{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.5023863350916855,
  "eval_steps": 500,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0025119316754584277,
      "grad_norm": 244.78928615182636,
      "learning_rate": 3.75e-07,
      "loss": 14.0363,
      "step": 10
    },
    {
      "epoch": 0.005023863350916855,
      "grad_norm": 318.1891497601652,
      "learning_rate": 7.916666666666667e-07,
      "loss": 13.4987,
      "step": 20
    },
    {
      "epoch": 0.007535795026375283,
      "grad_norm": 211.3992791294665,
      "learning_rate": 1.2083333333333333e-06,
      "loss": 11.4405,
      "step": 30
    },
    {
      "epoch": 0.01004772670183371,
      "grad_norm": 425.154938297453,
      "learning_rate": 1.6250000000000001e-06,
      "loss": 8.7926,
      "step": 40
    },
    {
      "epoch": 0.012559658377292138,
      "grad_norm": 136.69306564524402,
      "learning_rate": 2.041666666666667e-06,
      "loss": 7.1601,
      "step": 50
    },
    {
      "epoch": 0.015071590052750565,
      "grad_norm": 173.2049647337299,
      "learning_rate": 2.4583333333333332e-06,
      "loss": 6.2254,
      "step": 60
    },
    {
      "epoch": 0.017583521728208994,
      "grad_norm": 163.0490968387617,
      "learning_rate": 2.875e-06,
      "loss": 5.7153,
      "step": 70
    },
    {
      "epoch": 0.02009545340366742,
      "grad_norm": 70.98311735934806,
      "learning_rate": 3.2916666666666668e-06,
      "loss": 5.3196,
      "step": 80
    },
    {
      "epoch": 0.02260738507912585,
      "grad_norm": 101.87087924315038,
      "learning_rate": 3.708333333333334e-06,
      "loss": 5.1087,
      "step": 90
    },
    {
      "epoch": 0.025119316754584276,
      "grad_norm": 215.4312626681897,
      "learning_rate": 4.125e-06,
      "loss": 4.9029,
      "step": 100
    },
    {
      "epoch": 0.027631248430042703,
      "grad_norm": 78.6441119779914,
      "learning_rate": 4.541666666666667e-06,
      "loss": 4.8007,
      "step": 110
    },
    {
      "epoch": 0.03014318010550113,
      "grad_norm": 112.82426912718005,
      "learning_rate": 4.958333333333334e-06,
      "loss": 4.8032,
      "step": 120
    },
    {
      "epoch": 0.032655111780959555,
      "grad_norm": 44.43109383036881,
      "learning_rate": 4.999932966293553e-06,
      "loss": 4.5469,
      "step": 130
    },
    {
      "epoch": 0.03516704345641799,
      "grad_norm": 42.94277182614008,
      "learning_rate": 4.999701249454364e-06,
      "loss": 4.6328,
      "step": 140
    },
    {
      "epoch": 0.037678975131876416,
      "grad_norm": 102.19912456135354,
      "learning_rate": 4.999304037243215e-06,
      "loss": 4.5481,
      "step": 150
    },
    {
      "epoch": 0.04019090680733484,
      "grad_norm": 103.48520496328905,
      "learning_rate": 4.998741355957963e-06,
      "loss": 4.0883,
      "step": 160
    },
    {
      "epoch": 0.04270283848279327,
      "grad_norm": 59.75671346819062,
      "learning_rate": 4.998013242851519e-06,
      "loss": 4.09,
      "step": 170
    },
    {
      "epoch": 0.0452147701582517,
      "grad_norm": 54.56786450011987,
      "learning_rate": 4.997119746129383e-06,
      "loss": 3.8574,
      "step": 180
    },
    {
      "epoch": 0.047726701833710125,
      "grad_norm": 41.17219526900339,
      "learning_rate": 4.996060924946452e-06,
      "loss": 3.3506,
      "step": 190
    },
    {
      "epoch": 0.05023863350916855,
      "grad_norm": 64.19755013449358,
      "learning_rate": 4.994836849403107e-06,
      "loss": 3.1208,
      "step": 200
    },
    {
      "epoch": 0.05275056518462698,
      "grad_norm": 45.5637995880343,
      "learning_rate": 4.9934476005405665e-06,
      "loss": 2.9927,
      "step": 210
    },
    {
      "epoch": 0.05526249686008541,
      "grad_norm": 45.04513390690093,
      "learning_rate": 4.991893270335526e-06,
      "loss": 2.6818,
      "step": 220
    },
    {
      "epoch": 0.057774428535543834,
      "grad_norm": 46.258300578654584,
      "learning_rate": 4.990173961694062e-06,
      "loss": 2.4673,
      "step": 230
    },
    {
      "epoch": 0.06028636021100226,
      "grad_norm": 53.325400128100654,
      "learning_rate": 4.988289788444829e-06,
      "loss": 2.1375,
      "step": 240
    },
    {
      "epoch": 0.06279829188646069,
      "grad_norm": 28.819163526585655,
      "learning_rate": 4.986240875331513e-06,
      "loss": 1.9451,
      "step": 250
    },
    {
      "epoch": 0.06531022356191911,
      "grad_norm": 63.6642536784459,
      "learning_rate": 4.984027358004582e-06,
      "loss": 1.7398,
      "step": 260
    },
    {
      "epoch": 0.06782215523737754,
      "grad_norm": 64.052007917374,
      "learning_rate": 4.981649383012295e-06,
      "loss": 1.5556,
      "step": 270
    },
    {
      "epoch": 0.07033408691283598,
      "grad_norm": 41.550046723661794,
      "learning_rate": 4.979107107791009e-06,
      "loss": 1.5029,
      "step": 280
    },
    {
      "epoch": 0.0728460185882944,
      "grad_norm": 28.030603583599877,
      "learning_rate": 4.976400700654752e-06,
      "loss": 1.4048,
      "step": 290
    },
    {
      "epoch": 0.07535795026375283,
      "grad_norm": 21.7743052603298,
      "learning_rate": 4.9735303407840775e-06,
      "loss": 1.3447,
      "step": 300
    },
    {
      "epoch": 0.07786988193921125,
      "grad_norm": 27.724313133031774,
      "learning_rate": 4.970496218214205e-06,
      "loss": 1.369,
      "step": 310
    },
    {
      "epoch": 0.08038181361466969,
      "grad_norm": 39.89486925273176,
      "learning_rate": 4.9672985338224355e-06,
      "loss": 1.2511,
      "step": 320
    },
    {
      "epoch": 0.0828937452901281,
      "grad_norm": 32.71736411454423,
      "learning_rate": 4.963937499314857e-06,
      "loss": 1.1617,
      "step": 330
    },
    {
      "epoch": 0.08540567696558654,
      "grad_norm": 17.47130667747679,
      "learning_rate": 4.960413337212321e-06,
      "loss": 1.1987,
      "step": 340
    },
    {
      "epoch": 0.08791760864104496,
      "grad_norm": 17.44490963530016,
      "learning_rate": 4.956726280835718e-06,
      "loss": 1.2543,
      "step": 350
    },
    {
      "epoch": 0.0904295403165034,
      "grad_norm": 52.67566693581501,
      "learning_rate": 4.952876574290525e-06,
      "loss": 1.1756,
      "step": 360
    },
    {
      "epoch": 0.09294147199196182,
      "grad_norm": 36.08376015056633,
      "learning_rate": 4.948864472450646e-06,
      "loss": 1.1497,
      "step": 370
    },
    {
      "epoch": 0.09545340366742025,
      "grad_norm": 33.06467813616241,
      "learning_rate": 4.9446902409415386e-06,
      "loss": 1.1862,
      "step": 380
    },
    {
      "epoch": 0.09796533534287867,
      "grad_norm": 79.8613840478269,
      "learning_rate": 4.940354156122624e-06,
      "loss": 1.1443,
      "step": 390
    },
    {
      "epoch": 0.1004772670183371,
      "grad_norm": 28.636783892356842,
      "learning_rate": 4.935856505068999e-06,
      "loss": 1.0705,
      "step": 400
    },
    {
      "epoch": 0.10298919869379553,
      "grad_norm": 28.248493771801982,
      "learning_rate": 4.931197585552418e-06,
      "loss": 1.0342,
      "step": 410
    },
    {
      "epoch": 0.10550113036925396,
      "grad_norm": 35.58893396345627,
      "learning_rate": 4.92637770602159e-06,
      "loss": 1.0984,
      "step": 420
    },
    {
      "epoch": 0.10801306204471238,
      "grad_norm": 29.5108168082451,
      "learning_rate": 4.92139718558175e-06,
      "loss": 1.121,
      "step": 430
    },
    {
      "epoch": 0.11052499372017081,
      "grad_norm": 17.994080670599825,
      "learning_rate": 4.916256353973535e-06,
      "loss": 1.1537,
      "step": 440
    },
    {
      "epoch": 0.11303692539562923,
      "grad_norm": 32.844817441284924,
      "learning_rate": 4.910955551551153e-06,
      "loss": 1.1486,
      "step": 450
    },
    {
      "epoch": 0.11554885707108767,
      "grad_norm": 30.914619782560816,
      "learning_rate": 4.90549512925985e-06,
      "loss": 1.0481,
      "step": 460
    },
    {
      "epoch": 0.11806078874654609,
      "grad_norm": 25.55475057225604,
      "learning_rate": 4.899875448612672e-06,
      "loss": 1.0072,
      "step": 470
    },
    {
      "epoch": 0.12057272042200452,
      "grad_norm": 201.00243368632337,
      "learning_rate": 4.894096881666538e-06,
      "loss": 1.1146,
      "step": 480
    },
    {
      "epoch": 0.12308465209746294,
      "grad_norm": 51.017788501532685,
      "learning_rate": 4.8881598109976e-06,
      "loss": 0.9959,
      "step": 490
    },
    {
      "epoch": 0.12559658377292138,
      "grad_norm": 30.076609186117565,
      "learning_rate": 4.882064629675917e-06,
      "loss": 1.0998,
      "step": 500
    },
    {
      "epoch": 0.1281085154483798,
      "grad_norm": 25.914605989320286,
      "learning_rate": 4.875811741239431e-06,
      "loss": 0.9488,
      "step": 510
    },
    {
      "epoch": 0.13062044712383822,
      "grad_norm": 174.16033120144348,
      "learning_rate": 4.869401559667253e-06,
      "loss": 1.0104,
      "step": 520
    },
    {
      "epoch": 0.13313237879929665,
      "grad_norm": 24.53663814061994,
      "learning_rate": 4.862834509352251e-06,
      "loss": 0.9866,
      "step": 530
    },
    {
      "epoch": 0.1356443104747551,
      "grad_norm": 106.69096534898281,
      "learning_rate": 4.8561110250729546e-06,
      "loss": 1.0145,
      "step": 540
    },
    {
      "epoch": 0.13815624215021352,
      "grad_norm": 13.875443655979048,
      "learning_rate": 4.849231551964771e-06,
      "loss": 1.0714,
      "step": 550
    },
    {
      "epoch": 0.14066817382567195,
      "grad_norm": 21.267058863118724,
      "learning_rate": 4.842196545490512e-06,
      "loss": 1.0451,
      "step": 560
    },
    {
      "epoch": 0.14318010550113036,
      "grad_norm": 28.25986035068646,
      "learning_rate": 4.8350064714102415e-06,
      "loss": 0.9979,
      "step": 570
    },
    {
      "epoch": 0.1456920371765888,
      "grad_norm": 18.771189948926747,
      "learning_rate": 4.827661805750438e-06,
      "loss": 1.0446,
      "step": 580
    },
    {
      "epoch": 0.14820396885204723,
      "grad_norm": 33.442572448255376,
      "learning_rate": 4.82016303477248e-06,
      "loss": 1.0067,
      "step": 590
    },
    {
      "epoch": 0.15071590052750566,
      "grad_norm": 18.758813933167882,
      "learning_rate": 4.812510654940453e-06,
      "loss": 0.9925,
      "step": 600
    },
    {
      "epoch": 0.15322783220296407,
      "grad_norm": 18.44228949074116,
      "learning_rate": 4.804705172888277e-06,
      "loss": 0.9199,
      "step": 610
    },
    {
      "epoch": 0.1557397638784225,
      "grad_norm": 18.907319724853284,
      "learning_rate": 4.796747105386169e-06,
      "loss": 0.9933,
      "step": 620
    },
    {
      "epoch": 0.15825169555388094,
      "grad_norm": 24.999775435118767,
      "learning_rate": 4.788636979306428e-06,
      "loss": 0.946,
      "step": 630
    },
    {
      "epoch": 0.16076362722933937,
      "grad_norm": 20.69959166678649,
      "learning_rate": 4.78037533158855e-06,
      "loss": 1.0123,
      "step": 640
    },
    {
      "epoch": 0.16327555890479778,
      "grad_norm": 88.2316410192213,
      "learning_rate": 4.771962709203684e-06,
      "loss": 0.9311,
      "step": 650
    },
    {
      "epoch": 0.1657874905802562,
      "grad_norm": 18.849075967996672,
      "learning_rate": 4.763399669118414e-06,
      "loss": 0.9532,
      "step": 660
    },
    {
      "epoch": 0.16829942225571465,
      "grad_norm": 9.168139689519153,
      "learning_rate": 4.754686778257891e-06,
      "loss": 0.8347,
      "step": 670
    },
    {
      "epoch": 0.17081135393117308,
      "grad_norm": 57.688726700615845,
      "learning_rate": 4.745824613468293e-06,
      "loss": 0.892,
      "step": 680
    },
    {
      "epoch": 0.1733232856066315,
      "grad_norm": 34.38325408055489,
      "learning_rate": 4.736813761478638e-06,
      "loss": 0.941,
      "step": 690
    },
    {
      "epoch": 0.17583521728208992,
      "grad_norm": 18.535830277017187,
      "learning_rate": 4.727654818861937e-06,
      "loss": 0.9297,
      "step": 700
    },
    {
      "epoch": 0.17834714895754836,
      "grad_norm": 48.80919542270613,
      "learning_rate": 4.718348391995698e-06,
      "loss": 0.9512,
      "step": 710
    },
    {
      "epoch": 0.1808590806330068,
      "grad_norm": 10.424093008575271,
      "learning_rate": 4.7088950970217825e-06,
      "loss": 0.9061,
      "step": 720
    },
    {
      "epoch": 0.1833710123084652,
      "grad_norm": 13.175028312423365,
      "learning_rate": 4.699295559805606e-06,
      "loss": 0.9454,
      "step": 730
    },
    {
      "epoch": 0.18588294398392363,
      "grad_norm": 65.05485953479175,
      "learning_rate": 4.68955041589471e-06,
      "loss": 0.9451,
      "step": 740
    },
    {
      "epoch": 0.18839487565938207,
      "grad_norm": 25.29768837033324,
      "learning_rate": 4.679660310476682e-06,
      "loss": 0.831,
      "step": 750
    },
    {
      "epoch": 0.1909068073348405,
      "grad_norm": 19.717151893244445,
      "learning_rate": 4.669625898336439e-06,
      "loss": 0.8221,
      "step": 760
    },
    {
      "epoch": 0.1934187390102989,
      "grad_norm": 41.31445498318851,
      "learning_rate": 4.659447843812876e-06,
      "loss": 0.8756,
      "step": 770
    },
    {
      "epoch": 0.19593067068575734,
      "grad_norm": 14.81608234505718,
      "learning_rate": 4.649126820754887e-06,
      "loss": 0.9191,
      "step": 780
    },
    {
      "epoch": 0.19844260236121578,
      "grad_norm": 19.95005110698172,
      "learning_rate": 4.638663512476748e-06,
      "loss": 0.9372,
      "step": 790
    },
    {
      "epoch": 0.2009545340366742,
      "grad_norm": 20.065073396172792,
      "learning_rate": 4.628058611712879e-06,
      "loss": 1.0179,
      "step": 800
    },
    {
      "epoch": 0.20346646571213264,
      "grad_norm": 16.42772007372194,
      "learning_rate": 4.617312820571981e-06,
      "loss": 0.9572,
      "step": 810
    },
    {
      "epoch": 0.20597839738759105,
      "grad_norm": 15.309541598862848,
      "learning_rate": 4.606426850490551e-06,
      "loss": 0.9536,
      "step": 820
    },
    {
      "epoch": 0.20849032906304948,
      "grad_norm": 15.317369217882426,
      "learning_rate": 4.595401422185783e-06,
      "loss": 0.8942,
      "step": 830
    },
    {
      "epoch": 0.21100226073850792,
      "grad_norm": 12.500948604145567,
      "learning_rate": 4.584237265607847e-06,
      "loss": 0.8162,
      "step": 840
    },
    {
      "epoch": 0.21351419241396635,
      "grad_norm": 37.708490424859214,
      "learning_rate": 4.5729351198915715e-06,
      "loss": 0.9038,
      "step": 850
    },
    {
      "epoch": 0.21602612408942476,
      "grad_norm": 34.796355808216596,
      "learning_rate": 4.561495733307496e-06,
      "loss": 0.9207,
      "step": 860
    },
    {
      "epoch": 0.2185380557648832,
      "grad_norm": 26.298494542320878,
      "learning_rate": 4.549919863212341e-06,
      "loss": 0.9457,
      "step": 870
    },
    {
      "epoch": 0.22104998744034163,
      "grad_norm": 16.186659690778107,
      "learning_rate": 4.538208275998861e-06,
      "loss": 1.0715,
      "step": 880
    },
    {
      "epoch": 0.22356191911580006,
      "grad_norm": 24.662608067526172,
      "learning_rate": 4.526361747045108e-06,
      "loss": 0.9606,
      "step": 890
    },
    {
      "epoch": 0.22607385079125847,
      "grad_norm": 40.835876651445666,
      "learning_rate": 4.514381060663092e-06,
      "loss": 0.933,
      "step": 900
    },
    {
      "epoch": 0.2285857824667169,
      "grad_norm": 21.30586385592757,
      "learning_rate": 4.50226701004686e-06,
      "loss": 0.8754,
      "step": 910
    },
    {
      "epoch": 0.23109771414217534,
      "grad_norm": 16.612023821577484,
      "learning_rate": 4.49002039721998e-06,
      "loss": 0.7735,
      "step": 920
    },
    {
      "epoch": 0.23360964581763377,
      "grad_norm": 21.74905032903839,
      "learning_rate": 4.477642032982439e-06,
      "loss": 0.856,
      "step": 930
    },
    {
      "epoch": 0.23612157749309218,
      "grad_norm": 59.36191090209393,
      "learning_rate": 4.4651327368569695e-06,
      "loss": 0.9044,
      "step": 940
    },
    {
      "epoch": 0.2386335091685506,
      "grad_norm": 15.835736575328223,
      "learning_rate": 4.452493337034785e-06,
      "loss": 0.9162,
      "step": 950
    },
    {
      "epoch": 0.24114544084400905,
      "grad_norm": 15.54957354753592,
      "learning_rate": 4.439724670320755e-06,
      "loss": 0.9259,
      "step": 960
    },
    {
      "epoch": 0.24365737251946748,
      "grad_norm": 28.022932138978874,
      "learning_rate": 4.426827582077999e-06,
      "loss": 0.8854,
      "step": 970
    },
    {
      "epoch": 0.2461693041949259,
      "grad_norm": 13.702537050157368,
      "learning_rate": 4.413802926171921e-06,
      "loss": 0.8087,
      "step": 980
    },
    {
      "epoch": 0.24868123587038432,
      "grad_norm": 30.82271055352541,
      "learning_rate": 4.400651564913676e-06,
      "loss": 0.9174,
      "step": 990
    },
    {
      "epoch": 0.25119316754584275,
      "grad_norm": 19.961883729903487,
      "learning_rate": 4.387374369003083e-06,
      "loss": 0.8261,
      "step": 1000
    },
    {
      "epoch": 0.2537050992213012,
      "grad_norm": 13.910420016384712,
      "learning_rate": 4.373972217470976e-06,
      "loss": 0.8063,
      "step": 1010
    },
    {
      "epoch": 0.2562170308967596,
      "grad_norm": 15.74547796378487,
      "learning_rate": 4.36044599762101e-06,
      "loss": 0.893,
      "step": 1020
    },
    {
      "epoch": 0.25872896257221806,
      "grad_norm": 22.817146904726147,
      "learning_rate": 4.346796604970913e-06,
      "loss": 0.825,
      "step": 1030
    },
    {
      "epoch": 0.26124089424767644,
      "grad_norm": 34.97383754680491,
      "learning_rate": 4.333024943193198e-06,
      "loss": 0.8637,
      "step": 1040
    },
    {
      "epoch": 0.26375282592313487,
      "grad_norm": 14.442348394650033,
      "learning_rate": 4.31913192405534e-06,
      "loss": 0.8512,
      "step": 1050
    },
    {
      "epoch": 0.2662647575985933,
      "grad_norm": 109.10193874564091,
      "learning_rate": 4.305118467359402e-06,
      "loss": 0.8808,
      "step": 1060
    },
    {
      "epoch": 0.26877668927405174,
      "grad_norm": 27.206026550683518,
      "learning_rate": 4.290985500881143e-06,
      "loss": 0.8114,
      "step": 1070
    },
    {
      "epoch": 0.2712886209495102,
      "grad_norm": 16.12708282763735,
      "learning_rate": 4.276733960308599e-06,
      "loss": 0.8569,
      "step": 1080
    },
    {
      "epoch": 0.2738005526249686,
      "grad_norm": 84.431643774864,
      "learning_rate": 4.262364789180123e-06,
      "loss": 0.9139,
      "step": 1090
    },
    {
      "epoch": 0.27631248430042704,
      "grad_norm": 14.872893259437928,
      "learning_rate": 4.247878938821929e-06,
      "loss": 0.7897,
      "step": 1100
    },
    {
      "epoch": 0.2788244159758855,
      "grad_norm": 44.05664559088591,
      "learning_rate": 4.2332773682851e-06,
      "loss": 0.8999,
      "step": 1110
    },
    {
      "epoch": 0.2813363476513439,
      "grad_norm": 14.79051628084432,
      "learning_rate": 4.218561044282099e-06,
      "loss": 0.8443,
      "step": 1120
    },
    {
      "epoch": 0.2838482793268023,
      "grad_norm": 33.45294178219554,
      "learning_rate": 4.20373094112276e-06,
      "loss": 0.8567,
      "step": 1130
    },
    {
      "epoch": 0.2863602110022607,
      "grad_norm": 46.4659449561201,
      "learning_rate": 4.18878804064979e-06,
      "loss": 0.9058,
      "step": 1140
    },
    {
      "epoch": 0.28887214267771916,
      "grad_norm": 30.620320847801434,
      "learning_rate": 4.173733332173759e-06,
      "loss": 0.8482,
      "step": 1150
    },
    {
      "epoch": 0.2913840743531776,
      "grad_norm": 32.099385967886,
      "learning_rate": 4.158567812407609e-06,
      "loss": 0.9224,
      "step": 1160
    },
    {
      "epoch": 0.293896006028636,
      "grad_norm": 26.04095063399648,
      "learning_rate": 4.1432924854006554e-06,
      "loss": 0.7986,
      "step": 1170
    },
    {
      "epoch": 0.29640793770409446,
      "grad_norm": 21.53191952152889,
      "learning_rate": 4.127908362472121e-06,
      "loss": 0.8285,
      "step": 1180
    },
    {
      "epoch": 0.2989198693795529,
      "grad_norm": 14.804545412809087,
      "learning_rate": 4.112416462144181e-06,
      "loss": 0.8189,
      "step": 1190
    },
    {
      "epoch": 0.30143180105501133,
      "grad_norm": 28.04536852758184,
      "learning_rate": 4.096817810074521e-06,
      "loss": 0.9102,
      "step": 1200
    },
    {
      "epoch": 0.3039437327304697,
      "grad_norm": 16.38088643600367,
      "learning_rate": 4.081113438988443e-06,
      "loss": 0.8876,
      "step": 1210
    },
    {
      "epoch": 0.30645566440592814,
      "grad_norm": 26.286121352348058,
      "learning_rate": 4.065304388610491e-06,
      "loss": 0.8078,
      "step": 1220
    },
    {
      "epoch": 0.3089675960813866,
      "grad_norm": 49.451144669082176,
      "learning_rate": 4.049391705595605e-06,
      "loss": 0.874,
      "step": 1230
    },
    {
      "epoch": 0.311479527756845,
      "grad_norm": 43.93733084504646,
      "learning_rate": 4.033376443459842e-06,
      "loss": 0.8575,
      "step": 1240
    },
    {
      "epoch": 0.31399145943230344,
      "grad_norm": 33.5919976090378,
      "learning_rate": 4.017259662510613e-06,
      "loss": 0.8988,
      "step": 1250
    },
    {
      "epoch": 0.3165033911077619,
      "grad_norm": 41.047784413490206,
      "learning_rate": 4.00104242977649e-06,
      "loss": 0.8776,
      "step": 1260
    },
    {
      "epoch": 0.3190153227832203,
      "grad_norm": 21.485503202698204,
      "learning_rate": 3.9847258189365664e-06,
      "loss": 0.9317,
      "step": 1270
    },
    {
      "epoch": 0.32152725445867875,
      "grad_norm": 18.745800000012018,
      "learning_rate": 3.968310910249364e-06,
      "loss": 0.8406,
      "step": 1280
    },
    {
      "epoch": 0.3240391861341371,
      "grad_norm": 36.22488794899786,
      "learning_rate": 3.951798790481318e-06,
      "loss": 0.8217,
      "step": 1290
    },
    {
      "epoch": 0.32655111780959556,
      "grad_norm": 16.051366022833278,
      "learning_rate": 3.9351905528348285e-06,
      "loss": 0.8654,
      "step": 1300
    },
    {
      "epoch": 0.329063049485054,
      "grad_norm": 28.70539547850433,
      "learning_rate": 3.918487296875881e-06,
      "loss": 0.8681,
      "step": 1310
    },
    {
      "epoch": 0.3315749811605124,
      "grad_norm": 14.816194644055994,
      "learning_rate": 3.901690128461248e-06,
      "loss": 0.8818,
      "step": 1320
    },
    {
      "epoch": 0.33408691283597086,
      "grad_norm": 13.007885823046339,
      "learning_rate": 3.8848001596652765e-06,
      "loss": 0.8706,
      "step": 1330
    },
    {
      "epoch": 0.3365988445114293,
      "grad_norm": 12.482101372342157,
      "learning_rate": 3.867818508706261e-06,
      "loss": 0.8012,
      "step": 1340
    },
    {
      "epoch": 0.33911077618688773,
      "grad_norm": 26.169043780262097,
      "learning_rate": 3.850746299872412e-06,
      "loss": 0.9217,
      "step": 1350
    },
    {
      "epoch": 0.34162270786234616,
      "grad_norm": 46.59460344018054,
      "learning_rate": 3.833584663447418e-06,
      "loss": 0.876,
      "step": 1360
    },
    {
      "epoch": 0.3441346395378046,
      "grad_norm": 16.84714833096335,
      "learning_rate": 3.816334735635618e-06,
      "loss": 0.8291,
      "step": 1370
    },
    {
      "epoch": 0.346646571213263,
      "grad_norm": 37.545682101952806,
      "learning_rate": 3.7989976584867743e-06,
      "loss": 0.7684,
      "step": 1380
    },
    {
      "epoch": 0.3491585028887214,
      "grad_norm": 27.139012954460355,
      "learning_rate": 3.7815745798204646e-06,
      "loss": 0.761,
      "step": 1390
    },
    {
      "epoch": 0.35167043456417985,
      "grad_norm": 32.23972648827027,
      "learning_rate": 3.7640666531500868e-06,
      "loss": 0.796,
      "step": 1400
    },
    {
      "epoch": 0.3541823662396383,
      "grad_norm": 31.529769235569418,
      "learning_rate": 3.7464750376064934e-06,
      "loss": 0.8033,
      "step": 1410
    },
    {
      "epoch": 0.3566942979150967,
      "grad_norm": 20.56253529260212,
      "learning_rate": 3.7288008978612457e-06,
      "loss": 0.8213,
      "step": 1420
    },
    {
      "epoch": 0.35920622959055515,
      "grad_norm": 10.619817778124629,
      "learning_rate": 3.711045404049507e-06,
      "loss": 0.829,
      "step": 1430
    },
    {
      "epoch": 0.3617181612660136,
      "grad_norm": 19.304755989056705,
      "learning_rate": 3.693209731692573e-06,
      "loss": 0.8201,
      "step": 1440
    },
    {
      "epoch": 0.364230092941472,
      "grad_norm": 18.94828073608545,
      "learning_rate": 3.675295061620047e-06,
      "loss": 0.851,
      "step": 1450
    },
    {
      "epoch": 0.3667420246169304,
      "grad_norm": 29.832008452352895,
      "learning_rate": 3.6573025798916566e-06,
      "loss": 0.7624,
      "step": 1460
    },
    {
      "epoch": 0.36925395629238883,
      "grad_norm": 34.613004229596356,
      "learning_rate": 3.6392334777187354e-06,
      "loss": 0.7536,
      "step": 1470
    },
    {
      "epoch": 0.37176588796784726,
      "grad_norm": 15.983386653537558,
      "learning_rate": 3.621088951385353e-06,
      "loss": 0.769,
      "step": 1480
    },
    {
      "epoch": 0.3742778196433057,
      "grad_norm": 18.541258614645898,
      "learning_rate": 3.6028702021691164e-06,
      "loss": 0.8561,
      "step": 1490
    },
    {
      "epoch": 0.37678975131876413,
      "grad_norm": 13.569243914961227,
      "learning_rate": 3.5845784362616375e-06,
      "loss": 0.8026,
      "step": 1500
    },
    {
      "epoch": 0.37930168299422257,
      "grad_norm": 13.589477313526036,
      "learning_rate": 3.566214864688674e-06,
      "loss": 0.7856,
      "step": 1510
    },
    {
      "epoch": 0.381813614669681,
      "grad_norm": 9.08955551951589,
      "learning_rate": 3.5477807032299565e-06,
      "loss": 0.8099,
      "step": 1520
    },
    {
      "epoch": 0.38432554634513943,
      "grad_norm": 30.698836150904377,
      "learning_rate": 3.5292771723386916e-06,
      "loss": 0.7516,
      "step": 1530
    },
    {
      "epoch": 0.3868374780205978,
      "grad_norm": 13.203900449832538,
      "learning_rate": 3.5107054970607624e-06,
      "loss": 0.8014,
      "step": 1540
    },
    {
      "epoch": 0.38934940969605625,
      "grad_norm": 22.948546424383103,
      "learning_rate": 3.4920669069536265e-06,
      "loss": 0.8049,
      "step": 1550
    },
    {
      "epoch": 0.3918613413715147,
      "grad_norm": 31.428429983393126,
      "learning_rate": 3.4733626360049065e-06,
      "loss": 0.8858,
      "step": 1560
    },
    {
      "epoch": 0.3943732730469731,
      "grad_norm": 17.236743198133453,
      "learning_rate": 3.4545939225506935e-06,
      "loss": 0.8511,
      "step": 1570
    },
    {
      "epoch": 0.39688520472243155,
      "grad_norm": 22.980911438080426,
      "learning_rate": 3.4357620091935656e-06,
      "loss": 0.7983,
      "step": 1580
    },
    {
      "epoch": 0.39939713639789,
      "grad_norm": 11.441966137764789,
      "learning_rate": 3.416868142720316e-06,
      "loss": 0.7239,
      "step": 1590
    },
    {
      "epoch": 0.4019090680733484,
      "grad_norm": 14.91445014623267,
      "learning_rate": 3.39791357401941e-06,
      "loss": 0.7582,
      "step": 1600
    },
    {
      "epoch": 0.40442099974880685,
      "grad_norm": 41.79040186628522,
      "learning_rate": 3.3788995579981694e-06,
      "loss": 0.8557,
      "step": 1610
    },
    {
      "epoch": 0.4069329314242653,
      "grad_norm": 14.432086044310571,
      "learning_rate": 3.359827353499685e-06,
      "loss": 0.7389,
      "step": 1620
    },
    {
      "epoch": 0.40944486309972367,
      "grad_norm": 15.440731757695739,
      "learning_rate": 3.340698223219484e-06,
      "loss": 0.7378,
      "step": 1630
    },
    {
      "epoch": 0.4119567947751821,
      "grad_norm": 9.55068956773867,
      "learning_rate": 3.321513433621919e-06,
      "loss": 0.7992,
      "step": 1640
    },
    {
      "epoch": 0.41446872645064053,
      "grad_norm": 21.998448239749745,
      "learning_rate": 3.3022742548563293e-06,
      "loss": 0.7342,
      "step": 1650
    },
    {
      "epoch": 0.41698065812609897,
      "grad_norm": 10.474200018473434,
      "learning_rate": 3.282981960672948e-06,
      "loss": 0.7554,
      "step": 1660
    },
    {
      "epoch": 0.4194925898015574,
      "grad_norm": 20.24084901441791,
      "learning_rate": 3.2636378283385684e-06,
      "loss": 0.7177,
      "step": 1670
    },
    {
      "epoch": 0.42200452147701584,
      "grad_norm": 15.252907769914906,
      "learning_rate": 3.2442431385519853e-06,
      "loss": 0.8415,
      "step": 1680
    },
    {
      "epoch": 0.42451645315247427,
      "grad_norm": 31.941655660403704,
      "learning_rate": 3.2247991753592018e-06,
      "loss": 0.7944,
      "step": 1690
    },
    {
      "epoch": 0.4270283848279327,
      "grad_norm": 12.168344877357876,
      "learning_rate": 3.2053072260684206e-06,
      "loss": 0.7722,
      "step": 1700
    },
    {
      "epoch": 0.4295403165033911,
      "grad_norm": 24.94079803525497,
      "learning_rate": 3.1857685811648143e-06,
      "loss": 0.7928,
      "step": 1710
    },
    {
      "epoch": 0.4320522481788495,
      "grad_norm": 15.210129562675117,
      "learning_rate": 3.1661845342250874e-06,
      "loss": 0.7647,
      "step": 1720
    },
    {
      "epoch": 0.43456417985430795,
      "grad_norm": 29.737670350280883,
      "learning_rate": 3.1465563818318356e-06,
      "loss": 0.757,
      "step": 1730
    },
    {
      "epoch": 0.4370761115297664,
      "grad_norm": 34.11288282754126,
      "learning_rate": 3.1268854234877023e-06,
      "loss": 0.7198,
      "step": 1740
    },
    {
      "epoch": 0.4395880432052248,
      "grad_norm": 25.982392999301357,
      "learning_rate": 3.107172961529343e-06,
      "loss": 0.8304,
      "step": 1750
    },
    {
      "epoch": 0.44209997488068326,
      "grad_norm": 19.060965746755112,
      "learning_rate": 3.0874203010412057e-06,
      "loss": 0.7495,
      "step": 1760
    },
    {
      "epoch": 0.4446119065561417,
      "grad_norm": 9.450290682003281,
      "learning_rate": 3.067628749769125e-06,
      "loss": 0.6833,
      "step": 1770
    },
    {
      "epoch": 0.4471238382316001,
      "grad_norm": 81.52825084551178,
      "learning_rate": 3.047799618033739e-06,
      "loss": 0.7265,
      "step": 1780
    },
    {
      "epoch": 0.4496357699070585,
      "grad_norm": 112.89167490869183,
      "learning_rate": 3.027934218643742e-06,
      "loss": 0.7124,
      "step": 1790
    },
    {
      "epoch": 0.45214770158251694,
      "grad_norm": 12.80679045659251,
      "learning_rate": 3.008033866808967e-06,
      "loss": 0.795,
      "step": 1800
    },
    {
      "epoch": 0.45465963325797537,
      "grad_norm": 7.979181961331163,
      "learning_rate": 2.9880998800533095e-06,
      "loss": 0.7602,
      "step": 1810
    },
    {
      "epoch": 0.4571715649334338,
      "grad_norm": 104.22429484767227,
      "learning_rate": 2.968133578127501e-06,
      "loss": 0.7924,
      "step": 1820
    },
    {
      "epoch": 0.45968349660889224,
      "grad_norm": 25.072483756192565,
      "learning_rate": 2.9481362829217335e-06,
      "loss": 0.7684,
      "step": 1830
    },
    {
      "epoch": 0.4621954282843507,
      "grad_norm": 28.08445712014526,
      "learning_rate": 2.9281093183781406e-06,
      "loss": 0.7569,
      "step": 1840
    },
    {
      "epoch": 0.4647073599598091,
      "grad_norm": 17.77498223316829,
      "learning_rate": 2.9080540104031487e-06,
      "loss": 0.8381,
      "step": 1850
    },
    {
      "epoch": 0.46721929163526754,
      "grad_norm": 32.223471519237506,
      "learning_rate": 2.887971686779688e-06,
      "loss": 0.7015,
      "step": 1860
    },
    {
      "epoch": 0.4697312233107259,
      "grad_norm": 48.40756577090116,
      "learning_rate": 2.8678636770792907e-06,
      "loss": 0.7002,
      "step": 1870
    },
    {
      "epoch": 0.47224315498618435,
      "grad_norm": 24.520924633617422,
      "learning_rate": 2.847731312574061e-06,
      "loss": 0.7943,
      "step": 1880
    },
    {
      "epoch": 0.4747550866616428,
      "grad_norm": 22.88033862082146,
      "learning_rate": 2.8275759261485407e-06,
      "loss": 0.6797,
      "step": 1890
    },
    {
      "epoch": 0.4772670183371012,
      "grad_norm": 21.004816938320573,
      "learning_rate": 2.80739885221146e-06,
      "loss": 0.8262,
      "step": 1900
    },
    {
      "epoch": 0.47977895001255966,
      "grad_norm": 40.505775555323616,
      "learning_rate": 2.7872014266073972e-06,
      "loss": 0.856,
      "step": 1910
    },
    {
      "epoch": 0.4822908816880181,
      "grad_norm": 15.196929474940113,
      "learning_rate": 2.7669849865283303e-06,
      "loss": 0.6539,
      "step": 1920
    },
    {
      "epoch": 0.4848028133634765,
      "grad_norm": 21.136530968784285,
      "learning_rate": 2.746750870425114e-06,
      "loss": 0.6839,
      "step": 1930
    },
    {
      "epoch": 0.48731474503893496,
      "grad_norm": 66.94533407380263,
      "learning_rate": 2.7265004179188607e-06,
      "loss": 0.7212,
      "step": 1940
    },
    {
      "epoch": 0.4898266767143934,
      "grad_norm": 17.537525895857943,
      "learning_rate": 2.7062349697122536e-06,
      "loss": 0.7541,
      "step": 1950
    },
    {
      "epoch": 0.4923386083898518,
      "grad_norm": 54.32967278575285,
      "learning_rate": 2.68595586750078e-06,
      "loss": 0.7619,
      "step": 1960
    },
    {
      "epoch": 0.4948505400653102,
      "grad_norm": 21.831994544086985,
      "learning_rate": 2.665664453883907e-06,
      "loss": 0.7713,
      "step": 1970
    },
    {
      "epoch": 0.49736247174076864,
      "grad_norm": 27.261105082644335,
      "learning_rate": 2.6453620722761897e-06,
      "loss": 0.7319,
      "step": 1980
    },
    {
      "epoch": 0.4998744034162271,
      "grad_norm": 303.80383252156145,
      "learning_rate": 2.6250500668183325e-06,
      "loss": 0.7719,
      "step": 1990
    },
    {
      "epoch": 0.5023863350916855,
      "grad_norm": 18.975055416205706,
      "learning_rate": 2.6047297822881962e-06,
      "loss": 0.8072,
      "step": 2000
    }
  ],
  "logging_steps": 10,
  "max_steps": 3981,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 250,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2517310013177856.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}