{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.2125851845296043,
  "eval_steps": 500,
  "global_step": 25000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008850340738118417,
      "grad_norm": 1.517469048500061,
      "learning_rate": 1.9919999999999997e-06,
      "loss": 1.519,
      "step": 100
    },
    {
      "epoch": 0.017700681476236834,
      "grad_norm": 1.6662976741790771,
      "learning_rate": 1.984e-06,
      "loss": 1.4776,
      "step": 200
    },
    {
      "epoch": 0.026551022214355253,
      "grad_norm": 0.8021761775016785,
      "learning_rate": 1.976e-06,
      "loss": 1.4026,
      "step": 300
    },
    {
      "epoch": 0.03540136295247367,
      "grad_norm": 0.8772447109222412,
      "learning_rate": 1.968e-06,
      "loss": 1.3306,
      "step": 400
    },
    {
      "epoch": 0.04425170369059209,
      "grad_norm": 0.4179963767528534,
      "learning_rate": 1.96e-06,
      "loss": 1.286,
      "step": 500
    },
    {
      "epoch": 0.053102044428710506,
      "grad_norm": 0.4829399287700653,
      "learning_rate": 1.9519999999999997e-06,
      "loss": 1.2756,
      "step": 600
    },
    {
      "epoch": 0.061952385166828924,
      "grad_norm": 0.5651261806488037,
      "learning_rate": 1.944e-06,
      "loss": 1.2721,
      "step": 700
    },
    {
      "epoch": 0.07080272590494734,
      "grad_norm": 0.5160872936248779,
      "learning_rate": 1.9359999999999998e-06,
      "loss": 1.2579,
      "step": 800
    },
    {
      "epoch": 0.07965306664306576,
      "grad_norm": 0.7501122355461121,
      "learning_rate": 1.928e-06,
      "loss": 1.2853,
      "step": 900
    },
    {
      "epoch": 0.08850340738118417,
      "grad_norm": 0.591393232345581,
      "learning_rate": 1.92e-06,
      "loss": 1.2472,
      "step": 1000
    },
    {
      "epoch": 0.0973537481193026,
      "grad_norm": 0.5573005080223083,
      "learning_rate": 1.9119999999999997e-06,
      "loss": 1.2348,
      "step": 1100
    },
    {
      "epoch": 0.10620408885742101,
      "grad_norm": 0.607552707195282,
      "learning_rate": 1.904e-06,
      "loss": 1.2679,
      "step": 1200
    },
    {
      "epoch": 0.11505442959553942,
      "grad_norm": 0.6899151802062988,
      "learning_rate": 1.8959999999999997e-06,
      "loss": 1.2513,
      "step": 1300
    },
    {
      "epoch": 0.12390477033365785,
      "grad_norm": 0.5849478244781494,
      "learning_rate": 1.8879999999999998e-06,
      "loss": 1.226,
      "step": 1400
    },
    {
      "epoch": 0.13275511107177626,
      "grad_norm": 0.6027043461799622,
      "learning_rate": 1.8799999999999998e-06,
      "loss": 1.2163,
      "step": 1500
    },
    {
      "epoch": 0.14160545180989467,
      "grad_norm": 0.6785966157913208,
      "learning_rate": 1.872e-06,
      "loss": 1.2076,
      "step": 1600
    },
    {
      "epoch": 0.1504557925480131,
      "grad_norm": 0.5975379347801208,
      "learning_rate": 1.864e-06,
      "loss": 1.2146,
      "step": 1700
    },
    {
      "epoch": 0.15930613328613152,
      "grad_norm": 0.685453474521637,
      "learning_rate": 1.856e-06,
      "loss": 1.198,
      "step": 1800
    },
    {
      "epoch": 0.16815647402424994,
      "grad_norm": 1.1250847578048706,
      "learning_rate": 1.848e-06,
      "loss": 1.218,
      "step": 1900
    },
    {
      "epoch": 0.17700681476236835,
      "grad_norm": 0.6129019260406494,
      "learning_rate": 1.84e-06,
      "loss": 1.1894,
      "step": 2000
    },
    {
      "epoch": 0.18585715550048676,
      "grad_norm": 0.7388463616371155,
      "learning_rate": 1.832e-06,
      "loss": 1.1962,
      "step": 2100
    },
    {
      "epoch": 0.1947074962386052,
      "grad_norm": 0.5909475684165955,
      "learning_rate": 1.824e-06,
      "loss": 1.241,
      "step": 2200
    },
    {
      "epoch": 0.2035578369767236,
      "grad_norm": 0.5264613032341003,
      "learning_rate": 1.816e-06,
      "loss": 1.1915,
      "step": 2300
    },
    {
      "epoch": 0.21240817771484202,
      "grad_norm": 0.6419306397438049,
      "learning_rate": 1.8079999999999999e-06,
      "loss": 1.1938,
      "step": 2400
    },
    {
      "epoch": 0.22125851845296043,
      "grad_norm": 0.6240788698196411,
      "learning_rate": 1.8e-06,
      "loss": 1.214,
      "step": 2500
    },
    {
      "epoch": 0.23010885919107885,
      "grad_norm": 0.7954173684120178,
      "learning_rate": 1.792e-06,
      "loss": 1.195,
      "step": 2600
    },
    {
      "epoch": 0.23895919992919729,
      "grad_norm": 0.7451553344726562,
      "learning_rate": 1.784e-06,
      "loss": 1.1793,
      "step": 2700
    },
    {
      "epoch": 0.2478095406673157,
      "grad_norm": 0.8471085429191589,
      "learning_rate": 1.776e-06,
      "loss": 1.1895,
      "step": 2800
    },
    {
      "epoch": 0.2566598814054341,
      "grad_norm": 1.0085513591766357,
      "learning_rate": 1.7679999999999998e-06,
      "loss": 1.2087,
      "step": 2900
    },
    {
      "epoch": 0.2655102221435525,
      "grad_norm": 0.7709102630615234,
      "learning_rate": 1.7599999999999999e-06,
      "loss": 1.1959,
      "step": 3000
    },
    {
      "epoch": 0.27436056288167093,
      "grad_norm": 0.7612162232398987,
      "learning_rate": 1.752e-06,
      "loss": 1.1803,
      "step": 3100
    },
    {
      "epoch": 0.28321090361978934,
      "grad_norm": 1.0669310092926025,
      "learning_rate": 1.744e-06,
      "loss": 1.1548,
      "step": 3200
    },
    {
      "epoch": 0.29206124435790776,
      "grad_norm": 0.7489383220672607,
      "learning_rate": 1.736e-06,
      "loss": 1.1654,
      "step": 3300
    },
    {
      "epoch": 0.3009115850960262,
      "grad_norm": 0.8011970520019531,
      "learning_rate": 1.7279999999999998e-06,
      "loss": 1.1789,
      "step": 3400
    },
    {
      "epoch": 0.30976192583414464,
      "grad_norm": 0.7374864816665649,
      "learning_rate": 1.7199999999999998e-06,
      "loss": 1.171,
      "step": 3500
    },
    {
      "epoch": 0.31861226657226305,
      "grad_norm": 0.8267546892166138,
      "learning_rate": 1.7119999999999999e-06,
      "loss": 1.1435,
      "step": 3600
    },
    {
      "epoch": 0.32746260731038146,
      "grad_norm": 0.8837570548057556,
      "learning_rate": 1.704e-06,
      "loss": 1.1674,
      "step": 3700
    },
    {
      "epoch": 0.33631294804849987,
      "grad_norm": 0.7216938138008118,
      "learning_rate": 1.696e-06,
      "loss": 1.1446,
      "step": 3800
    },
    {
      "epoch": 0.3451632887866183,
      "grad_norm": 0.8478527069091797,
      "learning_rate": 1.6879999999999998e-06,
      "loss": 1.1721,
      "step": 3900
    },
    {
      "epoch": 0.3540136295247367,
      "grad_norm": 0.8281861543655396,
      "learning_rate": 1.6799999999999998e-06,
      "loss": 1.1741,
      "step": 4000
    },
    {
      "epoch": 0.3628639702628551,
      "grad_norm": 0.6918274760246277,
      "learning_rate": 1.6719999999999998e-06,
      "loss": 1.1433,
      "step": 4100
    },
    {
      "epoch": 0.3717143110009735,
      "grad_norm": 0.6722283363342285,
      "learning_rate": 1.6639999999999999e-06,
      "loss": 1.189,
      "step": 4200
    },
    {
      "epoch": 0.38056465173909193,
      "grad_norm": 0.8337692022323608,
      "learning_rate": 1.656e-06,
      "loss": 1.1694,
      "step": 4300
    },
    {
      "epoch": 0.3894149924772104,
      "grad_norm": 0.7980256080627441,
      "learning_rate": 1.648e-06,
      "loss": 1.1794,
      "step": 4400
    },
    {
      "epoch": 0.3982653332153288,
      "grad_norm": 0.9903015494346619,
      "learning_rate": 1.6399999999999998e-06,
      "loss": 1.2027,
      "step": 4500
    },
    {
      "epoch": 0.4071156739534472,
      "grad_norm": 0.7119297385215759,
      "learning_rate": 1.6319999999999998e-06,
      "loss": 1.1638,
      "step": 4600
    },
    {
      "epoch": 0.41596601469156563,
      "grad_norm": 0.9170375466346741,
      "learning_rate": 1.624e-06,
      "loss": 1.1706,
      "step": 4700
    },
    {
      "epoch": 0.42481635542968404,
      "grad_norm": 0.7098402976989746,
      "learning_rate": 1.616e-06,
      "loss": 1.1738,
      "step": 4800
    },
    {
      "epoch": 0.43366669616780246,
      "grad_norm": 0.9326470494270325,
      "learning_rate": 1.608e-06,
      "loss": 1.1526,
      "step": 4900
    },
    {
      "epoch": 0.44251703690592087,
      "grad_norm": 0.7765191197395325,
      "learning_rate": 1.6e-06,
      "loss": 1.1687,
      "step": 5000
    },
    {
      "epoch": 0.4513673776440393,
      "grad_norm": 0.8953288197517395,
      "learning_rate": 1.592e-06,
      "loss": 1.1763,
      "step": 5100
    },
    {
      "epoch": 0.4602177183821577,
      "grad_norm": 1.0437840223312378,
      "learning_rate": 1.584e-06,
      "loss": 1.1883,
      "step": 5200
    },
    {
      "epoch": 0.4690680591202761,
      "grad_norm": 0.8432090282440186,
      "learning_rate": 1.576e-06,
      "loss": 1.1826,
      "step": 5300
    },
    {
      "epoch": 0.47791839985839457,
      "grad_norm": 0.7580952048301697,
      "learning_rate": 1.568e-06,
      "loss": 1.1813,
      "step": 5400
    },
    {
      "epoch": 0.486768740596513,
      "grad_norm": 0.9914817214012146,
      "learning_rate": 1.5599999999999999e-06,
      "loss": 1.1751,
      "step": 5500
    },
    {
      "epoch": 0.4956190813346314,
      "grad_norm": 0.7643041610717773,
      "learning_rate": 1.552e-06,
      "loss": 1.1615,
      "step": 5600
    },
    {
      "epoch": 0.5044694220727498,
      "grad_norm": 0.8275469541549683,
      "learning_rate": 1.544e-06,
      "loss": 1.188,
      "step": 5700
    },
    {
      "epoch": 0.5133197628108682,
      "grad_norm": 1.950104832649231,
      "learning_rate": 1.536e-06,
      "loss": 1.2018,
      "step": 5800
    },
    {
      "epoch": 0.5221701035489866,
      "grad_norm": 0.8191819787025452,
      "learning_rate": 1.528e-06,
      "loss": 1.1865,
      "step": 5900
    },
    {
      "epoch": 0.531020444287105,
      "grad_norm": 0.7783402800559998,
      "learning_rate": 1.5199999999999998e-06,
      "loss": 1.1687,
      "step": 6000
    },
    {
      "epoch": 0.5398707850252235,
      "grad_norm": 1.0385206937789917,
      "learning_rate": 1.5119999999999999e-06,
      "loss": 1.1883,
      "step": 6100
    },
    {
      "epoch": 0.5487211257633419,
      "grad_norm": 0.9720994234085083,
      "learning_rate": 1.504e-06,
      "loss": 1.1824,
      "step": 6200
    },
    {
      "epoch": 0.5575714665014603,
      "grad_norm": 0.9594699740409851,
      "learning_rate": 1.496e-06,
      "loss": 1.1908,
      "step": 6300
    },
    {
      "epoch": 0.5664218072395787,
      "grad_norm": 0.9793533682823181,
      "learning_rate": 1.488e-06,
      "loss": 1.1385,
      "step": 6400
    },
    {
      "epoch": 0.5752721479776971,
      "grad_norm": 0.9187788367271423,
      "learning_rate": 1.48e-06,
      "loss": 1.1536,
      "step": 6500
    },
    {
      "epoch": 0.5841224887158155,
      "grad_norm": 0.9121326208114624,
      "learning_rate": 1.4719999999999998e-06,
      "loss": 1.1577,
      "step": 6600
    },
    {
      "epoch": 0.5929728294539339,
      "grad_norm": 1.3253475427627563,
      "learning_rate": 1.4639999999999999e-06,
      "loss": 1.1462,
      "step": 6700
    },
    {
      "epoch": 0.6018231701920524,
      "grad_norm": 0.9218689799308777,
      "learning_rate": 1.456e-06,
      "loss": 1.1745,
      "step": 6800
    },
    {
      "epoch": 0.6106735109301709,
      "grad_norm": 0.8204641938209534,
      "learning_rate": 1.448e-06,
      "loss": 1.1871,
      "step": 6900
    },
    {
      "epoch": 0.6195238516682893,
      "grad_norm": 0.8714615702629089,
      "learning_rate": 1.44e-06,
      "loss": 1.183,
      "step": 7000
    },
    {
      "epoch": 0.6283741924064077,
      "grad_norm": 1.2564398050308228,
      "learning_rate": 1.4319999999999998e-06,
      "loss": 1.1614,
      "step": 7100
    },
    {
      "epoch": 0.6372245331445261,
      "grad_norm": 0.8741295337677002,
      "learning_rate": 1.4239999999999998e-06,
      "loss": 1.1314,
      "step": 7200
    },
    {
      "epoch": 0.6460748738826445,
      "grad_norm": 0.8480414152145386,
      "learning_rate": 1.4159999999999999e-06,
      "loss": 1.1522,
      "step": 7300
    },
    {
      "epoch": 0.6549252146207629,
      "grad_norm": 1.0986804962158203,
      "learning_rate": 1.408e-06,
      "loss": 1.1885,
      "step": 7400
    },
    {
      "epoch": 0.6637755553588813,
      "grad_norm": 0.9965903759002686,
      "learning_rate": 1.4e-06,
      "loss": 1.1804,
      "step": 7500
    },
    {
      "epoch": 0.6726258960969997,
      "grad_norm": 1.079156756401062,
      "learning_rate": 1.3919999999999998e-06,
      "loss": 1.1797,
      "step": 7600
    },
    {
      "epoch": 0.6814762368351182,
      "grad_norm": 1.0092281103134155,
      "learning_rate": 1.3839999999999998e-06,
      "loss": 1.1613,
      "step": 7700
    },
    {
      "epoch": 0.6903265775732366,
      "grad_norm": 0.9988503456115723,
      "learning_rate": 1.3759999999999998e-06,
      "loss": 1.1526,
      "step": 7800
    },
    {
      "epoch": 0.699176918311355,
      "grad_norm": 1.0048474073410034,
      "learning_rate": 1.368e-06,
      "loss": 1.1804,
      "step": 7900
    },
    {
      "epoch": 0.7080272590494734,
      "grad_norm": 0.7530746459960938,
      "learning_rate": 1.3600000000000001e-06,
      "loss": 1.1465,
      "step": 8000
    },
    {
      "epoch": 0.7168775997875918,
      "grad_norm": 0.6247321963310242,
      "learning_rate": 1.352e-06,
      "loss": 1.1385,
      "step": 8100
    },
    {
      "epoch": 0.7257279405257102,
      "grad_norm": 0.7886361479759216,
      "learning_rate": 1.344e-06,
      "loss": 1.1857,
      "step": 8200
    },
    {
      "epoch": 0.7345782812638286,
      "grad_norm": 0.8461400270462036,
      "learning_rate": 1.336e-06,
      "loss": 1.1612,
      "step": 8300
    },
    {
      "epoch": 0.743428622001947,
      "grad_norm": 1.109844446182251,
      "learning_rate": 1.328e-06,
      "loss": 1.1406,
      "step": 8400
    },
    {
      "epoch": 0.7522789627400654,
      "grad_norm": 0.8707802891731262,
      "learning_rate": 1.32e-06,
      "loss": 1.1719,
      "step": 8500
    },
    {
      "epoch": 0.7611293034781839,
      "grad_norm": 0.8593458533287048,
      "learning_rate": 1.312e-06,
      "loss": 1.1696,
      "step": 8600
    },
    {
      "epoch": 0.7699796442163023,
      "grad_norm": 0.8872305750846863,
      "learning_rate": 1.304e-06,
      "loss": 1.1772,
      "step": 8700
    },
    {
      "epoch": 0.7788299849544208,
      "grad_norm": 0.9227966070175171,
      "learning_rate": 1.296e-06,
      "loss": 1.173,
      "step": 8800
    },
    {
      "epoch": 0.7876803256925392,
      "grad_norm": 0.8818754553794861,
      "learning_rate": 1.288e-06,
      "loss": 1.1889,
      "step": 8900
    },
    {
      "epoch": 0.7965306664306576,
      "grad_norm": 1.0658597946166992,
      "learning_rate": 1.28e-06,
      "loss": 1.1547,
      "step": 9000
    },
    {
      "epoch": 0.805381007168776,
      "grad_norm": 0.8553236126899719,
      "learning_rate": 1.272e-06,
      "loss": 1.1172,
      "step": 9100
    },
    {
      "epoch": 0.8142313479068944,
      "grad_norm": 1.120353102684021,
      "learning_rate": 1.2639999999999999e-06,
      "loss": 1.1649,
      "step": 9200
    },
    {
      "epoch": 0.8230816886450129,
      "grad_norm": 0.734362006187439,
      "learning_rate": 1.256e-06,
      "loss": 1.1938,
      "step": 9300
    },
    {
      "epoch": 0.8319320293831313,
      "grad_norm": 0.8664830923080444,
      "learning_rate": 1.248e-06,
      "loss": 1.1666,
      "step": 9400
    },
    {
      "epoch": 0.8407823701212497,
      "grad_norm": 2.742342948913574,
      "learning_rate": 1.24e-06,
      "loss": 1.1807,
      "step": 9500
    },
    {
      "epoch": 0.8496327108593681,
      "grad_norm": 1.302372694015503,
      "learning_rate": 1.232e-06,
      "loss": 1.1709,
      "step": 9600
    },
    {
      "epoch": 0.8584830515974865,
      "grad_norm": 0.9084349870681763,
      "learning_rate": 1.2239999999999998e-06,
      "loss": 1.1405,
      "step": 9700
    },
    {
      "epoch": 0.8673333923356049,
      "grad_norm": 0.8565220832824707,
      "learning_rate": 1.2159999999999999e-06,
      "loss": 1.1527,
      "step": 9800
    },
    {
      "epoch": 0.8761837330737233,
      "grad_norm": 0.7832551002502441,
      "learning_rate": 1.208e-06,
      "loss": 1.1787,
      "step": 9900
    },
    {
      "epoch": 0.8850340738118417,
      "grad_norm": 0.8366639018058777,
      "learning_rate": 1.2e-06,
      "loss": 1.1925,
      "step": 10000
    },
    {
      "epoch": 0.8938844145499601,
      "grad_norm": 1.0785573720932007,
      "learning_rate": 1.192e-06,
      "loss": 1.1515,
      "step": 10100
    },
    {
      "epoch": 0.9027347552880786,
      "grad_norm": 1.1639350652694702,
      "learning_rate": 1.1839999999999998e-06,
      "loss": 1.1742,
      "step": 10200
    },
    {
      "epoch": 0.911585096026197,
      "grad_norm": 1.7552626132965088,
      "learning_rate": 1.1759999999999998e-06,
      "loss": 1.1719,
      "step": 10300
    },
    {
      "epoch": 0.9204354367643154,
      "grad_norm": 1.1693079471588135,
      "learning_rate": 1.1679999999999999e-06,
      "loss": 1.1368,
      "step": 10400
    },
    {
      "epoch": 0.9292857775024338,
      "grad_norm": 1.0168983936309814,
      "learning_rate": 1.16e-06,
      "loss": 1.1883,
      "step": 10500
    },
    {
      "epoch": 0.9381361182405522,
      "grad_norm": 0.775669515132904,
      "learning_rate": 1.152e-06,
      "loss": 1.1538,
      "step": 10600
    },
    {
      "epoch": 0.9469864589786707,
      "grad_norm": 0.7753804922103882,
      "learning_rate": 1.1439999999999998e-06,
      "loss": 1.1433,
      "step": 10700
    },
    {
      "epoch": 0.9558367997167891,
      "grad_norm": 0.809089183807373,
      "learning_rate": 1.1359999999999998e-06,
      "loss": 1.1857,
      "step": 10800
    },
    {
      "epoch": 0.9646871404549076,
      "grad_norm": 1.0428608655929565,
      "learning_rate": 1.1279999999999998e-06,
      "loss": 1.1477,
      "step": 10900
    },
    {
      "epoch": 0.973537481193026,
      "grad_norm": 0.9207476377487183,
      "learning_rate": 1.12e-06,
      "loss": 1.1733,
      "step": 11000
    },
    {
      "epoch": 0.9823878219311444,
      "grad_norm": 1.4796310663223267,
      "learning_rate": 1.1120000000000001e-06,
      "loss": 1.1619,
      "step": 11100
    },
    {
      "epoch": 0.9912381626692628,
      "grad_norm": 0.9674447774887085,
      "learning_rate": 1.1040000000000001e-06,
      "loss": 1.1666,
      "step": 11200
    },
    {
      "epoch": 1.0000885034073812,
      "grad_norm": 0.8573771715164185,
      "learning_rate": 1.096e-06,
      "loss": 1.1413,
      "step": 11300
    },
    {
      "epoch": 1.0089388441454996,
      "grad_norm": 0.7278480529785156,
      "learning_rate": 1.088e-06,
      "loss": 1.1567,
      "step": 11400
    },
    {
      "epoch": 1.017789184883618,
      "grad_norm": 1.0488468408584595,
      "learning_rate": 1.08e-06,
      "loss": 1.1742,
      "step": 11500
    },
    {
      "epoch": 1.0266395256217364,
      "grad_norm": 0.9396333694458008,
      "learning_rate": 1.072e-06,
      "loss": 1.1516,
      "step": 11600
    },
    {
      "epoch": 1.0354898663598548,
      "grad_norm": 1.034722089767456,
      "learning_rate": 1.064e-06,
      "loss": 1.1485,
      "step": 11700
    },
    {
      "epoch": 1.0443402070979733,
      "grad_norm": 1.171837568283081,
      "learning_rate": 1.056e-06,
      "loss": 1.1531,
      "step": 11800
    },
    {
      "epoch": 1.0531905478360917,
      "grad_norm": 0.7825940847396851,
      "learning_rate": 1.048e-06,
      "loss": 1.1721,
      "step": 11900
    },
    {
      "epoch": 1.06204088857421,
      "grad_norm": 0.7846807837486267,
      "learning_rate": 1.04e-06,
      "loss": 1.1392,
      "step": 12000
    },
    {
      "epoch": 1.0708912293123285,
      "grad_norm": 1.3031669855117798,
      "learning_rate": 1.032e-06,
      "loss": 1.1797,
      "step": 12100
    },
    {
      "epoch": 1.079741570050447,
      "grad_norm": 0.7465079426765442,
      "learning_rate": 1.024e-06,
      "loss": 1.1756,
      "step": 12200
    },
    {
      "epoch": 1.0885919107885653,
      "grad_norm": 0.8680482506752014,
      "learning_rate": 1.0159999999999999e-06,
      "loss": 1.1584,
      "step": 12300
    },
    {
      "epoch": 1.0974422515266837,
      "grad_norm": 0.809716522693634,
      "learning_rate": 1.008e-06,
      "loss": 1.1791,
      "step": 12400
    },
    {
      "epoch": 1.1062925922648021,
      "grad_norm": 1.1684801578521729,
      "learning_rate": 1e-06,
      "loss": 1.1717,
      "step": 12500
    },
    {
      "epoch": 1.1151429330029206,
      "grad_norm": 0.9361375570297241,
      "learning_rate": 9.92e-07,
      "loss": 1.1436,
      "step": 12600
    },
    {
      "epoch": 1.123993273741039,
      "grad_norm": 1.0289329290390015,
      "learning_rate": 9.84e-07,
      "loss": 1.1497,
      "step": 12700
    },
    {
      "epoch": 1.1328436144791574,
      "grad_norm": 1.1782183647155762,
      "learning_rate": 9.759999999999998e-07,
      "loss": 1.1494,
      "step": 12800
    },
    {
      "epoch": 1.1416939552172758,
      "grad_norm": 1.1286753416061401,
      "learning_rate": 9.679999999999999e-07,
      "loss": 1.1423,
      "step": 12900
    },
    {
      "epoch": 1.1505442959553942,
      "grad_norm": 0.9268887639045715,
      "learning_rate": 9.6e-07,
      "loss": 1.1538,
      "step": 13000
    },
    {
      "epoch": 1.1593946366935126,
      "grad_norm": 0.9235917329788208,
      "learning_rate": 9.52e-07,
      "loss": 1.1815,
      "step": 13100
    },
    {
      "epoch": 1.168244977431631,
      "grad_norm": 0.9781365394592285,
      "learning_rate": 9.439999999999999e-07,
      "loss": 1.1706,
      "step": 13200
    },
    {
      "epoch": 1.1770953181697494,
      "grad_norm": 0.9572964310646057,
      "learning_rate": 9.36e-07,
      "loss": 1.1543,
      "step": 13300
    },
    {
      "epoch": 1.185945658907868,
      "grad_norm": 1.019444465637207,
      "learning_rate": 9.28e-07,
      "loss": 1.1272,
      "step": 13400
    },
    {
      "epoch": 1.1947959996459865,
      "grad_norm": 0.8430559039115906,
      "learning_rate": 9.2e-07,
      "loss": 1.1859,
      "step": 13500
    },
    {
      "epoch": 1.203646340384105,
      "grad_norm": 0.7822268009185791,
      "learning_rate": 9.12e-07,
      "loss": 1.148,
      "step": 13600
    },
    {
      "epoch": 1.2124966811222233,
      "grad_norm": 1.0175526142120361,
      "learning_rate": 9.039999999999999e-07,
      "loss": 1.1634,
      "step": 13700
    },
    {
      "epoch": 1.2213470218603417,
      "grad_norm": 1.0795433521270752,
      "learning_rate": 8.96e-07,
      "loss": 1.1476,
      "step": 13800
    },
    {
      "epoch": 1.2301973625984601,
      "grad_norm": 0.8210746645927429,
      "learning_rate": 8.88e-07,
      "loss": 1.1392,
      "step": 13900
    },
    {
      "epoch": 1.2390477033365785,
      "grad_norm": 0.9926521182060242,
      "learning_rate": 8.799999999999999e-07,
      "loss": 1.1729,
      "step": 14000
    },
    {
      "epoch": 1.247898044074697,
      "grad_norm": 0.968110978603363,
      "learning_rate": 8.72e-07,
      "loss": 1.1672,
      "step": 14100
    },
    {
      "epoch": 1.2567483848128154,
      "grad_norm": 0.7728704214096069,
      "learning_rate": 8.639999999999999e-07,
      "loss": 1.1542,
      "step": 14200
    },
    {
      "epoch": 1.2655987255509338,
      "grad_norm": 0.9193831086158752,
      "learning_rate": 8.559999999999999e-07,
      "loss": 1.1604,
      "step": 14300
    },
    {
      "epoch": 1.2744490662890522,
      "grad_norm": 0.9755778908729553,
      "learning_rate": 8.48e-07,
      "loss": 1.1698,
      "step": 14400
    },
    {
      "epoch": 1.2832994070271706,
      "grad_norm": 0.8752692341804504,
      "learning_rate": 8.399999999999999e-07,
      "loss": 1.1445,
      "step": 14500
    },
    {
      "epoch": 1.292149747765289,
      "grad_norm": 1.2579710483551025,
      "learning_rate": 8.319999999999999e-07,
      "loss": 1.1514,
      "step": 14600
    },
    {
      "epoch": 1.3010000885034074,
      "grad_norm": 0.8720273971557617,
      "learning_rate": 8.24e-07,
      "loss": 1.1835,
      "step": 14700
    },
    {
      "epoch": 1.3098504292415258,
      "grad_norm": 1.3184126615524292,
      "learning_rate": 8.159999999999999e-07,
      "loss": 1.1576,
      "step": 14800
    },
    {
      "epoch": 1.3187007699796442,
      "grad_norm": 1.0269840955734253,
      "learning_rate": 8.08e-07,
      "loss": 1.133,
      "step": 14900
    },
    {
      "epoch": 1.3275511107177627,
      "grad_norm": 1.114967942237854,
      "learning_rate": 8e-07,
      "loss": 1.1486,
      "step": 15000
    },
    {
      "epoch": 1.336401451455881,
      "grad_norm": 0.7998682260513306,
      "learning_rate": 7.92e-07,
      "loss": 1.161,
      "step": 15100
    },
    {
      "epoch": 1.3452517921939995,
      "grad_norm": 1.1405525207519531,
      "learning_rate": 7.84e-07,
      "loss": 1.161,
      "step": 15200
    },
    {
      "epoch": 1.354102132932118,
      "grad_norm": 0.8685086965560913,
      "learning_rate": 7.76e-07,
      "loss": 1.1438,
      "step": 15300
    },
    {
      "epoch": 1.3629524736702363,
      "grad_norm": 1.1836611032485962,
      "learning_rate": 7.68e-07,
      "loss": 1.1813,
      "step": 15400
    },
    {
      "epoch": 1.3718028144083547,
      "grad_norm": 1.2910044193267822,
      "learning_rate": 7.599999999999999e-07,
      "loss": 1.1477,
      "step": 15500
    },
    {
      "epoch": 1.3806531551464731,
      "grad_norm": 1.1127492189407349,
      "learning_rate": 7.52e-07,
      "loss": 1.1391,
      "step": 15600
    },
    {
      "epoch": 1.3895034958845915,
      "grad_norm": 1.0175690650939941,
      "learning_rate": 7.44e-07,
      "loss": 1.1613,
      "step": 15700
    },
    {
      "epoch": 1.39835383662271,
      "grad_norm": 1.0459816455841064,
      "learning_rate": 7.359999999999999e-07,
      "loss": 1.1479,
      "step": 15800
    },
    {
      "epoch": 1.4072041773608284,
      "grad_norm": 1.2192277908325195,
      "learning_rate": 7.28e-07,
      "loss": 1.1424,
      "step": 15900
    },
    {
      "epoch": 1.4160545180989468,
      "grad_norm": 1.0057787895202637,
      "learning_rate": 7.2e-07,
      "loss": 1.1452,
      "step": 16000
    },
    {
      "epoch": 1.4249048588370652,
      "grad_norm": 0.9277182221412659,
      "learning_rate": 7.119999999999999e-07,
      "loss": 1.1312,
      "step": 16100
    },
    {
      "epoch": 1.4337551995751836,
      "grad_norm": 1.6753654479980469,
      "learning_rate": 7.04e-07,
      "loss": 1.1622,
      "step": 16200
    },
    {
      "epoch": 1.442605540313302,
      "grad_norm": 1.0254045724868774,
      "learning_rate": 6.959999999999999e-07,
      "loss": 1.1605,
      "step": 16300
    },
    {
      "epoch": 1.4514558810514204,
      "grad_norm": 0.7054121494293213,
      "learning_rate": 6.879999999999999e-07,
      "loss": 1.1256,
      "step": 16400
    },
    {
      "epoch": 1.4603062217895388,
      "grad_norm": 1.7411776781082153,
      "learning_rate": 6.800000000000001e-07,
      "loss": 1.1628,
      "step": 16500
    },
    {
      "epoch": 1.4691565625276573,
      "grad_norm": 1.0013039112091064,
      "learning_rate": 6.72e-07,
      "loss": 1.1338,
      "step": 16600
    },
    {
      "epoch": 1.4780069032657757,
      "grad_norm": 1.111803650856018,
      "learning_rate": 6.64e-07,
      "loss": 1.1636,
      "step": 16700
    },
    {
      "epoch": 1.486857244003894,
      "grad_norm": 0.8137317895889282,
      "learning_rate": 6.56e-07,
      "loss": 1.1792,
      "step": 16800
    },
    {
      "epoch": 1.4957075847420125,
      "grad_norm": 1.5840669870376587,
      "learning_rate": 6.48e-07,
      "loss": 1.1615,
      "step": 16900
    },
    {
      "epoch": 1.504557925480131,
      "grad_norm": 0.8550547361373901,
      "learning_rate": 6.4e-07,
      "loss": 1.1632,
      "step": 17000
    },
    {
      "epoch": 1.5134082662182493,
      "grad_norm": 0.9156408309936523,
      "learning_rate": 6.319999999999999e-07,
      "loss": 1.1279,
      "step": 17100
    },
    {
      "epoch": 1.5222586069563677,
      "grad_norm": 1.0764663219451904,
      "learning_rate": 6.24e-07,
      "loss": 1.1249,
      "step": 17200
    },
    {
      "epoch": 1.5311089476944861,
      "grad_norm": 0.8155365586280823,
      "learning_rate": 6.16e-07,
      "loss": 1.1288,
      "step": 17300
    },
    {
      "epoch": 1.5399592884326045,
      "grad_norm": 1.1421250104904175,
      "learning_rate": 6.079999999999999e-07,
      "loss": 1.1768,
      "step": 17400
    },
    {
      "epoch": 1.548809629170723,
      "grad_norm": 1.4402973651885986,
      "learning_rate": 6e-07,
      "loss": 1.1719,
      "step": 17500
    },
    {
      "epoch": 1.5576599699088414,
      "grad_norm": 1.2016717195510864,
      "learning_rate": 5.919999999999999e-07,
      "loss": 1.1962,
      "step": 17600
    },
    {
      "epoch": 1.5665103106469598,
      "grad_norm": 1.2221474647521973,
      "learning_rate": 5.839999999999999e-07,
      "loss": 1.1949,
      "step": 17700
    },
    {
      "epoch": 1.5753606513850782,
      "grad_norm": 1.1164647340774536,
      "learning_rate": 5.76e-07,
      "loss": 1.165,
      "step": 17800
    },
    {
      "epoch": 1.5842109921231966,
      "grad_norm": 1.3106275796890259,
      "learning_rate": 5.679999999999999e-07,
      "loss": 1.1422,
      "step": 17900
    },
    {
      "epoch": 1.593061332861315,
      "grad_norm": 0.7900591492652893,
      "learning_rate": 5.6e-07,
      "loss": 1.1315,
      "step": 18000
    },
    {
      "epoch": 1.6019116735994334,
      "grad_norm": 0.8194898962974548,
      "learning_rate": 5.520000000000001e-07,
      "loss": 1.1442,
      "step": 18100
    },
    {
      "epoch": 1.6107620143375518,
      "grad_norm": 1.4453125,
      "learning_rate": 5.44e-07,
      "loss": 1.1772,
      "step": 18200
    },
    {
      "epoch": 1.6196123550756703,
      "grad_norm": 1.2650878429412842,
      "learning_rate": 5.36e-07,
      "loss": 1.1592,
      "step": 18300
    },
    {
      "epoch": 1.6284626958137889,
      "grad_norm": 1.273959755897522,
      "learning_rate": 5.28e-07,
      "loss": 1.1608,
      "step": 18400
    },
    {
      "epoch": 1.6373130365519073,
      "grad_norm": 0.9476118087768555,
      "learning_rate": 5.2e-07,
      "loss": 1.1736,
      "step": 18500
    },
    {
      "epoch": 1.6461633772900257,
      "grad_norm": 0.8563472032546997,
      "learning_rate": 5.12e-07,
      "loss": 1.1609,
      "step": 18600
    },
    {
      "epoch": 1.6550137180281441,
      "grad_norm": 0.9104204773902893,
      "learning_rate": 5.04e-07,
      "loss": 1.1582,
      "step": 18700
    },
    {
      "epoch": 1.6638640587662625,
      "grad_norm": 1.0252381563186646,
      "learning_rate": 4.96e-07,
      "loss": 1.161,
      "step": 18800
    },
    {
      "epoch": 1.672714399504381,
      "grad_norm": 1.0324159860610962,
      "learning_rate": 4.879999999999999e-07,
      "loss": 1.1629,
      "step": 18900
    },
    {
      "epoch": 1.6815647402424994,
      "grad_norm": 0.901113748550415,
      "learning_rate": 4.8e-07,
      "loss": 1.196,
      "step": 19000
    },
    {
      "epoch": 1.6904150809806178,
      "grad_norm": 0.9018805623054504,
      "learning_rate": 4.7199999999999994e-07,
      "loss": 1.1459,
      "step": 19100
    },
    {
      "epoch": 1.6992654217187362,
      "grad_norm": 0.9644899964332581,
      "learning_rate": 4.64e-07,
      "loss": 1.1395,
      "step": 19200
    },
    {
      "epoch": 1.7081157624568546,
      "grad_norm": 0.940774142742157,
      "learning_rate": 4.56e-07,
      "loss": 1.1608,
      "step": 19300
    },
    {
      "epoch": 1.716966103194973,
      "grad_norm": 0.9524922370910645,
      "learning_rate": 4.48e-07,
      "loss": 1.1544,
      "step": 19400
    },
    {
      "epoch": 1.7258164439330914,
      "grad_norm": 1.1308525800704956,
      "learning_rate": 4.3999999999999997e-07,
      "loss": 1.1622,
      "step": 19500
    },
    {
      "epoch": 1.7346667846712098,
      "grad_norm": 1.1476516723632812,
      "learning_rate": 4.3199999999999995e-07,
      "loss": 1.1431,
      "step": 19600
    },
    {
      "epoch": 1.7435171254093282,
      "grad_norm": 0.7359842658042908,
      "learning_rate": 4.24e-07,
      "loss": 1.1522,
      "step": 19700
    },
    {
      "epoch": 1.7523674661474467,
      "grad_norm": 1.2771000862121582,
      "learning_rate": 4.1599999999999997e-07,
      "loss": 1.1637,
      "step": 19800
    },
    {
      "epoch": 1.761217806885565,
      "grad_norm": 0.7793694734573364,
      "learning_rate": 4.0799999999999995e-07,
      "loss": 1.1678,
      "step": 19900
    },
    {
      "epoch": 1.7700681476236835,
      "grad_norm": 1.088559865951538,
      "learning_rate": 4e-07,
      "loss": 1.1394,
      "step": 20000
    },
    {
      "epoch": 1.7789184883618019,
      "grad_norm": 1.0532058477401733,
      "learning_rate": 3.92e-07,
      "loss": 1.147,
      "step": 20100
    },
    {
      "epoch": 1.7877688290999203,
      "grad_norm": 1.7931568622589111,
      "learning_rate": 3.84e-07,
      "loss": 1.1346,
      "step": 20200
    },
    {
      "epoch": 1.7966191698380387,
      "grad_norm": 0.8389997482299805,
      "learning_rate": 3.76e-07,
      "loss": 1.1607,
      "step": 20300
    },
    {
      "epoch": 1.8054695105761573,
      "grad_norm": 0.7847001552581787,
      "learning_rate": 3.6799999999999996e-07,
      "loss": 1.1491,
      "step": 20400
    },
    {
      "epoch": 1.8143198513142758,
      "grad_norm": 1.1962199211120605,
      "learning_rate": 3.6e-07,
      "loss": 1.1521,
      "step": 20500
    },
    {
      "epoch": 1.8231701920523942,
      "grad_norm": 0.910849392414093,
      "learning_rate": 3.52e-07,
      "loss": 1.1872,
      "step": 20600
    },
    {
      "epoch": 1.8320205327905126,
      "grad_norm": 1.1673359870910645,
      "learning_rate": 3.4399999999999996e-07,
      "loss": 1.1706,
      "step": 20700
    },
    {
      "epoch": 1.840870873528631,
      "grad_norm": 0.889744758605957,
      "learning_rate": 3.36e-07,
      "loss": 1.1501,
      "step": 20800
    },
    {
      "epoch": 1.8497212142667494,
      "grad_norm": 1.2205482721328735,
      "learning_rate": 3.28e-07,
      "loss": 1.1644,
      "step": 20900
    },
    {
      "epoch": 1.8585715550048678,
      "grad_norm": 1.288896083831787,
      "learning_rate": 3.2e-07,
      "loss": 1.1638,
      "step": 21000
    },
    {
      "epoch": 1.8674218957429862,
      "grad_norm": 0.7837293148040771,
      "learning_rate": 3.12e-07,
      "loss": 1.142,
      "step": 21100
    },
    {
      "epoch": 1.8762722364811046,
      "grad_norm": 1.148685097694397,
      "learning_rate": 3.0399999999999997e-07,
      "loss": 1.1692,
      "step": 21200
    },
    {
      "epoch": 1.885122577219223,
      "grad_norm": 1.1871728897094727,
      "learning_rate": 2.9599999999999995e-07,
      "loss": 1.1607,
      "step": 21300
    },
    {
      "epoch": 1.8939729179573415,
      "grad_norm": 1.3889893293380737,
      "learning_rate": 2.88e-07,
      "loss": 1.1653,
      "step": 21400
    },
    {
      "epoch": 1.9028232586954599,
      "grad_norm": 1.136010766029358,
      "learning_rate": 2.8e-07,
      "loss": 1.1792,
      "step": 21500
    },
    {
      "epoch": 1.9116735994335783,
      "grad_norm": 1.1500000953674316,
      "learning_rate": 2.72e-07,
      "loss": 1.1509,
      "step": 21600
    },
    {
      "epoch": 1.9205239401716967,
      "grad_norm": 0.9688245058059692,
      "learning_rate": 2.64e-07,
      "loss": 1.1372,
      "step": 21700
    },
    {
      "epoch": 1.929374280909815,
      "grad_norm": 1.2906290292739868,
      "learning_rate": 2.56e-07,
      "loss": 1.1604,
      "step": 21800
    },
    {
      "epoch": 1.9382246216479335,
      "grad_norm": 0.7762517333030701,
      "learning_rate": 2.48e-07,
      "loss": 1.1765,
      "step": 21900
    },
    {
      "epoch": 1.947074962386052,
      "grad_norm": 1.6632424592971802,
      "learning_rate": 2.4e-07,
      "loss": 1.155,
      "step": 22000
    },
    {
      "epoch": 1.9559253031241703,
      "grad_norm": 0.9481187462806702,
      "learning_rate": 2.32e-07,
      "loss": 1.1496,
      "step": 22100
    },
    {
      "epoch": 1.9647756438622888,
      "grad_norm": 0.9449306130409241,
      "learning_rate": 2.24e-07,
      "loss": 1.1482,
      "step": 22200
    },
    {
      "epoch": 1.9736259846004072,
      "grad_norm": 0.9778609275817871,
      "learning_rate": 2.1599999999999998e-07,
      "loss": 1.1549,
      "step": 22300
    },
    {
      "epoch": 1.9824763253385256,
      "grad_norm": 1.0118569135665894,
      "learning_rate": 2.0799999999999998e-07,
      "loss": 1.1338,
      "step": 22400
    },
    {
      "epoch": 1.991326666076644,
      "grad_norm": 1.3175530433654785,
      "learning_rate": 2e-07,
      "loss": 1.1474,
      "step": 22500
    },
    {
      "epoch": 2.0001770068147624,
      "grad_norm": 0.9164048433303833,
      "learning_rate": 1.92e-07,
      "loss": 1.1343,
      "step": 22600
    },
    {
      "epoch": 2.009027347552881,
      "grad_norm": 1.1534018516540527,
      "learning_rate": 1.8399999999999998e-07,
      "loss": 1.1582,
      "step": 22700
    },
    {
      "epoch": 2.0178776882909992,
      "grad_norm": 0.7509279251098633,
      "learning_rate": 1.76e-07,
      "loss": 1.1517,
      "step": 22800
    },
    {
      "epoch": 2.0267280290291176,
      "grad_norm": 1.0024540424346924,
      "learning_rate": 1.68e-07,
      "loss": 1.155,
      "step": 22900
    },
    {
      "epoch": 2.035578369767236,
      "grad_norm": 1.0684221982955933,
      "learning_rate": 1.6e-07,
      "loss": 1.1751,
      "step": 23000
    },
    {
      "epoch": 2.0444287105053545,
      "grad_norm": 1.075598955154419,
      "learning_rate": 1.5199999999999998e-07,
      "loss": 1.154,
      "step": 23100
    },
    {
      "epoch": 2.053279051243473,
      "grad_norm": 0.9409734010696411,
      "learning_rate": 1.44e-07,
      "loss": 1.1438,
      "step": 23200
    },
    {
      "epoch": 2.0621293919815913,
      "grad_norm": 0.893460214138031,
      "learning_rate": 1.36e-07,
      "loss": 1.1881,
      "step": 23300
    },
    {
      "epoch": 2.0709797327197097,
      "grad_norm": 1.0118149518966675,
      "learning_rate": 1.28e-07,
      "loss": 1.1432,
      "step": 23400
    },
    {
      "epoch": 2.079830073457828,
      "grad_norm": 1.147308588027954,
      "learning_rate": 1.2e-07,
      "loss": 1.1339,
      "step": 23500
    },
    {
      "epoch": 2.0886804141959465,
      "grad_norm": 1.020508050918579,
      "learning_rate": 1.12e-07,
      "loss": 1.1547,
      "step": 23600
    },
    {
      "epoch": 2.097530754934065,
      "grad_norm": 0.9823341369628906,
      "learning_rate": 1.0399999999999999e-07,
      "loss": 1.1561,
      "step": 23700
    },
    {
      "epoch": 2.1063810956721833,
      "grad_norm": 0.832475483417511,
      "learning_rate": 9.6e-08,
      "loss": 1.1678,
      "step": 23800
    },
    {
      "epoch": 2.1152314364103018,
      "grad_norm": 0.8791053891181946,
      "learning_rate": 8.8e-08,
      "loss": 1.1792,
      "step": 23900
    },
    {
      "epoch": 2.12408177714842,
      "grad_norm": 0.8894505500793457,
      "learning_rate": 8e-08,
      "loss": 1.1602,
      "step": 24000
    },
    {
      "epoch": 2.1329321178865386,
      "grad_norm": 0.8826361298561096,
      "learning_rate": 7.2e-08,
      "loss": 1.1366,
      "step": 24100
    },
    {
      "epoch": 2.141782458624657,
      "grad_norm": 1.2695685625076294,
      "learning_rate": 6.4e-08,
      "loss": 1.1673,
      "step": 24200
    },
    {
      "epoch": 2.1506327993627754,
      "grad_norm": 1.1635528802871704,
      "learning_rate": 5.6e-08,
      "loss": 1.1615,
      "step": 24300
    },
    {
      "epoch": 2.159483140100894,
      "grad_norm": 0.919233500957489,
      "learning_rate": 4.8e-08,
      "loss": 1.1661,
      "step": 24400
    },
    {
      "epoch": 2.1683334808390122,
      "grad_norm": 0.945484459400177,
      "learning_rate": 4e-08,
      "loss": 1.1396,
      "step": 24500
    },
    {
      "epoch": 2.1771838215771306,
      "grad_norm": 0.9278846979141235,
      "learning_rate": 3.2e-08,
      "loss": 1.1462,
      "step": 24600
    },
    {
      "epoch": 2.186034162315249,
      "grad_norm": 1.0193525552749634,
      "learning_rate": 2.4e-08,
      "loss": 1.1288,
      "step": 24700
    },
    {
      "epoch": 2.1948845030533675,
      "grad_norm": 1.0619391202926636,
      "learning_rate": 1.6e-08,
      "loss": 1.171,
      "step": 24800
    },
    {
      "epoch": 2.203734843791486,
      "grad_norm": 0.9740051627159119,
      "learning_rate": 8e-09,
      "loss": 1.1333,
      "step": 24900
    },
    {
      "epoch": 2.2125851845296043,
      "grad_norm": 1.6167223453521729,
      "learning_rate": 0.0,
      "loss": 1.1462,
      "step": 25000
    }
  ],
  "logging_steps": 100,
  "max_steps": 25000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 5000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.464490326009037e+19,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}