{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.009,
  "eval_steps": 500,
  "global_step": 9000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 5e-05,
      "grad_norm": 4.426811695098877,
      "learning_rate": 4.999755000000001e-05,
      "loss": 6.3755,
      "step": 50
    },
    {
      "epoch": 0.0001,
      "grad_norm": 1.7730491161346436,
      "learning_rate": 4.999505e-05,
      "loss": 3.3233,
      "step": 100
    },
    {
      "epoch": 0.00015,
      "grad_norm": 1.4473446607589722,
      "learning_rate": 4.999255e-05,
      "loss": 2.7523,
      "step": 150
    },
    {
      "epoch": 0.0002,
      "grad_norm": 1.4645352363586426,
      "learning_rate": 4.9990050000000004e-05,
      "loss": 2.4618,
      "step": 200
    },
    {
      "epoch": 0.00025,
      "grad_norm": 1.1990822553634644,
      "learning_rate": 4.998755e-05,
      "loss": 2.3417,
      "step": 250
    },
    {
      "epoch": 0.0003,
      "grad_norm": 3.0459673404693604,
      "learning_rate": 4.998505e-05,
      "loss": 2.2581,
      "step": 300
    },
    {
      "epoch": 0.00035,
      "grad_norm": 1.2380167245864868,
      "learning_rate": 4.998255e-05,
      "loss": 2.1563,
      "step": 350
    },
    {
      "epoch": 0.0004,
      "grad_norm": 1.2060939073562622,
      "learning_rate": 4.998005e-05,
      "loss": 1.8521,
      "step": 400
    },
    {
      "epoch": 0.00045,
      "grad_norm": 0.9207198619842529,
      "learning_rate": 4.9977550000000004e-05,
      "loss": 1.9201,
      "step": 450
    },
    {
      "epoch": 0.0005,
      "grad_norm": 1.0762739181518555,
      "learning_rate": 4.997505e-05,
      "loss": 1.8104,
      "step": 500
    },
    {
      "epoch": 0.00055,
      "grad_norm": 1.0367337465286255,
      "learning_rate": 4.997255e-05,
      "loss": 1.8756,
      "step": 550
    },
    {
      "epoch": 0.0006,
      "grad_norm": 0.9291325211524963,
      "learning_rate": 4.997005e-05,
      "loss": 1.7404,
      "step": 600
    },
    {
      "epoch": 0.00065,
      "grad_norm": 0.7452297210693359,
      "learning_rate": 4.996755e-05,
      "loss": 1.6843,
      "step": 650
    },
    {
      "epoch": 0.0007,
      "grad_norm": 0.9317042231559753,
      "learning_rate": 4.9965050000000004e-05,
      "loss": 1.7006,
      "step": 700
    },
    {
      "epoch": 0.00075,
      "grad_norm": 1.0354136228561401,
      "learning_rate": 4.9962550000000005e-05,
      "loss": 1.5652,
      "step": 750
    },
    {
      "epoch": 0.0008,
      "grad_norm": 1.050106406211853,
      "learning_rate": 4.996005e-05,
      "loss": 1.5966,
      "step": 800
    },
    {
      "epoch": 0.00085,
      "grad_norm": 0.9646019339561462,
      "learning_rate": 4.995755e-05,
      "loss": 1.6297,
      "step": 850
    },
    {
      "epoch": 0.0009,
      "grad_norm": 0.9152287840843201,
      "learning_rate": 4.995505e-05,
      "loss": 1.6328,
      "step": 900
    },
    {
      "epoch": 0.00095,
      "grad_norm": 0.9403690099716187,
      "learning_rate": 4.995255e-05,
      "loss": 1.6056,
      "step": 950
    },
    {
      "epoch": 0.001,
      "grad_norm": 1.1822874546051025,
      "learning_rate": 4.9950050000000005e-05,
      "loss": 1.6016,
      "step": 1000
    },
    {
      "epoch": 0.00105,
      "grad_norm": 1.3227542638778687,
      "learning_rate": 4.9947550000000006e-05,
      "loss": 1.5448,
      "step": 1050
    },
    {
      "epoch": 0.0011,
      "grad_norm": 1.0503350496292114,
      "learning_rate": 4.994505000000001e-05,
      "loss": 1.6355,
      "step": 1100
    },
    {
      "epoch": 0.00115,
      "grad_norm": 1.1647204160690308,
      "learning_rate": 4.994255e-05,
      "loss": 1.4673,
      "step": 1150
    },
    {
      "epoch": 0.0012,
      "grad_norm": 0.7281339168548584,
      "learning_rate": 4.994005e-05,
      "loss": 1.4641,
      "step": 1200
    },
    {
      "epoch": 0.00125,
      "grad_norm": 1.2438446283340454,
      "learning_rate": 4.9937550000000004e-05,
      "loss": 1.5177,
      "step": 1250
    },
    {
      "epoch": 0.0013,
      "grad_norm": 0.7967873811721802,
      "learning_rate": 4.993505e-05,
      "loss": 1.4399,
      "step": 1300
    },
    {
      "epoch": 0.00135,
      "grad_norm": 0.8938255310058594,
      "learning_rate": 4.993255e-05,
      "loss": 1.4453,
      "step": 1350
    },
    {
      "epoch": 0.0014,
      "grad_norm": 1.4659217596054077,
      "learning_rate": 4.993005e-05,
      "loss": 1.6517,
      "step": 1400
    },
    {
      "epoch": 0.00145,
      "grad_norm": 0.7856793403625488,
      "learning_rate": 4.992755e-05,
      "loss": 1.4721,
      "step": 1450
    },
    {
      "epoch": 0.0015,
      "grad_norm": 0.6772142648696899,
      "learning_rate": 4.9925050000000004e-05,
      "loss": 1.5314,
      "step": 1500
    },
    {
      "epoch": 0.00155,
      "grad_norm": 0.7613831758499146,
      "learning_rate": 4.992255e-05,
      "loss": 1.4603,
      "step": 1550
    },
    {
      "epoch": 0.0016,
      "grad_norm": 1.1216529607772827,
      "learning_rate": 4.992005e-05,
      "loss": 1.54,
      "step": 1600
    },
    {
      "epoch": 0.00165,
      "grad_norm": 0.7359323501586914,
      "learning_rate": 4.991755e-05,
      "loss": 1.5121,
      "step": 1650
    },
    {
      "epoch": 0.0017,
      "grad_norm": 0.8343626260757446,
      "learning_rate": 4.991505e-05,
      "loss": 1.6107,
      "step": 1700
    },
    {
      "epoch": 0.00175,
      "grad_norm": 0.9381140470504761,
      "learning_rate": 4.9912550000000004e-05,
      "loss": 1.5059,
      "step": 1750
    },
    {
      "epoch": 0.0018,
      "grad_norm": 0.7317638993263245,
      "learning_rate": 4.9910050000000005e-05,
      "loss": 1.5621,
      "step": 1800
    },
    {
      "epoch": 0.00185,
      "grad_norm": 0.8273525834083557,
      "learning_rate": 4.990755e-05,
      "loss": 1.4737,
      "step": 1850
    },
    {
      "epoch": 0.0019,
      "grad_norm": 0.7206686735153198,
      "learning_rate": 4.990505e-05,
      "loss": 1.5242,
      "step": 1900
    },
    {
      "epoch": 0.00195,
      "grad_norm": 0.9094993472099304,
      "learning_rate": 4.990255e-05,
      "loss": 1.5336,
      "step": 1950
    },
    {
      "epoch": 0.002,
      "grad_norm": 0.7432169914245605,
      "learning_rate": 4.9900050000000003e-05,
      "loss": 1.4913,
      "step": 2000
    },
    {
      "epoch": 0.00205,
      "grad_norm": 1.337276816368103,
      "learning_rate": 4.9897550000000005e-05,
      "loss": 1.5643,
      "step": 2050
    },
    {
      "epoch": 0.0021,
      "grad_norm": 0.6664676666259766,
      "learning_rate": 4.9895050000000006e-05,
      "loss": 1.4728,
      "step": 2100
    },
    {
      "epoch": 0.00215,
      "grad_norm": 0.7946372628211975,
      "learning_rate": 4.989255000000001e-05,
      "loss": 1.4832,
      "step": 2150
    },
    {
      "epoch": 0.0022,
      "grad_norm": 0.9259161949157715,
      "learning_rate": 4.989005e-05,
      "loss": 1.5674,
      "step": 2200
    },
    {
      "epoch": 0.00225,
      "grad_norm": 0.9703247547149658,
      "learning_rate": 4.988755e-05,
      "loss": 1.3973,
      "step": 2250
    },
    {
      "epoch": 0.0023,
      "grad_norm": 0.7390087842941284,
      "learning_rate": 4.988505e-05,
      "loss": 1.3919,
      "step": 2300
    },
    {
      "epoch": 0.00235,
      "grad_norm": 0.768277645111084,
      "learning_rate": 4.988255e-05,
      "loss": 1.4815,
      "step": 2350
    },
    {
      "epoch": 0.0024,
      "grad_norm": 0.76382976770401,
      "learning_rate": 4.988005e-05,
      "loss": 1.4458,
      "step": 2400
    },
    {
      "epoch": 0.00245,
      "grad_norm": 0.7780851721763611,
      "learning_rate": 4.987755e-05,
      "loss": 1.4382,
      "step": 2450
    },
    {
      "epoch": 0.0025,
      "grad_norm": 0.7184464335441589,
      "learning_rate": 4.987505e-05,
      "loss": 1.521,
      "step": 2500
    },
    {
      "epoch": 0.00255,
      "grad_norm": 0.6212390065193176,
      "learning_rate": 4.9872550000000004e-05,
      "loss": 1.3889,
      "step": 2550
    },
    {
      "epoch": 0.0026,
      "grad_norm": 0.8539580702781677,
      "learning_rate": 4.987005e-05,
      "loss": 1.415,
      "step": 2600
    },
    {
      "epoch": 0.00265,
      "grad_norm": 0.7129775881767273,
      "learning_rate": 4.986755e-05,
      "loss": 1.4102,
      "step": 2650
    },
    {
      "epoch": 0.0027,
      "grad_norm": 0.5899195671081543,
      "learning_rate": 4.986505e-05,
      "loss": 1.3074,
      "step": 2700
    },
    {
      "epoch": 0.00275,
      "grad_norm": 0.6940101981163025,
      "learning_rate": 4.986255e-05,
      "loss": 1.3748,
      "step": 2750
    },
    {
      "epoch": 0.0028,
      "grad_norm": 0.6420891880989075,
      "learning_rate": 4.9860050000000004e-05,
      "loss": 1.4089,
      "step": 2800
    },
    {
      "epoch": 0.00285,
      "grad_norm": 0.8561428189277649,
      "learning_rate": 4.9857550000000005e-05,
      "loss": 1.4052,
      "step": 2850
    },
    {
      "epoch": 0.0029,
      "grad_norm": 0.6900970935821533,
      "learning_rate": 4.9855050000000006e-05,
      "loss": 1.4323,
      "step": 2900
    },
    {
      "epoch": 0.00295,
      "grad_norm": 0.8071371912956238,
      "learning_rate": 4.985255e-05,
      "loss": 1.3485,
      "step": 2950
    },
    {
      "epoch": 0.003,
      "grad_norm": 0.6493249535560608,
      "learning_rate": 4.985005e-05,
      "loss": 1.4321,
      "step": 3000
    },
    {
      "epoch": 0.00305,
      "grad_norm": 0.8712514042854309,
      "learning_rate": 4.9847550000000004e-05,
      "loss": 1.406,
      "step": 3050
    },
    {
      "epoch": 0.0031,
      "grad_norm": 0.8464570045471191,
      "learning_rate": 4.9845050000000005e-05,
      "loss": 1.3322,
      "step": 3100
    },
    {
      "epoch": 0.00315,
      "grad_norm": 0.7779001593589783,
      "learning_rate": 4.9842550000000006e-05,
      "loss": 1.3905,
      "step": 3150
    },
    {
      "epoch": 0.0032,
      "grad_norm": 0.8314748406410217,
      "learning_rate": 4.984005000000001e-05,
      "loss": 1.4134,
      "step": 3200
    },
    {
      "epoch": 0.00325,
      "grad_norm": 0.7180835604667664,
      "learning_rate": 4.983755e-05,
      "loss": 1.3629,
      "step": 3250
    },
    {
      "epoch": 0.0033,
      "grad_norm": 0.6165933012962341,
      "learning_rate": 4.9835049999999996e-05,
      "loss": 1.3833,
      "step": 3300
    },
    {
      "epoch": 0.00335,
      "grad_norm": 1.8662049770355225,
      "learning_rate": 4.983255e-05,
      "loss": 1.3557,
      "step": 3350
    },
    {
      "epoch": 0.0034,
      "grad_norm": 0.9091550707817078,
      "learning_rate": 4.983005e-05,
      "loss": 1.4092,
      "step": 3400
    },
    {
      "epoch": 0.00345,
      "grad_norm": 0.9908722639083862,
      "learning_rate": 4.982755e-05,
      "loss": 1.3465,
      "step": 3450
    },
    {
      "epoch": 0.0035,
      "grad_norm": 0.862427830696106,
      "learning_rate": 4.982505e-05,
      "loss": 1.3143,
      "step": 3500
    },
    {
      "epoch": 0.00355,
      "grad_norm": 0.755211591720581,
      "learning_rate": 4.982255e-05,
      "loss": 1.358,
      "step": 3550
    },
    {
      "epoch": 0.0036,
      "grad_norm": 0.7872670888900757,
      "learning_rate": 4.9820050000000004e-05,
      "loss": 1.3952,
      "step": 3600
    },
    {
      "epoch": 0.00365,
      "grad_norm": 1.5124619007110596,
      "learning_rate": 4.981755e-05,
      "loss": 1.2929,
      "step": 3650
    },
    {
      "epoch": 0.0037,
      "grad_norm": 0.7712079286575317,
      "learning_rate": 4.981505e-05,
      "loss": 1.3429,
      "step": 3700
    },
    {
      "epoch": 0.00375,
      "grad_norm": 0.7001494765281677,
      "learning_rate": 4.981255e-05,
      "loss": 1.272,
      "step": 3750
    },
    {
      "epoch": 0.0038,
      "grad_norm": 0.674104630947113,
      "learning_rate": 4.981005e-05,
      "loss": 1.3891,
      "step": 3800
    },
    {
      "epoch": 0.00385,
      "grad_norm": 1.0494478940963745,
      "learning_rate": 4.9807550000000004e-05,
      "loss": 1.333,
      "step": 3850
    },
    {
      "epoch": 0.0039,
      "grad_norm": 0.6674365401268005,
      "learning_rate": 4.9805050000000005e-05,
      "loss": 1.4418,
      "step": 3900
    },
    {
      "epoch": 0.00395,
      "grad_norm": 0.7624682784080505,
      "learning_rate": 4.9802550000000007e-05,
      "loss": 1.2837,
      "step": 3950
    },
    {
      "epoch": 0.004,
      "grad_norm": 1.5128511190414429,
      "learning_rate": 4.980005e-05,
      "loss": 1.2609,
      "step": 4000
    },
    {
      "epoch": 0.00405,
      "grad_norm": 0.5422250628471375,
      "learning_rate": 4.979755e-05,
      "loss": 1.3253,
      "step": 4050
    },
    {
      "epoch": 0.0041,
      "grad_norm": 0.63419508934021,
      "learning_rate": 4.9795050000000004e-05,
      "loss": 1.3369,
      "step": 4100
    },
    {
      "epoch": 0.00415,
      "grad_norm": 0.7608025670051575,
      "learning_rate": 4.9792550000000005e-05,
      "loss": 1.3186,
      "step": 4150
    },
    {
      "epoch": 0.0042,
      "grad_norm": 0.6282067894935608,
      "learning_rate": 4.9790050000000006e-05,
      "loss": 1.3755,
      "step": 4200
    },
    {
      "epoch": 0.00425,
      "grad_norm": 0.7882742881774902,
      "learning_rate": 4.978755000000001e-05,
      "loss": 1.3326,
      "step": 4250
    },
    {
      "epoch": 0.0043,
      "grad_norm": 0.6797521710395813,
      "learning_rate": 4.978505e-05,
      "loss": 1.2774,
      "step": 4300
    },
    {
      "epoch": 0.00435,
      "grad_norm": 0.6129476428031921,
      "learning_rate": 4.978255e-05,
      "loss": 1.324,
      "step": 4350
    },
    {
      "epoch": 0.0044,
      "grad_norm": 0.7544896006584167,
      "learning_rate": 4.978005e-05,
      "loss": 1.299,
      "step": 4400
    },
    {
      "epoch": 0.00445,
      "grad_norm": 0.6218116283416748,
      "learning_rate": 4.977755e-05,
      "loss": 1.3084,
      "step": 4450
    },
    {
      "epoch": 0.0045,
      "grad_norm": 0.5591554045677185,
      "learning_rate": 4.977505e-05,
      "loss": 1.3001,
      "step": 4500
    },
    {
      "epoch": 0.00455,
      "grad_norm": 0.854743242263794,
      "learning_rate": 4.977255e-05,
      "loss": 1.3357,
      "step": 4550
    },
    {
      "epoch": 0.0046,
      "grad_norm": 0.7265657186508179,
      "learning_rate": 4.977005e-05,
      "loss": 1.3362,
      "step": 4600
    },
    {
      "epoch": 0.00465,
      "grad_norm": 0.8089432120323181,
      "learning_rate": 4.9767550000000004e-05,
      "loss": 1.3087,
      "step": 4650
    },
    {
      "epoch": 0.0047,
      "grad_norm": 0.8999931216239929,
      "learning_rate": 4.976505e-05,
      "loss": 1.3894,
      "step": 4700
    },
    {
      "epoch": 0.00475,
      "grad_norm": 0.7705693244934082,
      "learning_rate": 4.976255e-05,
      "loss": 1.2585,
      "step": 4750
    },
    {
      "epoch": 0.0048,
      "grad_norm": 0.7201912999153137,
      "learning_rate": 4.976005e-05,
      "loss": 1.221,
      "step": 4800
    },
    {
      "epoch": 0.00485,
      "grad_norm": 0.6199264526367188,
      "learning_rate": 4.975755e-05,
      "loss": 1.2878,
      "step": 4850
    },
    {
      "epoch": 0.0049,
      "grad_norm": 0.6883070468902588,
      "learning_rate": 4.9755050000000004e-05,
      "loss": 1.3032,
      "step": 4900
    },
    {
      "epoch": 0.00495,
      "grad_norm": 0.524889349937439,
      "learning_rate": 4.9752550000000005e-05,
      "loss": 1.2706,
      "step": 4950
    },
    {
      "epoch": 0.005,
      "grad_norm": 0.7067034244537354,
      "learning_rate": 4.9750050000000007e-05,
      "loss": 1.2424,
      "step": 5000
    },
    {
      "epoch": 0.00505,
      "grad_norm": 0.7097471356391907,
      "learning_rate": 4.974755e-05,
      "loss": 1.3175,
      "step": 5050
    },
    {
      "epoch": 0.0051,
      "grad_norm": 0.6991677284240723,
      "learning_rate": 4.974505e-05,
      "loss": 1.2983,
      "step": 5100
    },
    {
      "epoch": 0.00515,
      "grad_norm": 1.4025177955627441,
      "learning_rate": 4.9742550000000004e-05,
      "loss": 1.2695,
      "step": 5150
    },
    {
      "epoch": 0.0052,
      "grad_norm": 0.6075018048286438,
      "learning_rate": 4.9740050000000005e-05,
      "loss": 1.2973,
      "step": 5200
    },
    {
      "epoch": 0.00525,
      "grad_norm": 0.5344257950782776,
      "learning_rate": 4.9737550000000006e-05,
      "loss": 1.2353,
      "step": 5250
    },
    {
      "epoch": 0.0053,
      "grad_norm": 0.5963959693908691,
      "learning_rate": 4.973505e-05,
      "loss": 1.2359,
      "step": 5300
    },
    {
      "epoch": 0.00535,
      "grad_norm": 0.9493517279624939,
      "learning_rate": 4.973255e-05,
      "loss": 1.252,
      "step": 5350
    },
    {
      "epoch": 0.0054,
      "grad_norm": 0.6466087102890015,
      "learning_rate": 4.9730050000000003e-05,
      "loss": 1.2012,
      "step": 5400
    },
    {
      "epoch": 0.00545,
      "grad_norm": 0.6407445669174194,
      "learning_rate": 4.972755e-05,
      "loss": 1.2615,
      "step": 5450
    },
    {
      "epoch": 0.0055,
      "grad_norm": 0.617421567440033,
      "learning_rate": 4.972505e-05,
      "loss": 1.2718,
      "step": 5500
    },
    {
      "epoch": 0.00555,
      "grad_norm": 0.5649735331535339,
      "learning_rate": 4.972255e-05,
      "loss": 1.2439,
      "step": 5550
    },
    {
      "epoch": 0.0056,
      "grad_norm": 0.675645649433136,
      "learning_rate": 4.972005e-05,
      "loss": 1.2847,
      "step": 5600
    },
    {
      "epoch": 0.00565,
      "grad_norm": 0.6887643337249756,
      "learning_rate": 4.971755e-05,
      "loss": 1.2464,
      "step": 5650
    },
    {
      "epoch": 0.0057,
      "grad_norm": 0.6572535037994385,
      "learning_rate": 4.9715050000000004e-05,
      "loss": 1.2517,
      "step": 5700
    },
    {
      "epoch": 0.00575,
      "grad_norm": 0.6076003313064575,
      "learning_rate": 4.971255e-05,
      "loss": 1.2183,
      "step": 5750
    },
    {
      "epoch": 0.0058,
      "grad_norm": 0.6134760975837708,
      "learning_rate": 4.971005e-05,
      "loss": 1.2303,
      "step": 5800
    },
    {
      "epoch": 0.00585,
      "grad_norm": 0.6791936159133911,
      "learning_rate": 4.970755e-05,
      "loss": 1.2351,
      "step": 5850
    },
    {
      "epoch": 0.0059,
      "grad_norm": 0.6307588219642639,
      "learning_rate": 4.970505e-05,
      "loss": 1.254,
      "step": 5900
    },
    {
      "epoch": 0.00595,
      "grad_norm": 1.1272459030151367,
      "learning_rate": 4.9702550000000004e-05,
      "loss": 1.265,
      "step": 5950
    },
    {
      "epoch": 0.006,
      "grad_norm": 0.5415930151939392,
      "learning_rate": 4.9700050000000005e-05,
      "loss": 1.2068,
      "step": 6000
    },
    {
      "epoch": 0.00605,
      "grad_norm": 0.5450507402420044,
      "learning_rate": 4.969755000000001e-05,
      "loss": 1.2408,
      "step": 6050
    },
    {
      "epoch": 0.0061,
      "grad_norm": 0.6528737545013428,
      "learning_rate": 4.969505e-05,
      "loss": 1.2035,
      "step": 6100
    },
    {
      "epoch": 0.00615,
      "grad_norm": 0.6139540672302246,
      "learning_rate": 4.969255e-05,
      "loss": 1.2077,
      "step": 6150
    },
    {
      "epoch": 0.0062,
      "grad_norm": 0.6503865718841553,
      "learning_rate": 4.9690050000000004e-05,
      "loss": 1.2055,
      "step": 6200
    },
    {
      "epoch": 0.00625,
      "grad_norm": 0.6888736486434937,
      "learning_rate": 4.9687550000000005e-05,
      "loss": 1.2161,
      "step": 6250
    },
    {
      "epoch": 0.0063,
      "grad_norm": 1.1457703113555908,
      "learning_rate": 4.968505e-05,
      "loss": 1.1939,
      "step": 6300
    },
    {
      "epoch": 0.00635,
      "grad_norm": 0.5861665606498718,
      "learning_rate": 4.968255e-05,
      "loss": 1.2177,
      "step": 6350
    },
    {
      "epoch": 0.0064,
      "grad_norm": 0.8666422963142395,
      "learning_rate": 4.968005e-05,
      "loss": 1.1945,
      "step": 6400
    },
    {
      "epoch": 0.00645,
      "grad_norm": 0.6094856858253479,
      "learning_rate": 4.9677550000000003e-05,
      "loss": 1.2394,
      "step": 6450
    },
    {
      "epoch": 0.0065,
      "grad_norm": 0.6510404348373413,
      "learning_rate": 4.967505e-05,
      "loss": 1.1764,
      "step": 6500
    },
    {
      "epoch": 0.00655,
      "grad_norm": 0.7044185996055603,
      "learning_rate": 4.967255e-05,
      "loss": 1.1951,
      "step": 6550
    },
    {
      "epoch": 0.0066,
      "grad_norm": 0.6577385663986206,
      "learning_rate": 4.967005e-05,
      "loss": 1.2186,
      "step": 6600
    },
    {
      "epoch": 0.00665,
      "grad_norm": 0.6468052268028259,
      "learning_rate": 4.966755e-05,
      "loss": 1.2491,
      "step": 6650
    },
    {
      "epoch": 0.0067,
      "grad_norm": 0.6153103709220886,
      "learning_rate": 4.966505e-05,
      "loss": 1.25,
      "step": 6700
    },
    {
      "epoch": 0.00675,
      "grad_norm": 0.6066887974739075,
      "learning_rate": 4.9662550000000004e-05,
      "loss": 1.1852,
      "step": 6750
    },
    {
      "epoch": 0.0068,
      "grad_norm": 0.5735403895378113,
      "learning_rate": 4.9660050000000006e-05,
      "loss": 1.2139,
      "step": 6800
    },
    {
      "epoch": 0.00685,
      "grad_norm": 0.6247462630271912,
      "learning_rate": 4.965755e-05,
      "loss": 1.1597,
      "step": 6850
    },
    {
      "epoch": 0.0069,
      "grad_norm": 0.6690845489501953,
      "learning_rate": 4.965505e-05,
      "loss": 1.2095,
      "step": 6900
    },
    {
      "epoch": 0.00695,
      "grad_norm": 0.5683197975158691,
      "learning_rate": 4.965255e-05,
      "loss": 1.3077,
      "step": 6950
    },
    {
      "epoch": 0.007,
      "grad_norm": 0.6314153671264648,
      "learning_rate": 4.9650050000000004e-05,
      "loss": 1.2556,
      "step": 7000
    },
    {
      "epoch": 0.00705,
      "grad_norm": 0.641153872013092,
      "learning_rate": 4.9647550000000005e-05,
      "loss": 1.1643,
      "step": 7050
    },
    {
      "epoch": 0.0071,
      "grad_norm": 0.6113337278366089,
      "learning_rate": 4.964505000000001e-05,
      "loss": 1.1688,
      "step": 7100
    },
    {
      "epoch": 0.00715,
      "grad_norm": 0.5102031230926514,
      "learning_rate": 4.964255e-05,
      "loss": 1.1835,
      "step": 7150
    },
    {
      "epoch": 0.0072,
      "grad_norm": 0.6926486492156982,
      "learning_rate": 4.964005e-05,
      "loss": 1.1704,
      "step": 7200
    },
    {
      "epoch": 0.00725,
      "grad_norm": 0.594326376914978,
      "learning_rate": 4.9637550000000004e-05,
      "loss": 1.1434,
      "step": 7250
    },
    {
      "epoch": 0.0073,
      "grad_norm": 0.5609344840049744,
      "learning_rate": 4.963505e-05,
      "loss": 1.2389,
      "step": 7300
    },
    {
      "epoch": 0.00735,
      "grad_norm": 0.6464629173278809,
      "learning_rate": 4.963255e-05,
      "loss": 1.2097,
      "step": 7350
    },
    {
      "epoch": 0.0074,
      "grad_norm": 0.5782963633537292,
      "learning_rate": 4.963005e-05,
      "loss": 1.2607,
      "step": 7400
    },
    {
      "epoch": 0.00745,
      "grad_norm": 0.7301307320594788,
      "learning_rate": 4.962755e-05,
      "loss": 1.1759,
      "step": 7450
    },
    {
      "epoch": 0.0075,
      "grad_norm": 0.5845964550971985,
      "learning_rate": 4.9625050000000004e-05,
      "loss": 1.1806,
      "step": 7500
    },
    {
      "epoch": 0.00755,
      "grad_norm": 0.618295431137085,
      "learning_rate": 4.962255e-05,
      "loss": 1.1339,
      "step": 7550
    },
    {
      "epoch": 0.0076,
      "grad_norm": 0.5799853801727295,
      "learning_rate": 4.962005e-05,
      "loss": 1.1641,
      "step": 7600
    },
    {
      "epoch": 0.00765,
      "grad_norm": 0.5227323174476624,
      "learning_rate": 4.961755e-05,
      "loss": 1.2375,
      "step": 7650
    },
    {
      "epoch": 0.0077,
      "grad_norm": 0.699111819267273,
      "learning_rate": 4.961505e-05,
      "loss": 1.0943,
      "step": 7700
    },
    {
      "epoch": 0.00775,
      "grad_norm": 0.5230436325073242,
      "learning_rate": 4.961255e-05,
      "loss": 1.1762,
      "step": 7750
    },
    {
      "epoch": 0.0078,
      "grad_norm": 0.5776082873344421,
      "learning_rate": 4.9610050000000005e-05,
      "loss": 1.18,
      "step": 7800
    },
    {
      "epoch": 0.00785,
      "grad_norm": 0.584697425365448,
      "learning_rate": 4.9607550000000006e-05,
      "loss": 1.2878,
      "step": 7850
    },
    {
      "epoch": 0.0079,
      "grad_norm": 0.66282057762146,
      "learning_rate": 4.960505e-05,
      "loss": 1.219,
      "step": 7900
    },
    {
      "epoch": 0.00795,
      "grad_norm": 1.2405346632003784,
      "learning_rate": 4.960255e-05,
      "loss": 1.0984,
      "step": 7950
    },
    {
      "epoch": 0.008,
      "grad_norm": 0.6882185339927673,
      "learning_rate": 4.960005e-05,
      "loss": 1.2291,
      "step": 8000
    },
    {
      "epoch": 0.00805,
      "grad_norm": 0.5376139879226685,
      "learning_rate": 4.9597550000000004e-05,
      "loss": 1.1768,
      "step": 8050
    },
    {
      "epoch": 0.0081,
      "grad_norm": 0.5613966584205627,
      "learning_rate": 4.9595050000000006e-05,
      "loss": 1.1882,
      "step": 8100
    },
    {
      "epoch": 0.00815,
      "grad_norm": 0.5976732969284058,
      "learning_rate": 4.959255000000001e-05,
      "loss": 1.1938,
      "step": 8150
    },
    {
      "epoch": 0.0082,
      "grad_norm": 0.5053662657737732,
      "learning_rate": 4.959005e-05,
      "loss": 1.1335,
      "step": 8200
    },
    {
      "epoch": 0.00825,
      "grad_norm": 0.5319002866744995,
      "learning_rate": 4.958755e-05,
      "loss": 1.23,
      "step": 8250
    },
    {
      "epoch": 0.0083,
      "grad_norm": 0.6441113948822021,
      "learning_rate": 4.9585050000000004e-05,
      "loss": 1.095,
      "step": 8300
    },
    {
      "epoch": 0.00835,
      "grad_norm": 0.7779256701469421,
      "learning_rate": 4.958255e-05,
      "loss": 1.1637,
      "step": 8350
    },
    {
      "epoch": 0.0084,
      "grad_norm": 0.6262242794036865,
      "learning_rate": 4.958005e-05,
      "loss": 1.1756,
      "step": 8400
    },
    {
      "epoch": 0.00845,
      "grad_norm": 0.6683831214904785,
      "learning_rate": 4.957755e-05,
      "loss": 1.2142,
      "step": 8450
    },
    {
      "epoch": 0.0085,
      "grad_norm": 0.8390682339668274,
      "learning_rate": 4.957505e-05,
      "loss": 1.147,
      "step": 8500
    },
    {
      "epoch": 0.00855,
      "grad_norm": 0.6620815396308899,
      "learning_rate": 4.9572550000000004e-05,
      "loss": 1.1619,
      "step": 8550
    },
    {
      "epoch": 0.0086,
      "grad_norm": 0.5890603065490723,
      "learning_rate": 4.957005e-05,
      "loss": 1.1076,
      "step": 8600
    },
    {
      "epoch": 0.00865,
      "grad_norm": 0.578610897064209,
      "learning_rate": 4.956755e-05,
      "loss": 1.0664,
      "step": 8650
    },
    {
      "epoch": 0.0087,
      "grad_norm": 0.5939948558807373,
      "learning_rate": 4.956505e-05,
      "loss": 1.1613,
      "step": 8700
    },
    {
      "epoch": 0.00875,
      "grad_norm": 0.6854498386383057,
      "learning_rate": 4.956255e-05,
      "loss": 1.218,
      "step": 8750
    },
    {
      "epoch": 0.0088,
      "grad_norm": 0.8202462196350098,
      "learning_rate": 4.956005e-05,
      "loss": 1.2239,
      "step": 8800
    },
    {
      "epoch": 0.00885,
      "grad_norm": 0.5740517973899841,
      "learning_rate": 4.9557550000000005e-05,
      "loss": 1.1104,
      "step": 8850
    },
    {
      "epoch": 0.0089,
      "grad_norm": 0.7225694060325623,
      "learning_rate": 4.9555050000000006e-05,
      "loss": 1.1617,
      "step": 8900
    },
    {
      "epoch": 0.00895,
      "grad_norm": 0.5652306079864502,
      "learning_rate": 4.955255e-05,
      "loss": 1.0913,
      "step": 8950
    },
    {
      "epoch": 0.009,
      "grad_norm": 0.9501689672470093,
      "learning_rate": 4.955005e-05,
      "loss": 1.1862,
      "step": 9000
    }
  ],
  "logging_steps": 50,
  "max_steps": 1000000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.67761573576704e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}