{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.8071860346873057,
  "eval_steps": 500,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.009039037342523022,
      "grad_norm": 7.855499744415283,
      "learning_rate": 6.000000000000001e-07,
      "loss": 1.4181,
      "step": 10
    },
    {
      "epoch": 0.018078074685046044,
      "grad_norm": 3.3778610229492188,
      "learning_rate": 1.2666666666666669e-06,
      "loss": 1.2494,
      "step": 20
    },
    {
      "epoch": 0.027117112027569064,
      "grad_norm": 2.6225311756134033,
      "learning_rate": 1.9333333333333336e-06,
      "loss": 1.0602,
      "step": 30
    },
    {
      "epoch": 0.03615614937009209,
      "grad_norm": 1.594939112663269,
      "learning_rate": 2.6e-06,
      "loss": 0.9123,
      "step": 40
    },
    {
      "epoch": 0.045195186712615104,
      "grad_norm": 0.9766483902931213,
      "learning_rate": 3.266666666666667e-06,
      "loss": 0.7841,
      "step": 50
    },
    {
      "epoch": 0.05423422405513813,
      "grad_norm": 0.43368658423423767,
      "learning_rate": 3.9333333333333335e-06,
      "loss": 0.6803,
      "step": 60
    },
    {
      "epoch": 0.06327326139766115,
      "grad_norm": 0.49014410376548767,
      "learning_rate": 4.600000000000001e-06,
      "loss": 0.5921,
      "step": 70
    },
    {
      "epoch": 0.07231229874018417,
      "grad_norm": 0.46368488669395447,
      "learning_rate": 5.2666666666666665e-06,
      "loss": 0.6214,
      "step": 80
    },
    {
      "epoch": 0.0813513360827072,
      "grad_norm": 0.46477940678596497,
      "learning_rate": 5.933333333333335e-06,
      "loss": 0.6157,
      "step": 90
    },
    {
      "epoch": 0.09039037342523021,
      "grad_norm": 0.4853752553462982,
      "learning_rate": 6.600000000000001e-06,
      "loss": 0.6628,
      "step": 100
    },
    {
      "epoch": 0.09942941076775323,
      "grad_norm": 0.36940082907676697,
      "learning_rate": 7.266666666666668e-06,
      "loss": 0.5775,
      "step": 110
    },
    {
      "epoch": 0.10846844811027626,
      "grad_norm": 0.34785982966423035,
      "learning_rate": 7.933333333333334e-06,
      "loss": 0.6343,
      "step": 120
    },
    {
      "epoch": 0.11750748545279928,
      "grad_norm": 0.4787227511405945,
      "learning_rate": 8.6e-06,
      "loss": 0.6016,
      "step": 130
    },
    {
      "epoch": 0.1265465227953223,
      "grad_norm": 0.4643513262271881,
      "learning_rate": 9.266666666666667e-06,
      "loss": 0.6388,
      "step": 140
    },
    {
      "epoch": 0.1355855601378453,
      "grad_norm": 0.384676456451416,
      "learning_rate": 9.933333333333334e-06,
      "loss": 0.5895,
      "step": 150
    },
    {
      "epoch": 0.14462459748036835,
      "grad_norm": 0.4324859380722046,
      "learning_rate": 9.99991503499922e-06,
      "loss": 0.5942,
      "step": 160
    },
    {
      "epoch": 0.15366363482289136,
      "grad_norm": 0.41830965876579285,
      "learning_rate": 9.99962133253095e-06,
      "loss": 0.6121,
      "step": 170
    },
    {
      "epoch": 0.1627026721654144,
      "grad_norm": 0.3795354962348938,
      "learning_rate": 9.999117855964797e-06,
      "loss": 0.6041,
      "step": 180
    },
    {
      "epoch": 0.1717417095079374,
      "grad_norm": 0.28756389021873474,
      "learning_rate": 9.998404626425627e-06,
      "loss": 0.5529,
      "step": 190
    },
    {
      "epoch": 0.18078074685046042,
      "grad_norm": 0.40732458233833313,
      "learning_rate": 9.997481673839125e-06,
      "loss": 0.6321,
      "step": 200
    },
    {
      "epoch": 0.18981978419298345,
      "grad_norm": 0.3699369430541992,
      "learning_rate": 9.996349036930533e-06,
      "loss": 0.5886,
      "step": 210
    },
    {
      "epoch": 0.19885882153550646,
      "grad_norm": 0.4334062337875366,
      "learning_rate": 9.995006763223028e-06,
      "loss": 0.646,
      "step": 220
    },
    {
      "epoch": 0.2078978588780295,
      "grad_norm": 0.3638211786746979,
      "learning_rate": 9.993454909035724e-06,
      "loss": 0.5996,
      "step": 230
    },
    {
      "epoch": 0.2169368962205525,
      "grad_norm": 0.377408891916275,
      "learning_rate": 9.991693539481317e-06,
      "loss": 0.5999,
      "step": 240
    },
    {
      "epoch": 0.22597593356307552,
      "grad_norm": 0.4337885081768036,
      "learning_rate": 9.989722728463345e-06,
      "loss": 0.6233,
      "step": 250
    },
    {
      "epoch": 0.23501497090559856,
      "grad_norm": 0.35624244809150696,
      "learning_rate": 9.98754255867309e-06,
      "loss": 0.6279,
      "step": 260
    },
    {
      "epoch": 0.24405400824812157,
      "grad_norm": 0.385580837726593,
      "learning_rate": 9.985153121586111e-06,
      "loss": 0.5896,
      "step": 270
    },
    {
      "epoch": 0.2530930455906446,
      "grad_norm": 0.36690962314605713,
      "learning_rate": 9.982554517458403e-06,
      "loss": 0.6212,
      "step": 280
    },
    {
      "epoch": 0.26213208293316764,
      "grad_norm": 0.3190907835960388,
      "learning_rate": 9.979746855322192e-06,
      "loss": 0.6019,
      "step": 290
    },
    {
      "epoch": 0.2711711202756906,
      "grad_norm": 0.33014318346977234,
      "learning_rate": 9.976730252981354e-06,
      "loss": 0.5906,
      "step": 300
    },
    {
      "epoch": 0.28021015761821366,
      "grad_norm": 0.4516271650791168,
      "learning_rate": 9.973504837006487e-06,
      "loss": 0.612,
      "step": 310
    },
    {
      "epoch": 0.2892491949607367,
      "grad_norm": 0.3498607873916626,
      "learning_rate": 9.97007074272958e-06,
      "loss": 0.5591,
      "step": 320
    },
    {
      "epoch": 0.2982882323032597,
      "grad_norm": 0.4618026316165924,
      "learning_rate": 9.966428114238353e-06,
      "loss": 0.5808,
      "step": 330
    },
    {
      "epoch": 0.3073272696457827,
      "grad_norm": 0.35378628969192505,
      "learning_rate": 9.962577104370206e-06,
      "loss": 0.5983,
      "step": 340
    },
    {
      "epoch": 0.31636630698830576,
      "grad_norm": 0.3605310916900635,
      "learning_rate": 9.958517874705793e-06,
      "loss": 0.5898,
      "step": 350
    },
    {
      "epoch": 0.3254053443308288,
      "grad_norm": 0.41437646746635437,
      "learning_rate": 9.954250595562267e-06,
      "loss": 0.5746,
      "step": 360
    },
    {
      "epoch": 0.3344443816733518,
      "grad_norm": 0.3715222477912903,
      "learning_rate": 9.949775445986112e-06,
      "loss": 0.5994,
      "step": 370
    },
    {
      "epoch": 0.3434834190158748,
      "grad_norm": 0.41986992955207825,
      "learning_rate": 9.945092613745642e-06,
      "loss": 0.5492,
      "step": 380
    },
    {
      "epoch": 0.35252245635839785,
      "grad_norm": 0.3996647000312805,
      "learning_rate": 9.940202295323116e-06,
      "loss": 0.6066,
      "step": 390
    },
    {
      "epoch": 0.36156149370092083,
      "grad_norm": 0.3808385729789734,
      "learning_rate": 9.935104695906506e-06,
      "loss": 0.5628,
      "step": 400
    },
    {
      "epoch": 0.37060053104344387,
      "grad_norm": 0.3926442563533783,
      "learning_rate": 9.92980002938087e-06,
      "loss": 0.6087,
      "step": 410
    },
    {
      "epoch": 0.3796395683859669,
      "grad_norm": 0.3604278862476349,
      "learning_rate": 9.924288518319394e-06,
      "loss": 0.5661,
      "step": 420
    },
    {
      "epoch": 0.3886786057284899,
      "grad_norm": 0.36008164286613464,
      "learning_rate": 9.918570393974041e-06,
      "loss": 0.5948,
      "step": 430
    },
    {
      "epoch": 0.3977176430710129,
      "grad_norm": 0.3590993285179138,
      "learning_rate": 9.912645896265858e-06,
      "loss": 0.5692,
      "step": 440
    },
    {
      "epoch": 0.40675668041353596,
      "grad_norm": 0.3519396483898163,
      "learning_rate": 9.906515273774903e-06,
      "loss": 0.5289,
      "step": 450
    },
    {
      "epoch": 0.415795717756059,
      "grad_norm": 0.33698147535324097,
      "learning_rate": 9.900178783729817e-06,
      "loss": 0.5643,
      "step": 460
    },
    {
      "epoch": 0.424834755098582,
      "grad_norm": 0.41117480397224426,
      "learning_rate": 9.89363669199703e-06,
      "loss": 0.5875,
      "step": 470
    },
    {
      "epoch": 0.433873792441105,
      "grad_norm": 0.3583988547325134,
      "learning_rate": 9.886889273069614e-06,
      "loss": 0.5494,
      "step": 480
    },
    {
      "epoch": 0.44291282978362806,
      "grad_norm": 0.3181304633617401,
      "learning_rate": 9.879936810055746e-06,
      "loss": 0.5652,
      "step": 490
    },
    {
      "epoch": 0.45195186712615104,
      "grad_norm": 0.3814559280872345,
      "learning_rate": 9.872779594666856e-06,
      "loss": 0.6046,
      "step": 500
    },
    {
      "epoch": 0.4609909044686741,
      "grad_norm": 0.33179372549057007,
      "learning_rate": 9.865417927205363e-06,
      "loss": 0.6156,
      "step": 510
    },
    {
      "epoch": 0.4700299418111971,
      "grad_norm": 0.4139130115509033,
      "learning_rate": 9.857852116552094e-06,
      "loss": 0.5993,
      "step": 520
    },
    {
      "epoch": 0.47906897915372015,
      "grad_norm": 0.32068437337875366,
      "learning_rate": 9.850082480153306e-06,
      "loss": 0.603,
      "step": 530
    },
    {
      "epoch": 0.48810801649624314,
      "grad_norm": 0.9201164841651917,
      "learning_rate": 9.842109344007386e-06,
      "loss": 0.6036,
      "step": 540
    },
    {
      "epoch": 0.4971470538387662,
      "grad_norm": 0.42034047842025757,
      "learning_rate": 9.833933042651156e-06,
      "loss": 0.5678,
      "step": 550
    },
    {
      "epoch": 0.5061860911812892,
      "grad_norm": 0.3811807632446289,
      "learning_rate": 9.825553919145845e-06,
      "loss": 0.5596,
      "step": 560
    },
    {
      "epoch": 0.5152251285238122,
      "grad_norm": 0.36168864369392395,
      "learning_rate": 9.816972325062694e-06,
      "loss": 0.5906,
      "step": 570
    },
    {
      "epoch": 0.5242641658663353,
      "grad_norm": 0.3304130733013153,
      "learning_rate": 9.808188620468204e-06,
      "loss": 0.5398,
      "step": 580
    },
    {
      "epoch": 0.5333032032088583,
      "grad_norm": 0.32785892486572266,
      "learning_rate": 9.799203173909028e-06,
      "loss": 0.5982,
      "step": 590
    },
    {
      "epoch": 0.5423422405513812,
      "grad_norm": 0.39726918935775757,
      "learning_rate": 9.790016362396506e-06,
      "loss": 0.5561,
      "step": 600
    },
    {
      "epoch": 0.5513812778939043,
      "grad_norm": 0.3174471855163574,
      "learning_rate": 9.780628571390853e-06,
      "loss": 0.5268,
      "step": 610
    },
    {
      "epoch": 0.5604203152364273,
      "grad_norm": 0.4243045151233673,
      "learning_rate": 9.771040194784973e-06,
      "loss": 0.5954,
      "step": 620
    },
    {
      "epoch": 0.5694593525789503,
      "grad_norm": 0.35231855511665344,
      "learning_rate": 9.761251634887949e-06,
      "loss": 0.5711,
      "step": 630
    },
    {
      "epoch": 0.5784983899214734,
      "grad_norm": 0.349280446767807,
      "learning_rate": 9.751263302408146e-06,
      "loss": 0.6263,
      "step": 640
    },
    {
      "epoch": 0.5875374272639964,
      "grad_norm": 0.7063365578651428,
      "learning_rate": 9.741075616435995e-06,
      "loss": 0.6093,
      "step": 650
    },
    {
      "epoch": 0.5965764646065194,
      "grad_norm": 0.3803994059562683,
      "learning_rate": 9.730689004426392e-06,
      "loss": 0.5552,
      "step": 660
    },
    {
      "epoch": 0.6056155019490425,
      "grad_norm": 0.36839503049850464,
      "learning_rate": 9.720103902180776e-06,
      "loss": 0.5987,
      "step": 670
    },
    {
      "epoch": 0.6146545392915654,
      "grad_norm": 0.39339619874954224,
      "learning_rate": 9.709320753828837e-06,
      "loss": 0.5621,
      "step": 680
    },
    {
      "epoch": 0.6236935766340884,
      "grad_norm": 0.35199737548828125,
      "learning_rate": 9.698340011809883e-06,
      "loss": 0.5732,
      "step": 690
    },
    {
      "epoch": 0.6327326139766115,
      "grad_norm": 0.37193241715431213,
      "learning_rate": 9.687162136853858e-06,
      "loss": 0.5534,
      "step": 700
    },
    {
      "epoch": 0.6417716513191345,
      "grad_norm": 0.3605504035949707,
      "learning_rate": 9.675787597962007e-06,
      "loss": 0.608,
      "step": 710
    },
    {
      "epoch": 0.6508106886616576,
      "grad_norm": 0.3748184144496918,
      "learning_rate": 9.664216872387202e-06,
      "loss": 0.5903,
      "step": 720
    },
    {
      "epoch": 0.6598497260041806,
      "grad_norm": 0.3527565896511078,
      "learning_rate": 9.652450445613913e-06,
      "loss": 0.5836,
      "step": 730
    },
    {
      "epoch": 0.6688887633467036,
      "grad_norm": 0.34412819147109985,
      "learning_rate": 9.64048881133784e-06,
      "loss": 0.5672,
      "step": 740
    },
    {
      "epoch": 0.6779278006892266,
      "grad_norm": 0.3731626868247986,
      "learning_rate": 9.628332471445206e-06,
      "loss": 0.6236,
      "step": 750
    },
    {
      "epoch": 0.6869668380317496,
      "grad_norm": 0.35358232259750366,
      "learning_rate": 9.615981935991683e-06,
      "loss": 0.5535,
      "step": 760
    },
    {
      "epoch": 0.6960058753742726,
      "grad_norm": 0.3264493942260742,
      "learning_rate": 9.603437723181002e-06,
      "loss": 0.6,
      "step": 770
    },
    {
      "epoch": 0.7050449127167957,
      "grad_norm": 0.3995952308177948,
      "learning_rate": 9.59070035934321e-06,
      "loss": 0.5739,
      "step": 780
    },
    {
      "epoch": 0.7140839500593187,
      "grad_norm": 0.4088019132614136,
      "learning_rate": 9.577770378912584e-06,
      "loss": 0.5463,
      "step": 790
    },
    {
      "epoch": 0.7231229874018417,
      "grad_norm": 0.36028265953063965,
      "learning_rate": 9.564648324405206e-06,
      "loss": 0.6317,
      "step": 800
    },
    {
      "epoch": 0.7321620247443648,
      "grad_norm": 0.36535605788230896,
      "learning_rate": 9.551334746396203e-06,
      "loss": 0.5675,
      "step": 810
    },
    {
      "epoch": 0.7412010620868877,
      "grad_norm": 0.339760959148407,
      "learning_rate": 9.537830203496642e-06,
      "loss": 0.5587,
      "step": 820
    },
    {
      "epoch": 0.7502400994294107,
      "grad_norm": 0.36886465549468994,
      "learning_rate": 9.524135262330098e-06,
      "loss": 0.5954,
      "step": 830
    },
    {
      "epoch": 0.7592791367719338,
      "grad_norm": 0.36694473028182983,
      "learning_rate": 9.510250497508873e-06,
      "loss": 0.5626,
      "step": 840
    },
    {
      "epoch": 0.7683181741144568,
      "grad_norm": 0.3380376100540161,
      "learning_rate": 9.496176491609893e-06,
      "loss": 0.6203,
      "step": 850
    },
    {
      "epoch": 0.7773572114569798,
      "grad_norm": 0.33642667531967163,
      "learning_rate": 9.481913835150256e-06,
      "loss": 0.5856,
      "step": 860
    },
    {
      "epoch": 0.7863962487995029,
      "grad_norm": 0.43253782391548157,
      "learning_rate": 9.467463126562461e-06,
      "loss": 0.586,
      "step": 870
    },
    {
      "epoch": 0.7954352861420259,
      "grad_norm": 0.35205066204071045,
      "learning_rate": 9.452824972169298e-06,
      "loss": 0.581,
      "step": 880
    },
    {
      "epoch": 0.804474323484549,
      "grad_norm": 0.33954545855522156,
      "learning_rate": 9.437999986158413e-06,
      "loss": 0.5961,
      "step": 890
    },
    {
      "epoch": 0.8135133608270719,
      "grad_norm": 0.29890018701553345,
      "learning_rate": 9.422988790556524e-06,
      "loss": 0.6118,
      "step": 900
    },
    {
      "epoch": 0.8225523981695949,
      "grad_norm": 0.3720478415489197,
      "learning_rate": 9.407792015203331e-06,
      "loss": 0.5773,
      "step": 910
    },
    {
      "epoch": 0.831591435512118,
      "grad_norm": 0.48966363072395325,
      "learning_rate": 9.392410297725099e-06,
      "loss": 0.5739,
      "step": 920
    },
    {
      "epoch": 0.840630472854641,
      "grad_norm": 0.33022603392601013,
      "learning_rate": 9.37684428350788e-06,
      "loss": 0.5572,
      "step": 930
    },
    {
      "epoch": 0.849669510197164,
      "grad_norm": 0.3322959840297699,
      "learning_rate": 9.36109462567046e-06,
      "loss": 0.5884,
      "step": 940
    },
    {
      "epoch": 0.8587085475396871,
      "grad_norm": 0.38541945815086365,
      "learning_rate": 9.345161985036937e-06,
      "loss": 0.5485,
      "step": 950
    },
    {
      "epoch": 0.86774758488221,
      "grad_norm": 0.3497406244277954,
      "learning_rate": 9.329047030109e-06,
      "loss": 0.5774,
      "step": 960
    },
    {
      "epoch": 0.876786622224733,
      "grad_norm": 0.3379673361778259,
      "learning_rate": 9.312750437037886e-06,
      "loss": 0.5682,
      "step": 970
    },
    {
      "epoch": 0.8858256595672561,
      "grad_norm": 0.3425869047641754,
      "learning_rate": 9.296272889595997e-06,
      "loss": 0.597,
      "step": 980
    },
    {
      "epoch": 0.8948646969097791,
      "grad_norm": 0.3111202120780945,
      "learning_rate": 9.279615079148227e-06,
      "loss": 0.5629,
      "step": 990
    },
    {
      "epoch": 0.9039037342523021,
      "grad_norm": 0.36501139402389526,
      "learning_rate": 9.262777704622939e-06,
      "loss": 0.5763,
      "step": 1000
    },
    {
      "epoch": 0.9129427715948252,
      "grad_norm": 0.4016779661178589,
      "learning_rate": 9.245761472482646e-06,
      "loss": 0.6107,
      "step": 1010
    },
    {
      "epoch": 0.9219818089373482,
      "grad_norm": 0.3700155019760132,
      "learning_rate": 9.228567096694366e-06,
      "loss": 0.5753,
      "step": 1020
    },
    {
      "epoch": 0.9310208462798711,
      "grad_norm": 0.32044124603271484,
      "learning_rate": 9.211195298699674e-06,
      "loss": 0.5664,
      "step": 1030
    },
    {
      "epoch": 0.9400598836223942,
      "grad_norm": 0.3247808516025543,
      "learning_rate": 9.193646807384421e-06,
      "loss": 0.5559,
      "step": 1040
    },
    {
      "epoch": 0.9490989209649172,
      "grad_norm": 0.37978535890579224,
      "learning_rate": 9.175922359048156e-06,
      "loss": 0.6026,
      "step": 1050
    },
    {
      "epoch": 0.9581379583074403,
      "grad_norm": 0.37707069516181946,
      "learning_rate": 9.15802269737323e-06,
      "loss": 0.5683,
      "step": 1060
    },
    {
      "epoch": 0.9671769956499633,
      "grad_norm": 0.37162524461746216,
      "learning_rate": 9.139948573393602e-06,
      "loss": 0.622,
      "step": 1070
    },
    {
      "epoch": 0.9762160329924863,
      "grad_norm": 0.35211777687072754,
      "learning_rate": 9.121700745463312e-06,
      "loss": 0.5754,
      "step": 1080
    },
    {
      "epoch": 0.9852550703350094,
      "grad_norm": 0.35246849060058594,
      "learning_rate": 9.103279979224676e-06,
      "loss": 0.6038,
      "step": 1090
    },
    {
      "epoch": 0.9942941076775323,
      "grad_norm": 0.30493950843811035,
      "learning_rate": 9.084687047576149e-06,
      "loss": 0.5873,
      "step": 1100
    },
    {
      "epoch": 1.0027117112027568,
      "grad_norm": 0.37721627950668335,
      "learning_rate": 9.065922730639906e-06,
      "loss": 0.5726,
      "step": 1110
    },
    {
      "epoch": 1.01175074854528,
      "grad_norm": 0.4114892780780792,
      "learning_rate": 9.046987815729108e-06,
      "loss": 0.547,
      "step": 1120
    },
    {
      "epoch": 1.020789785887803,
      "grad_norm": 0.3268738090991974,
      "learning_rate": 9.027883097314859e-06,
      "loss": 0.5896,
      "step": 1130
    },
    {
      "epoch": 1.0298288232303259,
      "grad_norm": 0.38347479701042175,
      "learning_rate": 9.008609376992879e-06,
      "loss": 0.5496,
      "step": 1140
    },
    {
      "epoch": 1.038867860572849,
      "grad_norm": 0.4380169212818146,
      "learning_rate": 8.989167463449874e-06,
      "loss": 0.5468,
      "step": 1150
    },
    {
      "epoch": 1.047906897915372,
      "grad_norm": 0.3695785105228424,
      "learning_rate": 8.9695581724296e-06,
      "loss": 0.5636,
      "step": 1160
    },
    {
      "epoch": 1.056945935257895,
      "grad_norm": 0.31363677978515625,
      "learning_rate": 8.94978232669863e-06,
      "loss": 0.5774,
      "step": 1170
    },
    {
      "epoch": 1.065984972600418,
      "grad_norm": 0.32090502977371216,
      "learning_rate": 8.929840756011847e-06,
      "loss": 0.5819,
      "step": 1180
    },
    {
      "epoch": 1.0750240099429411,
      "grad_norm": 0.2986926734447479,
      "learning_rate": 8.909734297077618e-06,
      "loss": 0.5565,
      "step": 1190
    },
    {
      "epoch": 1.084063047285464,
      "grad_norm": 0.3923097252845764,
      "learning_rate": 8.889463793522687e-06,
      "loss": 0.5971,
      "step": 1200
    },
    {
      "epoch": 1.093102084627987,
      "grad_norm": 0.45070353150367737,
      "learning_rate": 8.86903009585679e-06,
      "loss": 0.5481,
      "step": 1210
    },
    {
      "epoch": 1.1021411219705102,
      "grad_norm": 0.42040202021598816,
      "learning_rate": 8.848434061436954e-06,
      "loss": 0.5468,
      "step": 1220
    },
    {
      "epoch": 1.111180159313033,
      "grad_norm": 0.45057106018066406,
      "learning_rate": 8.827676554431534e-06,
      "loss": 0.6003,
      "step": 1230
    },
    {
      "epoch": 1.1202191966555561,
      "grad_norm": 0.34813055396080017,
      "learning_rate": 8.806758445783954e-06,
      "loss": 0.5673,
      "step": 1240
    },
    {
      "epoch": 1.1292582339980792,
      "grad_norm": 0.39887797832489014,
      "learning_rate": 8.785680613176153e-06,
      "loss": 0.5783,
      "step": 1250
    },
    {
      "epoch": 1.138297271340602,
      "grad_norm": 0.3870787024497986,
      "learning_rate": 8.764443940991776e-06,
      "loss": 0.5533,
      "step": 1260
    },
    {
      "epoch": 1.1473363086831252,
      "grad_norm": 0.355211466550827,
      "learning_rate": 8.743049320279053e-06,
      "loss": 0.605,
      "step": 1270
    },
    {
      "epoch": 1.1563753460256483,
      "grad_norm": 0.335464209318161,
      "learning_rate": 8.721497648713423e-06,
      "loss": 0.5681,
      "step": 1280
    },
    {
      "epoch": 1.1654143833681714,
      "grad_norm": 0.411471962928772,
      "learning_rate": 8.699789830559856e-06,
      "loss": 0.5502,
      "step": 1290
    },
    {
      "epoch": 1.1744534207106943,
      "grad_norm": 0.2810121178627014,
      "learning_rate": 8.677926776634933e-06,
      "loss": 0.5689,
      "step": 1300
    },
    {
      "epoch": 1.1834924580532173,
      "grad_norm": 0.44342437386512756,
      "learning_rate": 8.655909404268603e-06,
      "loss": 0.5694,
      "step": 1310
    },
    {
      "epoch": 1.1925314953957404,
      "grad_norm": 0.40787559747695923,
      "learning_rate": 8.633738637265716e-06,
      "loss": 0.6138,
      "step": 1320
    },
    {
      "epoch": 1.2015705327382633,
      "grad_norm": 0.40878427028656006,
      "learning_rate": 8.611415405867251e-06,
      "loss": 0.5663,
      "step": 1330
    },
    {
      "epoch": 1.2106095700807864,
      "grad_norm": 0.4044298231601715,
      "learning_rate": 8.588940646711293e-06,
      "loss": 0.5703,
      "step": 1340
    },
    {
      "epoch": 1.2196486074233095,
      "grad_norm": 0.38636693358421326,
      "learning_rate": 8.566315302793717e-06,
      "loss": 0.5878,
      "step": 1350
    },
    {
      "epoch": 1.2286876447658324,
      "grad_norm": 0.4851875901222229,
      "learning_rate": 8.543540323428646e-06,
      "loss": 0.5393,
      "step": 1360
    },
    {
      "epoch": 1.2377266821083555,
      "grad_norm": 0.3599594235420227,
      "learning_rate": 8.520616664208597e-06,
      "loss": 0.5637,
      "step": 1370
    },
    {
      "epoch": 1.2467657194508786,
      "grad_norm": 0.393518328666687,
      "learning_rate": 8.4975452869644e-06,
      "loss": 0.5846,
      "step": 1380
    },
    {
      "epoch": 1.2558047567934014,
      "grad_norm": 0.4419288635253906,
      "learning_rate": 8.474327159724837e-06,
      "loss": 0.5647,
      "step": 1390
    },
    {
      "epoch": 1.2648437941359245,
      "grad_norm": 0.46883851289749146,
      "learning_rate": 8.450963256676026e-06,
      "loss": 0.5978,
      "step": 1400
    },
    {
      "epoch": 1.2738828314784476,
      "grad_norm": 0.3780512511730194,
      "learning_rate": 8.427454558120545e-06,
      "loss": 0.5254,
      "step": 1410
    },
    {
      "epoch": 1.2829218688209707,
      "grad_norm": 0.39592650532722473,
      "learning_rate": 8.403802050436302e-06,
      "loss": 0.6072,
      "step": 1420
    },
    {
      "epoch": 1.2919609061634936,
      "grad_norm": 0.34795117378234863,
      "learning_rate": 8.380006726035152e-06,
      "loss": 0.5118,
      "step": 1430
    },
    {
      "epoch": 1.3009999435060167,
      "grad_norm": 0.3898769021034241,
      "learning_rate": 8.356069583321245e-06,
      "loss": 0.5666,
      "step": 1440
    },
    {
      "epoch": 1.3100389808485398,
      "grad_norm": 0.4135303199291229,
      "learning_rate": 8.33199162664915e-06,
      "loss": 0.5661,
      "step": 1450
    },
    {
      "epoch": 1.3190780181910626,
      "grad_norm": 0.4047456979751587,
      "learning_rate": 8.307773866281705e-06,
      "loss": 0.5509,
      "step": 1460
    },
    {
      "epoch": 1.3281170555335857,
      "grad_norm": 0.3716682493686676,
      "learning_rate": 8.283417318347632e-06,
      "loss": 0.5456,
      "step": 1470
    },
    {
      "epoch": 1.3371560928761088,
      "grad_norm": 0.3757040500640869,
      "learning_rate": 8.258923004798902e-06,
      "loss": 0.5614,
      "step": 1480
    },
    {
      "epoch": 1.3461951302186317,
      "grad_norm": 0.3806832432746887,
      "learning_rate": 8.234291953367855e-06,
      "loss": 0.5157,
      "step": 1490
    },
    {
      "epoch": 1.3552341675611548,
      "grad_norm": 0.34731006622314453,
      "learning_rate": 8.209525197524073e-06,
      "loss": 0.5892,
      "step": 1500
    },
    {
      "epoch": 1.3642732049036779,
      "grad_norm": 0.38066595792770386,
      "learning_rate": 8.184623776431036e-06,
      "loss": 0.5416,
      "step": 1510
    },
    {
      "epoch": 1.3733122422462007,
      "grad_norm": 0.3750848174095154,
      "learning_rate": 8.159588734902495e-06,
      "loss": 0.5464,
      "step": 1520
    },
    {
      "epoch": 1.3823512795887238,
      "grad_norm": 0.42894601821899414,
      "learning_rate": 8.134421123358653e-06,
      "loss": 0.5601,
      "step": 1530
    },
    {
      "epoch": 1.391390316931247,
      "grad_norm": 0.40111929178237915,
      "learning_rate": 8.109121997782088e-06,
      "loss": 0.5573,
      "step": 1540
    },
    {
      "epoch": 1.4004293542737698,
      "grad_norm": 0.4170910120010376,
      "learning_rate": 8.083692419673436e-06,
      "loss": 0.53,
      "step": 1550
    },
    {
      "epoch": 1.409468391616293,
      "grad_norm": 0.37612515687942505,
      "learning_rate": 8.058133456006867e-06,
      "loss": 0.4952,
      "step": 1560
    },
    {
      "epoch": 1.418507428958816,
      "grad_norm": 0.42817550897598267,
      "learning_rate": 8.032446179185304e-06,
      "loss": 0.5558,
      "step": 1570
    },
    {
      "epoch": 1.4275464663013389,
      "grad_norm": 0.39047130942344666,
      "learning_rate": 8.00663166699544e-06,
      "loss": 0.5615,
      "step": 1580
    },
    {
      "epoch": 1.436585503643862,
      "grad_norm": 0.4020339846611023,
      "learning_rate": 7.980691002562505e-06,
      "loss": 0.5043,
      "step": 1590
    },
    {
      "epoch": 1.445624540986385,
      "grad_norm": 0.3834945559501648,
      "learning_rate": 7.954625274304822e-06,
      "loss": 0.5715,
      "step": 1600
    },
    {
      "epoch": 1.454663578328908,
      "grad_norm": 0.4076308310031891,
      "learning_rate": 7.928435575888146e-06,
      "loss": 0.5898,
      "step": 1610
    },
    {
      "epoch": 1.463702615671431,
      "grad_norm": 0.3430909514427185,
      "learning_rate": 7.902123006179767e-06,
      "loss": 0.5598,
      "step": 1620
    },
    {
      "epoch": 1.472741653013954,
      "grad_norm": 0.4031376242637634,
      "learning_rate": 7.875688669202416e-06,
      "loss": 0.5884,
      "step": 1630
    },
    {
      "epoch": 1.481780690356477,
      "grad_norm": 0.44467347860336304,
      "learning_rate": 7.849133674087926e-06,
      "loss": 0.5741,
      "step": 1640
    },
    {
      "epoch": 1.490819727699,
      "grad_norm": 0.37968936562538147,
      "learning_rate": 7.822459135030709e-06,
      "loss": 0.5536,
      "step": 1650
    },
    {
      "epoch": 1.4998587650415232,
      "grad_norm": 0.5008074045181274,
      "learning_rate": 7.795666171241001e-06,
      "loss": 0.5338,
      "step": 1660
    },
    {
      "epoch": 1.508897802384046,
      "grad_norm": 0.37119391560554504,
      "learning_rate": 7.7687559068979e-06,
      "loss": 0.5386,
      "step": 1670
    },
    {
      "epoch": 1.5179368397265691,
      "grad_norm": 0.4206055998802185,
      "learning_rate": 7.741729471102204e-06,
      "loss": 0.54,
      "step": 1680
    },
    {
      "epoch": 1.5269758770690922,
      "grad_norm": 0.4098663330078125,
      "learning_rate": 7.71458799782903e-06,
      "loss": 0.5443,
      "step": 1690
    },
    {
      "epoch": 1.536014914411615,
      "grad_norm": 0.41614359617233276,
      "learning_rate": 7.687332625880237e-06,
      "loss": 0.5714,
      "step": 1700
    },
    {
      "epoch": 1.5450539517541382,
      "grad_norm": 0.429938942193985,
      "learning_rate": 7.659964498836645e-06,
      "loss": 0.5209,
      "step": 1710
    },
    {
      "epoch": 1.5540929890966613,
      "grad_norm": 0.3705986440181732,
      "learning_rate": 7.632484765010052e-06,
      "loss": 0.5164,
      "step": 1720
    },
    {
      "epoch": 1.5631320264391841,
      "grad_norm": 0.3999808728694916,
      "learning_rate": 7.6048945773950536e-06,
      "loss": 0.5602,
      "step": 1730
    },
    {
      "epoch": 1.5721710637817072,
      "grad_norm": 0.4704653322696686,
      "learning_rate": 7.577195093620664e-06,
      "loss": 0.5639,
      "step": 1740
    },
    {
      "epoch": 1.5812101011242303,
      "grad_norm": 0.40082478523254395,
      "learning_rate": 7.549387475901746e-06,
      "loss": 0.5766,
      "step": 1750
    },
    {
      "epoch": 1.5902491384667532,
      "grad_norm": 0.4126080274581909,
      "learning_rate": 7.521472890990247e-06,
      "loss": 0.5779,
      "step": 1760
    },
    {
      "epoch": 1.5992881758092763,
      "grad_norm": 0.39301300048828125,
      "learning_rate": 7.493452510126241e-06,
      "loss": 0.5263,
      "step": 1770
    },
    {
      "epoch": 1.6083272131517994,
      "grad_norm": 0.4135958254337311,
      "learning_rate": 7.465327508988791e-06,
      "loss": 0.5831,
      "step": 1780
    },
    {
      "epoch": 1.6173662504943223,
      "grad_norm": 0.4341680109500885,
      "learning_rate": 7.437099067646616e-06,
      "loss": 0.5506,
      "step": 1790
    },
    {
      "epoch": 1.6264052878368453,
      "grad_norm": 0.43189555406570435,
      "learning_rate": 7.408768370508577e-06,
      "loss": 0.5403,
      "step": 1800
    },
    {
      "epoch": 1.6354443251793684,
      "grad_norm": 0.3924044966697693,
      "learning_rate": 7.3803366062739875e-06,
      "loss": 0.5378,
      "step": 1810
    },
    {
      "epoch": 1.6444833625218913,
      "grad_norm": 0.44657233357429504,
      "learning_rate": 7.3518049678827316e-06,
      "loss": 0.5927,
      "step": 1820
    },
    {
      "epoch": 1.6535223998644144,
      "grad_norm": 0.3682084083557129,
      "learning_rate": 7.323174652465213e-06,
      "loss": 0.5395,
      "step": 1830
    },
    {
      "epoch": 1.6625614372069375,
      "grad_norm": 0.4883032739162445,
      "learning_rate": 7.2944468612921235e-06,
      "loss": 0.565,
      "step": 1840
    },
    {
      "epoch": 1.6716004745494604,
      "grad_norm": 0.47772058844566345,
      "learning_rate": 7.2656227997240494e-06,
      "loss": 0.5589,
      "step": 1850
    },
    {
      "epoch": 1.6806395118919835,
      "grad_norm": 0.419160932302475,
      "learning_rate": 7.236703677160881e-06,
      "loss": 0.5703,
      "step": 1860
    },
    {
      "epoch": 1.6896785492345066,
      "grad_norm": 0.4483725428581238,
      "learning_rate": 7.207690706991084e-06,
      "loss": 0.5528,
      "step": 1870
    },
    {
      "epoch": 1.6987175865770294,
      "grad_norm": 0.36597567796707153,
      "learning_rate": 7.178585106540778e-06,
      "loss": 0.5522,
      "step": 1880
    },
    {
      "epoch": 1.7077566239195525,
      "grad_norm": 0.39857447147369385,
      "learning_rate": 7.149388097022668e-06,
      "loss": 0.5597,
      "step": 1890
    },
    {
      "epoch": 1.7167956612620756,
      "grad_norm": 0.4677294194698334,
      "learning_rate": 7.120100903484797e-06,
      "loss": 0.5495,
      "step": 1900
    },
    {
      "epoch": 1.7258346986045985,
      "grad_norm": 0.3514731824398041,
      "learning_rate": 7.090724754759146e-06,
      "loss": 0.5952,
      "step": 1910
    },
    {
      "epoch": 1.7348737359471218,
      "grad_norm": 0.44050273299217224,
      "learning_rate": 7.061260883410086e-06,
      "loss": 0.5154,
      "step": 1920
    },
    {
      "epoch": 1.7439127732896447,
      "grad_norm": 0.4586310386657715,
      "learning_rate": 7.031710525682645e-06,
      "loss": 0.5421,
      "step": 1930
    },
    {
      "epoch": 1.7529518106321675,
      "grad_norm": 0.4656844139099121,
      "learning_rate": 7.002074921450652e-06,
      "loss": 0.5591,
      "step": 1940
    },
    {
      "epoch": 1.7619908479746909,
      "grad_norm": 0.4101289212703705,
      "learning_rate": 6.9723553141647045e-06,
      "loss": 0.5589,
      "step": 1950
    },
    {
      "epoch": 1.7710298853172137,
      "grad_norm": 0.4236315190792084,
      "learning_rate": 6.942552950800003e-06,
      "loss": 0.5853,
      "step": 1960
    },
    {
      "epoch": 1.7800689226597366,
      "grad_norm": 0.3590008318424225,
      "learning_rate": 6.912669081804027e-06,
      "loss": 0.5627,
      "step": 1970
    },
    {
      "epoch": 1.78910796000226,
      "grad_norm": 0.3790905177593231,
      "learning_rate": 6.8827049610440685e-06,
      "loss": 0.555,
      "step": 1980
    },
    {
      "epoch": 1.7981469973447828,
      "grad_norm": 0.436749130487442,
      "learning_rate": 6.85266184575462e-06,
      "loss": 0.5676,
      "step": 1990
    },
    {
      "epoch": 1.8071860346873057,
      "grad_norm": 0.455223023891449,
      "learning_rate": 6.822540996484631e-06,
      "loss": 0.5975,
      "step": 2000
    }
  ],
  "logging_steps": 10,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 6.328532049660088e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}