": 0.0403,
      "step": 180
    },
    {
      "epoch": 0.07589374875174755,
      "grad_norm": 1.4640872478485107,
      "learning_rate": 3.7800000000000002e-06,
      "loss": 0.0385,
      "step": 190
    },
    {
      "epoch": 0.07988815658078689,
      "grad_norm": 3.634939432144165,
      "learning_rate": 3.980000000000001e-06,
      "loss": 0.0468,
      "step": 200
    },
    {
      "epoch": 0.08388256440982625,
      "grad_norm": 3.357348918914795,
      "learning_rate": 4.18e-06,
      "loss": 0.0337,
      "step": 210
    },
    {
      "epoch": 0.08787697223886559,
      "grad_norm": 3.1094183921813965,
      "learning_rate": 4.38e-06,
      "loss": 0.0359,
      "step": 220
    },
    {
      "epoch": 0.09187138006790493,
      "grad_norm": 5.401888847351074,
      "learning_rate": 4.58e-06,
      "loss": 0.0628,
      "step": 230
    },
    {
      "epoch": 0.09586578789694428,
      "grad_norm": 3.3238024711608887,
      "learning_rate": 4.78e-06,
      "loss": 0.059,
      "step": 240
    },
    {
      "epoch": 0.09986019572598362,
      "grad_norm": 2.741633892059326,
      "learning_rate": 4.980000000000001e-06,
      "loss": 0.0408,
      "step": 250
    },
    {
      "epoch": 0.10385460355502296,
      "grad_norm": 2.8601815700531006,
      "learning_rate": 5.18e-06,
      "loss": 0.0328,
      "step": 260
    },
    {
      "epoch": 0.10784901138406232,
      "grad_norm": 2.5481112003326416,
      "learning_rate": 5.380000000000001e-06,
      "loss": 0.034,
      "step": 270
    },
    {
      "epoch": 0.11184341921310166,
      "grad_norm": 2.9293053150177,
      "learning_rate": 5.580000000000001e-06,
      "loss": 0.0345,
      "step": 280
    },
    {
      "epoch": 0.115837827042141,
      "grad_norm": 3.529905319213867,
      "learning_rate": 5.78e-06,
      "loss": 0.0361,
      "step": 290
    },
    {
      "epoch": 0.11983223487118035,
      "grad_norm": 3.364583969116211,
      "learning_rate": 5.98e-06,
      "loss": 0.0389,
      "step": 300
    },
    {
      "epoch": 0.12382664270021969,
      "grad_norm": 4.122386455535889,
      "learning_rate": 6.18e-06,
      "loss": 0.0588,
      "step": 310
    },
    {
      "epoch": 0.12782105052925904,
      "grad_norm": 4.145680904388428,
      "learning_rate": 6.380000000000001e-06,
      "loss": 0.0495,
      "step": 320
    },
    {
      "epoch": 0.1318154583582984,
      "grad_norm": 4.538552761077881,
      "learning_rate": 6.5800000000000005e-06,
      "loss": 0.0418,
      "step": 330
    },
    {
      "epoch": 0.13580986618733773,
      "grad_norm": 3.27231502532959,
      "learning_rate": 6.780000000000001e-06,
      "loss": 0.037,
      "step": 340
    },
    {
      "epoch": 0.13980427401637707,
      "grad_norm": 2.733621597290039,
      "learning_rate": 6.98e-06,
      "loss": 0.0401,
      "step": 350
    },
    {
      "epoch": 0.14379868184541642,
      "grad_norm": 3.7901926040649414,
      "learning_rate": 7.180000000000001e-06,
      "loss": 0.0414,
      "step": 360
    },
    {
      "epoch": 0.14779308967445576,
      "grad_norm": 3.2932889461517334,
      "learning_rate": 7.3800000000000005e-06,
      "loss": 0.04,
      "step": 370
    },
    {
      "epoch": 0.1517874975034951,
      "grad_norm": 2.3562254905700684,
      "learning_rate": 7.58e-06,
      "loss": 0.0424,
      "step": 380
    },
    {
      "epoch": 0.15578190533253444,
      "grad_norm": 3.2661616802215576,
      "learning_rate": 7.78e-06,
      "loss": 0.035,
      "step": 390
    },
    {
      "epoch": 0.15977631316157379,
      "grad_norm": 2.809990167617798,
      "learning_rate": 7.980000000000002e-06,
      "loss": 0.0404,
      "step": 400
    },
    {
      "epoch": 0.16377072099061313,
      "grad_norm": 3.149474620819092,
      "learning_rate": 8.18e-06,
      "loss": 0.0488,
      "step": 410
    },
    {
      "epoch": 0.1677651288196525,
      "grad_norm": 3.6923530101776123,
      "learning_rate": 8.380000000000001e-06,
      "loss": 0.0515,
      "step": 420
    },
    {
      "epoch": 0.17175953664869184,
      "grad_norm": 2.9962356090545654,
      "learning_rate": 8.580000000000001e-06,
      "loss": 0.0416,
      "step": 430
    },
    {
      "epoch": 0.17575394447773118,
      "grad_norm": 4.252007007598877,
      "learning_rate": 8.78e-06,
      "loss": 0.0448,
      "step": 440
    },
    {
      "epoch": 0.17974835230677053,
      "grad_norm": 2.9716057777404785,
      "learning_rate": 8.98e-06,
      "loss": 0.0368,
      "step": 450
    },
    {
      "epoch": 0.18374276013580987,
      "grad_norm": 3.339590072631836,
      "learning_rate": 9.180000000000002e-06,
      "loss": 0.041,
      "step": 460
    },
    {
      "epoch": 0.1877371679648492,
      "grad_norm": 4.591104507446289,
      "learning_rate": 9.38e-06,
      "loss": 0.0469,
      "step": 470
    },
    {
      "epoch": 0.19173157579388855,
      "grad_norm": 3.201215982437134,
      "learning_rate": 9.58e-06,
      "loss": 0.0403,
      "step": 480
    },
    {
      "epoch": 0.1957259836229279,
      "grad_norm": 3.1253278255462646,
      "learning_rate": 9.780000000000001e-06,
      "loss": 0.0428,
      "step": 490
    },
    {
      "epoch": 0.19972039145196724,
      "grad_norm": 2.59318208694458,
      "learning_rate": 9.980000000000001e-06,
      "loss": 0.0397,
      "step": 500
    },
    {
      "epoch": 0.20371479928100658,
      "grad_norm": 3.2373461723327637,
      "learning_rate": 9.98716486023959e-06,
      "loss": 0.0515,
      "step": 510
    },
    {
      "epoch": 0.20770920711004592,
      "grad_norm": 3.1935927867889404,
      "learning_rate": 9.972903593839133e-06,
      "loss": 0.0466,
      "step": 520
    },
    {
      "epoch": 0.2117036149390853,
      "grad_norm": 3.210402011871338,
      "learning_rate": 9.958642327438678e-06,
      "loss": 0.0441,
      "step": 530
    },
    {
      "epoch": 0.21569802276812464,
      "grad_norm": 2.792036533355713,
      "learning_rate": 9.944381061038221e-06,
      "loss": 0.0317,
      "step": 540
    },
    {
      "epoch": 0.21969243059716398,
      "grad_norm": 1.8875095844268799,
      "learning_rate": 9.930119794637765e-06,
      "loss": 0.035,
      "step": 550
    },
    {
      "epoch": 0.22368683842620332,
      "grad_norm": 4.238950252532959,
      "learning_rate": 9.91585852823731e-06,
      "loss": 0.0656,
      "step": 560
    },
    {
      "epoch": 0.22768124625524266,
      "grad_norm": 3.023404359817505,
      "learning_rate": 9.901597261836851e-06,
      "loss": 0.0511,
      "step": 570
    },
    {
      "epoch": 0.231675654084282,
      "grad_norm": 3.4290168285369873,
      "learning_rate": 9.887335995436396e-06,
      "loss": 0.0496,
      "step": 580
    },
    {
      "epoch": 0.23567006191332135,
      "grad_norm": 4.026778697967529,
      "learning_rate": 9.87307472903594e-06,
      "loss": 0.0394,
      "step": 590
    },
    {
      "epoch": 0.2396644697423607,
      "grad_norm": 3.315932512283325,
      "learning_rate": 9.858813462635483e-06,
      "loss": 0.044,
      "step": 600
    },
    {
      "epoch": 0.24365887757140003,
      "grad_norm": 2.7956554889678955,
      "learning_rate": 9.844552196235026e-06,
      "loss": 0.0414,
      "step": 610
    },
    {
      "epoch": 0.24765328540043938,
      "grad_norm": 4.850486755371094,
      "learning_rate": 9.83029092983457e-06,
      "loss": 0.0522,
      "step": 620
    },
    {
      "epoch": 0.2516476932294787,
      "grad_norm": 3.1265628337860107,
      "learning_rate": 9.816029663434114e-06,
      "loss": 0.0358,
      "step": 630
    },
    {
      "epoch": 0.2556421010585181,
      "grad_norm": 2.609273910522461,
      "learning_rate": 9.801768397033657e-06,
      "loss": 0.0397,
      "step": 640
    },
    {
      "epoch": 0.2596365088875574,
      "grad_norm": 4.073995590209961,
      "learning_rate": 9.787507130633202e-06,
      "loss": 0.0491,
      "step": 650
    },
    {
      "epoch": 0.2636309167165968,
      "grad_norm": 2.9083051681518555,
      "learning_rate": 9.773245864232744e-06,
      "loss": 0.034,
      "step": 660
    },
    {
      "epoch": 0.2676253245456361,
      "grad_norm": 4.328786849975586,
      "learning_rate": 9.758984597832289e-06,
      "loss": 0.0498,
      "step": 670
    },
    {
      "epoch": 0.27161973237467546,
      "grad_norm": 3.5342602729797363,
      "learning_rate": 9.744723331431832e-06,
      "loss": 0.0361,
      "step": 680
    },
    {
      "epoch": 0.2756141402037148,
      "grad_norm": 1.7005605697631836,
      "learning_rate": 9.730462065031375e-06,
      "loss": 0.0512,
      "step": 690
    },
    {
      "epoch": 0.27960854803275415,
      "grad_norm": 2.4722962379455566,
      "learning_rate": 9.71620079863092e-06,
      "loss": 0.0564,
      "step": 700
    },
    {
      "epoch": 0.2836029558617935,
      "grad_norm": 4.595683574676514,
      "learning_rate": 9.701939532230463e-06,
      "loss": 0.0394,
      "step": 710
    },
    {
      "epoch": 0.28759736369083283,
      "grad_norm": 4.185388088226318,
      "learning_rate": 9.687678265830007e-06,
      "loss": 0.0537,
      "step": 720
    },
    {
      "epoch": 0.2915917715198722,
      "grad_norm": 2.412919044494629,
      "learning_rate": 9.67341699942955e-06,
      "loss": 0.0479,
      "step": 730
    },
    {
      "epoch": 0.2955861793489115,
      "grad_norm": 3.0715839862823486,
      "learning_rate": 9.659155733029095e-06,
      "loss": 0.0524,
      "step": 740
    },
    {
      "epoch": 0.2995805871779509,
      "grad_norm": 3.375519275665283,
      "learning_rate": 9.644894466628636e-06,
      "loss": 0.0555,
      "step": 750
    },
    {
      "epoch": 0.3035749950069902,
      "grad_norm": 4.457313060760498,
      "learning_rate": 9.630633200228181e-06,
      "loss": 0.0502,
      "step": 760
    },
    {
      "epoch": 0.30756940283602957,
      "grad_norm": 2.67417573928833,
      "learning_rate": 9.616371933827725e-06,
      "loss": 0.0367,
      "step": 770
    },
    {
      "epoch": 0.3115638106650689,
      "grad_norm": 4.992038726806641,
      "learning_rate": 9.602110667427268e-06,
      "loss": 0.0459,
      "step": 780
    },
    {
      "epoch": 0.31555821849410826,
      "grad_norm": 3.235741138458252,
      "learning_rate": 9.587849401026813e-06,
      "loss": 0.0363,
      "step": 790
    },
    {
      "epoch": 0.31955262632314757,
      "grad_norm": 3.169708490371704,
      "learning_rate": 9.573588134626356e-06,
      "loss": 0.0441,
      "step": 800
    },
    {
      "epoch": 0.32354703415218694,
      "grad_norm": 3.8259077072143555,
      "learning_rate": 9.5593268682259e-06,
      "loss": 0.0449,
      "step": 810
    },
    {
      "epoch": 0.32754144198122626,
      "grad_norm": 2.690516948699951,
      "learning_rate": 9.545065601825442e-06,
      "loss": 0.0529,
      "step": 820
    },
    {
      "epoch": 0.3315358498102656,
      "grad_norm": 3.342012643814087,
      "learning_rate": 9.530804335424987e-06,
      "loss": 0.0381,
      "step": 830
    },
    {
      "epoch": 0.335530257639305,
      "grad_norm": 4.141724109649658,
      "learning_rate": 9.516543069024529e-06,
      "loss": 0.0505,
      "step": 840
    },
    {
      "epoch": 0.3395246654683443,
      "grad_norm": 6.260738372802734,
      "learning_rate": 9.502281802624074e-06,
      "loss": 0.0466,
      "step": 850
    },
    {
      "epoch": 0.3435190732973837,
      "grad_norm": 3.450930118560791,
      "learning_rate": 9.488020536223617e-06,
      "loss": 0.0499,
      "step": 860
    },
    {
      "epoch": 0.347513481126423,
      "grad_norm": 3.835606813430786,
      "learning_rate": 9.47375926982316e-06,
      "loss": 0.0698,
      "step": 870
    },
    {
      "epoch": 0.35150788895546237,
      "grad_norm": 5.289712429046631,
      "learning_rate": 9.459498003422705e-06,
      "loss": 0.0545,
      "step": 880
    },
    {
      "epoch": 0.3555022967845017,
      "grad_norm": 3.3518259525299072,
      "learning_rate": 9.445236737022249e-06,
      "loss": 0.05,
      "step": 890
    },
    {
      "epoch": 0.35949670461354105,
      "grad_norm": 2.9629595279693604,
      "learning_rate": 9.430975470621792e-06,
      "loss": 0.0458,
      "step": 900
    },
    {
      "epoch": 0.36349111244258037,
      "grad_norm": 4.516363620758057,
      "learning_rate": 9.416714204221335e-06,
      "loss": 0.0505,
      "step": 910
    },
    {
      "epoch": 0.36748552027161974,
      "grad_norm": 3.87882661819458,
      "learning_rate": 9.40245293782088e-06,
      "loss": 0.063,
      "step": 920
    },
    {
      "epoch": 0.37147992810065905,
      "grad_norm": 5.499217987060547,
      "learning_rate": 9.388191671420423e-06,
      "loss": 0.0516,
      "step": 930
    },
    {
      "epoch": 0.3754743359296984,
      "grad_norm": 2.694190740585327,
      "learning_rate": 9.373930405019966e-06,
      "loss": 0.0463,
      "step": 940
    },
    {
      "epoch": 0.3794687437587378,
      "grad_norm": 3.5198814868927,
      "learning_rate": 9.35966913861951e-06,
      "loss": 0.0504,
      "step": 950
    },
    {
      "epoch": 0.3834631515877771,
      "grad_norm": 3.1238009929656982,
      "learning_rate": 9.345407872219053e-06,
      "loss": 0.0371,
      "step": 960
    },
    {
      "epoch": 0.3874575594168165,
      "grad_norm": 4.268932819366455,
      "learning_rate": 9.331146605818598e-06,
      "loss": 0.0528,
      "step": 970
    },
    {
      "epoch": 0.3914519672458558,
      "grad_norm": 3.029754161834717,
      "learning_rate": 9.316885339418141e-06,
      "loss": 0.035,
      "step": 980
    },
    {
      "epoch": 0.39544637507489516,
      "grad_norm": 4.07280158996582,
      "learning_rate": 9.302624073017684e-06,
      "loss": 0.0372,
      "step": 990
    },
    {
      "epoch": 0.3994407829039345,
      "grad_norm": 2.339977741241455,
      "learning_rate": 9.288362806617228e-06,
      "loss": 0.0397,
      "step": 1000
    },
    {
      "epoch": 0.40343519073297385,
      "grad_norm": 4.685292720794678,
      "learning_rate": 9.274101540216773e-06,
      "loss": 0.0498,
      "step": 1010
    },
    {
      "epoch": 0.40742959856201316,
      "grad_norm": 3.811983346939087,
      "learning_rate": 9.259840273816316e-06,
      "loss": 0.042,
      "step": 1020
    },
    {
      "epoch": 0.41142400639105253,
      "grad_norm": 4.947136878967285,
      "learning_rate": 9.245579007415859e-06,
      "loss": 0.0509,
      "step": 1030
    },
    {
      "epoch": 0.41541841422009185,
      "grad_norm": 3.673635482788086,
      "learning_rate": 9.231317741015402e-06,
      "loss": 0.0465,
      "step": 1040
    },
    {
      "epoch": 0.4194128220491312,
      "grad_norm": 4.268950462341309,
      "learning_rate": 9.217056474614946e-06,
      "loss": 0.0461,
      "step": 1050
    },
    {
      "epoch": 0.4234072298781706,
      "grad_norm": 5.4644927978515625,
      "learning_rate": 9.20279520821449e-06,
      "loss": 0.0585,
      "step": 1060
    },
    {
      "epoch": 0.4274016377072099,
      "grad_norm": 2.472513437271118,
      "learning_rate": 9.188533941814034e-06,
      "loss": 0.0397,
      "step": 1070
    },
    {
      "epoch": 0.4313960455362493,
      "grad_norm": 5.088723659515381,
      "learning_rate": 9.174272675413579e-06,
      "loss": 0.0578,
      "step": 1080
    },
    {
      "epoch": 0.4353904533652886,
      "grad_norm": 5.186155796051025,
      "learning_rate": 9.16001140901312e-06,
      "loss": 0.0479,
      "step": 1090
    },
    {
      "epoch": 0.43938486119432796,
      "grad_norm": 2.81628155708313,
      "learning_rate": 9.145750142612665e-06,
      "loss": 0.0458,
      "step": 1100
    },
    {
      "epoch": 0.4433792690233673,
      "grad_norm": 2.2827396392822266,
      "learning_rate": 9.131488876212208e-06,
      "loss": 0.0395,
      "step": 1110
    },
    {
      "epoch": 0.44737367685240664,
      "grad_norm": 4.735171318054199,
      "learning_rate": 9.117227609811752e-06,
      "loss": 0.0512,
      "step": 1120
    },
    {
      "epoch": 0.45136808468144596,
      "grad_norm": 2.573768138885498,
      "learning_rate": 9.102966343411297e-06,
      "loss": 0.04,
      "step": 1130
    },
    {
      "epoch": 0.45536249251048533,
      "grad_norm": 3.1118862628936768,
      "learning_rate": 9.08870507701084e-06,
      "loss": 0.0662,
      "step": 1140
    },
    {
      "epoch": 0.45935690033952464,
      "grad_norm": 4.575135231018066,
      "learning_rate": 9.074443810610383e-06,
      "loss": 0.0412,
      "step": 1150
    },
    {
      "epoch": 0.463351308168564,
      "grad_norm": 3.333889961242676,
      "learning_rate": 9.060182544209926e-06,
      "loss": 0.0422,
      "step": 1160
    },
    {
      "epoch": 0.46734571599760333,
      "grad_norm": 4.380234241485596,
      "learning_rate": 9.045921277809471e-06,
      "loss": 0.0678,
      "step": 1170
    },
    {
      "epoch": 0.4713401238266427,
      "grad_norm": 3.1090171337127686,
      "learning_rate": 9.031660011409013e-06,
      "loss": 0.0419,
      "step": 1180
    },
    {
      "epoch": 0.47533453165568207,
      "grad_norm": 3.1724114418029785,
      "learning_rate": 9.017398745008558e-06,
      "loss": 0.0471,
      "step": 1190
    },
    {
      "epoch": 0.4793289394847214,
      "grad_norm": 3.701120138168335,
      "learning_rate": 9.003137478608101e-06,
      "loss": 0.0431,
      "step": 1200
    },
    {
      "epoch": 0.48332334731376075,
      "grad_norm": 1.9329041242599487,
      "learning_rate": 8.988876212207644e-06,
      "loss": 0.032,
      "step": 1210
    },
    {
      "epoch": 0.48731775514280007,
      "grad_norm": 4.282286167144775,
      "learning_rate": 8.97461494580719e-06,
      "loss": 0.0469,
      "step": 1220
    },
    {
      "epoch": 0.49131216297183944,
      "grad_norm": 2.6785926818847656,
      "learning_rate": 8.960353679406733e-06,
      "loss": 0.0643,
      "step": 1230
    },
    {
      "epoch": 0.49530657080087875,
      "grad_norm": 5.211093425750732,
      "learning_rate": 8.946092413006276e-06,
      "loss": 0.0579,
      "step": 1240
    },
    {
      "epoch": 0.4993009786299181,
      "grad_norm": 4.612974643707275,
      "learning_rate": 8.931831146605819e-06,
      "loss": 0.0618,
      "step": 1250
    },
    {
      "epoch": 0.5032953864589574,
      "grad_norm": 4.7832441329956055,
      "learning_rate": 8.917569880205364e-06,
      "loss": 0.0567,
      "step": 1260
    },
    {
      "epoch": 0.5072897942879968,
      "grad_norm": 3.8425779342651367,
      "learning_rate": 8.903308613804906e-06,
      "loss": 0.0489,
      "step": 1270
    },
    {
      "epoch": 0.5112842021170362,
      "grad_norm": 2.750922679901123,
      "learning_rate": 8.88904734740445e-06,
      "loss": 0.0438,
      "step": 1280
    },
    {
      "epoch": 0.5152786099460755,
      "grad_norm": 3.5208797454833984,
      "learning_rate": 8.874786081003994e-06,
      "loss": 0.0472,
      "step": 1290
    },
    {
      "epoch": 0.5192730177751148,
      "grad_norm": 5.142436981201172,
      "learning_rate": 8.860524814603537e-06,
      "loss": 0.0622,
      "step": 1300
    },
    {
      "epoch": 0.5232674256041542,
      "grad_norm": 3.287731885910034,
      "learning_rate": 8.846263548203082e-06,
      "loss": 0.0476,
      "step": 1310
    },
    {
      "epoch": 0.5272618334331935,
      "grad_norm": 1.9734798669815063,
      "learning_rate": 8.832002281802625e-06,
      "loss": 0.0485,
      "step": 1320
    },
    {
      "epoch": 0.5312562412622329,
      "grad_norm": 4.245563507080078,
      "learning_rate": 8.817741015402168e-06,
      "loss": 0.0572,
      "step": 1330
    },
    {
      "epoch": 0.5352506490912722,
      "grad_norm": 3.6561875343322754,
      "learning_rate": 8.803479749001712e-06,
      "loss": 0.05,
      "step": 1340
    },
    {
      "epoch": 0.5392450569203115,
      "grad_norm": 3.4770493507385254,
      "learning_rate": 8.789218482601257e-06,
      "loss": 0.0455,
      "step": 1350
    },
    {
      "epoch": 0.5432394647493509,
      "grad_norm": 2.5556652545928955,
      "learning_rate": 8.7749572162008e-06,
      "loss": 0.0396,
      "step": 1360
    },
    {
      "epoch": 0.5472338725783903,
      "grad_norm": 3.5269246101379395,
      "learning_rate": 8.760695949800343e-06,
      "loss": 0.0435,
      "step": 1370
    },
    {
      "epoch": 0.5512282804074295,
      "grad_norm": 3.7632479667663574,
      "learning_rate": 8.746434683399886e-06,
      "loss": 0.0458,
      "step": 1380
    },
    {
      "epoch": 0.5552226882364689,
      "grad_norm": 4.281711578369141,
      "learning_rate": 8.73217341699943e-06,
      "loss": 0.0511,
      "step": 1390
    },
    {
      "epoch": 0.5592170960655083,
      "grad_norm": 3.2763116359710693,
      "learning_rate": 8.717912150598975e-06,
      "loss": 0.0405,
      "step": 1400
    },
    {
      "epoch": 0.5632115038945477,
      "grad_norm": 3.267866849899292,
      "learning_rate": 8.703650884198518e-06,
      "loss": 0.0393,
      "step": 1410
    },
    {
      "epoch": 0.567205911723587,
      "grad_norm": 3.6679697036743164,
      "learning_rate": 8.689389617798061e-06,
      "loss": 0.06,
      "step": 1420
    },
    {
      "epoch": 0.5712003195526263,
      "grad_norm": 2.4105310440063477,
      "learning_rate": 8.675128351397604e-06,
      "loss": 0.0406,
      "step": 1430
    },
    {
      "epoch": 0.5751947273816657,
      "grad_norm": 4.512986183166504,
      "learning_rate": 8.66086708499715e-06,
      "loss": 0.0522,
      "step": 1440
    },
    {
      "epoch": 0.579189135210705,
      "grad_norm": 2.781480550765991,
      "learning_rate": 8.646605818596692e-06,
      "loss": 0.0575,
      "step": 1450
    },
    {
      "epoch": 0.5831835430397444,
      "grad_norm": 2.455273151397705,
      "learning_rate": 8.632344552196236e-06,
      "loss": 0.0515,
      "step": 1460
    },
    {
      "epoch": 0.5871779508687837,
      "grad_norm": 3.5020835399627686,
      "learning_rate": 8.618083285795779e-06,
      "loss": 0.0493,
      "step": 1470
    },
    {
      "epoch": 0.591172358697823,
      "grad_norm": 2.549257278442383,
      "learning_rate": 8.603822019395322e-06,
      "loss": 0.0305,
      "step": 1480
    },
    {
      "epoch": 0.5951667665268624,
      "grad_norm": 3.308424949645996,
      "learning_rate": 8.589560752994867e-06,
      "loss": 0.0394,
      "step": 1490
    },
    {
      "epoch": 0.5991611743559018,
      "grad_norm": 4.140067100524902,
      "learning_rate": 8.57529948659441e-06,
      "loss": 0.0452,
      "step": 1500
    },
    {
      "epoch": 0.603155582184941,
      "grad_norm": 4.411773204803467,
      "learning_rate": 8.561038220193954e-06,
      "loss": 0.0564,
      "step": 1510
    },
    {
      "epoch": 0.6071499900139804,
      "grad_norm": 2.6754167079925537,
      "learning_rate": 8.546776953793497e-06,
      "loss": 0.0398,
      "step": 1520
    },
    {
      "epoch": 0.6111443978430198,
      "grad_norm": 2.152740001678467,
      "learning_rate": 8.532515687393042e-06,
      "loss": 0.0417,
      "step": 1530
    },
    {
      "epoch": 0.6151388056720591,
      "grad_norm": 2.487994909286499,
      "learning_rate": 8.518254420992585e-06,
      "loss": 0.0475,
      "step": 1540
    },
    {
      "epoch": 0.6191332135010985,
      "grad_norm": 1.9966262578964233,
      "learning_rate": 8.503993154592128e-06,
      "loss": 0.0456,
      "step": 1550
    },
    {
      "epoch": 0.6231276213301378,
      "grad_norm": 2.2700722217559814,
      "learning_rate": 8.489731888191672e-06,
      "loss": 0.0498,
      "step": 1560
    },
    {
      "epoch": 0.6271220291591771,
      "grad_norm": 4.674302101135254,
      "learning_rate": 8.475470621791215e-06,
      "loss": 0.0531,
      "step": 1570
    },
    {
      "epoch": 0.6311164369882165,
      "grad_norm": 3.2880172729492188,
      "learning_rate": 8.46120935539076e-06,
      "loss": 0.0418,
      "step": 1580
    },
    {
      "epoch": 0.6351108448172559,
      "grad_norm": 4.65671968460083,
      "learning_rate": 8.446948088990303e-06,
      "loss": 0.0429,
      "step": 1590
    },
    {
      "epoch": 0.6391052526462951,
      "grad_norm": 2.744880199432373,
      "learning_rate": 8.432686822589846e-06,
      "loss": 0.038,
      "step": 1600
    },
    {
      "epoch": 0.6430996604753345,
      "grad_norm": 3.2633752822875977,
      "learning_rate": 8.41842555618939e-06,
      "loss": 0.0508,
      "step": 1610
    },
    {
      "epoch": 0.6470940683043739,
      "grad_norm": 4.094485759735107,
      "learning_rate": 8.404164289788934e-06,
      "loss": 0.0482,
      "step": 1620
    },
    {
      "epoch": 0.6510884761334133,
      "grad_norm": 3.6843981742858887,
      "learning_rate": 8.389903023388478e-06,
      "loss": 0.0426,
      "step": 1630
    },
    {
      "epoch": 0.6550828839624525,
      "grad_norm": 2.887274742126465,
      "learning_rate": 8.375641756988021e-06,
      "loss": 0.051,
      "step": 1640
    },
    {
      "epoch": 0.6590772917914919,
      "grad_norm": 3.3279778957366943,
      "learning_rate": 8.361380490587566e-06,
      "loss": 0.0378,
      "step": 1650
    },
    {
      "epoch": 0.6630716996205313,
      "grad_norm": 3.5682291984558105,
      "learning_rate": 8.347119224187107e-06,
      "loss": 0.054,
      "step": 1660
    },
    {
      "epoch": 0.6670661074495706,
      "grad_norm": 4.352251052856445,
      "learning_rate": 8.332857957786652e-06,
      "loss": 0.0431,
      "step": 1670
    },
    {
      "epoch": 0.67106051527861,
      "grad_norm": 2.8711681365966797,
      "learning_rate": 8.318596691386196e-06,
      "loss": 0.0351,
      "step": 1680
    },
    {
      "epoch": 0.6750549231076493,
      "grad_norm": 3.0659475326538086,
      "learning_rate": 8.304335424985739e-06,
      "loss": 0.0413,
      "step": 1690
    },
    {
      "epoch": 0.6790493309366886,
      "grad_norm": 5.653151988983154,
      "learning_rate": 8.290074158585282e-06,
      "loss": 0.0526,
      "step": 1700
    },
    {
      "epoch": 0.683043738765728,
      "grad_norm": 7.713840007781982,
      "learning_rate": 8.275812892184827e-06,
      "loss": 0.0495,
      "step": 1710
    },
    {
      "epoch": 0.6870381465947674,
      "grad_norm": 2.798320770263672,
      "learning_rate": 8.26155162578437e-06,
      "loss": 0.0436,
      "step": 1720
    },
    {
      "epoch": 0.6910325544238066,
      "grad_norm": 3.696972131729126,
      "learning_rate": 8.247290359383914e-06,
      "loss": 0.0413,
      "step": 1730
    },
    {
      "epoch": 0.695026962252846,
      "grad_norm": 2.2935080528259277,
      "learning_rate": 8.233029092983458e-06,
      "loss": 0.0517,
      "step": 1740
    },
    {
      "epoch": 0.6990213700818854,
      "grad_norm": 4.169670581817627,
      "learning_rate": 8.218767826583002e-06,
      "loss": 0.0499,
      "step": 1750
    },
    {
      "epoch": 0.7030157779109247,
      "grad_norm": 4.221024036407471,
      "learning_rate": 8.204506560182545e-06,
      "loss": 0.041,
      "step": 1760
    },
    {
      "epoch": 0.7070101857399641,
      "grad_norm": 4.128349781036377,
      "learning_rate": 8.190245293782088e-06,
      "loss": 0.0598,
      "step": 1770
    },
    {
      "epoch": 0.7110045935690034,
      "grad_norm": 3.2227325439453125,
      "learning_rate": 8.175984027381633e-06,
      "loss": 0.0468,
      "step": 1780
    },
    {
      "epoch": 0.7149990013980427,
      "grad_norm": 2.807412624359131,
      "learning_rate": 8.161722760981175e-06,
      "loss": 0.0419,
      "step": 1790
    },
    {
      "epoch": 0.7189934092270821,
      "grad_norm": 3.685307741165161,
      "learning_rate": 8.14746149458072e-06,
      "loss": 0.05,
      "step": 1800
    },
    {
      "epoch": 0.7229878170561215,
      "grad_norm": 3.1187121868133545,
      "learning_rate": 8.133200228180263e-06,
      "loss": 0.0483,
      "step": 1810
    },
    {
      "epoch": 0.7269822248851607,
      "grad_norm": 3.4114596843719482,
      "learning_rate": 8.118938961779806e-06,
      "loss": 0.0494,
      "step": 1820
    },
    {
      "epoch": 0.7309766327142001,
      "grad_norm": 1.188926100730896,
      "learning_rate": 8.104677695379351e-06,
      "loss": 0.0366,
      "step": 1830
    },
    {
      "epoch": 0.7349710405432395,
      "grad_norm": 3.5981297492980957,
      "learning_rate": 8.090416428978894e-06,
      "loss": 0.0503,
      "step": 1840
    },
    {
      "epoch": 0.7389654483722788,
      "grad_norm": 3.0543458461761475,
      "learning_rate": 8.076155162578438e-06,
      "loss": 0.0495,
      "step": 1850
    },
    {
      "epoch": 0.7429598562013181,
      "grad_norm": 3.7244555950164795,
      "learning_rate": 8.06189389617798e-06,
      "loss": 0.0496,
      "step": 1860
    },
    {
      "epoch": 0.7469542640303575,
      "grad_norm": 2.367532968521118,
      "learning_rate": 8.047632629777526e-06,
      "loss": 0.0418,
      "step": 1870
    },
    {
      "epoch": 0.7509486718593968,
      "grad_norm": 2.6962149143218994,
      "learning_rate": 8.033371363377069e-06,
      "loss": 0.0379,
      "step": 1880
    },
    {
      "epoch": 0.7549430796884362,
      "grad_norm": 3.7440872192382812,
      "learning_rate": 8.019110096976612e-06,
      "loss": 0.0411,
      "step": 1890
    },
    {
      "epoch": 0.7589374875174756,
      "grad_norm": 3.3427021503448486,
      "learning_rate": 8.004848830576156e-06,
      "loss": 0.0367,
      "step": 1900
    },
    {
      "epoch": 0.7629318953465148,
      "grad_norm": 3.3870105743408203,
      "learning_rate": 7.990587564175699e-06,
      "loss": 0.0575,
      "step": 1910
    },
    {
      "epoch": 0.7669263031755542,
      "grad_norm": 3.3747177124023438,
      "learning_rate": 7.976326297775244e-06,
      "loss": 0.0455,
      "step": 1920
    },
    {
      "epoch": 0.7709207110045936,
      "grad_norm": 2.9369425773620605,
      "learning_rate": 7.962065031374787e-06,
      "loss": 0.0474,
      "step": 1930
    },
    {
      "epoch": 0.774915118833633,
      "grad_norm": 2.0219178199768066,
      "learning_rate": 7.94780376497433e-06,
      "loss": 0.0496,
      "step": 1940
    },
    {
      "epoch": 0.7789095266626722,
      "grad_norm": 3.041201591491699,
      "learning_rate": 7.933542498573873e-06,
      "loss": 0.0386,
      "step": 1950
    },
    {
      "epoch": 0.7829039344917116,
      "grad_norm": 3.874758243560791,
      "learning_rate": 7.919281232173418e-06,
      "loss": 0.0331,
      "step": 1960
    },
    {
      "epoch": 0.786898342320751,
      "grad_norm": 2.3752293586730957,
      "learning_rate": 7.905019965772962e-06,
      "loss": 0.0477,
      "step": 1970
    },
    {
      "epoch": 0.7908927501497903,
      "grad_norm": 4.09476900100708,
      "learning_rate": 7.890758699372505e-06,
      "loss": 0.051,
      "step": 1980
    },
    {
      "epoch": 0.7948871579788296,
      "grad_norm": 1.380469799041748,
      "learning_rate": 7.876497432972048e-06,
      "loss": 0.0313,
      "step": 1990
    },
    {
      "epoch": 0.798881565807869,
      "grad_norm": 2.8417437076568604,
      "learning_rate": 7.862236166571591e-06,
      "loss": 0.0444,
      "step": 2000
    },
    {
      "epoch": 0.798881565807869,
      "eval_loss": 0.02616371586918831,
      "eval_runtime": 9253.1401,
      "eval_samples_per_second": 2.164,
      "eval_steps_per_second": 0.271,
      "eval_wer": 2.3495261116052184,
      "step": 2000
    }
  ],
  "logging_steps": 10,
  "max_steps": 7512,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 2000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.846946562048e+19,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}