{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.5030674846625764,
  "eval_steps": 500,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.035056967572304996,
      "grad_norm": 12.451324462890625,
      "learning_rate": 1.8446601941747574e-06,
      "loss": 3.237,
      "step": 20
    },
    {
      "epoch": 0.07011393514460999,
      "grad_norm": 4.131791114807129,
      "learning_rate": 3.7864077669902915e-06,
      "loss": 2.5266,
      "step": 40
    },
    {
      "epoch": 0.10517090271691498,
      "grad_norm": 2.2451958656311035,
      "learning_rate": 5.728155339805825e-06,
      "loss": 1.8755,
      "step": 60
    },
    {
      "epoch": 0.14022787028921999,
      "grad_norm": 1.4421552419662476,
      "learning_rate": 7.66990291262136e-06,
      "loss": 1.3649,
      "step": 80
    },
    {
      "epoch": 0.175284837861525,
      "grad_norm": 0.9530179500579834,
      "learning_rate": 9.611650485436894e-06,
      "loss": 1.0674,
      "step": 100
    },
    {
      "epoch": 0.21034180543382996,
      "grad_norm": 0.7194477319717407,
      "learning_rate": 9.99942798060303e-06,
      "loss": 0.9241,
      "step": 120
    },
    {
      "epoch": 0.24539877300613497,
      "grad_norm": 0.6556061506271362,
      "learning_rate": 9.997104376116195e-06,
      "loss": 0.8575,
      "step": 140
    },
    {
      "epoch": 0.28045574057843997,
      "grad_norm": 0.5718048810958862,
      "learning_rate": 9.992994265395959e-06,
      "loss": 0.829,
      "step": 160
    },
    {
      "epoch": 0.31551270815074495,
      "grad_norm": 0.4922148287296295,
      "learning_rate": 9.987099117840969e-06,
      "loss": 0.8034,
      "step": 180
    },
    {
      "epoch": 0.35056967572305,
      "grad_norm": 0.47302234172821045,
      "learning_rate": 9.979421041015336e-06,
      "loss": 0.7839,
      "step": 200
    },
    {
      "epoch": 0.38562664329535495,
      "grad_norm": 0.49009189009666443,
      "learning_rate": 9.969962779895172e-06,
      "loss": 0.768,
      "step": 220
    },
    {
      "epoch": 0.42068361086765993,
      "grad_norm": 0.4963654577732086,
      "learning_rate": 9.958727715887218e-06,
      "loss": 0.7628,
      "step": 240
    },
    {
      "epoch": 0.45574057843996496,
      "grad_norm": 0.5206854343414307,
      "learning_rate": 9.94571986561998e-06,
      "loss": 0.7488,
      "step": 260
    },
    {
      "epoch": 0.49079754601226994,
      "grad_norm": 0.48924869298934937,
      "learning_rate": 9.930943879507748e-06,
      "loss": 0.7436,
      "step": 280
    },
    {
      "epoch": 0.5258545135845749,
      "grad_norm": 0.43540337681770325,
      "learning_rate": 9.914405040088026e-06,
      "loss": 0.7375,
      "step": 300
    },
    {
      "epoch": 0.5609114811568799,
      "grad_norm": 0.44258421659469604,
      "learning_rate": 9.896109260132993e-06,
      "loss": 0.7277,
      "step": 320
    },
    {
      "epoch": 0.595968448729185,
      "grad_norm": 0.4955386519432068,
      "learning_rate": 9.876063080535627e-06,
      "loss": 0.7284,
      "step": 340
    },
    {
      "epoch": 0.6310254163014899,
      "grad_norm": 0.5027541518211365,
      "learning_rate": 9.85427366797129e-06,
      "loss": 0.7231,
      "step": 360
    },
    {
      "epoch": 0.6660823838737949,
      "grad_norm": 0.4675957262516022,
      "learning_rate": 9.830748812335576e-06,
      "loss": 0.7212,
      "step": 380
    },
    {
      "epoch": 0.7011393514461,
      "grad_norm": 0.4283595383167267,
      "learning_rate": 9.805496923959363e-06,
      "loss": 0.7164,
      "step": 400
    },
    {
      "epoch": 0.7361963190184049,
      "grad_norm": 0.452084481716156,
      "learning_rate": 9.778527030602049e-06,
      "loss": 0.711,
      "step": 420
    },
    {
      "epoch": 0.7712532865907099,
      "grad_norm": 0.4737929105758667,
      "learning_rate": 9.74984877422405e-06,
      "loss": 0.7084,
      "step": 440
    },
    {
      "epoch": 0.8063102541630149,
      "grad_norm": 0.4964485466480255,
      "learning_rate": 9.719472407539725e-06,
      "loss": 0.7028,
      "step": 460
    },
    {
      "epoch": 0.8413672217353199,
      "grad_norm": 0.44363030791282654,
      "learning_rate": 9.68740879035194e-06,
      "loss": 0.7045,
      "step": 480
    },
    {
      "epoch": 0.8764241893076249,
      "grad_norm": 0.5004998445510864,
      "learning_rate": 9.6536693856696e-06,
      "loss": 0.6937,
      "step": 500
    },
    {
      "epoch": 0.9114811568799299,
      "grad_norm": 0.4564264118671417,
      "learning_rate": 9.618266255609533e-06,
      "loss": 0.699,
      "step": 520
    },
    {
      "epoch": 0.9465381244522348,
      "grad_norm": 0.4558616280555725,
      "learning_rate": 9.58121205708418e-06,
      "loss": 0.691,
      "step": 540
    },
    {
      "epoch": 0.9815950920245399,
      "grad_norm": 0.413114458322525,
      "learning_rate": 9.542520037276636e-06,
      "loss": 0.6891,
      "step": 560
    },
    {
      "epoch": 1.0157756354075373,
      "grad_norm": 0.403679758310318,
      "learning_rate": 9.502204028904687e-06,
      "loss": 0.6812,
      "step": 580
    },
    {
      "epoch": 1.0508326029798423,
      "grad_norm": 0.40308722853660583,
      "learning_rate": 9.46027844527549e-06,
      "loss": 0.6791,
      "step": 600
    },
    {
      "epoch": 1.0858895705521472,
      "grad_norm": 0.4085083603858948,
      "learning_rate": 9.416758275132693e-06,
      "loss": 0.6803,
      "step": 620
    },
    {
      "epoch": 1.1209465381244523,
      "grad_norm": 0.4475920796394348,
      "learning_rate": 9.371659077297843e-06,
      "loss": 0.6789,
      "step": 640
    },
    {
      "epoch": 1.1560035056967572,
      "grad_norm": 0.4604188799858093,
      "learning_rate": 9.324996975107978e-06,
      "loss": 0.674,
      "step": 660
    },
    {
      "epoch": 1.1910604732690622,
      "grad_norm": 0.4190482795238495,
      "learning_rate": 9.276788650651392e-06,
      "loss": 0.6746,
      "step": 680
    },
    {
      "epoch": 1.2261174408413673,
      "grad_norm": 0.420953631401062,
      "learning_rate": 9.227051338803656e-06,
      "loss": 0.6692,
      "step": 700
    },
    {
      "epoch": 1.2611744084136722,
      "grad_norm": 0.4463854432106018,
      "learning_rate": 9.175802821066009e-06,
      "loss": 0.6737,
      "step": 720
    },
    {
      "epoch": 1.2962313759859772,
      "grad_norm": 0.44004735350608826,
      "learning_rate": 9.12306141920832e-06,
      "loss": 0.6673,
      "step": 740
    },
    {
      "epoch": 1.331288343558282,
      "grad_norm": 0.42015475034713745,
      "learning_rate": 9.068845988718906e-06,
      "loss": 0.6676,
      "step": 760
    },
    {
      "epoch": 1.3663453111305872,
      "grad_norm": 0.43683475255966187,
      "learning_rate": 9.013175912063534e-06,
      "loss": 0.6649,
      "step": 780
    },
    {
      "epoch": 1.4014022787028921,
      "grad_norm": 0.4281805753707886,
      "learning_rate": 8.956071091756036e-06,
      "loss": 0.6658,
      "step": 800
    },
    {
      "epoch": 1.4364592462751973,
      "grad_norm": 0.4270734190940857,
      "learning_rate": 8.89755194324299e-06,
      "loss": 0.6646,
      "step": 820
    },
    {
      "epoch": 1.4715162138475022,
      "grad_norm": 0.4163481593132019,
      "learning_rate": 8.837639387605031e-06,
      "loss": 0.6658,
      "step": 840
    },
    {
      "epoch": 1.5065731814198071,
      "grad_norm": 0.45280900597572327,
      "learning_rate": 8.776354844077389e-06,
      "loss": 0.6592,
      "step": 860
    },
    {
      "epoch": 1.541630148992112,
      "grad_norm": 0.40485361218452454,
      "learning_rate": 8.713720222392338e-06,
      "loss": 0.6579,
      "step": 880
    },
    {
      "epoch": 1.5766871165644172,
      "grad_norm": 0.42039763927459717,
      "learning_rate": 8.649757914946284e-06,
      "loss": 0.6616,
      "step": 900
    },
    {
      "epoch": 1.6117440841367223,
      "grad_norm": 0.4760454595088959,
      "learning_rate": 8.584490788794296e-06,
      "loss": 0.6572,
      "step": 920
    },
    {
      "epoch": 1.6468010517090272,
      "grad_norm": 0.43802690505981445,
      "learning_rate": 8.517942177474943e-06,
      "loss": 0.6548,
      "step": 940
    },
    {
      "epoch": 1.6818580192813322,
      "grad_norm": 0.5002708435058594,
      "learning_rate": 8.450135872668369e-06,
      "loss": 0.6557,
      "step": 960
    },
    {
      "epoch": 1.716914986853637,
      "grad_norm": 0.4160609543323517,
      "learning_rate": 8.38109611569056e-06,
      "loss": 0.6529,
      "step": 980
    },
    {
      "epoch": 1.751971954425942,
      "grad_norm": 0.43179649114608765,
      "learning_rate": 8.310847588826876e-06,
      "loss": 0.6529,
      "step": 1000
    },
    {
      "epoch": 1.7870289219982471,
      "grad_norm": 0.4322780668735504,
      "learning_rate": 8.239415406507934e-06,
      "loss": 0.6535,
      "step": 1020
    },
    {
      "epoch": 1.8220858895705523,
      "grad_norm": 0.4642186462879181,
      "learning_rate": 8.166825106330985e-06,
      "loss": 0.649,
      "step": 1040
    },
    {
      "epoch": 1.8571428571428572,
      "grad_norm": 0.42697349190711975,
      "learning_rate": 8.093102639930013e-06,
      "loss": 0.65,
      "step": 1060
    },
    {
      "epoch": 1.8921998247151621,
      "grad_norm": 0.4486387372016907,
      "learning_rate": 8.01827436369781e-06,
      "loss": 0.6492,
      "step": 1080
    },
    {
      "epoch": 1.927256792287467,
      "grad_norm": 0.42962825298309326,
      "learning_rate": 7.942367029363351e-06,
      "loss": 0.6518,
      "step": 1100
    },
    {
      "epoch": 1.962313759859772,
      "grad_norm": 0.43645408749580383,
      "learning_rate": 7.865407774427828e-06,
      "loss": 0.6475,
      "step": 1120
    },
    {
      "epoch": 1.997370727432077,
      "grad_norm": 0.4662039875984192,
      "learning_rate": 7.787424112462758e-06,
      "loss": 0.649,
      "step": 1140
    },
    {
      "epoch": 2.0315512708150747,
      "grad_norm": 0.44152551889419556,
      "learning_rate": 7.708443923273671e-06,
      "loss": 0.6401,
      "step": 1160
    },
    {
      "epoch": 2.0666082383873796,
      "grad_norm": 0.39342495799064636,
      "learning_rate": 7.628495442932838e-06,
      "loss": 0.6386,
      "step": 1180
    },
    {
      "epoch": 2.1016652059596845,
      "grad_norm": 0.4356766641139984,
      "learning_rate": 7.54760725368464e-06,
      "loss": 0.6391,
      "step": 1200
    },
    {
      "epoch": 2.1367221735319895,
      "grad_norm": 0.3897708058357239,
      "learning_rate": 7.465808273727182e-06,
      "loss": 0.6383,
      "step": 1220
    },
    {
      "epoch": 2.1717791411042944,
      "grad_norm": 0.4168529510498047,
      "learning_rate": 7.383127746873796e-06,
      "loss": 0.6361,
      "step": 1240
    },
    {
      "epoch": 2.2068361086765993,
      "grad_norm": 0.39462465047836304,
      "learning_rate": 7.2995952320981356e-06,
      "loss": 0.6371,
      "step": 1260
    },
    {
      "epoch": 2.2418930762489047,
      "grad_norm": 0.42870041728019714,
      "learning_rate": 7.215240592966603e-06,
      "loss": 0.6318,
      "step": 1280
    },
    {
      "epoch": 2.2769500438212096,
      "grad_norm": 0.46848800778388977,
      "learning_rate": 7.130093986961868e-06,
      "loss": 0.633,
      "step": 1300
    },
    {
      "epoch": 2.3120070113935145,
      "grad_norm": 0.415912002325058,
      "learning_rate": 7.044185854701321e-06,
      "loss": 0.6367,
      "step": 1320
    },
    {
      "epoch": 2.3470639789658194,
      "grad_norm": 0.4347931444644928,
      "learning_rate": 6.957546909054304e-06,
      "loss": 0.6374,
      "step": 1340
    },
    {
      "epoch": 2.3821209465381243,
      "grad_norm": 0.4282444417476654,
      "learning_rate": 6.870208124161998e-06,
      "loss": 0.6353,
      "step": 1360
    },
    {
      "epoch": 2.4171779141104293,
      "grad_norm": 0.43224233388900757,
      "learning_rate": 6.78220072436392e-06,
      "loss": 0.6348,
      "step": 1380
    },
    {
      "epoch": 2.4522348816827346,
      "grad_norm": 0.4176190495491028,
      "learning_rate": 6.693556173034953e-06,
      "loss": 0.633,
      "step": 1400
    },
    {
      "epoch": 2.4872918492550395,
      "grad_norm": 0.4181615710258484,
      "learning_rate": 6.6043061613369356e-06,
      "loss": 0.6324,
      "step": 1420
    },
    {
      "epoch": 2.5223488168273445,
      "grad_norm": 0.38148173689842224,
      "learning_rate": 6.514482596888807e-06,
      "loss": 0.6301,
      "step": 1440
    },
    {
      "epoch": 2.5574057843996494,
      "grad_norm": 0.41031816601753235,
      "learning_rate": 6.424117592359367e-06,
      "loss": 0.6332,
      "step": 1460
    },
    {
      "epoch": 2.5924627519719543,
      "grad_norm": 0.4005562961101532,
      "learning_rate": 6.333243453986734e-06,
      "loss": 0.6265,
      "step": 1480
    },
    {
      "epoch": 2.6275197195442592,
      "grad_norm": 0.3940238058567047,
      "learning_rate": 6.241892670028595e-06,
      "loss": 0.6315,
      "step": 1500
    },
    {
      "epoch": 2.662576687116564,
      "grad_norm": 0.4001730680465698,
      "learning_rate": 6.150097899147384e-06,
      "loss": 0.6299,
      "step": 1520
    },
    {
      "epoch": 2.6976336546888695,
      "grad_norm": 0.3857872188091278,
      "learning_rate": 6.057891958734538e-06,
      "loss": 0.6304,
      "step": 1540
    },
    {
      "epoch": 2.7326906222611744,
      "grad_norm": 0.4169263243675232,
      "learning_rate": 5.965307813178015e-06,
      "loss": 0.6315,
      "step": 1560
    },
    {
      "epoch": 2.7677475898334793,
      "grad_norm": 0.4010975658893585,
      "learning_rate": 5.872378562077241e-06,
      "loss": 0.6297,
      "step": 1580
    },
    {
      "epoch": 2.8028045574057843,
      "grad_norm": 0.4302142262458801,
      "learning_rate": 5.779137428409738e-06,
      "loss": 0.6302,
      "step": 1600
    },
    {
      "epoch": 2.8378615249780896,
      "grad_norm": 0.39000585675239563,
      "learning_rate": 5.685617746653629e-06,
      "loss": 0.6312,
      "step": 1620
    },
    {
      "epoch": 2.8729184925503946,
      "grad_norm": 0.4292212724685669,
      "learning_rate": 5.591852950870287e-06,
      "loss": 0.6312,
      "step": 1640
    },
    {
      "epoch": 2.9079754601226995,
      "grad_norm": 0.3838886320590973,
      "learning_rate": 5.497876562751384e-06,
      "loss": 0.6302,
      "step": 1660
    },
    {
      "epoch": 2.9430324276950044,
      "grad_norm": 0.36835259199142456,
      "learning_rate": 5.403722179634602e-06,
      "loss": 0.6292,
      "step": 1680
    },
    {
      "epoch": 2.9780893952673093,
      "grad_norm": 0.3884848654270172,
      "learning_rate": 5.309423462492314e-06,
      "loss": 0.6261,
      "step": 1700
    },
    {
      "epoch": 3.0122699386503067,
      "grad_norm": 0.3762246072292328,
      "learning_rate": 5.215014123897504e-06,
      "loss": 0.6202,
      "step": 1720
    },
    {
      "epoch": 3.0473269062226116,
      "grad_norm": 0.38138872385025024,
      "learning_rate": 5.120527915971235e-06,
      "loss": 0.6205,
      "step": 1740
    },
    {
      "epoch": 3.0823838737949165,
      "grad_norm": 0.38698920607566833,
      "learning_rate": 5.0259986183160006e-06,
      "loss": 0.6186,
      "step": 1760
    },
    {
      "epoch": 3.117440841367222,
      "grad_norm": 0.378830224275589,
      "learning_rate": 4.931460025939226e-06,
      "loss": 0.6214,
      "step": 1780
    },
    {
      "epoch": 3.152497808939527,
      "grad_norm": 0.3751004934310913,
      "learning_rate": 4.836945937171279e-06,
      "loss": 0.6209,
      "step": 1800
    },
    {
      "epoch": 3.1875547765118317,
      "grad_norm": 0.3829745054244995,
      "learning_rate": 4.742490141582279e-06,
      "loss": 0.6213,
      "step": 1820
    },
    {
      "epoch": 3.2226117440841366,
      "grad_norm": 0.4228389859199524,
      "learning_rate": 4.648126407902058e-06,
      "loss": 0.6193,
      "step": 1840
    },
    {
      "epoch": 3.2576687116564416,
      "grad_norm": 0.367960125207901,
      "learning_rate": 4.553888471947546e-06,
      "loss": 0.6198,
      "step": 1860
    },
    {
      "epoch": 3.292725679228747,
      "grad_norm": 0.39815646409988403,
      "learning_rate": 4.4598100245619505e-06,
      "loss": 0.6187,
      "step": 1880
    },
    {
      "epoch": 3.327782646801052,
      "grad_norm": 0.3625248074531555,
      "learning_rate": 4.3659246995699845e-06,
      "loss": 0.6176,
      "step": 1900
    },
    {
      "epoch": 3.3628396143733568,
      "grad_norm": 0.37671083211898804,
      "learning_rate": 4.2722660617535105e-06,
      "loss": 0.6182,
      "step": 1920
    },
    {
      "epoch": 3.3978965819456617,
      "grad_norm": 0.3727245032787323,
      "learning_rate": 4.178867594851849e-06,
      "loss": 0.616,
      "step": 1940
    },
    {
      "epoch": 3.4329535495179666,
      "grad_norm": 0.361914724111557,
      "learning_rate": 4.085762689591054e-06,
      "loss": 0.6157,
      "step": 1960
    },
    {
      "epoch": 3.4680105170902715,
      "grad_norm": 0.3587988018989563,
      "learning_rate": 3.992984631746469e-06,
      "loss": 0.6188,
      "step": 1980
    },
    {
      "epoch": 3.5030674846625764,
      "grad_norm": 0.3729381561279297,
      "learning_rate": 3.9005665902427695e-06,
      "loss": 0.6208,
      "step": 2000
    }
  ],
  "logging_steps": 20,
  "max_steps": 3426,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.538206465283588e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}