{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 105393,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.014232444279980643,
      "grad_norm": 125.48007202148438,
      "learning_rate": 4.976279259533366e-05,
      "loss": 1.262,
      "step": 500
    },
    {
      "epoch": 0.028464888559961286,
      "grad_norm": 3.935548782348633,
      "learning_rate": 4.952558519066731e-05,
      "loss": 1.3135,
      "step": 1000
    },
    {
      "epoch": 0.04269733283994193,
      "grad_norm": 432.8229675292969,
      "learning_rate": 4.9288377786000975e-05,
      "loss": 1.3926,
      "step": 1500
    },
    {
      "epoch": 0.05692977711992257,
      "grad_norm": 44.97547149658203,
      "learning_rate": 4.9051170381334626e-05,
      "loss": 1.5597,
      "step": 2000
    },
    {
      "epoch": 0.07116222139990322,
      "grad_norm": 26.80133056640625,
      "learning_rate": 4.8813962976668284e-05,
      "loss": 1.4146,
      "step": 2500
    },
    {
      "epoch": 0.08539466567988387,
      "grad_norm": 86.9864273071289,
      "learning_rate": 4.8576755572001934e-05,
      "loss": 1.4553,
      "step": 3000
    },
    {
      "epoch": 0.09962710995986451,
      "grad_norm": 1.6426622867584229,
      "learning_rate": 4.83395481673356e-05,
      "loss": 1.2304,
      "step": 3500
    },
    {
      "epoch": 0.11385955423984515,
      "grad_norm": 280.5526428222656,
      "learning_rate": 4.810234076266925e-05,
      "loss": 1.2387,
      "step": 4000
    },
    {
      "epoch": 0.1280919985198258,
      "grad_norm": 19.09530258178711,
      "learning_rate": 4.786513335800291e-05,
      "loss": 1.5957,
      "step": 4500
    },
    {
      "epoch": 0.14232444279980644,
      "grad_norm": 8.927694320678711,
      "learning_rate": 4.762792595333656e-05,
      "loss": 1.3537,
      "step": 5000
    },
    {
      "epoch": 0.15655688707978707,
      "grad_norm": 214.30294799804688,
      "learning_rate": 4.7390718548670216e-05,
      "loss": 1.4233,
      "step": 5500
    },
    {
      "epoch": 0.17078933135976773,
      "grad_norm": 7.38736629486084,
      "learning_rate": 4.715351114400387e-05,
      "loss": 1.3794,
      "step": 6000
    },
    {
      "epoch": 0.18502177563974836,
      "grad_norm": 153.38523864746094,
      "learning_rate": 4.691630373933753e-05,
      "loss": 1.6875,
      "step": 6500
    },
    {
      "epoch": 0.19925421991972903,
      "grad_norm": 0.02275083400309086,
      "learning_rate": 4.667909633467118e-05,
      "loss": 1.4558,
      "step": 7000
    },
    {
      "epoch": 0.21348666419970966,
      "grad_norm": 320.4385070800781,
      "learning_rate": 4.644188893000484e-05,
      "loss": 1.3569,
      "step": 7500
    },
    {
      "epoch": 0.2277191084796903,
      "grad_norm": 3.4960036277770996,
      "learning_rate": 4.62046815253385e-05,
      "loss": 1.5315,
      "step": 8000
    },
    {
      "epoch": 0.24195155275967095,
      "grad_norm": 37.32798767089844,
      "learning_rate": 4.5967474120672155e-05,
      "loss": 1.4474,
      "step": 8500
    },
    {
      "epoch": 0.2561839970396516,
      "grad_norm": 46.48209762573242,
      "learning_rate": 4.573026671600581e-05,
      "loss": 1.3646,
      "step": 9000
    },
    {
      "epoch": 0.2704164413196322,
      "grad_norm": 0.00015533728583250195,
      "learning_rate": 4.549305931133946e-05,
      "loss": 1.1472,
      "step": 9500
    },
    {
      "epoch": 0.2846488855996129,
      "grad_norm": 0.002173429122194648,
      "learning_rate": 4.525585190667312e-05,
      "loss": 1.5633,
      "step": 10000
    },
    {
      "epoch": 0.29888132987959354,
      "grad_norm": 186.76336669921875,
      "learning_rate": 4.501864450200678e-05,
      "loss": 1.3533,
      "step": 10500
    },
    {
      "epoch": 0.31311377415957414,
      "grad_norm": 10.35462474822998,
      "learning_rate": 4.4781437097340436e-05,
      "loss": 1.4215,
      "step": 11000
    },
    {
      "epoch": 0.3273462184395548,
      "grad_norm": 0.01028844341635704,
      "learning_rate": 4.454422969267409e-05,
      "loss": 1.3768,
      "step": 11500
    },
    {
      "epoch": 0.34157866271953546,
      "grad_norm": 0.011700475588440895,
      "learning_rate": 4.4307022288007744e-05,
      "loss": 1.3978,
      "step": 12000
    },
    {
      "epoch": 0.3558111069995161,
      "grad_norm": 487.6321105957031,
      "learning_rate": 4.4069814883341395e-05,
      "loss": 1.3882,
      "step": 12500
    },
    {
      "epoch": 0.37004355127949673,
      "grad_norm": 7.461075782775879,
      "learning_rate": 4.383260747867506e-05,
      "loss": 1.4489,
      "step": 13000
    },
    {
      "epoch": 0.3842759955594774,
      "grad_norm": 0.00012495643750298768,
      "learning_rate": 4.359540007400871e-05,
      "loss": 1.413,
      "step": 13500
    },
    {
      "epoch": 0.39850843983945805,
      "grad_norm": 70.00542449951172,
      "learning_rate": 4.335819266934237e-05,
      "loss": 1.3233,
      "step": 14000
    },
    {
      "epoch": 0.41274088411943866,
      "grad_norm": 167.84930419921875,
      "learning_rate": 4.3120985264676026e-05,
      "loss": 1.425,
      "step": 14500
    },
    {
      "epoch": 0.4269733283994193,
      "grad_norm": 285.5869445800781,
      "learning_rate": 4.2883777860009676e-05,
      "loss": 1.4104,
      "step": 15000
    },
    {
      "epoch": 0.4412057726794,
      "grad_norm": 99.11154174804688,
      "learning_rate": 4.264657045534334e-05,
      "loss": 1.3,
      "step": 15500
    },
    {
      "epoch": 0.4554382169593806,
      "grad_norm": 9.479164145886898e-05,
      "learning_rate": 4.240936305067699e-05,
      "loss": 1.2376,
      "step": 16000
    },
    {
      "epoch": 0.46967066123936124,
      "grad_norm": 33.981483459472656,
      "learning_rate": 4.217215564601065e-05,
      "loss": 1.5736,
      "step": 16500
    },
    {
      "epoch": 0.4839031055193419,
      "grad_norm": 2.699108839035034,
      "learning_rate": 4.19349482413443e-05,
      "loss": 1.4016,
      "step": 17000
    },
    {
      "epoch": 0.49813554979932256,
      "grad_norm": 0.3147715628147125,
      "learning_rate": 4.1697740836677964e-05,
      "loss": 1.3702,
      "step": 17500
    },
    {
      "epoch": 0.5123679940793032,
      "grad_norm": 0.14605668187141418,
      "learning_rate": 4.1460533432011615e-05,
      "loss": 1.3285,
      "step": 18000
    },
    {
      "epoch": 0.5266004383592838,
      "grad_norm": 0.1937469094991684,
      "learning_rate": 4.122332602734527e-05,
      "loss": 1.348,
      "step": 18500
    },
    {
      "epoch": 0.5408328826392644,
      "grad_norm": 89.56853485107422,
      "learning_rate": 4.0986118622678924e-05,
      "loss": 1.482,
      "step": 19000
    },
    {
      "epoch": 0.5550653269192451,
      "grad_norm": 0.07370008528232574,
      "learning_rate": 4.074891121801258e-05,
      "loss": 1.2254,
      "step": 19500
    },
    {
      "epoch": 0.5692977711992258,
      "grad_norm": 199.52772521972656,
      "learning_rate": 4.051170381334624e-05,
      "loss": 1.4102,
      "step": 20000
    },
    {
      "epoch": 0.5835302154792064,
      "grad_norm": 45.664859771728516,
      "learning_rate": 4.0274496408679896e-05,
      "loss": 1.3631,
      "step": 20500
    },
    {
      "epoch": 0.5977626597591871,
      "grad_norm": 3.712595798788243e-06,
      "learning_rate": 4.003728900401355e-05,
      "loss": 1.3419,
      "step": 21000
    },
    {
      "epoch": 0.6119951040391677,
      "grad_norm": 8.145997708197683e-05,
      "learning_rate": 3.9800081599347205e-05,
      "loss": 1.3845,
      "step": 21500
    },
    {
      "epoch": 0.6262275483191483,
      "grad_norm": 248.4174346923828,
      "learning_rate": 3.956287419468086e-05,
      "loss": 1.1569,
      "step": 22000
    },
    {
      "epoch": 0.640459992599129,
      "grad_norm": 3.249052679166198e-05,
      "learning_rate": 3.932566679001452e-05,
      "loss": 1.306,
      "step": 22500
    },
    {
      "epoch": 0.6546924368791096,
      "grad_norm": 3.5558118725020904e-06,
      "learning_rate": 3.908845938534818e-05,
      "loss": 1.4762,
      "step": 23000
    },
    {
      "epoch": 0.6689248811590902,
      "grad_norm": 4.918412208557129,
      "learning_rate": 3.885125198068183e-05,
      "loss": 1.2786,
      "step": 23500
    },
    {
      "epoch": 0.6831573254390709,
      "grad_norm": 4.046954154968262,
      "learning_rate": 3.8614044576015486e-05,
      "loss": 1.4232,
      "step": 24000
    },
    {
      "epoch": 0.6973897697190515,
      "grad_norm": 139.35830688476562,
      "learning_rate": 3.8376837171349144e-05,
      "loss": 1.2306,
      "step": 24500
    },
    {
      "epoch": 0.7116222139990322,
      "grad_norm": 0.0009823115542531013,
      "learning_rate": 3.81396297666828e-05,
      "loss": 1.2463,
      "step": 25000
    },
    {
      "epoch": 0.7258546582790129,
      "grad_norm": 2.3099827766418457,
      "learning_rate": 3.790242236201645e-05,
      "loss": 1.4433,
      "step": 25500
    },
    {
      "epoch": 0.7400871025589935,
      "grad_norm": 137.370361328125,
      "learning_rate": 3.766521495735011e-05,
      "loss": 1.3623,
      "step": 26000
    },
    {
      "epoch": 0.7543195468389742,
      "grad_norm": 0.5372253656387329,
      "learning_rate": 3.742800755268376e-05,
      "loss": 1.432,
      "step": 26500
    },
    {
      "epoch": 0.7685519911189548,
      "grad_norm": 3.4454299679964606e-07,
      "learning_rate": 3.7190800148017425e-05,
      "loss": 1.4002,
      "step": 27000
    },
    {
      "epoch": 0.7827844353989354,
      "grad_norm": 0.3542407155036926,
      "learning_rate": 3.6953592743351076e-05,
      "loss": 1.3797,
      "step": 27500
    },
    {
      "epoch": 0.7970168796789161,
      "grad_norm": 0.013401851058006287,
      "learning_rate": 3.6716385338684733e-05,
      "loss": 1.2473,
      "step": 28000
    },
    {
      "epoch": 0.8112493239588967,
      "grad_norm": 251.7378692626953,
      "learning_rate": 3.647917793401839e-05,
      "loss": 1.2836,
      "step": 28500
    },
    {
      "epoch": 0.8254817682388773,
      "grad_norm": 0.6462361216545105,
      "learning_rate": 3.624197052935205e-05,
      "loss": 1.3998,
      "step": 29000
    },
    {
      "epoch": 0.839714212518858,
      "grad_norm": 0.006176416762173176,
      "learning_rate": 3.6004763124685706e-05,
      "loss": 1.2388,
      "step": 29500
    },
    {
      "epoch": 0.8539466567988386,
      "grad_norm": 0.6315366625785828,
      "learning_rate": 3.576755572001936e-05,
      "loss": 1.3803,
      "step": 30000
    },
    {
      "epoch": 0.8681791010788192,
      "grad_norm": 0.3394842743873596,
      "learning_rate": 3.5530348315353015e-05,
      "loss": 1.3189,
      "step": 30500
    },
    {
      "epoch": 0.8824115453588,
      "grad_norm": 64.10663604736328,
      "learning_rate": 3.5293140910686666e-05,
      "loss": 1.3589,
      "step": 31000
    },
    {
      "epoch": 0.8966439896387806,
      "grad_norm": 0.08269359171390533,
      "learning_rate": 3.505593350602033e-05,
      "loss": 1.2997,
      "step": 31500
    },
    {
      "epoch": 0.9108764339187612,
      "grad_norm": 101.3696060180664,
      "learning_rate": 3.481872610135398e-05,
      "loss": 1.3141,
      "step": 32000
    },
    {
      "epoch": 0.9251088781987419,
      "grad_norm": 84.51553344726562,
      "learning_rate": 3.458151869668764e-05,
      "loss": 1.2003,
      "step": 32500
    },
    {
      "epoch": 0.9393413224787225,
      "grad_norm": 17.5670108795166,
      "learning_rate": 3.434431129202129e-05,
      "loss": 1.2641,
      "step": 33000
    },
    {
      "epoch": 0.9535737667587031,
      "grad_norm": 1.7186918258666992,
      "learning_rate": 3.4107103887354954e-05,
      "loss": 1.3127,
      "step": 33500
    },
    {
      "epoch": 0.9678062110386838,
      "grad_norm": 173.8558807373047,
      "learning_rate": 3.3869896482688604e-05,
      "loss": 1.2033,
      "step": 34000
    },
    {
      "epoch": 0.9820386553186644,
      "grad_norm": 0.0010143565014004707,
      "learning_rate": 3.363268907802226e-05,
      "loss": 1.3354,
      "step": 34500
    },
    {
      "epoch": 0.9962710995986451,
      "grad_norm": 0.00039917067624628544,
      "learning_rate": 3.339548167335591e-05,
      "loss": 1.2617,
      "step": 35000
    },
    {
      "epoch": 1.0105035438786256,
      "grad_norm": 327.86859130859375,
      "learning_rate": 3.315827426868957e-05,
      "loss": 1.1885,
      "step": 35500
    },
    {
      "epoch": 1.0247359881586064,
      "grad_norm": 244.52615356445312,
      "learning_rate": 3.292106686402323e-05,
      "loss": 1.2093,
      "step": 36000
    },
    {
      "epoch": 1.038968432438587,
      "grad_norm": 7.460901997546898e-06,
      "learning_rate": 3.2683859459356886e-05,
      "loss": 1.5839,
      "step": 36500
    },
    {
      "epoch": 1.0532008767185677,
      "grad_norm": 1.3049445152282715,
      "learning_rate": 3.244665205469054e-05,
      "loss": 1.2053,
      "step": 37000
    },
    {
      "epoch": 1.0674333209985483,
      "grad_norm": 3.011463718394225e-07,
      "learning_rate": 3.2209444650024194e-05,
      "loss": 1.2656,
      "step": 37500
    },
    {
      "epoch": 1.0816657652785289,
      "grad_norm": 3.12375807762146,
      "learning_rate": 3.197223724535785e-05,
      "loss": 1.281,
      "step": 38000
    },
    {
      "epoch": 1.0958982095585097,
      "grad_norm": 0.007164264563471079,
      "learning_rate": 3.173502984069151e-05,
      "loss": 1.2791,
      "step": 38500
    },
    {
      "epoch": 1.1101306538384903,
      "grad_norm": 1.040968345478177e-05,
      "learning_rate": 3.149782243602517e-05,
      "loss": 1.3826,
      "step": 39000
    },
    {
      "epoch": 1.124363098118471,
      "grad_norm": 0.4420815110206604,
      "learning_rate": 3.126061503135882e-05,
      "loss": 1.3168,
      "step": 39500
    },
    {
      "epoch": 1.1385955423984515,
      "grad_norm": 212.03660583496094,
      "learning_rate": 3.1023407626692475e-05,
      "loss": 1.2582,
      "step": 40000
    },
    {
      "epoch": 1.1528279866784321,
      "grad_norm": 90.19596099853516,
      "learning_rate": 3.078620022202613e-05,
      "loss": 1.301,
      "step": 40500
    },
    {
      "epoch": 1.1670604309584127,
      "grad_norm": 2.8100829124450684,
      "learning_rate": 3.054899281735979e-05,
      "loss": 1.1346,
      "step": 41000
    },
    {
      "epoch": 1.1812928752383933,
      "grad_norm": 218.48757934570312,
      "learning_rate": 3.0311785412693445e-05,
      "loss": 1.2074,
      "step": 41500
    },
    {
      "epoch": 1.1955253195183742,
      "grad_norm": 157.1456298828125,
      "learning_rate": 3.00745780080271e-05,
      "loss": 1.1437,
      "step": 42000
    },
    {
      "epoch": 1.2097577637983548,
      "grad_norm": 0.2799847722053528,
      "learning_rate": 2.9837370603360753e-05,
      "loss": 1.2164,
      "step": 42500
    },
    {
      "epoch": 1.2239902080783354,
      "grad_norm": 4.315948963165283,
      "learning_rate": 2.9600163198694414e-05,
      "loss": 1.1489,
      "step": 43000
    },
    {
      "epoch": 1.238222652358316,
      "grad_norm": 8.09383487701416,
      "learning_rate": 2.936295579402807e-05,
      "loss": 1.2479,
      "step": 43500
    },
    {
      "epoch": 1.2524550966382968,
      "grad_norm": 357.42626953125,
      "learning_rate": 2.9125748389361723e-05,
      "loss": 1.1489,
      "step": 44000
    },
    {
      "epoch": 1.2666875409182774,
      "grad_norm": 0.011236801743507385,
      "learning_rate": 2.8888540984695377e-05,
      "loss": 1.2963,
      "step": 44500
    },
    {
      "epoch": 1.280919985198258,
      "grad_norm": 356.8849792480469,
      "learning_rate": 2.8651333580029034e-05,
      "loss": 1.2021,
      "step": 45000
    },
    {
      "epoch": 1.2951524294782386,
      "grad_norm": 1.403478741645813,
      "learning_rate": 2.8414126175362692e-05,
      "loss": 1.2875,
      "step": 45500
    },
    {
      "epoch": 1.3093848737582192,
      "grad_norm": 0.10168521851301193,
      "learning_rate": 2.8176918770696346e-05,
      "loss": 1.1594,
      "step": 46000
    },
    {
      "epoch": 1.3236173180381998,
      "grad_norm": 0.0002031529729720205,
      "learning_rate": 2.7939711366030004e-05,
      "loss": 1.1856,
      "step": 46500
    },
    {
      "epoch": 1.3378497623181804,
      "grad_norm": 0.19232670962810516,
      "learning_rate": 2.7702503961363658e-05,
      "loss": 1.3044,
      "step": 47000
    },
    {
      "epoch": 1.3520822065981613,
      "grad_norm": 3.5715222358703613,
      "learning_rate": 2.746529655669732e-05,
      "loss": 1.1655,
      "step": 47500
    },
    {
      "epoch": 1.3663146508781419,
      "grad_norm": 187.40589904785156,
      "learning_rate": 2.7228089152030973e-05,
      "loss": 1.2214,
      "step": 48000
    },
    {
      "epoch": 1.3805470951581225,
      "grad_norm": 0.8012977242469788,
      "learning_rate": 2.6990881747364628e-05,
      "loss": 1.2209,
      "step": 48500
    },
    {
      "epoch": 1.394779539438103,
      "grad_norm": 57.39756774902344,
      "learning_rate": 2.6753674342698282e-05,
      "loss": 1.3786,
      "step": 49000
    },
    {
      "epoch": 1.4090119837180837,
      "grad_norm": 7.3047776222229,
      "learning_rate": 2.6516466938031936e-05,
      "loss": 1.3933,
      "step": 49500
    },
    {
      "epoch": 1.4232444279980645,
      "grad_norm": 510.1078796386719,
      "learning_rate": 2.6279259533365597e-05,
      "loss": 1.3421,
      "step": 50000
    },
    {
      "epoch": 1.437476872278045,
      "grad_norm": 0.0005555571406148374,
      "learning_rate": 2.604205212869925e-05,
      "loss": 1.4459,
      "step": 50500
    },
    {
      "epoch": 1.4517093165580257,
      "grad_norm": 357.9839172363281,
      "learning_rate": 2.5804844724032905e-05,
      "loss": 1.5949,
      "step": 51000
    },
    {
      "epoch": 1.4659417608380063,
      "grad_norm": 10.876310348510742,
      "learning_rate": 2.556763731936656e-05,
      "loss": 1.5693,
      "step": 51500
    },
    {
      "epoch": 1.480174205117987,
      "grad_norm": 0.9166773557662964,
      "learning_rate": 2.533042991470022e-05,
      "loss": 1.5855,
      "step": 52000
    },
    {
      "epoch": 1.4944066493979675,
      "grad_norm": 0.0008015409694053233,
      "learning_rate": 2.5093222510033875e-05,
      "loss": 1.5393,
      "step": 52500
    },
    {
      "epoch": 1.5086390936779481,
      "grad_norm": 565.9046630859375,
      "learning_rate": 2.485601510536753e-05,
      "loss": 1.4058,
      "step": 53000
    },
    {
      "epoch": 1.5228715379579287,
      "grad_norm": 0.002393967006355524,
      "learning_rate": 2.4618807700701187e-05,
      "loss": 1.5389,
      "step": 53500
    },
    {
      "epoch": 1.5371039822379096,
      "grad_norm": 396.0699157714844,
      "learning_rate": 2.4381600296034844e-05,
      "loss": 1.6004,
      "step": 54000
    },
    {
      "epoch": 1.5513364265178902,
      "grad_norm": 0.6404986381530762,
      "learning_rate": 2.41443928913685e-05,
      "loss": 1.5296,
      "step": 54500
    },
    {
      "epoch": 1.565568870797871,
      "grad_norm": 335.4157409667969,
      "learning_rate": 2.3907185486702156e-05,
      "loss": 1.4802,
      "step": 55000
    },
    {
      "epoch": 1.5798013150778516,
      "grad_norm": 210.5367431640625,
      "learning_rate": 2.366997808203581e-05,
      "loss": 1.2892,
      "step": 55500
    },
    {
      "epoch": 1.5940337593578322,
      "grad_norm": 0.0008832117891870439,
      "learning_rate": 2.3432770677369468e-05,
      "loss": 1.305,
      "step": 56000
    },
    {
      "epoch": 1.6082662036378128,
      "grad_norm": 157.38621520996094,
      "learning_rate": 2.3195563272703122e-05,
      "loss": 1.3269,
      "step": 56500
    },
    {
      "epoch": 1.6224986479177934,
      "grad_norm": 0.49598127603530884,
      "learning_rate": 2.2958355868036776e-05,
      "loss": 0.9746,
      "step": 57000
    },
    {
      "epoch": 1.636731092197774,
      "grad_norm": 5.005952630199317e-07,
      "learning_rate": 2.2721148463370434e-05,
      "loss": 1.3408,
      "step": 57500
    },
    {
      "epoch": 1.6509635364777546,
      "grad_norm": 0.018824387341737747,
      "learning_rate": 2.2483941058704088e-05,
      "loss": 1.2307,
      "step": 58000
    },
    {
      "epoch": 1.6651959807577352,
      "grad_norm": 0.004480020143091679,
      "learning_rate": 2.2246733654037746e-05,
      "loss": 1.2367,
      "step": 58500
    },
    {
      "epoch": 1.6794284250377158,
      "grad_norm": 11.437743186950684,
      "learning_rate": 2.20095262493714e-05,
      "loss": 1.3383,
      "step": 59000
    },
    {
      "epoch": 1.6936608693176967,
      "grad_norm": 0.06036483868956566,
      "learning_rate": 2.1772318844705058e-05,
      "loss": 1.3084,
      "step": 59500
    },
    {
      "epoch": 1.7078933135976773,
      "grad_norm": 1.3579503297805786,
      "learning_rate": 2.1535111440038712e-05,
      "loss": 1.4009,
      "step": 60000
    },
    {
      "epoch": 1.7221257578776579,
      "grad_norm": 21.6284236907959,
      "learning_rate": 2.129790403537237e-05,
      "loss": 1.3342,
      "step": 60500
    },
    {
      "epoch": 1.7363582021576387,
      "grad_norm": 29.55640983581543,
      "learning_rate": 2.1060696630706027e-05,
      "loss": 1.0924,
      "step": 61000
    },
    {
      "epoch": 1.7505906464376193,
      "grad_norm": 173.9644012451172,
      "learning_rate": 2.082348922603968e-05,
      "loss": 1.2447,
      "step": 61500
    },
    {
      "epoch": 1.7648230907176,
      "grad_norm": 6.281250476837158,
      "learning_rate": 2.058628182137334e-05,
      "loss": 1.3199,
      "step": 62000
    },
    {
      "epoch": 1.7790555349975805,
      "grad_norm": 18.333364486694336,
      "learning_rate": 2.0349074416706993e-05,
      "loss": 1.2088,
      "step": 62500
    },
    {
      "epoch": 1.7932879792775611,
      "grad_norm": 0.01794009655714035,
      "learning_rate": 2.011186701204065e-05,
      "loss": 1.2375,
      "step": 63000
    },
    {
      "epoch": 1.8075204235575417,
      "grad_norm": 2.168732166290283,
      "learning_rate": 1.9874659607374305e-05,
      "loss": 1.2492,
      "step": 63500
    },
    {
      "epoch": 1.8217528678375223,
      "grad_norm": 8.638180588604882e-05,
      "learning_rate": 1.963745220270796e-05,
      "loss": 1.3141,
      "step": 64000
    },
    {
      "epoch": 1.835985312117503,
      "grad_norm": 19.142189025878906,
      "learning_rate": 1.9400244798041617e-05,
      "loss": 1.2977,
      "step": 64500
    },
    {
      "epoch": 1.8502177563974838,
      "grad_norm": 308.3004455566406,
      "learning_rate": 1.916303739337527e-05,
      "loss": 1.1852,
      "step": 65000
    },
    {
      "epoch": 1.8644502006774644,
      "grad_norm": 1.4798052783149274e-09,
      "learning_rate": 1.892582998870893e-05,
      "loss": 1.2323,
      "step": 65500
    },
    {
      "epoch": 1.878682644957445,
      "grad_norm": 5.560269832611084,
      "learning_rate": 1.8688622584042583e-05,
      "loss": 1.4425,
      "step": 66000
    },
    {
      "epoch": 1.8929150892374258,
      "grad_norm": 0.2333667278289795,
      "learning_rate": 1.845141517937624e-05,
      "loss": 1.2208,
      "step": 66500
    },
    {
      "epoch": 1.9071475335174064,
      "grad_norm": 0.011270904913544655,
      "learning_rate": 1.8214207774709895e-05,
      "loss": 1.1584,
      "step": 67000
    },
    {
      "epoch": 1.921379977797387,
      "grad_norm": 291.5809020996094,
      "learning_rate": 1.7977000370043552e-05,
      "loss": 1.4176,
      "step": 67500
    },
    {
      "epoch": 1.9356124220773676,
      "grad_norm": 1.7415293455123901,
      "learning_rate": 1.773979296537721e-05,
      "loss": 1.1936,
      "step": 68000
    },
    {
      "epoch": 1.9498448663573482,
      "grad_norm": 85.77689361572266,
      "learning_rate": 1.7502585560710864e-05,
      "loss": 0.9786,
      "step": 68500
    },
    {
      "epoch": 1.9640773106373288,
      "grad_norm": 208.44168090820312,
      "learning_rate": 1.726537815604452e-05,
      "loss": 1.3196,
      "step": 69000
    },
    {
      "epoch": 1.9783097549173094,
      "grad_norm": 5.191640853881836,
      "learning_rate": 1.7028170751378176e-05,
      "loss": 1.2343,
      "step": 69500
    },
    {
      "epoch": 1.99254219919729,
      "grad_norm": 0.000520756293553859,
      "learning_rate": 1.6790963346711833e-05,
      "loss": 1.2603,
      "step": 70000
    },
    {
      "epoch": 2.0067746434772706,
      "grad_norm": 0.48437151312828064,
      "learning_rate": 1.6553755942045488e-05,
      "loss": 1.1047,
      "step": 70500
    },
    {
      "epoch": 2.0210070877572512,
      "grad_norm": 66.11140441894531,
      "learning_rate": 1.6316548537379145e-05,
      "loss": 1.2467,
      "step": 71000
    },
    {
      "epoch": 2.0352395320372323,
      "grad_norm": 6.315969949355349e-05,
      "learning_rate": 1.60793411327128e-05,
      "loss": 1.1051,
      "step": 71500
    },
    {
      "epoch": 2.049471976317213,
      "grad_norm": 0.015222056768834591,
      "learning_rate": 1.5842133728046454e-05,
      "loss": 1.1647,
      "step": 72000
    },
    {
      "epoch": 2.0637044205971935,
      "grad_norm": 0.0003228056593798101,
      "learning_rate": 1.560492632338011e-05,
      "loss": 1.1471,
      "step": 72500
    },
    {
      "epoch": 2.077936864877174,
      "grad_norm": 0.00011044983693864197,
      "learning_rate": 1.5367718918713766e-05,
      "loss": 1.0928,
      "step": 73000
    },
    {
      "epoch": 2.0921693091571547,
      "grad_norm": 0.002029678551480174,
      "learning_rate": 1.5130511514047425e-05,
      "loss": 1.2604,
      "step": 73500
    },
    {
      "epoch": 2.1064017534371353,
      "grad_norm": 437.6989440917969,
      "learning_rate": 1.4893304109381079e-05,
      "loss": 1.1682,
      "step": 74000
    },
    {
      "epoch": 2.120634197717116,
      "grad_norm": 0.008761188015341759,
      "learning_rate": 1.4656096704714737e-05,
      "loss": 1.2494,
      "step": 74500
    },
    {
      "epoch": 2.1348666419970965,
      "grad_norm": 22.72408676147461,
      "learning_rate": 1.4418889300048391e-05,
      "loss": 1.2037,
      "step": 75000
    },
    {
      "epoch": 2.149099086277077,
      "grad_norm": 0.01661345176398754,
      "learning_rate": 1.4181681895382045e-05,
      "loss": 1.1032,
      "step": 75500
    },
    {
      "epoch": 2.1633315305570577,
      "grad_norm": 194.62342834472656,
      "learning_rate": 1.3944474490715703e-05,
      "loss": 1.4635,
      "step": 76000
    },
    {
      "epoch": 2.1775639748370383,
      "grad_norm": 1.419067621231079,
      "learning_rate": 1.3707267086049359e-05,
      "loss": 1.1742,
      "step": 76500
    },
    {
      "epoch": 2.1917964191170194,
      "grad_norm": 0.3263187110424042,
      "learning_rate": 1.3470059681383016e-05,
      "loss": 1.2213,
      "step": 77000
    },
    {
      "epoch": 2.206028863397,
      "grad_norm": 341.4047546386719,
      "learning_rate": 1.323285227671667e-05,
      "loss": 1.3171,
      "step": 77500
    },
    {
      "epoch": 2.2202613076769806,
      "grad_norm": 0.010969799011945724,
      "learning_rate": 1.2995644872050328e-05,
      "loss": 1.1598,
      "step": 78000
    },
    {
      "epoch": 2.234493751956961,
      "grad_norm": 1.907973289489746,
      "learning_rate": 1.2758437467383982e-05,
      "loss": 1.233,
      "step": 78500
    },
    {
      "epoch": 2.248726196236942,
      "grad_norm": 0.5124784708023071,
      "learning_rate": 1.2521230062717636e-05,
      "loss": 1.034,
      "step": 79000
    },
    {
      "epoch": 2.2629586405169224,
      "grad_norm": 3.334690177325683e-07,
      "learning_rate": 1.2284022658051294e-05,
      "loss": 1.1803,
      "step": 79500
    },
    {
      "epoch": 2.277191084796903,
      "grad_norm": 147.6239471435547,
      "learning_rate": 1.204681525338495e-05,
      "loss": 1.1318,
      "step": 80000
    },
    {
      "epoch": 2.2914235290768836,
      "grad_norm": 0.00035827644751407206,
      "learning_rate": 1.1809607848718608e-05,
      "loss": 1.1722,
      "step": 80500
    },
    {
      "epoch": 2.3056559733568642,
      "grad_norm": 4.226629810033522e-11,
      "learning_rate": 1.1572400444052262e-05,
      "loss": 1.102,
      "step": 81000
    },
    {
      "epoch": 2.319888417636845,
      "grad_norm": 0.029256457462906837,
      "learning_rate": 1.1335193039385918e-05,
      "loss": 1.3087,
      "step": 81500
    },
    {
      "epoch": 2.3341208619168254,
      "grad_norm": 0.40314897894859314,
      "learning_rate": 1.1097985634719574e-05,
      "loss": 1.2656,
      "step": 82000
    },
    {
      "epoch": 2.348353306196806,
      "grad_norm": 0.001224256120622158,
      "learning_rate": 1.086077823005323e-05,
      "loss": 1.1994,
      "step": 82500
    },
    {
      "epoch": 2.3625857504767867,
      "grad_norm": 0.288789302110672,
      "learning_rate": 1.0623570825386885e-05,
      "loss": 1.4443,
      "step": 83000
    },
    {
      "epoch": 2.3768181947567677,
      "grad_norm": 329.513671875,
      "learning_rate": 1.0386363420720541e-05,
      "loss": 1.1175,
      "step": 83500
    },
    {
      "epoch": 2.3910506390367483,
      "grad_norm": 0.004081339109688997,
      "learning_rate": 1.0149156016054199e-05,
      "loss": 1.0974,
      "step": 84000
    },
    {
      "epoch": 2.405283083316729,
      "grad_norm": 30.36007308959961,
      "learning_rate": 9.911948611387853e-06,
      "loss": 1.1901,
      "step": 84500
    },
    {
      "epoch": 2.4195155275967095,
      "grad_norm": 235.61441040039062,
      "learning_rate": 9.674741206721509e-06,
      "loss": 1.2007,
      "step": 85000
    },
    {
      "epoch": 2.43374797187669,
      "grad_norm": 312.31610107421875,
      "learning_rate": 9.437533802055165e-06,
      "loss": 1.1889,
      "step": 85500
    },
    {
      "epoch": 2.4479804161566707,
      "grad_norm": 17.370275497436523,
      "learning_rate": 9.200326397388821e-06,
      "loss": 1.254,
      "step": 86000
    },
    {
      "epoch": 2.4622128604366513,
      "grad_norm": 199.55438232421875,
      "learning_rate": 8.963118992722477e-06,
      "loss": 1.1899,
      "step": 86500
    },
    {
      "epoch": 2.476445304716632,
      "grad_norm": 0.0002413564798189327,
      "learning_rate": 8.725911588056133e-06,
      "loss": 1.248,
      "step": 87000
    },
    {
      "epoch": 2.4906777489966125,
      "grad_norm": 230.59945678710938,
      "learning_rate": 8.48870418338979e-06,
      "loss": 1.0638,
      "step": 87500
    },
    {
      "epoch": 2.5049101932765936,
      "grad_norm": 9.951874879732259e-09,
      "learning_rate": 8.251496778723446e-06,
      "loss": 1.4321,
      "step": 88000
    },
    {
      "epoch": 2.519142637556574,
      "grad_norm": 21.62441062927246,
      "learning_rate": 8.0142893740571e-06,
      "loss": 1.101,
      "step": 88500
    },
    {
      "epoch": 2.533375081836555,
      "grad_norm": 2.4028645384532865e-06,
      "learning_rate": 7.777081969390756e-06,
      "loss": 1.2014,
      "step": 89000
    },
    {
      "epoch": 2.5476075261165354,
      "grad_norm": 164.88087463378906,
      "learning_rate": 7.539874564724412e-06,
      "loss": 1.0345,
      "step": 89500
    },
    {
      "epoch": 2.561839970396516,
      "grad_norm": 6.1404312745594325e-09,
      "learning_rate": 7.302667160058069e-06,
      "loss": 1.3437,
      "step": 90000
    },
    {
      "epoch": 2.5760724146764966,
      "grad_norm": 0.0006744582788087428,
      "learning_rate": 7.065459755391725e-06,
      "loss": 1.2536,
      "step": 90500
    },
    {
      "epoch": 2.590304858956477,
      "grad_norm": 213.77198791503906,
      "learning_rate": 6.828252350725381e-06,
      "loss": 1.2701,
      "step": 91000
    },
    {
      "epoch": 2.604537303236458,
      "grad_norm": 323.82061767578125,
      "learning_rate": 6.591044946059037e-06,
      "loss": 1.2165,
      "step": 91500
    },
    {
      "epoch": 2.6187697475164384,
      "grad_norm": 0.0007768824580125511,
      "learning_rate": 6.353837541392692e-06,
      "loss": 0.9871,
      "step": 92000
    },
    {
      "epoch": 2.633002191796419,
      "grad_norm": 1.1964664814456683e-08,
      "learning_rate": 6.116630136726349e-06,
      "loss": 1.1809,
      "step": 92500
    },
    {
      "epoch": 2.6472346360763996,
      "grad_norm": 0.01886417344212532,
      "learning_rate": 5.879422732060004e-06,
      "loss": 1.2455,
      "step": 93000
    },
    {
      "epoch": 2.6614670803563802,
      "grad_norm": 1.3518683910369873,
      "learning_rate": 5.6422153273936604e-06,
      "loss": 1.0477,
      "step": 93500
    },
    {
      "epoch": 2.675699524636361,
      "grad_norm": 89.34034729003906,
      "learning_rate": 5.405007922727316e-06,
      "loss": 1.3728,
      "step": 94000
    },
    {
      "epoch": 2.6899319689163415,
      "grad_norm": 24.910266876220703,
      "learning_rate": 5.167800518060972e-06,
      "loss": 1.2636,
      "step": 94500
    },
    {
      "epoch": 2.7041644131963225,
      "grad_norm": 50.8535270690918,
      "learning_rate": 4.930593113394627e-06,
      "loss": 1.396,
      "step": 95000
    },
    {
      "epoch": 2.718396857476303,
      "grad_norm": 0.004643031395971775,
      "learning_rate": 4.693385708728284e-06,
      "loss": 1.1522,
      "step": 95500
    },
    {
      "epoch": 2.7326293017562837,
      "grad_norm": 0.933087944984436,
      "learning_rate": 4.45617830406194e-06,
      "loss": 1.3026,
      "step": 96000
    },
    {
      "epoch": 2.7468617460362643,
      "grad_norm": 3.416786000798311e-07,
      "learning_rate": 4.218970899395596e-06,
      "loss": 1.0283,
      "step": 96500
    },
    {
      "epoch": 2.761094190316245,
      "grad_norm": 130.11692810058594,
      "learning_rate": 3.981763494729252e-06,
      "loss": 1.3739,
      "step": 97000
    },
    {
      "epoch": 2.7753266345962255,
      "grad_norm": 2.1324740373529494e-05,
      "learning_rate": 3.7445560900629077e-06,
      "loss": 1.3277,
      "step": 97500
    },
    {
      "epoch": 2.789559078876206,
      "grad_norm": 0.00010976188787026331,
      "learning_rate": 3.5073486853965636e-06,
      "loss": 1.2325,
      "step": 98000
    },
    {
      "epoch": 2.8037915231561867,
      "grad_norm": 238.07444763183594,
      "learning_rate": 3.270141280730219e-06,
      "loss": 1.1932,
      "step": 98500
    },
    {
      "epoch": 2.8180239674361673,
      "grad_norm": 0.0008449407760053873,
      "learning_rate": 3.032933876063875e-06,
      "loss": 1.2113,
      "step": 99000
    },
    {
      "epoch": 2.8322564117161484,
      "grad_norm": 2.691925048828125,
      "learning_rate": 2.7957264713975314e-06,
      "loss": 1.0264,
      "step": 99500
    },
    {
      "epoch": 2.846488855996129,
      "grad_norm": 1.2127447128295898,
      "learning_rate": 2.5585190667311873e-06,
      "loss": 1.2739,
      "step": 100000
    },
    {
      "epoch": 2.8607213002761096,
      "grad_norm": 2.0609796047210693,
      "learning_rate": 2.321311662064843e-06,
      "loss": 1.0268,
      "step": 100500
    },
    {
      "epoch": 2.87495374455609,
      "grad_norm": 0.4568251669406891,
      "learning_rate": 2.084104257398499e-06,
      "loss": 1.3471,
      "step": 101000
    },
    {
      "epoch": 2.889186188836071,
      "grad_norm": 239.36712646484375,
      "learning_rate": 1.846896852732155e-06,
      "loss": 1.1087,
      "step": 101500
    },
    {
      "epoch": 2.9034186331160514,
      "grad_norm": 0.14541205763816833,
      "learning_rate": 1.609689448065811e-06,
      "loss": 1.0582,
      "step": 102000
    },
    {
      "epoch": 2.917651077396032,
      "grad_norm": 247.89581298828125,
      "learning_rate": 1.3724820433994668e-06,
      "loss": 1.1325,
      "step": 102500
    },
    {
      "epoch": 2.9318835216760126,
      "grad_norm": 0.002974768402054906,
      "learning_rate": 1.1352746387331228e-06,
      "loss": 1.3426,
      "step": 103000
    },
    {
      "epoch": 2.9461159659559932,
      "grad_norm": 0.0011548080947250128,
      "learning_rate": 8.980672340667787e-07,
      "loss": 1.1403,
      "step": 103500
    },
    {
      "epoch": 2.960348410235974,
      "grad_norm": 0.3729552924633026,
      "learning_rate": 6.608598294004346e-07,
      "loss": 1.1825,
      "step": 104000
    },
    {
      "epoch": 2.9745808545159544,
      "grad_norm": 268.66912841796875,
      "learning_rate": 4.236524247340905e-07,
      "loss": 1.3315,
      "step": 104500
    },
    {
      "epoch": 2.988813298795935,
      "grad_norm": 43.98519515991211,
      "learning_rate": 1.8644502006774646e-07,
      "loss": 1.2468,
      "step": 105000
    }
  ],
  "logging_steps": 500,
  "max_steps": 105393,
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}