{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.990259740259741,
  "eval_steps": 500,
  "global_step": 1537,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.032467532467532464,
      "grad_norm": 11.862669944763184,
      "learning_rate": 2.5974025974025972e-05,
      "loss": 2.2746,
      "step": 10
    },
    {
      "epoch": 0.06493506493506493,
      "grad_norm": 14.773530006408691,
      "learning_rate": 5.1948051948051944e-05,
      "loss": 0.6306,
      "step": 20
    },
    {
      "epoch": 0.09740259740259741,
      "grad_norm": 2.444714069366455,
      "learning_rate": 7.792207792207793e-05,
      "loss": 0.2859,
      "step": 30
    },
    {
      "epoch": 0.12987012987012986,
      "grad_norm": 1.5860096216201782,
      "learning_rate": 0.00010389610389610389,
      "loss": 0.1714,
      "step": 40
    },
    {
      "epoch": 0.16233766233766234,
      "grad_norm": 1.3271065950393677,
      "learning_rate": 0.00012987012987012987,
      "loss": 0.1621,
      "step": 50
    },
    {
      "epoch": 0.19480519480519481,
      "grad_norm": 1.9922699928283691,
      "learning_rate": 0.00015584415584415587,
      "loss": 0.164,
      "step": 60
    },
    {
      "epoch": 0.22727272727272727,
      "grad_norm": 1.0023343563079834,
      "learning_rate": 0.00018181818181818183,
      "loss": 0.1308,
      "step": 70
    },
    {
      "epoch": 0.2597402597402597,
      "grad_norm": 0.9265744686126709,
      "learning_rate": 0.00019999791644466247,
      "loss": 0.1217,
      "step": 80
    },
    {
      "epoch": 0.2922077922077922,
      "grad_norm": 1.0755594968795776,
      "learning_rate": 0.00019996087787618624,
      "loss": 0.1126,
      "step": 90
    },
    {
      "epoch": 0.3246753246753247,
      "grad_norm": 1.6246726512908936,
      "learning_rate": 0.00019987755781707668,
      "loss": 0.1066,
      "step": 100
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 1.3438688516616821,
      "learning_rate": 0.00019974799484419106,
      "loss": 0.0931,
      "step": 110
    },
    {
      "epoch": 0.38961038961038963,
      "grad_norm": 1.0927611589431763,
      "learning_rate": 0.0001995722489446729,
      "loss": 0.0955,
      "step": 120
    },
    {
      "epoch": 0.42207792207792205,
      "grad_norm": 0.5384888648986816,
      "learning_rate": 0.0001993504014881777,
      "loss": 0.0543,
      "step": 130
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 1.163609504699707,
      "learning_rate": 0.0001990825551891994,
      "loss": 0.0649,
      "step": 140
    },
    {
      "epoch": 0.487012987012987,
      "grad_norm": 1.2434951066970825,
      "learning_rate": 0.00019876883405951377,
      "loss": 0.0798,
      "step": 150
    },
    {
      "epoch": 0.5194805194805194,
      "grad_norm": 0.865424633026123,
      "learning_rate": 0.00019840938335076162,
      "loss": 0.082,
      "step": 160
    },
    {
      "epoch": 0.551948051948052,
      "grad_norm": 0.8742974400520325,
      "learning_rate": 0.00019800436948719775,
      "loss": 0.0725,
      "step": 170
    },
    {
      "epoch": 0.5844155844155844,
      "grad_norm": 0.5820501446723938,
      "learning_rate": 0.0001975539799886372,
      "loss": 0.0602,
      "step": 180
    },
    {
      "epoch": 0.6168831168831169,
      "grad_norm": 0.6646164655685425,
      "learning_rate": 0.00019705842338363434,
      "loss": 0.0614,
      "step": 190
    },
    {
      "epoch": 0.6493506493506493,
      "grad_norm": 1.1510143280029297,
      "learning_rate": 0.00019651792911293508,
      "loss": 0.066,
      "step": 200
    },
    {
      "epoch": 0.6818181818181818,
      "grad_norm": 0.9399957060813904,
      "learning_rate": 0.0001959327474232464,
      "loss": 0.0627,
      "step": 210
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 0.9072800874710083,
      "learning_rate": 0.00019530314925137355,
      "loss": 0.0666,
      "step": 220
    },
    {
      "epoch": 0.7467532467532467,
      "grad_norm": 0.8320746421813965,
      "learning_rate": 0.00019462942609877696,
      "loss": 0.0518,
      "step": 230
    },
    {
      "epoch": 0.7792207792207793,
      "grad_norm": 0.820354163646698,
      "learning_rate": 0.00019391188989660848,
      "loss": 0.0857,
      "step": 240
    },
    {
      "epoch": 0.8116883116883117,
      "grad_norm": 0.9190356731414795,
      "learning_rate": 0.0001931508728612885,
      "loss": 0.0651,
      "step": 250
    },
    {
      "epoch": 0.8441558441558441,
      "grad_norm": 0.6117440462112427,
      "learning_rate": 0.0001923467273406908,
      "loss": 0.0526,
      "step": 260
    },
    {
      "epoch": 0.8766233766233766,
      "grad_norm": 0.8513985276222229,
      "learning_rate": 0.00019149982565100728,
      "loss": 0.0549,
      "step": 270
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.5364561080932617,
      "learning_rate": 0.00019061055990436665,
      "loss": 0.0612,
      "step": 280
    },
    {
      "epoch": 0.9415584415584416,
      "grad_norm": 0.8399089574813843,
      "learning_rate": 0.00018967934182728837,
      "loss": 0.0417,
      "step": 290
    },
    {
      "epoch": 0.974025974025974,
      "grad_norm": 0.7521604299545288,
      "learning_rate": 0.00018870660257005466,
      "loss": 0.0535,
      "step": 300
    },
    {
      "epoch": 1.0064935064935066,
      "grad_norm": 0.5698351263999939,
      "learning_rate": 0.00018769279250708976,
      "loss": 0.0489,
      "step": 310
    },
    {
      "epoch": 1.0389610389610389,
      "grad_norm": 0.6740209460258484,
      "learning_rate": 0.0001866383810284384,
      "loss": 0.0569,
      "step": 320
    },
    {
      "epoch": 1.0714285714285714,
      "grad_norm": 0.49634265899658203,
      "learning_rate": 0.00018554385632244,
      "loss": 0.0561,
      "step": 330
    },
    {
      "epoch": 1.103896103896104,
      "grad_norm": 0.6321184039115906,
      "learning_rate": 0.00018440972514969926,
      "loss": 0.0413,
      "step": 340
    },
    {
      "epoch": 1.1363636363636362,
      "grad_norm": 0.41812852025032043,
      "learning_rate": 0.00018323651260845832,
      "loss": 0.0411,
      "step": 350
    },
    {
      "epoch": 1.1688311688311688,
      "grad_norm": 0.3897482454776764,
      "learning_rate": 0.00018202476189147803,
      "loss": 0.0505,
      "step": 360
    },
    {
      "epoch": 1.2012987012987013,
      "grad_norm": 0.6317518353462219,
      "learning_rate": 0.00018077503403454216,
      "loss": 0.0387,
      "step": 370
    },
    {
      "epoch": 1.2337662337662338,
      "grad_norm": 0.49315449595451355,
      "learning_rate": 0.0001794879076566998,
      "loss": 0.0458,
      "step": 380
    },
    {
      "epoch": 1.2662337662337662,
      "grad_norm": 0.522301971912384,
      "learning_rate": 0.00017816397869236717,
      "loss": 0.0509,
      "step": 390
    },
    {
      "epoch": 1.2987012987012987,
      "grad_norm": 0.6012665629386902,
      "learning_rate": 0.00017680386011541222,
      "loss": 0.0379,
      "step": 400
    },
    {
      "epoch": 1.3311688311688312,
      "grad_norm": 0.783263087272644,
      "learning_rate": 0.0001754081816553504,
      "loss": 0.047,
      "step": 410
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 0.7606521248817444,
      "learning_rate": 0.00017397758950578207,
      "loss": 0.0733,
      "step": 420
    },
    {
      "epoch": 1.396103896103896,
      "grad_norm": 0.4706479012966156,
      "learning_rate": 0.00017251274602520766,
      "loss": 0.0556,
      "step": 430
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.5615910887718201,
      "learning_rate": 0.00017101432943035825,
      "loss": 0.0491,
      "step": 440
    },
    {
      "epoch": 1.4610389610389611,
      "grad_norm": 0.6203674674034119,
      "learning_rate": 0.0001694830334821838,
      "loss": 0.0414,
      "step": 450
    },
    {
      "epoch": 1.4935064935064934,
      "grad_norm": 0.6230823397636414,
      "learning_rate": 0.00016791956716464472,
      "loss": 0.0484,
      "step": 460
    },
    {
      "epoch": 1.525974025974026,
      "grad_norm": 0.49440866708755493,
      "learning_rate": 0.0001663246543564551,
      "loss": 0.056,
      "step": 470
    },
    {
      "epoch": 1.5584415584415585,
      "grad_norm": 0.5235504508018494,
      "learning_rate": 0.00016469903349592956,
      "loss": 0.036,
      "step": 480
    },
    {
      "epoch": 1.5909090909090908,
      "grad_norm": 0.7212300300598145,
      "learning_rate": 0.0001630434572390895,
      "loss": 0.0464,
      "step": 490
    },
    {
      "epoch": 1.6233766233766234,
      "grad_norm": 0.45477214455604553,
      "learning_rate": 0.00016135869211118603,
      "loss": 0.04,
      "step": 500
    },
    {
      "epoch": 1.655844155844156,
      "grad_norm": 0.6098753809928894,
      "learning_rate": 0.00015964551815180213,
      "loss": 0.0396,
      "step": 510
    },
    {
      "epoch": 1.6883116883116882,
      "grad_norm": 1.0758476257324219,
      "learning_rate": 0.00015790472855369716,
      "loss": 0.0567,
      "step": 520
    },
    {
      "epoch": 1.7207792207792207,
      "grad_norm": 0.8209078311920166,
      "learning_rate": 0.00015613712929556193,
      "loss": 0.0506,
      "step": 530
    },
    {
      "epoch": 1.7532467532467533,
      "grad_norm": 0.7815719246864319,
      "learning_rate": 0.00015434353876885362,
      "loss": 0.0474,
      "step": 540
    },
    {
      "epoch": 1.7857142857142856,
      "grad_norm": 0.3328838050365448,
      "learning_rate": 0.00015252478739888385,
      "loss": 0.0333,
      "step": 550
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.6883902549743652,
      "learning_rate": 0.0001506817172603351,
      "loss": 0.0342,
      "step": 560
    },
    {
      "epoch": 1.8506493506493507,
      "grad_norm": 0.4302406311035156,
      "learning_rate": 0.0001488151816873834,
      "loss": 0.0415,
      "step": 570
    },
    {
      "epoch": 1.883116883116883,
      "grad_norm": 0.5647683143615723,
      "learning_rate": 0.00014692604487860785,
      "loss": 0.0492,
      "step": 580
    },
    {
      "epoch": 1.9155844155844157,
      "grad_norm": 0.46459880471229553,
      "learning_rate": 0.00014501518149687042,
      "loss": 0.035,
      "step": 590
    },
    {
      "epoch": 1.948051948051948,
      "grad_norm": 0.49570703506469727,
      "learning_rate": 0.0001430834762643502,
      "loss": 0.0402,
      "step": 600
    },
    {
      "epoch": 1.9805194805194806,
      "grad_norm": 0.44045427441596985,
      "learning_rate": 0.00014113182355292078,
      "loss": 0.045,
      "step": 610
    },
    {
      "epoch": 2.012987012987013,
      "grad_norm": 0.22453299164772034,
      "learning_rate": 0.0001391611269700594,
      "loss": 0.0419,
      "step": 620
    },
    {
      "epoch": 2.0454545454545454,
      "grad_norm": 0.4719075560569763,
      "learning_rate": 0.00013717229894048038,
      "loss": 0.037,
      "step": 630
    },
    {
      "epoch": 2.0779220779220777,
      "grad_norm": 0.4710325598716736,
      "learning_rate": 0.0001351662602836861,
      "loss": 0.0368,
      "step": 640
    },
    {
      "epoch": 2.1103896103896105,
      "grad_norm": 0.5058510303497314,
      "learning_rate": 0.0001331439397876312,
      "loss": 0.0385,
      "step": 650
    },
    {
      "epoch": 2.142857142857143,
      "grad_norm": 0.5033309459686279,
      "learning_rate": 0.0001311062737786974,
      "loss": 0.0357,
      "step": 660
    },
    {
      "epoch": 2.175324675324675,
      "grad_norm": 0.36022046208381653,
      "learning_rate": 0.0001290542056881781,
      "loss": 0.0243,
      "step": 670
    },
    {
      "epoch": 2.207792207792208,
      "grad_norm": 0.3985787034034729,
      "learning_rate": 0.0001269886856154735,
      "loss": 0.0471,
      "step": 680
    },
    {
      "epoch": 2.24025974025974,
      "grad_norm": 0.47278791666030884,
      "learning_rate": 0.0001249106698881982,
      "loss": 0.0353,
      "step": 690
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 0.5603681802749634,
      "learning_rate": 0.0001228211206194055,
      "loss": 0.0324,
      "step": 700
    },
    {
      "epoch": 2.3051948051948052,
      "grad_norm": 0.28177714347839355,
      "learning_rate": 0.0001207210052621327,
      "loss": 0.0432,
      "step": 710
    },
    {
      "epoch": 2.3376623376623376,
      "grad_norm": 0.5803484320640564,
      "learning_rate": 0.00011861129616147418,
      "loss": 0.0322,
      "step": 720
    },
    {
      "epoch": 2.3701298701298703,
      "grad_norm": 0.298414409160614,
      "learning_rate": 0.00011649297010438956,
      "loss": 0.0428,
      "step": 730
    },
    {
      "epoch": 2.4025974025974026,
      "grad_norm": 0.21691694855690002,
      "learning_rate": 0.00011436700786745515,
      "loss": 0.0243,
      "step": 740
    },
    {
      "epoch": 2.435064935064935,
      "grad_norm": 0.5338732600212097,
      "learning_rate": 0.00011223439376276835,
      "loss": 0.0325,
      "step": 750
    },
    {
      "epoch": 2.4675324675324677,
      "grad_norm": 0.5570558905601501,
      "learning_rate": 0.00011009611518221489,
      "loss": 0.0368,
      "step": 760
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.4559939503669739,
      "learning_rate": 0.00010795316214031048,
      "loss": 0.0268,
      "step": 770
    },
    {
      "epoch": 2.5324675324675323,
      "grad_norm": 0.33397603034973145,
      "learning_rate": 0.0001058065268158279,
      "loss": 0.0359,
      "step": 780
    },
    {
      "epoch": 2.564935064935065,
      "grad_norm": 0.22799064218997955,
      "learning_rate": 0.00010365720309242217,
      "loss": 0.0264,
      "step": 790
    },
    {
      "epoch": 2.5974025974025974,
      "grad_norm": 0.38165828585624695,
      "learning_rate": 0.00010150618609846638,
      "loss": 0.0301,
      "step": 800
    },
    {
      "epoch": 2.62987012987013,
      "grad_norm": 0.5442892909049988,
      "learning_rate": 9.935447174631119e-05,
      "loss": 0.0238,
      "step": 810
    },
    {
      "epoch": 2.6623376623376624,
      "grad_norm": 0.22587241232395172,
      "learning_rate": 9.720305627118126e-05,
      "loss": 0.0337,
      "step": 820
    },
    {
      "epoch": 2.6948051948051948,
      "grad_norm": 0.2878405451774597,
      "learning_rate": 9.505293576992251e-05,
      "loss": 0.0253,
      "step": 830
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.5154989957809448,
      "learning_rate": 9.290510573981298e-05,
      "loss": 0.0279,
      "step": 840
    },
    {
      "epoch": 2.75974025974026,
      "grad_norm": 0.42716383934020996,
      "learning_rate": 9.076056061765173e-05,
      "loss": 0.0358,
      "step": 850
    },
    {
      "epoch": 2.792207792207792,
      "grad_norm": 0.6430768370628357,
      "learning_rate": 8.862029331933828e-05,
      "loss": 0.0315,
      "step": 860
    },
    {
      "epoch": 2.824675324675325,
      "grad_norm": 0.6324707269668579,
      "learning_rate": 8.648529478015685e-05,
      "loss": 0.0305,
      "step": 870
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.27602899074554443,
      "learning_rate": 8.435655349597689e-05,
      "loss": 0.026,
      "step": 880
    },
    {
      "epoch": 2.8896103896103895,
      "grad_norm": 0.5011958479881287,
      "learning_rate": 8.223505506558375e-05,
      "loss": 0.0274,
      "step": 890
    },
    {
      "epoch": 2.9220779220779223,
      "grad_norm": 0.3036370277404785,
      "learning_rate": 8.012178173434986e-05,
      "loss": 0.0306,
      "step": 900
    },
    {
      "epoch": 2.9545454545454546,
      "grad_norm": 0.3664824366569519,
      "learning_rate": 7.80177119394596e-05,
      "loss": 0.0241,
      "step": 910
    },
    {
      "epoch": 2.987012987012987,
      "grad_norm": 0.43233174085617065,
      "learning_rate": 7.592381985689628e-05,
      "loss": 0.0309,
      "step": 920
    },
    {
      "epoch": 3.0194805194805197,
      "grad_norm": 0.3537318706512451,
      "learning_rate": 7.384107495040284e-05,
      "loss": 0.0321,
      "step": 930
    },
    {
      "epoch": 3.051948051948052,
      "grad_norm": 0.2594110071659088,
      "learning_rate": 7.177044152262369e-05,
      "loss": 0.0245,
      "step": 940
    },
    {
      "epoch": 3.0844155844155843,
      "grad_norm": 0.3139798939228058,
      "learning_rate": 6.97128782686367e-05,
      "loss": 0.0218,
      "step": 950
    },
    {
      "epoch": 3.116883116883117,
      "grad_norm": 0.2899845540523529,
      "learning_rate": 6.766933783208092e-05,
      "loss": 0.029,
      "step": 960
    },
    {
      "epoch": 3.1493506493506493,
      "grad_norm": 0.320487916469574,
      "learning_rate": 6.564076636408656e-05,
      "loss": 0.0226,
      "step": 970
    },
    {
      "epoch": 3.1818181818181817,
      "grad_norm": 0.4478653371334076,
      "learning_rate": 6.362810308521054e-05,
      "loss": 0.0224,
      "step": 980
    },
    {
      "epoch": 3.2142857142857144,
      "grad_norm": 0.20480212569236755,
      "learning_rate": 6.16322798505813e-05,
      "loss": 0.0224,
      "step": 990
    },
    {
      "epoch": 3.2467532467532467,
      "grad_norm": 0.4546569287776947,
      "learning_rate": 5.9654220718453424e-05,
      "loss": 0.0311,
      "step": 1000
    },
    {
      "epoch": 3.279220779220779,
      "grad_norm": 0.2200387865304947,
      "learning_rate": 5.769484152237249e-05,
      "loss": 0.026,
      "step": 1010
    },
    {
      "epoch": 3.311688311688312,
      "grad_norm": 0.40406161546707153,
      "learning_rate": 5.5755049447147444e-05,
      "loss": 0.028,
      "step": 1020
    },
    {
      "epoch": 3.344155844155844,
      "grad_norm": 0.41747692227363586,
      "learning_rate": 5.383574260882802e-05,
      "loss": 0.0255,
      "step": 1030
    },
    {
      "epoch": 3.3766233766233764,
      "grad_norm": 0.4012701213359833,
      "learning_rate": 5.1937809638880374e-05,
      "loss": 0.0289,
      "step": 1040
    },
    {
      "epoch": 3.409090909090909,
      "grad_norm": 0.49661171436309814,
      "learning_rate": 5.0062129272754146e-05,
      "loss": 0.0249,
      "step": 1050
    },
    {
      "epoch": 3.4415584415584415,
      "grad_norm": 0.225033700466156,
      "learning_rate": 4.8209569943031516e-05,
      "loss": 0.0274,
      "step": 1060
    },
    {
      "epoch": 3.474025974025974,
      "grad_norm": 0.2046009600162506,
      "learning_rate": 4.638098937734648e-05,
      "loss": 0.0232,
      "step": 1070
    },
    {
      "epoch": 3.5064935064935066,
      "grad_norm": 0.4959232211112976,
      "learning_rate": 4.4577234201260196e-05,
      "loss": 0.0233,
      "step": 1080
    },
    {
      "epoch": 3.538961038961039,
      "grad_norm": 0.42567142844200134,
      "learning_rate": 4.279913954627667e-05,
      "loss": 0.0188,
      "step": 1090
    },
    {
      "epoch": 3.571428571428571,
      "grad_norm": 0.22492988407611847,
      "learning_rate": 4.104752866318026e-05,
      "loss": 0.0178,
      "step": 1100
    },
    {
      "epoch": 3.603896103896104,
      "grad_norm": 0.2631309926509857,
      "learning_rate": 3.932321254087389e-05,
      "loss": 0.0187,
      "step": 1110
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 0.45389196276664734,
      "learning_rate": 3.7626989530894285e-05,
      "loss": 0.0211,
      "step": 1120
    },
    {
      "epoch": 3.6688311688311686,
      "grad_norm": 0.20797108113765717,
      "learning_rate": 3.5959644977778386e-05,
      "loss": 0.024,
      "step": 1130
    },
    {
      "epoch": 3.7012987012987013,
      "grad_norm": 0.22460906207561493,
      "learning_rate": 3.432195085545191e-05,
      "loss": 0.0161,
      "step": 1140
    },
    {
      "epoch": 3.7337662337662336,
      "grad_norm": 0.3012770116329193,
      "learning_rate": 3.271466540980862e-05,
      "loss": 0.0232,
      "step": 1150
    },
    {
      "epoch": 3.7662337662337664,
      "grad_norm": 0.38262003660202026,
      "learning_rate": 3.1138532807645394e-05,
      "loss": 0.0174,
      "step": 1160
    },
    {
      "epoch": 3.7987012987012987,
      "grad_norm": 0.22977425158023834,
      "learning_rate": 2.9594282792115857e-05,
      "loss": 0.0227,
      "step": 1170
    },
    {
      "epoch": 3.8311688311688314,
      "grad_norm": 0.16646578907966614,
      "learning_rate": 2.808263034486226e-05,
      "loss": 0.0155,
      "step": 1180
    },
    {
      "epoch": 3.8636363636363638,
      "grad_norm": 0.257495254278183,
      "learning_rate": 2.660427535498191e-05,
      "loss": 0.0218,
      "step": 1190
    },
    {
      "epoch": 3.896103896103896,
      "grad_norm": 0.3283085227012634,
      "learning_rate": 2.5159902294981197e-05,
      "loss": 0.0224,
      "step": 1200
    },
    {
      "epoch": 3.928571428571429,
      "grad_norm": 0.36105039715766907,
      "learning_rate": 2.3750179903867443e-05,
      "loss": 0.0278,
      "step": 1210
    },
    {
      "epoch": 3.961038961038961,
      "grad_norm": 0.3740982115268707,
      "learning_rate": 2.237576087752554e-05,
      "loss": 0.0241,
      "step": 1220
    },
    {
      "epoch": 3.9935064935064934,
      "grad_norm": 0.19882728159427643,
      "learning_rate": 2.1037281566522304e-05,
      "loss": 0.0174,
      "step": 1230
    },
    {
      "epoch": 4.025974025974026,
      "grad_norm": 0.12800416350364685,
      "learning_rate": 1.973536168147867e-05,
      "loss": 0.0137,
      "step": 1240
    },
    {
      "epoch": 4.058441558441558,
      "grad_norm": 0.352993905544281,
      "learning_rate": 1.8470604006146064e-05,
      "loss": 0.0221,
      "step": 1250
    },
    {
      "epoch": 4.090909090909091,
      "grad_norm": 0.3981715738773346,
      "learning_rate": 1.7243594118319985e-05,
      "loss": 0.0162,
      "step": 1260
    },
    {
      "epoch": 4.123376623376624,
      "grad_norm": 0.23238126933574677,
      "learning_rate": 1.6054900118719807e-05,
      "loss": 0.0201,
      "step": 1270
    },
    {
      "epoch": 4.1558441558441555,
      "grad_norm": 0.3487814962863922,
      "learning_rate": 1.4905072367960437e-05,
      "loss": 0.0182,
      "step": 1280
    },
    {
      "epoch": 4.188311688311688,
      "grad_norm": 0.2413235455751419,
      "learning_rate": 1.3794643231737348e-05,
      "loss": 0.0161,
      "step": 1290
    },
    {
      "epoch": 4.220779220779221,
      "grad_norm": 0.3678736984729767,
      "learning_rate": 1.2724126834343564e-05,
      "loss": 0.02,
      "step": 1300
    },
    {
      "epoch": 4.253246753246753,
      "grad_norm": 0.0916026160120964,
      "learning_rate": 1.1694018820632068e-05,
      "loss": 0.0179,
      "step": 1310
    },
    {
      "epoch": 4.285714285714286,
      "grad_norm": 0.16847121715545654,
      "learning_rate": 1.0704796126534234e-05,
      "loss": 0.0182,
      "step": 1320
    },
    {
      "epoch": 4.318181818181818,
      "grad_norm": 0.10810457170009613,
      "learning_rate": 9.756916758240286e-06,
      "loss": 0.0219,
      "step": 1330
    },
    {
      "epoch": 4.35064935064935,
      "grad_norm": 0.31443580985069275,
      "learning_rate": 8.850819580144387e-06,
      "loss": 0.0254,
      "step": 1340
    },
    {
      "epoch": 4.383116883116883,
      "grad_norm": 0.2282458394765854,
      "learning_rate": 7.986924111652006e-06,
      "loss": 0.0177,
      "step": 1350
    },
    {
      "epoch": 4.415584415584416,
      "grad_norm": 0.34387949109077454,
      "learning_rate": 7.16563033294424e-06,
      "loss": 0.0167,
      "step": 1360
    },
    {
      "epoch": 4.448051948051948,
      "grad_norm": 0.12599095702171326,
      "learning_rate": 6.387318499788497e-06,
      "loss": 0.0205,
      "step": 1370
    },
    {
      "epoch": 4.48051948051948,
      "grad_norm": 0.14966098964214325,
      "learning_rate": 5.652348967481569e-06,
      "loss": 0.0201,
      "step": 1380
    },
    {
      "epoch": 4.512987012987013,
      "grad_norm": 0.28549131751060486,
      "learning_rate": 4.961062024006591e-06,
      "loss": 0.019,
      "step": 1390
    },
    {
      "epoch": 4.545454545454545,
      "grad_norm": 0.22848688066005707,
      "learning_rate": 4.313777732481039e-06,
      "loss": 0.0182,
      "step": 1400
    },
    {
      "epoch": 4.577922077922078,
      "grad_norm": 0.20382894575595856,
      "learning_rate": 3.7107957829688234e-06,
      "loss": 0.0228,
      "step": 1410
    },
    {
      "epoch": 4.6103896103896105,
      "grad_norm": 0.1720917969942093,
      "learning_rate": 3.1523953537248684e-06,
      "loss": 0.0189,
      "step": 1420
    },
    {
      "epoch": 4.642857142857143,
      "grad_norm": 0.10501822084188461,
      "learning_rate": 2.638834981936744e-06,
      "loss": 0.015,
      "step": 1430
    },
    {
      "epoch": 4.675324675324675,
      "grad_norm": 0.13112539052963257,
      "learning_rate": 2.1703524440230383e-06,
      "loss": 0.0118,
      "step": 1440
    },
    {
      "epoch": 4.707792207792208,
      "grad_norm": 0.17099051177501678,
      "learning_rate": 1.7471646455437085e-06,
      "loss": 0.0151,
      "step": 1450
    },
    {
      "epoch": 4.740259740259741,
      "grad_norm": 0.3271525204181671,
      "learning_rate": 1.3694675207737151e-06,
      "loss": 0.0223,
      "step": 1460
    },
    {
      "epoch": 4.7727272727272725,
      "grad_norm": 0.35389673709869385,
      "learning_rate": 1.03743594198622e-06,
      "loss": 0.0133,
      "step": 1470
    },
    {
      "epoch": 4.805194805194805,
      "grad_norm": 0.25904619693756104,
      "learning_rate": 7.512236384874305e-07,
      "loss": 0.0166,
      "step": 1480
    },
    {
      "epoch": 4.837662337662338,
      "grad_norm": 0.14534501731395721,
      "learning_rate": 5.109631254405445e-07,
      "loss": 0.0133,
      "step": 1490
    },
    {
      "epoch": 4.87012987012987,
      "grad_norm": 0.17412607371807098,
      "learning_rate": 3.1676564251171824e-07,
      "loss": 0.0188,
      "step": 1500
    },
    {
      "epoch": 4.902597402597403,
      "grad_norm": 0.20664988458156586,
      "learning_rate": 1.687211023665647e-07,
      "loss": 0.0114,
      "step": 1510
    },
    {
      "epoch": 4.935064935064935,
      "grad_norm": 0.19708259403705597,
      "learning_rate": 6.68980490409421e-08,
      "loss": 0.0179,
      "step": 1520
    },
    {
      "epoch": 4.967532467532467,
      "grad_norm": 0.5721464157104492,
      "learning_rate": 1.134362620534013e-08,
      "loss": 0.0192,
      "step": 1530
    },
    {
      "epoch": 4.990259740259741,
      "step": 1537,
      "total_flos": 5.134548117135398e+16,
      "train_loss": 0.05982626907687413,
      "train_runtime": 715.8651,
      "train_samples_per_second": 34.353,
      "train_steps_per_second": 2.147
    }
  ],
  "logging_steps": 10,
  "max_steps": 1537,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.134548117135398e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}