{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.996423573316747,
  "eval_steps": 500,
  "global_step": 2409,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.012439744985227803,
      "grad_norm": 1.897183168042883,
      "learning_rate": 7.468879668049793e-07,
      "loss": 0.637,
      "step": 10
    },
    {
      "epoch": 0.024879489970455606,
      "grad_norm": 1.3113204984682525,
      "learning_rate": 1.5767634854771784e-06,
      "loss": 0.5821,
      "step": 20
    },
    {
      "epoch": 0.03731923495568341,
      "grad_norm": 1.0571866692485175,
      "learning_rate": 2.4066390041493776e-06,
      "loss": 0.425,
      "step": 30
    },
    {
      "epoch": 0.04975897994091121,
      "grad_norm": 1.0021561734833917,
      "learning_rate": 3.2365145228215773e-06,
      "loss": 0.2552,
      "step": 40
    },
    {
      "epoch": 0.06219872492613902,
      "grad_norm": 0.3660702758245685,
      "learning_rate": 4.0663900414937765e-06,
      "loss": 0.1621,
      "step": 50
    },
    {
      "epoch": 0.07463846991136681,
      "grad_norm": 1.0078512374174537,
      "learning_rate": 4.896265560165976e-06,
      "loss": 0.1242,
      "step": 60
    },
    {
      "epoch": 0.08707821489659462,
      "grad_norm": 0.22907555295988605,
      "learning_rate": 5.726141078838174e-06,
      "loss": 0.0994,
      "step": 70
    },
    {
      "epoch": 0.09951795988182242,
      "grad_norm": 0.2073329739535906,
      "learning_rate": 6.556016597510374e-06,
      "loss": 0.0867,
      "step": 80
    },
    {
      "epoch": 0.11195770486705023,
      "grad_norm": 0.16148748791334328,
      "learning_rate": 7.385892116182573e-06,
      "loss": 0.0783,
      "step": 90
    },
    {
      "epoch": 0.12439744985227803,
      "grad_norm": 0.3401530106760488,
      "learning_rate": 8.215767634854772e-06,
      "loss": 0.0659,
      "step": 100
    },
    {
      "epoch": 0.13683719483750584,
      "grad_norm": 0.43710710342722814,
      "learning_rate": 9.045643153526971e-06,
      "loss": 0.0622,
      "step": 110
    },
    {
      "epoch": 0.14927693982273363,
      "grad_norm": 0.20444416169246796,
      "learning_rate": 9.875518672199172e-06,
      "loss": 0.0594,
      "step": 120
    },
    {
      "epoch": 0.16171668480796145,
      "grad_norm": 0.2708103656836858,
      "learning_rate": 1.070539419087137e-05,
      "loss": 0.0586,
      "step": 130
    },
    {
      "epoch": 0.17415642979318924,
      "grad_norm": 0.36138807011425644,
      "learning_rate": 1.1535269709543569e-05,
      "loss": 0.0484,
      "step": 140
    },
    {
      "epoch": 0.18659617477841703,
      "grad_norm": 0.3463799471224063,
      "learning_rate": 1.236514522821577e-05,
      "loss": 0.0509,
      "step": 150
    },
    {
      "epoch": 0.19903591976364485,
      "grad_norm": 0.2736460915009888,
      "learning_rate": 1.3195020746887967e-05,
      "loss": 0.0481,
      "step": 160
    },
    {
      "epoch": 0.21147566474887264,
      "grad_norm": 0.26486302971834347,
      "learning_rate": 1.4024896265560166e-05,
      "loss": 0.0488,
      "step": 170
    },
    {
      "epoch": 0.22391540973410046,
      "grad_norm": 0.23132916477830964,
      "learning_rate": 1.4854771784232367e-05,
      "loss": 0.0413,
      "step": 180
    },
    {
      "epoch": 0.23635515471932825,
      "grad_norm": 0.3293544356167722,
      "learning_rate": 1.5684647302904566e-05,
      "loss": 0.0422,
      "step": 190
    },
    {
      "epoch": 0.24879489970455607,
      "grad_norm": 0.211688759676043,
      "learning_rate": 1.6514522821576764e-05,
      "loss": 0.0397,
      "step": 200
    },
    {
      "epoch": 0.26123464468978386,
      "grad_norm": 0.18497421951025278,
      "learning_rate": 1.7344398340248965e-05,
      "loss": 0.0425,
      "step": 210
    },
    {
      "epoch": 0.2736743896750117,
      "grad_norm": 0.1670272600636411,
      "learning_rate": 1.8174273858921162e-05,
      "loss": 0.0375,
      "step": 220
    },
    {
      "epoch": 0.28611413466023944,
      "grad_norm": 0.19117537641121957,
      "learning_rate": 1.9004149377593363e-05,
      "loss": 0.0388,
      "step": 230
    },
    {
      "epoch": 0.29855387964546726,
      "grad_norm": 0.16271436719429785,
      "learning_rate": 1.983402489626556e-05,
      "loss": 0.0372,
      "step": 240
    },
    {
      "epoch": 0.3109936246306951,
      "grad_norm": 0.2669250204984381,
      "learning_rate": 1.9999328066483867e-05,
      "loss": 0.0341,
      "step": 250
    },
    {
      "epoch": 0.3234333696159229,
      "grad_norm": 0.17368593835356172,
      "learning_rate": 1.99965984913342e-05,
      "loss": 0.0367,
      "step": 260
    },
    {
      "epoch": 0.33587311460115066,
      "grad_norm": 0.4319208125818109,
      "learning_rate": 1.999176985141578e-05,
      "loss": 0.0337,
      "step": 270
    },
    {
      "epoch": 0.3483128595863785,
      "grad_norm": 0.22229799036098782,
      "learning_rate": 1.998484316063629e-05,
      "loss": 0.0308,
      "step": 280
    },
    {
      "epoch": 0.3607526045716063,
      "grad_norm": 0.18260039982968104,
      "learning_rate": 1.997581987344772e-05,
      "loss": 0.0328,
      "step": 290
    },
    {
      "epoch": 0.37319234955683406,
      "grad_norm": 0.2312512020216777,
      "learning_rate": 1.9964701884540964e-05,
      "loss": 0.0301,
      "step": 300
    },
    {
      "epoch": 0.3856320945420619,
      "grad_norm": 0.18105099802902547,
      "learning_rate": 1.9951491528448005e-05,
      "loss": 0.0314,
      "step": 310
    },
    {
      "epoch": 0.3980718395272897,
      "grad_norm": 0.22913187442874947,
      "learning_rate": 1.993619157905169e-05,
      "loss": 0.0293,
      "step": 320
    },
    {
      "epoch": 0.4105115845125175,
      "grad_norm": 0.31994143894737154,
      "learning_rate": 1.9918805249003272e-05,
      "loss": 0.0312,
      "step": 330
    },
    {
      "epoch": 0.4229513294977453,
      "grad_norm": 0.6585694791305013,
      "learning_rate": 1.9899336189047846e-05,
      "loss": 0.0287,
      "step": 340
    },
    {
      "epoch": 0.4353910744829731,
      "grad_norm": 0.18850124632007803,
      "learning_rate": 1.987778848725775e-05,
      "loss": 0.0269,
      "step": 350
    },
    {
      "epoch": 0.4478308194682009,
      "grad_norm": 0.24710399710319328,
      "learning_rate": 1.985416666817419e-05,
      "loss": 0.03,
      "step": 360
    },
    {
      "epoch": 0.46027056445342873,
      "grad_norm": 0.12842349092107871,
      "learning_rate": 1.9828475691857148e-05,
      "loss": 0.0271,
      "step": 370
    },
    {
      "epoch": 0.4727103094386565,
      "grad_norm": 0.14250982750272018,
      "learning_rate": 1.9800720952843906e-05,
      "loss": 0.0263,
      "step": 380
    },
    {
      "epoch": 0.4851500544238843,
      "grad_norm": 0.3282136757807734,
      "learning_rate": 1.977090827901631e-05,
      "loss": 0.0263,
      "step": 390
    },
    {
      "epoch": 0.49758979940911213,
      "grad_norm": 0.1560413238804571,
      "learning_rate": 1.973904393037703e-05,
      "loss": 0.0265,
      "step": 400
    },
    {
      "epoch": 0.5100295443943399,
      "grad_norm": 0.15169578616837004,
      "learning_rate": 1.9705134597735113e-05,
      "loss": 0.0274,
      "step": 410
    },
    {
      "epoch": 0.5224692893795677,
      "grad_norm": 0.1339115650032496,
      "learning_rate": 1.966918740130106e-05,
      "loss": 0.0285,
      "step": 420
    },
    {
      "epoch": 0.5349090343647955,
      "grad_norm": 0.17110253852140866,
      "learning_rate": 1.9631209889191712e-05,
      "loss": 0.0256,
      "step": 430
    },
    {
      "epoch": 0.5473487793500234,
      "grad_norm": 0.5948554985143815,
      "learning_rate": 1.959121003584536e-05,
      "loss": 0.025,
      "step": 440
    },
    {
      "epoch": 0.5597885243352512,
      "grad_norm": 0.2206397525799233,
      "learning_rate": 1.954919624034725e-05,
      "loss": 0.0258,
      "step": 450
    },
    {
      "epoch": 0.5722282693204789,
      "grad_norm": 0.13697762099163308,
      "learning_rate": 1.950517732466597e-05,
      "loss": 0.0226,
      "step": 460
    },
    {
      "epoch": 0.5846680143057067,
      "grad_norm": 0.10557073555012622,
      "learning_rate": 1.9459162531801048e-05,
      "loss": 0.0243,
      "step": 470
    },
    {
      "epoch": 0.5971077592909345,
      "grad_norm": 0.10166391741425912,
      "learning_rate": 1.94111615238421e-05,
      "loss": 0.0232,
      "step": 480
    },
    {
      "epoch": 0.6095475042761623,
      "grad_norm": 0.14115013264667553,
      "learning_rate": 1.936118437994003e-05,
      "loss": 0.025,
      "step": 490
    },
    {
      "epoch": 0.6219872492613902,
      "grad_norm": 0.28777045307662136,
      "learning_rate": 1.9309241594190614e-05,
      "loss": 0.0247,
      "step": 500
    },
    {
      "epoch": 0.634426994246618,
      "grad_norm": 0.5046518768661232,
      "learning_rate": 1.925534407343097e-05,
      "loss": 0.0219,
      "step": 510
    },
    {
      "epoch": 0.6468667392318458,
      "grad_norm": 0.10703238283095413,
      "learning_rate": 1.919950313494939e-05,
      "loss": 0.0222,
      "step": 520
    },
    {
      "epoch": 0.6593064842170735,
      "grad_norm": 0.12518845368858994,
      "learning_rate": 1.9141730504108923e-05,
      "loss": 0.0234,
      "step": 530
    },
    {
      "epoch": 0.6717462292023013,
      "grad_norm": 0.10271623731558199,
      "learning_rate": 1.9082038311885326e-05,
      "loss": 0.0243,
      "step": 540
    },
    {
      "epoch": 0.6841859741875291,
      "grad_norm": 0.10981631074412306,
      "learning_rate": 1.902043909231984e-05,
      "loss": 0.0229,
      "step": 550
    },
    {
      "epoch": 0.696625719172757,
      "grad_norm": 0.13369798785758763,
      "learning_rate": 1.8956945779887288e-05,
      "loss": 0.0233,
      "step": 560
    },
    {
      "epoch": 0.7090654641579848,
      "grad_norm": 0.143620544527637,
      "learning_rate": 1.889157170678015e-05,
      "loss": 0.0219,
      "step": 570
    },
    {
      "epoch": 0.7215052091432126,
      "grad_norm": 0.177237676722625,
      "learning_rate": 1.882433060010907e-05,
      "loss": 0.0245,
      "step": 580
    },
    {
      "epoch": 0.7339449541284404,
      "grad_norm": 0.130978891364368,
      "learning_rate": 1.8755236579020503e-05,
      "loss": 0.0212,
      "step": 590
    },
    {
      "epoch": 0.7463846991136681,
      "grad_norm": 0.45993941931340915,
      "learning_rate": 1.868430415173196e-05,
      "loss": 0.0226,
      "step": 600
    },
    {
      "epoch": 0.7588244440988959,
      "grad_norm": 0.11832698219980749,
      "learning_rate": 1.861154821248565e-05,
      "loss": 0.0237,
      "step": 610
    },
    {
      "epoch": 0.7712641890841238,
      "grad_norm": 0.10265394845541156,
      "learning_rate": 1.8536984038420972e-05,
      "loss": 0.022,
      "step": 620
    },
    {
      "epoch": 0.7837039340693516,
      "grad_norm": 0.1663599434930081,
      "learning_rate": 1.84606272863667e-05,
      "loss": 0.0218,
      "step": 630
    },
    {
      "epoch": 0.7961436790545794,
      "grad_norm": 0.1117446647179745,
      "learning_rate": 1.8382493989553352e-05,
      "loss": 0.0213,
      "step": 640
    },
    {
      "epoch": 0.8085834240398072,
      "grad_norm": 0.11598938811295172,
      "learning_rate": 1.83026005542466e-05,
      "loss": 0.0217,
      "step": 650
    },
    {
      "epoch": 0.821023169025035,
      "grad_norm": 0.09914344844457808,
      "learning_rate": 1.8220963756302294e-05,
      "loss": 0.019,
      "step": 660
    },
    {
      "epoch": 0.8334629140102627,
      "grad_norm": 0.0857295749543062,
      "learning_rate": 1.8137600737643915e-05,
      "loss": 0.0217,
      "step": 670
    },
    {
      "epoch": 0.8459026589954906,
      "grad_norm": 0.09784476893759307,
      "learning_rate": 1.8052529002663142e-05,
      "loss": 0.0216,
      "step": 680
    },
    {
      "epoch": 0.8583424039807184,
      "grad_norm": 0.16782193537664652,
      "learning_rate": 1.7965766414544328e-05,
      "loss": 0.0207,
      "step": 690
    },
    {
      "epoch": 0.8707821489659462,
      "grad_norm": 0.19743401745663816,
      "learning_rate": 1.7877331191513615e-05,
      "loss": 0.0204,
      "step": 700
    },
    {
      "epoch": 0.883221893951174,
      "grad_norm": 0.11462401822486391,
      "learning_rate": 1.778724190301351e-05,
      "loss": 0.0208,
      "step": 710
    },
    {
      "epoch": 0.8956616389364018,
      "grad_norm": 0.11484923933550127,
      "learning_rate": 1.769551746580372e-05,
      "loss": 0.0203,
      "step": 720
    },
    {
      "epoch": 0.9081013839216296,
      "grad_norm": 0.10647239094783058,
      "learning_rate": 1.7602177139989046e-05,
      "loss": 0.02,
      "step": 730
    },
    {
      "epoch": 0.9205411289068575,
      "grad_norm": 0.19684589529140026,
      "learning_rate": 1.750724052497518e-05,
      "loss": 0.0193,
      "step": 740
    },
    {
      "epoch": 0.9329808738920852,
      "grad_norm": 0.11502794879173585,
      "learning_rate": 1.7410727555353282e-05,
      "loss": 0.0187,
      "step": 750
    },
    {
      "epoch": 0.945420618877313,
      "grad_norm": 0.16018525075457607,
      "learning_rate": 1.731265849671413e-05,
      "loss": 0.0206,
      "step": 760
    },
    {
      "epoch": 0.9578603638625408,
      "grad_norm": 0.15612624834062397,
      "learning_rate": 1.721305394139282e-05,
      "loss": 0.0208,
      "step": 770
    },
    {
      "epoch": 0.9703001088477686,
      "grad_norm": 0.09254252754570763,
      "learning_rate": 1.7111934804144804e-05,
      "loss": 0.0196,
      "step": 780
    },
    {
      "epoch": 0.9827398538329964,
      "grad_norm": 0.09488119858264474,
      "learning_rate": 1.700932231775428e-05,
      "loss": 0.0197,
      "step": 790
    },
    {
      "epoch": 0.9951795988182243,
      "grad_norm": 0.20941185378685676,
      "learning_rate": 1.6905238028575764e-05,
      "loss": 0.0176,
      "step": 800
    },
    {
      "epoch": 1.0074638469911368,
      "grad_norm": 0.181375037457905,
      "learning_rate": 1.679970379200983e-05,
      "loss": 0.0194,
      "step": 810
    },
    {
      "epoch": 1.0199035919763644,
      "grad_norm": 0.09468153651297802,
      "learning_rate": 1.669274176791395e-05,
      "loss": 0.0179,
      "step": 820
    },
    {
      "epoch": 1.0323433369615922,
      "grad_norm": 0.08407573052276653,
      "learning_rate": 1.6584374415949446e-05,
      "loss": 0.0174,
      "step": 830
    },
    {
      "epoch": 1.04478308194682,
      "grad_norm": 0.11134839001053291,
      "learning_rate": 1.6474624490865417e-05,
      "loss": 0.0172,
      "step": 840
    },
    {
      "epoch": 1.0572228269320478,
      "grad_norm": 0.15157262336734084,
      "learning_rate": 1.6363515037720774e-05,
      "loss": 0.0178,
      "step": 850
    },
    {
      "epoch": 1.0696625719172757,
      "grad_norm": 0.16013376889162095,
      "learning_rate": 1.6251069387045277e-05,
      "loss": 0.017,
      "step": 860
    },
    {
      "epoch": 1.0821023169025035,
      "grad_norm": 0.10948196093183724,
      "learning_rate": 1.6137311149940633e-05,
      "loss": 0.0197,
      "step": 870
    },
    {
      "epoch": 1.0945420618877313,
      "grad_norm": 0.1307924047930049,
      "learning_rate": 1.6022264213122694e-05,
      "loss": 0.0169,
      "step": 880
    },
    {
      "epoch": 1.106981806872959,
      "grad_norm": 1.088622998132483,
      "learning_rate": 1.5905952733905777e-05,
      "loss": 0.0182,
      "step": 890
    },
    {
      "epoch": 1.119421551858187,
      "grad_norm": 0.272945292297306,
      "learning_rate": 1.578840113513015e-05,
      "loss": 0.019,
      "step": 900
    },
    {
      "epoch": 1.1318612968434147,
      "grad_norm": 0.11060732816349034,
      "learning_rate": 1.5669634100033798e-05,
      "loss": 0.0177,
      "step": 910
    },
    {
      "epoch": 1.1443010418286426,
      "grad_norm": 0.10272292133621486,
      "learning_rate": 1.5549676567069448e-05,
      "loss": 0.0174,
      "step": 920
    },
    {
      "epoch": 1.1567407868138704,
      "grad_norm": 0.10169745449883333,
      "learning_rate": 1.5428553724668103e-05,
      "loss": 0.0181,
      "step": 930
    },
    {
      "epoch": 1.1691805317990982,
      "grad_norm": 0.09577367827512895,
      "learning_rate": 1.5306291005949964e-05,
      "loss": 0.0171,
      "step": 940
    },
    {
      "epoch": 1.181620276784326,
      "grad_norm": 0.18860005518978146,
      "learning_rate": 1.518291408338409e-05,
      "loss": 0.0159,
      "step": 950
    },
    {
      "epoch": 1.1940600217695536,
      "grad_norm": 0.11492487539850668,
      "learning_rate": 1.5058448863397716e-05,
      "loss": 0.0174,
      "step": 960
    },
    {
      "epoch": 1.2064997667547814,
      "grad_norm": 0.10088384003615597,
      "learning_rate": 1.4932921480936491e-05,
      "loss": 0.0154,
      "step": 970
    },
    {
      "epoch": 1.2189395117400093,
      "grad_norm": 0.12847829935015723,
      "learning_rate": 1.48063582939767e-05,
      "loss": 0.0169,
      "step": 980
    },
    {
      "epoch": 1.231379256725237,
      "grad_norm": 0.28169364450115963,
      "learning_rate": 1.4678785877990699e-05,
      "loss": 0.0166,
      "step": 990
    },
    {
      "epoch": 1.2438190017104649,
      "grad_norm": 0.08429180623697989,
      "learning_rate": 1.4550231020366619e-05,
      "loss": 0.0174,
      "step": 1000
    },
    {
      "epoch": 1.2562587466956927,
      "grad_norm": 0.08463281186284626,
      "learning_rate": 1.4420720714783635e-05,
      "loss": 0.0166,
      "step": 1010
    },
    {
      "epoch": 1.2686984916809205,
      "grad_norm": 0.08762626199625026,
      "learning_rate": 1.4290282155543876e-05,
      "loss": 0.0175,
      "step": 1020
    },
    {
      "epoch": 1.2811382366661483,
      "grad_norm": 0.07933117353597849,
      "learning_rate": 1.4158942731862229e-05,
      "loss": 0.0153,
      "step": 1030
    },
    {
      "epoch": 1.2935779816513762,
      "grad_norm": 0.10533333804089806,
      "learning_rate": 1.40267300221152e-05,
      "loss": 0.016,
      "step": 1040
    },
    {
      "epoch": 1.306017726636604,
      "grad_norm": 0.07414521830795055,
      "learning_rate": 1.3893671788050073e-05,
      "loss": 0.0155,
      "step": 1050
    },
    {
      "epoch": 1.3184574716218318,
      "grad_norm": 0.08946617386194489,
      "learning_rate": 1.3759795968955555e-05,
      "loss": 0.0161,
      "step": 1060
    },
    {
      "epoch": 1.3308972166070596,
      "grad_norm": 0.08424657087243657,
      "learning_rate": 1.3625130675795135e-05,
      "loss": 0.0159,
      "step": 1070
    },
    {
      "epoch": 1.3433369615922874,
      "grad_norm": 0.154667760134294,
      "learning_rate": 1.3489704185304417e-05,
      "loss": 0.0166,
      "step": 1080
    },
    {
      "epoch": 1.3557767065775153,
      "grad_norm": 0.25222266424133033,
      "learning_rate": 1.3353544934053618e-05,
      "loss": 0.0159,
      "step": 1090
    },
    {
      "epoch": 1.368216451562743,
      "grad_norm": 0.15134319898088155,
      "learning_rate": 1.3216681512476522e-05,
      "loss": 0.0169,
      "step": 1100
    },
    {
      "epoch": 1.380656196547971,
      "grad_norm": 0.2540968932972435,
      "learning_rate": 1.3079142658867124e-05,
      "loss": 0.0166,
      "step": 1110
    },
    {
      "epoch": 1.3930959415331985,
      "grad_norm": 0.1895056955473711,
      "learning_rate": 1.2940957253345214e-05,
      "loss": 0.0154,
      "step": 1120
    },
    {
      "epoch": 1.4055356865184263,
      "grad_norm": 0.16105235972760468,
      "learning_rate": 1.2802154311792196e-05,
      "loss": 0.0168,
      "step": 1130
    },
    {
      "epoch": 1.4179754315036541,
      "grad_norm": 0.08102283676505807,
      "learning_rate": 1.26627629797584e-05,
      "loss": 0.0168,
      "step": 1140
    },
    {
      "epoch": 1.430415176488882,
      "grad_norm": 0.12577562095667424,
      "learning_rate": 1.2522812526343149e-05,
      "loss": 0.0153,
      "step": 1150
    },
    {
      "epoch": 1.4428549214741098,
      "grad_norm": 0.08130680092717123,
      "learning_rate": 1.2382332338048899e-05,
      "loss": 0.015,
      "step": 1160
    },
    {
      "epoch": 1.4552946664593376,
      "grad_norm": 0.1058774954009169,
      "learning_rate": 1.2241351912610726e-05,
      "loss": 0.0165,
      "step": 1170
    },
    {
      "epoch": 1.4677344114445654,
      "grad_norm": 0.07782679016018124,
      "learning_rate": 1.2099900852802449e-05,
      "loss": 0.0161,
      "step": 1180
    },
    {
      "epoch": 1.4801741564297932,
      "grad_norm": 0.0952879760193118,
      "learning_rate": 1.1958008860220711e-05,
      "loss": 0.015,
      "step": 1190
    },
    {
      "epoch": 1.492613901415021,
      "grad_norm": 0.12845390703143658,
      "learning_rate": 1.1815705729048283e-05,
      "loss": 0.0153,
      "step": 1200
    },
    {
      "epoch": 1.5050536464002486,
      "grad_norm": 0.3195211355717146,
      "learning_rate": 1.1673021339797967e-05,
      "loss": 0.0157,
      "step": 1210
    },
    {
      "epoch": 1.5174933913854765,
      "grad_norm": 0.11061694405451432,
      "learning_rate": 1.1529985653038325e-05,
      "loss": 0.0161,
      "step": 1220
    },
    {
      "epoch": 1.5299331363707043,
      "grad_norm": 0.3459773839594,
      "learning_rate": 1.1386628703102634e-05,
      "loss": 0.0155,
      "step": 1230
    },
    {
      "epoch": 1.542372881355932,
      "grad_norm": 0.34577546641204,
      "learning_rate": 1.1242980591782344e-05,
      "loss": 0.0149,
      "step": 1240
    },
    {
      "epoch": 1.55481262634116,
      "grad_norm": 0.17666476314020332,
      "learning_rate": 1.1099071482006361e-05,
      "loss": 0.0153,
      "step": 1250
    },
    {
      "epoch": 1.5672523713263877,
      "grad_norm": 0.08168049687427667,
      "learning_rate": 1.0954931591507489e-05,
      "loss": 0.0148,
      "step": 1260
    },
    {
      "epoch": 1.5796921163116155,
      "grad_norm": 0.06905046308060793,
      "learning_rate": 1.0810591186477402e-05,
      "loss": 0.0145,
      "step": 1270
    },
    {
      "epoch": 1.5921318612968434,
      "grad_norm": 0.07864628972274919,
      "learning_rate": 1.0666080575211373e-05,
      "loss": 0.015,
      "step": 1280
    },
    {
      "epoch": 1.6045716062820712,
      "grad_norm": 0.06955473080486402,
      "learning_rate": 1.0521430101744238e-05,
      "loss": 0.0143,
      "step": 1290
    },
    {
      "epoch": 1.617011351267299,
      "grad_norm": 0.06905190911851168,
      "learning_rate": 1.03766701394788e-05,
      "loss": 0.0145,
      "step": 1300
    },
    {
      "epoch": 1.6294510962525268,
      "grad_norm": 0.13076128523192543,
      "learning_rate": 1.023183108480809e-05,
      "loss": 0.0153,
      "step": 1310
    },
    {
      "epoch": 1.6418908412377546,
      "grad_norm": 0.16263053965593313,
      "learning_rate": 1.0086943350732797e-05,
      "loss": 0.015,
      "step": 1320
    },
    {
      "epoch": 1.6543305862229825,
      "grad_norm": 0.09571388113006955,
      "learning_rate": 9.942037360475205e-06,
      "loss": 0.0154,
      "step": 1330
    },
    {
      "epoch": 1.6667703312082103,
      "grad_norm": 0.08754424890443553,
      "learning_rate": 9.797143541091e-06,
      "loss": 0.0145,
      "step": 1340
    },
    {
      "epoch": 1.679210076193438,
      "grad_norm": 0.10644538613606132,
      "learning_rate": 9.65229231708025e-06,
      "loss": 0.0151,
      "step": 1350
    },
    {
      "epoch": 1.691649821178666,
      "grad_norm": 0.06371833554085514,
      "learning_rate": 9.507514103998936e-06,
      "loss": 0.0155,
      "step": 1360
    },
    {
      "epoch": 1.7040895661638937,
      "grad_norm": 0.11242274171252464,
      "learning_rate": 9.362839302072354e-06,
      "loss": 0.014,
      "step": 1370
    },
    {
      "epoch": 1.7165293111491216,
      "grad_norm": 0.0744984110285503,
      "learning_rate": 9.218298289811748e-06,
      "loss": 0.0144,
      "step": 1380
    },
    {
      "epoch": 1.7289690561343494,
      "grad_norm": 0.084129194637179,
      "learning_rate": 9.073921417635485e-06,
      "loss": 0.0142,
      "step": 1390
    },
    {
      "epoch": 1.7414088011195772,
      "grad_norm": 0.08826537552746329,
      "learning_rate": 8.92973900149615e-06,
      "loss": 0.0149,
      "step": 1400
    },
    {
      "epoch": 1.753848546104805,
      "grad_norm": 0.07923942003502486,
      "learning_rate": 8.785781316514841e-06,
      "loss": 0.0149,
      "step": 1410
    },
    {
      "epoch": 1.7662882910900326,
      "grad_norm": 0.07975941139506229,
      "learning_rate": 8.642078590624097e-06,
      "loss": 0.0157,
      "step": 1420
    },
    {
      "epoch": 1.7787280360752604,
      "grad_norm": 0.13647611265750945,
      "learning_rate": 8.498660998220669e-06,
      "loss": 0.0148,
      "step": 1430
    },
    {
      "epoch": 1.7911677810604882,
      "grad_norm": 0.127909426802921,
      "learning_rate": 8.355558653829586e-06,
      "loss": 0.0148,
      "step": 1440
    },
    {
      "epoch": 1.803607526045716,
      "grad_norm": 0.08034649199561664,
      "learning_rate": 8.212801605780754e-06,
      "loss": 0.0137,
      "step": 1450
    },
    {
      "epoch": 1.8160472710309439,
      "grad_norm": 0.1156382724094793,
      "learning_rate": 8.07041982989948e-06,
      "loss": 0.0136,
      "step": 1460
    },
    {
      "epoch": 1.8284870160161717,
      "grad_norm": 0.07192051958747497,
      "learning_rate": 7.928443223212216e-06,
      "loss": 0.0131,
      "step": 1470
    },
    {
      "epoch": 1.8409267610013995,
      "grad_norm": 0.07316406629339688,
      "learning_rate": 7.786901597668824e-06,
      "loss": 0.0139,
      "step": 1480
    },
    {
      "epoch": 1.853366505986627,
      "grad_norm": 0.06571537931190587,
      "learning_rate": 7.64582467388275e-06,
      "loss": 0.0128,
      "step": 1490
    },
    {
      "epoch": 1.865806250971855,
      "grad_norm": 0.0762378122440442,
      "learning_rate": 7.505242074890333e-06,
      "loss": 0.0149,
      "step": 1500
    },
    {
      "epoch": 1.8782459959570827,
      "grad_norm": 0.11357166687216505,
      "learning_rate": 7.3651833199306355e-06,
      "loss": 0.0141,
      "step": 1510
    },
    {
      "epoch": 1.8906857409423106,
      "grad_norm": 0.07835172216036307,
      "learning_rate": 7.225677818247041e-06,
      "loss": 0.0143,
      "step": 1520
    },
    {
      "epoch": 1.9031254859275384,
      "grad_norm": 0.07485227727372203,
      "learning_rate": 7.086754862911982e-06,
      "loss": 0.0147,
      "step": 1530
    },
    {
      "epoch": 1.9155652309127662,
      "grad_norm": 0.11434773863926712,
      "learning_rate": 6.948443624676017e-06,
      "loss": 0.0141,
      "step": 1540
    },
    {
      "epoch": 1.928004975897994,
      "grad_norm": 0.0944371868336863,
      "learning_rate": 6.810773145842653e-06,
      "loss": 0.0145,
      "step": 1550
    },
    {
      "epoch": 1.9404447208832218,
      "grad_norm": 0.10904635124038897,
      "learning_rate": 6.673772334170063e-06,
      "loss": 0.0138,
      "step": 1560
    },
    {
      "epoch": 1.9528844658684497,
      "grad_norm": 0.10348119556114112,
      "learning_rate": 6.537469956801128e-06,
      "loss": 0.0142,
      "step": 1570
    },
    {
      "epoch": 1.9653242108536775,
      "grad_norm": 0.08421933767316382,
      "learning_rate": 6.4018946342229495e-06,
      "loss": 0.0126,
      "step": 1580
    },
    {
      "epoch": 1.9777639558389053,
      "grad_norm": 0.07815653103052027,
      "learning_rate": 6.267074834257199e-06,
      "loss": 0.0132,
      "step": 1590
    },
    {
      "epoch": 1.9902037008241331,
      "grad_norm": 0.07496826979475398,
      "learning_rate": 6.1330388660824906e-06,
      "loss": 0.0139,
      "step": 1600
    },
    {
      "epoch": 2.0024879489970457,
      "grad_norm": 0.0785889612753055,
      "learning_rate": 5.999814874290084e-06,
      "loss": 0.013,
      "step": 1610
    },
    {
      "epoch": 2.0149276939822736,
      "grad_norm": 0.11822070670557466,
      "learning_rate": 5.867430832974145e-06,
      "loss": 0.0113,
      "step": 1620
    },
    {
      "epoch": 2.0273674389675014,
      "grad_norm": 0.08387782280109485,
      "learning_rate": 5.735914539857799e-06,
      "loss": 0.0122,
      "step": 1630
    },
    {
      "epoch": 2.0398071839527288,
      "grad_norm": 0.07657983351411304,
      "learning_rate": 5.60529361045623e-06,
      "loss": 0.0125,
      "step": 1640
    },
    {
      "epoch": 2.0522469289379566,
      "grad_norm": 0.0751504296879094,
      "learning_rate": 5.4755954722780236e-06,
      "loss": 0.0116,
      "step": 1650
    },
    {
      "epoch": 2.0646866739231844,
      "grad_norm": 0.0670877568099031,
      "learning_rate": 5.346847359066006e-06,
      "loss": 0.0119,
      "step": 1660
    },
    {
      "epoch": 2.077126418908412,
      "grad_norm": 0.1033921415346376,
      "learning_rate": 5.21907630507875e-06,
      "loss": 0.0118,
      "step": 1670
    },
    {
      "epoch": 2.08956616389364,
      "grad_norm": 0.07925266750365705,
      "learning_rate": 5.092309139413982e-06,
      "loss": 0.0116,
      "step": 1680
    },
    {
      "epoch": 2.102005908878868,
      "grad_norm": 0.06476533507187097,
      "learning_rate": 4.966572480375076e-06,
      "loss": 0.0117,
      "step": 1690
    },
    {
      "epoch": 2.1144456538640957,
      "grad_norm": 0.0703877565719974,
      "learning_rate": 4.84189272988177e-06,
      "loss": 0.0112,
      "step": 1700
    },
    {
      "epoch": 2.1268853988493235,
      "grad_norm": 0.07162026622912554,
      "learning_rate": 4.71829606792639e-06,
      "loss": 0.0118,
      "step": 1710
    },
    {
      "epoch": 2.1393251438345513,
      "grad_norm": 0.07209478907389254,
      "learning_rate": 4.595808447076578e-06,
      "loss": 0.0122,
      "step": 1720
    },
    {
      "epoch": 2.151764888819779,
      "grad_norm": 0.07615553579496778,
      "learning_rate": 4.47445558702587e-06,
      "loss": 0.0113,
      "step": 1730
    },
    {
      "epoch": 2.164204633805007,
      "grad_norm": 0.07718560204490735,
      "learning_rate": 4.354262969193097e-06,
      "loss": 0.0119,
      "step": 1740
    },
    {
      "epoch": 2.1766443787902348,
      "grad_norm": 0.07348631771208182,
      "learning_rate": 4.235255831371879e-06,
      "loss": 0.0115,
      "step": 1750
    },
    {
      "epoch": 2.1890841237754626,
      "grad_norm": 0.08594241606944986,
      "learning_rate": 4.1174591624312235e-06,
      "loss": 0.0135,
      "step": 1760
    },
    {
      "epoch": 2.2015238687606904,
      "grad_norm": 0.09192670032838215,
      "learning_rate": 4.000897697068418e-06,
      "loss": 0.0111,
      "step": 1770
    },
    {
      "epoch": 2.213963613745918,
      "grad_norm": 0.14183042776581714,
      "learning_rate": 3.88559591061531e-06,
      "loss": 0.0112,
      "step": 1780
    },
    {
      "epoch": 2.226403358731146,
      "grad_norm": 0.10661299995011776,
      "learning_rate": 3.7715780138989965e-06,
      "loss": 0.0115,
      "step": 1790
    },
    {
      "epoch": 2.238843103716374,
      "grad_norm": 0.07655545886588545,
      "learning_rate": 3.658867948158119e-06,
      "loss": 0.0104,
      "step": 1800
    },
    {
      "epoch": 2.2512828487016017,
      "grad_norm": 0.06029047134287144,
      "learning_rate": 3.5474893800157005e-06,
      "loss": 0.0111,
      "step": 1810
    },
    {
      "epoch": 2.2637225936868295,
      "grad_norm": 0.06953163791899282,
      "learning_rate": 3.4374656965097086e-06,
      "loss": 0.0119,
      "step": 1820
    },
    {
      "epoch": 2.2761623386720573,
      "grad_norm": 0.07575417535866263,
      "learning_rate": 3.3288200001822624e-06,
      "loss": 0.0112,
      "step": 1830
    },
    {
      "epoch": 2.288602083657285,
      "grad_norm": 0.06786765717592823,
      "learning_rate": 3.22157510422864e-06,
      "loss": 0.0127,
      "step": 1840
    },
    {
      "epoch": 2.301041828642513,
      "grad_norm": 0.10452868686240589,
      "learning_rate": 3.115753527706986e-06,
      "loss": 0.0121,
      "step": 1850
    },
    {
      "epoch": 2.3134815736277408,
      "grad_norm": 0.07244062565338512,
      "learning_rate": 3.0113774908098257e-06,
      "loss": 0.0118,
      "step": 1860
    },
    {
      "epoch": 2.3259213186129686,
      "grad_norm": 0.07590029743248466,
      "learning_rate": 2.9084689101983076e-06,
      "loss": 0.0117,
      "step": 1870
    },
    {
      "epoch": 2.3383610635981964,
      "grad_norm": 0.1126832171191715,
      "learning_rate": 2.8070493944001843e-06,
      "loss": 0.0113,
      "step": 1880
    },
    {
      "epoch": 2.350800808583424,
      "grad_norm": 0.07204328570173935,
      "learning_rate": 2.70714023927251e-06,
      "loss": 0.0114,
      "step": 1890
    },
    {
      "epoch": 2.363240553568652,
      "grad_norm": 0.07240113234919121,
      "learning_rate": 2.608762423529969e-06,
      "loss": 0.0133,
      "step": 1900
    },
    {
      "epoch": 2.3756802985538794,
      "grad_norm": 0.09682767647010854,
      "learning_rate": 2.5119366043398265e-06,
      "loss": 0.0106,
      "step": 1910
    },
    {
      "epoch": 2.3881200435391072,
      "grad_norm": 0.08470381089691305,
      "learning_rate": 2.416683112984355e-06,
      "loss": 0.0115,
      "step": 1920
    },
    {
      "epoch": 2.400559788524335,
      "grad_norm": 0.0843703713597429,
      "learning_rate": 2.323021950591743e-06,
      "loss": 0.0102,
      "step": 1930
    },
    {
      "epoch": 2.412999533509563,
      "grad_norm": 0.0666017917756727,
      "learning_rate": 2.2309727839362716e-06,
      "loss": 0.0114,
      "step": 1940
    },
    {
      "epoch": 2.4254392784947907,
      "grad_norm": 0.06152131645691254,
      "learning_rate": 2.1405549413087543e-06,
      "loss": 0.0115,
      "step": 1950
    },
    {
      "epoch": 2.4378790234800185,
      "grad_norm": 0.09394368820589998,
      "learning_rate": 2.051787408457997e-06,
      "loss": 0.0105,
      "step": 1960
    },
    {
      "epoch": 2.4503187684652463,
      "grad_norm": 0.07889036601155558,
      "learning_rate": 1.964688824604234e-06,
      "loss": 0.0112,
      "step": 1970
    },
    {
      "epoch": 2.462758513450474,
      "grad_norm": 0.08801513247697691,
      "learning_rate": 1.8792774785252798e-06,
      "loss": 0.0122,
      "step": 1980
    },
    {
      "epoch": 2.475198258435702,
      "grad_norm": 0.06686354345628483,
      "learning_rate": 1.795571304716316e-06,
      "loss": 0.0104,
      "step": 1990
    },
    {
      "epoch": 2.4876380034209298,
      "grad_norm": 0.06685944863330902,
      "learning_rate": 1.713587879624019e-06,
      "loss": 0.0114,
      "step": 2000
    },
    {
      "epoch": 2.5000777484061576,
      "grad_norm": 0.06943091210316915,
      "learning_rate": 1.6333444179559078e-06,
      "loss": 0.0109,
      "step": 2010
    },
    {
      "epoch": 2.5125174933913854,
      "grad_norm": 0.06851292056328774,
      "learning_rate": 1.5548577690656386e-06,
      "loss": 0.0107,
      "step": 2020
    },
    {
      "epoch": 2.5249572383766132,
      "grad_norm": 0.06369399303852112,
      "learning_rate": 1.4781444134150048e-06,
      "loss": 0.0103,
      "step": 2030
    },
    {
      "epoch": 2.537396983361841,
      "grad_norm": 0.06578267762761443,
      "learning_rate": 1.4032204591134092e-06,
      "loss": 0.0115,
      "step": 2040
    },
    {
      "epoch": 2.549836728347069,
      "grad_norm": 0.06177272223447156,
      "learning_rate": 1.3301016385355093e-06,
      "loss": 0.0109,
      "step": 2050
    },
    {
      "epoch": 2.5622764733322967,
      "grad_norm": 0.06867656463826503,
      "learning_rate": 1.2588033050177828e-06,
      "loss": 0.0118,
      "step": 2060
    },
    {
      "epoch": 2.5747162183175245,
      "grad_norm": 0.05808795434357701,
      "learning_rate": 1.1893404296346422e-06,
      "loss": 0.0101,
      "step": 2070
    },
    {
      "epoch": 2.5871559633027523,
      "grad_norm": 0.06342258704121058,
      "learning_rate": 1.121727598054867e-06,
      "loss": 0.0109,
      "step": 2080
    },
    {
      "epoch": 2.59959570828798,
      "grad_norm": 0.06881727205540568,
      "learning_rate": 1.0559790074789134e-06,
      "loss": 0.0102,
      "step": 2090
    },
    {
      "epoch": 2.612035453273208,
      "grad_norm": 0.06558270498757422,
      "learning_rate": 9.921084636578237e-07,
      "loss": 0.0105,
      "step": 2100
    },
    {
      "epoch": 2.624475198258436,
      "grad_norm": 0.06250403568916518,
      "learning_rate": 9.301293779943321e-07,
      "loss": 0.0107,
      "step": 2110
    },
    {
      "epoch": 2.6369149432436636,
      "grad_norm": 0.07965076756190062,
      "learning_rate": 8.700547647267521e-07,
      "loss": 0.0127,
      "step": 2120
    },
    {
      "epoch": 2.6493546882288914,
      "grad_norm": 0.06752093682311762,
      "learning_rate": 8.118972381962853e-07,
      "loss": 0.0113,
      "step": 2130
    },
    {
      "epoch": 2.6617944332141192,
      "grad_norm": 0.055232374525591425,
      "learning_rate": 7.5566901019828e-07,
      "loss": 0.0109,
      "step": 2140
    },
    {
      "epoch": 2.674234178199347,
      "grad_norm": 0.06071359581593855,
      "learning_rate": 7.013818874180323e-07,
      "loss": 0.0109,
      "step": 2150
    },
    {
      "epoch": 2.686673923184575,
      "grad_norm": 0.06935994936974366,
      "learning_rate": 6.490472689516358e-07,
      "loss": 0.0121,
      "step": 2160
    },
    {
      "epoch": 2.6991136681698027,
      "grad_norm": 0.06724035806496545,
      "learning_rate": 5.986761439124289e-07,
      "loss": 0.011,
      "step": 2170
    },
    {
      "epoch": 2.7115534131550305,
      "grad_norm": 0.05962193536771214,
      "learning_rate": 5.502790891235188e-07,
      "loss": 0.0104,
      "step": 2180
    },
    {
      "epoch": 2.7239931581402583,
      "grad_norm": 0.06534442932179661,
      "learning_rate": 5.038662668968885e-07,
      "loss": 0.0118,
      "step": 2190
    },
    {
      "epoch": 2.736432903125486,
      "grad_norm": 0.06386278747449284,
      "learning_rate": 4.594474228995305e-07,
      "loss": 0.0105,
      "step": 2200
    },
    {
      "epoch": 2.748872648110714,
      "grad_norm": 0.07965498231812315,
      "learning_rate": 4.1703188410707087e-07,
      "loss": 0.0112,
      "step": 2210
    },
    {
      "epoch": 2.761312393095942,
      "grad_norm": 0.06905252277698515,
      "learning_rate": 3.7662855684532094e-07,
      "loss": 0.0103,
      "step": 2220
    },
    {
      "epoch": 2.7737521380811696,
      "grad_norm": 0.06868620310161447,
      "learning_rate": 3.3824592492013087e-07,
      "loss": 0.0116,
      "step": 2230
    },
    {
      "epoch": 2.786191883066397,
      "grad_norm": 0.06445402999949239,
      "learning_rate": 3.018920478359921e-07,
      "loss": 0.0108,
      "step": 2240
    },
    {
      "epoch": 2.798631628051625,
      "grad_norm": 0.06986517506942372,
      "learning_rate": 2.6757455910370486e-07,
      "loss": 0.0103,
      "step": 2250
    },
    {
      "epoch": 2.8110713730368526,
      "grad_norm": 0.06071637832773147,
      "learning_rate": 2.3530066463752665e-07,
      "loss": 0.0114,
      "step": 2260
    },
    {
      "epoch": 2.8235111180220804,
      "grad_norm": 0.06422168949049402,
      "learning_rate": 2.0507714124207157e-07,
      "loss": 0.0111,
      "step": 2270
    },
    {
      "epoch": 2.8359508630073083,
      "grad_norm": 0.05843126847655036,
      "learning_rate": 1.7691033518934864e-07,
      "loss": 0.0108,
      "step": 2280
    },
    {
      "epoch": 2.848390607992536,
      "grad_norm": 0.06903086028558499,
      "learning_rate": 1.5080616088616884e-07,
      "loss": 0.0107,
      "step": 2290
    },
    {
      "epoch": 2.860830352977764,
      "grad_norm": 0.07620555195872948,
      "learning_rate": 1.2677009963226073e-07,
      "loss": 0.0105,
      "step": 2300
    },
    {
      "epoch": 2.8732700979629917,
      "grad_norm": 0.0644333060708327,
      "learning_rate": 1.0480719846931775e-07,
      "loss": 0.0114,
      "step": 2310
    },
    {
      "epoch": 2.8857098429482195,
      "grad_norm": 0.06561279672863249,
      "learning_rate": 8.492206912122492e-08,
      "loss": 0.0113,
      "step": 2320
    },
    {
      "epoch": 2.8981495879334473,
      "grad_norm": 0.059740986829645897,
      "learning_rate": 6.711888702570556e-08,
      "loss": 0.01,
      "step": 2330
    },
    {
      "epoch": 2.910589332918675,
      "grad_norm": 0.06268877873361439,
      "learning_rate": 5.140139045756609e-08,
      "loss": 0.0114,
      "step": 2340
    },
    {
      "epoch": 2.923029077903903,
      "grad_norm": 0.06210995212416848,
      "learning_rate": 3.7772879743749324e-08,
      "loss": 0.0105,
      "step": 2350
    },
    {
      "epoch": 2.935468822889131,
      "grad_norm": 0.057688802256361,
      "learning_rate": 2.62362165703256e-08,
      "loss": 0.0113,
      "step": 2360
    },
    {
      "epoch": 2.9479085678743586,
      "grad_norm": 0.06021148243668492,
      "learning_rate": 1.6793823381614506e-08,
      "loss": 0.0107,
      "step": 2370
    },
    {
      "epoch": 2.9603483128595864,
      "grad_norm": 0.06350014557184043,
      "learning_rate": 9.447682871519537e-09,
      "loss": 0.0112,
      "step": 2380
    },
    {
      "epoch": 2.9727880578448143,
      "grad_norm": 0.06466421436261854,
      "learning_rate": 4.199337567203365e-09,
      "loss": 0.0112,
      "step": 2390
    },
    {
      "epoch": 2.985227802830042,
      "grad_norm": 0.061480021578027864,
      "learning_rate": 1.0498895051991309e-09,
      "loss": 0.0119,
      "step": 2400
    },
    {
      "epoch": 2.996423573316747,
      "step": 2409,
      "total_flos": 5238888339603456.0,
      "train_loss": 0.027854068378731143,
      "train_runtime": 162000.0754,
      "train_samples_per_second": 1.905,
      "train_steps_per_second": 0.015
    }
  ],
  "logging_steps": 10,
  "max_steps": 2409,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 5238888339603456.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}