{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 1000,
  "global_step": 2975,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01680672268907563,
      "grad_norm": 1.6710422039031982,
      "learning_rate": 6.7114093959731546e-06,
      "loss": 1.2952,
      "step": 10
    },
    {
      "epoch": 0.03361344537815126,
      "grad_norm": 1.708608865737915,
      "learning_rate": 1.3422818791946309e-05,
      "loss": 1.4041,
      "step": 20
    },
    {
      "epoch": 0.05042016806722689,
      "grad_norm": 1.9668169021606445,
      "learning_rate": 2.013422818791946e-05,
      "loss": 1.2144,
      "step": 30
    },
    {
      "epoch": 0.06722689075630252,
      "grad_norm": 1.0957667827606201,
      "learning_rate": 2.6845637583892618e-05,
      "loss": 0.8496,
      "step": 40
    },
    {
      "epoch": 0.08403361344537816,
      "grad_norm": 0.7974053621292114,
      "learning_rate": 3.3557046979865775e-05,
      "loss": 0.642,
      "step": 50
    },
    {
      "epoch": 0.10084033613445378,
      "grad_norm": 0.6723673939704895,
      "learning_rate": 4.026845637583892e-05,
      "loss": 0.5978,
      "step": 60
    },
    {
      "epoch": 0.11764705882352941,
      "grad_norm": 0.7411757707595825,
      "learning_rate": 4.697986577181208e-05,
      "loss": 0.5342,
      "step": 70
    },
    {
      "epoch": 0.13445378151260504,
      "grad_norm": 0.7030589580535889,
      "learning_rate": 5.3691275167785237e-05,
      "loss": 0.536,
      "step": 80
    },
    {
      "epoch": 0.15126050420168066,
      "grad_norm": 0.5388085842132568,
      "learning_rate": 6.04026845637584e-05,
      "loss": 0.5294,
      "step": 90
    },
    {
      "epoch": 0.16806722689075632,
      "grad_norm": 0.5648273229598999,
      "learning_rate": 6.711409395973155e-05,
      "loss": 0.5288,
      "step": 100
    },
    {
      "epoch": 0.18487394957983194,
      "grad_norm": 0.5652063488960266,
      "learning_rate": 7.382550335570471e-05,
      "loss": 0.5467,
      "step": 110
    },
    {
      "epoch": 0.20168067226890757,
      "grad_norm": 0.5379688739776611,
      "learning_rate": 8.053691275167784e-05,
      "loss": 0.5057,
      "step": 120
    },
    {
      "epoch": 0.2184873949579832,
      "grad_norm": 0.7668149471282959,
      "learning_rate": 8.7248322147651e-05,
      "loss": 0.5094,
      "step": 130
    },
    {
      "epoch": 0.23529411764705882,
      "grad_norm": 0.5256843566894531,
      "learning_rate": 9.395973154362417e-05,
      "loss": 0.5195,
      "step": 140
    },
    {
      "epoch": 0.25210084033613445,
      "grad_norm": 0.5416532754898071,
      "learning_rate": 0.00010067114093959733,
      "loss": 0.4653,
      "step": 150
    },
    {
      "epoch": 0.2689075630252101,
      "grad_norm": 0.474119633436203,
      "learning_rate": 0.00010738255033557047,
      "loss": 0.4979,
      "step": 160
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 0.5296077132225037,
      "learning_rate": 0.00011409395973154362,
      "loss": 0.5004,
      "step": 170
    },
    {
      "epoch": 0.3025210084033613,
      "grad_norm": 0.4529200494289398,
      "learning_rate": 0.0001208053691275168,
      "loss": 0.4975,
      "step": 180
    },
    {
      "epoch": 0.31932773109243695,
      "grad_norm": 0.4488178491592407,
      "learning_rate": 0.00012751677852348994,
      "loss": 0.4779,
      "step": 190
    },
    {
      "epoch": 0.33613445378151263,
      "grad_norm": 0.4906866252422333,
      "learning_rate": 0.0001342281879194631,
      "loss": 0.5172,
      "step": 200
    },
    {
      "epoch": 0.35294117647058826,
      "grad_norm": 0.647227942943573,
      "learning_rate": 0.00014093959731543624,
      "loss": 0.4858,
      "step": 210
    },
    {
      "epoch": 0.3697478991596639,
      "grad_norm": 0.5192333459854126,
      "learning_rate": 0.00014765100671140942,
      "loss": 0.5518,
      "step": 220
    },
    {
      "epoch": 0.3865546218487395,
      "grad_norm": 0.4241911470890045,
      "learning_rate": 0.00015436241610738256,
      "loss": 0.4668,
      "step": 230
    },
    {
      "epoch": 0.40336134453781514,
      "grad_norm": 0.49162909388542175,
      "learning_rate": 0.0001610738255033557,
      "loss": 0.5524,
      "step": 240
    },
    {
      "epoch": 0.42016806722689076,
      "grad_norm": 0.45716652274131775,
      "learning_rate": 0.00016778523489932888,
      "loss": 0.4681,
      "step": 250
    },
    {
      "epoch": 0.4369747899159664,
      "grad_norm": 0.46797800064086914,
      "learning_rate": 0.000174496644295302,
      "loss": 0.4922,
      "step": 260
    },
    {
      "epoch": 0.453781512605042,
      "grad_norm": 0.5456430315971375,
      "learning_rate": 0.00018120805369127517,
      "loss": 0.4656,
      "step": 270
    },
    {
      "epoch": 0.47058823529411764,
      "grad_norm": 0.41116565465927124,
      "learning_rate": 0.00018791946308724833,
      "loss": 0.4973,
      "step": 280
    },
    {
      "epoch": 0.48739495798319327,
      "grad_norm": 0.47659677267074585,
      "learning_rate": 0.0001946308724832215,
      "loss": 0.4703,
      "step": 290
    },
    {
      "epoch": 0.5042016806722689,
      "grad_norm": 0.4758327007293701,
      "learning_rate": 0.00019999972455632807,
      "loss": 0.4574,
      "step": 300
    },
    {
      "epoch": 0.5210084033613446,
      "grad_norm": 0.38325995206832886,
      "learning_rate": 0.00019999008418713482,
      "loss": 0.4596,
      "step": 310
    },
    {
      "epoch": 0.5378151260504201,
      "grad_norm": 0.5056403875350952,
      "learning_rate": 0.00019996667315169163,
      "loss": 0.4853,
      "step": 320
    },
    {
      "epoch": 0.5546218487394958,
      "grad_norm": 0.49656251072883606,
      "learning_rate": 0.00019992949467417378,
      "loss": 0.4407,
      "step": 330
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 0.49380919337272644,
      "learning_rate": 0.00019987855387481305,
      "loss": 0.4614,
      "step": 340
    },
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 0.47169238328933716,
      "learning_rate": 0.0001998138577691925,
      "loss": 0.4619,
      "step": 350
    },
    {
      "epoch": 0.6050420168067226,
      "grad_norm": 0.34966766834259033,
      "learning_rate": 0.00019973541526728048,
      "loss": 0.4828,
      "step": 360
    },
    {
      "epoch": 0.6218487394957983,
      "grad_norm": 0.4971473217010498,
      "learning_rate": 0.0001996432371722033,
      "loss": 0.4406,
      "step": 370
    },
    {
      "epoch": 0.6386554621848739,
      "grad_norm": 0.3743559420108795,
      "learning_rate": 0.00019953733617875759,
      "loss": 0.4854,
      "step": 380
    },
    {
      "epoch": 0.6554621848739496,
      "grad_norm": 0.3752104341983795,
      "learning_rate": 0.00019941772687166192,
      "loss": 0.4921,
      "step": 390
    },
    {
      "epoch": 0.6722689075630253,
      "grad_norm": 0.4422326982021332,
      "learning_rate": 0.00019928442572354814,
      "loss": 0.4802,
      "step": 400
    },
    {
      "epoch": 0.6890756302521008,
      "grad_norm": 0.4718156158924103,
      "learning_rate": 0.00019913745109269288,
      "loss": 0.5042,
      "step": 410
    },
    {
      "epoch": 0.7058823529411765,
      "grad_norm": 0.4682941734790802,
      "learning_rate": 0.00019897682322048908,
      "loss": 0.4865,
      "step": 420
    },
    {
      "epoch": 0.7226890756302521,
      "grad_norm": 0.5025270581245422,
      "learning_rate": 0.0001988025642286585,
      "loss": 0.4781,
      "step": 430
    },
    {
      "epoch": 0.7394957983193278,
      "grad_norm": 0.6233184337615967,
      "learning_rate": 0.000198614698116205,
      "loss": 0.5091,
      "step": 440
    },
    {
      "epoch": 0.7563025210084033,
      "grad_norm": 0.354615718126297,
      "learning_rate": 0.00019841325075610944,
      "loss": 0.5004,
      "step": 450
    },
    {
      "epoch": 0.773109243697479,
      "grad_norm": 0.42839425802230835,
      "learning_rate": 0.00019819824989176645,
      "loss": 0.4652,
      "step": 460
    },
    {
      "epoch": 0.7899159663865546,
      "grad_norm": 0.379213809967041,
      "learning_rate": 0.00019796972513316355,
      "loss": 0.4739,
      "step": 470
    },
    {
      "epoch": 0.8067226890756303,
      "grad_norm": 0.4357176423072815,
      "learning_rate": 0.00019772770795280329,
      "loss": 0.5072,
      "step": 480
    },
    {
      "epoch": 0.8235294117647058,
      "grad_norm": 0.4875657856464386,
      "learning_rate": 0.00019747223168136886,
      "loss": 0.5094,
      "step": 490
    },
    {
      "epoch": 0.8403361344537815,
      "grad_norm": 0.4749436676502228,
      "learning_rate": 0.00019720333150313377,
      "loss": 0.4812,
      "step": 500
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 0.446424663066864,
      "learning_rate": 0.0001969210444511162,
      "loss": 0.4823,
      "step": 510
    },
    {
      "epoch": 0.8739495798319328,
      "grad_norm": 0.4699801802635193,
      "learning_rate": 0.00019662540940197892,
      "loss": 0.4559,
      "step": 520
    },
    {
      "epoch": 0.8907563025210085,
      "grad_norm": 0.4818078875541687,
      "learning_rate": 0.00019631646707067511,
      "loss": 0.4803,
      "step": 530
    },
    {
      "epoch": 0.907563025210084,
      "grad_norm": 0.4135447144508362,
      "learning_rate": 0.000195994260004841,
      "loss": 0.4481,
      "step": 540
    },
    {
      "epoch": 0.9243697478991597,
      "grad_norm": 0.3625709116458893,
      "learning_rate": 0.00019565883257893635,
      "loss": 0.4606,
      "step": 550
    },
    {
      "epoch": 0.9411764705882353,
      "grad_norm": 0.6398768424987793,
      "learning_rate": 0.00019531023098813316,
      "loss": 0.4819,
      "step": 560
    },
    {
      "epoch": 0.957983193277311,
      "grad_norm": 0.40339210629463196,
      "learning_rate": 0.00019494850324195353,
      "loss": 0.4742,
      "step": 570
    },
    {
      "epoch": 0.9747899159663865,
      "grad_norm": 0.37650129199028015,
      "learning_rate": 0.00019457369915765798,
      "loss": 0.4514,
      "step": 580
    },
    {
      "epoch": 0.9915966386554622,
      "grad_norm": 0.41677525639533997,
      "learning_rate": 0.00019418587035338432,
      "loss": 0.4849,
      "step": 590
    },
    {
      "epoch": 1.0084033613445378,
      "grad_norm": 0.44225984811782837,
      "learning_rate": 0.0001937850702410391,
      "loss": 0.4626,
      "step": 600
    },
    {
      "epoch": 1.0252100840336134,
      "grad_norm": 0.4943791627883911,
      "learning_rate": 0.00019337135401894146,
      "loss": 0.4441,
      "step": 610
    },
    {
      "epoch": 1.0420168067226891,
      "grad_norm": 0.38774925470352173,
      "learning_rate": 0.0001929447786642213,
      "loss": 0.4287,
      "step": 620
    },
    {
      "epoch": 1.0588235294117647,
      "grad_norm": 0.4631214737892151,
      "learning_rate": 0.00019250540292497235,
      "loss": 0.4369,
      "step": 630
    },
    {
      "epoch": 1.0756302521008403,
      "grad_norm": 0.5732137560844421,
      "learning_rate": 0.00019205328731216147,
      "loss": 0.4119,
      "step": 640
    },
    {
      "epoch": 1.092436974789916,
      "grad_norm": 0.5002462863922119,
      "learning_rate": 0.00019158849409129484,
      "loss": 0.4172,
      "step": 650
    },
    {
      "epoch": 1.1092436974789917,
      "grad_norm": 0.6402649879455566,
      "learning_rate": 0.00019111108727384296,
      "loss": 0.3912,
      "step": 660
    },
    {
      "epoch": 1.1260504201680672,
      "grad_norm": 0.5174903869628906,
      "learning_rate": 0.00019062113260842476,
      "loss": 0.4656,
      "step": 670
    },
    {
      "epoch": 1.1428571428571428,
      "grad_norm": 0.4554598033428192,
      "learning_rate": 0.0001901186975717529,
      "loss": 0.4059,
      "step": 680
    },
    {
      "epoch": 1.1596638655462184,
      "grad_norm": 0.480656236410141,
      "learning_rate": 0.0001896038513593407,
      "loss": 0.4013,
      "step": 690
    },
    {
      "epoch": 1.1764705882352942,
      "grad_norm": 0.40827691555023193,
      "learning_rate": 0.00018907666487597257,
      "loss": 0.4278,
      "step": 700
    },
    {
      "epoch": 1.1932773109243697,
      "grad_norm": 0.4276060163974762,
      "learning_rate": 0.00018853721072593894,
      "loss": 0.4112,
      "step": 710
    },
    {
      "epoch": 1.2100840336134453,
      "grad_norm": 0.5786320567131042,
      "learning_rate": 0.00018798556320303718,
      "loss": 0.4518,
      "step": 720
    },
    {
      "epoch": 1.226890756302521,
      "grad_norm": 0.37131550908088684,
      "learning_rate": 0.00018742179828033992,
      "loss": 0.4132,
      "step": 730
    },
    {
      "epoch": 1.2436974789915967,
      "grad_norm": 0.4482327401638031,
      "learning_rate": 0.00018684599359973194,
      "loss": 0.4196,
      "step": 740
    },
    {
      "epoch": 1.2605042016806722,
      "grad_norm": 0.6399847269058228,
      "learning_rate": 0.00018625822846121727,
      "loss": 0.427,
      "step": 750
    },
    {
      "epoch": 1.2773109243697478,
      "grad_norm": 0.5624536871910095,
      "learning_rate": 0.0001856585838119981,
      "loss": 0.3879,
      "step": 760
    },
    {
      "epoch": 1.2941176470588236,
      "grad_norm": 0.4050937592983246,
      "learning_rate": 0.0001850471422353265,
      "loss": 0.44,
      "step": 770
    },
    {
      "epoch": 1.3109243697478992,
      "grad_norm": 0.4883362054824829,
      "learning_rate": 0.00018442398793913118,
      "loss": 0.416,
      "step": 780
    },
    {
      "epoch": 1.3277310924369747,
      "grad_norm": 0.5154753923416138,
      "learning_rate": 0.00018378920674442045,
      "loss": 0.4266,
      "step": 790
    },
    {
      "epoch": 1.3445378151260505,
      "grad_norm": 0.5751391649246216,
      "learning_rate": 0.00018314288607346258,
      "loss": 0.4331,
      "step": 800
    },
    {
      "epoch": 1.361344537815126,
      "grad_norm": 0.5200420618057251,
      "learning_rate": 0.0001824851149377463,
      "loss": 0.4275,
      "step": 810
    },
    {
      "epoch": 1.3781512605042017,
      "grad_norm": 0.5671929121017456,
      "learning_rate": 0.00018181598392572203,
      "loss": 0.4285,
      "step": 820
    },
    {
      "epoch": 1.3949579831932772,
      "grad_norm": 0.5295830368995667,
      "learning_rate": 0.0001811355851903259,
      "loss": 0.3888,
      "step": 830
    },
    {
      "epoch": 1.4117647058823528,
      "grad_norm": 0.4832970201969147,
      "learning_rate": 0.0001804440124362885,
      "loss": 0.4299,
      "step": 840
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.4614173173904419,
      "learning_rate": 0.00017974136090722985,
      "loss": 0.4012,
      "step": 850
    },
    {
      "epoch": 1.4453781512605042,
      "grad_norm": 0.5086454153060913,
      "learning_rate": 0.00017902772737254238,
      "loss": 0.4103,
      "step": 860
    },
    {
      "epoch": 1.46218487394958,
      "grad_norm": 0.6145858764648438,
      "learning_rate": 0.00017830321011406386,
      "loss": 0.4378,
      "step": 870
    },
    {
      "epoch": 1.4789915966386555,
      "grad_norm": 0.4497447907924652,
      "learning_rate": 0.00017756790891254194,
      "loss": 0.4021,
      "step": 880
    },
    {
      "epoch": 1.495798319327731,
      "grad_norm": 0.4918227195739746,
      "learning_rate": 0.00017682192503389236,
      "loss": 0.4364,
      "step": 890
    },
    {
      "epoch": 1.5126050420168067,
      "grad_norm": 0.7057837247848511,
      "learning_rate": 0.00017606536121525255,
      "loss": 0.3721,
      "step": 900
    },
    {
      "epoch": 1.5294117647058822,
      "grad_norm": 0.5684515237808228,
      "learning_rate": 0.00017529832165083265,
      "loss": 0.4458,
      "step": 910
    },
    {
      "epoch": 1.5462184873949578,
      "grad_norm": 0.6396010518074036,
      "learning_rate": 0.0001745209119775658,
      "loss": 0.4156,
      "step": 920
    },
    {
      "epoch": 1.5630252100840336,
      "grad_norm": 0.5182276368141174,
      "learning_rate": 0.0001737332392605599,
      "loss": 0.4167,
      "step": 930
    },
    {
      "epoch": 1.5798319327731094,
      "grad_norm": 0.4409577548503876,
      "learning_rate": 0.0001729354119783525,
      "loss": 0.4272,
      "step": 940
    },
    {
      "epoch": 1.596638655462185,
      "grad_norm": 0.4636439085006714,
      "learning_rate": 0.00017212754000797107,
      "loss": 0.3888,
      "step": 950
    },
    {
      "epoch": 1.6134453781512605,
      "grad_norm": 0.5916410684585571,
      "learning_rate": 0.00017130973460980077,
      "loss": 0.4319,
      "step": 960
    },
    {
      "epoch": 1.6302521008403361,
      "grad_norm": 0.48982053995132446,
      "learning_rate": 0.00017048210841226155,
      "loss": 0.4168,
      "step": 970
    },
    {
      "epoch": 1.6470588235294117,
      "grad_norm": 0.8442604541778564,
      "learning_rate": 0.00016964477539629702,
      "loss": 0.427,
      "step": 980
    },
    {
      "epoch": 1.6638655462184873,
      "grad_norm": 0.5364897847175598,
      "learning_rate": 0.0001687978508796769,
      "loss": 0.4284,
      "step": 990
    },
    {
      "epoch": 1.680672268907563,
      "grad_norm": 0.5252152681350708,
      "learning_rate": 0.00016794145150111542,
      "loss": 0.4081,
      "step": 1000
    },
    {
      "epoch": 1.680672268907563,
      "eval_loss": 0.469873309135437,
      "eval_runtime": 8.5876,
      "eval_samples_per_second": 116.564,
      "eval_steps_per_second": 14.672,
      "step": 1000
    },
    {
      "epoch": 1.6974789915966386,
      "grad_norm": 0.5419629812240601,
      "learning_rate": 0.00016707569520420785,
      "loss": 0.4571,
      "step": 1010
    },
    {
      "epoch": 1.7142857142857144,
      "grad_norm": 0.447343647480011,
      "learning_rate": 0.00016620070122118724,
      "loss": 0.417,
      "step": 1020
    },
    {
      "epoch": 1.73109243697479,
      "grad_norm": 0.7582902908325195,
      "learning_rate": 0.00016531659005650367,
      "loss": 0.4514,
      "step": 1030
    },
    {
      "epoch": 1.7478991596638656,
      "grad_norm": 0.5247509479522705,
      "learning_rate": 0.00016442348347022842,
      "loss": 0.4377,
      "step": 1040
    },
    {
      "epoch": 1.7647058823529411,
      "grad_norm": 0.525404155254364,
      "learning_rate": 0.0001635215044612851,
      "loss": 0.4159,
      "step": 1050
    },
    {
      "epoch": 1.7815126050420167,
      "grad_norm": 0.5584174990653992,
      "learning_rate": 0.00016261077725051015,
      "loss": 0.4243,
      "step": 1060
    },
    {
      "epoch": 1.7983193277310925,
      "grad_norm": 0.4921722114086151,
      "learning_rate": 0.00016169142726354515,
      "loss": 0.3873,
      "step": 1070
    },
    {
      "epoch": 1.815126050420168,
      "grad_norm": 0.5573268532752991,
      "learning_rate": 0.00016076358111356332,
      "loss": 0.418,
      "step": 1080
    },
    {
      "epoch": 1.8319327731092439,
      "grad_norm": 0.5113047361373901,
      "learning_rate": 0.0001598273665838319,
      "loss": 0.4322,
      "step": 1090
    },
    {
      "epoch": 1.8487394957983194,
      "grad_norm": 0.48914453387260437,
      "learning_rate": 0.0001588829126101143,
      "loss": 0.431,
      "step": 1100
    },
    {
      "epoch": 1.865546218487395,
      "grad_norm": 0.4903518259525299,
      "learning_rate": 0.0001579303492629126,
      "loss": 0.4301,
      "step": 1110
    },
    {
      "epoch": 1.8823529411764706,
      "grad_norm": 0.6053085923194885,
      "learning_rate": 0.0001569698077295545,
      "loss": 0.4376,
      "step": 1120
    },
    {
      "epoch": 1.8991596638655461,
      "grad_norm": 0.6364733576774597,
      "learning_rate": 0.00015600142029612596,
      "loss": 0.4321,
      "step": 1130
    },
    {
      "epoch": 1.9159663865546217,
      "grad_norm": 0.5780381560325623,
      "learning_rate": 0.00015502532032925283,
      "loss": 0.4931,
      "step": 1140
    },
    {
      "epoch": 1.9327731092436975,
      "grad_norm": 0.5068004131317139,
      "learning_rate": 0.0001540416422577335,
      "loss": 0.4143,
      "step": 1150
    },
    {
      "epoch": 1.949579831932773,
      "grad_norm": 0.44138628244400024,
      "learning_rate": 0.00015305052155402545,
      "loss": 0.4577,
      "step": 1160
    },
    {
      "epoch": 1.9663865546218489,
      "grad_norm": 0.542020857334137,
      "learning_rate": 0.0001520520947155877,
      "loss": 0.4184,
      "step": 1170
    },
    {
      "epoch": 1.9831932773109244,
      "grad_norm": 0.38625264167785645,
      "learning_rate": 0.00015104649924608252,
      "loss": 0.3933,
      "step": 1180
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.42412513494491577,
      "learning_rate": 0.00015003387363643836,
      "loss": 0.3885,
      "step": 1190
    },
    {
      "epoch": 2.0168067226890756,
      "grad_norm": 0.4694175720214844,
      "learning_rate": 0.00014901435734577677,
      "loss": 0.3581,
      "step": 1200
    },
    {
      "epoch": 2.033613445378151,
      "grad_norm": 0.5906610488891602,
      "learning_rate": 0.0001479880907822061,
      "loss": 0.3139,
      "step": 1210
    },
    {
      "epoch": 2.0504201680672267,
      "grad_norm": 0.6226663589477539,
      "learning_rate": 0.00014695521528348447,
      "loss": 0.3298,
      "step": 1220
    },
    {
      "epoch": 2.0672268907563027,
      "grad_norm": 0.5035540461540222,
      "learning_rate": 0.00014591587309755464,
      "loss": 0.3193,
      "step": 1230
    },
    {
      "epoch": 2.0840336134453783,
      "grad_norm": 0.5115870237350464,
      "learning_rate": 0.00014487020736295354,
      "loss": 0.3035,
      "step": 1240
    },
    {
      "epoch": 2.100840336134454,
      "grad_norm": 0.5900698304176331,
      "learning_rate": 0.00014381836208909927,
      "loss": 0.3432,
      "step": 1250
    },
    {
      "epoch": 2.1176470588235294,
      "grad_norm": 0.4748307466506958,
      "learning_rate": 0.00014276048213645809,
      "loss": 0.3214,
      "step": 1260
    },
    {
      "epoch": 2.134453781512605,
      "grad_norm": 0.9723687767982483,
      "learning_rate": 0.0001416967131965941,
      "loss": 0.3571,
      "step": 1270
    },
    {
      "epoch": 2.1512605042016806,
      "grad_norm": 0.46582403779029846,
      "learning_rate": 0.00014062720177210455,
      "loss": 0.3357,
      "step": 1280
    },
    {
      "epoch": 2.168067226890756,
      "grad_norm": 0.6172323226928711,
      "learning_rate": 0.00013955209515644363,
      "loss": 0.3249,
      "step": 1290
    },
    {
      "epoch": 2.184873949579832,
      "grad_norm": 0.6465519070625305,
      "learning_rate": 0.00013847154141363691,
      "loss": 0.3485,
      "step": 1300
    },
    {
      "epoch": 2.2016806722689077,
      "grad_norm": 0.560625433921814,
      "learning_rate": 0.00013738568935789015,
      "loss": 0.3311,
      "step": 1310
    },
    {
      "epoch": 2.2184873949579833,
      "grad_norm": 0.6312292218208313,
      "learning_rate": 0.00013629468853309433,
      "loss": 0.3137,
      "step": 1320
    },
    {
      "epoch": 2.235294117647059,
      "grad_norm": 0.7188405990600586,
      "learning_rate": 0.00013519868919223069,
      "loss": 0.3484,
      "step": 1330
    },
    {
      "epoch": 2.2521008403361344,
      "grad_norm": 0.4797142446041107,
      "learning_rate": 0.00013409784227667762,
      "loss": 0.3316,
      "step": 1340
    },
    {
      "epoch": 2.26890756302521,
      "grad_norm": 0.5650676488876343,
      "learning_rate": 0.00013299229939542302,
      "loss": 0.3169,
      "step": 1350
    },
    {
      "epoch": 2.2857142857142856,
      "grad_norm": 0.7440696358680725,
      "learning_rate": 0.00013188221280418472,
      "loss": 0.3573,
      "step": 1360
    },
    {
      "epoch": 2.302521008403361,
      "grad_norm": 0.5263564586639404,
      "learning_rate": 0.0001307677353844417,
      "loss": 0.3157,
      "step": 1370
    },
    {
      "epoch": 2.3193277310924367,
      "grad_norm": 0.8205832839012146,
      "learning_rate": 0.0001296490206223793,
      "loss": 0.3192,
      "step": 1380
    },
    {
      "epoch": 2.3361344537815127,
      "grad_norm": 0.6581016778945923,
      "learning_rate": 0.000128526222587751,
      "loss": 0.3136,
      "step": 1390
    },
    {
      "epoch": 2.3529411764705883,
      "grad_norm": 0.6751677393913269,
      "learning_rate": 0.00012739949591265975,
      "loss": 0.3157,
      "step": 1400
    },
    {
      "epoch": 2.369747899159664,
      "grad_norm": 0.7323719263076782,
      "learning_rate": 0.0001262689957702623,
      "loss": 0.3448,
      "step": 1410
    },
    {
      "epoch": 2.3865546218487395,
      "grad_norm": 0.6688811182975769,
      "learning_rate": 0.00012513487785339832,
      "loss": 0.3492,
      "step": 1420
    },
    {
      "epoch": 2.403361344537815,
      "grad_norm": 0.7764780521392822,
      "learning_rate": 0.0001239972983531487,
      "loss": 0.3318,
      "step": 1430
    },
    {
      "epoch": 2.4201680672268906,
      "grad_norm": 0.7399850487709045,
      "learning_rate": 0.00012285641393732462,
      "loss": 0.3171,
      "step": 1440
    },
    {
      "epoch": 2.4369747899159666,
      "grad_norm": 0.9961610436439514,
      "learning_rate": 0.00012171238172889135,
      "loss": 0.3426,
      "step": 1450
    },
    {
      "epoch": 2.453781512605042,
      "grad_norm": 0.6966540813446045,
      "learning_rate": 0.00012056535928432902,
      "loss": 0.3198,
      "step": 1460
    },
    {
      "epoch": 2.4705882352941178,
      "grad_norm": 0.5827167630195618,
      "learning_rate": 0.00011941550457193418,
      "loss": 0.3535,
      "step": 1470
    },
    {
      "epoch": 2.4873949579831933,
      "grad_norm": 0.6033148765563965,
      "learning_rate": 0.00011826297595006417,
      "loss": 0.3141,
      "step": 1480
    },
    {
      "epoch": 2.504201680672269,
      "grad_norm": 0.7176485061645508,
      "learning_rate": 0.00011710793214532791,
      "loss": 0.3435,
      "step": 1490
    },
    {
      "epoch": 2.5210084033613445,
      "grad_norm": 0.626823365688324,
      "learning_rate": 0.00011595053223072623,
      "loss": 0.3532,
      "step": 1500
    },
    {
      "epoch": 2.53781512605042,
      "grad_norm": 0.6843551993370056,
      "learning_rate": 0.00011479093560374406,
      "loss": 0.327,
      "step": 1510
    },
    {
      "epoch": 2.5546218487394956,
      "grad_norm": 0.7965649962425232,
      "learning_rate": 0.00011362930196439832,
      "loss": 0.3564,
      "step": 1520
    },
    {
      "epoch": 2.571428571428571,
      "grad_norm": 0.5511319041252136,
      "learning_rate": 0.00011246579129324393,
      "loss": 0.3281,
      "step": 1530
    },
    {
      "epoch": 2.588235294117647,
      "grad_norm": 0.6023262143135071,
      "learning_rate": 0.00011130056382934115,
      "loss": 0.3361,
      "step": 1540
    },
    {
      "epoch": 2.6050420168067228,
      "grad_norm": 0.7548812627792358,
      "learning_rate": 0.00011013378004818762,
      "loss": 0.3317,
      "step": 1550
    },
    {
      "epoch": 2.6218487394957983,
      "grad_norm": 0.5712910294532776,
      "learning_rate": 0.00010896560063961732,
      "loss": 0.3243,
      "step": 1560
    },
    {
      "epoch": 2.638655462184874,
      "grad_norm": 0.686442494392395,
      "learning_rate": 0.00010779618648567058,
      "loss": 0.3443,
      "step": 1570
    },
    {
      "epoch": 2.6554621848739495,
      "grad_norm": 0.6805222034454346,
      "learning_rate": 0.00010662569863843725,
      "loss": 0.3314,
      "step": 1580
    },
    {
      "epoch": 2.6722689075630255,
      "grad_norm": 0.776661217212677,
      "learning_rate": 0.00010545429829787659,
      "loss": 0.3219,
      "step": 1590
    },
    {
      "epoch": 2.689075630252101,
      "grad_norm": 0.5905969142913818,
      "learning_rate": 0.00010428214678961668,
      "loss": 0.3445,
      "step": 1600
    },
    {
      "epoch": 2.7058823529411766,
      "grad_norm": 0.5671495199203491,
      "learning_rate": 0.00010310940554273663,
      "loss": 0.3459,
      "step": 1610
    },
    {
      "epoch": 2.722689075630252,
      "grad_norm": 0.7796810269355774,
      "learning_rate": 0.00010193623606753464,
      "loss": 0.3121,
      "step": 1620
    },
    {
      "epoch": 2.7394957983193278,
      "grad_norm": 0.6266621947288513,
      "learning_rate": 0.00010076279993328455,
      "loss": 0.34,
      "step": 1630
    },
    {
      "epoch": 2.7563025210084033,
      "grad_norm": 0.6675502061843872,
      "learning_rate": 9.958925874598464e-05,
      "loss": 0.298,
      "step": 1640
    },
    {
      "epoch": 2.773109243697479,
      "grad_norm": 0.6532730460166931,
      "learning_rate": 9.841577412610108e-05,
      "loss": 0.3519,
      "step": 1650
    },
    {
      "epoch": 2.7899159663865545,
      "grad_norm": 0.8185803890228271,
      "learning_rate": 9.724250768630959e-05,
      "loss": 0.3373,
      "step": 1660
    },
    {
      "epoch": 2.80672268907563,
      "grad_norm": 0.6204152703285217,
      "learning_rate": 9.606962100923807e-05,
      "loss": 0.3061,
      "step": 1670
    },
    {
      "epoch": 2.8235294117647056,
      "grad_norm": 0.5997849702835083,
      "learning_rate": 9.489727562521332e-05,
      "loss": 0.3478,
      "step": 1680
    },
    {
      "epoch": 2.8403361344537816,
      "grad_norm": 0.6466342806816101,
      "learning_rate": 9.372563299001514e-05,
      "loss": 0.3242,
      "step": 1690
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.8826155662536621,
      "learning_rate": 9.255485446264046e-05,
      "loss": 0.3093,
      "step": 1700
    },
    {
      "epoch": 2.8739495798319328,
      "grad_norm": 0.9787838459014893,
      "learning_rate": 9.138510128308117e-05,
      "loss": 0.3608,
      "step": 1710
    },
    {
      "epoch": 2.8907563025210083,
      "grad_norm": 0.7216944098472595,
      "learning_rate": 9.021653455011778e-05,
      "loss": 0.3247,
      "step": 1720
    },
    {
      "epoch": 2.907563025210084,
      "grad_norm": 0.7170887589454651,
      "learning_rate": 8.904931519913304e-05,
      "loss": 0.354,
      "step": 1730
    },
    {
      "epoch": 2.92436974789916,
      "grad_norm": 0.6556657552719116,
      "learning_rate": 8.788360397994785e-05,
      "loss": 0.3406,
      "step": 1740
    },
    {
      "epoch": 2.9411764705882355,
      "grad_norm": 0.7959017157554626,
      "learning_rate": 8.671956143468267e-05,
      "loss": 0.3388,
      "step": 1750
    },
    {
      "epoch": 2.957983193277311,
      "grad_norm": 0.7491987347602844,
      "learning_rate": 8.55573478756477e-05,
      "loss": 0.3676,
      "step": 1760
    },
    {
      "epoch": 2.9747899159663866,
      "grad_norm": 0.7010266780853271,
      "learning_rate": 8.439712336326452e-05,
      "loss": 0.3152,
      "step": 1770
    },
    {
      "epoch": 2.991596638655462,
      "grad_norm": 0.7219695448875427,
      "learning_rate": 8.32390476840227e-05,
      "loss": 0.3575,
      "step": 1780
    },
    {
      "epoch": 3.008403361344538,
      "grad_norm": 0.8448328971862793,
      "learning_rate": 8.208328032847366e-05,
      "loss": 0.2936,
      "step": 1790
    },
    {
      "epoch": 3.0252100840336134,
      "grad_norm": 0.575706958770752,
      "learning_rate": 8.092998046926583e-05,
      "loss": 0.2582,
      "step": 1800
    },
    {
      "epoch": 3.042016806722689,
      "grad_norm": 0.8433834910392761,
      "learning_rate": 7.977930693922329e-05,
      "loss": 0.2324,
      "step": 1810
    },
    {
      "epoch": 3.0588235294117645,
      "grad_norm": 0.8530866503715515,
      "learning_rate": 7.863141820947114e-05,
      "loss": 0.2219,
      "step": 1820
    },
    {
      "epoch": 3.0756302521008405,
      "grad_norm": 0.7315991520881653,
      "learning_rate": 7.748647236761094e-05,
      "loss": 0.2201,
      "step": 1830
    },
    {
      "epoch": 3.092436974789916,
      "grad_norm": 0.7668396234512329,
      "learning_rate": 7.63446270959488e-05,
      "loss": 0.2664,
      "step": 1840
    },
    {
      "epoch": 3.1092436974789917,
      "grad_norm": 0.7540835738182068,
      "learning_rate": 7.520603964977939e-05,
      "loss": 0.2426,
      "step": 1850
    },
    {
      "epoch": 3.1260504201680672,
      "grad_norm": 0.756682276725769,
      "learning_rate": 7.407086683572846e-05,
      "loss": 0.2297,
      "step": 1860
    },
    {
      "epoch": 3.142857142857143,
      "grad_norm": 0.7847204804420471,
      "learning_rate": 7.293926499015763e-05,
      "loss": 0.257,
      "step": 1870
    },
    {
      "epoch": 3.1596638655462184,
      "grad_norm": 0.909298837184906,
      "learning_rate": 7.181138995763362e-05,
      "loss": 0.2374,
      "step": 1880
    },
    {
      "epoch": 3.176470588235294,
      "grad_norm": 1.1145635843276978,
      "learning_rate": 7.068739706946517e-05,
      "loss": 0.2389,
      "step": 1890
    },
    {
      "epoch": 3.19327731092437,
      "grad_norm": 0.7492973804473877,
      "learning_rate": 6.956744112231095e-05,
      "loss": 0.2466,
      "step": 1900
    },
    {
      "epoch": 3.2100840336134455,
      "grad_norm": 0.7049612998962402,
      "learning_rate": 6.845167635686083e-05,
      "loss": 0.2369,
      "step": 1910
    },
    {
      "epoch": 3.226890756302521,
      "grad_norm": 0.6115100979804993,
      "learning_rate": 6.73402564365938e-05,
      "loss": 0.216,
      "step": 1920
    },
    {
      "epoch": 3.2436974789915967,
      "grad_norm": 0.7355627417564392,
      "learning_rate": 6.623333442661532e-05,
      "loss": 0.2449,
      "step": 1930
    },
    {
      "epoch": 3.2605042016806722,
      "grad_norm": 0.8105019927024841,
      "learning_rate": 6.513106277257728e-05,
      "loss": 0.2361,
      "step": 1940
    },
    {
      "epoch": 3.277310924369748,
      "grad_norm": 0.6543210744857788,
      "learning_rate": 6.403359327968303e-05,
      "loss": 0.2136,
      "step": 1950
    },
    {
      "epoch": 3.2941176470588234,
      "grad_norm": 1.191417932510376,
      "learning_rate": 6.294107709178079e-05,
      "loss": 0.2457,
      "step": 1960
    },
    {
      "epoch": 3.310924369747899,
      "grad_norm": 0.8825222849845886,
      "learning_rate": 6.185366467054799e-05,
      "loss": 0.2184,
      "step": 1970
    },
    {
      "epoch": 3.327731092436975,
      "grad_norm": 0.8879269361495972,
      "learning_rate": 6.0771505774769734e-05,
      "loss": 0.2419,
      "step": 1980
    },
    {
      "epoch": 3.3445378151260505,
      "grad_norm": 0.692824125289917,
      "learning_rate": 5.9694749439713936e-05,
      "loss": 0.2559,
      "step": 1990
    },
    {
      "epoch": 3.361344537815126,
      "grad_norm": 0.7757511138916016,
      "learning_rate": 5.8623543956606074e-05,
      "loss": 0.2383,
      "step": 2000
    },
    {
      "epoch": 3.361344537815126,
      "eval_loss": 0.5869996547698975,
      "eval_runtime": 8.7439,
      "eval_samples_per_second": 114.479,
      "eval_steps_per_second": 14.41,
      "step": 2000
    },
    {
      "epoch": 3.3781512605042017,
      "grad_norm": 0.8735210299491882,
      "learning_rate": 5.7558036852206596e-05,
      "loss": 0.2554,
      "step": 2010
    },
    {
      "epoch": 3.3949579831932772,
      "grad_norm": 0.9418761134147644,
      "learning_rate": 5.6498374868493296e-05,
      "loss": 0.2682,
      "step": 2020
    },
    {
      "epoch": 3.411764705882353,
      "grad_norm": 0.7112303972244263,
      "learning_rate": 5.544470394245225e-05,
      "loss": 0.2347,
      "step": 2030
    },
    {
      "epoch": 3.4285714285714284,
      "grad_norm": 1.082740068435669,
      "learning_rate": 5.439716918597906e-05,
      "loss": 0.2561,
      "step": 2040
    },
    {
      "epoch": 3.4453781512605044,
      "grad_norm": 0.7755303382873535,
      "learning_rate": 5.335591486589406e-05,
      "loss": 0.2087,
      "step": 2050
    },
    {
      "epoch": 3.46218487394958,
      "grad_norm": 0.8446845412254333,
      "learning_rate": 5.232108438407399e-05,
      "loss": 0.2323,
      "step": 2060
    },
    {
      "epoch": 3.4789915966386555,
      "grad_norm": 0.8153941035270691,
      "learning_rate": 5.129282025770252e-05,
      "loss": 0.2275,
      "step": 2070
    },
    {
      "epoch": 3.495798319327731,
      "grad_norm": 0.7065542340278625,
      "learning_rate": 5.027126409964271e-05,
      "loss": 0.2136,
      "step": 2080
    },
    {
      "epoch": 3.5126050420168067,
      "grad_norm": 1.1057276725769043,
      "learning_rate": 4.9256556598934054e-05,
      "loss": 0.2381,
      "step": 2090
    },
    {
      "epoch": 3.5294117647058822,
      "grad_norm": 0.6242520809173584,
      "learning_rate": 4.82488375014169e-05,
      "loss": 0.2394,
      "step": 2100
    },
    {
      "epoch": 3.546218487394958,
      "grad_norm": 0.6137174367904663,
      "learning_rate": 4.7248245590486375e-05,
      "loss": 0.2411,
      "step": 2110
    },
    {
      "epoch": 3.5630252100840334,
      "grad_norm": 0.8595492839813232,
      "learning_rate": 4.625491866797917e-05,
      "loss": 0.2383,
      "step": 2120
    },
    {
      "epoch": 3.5798319327731094,
      "grad_norm": 0.7664365768432617,
      "learning_rate": 4.526899353519557e-05,
      "loss": 0.264,
      "step": 2130
    },
    {
      "epoch": 3.596638655462185,
      "grad_norm": 1.2114841938018799,
      "learning_rate": 4.4290605974058995e-05,
      "loss": 0.2412,
      "step": 2140
    },
    {
      "epoch": 3.6134453781512605,
      "grad_norm": 0.8489940166473389,
      "learning_rate": 4.331989072841604e-05,
      "loss": 0.2408,
      "step": 2150
    },
    {
      "epoch": 3.630252100840336,
      "grad_norm": 0.7826299071311951,
      "learning_rate": 4.235698148547955e-05,
      "loss": 0.2347,
      "step": 2160
    },
    {
      "epoch": 3.6470588235294117,
      "grad_norm": 0.9073792099952698,
      "learning_rate": 4.140201085741731e-05,
      "loss": 0.2454,
      "step": 2170
    },
    {
      "epoch": 3.6638655462184873,
      "grad_norm": 0.8302606344223022,
      "learning_rate": 4.0455110363088424e-05,
      "loss": 0.2381,
      "step": 2180
    },
    {
      "epoch": 3.6806722689075633,
      "grad_norm": 0.7296493649482727,
      "learning_rate": 3.9516410409930594e-05,
      "loss": 0.2432,
      "step": 2190
    },
    {
      "epoch": 3.697478991596639,
      "grad_norm": 0.8790242075920105,
      "learning_rate": 3.858604027600049e-05,
      "loss": 0.2387,
      "step": 2200
    },
    {
      "epoch": 3.7142857142857144,
      "grad_norm": 0.8583736419677734,
      "learning_rate": 3.766412809216937e-05,
      "loss": 0.2482,
      "step": 2210
    },
    {
      "epoch": 3.73109243697479,
      "grad_norm": 0.8858778476715088,
      "learning_rate": 3.675080082447688e-05,
      "loss": 0.2524,
      "step": 2220
    },
    {
      "epoch": 3.7478991596638656,
      "grad_norm": 0.896706759929657,
      "learning_rate": 3.584618425664526e-05,
      "loss": 0.2477,
      "step": 2230
    },
    {
      "epoch": 3.764705882352941,
      "grad_norm": 0.7617205381393433,
      "learning_rate": 3.495040297275651e-05,
      "loss": 0.2194,
      "step": 2240
    },
    {
      "epoch": 3.7815126050420167,
      "grad_norm": 0.7257958650588989,
      "learning_rate": 3.40635803400944e-05,
      "loss": 0.2414,
      "step": 2250
    },
    {
      "epoch": 3.7983193277310923,
      "grad_norm": 0.83749920129776,
      "learning_rate": 3.3185838492154386e-05,
      "loss": 0.2218,
      "step": 2260
    },
    {
      "epoch": 3.815126050420168,
      "grad_norm": 1.028795599937439,
      "learning_rate": 3.231729831182344e-05,
      "loss": 0.2312,
      "step": 2270
    },
    {
      "epoch": 3.831932773109244,
      "grad_norm": 0.8044925332069397,
      "learning_rate": 3.1458079414731955e-05,
      "loss": 0.2317,
      "step": 2280
    },
    {
      "epoch": 3.8487394957983194,
      "grad_norm": 0.7760542035102844,
      "learning_rate": 3.060830013278012e-05,
      "loss": 0.2222,
      "step": 2290
    },
    {
      "epoch": 3.865546218487395,
      "grad_norm": 0.7686411738395691,
      "learning_rate": 2.9768077497841352e-05,
      "loss": 0.2313,
      "step": 2300
    },
    {
      "epoch": 3.8823529411764706,
      "grad_norm": 0.700995147228241,
      "learning_rate": 2.8937527225644677e-05,
      "loss": 0.231,
      "step": 2310
    },
    {
      "epoch": 3.899159663865546,
      "grad_norm": 0.9482907652854919,
      "learning_rate": 2.8116763699838167e-05,
      "loss": 0.2504,
      "step": 2320
    },
    {
      "epoch": 3.9159663865546217,
      "grad_norm": 0.9015063047409058,
      "learning_rate": 2.7305899956236046e-05,
      "loss": 0.24,
      "step": 2330
    },
    {
      "epoch": 3.9327731092436977,
      "grad_norm": 0.9054192900657654,
      "learning_rate": 2.6505047667251437e-05,
      "loss": 0.2478,
      "step": 2340
    },
    {
      "epoch": 3.9495798319327733,
      "grad_norm": 0.9170212745666504,
      "learning_rate": 2.5714317126516763e-05,
      "loss": 0.2419,
      "step": 2350
    },
    {
      "epoch": 3.966386554621849,
      "grad_norm": 1.029768466949463,
      "learning_rate": 2.493381723369398e-05,
      "loss": 0.2341,
      "step": 2360
    },
    {
      "epoch": 3.9831932773109244,
      "grad_norm": 0.6440130472183228,
      "learning_rate": 2.4163655479476943e-05,
      "loss": 0.213,
      "step": 2370
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.8131651282310486,
      "learning_rate": 2.340393793078789e-05,
      "loss": 0.2302,
      "step": 2380
    },
    {
      "epoch": 4.016806722689076,
      "grad_norm": 0.7478677034378052,
      "learning_rate": 2.2654769216169602e-05,
      "loss": 0.1944,
      "step": 2390
    },
    {
      "epoch": 4.033613445378151,
      "grad_norm": 0.7543561458587646,
      "learning_rate": 2.1916252511376133e-05,
      "loss": 0.172,
      "step": 2400
    },
    {
      "epoch": 4.050420168067227,
      "grad_norm": 0.7355443239212036,
      "learning_rate": 2.1188489525163367e-05,
      "loss": 0.166,
      "step": 2410
    },
    {
      "epoch": 4.067226890756302,
      "grad_norm": 0.7719018459320068,
      "learning_rate": 2.047158048528165e-05,
      "loss": 0.1421,
      "step": 2420
    },
    {
      "epoch": 4.084033613445378,
      "grad_norm": 0.7346205115318298,
      "learning_rate": 1.976562412467232e-05,
      "loss": 0.174,
      "step": 2430
    },
    {
      "epoch": 4.100840336134453,
      "grad_norm": 0.960121214389801,
      "learning_rate": 1.9070717667870263e-05,
      "loss": 0.1725,
      "step": 2440
    },
    {
      "epoch": 4.117647058823529,
      "grad_norm": 0.8343680500984192,
      "learning_rate": 1.838695681761412e-05,
      "loss": 0.1777,
      "step": 2450
    },
    {
      "epoch": 4.1344537815126055,
      "grad_norm": 0.9134616851806641,
      "learning_rate": 1.7714435741666037e-05,
      "loss": 0.1541,
      "step": 2460
    },
    {
      "epoch": 4.151260504201681,
      "grad_norm": 0.9840583801269531,
      "learning_rate": 1.7053247059842793e-05,
      "loss": 0.1905,
      "step": 2470
    },
    {
      "epoch": 4.168067226890757,
      "grad_norm": 0.7404395341873169,
      "learning_rate": 1.6403481831260338e-05,
      "loss": 0.1803,
      "step": 2480
    },
    {
      "epoch": 4.184873949579832,
      "grad_norm": 1.0390299558639526,
      "learning_rate": 1.5765229541792982e-05,
      "loss": 0.1691,
      "step": 2490
    },
    {
      "epoch": 4.201680672268908,
      "grad_norm": 0.8568964600563049,
      "learning_rate": 1.5138578091749334e-05,
      "loss": 0.1706,
      "step": 2500
    },
    {
      "epoch": 4.218487394957983,
      "grad_norm": 0.9740147590637207,
      "learning_rate": 1.4523613783766732e-05,
      "loss": 0.1795,
      "step": 2510
    },
    {
      "epoch": 4.235294117647059,
      "grad_norm": 0.9405394792556763,
      "learning_rate": 1.3920421310925636e-05,
      "loss": 0.1614,
      "step": 2520
    },
    {
      "epoch": 4.2521008403361344,
      "grad_norm": 0.8562027215957642,
      "learning_rate": 1.3329083745085547e-05,
      "loss": 0.1621,
      "step": 2530
    },
    {
      "epoch": 4.26890756302521,
      "grad_norm": 0.8962126970291138,
      "learning_rate": 1.2749682525444407e-05,
      "loss": 0.1642,
      "step": 2540
    },
    {
      "epoch": 4.285714285714286,
      "grad_norm": 0.9526483416557312,
      "learning_rate": 1.2182297447322787e-05,
      "loss": 0.1581,
      "step": 2550
    },
    {
      "epoch": 4.302521008403361,
      "grad_norm": 0.7188050746917725,
      "learning_rate": 1.1627006651174444e-05,
      "loss": 0.1861,
      "step": 2560
    },
    {
      "epoch": 4.319327731092437,
      "grad_norm": 0.8526328206062317,
      "learning_rate": 1.10838866118247e-05,
      "loss": 0.1589,
      "step": 2570
    },
    {
      "epoch": 4.336134453781512,
      "grad_norm": 0.845211923122406,
      "learning_rate": 1.0553012127938389e-05,
      "loss": 0.1958,
      "step": 2580
    },
    {
      "epoch": 4.352941176470588,
      "grad_norm": 0.9076491594314575,
      "learning_rate": 1.0034456311718609e-05,
      "loss": 0.1664,
      "step": 2590
    },
    {
      "epoch": 4.369747899159664,
      "grad_norm": 1.0046247243881226,
      "learning_rate": 9.528290578837551e-06,
      "loss": 0.1713,
      "step": 2600
    },
    {
      "epoch": 4.38655462184874,
      "grad_norm": 0.7860477566719055,
      "learning_rate": 9.034584638601196e-06,
      "loss": 0.1494,
      "step": 2610
    },
    {
      "epoch": 4.4033613445378155,
      "grad_norm": 1.0020995140075684,
      "learning_rate": 8.553406484348914e-06,
      "loss": 0.1823,
      "step": 2620
    },
    {
      "epoch": 4.420168067226891,
      "grad_norm": 1.2964270114898682,
      "learning_rate": 8.084822384089453e-06,
      "loss": 0.1825,
      "step": 2630
    },
    {
      "epoch": 4.436974789915967,
      "grad_norm": 1.0893545150756836,
      "learning_rate": 7.628896871374358e-06,
      "loss": 0.1799,
      "step": 2640
    },
    {
      "epoch": 4.453781512605042,
      "grad_norm": 0.8512413501739502,
      "learning_rate": 7.185692736410477e-06,
      "loss": 0.1707,
      "step": 2650
    },
    {
      "epoch": 4.470588235294118,
      "grad_norm": 1.3223856687545776,
      "learning_rate": 6.755271017412501e-06,
      "loss": 0.1748,
      "step": 2660
    },
    {
      "epoch": 4.487394957983193,
      "grad_norm": 1.0431852340698242,
      "learning_rate": 6.337690992196688e-06,
      "loss": 0.1703,
      "step": 2670
    },
    {
      "epoch": 4.504201680672269,
      "grad_norm": 1.3509666919708252,
      "learning_rate": 5.933010170017139e-06,
      "loss": 0.1717,
      "step": 2680
    },
    {
      "epoch": 4.5210084033613445,
      "grad_norm": 0.9030164480209351,
      "learning_rate": 5.54128428364562e-06,
      "loss": 0.1866,
      "step": 2690
    },
    {
      "epoch": 4.53781512605042,
      "grad_norm": 0.9294901490211487,
      "learning_rate": 5.162567281696018e-06,
      "loss": 0.1659,
      "step": 2700
    },
    {
      "epoch": 4.554621848739496,
      "grad_norm": 0.887606680393219,
      "learning_rate": 4.796911321194475e-06,
      "loss": 0.166,
      "step": 2710
    },
    {
      "epoch": 4.571428571428571,
      "grad_norm": 0.8213235139846802,
      "learning_rate": 4.444366760396357e-06,
      "loss": 0.1768,
      "step": 2720
    },
    {
      "epoch": 4.588235294117647,
      "grad_norm": 0.9339560270309448,
      "learning_rate": 4.104982151850911e-06,
      "loss": 0.1769,
      "step": 2730
    },
    {
      "epoch": 4.605042016806722,
      "grad_norm": 0.8878592252731323,
      "learning_rate": 3.7788042357145616e-06,
      "loss": 0.1708,
      "step": 2740
    },
    {
      "epoch": 4.621848739495798,
      "grad_norm": 0.8466598987579346,
      "learning_rate": 3.465877933313799e-06,
      "loss": 0.176,
      "step": 2750
    },
    {
      "epoch": 4.6386554621848735,
      "grad_norm": 0.8101462721824646,
      "learning_rate": 3.166246340958712e-06,
      "loss": 0.1722,
      "step": 2760
    },
    {
      "epoch": 4.65546218487395,
      "grad_norm": 1.1340394020080566,
      "learning_rate": 2.8799507240076406e-06,
      "loss": 0.1851,
      "step": 2770
    },
    {
      "epoch": 4.6722689075630255,
      "grad_norm": 1.2405331134796143,
      "learning_rate": 2.607030511184183e-06,
      "loss": 0.1671,
      "step": 2780
    },
    {
      "epoch": 4.689075630252101,
      "grad_norm": 0.8664318323135376,
      "learning_rate": 2.347523289146991e-06,
      "loss": 0.1728,
      "step": 2790
    },
    {
      "epoch": 4.705882352941177,
      "grad_norm": 0.9817457795143127,
      "learning_rate": 2.101464797313357e-06,
      "loss": 0.1549,
      "step": 2800
    },
    {
      "epoch": 4.722689075630252,
      "grad_norm": 0.8827335834503174,
      "learning_rate": 1.868888922937162e-06,
      "loss": 0.1423,
      "step": 2810
    },
    {
      "epoch": 4.739495798319328,
      "grad_norm": 0.7880733609199524,
      "learning_rate": 1.649827696441919e-06,
      "loss": 0.1571,
      "step": 2820
    },
    {
      "epoch": 4.756302521008403,
      "grad_norm": 1.198318600654602,
      "learning_rate": 1.4443112870095588e-06,
      "loss": 0.1624,
      "step": 2830
    },
    {
      "epoch": 4.773109243697479,
      "grad_norm": 0.7685941457748413,
      "learning_rate": 1.25236799842543e-06,
      "loss": 0.1609,
      "step": 2840
    },
    {
      "epoch": 4.7899159663865545,
      "grad_norm": 0.929218053817749,
      "learning_rate": 1.0740242651803956e-06,
      "loss": 0.1664,
      "step": 2850
    },
    {
      "epoch": 4.80672268907563,
      "grad_norm": 0.9220144152641296,
      "learning_rate": 9.093046488302337e-07,
      "loss": 0.1793,
      "step": 2860
    },
    {
      "epoch": 4.823529411764706,
      "grad_norm": 0.8192142844200134,
      "learning_rate": 7.582318346129657e-07,
      "loss": 0.1641,
      "step": 2870
    },
    {
      "epoch": 4.840336134453781,
      "grad_norm": 0.8484081029891968,
      "learning_rate": 6.20826628324711e-07,
      "loss": 0.1508,
      "step": 2880
    },
    {
      "epoch": 4.857142857142857,
      "grad_norm": 1.0260757207870483,
      "learning_rate": 4.971079534542345e-07,
      "loss": 0.1628,
      "step": 2890
    },
    {
      "epoch": 4.873949579831933,
      "grad_norm": 0.9038840532302856,
      "learning_rate": 3.8709284857687546e-07,
      "loss": 0.1684,
      "step": 2900
    },
    {
      "epoch": 4.890756302521009,
      "grad_norm": 0.7963437438011169,
      "learning_rate": 2.9079646500792534e-07,
      "loss": 0.1619,
      "step": 2910
    },
    {
      "epoch": 4.907563025210084,
      "grad_norm": 0.8787494897842407,
      "learning_rate": 2.0823206471599632e-07,
      "loss": 0.1612,
      "step": 2920
    },
    {
      "epoch": 4.92436974789916,
      "grad_norm": 1.0699881315231323,
      "learning_rate": 1.3941101849660466e-07,
      "loss": 0.1647,
      "step": 2930
    },
    {
      "epoch": 4.9411764705882355,
      "grad_norm": 0.6501001715660095,
      "learning_rate": 8.434280440616782e-08,
      "loss": 0.1611,
      "step": 2940
    },
    {
      "epoch": 4.957983193277311,
      "grad_norm": 0.9485645890235901,
      "learning_rate": 4.3035006456670915e-08,
      "loss": 0.1694,
      "step": 2950
    },
    {
      "epoch": 4.974789915966387,
      "grad_norm": 0.8494912385940552,
      "learning_rate": 1.5493313571202093e-08,
      "loss": 0.1695,
      "step": 2960
    },
    {
      "epoch": 4.991596638655462,
      "grad_norm": 0.9211341738700867,
      "learning_rate": 1.7215188004904293e-09,
      "loss": 0.155,
      "step": 2970
    },
    {
      "epoch": 5.0,
      "step": 2975,
      "total_flos": 2.1667263896407245e+17,
      "train_loss": 0.34065224741687294,
      "train_runtime": 1015.5372,
      "train_samples_per_second": 93.62,
      "train_steps_per_second": 2.929
    }
  ],
  "logging_steps": 10,
  "max_steps": 2975,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.1667263896407245e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}