{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 61986,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.024199012680282645,
      "grad_norm": 3.446533203125,
      "learning_rate": 4.959668312199529e-05,
      "loss": 2.3814,
      "step": 500
    },
    {
      "epoch": 0.04839802536056529,
      "grad_norm": 2.858267307281494,
      "learning_rate": 4.919336624399058e-05,
      "loss": 1.6485,
      "step": 1000
    },
    {
      "epoch": 0.07259703804084794,
      "grad_norm": 3.279078960418701,
      "learning_rate": 4.8790049365985867e-05,
      "loss": 1.4452,
      "step": 1500
    },
    {
      "epoch": 0.09679605072113058,
      "grad_norm": 2.5900914669036865,
      "learning_rate": 4.838673248798116e-05,
      "loss": 1.3332,
      "step": 2000
    },
    {
      "epoch": 0.12099506340141322,
      "grad_norm": 3.089566707611084,
      "learning_rate": 4.798341560997645e-05,
      "loss": 1.2338,
      "step": 2500
    },
    {
      "epoch": 0.14519407608169588,
      "grad_norm": 3.1429877281188965,
      "learning_rate": 4.758009873197174e-05,
      "loss": 1.1925,
      "step": 3000
    },
    {
      "epoch": 0.1693930887619785,
      "grad_norm": 2.609388589859009,
      "learning_rate": 4.717678185396703e-05,
      "loss": 1.1659,
      "step": 3500
    },
    {
      "epoch": 0.19359210144226116,
      "grad_norm": 2.7544608116149902,
      "learning_rate": 4.6773464975962316e-05,
      "loss": 1.1226,
      "step": 4000
    },
    {
      "epoch": 0.2177911141225438,
      "grad_norm": 2.377077102661133,
      "learning_rate": 4.637014809795761e-05,
      "loss": 1.0864,
      "step": 4500
    },
    {
      "epoch": 0.24199012680282644,
      "grad_norm": 2.427476167678833,
      "learning_rate": 4.5966831219952894e-05,
      "loss": 1.0755,
      "step": 5000
    },
    {
      "epoch": 0.26618913948310907,
      "grad_norm": 2.287100315093994,
      "learning_rate": 4.556351434194818e-05,
      "loss": 1.046,
      "step": 5500
    },
    {
      "epoch": 0.29038815216339176,
      "grad_norm": 2.4845123291015625,
      "learning_rate": 4.516019746394347e-05,
      "loss": 1.0138,
      "step": 6000
    },
    {
      "epoch": 0.3145871648436744,
      "grad_norm": 2.4564921855926514,
      "learning_rate": 4.4756880585938765e-05,
      "loss": 1.0102,
      "step": 6500
    },
    {
      "epoch": 0.338786177523957,
      "grad_norm": 2.1205663681030273,
      "learning_rate": 4.435356370793405e-05,
      "loss": 0.9995,
      "step": 7000
    },
    {
      "epoch": 0.3629851902042397,
      "grad_norm": 2.4256932735443115,
      "learning_rate": 4.3950246829929344e-05,
      "loss": 0.9885,
      "step": 7500
    },
    {
      "epoch": 0.3871842028845223,
      "grad_norm": 2.172882080078125,
      "learning_rate": 4.354692995192463e-05,
      "loss": 0.9622,
      "step": 8000
    },
    {
      "epoch": 0.41138321556480495,
      "grad_norm": 2.3957536220550537,
      "learning_rate": 4.314361307391992e-05,
      "loss": 0.9523,
      "step": 8500
    },
    {
      "epoch": 0.4355822282450876,
      "grad_norm": 2.0573976039886475,
      "learning_rate": 4.274029619591521e-05,
      "loss": 0.9413,
      "step": 9000
    },
    {
      "epoch": 0.45978124092537026,
      "grad_norm": 1.9949113130569458,
      "learning_rate": 4.2336979317910493e-05,
      "loss": 0.9375,
      "step": 9500
    },
    {
      "epoch": 0.4839802536056529,
      "grad_norm": 1.9614790678024292,
      "learning_rate": 4.1933662439905786e-05,
      "loss": 0.9216,
      "step": 10000
    },
    {
      "epoch": 0.5081792662859356,
      "grad_norm": 1.9464462995529175,
      "learning_rate": 4.153034556190107e-05,
      "loss": 0.9116,
      "step": 10500
    },
    {
      "epoch": 0.5323782789662181,
      "grad_norm": 2.082477331161499,
      "learning_rate": 4.112702868389637e-05,
      "loss": 0.9012,
      "step": 11000
    },
    {
      "epoch": 0.5565772916465008,
      "grad_norm": 1.8631994724273682,
      "learning_rate": 4.072371180589166e-05,
      "loss": 0.8909,
      "step": 11500
    },
    {
      "epoch": 0.5807763043267835,
      "grad_norm": 2.108043670654297,
      "learning_rate": 4.032039492788694e-05,
      "loss": 0.8833,
      "step": 12000
    },
    {
      "epoch": 0.6049753170070661,
      "grad_norm": 2.236179828643799,
      "learning_rate": 3.9917078049882235e-05,
      "loss": 0.8749,
      "step": 12500
    },
    {
      "epoch": 0.6291743296873488,
      "grad_norm": 2.340641498565674,
      "learning_rate": 3.951376117187752e-05,
      "loss": 0.8711,
      "step": 13000
    },
    {
      "epoch": 0.6533733423676314,
      "grad_norm": 2.0404820442199707,
      "learning_rate": 3.9110444293872814e-05,
      "loss": 0.8621,
      "step": 13500
    },
    {
      "epoch": 0.677572355047914,
      "grad_norm": 2.5088083744049072,
      "learning_rate": 3.87071274158681e-05,
      "loss": 0.8675,
      "step": 14000
    },
    {
      "epoch": 0.7017713677281967,
      "grad_norm": 1.8882901668548584,
      "learning_rate": 3.8303810537863385e-05,
      "loss": 0.8642,
      "step": 14500
    },
    {
      "epoch": 0.7259703804084794,
      "grad_norm": 2.184558629989624,
      "learning_rate": 3.790049365985868e-05,
      "loss": 0.8546,
      "step": 15000
    },
    {
      "epoch": 0.750169393088762,
      "grad_norm": 1.7006648778915405,
      "learning_rate": 3.749717678185397e-05,
      "loss": 0.8608,
      "step": 15500
    },
    {
      "epoch": 0.7743684057690446,
      "grad_norm": 1.9007302522659302,
      "learning_rate": 3.7093859903849256e-05,
      "loss": 0.846,
      "step": 16000
    },
    {
      "epoch": 0.7985674184493272,
      "grad_norm": 2.195310354232788,
      "learning_rate": 3.669054302584455e-05,
      "loss": 0.8256,
      "step": 16500
    },
    {
      "epoch": 0.8227664311296099,
      "grad_norm": 1.8752928972244263,
      "learning_rate": 3.6287226147839835e-05,
      "loss": 0.8325,
      "step": 17000
    },
    {
      "epoch": 0.8469654438098926,
      "grad_norm": 2.0472004413604736,
      "learning_rate": 3.588390926983513e-05,
      "loss": 0.8341,
      "step": 17500
    },
    {
      "epoch": 0.8711644564901752,
      "grad_norm": 1.9115586280822754,
      "learning_rate": 3.548059239183041e-05,
      "loss": 0.8077,
      "step": 18000
    },
    {
      "epoch": 0.8953634691704578,
      "grad_norm": 1.7010232210159302,
      "learning_rate": 3.5077275513825705e-05,
      "loss": 0.8204,
      "step": 18500
    },
    {
      "epoch": 0.9195624818507405,
      "grad_norm": 2.0516223907470703,
      "learning_rate": 3.467395863582099e-05,
      "loss": 0.815,
      "step": 19000
    },
    {
      "epoch": 0.9437614945310231,
      "grad_norm": 2.2772953510284424,
      "learning_rate": 3.4270641757816284e-05,
      "loss": 0.7937,
      "step": 19500
    },
    {
      "epoch": 0.9679605072113058,
      "grad_norm": 2.2498035430908203,
      "learning_rate": 3.3867324879811576e-05,
      "loss": 0.8024,
      "step": 20000
    },
    {
      "epoch": 0.9921595198915885,
      "grad_norm": 2.2339396476745605,
      "learning_rate": 3.346400800180686e-05,
      "loss": 0.8007,
      "step": 20500
    },
    {
      "epoch": 1.0163585325718711,
      "grad_norm": 1.7616324424743652,
      "learning_rate": 3.306069112380215e-05,
      "loss": 0.736,
      "step": 21000
    },
    {
      "epoch": 1.0405575452521538,
      "grad_norm": 2.0441577434539795,
      "learning_rate": 3.265737424579744e-05,
      "loss": 0.7232,
      "step": 21500
    },
    {
      "epoch": 1.0647565579324363,
      "grad_norm": 2.0613925457000732,
      "learning_rate": 3.2254057367792726e-05,
      "loss": 0.7219,
      "step": 22000
    },
    {
      "epoch": 1.088955570612719,
      "grad_norm": 2.029736280441284,
      "learning_rate": 3.185074048978802e-05,
      "loss": 0.7159,
      "step": 22500
    },
    {
      "epoch": 1.1131545832930017,
      "grad_norm": 2.0547034740448,
      "learning_rate": 3.1447423611783305e-05,
      "loss": 0.7145,
      "step": 23000
    },
    {
      "epoch": 1.1373535959732843,
      "grad_norm": 1.8258686065673828,
      "learning_rate": 3.10441067337786e-05,
      "loss": 0.7141,
      "step": 23500
    },
    {
      "epoch": 1.161552608653567,
      "grad_norm": 2.208686113357544,
      "learning_rate": 3.064078985577389e-05,
      "loss": 0.7139,
      "step": 24000
    },
    {
      "epoch": 1.1857516213338495,
      "grad_norm": 1.617804765701294,
      "learning_rate": 3.0237472977769176e-05,
      "loss": 0.7022,
      "step": 24500
    },
    {
      "epoch": 1.2099506340141322,
      "grad_norm": 1.8249517679214478,
      "learning_rate": 2.9834156099764465e-05,
      "loss": 0.7109,
      "step": 25000
    },
    {
      "epoch": 1.2341496466944148,
      "grad_norm": 2.5547683238983154,
      "learning_rate": 2.9430839221759754e-05,
      "loss": 0.7052,
      "step": 25500
    },
    {
      "epoch": 1.2583486593746975,
      "grad_norm": 2.1314806938171387,
      "learning_rate": 2.9027522343755043e-05,
      "loss": 0.6946,
      "step": 26000
    },
    {
      "epoch": 1.2825476720549802,
      "grad_norm": 1.804982304573059,
      "learning_rate": 2.8624205465750332e-05,
      "loss": 0.7012,
      "step": 26500
    },
    {
      "epoch": 1.3067466847352627,
      "grad_norm": 1.6890604496002197,
      "learning_rate": 2.8220888587745618e-05,
      "loss": 0.7059,
      "step": 27000
    },
    {
      "epoch": 1.3309456974155456,
      "grad_norm": 2.106729030609131,
      "learning_rate": 2.7817571709740907e-05,
      "loss": 0.7026,
      "step": 27500
    },
    {
      "epoch": 1.355144710095828,
      "grad_norm": 1.710942268371582,
      "learning_rate": 2.7414254831736203e-05,
      "loss": 0.6938,
      "step": 28000
    },
    {
      "epoch": 1.3793437227761107,
      "grad_norm": 1.7750734090805054,
      "learning_rate": 2.7010937953731493e-05,
      "loss": 0.7017,
      "step": 28500
    },
    {
      "epoch": 1.4035427354563934,
      "grad_norm": 1.5013315677642822,
      "learning_rate": 2.660762107572678e-05,
      "loss": 0.7082,
      "step": 29000
    },
    {
      "epoch": 1.427741748136676,
      "grad_norm": 1.9315123558044434,
      "learning_rate": 2.6204304197722067e-05,
      "loss": 0.6967,
      "step": 29500
    },
    {
      "epoch": 1.4519407608169588,
      "grad_norm": 1.6147435903549194,
      "learning_rate": 2.5800987319717357e-05,
      "loss": 0.6935,
      "step": 30000
    },
    {
      "epoch": 1.4761397734972412,
      "grad_norm": 1.6625306606292725,
      "learning_rate": 2.5397670441712646e-05,
      "loss": 0.6966,
      "step": 30500
    },
    {
      "epoch": 1.500338786177524,
      "grad_norm": 2.0794835090637207,
      "learning_rate": 2.4994353563707935e-05,
      "loss": 0.6915,
      "step": 31000
    },
    {
      "epoch": 1.5245377988578066,
      "grad_norm": 1.884932518005371,
      "learning_rate": 2.4591036685703224e-05,
      "loss": 0.6828,
      "step": 31500
    },
    {
      "epoch": 1.5487368115380893,
      "grad_norm": 1.472744107246399,
      "learning_rate": 2.4187719807698513e-05,
      "loss": 0.6769,
      "step": 32000
    },
    {
      "epoch": 1.572935824218372,
      "grad_norm": 2.082698345184326,
      "learning_rate": 2.3784402929693803e-05,
      "loss": 0.6916,
      "step": 32500
    },
    {
      "epoch": 1.5971348368986544,
      "grad_norm": 2.057123899459839,
      "learning_rate": 2.3381086051689092e-05,
      "loss": 0.675,
      "step": 33000
    },
    {
      "epoch": 1.6213338495789373,
      "grad_norm": 2.084205389022827,
      "learning_rate": 2.297776917368438e-05,
      "loss": 0.6836,
      "step": 33500
    },
    {
      "epoch": 1.6455328622592198,
      "grad_norm": 2.014021396636963,
      "learning_rate": 2.257445229567967e-05,
      "loss": 0.6826,
      "step": 34000
    },
    {
      "epoch": 1.6697318749395025,
      "grad_norm": 2.0385689735412598,
      "learning_rate": 2.217113541767496e-05,
      "loss": 0.669,
      "step": 34500
    },
    {
      "epoch": 1.6939308876197852,
      "grad_norm": 1.8431167602539062,
      "learning_rate": 2.176781853967025e-05,
      "loss": 0.6752,
      "step": 35000
    },
    {
      "epoch": 1.7181299003000676,
      "grad_norm": 1.9194872379302979,
      "learning_rate": 2.136450166166554e-05,
      "loss": 0.672,
      "step": 35500
    },
    {
      "epoch": 1.7423289129803505,
      "grad_norm": 1.92533278465271,
      "learning_rate": 2.0961184783660827e-05,
      "loss": 0.6707,
      "step": 36000
    },
    {
      "epoch": 1.766527925660633,
      "grad_norm": 1.8021315336227417,
      "learning_rate": 2.0557867905656116e-05,
      "loss": 0.6672,
      "step": 36500
    },
    {
      "epoch": 1.7907269383409157,
      "grad_norm": 1.519113302230835,
      "learning_rate": 2.0154551027651405e-05,
      "loss": 0.662,
      "step": 37000
    },
    {
      "epoch": 1.8149259510211984,
      "grad_norm": 2.1313390731811523,
      "learning_rate": 1.9751234149646694e-05,
      "loss": 0.6669,
      "step": 37500
    },
    {
      "epoch": 1.8391249637014808,
      "grad_norm": 1.543104648590088,
      "learning_rate": 1.9347917271641987e-05,
      "loss": 0.6673,
      "step": 38000
    },
    {
      "epoch": 1.8633239763817637,
      "grad_norm": 1.8662470579147339,
      "learning_rate": 1.8944600393637273e-05,
      "loss": 0.6506,
      "step": 38500
    },
    {
      "epoch": 1.8875229890620462,
      "grad_norm": 2.1024982929229736,
      "learning_rate": 1.8541283515632562e-05,
      "loss": 0.6632,
      "step": 39000
    },
    {
      "epoch": 1.9117220017423289,
      "grad_norm": 1.7612417936325073,
      "learning_rate": 1.813796663762785e-05,
      "loss": 0.664,
      "step": 39500
    },
    {
      "epoch": 1.9359210144226116,
      "grad_norm": 1.5672264099121094,
      "learning_rate": 1.7734649759623144e-05,
      "loss": 0.6495,
      "step": 40000
    },
    {
      "epoch": 1.9601200271028942,
      "grad_norm": 1.762838363647461,
      "learning_rate": 1.733133288161843e-05,
      "loss": 0.6662,
      "step": 40500
    },
    {
      "epoch": 1.984319039783177,
      "grad_norm": 1.5966898202896118,
      "learning_rate": 1.692801600361372e-05,
      "loss": 0.6522,
      "step": 41000
    },
    {
      "epoch": 2.0085180524634594,
      "grad_norm": 1.9735461473464966,
      "learning_rate": 1.6524699125609008e-05,
      "loss": 0.6247,
      "step": 41500
    },
    {
      "epoch": 2.0327170651437423,
      "grad_norm": 1.692125678062439,
      "learning_rate": 1.61213822476043e-05,
      "loss": 0.5777,
      "step": 42000
    },
    {
      "epoch": 2.0569160778240247,
      "grad_norm": 2.1715869903564453,
      "learning_rate": 1.571806536959959e-05,
      "loss": 0.5773,
      "step": 42500
    },
    {
      "epoch": 2.0811150905043077,
      "grad_norm": 2.170106887817383,
      "learning_rate": 1.5314748491594875e-05,
      "loss": 0.5768,
      "step": 43000
    },
    {
      "epoch": 2.10531410318459,
      "grad_norm": 1.8080190420150757,
      "learning_rate": 1.4911431613590166e-05,
      "loss": 0.5746,
      "step": 43500
    },
    {
      "epoch": 2.1295131158648726,
      "grad_norm": 1.7945517301559448,
      "learning_rate": 1.4508114735585457e-05,
      "loss": 0.5798,
      "step": 44000
    },
    {
      "epoch": 2.1537121285451555,
      "grad_norm": 1.9387438297271729,
      "learning_rate": 1.4104797857580745e-05,
      "loss": 0.5701,
      "step": 44500
    },
    {
      "epoch": 2.177911141225438,
      "grad_norm": 1.6410374641418457,
      "learning_rate": 1.3701480979576034e-05,
      "loss": 0.5748,
      "step": 45000
    },
    {
      "epoch": 2.202110153905721,
      "grad_norm": 1.6036646366119385,
      "learning_rate": 1.3298164101571323e-05,
      "loss": 0.5736,
      "step": 45500
    },
    {
      "epoch": 2.2263091665860033,
      "grad_norm": 1.8704679012298584,
      "learning_rate": 1.289484722356661e-05,
      "loss": 0.5661,
      "step": 46000
    },
    {
      "epoch": 2.2505081792662858,
      "grad_norm": 1.515893816947937,
      "learning_rate": 1.2491530345561901e-05,
      "loss": 0.5733,
      "step": 46500
    },
    {
      "epoch": 2.2747071919465687,
      "grad_norm": 2.0149569511413574,
      "learning_rate": 1.208821346755719e-05,
      "loss": 0.5816,
      "step": 47000
    },
    {
      "epoch": 2.298906204626851,
      "grad_norm": 1.8056936264038086,
      "learning_rate": 1.168489658955248e-05,
      "loss": 0.5707,
      "step": 47500
    },
    {
      "epoch": 2.323105217307134,
      "grad_norm": 2.362672805786133,
      "learning_rate": 1.128157971154777e-05,
      "loss": 0.5729,
      "step": 48000
    },
    {
      "epoch": 2.3473042299874165,
      "grad_norm": 2.264782667160034,
      "learning_rate": 1.0878262833543058e-05,
      "loss": 0.5677,
      "step": 48500
    },
    {
      "epoch": 2.371503242667699,
      "grad_norm": 1.970468521118164,
      "learning_rate": 1.0474945955538349e-05,
      "loss": 0.5711,
      "step": 49000
    },
    {
      "epoch": 2.395702255347982,
      "grad_norm": 1.8722805976867676,
      "learning_rate": 1.0071629077533636e-05,
      "loss": 0.5743,
      "step": 49500
    },
    {
      "epoch": 2.4199012680282643,
      "grad_norm": 1.9151874780654907,
      "learning_rate": 9.668312199528927e-06,
      "loss": 0.5713,
      "step": 50000
    },
    {
      "epoch": 2.4441002807085472,
      "grad_norm": 1.5819072723388672,
      "learning_rate": 9.264995321524215e-06,
      "loss": 0.5651,
      "step": 50500
    },
    {
      "epoch": 2.4682992933888297,
      "grad_norm": 1.572161078453064,
      "learning_rate": 8.861678443519506e-06,
      "loss": 0.5618,
      "step": 51000
    },
    {
      "epoch": 2.4924983060691126,
      "grad_norm": 1.68492591381073,
      "learning_rate": 8.458361565514795e-06,
      "loss": 0.5641,
      "step": 51500
    },
    {
      "epoch": 2.516697318749395,
      "grad_norm": 1.9714465141296387,
      "learning_rate": 8.055044687510082e-06,
      "loss": 0.561,
      "step": 52000
    },
    {
      "epoch": 2.5408963314296775,
      "grad_norm": 2.1967012882232666,
      "learning_rate": 7.651727809505373e-06,
      "loss": 0.5618,
      "step": 52500
    },
    {
      "epoch": 2.5650953441099604,
      "grad_norm": 1.7313326597213745,
      "learning_rate": 7.2484109315006616e-06,
      "loss": 0.5648,
      "step": 53000
    },
    {
      "epoch": 2.589294356790243,
      "grad_norm": 1.7489323616027832,
      "learning_rate": 6.845094053495952e-06,
      "loss": 0.5609,
      "step": 53500
    },
    {
      "epoch": 2.6134933694705254,
      "grad_norm": 1.8374857902526855,
      "learning_rate": 6.44177717549124e-06,
      "loss": 0.5592,
      "step": 54000
    },
    {
      "epoch": 2.6376923821508083,
      "grad_norm": 2.1268906593322754,
      "learning_rate": 6.03846029748653e-06,
      "loss": 0.5584,
      "step": 54500
    },
    {
      "epoch": 2.661891394831091,
      "grad_norm": 2.1709351539611816,
      "learning_rate": 5.635143419481819e-06,
      "loss": 0.5563,
      "step": 55000
    },
    {
      "epoch": 2.6860904075113736,
      "grad_norm": 1.9610742330551147,
      "learning_rate": 5.2318265414771075e-06,
      "loss": 0.5633,
      "step": 55500
    },
    {
      "epoch": 2.710289420191656,
      "grad_norm": 1.9506553411483765,
      "learning_rate": 4.828509663472397e-06,
      "loss": 0.555,
      "step": 56000
    },
    {
      "epoch": 2.734488432871939,
      "grad_norm": 1.5822830200195312,
      "learning_rate": 4.425192785467687e-06,
      "loss": 0.5508,
      "step": 56500
    },
    {
      "epoch": 2.7586874455522215,
      "grad_norm": 1.636152744293213,
      "learning_rate": 4.021875907462976e-06,
      "loss": 0.5504,
      "step": 57000
    },
    {
      "epoch": 2.782886458232504,
      "grad_norm": 1.5931339263916016,
      "learning_rate": 3.618559029458265e-06,
      "loss": 0.5528,
      "step": 57500
    },
    {
      "epoch": 2.807085470912787,
      "grad_norm": 1.7630141973495483,
      "learning_rate": 3.2152421514535542e-06,
      "loss": 0.5476,
      "step": 58000
    },
    {
      "epoch": 2.8312844835930693,
      "grad_norm": 1.8491230010986328,
      "learning_rate": 2.8119252734488434e-06,
      "loss": 0.5586,
      "step": 58500
    },
    {
      "epoch": 2.855483496273352,
      "grad_norm": 1.9479539394378662,
      "learning_rate": 2.4086083954441326e-06,
      "loss": 0.5508,
      "step": 59000
    },
    {
      "epoch": 2.8796825089536346,
      "grad_norm": 2.1013519763946533,
      "learning_rate": 2.0052915174394218e-06,
      "loss": 0.5482,
      "step": 59500
    },
    {
      "epoch": 2.9038815216339176,
      "grad_norm": 1.46125066280365,
      "learning_rate": 1.6019746394347112e-06,
      "loss": 0.5467,
      "step": 60000
    },
    {
      "epoch": 2.9280805343142,
      "grad_norm": 1.8522732257843018,
      "learning_rate": 1.1986577614300003e-06,
      "loss": 0.5542,
      "step": 60500
    },
    {
      "epoch": 2.9522795469944825,
      "grad_norm": 1.802648663520813,
      "learning_rate": 7.953408834252896e-07,
      "loss": 0.5501,
      "step": 61000
    },
    {
      "epoch": 2.9764785596747654,
      "grad_norm": 1.584294319152832,
      "learning_rate": 3.9202400542057887e-07,
      "loss": 0.5514,
      "step": 61500
    },
    {
      "epoch": 3.0,
      "step": 61986,
      "total_flos": 1.2743505814914662e+17,
      "train_loss": 0.7519291775680383,
      "train_runtime": 22698.2595,
      "train_samples_per_second": 43.694,
      "train_steps_per_second": 2.731
    }
  ],
  "logging_steps": 500,
  "max_steps": 61986,
  "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": 1.2743505814914662e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}