{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.8740629685157422,
  "eval_steps": 500,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.018740629685157422,
      "grad_norm": 1.3957158327102661,
      "learning_rate": 1.47e-05,
      "loss": 9.729989013671876,
      "step": 50
    },
    {
      "epoch": 0.037481259370314844,
      "grad_norm": 1.4920259714126587,
      "learning_rate": 2.97e-05,
      "loss": 9.0034130859375,
      "step": 100
    },
    {
      "epoch": 0.05622188905547226,
      "grad_norm": 1.3049159049987793,
      "learning_rate": 4.4699999999999996e-05,
      "loss": 8.466640014648437,
      "step": 150
    },
    {
      "epoch": 0.07496251874062969,
      "grad_norm": 0.8636816740036011,
      "learning_rate": 5.97e-05,
      "loss": 7.869417724609375,
      "step": 200
    },
    {
      "epoch": 0.0937031484257871,
      "grad_norm": 0.46240487694740295,
      "learning_rate": 7.47e-05,
      "loss": 7.4413623046875,
      "step": 250
    },
    {
      "epoch": 0.11244377811094453,
      "grad_norm": 0.42481327056884766,
      "learning_rate": 8.969999999999998e-05,
      "loss": 7.2437841796875,
      "step": 300
    },
    {
      "epoch": 0.13118440779610194,
      "grad_norm": 0.48112183809280396,
      "learning_rate": 0.00010469999999999998,
      "loss": 7.021725463867187,
      "step": 350
    },
    {
      "epoch": 0.14992503748125938,
      "grad_norm": 0.4746127724647522,
      "learning_rate": 0.0001197,
      "loss": 6.769576416015625,
      "step": 400
    },
    {
      "epoch": 0.1686656671664168,
      "grad_norm": 0.48956719040870667,
      "learning_rate": 0.0001347,
      "loss": 6.538475341796875,
      "step": 450
    },
    {
      "epoch": 0.1874062968515742,
      "grad_norm": 0.5688585638999939,
      "learning_rate": 0.00014969999999999998,
      "loss": 6.308619384765625,
      "step": 500
    },
    {
      "epoch": 0.1874062968515742,
      "eval_loss": 6.167688369750977,
      "eval_runtime": 1.0817,
      "eval_samples_per_second": 1849.023,
      "eval_steps_per_second": 14.792,
      "step": 500
    },
    {
      "epoch": 0.20614692653673164,
      "grad_norm": 0.7463164925575256,
      "learning_rate": 0.0001647,
      "loss": 6.135632934570313,
      "step": 550
    },
    {
      "epoch": 0.22488755622188905,
      "grad_norm": 0.6867733001708984,
      "learning_rate": 0.00017969999999999998,
      "loss": 5.97782470703125,
      "step": 600
    },
    {
      "epoch": 0.2436281859070465,
      "grad_norm": 0.6764475703239441,
      "learning_rate": 0.0001947,
      "loss": 5.829049072265625,
      "step": 650
    },
    {
      "epoch": 0.2623688155922039,
      "grad_norm": 0.8201554417610168,
      "learning_rate": 0.00020969999999999997,
      "loss": 5.679007568359375,
      "step": 700
    },
    {
      "epoch": 0.2811094452773613,
      "grad_norm": 0.7325769662857056,
      "learning_rate": 0.0002247,
      "loss": 5.54383056640625,
      "step": 750
    },
    {
      "epoch": 0.29985007496251875,
      "grad_norm": 0.8120824694633484,
      "learning_rate": 0.0002397,
      "loss": 5.398251953125,
      "step": 800
    },
    {
      "epoch": 0.31859070464767614,
      "grad_norm": 0.8338794708251953,
      "learning_rate": 0.00025469999999999996,
      "loss": 5.269853515625,
      "step": 850
    },
    {
      "epoch": 0.3373313343328336,
      "grad_norm": 0.747706413269043,
      "learning_rate": 0.0002697,
      "loss": 5.1615771484375,
      "step": 900
    },
    {
      "epoch": 0.356071964017991,
      "grad_norm": 1.213719367980957,
      "learning_rate": 0.0002847,
      "loss": 5.037232666015625,
      "step": 950
    },
    {
      "epoch": 0.3748125937031484,
      "grad_norm": 1.0340992212295532,
      "learning_rate": 0.00029969999999999997,
      "loss": 4.913594055175781,
      "step": 1000
    },
    {
      "epoch": 0.3748125937031484,
      "eval_loss": 4.758162021636963,
      "eval_runtime": 1.0759,
      "eval_samples_per_second": 1858.983,
      "eval_steps_per_second": 14.872,
      "step": 1000
    },
    {
      "epoch": 0.39355322338830584,
      "grad_norm": 0.8748279809951782,
      "learning_rate": 0.00029706,
      "loss": 4.810666809082031,
      "step": 1050
    },
    {
      "epoch": 0.4122938530734633,
      "grad_norm": 0.8297407031059265,
      "learning_rate": 0.00029405999999999996,
      "loss": 4.683534851074219,
      "step": 1100
    },
    {
      "epoch": 0.43103448275862066,
      "grad_norm": 0.9721106290817261,
      "learning_rate": 0.00029105999999999994,
      "loss": 4.601776733398437,
      "step": 1150
    },
    {
      "epoch": 0.4497751124437781,
      "grad_norm": 0.9634228348731995,
      "learning_rate": 0.00028806,
      "loss": 4.503483581542969,
      "step": 1200
    },
    {
      "epoch": 0.46851574212893554,
      "grad_norm": 1.0372576713562012,
      "learning_rate": 0.00028506,
      "loss": 4.401589050292968,
      "step": 1250
    },
    {
      "epoch": 0.487256371814093,
      "grad_norm": 1.0020194053649902,
      "learning_rate": 0.00028206,
      "loss": 4.325810546875,
      "step": 1300
    },
    {
      "epoch": 0.5059970014992504,
      "grad_norm": 0.969415545463562,
      "learning_rate": 0.00027906,
      "loss": 4.261092834472656,
      "step": 1350
    },
    {
      "epoch": 0.5247376311844077,
      "grad_norm": 1.1663305759429932,
      "learning_rate": 0.00027605999999999996,
      "loss": 4.167784423828125,
      "step": 1400
    },
    {
      "epoch": 0.5434782608695652,
      "grad_norm": 1.0733853578567505,
      "learning_rate": 0.00027306,
      "loss": 4.096864624023437,
      "step": 1450
    },
    {
      "epoch": 0.5622188905547226,
      "grad_norm": 1.0118414163589478,
      "learning_rate": 0.00027005999999999997,
      "loss": 4.057973937988281,
      "step": 1500
    },
    {
      "epoch": 0.5622188905547226,
      "eval_loss": 3.895841121673584,
      "eval_runtime": 1.0897,
      "eval_samples_per_second": 1835.392,
      "eval_steps_per_second": 14.683,
      "step": 1500
    },
    {
      "epoch": 0.5809595202398801,
      "grad_norm": 0.9456564784049988,
      "learning_rate": 0.00026705999999999995,
      "loss": 3.999591064453125,
      "step": 1550
    },
    {
      "epoch": 0.5997001499250375,
      "grad_norm": 1.0527812242507935,
      "learning_rate": 0.00026406,
      "loss": 3.928629150390625,
      "step": 1600
    },
    {
      "epoch": 0.618440779610195,
      "grad_norm": 0.9593741297721863,
      "learning_rate": 0.00026105999999999997,
      "loss": 3.86834228515625,
      "step": 1650
    },
    {
      "epoch": 0.6371814092953523,
      "grad_norm": 1.0123467445373535,
      "learning_rate": 0.00025805999999999995,
      "loss": 3.8236553955078123,
      "step": 1700
    },
    {
      "epoch": 0.6559220389805097,
      "grad_norm": 1.0930705070495605,
      "learning_rate": 0.00025506,
      "loss": 3.7675225830078123,
      "step": 1750
    },
    {
      "epoch": 0.6746626686656672,
      "grad_norm": 1.2183789014816284,
      "learning_rate": 0.00025205999999999997,
      "loss": 3.722022705078125,
      "step": 1800
    },
    {
      "epoch": 0.6934032983508246,
      "grad_norm": 1.1733542680740356,
      "learning_rate": 0.00024906,
      "loss": 3.6782516479492187,
      "step": 1850
    },
    {
      "epoch": 0.712143928035982,
      "grad_norm": 1.1061464548110962,
      "learning_rate": 0.00024606,
      "loss": 3.6340737915039063,
      "step": 1900
    },
    {
      "epoch": 0.7308845577211395,
      "grad_norm": 1.1385581493377686,
      "learning_rate": 0.00024306,
      "loss": 3.596466369628906,
      "step": 1950
    },
    {
      "epoch": 0.7496251874062968,
      "grad_norm": 1.2214317321777344,
      "learning_rate": 0.00024005999999999998,
      "loss": 3.5506155395507815,
      "step": 2000
    },
    {
      "epoch": 0.7496251874062968,
      "eval_loss": 3.386573553085327,
      "eval_runtime": 1.0773,
      "eval_samples_per_second": 1856.47,
      "eval_steps_per_second": 14.852,
      "step": 2000
    },
    {
      "epoch": 0.7683658170914542,
      "grad_norm": 1.081070899963379,
      "learning_rate": 0.00023705999999999998,
      "loss": 3.540574951171875,
      "step": 2050
    },
    {
      "epoch": 0.7871064467766117,
      "grad_norm": 1.143646001815796,
      "learning_rate": 0.00023406,
      "loss": 3.496670837402344,
      "step": 2100
    },
    {
      "epoch": 0.8058470764617691,
      "grad_norm": 1.1834206581115723,
      "learning_rate": 0.00023105999999999997,
      "loss": 3.4573333740234373,
      "step": 2150
    },
    {
      "epoch": 0.8245877061469266,
      "grad_norm": 1.1659979820251465,
      "learning_rate": 0.00022805999999999998,
      "loss": 3.454217529296875,
      "step": 2200
    },
    {
      "epoch": 0.843328335832084,
      "grad_norm": 1.1012392044067383,
      "learning_rate": 0.00022505999999999996,
      "loss": 3.410209045410156,
      "step": 2250
    },
    {
      "epoch": 0.8620689655172413,
      "grad_norm": 1.1149102449417114,
      "learning_rate": 0.00022205999999999997,
      "loss": 3.3710055541992188,
      "step": 2300
    },
    {
      "epoch": 0.8808095952023988,
      "grad_norm": 1.164297103881836,
      "learning_rate": 0.00021905999999999998,
      "loss": 3.3450372314453123,
      "step": 2350
    },
    {
      "epoch": 0.8995502248875562,
      "grad_norm": 1.15338933467865,
      "learning_rate": 0.00021605999999999996,
      "loss": 3.318499755859375,
      "step": 2400
    },
    {
      "epoch": 0.9182908545727136,
      "grad_norm": 1.1890783309936523,
      "learning_rate": 0.00021306,
      "loss": 3.288759460449219,
      "step": 2450
    },
    {
      "epoch": 0.9370314842578711,
      "grad_norm": 1.1333503723144531,
      "learning_rate": 0.00021006,
      "loss": 3.2676837158203127,
      "step": 2500
    },
    {
      "epoch": 0.9370314842578711,
      "eval_loss": 3.1189351081848145,
      "eval_runtime": 1.0777,
      "eval_samples_per_second": 1855.782,
      "eval_steps_per_second": 14.846,
      "step": 2500
    },
    {
      "epoch": 0.9557721139430285,
      "grad_norm": 1.0724334716796875,
      "learning_rate": 0.00020705999999999999,
      "loss": 3.264051818847656,
      "step": 2550
    },
    {
      "epoch": 0.974512743628186,
      "grad_norm": 1.0788660049438477,
      "learning_rate": 0.00020406,
      "loss": 3.240655822753906,
      "step": 2600
    },
    {
      "epoch": 0.9932533733133433,
      "grad_norm": 1.0767464637756348,
      "learning_rate": 0.00020105999999999998,
      "loss": 3.2095187377929686,
      "step": 2650
    },
    {
      "epoch": 1.0119940029985008,
      "grad_norm": 1.119018793106079,
      "learning_rate": 0.00019805999999999998,
      "loss": 3.1692465209960936,
      "step": 2700
    },
    {
      "epoch": 1.0307346326836582,
      "grad_norm": 1.1218299865722656,
      "learning_rate": 0.00019506,
      "loss": 3.13397705078125,
      "step": 2750
    },
    {
      "epoch": 1.0494752623688155,
      "grad_norm": 1.1980317831039429,
      "learning_rate": 0.00019205999999999997,
      "loss": 3.1120867919921875,
      "step": 2800
    },
    {
      "epoch": 1.068215892053973,
      "grad_norm": 1.0966095924377441,
      "learning_rate": 0.00018905999999999998,
      "loss": 3.122065124511719,
      "step": 2850
    },
    {
      "epoch": 1.0869565217391304,
      "grad_norm": 1.1618859767913818,
      "learning_rate": 0.00018605999999999996,
      "loss": 3.1155767822265625,
      "step": 2900
    },
    {
      "epoch": 1.105697151424288,
      "grad_norm": 1.0753134489059448,
      "learning_rate": 0.00018305999999999997,
      "loss": 3.072420959472656,
      "step": 2950
    },
    {
      "epoch": 1.1244377811094453,
      "grad_norm": 1.0856844186782837,
      "learning_rate": 0.00018005999999999998,
      "loss": 3.069416809082031,
      "step": 3000
    },
    {
      "epoch": 1.1244377811094453,
      "eval_loss": 2.9664721488952637,
      "eval_runtime": 1.0826,
      "eval_samples_per_second": 1847.477,
      "eval_steps_per_second": 14.78,
      "step": 3000
    },
    {
      "epoch": 1.1431784107946026,
      "grad_norm": 1.0895930528640747,
      "learning_rate": 0.00017705999999999996,
      "loss": 3.0642434692382814,
      "step": 3050
    },
    {
      "epoch": 1.1619190404797601,
      "grad_norm": 1.003448247909546,
      "learning_rate": 0.00017406,
      "loss": 3.0559292602539063,
      "step": 3100
    },
    {
      "epoch": 1.1806596701649175,
      "grad_norm": 1.0393710136413574,
      "learning_rate": 0.00017106,
      "loss": 3.0400439453125,
      "step": 3150
    },
    {
      "epoch": 1.199400299850075,
      "grad_norm": 1.065361738204956,
      "learning_rate": 0.00016806,
      "loss": 3.0358206176757814,
      "step": 3200
    },
    {
      "epoch": 1.2181409295352323,
      "grad_norm": 1.11861252784729,
      "learning_rate": 0.00016506,
      "loss": 3.0290408325195313,
      "step": 3250
    },
    {
      "epoch": 1.23688155922039,
      "grad_norm": 1.0914726257324219,
      "learning_rate": 0.00016205999999999998,
      "loss": 3.006223449707031,
      "step": 3300
    },
    {
      "epoch": 1.2556221889055472,
      "grad_norm": 1.0921860933303833,
      "learning_rate": 0.00015905999999999998,
      "loss": 3.0094760131835936,
      "step": 3350
    },
    {
      "epoch": 1.2743628185907045,
      "grad_norm": 1.1069430112838745,
      "learning_rate": 0.00015606,
      "loss": 2.9911233520507814,
      "step": 3400
    },
    {
      "epoch": 1.293103448275862,
      "grad_norm": 1.1226145029067993,
      "learning_rate": 0.00015305999999999997,
      "loss": 2.9855892944335936,
      "step": 3450
    },
    {
      "epoch": 1.3118440779610194,
      "grad_norm": 1.0400314331054688,
      "learning_rate": 0.00015005999999999998,
      "loss": 2.967692565917969,
      "step": 3500
    },
    {
      "epoch": 1.3118440779610194,
      "eval_loss": 2.855978012084961,
      "eval_runtime": 1.0725,
      "eval_samples_per_second": 1864.838,
      "eval_steps_per_second": 14.919,
      "step": 3500
    },
    {
      "epoch": 1.330584707646177,
      "grad_norm": 1.1032757759094238,
      "learning_rate": 0.00014706,
      "loss": 2.9611080932617186,
      "step": 3550
    },
    {
      "epoch": 1.3493253373313343,
      "grad_norm": 1.0784434080123901,
      "learning_rate": 0.00014406,
      "loss": 2.9634393310546874,
      "step": 3600
    },
    {
      "epoch": 1.3680659670164919,
      "grad_norm": 1.0988497734069824,
      "learning_rate": 0.00014105999999999998,
      "loss": 2.951357116699219,
      "step": 3650
    },
    {
      "epoch": 1.3868065967016492,
      "grad_norm": 1.1248743534088135,
      "learning_rate": 0.00013806,
      "loss": 2.935252685546875,
      "step": 3700
    },
    {
      "epoch": 1.4055472263868065,
      "grad_norm": 1.0736628770828247,
      "learning_rate": 0.00013506,
      "loss": 2.941317138671875,
      "step": 3750
    },
    {
      "epoch": 1.424287856071964,
      "grad_norm": 1.0715457201004028,
      "learning_rate": 0.00013205999999999998,
      "loss": 2.91337158203125,
      "step": 3800
    },
    {
      "epoch": 1.4430284857571214,
      "grad_norm": 1.0192804336547852,
      "learning_rate": 0.00012906,
      "loss": 2.9139144897460936,
      "step": 3850
    },
    {
      "epoch": 1.461769115442279,
      "grad_norm": 1.1134312152862549,
      "learning_rate": 0.00012606,
      "loss": 2.8937176513671874,
      "step": 3900
    },
    {
      "epoch": 1.4805097451274363,
      "grad_norm": 1.0220180749893188,
      "learning_rate": 0.00012306,
      "loss": 2.8912167358398437,
      "step": 3950
    },
    {
      "epoch": 1.4992503748125938,
      "grad_norm": 1.0451494455337524,
      "learning_rate": 0.00012005999999999999,
      "loss": 2.885633544921875,
      "step": 4000
    },
    {
      "epoch": 1.4992503748125938,
      "eval_loss": 2.7816965579986572,
      "eval_runtime": 1.0895,
      "eval_samples_per_second": 1835.711,
      "eval_steps_per_second": 14.686,
      "step": 4000
    },
    {
      "epoch": 1.5179910044977512,
      "grad_norm": 1.1047956943511963,
      "learning_rate": 0.00011705999999999998,
      "loss": 2.8766485595703126,
      "step": 4050
    },
    {
      "epoch": 1.5367316341829085,
      "grad_norm": 1.0666900873184204,
      "learning_rate": 0.00011405999999999999,
      "loss": 2.8759930419921873,
      "step": 4100
    },
    {
      "epoch": 1.555472263868066,
      "grad_norm": 1.054360032081604,
      "learning_rate": 0.00011105999999999998,
      "loss": 2.87041748046875,
      "step": 4150
    },
    {
      "epoch": 1.5742128935532234,
      "grad_norm": 0.9895801544189453,
      "learning_rate": 0.00010805999999999999,
      "loss": 2.8669659423828127,
      "step": 4200
    },
    {
      "epoch": 1.5929535232383807,
      "grad_norm": 1.037304162979126,
      "learning_rate": 0.00010506,
      "loss": 2.865623779296875,
      "step": 4250
    },
    {
      "epoch": 1.6116941529235382,
      "grad_norm": 1.0408546924591064,
      "learning_rate": 0.00010206,
      "loss": 2.8661175537109376,
      "step": 4300
    },
    {
      "epoch": 1.6304347826086958,
      "grad_norm": 1.0272372961044312,
      "learning_rate": 9.905999999999999e-05,
      "loss": 2.844149475097656,
      "step": 4350
    },
    {
      "epoch": 1.6491754122938531,
      "grad_norm": 1.111170768737793,
      "learning_rate": 9.605999999999998e-05,
      "loss": 2.8450665283203125,
      "step": 4400
    },
    {
      "epoch": 1.6679160419790104,
      "grad_norm": 1.0780202150344849,
      "learning_rate": 9.305999999999998e-05,
      "loss": 2.8447799682617188,
      "step": 4450
    },
    {
      "epoch": 1.686656671664168,
      "grad_norm": 1.010868787765503,
      "learning_rate": 9.006e-05,
      "loss": 2.8323443603515623,
      "step": 4500
    },
    {
      "epoch": 1.686656671664168,
      "eval_loss": 2.7236273288726807,
      "eval_runtime": 1.088,
      "eval_samples_per_second": 1838.168,
      "eval_steps_per_second": 14.705,
      "step": 4500
    },
    {
      "epoch": 1.7053973013493253,
      "grad_norm": 0.9945926666259766,
      "learning_rate": 8.706e-05,
      "loss": 2.8170339965820315,
      "step": 4550
    },
    {
      "epoch": 1.7241379310344827,
      "grad_norm": 1.0083485841751099,
      "learning_rate": 8.405999999999999e-05,
      "loss": 2.811573791503906,
      "step": 4600
    },
    {
      "epoch": 1.7428785607196402,
      "grad_norm": 0.9850161075592041,
      "learning_rate": 8.105999999999999e-05,
      "loss": 2.8162551879882813,
      "step": 4650
    },
    {
      "epoch": 1.7616191904047978,
      "grad_norm": 1.0177538394927979,
      "learning_rate": 7.806e-05,
      "loss": 2.8090859985351564,
      "step": 4700
    },
    {
      "epoch": 1.7803598200899549,
      "grad_norm": 1.0135115385055542,
      "learning_rate": 7.505999999999999e-05,
      "loss": 2.810616760253906,
      "step": 4750
    },
    {
      "epoch": 1.7991004497751124,
      "grad_norm": 0.9969961047172546,
      "learning_rate": 7.206e-05,
      "loss": 2.8046578979492187,
      "step": 4800
    },
    {
      "epoch": 1.81784107946027,
      "grad_norm": 1.0012859106063843,
      "learning_rate": 6.905999999999999e-05,
      "loss": 2.8064453125,
      "step": 4850
    },
    {
      "epoch": 1.8365817091454273,
      "grad_norm": 0.9606834650039673,
      "learning_rate": 6.606e-05,
      "loss": 2.8021035766601563,
      "step": 4900
    },
    {
      "epoch": 1.8553223388305846,
      "grad_norm": 1.109708309173584,
      "learning_rate": 6.306e-05,
      "loss": 2.7843258666992186,
      "step": 4950
    },
    {
      "epoch": 1.8740629685157422,
      "grad_norm": 1.040759563446045,
      "learning_rate": 6.005999999999999e-05,
      "loss": 2.7871945190429686,
      "step": 5000
    },
    {
      "epoch": 1.8740629685157422,
      "eval_loss": 2.683154582977295,
      "eval_runtime": 1.0812,
      "eval_samples_per_second": 1849.713,
      "eval_steps_per_second": 14.798,
      "step": 5000
    }
  ],
  "logging_steps": 50,
  "max_steps": 6000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3634717584900096.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": null
}