{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 30.0,
  "eval_steps": 500,
  "global_step": 5370,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.055865921787709494,
      "grad_norm": 9.474554061889648,
      "learning_rate": 5.018587360594795e-06,
      "loss": 0.9647,
      "step": 10
    },
    {
      "epoch": 0.11173184357541899,
      "grad_norm": 2.904832124710083,
      "learning_rate": 1.0594795539033457e-05,
      "loss": 0.3646,
      "step": 20
    },
    {
      "epoch": 0.16759776536312848,
      "grad_norm": 2.1869466304779053,
      "learning_rate": 1.617100371747212e-05,
      "loss": 0.2295,
      "step": 30
    },
    {
      "epoch": 0.22346368715083798,
      "grad_norm": 1.0514620542526245,
      "learning_rate": 2.1747211895910778e-05,
      "loss": 0.1933,
      "step": 40
    },
    {
      "epoch": 0.27932960893854747,
      "grad_norm": 1.2703338861465454,
      "learning_rate": 2.732342007434944e-05,
      "loss": 0.16,
      "step": 50
    },
    {
      "epoch": 0.33519553072625696,
      "grad_norm": 1.139123797416687,
      "learning_rate": 3.28996282527881e-05,
      "loss": 0.1487,
      "step": 60
    },
    {
      "epoch": 0.39106145251396646,
      "grad_norm": 1.1433963775634766,
      "learning_rate": 3.8475836431226764e-05,
      "loss": 0.1484,
      "step": 70
    },
    {
      "epoch": 0.44692737430167595,
      "grad_norm": 1.1463357210159302,
      "learning_rate": 4.4052044609665426e-05,
      "loss": 0.1315,
      "step": 80
    },
    {
      "epoch": 0.5027932960893855,
      "grad_norm": 1.1672674417495728,
      "learning_rate": 4.962825278810409e-05,
      "loss": 0.1267,
      "step": 90
    },
    {
      "epoch": 0.5586592178770949,
      "grad_norm": 1.0597262382507324,
      "learning_rate": 5.520446096654275e-05,
      "loss": 0.1244,
      "step": 100
    },
    {
      "epoch": 0.6145251396648045,
      "grad_norm": 1.1038429737091064,
      "learning_rate": 6.0780669144981406e-05,
      "loss": 0.1147,
      "step": 110
    },
    {
      "epoch": 0.6703910614525139,
      "grad_norm": 0.6325312852859497,
      "learning_rate": 6.635687732342006e-05,
      "loss": 0.1192,
      "step": 120
    },
    {
      "epoch": 0.7262569832402235,
      "grad_norm": 1.0795389413833618,
      "learning_rate": 7.193308550185873e-05,
      "loss": 0.1105,
      "step": 130
    },
    {
      "epoch": 0.7821229050279329,
      "grad_norm": 0.7824896574020386,
      "learning_rate": 7.750929368029739e-05,
      "loss": 0.1035,
      "step": 140
    },
    {
      "epoch": 0.8379888268156425,
      "grad_norm": 1.3778011798858643,
      "learning_rate": 8.308550185873605e-05,
      "loss": 0.1069,
      "step": 150
    },
    {
      "epoch": 0.8938547486033519,
      "grad_norm": 0.8529579639434814,
      "learning_rate": 8.866171003717471e-05,
      "loss": 0.1022,
      "step": 160
    },
    {
      "epoch": 0.9497206703910615,
      "grad_norm": 1.441953182220459,
      "learning_rate": 9.423791821561338e-05,
      "loss": 0.0985,
      "step": 170
    },
    {
      "epoch": 1.005586592178771,
      "grad_norm": 0.7136903405189514,
      "learning_rate": 9.981412639405203e-05,
      "loss": 0.1026,
      "step": 180
    },
    {
      "epoch": 1.0614525139664805,
      "grad_norm": 1.4089112281799316,
      "learning_rate": 0.0001053903345724907,
      "loss": 0.0973,
      "step": 190
    },
    {
      "epoch": 1.1173184357541899,
      "grad_norm": 0.8666443228721619,
      "learning_rate": 0.00011096654275092937,
      "loss": 0.0899,
      "step": 200
    },
    {
      "epoch": 1.1731843575418994,
      "grad_norm": 0.6051232218742371,
      "learning_rate": 0.00011654275092936801,
      "loss": 0.0891,
      "step": 210
    },
    {
      "epoch": 1.229050279329609,
      "grad_norm": 0.5882899165153503,
      "learning_rate": 0.00012211895910780668,
      "loss": 0.0859,
      "step": 220
    },
    {
      "epoch": 1.2849162011173183,
      "grad_norm": 0.662291407585144,
      "learning_rate": 0.00012769516728624534,
      "loss": 0.0835,
      "step": 230
    },
    {
      "epoch": 1.3407821229050279,
      "grad_norm": 1.036959171295166,
      "learning_rate": 0.000133271375464684,
      "loss": 0.0801,
      "step": 240
    },
    {
      "epoch": 1.3966480446927374,
      "grad_norm": 0.5812122225761414,
      "learning_rate": 0.00013884758364312265,
      "loss": 0.0723,
      "step": 250
    },
    {
      "epoch": 1.452513966480447,
      "grad_norm": 0.799187183380127,
      "learning_rate": 0.00014442379182156133,
      "loss": 0.0701,
      "step": 260
    },
    {
      "epoch": 1.5083798882681565,
      "grad_norm": 1.0647534132003784,
      "learning_rate": 0.00015,
      "loss": 0.075,
      "step": 270
    },
    {
      "epoch": 1.564245810055866,
      "grad_norm": 0.9664549827575684,
      "learning_rate": 0.0001499985776090078,
      "loss": 0.0696,
      "step": 280
    },
    {
      "epoch": 1.6201117318435754,
      "grad_norm": 0.621644139289856,
      "learning_rate": 0.0001499943104899832,
      "loss": 0.0657,
      "step": 290
    },
    {
      "epoch": 1.675977653631285,
      "grad_norm": 0.6787592768669128,
      "learning_rate": 0.00014998719880477977,
      "loss": 0.0731,
      "step": 300
    },
    {
      "epoch": 1.7318435754189943,
      "grad_norm": 0.563049852848053,
      "learning_rate": 0.00014997724282314682,
      "loss": 0.0707,
      "step": 310
    },
    {
      "epoch": 1.7877094972067038,
      "grad_norm": 0.4710414409637451,
      "learning_rate": 0.00014996444292271892,
      "loss": 0.0657,
      "step": 320
    },
    {
      "epoch": 1.8435754189944134,
      "grad_norm": 0.46238359808921814,
      "learning_rate": 0.0001499487995890018,
      "loss": 0.0649,
      "step": 330
    },
    {
      "epoch": 1.899441340782123,
      "grad_norm": 0.4369705021381378,
      "learning_rate": 0.00014993031341535375,
      "loss": 0.0562,
      "step": 340
    },
    {
      "epoch": 1.9553072625698324,
      "grad_norm": 0.5079113841056824,
      "learning_rate": 0.0001499089851029632,
      "loss": 0.0553,
      "step": 350
    },
    {
      "epoch": 2.011173184357542,
      "grad_norm": 0.7058802247047424,
      "learning_rate": 0.00014988481546082222,
      "loss": 0.0551,
      "step": 360
    },
    {
      "epoch": 2.0670391061452515,
      "grad_norm": 0.42536088824272156,
      "learning_rate": 0.0001498578054056956,
      "loss": 0.0553,
      "step": 370
    },
    {
      "epoch": 2.122905027932961,
      "grad_norm": 0.5014165043830872,
      "learning_rate": 0.00014982795596208618,
      "loss": 0.0539,
      "step": 380
    },
    {
      "epoch": 2.17877094972067,
      "grad_norm": 0.6823521852493286,
      "learning_rate": 0.0001497952682621962,
      "loss": 0.0507,
      "step": 390
    },
    {
      "epoch": 2.2346368715083798,
      "grad_norm": 0.4654916822910309,
      "learning_rate": 0.00014975974354588395,
      "loss": 0.058,
      "step": 400
    },
    {
      "epoch": 2.2905027932960893,
      "grad_norm": 0.33408060669898987,
      "learning_rate": 0.00014972138316061715,
      "loss": 0.0484,
      "step": 410
    },
    {
      "epoch": 2.346368715083799,
      "grad_norm": 0.40119659900665283,
      "learning_rate": 0.00014968018856142154,
      "loss": 0.051,
      "step": 420
    },
    {
      "epoch": 2.4022346368715084,
      "grad_norm": 0.3584083616733551,
      "learning_rate": 0.00014963616131082584,
      "loss": 0.0494,
      "step": 430
    },
    {
      "epoch": 2.458100558659218,
      "grad_norm": 0.6725554466247559,
      "learning_rate": 0.00014958930307880242,
      "loss": 0.0572,
      "step": 440
    },
    {
      "epoch": 2.5139664804469275,
      "grad_norm": 0.582353413105011,
      "learning_rate": 0.00014953961564270404,
      "loss": 0.0495,
      "step": 450
    },
    {
      "epoch": 2.5698324022346366,
      "grad_norm": 0.6442360877990723,
      "learning_rate": 0.00014948710088719632,
      "loss": 0.0581,
      "step": 460
    },
    {
      "epoch": 2.6256983240223466,
      "grad_norm": 0.6121304631233215,
      "learning_rate": 0.00014943176080418633,
      "loss": 0.0528,
      "step": 470
    },
    {
      "epoch": 2.6815642458100557,
      "grad_norm": 0.6058341860771179,
      "learning_rate": 0.00014937359749274705,
      "loss": 0.0504,
      "step": 480
    },
    {
      "epoch": 2.7374301675977653,
      "grad_norm": 0.4608060121536255,
      "learning_rate": 0.00014931261315903765,
      "loss": 0.0501,
      "step": 490
    },
    {
      "epoch": 2.793296089385475,
      "grad_norm": 0.33580687642097473,
      "learning_rate": 0.00014924881011621992,
      "loss": 0.0441,
      "step": 500
    },
    {
      "epoch": 2.8491620111731844,
      "grad_norm": 0.5434100031852722,
      "learning_rate": 0.0001491821907843705,
      "loss": 0.0469,
      "step": 510
    },
    {
      "epoch": 2.905027932960894,
      "grad_norm": 0.7853328585624695,
      "learning_rate": 0.00014911275769038902,
      "loss": 0.0435,
      "step": 520
    },
    {
      "epoch": 2.9608938547486034,
      "grad_norm": 0.6744305491447449,
      "learning_rate": 0.0001490405134679024,
      "loss": 0.0477,
      "step": 530
    },
    {
      "epoch": 3.016759776536313,
      "grad_norm": 0.6475287079811096,
      "learning_rate": 0.00014896546085716475,
      "loss": 0.0468,
      "step": 540
    },
    {
      "epoch": 3.0726256983240225,
      "grad_norm": 0.5036594271659851,
      "learning_rate": 0.00014888760270495365,
      "loss": 0.044,
      "step": 550
    },
    {
      "epoch": 3.1284916201117317,
      "grad_norm": 0.4917600452899933,
      "learning_rate": 0.000148806941964462,
      "loss": 0.0478,
      "step": 560
    },
    {
      "epoch": 3.184357541899441,
      "grad_norm": 0.5577938556671143,
      "learning_rate": 0.0001487234816951861,
      "loss": 0.0464,
      "step": 570
    },
    {
      "epoch": 3.2402234636871508,
      "grad_norm": 0.452462375164032,
      "learning_rate": 0.0001486372250628095,
      "loss": 0.041,
      "step": 580
    },
    {
      "epoch": 3.2960893854748603,
      "grad_norm": 0.3954116702079773,
      "learning_rate": 0.00014854817533908313,
      "loss": 0.0437,
      "step": 590
    },
    {
      "epoch": 3.35195530726257,
      "grad_norm": 0.529547393321991,
      "learning_rate": 0.00014845633590170092,
      "loss": 0.0433,
      "step": 600
    },
    {
      "epoch": 3.4078212290502794,
      "grad_norm": 0.3528311848640442,
      "learning_rate": 0.00014836171023417191,
      "loss": 0.0365,
      "step": 610
    },
    {
      "epoch": 3.463687150837989,
      "grad_norm": 0.4728206992149353,
      "learning_rate": 0.00014826430192568807,
      "loss": 0.0457,
      "step": 620
    },
    {
      "epoch": 3.5195530726256985,
      "grad_norm": 0.6676551103591919,
      "learning_rate": 0.00014816411467098806,
      "loss": 0.045,
      "step": 630
    },
    {
      "epoch": 3.5754189944134076,
      "grad_norm": 0.3946695327758789,
      "learning_rate": 0.00014806115227021714,
      "loss": 0.0465,
      "step": 640
    },
    {
      "epoch": 3.631284916201117,
      "grad_norm": 0.6944833993911743,
      "learning_rate": 0.00014795541862878308,
      "loss": 0.0481,
      "step": 650
    },
    {
      "epoch": 3.6871508379888267,
      "grad_norm": 0.6339210271835327,
      "learning_rate": 0.00014784691775720807,
      "loss": 0.0394,
      "step": 660
    },
    {
      "epoch": 3.7430167597765363,
      "grad_norm": 0.47451841831207275,
      "learning_rate": 0.00014773565377097631,
      "loss": 0.0392,
      "step": 670
    },
    {
      "epoch": 3.798882681564246,
      "grad_norm": 0.3181611895561218,
      "learning_rate": 0.0001476216308903784,
      "loss": 0.0436,
      "step": 680
    },
    {
      "epoch": 3.8547486033519553,
      "grad_norm": 0.4814484715461731,
      "learning_rate": 0.00014750485344035071,
      "loss": 0.0383,
      "step": 690
    },
    {
      "epoch": 3.910614525139665,
      "grad_norm": 0.3436708450317383,
      "learning_rate": 0.00014738532585031178,
      "loss": 0.044,
      "step": 700
    },
    {
      "epoch": 3.9664804469273744,
      "grad_norm": 0.367412269115448,
      "learning_rate": 0.00014726305265399403,
      "loss": 0.0403,
      "step": 710
    },
    {
      "epoch": 4.022346368715084,
      "grad_norm": 0.23561592400074005,
      "learning_rate": 0.00014713803848927196,
      "loss": 0.0422,
      "step": 720
    },
    {
      "epoch": 4.078212290502793,
      "grad_norm": 0.33305707573890686,
      "learning_rate": 0.00014701028809798619,
      "loss": 0.0431,
      "step": 730
    },
    {
      "epoch": 4.134078212290503,
      "grad_norm": 0.43914180994033813,
      "learning_rate": 0.00014687980632576347,
      "loss": 0.0432,
      "step": 740
    },
    {
      "epoch": 4.189944134078212,
      "grad_norm": 0.41326114535331726,
      "learning_rate": 0.0001467465981218331,
      "loss": 0.0408,
      "step": 750
    },
    {
      "epoch": 4.245810055865922,
      "grad_norm": 0.3312179148197174,
      "learning_rate": 0.0001466106685388391,
      "loss": 0.0437,
      "step": 760
    },
    {
      "epoch": 4.301675977653631,
      "grad_norm": 0.24370786547660828,
      "learning_rate": 0.00014647202273264848,
      "loss": 0.0381,
      "step": 770
    },
    {
      "epoch": 4.35754189944134,
      "grad_norm": 0.3902948796749115,
      "learning_rate": 0.00014633066596215577,
      "loss": 0.039,
      "step": 780
    },
    {
      "epoch": 4.41340782122905,
      "grad_norm": 0.48275288939476013,
      "learning_rate": 0.0001461866035890836,
      "loss": 0.0395,
      "step": 790
    },
    {
      "epoch": 4.4692737430167595,
      "grad_norm": 0.45421525835990906,
      "learning_rate": 0.00014603984107777924,
      "loss": 0.0386,
      "step": 800
    },
    {
      "epoch": 4.5251396648044695,
      "grad_norm": 0.2560478150844574,
      "learning_rate": 0.00014589038399500726,
      "loss": 0.0375,
      "step": 810
    },
    {
      "epoch": 4.581005586592179,
      "grad_norm": 0.3527555465698242,
      "learning_rate": 0.0001457382380097386,
      "loss": 0.046,
      "step": 820
    },
    {
      "epoch": 4.636871508379889,
      "grad_norm": 0.5826641917228699,
      "learning_rate": 0.00014558340889293534,
      "loss": 0.0419,
      "step": 830
    },
    {
      "epoch": 4.692737430167598,
      "grad_norm": 0.5426205992698669,
      "learning_rate": 0.00014542590251733192,
      "loss": 0.0449,
      "step": 840
    },
    {
      "epoch": 4.748603351955307,
      "grad_norm": 0.4121716618537903,
      "learning_rate": 0.00014526572485721234,
      "loss": 0.0404,
      "step": 850
    },
    {
      "epoch": 4.804469273743017,
      "grad_norm": 0.41459986567497253,
      "learning_rate": 0.00014510288198818356,
      "loss": 0.0434,
      "step": 860
    },
    {
      "epoch": 4.860335195530726,
      "grad_norm": 0.42841342091560364,
      "learning_rate": 0.00014493738008694503,
      "loss": 0.0418,
      "step": 870
    },
    {
      "epoch": 4.916201117318436,
      "grad_norm": 0.3888803720474243,
      "learning_rate": 0.00014476922543105443,
      "loss": 0.0391,
      "step": 880
    },
    {
      "epoch": 4.972067039106145,
      "grad_norm": 0.2680191695690155,
      "learning_rate": 0.00014459842439868963,
      "loss": 0.0412,
      "step": 890
    },
    {
      "epoch": 5.027932960893855,
      "grad_norm": 0.3152766823768616,
      "learning_rate": 0.00014442498346840658,
      "loss": 0.0351,
      "step": 900
    },
    {
      "epoch": 5.083798882681564,
      "grad_norm": 0.3976776897907257,
      "learning_rate": 0.00014424890921889373,
      "loss": 0.0369,
      "step": 910
    },
    {
      "epoch": 5.139664804469274,
      "grad_norm": 0.4556487500667572,
      "learning_rate": 0.00014407020832872246,
      "loss": 0.0337,
      "step": 920
    },
    {
      "epoch": 5.195530726256983,
      "grad_norm": 0.368202269077301,
      "learning_rate": 0.00014388888757609376,
      "loss": 0.0396,
      "step": 930
    },
    {
      "epoch": 5.251396648044693,
      "grad_norm": 0.34850746393203735,
      "learning_rate": 0.00014370495383858107,
      "loss": 0.0361,
      "step": 940
    },
    {
      "epoch": 5.307262569832402,
      "grad_norm": 0.21902212500572205,
      "learning_rate": 0.00014351841409286954,
      "loss": 0.0332,
      "step": 950
    },
    {
      "epoch": 5.363128491620111,
      "grad_norm": 0.37780269980430603,
      "learning_rate": 0.00014332927541449122,
      "loss": 0.0359,
      "step": 960
    },
    {
      "epoch": 5.418994413407821,
      "grad_norm": 0.20398513972759247,
      "learning_rate": 0.00014313754497755687,
      "loss": 0.0394,
      "step": 970
    },
    {
      "epoch": 5.4748603351955305,
      "grad_norm": 0.2890792787075043,
      "learning_rate": 0.00014294323005448374,
      "loss": 0.042,
      "step": 980
    },
    {
      "epoch": 5.5307262569832405,
      "grad_norm": 0.3031609356403351,
      "learning_rate": 0.0001427463380157197,
      "loss": 0.0429,
      "step": 990
    },
    {
      "epoch": 5.58659217877095,
      "grad_norm": 0.5294051766395569,
      "learning_rate": 0.0001425468763294638,
      "loss": 0.0424,
      "step": 1000
    },
    {
      "epoch": 5.64245810055866,
      "grad_norm": 0.31394124031066895,
      "learning_rate": 0.00014234485256138277,
      "loss": 0.0455,
      "step": 1010
    },
    {
      "epoch": 5.698324022346369,
      "grad_norm": 0.5516133904457092,
      "learning_rate": 0.00014214027437432439,
      "loss": 0.0375,
      "step": 1020
    },
    {
      "epoch": 5.754189944134078,
      "grad_norm": 0.45591840147972107,
      "learning_rate": 0.00014193314952802645,
      "loss": 0.0361,
      "step": 1030
    },
    {
      "epoch": 5.810055865921788,
      "grad_norm": 0.41406169533729553,
      "learning_rate": 0.00014172348587882276,
      "loss": 0.0347,
      "step": 1040
    },
    {
      "epoch": 5.865921787709497,
      "grad_norm": 0.3035840094089508,
      "learning_rate": 0.00014151129137934492,
      "loss": 0.0354,
      "step": 1050
    },
    {
      "epoch": 5.921787709497207,
      "grad_norm": 0.3065755367279053,
      "learning_rate": 0.00014129657407822082,
      "loss": 0.0364,
      "step": 1060
    },
    {
      "epoch": 5.977653631284916,
      "grad_norm": 0.3852936029434204,
      "learning_rate": 0.0001410793421197692,
      "loss": 0.0352,
      "step": 1070
    },
    {
      "epoch": 6.033519553072626,
      "grad_norm": 0.3598945140838623,
      "learning_rate": 0.00014085960374369096,
      "loss": 0.0404,
      "step": 1080
    },
    {
      "epoch": 6.089385474860335,
      "grad_norm": 0.3946482539176941,
      "learning_rate": 0.00014063736728475634,
      "loss": 0.0312,
      "step": 1090
    },
    {
      "epoch": 6.145251396648045,
      "grad_norm": 0.4066295027732849,
      "learning_rate": 0.00014041264117248907,
      "loss": 0.041,
      "step": 1100
    },
    {
      "epoch": 6.201117318435754,
      "grad_norm": 0.2542637586593628,
      "learning_rate": 0.00014018543393084633,
      "loss": 0.0366,
      "step": 1110
    },
    {
      "epoch": 6.256983240223463,
      "grad_norm": 0.34908485412597656,
      "learning_rate": 0.0001399557541778958,
      "loss": 0.0358,
      "step": 1120
    },
    {
      "epoch": 6.312849162011173,
      "grad_norm": 0.39812198281288147,
      "learning_rate": 0.00013972361062548837,
      "loss": 0.0405,
      "step": 1130
    },
    {
      "epoch": 6.368715083798882,
      "grad_norm": 0.45712536573410034,
      "learning_rate": 0.00013948901207892807,
      "loss": 0.037,
      "step": 1140
    },
    {
      "epoch": 6.424581005586592,
      "grad_norm": 0.5260054469108582,
      "learning_rate": 0.0001392519674366377,
      "loss": 0.0381,
      "step": 1150
    },
    {
      "epoch": 6.4804469273743015,
      "grad_norm": 0.39201897382736206,
      "learning_rate": 0.00013901248568982177,
      "loss": 0.0366,
      "step": 1160
    },
    {
      "epoch": 6.5363128491620115,
      "grad_norm": 0.22905181348323822,
      "learning_rate": 0.00013877057592212496,
      "loss": 0.0397,
      "step": 1170
    },
    {
      "epoch": 6.592178770949721,
      "grad_norm": 0.4780491590499878,
      "learning_rate": 0.00013852624730928794,
      "loss": 0.0414,
      "step": 1180
    },
    {
      "epoch": 6.648044692737431,
      "grad_norm": 0.3810228705406189,
      "learning_rate": 0.00013827950911879922,
      "loss": 0.035,
      "step": 1190
    },
    {
      "epoch": 6.70391061452514,
      "grad_norm": 0.3878680467605591,
      "learning_rate": 0.0001380303707095436,
      "loss": 0.0365,
      "step": 1200
    },
    {
      "epoch": 6.759776536312849,
      "grad_norm": 0.3587484657764435,
      "learning_rate": 0.00013777884153144714,
      "loss": 0.0349,
      "step": 1210
    },
    {
      "epoch": 6.815642458100559,
      "grad_norm": 0.3499602675437927,
      "learning_rate": 0.00013752493112511888,
      "loss": 0.0418,
      "step": 1220
    },
    {
      "epoch": 6.871508379888268,
      "grad_norm": 0.3068997859954834,
      "learning_rate": 0.00013726864912148878,
      "loss": 0.0364,
      "step": 1230
    },
    {
      "epoch": 6.927374301675978,
      "grad_norm": 0.24322666227817535,
      "learning_rate": 0.00013701000524144252,
      "loss": 0.0413,
      "step": 1240
    },
    {
      "epoch": 6.983240223463687,
      "grad_norm": 0.27041393518447876,
      "learning_rate": 0.00013674900929545284,
      "loss": 0.0362,
      "step": 1250
    },
    {
      "epoch": 7.039106145251397,
      "grad_norm": 0.4791170358657837,
      "learning_rate": 0.00013648567118320723,
      "loss": 0.0364,
      "step": 1260
    },
    {
      "epoch": 7.094972067039106,
      "grad_norm": 0.39187201857566833,
      "learning_rate": 0.00013622000089323256,
      "loss": 0.0406,
      "step": 1270
    },
    {
      "epoch": 7.150837988826815,
      "grad_norm": 0.3253427743911743,
      "learning_rate": 0.00013595200850251627,
      "loss": 0.0337,
      "step": 1280
    },
    {
      "epoch": 7.206703910614525,
      "grad_norm": 0.2595740854740143,
      "learning_rate": 0.000135681704176124,
      "loss": 0.0367,
      "step": 1290
    },
    {
      "epoch": 7.262569832402234,
      "grad_norm": 0.4414283335208893,
      "learning_rate": 0.00013540909816681416,
      "loss": 0.0391,
      "step": 1300
    },
    {
      "epoch": 7.318435754189944,
      "grad_norm": 0.329851895570755,
      "learning_rate": 0.0001351342008146489,
      "loss": 0.0306,
      "step": 1310
    },
    {
      "epoch": 7.374301675977653,
      "grad_norm": 0.3009398877620697,
      "learning_rate": 0.000134857022546602,
      "loss": 0.032,
      "step": 1320
    },
    {
      "epoch": 7.430167597765363,
      "grad_norm": 0.670881986618042,
      "learning_rate": 0.00013457757387616348,
      "loss": 0.04,
      "step": 1330
    },
    {
      "epoch": 7.4860335195530725,
      "grad_norm": 0.38917240500450134,
      "learning_rate": 0.00013429586540294049,
      "loss": 0.0337,
      "step": 1340
    },
    {
      "epoch": 7.5418994413407825,
      "grad_norm": 0.3102369010448456,
      "learning_rate": 0.00013401190781225553,
      "loss": 0.031,
      "step": 1350
    },
    {
      "epoch": 7.597765363128492,
      "grad_norm": 0.37413033843040466,
      "learning_rate": 0.00013372571187474121,
      "loss": 0.0334,
      "step": 1360
    },
    {
      "epoch": 7.653631284916202,
      "grad_norm": 0.28566494584083557,
      "learning_rate": 0.0001334372884459314,
      "loss": 0.0284,
      "step": 1370
    },
    {
      "epoch": 7.709497206703911,
      "grad_norm": 0.35984960198402405,
      "learning_rate": 0.0001331466484658498,
      "loss": 0.028,
      "step": 1380
    },
    {
      "epoch": 7.76536312849162,
      "grad_norm": 0.41864386200904846,
      "learning_rate": 0.00013285380295859482,
      "loss": 0.0316,
      "step": 1390
    },
    {
      "epoch": 7.82122905027933,
      "grad_norm": 0.44935837388038635,
      "learning_rate": 0.00013255876303192138,
      "loss": 0.0341,
      "step": 1400
    },
    {
      "epoch": 7.877094972067039,
      "grad_norm": 0.2935531437397003,
      "learning_rate": 0.00013226153987681974,
      "loss": 0.035,
      "step": 1410
    },
    {
      "epoch": 7.932960893854749,
      "grad_norm": 0.5159788131713867,
      "learning_rate": 0.00013196214476709096,
      "loss": 0.0303,
      "step": 1420
    },
    {
      "epoch": 7.988826815642458,
      "grad_norm": 0.3987308144569397,
      "learning_rate": 0.00013166058905891918,
      "loss": 0.0342,
      "step": 1430
    },
    {
      "epoch": 8.044692737430168,
      "grad_norm": 0.35137230157852173,
      "learning_rate": 0.00013135688419044109,
      "loss": 0.0298,
      "step": 1440
    },
    {
      "epoch": 8.100558659217878,
      "grad_norm": 0.2903079092502594,
      "learning_rate": 0.00013105104168131178,
      "loss": 0.0293,
      "step": 1450
    },
    {
      "epoch": 8.156424581005586,
      "grad_norm": 0.26529935002326965,
      "learning_rate": 0.00013074307313226807,
      "loss": 0.0327,
      "step": 1460
    },
    {
      "epoch": 8.212290502793296,
      "grad_norm": 0.46690046787261963,
      "learning_rate": 0.0001304329902246884,
      "loss": 0.0336,
      "step": 1470
    },
    {
      "epoch": 8.268156424581006,
      "grad_norm": 0.37878528237342834,
      "learning_rate": 0.00013012080472014968,
      "loss": 0.0304,
      "step": 1480
    },
    {
      "epoch": 8.324022346368714,
      "grad_norm": 0.2849811315536499,
      "learning_rate": 0.00012980652845998123,
      "loss": 0.032,
      "step": 1490
    },
    {
      "epoch": 8.379888268156424,
      "grad_norm": 0.2862546741962433,
      "learning_rate": 0.0001294901733648156,
      "loss": 0.0286,
      "step": 1500
    },
    {
      "epoch": 8.435754189944134,
      "grad_norm": 0.21827903389930725,
      "learning_rate": 0.00012917175143413646,
      "loss": 0.0373,
      "step": 1510
    },
    {
      "epoch": 8.491620111731844,
      "grad_norm": 0.16881871223449707,
      "learning_rate": 0.00012885127474582348,
      "loss": 0.0334,
      "step": 1520
    },
    {
      "epoch": 8.547486033519553,
      "grad_norm": 0.5263266563415527,
      "learning_rate": 0.00012852875545569408,
      "loss": 0.037,
      "step": 1530
    },
    {
      "epoch": 8.603351955307263,
      "grad_norm": 0.41702142357826233,
      "learning_rate": 0.00012820420579704245,
      "loss": 0.0398,
      "step": 1540
    },
    {
      "epoch": 8.659217877094973,
      "grad_norm": 0.20871484279632568,
      "learning_rate": 0.00012787763808017565,
      "loss": 0.0321,
      "step": 1550
    },
    {
      "epoch": 8.71508379888268,
      "grad_norm": 0.4315951466560364,
      "learning_rate": 0.00012754906469194632,
      "loss": 0.0316,
      "step": 1560
    },
    {
      "epoch": 8.77094972067039,
      "grad_norm": 0.2898998558521271,
      "learning_rate": 0.0001272184980952833,
      "loss": 0.0314,
      "step": 1570
    },
    {
      "epoch": 8.8268156424581,
      "grad_norm": 0.4114074409008026,
      "learning_rate": 0.00012688595082871852,
      "loss": 0.0316,
      "step": 1580
    },
    {
      "epoch": 8.88268156424581,
      "grad_norm": 0.3064340353012085,
      "learning_rate": 0.00012655143550591164,
      "loss": 0.0335,
      "step": 1590
    },
    {
      "epoch": 8.938547486033519,
      "grad_norm": 0.2396443486213684,
      "learning_rate": 0.0001262149648151715,
      "loss": 0.0292,
      "step": 1600
    },
    {
      "epoch": 8.994413407821229,
      "grad_norm": 0.280146986246109,
      "learning_rate": 0.00012587655151897488,
      "loss": 0.031,
      "step": 1610
    },
    {
      "epoch": 9.050279329608939,
      "grad_norm": 0.2552490234375,
      "learning_rate": 0.00012553620845348246,
      "loss": 0.0249,
      "step": 1620
    },
    {
      "epoch": 9.106145251396647,
      "grad_norm": 0.1910940706729889,
      "learning_rate": 0.00012519394852805182,
      "loss": 0.0297,
      "step": 1630
    },
    {
      "epoch": 9.162011173184357,
      "grad_norm": 0.2914765477180481,
      "learning_rate": 0.0001248497847247479,
      "loss": 0.0321,
      "step": 1640
    },
    {
      "epoch": 9.217877094972067,
      "grad_norm": 0.34026339650154114,
      "learning_rate": 0.00012450373009785054,
      "loss": 0.0272,
      "step": 1650
    },
    {
      "epoch": 9.273743016759777,
      "grad_norm": 0.21475638449192047,
      "learning_rate": 0.00012415579777335932,
      "loss": 0.026,
      "step": 1660
    },
    {
      "epoch": 9.329608938547485,
      "grad_norm": 0.30112239718437195,
      "learning_rate": 0.00012380600094849566,
      "loss": 0.0374,
      "step": 1670
    },
    {
      "epoch": 9.385474860335195,
      "grad_norm": 0.17926353216171265,
      "learning_rate": 0.00012345435289120234,
      "loss": 0.026,
      "step": 1680
    },
    {
      "epoch": 9.441340782122905,
      "grad_norm": 0.16974708437919617,
      "learning_rate": 0.00012310086693964012,
      "loss": 0.0308,
      "step": 1690
    },
    {
      "epoch": 9.497206703910614,
      "grad_norm": 0.3298158347606659,
      "learning_rate": 0.00012274555650168187,
      "loss": 0.03,
      "step": 1700
    },
    {
      "epoch": 9.553072625698324,
      "grad_norm": 0.33263733983039856,
      "learning_rate": 0.000122388435054404,
      "loss": 0.03,
      "step": 1710
    },
    {
      "epoch": 9.608938547486034,
      "grad_norm": 0.5304810404777527,
      "learning_rate": 0.00012202951614357541,
      "loss": 0.028,
      "step": 1720
    },
    {
      "epoch": 9.664804469273744,
      "grad_norm": 0.2756952941417694,
      "learning_rate": 0.00012166881338314337,
      "loss": 0.0271,
      "step": 1730
    },
    {
      "epoch": 9.720670391061452,
      "grad_norm": 0.329500675201416,
      "learning_rate": 0.0001213063404547174,
      "loss": 0.0392,
      "step": 1740
    },
    {
      "epoch": 9.776536312849162,
      "grad_norm": 0.20280927419662476,
      "learning_rate": 0.0001209421111070503,
      "loss": 0.0293,
      "step": 1750
    },
    {
      "epoch": 9.832402234636872,
      "grad_norm": 0.26771217584609985,
      "learning_rate": 0.00012057613915551652,
      "loss": 0.0283,
      "step": 1760
    },
    {
      "epoch": 9.888268156424582,
      "grad_norm": 0.2704836428165436,
      "learning_rate": 0.00012020843848158826,
      "loss": 0.0295,
      "step": 1770
    },
    {
      "epoch": 9.94413407821229,
      "grad_norm": 0.34217214584350586,
      "learning_rate": 0.00011983902303230892,
      "loss": 0.0303,
      "step": 1780
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.27921390533447266,
      "learning_rate": 0.00011946790681976399,
      "loss": 0.0292,
      "step": 1790
    },
    {
      "epoch": 10.05586592178771,
      "grad_norm": 0.2127024084329605,
      "learning_rate": 0.0001190951039205497,
      "loss": 0.0256,
      "step": 1800
    },
    {
      "epoch": 10.111731843575418,
      "grad_norm": 0.2055603712797165,
      "learning_rate": 0.00011872062847523899,
      "loss": 0.028,
      "step": 1810
    },
    {
      "epoch": 10.167597765363128,
      "grad_norm": 0.27974414825439453,
      "learning_rate": 0.00011834449468784523,
      "loss": 0.03,
      "step": 1820
    },
    {
      "epoch": 10.223463687150838,
      "grad_norm": 0.3269798159599304,
      "learning_rate": 0.00011796671682528334,
      "loss": 0.0287,
      "step": 1830
    },
    {
      "epoch": 10.279329608938548,
      "grad_norm": 0.3258807063102722,
      "learning_rate": 0.00011758730921682876,
      "loss": 0.0284,
      "step": 1840
    },
    {
      "epoch": 10.335195530726256,
      "grad_norm": 0.21946227550506592,
      "learning_rate": 0.00011720628625357392,
      "loss": 0.0306,
      "step": 1850
    },
    {
      "epoch": 10.391061452513966,
      "grad_norm": 0.3200410008430481,
      "learning_rate": 0.00011682366238788225,
      "loss": 0.0283,
      "step": 1860
    },
    {
      "epoch": 10.446927374301676,
      "grad_norm": 0.32745105028152466,
      "learning_rate": 0.0001164394521328402,
      "loss": 0.0269,
      "step": 1870
    },
    {
      "epoch": 10.502793296089386,
      "grad_norm": 0.30036798119544983,
      "learning_rate": 0.00011605367006170654,
      "loss": 0.0267,
      "step": 1880
    },
    {
      "epoch": 10.558659217877095,
      "grad_norm": 0.3197910785675049,
      "learning_rate": 0.0001156663308073598,
      "loss": 0.0257,
      "step": 1890
    },
    {
      "epoch": 10.614525139664805,
      "grad_norm": 0.23818261921405792,
      "learning_rate": 0.00011527744906174306,
      "loss": 0.0309,
      "step": 1900
    },
    {
      "epoch": 10.670391061452515,
      "grad_norm": 0.1711827963590622,
      "learning_rate": 0.00011488703957530675,
      "loss": 0.0275,
      "step": 1910
    },
    {
      "epoch": 10.726256983240223,
      "grad_norm": 0.3193713426589966,
      "learning_rate": 0.00011449511715644922,
      "loss": 0.0286,
      "step": 1920
    },
    {
      "epoch": 10.782122905027933,
      "grad_norm": 0.3227252662181854,
      "learning_rate": 0.00011410169667095494,
      "loss": 0.0274,
      "step": 1930
    },
    {
      "epoch": 10.837988826815643,
      "grad_norm": 0.27391719818115234,
      "learning_rate": 0.0001137067930414307,
      "loss": 0.0314,
      "step": 1940
    },
    {
      "epoch": 10.893854748603353,
      "grad_norm": 0.407885879278183,
      "learning_rate": 0.0001133104212467396,
      "loss": 0.0272,
      "step": 1950
    },
    {
      "epoch": 10.949720670391061,
      "grad_norm": 0.2611219584941864,
      "learning_rate": 0.0001129125963214328,
      "loss": 0.0266,
      "step": 1960
    },
    {
      "epoch": 11.005586592178771,
      "grad_norm": 0.335827499628067,
      "learning_rate": 0.00011251333335517944,
      "loss": 0.0273,
      "step": 1970
    },
    {
      "epoch": 11.061452513966481,
      "grad_norm": 0.24541901051998138,
      "learning_rate": 0.00011211264749219403,
      "loss": 0.0295,
      "step": 1980
    },
    {
      "epoch": 11.11731843575419,
      "grad_norm": 0.27788233757019043,
      "learning_rate": 0.00011171055393066224,
      "loss": 0.0254,
      "step": 1990
    },
    {
      "epoch": 11.1731843575419,
      "grad_norm": 0.23598261177539825,
      "learning_rate": 0.00011130706792216439,
      "loss": 0.0298,
      "step": 2000
    },
    {
      "epoch": 11.22905027932961,
      "grad_norm": 0.20312641561031342,
      "learning_rate": 0.0001109022047710968,
      "loss": 0.0278,
      "step": 2010
    },
    {
      "epoch": 11.28491620111732,
      "grad_norm": 0.2139887809753418,
      "learning_rate": 0.00011049597983409147,
      "loss": 0.0271,
      "step": 2020
    },
    {
      "epoch": 11.340782122905027,
      "grad_norm": 0.22644950449466705,
      "learning_rate": 0.00011008840851943348,
      "loss": 0.0298,
      "step": 2030
    },
    {
      "epoch": 11.396648044692737,
      "grad_norm": 0.20427192747592926,
      "learning_rate": 0.00010967950628647664,
      "loss": 0.03,
      "step": 2040
    },
    {
      "epoch": 11.452513966480447,
      "grad_norm": 0.4383834898471832,
      "learning_rate": 0.00010926928864505698,
      "loss": 0.0303,
      "step": 2050
    },
    {
      "epoch": 11.508379888268156,
      "grad_norm": 0.2548943758010864,
      "learning_rate": 0.00010885777115490463,
      "loss": 0.0288,
      "step": 2060
    },
    {
      "epoch": 11.564245810055866,
      "grad_norm": 0.2970632016658783,
      "learning_rate": 0.00010844496942505342,
      "loss": 0.0287,
      "step": 2070
    },
    {
      "epoch": 11.620111731843576,
      "grad_norm": 0.28128278255462646,
      "learning_rate": 0.00010803089911324907,
      "loss": 0.0298,
      "step": 2080
    },
    {
      "epoch": 11.675977653631286,
      "grad_norm": 0.2895059287548065,
      "learning_rate": 0.00010761557592535509,
      "loss": 0.0312,
      "step": 2090
    },
    {
      "epoch": 11.731843575418994,
      "grad_norm": 0.3408230245113373,
      "learning_rate": 0.00010719901561475706,
      "loss": 0.031,
      "step": 2100
    },
    {
      "epoch": 11.787709497206704,
      "grad_norm": 0.2545909881591797,
      "learning_rate": 0.00010678123398176526,
      "loss": 0.0274,
      "step": 2110
    },
    {
      "epoch": 11.843575418994414,
      "grad_norm": 0.2746656537055969,
      "learning_rate": 0.00010636224687301515,
      "loss": 0.0214,
      "step": 2120
    },
    {
      "epoch": 11.899441340782122,
      "grad_norm": 0.35347574949264526,
      "learning_rate": 0.00010594207018086647,
      "loss": 0.0306,
      "step": 2130
    },
    {
      "epoch": 11.955307262569832,
      "grad_norm": 0.2086307555437088,
      "learning_rate": 0.00010552071984280035,
      "loss": 0.0253,
      "step": 2140
    },
    {
      "epoch": 12.011173184357542,
      "grad_norm": 0.2846223711967468,
      "learning_rate": 0.00010509821184081479,
      "loss": 0.0267,
      "step": 2150
    },
    {
      "epoch": 12.067039106145252,
      "grad_norm": 0.26856729388237,
      "learning_rate": 0.00010467456220081847,
      "loss": 0.0218,
      "step": 2160
    },
    {
      "epoch": 12.12290502793296,
      "grad_norm": 0.24002467095851898,
      "learning_rate": 0.00010424978699202294,
      "loss": 0.0268,
      "step": 2170
    },
    {
      "epoch": 12.17877094972067,
      "grad_norm": 0.37470901012420654,
      "learning_rate": 0.00010382390232633298,
      "loss": 0.0258,
      "step": 2180
    },
    {
      "epoch": 12.23463687150838,
      "grad_norm": 0.3723151981830597,
      "learning_rate": 0.00010339692435773564,
      "loss": 0.0303,
      "step": 2190
    },
    {
      "epoch": 12.29050279329609,
      "grad_norm": 0.44261157512664795,
      "learning_rate": 0.00010296886928168728,
      "loss": 0.0302,
      "step": 2200
    },
    {
      "epoch": 12.346368715083798,
      "grad_norm": 0.23005405068397522,
      "learning_rate": 0.00010253975333449953,
      "loss": 0.0302,
      "step": 2210
    },
    {
      "epoch": 12.402234636871508,
      "grad_norm": 0.35475534200668335,
      "learning_rate": 0.00010210959279272327,
      "loss": 0.0305,
      "step": 2220
    },
    {
      "epoch": 12.458100558659218,
      "grad_norm": 0.1821463704109192,
      "learning_rate": 0.00010167840397253129,
      "loss": 0.0293,
      "step": 2230
    },
    {
      "epoch": 12.513966480446927,
      "grad_norm": 0.3717893958091736,
      "learning_rate": 0.00010124620322909939,
      "loss": 0.0265,
      "step": 2240
    },
    {
      "epoch": 12.569832402234637,
      "grad_norm": 0.35059571266174316,
      "learning_rate": 0.00010081300695598612,
      "loss": 0.0257,
      "step": 2250
    },
    {
      "epoch": 12.625698324022347,
      "grad_norm": 0.3039356470108032,
      "learning_rate": 0.00010037883158451083,
      "loss": 0.0288,
      "step": 2260
    },
    {
      "epoch": 12.681564245810057,
      "grad_norm": 0.49444425106048584,
      "learning_rate": 9.994369358313057e-05,
      "loss": 0.029,
      "step": 2270
    },
    {
      "epoch": 12.737430167597765,
      "grad_norm": 0.2746371626853943,
      "learning_rate": 9.950760945681525e-05,
      "loss": 0.0265,
      "step": 2280
    },
    {
      "epoch": 12.793296089385475,
      "grad_norm": 0.21077698469161987,
      "learning_rate": 9.907059574642177e-05,
      "loss": 0.0279,
      "step": 2290
    },
    {
      "epoch": 12.849162011173185,
      "grad_norm": 0.3581600487232208,
      "learning_rate": 9.863266902806654e-05,
      "loss": 0.0256,
      "step": 2300
    },
    {
      "epoch": 12.905027932960895,
      "grad_norm": 0.3539135754108429,
      "learning_rate": 9.81938459124967e-05,
      "loss": 0.0244,
      "step": 2310
    },
    {
      "epoch": 12.960893854748603,
      "grad_norm": 0.24212266504764557,
      "learning_rate": 9.775414304446024e-05,
      "loss": 0.0291,
      "step": 2320
    },
    {
      "epoch": 13.016759776536313,
      "grad_norm": 0.24756573140621185,
      "learning_rate": 9.731357710207442e-05,
      "loss": 0.0254,
      "step": 2330
    },
    {
      "epoch": 13.072625698324023,
      "grad_norm": 0.20405328273773193,
      "learning_rate": 9.687216479619328e-05,
      "loss": 0.0274,
      "step": 2340
    },
    {
      "epoch": 13.128491620111731,
      "grad_norm": 0.3089810013771057,
      "learning_rate": 9.64299228697739e-05,
      "loss": 0.0233,
      "step": 2350
    },
    {
      "epoch": 13.184357541899441,
      "grad_norm": 0.3249969780445099,
      "learning_rate": 9.598686809724109e-05,
      "loss": 0.0251,
      "step": 2360
    },
    {
      "epoch": 13.240223463687151,
      "grad_norm": 0.30293509364128113,
      "learning_rate": 9.554301728385133e-05,
      "loss": 0.0257,
      "step": 2370
    },
    {
      "epoch": 13.296089385474861,
      "grad_norm": 0.27823933959007263,
      "learning_rate": 9.509838726505527e-05,
      "loss": 0.0245,
      "step": 2380
    },
    {
      "epoch": 13.35195530726257,
      "grad_norm": 0.31522759795188904,
      "learning_rate": 9.465299490585914e-05,
      "loss": 0.0254,
      "step": 2390
    },
    {
      "epoch": 13.40782122905028,
      "grad_norm": 0.28887462615966797,
      "learning_rate": 9.420685710018505e-05,
      "loss": 0.0237,
      "step": 2400
    },
    {
      "epoch": 13.46368715083799,
      "grad_norm": 0.23696090281009674,
      "learning_rate": 9.375999077023024e-05,
      "loss": 0.0264,
      "step": 2410
    },
    {
      "epoch": 13.519553072625698,
      "grad_norm": 0.2132495492696762,
      "learning_rate": 9.331241286582515e-05,
      "loss": 0.0238,
      "step": 2420
    },
    {
      "epoch": 13.575418994413408,
      "grad_norm": 0.3045707046985626,
      "learning_rate": 9.28641403637906e-05,
      "loss": 0.0292,
      "step": 2430
    },
    {
      "epoch": 13.631284916201118,
      "grad_norm": 0.2944694459438324,
      "learning_rate": 9.241519026729374e-05,
      "loss": 0.0252,
      "step": 2440
    },
    {
      "epoch": 13.687150837988828,
      "grad_norm": 0.18707014620304108,
      "learning_rate": 9.19655796052032e-05,
      "loss": 0.0264,
      "step": 2450
    },
    {
      "epoch": 13.743016759776536,
      "grad_norm": 0.4140284061431885,
      "learning_rate": 9.151532543144313e-05,
      "loss": 0.0284,
      "step": 2460
    },
    {
      "epoch": 13.798882681564246,
      "grad_norm": 0.24894434213638306,
      "learning_rate": 9.106444482434638e-05,
      "loss": 0.0288,
      "step": 2470
    },
    {
      "epoch": 13.854748603351956,
      "grad_norm": 0.24449239671230316,
      "learning_rate": 9.061295488600659e-05,
      "loss": 0.0232,
      "step": 2480
    },
    {
      "epoch": 13.910614525139664,
      "grad_norm": 0.2831474244594574,
      "learning_rate": 9.016087274162972e-05,
      "loss": 0.0238,
      "step": 2490
    },
    {
      "epoch": 13.966480446927374,
      "grad_norm": 0.27097368240356445,
      "learning_rate": 8.970821553888428e-05,
      "loss": 0.0314,
      "step": 2500
    },
    {
      "epoch": 14.022346368715084,
      "grad_norm": 0.2659156024456024,
      "learning_rate": 8.925500044725104e-05,
      "loss": 0.0254,
      "step": 2510
    },
    {
      "epoch": 14.078212290502794,
      "grad_norm": 0.200448676943779,
      "learning_rate": 8.880124465737168e-05,
      "loss": 0.0242,
      "step": 2520
    },
    {
      "epoch": 14.134078212290502,
      "grad_norm": 0.21925762295722961,
      "learning_rate": 8.834696538039678e-05,
      "loss": 0.0238,
      "step": 2530
    },
    {
      "epoch": 14.189944134078212,
      "grad_norm": 0.2610517144203186,
      "learning_rate": 8.789217984733312e-05,
      "loss": 0.0222,
      "step": 2540
    },
    {
      "epoch": 14.245810055865922,
      "grad_norm": 0.29856961965560913,
      "learning_rate": 8.743690530838983e-05,
      "loss": 0.0267,
      "step": 2550
    },
    {
      "epoch": 14.30167597765363,
      "grad_norm": 0.22454680502414703,
      "learning_rate": 8.69811590323244e-05,
      "loss": 0.023,
      "step": 2560
    },
    {
      "epoch": 14.35754189944134,
      "grad_norm": 0.3834502398967743,
      "learning_rate": 8.652495830578744e-05,
      "loss": 0.0248,
      "step": 2570
    },
    {
      "epoch": 14.41340782122905,
      "grad_norm": 0.3710498511791229,
      "learning_rate": 8.6068320432667e-05,
      "loss": 0.0263,
      "step": 2580
    },
    {
      "epoch": 14.46927374301676,
      "grad_norm": 0.19465550780296326,
      "learning_rate": 8.561126273343239e-05,
      "loss": 0.0219,
      "step": 2590
    },
    {
      "epoch": 14.525139664804469,
      "grad_norm": 0.1909979283809662,
      "learning_rate": 8.515380254447703e-05,
      "loss": 0.0255,
      "step": 2600
    },
    {
      "epoch": 14.581005586592179,
      "grad_norm": 0.3466046452522278,
      "learning_rate": 8.4695957217461e-05,
      "loss": 0.0249,
      "step": 2610
    },
    {
      "epoch": 14.636871508379889,
      "grad_norm": 0.3462992012500763,
      "learning_rate": 8.42377441186528e-05,
      "loss": 0.0245,
      "step": 2620
    },
    {
      "epoch": 14.692737430167599,
      "grad_norm": 0.28605830669403076,
      "learning_rate": 8.377918062827068e-05,
      "loss": 0.0254,
      "step": 2630
    },
    {
      "epoch": 14.748603351955307,
      "grad_norm": 0.2495012879371643,
      "learning_rate": 8.332028413982336e-05,
      "loss": 0.0268,
      "step": 2640
    },
    {
      "epoch": 14.804469273743017,
      "grad_norm": 0.23009264469146729,
      "learning_rate": 8.286107205945036e-05,
      "loss": 0.0235,
      "step": 2650
    },
    {
      "epoch": 14.860335195530727,
      "grad_norm": 0.5348438024520874,
      "learning_rate": 8.240156180526173e-05,
      "loss": 0.0282,
      "step": 2660
    },
    {
      "epoch": 14.916201117318435,
      "grad_norm": 0.23143941164016724,
      "learning_rate": 8.194177080667739e-05,
      "loss": 0.0227,
      "step": 2670
    },
    {
      "epoch": 14.972067039106145,
      "grad_norm": 0.2876596450805664,
      "learning_rate": 8.148171650376595e-05,
      "loss": 0.0217,
      "step": 2680
    },
    {
      "epoch": 15.027932960893855,
      "grad_norm": 0.21062903106212616,
      "learning_rate": 8.102141634658339e-05,
      "loss": 0.0226,
      "step": 2690
    },
    {
      "epoch": 15.083798882681565,
      "grad_norm": 0.22216618061065674,
      "learning_rate": 8.056088779451088e-05,
      "loss": 0.0244,
      "step": 2700
    },
    {
      "epoch": 15.139664804469273,
      "grad_norm": 0.2092180848121643,
      "learning_rate": 8.010014831559288e-05,
      "loss": 0.0267,
      "step": 2710
    },
    {
      "epoch": 15.195530726256983,
      "grad_norm": 0.31914299726486206,
      "learning_rate": 7.963921538587422e-05,
      "loss": 0.0186,
      "step": 2720
    },
    {
      "epoch": 15.251396648044693,
      "grad_norm": 0.28938183188438416,
      "learning_rate": 7.917810648873758e-05,
      "loss": 0.0254,
      "step": 2730
    },
    {
      "epoch": 15.307262569832401,
      "grad_norm": 0.2850409746170044,
      "learning_rate": 7.871683911424e-05,
      "loss": 0.0254,
      "step": 2740
    },
    {
      "epoch": 15.363128491620111,
      "grad_norm": 0.3008069097995758,
      "learning_rate": 7.82554307584498e-05,
      "loss": 0.021,
      "step": 2750
    },
    {
      "epoch": 15.418994413407821,
      "grad_norm": 0.2249889373779297,
      "learning_rate": 7.779389892278258e-05,
      "loss": 0.0215,
      "step": 2760
    },
    {
      "epoch": 15.474860335195531,
      "grad_norm": 0.3198390305042267,
      "learning_rate": 7.733226111333775e-05,
      "loss": 0.0214,
      "step": 2770
    },
    {
      "epoch": 15.53072625698324,
      "grad_norm": 0.2054470032453537,
      "learning_rate": 7.687053484023428e-05,
      "loss": 0.0248,
      "step": 2780
    },
    {
      "epoch": 15.58659217877095,
      "grad_norm": 0.3050447106361389,
      "learning_rate": 7.640873761694661e-05,
      "loss": 0.0254,
      "step": 2790
    },
    {
      "epoch": 15.64245810055866,
      "grad_norm": 0.3348614275455475,
      "learning_rate": 7.594688695964039e-05,
      "loss": 0.0228,
      "step": 2800
    },
    {
      "epoch": 15.69832402234637,
      "grad_norm": 0.2211858332157135,
      "learning_rate": 7.548500038650796e-05,
      "loss": 0.0211,
      "step": 2810
    },
    {
      "epoch": 15.754189944134078,
      "grad_norm": 0.25668853521347046,
      "learning_rate": 7.502309541710404e-05,
      "loss": 0.0247,
      "step": 2820
    },
    {
      "epoch": 15.810055865921788,
      "grad_norm": 0.18035383522510529,
      "learning_rate": 7.456118957168111e-05,
      "loss": 0.022,
      "step": 2830
    },
    {
      "epoch": 15.865921787709498,
      "grad_norm": 0.3626658320426941,
      "learning_rate": 7.409930037052485e-05,
      "loss": 0.0239,
      "step": 2840
    },
    {
      "epoch": 15.921787709497206,
      "grad_norm": 0.21136699616909027,
      "learning_rate": 7.363744533328964e-05,
      "loss": 0.0222,
      "step": 2850
    },
    {
      "epoch": 15.977653631284916,
      "grad_norm": 0.35462382435798645,
      "learning_rate": 7.317564197833401e-05,
      "loss": 0.0213,
      "step": 2860
    },
    {
      "epoch": 16.033519553072626,
      "grad_norm": 0.19702385365962982,
      "learning_rate": 7.271390782205617e-05,
      "loss": 0.0245,
      "step": 2870
    },
    {
      "epoch": 16.089385474860336,
      "grad_norm": 0.2611335813999176,
      "learning_rate": 7.225226037822956e-05,
      "loss": 0.0249,
      "step": 2880
    },
    {
      "epoch": 16.145251396648046,
      "grad_norm": 0.3108862340450287,
      "learning_rate": 7.179071715733858e-05,
      "loss": 0.0238,
      "step": 2890
    },
    {
      "epoch": 16.201117318435756,
      "grad_norm": 0.3305630385875702,
      "learning_rate": 7.132929566591445e-05,
      "loss": 0.0254,
      "step": 2900
    },
    {
      "epoch": 16.256983240223462,
      "grad_norm": 0.29176512360572815,
      "learning_rate": 7.086801340587114e-05,
      "loss": 0.0224,
      "step": 2910
    },
    {
      "epoch": 16.312849162011172,
      "grad_norm": 0.33446812629699707,
      "learning_rate": 7.040688787384144e-05,
      "loss": 0.0223,
      "step": 2920
    },
    {
      "epoch": 16.368715083798882,
      "grad_norm": 0.19562651216983795,
      "learning_rate": 6.994593656051346e-05,
      "loss": 0.0254,
      "step": 2930
    },
    {
      "epoch": 16.424581005586592,
      "grad_norm": 0.21511122584342957,
      "learning_rate": 6.94851769499671e-05,
      "loss": 0.0226,
      "step": 2940
    },
    {
      "epoch": 16.480446927374302,
      "grad_norm": 0.23985400795936584,
      "learning_rate": 6.902462651901085e-05,
      "loss": 0.0269,
      "step": 2950
    },
    {
      "epoch": 16.536312849162012,
      "grad_norm": 0.3013317584991455,
      "learning_rate": 6.856430273651896e-05,
      "loss": 0.021,
      "step": 2960
    },
    {
      "epoch": 16.592178770949722,
      "grad_norm": 0.35089585185050964,
      "learning_rate": 6.810422306276884e-05,
      "loss": 0.0232,
      "step": 2970
    },
    {
      "epoch": 16.64804469273743,
      "grad_norm": 0.17514319717884064,
      "learning_rate": 6.764440494877874e-05,
      "loss": 0.0203,
      "step": 2980
    },
    {
      "epoch": 16.70391061452514,
      "grad_norm": 0.15833304822444916,
      "learning_rate": 6.718486583564577e-05,
      "loss": 0.0207,
      "step": 2990
    },
    {
      "epoch": 16.75977653631285,
      "grad_norm": 0.22394374012947083,
      "learning_rate": 6.672562315388446e-05,
      "loss": 0.0219,
      "step": 3000
    },
    {
      "epoch": 16.81564245810056,
      "grad_norm": 0.33897238969802856,
      "learning_rate": 6.62666943227656e-05,
      "loss": 0.0231,
      "step": 3010
    },
    {
      "epoch": 16.87150837988827,
      "grad_norm": 0.3397151529788971,
      "learning_rate": 6.580809674965549e-05,
      "loss": 0.0203,
      "step": 3020
    },
    {
      "epoch": 16.92737430167598,
      "grad_norm": 0.2227960079908371,
      "learning_rate": 6.53498478293556e-05,
      "loss": 0.0197,
      "step": 3030
    },
    {
      "epoch": 16.98324022346369,
      "grad_norm": 0.29340264201164246,
      "learning_rate": 6.489196494344294e-05,
      "loss": 0.0238,
      "step": 3040
    },
    {
      "epoch": 17.039106145251395,
      "grad_norm": 0.35238850116729736,
      "learning_rate": 6.44344654596106e-05,
      "loss": 0.0271,
      "step": 3050
    },
    {
      "epoch": 17.094972067039105,
      "grad_norm": 0.18793293833732605,
      "learning_rate": 6.397736673100918e-05,
      "loss": 0.021,
      "step": 3060
    },
    {
      "epoch": 17.150837988826815,
      "grad_norm": 0.25130748748779297,
      "learning_rate": 6.35206860955883e-05,
      "loss": 0.0249,
      "step": 3070
    },
    {
      "epoch": 17.206703910614525,
      "grad_norm": 0.2051066905260086,
      "learning_rate": 6.306444087543927e-05,
      "loss": 0.019,
      "step": 3080
    },
    {
      "epoch": 17.262569832402235,
      "grad_norm": 0.16360081732273102,
      "learning_rate": 6.260864837613788e-05,
      "loss": 0.0213,
      "step": 3090
    },
    {
      "epoch": 17.318435754189945,
      "grad_norm": 0.3169659674167633,
      "learning_rate": 6.215332588608793e-05,
      "loss": 0.018,
      "step": 3100
    },
    {
      "epoch": 17.374301675977655,
      "grad_norm": 0.24224449694156647,
      "learning_rate": 6.16984906758657e-05,
      "loss": 0.0194,
      "step": 3110
    },
    {
      "epoch": 17.43016759776536,
      "grad_norm": 0.3294720947742462,
      "learning_rate": 6.124415999756466e-05,
      "loss": 0.025,
      "step": 3120
    },
    {
      "epoch": 17.48603351955307,
      "grad_norm": 0.3010692298412323,
      "learning_rate": 6.079035108414123e-05,
      "loss": 0.019,
      "step": 3130
    },
    {
      "epoch": 17.54189944134078,
      "grad_norm": 0.1720583438873291,
      "learning_rate": 6.033708114876097e-05,
      "loss": 0.0197,
      "step": 3140
    },
    {
      "epoch": 17.59776536312849,
      "grad_norm": 0.13594307005405426,
      "learning_rate": 5.988436738414584e-05,
      "loss": 0.0208,
      "step": 3150
    },
    {
      "epoch": 17.6536312849162,
      "grad_norm": 0.2565588057041168,
      "learning_rate": 5.943222696192204e-05,
      "loss": 0.0246,
      "step": 3160
    },
    {
      "epoch": 17.70949720670391,
      "grad_norm": 0.1882692277431488,
      "learning_rate": 5.898067703196857e-05,
      "loss": 0.023,
      "step": 3170
    },
    {
      "epoch": 17.76536312849162,
      "grad_norm": 0.16729940474033356,
      "learning_rate": 5.852973472176685e-05,
      "loss": 0.0188,
      "step": 3180
    },
    {
      "epoch": 17.821229050279328,
      "grad_norm": 0.2922714948654175,
      "learning_rate": 5.807941713575098e-05,
      "loss": 0.0226,
      "step": 3190
    },
    {
      "epoch": 17.877094972067038,
      "grad_norm": 0.33513203263282776,
      "learning_rate": 5.762974135465907e-05,
      "loss": 0.0234,
      "step": 3200
    },
    {
      "epoch": 17.932960893854748,
      "grad_norm": 0.2364656627178192,
      "learning_rate": 5.7180724434885294e-05,
      "loss": 0.0195,
      "step": 3210
    },
    {
      "epoch": 17.988826815642458,
      "grad_norm": 0.23285451531410217,
      "learning_rate": 5.673238340783285e-05,
      "loss": 0.0193,
      "step": 3220
    },
    {
      "epoch": 18.044692737430168,
      "grad_norm": 0.20390582084655762,
      "learning_rate": 5.6284735279268134e-05,
      "loss": 0.0206,
      "step": 3230
    },
    {
      "epoch": 18.100558659217878,
      "grad_norm": 0.3086446225643158,
      "learning_rate": 5.583779702867555e-05,
      "loss": 0.0201,
      "step": 3240
    },
    {
      "epoch": 18.156424581005588,
      "grad_norm": 0.17210273444652557,
      "learning_rate": 5.5391585608613514e-05,
      "loss": 0.0231,
      "step": 3250
    },
    {
      "epoch": 18.212290502793294,
      "grad_norm": 0.18998980522155762,
      "learning_rate": 5.4946117944071514e-05,
      "loss": 0.0203,
      "step": 3260
    },
    {
      "epoch": 18.268156424581004,
      "grad_norm": 0.2326553910970688,
      "learning_rate": 5.450141093182803e-05,
      "loss": 0.0194,
      "step": 3270
    },
    {
      "epoch": 18.324022346368714,
      "grad_norm": 0.24682560563087463,
      "learning_rate": 5.405748143980972e-05,
      "loss": 0.0208,
      "step": 3280
    },
    {
      "epoch": 18.379888268156424,
      "grad_norm": 0.2583821415901184,
      "learning_rate": 5.361434630645148e-05,
      "loss": 0.0204,
      "step": 3290
    },
    {
      "epoch": 18.435754189944134,
      "grad_norm": 0.23446252942085266,
      "learning_rate": 5.31720223400579e-05,
      "loss": 0.0176,
      "step": 3300
    },
    {
      "epoch": 18.491620111731844,
      "grad_norm": 0.2943982779979706,
      "learning_rate": 5.2730526318165684e-05,
      "loss": 0.0192,
      "step": 3310
    },
    {
      "epoch": 18.547486033519554,
      "grad_norm": 0.2999999225139618,
      "learning_rate": 5.2289874986907236e-05,
      "loss": 0.0204,
      "step": 3320
    },
    {
      "epoch": 18.60335195530726,
      "grad_norm": 0.21415217220783234,
      "learning_rate": 5.1850085060375417e-05,
      "loss": 0.0201,
      "step": 3330
    },
    {
      "epoch": 18.65921787709497,
      "grad_norm": 0.2961329221725464,
      "learning_rate": 5.1411173219989736e-05,
      "loss": 0.0199,
      "step": 3340
    },
    {
      "epoch": 18.71508379888268,
      "grad_norm": 0.25467586517333984,
      "learning_rate": 5.097315611386345e-05,
      "loss": 0.021,
      "step": 3350
    },
    {
      "epoch": 18.77094972067039,
      "grad_norm": 0.2654760777950287,
      "learning_rate": 5.0536050356172234e-05,
      "loss": 0.0171,
      "step": 3360
    },
    {
      "epoch": 18.8268156424581,
      "grad_norm": 0.279527485370636,
      "learning_rate": 5.0099872526523826e-05,
      "loss": 0.0205,
      "step": 3370
    },
    {
      "epoch": 18.88268156424581,
      "grad_norm": 0.41244757175445557,
      "learning_rate": 4.966463916932934e-05,
      "loss": 0.0162,
      "step": 3380
    },
    {
      "epoch": 18.93854748603352,
      "grad_norm": 0.25530514121055603,
      "learning_rate": 4.9230366793175654e-05,
      "loss": 0.0181,
      "step": 3390
    },
    {
      "epoch": 18.994413407821227,
      "grad_norm": 0.18296165764331818,
      "learning_rate": 4.879707187019914e-05,
      "loss": 0.0229,
      "step": 3400
    },
    {
      "epoch": 19.050279329608937,
      "grad_norm": 0.18316328525543213,
      "learning_rate": 4.8364770835461054e-05,
      "loss": 0.0195,
      "step": 3410
    },
    {
      "epoch": 19.106145251396647,
      "grad_norm": 0.1494152843952179,
      "learning_rate": 4.793348008632396e-05,
      "loss": 0.0167,
      "step": 3420
    },
    {
      "epoch": 19.162011173184357,
      "grad_norm": 0.3076527416706085,
      "learning_rate": 4.750321598182995e-05,
      "loss": 0.0177,
      "step": 3430
    },
    {
      "epoch": 19.217877094972067,
      "grad_norm": 0.27527594566345215,
      "learning_rate": 4.707399484207995e-05,
      "loss": 0.0216,
      "step": 3440
    },
    {
      "epoch": 19.273743016759777,
      "grad_norm": 0.16726377606391907,
      "learning_rate": 4.664583294761485e-05,
      "loss": 0.0211,
      "step": 3450
    },
    {
      "epoch": 19.329608938547487,
      "grad_norm": 0.2316291779279709,
      "learning_rate": 4.6218746538797926e-05,
      "loss": 0.0197,
      "step": 3460
    },
    {
      "epoch": 19.385474860335197,
      "grad_norm": 0.2616920471191406,
      "learning_rate": 4.579275181519879e-05,
      "loss": 0.0193,
      "step": 3470
    },
    {
      "epoch": 19.441340782122904,
      "grad_norm": 0.17275288701057434,
      "learning_rate": 4.536786493497894e-05,
      "loss": 0.0186,
      "step": 3480
    },
    {
      "epoch": 19.497206703910614,
      "grad_norm": 0.18898899853229523,
      "learning_rate": 4.494410201427896e-05,
      "loss": 0.0198,
      "step": 3490
    },
    {
      "epoch": 19.553072625698324,
      "grad_norm": 0.3262523412704468,
      "learning_rate": 4.452147912660715e-05,
      "loss": 0.0212,
      "step": 3500
    },
    {
      "epoch": 19.608938547486034,
      "grad_norm": 0.1693217158317566,
      "learning_rate": 4.4100012302229915e-05,
      "loss": 0.0203,
      "step": 3510
    },
    {
      "epoch": 19.664804469273744,
      "grad_norm": 0.1959124356508255,
      "learning_rate": 4.3679717527563576e-05,
      "loss": 0.0184,
      "step": 3520
    },
    {
      "epoch": 19.720670391061454,
      "grad_norm": 0.21138575673103333,
      "learning_rate": 4.326061074456817e-05,
      "loss": 0.0172,
      "step": 3530
    },
    {
      "epoch": 19.776536312849164,
      "grad_norm": 0.6261563301086426,
      "learning_rate": 4.2842707850142806e-05,
      "loss": 0.018,
      "step": 3540
    },
    {
      "epoch": 19.83240223463687,
      "grad_norm": 0.2905479073524475,
      "learning_rate": 4.242602469552234e-05,
      "loss": 0.017,
      "step": 3550
    },
    {
      "epoch": 19.88826815642458,
      "grad_norm": 0.4615681767463684,
      "learning_rate": 4.201057708567664e-05,
      "loss": 0.0206,
      "step": 3560
    },
    {
      "epoch": 19.94413407821229,
      "grad_norm": 0.1858612447977066,
      "learning_rate": 4.1596380778710686e-05,
      "loss": 0.0193,
      "step": 3570
    },
    {
      "epoch": 20.0,
      "grad_norm": 0.21462394297122955,
      "learning_rate": 4.1183451485267034e-05,
      "loss": 0.0175,
      "step": 3580
    },
    {
      "epoch": 20.05586592178771,
      "grad_norm": 0.2793791890144348,
      "learning_rate": 4.077180486792987e-05,
      "loss": 0.0166,
      "step": 3590
    },
    {
      "epoch": 20.11173184357542,
      "grad_norm": 0.22271260619163513,
      "learning_rate": 4.036145654063093e-05,
      "loss": 0.0146,
      "step": 3600
    },
    {
      "epoch": 20.16759776536313,
      "grad_norm": 0.15600596368312836,
      "learning_rate": 3.995242206805731e-05,
      "loss": 0.0193,
      "step": 3610
    },
    {
      "epoch": 20.223463687150836,
      "grad_norm": 0.20925801992416382,
      "learning_rate": 3.9544716965060996e-05,
      "loss": 0.0209,
      "step": 3620
    },
    {
      "epoch": 20.279329608938546,
      "grad_norm": 0.16699257493019104,
      "learning_rate": 3.9138356696070315e-05,
      "loss": 0.0136,
      "step": 3630
    },
    {
      "epoch": 20.335195530726256,
      "grad_norm": 0.22640997171401978,
      "learning_rate": 3.8733356674503676e-05,
      "loss": 0.0161,
      "step": 3640
    },
    {
      "epoch": 20.391061452513966,
      "grad_norm": 0.5254058837890625,
      "learning_rate": 3.832973226218457e-05,
      "loss": 0.0176,
      "step": 3650
    },
    {
      "epoch": 20.446927374301676,
      "grad_norm": 0.148833766579628,
      "learning_rate": 3.792749876875907e-05,
      "loss": 0.0147,
      "step": 3660
    },
    {
      "epoch": 20.502793296089386,
      "grad_norm": 0.26952382922172546,
      "learning_rate": 3.752667145111514e-05,
      "loss": 0.0193,
      "step": 3670
    },
    {
      "epoch": 20.558659217877096,
      "grad_norm": 0.24985520541667938,
      "learning_rate": 3.712726551280385e-05,
      "loss": 0.0164,
      "step": 3680
    },
    {
      "epoch": 20.614525139664803,
      "grad_norm": 0.12173712253570557,
      "learning_rate": 3.6729296103462835e-05,
      "loss": 0.016,
      "step": 3690
    },
    {
      "epoch": 20.670391061452513,
      "grad_norm": 0.24871109426021576,
      "learning_rate": 3.63327783182414e-05,
      "loss": 0.0174,
      "step": 3700
    },
    {
      "epoch": 20.726256983240223,
      "grad_norm": 0.21813338994979858,
      "learning_rate": 3.593772719722832e-05,
      "loss": 0.0175,
      "step": 3710
    },
    {
      "epoch": 20.782122905027933,
      "grad_norm": 0.15502704679965973,
      "learning_rate": 3.5544157724881045e-05,
      "loss": 0.0171,
      "step": 3720
    },
    {
      "epoch": 20.837988826815643,
      "grad_norm": 0.2693827152252197,
      "learning_rate": 3.515208482945758e-05,
      "loss": 0.0184,
      "step": 3730
    },
    {
      "epoch": 20.893854748603353,
      "grad_norm": 0.2545803487300873,
      "learning_rate": 3.476152338244995e-05,
      "loss": 0.0211,
      "step": 3740
    },
    {
      "epoch": 20.949720670391063,
      "grad_norm": 0.14349234104156494,
      "learning_rate": 3.437248819802042e-05,
      "loss": 0.016,
      "step": 3750
    },
    {
      "epoch": 21.00558659217877,
      "grad_norm": 0.42411503195762634,
      "learning_rate": 3.3984994032439496e-05,
      "loss": 0.0178,
      "step": 3760
    },
    {
      "epoch": 21.06145251396648,
      "grad_norm": 0.25866132974624634,
      "learning_rate": 3.359905558352609e-05,
      "loss": 0.0165,
      "step": 3770
    },
    {
      "epoch": 21.11731843575419,
      "grad_norm": 0.22836929559707642,
      "learning_rate": 3.321468749009017e-05,
      "loss": 0.0196,
      "step": 3780
    },
    {
      "epoch": 21.1731843575419,
      "grad_norm": 0.1367328017950058,
      "learning_rate": 3.283190433137742e-05,
      "loss": 0.0177,
      "step": 3790
    },
    {
      "epoch": 21.22905027932961,
      "grad_norm": 0.16710641980171204,
      "learning_rate": 3.2450720626516294e-05,
      "loss": 0.0153,
      "step": 3800
    },
    {
      "epoch": 21.28491620111732,
      "grad_norm": 0.4859740734100342,
      "learning_rate": 3.207115083396728e-05,
      "loss": 0.0166,
      "step": 3810
    },
    {
      "epoch": 21.34078212290503,
      "grad_norm": 0.2656506299972534,
      "learning_rate": 3.1693209350974466e-05,
      "loss": 0.0174,
      "step": 3820
    },
    {
      "epoch": 21.39664804469274,
      "grad_norm": 0.2200634777545929,
      "learning_rate": 3.131691051301952e-05,
      "loss": 0.0192,
      "step": 3830
    },
    {
      "epoch": 21.452513966480446,
      "grad_norm": 0.15455004572868347,
      "learning_rate": 3.09422685932778e-05,
      "loss": 0.0185,
      "step": 3840
    },
    {
      "epoch": 21.508379888268156,
      "grad_norm": 0.32110223174095154,
      "learning_rate": 3.056929780207711e-05,
      "loss": 0.0178,
      "step": 3850
    },
    {
      "epoch": 21.564245810055866,
      "grad_norm": 0.20172889530658722,
      "learning_rate": 3.0198012286358618e-05,
      "loss": 0.0154,
      "step": 3860
    },
    {
      "epoch": 21.620111731843576,
      "grad_norm": 0.1673143357038498,
      "learning_rate": 2.9828426129140237e-05,
      "loss": 0.0173,
      "step": 3870
    },
    {
      "epoch": 21.675977653631286,
      "grad_norm": 0.13557426631450653,
      "learning_rate": 2.94605533489826e-05,
      "loss": 0.0146,
      "step": 3880
    },
    {
      "epoch": 21.731843575418996,
      "grad_norm": 0.11623450368642807,
      "learning_rate": 2.9094407899457046e-05,
      "loss": 0.0139,
      "step": 3890
    },
    {
      "epoch": 21.787709497206706,
      "grad_norm": 0.16807802021503448,
      "learning_rate": 2.8730003668616698e-05,
      "loss": 0.0135,
      "step": 3900
    },
    {
      "epoch": 21.843575418994412,
      "grad_norm": 0.37868157029151917,
      "learning_rate": 2.8367354478469396e-05,
      "loss": 0.0151,
      "step": 3910
    },
    {
      "epoch": 21.899441340782122,
      "grad_norm": 0.29837918281555176,
      "learning_rate": 2.8006474084453595e-05,
      "loss": 0.0151,
      "step": 3920
    },
    {
      "epoch": 21.955307262569832,
      "grad_norm": 0.24290640652179718,
      "learning_rate": 2.764737617491653e-05,
      "loss": 0.0141,
      "step": 3930
    },
    {
      "epoch": 22.011173184357542,
      "grad_norm": 0.1992618888616562,
      "learning_rate": 2.729007437059502e-05,
      "loss": 0.0145,
      "step": 3940
    },
    {
      "epoch": 22.067039106145252,
      "grad_norm": 0.13898395001888275,
      "learning_rate": 2.6934582224098963e-05,
      "loss": 0.0145,
      "step": 3950
    },
    {
      "epoch": 22.122905027932962,
      "grad_norm": 0.257982075214386,
      "learning_rate": 2.658091321939697e-05,
      "loss": 0.0144,
      "step": 3960
    },
    {
      "epoch": 22.178770949720672,
      "grad_norm": 0.1655452847480774,
      "learning_rate": 2.6229080771305293e-05,
      "loss": 0.0141,
      "step": 3970
    },
    {
      "epoch": 22.23463687150838,
      "grad_norm": 0.15178877115249634,
      "learning_rate": 2.5879098224978706e-05,
      "loss": 0.0181,
      "step": 3980
    },
    {
      "epoch": 22.29050279329609,
      "grad_norm": 0.12818969786167145,
      "learning_rate": 2.5530978855404448e-05,
      "loss": 0.0112,
      "step": 3990
    },
    {
      "epoch": 22.3463687150838,
      "grad_norm": 0.18601444363594055,
      "learning_rate": 2.5184735866898657e-05,
      "loss": 0.0131,
      "step": 4000
    },
    {
      "epoch": 22.40223463687151,
      "grad_norm": 0.2364937663078308,
      "learning_rate": 2.4840382392605515e-05,
      "loss": 0.0169,
      "step": 4010
    },
    {
      "epoch": 22.45810055865922,
      "grad_norm": 0.1969766914844513,
      "learning_rate": 2.449793149399922e-05,
      "loss": 0.0117,
      "step": 4020
    },
    {
      "epoch": 22.51396648044693,
      "grad_norm": 0.31158965826034546,
      "learning_rate": 2.4157396160388358e-05,
      "loss": 0.0164,
      "step": 4030
    },
    {
      "epoch": 22.56983240223464,
      "grad_norm": 0.17856691777706146,
      "learning_rate": 2.381878930842337e-05,
      "loss": 0.015,
      "step": 4040
    },
    {
      "epoch": 22.625698324022345,
      "grad_norm": 0.2636869251728058,
      "learning_rate": 2.3482123781606547e-05,
      "loss": 0.0185,
      "step": 4050
    },
    {
      "epoch": 22.681564245810055,
      "grad_norm": 0.12141893804073334,
      "learning_rate": 2.3147412349804862e-05,
      "loss": 0.014,
      "step": 4060
    },
    {
      "epoch": 22.737430167597765,
      "grad_norm": 0.22012671828269958,
      "learning_rate": 2.2814667708765745e-05,
      "loss": 0.0162,
      "step": 4070
    },
    {
      "epoch": 22.793296089385475,
      "grad_norm": 0.16044989228248596,
      "learning_rate": 2.2483902479635238e-05,
      "loss": 0.0131,
      "step": 4080
    },
    {
      "epoch": 22.849162011173185,
      "grad_norm": 0.15903018414974213,
      "learning_rate": 2.2155129208479606e-05,
      "loss": 0.0152,
      "step": 4090
    },
    {
      "epoch": 22.905027932960895,
      "grad_norm": 0.1807982474565506,
      "learning_rate": 2.182836036580923e-05,
      "loss": 0.0152,
      "step": 4100
    },
    {
      "epoch": 22.960893854748605,
      "grad_norm": 0.2643374502658844,
      "learning_rate": 2.1503608346105598e-05,
      "loss": 0.0171,
      "step": 4110
    },
    {
      "epoch": 23.01675977653631,
      "grad_norm": 0.13949626684188843,
      "learning_rate": 2.118088546735138e-05,
      "loss": 0.0132,
      "step": 4120
    },
    {
      "epoch": 23.07262569832402,
      "grad_norm": 0.6042327880859375,
      "learning_rate": 2.0860203970562954e-05,
      "loss": 0.0146,
      "step": 4130
    },
    {
      "epoch": 23.12849162011173,
      "grad_norm": 0.28848835825920105,
      "learning_rate": 2.054157601932631e-05,
      "loss": 0.0157,
      "step": 4140
    },
    {
      "epoch": 23.18435754189944,
      "grad_norm": 0.5369081497192383,
      "learning_rate": 2.0225013699335435e-05,
      "loss": 0.0128,
      "step": 4150
    },
    {
      "epoch": 23.24022346368715,
      "grad_norm": 0.15896014869213104,
      "learning_rate": 1.9910529017934196e-05,
      "loss": 0.0171,
      "step": 4160
    },
    {
      "epoch": 23.29608938547486,
      "grad_norm": 0.22632993757724762,
      "learning_rate": 1.9598133903660642e-05,
      "loss": 0.0117,
      "step": 4170
    },
    {
      "epoch": 23.35195530726257,
      "grad_norm": 0.14386065304279327,
      "learning_rate": 1.9287840205794693e-05,
      "loss": 0.0148,
      "step": 4180
    },
    {
      "epoch": 23.407821229050278,
      "grad_norm": 0.2815168797969818,
      "learning_rate": 1.8979659693908616e-05,
      "loss": 0.0143,
      "step": 4190
    },
    {
      "epoch": 23.463687150837988,
      "grad_norm": 0.17017313838005066,
      "learning_rate": 1.867360405742066e-05,
      "loss": 0.0126,
      "step": 4200
    },
    {
      "epoch": 23.519553072625698,
      "grad_norm": 0.2865433990955353,
      "learning_rate": 1.8369684905151677e-05,
      "loss": 0.0129,
      "step": 4210
    },
    {
      "epoch": 23.575418994413408,
      "grad_norm": 0.2120775431394577,
      "learning_rate": 1.806791376488467e-05,
      "loss": 0.0111,
      "step": 4220
    },
    {
      "epoch": 23.631284916201118,
      "grad_norm": 0.2156466841697693,
      "learning_rate": 1.7768302082927673e-05,
      "loss": 0.0134,
      "step": 4230
    },
    {
      "epoch": 23.687150837988828,
      "grad_norm": 0.13205291330814362,
      "learning_rate": 1.74708612236796e-05,
      "loss": 0.0116,
      "step": 4240
    },
    {
      "epoch": 23.743016759776538,
      "grad_norm": 0.1907736212015152,
      "learning_rate": 1.7175602469199075e-05,
      "loss": 0.0107,
      "step": 4250
    },
    {
      "epoch": 23.798882681564244,
      "grad_norm": 0.19241803884506226,
      "learning_rate": 1.6882537018776567e-05,
      "loss": 0.0136,
      "step": 4260
    },
    {
      "epoch": 23.854748603351954,
      "grad_norm": 0.26086995005607605,
      "learning_rate": 1.6591675988509588e-05,
      "loss": 0.0169,
      "step": 4270
    },
    {
      "epoch": 23.910614525139664,
      "grad_norm": 0.1134791448712349,
      "learning_rate": 1.6303030410881034e-05,
      "loss": 0.0126,
      "step": 4280
    },
    {
      "epoch": 23.966480446927374,
      "grad_norm": 0.19399867951869965,
      "learning_rate": 1.601661123434082e-05,
      "loss": 0.0146,
      "step": 4290
    },
    {
      "epoch": 24.022346368715084,
      "grad_norm": 0.218746617436409,
      "learning_rate": 1.5732429322890384e-05,
      "loss": 0.0196,
      "step": 4300
    },
    {
      "epoch": 24.078212290502794,
      "grad_norm": 0.173094242811203,
      "learning_rate": 1.5450495455670868e-05,
      "loss": 0.0131,
      "step": 4310
    },
    {
      "epoch": 24.134078212290504,
      "grad_norm": 0.28099843859672546,
      "learning_rate": 1.5170820326554087e-05,
      "loss": 0.0125,
      "step": 4320
    },
    {
      "epoch": 24.189944134078214,
      "grad_norm": 0.2801670432090759,
      "learning_rate": 1.489341454373694e-05,
      "loss": 0.0129,
      "step": 4330
    },
    {
      "epoch": 24.24581005586592,
      "grad_norm": 0.21972441673278809,
      "learning_rate": 1.4618288629339079e-05,
      "loss": 0.016,
      "step": 4340
    },
    {
      "epoch": 24.30167597765363,
      "grad_norm": 0.132466122508049,
      "learning_rate": 1.4345453019003745e-05,
      "loss": 0.0134,
      "step": 4350
    },
    {
      "epoch": 24.35754189944134,
      "grad_norm": 0.1710919737815857,
      "learning_rate": 1.4074918061502045e-05,
      "loss": 0.0131,
      "step": 4360
    },
    {
      "epoch": 24.41340782122905,
      "grad_norm": 0.2302296757698059,
      "learning_rate": 1.3806694018340223e-05,
      "loss": 0.0138,
      "step": 4370
    },
    {
      "epoch": 24.46927374301676,
      "grad_norm": 0.34001755714416504,
      "learning_rate": 1.3540791063370673e-05,
      "loss": 0.0132,
      "step": 4380
    },
    {
      "epoch": 24.52513966480447,
      "grad_norm": 0.18118420243263245,
      "learning_rate": 1.3277219282405858e-05,
      "loss": 0.0114,
      "step": 4390
    },
    {
      "epoch": 24.58100558659218,
      "grad_norm": 0.2221328318119049,
      "learning_rate": 1.3015988672835823e-05,
      "loss": 0.0135,
      "step": 4400
    },
    {
      "epoch": 24.636871508379887,
      "grad_norm": 0.2257193773984909,
      "learning_rate": 1.2757109143249002e-05,
      "loss": 0.0155,
      "step": 4410
    },
    {
      "epoch": 24.692737430167597,
      "grad_norm": 0.3096782863140106,
      "learning_rate": 1.2500590513056333e-05,
      "loss": 0.0099,
      "step": 4420
    },
    {
      "epoch": 24.748603351955307,
      "grad_norm": 0.20636042952537537,
      "learning_rate": 1.2246442512118886e-05,
      "loss": 0.0124,
      "step": 4430
    },
    {
      "epoch": 24.804469273743017,
      "grad_norm": 0.24674853682518005,
      "learning_rate": 1.199467478037871e-05,
      "loss": 0.0134,
      "step": 4440
    },
    {
      "epoch": 24.860335195530727,
      "grad_norm": 0.3628787100315094,
      "learning_rate": 1.1745296867493226e-05,
      "loss": 0.0168,
      "step": 4450
    },
    {
      "epoch": 24.916201117318437,
      "grad_norm": 0.17266486585140228,
      "learning_rate": 1.1498318232473033e-05,
      "loss": 0.0112,
      "step": 4460
    },
    {
      "epoch": 24.972067039106147,
      "grad_norm": 0.20120470225811005,
      "learning_rate": 1.1253748243323055e-05,
      "loss": 0.0126,
      "step": 4470
    },
    {
      "epoch": 25.027932960893853,
      "grad_norm": 0.30043306946754456,
      "learning_rate": 1.1011596176687343e-05,
      "loss": 0.011,
      "step": 4480
    },
    {
      "epoch": 25.083798882681563,
      "grad_norm": 0.23731638491153717,
      "learning_rate": 1.0771871217496971e-05,
      "loss": 0.0164,
      "step": 4490
    },
    {
      "epoch": 25.139664804469273,
      "grad_norm": 0.1487792432308197,
      "learning_rate": 1.0534582458621932e-05,
      "loss": 0.0171,
      "step": 4500
    },
    {
      "epoch": 25.195530726256983,
      "grad_norm": 0.16674961149692535,
      "learning_rate": 1.0299738900526022e-05,
      "loss": 0.0118,
      "step": 4510
    },
    {
      "epoch": 25.251396648044693,
      "grad_norm": 0.1953108161687851,
      "learning_rate": 1.0067349450925548e-05,
      "loss": 0.0141,
      "step": 4520
    },
    {
      "epoch": 25.307262569832403,
      "grad_norm": 0.2785939872264862,
      "learning_rate": 9.837422924451454e-06,
      "loss": 0.013,
      "step": 4530
    },
    {
      "epoch": 25.363128491620113,
      "grad_norm": 0.18461591005325317,
      "learning_rate": 9.609968042314908e-06,
      "loss": 0.0101,
      "step": 4540
    },
    {
      "epoch": 25.41899441340782,
      "grad_norm": 0.195012629032135,
      "learning_rate": 9.384993431976647e-06,
      "loss": 0.0142,
      "step": 4550
    },
    {
      "epoch": 25.47486033519553,
      "grad_norm": 0.11832192540168762,
      "learning_rate": 9.162507626819516e-06,
      "loss": 0.0118,
      "step": 4560
    },
    {
      "epoch": 25.53072625698324,
      "grad_norm": 0.24214394390583038,
      "learning_rate": 8.942519065825041e-06,
      "loss": 0.0159,
      "step": 4570
    },
    {
      "epoch": 25.58659217877095,
      "grad_norm": 0.18261413276195526,
      "learning_rate": 8.725036093253124e-06,
      "loss": 0.011,
      "step": 4580
    },
    {
      "epoch": 25.64245810055866,
      "grad_norm": 0.1193467229604721,
      "learning_rate": 8.510066958325679e-06,
      "loss": 0.0109,
      "step": 4590
    },
    {
      "epoch": 25.69832402234637,
      "grad_norm": 0.25043681263923645,
      "learning_rate": 8.297619814913667e-06,
      "loss": 0.0124,
      "step": 4600
    },
    {
      "epoch": 25.75418994413408,
      "grad_norm": 0.17554740607738495,
      "learning_rate": 8.087702721227812e-06,
      "loss": 0.017,
      "step": 4610
    },
    {
      "epoch": 25.810055865921786,
      "grad_norm": 0.13422702252864838,
      "learning_rate": 7.880323639513029e-06,
      "loss": 0.0122,
      "step": 4620
    },
    {
      "epoch": 25.865921787709496,
      "grad_norm": 0.16326576471328735,
      "learning_rate": 7.675490435746326e-06,
      "loss": 0.0106,
      "step": 4630
    },
    {
      "epoch": 25.921787709497206,
      "grad_norm": 0.17265120148658752,
      "learning_rate": 7.473210879338462e-06,
      "loss": 0.0123,
      "step": 4640
    },
    {
      "epoch": 25.977653631284916,
      "grad_norm": 0.18150538206100464,
      "learning_rate": 7.273492642839273e-06,
      "loss": 0.0117,
      "step": 4650
    },
    {
      "epoch": 26.033519553072626,
      "grad_norm": 0.2668885290622711,
      "learning_rate": 7.0763433016466535e-06,
      "loss": 0.0125,
      "step": 4660
    },
    {
      "epoch": 26.089385474860336,
      "grad_norm": 0.20581719279289246,
      "learning_rate": 6.881770333719177e-06,
      "loss": 0.0148,
      "step": 4670
    },
    {
      "epoch": 26.145251396648046,
      "grad_norm": 0.15722183883190155,
      "learning_rate": 6.689781119292487e-06,
      "loss": 0.0124,
      "step": 4680
    },
    {
      "epoch": 26.201117318435756,
      "grad_norm": 0.1610933542251587,
      "learning_rate": 6.5003829405993885e-06,
      "loss": 0.011,
      "step": 4690
    },
    {
      "epoch": 26.256983240223462,
      "grad_norm": 0.2166851907968521,
      "learning_rate": 6.3135829815935736e-06,
      "loss": 0.0135,
      "step": 4700
    },
    {
      "epoch": 26.312849162011172,
      "grad_norm": 0.21534748375415802,
      "learning_rate": 6.1293883276770985e-06,
      "loss": 0.0092,
      "step": 4710
    },
    {
      "epoch": 26.368715083798882,
      "grad_norm": 0.1989695280790329,
      "learning_rate": 5.947805965431792e-06,
      "loss": 0.0127,
      "step": 4720
    },
    {
      "epoch": 26.424581005586592,
      "grad_norm": 0.44271120429039,
      "learning_rate": 5.768842782354055e-06,
      "loss": 0.01,
      "step": 4730
    },
    {
      "epoch": 26.480446927374302,
      "grad_norm": 0.17514117062091827,
      "learning_rate": 5.592505566593786e-06,
      "loss": 0.0106,
      "step": 4740
    },
    {
      "epoch": 26.536312849162012,
      "grad_norm": 0.15244068205356598,
      "learning_rate": 5.418801006696736e-06,
      "loss": 0.0124,
      "step": 4750
    },
    {
      "epoch": 26.592178770949722,
      "grad_norm": 0.3087766468524933,
      "learning_rate": 5.247735691350941e-06,
      "loss": 0.0099,
      "step": 4760
    },
    {
      "epoch": 26.64804469273743,
      "grad_norm": 0.12699635326862335,
      "learning_rate": 5.079316109136808e-06,
      "loss": 0.0116,
      "step": 4770
    },
    {
      "epoch": 26.70391061452514,
      "grad_norm": 0.14864447712898254,
      "learning_rate": 4.9135486482808515e-06,
      "loss": 0.0137,
      "step": 4780
    },
    {
      "epoch": 26.75977653631285,
      "grad_norm": 0.14272750914096832,
      "learning_rate": 4.750439596413605e-06,
      "loss": 0.011,
      "step": 4790
    },
    {
      "epoch": 26.81564245810056,
      "grad_norm": 0.18606601655483246,
      "learning_rate": 4.589995140330952e-06,
      "loss": 0.0127,
      "step": 4800
    },
    {
      "epoch": 26.87150837988827,
      "grad_norm": 0.1599322408437729,
      "learning_rate": 4.432221365759523e-06,
      "loss": 0.0102,
      "step": 4810
    },
    {
      "epoch": 26.92737430167598,
      "grad_norm": 0.2554507553577423,
      "learning_rate": 4.277124257125872e-06,
      "loss": 0.0145,
      "step": 4820
    },
    {
      "epoch": 26.98324022346369,
      "grad_norm": 0.15352533757686615,
      "learning_rate": 4.124709697329476e-06,
      "loss": 0.0137,
      "step": 4830
    },
    {
      "epoch": 27.039106145251395,
      "grad_norm": 0.10090656578540802,
      "learning_rate": 3.974983467519593e-06,
      "loss": 0.0099,
      "step": 4840
    },
    {
      "epoch": 27.094972067039105,
      "grad_norm": 0.203408882021904,
      "learning_rate": 3.8279512468759684e-06,
      "loss": 0.0133,
      "step": 4850
    },
    {
      "epoch": 27.150837988826815,
      "grad_norm": 0.0952206403017044,
      "learning_rate": 3.683618612393441e-06,
      "loss": 0.0127,
      "step": 4860
    },
    {
      "epoch": 27.206703910614525,
      "grad_norm": 0.09654044359922409,
      "learning_rate": 3.5419910386703694e-06,
      "loss": 0.0092,
      "step": 4870
    },
    {
      "epoch": 27.262569832402235,
      "grad_norm": 0.42048877477645874,
      "learning_rate": 3.403073897701028e-06,
      "loss": 0.0126,
      "step": 4880
    },
    {
      "epoch": 27.318435754189945,
      "grad_norm": 0.16950425505638123,
      "learning_rate": 3.266872458671846e-06,
      "loss": 0.0112,
      "step": 4890
    },
    {
      "epoch": 27.374301675977655,
      "grad_norm": 0.2811785042285919,
      "learning_rate": 3.133391887761441e-06,
      "loss": 0.0113,
      "step": 4900
    },
    {
      "epoch": 27.43016759776536,
      "grad_norm": 0.20062536001205444,
      "learning_rate": 3.002637247944828e-06,
      "loss": 0.0101,
      "step": 4910
    },
    {
      "epoch": 27.48603351955307,
      "grad_norm": 0.176531121134758,
      "learning_rate": 2.8746134988012453e-06,
      "loss": 0.0122,
      "step": 4920
    },
    {
      "epoch": 27.54189944134078,
      "grad_norm": 0.2781747877597809,
      "learning_rate": 2.749325496326099e-06,
      "loss": 0.0114,
      "step": 4930
    },
    {
      "epoch": 27.59776536312849,
      "grad_norm": 0.23316988348960876,
      "learning_rate": 2.6267779927467423e-06,
      "loss": 0.009,
      "step": 4940
    },
    {
      "epoch": 27.6536312849162,
      "grad_norm": 0.17585094273090363,
      "learning_rate": 2.5069756363422454e-06,
      "loss": 0.0131,
      "step": 4950
    },
    {
      "epoch": 27.70949720670391,
      "grad_norm": 0.22221805155277252,
      "learning_rate": 2.3899229712671027e-06,
      "loss": 0.0129,
      "step": 4960
    },
    {
      "epoch": 27.76536312849162,
      "grad_norm": 0.18307988345623016,
      "learning_rate": 2.2756244373787867e-06,
      "loss": 0.0099,
      "step": 4970
    },
    {
      "epoch": 27.821229050279328,
      "grad_norm": 0.32926902174949646,
      "learning_rate": 2.164084370069452e-06,
      "loss": 0.0106,
      "step": 4980
    },
    {
      "epoch": 27.877094972067038,
      "grad_norm": 0.1888887584209442,
      "learning_rate": 2.055307000101414e-06,
      "loss": 0.0122,
      "step": 4990
    },
    {
      "epoch": 27.932960893854748,
      "grad_norm": 0.13414178788661957,
      "learning_rate": 1.9492964534467026e-06,
      "loss": 0.0109,
      "step": 5000
    },
    {
      "epoch": 27.988826815642458,
      "grad_norm": 0.27602723240852356,
      "learning_rate": 1.8460567511305645e-06,
      "loss": 0.0125,
      "step": 5010
    },
    {
      "epoch": 28.044692737430168,
      "grad_norm": 0.1933746188879013,
      "learning_rate": 1.745591809078925e-06,
      "loss": 0.0105,
      "step": 5020
    },
    {
      "epoch": 28.100558659217878,
      "grad_norm": 0.09113660454750061,
      "learning_rate": 1.6479054379698985e-06,
      "loss": 0.0095,
      "step": 5030
    },
    {
      "epoch": 28.156424581005588,
      "grad_norm": 0.11389518529176712,
      "learning_rate": 1.5530013430891886e-06,
      "loss": 0.0123,
      "step": 5040
    },
    {
      "epoch": 28.212290502793294,
      "grad_norm": 0.12854307889938354,
      "learning_rate": 1.4608831241895823e-06,
      "loss": 0.0111,
      "step": 5050
    },
    {
      "epoch": 28.268156424581004,
      "grad_norm": 0.1454714685678482,
      "learning_rate": 1.3715542753544195e-06,
      "loss": 0.0116,
      "step": 5060
    },
    {
      "epoch": 28.324022346368714,
      "grad_norm": 0.11753421276807785,
      "learning_rate": 1.2850181848650304e-06,
      "loss": 0.0085,
      "step": 5070
    },
    {
      "epoch": 28.379888268156424,
      "grad_norm": 0.19782139360904694,
      "learning_rate": 1.2012781350722145e-06,
      "loss": 0.0099,
      "step": 5080
    },
    {
      "epoch": 28.435754189944134,
      "grad_norm": 0.1429385244846344,
      "learning_rate": 1.1203373022717899e-06,
      "loss": 0.0096,
      "step": 5090
    },
    {
      "epoch": 28.491620111731844,
      "grad_norm": 0.17323465645313263,
      "learning_rate": 1.0421987565840562e-06,
      "loss": 0.0107,
      "step": 5100
    },
    {
      "epoch": 28.547486033519554,
      "grad_norm": 0.1123204305768013,
      "learning_rate": 9.66865461837371e-07,
      "loss": 0.0139,
      "step": 5110
    },
    {
      "epoch": 28.60335195530726,
      "grad_norm": 0.23444265127182007,
      "learning_rate": 8.943402754557405e-07,
      "loss": 0.0148,
      "step": 5120
    },
    {
      "epoch": 28.65921787709497,
      "grad_norm": 0.10738368332386017,
      "learning_rate": 8.246259483504047e-07,
      "loss": 0.0115,
      "step": 5130
    },
    {
      "epoch": 28.71508379888268,
      "grad_norm": 0.1469849944114685,
      "learning_rate": 7.577251248155142e-07,
      "loss": 0.0092,
      "step": 5140
    },
    {
      "epoch": 28.77094972067039,
      "grad_norm": 0.1442977786064148,
      "learning_rate": 6.936403424278591e-07,
      "loss": 0.0126,
      "step": 5150
    },
    {
      "epoch": 28.8268156424581,
      "grad_norm": 0.11567971855401993,
      "learning_rate": 6.323740319505388e-07,
      "loss": 0.0108,
      "step": 5160
    },
    {
      "epoch": 28.88268156424581,
      "grad_norm": 0.12551957368850708,
      "learning_rate": 5.739285172408598e-07,
      "loss": 0.0129,
      "step": 5170
    },
    {
      "epoch": 28.93854748603352,
      "grad_norm": 0.16800257563591003,
      "learning_rate": 5.183060151620983e-07,
      "loss": 0.0124,
      "step": 5180
    },
    {
      "epoch": 28.994413407821227,
      "grad_norm": 0.17249980568885803,
      "learning_rate": 4.655086354994925e-07,
      "loss": 0.0118,
      "step": 5190
    },
    {
      "epoch": 29.050279329608937,
      "grad_norm": 0.3741263151168823,
      "learning_rate": 4.155383808801649e-07,
      "loss": 0.0147,
      "step": 5200
    },
    {
      "epoch": 29.106145251396647,
      "grad_norm": 0.22194480895996094,
      "learning_rate": 3.68397146697183e-07,
      "loss": 0.0138,
      "step": 5210
    },
    {
      "epoch": 29.162011173184357,
      "grad_norm": 0.2826599180698395,
      "learning_rate": 3.240867210376752e-07,
      "loss": 0.0145,
      "step": 5220
    },
    {
      "epoch": 29.217877094972067,
      "grad_norm": 0.21220660209655762,
      "learning_rate": 2.8260878461495995e-07,
      "loss": 0.0153,
      "step": 5230
    },
    {
      "epoch": 29.273743016759777,
      "grad_norm": 0.2775328755378723,
      "learning_rate": 2.439649107048719e-07,
      "loss": 0.0112,
      "step": 5240
    },
    {
      "epoch": 29.329608938547487,
      "grad_norm": 0.2872767150402069,
      "learning_rate": 2.0815656508601785e-07,
      "loss": 0.0114,
      "step": 5250
    },
    {
      "epoch": 29.385474860335197,
      "grad_norm": 0.2828880250453949,
      "learning_rate": 1.7518510598422963e-07,
      "loss": 0.0094,
      "step": 5260
    },
    {
      "epoch": 29.441340782122904,
      "grad_norm": 0.27305498719215393,
      "learning_rate": 1.4505178402099692e-07,
      "loss": 0.0119,
      "step": 5270
    },
    {
      "epoch": 29.497206703910614,
      "grad_norm": 0.1336456537246704,
      "learning_rate": 1.1775774216608858e-07,
      "loss": 0.0082,
      "step": 5280
    },
    {
      "epoch": 29.553072625698324,
      "grad_norm": 0.12913921475410461,
      "learning_rate": 9.330401569413725e-08,
      "loss": 0.0094,
      "step": 5290
    },
    {
      "epoch": 29.608938547486034,
      "grad_norm": 0.193023219704628,
      "learning_rate": 7.169153214543754e-08,
      "loss": 0.0094,
      "step": 5300
    },
    {
      "epoch": 29.664804469273744,
      "grad_norm": 0.1686532348394394,
      "learning_rate": 5.292111129068244e-08,
      "loss": 0.0113,
      "step": 5310
    },
    {
      "epoch": 29.720670391061454,
      "grad_norm": 0.18468423187732697,
      "learning_rate": 3.69934650999465e-08,
      "loss": 0.0115,
      "step": 5320
    },
    {
      "epoch": 29.776536312849164,
      "grad_norm": 0.21080362796783447,
      "learning_rate": 2.3909197715657468e-08,
      "loss": 0.0126,
      "step": 5330
    },
    {
      "epoch": 29.83240223463687,
      "grad_norm": 0.15631383657455444,
      "learning_rate": 1.3668805429639662e-08,
      "loss": 0.0107,
      "step": 5340
    },
    {
      "epoch": 29.88826815642458,
      "grad_norm": 0.16603289544582367,
      "learning_rate": 6.27267666434561e-09,
      "loss": 0.0123,
      "step": 5350
    },
    {
      "epoch": 29.94413407821229,
      "grad_norm": 0.12879528105258942,
      "learning_rate": 1.7210919580928906e-09,
      "loss": 0.0119,
      "step": 5360
    },
    {
      "epoch": 30.0,
      "grad_norm": 0.17963722348213196,
      "learning_rate": 1.4223954430958584e-11,
      "loss": 0.0121,
      "step": 5370
    },
    {
      "epoch": 30.0,
      "step": 5370,
      "total_flos": 0.0,
      "train_loss": 0.03160388583327805,
      "train_runtime": 6237.9083,
      "train_samples_per_second": 42.139,
      "train_steps_per_second": 0.861
    }
  ],
  "logging_steps": 10,
  "max_steps": 5370,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 30,
  "save_steps": 20000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 49,
  "trial_name": null,
  "trial_params": null
}