{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9039037342523021,
  "eval_steps": 500,
  "global_step": 1000,
  "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
    }
  ],
  "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": 3.1716984075329126e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}